UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Enzymatic hydrolysis of lignocellulose : cellulase enzyme adsorption and recycle Tu, Maobing 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2007-268131.pdf [ 10.09MB ]
Metadata
JSON: 831-1.0074969.json
JSON-LD: 831-1.0074969-ld.json
RDF/XML (Pretty): 831-1.0074969-rdf.xml
RDF/JSON: 831-1.0074969-rdf.json
Turtle: 831-1.0074969-turtle.txt
N-Triples: 831-1.0074969-rdf-ntriples.txt
Original Record: 831-1.0074969-source.json
Full Text
831-1.0074969-fulltext.txt
Citation
831-1.0074969.ris

Full Text

E N Z Y M A T I C HYDROLYSIS OF L I G N O C E L L U L O S E : C E L L U L A S E E N Z Y M E ADSORPTION AND R E C Y C L E by MAOBING T U i A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF D O C T O R OF PHILOSOPHY in T H E F A C U L T Y OF G R A D U A T E STUDIES (Forestry) T H E UNIVERSITY OF BRITISH C O L U M B I A December, 2006 ©Maobing Tu, 2006 11 Abstract Producing ethanol from the bioconversion of lignocellulosic substrates is one of the most promising technologies to decrease fossil fuel utilization. However, the current economics of the bioconversion process prohibit its commercialization due to the high cost of cellulase enzymes. One potential means to decrease enzyme costs is to recycle enzymes during the bioconversion process. The initial work focused on comparing the distribution of cellulases among the solid and liquid phases after a typical enzymatic hydrolysis of Av ice l and an organosolv pretreated Douglas fir substrate. It was shown that 50% of the applied cellulases desorbed into the liquid phase after the hydrolysis of an ethanol pretreated D . fir substrate compared to 76% in the case of Av ice l . B y exploiting the natural affinity of cellulases for cellulosic substrates, the free enzymes were recovered via readsorption onto fresh substrates. Using this approach, 85% of the free enzymes could be recovered, compared to an 82% recovery predicted by the Langmuir isotherm model. A novel recycling strategy for recovering both the free and bound enzymes was developed where Tween 80 was added at the beginning of the hydrolysis, followed by the addition of fresh substrate to recover free enzymes after hydrolysis. The cellulases from T. reesei (Hypocrea jecorina) could be recycled for four consecutive rounds of hydrolysis of an ethanol pretreated (EPLP) substrate with the addition of 0.2% Tween 80, compared to one round with a steam exploded (SELP) substrate, presumably due to the higher lignin content of the S E L P substrate. Comparing isolated lignin preparations from S E L P and E P L P , it was shown that C E L - S E L P lignin Ill exhibited a greater capacity to bind cellulases than C E L - E P L P lignin. A reduction in the adsorption of cellulases to lignin was achieved by the addition of Tween 80. The recycling of P-glucosidase was achieved by immobilization on an inert carrier, Eupergit C . The immobilized P-glucosidase exhibited improved operational stability and an increase in the apparent K m and V m a x . Overall, the results demonstrated that enzyme recycling using a combination of surfactants, readsorption onto substrates and enzyme immobilization could potentially decrease enzyme costs in the hydrolysis of softwoods during the bioconversion process. IV Table of Contents A B S T R A C T ii T A B L E OF CONTENTS iv LIST OF T A B L E S vii LIST OF FIGURES ix LIST OF ABBREVIATIONS xiv ACKNOWLEDGEMENTS xvii 1 INTRODUCTION 1 1.1 Background-bioethanol 1 1.1.1 Energy and environmental issues 1 1.1.2 Bioconversion of biomass to ethanol 2 1.1.3 Problem definition 3 1.2 Lignocellulosic materials 6 1.3 Bioconversion of lignocellulosic materials 10 1.3.1 Pretreatment 13 1.3.2 Enzymatic hydrolysis 17 1.3.3 Cellulase enzyme system from Trichoderma reesei 19 1.3.4 Cellulases from Penicillium sp 22 1.3.5 Fermentation 26 1.4 Strategies to reduce the cost of enzymatic hydrolysis 28 1.4.1 Improving cellulase production 28 1.4.2 Improving cellulase activity 30 1.4.3 Addition of surfactants to improve enzymatic hydrolysis 32 1.4.4 Recovery and reuse of cellulases 34 1.4.4.1 Adsorption of cellulase on substrates 34 1.4.4.2 Cellulase recycling during lignocellulosic hydrolysis 39 1.4.5 Limitations of current cellulase recycling strategies 42 1.5 Thesis Objectives 43 2 MATERIALS AND METHODS 49 2.1 Enzymes and immobilization 49 2.1.1 Cellulases 49 2.1.2 Enzyme activity assays 49 2.1.3 Immobilization of P-glucosidase 50 V 2.2 Protein content assay 51 2.3 Lignocellulosic substrates 52 2.3.1 Steam explosion pretreatment 52 2.3.2 Organosolv pretreatment 53 2.3.3 Other cellulose substrates and chemicals 53 2.3.4 Characterization of substrates 54 2.3.5 Crystallinity by solid state nuclear magnetic resonance (NMR) 54 2.3.6 Cellulolytic enzyme lignin (CEL) isolation 55 2.3.6.1 Elemental analysis of isolated lignin 55 2.4 Cellulase adsorption kinetics and isotherm 56 2.4.1 Cellulase adsorption kinetics and isotherm on lignocellulosic materials 56 2.4.2 Cellulase adsorption kinetics and isotherm on cellulolytic enzyme lignin 57 2.5 Enzymatic hydrolysis of lignocellulosic substrates 58 2.6 Cellulase recovery strategies 59 2.6.1 Recovering free cellulases from supernatants and bound enzymes with hydrolysis residues... 59 2.6.2 Recovering free cellulase from supernatants and bound enzymes via desorption 60 2.6.3 Addition of surfactant to lignocellulosic hydrolysis 60 2.7 Experimental design and statistical analysis 61 2.7.1 Experimental design 61 2.7.2 Statistical analysis 62 3 RESULTS AND DISCUSSION 63 3.1 Cellulase distribution and free cellulase recovery 63 3.1.1 Background : : 63 3.1.2 Enzymatic hydrolysis and changes in enzyme distribution 64 3.1.3 Cellulase distribution during lignocellulosic hydrolysis 71 3.1.4 Theoretical prediction of free cellulase recovery 73 3.1.5 Experimental verification of free cellulase recovery 79 3.1.6 The activity of recovered free cellulases 80 3.1.7 Conclusions 84 3.2 Optimizing the recovery of bound cellulases using Response Surface Methodology 86 3.2.1 Background 86 3.2.2 Experimental design and optimization of enzyme desorption from residual substrate 87 3.2.3 Statistical Analysis and Response Surface Methodology 89 3.2.4 Effect of temperature, pH, ionic strength and surfactant concentration on cellulase desorption96 3.2.5 The comparison of two potential recycling strategies 100 3.2.6 Conclusions 102 3.3 Application of enzyme recovery to lignin rich substrates 104 3.3.1 Background 104 3.3.2 Effect of various surfactants on enzymatic hydrolysis and recycling 107 3.3.3 Evaluating the adsorption kinetics and isotherms of three cellulase preparations 112 3.3.4 Effect of varying cellulase preparations on enzyme recycling 115 3.3.5 Effect of different pretreatments on enzyme recycling 117 3.3.6 Economic analysis for enzyme recycling with the addition of Tween 121 3.3.7 Conclusions 132 V I 3.4 Adsorption of cellulases on enzymatic lignin 133 3.4.1 Background 133 3.4.2 Adsorption kinetics and isotherm 134 3.4.3 Effect of temperature on cellulase adsorption onto lignin 142 3.4.4 Effect of ionic strength on cellulase adsorption onto lignin 145 3.4.5 Effect of Tween 80 on cellulase adsorption onto lignin 147 3.4.6 Conclusions : 148 3.5 Immobilization of p-glucosidase 150 3.5.1 Background 150 3.5.2 Immobilization of (5-glucosidase on Eupergit C 152 3.5.3 Thermal stability of free and immobilized enzyme 154 3.5.4 Effect of immobilization on pH optimum 156 3.5.5 Determination of kinetic parameters 157 3.5.6 Hydrolysis of cellulosic and lignocellulosic substrates using immobilized P-glucosidase 158 3.5.7 Conclusions 162 4 CONCLUSIONS AND FUTURE WORK 164 R E F E R E N C E S 169 Vll List of Tables Table 1. Cellulase action models and products 19 Table 2. Physical features of cellulases from Trichoderma reesei 21 Table 3. Physical features of cellulases from Penicillium sp 23 Table 4. Substrate and enzyme factors limiting enzymatic hydrolysis 26 Table 5. Fermentation approaches for bioconversion 28 Table 6. Adsorption parameters for Celluclast and CBDcex on different cellulosic substrates 38 Table 7. Cellulase recycling strategies during the hydrolysis of lignocellulosic materials 42 Table 8. Major substrates used in this research thesis 3 52 Table 9. Experimental design (small center composite design) for enzyme desorption from hydrolysis residues 62 Table 10. Experimental design with variables, levels and response values in the small central composite design 90 Table 11. Analysis of variance ( A N O V A ) for cellulase activity from the desorption for the master model and predictive model (without non-significant terms) 91 Table 12. Analysis of variance ( A N O V A ) for protein content from the desorption for the master model and predictive model (without non-significant terms) 92 Table 13. Estimates for cellulase activity and protein content from desorption in the polynomial model 94 Table 14. Surfactants with corresponding hydrophilic-lypophilic balance ( H L B ) and the increased performance of enzyme recycling during hydrolysis of E P L P after surfactant addition 112 Table 15. Langmuir constants from Celluclast, Spezyme and M S U B C on E P L P substrate at 25 °C 114 Table 16. Characteristics of S E L P and E P L P substrates by Klason analysis 117 Table 17. Assumptions for bioconversion of an ethanol pretreated Lodgepole pine (EPLP) to produce ethanol (SHF) 123 Table 18. Effect of Tween 80 concentration on the number of potential recycling rounds 125 Table 19. Enzyme cost savings with different concentrations of Tween 80 127 Table 20. Various costs used in the economic analysis of enzyme recycling and Tween 80 recirculation for producing 1 gallon of ethanol 131 Table 21. Characteristics of C E L - S E L P lignin and C E L - E P L P lignin by Klason analysis 134 Table 22. Adsorption isotherm parameters for different cellulases on C E L - S E L P and C E L - E P L P lignin at 25°C 141 Table 23. Thermodynamic parameters of cellulase adsorption on C E L - S E L P and C E L - E P L P lignin 145 Table 24. Effect of glucose and blocking agents on the activity of immobilized ( 3 -glucosidase 153 Table 25. Kinetic constants for the free and immobilized P-glucosidase 158 ix List of Figures Figure 1. Microscopic structure of cellulose (Bruley et al. 1974) 7 Figure 2. Cellulose is a linear polymer of (3-(l, 4)-D-glucopyranose units 8 Figure 3. The basic structural units (precursors) of lignin (Sjostrom 1993) 9 Figure 4. The structure of softwood lignin showing functional groups and linkages (Adler 1977) 10 Figure 5. Schematic diagram of a bioconversion process 12 Figure 6. Cellulose hydrolysis by cellulases from Trichoderma reesei. The solid squares represent reducing ends, and the open squares represent non-reducing ends (Lynd et al. 2002) 20 Figure 7. Cellulase adsorption and hydrolysis profile of Avice l . Hydrolysis conditions: 45°C, 48 hr, 150 rpm and 20 F P U g" f cellulose, 40 IU (3-Glucosidase g"1 cellulose, 2% Avice l in N a A c buffer pH4.8 66 Figure 8. Cellulase adsorption and hydrolysis profile of acetic acid pretreated Douglas fir (3% lignin). Hydrolysis conditions: 45°C, 48 hr, 150 rpm and 20 F P U g"1 cellulose, 40 IU P-Glucosidase g 1 cellulose, 2% in N a A c buffer pH4.8 66 Figure 9. S D S - P A G E analysis of hydrolysate over 48 hr hydrolysis of Av ice l . A , 0 hr; B , 2 hr; C , 6 hr; D , 12 hr; E , Protein standard; F, 24 hr; G , 36 hr; H , 48 hr. 0.85mL hydrolysate sample was put into -20°C overnight with 3 m L acetone. The mixture was centrifuged at 4000g to remove supernatant. 0.5 m L acetate buffer was added further. For S D S - P A G E running, 10 ul sample was mixed with 20 u L running sample buffer, 20 u L loading sample at 200 V for running 2 hr 67 Figure 10. S D S - P A G E analysis of hydrolysate over 48 hr hydrolysis of acetic acid pretreated Douglas fir. (The conditions were the same as above) 67 Figure 11. Change in the hydrolysis of Avice l upon supplementation either additional Avice l or cellulases. After the first 48 hr hydrolysis, an additional 20 F P U of Celluclast was added to one of the experiments and the hydrolysis was allowed to continue, while another 2 g Avice l was added into second hydrolysis experiment. For the control, the hydrolysis was continued without adding any additional Avice l or enzyme. Hydrolysis conditions: 45°C, 48 hr, 150 rpm and 20 F P U g~' cellulose, 40 IU p-Glucosidase g"1 cellulose, 2% Avice l in 100 m L N a A c buffer (pH4.8) , 70 X Figure 12. I 3 C C P M A S N M R spectrum of Avice l to obtain post-hydrolysis crystallinity. Hydrolysis conditions were 2% Avice l in 100 m L acetate buffer (pH 4.8) and 45°C, with 20 F P U Celluclast g"1 cellulose and 40 IU Novozym 188 g 1 cellulose 70 Figure 13. Hydrolysis yield, protein content and F P U activity in the supernatant for ethanol pretreated mixed softwood ( E P M S substrate). Hydrolysis conditions: The hydrolysis was carried out in 100 m L of 50 m M acetate buffer (pH 4.8) at 2% consistency (based on cellulose content). The reaction was incubated at 45 °C at 150 rpm. Enzyme loading was 10 F P U g"1 cellulose with 20 IU g"1 cellulose of P-glucosidase (Novozym 188). The substrate was E P M S with a 5.96% lignin content 73 Figure 14. Cellulase adsorption kinetics on ethanol pretreated mixed softwood (EPMS) at 4°C. Celluclast in 50 m M sodium acetate (pH 4.8) at a final protein concentration of 3.24 mg/mL was added to microtubes with the substrate at a 2% consistency. The microtubes were incubated at 4°C with inverse shaking. Aliquots were taken at 10, 20, 30, 60, 90, 120 min during the incubation, and collected the supernatant by centrifugation. The protein content was determined by the Bradford assay 75 Figure 15. Cellulases adsorption isotherm on ethanol pretreated softwood at 4°C. Different concentration of Celluclast was incubated with 2% of E P M S in 50 m M acetate buffer at 4°C for 1 hr. The protein content in the supernatant was determined for the non-adsorbed cellulase; the adsorbed cellulase was calculated from the difference in the initial cellulase content and the non-adsorbed cellulase content in the supernatant 75 Figure 16. Graphic solution for cellulase readsorption. The intersection of the equilibrium line and operating line indicates the equilibrium concentration of free cellulases 78 Figure 17. Isoelectric focusing of cellulases during E P M S hydrolysis and enzyme recycling. A , P-glucosidase; B sample after hydrolysis; C , sample before hydrolysis; D , sample after hydrolysis and readsorption; E , original cellulases (Celluclast). Samples were centrifuged at 4000g for 10 min, and then the supernatant was concentrated 16 fold by Amicon ultra centrifugal filter device (from 0.8 m L to 50 uL). Sample loadings were 3 u L on IEF gels (pH 3.0-9.0), which were run by Pharmacia L K B , Phast system 80 Figure 18. Schematic diagram of a free cellulase recycling process 82 x i Figure 19. Comparison of the activities of different sets of recycled cellulases based on E P M S hydrolysis yields. After the first round of hydrolysis, free cellulases in the supernatant were collected by centrifugation (4000g) at 4°C. The supernatant containing free cellulases was then added to fresh E P M S (2%). After 2 hr incubation, the readsorbed cellulase on fresh E P M S substrate was collected by centrifugation at 4°C. The non-adsorbed cellulase in the supernatant was recovered by ultrafiltration to remove sugars. The recycled bound enzyme on the hydrolysis residue and readsorbed cellulases were supplemented with fresh buffer containing fl-glucosidase and used to hydrolyze fresh ethanol pulp separately to compare their hydrolysis yields (as cellulase relative activities) 84 Figure 20. Process optimization of temperature, p H , ionic strength and surfactant on desorption of bound enzyme (protein content and enzyme activity) from the lignin rich hydrolysis residues at the end of hydrolysis (EPMS) using the prediction profiler 95 Figure 21. Three dimensional surface plots of cellulase activity as the function of p H and temperature (Fixed levels, NaCl=0, Tween 80=0.5%) 96 Figure 22. Comparison of two recycling strategies, both involving recovery of the bound and free cellulases 101 Figure 23. Comparison of cellulose conversion over three rounds of hydrolysis (EPMS) using recycling Strategy 1 and Strategy 2 102 Figure 24. Schematic diagram of a one-step cellulase recycling process 106 Figure 25. Effect of different surfactants on the hydrolysis of E P L P (14.6% lignin) and Sigmacell 50. Hydrolysis conditions: 20 F P U Celluclast, 40 IU p-glucosidase at 45°C and 150 rpm for 48 hr with 0.2% of each surfactant 109 Figure 26. Effect of Tween 80 concentration on the hydrolysis of S E L P (41.7% lignin) Hydrolysis conditions: 10 F P U Celluclast, 20 IU P-glucosidase g"1 cellulose at 45°C and 150 rpm for 24 hr 109 Figure 27. Effect of various surfactants on enzyme recycling during the hydrolysis of E P L P . Hydrolysis conditions: 20 F P U Celluclast, 40 IU P-glucosidase at 45°C and 150 rpm for 48 hr with 0.2% of surfactants (except E P L P control). After hydrolysis, free enzymes were first collected by filtration; then fresh E P L P was added to the supernatant containing free enzymes for 2.5 hr readsorption at 25°C. Readsorbed enzyme with substrates were filtered and transferred to fresh buffer for further hydrolysis (supplemented new P-glucosidase) I l l Figure 28. Adsorption kinetics of Celluclast, Spezyme and M S U B C on E P L P substrate at 25°C 114 Figure 29. Cellulase adsorption isotherms on E P L P substrate (14.45% lignin) at 25°C 115 Figure 30. Enzyme recycling for E P L P substrate (ethanol pretreated Lodgepole pine, Klason lignin 14.45%) Standard hydrolysis: 2% substrate in 50 m L buffer with 20 F P U Cellulase and 40 IU beta-glucosidase under 45°C, 150 rpm, 48 hr 117 Figure 31. Enzyme recycling for S E L P substrate (steam exploded Lodgepole pine, 45.6% Klason lignin) Standard hydrolysis: 2% substrate in 50 m L buffer with 20 F P U Cellulase and 40 IU beta-glucosidase at 45°C, 150 rpm, 48 hr 120 Figure 32. Effect of Tween 80 concentration on cellulase distribution during the hydrolysis of E P M S . Hydrolysis was carried out at 45°C for 24 hr with 2% of E P M S , the enzyme loading was 10 F P U Celluclast, 20 IU P-glucosidase per gram of cellulose 124 Figure 33. Enzyme cost evaluation vs. concentration of Tween 80 128 Figure 34. Schematic flowsheet of ethanol production with enzyme recycling and recirculation of process streams 130 Figure 35. Cellulase adsorption kinetics on isolated C E L - S E L P lignin and C E L -E P L P lignin at 25°C. For adsorption kinetics, 30 mg lignin ( C E L - S E L P and C E L - E P L P ) samples were placed into microtubes in 5 m L acetate buffer (50 m M , p H 4.8) with 0.10-0.15 mg/mL of commercial cellulases (Celluclast, Spezyme and M S U B C ) for 6 hr incubation at 25°C. The microtubes were incubated on a rotary shaker. Aliquots of 0.15 m L were taken at 0, 15, 30, 60, 120, 180 and 360 min during the incubation. The protein content was determined using the ninhydrin assay 136 Figure 36. Cellulase adsorption isotherms on C E L - S E L P and C E L - E P L P lignins at 25°C (solid line from Langmuir model, dotted line from Freundlich). For the adsorption isotherm, different concentrations of cellulases were incubated with 2% (2g/100ml) of lignin in 50 m M acetate buffer at 25°C for 3 hr to reach equilibrium. The protein content in the supernatant was determined for the non-adsorbed cellulase. The adsorbed cellulase was calculated from the difference of initial cellulase content and non-adsorbed cellulase content in the supernatant. The classical Langmuir theory of adsorption was also applied to cellulase adsorption on lignin samples 141 Figure 37. Cellulase adsorption isotherms on C E L - S E L P lignin at different temperatures 144 Figure 38. Cellulase adsorption isotherms on C E L - E P L P lignin at different temperatures 144 Figure 39. Effect of ionic strength on cellulase adsorption on C E L - S E L P lignin (Celluclast) 146 Figure 40. Effect of surfactant concentration on cellulase adsorption on C E L - S E L P lignin (Celluclast). -0.13 mg/mL of cellulase was incubated with -20 mg lignin in 1.0 m L of 50 m M acetate buffer at 25 °C for 3 hr to reach equilibrium. The protein content in supernatant was determined for the non-adsorbed cellulase 148 Figure 41. Time course of P-glucosidase immobilization. Eupergit C (0.5g dry wt) was added to 10 m L 1.0 M potassium phosphate buffer, p H 7.0, containing 5.5mg of Novozym 188. The reaction mixture was incubated at 25°C for up to 36hr with gentle shaking. The amount of immobilized protein was determined from the difference between the total protein added and the amount remaining in solution after immobilization 153 Figure 42. Thermal stabilities of free and immobilized P-glucosidase. Enzyme was incubated in 50 m M sodium acetate buffer, p H 4.8, and then assayed at 50°C using /?-NPG. The activity of the free enzyme act 0 hr was 1.93 IU mg"1 protein; the activity of the immobilized enzyme was 3.50 IU mg"1 biocatalyst 155 Figure 43. p H optima of free and immobilized P-glucosidase. P-glucosidase activity was determined using p - N P G in sodium acetate buffer, p H 3.0-8.0 157 Figure 44. Effect of free and immobilized P-glucosidase on the hydrolysis of filter paper. Reaction mixtures containing 2% (w/v) cellulose in 25 m L 50 m M sodium acetate buffer (pH 4.8), were incubated at 45°C for up to 48 hr. The cellulase loading was 20 F P U Celluclast g"1 cellulose. The loading of free P-glucosidase was 40 IU g"1 cellulose. The loading of immobilized P-glucosidase was 5.73, 2.87 or 1.43 mg total protein g"1 cellulose (i.e., 100%, 50% and 25% of the free P-glucosidase protein loading). These quantities of immobilized P-. glucosidase correspond to 4.4, 2.2 and 1.1 IU respectively 159 Figure 45. Effect of free and immobilized P-glucosidase on the hydrolysis of various cellulosic and lignocellulosic substrates. Reaction mixtures containing 2% (w/v) cellulose in 25 m L 50 m M sodium acetate buffer (pH 4.8) were incubated at 45 °C for 48 hr. The cellulase loading was 20 F P U Celluclast g"1 cellulose 161 Figure 46. Operational stability of immobilized P-glucosidase during hydrolysis of acetic acid pretreated Douglas fir. The reaction mixture contained 20 F P U cellulase and 4.4 IU immobilized P-glucosidase g"1 cellulose. After 48 hr hydrolysis, the immobilized P-glucosidase was recovered by brief centrifugation at lOOOg. Residual substrate remained in suspension and was removed by decantation. The recovered immobilized enzyme was then used to supplement a second reaction mixture containing fresh substrate and cellulase. The process was repeated for a total of six rounds of hydrolysis. In the free enzyme control (free P-g), immobilized P-glucosidase was replaced by 40 IU of P-glucosidase. A second control without P-glucosidase supplementation (no P-g) was also included 162 List of Abbreviations a alpha P beta °C degrees Celsius A F E X ammonia fiber explosion pretreatment Ara arabinose B M C C bacterial microcrystalline cellulose B S A bovine serum albumin C concentration of unabsorbed cellulase in bulk solution C B D cellulose binding domain C B H cellobiohydrolase C D catalytic domain C E L - E P L P cellulolytic enzyme lignin from ethanol pretreated Lodgepole pine C E L - S E L P cellulolytic enzyme lignin from steam exploded Lodgepole pine C E L - S P S cellulolytic enzyme lignin from steam exploded Spruce cm centimeter C M C carboxymethylcellulose D N S dinitrosalicylic acid D P degree of polymerization E C Enzyme Commission E G endoglucanase E P L P ethanol pretreated Lodgepole pine E P M S ethanol pretreated mixed (mixture) softwood F P L C fast protein liquid chromatography F P U filter paper units FT-IR Fourier transformed infrared spectroscopy g gram g acceleration due to gravity A G change in Gibbs free energy Gal galactose G C gas chromatography G l u glucose A H change in enthalpy h(r) hour(s) HC1 hydrochloric acid H 2 0 2 hydrogen peroxide H 2 S 0 4 sulphuric acid H M F 5-hydroxymethylfurfural H P L C high performance liquid chromatography IU international units K Langmuir constant Kcat turnover number K m apparent dissociation constant kDa kilodalton L liter m meter M molar 2 - M E 2-mercaptoethanol mg milligram min minute(s) m L milliliter mm millimeter m M millimolar M W molecular weight nm nanometer O D optical density P A G E polyacrylamide gel electrophoresis rpm revolutions per minute R S M response surface methodology sec second(s) AS change in entropy SDS sodium dodecylsulphate xv i S E L P steam exploded Lodgepole pine S H F separate hydrolysis and fermentation SO2 sulphur dioxide SSF simultaneous saccharification and fermentation t time T temperature T concentration of adsorbed cellulase rm a x maximum adsorbed cellulase T C A trichloroacetic acid |j,L microliter (xm micrometer u M micromolar U V ultraviolet light Vmax maximum enzyme velocity v/v volume per volume w/v weight per volume X y l xylose XVll Acknowledgements First I would like to express my appreciation to my supervisor Dr. Jack Saddler for his support and guidance throughout my studies. I would also like to thank my committee members Dr. Paul McFarlane, Dr. Rodger Beatson and Dr. Douglas Kilburn for their continuing advice, guidance and support. I have also benefited from all the members in Forest Products Biotechnology group. I would like to thank Dr. Richard Chandra, Dr. Nei l Gilkes, Dr. Xuejun Pan, Dr. Warren Mabee and Dr. Alex Berl in for their help and helpful discussion. Finally, I would like to express my deep gratitude to my parents and my wife for encouragement and support through all these years. 1 1 Introduction 1.1 Background-bioethanol 1.1.1 Energy and environmental issues With the rapid rise in energy demand throughout the world and particularly in developing nations (Asia, Central and South America), world energy consumption is projected to rise from 380 quadrillion British thermal units (Btu) to 608 quadrillion Btu from 1997-2020 (Energy Information Administration 2000). O i l is expected to remain the dominant energy fuel (39% in 2002) as it has been for decades (Energy Information Administration 2005). World oi l prices continue to rise, recently (July, 2006) ascending to the $70 per barrel range. Crude oil consumption represents 40-60% of the total energy utilization in the world and is predicted to remain at or near 40% for the next 20 years (Energy Information Administration 2005). Escalating oi l prices have driven research focusing on alternative fuel options for the future. Indeed, it has been predicted that alternative and renewable energy consumption wi l l increase by 53% from 1999 to 2020 (Energy Information Administration 2005). In addition to economic hardship, there wi l l be profound environmental consequences if the current pattern of worldwide energy use continues. Human activities contribute about 3 bi l l ion metric tons of carbon dioxide per year into the atmosphere (Vitousek et al. 1997). A t the root of this problem is the 98% of C O 2 emitted into the atmosphere resulting from the combustion of fossil fuels (Ahrens 1985). When emitted into the earth's atmosphere, carbon dioxide and other greenhouse gases such as methane and nitrous oxide trap the heat that normally radiates from the earth resulting in an increase in 2 temperature or "greenhouse effect" (Australian Greenhouse Office 2006). Wi th a mandate of reducing emissions of carbon dioxide and other greenhouse gases, a group of countries in 1997 reached an international treaty on climate change referred to as the Kyoto Protocol (Morlot 1999). The treaty's objective is the "stabilization of greenhouse gas concentrations in the atmosphere at a level that would prevent dangerous anthropogenic interference with the climate system". In 1997, Canada signed the Kyoto Protocol and made a commitment to reduce its carbon emissions 6% below levels measured in 1990 by the years 2008-2012 (Martineau 2002). Ethanol is a potential transportation fuel with low greenhouse gas emissions (Wyman 1996). Uti l iz ing ethanol in Canada to replace gasoline could potentially enable us to reach the objectives of the Kyoto Protocol. Bioconversion of biomass such as agricultural and wood waste to ethanol is one potential method of producing a renewable and environmentally friendly fuel that is currently being explored by researchers worldwide (Kadam and Newman 1997; Lee 1997; Wyman 1999). 1.1.2 Bioconversion of biomass to ethanol Ethanol has been shown to be a promising alternative fuel to replace gasoline (Farrell et al. 2006). The manufacture of fuel ethanol has become a fast-growing and expanding industry (Sheehan and Himmel 2001). Currently, 4,227 mill ion gallons, 4,264 mil l ion gallons, and 1,004 mill ion gallons of fuel ethanol is produced mainly from sugar and starch crops per year in Brazi l , the United States, and China respectively (Renewable Fuels Association 2005). The ethanol is produced either by converting the carbohydrate portion of starch crops, such as corn, into monomeric sugars, or by using the sugars themselves from crops such as sugar cane & sugar beet that are subsequently fermented 3 resulting in the formation of ethanol. Although the production of ethanol from starch crops is expected to increase in the next few years, significant research efforts are / currently underway, exploring the viability of large-scale production of ethanol from lignocellulosic biomass. Lignocellulosic substrates such as forestry and agricultural residues represent a readily available and renewable source of biomass for the production of ethanol. There have been an increasing number of investigations assessing the potential of converting wood and agricultural residues to ethanol'during the last few decades (Boussaid et al. 1999; Galbe and Zacchi 2002; Kadam et al. 2000; Saddler et al. 1983). Bioconversion of lignocellulosic biomass to ethanol shows considerable potential for enabling the eventual utilization of ethanol as a fuel for automobiles. The United States plans to increase the inclusion of biomass derived transportation fuels in their fuel supply to at least 5% in 2010, and 8% in 2012 (Governors' Ethanol Coalition 2005). The existence of multinational research organizations such as the International Energy Agency (IEA) dedicated to the development of new fuel technologies, demonstrates the significance of bioconversion to the overall goal of securing sustainable fuel sources. 1.1.3 Problem definition As mentioned earlier, bioethanol from lignocellulosic substrates is a promising alternative fuel to replace fossil fuels. However, at present, the production of bioethanol from lignocellulose is technically difficult and economically challenging. The United States Department of Energy (DOE) has chosen a target ethanol price of U S $1.07 per gallon as a goal for the year 2010 (Aden et al. 2002). Upon techno-economic analysis of 4 the entire bioconversion process (Gregg et al. 1998), it was estimated that the enzymatic hydrolysis step accounts for a considerable portion (approximately 60%) of the total process cost. In an effort to reduce the cost associated with enzymatic hydrolysis, a considerable amount of work has focused on improving the production and efficiency of the enzymes used in the hydrolysis step (Chand et al. 2005; Chen and Wayman 1993; Deshpande and Eriksson 1984; Hogan and Meshartree 1990). In 2003, a twelve-fold reduction in the cost of enzymes was achieved by a joint effort between Genencor International and Novozymes as sponsored by the U.S . Department of Energy (DOE) . According to Genencor International, enzymes were made less expensive in two areas: a 4.9 fold reduction in the cost of enzyme production and a 2.4 fold improvement in cellulase performance (Jensen 2003). Enzyme production costs were reduced by eliminating downstream processing steps, replacing lactose with a cheaper carbon source (glucose), adding an inducer (sophorose), optimizing fermentation conditions, and by improving the production strains through random mutations. Cellulase enzyme performance was improved through optimizing the mono-component enzyme ratios, adding additional enzymes, and by improving specific activity through site-directed mutagenesis (Jensen 2003). In a parallel effort, Novozymes stated that they reduced enzyme production costs by reducing the cost of feedstocks, improving enzyme recovery, on-site production and increasing fermentation yield. To increase enzyme activity, Novozymes improved enzyme thermostability, increased specific activity and optimized the proportions of each component enzyme in the cellulase mixture (Cherry 2003). A s of early 2004, both Genencor International and Novozymes reported greater than a ten-fold 5 reduction in the cost of the enzyme to bring it to $0.50 per gallon of ethanol (U.S. Department of Energy 2006). Although the research efforts expanded by industrial and government researchers have resulted in significant reductions in the overall cost of enzyme hydrolysis, it has been stated that continuing work is expected to further reduce cellulase costs to about $0.10 U.S . per gallon of ethanol or less (U.S. Department of Energy 2006). Reducing the cost of cellulase production, improving enzyme performance (increasing specific activity) and recycling the enzymes during hydrolysis are the three main strategies that have been suggested for reducing the cost of the enzymatic hydrolysis step. However, it is unlikely that increasing the already high levels of cellulase production from T. reesei w i l l allow the enzyme cost target of $0.10 per gallon to be achieved (U.S. Department of Energy). Further improvements in cellulase specific activity by genetic engineering remains a relatively long term prospect (Mabee and Saddler 2005). It should also be noted that the reported 12 fold enzyme cost reduction from Genencor International and Novozymes was based on the hydrolysis of corn stover (Cherry 2003). Due to their recalcitrant structure and high lignin content, obtaining a 12 fold enzyme cost reduction during the enzymatic hydrolysis of softwood substrates is expected to be a challenge (Galbe and Zacchi 2002). On the other hand, the high affinity of cellulases for cellulose presents opportunities for the design of enzyme recycling strategies utilizing fresh substrates to readsorb cellulases (Castanon and Wilke 1980) in a potentially simple, effective and economical recovery process. Considering all of the factors presented, in addition to the previous work on maximizing enzyme production and improving specific activity, enzyme recycling could 6 be one of the most promising options to further decrease enzyme costs, and reach cost targets required to improve the economic viability of lignocellulosic bioconversion. In previous work cellulase recycling has been evaluated during the hydrolysis of steam pretreated hardwood substrates (Lee et al. 1995; Ramos and Saddler 1994) while there has been little work assessing the feasibility of cellulase recycling during the hydrolysis of softwood substrates. Softwoods are an abundant, readily available source of biomass in British Columbia. However, their high lignin content presents challenges for the development of enzyme recycling strategies. Organosolv and steam explosion pretreatments produce substrates that are easily hydrolyzed. Therefore, in this research project, we focused on developing cellulase recycling protocols for the hydrolysis of softwoods (Douglas fir and Lodgepole pine) pretreated using organosolv and steam explosion. 1.2 Lignocellulosic materials Straw, bagasse, corn stalk, grass, wood pulp, wood waste and bark are all examples of lignocellulosic materials. The three major components of lignocellulosics are cellulose, hemicellulose and lignin. Cellulose is comprised of long chains of glucose that can be broken down by cellulases enzymatic hydrolysis. Hemicelluloses are amorphous carbohydrate chains composed of a mixture of sugars, including arabinose, galactose, glucose, mannose and xylose. Lignin is an aromatic network polymer associated with cellulose and hemicelluloses (Sjostrom 1993; Wyman 1999). Cellulose is the most abundant biopolymer in the world composed of R - l , 4 linked D -glucose with a repeating unit of cellobiose (Figure 1 and Figure 2). Cellulose microfibrils 7 are molecular chains of cellulose held together in crystalline and amorphous regions by inter and intra-molecular hydrogen bonding. Microfibrils form fibrils that assemble to form fibers. Cellulose makes up approximately 50% of the dry weight of wood and 90% of the cotton fiber. The average degree of polymerization of (DP) of cotton cellulose is about 15,000 and about 10,000 for wood (Sjostrom 1993). Plants synthesize cellulose as one of their structural components along with lignin and hemicellulose. Fiber I Microfibril I Microfibril I Cellulose molecules I Two glucose residues Figure 1. Microscopic structure of cellulose adapted from (Bruley et al. 1974). Cellobiose Figure 2. Cellulose is a linear polymer of P-(l, 4)-D-glucopyranose units. Hemicelluloses are a group of branched polysaccharides present in almost all plant cell walls along with cellulose and lignin. Hemicelluloses include galactoglucomannans, arabinoglucuronoxylan, arabinogalactan, glucuronoxylan, glucomannan. Most hemicelluloses have a degree of polymerization of 150-200 and are relatively easily hydrolyzed to monomers such as glucose, mannose, galactose, xylose, and arabinose. L ike cellulose, hemicelluloses function as supporting materials in the cell walls. The amount of hemicelluloses in wood typically varies between 20% and 30% (Sjostrom 1993). The composition and structure of the hemicelluloses in hardwoods differ from those in softwoods. Hardwood hemicelluloses are rich in xylan polymers with small amounts of mannan, while softwood hemicelluloses are rich in mannan polymers (Sjostrom 1993). Lignin is the second most abundant biopolymer on earth next to cellulose. It has been generally accepted that lignin is the "glue" that holds cellulose and hemicellulose together within the wood structure providing rigidity while limiting water permeability. Lignin also plays a role in a plant's natural defense against degradation by microbial enzymes (Bucciarelli et al. 1998; Vance et al. 1980). Lignin is composed of phenylpropane monomers with three basic structural units (Figure 3). The phenylpropane units in lignin are connected together by C - O - C (ether) and C - C linkage. The P-O-4 aryl ether linkages dominate lignin structure (Figure 4) (Adler 1977). The other linkages include a-O-4, carbon 5-5, and 4-0-5. C h L O H I 2 C H 2 O H C H 2 O H M e M e O O H O H O M e Coumaryl alcolhol Coniferyl alcohol Sinapyl alcohol Figure 3. The basic structural units (precursors) of lignin (Sjostrom 1993). The chemical constituents of lignin vary between softwoods and hardwoods. Lignin in softwoods can be described as "guaiacyl l ignin" since it is composed mainly of guaiacyl (G) lignin sub-units, while hardwoods contain mainly a "guaiacyl-syringyl" lignin that is composed of a higher amount of syringyl (S) units than G units. In the case of softwoods, the lignin concentration is higher in the middle lamella than the secondary wall . However, approximately 70% of the total lignin content is located within the secondary wall. Due to the various cells that constitute the structure of hardwoods, there have not been any definite measurements of lignin content in the secondary wall (Sjostrom 1993). However, it has been shown that the lignin in the secondary wall of hardwood fibers possesses a high content of syringyl units, whereas guaiacyl units are present in the middle lamella and vessel walls. Both the location and type (G or S) of lignin in both hardwoods and softwoods may have profound effects on the susceptibility of the substrate to cellulolytic enzymatic hydrolysis during the bioconversion process. 10 CH20H HCOH c=o < Carbonyl Group 5-5' linkage Methoxyl Group Figure 4. The structure of softwood lignin showing functional groups and linkages adapted from (Adler 1977). 1.3 Bioconversion of lignocellulosic materials The bioconversion process consists of multiple stages converting biomass to fermentable sugars for the production of ethanol. The feedstock is first pretreated, and then subjected to enzymatic hydrolysis with subsequent conversion of glucose to ethanol by yeast or bacteria. With the increased research interest in the production of biomass derived fuels, lignocellulosic materials have been widely explored as a potential feedstock for bioconversion to ethanol (Chandrakant and Bisaria 1998; Gregg and Saddler 1995; Knauf 11 and Moniruzzaman 2004; Lee 1997; Lynd et al. 1996; Wyman 1999; Zyabreva et al. 2001). The bioconversion process involves three major stages: pretreatment, enzymatic hydrolysis and fermentation (Figure 5). The typical pretreatment methods include ammonia fiber explosion, lime pretreatment, steam explosion and organosolv (Sun and Cheng 2002; Wyman et al. 2005). The enzymatic hydrolysis is carried out by the synergetic action of exoglucanases, endoglucanases and P-glucosidase. Ethanologenic microorganisms, including yeast (Saccharomyces cerevisiae) and bacteria (Zymomonas mobilis, Klebsiella oxytoca), can efficiently ferment sugars to ethanol (Lynd et al. 2002b; Zhou et al. 2001; Zhou and Ingram 2001). The enzymatic hydrolysis of cellulose and subsequent fermentation stages can either be performed separately in separate hydrolysis and fermentation (SHF) or the hydrolysis and fermentation can be performed simultaneously in simultaneous saccharification and fermentation (SSF). Prior to the enzymatic hydrolysis stage of bioconversion, lignocellulosic substrates require pretreatment to increase accessibility to cellulase enzymes (Lynd et al. 2002b). Lignocellulosic materials 1 Pretreatment 1 Enzymatic hydrolysis 1 Fermentation Distillation * I Ethanol Methods & biocatalyst • Steam explosion • Organosolv • Ammonia fiber explosion • Lime pretreatment • Exoglucanases • Endoglucanases • Beta-glucosidase • Saccharomyces cerevisiae • Zymomonas mobilis • Klebsiella oxytoca Figure 5. Schematic diagram of a bioconversion process. 13 1.3.1 Pretreatment Lignocellulosic substrates must be either physically or chemically pretreated to fractionate the lignin, hemicellulose and cellulose components and to increase the cellulose surface area and improve accessibility to cellulase enzymes for subsequent hydrolysis. In general, pretreatment methods solubilize hemicellulose, modify and/or partially remove lignin and increase the accessibility of cellulose to cellulases. Pretreatment techniques can generally be classified into three categories: physical, chemical and biological methods. Physical pretreatment methods do not generally involve the addition of chemicals (other than steam or water) and usually involve ball-mill ing, irradiation, steaming and hydrothermolysis (Hsu 1996). Chemical pretreatment methods have attracted the most attention by researchers for their potential application in the bioconversion process. Chemical pretreatments include acid, alkaline, solvent, and ammonia ( M c M i l l a n 1994). Biological methods involve the application of lignin degrading fungi to improve substrate accessibility to cellulase enzymes (Taniguchi et al. 2005). The chief physiochemical pretreatment methods that have been explored over the past few decades include steam explosion, dilute acid pretreatment (Allen et al. 2001; Kalman et al. 2002; Nguyen et al. 1999; Nguyen et al. 2000; Saha et al. 2005b; Sun and Cheng 2005), the organosolv processes (Arato et al. 2005; Chum et al. 1990; Holtzapple and Humphrey 1984; Neilson et al. 1983; Rughani and Mcginnis 1987; Shevchenko et al. 2001b; Sidiras and Koukios 2004), lime treatment (Chang et al. 2001a; Kaar and Holtzapple 2000; K i m and Holtzapple 2005) (Schell et al. 1998; Zimbardi et al. 1999), ammonia fiber explosion and aqueous ammonia pretreatment (Arato et al. 2005; Chum et 14 al. 1990; Holtzapple and Humphrey 1984; K i m and Lee 2005b; Neilson et al. 1983; Rughani and Mcginnis 1987; Shevchenko et al. 2001b; Sidiras and Koukios 2004). Lime pretreatment Lime has been employed as a pretreatment method mainly in investigations (Adams 1995) with corn stover and poplar (Chang et al. 2001b; Kaar and Holtzapple 2000; K i m and Holtzapple 2005). In lime pretreatment, feedstocks are subjected to temperatures of 25-100°C for 1 day to 2 weeks in the presence of dilute N a O H or Ca(OH)2. L ime pretreatment removes 60-80% of the total lignin, decreases cellulose D P and crystallinity and increases surface area. A major disadvantage of the process is the difficulty in recovering N a O H or Ca(OH)2 (Hsu et al. 1996), therefore, the ammonia fiber explosion process was developed as another alkaline based pretreatment method to improve lignocellulosic hydrolysis. A F E X (Ammonia Fiber Explosion) The A F E X process is a physiochemical pretreatment where feedstocks are combined with liquid anhydrous ammonia at temperatures from 60 to 100°C at pressures from 250-300 psi for 5 min (Wyman et al. 2005). A t the end of the treatment, the pressure is released resulting in a physical action on the substrate thus facilitating its disintegration. Ammonia aids enzymatic hydrolysis by degrading crystalline cellulose (Gollapalli et al. 2002), while preserving hemicellulose (Wyman et al. 2005). The A F E X process has been shown to modify lignin, thereby reducing non-productive binding between cellulase and lignin (Martinez et al. 1991). After the treatment, the ammonia can be recovered for reuse. A F E X has been modified to the ammonia recycle percolation (ARP) process where aqueous ammonia is used as the pretreatment reagent and the reactor has been altered to 15 operate in a continuous mode utilizing a flow-through percolation reactor (Iyer et al. 1996; K i m and Lee 2005a; Oh et al. 2002; W u and Lee 1997; Yoon et al. 1995). Although both the A F E X and A R P processes are effective for the pretreatment of hardwoods and agricultural residues, these processes have been shown to be less effective for softwoods (Iyer et al. 1996; K i m et al. 2000). Dilute acid pretreatment Dilute acid pretreatment is another process that has been investigated for its potential use in bioconversion. In this process, lignocellulosic materials are suspended in water with dilute HC1 or H 2 S 0 4 and autoclaved at 121°C for 1 hr (Saha et al. 2005a). Dilute acid pretreatment is similar to steam explosion with an acid catalyst (Grethlein and Converse 1991). However, dilute acid pretreatment is generally performed at lower temperatures (120-190°C), longer retention times (15 to 60 min), and the feedstock is suspended in an aqueous acid solution (Saha et al. 2005b). Steam explosion Steam explosion has been one of the most thoroughly investigated pretreatment methods for application in the bioconversion process (Boussaid et al. 2000; Grous et al. 1986; Pan et al. 2004; Yang et al. 2002). The process consists of subjecting the feedstock to high-pressure steam at temperatures of 200-240°C for retention times of 20 sec to 5 min, concluding with a final pressure release (explosion) that facilitates the disintegration of the feedstock (Brownell and Saddler 1987). Depending on the nature of the lignocellulosic material, stream pretreatment may require the addition of an acid catalyst such as SO2 and H2SO4 to improve pretreatment efficiency. Even in the absence of an acid catalyst, steam exploded poplar has been shown to be highly susceptible to 16 subsequent enzymatic hydrolysis resulting in a 90% conversion in 24 hr (Brownell and Saddler 1987; Grous et al. 1986). Steam explosion is an inexpensive and effective process for the treatment of hardwoods and agricultural residues, however, due to their recalcitrant structure, softwoods generally require the addition of sulfur dioxide (SO2) or sulfuric acid ( H 2 S O 4 ) to facilitate their degradation (Boussaid et al. 2000; Mais et al. 2002; Robinson et al. 2002; Shahbazi et al. 2005; Shevchenko et al. 2001a; Tucker et al. 2003; Varga et al. 2004; W u et al. 1999). The typical steam explosion pretreatment produces both solid and liquid product streams. Although the solid stream has a high yield of cellulose, the lignin content in the solid stream increases to 36-53% (Schell et al. 1998; Shevchenko et al. 2001a). The hemicelluloses, extractives and low molecular weight lignin fragments are solubilized in the liquid stream (Schell et al. 1998; Shevchenko et al. 2001a). Steam explosion usually results in lignin condensation and enrichment in lignin phenolic groups (Shevchenko et al. 2001a). Changes to the lignin structure w i l l most likely affect enzyme recycling strategies based on the recovery of adsorbed enzymes via desorption of cellulases from hydrolysis residues. The advantages of the steam explosion pretreatment include favourable economics, low water usage, and energy savings (Duff and Murray 1996). Organosolv Organosolv was originally developed as a pulping process for the pulp and paper industry (Pye and Lora 1991; Stockburger 1993). A n aqueous organic solvent mixture with or without an acid catalyst (HC1 or H2SO4) is employed to remove lignin and enrich carbohydrates in the pulp. The organic solvent utilized can be methanol, ethanol, acetone or ethylene glycol. During the process, the solvents can be evaporated, condensed and 17 recycled. Organosolv pretreatment breaks down bonds such as a-aryl ether and arylglycerol-P-aryl ether in the lignin macromolecule (Sarkanen et al. 1981) imparting significant changes in the lignin structure, including increases in phenolic and methoxyl groups and decreases in the average molecular weight (Gilarranz et al. 2000). The solubilized lignin from the process can be subsequently used to produce valuable co-products such as strand binder, adhesives and emulsifiers (Sellers et al. 1994). In addition to producing valuable co-products, work by Neilson et al. (1983) and Pan et al. (2005) has demonstrated that organosolv pretreatment produces a substrate highly susceptible to subsequent enzymatic hydrolysis (Neilson et al. 1983; Pan et al. 2005a). The appealing aspects of utilizing the organosolv process for pretreatment of lignocellulosic substrates include the potential for obtaining valuable lignin-based co-products and more importantly, the production of substrates readily hydrolyzed by cellulases (Neilson et al. 1983; Pan et al. 2005a). 1.3.2 Enzymatic hydrolysis After pretreatment, lignocellulosic substrates are hydrolyzed by cellulases to yield sugars for subsequent fermentation. Depending on the process employed, 80-100% of the hemicelluloses and 20-50% of the lignin are generally solubilized during pretreatment (Lynd et al. 2002b). Enzymatic hydrolysis of cellulose is a complex reaction system, usually involving five steps: 1. transfer of the enzymes from bulk solution to the surface of the cellulose substrates, 2. adsorption of the enzymes onto the cellulose substrate, 3. hydrolysis of cellulose by different cellulase components, 4. transfer of the glucose, cellodextrin and cellobiose from the surface of substrate to the bulk solution, 5. hydrolysis of cellobiose and cellodextrin into glucose by P-glucosidase in the bulk 18 solution (Walker and Wilson 1991); Cellulase adsorption onto substrates is considered to be the rate-limiting step during enzymatic hydrolysis (Converse 1993). Enzymatic hydrolysis is influenced by the structural features of substrates and the nature of cellulases employed in the hydrolysis system (Mansfield et al. 1999). General fungal cellulase action model The cellulose macromolecule is hydrolyzed to glucose monomers by the synergistic action of each enzyme of the cellulase system (Reese 1976). Cellulase enzymes from Trichoderma reesei, the most frequently studied cellulolytic fungus, are composed of three major component enzymes (endoglucanases, exoglucanases and P-glucosidase). The roles of these enzymes are summarized in Table 1. Cellulases from Penicillium sp. are another group of enzymes that have been investigated for their potential use in lignocellulosic hydrolysis (Chaabouni et al. 1994; Krogh et al. 2004). Penicillium cellulases possess major components enzymes corresponding to those produced by T. reesei (Jorgensen et al. 2003). A schematic illustrating the action of each enzyme on the amorphous and crystalline cellulose regions of cellulose is shown in Figure 6 (Lynd et al. 2002b). Endoglucanases attack randomly at internal amorphous sites in the cellulose polysaccharide chain, producing oligosaccharides and new chain ends. Exoglucanases attack the reducing or nonreducing ends of cellulose polysaccharide chains, releasing glucose and cellobiose. P-glucosidase hydrolyzes small oligosaccharides and cellobiose to glucose (Lynd et al. 2002b). Cellulase enzymes can act synergistically to hydrolyze the cellulose substrate. Four kinds of synergism between the component enzymes of T. reesei have been reported during lignocellulosic hydrolysis: 19 1. exo+exo synergism between exoglucanases (Baker et al. 1995), 2. endo+exo synergism between endoglucanases and exoglucanases (Beldman et al. 1988; Medve et al. 1994), 3. exo+P-glucosidase (Lynd et al. 2002a), 4. intramolecular synergism between catalytic domains and cellulose binding domains. The next section wi l l discuss the general functioning of cellulases with particular focus on cellulases from T. reesei and Penicillium sp. Table 1. Cellulase action models and products. Cellulases name Action sites Products E C Number Endoglucanases Soluble and insoluble 1, 4-P-glucan substrates. Oligosaccharides and new chain ends E C 3.2.1.4 Exoglucanases Reducing or non-reducing ends of cellulose chain Cellobiose and glucose E C 3.2.1.74 E C 3.2.1.91 P-glucosidase Soluble cellodextrins and cellobiose Glucose E C 3.2.1.21 E C , Enzyme Commission. 1.3.3 Cellulase enzyme system from Trichoderma reesei In nature, cellulases are produced by both fungi and bacteria. Trichoderma reesei and Penicillium sp. (L i et al. 2005; Murray et al. 2004; Ortega et al. 2000; Palonen et al. 2004; Reczey et al. 1996; Sorensen et al. 2005) have been among the most thoroughly studied cellulolytic fungi because of their potential to be applied to lignocellulosic substrates (Berlin et al. 2005a; Jorgensen et al. 2005; Thygesen et al. 2003; van W y k 1998). 20 Crystalline cellulose Amorphous cellulose Crystalline cellulose C B H I Q C B H II Q Engoglucanase Q beta-glucosidase Cellobiose / Glucose Figure 6. Cellulose hydrolysis by cellulases from Trichoderma reesei. Adapted from (Lynd et al. 2002). The cellulase system of T. reesei has been studied extensively for over 30 years (Bailey and Nevalainen 1981; Lemos et al. 2003; Nevalainen et al. 1980; Olsson et al. 2003). T. reesei belongs to the deuteromycetes and produces three groups of enzymes involved in the hydrolysis of cellulosic substrates. The enzymes include cellobiohydrolases (E.C.3.2.1.91; C B H , exoglucanases), endoglucanases (E.C.3.2.1.4, E G ) and (3-glucosidases (E.C. 3.2.1.21) (Goyal et al. 1991). T. reesei produces at least two 21 exoglucanases ( C B H I and C B H II), five endoglucanases ( E G I, EGII , EGIII, E G I V and E G V ) , and two p-glucosidases ((3-G I and II) (Nogawa et al. 2001). The cellobiohydrolases ( C B H I and C B H II), hydrolyze the reducing and non-reducing ends of cellulose chains respectively. C B H ' s are the major cellulase components of T. reesei, comprising 60% and 20% of the total protein content respectively (Goyal et al. 1991). Endoglucanases (EG), accounting for less than 20% of the total cellulases, randomly cut the cellulose chains at amorphous regions to produce new sites for cellobiohydrolases. 0-glucosidases hydrolyze cellobiose and cellotriose to glucose. Strictly speaking, 0-glucosidases should be considered as an "accessory enzyme" because they do hot act on cellulose directly. Some physical features of the cellulases from Trichoderma reesei are summarized in Table 2. A l l of the component enzymes of the cellulase system of T. reesei have been purified and characterized (Hui et al. 2001; Juhasz et al. 2005; Shoemaker et al. 1983) and some of their genes have been cloned (Jia et al. 1999; Yasokawa et al. 2003; Zandona et al. 2003). Table 2. Physical features of cellulases from Trichoderma reesei. Enzyme Number of total amino acids Amino acids Core Molecular weight (kDa) Isoelectric point (pi) Position of C B D Molar extinction coefficients C B H I 497 430 59-68 3.5-4.2 C 78,800M"' C B H II 447 367 50-58 5.1-6.3 N 92,000M"' E G I 437 368 50-55 4.0-6.0 C 67,000M"' EGII 397 327 48 5.5 N 78,000M"' EGIII, 218 218 25 7.5 - 38,200M"' E G V 225 166 23 - C -p-G I 713 713 75 7.4-7.8 - -Data adapted from (Medve 1997; Valjamae et al. 2001). 22 The various component enzymes of the T. reesei cellulase system have been shown to act synergistically to hydrolyze microcrystalline cellulose. C B H I and C B H II have been shown to demonstrate exo-exo synergism (Medve et al. 1994), and endo-exo synergism also has been found during the hydrolysis of Avice l and bacterial microcrystalline cellulose ( B M C C ) (Henrissat et al. 1985). The activity of C B H is strongly inhibited by the reaction product cellobiose (end product inhibition). Results from FT-IR and circular dichroism studies have shown that cellobiose combines with the tryptophan residue located at the active site of C B H resulting in steric hindrance that prevents cellulose chains from diffusing into the active site of the enzyme (Zhao et al. 2004). Furthermore, cellobiose binding has also been shown to cause a change in the conformation of C B H , thus reducing its activity during hydrolysis (Zhao et al. 2004). Similar to T. reesei, Penicillium sp. has also demonstrated the ability to produce a complete cellulase enzyme system (Brown et al. 1987). 1.3.4 Cellulases from Penicillium sp. In some cases, it has been shown that cellulases from Penicillium sp. are more effective at hydrolyzing lignocellulosics than those produced by Trichoderma reesei (Chaabouni et al. 1994; Krogh et al. 2004). Although there have been differences in activities between the two cellulase systems, they are quite similar with corresponding enzymes participating in the various roles of the hydrolysis process. Two exoglucanases ( C B H a and C B H b ) , three endoglucanases (EGa, E G b l a n d EGb2), and one xylanase have been purified and characterized from a culture of Penicillium brasilianum (Jorgensen et al. 2003). It has been postulated that C B H a and C B H b are analogous to the C B H II and C B H I from T. reesei in terms of their hydrolytic properties 23 (Jorgensen et al. 2003). They both cannot hydrolyze carboxymethyl cellulose ( C M C ) , but can convert Av ice l , mainly to cellobiose. The E G a (21kDa) of P. Brasilianum is similar in size to E G III of T. reesei. Furthermore, E G a and E G III possess C M C activity and hydrolyze Avice l , producing cellobiose, glucose and cellotriose (Jorgensen et al. 2003; Karlsson et al. 2002). E G b 2 also has a similar sequence to E G II CT. reesei). The physical features of Penicillium cellulases are shown in Table 3. In addition to having corresponding enzymes to those of T. reesei, the cellulases of Penicillium are also thought to work in unison to degrade lignocellulosic substrates. Synergism was found between the two exoglucanases ( C B H I and C B H I I ) from Penicillium occitanis during hydrolysis of Av ice l and H3PO4 swollen cellulose (Limam et al. 1995). Cellulase components from other Penicillium sp. {Penicillium pinophilium, Penicillium funiculosum) have also been investigated for their synergism during lignocellulosic hydrolysis (Mishra and Rao 1988; Morris 1982; Wood 1992; Wood and Mccrae 1986; Wood et al. 1989; Woodward 1991). Table 3. Physical features of cellulases from Penicillium sp. Enzyme Molecular weight (kDa) Isoelectric point (PD Total protein (%) Adsorption (%) C B H a 63 18 64 C B H b 70 4.1 25 79 E G a 21 5.4 21 9 E G b l 53 4.1 4 4 E G b2 54 4.4 10 72 X Y L >9 >9 - 4 Adapted from (Jorgensen et al. 2003), Adsorption (%)= percent adsorption of enzymes on cellulose after 60 min of incubation at 4°C. The enzyme/substrate ratio was 0.2 umol g"1 Avice l . 24 Comparisons between the hydrolytic performance of cellulase preparations from Penicillium sp. and those derived from T. reesei revealed that T. reesei cellulases typically possess higher "specific activities" measured in filter paper units than those from Penicillium sp. (170 F P U vs. 120 FPU) (Castellanos et al. 1995). T. reesei cellulase preparations have lower beta-glucosidase activity (0.8 IU vs. 56 IU) (Castellanos et al. 1995). Therefore, T. reesei requires additional P-glucosidase to decrease end product inhibition, thus improving the efficiency of cellulose hydrolysis. Previous studies have demonstrated that Penicillium cellulases were more effective in cellulose hydrolysis, however, upon supplementation with beta-glucosidase, T. reesei cellulases showed similar hydrolytic performance to Penicillium on cellolginin, steam pretreated spruce and Sigmacell (Castellanos et al. 1995; Jorgensen and Olsson 2006). Enzymatic factors that limit enzymatic hydrolysis Enzymatic factors that limit hydrolysis include lack of synergism, non-productive binding and end product inhibition (Mansfield et al. 1999). A major enzymatic factor that limits hydrolysis by cellulases is end product inhibition (Xiao et al. 2004). Bezerra and Dias (2005) showed that cellobiose is a strong inhibitor of both crude cellulases and exoglucanase (Cel7A). The strategies suggested to reduce end product inhibition during enzymatic hydrolysis include increasing cellulase loading, supplementing with P-glucosidase, removing sugars during hydrolysis by filtration (Gan et al. 2005) and using simultaneous saccharification and fermentation (SSF) (Vinzant et al. 1994). In addition to enzyme factors, the properties of the substrate are also known to play a major role in enzymatic hydrolysis. 25 Substrate factors that limit enzymatic hydrolysis Due to their inherent structural characteristics, lignocellulosic substrates are naturally resistant to enzymatic hydrolysis. Substrate factors that limit enzymatic hydrolysis are summarized in Table 4 (Converse 1993; Mansfield et al. 1999). O f the factors that affect hydrolysis, it is still unclear i f the degree of polymerization (DP) of cellulose in the substrate plays a significant role. Puri (1984) showed that the hydrolysis rate and extent of saccharification was controlled by cellulose D P and surface area, while experiments by Sinitsyn and others indicated the cellulose D P did not affect hydrolysis (Sinitsyn et al. 1991). It has been assumed that a high level of crystallinity results in low rates of hydrolysis, since amorphous cellulose is easier to hydrolyze (Converse 1993). Puis and Wood (1991) showed that the crystallinity index of a-cellulose remained constant during a 7 day hydrolysis treatment, therefore, it was concluded that there was no direct relationship between crystallinity and hydrolysis. In contrast to Puis and Wood (1991), Mansfield et al. (1999) hypothesized that the enzymatic hydrolysis is most likely governed by other substrate associated factors such as accessible surface area, particle size and lignin distribution, rather than crystallinity itself. Accessible surface area is another significant factor that affects enzymatic hydrolysis. Mooney et al. (1999) showed that the increased specific surface area provided by the presence of small fibers and fines enhanced hydrolysis rates and yields. In addition to physical attributes of cellulose such as crystallinity, lignin also has a profound influence on the hydrolyzability of lignocellulosic substrates. Koullas et al. (1992) found an empirical correlation between the extent of hydrolysis and lignin content, and suggested a complete hydrolysis of cellulose would be possible at a 26 lignin content of less than 10% (Koullas et al. 1992). It has been hypothesized that lignin has two major effects on cellulose hydrolysis: Lignin can cover the surface of cellulose or prevent the swelling of fibers thereby reducing the accessibility of cellulases to cellulose; and it can bind with cellulase enzymes thus limiting their action on cellulose (Mansfield et al. 1999). Table 4. Substrate and enzyme factors limiting enzymatic hydrolysis. Impact factors Characteristics Reference: Degree of polymerization Crystallinity Accessible surface area Lign in Inhibition Synergism Non-productive binding D P affects the extent of hydrolysis, but the (Sinitsyn et al. 1991; correlation between hydrolysis and D P are Zhang and L y n d 2005) low. Amorphous cellulose hydrolyzes rapidly and crystalline hydrolyzes slowly. (Koullas et al. 1992; Puis and Wood 1991) Increasing the accessible surface area Mooney(1999) improves lignocellulosic hydrolysis. L ign in reduces the accessibility of cellulose (Koullas et al. 1992) to enzymes, and increases non-productive binding with enzymes. Cellulases undergo strong end product (Bezerra and Dias 2005; inhibition thus requiring supplemental beta- X i a o et al. 2004) glucosidase. Exoglucanases ( C B H I and CBHII ) and (Baker et al. endoglucanases (EGI and EGII) act Beldman et al synergistically on lignocellulosic substrates. 1995; 1988; Medve et al. 1994; Riedel and Bronnenmeier 1998; Watson et al. 2002) Non-productive binding decreases the (Gan et al. 2003; Nidetzky accessibility of cellulases to cellulose. et al. 1994; Valjamae et al. 1998) 1.3.5 Fermentation Following enzymatic hydrolysis, fermentation is the phase of the bioconversion process where sugars from the hydrolysis step are fermented to ethanol. Three approaches have 27 been investigated for ethanol fermentation: separate hydrolysis and fermentation (SHF), simultaneous saccharification and fermentation (SSF) and direct microbial conversion ( D M C ) (Table 5). Separate hydrolysis and fermentation (SHF) is the most frequently applied process in bioconversion, and involves the fermentation of a pre-hydrolyzed substrate using Saccharomyces cerevisiae or Zymomonas mobilis (Eklund and Zacchi 1995; Holtzapple et al. 1994). Factors limiting the effectiveness of S H F include end product inhibition and the requirement for high enzyme loadings with low substrate concentrations. In simultaneous saccharification and fermentation (SSF), the enzymatic hydrolysis and fermentation processes are coupled together to minimize end product inhibition, since sugars are simultaneously produced and removed during hydrolysis/fermentation. SSF has been shown to improve hydrolysis yield and reduce required enzyme loadings (Wyman et al. 1992), although this a compromise in the temperature and p H optimum for the combined process. Direct microbial conversion integrates enzyme production, lignocellulosic hydrolysis and fermentation (Ravinder et al. 2000), thus simplifying the process and reducing the capital cost. However, the problems of this process include low ethanol tolerance and an increase in the production of by-products. Although SSF is currently the most promising approach for bioconversion to ethanol (Wingren et al. 2003), the high cost of cellulases remains a limitation in commercializing an SSF process. Laboratory studies on direct microbial conversion with Klebsiella oxytoca show promise in commercializing cellulose to ethanol processes (Zhou et al. 2001; Zhou and Ingram 2001). 28 Table 5. Fermentation approaches for bioconversion. Fermentation Characteristics Reference: approaches Separate hydrolysis Optimal conditions for hydrolysis and (Alfani et al. 2000; Saha et al. fermentation (SHF) fermentation respectively. Strong end 2005b) product inhibition, lower substrate concentration and higher enzyme loading. Simultaneous Reduces end product inhibition, lower (Borden et al. 2000; Chang et saccharification and enzyme loadings, and higher product al. 2001a; Deshpande 1992; fermentation (SSF) yields. Duff et al. 1994; Stenberg et al. 2000) Direct Microbia l Simplified process, low ethanol (Ravinderetal . 2000; South et conversion ( P M C ) tolerance, more by-products. al. 1993) 1.4 Strategies to reduce the cost of enzymatic hydrolysis 1.4.1 Improving cellulase production Bioconversion of lignocellulosics to ethanol is an intriguing option for securing a renewable and sustainable energy supply. However, one of the bottlenecks in this process remains the high cost of enzymes utilized in the hydrolysis step (Steele et al. 2005). Considerable efforts are currently being devoted to reducing the cost of enzyme production. Several factors have a significant role in microbial cellulase production including carbon sources, p H , temperature, and agitation. Most of the work in this area has focused on Trichoderma reesei since it is considered as the microorganism with the greatest potential for the production of cellulases (Egyhazi et al. 2004; Gadgil et al. 1995; Juhasz et al. 2004; Kadam and Keutzer 1995). The carbon source has been shown to be one of the most cost intensive aspects of cellulase production by T. reesei (Ryu and Mandels 1980). Pure cellulose (Solka-Floc, Avicel) and lactose have been often used as carbon sources to produce cellulases 29 (Esterbauer et al. 1991; Hayward et al. 2000; Lejeune and Baron 1995). The cost of cellulase production can be reduced by using lignocellulosic materials such as wheat straw (Maheswari et al. 1993), sawdust, corn cob residue (Xia and Shen 2004) or bagasse (Bigelow and Wyman 2002) as the carbon source. L o et al. (2005) demonstrated a production rate of 0.046 F P U / g cells per hour by T. reesei using sawdust hydrolysate compared to a rate of to 0.017 F P U / g cells per hour on cellulose controls (Lo et al. 2005; Maheswari et al. 1993). Mutation of cellulolytic microbial strains has also been employed frequently to achieve improvements in cellulase production (Araujo et al. 1991; Chand et al. 2005; Withers et al. 1992). It has been shown that a mutant of T. reesei capable of a 1.5 fold increase in cellulase production compared to the wi ld type, could be created by using a combination of ultraviolet light and sodium nitrite (Gadgil et al. 1995). In later work, enhanced production of cellulase by T. reesei was also attained by using sub-lethal concentrations of nitrosoguanidine and ethidium bromide for a 30 min to 1 hr (Chand et al. 2005). In addition to mutations and the alteration of the carbon source, cellulase production by T. reesei can also be enhanced at low p H and with increased agitation. On either a cellulose (Bailey et al. 1993) or lactose (Xiong et al. 2004) based growth medium, it was shown that cellulase production was favoured at p H 4.0 while xylanase was favoured at p H 6.0-7.0. Buffering the growth medium at a p H of 3.5 using citrate resulted in a two-fold increase in cellulase production compared to the system without buffer (Kadam and Keutzer 1995). Culture agitation influences the production of cellulases by T. reesei, primarily due to the improvement in aeration at higher agitation rates. Lejeune and Baron (1995) showed that agitation rates of 130-200 rpm were beneficial to cellulase production, while higher rates were detrimental (Lejeune and 30 Baron 1995). While boosting cellulase production is an effective approach to reduce enzyme costs for bioconversion, an alternative scheme is to improve cellulase specific activity. 1.4.2 Improving cellulase activity A range of methods have been applied to improve the specific activity of cellulases, including increasing thermostability, decreasing non-productive binding of the enzyme, decreasing product inhibition and increasing K c a t (the number of substrate molecule each enzyme site converts to product per unit time). Increasing the thermostability of cellulases can be quite beneficial to improving specific activity as a 10°C increase in temperature can result in a 2-3 fold increase in the reaction rate (Mozhaev et al. 1988). Ozawa reported that an endo-l,4-R-glucanase from Bacillus sp. was thermostabilized by the replacement of both A s n l 7 9 and A s p l 9 4 with lysine by site-directed mutagenesis (Ozawa et al. 2001). The residual activity of the thermostabilized enzyme was improved from 15% to 50% at 80°C after 30 min incubation. Both thermostability and binding of cellulases are integral to their overall activity on lignocellulosic substrates. Enzymatic hydrolysis of lignocellulosics is a heterogeneous phase process where cellulase enzymes react with solid substrates in a liquid environment. Therefore, effective hydrolysis requires the direct binding of cellulases to cellulose. "Non-productive" binding occurs when cellulases bind to lignin or cellobiose thus decreasing the overall efficiency of the hydrolysis. In an effort to improve specific activity, investigations striving to decrease non-productive binding have focused on either mutation (Becker et al. 2001) or the selection of enzymes with a lower binding for lignin (Berlin et al. 2005b). 31 Another major factor that affects cellulase activity is end product inhibition. Traditionally, cellulases are inhibited by cellobiose, glucose, and by-products of the pretreatment process, such as furfural, acetic acid, lactic acid and aldehydes (Gruno et al. 2004; Iyer and Lee 1999; Szengyel and Zacchi 2000). Enzymes can be modified to reduce or eliminate end product inhibition, for example, the enzyme dihydrodipicolinate synthase, catalyzing the first step in lysine synthesis, has been successfully mutated to eliminate end product inhibition by lysine (Shaver et al. 1996). B y identifying the mechanisms whereby cellulases bind inhibitors, key amino acids residues responsible for inhibitor binding could be substituted, thus reducing or eliminating end product inhibition from cellulases. Another approach to improving enzyme activity is to increase the Kcat of the reaction. Kcat, often referred to,as turnover number, represents the maximum number of substrate molecules that can be converted to products per enzyme molecule per unit time. Increasing Kcat w i l l improve the enzyme's catalytic efficiency (Kcat/Km). For example, alkaline phosphatase from Escherichia coli exhibited a 6 fold increase in Kca t /Km and a 35 fold increase of specific activity compared to the wi ld type after modification using site-directed mutagenesis (Mandecki et al. 1991). Decreasing non-productive binding and end product inhibition, and increasing thermal stability and turnover number show great potential for improving cellulase activity. A s mentioned earlier, Novozymes increased cellulase activity by improving thermostability and specific activity of cellulases under a contract with the U.S . D O E (Cherry 2003). The improved cellulase activity and reduced enzyme production costs have resulted in considerable reductions in the cost of the enzymatic hydrolysis stage. However, it has been stated that a further 2-3 fold enzyme cost reduction is necessary to reach cost targets 32 for the eventual commercialization of bioconversion of lignocellulosics to ethanol (Cherry 2003). Increasing cellulase activity and production are two possible approaches to improve lignocellulosic hydrolysis and reduce enzyme cost. On the other hand, the addition of surfactants during lignocellulosic hydrolysis can also reduce enzyme loading and improve hydrolysis yields (Eriksson et al. 2002) 1.4.3 Addition of surfactants to improve enzymatic hydrolysis There have been many reports of improved enzymatic hydrolysis upon addition of surfactants (Eriksson et al. 2002; Helle et al. 1993; Kaar and Holtzapple 1998; K i m and Chun 2004). One of the main surfactants that have been investigated for its potential to improve hydrolysis is Tween 80 (polyoxyethylene sorbitan monolaurate). The hydrolysis yield of newspaper was increased from 41% to 55% by the addition of Tween 80 (Castanon and Wilke 1981). Ooshima (1986) later showed that Tween 20 enhanced the enzymatic hydrolysis of Av ice l from 33% to 45% within 72 hr. Many ideas have been presented to explain the effectiveness of surfactants such as Tween in improving hydrolysis, including decreasing the adsorption of cellulases to substrates (Ooshima et al. 1986), increasing the desorption of enzymes from their binding sites (Park et al. 1992), and enzyme stabilization (Kaar and Holtzapple 1998). Eriksson et al. (2002) found that regardless of the addition of Tween 20, cellulase Ce l7A ( C B H I) retained 100% of its activity during 96 hr of incubation. Therefore, it was unlikely that surfactants function as stabilizers during hydrolysis. Eriksson et al (2002) also observed a 15% improvement in the hydrolysis of steam exploded spruce compared to only a 5% improvement during the hydrolysis of Avice l with the addition of Tween. Therefore, it was suspected that lignin played a role in the effect of the surfactant. Furthermore, 33 Ericksson et al (2002) also noted that proteins such as bovine serum albumin (BSA) could be used as a substitute for Tween. This was confirmed in later work investigating the hydrolysis of steam exploded Douglas fir (Pan et al. 2005b). It was concluded that improvements in hydrolysis resulting from surfactant addition were due to a decrease in non-productive binding of cellulases (Eriksson et al. 2002). In addition to aiding in the enzyme hydrolysis stages, Tween 20 and 80 have also been shown to enhance simultaneous saccharification and fermentation (SSF). Lee et al. (1996) compared various surfactants as additives during SSF and reported that Tween 80 addition gave the best performance for ethanol production from steam exploded wood (Lee et al. 1996). Although Triton X-100 improved the enzymatic hydrolysis step, it had a negative effect on fermentation. Alkasrawi et al (2003) found that Tween 20 (2.5 g/L) increased ethanol yield by 8% in simultaneous saccharification and fermentation (SSF) of steam explosion pretreated spruce. Similar to the enzymatic hydrolysis results, the most likely explanation for the improvement in SSF performance with the addition of Tween is the reduction of non-productive binding of the enzyme to lignin (Eriksson et al. 2002). It was suggested that the hydrophobic sites on the lignin surface may be occupied by Tween which could lead to the displacement of adsorbed enzymes and thus a reduction in non-productive enzyme adsorption (Eriksson et al. 2002). In conclusion, the addition of surfactants would improve enzymatic hydrolysis by reducing the non-productive binding of cellulases and further improve the economics of the enzymatic hydrolysis stage when used along with recovery and reuse of enzymes after hydrolysis. 34 1.4.4 Recovery and reuse of cellulases After a typical enzymatic hydrolysis, cellulases are distributed between the liquid phases as free enzymes and bound enzymes on the solid lignocellulosic hydrolysis residue (Boussaid and Saddler 1999). It has been shown that both free and bound enzymes are recoverable from the two phases (Lee et al. 1995). Early work establishing that cellulases from T. reesei are highly stable and have a high affinity for cellulose (Reese and Mandels 1980), has suggested the possibility of recovering and reusing cellulases in the bioconversion process (Castanon and Wi lke 1980). Free enzymes have been found to be effectively recycled by readsorption to fresh substrates after a complete hydrolysis of SO2 catalyzed steam exploded Eucalyptus (0.4% lignin after alkali and peroxide treatment) (Ramos et al. 1993). Similar results were reported during the evaluation of cellulase recycling in the hydrolysis of steam exploded Birch (4% lignin after alkali and peroxide treatment) (Lee et al. 1995). Lee et al. (1995) recycled the bound enzymes in the solid phase by collecting the lignin-rich post-hydrolysis residue. Lee et al. (1995) hypothesized that, upon addition of fresh substrates, the bound enzyme could transfer from the lignin-rich hydrolysis residue to the added substrate. These early studies indicated that the recycling of bound enzymes would likely be governed mainly by the enzyme desorption process, while the free enzymes could be recovered by readsorption. Thus, any factor affecting the adsorption and desorption processes w i l l affect the efficiency of cellulase recycling during lignocellulosic hydrolysis. 1.4.4.1 Adsorption of cellulase on substrates The adsorption of cellulase enzymes on the surface of lignocellulosics is very important for the effective hydrolysis of cellulose. The adsorption rate of cellulase on cellulose 35 substrates has been directly correlated with the rate of cellulase hydrolysis (Converse et al. 1990). Both fungal and bacterial cellulases typically have a catalytic domain and a binding domain connected by a flexible linker (Gilkes et al. 1992; Tomme et al. 1995). Most cellulase components from T. reesei have been shown to possess a cellulose binding domain (CBD) (Arunachalam and Kell is 1996). The purpose of the C B D is to improve to the binding of the catalytic domain to crystalline cellulose to enhance hydrolysis. This is especially important, since cellulases function in heterogeneous phase systems where the liquid phase enzymes hydrolyze solid lignocellulosics. Using papain to remove the C B D from C B H I decreased the adsorption of C B H I on Avice l (crystalline cellulose) from 0.293 nmol/mg to 0.103 nmol/mg, however, the adsorption of C B H I on amorphous cellulose was unaffected (Tomme et al. 1988). Similarly, in the absence of the C B D , the catalytic domain lost 50-60% of its ability to hydrolyze insoluble cellulose but retained its activity on soluble cellulose (Tomme et al. 1988). Furthermore, the hydrolytic activities of C B H I and endoglucanases (Lemos et al. 2003) have also been shown to be improved upon the addition of isolated C B D from T. reesei. Further insight into the substrate binding character of cellulases can be gained by a discussion of the driving forces that affect general protein adsorption. Adsorption kinetics and isotherm In general, the adsorption of protein at constant temperature and pressure occurs at a low value of Gibbs free energy G as shown below: \ l l s G = AadsH-TAadsS<0 Where H , S and T are the enthalpy, entropy and absolute temperature respectively (Haynes et al. 1994). 36 The adsorption of cellulases to a lignocellulosic substrate can be affected by the following factors: hydration changes (hydrophobic interactions), electrostatic interactions, van der Waals forces and hydrogen bonding (Norde 1998). If the surfaces of the cellulase and lignocellulosic substrate are both hydrophilic, their hydration is favorable in the system, thus decreasing the entropy of water molecules, resulting in an increase in G , and less favourable adsorption. However, when both surfaces are hydrophobic, the adsorption process is promoted, since dehydration of non-polar components in an aqueous phase releases water molecules resulting in an increase in entropy and a lower Gibbs free energy. Creagh et al. (1996) postulated that substrate dehydration (i.e. hydrophobic interaction) is the major driving force for binding of C B D c e x to bacterial microcrystalline cellulose ( B M C C ) (Creagh et al. 1996). Electrostatic interactions can also occur between the charged amino acid residues in proteins and the charged groups on lignocellulose. According to Creagh et al (1996), cellulose is a neutral polymer, thus electrostatic interaction cannot be regarded as a driving force for adsorption of C B D c e x to B M C C (Creagh et al. 1996). Structural rearrangements in the protein may affect intramolecular hydrogen bonds, resulting in either an increase or a decrease in protein adsorption, depending on the balance between energetically favorable interactions and protein conformational entropy (Norde 1998). The classic Langmuir model of gas adsorption can be applied to protein adsorption in solution on solid particles providing that four requirements are met: 1. The protein solution is sufficiently dilute, 2. Only one class of adsorption site is present, 3. There are no lateral interactions between adsorbed proteins, 37 4. The adsorption is reversible (Andrade 1985; Basmadjian 1997; Fogler 1999). Adsorption of the major cellulases components on different lignocellulosic substrates at dilute concentrations has been shown to follow the Langmuir adsorption isotherm (Gilkes et al. 1992). The Langmuir isotherm has the following form: r _r m a x ^c i+*:c Where: C=concentration of unadsorbed cellulase in bulk solution (mg/mL) r concentration of adsorbed cellulase (mg/g insoluble substrate) r ^ m a x = t n e maximal adsorbed cellulase (mg/g insoluble substrate) K = Langmuir constant (mL/mg enzyme) The linearized form of the Langmuir model can be expressed as: 1 1 1 1 — = + • — rmax and K could be determined by the non-linear curve fit or from the linear plot of -p 1 versus — . C If two classes of adsorption sites on lignocellulosic substrates are present, a two component Langmuir model may be used to fit the adsorption process (Medve et al. 1997). r _ r ^ c F2K2C l + K{C 1 + K2C Where Kx and T{ are the Langmuir constant and maximal adsorbed cellulase on site 1; K2 and T2 are the Langmuir constant and maximal adsorbed cellulase on site 2. Previous work in the Forest Products Biotechnology group examined the adsorption isotherms of cellulase (Celluclast) and C B D c e x from Cellulomonas fimi (C. fimi) on different lignocellulosic substrates (Lee 1994). The adsorption parameters are shown in 38 Table 6. The experimental results fit the Langmuir isotherm model, thus indicating the Langmuir model is suitable for analyzing cellulase adsorption. However, Medve et al. (1997) suggested that the two-site Langmuir isotherm model and the Langmuir-Freundlich isotherm model may be more suitable than the one-site Langmuir model to analyze the adsorption of C B H I and C B H II on Avice l (Medve et al. 1997). The use of adsorption isotherm models may give additional information for cellulases adsorption on lignin. Table 6. Adsorption parameters for Celluclast and C B D c e x on different cellulosic substrates. Substrates rimx of Celluclast K of Celluclast (mg enzyme /mg cellulose) (mL/mg enzyme) r m a x 0 f CBDcex (mg enzyme /mg cellulose) K o f CBDcex (mL/mg enzyme) Acid Swollen Avicel (ASA) 0.513 Air-dried A S A 0.192 Oven-dried A S A 0.113 Avicel 0.111 Water-washed steam-exploded 0.214 Birch (WB) Alkaline-washed W B 0.237 Peroxide-treated W B 0.233 Solka Floe 0.048 1.0 14.1 3.5 0.6 2.1 2.8 1.9 1.3 0.396 0.186 0.077 0.05 0.374 0.434 0.321 0.051 12.9 22.4 29.5 13.4 13.9 18.4 43.8 14.8 Data from Lee (Lee 1994). Adsorption of cellulases onto lignin Although there have been several studies focusing on the adsorption of cellulases from T. reesei on different cellulosic substrates, only a few studies have investigated the interaction between cellulases and lignin (Palonen et al. 2004). Palonen et al. (2004) 39 showed that significant amounts of C B H I and E G II could bind to alkali and enzymatic lignin. The Langmuir adsorption of the binding of C B H I were 5.5 Lg" 1 (partition coefficients) on steam exploded spruce lignin, 1.7 Lg" 1 on alkali lignin and 0.6 Lg" 1 on enzymatic lignin (Palonen et al. 2004). The low amount of adsorption on the enzymatic lignin preparation was most likely due to a lack of sites for cellulase binding, since the lignin was isolated using an cellulolytic enzyme hydrolysis process. Residual lignin in softwood Kraft pulps has been shown to bind to cellulases, thus reducing their accessibility to cellulose (Mooney 1998). It is likely that the lignin content of lignocellulosic substrates wi l l play a significant role in the development of strategies to recycle cellulase enzymes during the hydrolysis process. 1.4.4.2 Cellulase recycling during lignocellulosic hydrolysis Previously it was estimated that the enzymatic hydrolysis step accounted for up to 60% of the cost of bioconversion of lignocellulosic biomass to ethanol (Gregg et al. 1998). In order to move this technology towards commercialization, the cost of the enzymes used in hydrolysis needed to be reduced significantly. Therefore, recovery and reuse of cellulases from the liquid and solid hydrolysis streams is worth pursuing. As mentioned earlier, several previous studies in the Forest Products Biotechnology group had evaluated specific aspects of enzyme recycling during cellulose hydrolysis (Lee et al. 1995; Lee et al. 1994; Meshartree et al. 1987; Ramos et al. 1993; Ramos and Saddler 1994). These studies showed that, during hydrolysis, enzymes are quickly adsorbed on the substrate. However, as the process proceeds, some cellulases gradually desorb into the bulk solution. Therefore, after hydrolysis, cellulases are distributed between the residual solid substrate and soluble phase. This past work suggested that an 40 approach for recycling cellulases should include strategies for recovering enzymes from both the liquid and solid phases. The recovery of free enzymes from the bulk solution has been achieved in the laboratory by applying methods (Table 7) such as ultrafiltration (Lu et al. 2002) and the addition of fresh substrates (Castanon and Wilke 1980; Lee et al. 1995). Ultrafiltration has also been shown to be a viable approach to recycle cellulases (Lu et al. 2002; Mores et al. 2001; Tan et al. 1986). Mores suggested a combined sedimentation and membrane filtration process could recycle 75% of the active enzymes in hydrolysis of ground Yel low poplar (Mores et al. 2001). B y exploiting the natural affinity of cellulases for cellulose, enzymes in the hydrolysate could also be recovered by readsorption onto fresh substrates (Castanon and Wilke 1980). A n enzyme mixture applied to hydrolyze steam-exploded Eucalyptus viminalis was recovered successfully using readsorption and applied for five consecutive cycles (Ramos and Saddler 1994). Castanon and Wi lke reported a 40% recovery of free enzymes in solution after typical hydrolysis, via the addition of fresh substrate (Castanon and Wilke 1980). Bound enzymes have also been recovered from the solid phase via the addition of fresh substrates, desorption by buffers and surfactants (Jackson et al. 1996) and by recovering enzyme-containing lignin rich post hydrolysis residue (Ramos and Saddler 1994). Enzymes adsorbed on lignocellulosic hydrolysis residues possessed lower activity when re-combined with fresh substrates (Girard and Converse 1993). For example, recycling of enzymes adsorbed on hydrolysis residues was carried out five times for alkali and peroxide-treated steam-exploded Birch (PB, 4% Klason lignin), however, the same recycling strategy was not effective on steam-exploded Birch (32% Klason lignin), as 41 conversion dropped from 68.4% to 24.9% after 4 hydrolysis rounds. It has also been shown that 45% of the bound cellulases could be recovered from crystalline cellulose by simply increasing the p H from 5.0 to 10.0 (Otter et al. 1984). The amount of recovered cellulases could be increased further to 65% by adding a detergent (Otter et al. 1984). Rao et al. (1983) reported that 90% of the bound cellulases (from Penicillium funiculosum) could be recovered during hydrolysis of bagasse by adding 0.4% Tween 80 (Rao et al. 1983). Among the enzyme recycling methods applied to both the solid and liquid products of hydrolysis, only ultrafiltration was capable of recovering all of the cellulase component enzymes including P-glucosidases, endoglucanases and exoglucanases (Lu et al., 2002). The addition of fresh substrates, and/or surfactants were not capable of recovering p-glucosidases. p-glucosidases do not typically adsorb onto lignocellulosic substrates, most likely because their primary role is to hydrolyze soluble cellobiose to glucose. Up to date, there has been limited published work or strategies developed to recycle P-glucosidases separately during lignocellulosic hydrolysis (Ong et al. 1993), albeit there are many reports on immobilized glucosidase being used in starch hydrolysis (Ram and Venkatasubramanian 1982; Rani et al. 2000; Ray and Majumdar 1994). A few reports have attempted to immobilize P-glucosidase on alumina for cellulose hydrolysis (Sundstrom et al. 1981). However, these approaches have not combined immobilization with recycling. 42 Table 7. Cellulase recycling strategies during the hydrolysis of lignocellulosic materials. Cellulase recycling approaches Characteristics Reference: Desorption of bound cellulases Readsorption of free cellulases Ultra-filtration of free cellulases Recovery of bound cellulases with lignin residues Immobilization of cellulases and 0-glucosidase Elute enzymes from lignin residues by water, acetate and phosphate buffers, urea, glycerol solution, and surfactants Free enzymes were recovered readsorption onto fresh substrates. Recycle both cellulases and P-glucosidase The bound cellulases were recovered together with lignin rich hydrolysis residues Immobilization of cellulases likely reduced the access of enzymes to insoluble cellulose. Immobilized P-glucosidase could be recycled. (Deshpande and Eriksson 1984) (Castanon and Wilke 1980; Girard and Converse 1993; Vallander and Eriksson 1987) (Knutsen and Davis 2004; Steele et al. 2005) (Ramos and Saddler 1994) (Busto et al. 1998; Dourado et al. 2002; Fujishima et al. 1991; Mao et al. 2006; Saville et al. 2004; Shen and Xia 2004; Tu et al. 2006; Yuan et al. 1999) 1.4.5 Limitations of current cellulase recycling strategies Although there has been significant work devoted to the development of recycling strategies, many challenges remain that hinder the application of the existing technologies to the hydrolysis process. One of the major obstacles of recycling is the difficulty in recovering P-glucosidase, since it does not bind to lignocellulosic substrates. The second challenge is that most of the existing recycling strategies involve the removal of cellulase-laden hydrolysis residues that also contain a high amount of lignin, therefore, the lignin concentration increases with each recycling round. It has been reported that the lignin content in the hydrolysis step reached 70% for steam exploded Birch after recycling the residues for five consecutive rounds (Lee et al. 1995). The increase in lignin content during recycling decreases the efficiency of subsequent hydrolysis steps, 43 especially in the case of softwood residues. Although ultrafiltration seems to be the most attractive method assessed so far, this technique requires a specialized membrane to collect the cellulases. The high-cost of this membrane (12 cents/gal EtOH) (Mores et al. 2001), combined with the fact that it cannot be reused after several cycles due to fouling by protein and lignin, likely renders this process inefficient and impractical for large scale applications. In addition to practical limitations, previous studies investigating recycling strategies do not address some critical issues such as reducing enzyme-lignin interactions that are necessary for achieving further improvements of current recycling technologies. To date, most studies on cellulase recycling have dealt with evaluating the feasibility of recycling to improve the economics of the hydrolysis process. However, only a few studies have elucidated the mechanisms responsible for the loss in enzyme activity during each recycling step (Castanon and Wilke 1980; Ramos and Saddler 1994). Although lignin has been shown to be a major factor limiting the hydrolysis of lignocellulosic substrates (Lee et al. 1995; L u et al. 2002), there is no direct evidence supporting non-productive irreversible binding between cellulases and lignin. Most of the previous studies have assessed the ability to recover cellulases from hardwood biomass such as Aspen (Meshartree et al. 1987), Birch (Lee et al. 1995) and Eucalyptus (Ramos and Saddler 1994). However, differences in lignin content and structure of softwoods are expected to have a considerable effect on enzyme recycling. 1.5 Thesis Objectives Increasing enzyme production yield, specific activity or recycling cellulase enzymes have all been suggested as methods to reduce the cost of cellulases required for lignocellulosic hydrolysis in a biomass to ethanol process (Cherry 2003; Lee et al. 1995). Although 44 several industrial groups such as Genencor International, Novozymes, and Iogen are continually striving to increase the productivity and specific activity of cellulolytic microorganisms and enzymes, current research efforts assessing the ability to recover and reuse cellulase enzymes have been quite limited. Softwoods, especially Douglas fir and Lodgepole pine, are an abundant, readily available natural resource in British Columbia. Pretreated Douglas fir and Lodgepole pine substrates were the main materials employed in this study. Steam explosion pretreatment possesses great potential to be used in the bioconversion process since it is a quick, simple, and effective method capable of processing lignocellulosics. However, in the case of softwoods, steam explosion results in a substrate with high lignin content (-45%), thus presenting a challenge for subsequent hydrolysis and enzyme recycling. In contrast to steam explosion, it has been demonstrated that organosolv pretreated softwoods are readily hydrolysable by cellulases (Pan et al. 2005a). It was expected that the relatively low lignin content (10-25%) of organosolv pretreated substrates (acetic acid or ethanol) would result in substrates more amenable to the application of enzyme recycling. The primary objective of this thesis was to develop effective recycling strategies to recover cellulase enzymes during, and after the completion of hydrolysis and to better understand the enzyme-substrate interactions during hydrolysis. Most earlier attempts to recover cellulases have focused on the recovery of cellulases from the residual hydrolyzed substrates (Otter et al. 1989) or from the hydrolysis supernatant (Ramos et al. 1993). Although these were substantial efforts, they either resulted in only a partial recovery of total cellulase enzymes or in the recovery of only specific component enzymes of the cellulase mixture (endoglucanase, exoglucanase and P-glucosidase). A n 45 effective and practical method for the recovery of the entire cellulase family has not yet been established. Since the substrate specificity, hydrophobicity and binding capacity of each component enzyme is considered to be different; it is predicted that specific recycling strategies wi l l be required in order to address the recovery of each enzyme component. For example, during hydrolysis, the P-glucosidase component and its substrate, cellobiose, remain soluble in the water phase, therefore immobilization of P-glucosidase on an inert insoluble carrier may enable its recovery and reuse, p-glucosidase has been immobilized on various materials via either physical adsorption or covalent bonding (Abdel-Fattah et al. 1997; Bissett and Sternberg 1978; Calsavara et al. 2000; Fujishima et al. 1987; Gomez et al. 2005; Martino et al. 1996; Sardar et al. 1997). However, there have been no reports of the application of immobilized p-glucosidase to lignocellulosic substrates. Further complicating enzyme recycling strategies, endoglucanases and exoglucanases also exhibit different adsorption profiles during the hydrolysis of lignocellulosic substrates. These factors further emphasize the necessity to consider the behavior of each component enzyme in order to develop approaches for the effective recovery of cellulases. In the process of developing methodologies for cellulase recycling, some key issues need to be addressed: How are the free enzymes and bound enzymes distributed between the solid and liquid phases after hydrolysis? What proportion of the free enzymes can be recovered from the liquid phase? How accurate are theoretical models in predicting the amount of free enzymes recovery compared to actual experimental data? What are the most significant experimental conditions affecting the recovery of bound enzymes? Does the addition of surfactants to enzyme hydrolysis influence enzyme recovery? Which 46 combination of pretreatment method and cellulase preparation results in the greatest amount of cellulase recovery? How do cellulases adsorb on lignin? With these questions in mind the following are the specific objectives of this study: • To determine the cellulase distribution between the liquid and solid phases during the enzymatic hydrolysis of lignocellulosic materials. • To compare the theoretical and experimental recovery of free cellulases based on their adsorption isotherms. • To optimize the process conditions to maximize the amount of bound enzyme desorption from lignocellulosic residues. • To evaluate the impact of surfactants on cellulase recycling strategies. • To assess enzymatic hydrolysis and apply enzyme recycling strategies to steam exploded Lodgepole pine and ethanol pretreated Lodgepole pine with different commercial cellulase preparations. • To study the adsorption of cellulase on lignin. • To evaluate the P-glucosidase recycling by immobilization The research objectives were derived from the fact that, after a typical hydrolysis of a lignocellulosic substrate, cellulases are distributed between the solid and liquid phases. In order to develop a complete recycling strategy, the recovery of free cellulases, adsorbed cellulases and p-glucosidases must all be considered. To recycle free cellulases, the process of readsorption onto fresh substrates wi l l be developed and examined using the Langmuir adsorption isotherm model. This process model w i l l be used to predict the amount of free cellulases recoverable during a batch readsorption. The development of a predictive model wi l l be potentially useful for future practical applications. 47 The approach of improving enzyme desorption wi l l be explored to recover bound enzymes. The cellulase desorption process w i l l be optimized by response surface methodology to ascertain which factors (pH, temperature, ionic strength, and surfactant) have a significant effect on the recovery of bound enzymes. We hypothesized that a surfactant would reduce the interaction between cellulases and lignin. Different surfactants w i l l be compared and evaluated for their ability to improve the recovery of enzymes after the enzyme hydrolysis of a lignocellulosic substrate. Once the protocols for recovery of free and bound enzymes are developed, the recycling performance of cellulase preparations from T. reesei and Penicillium sp. wi l l be compared. In addition to employing different cellulase preparations, Lodgepole pine substrates pretreated using steam explosion and organosolv wi l l be also be compared for their performance during enzyme recycling. It was expected that the different lignin contents and physical properties of the two substrates would affect enzyme recycling performance. The results should indicate which combination of cellulase preparation and pretreatment method would be most suitable for enzyme recycling during lignocellulosic hydrolysis. In addition to the studies on free and bound enzymes, the recovery and reuse of (3-glucosidase wi l l also be explored. To address the recycling of P-glucosidase the feasibility of immobilizing P-glucosidase on an inert carrier while maintaining sufficient activity and stability wi l l be evaluated. In summary, the work planned for the research project fully investigates the possibility of recycling the entire "family" of cellulases during lignocellulosic hydrolysis. The goal of this work was to develop novel enzyme recycling processes to potentially reduce the 48 enzyme costs associated with bioconversion, while gaining a better understanding of cellulase adsorption processes on softwood lignocellulosic substrates. 49 2 Materials and methods 2.1 Enzymes and immobilization 2.1.1 Cellulases Two commercial cellulase preparations (assuming protease free) derived from T. reesei were used in this study: Celluclast 1.5L (Novozymes, Franklinton, N C , U S A ) ; and Spezyme C P (Genencor International, San Francisco, C A , U S A ) . In addition to the commercial preparations, another cellulase preparation produced in the Forest Products Biotechnology (FPB) laboratory derived from Penicillium sp. was also used. This F P B lab enzyme preparation is referred to as M S U B C . The P-glucosidase preparation used was Novozym 188 (Novozymes, Franklinton, N C , U S A ) . 2.1.2 Enzyme activity assays The total cellulase activity of each enzyme preparation was measured using the filter paper assayed according to Ghose (1987). A 1.0 x 6.0 cm Whatman no . l filter paper strip was used as the substrate for the filter paper assay. Prior to the assay, the enzyme preparation (0.5 mL) was diluted in acetate buffer (0.05 M , p H 4.8) and added to test tubes containing a filter paper strip. The test tube was then incubated for 1 hr at 50°C in water bath. Dinitrosalicylic acid (DNS) (3 mL) was added after the incubation to end the reaction. The reaction tubes were placed in a boiling water bath for 5 min and the cellulase activity was determined using a Perkin Elmer U V - v i s spectrometer (Lambda 45) at 540 nm. The absorbance value was corrected using the enzyme and reagent blanks. 50 Enzyme activity was expressed in filter paper units (FPU), where 1 F P U is equivalent to the enzyme concentration required to release 2.0 mg of glucose in 60 min (Ghose 1987). P-glucosidase (Novozym 188) activity was determined using /?-nitrophenyl-P-D-glucopyranoside (p-NPG), as described previously by Wood and Bhat (1988). A 200 ul aliquot of free or immobilized p-glucosidase was added to a reaction mixture containing 1.8 m L of 0.1 M sodium acetate buffer (pH4.8), and 1.0 m L of 5 m M p - N P G in sodium acetate buffer (pH 4.8). The reaction mixtures were incubated at 50°C for 30 min with gentle shaking, followed by addition of 4.0 m L 0.4 M glycine buffer (pH 10.8) to stop the reaction. The extent of hydrolysis was determined by monitoring the release of p-nitrophenyl at 405 nm on a Perkin Elmer U V - v i s spectrometer (Wood and Bhat 1988). The absorbance of the reaction was compared to a standard curve prepared using p-nitrophenol. The activity is expressed in International Units (IU) where one I U is defined as the amount of enzyme required to release 1 umol of p-nitrophenol per min. For determination of kinetic parameters, P-glucosidase activity was determined by monitoring the hydrolysis of cellobiose. This technique is described in detail within section 3.5. 2.1.3 Immobilization of p-glucosidase P-glucosidase was immobilized on Eupergit C ® (Rohm G m b H & Co.) according to the manufacturer's instructions (Rohm Pharm). Briefly, 0.5 g (dry weight) Eupergit C was added to 10 m L of 1 M potassium phosphate buffer (pH 7.0), containing 30 u L Novozym 188 (5.5 mg total protein equivalent to 40 I U P-glucosidase). The immobilization reaction was incubated at 25°C for up to 36 hr with gentle agitation. The amount of immobilized 51 protein was determined from the difference between the total protein added and the amount remaining in solution after the immobilization reaction. The ability of glucose (a competitive inhibitor of P-glucosidase) to increase the efficiency of immobilization by protecting the p-glucosidase active site was evaluated by incubation in 1% glucose prior to addition of Eupergit C . After the reaction, the Eupergit C-immobilized enzyme was washed extensively with distilled water and resuspended in 50 m M sodium acetate buffer (pH 4.8). In the reactions evaluating the ability of bovine serum albumin (BSA) or 2-mercaptoethanol (2-ME) to improve immobilization efficiency by blocking unreacted epoxy groups, the freshly washed Eupergit C -immobilized enzyme was incubated in phosphate buffer containing 1% B S A or 1% 2-M E , at 25 °C for 24 hr, then further washed with distilled water and resuspended in acetate buffer. 2.2 Protein content assay In order to minimize the interference from reducing agents, detergents and lignin, different protein assay methods were utilized in this work. The protein assay used for each experiment is specified. Protein concentrations were determined using the Pierce B C A ™ protein assay (Pierce Biotechnology Inc. 2003) in section 3.5, with reference to a calibration curve prepared with bovine serum albumin (BSA) . In order to eliminate the effects of glucose and lignin on the assay, the protein content was measured using the R C D C Protein Assay (Bio-Rad, Hercules) in section 3.2. Bovine serum albumin (BSA) was used as the standard. 52 A protein assay utilizing ninhydrin (Hirs 1967) modified by others (Chauvet and Lamy 1990; Conklin-Brittain et al. 1999; Hirs 1967; Makkar et al. 1987; Starcher 2001) was used in section 3.3 and 3.4. The protein samples (0.1 mL) were pipetted into tubes and placed in an oven at 105°C for 2 hr. After the tubes were completely dry, 0.15 m L 13.5N N a O H was added to each tube. The tubes were then autoclaved at 121°C for 20 min. After cooling, the alkali was neutralized by adding 0.25 m L glacial acetic acid. Ninhydrin reagent (0.5 mL) was added to each tube and then heated for 20 min in a boiling water bath. After the reaction, the tubes were cooled in an ice water for 10 min, adding 2.5 m L 50% ethanol. The tubes were then shaken vigorously on a vortex. Finally the solution was read at 570 nm on a U V - v i s Spectrometer (Perkin-Elmer). B S A was used as the protein standard. 2.3 Lignocellulosic substrates Table 8. Major substrates used in this research thesis". Substrates Pretreatment methods Lignin content ( %) Abbreviations Avice l (pure cellulose) - - -Douglas fir Organosolv (acetic acid) 3 M i x e d softwood Organosolv (ethanol) 6 E P M S Lodgepole pine Steam explosion 45 S E L P Lodgepole pine Organosolv (ethanol) 14 E P L P aSoftwoods Douglas fir (Pseudotsuga menziesii) and Lodgepole pine (Pinus contorta) are two major species which grow in British Columbia, Canada. 2.3.1 Steam explosion pretreatment Softwoods substrates Douglas fir (Pseudotsuga menziesii) and Lodgepole pine (Pinus contorta) were used for the experiments due to their availability. The samples were chipped to a 4 x 4 cm size. Chip samples of 50 g O D (Oven-dry weight) were impregnated with 4.0% anhydrous SO2 (w/w) in a plastic bag. The samples were loaded 53 in 50 g O D batches into a pre-heated 2 L Stake Tech III (Stake Technologies, Norvall O N , Canada) steam gun in the Forestry Product Biotechnology laboratory at the University of British Columbia. The conditions used for steam exploded Lodgepole pine (SELP) and Douglas fir was: 200°C, 4.0% S 0 2 (w/w) at a residence time of 5 min. After the pretreatment, the substrates were washed with water and stored at 4°C for subsequent experiments. 2.3.2 Organosolv pretreatment Acetic acid pulp (-3% w/w lignin) was prepared from softwood (Douglas fir) as described previously by Pan and Sana (Pan and Sana 1999). Briefly, a 30 g sample of Douglas fir chips were cooked in 180 m L of 95% acetic acid at 160°C, 60 psi, for 1 hr. After cooking the pulp was then treated with hot alkali peroxide (4% H2O2) at ~80°C for 1 hr. Pulps were washed with water and stored at 4°C. Ethanol pretreated Lodgepole pine (EPLP) was prepared by auto-catalyzed ethanol pulping with wood chips (4 x4 cm) at the University of British Columbia in a four-vessel (2 L each), rotating digester manufactured by Aurora Products Ltd (Savona, B C , Canada). The cooking conditions were 170 °C, 1.1% H 2 S 0 4 (w/v) in 65% Ethanol (v/v) for 60 min. The substrates were washed with ethanol and stored at 4 °C. 2.3.3 Other cellulose substrates and chemicals Avice l PH101, a microcrystalline cellulose, was purchased from Fluka (Switzerland). Sigmacell 50, a microcrystalline cellulose, was purchased from Sigma (USA) . Tween 20, Tween 80, Span 80 and Triton X-100 were obtained from Sigma (USA) . Sodium dodecyl sulfate (SDS) was purchased from BioShop (Canada). 54 2.3.4 Characterization of substrates The carbohydrate and lignin composition of lignocellulosic substrates were determined using sulfuric acid hydrolysis. The lignin content was determined using the Klason lignin technique (TAPPI Method T249 cm-85). The acid soluble lignin content was measured at 205 nm (TAPPI Useful Method UM250) . The carbohydrate composition of acid hydrolysates was quantified using a Dionex DX-500 and 2500 High Performance Liquid Chromatography ( H P L C ) on a CarboPac PA-1 column. 2.3.5 Crystallinity by solid state nuclear magnetic resonance (NMR) The degree of crystallinity of samples (Avicel) was measured by solid state N M R (Will is and Herring 1987). The Avice l samples were tested for their crystallinity before and after enzymatic hydrolysis. Solid state 1 3 C C P M A S N M R measurements were done with a Varian I N O V A - 4 0 0 N M R spectrometer. The spinning speed was 5000HZ, acquisition time 0.020 sec, relaxation delay 2.0 sec. The total time of accumulation was 2 hr, 49 min. For the N M R experiments the hydrolysis conditions employed were 2% Avice l substrate in 100 m L acetate buffer (pH 4.8) and 45°C, with 20 F P U Celluclast g"1 cellulose and 40 IU Novozym 188 g"1 cellulose. After 48 hr hydrolysis with 59.25% hydrolysis yield, the cellulose residue was filtered washed with distilled water; and freeze dried overnight on a Savant freeze dryer. Before N M R measurement, the samples were moistened with distilled water to improve the signal/noise ratio. The crystallinity of cellulose (CrI) was determined from the areas of the crystalline of cellulose (86-92 ppm) and amorphous (79-86 ppm) C 4 signals. The chemical shifts in this solid state N M R determined using the C I signal as an internal standard (5 105). 55 2.3.6 Cellulolytic enzyme lignin (CEL) isolation Enzymatic lignin was prepared by hydrolyzing the ethanol pretreated Lodgepole pine and steam exploded Lodgepole pine. The substrates were suspended at a 2% consistency (based on cellulose) in 100 m L acetate buffer and hydrolyzed with 20 F P U Celluclast g"1 cellulose, 40 IU P-glucosidase g"1 cellulose, and 0.2% Tween 80 for 48hr. The hydrolysate was filtered using glass microfiber (Whatman G F / A ) . After the reaction, the residue was collected and further hydrolyzed for another 48 hr using identical amounts of cellulase, P-glucosidase and Tween 80 for 48 hr. After the second hydrolysis stage, the residues were recovered by filtration, resuspended in 0.2% Tween 80 and incubated for 2 hr at 45°C. The resulting lignin preparation was washed with 500 m L distilled water (-50 °C). The final lignin was put into vacuum oven for air dry at room temperature. The lignin samples were ground and screened by 60 mesh. 2.3.6.1 Elemental analysis of isolated lignin A n elemental analysis of steam exploded Lodgepole pine (SELP) lignin and ethanol pretreated Lodgepole pine (EPLP) lignin was performed using Perkin Elmer series II C H N S / O 2400 analyzer (Norwalk, C T , U S A ) . The protein content of purified lignin was calculated by multiplying the nitrogen content value by 6.25. A l l samples were measured in triplicate. 56 2.4 Cellulase adsorption kinetics and isotherm 2.4.1 Cellulase adsorption kinetics and isotherm on lignocellulosic materials Cellulase adsorption kinetics and isotherm experiments were performed at 4°C or 25°C using 5 0 m M sodium acetate buffer (pH 4.8). Cellulase preparations and pretreated substrates were varied in cellulase adsorption for different parts of experiments. In section 3.1, the adsorption of Celluclast on ethanol (organosolv) pretreated softwood (EPMS) at 4°C was studied. To measure the adsorption kinetics, Celluclast was suspended in 50 m M sodium acetate (pH 4.8) at a final protein concentration of 3.24 mg/mL. The Celluclast mixture was added to micro-centrifuge tubes containing the E P M S at a 2% of substrate. The micro-centrifuge was incubated at 4°C with agitation. Aliquots (0.2 mL) were removed from the reaction at 10, 20, 30, 60, 90, 120 min during the incubation. The supernatants were collected from the samples by centrifugation. A n d protein content in the supernatant was assessed by the Bradford assay (Lee et al. 1995). To determine the adsorption isotherm, a range of concentrations of Celluclast were incubated with ethanol pretreated softwood suspended at a 2% substrate in 50 m M acetate buffer at 4°C for 1 hr to reach equilibrium. The protein content in supernatant was determined for the non-adsorbed cellulase using the Bradford assay (Lee et al. 1995). The adsorbed cellulase was calculated from the difference of initial cellulase content and non-adsorbed cellulase content in the supernatant. In section 3.3, the cellulase adsorption was studied on the S E L P and E P L P substrates at 25°C. To measure adsorption kinetics, 20 F P U each of the three cellulase preparations (Celluclast, Spezyme and M S U B C ) in 5 0 m M sodium acetate (pH 4.8) was added to a 250 57 m L flask with 2% of substrate (EPLP) . The flasks were incubated at 25°C with agitation. Aliquots (0.2 mL) were removed at 0, 15, 30, 60, 90, 120, 180 min during the incubation. The supernatant from each sample was collected by centrifugation. The protein content of the supernatant was determined using the ninhydrin assay (Starcher 2001). To determine the adsorption isotherm, a range of cellulase concentrations (0.025 mg/mL-0.6 mg/mL concentrations) were incubated with 2% of E P L P pulp in 50 m M acetate buffer at 25°C for 1 hr to reach equilibrium. The protein content in the supernatant was determined for the non-adsorbed cellulase and the adsorbed cellulase was calculated from the difference between the initial cellulase dosage and the non-adsorbed cellulase. The classical Langmuir theory of adsorption was applied to the cellulase in solution. In this case, the surface concentration of adsorbed enzyme (T) was given by l + KC Where T m a x and is the surface concentration of protein at full coverage, K is the Langmuir constant. C is the protein concentration in bulk solution. 2.4.2 Cellulase adsorption kinetics and isotherm on cellulolytic enzyme lignin To measure adsorption kinetics, 30 mg lignin ( C E L - S E L P lignin or C E L - E P L P lignin) samples were suspended in 5 m L acetate buffer (50 m M , p H 4.8) with 0.18 mg/mL of commercial cellulases (Celluclast, Spezyme and M S U B C ) . The reactions were incubated with shaking for 6 hr at 25°C. Aliquots (0.15 mL) were taken at 0, 15, 30, 60, 120, 180, 360 min during the incubation. The protein content was determined using the ninhydrin assay. 58 To determine the adsorption isotherm, different concentration (0.05 mg/mL-0.4 mg/mL) of cellulase was incubated with -20 mg lignin in 1 m L of 50 m M acetate buffer at 25°C for 3 hr to reach equilibrium. The protein content that in supernatant was determined for the non-adsorbed cellulase. The adsorbed cellulase was calculated from the difference between the initial cellulase dosage and the non-adsorbed cellulase. The classical Langmuir theory of adsorption was also applied to cellulase adsorption on lignin samples. 2.5 Enzymatic hydrolysis of lignocellulosic substrates Unless otherwise stated, all enzymatic hydrolysis experiments were performed in 50mL of 50 m M acetate buffer (pH 4.8) at a 2% cellulose consistency. The reaction was incubated at 45°C, with shaking (150 rpm) for 24 or 48 hr. Enzyme loadings were 20 F P U g"1 cellulose with 40 I U g"1 cellulose of B-glucosidase (Novozym 188). In section 3.1 and 3.2 involving the hydrolysis of E P M S , the enzyme loadings were 10 F P U g"1 cellulose and 20 I U g"1 cellulose of (3-glucosidase (Novozym 188) due to the low lignin content (5.96%). Samples were removed from the reaction at different times and centrifuged to remove the insoluble materials. The reducing sugar content was measured by H P L C (see above); the hydrolysis yield of the substrate was calculated from the measured reducing sugar content, as a percentage of the theoretical reducing sugar available in the substrates. The protein content in supernatant was measured by the ninhydrin assay using B S A as the protein standard (Bio-Rad, U S A ) . 59 2.6 Cellulase recovery strategies After enzymatic hydrolysis, three major recycling strategies were assessed for their effectiveness in recovering active enzymes. The experimental protocols for all three strategies are outlined below. Ultrafiltration In the comparison of the activities of different sets of recycled cellulases based on ethanol pretreated mixed softwood (EPMS) hydrolysis yield (section 3.1.6). After first round of hydrolysis, free cellulases in the supernatant were collected by centrifugation (4000g) at 4°C. The supernatant with free enzymes was then combined with fresh E P M S (2% consistency based on cellulose content). After 2 hr incubation, the readsorbed cellulase on fresh E P M S substrate was collected by centrifugation at 4°C. The non-adsorbed cellulase in the supernatant was recovered by ultrafiltration to remove sugars. The recycled bound enzyme on the hydrolysis residue and readsorbed cellulases were supplemented with fresh buffer containing J3-glucosidase and used to hydrolyze fresh ethanol pulp separately to compare their hydrolysis yield (as cellulase relative activities). Ultrafiltration was performed using an Amicon ultrafiltration cell with a 10,000 molecular weight cutoffs of polyethersulfone membrane (Millipore, Bedford, M A ) at a transmembrane pressure of ~40 psi and room temperature (~25°C). 2.6.1 Recovering free cellulases from supernatants and bound enzymes with hydrolysis residues Ethanol pretreated mixed softwood (EPMS) (2% of substrates) was hydrolyzed in 100 m L buffer, with Celluclast (10 FPUg" 1 cellulose) and P-glucosidase (20 IUg" 1 cellulose). 60 The hydrolysis reactions were incubated at 45°C, with shaking (150 rpm) for 24 hr. After hydrolysis, the mixture was centrifuged at 4°C (4000g, 15 min), and the supernatant with free enzymes were separated and collected. To readsorb free enzymes in the supernatant, fresh substrates (EPMS) were added and allowed to stand at room temperature for 2.5 hr. The bound enzymes were recovered with the residual lignin via filtration. The recovered free enzymes and the cellulase-containing residual lignin and fresh P-glucosidase were combined together and used for further hydrolysis experiments. 2.6.2 Recovering free cellulase from supernatants and bound enzymes via desorption After a standard hydrolysis (section 2.4), the free enzymes were recovered by readsorption to fresh substrate as described above. To recover bound enzymes, solid lignocellulosic residues from the hydrolysis were suspended in 40 m L of acetate buffer (pH 5.3) and Tween 80. The suspension was incubated at 45°C for 2.5 hr. The desorbed enzyme was recovered by centrifugation. A combination of the desorbed cellulases, free cellulases and fresh p-glucosidase was used for further hydrolysis. 2.6.3 Addition of surfactant to lignocellulosic hydrolysis Surfactants (0.2%) were added to the hydrolysis of ethanol-pretreated (organosolv) Lodgepole pine (EPLP) substrate and steam exploded Lodgepole pine (SELP) substrate with Celluclast, Spezyme and M S U B C cellulase preparations. The filtrate was recovered from the hydrolyzed samples by filtration using a glass microfiber (Whatman G F / A ) filter. The filter cake was rinsed with an additional lOmL of acetate (pH 4.8) buffer. The free cellulase was then readsorbed from the filtrate using fresh substrate at 25°C for 2.5 61 hr. The free cellulase on the fresh substrates was recovered by filtration and re-suspended in acetate buffer containing fresh p-glucosidase, (P-glucosidase did not adsorb onto the substrates). A second round of hydrolysis was performed. Cellulase recycling was carried out four times for a total of five hydrolytic rounds. The protein content and reducing sugar content in supernatant was determined by ninhydrin and H P L C as discussed earlier. 2.7 Experimental design and statistical analysis 2.7.1 Experimental design Four factors were considered to have significant effects on enzyme desorption from lignin residues after enzymatic hydrolysis. The factors investigated were temperature ( X | ) , p H ( X 2 ) , ionic strength ( X 3 j NaCl ) and surfactant concentration ( X 4 , Tween 80). Response surface methodology (RSM) was used to optimize these factors. A second order polynomial model was fitted to the desorbed enzyme activity ( Y l ) and protein content (Y2) from a small composite design. Y = &+£/3lXi+jrj3liX?+± f4f3ijXiX] i=\ 1=1 i=l j=i+l Where Y is the value of the response variable, and X (1=1-4) is the value the coded values of the independent variables (factors) (Table 9), / J 0 is a constant, /?,,. and Ptj are the interaction coefficients and the quadratic coefficients. The range/ and levels of each variable (X) are shown in Table 9. A small composite design, consisting of 8 factorial components, 8 axial runs and 5 center points, was selected to assess the effects of the four independent variables (Montgomery 2000). 62 Table 9. Experimental design (small center composite design) for enzyme desorption from hydrolysis residues. Variables Variable name -1.68 -1 Levels (a) 0 + 1 + 1.68 X I (°C) Temperature 25 35 50 65 75 X2 pH 2.20 3.25 4.80 6.35 7.40 X3 (M) NaCl 0.0 0.20 0.50 0.80 1.00 X4 (%) Tween 80 0.0 0.405 1.00 1.595 2.00 2.7.2 Statistical analysis Statistical analysis was conducted using the S A S software package (SAS Institute Inc., Cary, N C ) Statistics. The analysis of variance for variables was used to determine the significant factors at a 95% confidence level. 63 3 Results and discussion 3.1 Cellulase distribution and free cellulase recovery 3.1.1 Background As mentioned in the introduction, the adsorption of cellulases on lignocellulosic substrates affects both the rate of cellulose hydrolysis and the ability to recover enzymes. After a typical hydrolysis, cellulases are partitioned between the supernatant and the solid residue of the resulting hydrolysate. Currently, there is little information to indicate the distribution of cellulases between the solid and liquid phases. To further complicate the system, the presence of lignin in substrates has also been shown to affect cellulase adsorption and distribution (Wu and Lee 1997). In order to develop effective recycling strategies, a more detailed understanding of cellulase adsorption is required. Therefore, in this chapter of the thesis, several experimental approaches were employed to try to determine cellulase distribution, enzyme adsorption, and substrate changes. Due to the initial availability of pretreated lignocellulosic substrates, the first set of experiments in this study investigated the hydrolysis of organosolv (acetic acid) pretreated Douglas fir substrates. The second portion of the work elucidated the adsorption characteristics of cellulases on an organosolv (ethanol) pretreated mixed softwood substrate (EPMS) . The results of this work were used to assess the potential to apply enzyme recycling strategies in later chapters. 64 3.1.2 Enzymatic hydrolysis and changes in enzyme distribution To accurately assess the potential of enzyme recycling, it was imperative to gain a fundamental understanding of the interaction between cellulases and substrates during hydrolysis. The adsorption behaviour of cellulases on a pure cellulose (Avicel) substrate was compared to an acetic acid (organosolv) pretreated Douglas fir (D. fir) substrate with 3% lignin content. During the hydrolysis of the two substrates with a commercial cellulase preparation (Celluclast), the protein content in the liquid phase was monitored and compared to determine the enzyme distribution. Within the first 6-12 hr, the total protein content in the solution decreased continuously to 70% and 50% for the Avice l and D . fir substrates respectively (Figure 7 and Figure 8). After 12 hr, the amount of protein in the solution increased to approximately 90% and 65% for the Avice l and D. fir substrates, indicating the desorption of cellulases from the substrate. L u et al. (2002) observed similar results during the hydrolysis of Avice l , steam exploded Douglas fir (D. fir, lignin 46.1%) and hot alkali peroxide treated Douglas fir (lignin 8.2%), as after 48 hr hydrolysis, the protein contents of the hydrolysis solutions of each substrate were 85%, 30%, 65% respectively (Lu et al. 2002). The adsorption behaviour of cellulases over the 48 hr hydrolysis period was also demonstrated visually by sodium dodecyl sulfate polyacrylamide gel electrophoresis ( S D S - P A G E ) (Figure 9 and Figure 10). S D S - P A G E gels indicated the major cellulase components ( C B H I and C B H II) were in the 65,000 kDa molecular weight band. The relative intensity of the 65,000 kDa band ( C B H I and C B H II) decreased in the first 12 hr of hydrolysis, then gradually increased until at the end of the hydrolysis (Figure 9). Furthermore, the final intensity of the 65,000kDa band in the hydrolysis of Avice l was much higher than that from the 65 hydrolysis of acetic acid pretreated D . fir (Figure 10). These results also suggested a greater amount of cellulases adsorbed on the hydrolysis residue of the acetic acid pretreated Douglas fir than pure cellulose (Avicel), thus further demonstrating the effects of lignin on the desorption of cellulases from the substrate. Although most of the adsorbed cellulases desorb after 48 hr of hydrolysis when using pure cellulose substrates such as Avice l (Figure 7), the presence of lignin is a significant hindrance to the desorption process. Therefore, due to potential enzyme lignin interactions, strategies for recycling cellulases during lignocellulosic hydrolysis must consider the effects of pretreatment on the lignin content of the resulting substrate. However, in order to gain insight into the interaction of cellulases with lignocellulosic substrates, it is necessary to first fully understand the interaction of cellulases with their preferred substrate, pure cellulose (Avicel). The next set of experiments focused on studying the properties of cellulases and Avice l during the hydrolysis process. 66 Time (hr) Figure 7. Cellulase adsorption and hydrolysis profile of Av ice l . Hydrolysis conditions: 45°C, 48 hr, 150 rpm and 20 F P U g 1 cellulose, 40 I U p-Glucosidase g"1 cellulose, 2% Avice l in N a A c buffer pH4.8. i 1 1 , 1 , r 1 1 1 ' 1 1 1 • 1 1 1 1 0 10 20 30 40 50 Time (hr) Figure 8. Cellulase adsorption and hydrolysis profile of acetic acid pretreated Douglas fir (3% lignin). Hydrolysis conditions: 45°C, 48 hr, 150 rpm and 20 F P U g"1 cellulose, 40 I U p-Glucosidase g"1 cellulose, 2% in N a A c buffer pH4.8. 67 A B C D E F G H 200000 116250 97400 66200 45000 3 K 21500 Figure 9. S D S - P A G E analysis of hydrolysate over 48 hr hydrolysis of Avice l . A , 0 hr; B , 2 hr; C , 6 hr; D , 12 hr; E , Protein standard; F, 24 hr; G , 36 hr; H , 48 hr. 0.85mL hydrolysate sample was put into -20°C overnight with 3 m L acetone. The mixture was centrifuged at 4000g to remove supernatant. 0.5 m L acetate buffer was added further. For S D S - P A G E running, 10 ul sample was mixed with 20 u L running sample buffer, 20 uL loading sample at 200 V for running 2 hr. E 200000 j p i r r n r c i f 116250 97400 6 6 2 0 0 45000 31000 21500 1 4 4 0 0 6 5 0 0 Figure 10. S D S - P A G E analysis of hydrolysate over 48 hr hydrolysis of acetic acid pretreated Douglas fir. (The conditions were the same as above). 68 To determine i f the cellulases that remained in solution after hydrolysis possessed hydrolytic activity and substrate reactivity, two standard 48 hr hydrolysis experiments of Avice l were performed and then restarted after the addition of either cellulase (Celluclast) or Avice l . The released sugars were monitored for another 48 hr (Figure 11). The addition of fresh cellulases after a 48 hr hydrolysis improved sugar release marginally from 15.2 g/L to 17.1 g/L, however, the addition of fresh Avice l after 48 hr hydrolysis increased the sugar release significantly from 15.2 g/L to 21.5 g/L. The results show that after 48 hr of hydrolysis, the cellulases in solution retained their activity. Therefore, enzyme recycling strategies involving cellulases in hydrolysis solutions are feasible. It was also of interest to determine the reasons why desorbed cellulases did not readsorb to the hydrolyzed solid substrates during the 48 hr hydrolysis. It was hypothesized that the desorption of cellulases from pure cellulose substrates was due to a decrease in the available surface area and/or changes in the crystallinity of the substrate (Burns et al. 1989). To test this hypothesis, solid state l 3 C cross polarized magic angle spinning (CP M A S ) N M R was employed to measure changes in the crystallinity of Avice l after a 48 hr hydrolysis by cellulases. The chemical shifts in this solid state N M R were determined using the C I signal as an internal standard (5 105). The crystallinity of Avice l did not undergo appreciable changes during the 48 hr hydrolysis, as the crystallinity index increased only from 63.75 to 68.89. Similar results were shown during the hydrolysis of Sigmacell by cellulase component enzymes (Mansfield and Meder 2003; Ramos et al. 1999). These results suggest that the decrease in cellulase adsorption is not a result of increased crystallinity of the Avice l substrate. Therefore, the desorption may be attributed to changes in available surface area for cellulase adsorption. 69 The effect of changing available surface area of cellulose substrates on cellulase adsorption was demonstrated by Ooshima et al. (1991). Comparing the adsorption of cellulases from Trichoderma viride on fresh vs. pre-hydrolyzed Avice l , Ooshima et al (1991) postulated that the decreased enzyme adsorption on the pre-hydrolyzed Av ice l was most likely due to a decrease in available surface area that occurred during the pre-hydrolysis treatment. Therefore, similar to the work by Ooshima et al (1991), it is also possible that the cellulases in this work did not readsorb to the hydrolyzed substrate due to a lack of available surface area. Overall, the data from the hydrolysis of pure cellulose substrates shows that the cellulase activity of desorbed enzymes does not decline after 48 hr hydrolysis; and the desorbed cellulases in solution can be readsorbed by the addition of fresh substrates. 70 Time (hr) Figure 11. Change in the hydrolysis of Avice l upon supplementation either additional Avice l or cellulases. After the first 48 hr hydrolysis, an additional 20 F P U of Celluclast was added to one of the experiments and the hydrolysis was allowed to continue, while another 2 g Av ice l was added into second hydrolysis experiment. For the control, the hydrolysis was continued without adding any additional Av ice l or enzyme. Hydrolysis conditions: 45°C, 48 hr, 150 rpm and 20 F P U g"1 cellulose, 40 IU p-Glucosidase g"1 cellulose, 2% Avice l in 100 m L N a A c buffer (pH 4.8). ci C 4 crystall ine amorphoiu A 140 133 126 119 112 105 ' I 1 1 ' I 1 , 1 l ' " 1 1 1 1 I " ' 1 I * 91 84 77 70 63 56 PPM Y^ ,.x Avicel control Avicel Hydrolysis i I i i i I i i T i 42 35 26 21 14 7 0 4ft Figure 12. 1 3 C C P M A S N M R spectrum of Avice l to obtain post-hydrolysis crystallinity. Hydrolysis conditions were 2% Avice l in 100 m L acetate buffer (pH 4.8) and 45°C, with 20 F P U Celluclast g"1 cellulose and 40 IU Novozym 188 g"1 cellulose. 71 3.1.3 Cellulase distribution during lignocellulosic hydrolysis Lignocellulosic materials are composed of cellulose, hemicellulose and lignin. The pretreatment step removes only a fraction of the lignin, thus a significant amount remains in the substrate. Since lignin has already been shown to influence the amount of cellulases that desorb into the liquid phase during hydrolysis, it is expected that the distribution of cellulases after lignocellulosic hydrolysis w i l l differ from pure cellulose. The hydrolysis of lignocellulosic substrates was studied by treating an ethanol (organosolv) pretreated mixture softwood (EPMS) with a commercial cellulase preparation (Celluclast). The hydrolysis yield, protein content and cellulase (FPU) activity were all monitored during the hydrolytic process. The results from the hydrolysis of the ethanol (organosolv) pretreated mixture softwood (EPMS) show the substrate was almost completely hydrolyzed within 24 hr (Figure 13). With the exception of P-glucosidase (a negligible amount of P-glucosidase adsorbed on the substrate or lignin-rich hydrolysis residue in Figure 17), more than 75% of the added cellulases were adsorbed onto the substrate during the first 3 hr of hydrolysis. As the hydrolysis proceeded, cellulases began to desorb into solution, finally reaching 50% in the liquid phase after 24 hr. It was shown earlier that approximately 90% of the total protein (added cellulase + P-glucosidase) could be recovered after the hydrolysis of pure cellulose. Therefore, the results suggest that the inability to recover of 50% of the added cellulases in the liquid phase can most likely be attributed to adsorption of cellulases to the lignin fraction of the substrate. Initially, the cellulase activity in the liquid phase decreased quickly to 30% of the initial activity. However, as the hydrolysis proceeded the activity only increased to 35% even though the protein content in solution doubled from 72 25 to 50%. The activity did not surpass 35%, most likely due to the effect of end product inhibition on F P U activity. The results have shown that during the hydrolysis of lignocellulosics, a significant amount of active cellulases remain bound to the residual substrate after hydrolysis, therefore, enzyme recovery strategies must be devised to recover both the free and bound enzymes. It has been suggested earlier that by exploiting their high affinity for cellulose, free cellulases can be recovered in a straightforward fashion by readsorption onto fresh substrates (Castanon and Wilke 1980). However, a complete recovery of free enzymes may not be guaranteed by applying this method. Therefore, prior to practical application, theoretical models predicting the amount of free cellulases that can be readsorbed to fresh substrates would be helpful in developing potential enzyme recycling methods. 73 T 1 1 1 1 1 1 1 r 0 5 10 15 20 Time (hr) Figure 13. Hydrolysis yield, protein content and F P U activity in the supernatant for ethanol pretreated mixed softwood ( E P M S substrate). Hydrolysis conditions: The hydrolysis was carried out in 100 m L of 50 m M acetate buffer (pH 4.8) at 2% consistency (based on cellulose content). The reaction was incubated at 45°C at 150 rpm. Enzyme loading was 10 F P U g"1 cellulose with 20 I U g"1 cellulose of P-glucosidase (Novozym 188). The substrate was E P M S with a 5.96% lignin content. 3.1.4 Theoretical prediction of free cellulase recovery Based on the adsorption kinetics and isotherms of cellulase adsorption it is possible to calculate the amount of free enzymes that can be recovered by readsorption onto fresh substrates. The kinetics of cellulase adsorption was studied on ethanol pretreated mixed softwood (EPMS) substrate. The hydrolytic activity of cellulases during the adsorption study was limited by decreasing the temperature to 4°C (less than 1% of substrate being hydrolyzed). The results show that the enzyme quickly adsorbs onto the substrate in the first 20 min, followed by a period of 30 min where the adsorption rate decreases and 74 reaches a plateau, therefore, cellulase adsorption reached equilibrium after 1 hr at 4°C (Figure 14). The adsorption isotherm of cellulases on the ethanol pretreated softwood substrate ( E P M S ) was determined and analyzed using the Langmuir model (Gilkes et al. 1992). T^KC r = -max l+KC The adsorption data was fitted using non-linear regression. The high coefficient of determination (R 2=0.99, the proportion of variability in a data set) indicated that this model was suitable for analyzing the adsorption of cellulases on E P M S at low enzyme concentrations. T m a x = 54.77 ± 3.45 mg/g substrate (dry weight), K = 4.53 ± 0 . 8 0 mL/mg. 1 1 Based on the linear plot of — versus — from the linear form of Langmuir model, r c 1 1 1 1 — + • — rmax and K are 45.25 mg/g substrate, 6.14 mL/mg respectively (R 2=0.98, when the first two points are removed in the linear plot). 75 Figure 14. Cellulase adsorption kinetics on ethanol pretreated mixed softwood (EPMS) at 4°C. Celluclast in 50 m M sodium acetate (pH 4.8) at a final protein concentration of 3.24 mg/mL was added to microtubes with the substrate at a 2% consistency. The microtubes were incubated at 4°C with inverse shaking. Aliquots were taken at 10, 20, 30, 60, 90, 120 min during the incubation, and collected the supernatant by centrifugation. The protein content was determined by the Bradford assay. Free cellulases in solution (mg/mL) Figure 15. Cellulases adsorption isotherm on ethanol pretreated softwood at 4°C. Different concentration of Celluclast was incubated with 2% of E P M S in 50 m M acetate buffer at 4°C for 1 hr. The protein content in the supernatant was determined for the non-adsorbed cellulase; the adsorbed cellulase was calculated from the difference in the initial cellulase content and the non-adsorbed cellulase content in the supernatant. 76 In the typical batch adsorption, the protein adsorption process can be monitored by controlling the equilibrium line and operating line (Belter et al. 1988). The equilibrium line is the adsorption isotherm shown in Figure 15. The equilibrium line is given in Figure 15 by the Langmuir model, i + KC Since T m a x = 54.77 ± 3.45 mg/g substrate (dry weight), K = 4.53 ± 0.80 mL/mg. _ rmax KC _ 54.77 x 4.53C = 54.77C ~ l+KC ~ 1 + 4.53C ~ 0.22053 + C In order to get agreement with the mass balance equation, V was replaced by q, and C by y. Therefore, the equation for the equilibrium line becomes: 54.77y q ~ 0.22053 + y y=concentration of unadsorbed cellulase in bulk solution (mg/mL) q=concentration of adsorbed cellulase (mg/g substrate) The mass balance for the free cellulases in the system is given by yFH+qFW = yH + qW Where: • y and yF are the final and feed cellulases concentrations in solution, • q and qF are the final and free cellulases content on the substrates, • H is the amount of feed volume, W is the amount of substrate. • yFH +qFW is the total cellulase in system before readsorption, • yF H is total free cellulase in solution, 77 • qFW is the total cellulase adsorbed onto fresh substrates before the readsorption experiment. This value is zero, since no cellulase adsorbed onto fresh substrate before the readsorption process. • yH + qW is the total cellulase in system after readsorption, • yH is total free cellulase in solution, qW is the total cellulase adsorbed onto fresh substrates after readsorption. This equation yFH + qFW = yH + qW can be rearranged as shown below: q — qF + — (yF - y) = ^®®m^ (o. 1 \mg I mL - y) , thus giving us the operating line with a W 2.0g negative slope presented in Figure 16. The original cellulase protein content (not including p-glucosidase) in the buffer was 0.21 mg/mL before hydrolysis, and 51% was left in solution, so yF = Q.Wmg I mL. There are two approaches to obtain the y value (the cellulase concentration in solution after readsorption) from the equilibrium and operating equations. One is a graphic solution; the other is a numerical solution. According to the equilibrium and operating equations, two lines could be drawn in the graph using graphic solution (Figure 16). The intersection of the two lines gave us the desired value y, equilibrium concentration of y=0.02mg/mL. The percentage recovery of free cellulases is calculated according to the following: Percentage recovery = y p ~ y x\Q0% = Q - 1 1 " 0 - 0 2 X 1 0 0 % = 81.8%. yF 0.11 In the numerical solution, combining equilibrium and operating equations, 78 54.77y 100 0.22053+y 2.0 ( 0 . 1 1 - y ) , y 2 + 1.20593y - 0.0242583 = 0 -b + jb2 + 4ac -1.20593+ V l . 2 0 5 9 3 2 +4x0.0242583 y = = = 0.01979 (mg/mL). 2a 2 0.11-0.01979 i n n m o n n m Percent recovery = x 100% = 82.0% . 0.11 As a result, based on the equilibrium line (adsorption isotherm) and operation line, in theory, approximately 82% of the free cellulases remaining in the supernatant after hydrolysis of ethanol pretreated softwood can be recovered. The objective of the next set of experiments was to verify the theoretical prediction for free cellulase recovery. a- 20 3 a at ~B> E * 1 5 - l 3 re « (0 A w 10 c o (0 ai re I 5 o> o •o a> n 1-o (0 •o < 04 — 1 — 0.02 — 1 — 0.08 0.00 2 0.04 0.06 0.10 Free cellulases in solution (mg/mL) 0.12 Figure 16. Graphic solution for cellulase readsorption. The intersection of the equilibrium line and operating line indicates the equilibrium concentration of free cellulases. 79 3.1.5 Experimental verification of free cellulase recovery These experiments were performed to assess the accuracy of the theoretical prediction of free cellulase recovery using readsorption on fresh substrates. After the hydrolysis of an ethanol pretreated mixed softwood substrate with subsequent centrifugation, 51% of the added cellulase was shown to remain in the supernatant (Figure 13). The free cellulases (51% of the protein from the original hydrolysis) in the supernatant were added to a suspension of fresh ethanol pretreated mixed softwood substrate (EPMS) for readsorption. After incubation of the enzyme-substrate mixture, the protein content in the liquid phase was shown to be 6.3% of the total free cellulase (51%). Based on this data, 87.6% (Calculation: [51%- 6.3%]/51%) of the free cellulases in the original hydrolysis of ethanol pretreated softwood were recovered by readsorption onto the fresh substrate while 12.4% of the free enzymes were not recovered. The experimental data agreed with the 82% free cellulase protein recovery predicted by the theoretical calculations. The distribution of the cellulase component enzymes in the hydrolysis supernatant was also analyzed by isoelectric focusing (IEF, Figure 17). Similar IEF analysis of the cellulase major components has been performed previously (Medve et al. 1998), revealing the band position of C B H I , C B H I I , E G I and EGII from Celluclast in IEF gel (pH 3.0-9.0). After a 24 hr hydrolysis of the ethanol pretreated mixed softwood substrate (EPMS) , the P-glucosidase band remained unchanged, however, the C B H I , C B H I I , E G I and EGII bands decreased. After further readsorption on fresh substrate, only the band of P-glucosidase remained, while the other four components disappeared. The results indicate P-glucosidase does not adsorb onto cellulose and lignin, while the four major cellulase 80 components C B H I , C B H I I , E G I and EGII were recovered using readsorption onto fresh substrates due to their affinity for cellulose. A B *. C D E p H 3 g CBHI EX3I Figure 17. Isoelectric focusing of cellulases during E P M S hydrolysis and enzyme recycling. A , P-glucosidase; B sample after hydrolysis; C , sample before hydrolysis; D , sample after hydrolysis and readsorption; E , original cellulases (Celluclast). Samples were centrifuged at 4000g for 10 min, and then the supernatant was concentrated 16 fold by Amicon ultra centrifugal filter device (from 0.8 m L to 50 uL). Sample loadings were 3 uL on IEF gels (pH 3.0-9.0), which were run by Pharmacia L K B , Phast system. Although a large proportion of the free cellulase proteins were recovered, the proportion of the cellulase component enzymes and the specific activity wi l l most likely be altered in the recovered enzyme mixture. Therefore in subsequent work we quantified the amount of enzyme activity that could be recovered during the readsorption process. 3.1.6 The activity of recovered free cellulases Similar to the previous experiments, Celluclast was applied to ethanol pretreated mixture softwood (EPMS) substrate, with subsequent isolation of free enzymes in the supernatant 81 by centrifugation. The bound enzymes in the centrifuge pellet (solid phase) were also retained. The free enzymes in the supernatant were readsorbed onto fresh substrate. The enzymes that were not adsorbed upon addition of fresh substrate were recovered by ultrafiltration; therefore the final enzyme inventory consisted of three sets of cellulase enzymes (Figure 18): 1. Enzymes bound to the lignin-rich residues of the initial hydrolysis (Bound enzyme) 2. Free enzymes readsorbed onto fresh-substrates (Readsorbed free enzyme) 3. Enzymes were not readsorbed, recovered by ultrafiltration (Non-adsorbed free enzyme) 82 Lignocellulosic materials (Ethanol pretreated softwood) 1 Celluclast +P-glucosidase Enzymatic hydrolysis Centrifuge Bound enzyme #1 (49% total cellulase) I Free enzyme (51% total cellulase) I .Readsorption onto fresh substrates 1 Readsorbed free cellulase #2 -Non-adsorbed free cellulase #3 87.6% free cellulase 12.4% free cellulase Figure 18. Schematic diagram of a free cellulase recycling process. Cellulase mixtures #1 and #2 (Figure 18) were supplemented with P-glucosidase and evaluated for their ability to hydrolyze the fresh ethanol pretreated mixed softwood (EPMS) substrate. The hydrolysis with non-adsorbed cellulases (#3) was applied to the E P M S substrate without additional P-glucosidase since the P-glucosidases from the original hydrolysis were also recovered in the ultrafiltration process. The results are shown in Figure 19. 83 The hydrolysis yields were 35%, 70% and 12%, for the bound enzyme (#1), readsorbed free enzyme (#2), and non-adsorbed free enzymes (#3) respectively. Similar to the results of protein analysis in the previous section, the recoverable free-cellulase activity during readsorption on fresh substrates was 85% [=70%/(70%+12%)], thus further demonstrating the accuracy of the theoretical prediction of 82% free cellulase recovery by readsorption to fresh substrates. The results shown here strongly support the application of fresh substrates to recover free enzymes in the supernatant. Although 50% of the cellulase protein was adsorbed to the lignin-rich hydrolysis residues, this cellulase mixture (cellulase mixture #1) performed poorly (35% hydrolysis yield) when recycled for use in a second hydrolysis. A likely explanation for this phenomenon is the inhibitory effect of lignin on the activity of the bound enzymes. Similar observations have been made previously during efforts to recover enzymes from the hydrolysis of steam exploded softwoods (Lu et al. 2002). 84 ioo4 80 4 1 1 hydrolysis 2 Recycled bound enzyme 3 Recycled readsorbed enzyme _ 4 Recycled nonadorbed enzyme 5 Glucosidase control 60 4 o •o 40 4 20 4 0 I < I 1 —• ' 'I ' ' 2 3 4 Different recycled cellulase 5 Figure 19. Comparison of the activities of different sets of recycled cellulases based on E P M S hydrolysis yields. After the first round of hydrolysis, free cellulases in the supernatant were collected by centrifugation (4000g) at 4°C. The supernatant containing free cellulases was then added to fresh E P M S (2%). After 2 hr incubation, the readsorbed cellulase on fresh E P M S substrate was collected by centrifugation at 4°C. The non-adsorbed cellulase in the supernatant was recovered by ultrafiltration to remove sugars. The recycled bound enzyme on the hydrolysis residue and readsorbed cellulases were supplemented with fresh buffer containing P-glucosidase and used to hydrolyze fresh ethanol pulp separately to compare their hydrolysis yields (as cellulase relative activities). 3.1.7 Conclusions The adsorption profiles of cellulase enzymes on pure cellulose and lignocellulosic substrates were studied to gain a fundamental understanding for the subsequent development of an enzyme recycling methodology. Since cellulases are partitioned between the solid and liquid phases after hydrolysis, it was shown that the continuous adsorption and desorption of cellulases has a profound effect on the recyclability of cellulase enzymes during the hydrolysis of lignocellulosic substrates. After hydrolysis of pure cellulose and lignocellulosic residues, cellulases retained most of their activity, thus 85 providing opportunities to recover and reuse these enzymes. Approximately 76% of the cellulases could be recovered in the liquid phase after 48 hr hydrolysis of pure cellulose substrates, while only 51% of the applied cellulases were recovered from the hydrolysis of an ethanol pretreated mixed softwood substrate. Readsorption of free-cellulases on fresh lignocellulosic substrates was shown to be an effective method for enzyme recovery Using calculations based on the adsorption isotherm of cellulases, 82% of the free cellulases in solution could be recovered by readsorption onto fresh substrates. It was shown that during the hydrolysis of an ethanol pretreated softwood substrate, 87% and 85% of the free cellulases could be recovered on a protein and activity basis respectively. Therefore, it is possible to accurately predict the recovery of free cellulases using the cellulase adsorption isotherm. Although readsorption of free-cellulases on fresh substrate was effective in recovering 50% of the protein, the enzyme activity of the recovered cellulases was only 35% of the activity added to the initial hydrolysis. Attempts to reuse the cellulases adsorbed on lignin-rich hydrolysis residues illustrated the detrimental effects of lignin on enzyme recycling. However, pretreatment methods that result in substrates with lower lignin contents or methods that release enzymes adsorbed to lignin may alleviate these problems. Since we had established that it was possible to recover a large proportion of free cellulases using readsorption onto fresh substrates, we next wanted to address the recovery of the enzymes that are bound to the solid lignin-rich hydrolysis residue. 86 3.2 Optimizing the recovery of bound cellulases using Response Surface Methodology 3.2.1 Background As mentioned earlier, bioconversion of lignocellulosics to produce fuel-grade ethanol remains an attractive option for the production of renewable, environmentally friendly fuels (Farrell et al. 2006). Although there have been recent advances that reduce the price of the bioconversion process (Cherry 2003), the high-cost of hydrolytic enzymes required for the hydrolysis step remains a significant obstacle to the application of this technology at commercial scale. Methods to reduce enzyme cost include increasing enzyme production yield, increasing specific activity and recycling cellulase enzymes to be used in subsequent hydrolysis. As part of a research effort to bring the bioconversion of lignocellulosics closer to commercialization, we assessed the potential of enzyme recycling during the hydrolysis of a mixed softwood feedstock (spruce/pine/ Douglas-fir) pretreated using an ethanol organosolv process. It was demonstrated in a previous section, that after the hydrolysis of an ethanol pretreated mixed softwood substrate (EPMS) (Figure 13), approximately 50% of the added cellulase enzymes remain free in solution (free enzymes). However, another 50%, possessing significant enzyme activity remains bound to the lignin-rich hydrolysis residue. Due to the large proportion of bound enzyme, an efficient enzyme recycling strategy requires the recovery of both the free and bound cellulases. We showed that, after lignocellulosic hydrolysis, the addition of fresh substrate could be an effective method for recovering up to 87% of the free cellulases in the liquid phase. Although the bound cellulases have been shown to be recoverable in the lignin-rich hydrolysis residue, 87 they only possess a fraction of their original activity (section 3.1.6). A n effective enzyme recycling strategy should aim to liberate these bound enzymes in order to achieve a complete recovery of cellulase activity. It has been shown that the surfactant Tween 80 is an effective additive to enzymatic hydrolysis, capable of liberating bound cellulases from the hydrolysis residue (Eriksson et al. 2002; Helle et al. 1993; Kaar and Holtzapple 1998; Lee et al. 1996; Otter et al. 1989). Therefore, Tween 80 was selected for cellulase desorption in this study. The work in this chapter focuses on optimizing the treatment conditions (temperature, p H , ionic strength, and surfactant concentration) using Response Surface Methodology ( R S M ) to maximize the desorption of bound cellulases from a hydrolysis residue taken from the hydrolysis of the E P M S substrate. The optimum conditions were then evaluated for the ability to facilitate the recovery and reuse cellulases in three rounds of hydrolysis of the E P M S substrate. It was hoped that the process analysis would then be useful in determining the factors that control cellulase desorption from hydrolysis residues, thus providing valuable information for further development of enzyme recycling strategies. 3.2.2 Experimental design and optimization of enzyme desorption from residual substrate A typical hydrolysis of the E P M S substrate was used to supply the hydrolysis residue for these experiments. The E P M S substrate was hydrolyzed in 50 m M acetate buffer (pH 4.8) for 24 hr at 45°C and 150 rpm with Celluclast at 10 F P U g"1 cellulose and 20 p-glucosidase IU g"1 cellulose. The hydrolysis residue was then treated to desorb enzymes under the conditions outlined in the experimental design (Table 9). After each desorption 88 experiment, samples were centrifuged and the recovered supernatant was tested for cellulase activity and protein content to measure the desorbed enzymes. The four variables evaluated for the optimization of enzyme desorption from the hydrolysis residue [symbol in brackets] were temperature [X i ] , p H [ X 2 ] , ionic strength (NaCl concentration) [ X 3 ] , and surfactant (Tween 80) concentration [ X 4 ] . The incubation time was fixed at 2.5 hr for the optimization experiments, since it was demonstrated in preliminary experiments that the desorption of bound cellulases from softwood substrates was unaffected by the incubation time. Response surface methodology (RSM) has been applied previously for the optimizations of cellulase production (Hao et al. 2006) and ethanol production during the simultaneous saccharification and fermentation of lignocellulosic biomass (Krishna and Chowdary 2000). In this work, R S M was used to optimize the desorption of enzymes from hydrolysis residues using a small center composite design (Ismail et al. 1998). The small center composite design has been frequently used in conjunction with R S M (Bandi et al. 2005; L i u and Tzeng 1998; Tanyildizi et al. 2006) A second order polynomial model was fitted to the recovered desorbed enzyme activity ( Y l ) and protein content (Y2): Y=/30 +f#Xt +fj3aX; f^X, X,. Equation 1 1=1 1=1 1=1 y=i+i Where Y is the value of the response, X (i=l-4) denotes the coded values of factors (Table 10), /? 0 is a constant; /?. and jSy are the interaction coefficients and quadratic coefficients. 89 3.2.3 Statistical Analysis and Response Surface Methodology After the experiments, the data was analyzed using analysis of variance ( A N O V A ) by S A S A D X software. The results of the measured cellulase activity and protein content recovered via desorption in the supernatants at the experimental conditions tested (Table 10) can be described by a polynomial model (Equation 1). The statistical program (SAS A D X ) eliminated the variables that were not regarded as having a significant influence on the desorption process at a 95% confidence (P-value>0.05). From the significant variables affected, a predictive model was developed where the p H and temperature, possessing a P-value greater than 0.05 were included due to quadratic effects based on the principals of hierarchy (Eon-Duval et al. 2003). 90 Table 10. Experimental design with variables, levels and response values in the small central composite design. R U N T E M P P H NaCl T W E E N A C T I V I T Y PROTEIN (°C) (M) (%) (IU) (mg/mL) 1 35 3.25 0.2 1.595 0.541 0.297 2 65 3.25 0.2 1.595 0.035 0.128 3 35 6.35 0.2 0.405 0.463 0.046 4 65 6.35 0.2 0.405 0.055 0.06 5 35 3.25 0.8 0.405 0.472 0.116 6 65 3.25 0.8 0.405 0.04 0.027 7 35 6.35 0.8 1.595 0.363 0.212 8 65 6.35 0.8 1.595 0.059 0.154 9 25 4.8 0.5 1 0.29 0.224 10 75 4.8 0.5 1 0.034 0 11 50 2.2 0.5 1 0.042 0.02 12 50 7.4 0.5 1 0.534 0.195 13 50 4.8 0 1 0.854 0.306 14 50 4.8 1 1 0.596 0.313 15 50 4.8 0.5 0 0.374 0.201 16 50 4.8 0.5 2 0.815 0.473 17 50 4.8 0.5 1 0.649 0.218 18 50 4.8 0.5 1 0.698 0.253 19 50 4.8 0.5 1 0.808 0.287 20 50 4.8 0.5 1 0.675 0.364 21 50 4.8 0.5 1 0.619 0.362 Ethanol pretreated mixed softwood (EPMS) was hydrolyzed in 25mL buffer with 10 F P U Celluclast and 20 IU Novozym 188 per gram cellulose. E P M S substrate at 2% consistency (based on cellulose) was hydrolyzed for 24 hr at 45°C and 150 rpm. After hydrolysis, the mixture was centrifuged at 4 °C (4000g, 15min), the supernatant was removed, and the lignin-rich hydrolysis residues were collected for further desorption. Acetate buffer (10 mL) at different temperatures, p H , N a C l concentration and Tween 80 concentration were added to the lignin-rich hydrolysis residues to desorb cellulases. After incubation, the mixture was centrifuged and the supernatant with desorbed cellulases was recovered. Cellulase activity and protein content were determined. 91 Table 11. Analysis of variance ( A N O V A ) for cellulase activity from the desorption for the master model and predictive model (without non-significant terms). Master Model Predictive Model Source DF SS MS F Pr>F DF SS MS F Pr>F T E M P I 0.317 0.317 20.076 0.004 1 0.317 0.317 14.228 0.002 P H 1 0.121 0.121 7.663 0.033 1 0.034 0.034 1.514 0.237 N A C L 1 0.033 0.033 2.107 0.197 T W E E N 1 0.097 0.097 6.157 0.048 1 0.037 0.037 1.652 0.218 T E M P * T E M P 1 0.634 0.634 40.122 0.001 1 0.616 0.616 27.658 <0.001 T E M P * P H 1 0.006 0.006 0.404 0.548 T E M P * N A C L 1 0.004 0.004 0.251 0.634 T E M P * T W E E N 1 0.000 0.000 0.007 0.935 PH*PH 1 0.389 0.389 24.633 0.003 1 0.375 0.375 16.812 0.001 P H * N A C L 1 0.061 0.061 3.833 0.098 P H * T W E E N 1 0.011 0.011 0.676 0.443 N A C L * N A C L 1 0.001 0.001 0.043 0.842 N A C L * T W E E N 1 0.090 0.090 5.700 0.054 T W E E N * T W E E N 0.042 0.042 2.651 0.155 Model 14 1.566 0.112 7.083 0.012 5 1.327 0.265 11.905 0.000 (Linear) 4 0.413 0.103 6.545 0.022 (Quadratic) 4 0.981 0.245 15.527 0.003 (Cross Product) 6 0.172 0.029 1.812 0.244 Enor 6 0.095 0.016 15 0.334 0.022 (Lack of fit) 2 0.074 0.037 7.054 0.049 9 0.278 0.031 3.314 0.079 (Pure Error) 4 0.021 0.005 6 0.056 0.009 Total 20 1.661 20 1.661 R 2 0.943 0.799 D F , degree of freedom; SS, sum of squares; M S , mean squares; F , F value; Pr, P value. 92 Table 12. Analysis of variance ( A N O V A ) for protein content from the desorption for the master model and predictive model (without non-significant terms). Master Model Predictive Model Source DF SS MS F Pr>F DF SS MS F Pr>F T E M P ! 0.034 0.034 7.608 0.033 1 0.034 0.034 10.035 0.006 P H 1 0.015 0.015 3.455 0.112 1 0.003 0.003 0.855 0.370 N A C L 1 0.000 0.000 0.006 0.943 T W E E N 1 0.037 0.037 8.346 0.028 1 0.073 0.073 21.765 <0.001 T E M P * T E M P 1 0.089 0.089 20.128 0.004 1 0.089 0.089 26.548 <0.001 T E M P * P H 1 0.006 0.006 1.292 0.299 T E M P * N A C L 1 0.000 0.000 0.002 0.967 T E M P * T W E E N 1 0.003 0.003 0.652 0.450 PH*PH 1 0.093 0.093 20.966 0.004 1 0.093 0.093 27.658 <0.001 P H * N A C L 1 0.001 0.001 0.130 0.731 P H * T W E E N 1 0.000 0.000 0.017 0.899 N A C L * N A C L 1 0.001 0.001 0.185 0.683 N A C L * T W E E N 1 0.014 0.014 3.066 0.130 T W E E N * T W E E N 1 0.000 0.000 0.018 0.897 Model 14 0.306 0.022 4.928 0.030 5 0.282 0.056 16.783 0.000 (Linear) 4 0.110 0.027 6.190 0.025 (Quadratic) 4 0.173 0.043 9.768 0.008 (Cross Product) 6 0.023 0.004 0.860 0.570 Error 6 0.027 0.004 15 0.050 0.003 (Lack of fit) 2 0.010 0.005 1.130 0.408 9 0.033 0.004 1.282 0.394 (Pure Error) 4 0.017 0.004 6 0.017 0.003 Total 20 0.332 20 0.332 R 2 0.920 0.848 D F , degree of freedom; SS, sum of squares; M S , mean squares; F, F value; Pr, P value. 93 The fit statistics for cellulase activity and protein content showed that there was a satisfactory coefficient of determination (R =0.80 for activity, and 0.85 for protein content), therefore there was no lack of fit in the analysis of both response variables. The final predictive models were based on the estimates of coefficients in Table 13 and are shown in the following equations: Activi ty=0.664-0.152*TEMP+0.050*PH+0.052*TWEEN-0.202*TEMP*TEMP-0.158*PH*PH (coded) Equation 2 Protein=0.304-0.050*TEMP+0.015*PH+0.073*TWEEN-0.077*TEMP*TEMP-0.079*PH*PH (coded) Equation 3 The effects of the four variables (temperature, Tween concentration, p H and ionic strength) on the two responses (desorbed enzyme activity and protein content) could be understood by examining the prediction profiler. As part of the S A S A D X software, the prediction profiler shows the slices of the fitted dependent variable values based on the values of the independent variables (SAS Institute Inc. 2002). In this case, based on predictive model (Equation 2 and 3), we could calculate the optimal conditions for each independent variable (pH and Temperature) by finding extreme values of either the desorbed enzyme activity or protein content (Adams 1995). Following this protocol, the optimal temperature and p H for enzyme desorption based on protein content were also obtained. In terms of desorbed protein (equation 3), the optimal temperature was 45.1 °C, and the p H was 4.9. Based on desorbed enzyme activity (Equation 2), the optimal temperature was 44.4 °C, and the pH was 5.3. Therefore, the optimal conditions for activity and protein content were quite close. 94 The ionic strength was found to have no significant effect on cellulase desorption, while the effects of temperature, p H and surfactant w i l l be discussed in detail below. To determine the optimal levels of the other three variables (temperature, p H and Tween 80 concentration), three dimensional response surface plots were constructed according to the predictive model shown above (Equation 2 and 3). The prediction profiler was generated by drawing the response vs. each variable, while the other variables remained fixed at the center point values (Figure 20). The plots were generated by drawing contours of the response vs. two selected (pH and temperature in Figure 21) variables, while the other variables (concentration of Tween 80) were held constant (Figure 21). The subsequent discussion wi l l focus on the effects of temperature, p H and the concentration of Tween 80 on the desorption of cellulase enzymes from hydrolysis residues. Table 13. Estimates for cellulase activity and protein content from desorption in the polynomial model. Term Estimate S t d E n t Pr>|t| Estimate S t d E n t Pr>|t| For Activity For Protein T E M P -0.152 0.040 -3.772 0.002 -0.050 0.016 -3.168 0.006 P H 0.050 0.040 1.230 0.237 0.015 0.016 0.925 0.370 T W E E N 0.052 0.040 1.285 0.218 0.073 0.016 4.665 <0.001 T E M P * T E M P -0.203 0.039 -5.259 <0.001 -0.077 0.015 -5.152 <0.001 PH*PH -0.158 0.039 -4.100 0.001 -0.079 0.015 -5.259 <0.001 95 Figure 20. Process optimization of temperature, pH, ionic strength and surfactant on desorption of bound enzyme (protein content and enzyme activity) from the lignin rich hydrolysis residues at the end of hydrolysis (EPMS) using the prediction profiler. 96 Figure 21. Three dimensional surface plots of cellulase activity as the function of p H and temperature (Fixed levels, NaCl=0, Tween 80=0.5%). -3.2.4 Effect of temperature, pH, ionic strength and surfactant concentration on cellulase desorption There has been only limited work chronicling the effects of environmental factors on the desorption of cellulases from lignin-rich hydrolysis residues (Converse et al. 1990). However, there have been numerous studies investigating the adsorption of cellulases on cellulose (Andreaus et al. 1999; Lee et al. 1994; Lee et al. 1982; Ooshima et al. 1990; Ryu et al. 1984). In this study, we have shown that the desorption of cellulases from hydrolysis residues was strongly dependent on temperature, p H , and concentration of Tween 80. 97 Desorption of cellulases increased as the temperature was raised from 25-45°C and then dropped continuously as the temperature surpassed 45°C (Figure 20 A ) . The results of the cellulase activity measurements indicated the drop in desorption at elevated temperatures was most likely due to the denaturation of the cellulase enzymes. Denatured enzymes would result in a lower amount of cellulase activity and protein content in liquid phase during the desorption treatment. The optimal temperature for both enzyme activity and desorption was observed to be approximately 45 °C. Similar to the results reported here, previous work investigating the desorption of cellulases from Avice l showed effective desorption in the 20-55°C range while further temperature increases reduced both enzyme stability and desorption (Otter et al. 1984). Carrard and Linder (1999) also reported that the desorption of cellulose binding domains (CBDCBHI and CBDCBHII) from cellulose occurred when the temperature was increased from 4 ° C to 3 4 ° C (Carrard and Linder 1999). The Effect of pH on the desorption of cellulases The linear effect of p H on desorbed enzyme activity and desorbed protein was not significant (P-value>0.05). However the quadratic effect of p H was significant. A s a result, p H was one of the significant factors affecting the enzyme desorption from the hydrolysis residues. The desorbed enzyme activity and protein content were enhanced with an increase of p H from 3.0-5.0. However, the desorption decreased as the p H was raised to 7.4 (Figure 20 B ) . Previous reports have suggested that more than 45% of the bound cellulases could be recovered from cellulose by p H adjustment from pH5.0 to pHIO (Otter et al. 1984). Although the p H swing to 10 was effective in desorbing bound cellulases, the desorbed enzymes were inactivated within 1 min due to denaturation at 98 high p H . The results demonstrate that p H control must be considered in the development of methods for the desorption of cellulases. The effect of Tween 80 on the desorption of cellulases From previous work utilizing surfactants to improve the productive binding of cellulases to substrates, it was hypothesized that the surfactant (Tween 80) competes with cellulase enzymes for hydrophobic adsorption sites on lignin-rich hydrolysis residues (Eriksson et al. 2002). The results show that the addition of Tween 80 was effective in facilitating cellulase desorption (Figure 20 D). Similar to the results here, Rao et al. (1983) obtained a 90% recovery of CMCase , P-glucosidase and filter paper activity from the hydrolysis of sugarcane bagasse using a Penicillium funiculosum cellulase, via the addition of 0.4% of Tween 80. It has also been reported that 68% of initial cellulase activity could be recovered from Avice l hydrolysis at pHIO and 40°C for 5 min using Tween 80 (Otter et al. 1989). In this study, the concentration of Tween 80 was not optimized from the data analysis using the conditions employed. There was a linear increase in desorbed enzyme and protein content with increases in Tween 80 concentration. The conditions employed in the experimental design (Table 9) did not maximize the enzyme desorption with Tween 80 addition, therefore, further increases in the Tween 80 concentration may result in an increase in enzyme desorption. Although the recovery of bound cellulases via the addition of Tween 80 was not optimized, the economic impact of Tween addition should be considered when suggesting "optimal" cellulase desorption conditions (Alkasrawi et al. 2003). The costs 99 associated with Tween 80 addition must be integrated into the economics of the entire bioconversion process in order to be weighed against the potential benefits. The applied Tween 80 concentration wi l l be examined alongside the decrease in cellulase cost during recycling and any other alterations to the bioconversion process that arise from Tween 80 addition (Alkasrawi et al. 2003). Previous work by Eriksson et al. (2002) has shown the benefits of adding 0.25% Tween 80 as the enzymatic hydrolysis of steam exploded spruce improved from 40% to 65%. Considering the previous work and the results obtained in the small composite design, the addition of 0.2% Tween 80 was chosen as the preferred concentration in the subsequent experiments (after the desorption process, for further hydrolysis with 25 m L buffer, the Tween 80 concentration was 0.2%). As a result, the preferred conditions for cellulase desorption determined by response surface methodology were 44.4°C, pH5.3, and 0.2% Tween 80. The effects of temperature, p H and surfactant concentration on the desorption of cellulase activity are described by the following equation: Activity=-2.830+0.080*TEMP+ 0.662*PH +0 .087*TWEEN-0 .001*TEMP*TEMP-0.066*PH*PH. The next set of experiments focused on evaluating the recyclability of cellulases, and utilizing the information gained in the Response Surface Methodology experiments. 100 3.2.5 The comparison of two potential recycling strategies Two strategies for recycling cellulases were developed based on the results of the response surface methodology (RSM) . Strategy 1 was compared with Strategy 2 (Figure 22 A and Figure 22 B) for the hydrolysis of E P M S substrate. The preferred conditions for enzyme desorption (44.4°C, pH5.3 and 0.2% Tween 80) were used for desorption of bound enzymes from the hydrolysis residue in Strategy 1 (Figure 22). Strategy 1 involves the separation of the hydrolysis residues and supernatant, with subsequent treatment of the hydrolysis residues with Tween 80 under the preferred conditions obtained from R S M . The enzymes in the supernatant were recovered using fresh substrates. The desorbed and readsorbed enzymes were combined and reused for hydrolysis of E P M S substrate. Strategy 1 was compared with Strategy 2 in which bound enzymes were recycled as part of the lignin-rich hydrolysis residues (without desorption treatment) while free cellulases were captured by readsorption onto fresh substrates. 101 hydrolysis round 1 Celluclast (10 FPU/g cellulose) fj-gtucosidase (20 lU/g cellulose) substrate (2 g pulp) buffer (100 mL) subsequent rounds fresh [5-glucosidase + buffer PELLET fresh substrate bound cellulase EXTRACT extracted cellulases buffer SUPERNATANT soluble sugars centrifuge * ^ p-glucosidase + unrecovered cellulases PELLET residual substrate discard extract with buffer (RSM) centrifuge PELLET residual substrate ' bound cellulase SUPERNATANT soluble sugars + p-glucosidase + unbound cellulases \ fwsh substrate Strategy 1, A hydrolysis round 1 Celluclast (10 FPU/g cellulose) p-glucosidase (20 lU/g cellulose) substrate (2 g pulp) buffer {100 mL) RETENTATE fresh substrate bound cellulase FILTRATE soluble sugars fitter + (5-glucosidase + unrecovered cellulases subsequent rounds fresh (5-glucosidase + buffer incubate 2.5h 4'C 150 rpm centrifuge PELLET residual substrate bound cellulase SUPERNATANT soluble sugars (5-glucosidase + unbound cellulases fr&sh substrata Strategy 2, B Figure 22. Comparison of two recycling strategies, both involving recovery of the bound and free cellulases. 102 Strategy 1 was compared to Strategy 2 in three rounds of hydrolysis. The "preferred" desorption conditions obtained from the optimization experiments utilized in Strategy 1 (Figure 23) resulted in an increase in the hydrolysis yield from 59% to 88% in the third round of hydrolysis of E P M S substrate. The continuous accumulation of lignin that occurred during each round of hydrolysis and recycling was the factor most likely responsible for the ineffectiveness of recycling Strategy 2 (Figure 23). 1 1 • 1 1 1 • 1 1 1 0 0 _ strategy 1 1 2 3 Hydrolysis rounds Figure 23. Comparison of cellulose conversion over three rounds of hydrolysis (EPMS) using recycling Strategy 1 and Strategy 2. 3.2.6 Conclusions These experiments were performed to gauge the effects of temperature, pH, ionic strength, and Tween 80 on the desorption of cellulases from the hydrolysis residues from the hydrolysis of an ethanol pretreated mixed softwood ( E P M S ) substrate. Response surface methodology with a small centre composite design was used to obtain preferred conditions (temperature 44.4°C, p H 5.3, and 0.2% Tween80) for enzyme desorption 103 based on our data and previous literature. The results indicated that the temperature, p H and surfactant were the three significant factors affecting the desorption of cellulases from the lignin rich hydrolysis residues. Increasing the temperature improves cellulase desorption within the 20-50°C range, while higher temperatures result in the inactivation of desorbed enzymes. The optimal temperature for cellulase desorption was shown to be 44.4°C, which has implications for the practical application of this technology. The optimal desorption temperature is almost identical to the temperature for optimal cellulase activity (45°C). Therefore, the enzyme desorption does not require a change in temperature and could result in savings of both energy and processing time. Similar to the results for temperature, the optimal p H was shown to be p H 5.3, which is close to 4.8, the optimum for cellulase activity. The surfactant Tween 80 enhanced the desorption of cellulases. A recycling strategy employing the preferred conditions for cellulase desorption from hydrolysis residues (Strategy 1) was compared to a strategy where hydrolysis residues were reused without a desorption treatment (Strategy 2). The results showed that the combination of enzyme desorption from hydrolysis residues and free enzymes readsorption is an effective method to recover cellulases during the hydrolysis of lignocellulosic substrates. 104 3.3 Application of enzyme recovery to lignin rich substrates 3.3.1 Background A s mentioned earlier, after a typical enzymatic hydrolysis of lignocellulosic substrates, cellulases are distributed between the solid and liquid phases of the hydrolysate. Enzymes either remain "free" in the liquid phase, or bound to the lignin-rich hydrolysis residues. Free enzymes can be recovered by for use in subsequent hydrolysis either by ultrafiltration (Tjerneld 1994) or by adsorption onto fresh substrates (Lee et al. 1995; Ramos et al. 1993). Although bound enzymes can easily be recovered as part of the lignin-rich hydrolysis residue, they are relatively ineffective in catalyzing further hydrolysis (Deshpande and Eriksson 1984; Lee et al. 1995). Since the presence of lignin has been shown to increase the amount of bound cellulases during hydrolysis, methods such as the addition of surfactant have been applied to decrease the enzyme-lignin interaction. Many studies have reported improvements in enzymatic hydrolysis of lignocellulosics after the addition of surfactants (Alkasrawi et al. 2003; Duff et al. 1995; Eriksson et al. 2002; Helle et al. 1993; Kaar and Holtzapple 1998). The positive effects of surfactant addition have been attributed to an interaction of the surfactant with lignin at the substrate surface, thus decreasing non-productive binding of cellulases to lignin (Alkasrawi et al. 2003; Eriksson et al. 2002). The ability of surfactants such as Tween to decrease the enzyme-lignin interactions during hydrolysis may also be exploited for the recycling of bound cellulase enzymes. Indeed, in the previous chapter effective conditions for enzyme recycling using the surfactant Tween 80 were established for the recovery of bound 105 cellulase enzymes. Using the information from the previous chapter, it was hypothesized that a simplified approach of adding surfactant during the hydrolysis, followed by readsorption to fresh substrate, may be more efficient than a two-step process where surfactant was added after the hydrolysis step. The one-step cellulase recycling approach is shown in Figure 24. In this part of the thesis, various surfactants were tested for their effectiveness in improving the recycling of cellulase preparations using the one-step enzyme recycling strategy (Figure 24). Using the most efficient surfactant, three cellulase preparations (Celluclast, Spezyme and M S U B C ) were recycled during the hydrolysis of steam exploded Lodgepole pine (SELP) and ethanol pretreated Lodgepole pine substrates (EPLP) . 106 Lignocellulosics (Ethanol pretreated Lodgepole pine) Celluclast p-glucosidase +0.2% Tween 80 Enzymatic hydrolysis Filtration Adsorbed cellulases onto fresh substrate Free enzymes in solution Hydrolysis Residue Filtration Incubation for readsorption Addition of fresh substrates Figure 24. Schematic diagram of a one-step cellulase recycling process. 107 3.3.2 Effect of various surfactants on enzymatic hydrolysis and recycling The effect of surfactants Span 80, Triton X-100, Tween 80, Tween 20 and SDS on the hydrolysis of E P L P and Sigmacell 50 using Celluclast was investigated (Figure 25). The substrate hydrolysis was carried out in acetate buffer (pH 4.8) with 20 F P U g"1 cellulose of Celluclast and 40 I U g"1 cellulose of P-glucosidase (Novozym 188). The effects of various surfactants (added at 0.2% to the above mixture) were compared using the conditions from a typical hydrolysis with cellulases (pH 4.8, Temperature 45°C) since the best-case scenario for enzyme recycling would involve surfactant addition at equivalent conditions as enzymatic hydrolysis. In the case of the E P L P , the addition of Span 80, Triton X-100, Tween 80 or Tween 20 had negligible effects on the hydrolytic performance on E P L P while SDS had a negative effect on the performance. The addition of Tween 80 to Sigmacell resulted in a 5% improvement in hydrolysis yield. The effects of Tween 80 on the hydrolysis of steam exploded Lodgepole pine (SELP) were also investigated. During the hydrolysis of S E L P with increasing concentrations of Tween 80 at low enzyme loading (10FPU g"1 cellulose), the hydrolysis yield increased from 48% to 70%, 81%, 89% by the addition of 0.1%, 0.2%, 0.5% of Tween 80 respectively (Figure 26). Previous studies have focused on the improvement of the enzymatic hydrolysis of various lignocellulosic substrates using Tween (Castanon and Wilke 1981; Eriksson et al. 2002; Ooshima et al. 1986).The hydrolysis yield of newsprint pulp ( with - 1 3 F P U g"1 cellulase) was increased from 41% to 55% by using 0.1% Tween 80 (Castanon and Wilke 1981). Similar to the results observed above for the comparison of Sigmacell to S E L P , Eriksson 108 et al (2002) observed a 15% improvement in the hydrolysis of steam exploded spruce compared to only a 5% improvement during the hydrolysis of Av ice l with the addition of Tween. These results suggest the improvement of enzymatic hydrolysis via the addition of surfactant most likely depends on the enzyme loading and substrate properties. Helle et al. (1993) reported that surfactant (Tween 80/sophorolipdid) improved the enzymatic hydrolysis of Sigmacell 100 only when the enzyme loading was low (~3 F P U g"1 cellulose). The hydrolysis yield of S E L P substrate (>40% lignin) increased from 48% to 81% with the addition of 0.2% of Tween 80 (Celluclast was added at 10 F P U g"1 cellulose for the hydrolysis) (Figure 26). Eriksson et al. (2002) also found the addition of 0.5% Tween 20 improved the hydrolysis of steam exploded spruce (>40% lignin) from 40% to 65%. However, only a 2% improvement was observed with delignified steam exploded spruce (Eriksson et al. 2002). Consequently, it was suspected that lignin played a role in the effect of the surfactant (Eriksson et al. 2002). Surfactants did not have a significant effect on hydrolysis of E P L P substrate at the conditions used in these experiments. Therefore, in the next set of experiments various surfactants were evaluated for their ability to aid in a single round of enzyme recycling during the hydrolysis of E P L P . 109 Time (hr) Figure 25. Effect of different surfactants on the hydrolysis of E P L P (14.6% lignin) and Sigmacell 50. Hydrolysis conditions: 20 F P U Celluclast, 40 I U P-glucosidase at 45°C and 150 rpm for 48 hr with 0.2% of each surfactant. Figure 26. Effect of Tween 80 concentration on the hydrolysis of S E L P (41.7% lignin) Hydrolysis conditions: 10 F P U Celluclast, 20 IU P-glucosidase g"1 cellulose at 45°C and 150 rpm for 24 hr. 110 A l l the surfactants shown in Figure 27 were each added into the hydrolysis of E P L P substrate with subsequent addition of fresh E P L P substrate to readsorb free enzyme (Figure 24). The surfactants Triton X-100, Tween 80 and Tween 20, enhanced cellulase recycling performance (based on the hydrolysis yield) for E P L P substrate by 50%. The addition of Span 80 had a negligible effect on the second round of hydrolysis of the E P L P substrate. For the pure cellulose Sigmacell 50, the addition of 0.2% Tween 80 only increased the second hydrolysis yield from 60.8% to 63.7% under the same condition in Figure 27. While , as expected the addition of SDS had a negative effect on enzyme recycling performance (Figure 27). The hydrophilic-lyophilic balance ( H L B ) reflects the surfactant's partitioning behavior between a polar (water) and non-polar (oil) medium. A positive correlation was found between H L B and increased performance of surfactants with H L B values up to 15 (Table 14), Our results agree with previous findings from Park et al. (1992) who concluded that the increased effectiveness of non-ionic surfactants with higher H L B values during the hydrolysis of newsprint pulp was due to a desorption of cellulases from the substrate, allowing the enzyme to subsequently readsorb for further hydrolysis. K i m and Chun (2004) did not find a relationship between H L B value and enzyme hydrolysis. This difference in results can most likely be attributed to a difference in methodology as K i m and Chun applied surfactant only in the pretreatment step (K im and Chun 2004), while in the current study the surfactant was added during the enzymatic hydrolysis step. The best surfactant for enzyme recycling at the conditions tested was shown to be Tween 80, thus Tween 80 was utilized in the subsequent recycling experiments. I l l F igure 27. Effect of various surfactants on enzyme recycling during the hydrolysis of E P L P . Hydrolysis conditions: 20 F P U Celluclast, 40 IU P-glucosidase at 45°C and 150 rpm for 48 hr with 0.2% of surfactants (except E P L P control). After hydrolysis, free enzymes were first collected by filtration; then fresh E P L P was added to the supernatant containing free enzymes for 2.5 hr readsorption at 25°C. Readsorbed enzyme with substrates were filtered and transferred to fresh buffer for further hydrolysis (supplemented new P-glucosidase). 112 Table 14. Surfactants with corresponding hydrophilic-lypophilic balance ( H L B ) and the increased performance of enzyme recycling during hydrolysis of E P L P after surfactant addition. Surfactants H L B Increased performance (%) Span 80 4.3 4.0 Triton X-100 13.0 54.2 Tween 80 15.0 55.1 Tween 20 16.7 47.2 SDS 40.0 -91.0 Increased performance= (Hydrolysis yield with surfactant - yield without surfactant)/yield without surfactant. Hydrolysis conditions: 2% of E P L P substrate, 20 F P U Celluclast, 40 IU P-glucosidase per gram cellulose at 45°C and 150 rpm for 48 hr with 0.2% of surfactants. After hydrolysis, free enzymes were first collected by filtration; then fresh E P L P was added into the supernatant containing free enzymes for 2.5 hr readsorption at 25°C. Readsorbed enzyme with substrates were filtered and transferred to fresh buffer for further hydrolysis (supplemented new P-glucosidase). 3.3.3 Evaluating the adsorption kinetics and isotherms of three cellulase preparations The adsorption kinetics were determined for the cellulase preparations derived from two Trichoderma reesei (Celluclast and Spezyme), and Penicillium ( M S U B C ) on ethanol pretreated Lodgepole pine (EPLP) . Cellulases over a range of concentrations were incubated with 2% of E P L P substrate in 50 m M acetate buffer. The cellulase adsorption on E P L P reached equilibrium within 1 hr (Figure 28) ( E P L P was hydrolyzed less than 3% during the incubation). From the adsorption isotherm (Figure 29), it was apparent that Celluclast and Spezyme (3.48 mL/mg and 3.17 mL/mg Langmuir constants) (Table 15) had a similar affinity for E P L P while the M S U B C had a much lower substrate affinity (0.62 mL/mg Langmuir constant). These Langmuir constants are in the same range as those shown in previous work by Lee (1994), where the Langmuir constants were 2.1 113 mL/mg for Celluclast on water-washed steam-exploded birch (WB) and 2.8 mL/mg on alkaline-washed W B (Lee 1994). As was shown in the previous section (3.1), Celluclast adsorption on ethanol pretreated mixed softwood ( E P M S ) at 4°C resulted in rmax =54.77 mg/g substrate, K=4.53 mL/mg similar to the results obtained with E P L P (Table 15). However, it was apparent that the adsorption capacity of E P L P was greater than that for E P M S and the Langmuir constant of E P M S was higher than that for E P L P . These differences were probably due to the different experimental conditions employed during the determination of the adsorption isotherms. Similar results were found for the adsorption of cellulases from T. reesei on pure cellulose (Solka Floe) at different temperatures (Kyriacou et al. 1988). The results of Kyriacou et al. (1988) indicated that, as the temperature was raised from 5.0°C to 50°C, the rmax of E G I (endoglucanase) adsorption on cellulose increased from ~5.0mg/g cellulose to -27.0 mg/cellulose while the Langmuir constant (K) decreased from -5.4 mL/mg to -3.0 mL/mg. Jorgenson and Olsson (2006) also found similar differences between the cellulases derived from Trichoderma and Penicillium during their work on steam exploded spruce (SPS). Greater than 94% of the Celluclast enzymes (Trichoderma) adsorbed onto SPS with a 25 F P U g"1 cellulase loading, while only 70% of the cellulases from Penicillium adsorbed onto SPS (Jorgensen and Olsson 2006). A possible explanation for the difference in adsorption is the fact that most of the cellulase components ( C B H I, C B H II, E G I, EGII , E G I V and E G V ) from Trichoderma possess a cellulose binding domain (CBD) except EGIII (<3% total protein) (Palonen 2004), while it is suspected that the 114 two significant cellulase components ( E G a and E G b l 25% of total protein) from Penicillium do not have a C B D (Jorgensen et al. 2003). Since the main recycling method employed in this study is the readsorption onto fresh substrates, it is expected that the affinity of each enzyme preparation for cellulose wi l l have significant effects on the recycling efficiency. Table 15. Langmuir constants from Celluclast, Spezyme and M S U B C on E P L P substrate at 25°C. Cellulase Tmax (mg/g) K (mL/mg) Celluclast on E P L P 87.69 3.48 Spezyme on E P L P 60.35 3.17 M S U B C on E P L P 75.24 0.62 Celluclast on E P M S * 54.77 4.53 *Celluclast adsorption on E P M S was at 4°C from section 3.1. ~i 1 — i — 1 — i — 1 — i — ' — i — ' — i — < — i — > — i — < — i — i — i — i --20 0 20 40 60 80 100 120 140 160 180 200 Time (min) Figure 28. Adsorption kinetics of Celluclast, Spezyme and M S U B C on E P L P substrate at 25°C. 115 -i 1 1 1 1 1 1 1 1 1 1 0.0 0.1 0.2 0.3 0.4 0.5 Free cellulases in solution(mg/mL) Figure 29. Cellulase adsorption isotherms on E P L P substrate (14.45% lignin) at 25°C. 3.3.4 Effect of varying cellulase preparations on enzyme recycling Celluclast, Spezyme and M S U B C were compared for their enzyme recycling performance by applying 0.2% Tween 80 and adding fresh substrate after 48 hr hydrolysis (Figure 30). The recycling process is shown in Figure 24. The recycling of the enzymes was tested on the E P L P substrate. The cellulases from the Trichoderma cellulases, Celluclast and Spezyme were the most effective for recycling from the hydrolysis of E P L P substrate, achieving an 80% hydrolysis yield after 4 recycling rounds. In contrast to the Trichoderma cellulase preparations, the hydrolysis yield of M S U B C cellulase derived from Penicillium was only 18% after 4 recycling rounds. As shown in the previous section, this could be due to the lower affinity of cellulases from Penicillium 116 on lignocellulosic substrates resulting in the inability to retrieve an appreciable amount of enzyme during readsorption. The higher substrate affinity of the Trichoderma cellulase preparations (Figure 29) for E P L P most likely facilitated their readsorption upon addition of fresh substrates, thus improving their recovery for subsequent hydrolysis. The difference in recycling performance of cellulases from Trichoderma and Penicillium could also be explained by the equilibrium line and operating line in Figure 16. In contrast to the Penicillium cellulases, the higher slope (relative higher affinity) of the equilibrium line in the adsorption isotherm for the Trichoderma cellulases would result in the intersection point of the two lines being located at a lower range of values, suggesting a higher amount of free enzymes were recovered. The results indicate that enzymes from Trichoderma sp. were suitable for use in cellulase recycling using readsorption to fresh substrates. The subsequent portion of our work next investigated the effects of changing the substrate on the recycling of the Celluclast (from Trichoderma) enzyme preparation. 117 T 1 1 1 1 1 r r M H Cel luc last EZ3 S p e z y m e K § M S U B C RO R1 R2 R3 R4 Enzyme recycling rounds Figure 30. Enzyme recycling for E P L P substrate (ethanol pretreated Lodgepole pine, Klason lignin 14.45%) Standard hydrolysis: 2% substrate in 50 m L buffer with 20 F P U Cellulase and 40 IU beta-glucosidase under 45°C, 150 rpm, 48 hr. Table 16. Characteristics of S E L P arid E P L P substrates by Klason analysis. Substrate Arabinose Galactose Glucose Xylose Mannose Klason lignin S E L P 0.077% 0.058% 53.44% 0.40% 0.47% 45.60% E P L P 0.04% 0.06% 90.46% 1.53% 1.53% 14.45% 3.3.5 Effect of different pretreatments on enzyme recycling Due to the inherent differences between steam explosion and organosolv pretreatment, it is expected that these two pretreatment methods would results in significantly different substrates (Table 16). It was of interest to compare the ability to recycle the Celluclast enzyme preparation during the hydrolysis of steam exploded Lodgepole pine (SELP) and ethanol pretreated Lodgepole pine (EPLP) . As shown in the previous section, when 118 hydrolyzing E P L P , cellulases retained the ability to achieve an 80% hydrolysis yield after four consecutive recycling rounds. However, only one round of hydrolysis could be performed on the S E L P substrate, since the hydrolysis yields decreased to 60% after the first recycling round (Figure 31). The higher lignin content of the S E L P substrate (45.60%) (Table 16) most likely resulted in an increase in enzyme adsorption that limited the ability of the surfactant to desorb cellulases into the supernatant for subsequent readsorption. Although post-treatments such as peroxide or oxygen can delignify the substrate to facilitate enzyme recycling (Pan et al. 2004), their addition would add considerable cost to the overall bioconversion process. In subsequent work Pan et al (2005) found that ethanol organosolv pretreatment is suitable for softwood substrates, as the organosolv process produces pulp substrates with low lignin content in the range of 6.4-27.4%, without the need for additional post-treatment steps. The organosolv softwood substrates could be effectively hydrolyzed within 48 hr by cellulases (20 F P U g"1 cellulose) (Pan et al. 2005a). As suggested by Eriksson et al (2002) and Pan et al. (2005), the effect of lignin on the enzymatic hydrolysis was mainly due to the non-productive binding of cellulases on lignin. The addition of a surfactant or protein (Bovine Serum Albumin) has been shown to reduce non-productive binding and improve enzymatic hydrolysis for lignin-rich substrates (Eriksson et al. 2002; Pan et al. 2005b; Yang and Wyman 2006). This approach also works for enzyme recycling, since the surfactant most likely reduces the adsorption of cellulases on lignin-rich hydrolysis residues, thus increasing the free 119 enzymes concentration in the solution after hydrolysis for subsequent readsorption on fresh substrates. Overall, the results here show that organosolv pretreatment (ethanol) was a good approach to produce a substrate (EPLP) suitable for the application of enzyme recycling. The higher lignin content and the nature of the S E L P substrate were most likely responsible for a greater amount of non-productive binding of cellulases to lignin. In the next part of our work we looked at the mechanism of the adsorption of cellulases to the lignin in S E L P and E P L P in much greater detail. Since E P L P was shown to be an effective substrate for the application of enzyme recycling with Tween 80, a simple economic analysis was performed to gain insight into the potential cost benefits of the recycling protocols developed in this work. 120 R0 R1 R2 R3 R4 Enzyme recycling rounds Figure 31. Enzyme recycling for S E L P substrate (steam exploded Lodgepole pine, 45.6% Klason lignin) Standard hydrolysis: 2% substrate in 50 m L buffer with 20 F P U Cellulase and 40 IU beta-glucosidase at 45°C, 150 rpm, 48 hr. 121 3.3.6 Economic analysis for enzyme recycling with the addition of Tween The work described so far in this thesis has shown that it is possible to use Tween to facilitate the recycling of cellulases during the hydrolysis of an ethanol pretreated Lodgepole pine (EPLP) substrate. In order to assess the value of enzyme recycling strategies, further analysis should be performed, including careful consideration of the some of the costs associated with implementing this technology. The following section summarizes a simple economic analysis of the hydrolysis process, with specific assumptions. In order to quantify the possible cost benefits of Tween addition, a simple economic analysis of the benefits of using Tween and enzyme recycling on the enzymatic hydrolysis was undertaken. As described in section 3.3 of this thesis, cellulases could be recycled for five consecutive rounds of hydrolysis. From these results, it can be inferred that it would be possible to reduce the cost of cellulases five fold during the hydrolysis of E P L P (Figure 30). However, the impact of the cost of Tween 80 and other necessary process and equipment modifications required to carry out enzyme recycling must also be taken into account. Cellulase enzymes typically cost approximately US $0.40-0.50 per gallon of ethanol production (Steele et al. 2005), while the cost of Tween 80 was estimated to be over the range of U S $0.25-$ 1.00 per kg. The desorption of cellulases from the hydrolysis residue showed a positive linear correlation with the concentration of Tween 80 in the system (Figure 20 D). Increasing Tween concentration wi l l therefore increase the concentration of free enzymes in the supernatant after lignocellulosic hydrolysis. In order to truly evaluate the economic 122 impact of adding Tween to the hydrolysis step, it is necessary to carry out a complete economic evaluation of the bioconversion process, including any capital and operation costs that become necessary as a result of Tween addition. However, at this stage of the thesis work, by making a few assumptions, a simplified economic analysis could be made to assess the potential enzyme cost savings that may result by adding Tween 80 to the hydrolysis. In this initial analysis it was assumed that the enzyme recycling process does not require additional capital and operational costs. Furthermore, the economic analysis focused mainly on the cellulases without considering P-glucosidase, since it has been shown that P-glucosidase could be effectively "recycled" by immobilization (Figure 46). Based on our experiments and literature values (Demirbas 2005), the cellulose conversion, ethanol stoichiometric yield and fermentation efficiency are summarized in Table 17. Using this information, it was calculated that, the production of 1 gallon of ethanol requires 10.6 kg of ethanol pre-treated Lodgepole pine (EPLP) substrate. Assuming the hydrolysis of E P L P is performed at a concentration of 6% (w/v) and a Tween 80 concentration of 0.2% (w/v), theoretically, 353 g of fresh Tween 80 would be required for four consecutive enzyme recycling rounds Table 17. Decreasing the concentration of Tween 80 in the hydrolysis system would result in a decrease in the number of effective enzyme recycling rounds (Table 18). 123 Table 17. Assumptions for bioconversion of an ethanol pretreated Lodgepole pine (EPLP) to produce ethanol (SHF). Technical assumption Value Units Ethanol Pretreated Lodgepole pine (Oven- Dry) 10.60 K g Cellulose (glucose) content 90 % Cellulose conversion and recovery efficiency (hydrolysis yield) 90 % Ethanol stoichiometric yield 51 % glucose fermentation efficiency 75 % Ethanol produced * 3.28* kg Ethanol stoichiometric yield and glucose fermentation efficiency obtained from Demirbas (2005). *3.28 kg of ethanol is equivalent to 1 U S gallon. Clearly there is a tradeoff between the initial enzyme costs and the cost of Tween 80 addition. In order to create a relationship between Tween 80 concentration and enzyme cost savings, a correlation between Tween 80 concentration and cellulase distribution from the hydrolysis of ethanol pretreated mixed softwood ( E P M S section 3.1) w i l l be used for the economic analysis during the hydrolysis of E P L P . The correlation between Tween 80 concentration and cellulase distribution is shown in Figure 32. After the first round of hydrolysis using a Tween 80 concentration of 0.2%, 96.4% of the initial cellulases remained in the liquid phase (supernatant). Using non-linear curve fitting (R 2=0.99), a relationship between Tween 80 concentration and the percentage of cellulases in the supernatant was established as: y = a - b \n{x + c), where y= percentage of cellulase (%), x=Tween 80 concentration (%) a=104.86±0.37, b=-5.39±0.26, c=0.00015±0.00006. This equation was used to establish a relationship between the proportion of the initial cellulase in the supernatant and the concentration of Tween 80. The values calculated using the equation are shown in Table 18. In section 3.1, free cellulase recovery by 124 readsorption on fresh substrates was 85% from the Langmuir model analysis. In the present example, with the addition of 0.2% Tween 80, 96.4% of the free cellulases remain in the supernatant after the first round hydrolysis, of which 81.9% (=96.4%x85%) can be recovered for use in a second round of hydrolysis (Table 18). Tween 80 (%) Figure 32. Effect of Tween 80 concentration on cellulase distribution during the hydrolysis of E P M S . Hydrolysis was carried out at 45°C for 24 hr with 2% of E P M S , the enzyme loading was 10 F P U Celluclast, 20 IU P-glucosidase per gram of cellulose. Table 18. Effect of Tween 80 concentration on the number of potential recycling rounds. 0.20% Tween 80 0.15% Tween 80 0.1% Tween 80 0.05% Tween 80 0.025% Tween 80 0.005% Tween 80 Free Readsorbed Free Readsorbed Free Readsorbed Free Readsorbed Free Readsorbed Free Readsorbed Hydrolysis cellulase cellulase cellulase cellulase cellulase cellulase cellulase cellulase cellulase cellulase cellulase cellulase rounds (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) 1 100 - 100 - 100 - 100 - 100 - 100 -2 96.4 81.9 94.6 80.4 92.2 78.4 88.7 75.4 85.0 72.3 76.5 65.0 3 79.0 67.1 76.1 64.7 72.3 61.4 66.9 56.8 61.4 52.2 49.7 42.2 4 64.7 55.0 61.2 52.0 56.6 48.1 50.4 42.9 44.4 37.7 21.0 17.8 5 53.0 45.1 49.2 41.8 44.4 37.7 38.0 32.3 32.1 27.2 3.7 3.2 6 43.5 36.9 39.5 33.6 34.8 29.6 28.7 24.4 23.2 19.7 0.1 0.1 Percentage of free cellulase was calculated from the free cellulase in the supernatant obtained from the previous round of hydrolysis; readsorbed cellulase was calculated assuming an 85% recovery using re-adsorption on fresh substrates as established in section 3.1. 125 126 In section 3.4, it was shown that the addition 0.2% Tween 80 permitted the recycling of cellulases in five rounds of hydrolysis of E P L P . According to the data in Table 18, Tween 80 at a 0.2% concentration results in 5 rounds of hydrolysis, which suggests that the threshold amount of free cellulases necessary to carry out an additional successful round of hydrolysis could be set at greater than or equal to 53%. According to this criterion, the addition of 0.15% or 0.10% Tween 80 enables four successful recycling rounds, while 0.05% or 0.025% Tween 80 would attain three successful recycling rounds. Using this information, the cost of cellulase enzymes in the entire process can be estimated from dividing the cost of cellulase enzymes (US $0.50/gallon ethanol) by the number of possible hydrolysis rounds. Considering the total cost of enzyme and Tween 80, the potential cost savings as a result of enzyme recycling using a range of Tween concentrations is shown in Table 19. Without the addition of Tween 80, the total initial cost of enzymes was set to US $0.50 per gallon ethanol. Increasing the Tween 80 concentration from 0.005% to 0.20%, allows the number of effective hydrolysis rounds to be increased from 2 to 5. As a result, the cost of enzymes could potentially be decreased further from $0.25 to $0.10. However the cost of Tween wi l l also increase from $0.01 to $0.35. Therefore, the highest enzyme cost savings that can be obtained in this economic analysis was 58% using a Tween 80 concentration of 0.025% Tween 80 assuming the price of Tween 80 is $1.0/kg. 127 Table 19. Enzyme cost savings with different concentrations of Tween 80. Tween Hydrolysis Cost of Tween Cost of Total cost of enzyme Enzyme cost 80 (%) rounds enzyme ($) (g) Tween ($) and Tween ($) savings (%) 0.000 1 0.50 0 0.00 0.50 0 0.005 2 0.25 8.8 0.01 0.26 48 0.025 3 0.17 44 0.04 0.21 58 0.050 3 0.17 88 0.09 0.25 49 0.100 4 0.13 176 0.18 0.30 40 0.150 4 0.13 264 ' 0.26 0.39 22 0.200 5 0.10 353 0.35 0.45 9 Enzyme cost savings= (Initial enzyme cost without recycling-Total cost of enzyme and Tween)/Initial enzyme cost without recycling. Price of Tween 80: U S $1.0/kg. Alkasrawi et al. (2003) performed a similar economic evaluation of the effect of Tween 20 on the production cost of ethanol in the simultaneous saccharification and fermentation process (SSF) (Alkasrawi et al. 2003). In their evaluation it was suggested that the maximum allowable price for Tween 20 could be U S $0.429/kg for one enzyme recycling round in SSF. Since an accurate industrial price for Tween 80 is unavailable, we have assumed the price of Tween 80 is in the range of $0.25-$1.00/kg. Following the same approach as above, the enzyme cost savings using Tween prices of $0.25/kg and $0.50/kg were recalculated (Figure 33), thus covering a range of Tween 80 prices. 128 100 • 90 -\ 80-\ 70 - | in ? 60-> CO 50-• CO 8 40-V 1. 30-N UJ 20 10-0-T 0.00 - • — T w e e n 80 $1.00/kg - • — T w e e n 80 $0.50/kg - A — T w e e n 80 $0.25/kg 0.05 0.10 0.15 Tween 80 (%) 0.20 Figure 33. Enzyme cost evaluation vs. concentration of Tween 80. From the economic analysis of enzyme recycling with the addition of Tween (Figure 33), if the price of Tween 80 is approximately $0.25/kg, Tween 80 addition could save -60% of the total enzyme cost (including cost of Tween) at concentrations in the 0.025% to 0.2% range. However, i f Tween 80 costs $1.00/kg, the potential cost savings decreases from 58% to 10% as the Tween concentration is raised from 0.025% to 0.2% (Figure 33). From the economic analysis it is apparent that, although the addition of Tween 80 wi l l result in savings in enzyme costs, Tween itself imposes an additional cost on the process. However, it may be possible to recycle both cellulase enzymes and Tween. A process was envisioned where the ethanol production cost could be reduced significantly by recirculating either the SSF product stream or the post-distillation stream in order to re-use buffer (Alkasrawi et al. 2002). It is likely that Tween 80 could be recycled as part of the product stream after distillation. Alkasrawi et al. (2002) compared ethanol yields and productivity of SSF with recirculation of the product stream after distillation to the same 129 process without recirculation. Using recirculation to recycle Tween, the ethanol production cost was reduced by 17% (Alkasrawi et al. 2002), therefore it is possible that the inclusion of surfactant recycling may partially offset the cost of Tween addition. In contrast to the work of Alkasrawi, the enzyme recycling protocol proposed in the section 3.3 of this thesis is based on a separate hydrolysis and fermentation. As shown in the schematic in Figure 34, our recycling protocol consists of a separate hydrolysis followed by readsorption of cellulases onto fresh substrates for enzyme recovery. The hydrolysis products are subsequently fermented to ethanol. Tween 80 may be recovered in the distillation step for subsequent re-use in a similar fashion as that shown in Figure 34. 130 Cellulase + Tween Fresh substrate Pretreatment I Enzymatic hydrolysis Recovered enzyme Filtration Recirculation streams (Tween) Distillation Ethanol Figure 34. Schematic flowsheet of ethanol production with enzyme recycling and recirculation of process streams. The potential to recycle both cellulases and Tween 80, enables a further reduction in Tween cost, assuming the process does not require additional capital and operation costs. If Tween 80 could be recycled for five rounds by recirculation during the bioconversion process, the cost of Tween would be $0.07 to produce 1 gallon of ethanol from the E P L P substrate (Table 20). Performing another cost analysis with integration of Tween 80 recycling, a 66% reduction in total enzyme cost could potentially be achieved under the assumed conditions. 131 Table 20. Various costs used in the economic analysis of enzyme recycling and Tween 80 recirculation for producing 1 gallon of ethanol. Economic analysis Value Units Enzyme cost without recycling 0.50 S Enzyme cost with recycling 0.10 $ Tween cost without recycling 0.35 $ Tween 80 cost with recycling 0.07 $ Total enzyme and Tween 80 cost 0.17 $ Enzyme cost savings 66% -Price of Tween 80: $1.0/kg, assuming hydrolysis of E P L P substrate is carried out at 6% consistency and 0.2% of Tween 80. Price of Tween 80: $1.00/kg. Assuming both enzyme and Tween was used for five recycling rounds. A simple economic evaluation for enzyme recycling with Tween 80 was presented here, which suggested that the addition of Tween could reduce the total enzyme costs for the overall process. However, an important factor governing the maximum achievable enzyme cost savings obtained during enzyme recycling is the cost of Tween. The possibility of recycling Tween 80 using a recirculation step improves the prospect for practical application of the enzyme recycling strategies developed in this thesis. Furthermore, the costs of immobilization of P-glucosidases wi l l also need to be considered in the overall recycling scheme. Once the specific capital and operational costs of enzyme recycling are established, a thorough economic analysis must be performed considering the impact of enzyme recycling on the entire bioconversion process. 132 3.3.7 Conclusions The recycling of cellulase enzymes was accomplished on ethanol pretreated Lodgepole pine (EPLP) with a 14.5% Klason lignin content with the addition of surfactants and subsequent readsorption of desorbed cellulases onto fresh substrates. Comparisons of various surfactants revealed that Tween 80 was the most effective detergent used in enzyme recycling. The efficient recycling of cellulases from Trichoderma was shown to be due to their high affinity for the E P L P substrate which facilitated subsequent readsorption to fresh substrate. The inability to effectively recycle cellulases during the hydrolysis of high-lignin content S E L P indicated the detrimental effect which lignin had on the proposed cellulase recycling strategies. The results from this study show that, cellulase recycling is feasible for ethanol-pretreated lignocellulosics with the addition of surfactant, as cellulases could be recycled and used for a total of five hydrolysis rounds. Based on a simple economic analysis, the addition of Tween 80 could save 60% of the total enzyme cost at concentrations in the 0.025% to 0.2% range (the price of Tween 80 is US $0.25/kg) assuming there are no additional capital or operation costs. However, it is recognized that the benefits of Tween 80 w i l l be heavily dependent on the price of Tween and cellulase enzymes. 133 3.4 Adsorption of cellulases on enzymatic lignin 3.4.1 Background As discussed throughout the thesis, recycling cellulases after enzymatic hydrolysis is one of the major strategies being investigated for its potential to decrease the cost of enzymatic hydrolysis of lignocellulosic substrates. From the results of the previous chapter, it was hypothesized that the amount, and possibly the nature of the residual lignin within the pretreated substrates could be responsible for the inability of cellulases to be recycled after hydrolysis. This was suggested as the cellulases were effectively recycled during hydrolysis of ethanol pretreated Lodgepole pine (EPLP) , but not when using steam exploded Lodgepole pine (SELP) . Since the steam exploded Lodgepole pine (SELP) substrate contained approximately 45% lignin, compared to 15% for ethanol (organosolv) pretreated Lodgepole pine (EPLP) (Table 15), it was of interest to compare the adsorption of cellulases to purified lignin preparations isolated from the two substrates at an equal dosage to determine if the two pretreatment methods would result in differences in cellulase adsorption. A summary of characteristics of S E L P lignin ( C E L - S E L P ) and E P L P lignin ( C E L - E P L P ) is shown in Table 21. 134 Table 21. Characteristics of CEL-SELP lignin and CEL-EPLP lignin by Klason analysis. Substrates Arabinose Galactose Glucose Xylose Mannose Nitrogen Klason < content lignin C E L - S E L P 0.074% 0.090% 3.31% 0.09% 0.32% 0.14% 85.76% C E L - E P L P 0.053% 0.139% 1.72% 0.058% 0.21% 0.29% 80.23% Enzymatic lignin was prepared by hydrolyzing the ethanol pretreated Lodgepole pine (EPLP) and steam exploded Lodgepole pine (SELP) . - 2 % of substrate in lOOmL Acetate buffer was hydrolyzed with 20 F P U Celluclast g"1 cellulose, 40 IU P- glucosidase g"1 cellulose, 0.2% Tween 80 for 48 hr. The hydrolysate was filtered by glass microfiber (Whatman G F / A ) . The collected residues were hydrolyzed again with the same amount of cellulase, P-glucosidase and Tween 80 for another 48 hr. The residues were recovered by filtration and combined with 0.2% Tween 80 for 2 hr incubation at 45°C. The collected lignin was washed with 0.5 L distilled water (~50°C). The final lignin was put into vacuum oven to air dry at room temperature. The lignin samples were ground and screened to a 60 mesh particle size. 3.4.2 Adsorption kinetics and isotherm The adsorption of cellulases to the purified lignin preparations were determined in a similar fashion as was reported in the previous work, characterizing the adsorption of cellulases to E P L P substrates (Table 14, section 3.3.3). The enzymatic lignin preparations ( C E L - E P L P and C E L - S E L P ) were suspended in acetate buffer (pH 4.8, 50mM). To a fresh suspension of each lignin preparation, Celluclast (T. reesei), Spezyme (T. reesei) or M S U B C (Penicillium sp.) was applied to study the enzyme adsorption kinetics. The adsorption of cellulases onto the C E L - E P L P and C E L - S E L P lignin reached equilibrium within approximately 3 hr (Figure 35). A s shown in Figure 35 A and Figure 35 B , a greater amount of the Celluclast from T. reesei adsorbed to the C E L - S E L P lignin than adsorbed to the C E L - E P L P lignin. The data was analyzed further using adsorption isotherm models. Previously, a similar approach using the Langmuir adsorption model was suggested to monitor the cellulase adsorption on enzymatic lignin isolated from dilute acid hydrolysis pretreated with explosion mixed hardwood (Bernardez et al. 1993) and enzymatic lignin from steam exploded spruce (Palonen et al. 2004). 136 0.16--j 0.14-"|» 0.12-~ 0.10-3 V) 0.08-O 0.06-c o u 0.04-0.02-- •— Celluclast on S E L P lignin - •— Spezyme on S E L P lignin - A — M S U B C on S E L P lignin 0.00--50 50 100 —I— 150 —I— 200 250 300 350 400 Time(min) A Cellulase adsorption kinetics on cellulolytic enzyme lignin from steam exploded Lodgepole pine ( C E L - S E L P ) . 0.14-^ 0.12-_l E "5) £ 0.10-"sf o 3 0.08-~ 0.06-c c g 0.04-I c o g 0.02 4 a. - •— Celluclast on E P L P lignin - •— Spezyme on E P L P lignin - A — M S U B C on E P L P Iginin 0.00--50 50 100 —I • 1— 150 200 Time(min) — i — 250 300 350 400 B Cellulase adsorption kinetics on cellulolytic enzyme lignin from ethanol pretreated Lodgepole pine ( C E L - E P L P ) . Figure 35. Cellulase adsorption kinetics on isolated C E L - S E L P lignin and C E L - E P L P lignin at 25°C. For adsorption kinetics, 30 mg lignin ( C E L - S E L P and C E L - E P L P ) samples were placed into microtubes in 5 m L acetate buffer (50 m M , p H 4.8) with 0.10-0.15 mg/mL of commercial cellulases (Celluclast, Spezyme and M S U B C ) for 6 hr incubation at 25°C. The microtubes were incubated on a rotary shaker. Aliquots of 0.15 m L were taken at 0, 15, 30, 60, 120, 180 and 360 min during the incubation. The protein content was determined using the ninhydrin assay. 137 Cellulase adsorption isotherms on CEL-lignin To gain an understanding of the adsorption of cellulases onto the isolated lignin preparations ( C E L - E P L P and C E L - S E L P ) , the system was analyzed using two different models. For the adsorption isotherm, different concentrations of cellulases were incubated with a - 2 % (2.0 g/100 mL) suspension of lignin in an acetate buffer at 25°C for 3 hr in order to reach equilibrium. The protein content in the supernatant was then determined. The Langmuir and Freundlich isotherms haven been applied extensively to study protein adsorption onto various particle surfaces (Luey et al. 1991; Moreno et al. 1982; Palonen et al. 2004). The Langmuir model is a non-linear model that assumes a monolayer and reversible adsorption, without interaction of the adsorbate (cellulase), uniform binding sites on the adsorbent (lignin), and a dilute adsorbate. r_r m a x*c l+KC Where C=concentration of unadsorbed cellulase in bulk solution (mg/mL) r concentration of adsorbed cellulase (mg/g lignin) T m a x =the maximal adsorbed cellulase (mg/g lignin) K = Langmuir constant (mL/mg enzyme) The linearized form of the Langmuir model can be expressed as: 1 1 1 1 - + • — rmax and K could be determined by non-linear curve fit or from the linear plot of ^ 1 versus — . C 138 Alternatively, the adsorption process can be monitored using the Freundlich model. The Freundlich isotherm is also a non-linear model, that assumes a heterogeneous surface on the adsorbent with varying affinities of the binding sites for the adsorbate, and it accounts for interactions between the adsorbed molecules (Sawalha et al. 2005). r = KFC" Where T concentration of adsorbed cellulase (mg/g lignin) C=concentration of unadsorbed cellulase in bulk solution (mg/mL) KF =Freundlich constant, n= the heterogeneity factor which has a lower value for most heterogeneous surfaces. The linearized form of the Freundlich isotherm is XnT = \nKF + n l n C KF and n are determined by data fit or from the linear plot of In Y versus In C. The analysis of the data using the Langmuir model results in a decent fit for the data (Table 22). Therefore, it was used to compare the adsorption of the various cellulase preparations. The adsorption isotherms of cellulase preparations from the Trichoderma based Celluclast and the Penicillium based M S U B C were studied on C E L - S E L P and C E L -E P L P at room temperature (Figure 36). The classical Langmuir adsorption was used to fit the experimental data (R 2 >0.85). The M S U B C showed no detectable enzyme adsorption on E P L P lignin, therefore, no adsorption isotherm was presented in Figure 36. The analysis of the adsorption data using the Langmuir model revealed significant differences between the various cellulases on the two lignin preparations (Figure 36). The data of the 139 Spezyme preparation is not shown since the results are similar to those obtained with Celluclast. The adsorption capacity of Celluclast on C E L - S E L P was rmax = 10.20 ± 5 . 9 3 m g / g lignin compared to T m a x = 2.73 ± 0 . 8 2 mg/g lignin which was obtained when C E L - E P L P was used. This indicated that varying the pretreatment method affected the adsorption of cellulases on lignin. A similar cellulase adsorption capacity (15 mg/g lignin) on enzymatic lignin from steam exploded spruce (CEL-SPS) could be calculated from previous work (Palonen et al. 2004). Ooshima et al. (1990) also reported the maximum adsorbed cellulase (Trichoderma) was 12.3mg/g lignin on enzymatic lignin from steam exploded hardwood at 220°C. For the different cellulases on C E L - S E L P , the adsorption capacity of Celluclast was rmax = 10.20 ± 5 . 9 3 mg/g; compared to rmax = 4.87 ±1 .99 mg/g for M S U B C , further demonstrating the lower overall adsorption of Penicillium cellulases on lignocellulosic substrates. Similar results have been reported by Berlin et al. (2005), where they found that Penicillium cellulases had lower adsorption on enzymatic lignin from organosolv-pretreated Douglas fir than Trichoderma cellulases. Since Langmuir constants represent equilibrium affinity constants, they have often been used to evaluate the affinity of cellulose binding domains (CBD) on different cellulose substrates (Creagh et al. 1996; Tomme et al. 1998). Using Langmuir constants, the relative affinity and adsorption capacities of the substrates can be compared. Although the C E L - S E L P had a higher adsorption capacity (higher rm a x) than did the C E L - E P L P , it is interesting to note that the Celluclast enzyme preparation had a higher affinity for the C E L - E P L P than the C E L - S E L P (higher K value) (Table 22). Since the lignin preparations ( C E L - S E L P and C E L - E P L P ) were isolated using a cellulase hydrolysis, nitrogen content measurements could be used to confirm the calculation. The C E L - S E L P 140 had a nitrogen content of 0.14%, while C E L - E P L P had nitrogen content 0.29%, thus indicating that the C E L - E P L P bound almost twice the amount of cellulase as did the C E L - S E L P at equivalent conditions. Since both lignin preparations underwent the same washing step, the results suggest that cellulase had a higher affinity for the C E L - E P L P lignin. Although the C E L - S E L P bound a greater amount of cellulases, the C E L - E P L P had a higher affinity for cellulases. One possible reason for these results is that the C E L - S E L P from steam explosion had more binding sites on the particle surface due to the preservation of functional groups (phenolic hydroxyl and benzyl) during the pretreatment. It is possible that the C E L - E P L P obtained from organosolv pretreatment possessed fewer binding sites for cellulases since a larger proportion of the lignin is solubilized during the pretreatment. Due to a higher solubilization of lignin during the pretreatment, the potentially hydrophobic nature of the resulting C E L - E P L P lignin may show a greater affinity for cellulase due to hydrophobic interactions. Since the experiments here were performed at room temperature, the next set of experiments evaluated the effects of varying the temperature on cellulase adsorption. 141 Figure 36. Cellulase adsorption isotherms on C E L - S E L P and C E L - E P L P lignins at 25°C (solid line from Langmuir model, dotted line from Freundlich). For the adsorption isotherm, different concentrations of cellulases were incubated with 2% (2g/100ml) of lignin in 50 m M acetate buffer at 25°C for 3 hr to reach equilibrium. The protein content in the supernatant was determined for the non-adsorbed cellulase. The adsorbed cellulase was calculated from the difference of initial cellulase content and non-adsorbed cellulase content in the supernatant. The classical Langmuir theory of adsorption was also applied to cellulase adsorption on lignin samples. Table 22. Adsorption isotherm parameters for different cellulases on C E L - S E L P and C E L - E P L P lignin at 25°C. Cellulase Langmuir Freundlich adsorption rmax (mg/g) K (mL/mg) R2 KF N R2 Celluclast on C E L - 10.20±5.93 3.21 ±3.45 0.847 10.34±2.68 0.67±0.20 0.887 SELP lignin M S U B C on C E L - S E L P 4.87±2.30 4.95±5.31 0.842 5.74±2.25 0.60±0.28 0.775 lignin Celluclast on C E L - 2.73±0.95 6.44±5.57 0.819 3.12+1.05 0.49±0.23 0.732 EPLP lignin —2 R coefficient of determination 142 3.4.3 Effect of temperature on cellulase adsorption onto lignin Since it was shown previously that temperature had a significant effect on the desorption of cellulases from lignin-rich hydrolysis residues (section 3.2.4), it was also of interest to gauge the effects of varying temperature on cellulase adsorption on isolated lignin. Different concentrations of cellulases (0.05 mg/mL-0.4 mg/mL) were incubated with -20 mg lignin in l m L 50 m M acetate buffer for 3hr. Cellulase adsorption on lignin was measured at 4°C, 25°C, and 45°C (Figure 37 and Figure 38). Although temperature had only minimal effects on the adsorption of cellulases on the C E L - E P L P and C E L - S E L P , increasing the temperature to 45°C resulted in a slight increase in the adsorption on the C E L - S E L P and C E L - E P L P . The results of these experiments were used for a thermodynamic analysis of the adsorption process. Thermodynamic analysis The adsorption process can be analyzed using thermodynamic parameters such as free energy change ( A G 0 ) enthalpy change (AH0 ) and entropy change ( A S 0 ) . The value A G 0 can be obtained using equilibrium constants from the adsorption isotherms of Celluclast with S E L P and E P L P at increasing temperatures. The Gibbs free energy change of the adsorption process is shown in the following equation (Butt et al. 2003; Ozer et al. 2004): A G 0 = -RTIn K Where A G 0 is the free energy change (J mol" 1); R is the universal gas constant, 8.314 (J mof 1 K" 1 ) ; and T is the absolute temperature, K is the Langmuir constant. A negative A G 0 value indicates an exothermic reaction while a positive A G 0 value indicates an endothermic reaction. 143 The equation used to calculate A G 0 shows that the spontaneity of the enzyme adsorption process is governed by both the entropy and enthalpy of the reaction (Ozer et al . 2004). The A G 0 values were calculated using Langmuir constants. Typically, the A G 0 for physical adsorption is in the range of 0 to -20 kJ/mol, and the A G 0 for chemical adsorption (adsorption which results from chemical bond formation) in the range of -80 to -400 kJ/mol (Jaycock and Parfitt 1981). The results obtained for cellulase adsorption on lignin were from -29.9 to -37.1kJ/mol (Table 23) indicating that the binding of cellulases on lignin was the result of physical adsorption (Zheng et al. 2005). The negative values of A G 0 suggests that the process of cellulase adsorption on both S E L P and E P L P lignins is a spontaneous process. In the current discussion, it is difficult to make comparisons to previous work, since these results represent the first occasion where the adsorption of cellulases on S E L P and E P L P lignins were analyzed using this thermodynamic approach. The use of thermodynamic parameters to determine cellulase adsorption is a potentially useful method for evaluating the ability of pretreatments to produce substrates with a lower tendency to adsorb cellulases. Since the C E L - S E L P showed a greater capacity to adsorb cellulases, further work was done to determine i f the interaction of cellulases with the C E L - S E L P could be decreased by varying the temperature and ionic strength of the reaction and/or by the addition of a surfactant. 144 10-E 8 94 74 c 'c Dl a. 6-_i </) 5-c o 4H (A «; 3. 0> 2 4 i ' 8 0. •a Celluclast adsorption at 48C Celluclast adsorption at 25-C Celluclast adsorption at 459C —i 1 1 , 1 1 1 1 1— 0.0 0.1 0.2 0.3 0.4 Free cellulase (mg/mL) 0.5 Figure 37. Cellulase adsorption isotherms on C E L - S E L P lignin at different temperatures. 10 E Q. 6 . c o <u (0 n 4. a> u •o 5 2 i_ o (A •o < 04 Celluclast adsorption at 4 9C • Celluclast adsorption at 25SC A Celluclast adsorption at 452C —1— 0.0 —r 0.1 0.2 0.3 Free cellulase (mg/mL) 0.4 0.5 Figure 38. Cellulase adsorption isotherms on C E L - E P L P lignin at different temperatures. 145 Table 23. Thermodynamic parameters of cellulase adsorption on C E L - S E L P and C E L -E P L P lignin. Cellulase on C E L - S E L P lignin Cellulase on C E L - E P L P lignin Temperature K A G 0 K A G 0 R2 (K) (mL/mg) (kJ/mol) (mL/mg) (kJ/mol) 277 7.41 ±2.46 -29.95 0.98 7.92+1.95 -30.10 0.99 298 4.66±2.51 -31.07 0.95 6.44±4.80 -31.87 0.88 318 5.89±3.38 -33.78 0.96 20.91 ±12.05 -37.13 0.92 Before calculating the Gibbs free energy change from Langmuir constants, the volumetric concentration was converted to molar concentration (Liu 2006). Tiered = KMA, MAis the molecular weight of cellulases (assume average MA =60000), A G 0 =RT\nK . R 2 : coefficient of determination. 3.4.4 Effect of ionic strength on cellulase adsorption onto lignin The effect of ionic strength on cellulase adsorption onto C E L - S E L P lignin was evaluated by increasing the N a C l concentration from 0 M to 0 .4M (Figure 26). Different concentrations of cellulases were incubated with -20 mg of C E L - S E L P lignin in acetate buffer at 25°C. The cellulase adsorption onto C E L - S E L P lignin decreased with increasing N a C l concentration as the N a C l concentration was raised to 0 .1M. The effect of ionic strength on protein adsorption is a complex process. The salt ions could either compete against the protein for binding sites, or change the configuration of the protein resulting in a decrease in hydrophobic interactions between the protein and the solid surface (Can and Guner 2006). Similarly, the adsorption capacity of P-lactoglobulin on a hydrophobic silicon surface decreased from 1.06 ug/cm 2 to 0.486 ng/cm 2 as the N a C l concentration was raised from 0 .1M to 0 .5M (Luey et al. 1991). In the present study, the decrease in cellulase adsorption onto C E L - S E L P occurred at a low concentration of N a C l (<0.1M). As the concentration of N a C l increased from 0 .1M 146 to 0 .4M, cellulase adsorption did not decrease further. The ineffectiveness of N a C l concentrations >0.1 M could be due to a saturation of the lignin surface at low salt concentrations. A s the salt concentration is increased past the saturation point, the salt cannot further compete with the cellulase protein for binding sites since all the binding sites are depleted. Since the previous chapters showed that Tween 80 addition was highly successful in improving cellulase desorption from substrates, the next set of experiments examined the effects of Tween 80 on the adsorption of cellulases on C E L - S E L P lignin. Figure 39. Effect of ionic strength on cellulase adsorption on C E L - S E L P lignin (Celluclast). 147 3.4.5 Effect of Tween 80 on cellulase adsorption onto lignin The effect of Tween 80 on the adsorption of cellulases (Celluclast) to C E L - S E L P was studied by applying Tween 80 at 0% to 0.2% to the adsorption process. The adsorption capacity decreased from 2.5 mg/g lignin to 1.0 mg/g lignin as the Tween 80 concentration was raised (Figure 40). The decrease in cellulase adsorption was most likely due to the competitive adsorption of Tween 80 onto the hydrophobic lignin surface. Eriksson et al. (2002) also suggested that non-ionic surfactants such as Tween 20 and Tween 80 could reduce enzyme adsorption because the hydrophobic portion of the surfactant binds to lignin to prevent the non-productive binding of cellulases to lignin. A similar effect of Tween 80 on the adsorption of insulin onto the hydrophobic surface of silane-modified quartz slides has been reported (Mollmann et al. 2005), as it was hypothesized that Tween 80 displaced adsorbed insulin from the silane-modified quartz (Mollmann et al. 2005). A t a 0.2% concentration of Tween 80, the cellulase adsorption onto C E L - S E L P decreased to -1.0 mg/g lignin. The results indicate that the addition of Tween 80 to a lignocellulosic hydrolysis system could reduce enzyme adsorption onto lignin and potentially improve the enzyme recovery efficiency. 148 3.00 0 0.05 0.1 0.2 Tween 80 concentration (%) Figure 40. Effect of surfactant concentration on cellulase adsorption on C E L - S E L P lignin (Celluclast). -0.13 mg/mL of cellulase was incubated with -20 mg lignin in 1.0 m L of 50 m M acetate buffer at 25°C for 3 hr to reach equilibrium. The protein content in supernatant was determined for the non-adsorbed cellulase. 3.4.6 Conclusions The adsorption of cellulases on lignin plays an important role for lignocellulosic hydrolysis because the adsorption influences the efficiency of hydrolysis and the potential to recover enzymes. Cellulase adsorption on C E L - S E L P and C E L - E P L P lignin preparations followed the Langmuir adsorption model. The adsorption capacity of cellulases (Celluclast) on C E L - S E L P was 3 times higher than that obtained with C E L -E P L P lignin. The negative value of Gibbs free energy change (~ -30kJ/mol) suggested that cellulase adsorption onto both C E L - S E L P and C E L - E P L P lignin was a spontaneous process. The different results for the change in free energy indicate that the adsorption of cellulases on lignin differs significantly depending on the pretreatment applied to the substrate. 149 Ionic strength had a significant effect on cellulase adsorption on C E L - S E L P lignin, as cellulase adsorption decreased with increasing N a C l concentration from 0 to 0.4 M . Similarly, the addition of Tween 80 affected cellulase adsorption onto lignin as increasing Tween 80 concentrations (0-0.2%) resulted in a decrease in cellulase adsorption. In fact, the adsorption of cellulases onto C E L - S E L P was reduced by 60% with the addition of 0.1%-0.2% Tween 80. Overall, the results show that each pretreatment method has unique effects on the downstream cellulase-substrate lignin interactions that occur during hydrolysis. The finding of interactions between cellulases and lignins from different pretreatments w i l l help us design suitable strategies for cellulase recovery. The next part of our work focuses on evaluating the potential of enzyme immobilization techniques to enable the recovery of beta-glucosidases for subsequent reuse during lignocellulosic hydrolysis. 150 3.5 Immobilization of p-glucosidase 3.5.1 Background Lignocellulosic feedstocks currently under consideration for a biomass-to-ethanol process include agricultural residues such as corn stover, fast-growing hardwoods such as poplar, and softwood residues from forest industries. Typical bioconversion schemes include a pretreatment step to improve substrate accessibility, followed by enzymatic hydrolysis of the cellulose component to produce glucose for fermentation. As mentioned in the introduction, efficient hydrolysis of cellulose requires the synergistic activities of three types of enzymes. Endo-P-1, 4-glucanases hydrolyze accessible regions on cellulose chains to provide new sites for attack by exo-acting cellobiohydrolases which remove successive cellobiose units from newly-created chains ends. Finally, p-glucosidase hydrolyzes cellobiose, and smaller amounts of higher cellooligomers, to glucose. Although fungal cellulase preparations contain endogenous P-glucosidase activity, the levels of this enzyme are generally insufficient to prevent the accumulation of cellobiose, resulting in product inhibition of endoglucanases and cellobiohydrolases. Product inhibition is particularly problematic when high substrate consistencies are used in order to produce more concentrated glucose syrup for fermentation. Consequently, cellulase preparations are typically supplemented with extra P-glucosidase (Duff and Murray 1996; Tengborg et al. 2001). However, supplementation adds to the already substantial cost for enzymes in the bioconversion process. Although the cost of commercial cellulase preparations has been reduced significantly in recent years, enzyme costs are still an obstacle to full-scale process commercialization (Anonymous 2004). 151 Immobilization on an inert carrier offers the prospect of significant cost savings by facilitating enzyme recycling through multiple cycles of batch-wise hydrolysis. Also , enzyme immobilization frequently results in improved thermostability or resistance to shear inactivation. In the investigation described below, we have examined the use of Eupergit® C to immobilize of a commercial p-glucosidase preparation, Novozym 188, for use in lignocelluloses hydrolysis. Eupergit® C , an epoxy-activated immobilization support, has been identified as one of the most useful carriers for covalent immobilization of enzymes for industrial applications because of its ability to stabilize protein conformation by multi-point attachment (Katchalski-Katzir and Kraemer 2000; Mateo et al. 2000). Eupergit C is a polymer containing a high density of oxirane groups on its surface to bind with the amino groups on protein molecules (Katchalski-Katzir and Kraemer 2000). Moreover, Eupergit C can also effectively bind protein through their carboxyl groups. This provides a variety of binding sites for enzyme immobilization thus facilitating a multi-point attachment between the immobilized enzymes and Eupergit C (Mateo et al. 2002b). Eupergit C can covalently bind enzymes within a wide p H range from 1.0 to 12.0 (Katchalski-Katzir and Kraemer 2000). The immobilization procedure is simple compared to other carriers such as Amberlite X A D , agarose and silica gel. Considering these characteristics Eupergit C can be regarded as an ideal enzyme carrier for immobilization of various industrial enzymes (Martin et al. 2003; Mateo et al. 2002b; Seip et al. 1994; Torres-Bacete et al. 2000). 3.5.2 Immobilization of p-glucosidase on Eupergit C The kinetics of immobilization was determined from a plot of protein remaining in solution vs. time of incubation with Eupergit C (Figure 41). There was a rapid decline in the level of soluble protein during the first 3 hr of incubation; followed by a slow decline over the subsequent 34 hr. Approximately 96% of the added protein was immobilized at 36 hr (Figure 38). Based on this preliminary experiment and published data for other enzymes, P-glucosidase was routinely immobilized by incubation with Eupergit C at 25°C for 36 hr. Extended incubation with Eupergit C was previously shown to enhance multi-point attachment (Mateo et al. 2000). In the absence of added glucose or blocking agents (control), the immobilization efficiency was estimated to be 12%, thus indicating significant enzyme deactivation (Table 24). In a previous study, a similar loss of activity was attributed to interactions between oxirane groups on Eupergit C and the active site of the enzyme (Levitsky et al. 1999). 153 Figure 41. Time course of P-glucosidase immobilization. Eupergit C (0.5g dry wt) was added to 10 m L 1.0 M potassium phosphate buffer, p H 7.0, containing 5.5mg of Novozym 188. The reaction mixture was incubated at 25°C for up to 36hr with gentle shaking. The amount of immobilized protein was determined from the difference between the total protein added and the amount remaining in solution after immobilization. Table 24. Effect of glucose and blocking agents on the activity of immobilized P~ glucosidase. Immobilization conditions Relative activity 3 Immobilization efficiency (%) b Control (no additives) 1.00 12 plus 1% glucose 1.68 20 plus 1% glucose and 1% B S A 2.51 30 plus 1% glucose and 2 - M E 1.56 19 "Observed activity/activity of control (0.86 IU g"1 Eupergit C). b (Activi ty of immobilized enzyme/enzyme activity added) xlOO. In attempts to improve the immobilization yield, two strategies were evaluated. In the first, the immobilization reaction was performed in the presence of glucose, a competitive 154 inhibitor of fj-glucosidases (Riou et al. 1998). Previous reports have demonstrated the effectiveness of this strategy: for example, improved immobilization efficiency for penicillin acylase was obtained by addition of the competitive inhibitor phenoxyacetic acid (Torees-Bacete et al. 2000). Addition of 1% glucose to p-glucosidases immediately prior to incubation with Eupergit C increased the immobilization yield to 20%, a 1.7-fold improvement over the control. Presumably, glucose reduces the reaction of oxirane groups with active site by steric hindrance. In an attempt to improve the yield further, a second strategy involving addition of B S A or 2-mercaptoethanol was tested. B S A was previously shown to improve the immobilization yield of penicillin acylase, presumably by reacting with residual oxirane groups. Similarly, thiols such as 2 - M E or dithiothreitol have been shown to improve immobilization yields for penicillin amidase (Katchalske-Katzir and Kraemer 2000), although in the case of penicillin acylase, negative effects were reported (Torees-Bacete et al. 2000). Preincubation with 1% glucose, followed by incubation in presence of 1% B S A after immobilization, increased the immobilization yield to 30%, a 2.5-fold improvement over the control. In contrast, preincubation with 1% glucose, followed by incubation in presence of 1% 2-mercaptoethanol, gave no improvement over incubation with glucose alone (Table 24). As a result of these experiments, 1% glucose and 1% B S A were included in all subsequent immobilization reactions. 3.5.3 Thermal stability of free and immobilized enzyme Typically, P-glucosidases are prone to thermal deactivation under process operating conditions, but immobilization has been shown to improve thermal stability substantially. For example, the half-life of an Aspergillus phoenicis P-glucosidase at 65°C was 155 improved ~ 100-fold by covalent immobilization on chitosan by glutaraldehyde cross-linking (Bissett and Sternberg 1978). In the current study, the stabilities of both free and Eupergit C-immobilized P-glucosidase were determined by preincubation at 45°C, 65°C and 85°C in the absence of substrate, followed by assay at 45°C using p - N P G . The free and immobilized enzyme were both stable at 45 °C with no significant loss of activity after 60 hr of preincubation (Figure 42). Instability was observed for the free enzyme at 65°C (>90% loss of activity by 8 hr). However, stability was significantly improved as a result of immobilization (Figure 42). Although there was a rapid decline in activity during the first 8 hr of preincubation, the immobilized enzyme showed good stability during the remainder of the 60 hr preincubation period. Neither the free nor the immobilized enzyme was stable at 85°C. 80->^  60-> o CO 0) 40 • > a> 1 20-OH -A— Free enzyme 45°C -V— Immobilized enzyme 45°C -•— Free enzyme 65°C -o— Immobilized enzyme 65°C -A— Free enzyme 85°C -V— Immobilized enzyme 85°C 10 - r -20 ~r— 30 - 1 40 50 60 Time (hr) Figure 42. Thermal stabilities of free and immobilized P-glucosidase. Enzyme was incubated in 50 m M sodium acetate buffer, p H 4.8, and then assayed at 50°C using p-N P G . The activity of the free enzyme act 0 hr was 1.93 IU mg"1 protein; the activity of the immobilized enzyme was 3.50 I U mg"1 biocatalyst. 156 Improved thermostability following immobilization has been demonstrated for numerous enzymes (Bissett and Sternberg 1978; De Queiroz et al. 2002; Dourado et al. 2002; Saville et al. 2004). Immobilization can involve non-covalent interaction, or covalent interaction through single point or multi-point attachment. The efficacy of Eupergit C immobilization is related to its ability to achieve multi-point attachments which tend to stabilize the three-dimensional structure of the enzyme at elevated temperatures (Katchalski-Katzir and Kraemer 2000; Mateo et al. 2000). Therefore, it is likely that the initial rapid loss of activity for P-glucosidase seen at 65°C (Figure 42) was due to denaturation of a population of enzyme molecules immobilized by single point attachment. This is consistent with a previous report on penicillin G acylase which showed that the half life of the multipoint immobilized enzyme was 100 fold higher than that of single-point immobilized enzyme (Mateo et al. 2002a). In this context, it is noted that A. niger P-glucosidase is a dimeric enzyme composed of two identical -100 kDa subunits (Seidle et al. 2004); presumably, thermal stabilization requires multipoint attachment of both monomers. 3.5.4 Effect of immobilization on pH optimum A shift in p H optimum following immobilization has been reported for several enzymes (Lamb and Stuckey 2000; Martin et al. 2003). For example, the p H optimum of a P-glucosidase changed from 6.0 to 5.3 following immobilization on Eupergit® C (Torees-Bacete et al. 2000). Similar results were observed for immobilization of pectinase on Eupergit® C carriers (Spagna et al. 1993). Shifts in p H optima are attributed to charge effects and are related to the ionization of groups on the carrier. For example, the p H optimum of an A. phoenicis P-glucosidase changed from 4.8 to 3.5 after immobilization 157 on an Alumina-based carrier, but remained unchanged when immobilized on chitosan (Bissett and Sternberg 1978). In the current study, the p H optimum of Eupergit C -immobilized P-glucosidase was the same as that of the free enzyme (pH 5.0) and the p H profiles were indistinguishable in the range p H 3.0-8.0 (Figure 43). A n acidic p H optimum is advantageous when P-glucosidase is used in conjunction with other enzymes having similar optima, such as fungal cellulases. 3 4 5 6 7 8 PH Figure 43. p H optima of free and immobilized P-glucosidase. p-glucosidase activity was determined using p-NPG in sodium acetate buffer, p H 3.0-8.0. 3.5.5 Determination of kinetic parameters When the kinetic parameters for free and immobilized P-glucosidase are assessed (Table 25), the K m of the free enzyme was shown to be -1.1 m M , typical of fungal p-glucosidase (Bissett and Sternberg 1978). The K m value for the immobilized enzyme was approximately ten-fold higher than the free enzyme, suggesting a reduced affinity for the 158 substrate. Similar increases in apparent K m have been reported for P-glucosidases immobilized on other carriers (Levitsky et al. 1999; Torres-Bacete et al. 2000) and are attributed to alteration of the enzymes three-dimensional structure and mass transfer limitations. Interestingly, immobilization resulted in an eight-fold increase in V m a x . Typically, immobilized enzymes have lower V m a x values than their free counterparts (Petzelbauer et al. 2002; Saville et al. 2004), but higher V m a x values following immobilization have been reported for P-galactosidases (Lamb and Stuckey 2000; Petzelbauer et al. 2002) and a cellulase complex (Yuan et al. 1999). The mechanistic basis for this positive feature requires further investigation. Table 25. Kinetic constants for the free and immobilized P-glucosidase. Enzyme K m , cellobiose (mM) V m a x , cellobiose (u.mol mg'min" 1 ) Free P-glucosidase 1.1 296 Immobilized p-glucosidase 10.8 2430 Initial rates were estimated from the extent of cellobiose hydrolysis during 5 min incubation at 45°C. Reaction mixtures contained 0.625-10 m M cellobiose and 0.185 mg free P-glucosidase, or 0.1 mg P-glucosidase immobilized on 25 mg Eupergit C , in 2.1 m L in 50 m M sodium acetate buffer, p H 4.8. Kinetic parameters were calculated from Lineweaver-Burk plots. 3.5.6 Hydrolysis of cellulosic and lignocellulosic substrates using immobilized p-glucosidase The hydrolysis of Whatman N o . l filter paper (2% w/v) by Celluclast in the presence of immobilized P-glucosidase is shown in Figure 44. Data for hydrolysis in the presence of free P-glucosidase, and in the absence of any added P-glucosidase, are also included. A l l reaction mixtures contained 20 F P U Celluclast g"1 cellulose. The loading of free P-159 glucosidase was 40 I U (5.73 mg total protein)/ g cellulose. The loading of immobilized (3-glucosidase was 5.73, 2.87 or 1.43 mg total protein g"1 cellulose (i.e., 100, 50 and 25% of the free p-glucosidase protein loading). These quantities of immobilized P-glucosidase correspond to 4.4, 2.2 and 1.1 IU, respectively. Time(h) Figure 44. Effect of free and immobilized P-glucosidase on the hydrolysis of filter paper. Reaction mixtures containing 2% (w/v) cellulose in 25 m L 50 m M sodium acetate buffer (pH 4.8), were incubated at 45°C for up to 48 hr. The cellulase loading was 20 F P U Celluclast g"1 cellulose. The loading of free P-glucosidase was 40 IU g"1 cellulose. The loading of immobilized P-glucosidase was 5.73, 2.87 or 1.43 mg total protein g"1 cellulose (i.e., 100%, 50% and 25% of the free P-glucosidase protein loading). These quantities of immobilized P-glucosidase correspond to 4.4, 2.2 and 1.1 I U respectively. In the absence of added P-glucosidase, Celluclast hydrolyzed -36% of the cellulose to glucose in 48 hr. Conversions of 70% and 80% were obtained with the addition of 2.87 and 1.43 mg immobilized P-glucosidase g"1 cellulose, respectively. The conversion of cellulose to glucose in the presence of 5.7 mg immobilized P-glucosidase g"1 cellulose 160 (82%) was similar to that seen when free P-glucosidase was used at the same protein loading (88% conversion). The hydrolysis of other cellulosic (Avicel) and lignocellulosic substrates including organosolv pulp (ethanol pretreated mixed softwood, E P M S ) and acetic acid pulp (acetic acid pretreated Douglas fir) in the presence of free or immobilized P-glucosidase is shown in Figure 45; data for filter paper hydrolysis is included for comparison. The loading of free P-glucosidase was 40 I U (5.73 mg total protein)/ g cellulose. The loading of immobilized P-glucosidase was about 6.0 mg total protein g"1 cellulose, corresponding to 5 IU of activity. The results demonstrate that immobilized P-glucosidase was as effective as the free enzyme, when used at the same protein loading, with typical lignocellulosic substrates (organosolv and acetic acid softwood pulps), as well as with model cellulosic substrates. 161 T ' 1 ' r V///A Free enzyme OCsN^ I Immobilized enzyme Filter paper Av ice l Organoso lv pulp Acet ic ac id pulp Different substrates Figure 45. Effect of free and immobilized P-glucosidase on the hydrolysis of various cellulosic and lignocellulosic substrates. Reaction mixtures containing 2% (w/v) cellulose in 25 m L 50 m M sodium acetate buffer (pH 4.8) were incubated at 45°C for 48 hr. The cellulase loading was 20 F P U Celluclast g"1 cellulose. The recovery and operational stability of immobilized P-glucosidase was examined during six successive rounds of hydrolysis (i.e., initial hydrolysis plus five rounds of recycle) using Douglas fir pretreated by acetic pulping as substrate. Cellulose to glucose conversion in the presence of immobilized P-glucosidase during the first round of hydrolysis (-80%) was similar to that seen in the free P-glucosidase control (Figure 46), while only -34% conversion was seen without added P-glucosidase. Conversion decreased slightly in the second round to -75% and then remained relatively stable during the subsequent four rounds, thus demonstrating good recovery and operational stability of the immobilized enzyme over the total 288 hr experimental period. 162 20 j 18 -Free No 1 2 3 4 5 6 3-g P-g Hydrolysis round Figure 46. Operational stability of immobilized p-glucosidase during hydrolysis of acetic acid pretreated Douglas fir. The reaction mixture contained 20 F P U cellulase and 4.4 I U immobilized p-glucosidase g"1 cellulose. After 48 hr hydrolysis, the immobilized p-glucosidase was recovered by brief centrifugation at lOOOg. Residual substrate remained in suspension and was removed by decantation. The recovered immobilized enzyme was then used to supplement a second reaction mixture containing fresh substrate and cellulase. The process was repeated for a total of six rounds of hydrolysis. In the free enzyme control (free P-g), immobilized P-glucosidase was replaced by 40 IU of p-glucosidase. A second control without P-glucosidase supplementation (no p-g) was also included. 3.5.7 Conclusions The applicability of the immobilized p-glucosidase to industrial-scale bioconversion wi l l be heavily influenced by enzyme cost and process design. The cost of cellulase preparations for bioconversion has been reduced substantially in recent years, largely as a result of lowered production costs (Anonymous 2004). However, commercial cellulase preparations are generally deficient in p-glucosidase and typically require 163 supplementation, thereby introducing additional costs (Xiao et al. 2004). In general, immobilized enzymes are considered inappropriate for use with insoluble substrates (Katchalske-Katzir and Kraemer 2000). However, there is a potential for using recycled immobilized P-glucosidase to relieve the end product inhibition due to the accumulation of cellobiose and other soluble sugars in a sequential batch-wise or semi-batch-wise saccharification processes. Our results have demonstrated good operational stability for Eupergit C-immobilized A. niger P-glucosidase over at least six rounds of batch-wise lignocelluloses hydrolysis. Commercial scale lignocellulosic saccharification wi l l likely involve hydrolysis at high substrate consistency (>10% cellulose w/v) in order to avoid the generation of dilute product streams. However, batch hydrolysis at high consistency is severely limited by product inhibition due to the accumulation of both glucose and cellobiose. A possible solution is provided by saccharification schemes involving batch-wise or continuous removal of soluble sugar products by ultrafiltration (Duff and Murray 1996). Since conventional yeasts are unable to ferment cellobiose, immobilized P-glucosidase would provide a useful means to convert the cellobiose in the filtrate to glucose, thereby significantly improving fermentation efficiency. 164 4 Conclusions and future work Conclusions As was hypothesized for this project, effective strategies were developed to recycle both free and bound cellulases and beta-glucosidases during lignocellulosic hydrolysis. Free cellulases in solution were recovered via readsorption onto fresh substrates. Bound cellulases were recovered by adding surfactants during the enzymatic hydrolysis, and P~ glucosidases were recycled by immobilization on an inert carrier. Early on it was evident that the presence of lignin was detrimental to the recovery of cellulases, as the hydrolysis of pure cellulose resulted in a 90% recovery of cellulases compared to only a 50% recovery during the hydrolysis of ethanol pretreated mixed softwood (EPMS) , which contained an appreciable amount of lignin. The Langmuir adsorption isotherm was used to determine that 82% was the theoretical maximum recoverable enzyme in the liquid phase using readsorption to fresh substrates. In practice it was possible to recover 85% of the enzyme protein in the liquid phase; therefore, the theoretical calculations were shown to be capable of predicting the amount of recoverable free enzymes. This was the first occasion where a theoretical analysis with practical confirmation has been presented for the readsorption of cellulases on fresh substrates. Although free enzymes were recovered effectively, approximately 50% of the protein remained bound to the lignin-rich hydrolysis residue. The experimental conditions for the surfactant-facilitated desorption of bound enzymes were optimized to maximize enzyme recovery using response surface methodology. The data analysis revealed that temperature, p H and the concentration of surfactant (Tween 80) were significant factors affecting the desorption of bound enzymes. The optimal 165 conditions were 45°C, and p H 5.3, while a 0.2% Tween 80 concentration was chosen based on the experimental data and previous results in the literature. Using the information from the optimization experiments, a novel integrated recovery process was developed where surfactant was added at the beginning of hydrolysis to aid enzyme desorption from substrates, and fresh substrate was added at the end of hydrolysis to recover enzymes desorbed into the liquid phase. Various surfactants and cellulase preparations from T. reesei and Penicillium sp. were compared for their performance in the integrated recycling process. None of the surfactants tested could outperform the nonionic surfactants, Tween 80 and 20. Due to their higher affinity for lignocellulosic substrates, T. reesei cellulases were easier to recycle than Penicillium cellulases. Comparing the performance of the integrated recycling process on S E L P and E P L P substrates showed that the E P L P substrate was more responsive to cellulase recycling. Therefore, organosolv pretreatment with ethanol is effective in producing a substrate suitable for the downstream application of enzyme recovery protocols developed in this study. Since lignin was shown to influence the desorption of cellulases for subsequent recovery, lignin preparations isolated from ethanol (organosolv) pretreated ( C E L - E P L P ) and steam-pretreated Lodgepole pine ( C E L - S E L P ) were compared for their ability to adsorb cellulases. The C E L - S E L P lignin showed a higher adsorption capacity than did the C E L -E P L P lignin. The addition of N a C l and Tween 80 reduced the adsorption of cellulases to C E L - S E L P , most likely due to competition with cellulases for binding sites on the lignin. Elucidating cellulase adsorption characteristics was an effective approach for developing 166 methods for recovering enzymes using readsorption to fresh substrates. However, the recycling of P-glucosidase had yet to be addressed. Since P-glucosidase does not possess a cellulose binding domain it cannot be readsorbed on fresh substrate. The recovery of P-glucosidase was accomplished by immobilization on an inert carrier Eupergit C. The immobilized P-glucosidase was shown to be recoverable, and stable over six consecutive rounds of hydrolysis. In summary, enzyme recycling during the hydrolysis of lignocellulosic substrates has been evaluated during this study. A theoretical model to predict enzyme recovery via readsorption to fresh substrate was developed and verified experimentally. The enzyme recycling strategies in this study were applied successfully to an organosolv pretreated Lodgepole pine substrate (EPLP) for five consecutive rounds of hydrolysis. The influence of lignin on cellulase adsorption and recycling was also investigated and an approach to reduce the effect of lignin on cellulase recycling was proposed. With an appropriate combination of cellulase preparation, surfactant and pretreatment method, it is possible to successfully recycle cellulase enzymes to potentially reduce the cost of enzyme hydrolysis during bioconversion of lignocellulosics to ethanol. Future work This study has addressed some aspects of enzyme recycling during lignocellulosic hydrolysis; however some questions remain unanswered in terms of application and understanding. The following suggestions would be helpful in guiding the direction of further work in the field of enzymatic hydrolysis and cellulase recycling. 167 Pilot-scale assessment of one-step enzyme recycling strategy In this study enzyme recycling was studied at the laboratory scale utilizing lignocellulosic substrates suspended at approximately 2% consistency (g pulp/g pulp+g water) in a total volume of 100 m L . It would be a worthwhile study to scale-up the hydrolysis and enzyme recycling to volumes ranging from 5-10 L at consistencies in the 9-10% range to increase glucose concentrations for downstream fermentation. Challenges that may be encountered during the scale-up to high-consistency include energy consumption, diffusion limitations, and the removal of lignin-rich hydrolysis residues. The results from the pilot-trial should also provide realistic data for a techno-economic assessment of the process. Techno-economic assessment of enzymatic hydrolysis and enzyme recycling The research carried out in the thesis has showed that it is feasible to recycle enzymes during the hydrolysis of lignocellulosic substrates. According to this study, effective enzyme recycling requires the addition of a surfactant and a substrate pretreated by a suitable method (organosolv). In order to evaluate the benefits of the approach of adding surfactant to an organosolv pretreated substrate, a basic techno-economic assessment of this process must be undertaken. This techno-economic assessment wi l l provide information regarding the potential reductions in overall cost that could be realized by applying this enzyme recycling strategy. In addition to the economic assessment and pilot trial, some fundamental aspects of the enzyme recycling results in this study should be examined in greater detail. 168 Chemical and physical nature of lignin Since this work has shown that various pretreatment methods affect the adsorption of cellulases to lignin, fundamental studies would be indispensable for understanding the role of lignin in enzymatic hydrolysis and enzyme recycling. Explanations for the variations in cellulase adsorption on lignin with different pretreatment methods may include lignin content differences, changes in lignin molecular weight, and the depletion or fortification of specific functional groups (hydroxyl groups, phenolic groups and methoxyl groups). The chemical constituents of lignin located at the surface of substrates also govern its hydrophilicity, thereby affecting enzyme adsorption. The fundamental information gained from these studies should aid in the selection of pretreatment processes and surfactants that enable effective enzymatic hydrolysis and recycling. Producing p-glucosidase with CBD In this study, P-glucosidase was recycled by immobilization while other cellulases (endoglucanases and exoglucanases) were recovered by a one step readsorption process. If p-glucosidase could also be recovered in the same readsorption step as the other cellulases, the recycling process could be simplified while retaining the ability to recycle all the component enzymes. A C B D (Cellulose Binding Domain) gene has already been combined with a P-glucosidase gene, and the P-glucosidase possessing the C B D was shown to have a high affinity for cellulose (Ong et al. 1993; Ong et al. 1994). This new P-glucosidase can potentially facilitate the application of a one-step recycling process where P-glucosidase can be recycled together with other cellulase components. This type of innovation may help improve the prospects of realizing the major goal of reducing the cost of enzymes utilized for bioconversion. 169 References Abdel-Fattah, A . F . , Osman, M . Y . and Abdel-Naby, M . A . , 1997. Production and immobilization of cellobiase from Aspergillus niger A20 . Chemical Engineering Journal. 68(2-3): 189-196. Adams, R . A . , 1995. A complete courses Calculus. Addison-Wesley Publishers Limited, Don M i l l s , Ontartio, 170-173 pp. Aden, A . et al., 2002. Lignocellulosic biomass to ethanol process design and economics utilizing co-current dilute acid prehydrolysis and enzymatic hydrolysis for corn stover. http://wwwl.eere.energy.gov/biomass/pdfs/32438.pdf.. National Renewable Energy Laboratory and Harris Group. Adler, E . , 1977. Lignin Chemistry - Past, Present and Future. Wood Science and Technology. 11(3): 169-218. Ahrens, C D . , 1985. Meteorology Today. West Publishing Co. , St. Paul, Minnesot, 18-19 pp. Alfani , F. , Gallifuoco, A . , Saporosi, A . , Spera, A . and Cantarella, M . , 2000. Comparison of S H F and SSF processes for the bioconversion of steam-exploded wheat straw. Journal of Industrial Microbiology & Biotechnology. 25(4): 184-192. Alkasrawi, M . et al., 2003. The effect of Tween-20 on simultaneous saccharification and fermentation of softwood to ethanol. Enzyme and Microbial Technology. 33(1): 71-78. Alkasrawi, M . , Galbe, M . and Zacchi, G . , 2002. Recirculation of process streams in fuel ethanol production from softwood based on simultaneous saccharification and fermentation. Applied Biochemistry and Biotechnology. 98: 849-861. Al len , S.G. et al., 2001. A comparison of aqueous and dilute-acid single-temperature pretreatment of yellow poplar sawdust. Industrial & Engineering Chemistry Research. 40(10): 2352-2361. Andrade, J.D., 1985. Principles of protein adsorption. In: J.D. Andrade (Editor), Surface and Interfacial Aspects of Biomedical Polymers. Plenum Press, New York, pp. 1-80. Andreaus, J. , Azevedo, H . and Cavaco-Paulo, A . , 1999. Effects of temperature on the cellulose binding ability of cellulase enzymes. Journal of Molecular Catalysis B -Enzymatic. 7(1-4): 233-239. 170 Anonymous, 2004. N R E L make further biomass-to-ethanol progress, Ethanol Producer Magazine, B B I International, pp. p l4 . Arato, C , Pye, E . K . and Gjennestad, G . , 2005. The lignol approach to biorefining of woody biomass to produce ethanol and chemicals. Applied Biochemistry and Biotechnology. 121: 871-882. Araujo, A . , Ward, O.P. and Dsouza, J., 1991. Use of Mutation Strategies Applied to Aspergillus-Terreus Atcc 52430 to Obtain Mutants with Improved Cellulase Productivity. Biotechnology Techniques. 5(4): 283-288. Arunachalam, U . and Kel l is , J.T., 1996. Folding and stability of endoglucanase III, a single-domain cellulase from Trichoderma reesei. Biochemistry. 35(35): 11379-11385. Australian Greenhouse Office, 2006. Cooperative Research Centre for Greenhouse Accounting Glossary, http://www.greenhouse.crc.org.au/glossary.cfm. Bailey, M . J . , Buchert, J. and Vi ikar i , L . , 1993. Effect of p H on Production of Xylanase by Trichoderma-Reesei on Xylan-Based and Cellulose-Based Media. Applied Microbiology and Biotechnology. 40(2-3): 224-229. Bailey, M . J . and Nevalainen, K . M . H . , 1981. Induction, Isolation and Testing of Stable Trichoderma-Reesei Mutants with Improved Production of Solubilizing Cellulase. Enzyme and Microbial Technology. 3(2): 153-157. Baker, J.O. et al., 1995. Synergism between purified bacterial and fungal cellulases. Enzymatic Degradation of Insoluble Carbohydrates. 618: 113-141. Bandi, S., K i m , Y . J . , Sa, S.O. and Chang, Y . K . , 2005. Statistical approach to development of culture medium for ansamitocin P-3 production with Actinosynnema pretiosum A T C C 31565. Journal of Microbiology and Biotechnology. 15(5): 930-937. Basmadjian, D . , 1997. The little adsorption book: a practical guide for engineers and scientists. C R C Press, Boca Raton, Fla., 7 pp. Becker, D . et al., 2001. Engineering of a glycosidase Family 7 cellobiohydrolase to more alkaline p H optimum: the p H behaviour of Trichoderma reesei CeI7A and its E223S/A224H/L225V/T226A/D262G mutant. Biochemical Journal. 356: 19-30. Beldman, G . , Voragen, A . G . J . , Rombouts, F . M . and Pilnik, W. , 1988. Synergism in Cellulose Hydrolysis by Endoglucanases and Exoglucanases Purified from Trichoderma-Viride. Biotechnology and Bioengineering. 31(2): 173-178. Belter, P .A . , Cussler, E . L . and Hu, W.-S. , 1988. Bioseparations: downstream processing for biotechnology. Wiley, New York, 155 pp. 171 Berlin, A . et al., 2005a. Evaluation of novel fungal cellulase preparations for ability to hydrolyze softwood substrates - evidence for the role of accessory enzymes. Enzyme and Microbial Technology. 37(2): 175-184. Berlin, A . et al., 2005b. Weak lignin-binding enzymes - A novel approach to improve activity of cellulases for hydrolysis of lignocellulosics. Applied Biochemistry and Biotechnology. 121: 163-170. Bernardez, T .D. , Lyford, K . , Hogsett, D . A . and Lynd, L .R . , 1993. Adsorption of Clostridium-Thermocellum Cellulases onto Pretreated Mixed Hardwood, Avice l , and Lignin. Biotechnology and Bioengineering. 42(7): 899-907. Bezerra, R . M . F . and Dias, A . A . , 2005. Enzymatic kinetic of cellulose hydrolysis -Inhibition by ethanol and cellobiose. Applied Biochemistry and Biotechnology. 126(1): 49-59. Bigelow, M . and Wyman, C .E . , 2002. Cellulase production on Bagasse pretreated with hot water. Applied Biochemistry and Biotechnology. 98: 921-934. Bissett, F. and Sternberg, D . , 1978. Immobilization of Aspergillus Beta-Glucosidase on Chitosan. Applied and Environmental Microbiology. 35(4): 750-755. Borden, J.R., Lee, Y . Y . and Yoon, H . H . , 2000. Simultaneous saccharification and fermentation of cellulosic biomass to acetic acid. Applied Biochemistry and Biotechnology. 84-6: 963-970. Boussaid, A . , Robinson, J., Ca i , Y . J . , Gregg, D.J . and Saddler, J.R., 1999. Fermentability of the hemicellulose-derived sugars from steam-exploded softwood (Douglas fir). Biotechnology and Bioengineering. 64(3): 284-289. Boussaid, A . and Saddler, J .N. , 1999. Adsorption and activity profiles of cellulases during the hydrolysis of two Douglas fir pulps. Enzyme and Microbial Technology. 24(3-4): 138-143. Boussaid, A . L . , Esteghlalian, A . R . , Gregg, D.J . , Lee, K . H . and Saddler, J .N. , 2000. Steam pretreatment of Douglas-fir wood chips - Can conditions for optimum hemicellulose recovery still provide adequate access for efficient enzymatic hydrolysis? Applied Biochemistry and Biotechnology. 84-6: 693-705. Brown, J .A. , Col l in , S.A. and Wood, T . M . , 1987. Development of a Medium for High Cellulase, Xylanase and Beta-Glucosidase Production by a Mutant Strain (Ntg-Iii/6) of the Cellulolytic Fungus Penicillium-Pinophilum. Enzyme and Microbial Technology. 9(6): 355-360. Brownell, H . H . and Saddler, J .N. , 1987. Steam Pretreatment of Lignocellulosic Material for Enhanced Enzymatic-Hydrolysis. Biotechnology and Bioengineering. 29(2): 228-235. 172 Bruley, A . J . , Jones, D . and Canada. Water Pollution Control Directorate, 1974. Training manual on the basic technology of the pulp and paper industry and its waste reduction practices. Water Pollution and Control Directorate, [Ottawa], 1370 pp. Bucciarelli, B . , Jung, H . G . , Ostry, M . E . , Anderson, N . A . and Vance, C P . , 1998. Wound response characteristics as related to phenylpropanoid enzyme activity and lignin deposition in resistant and susceptible Populus tremuloides inoculated with Entoleuca mammata (Hypoxylon mammatum). Canadian Journal of Botany-Revue Canadienne De Botanique. 76(7): 1282-1289. Burns, D.S. , Ooshima, H . and Converse, A . O . , 1989. Surface-Area of Pretreated Lignocellulosics as a Function of the Extent of Enzymatic-Hydrolysis. Applied Biochemistry and Biotechnology. 20-1: 79-94. Busto, M . D . , Ortega, N . and Perez-Mateos, M . , 1998. Characterization of microbial endo-beta-glucanase immobilized in alginate beads. Acta Biotechnologica. 18(3): 189-200. Butt, H.-J . , Graf, K . and Kappl, M . , 2003. Physics and chemistry of interfaces. Wiley-V C H ;John Wiley, Weinheim Chichester. Calsavara, L . P . V . , Moraes, F.F. and Zanin, G . M . , 2000. Thermal stability and energy of deactivation of free and immobilized cellobiase. Brazilian Journal of Chemical Engineering. 17(4-7): 841-847. Can, H . K . and Guner, A . , 2006. Investigation of adsorption-desorption dynamism of bovine serum albumin on crosslinked N,N'-diethylaminoethyl dextran microbeads: Solution phase. Journal of Applied Polymer Science. 99(5): 2288-2299. Carrard, G . and Linder, M . , 1999. Widely different off rates of two closely related cellulose-binding domains from Trichoderma reesei. European Journal of Biochemistry. 262(3): 637-643. Castanon, M . and Wilke , C.R. , 1980. Adsorption and Recovery of Cellulases During Hydrolysis of Newspaper. Biotechnology and Bioengineering. 22(5): 1037-1053. Castanon, M . and Wilke, C.R. , 1981. Effects of the Surfactant Tween-80 on Enzymatic-Hydrolysis of Newspaper. Biotechnology and Bioengineering. 23(6): 1365-1372. Castellanos, O.F., Sinitsyn, A . P . and Vlasenko, E . Y . , 1995. Evaluation of Hydrolysis Conditions of Cellulosic Materials by Penicillium Cellulase. Bioresource Technology. 52(2): 109-117. Chaabouni, S.E., Hadjtaieb, N . , Mosrati, R. and Ellouz, R., 1994. Preliminary Assessment of Penicillium-Occitanis Cellulase - a Further Useful System. Enzyme and Microbial Technology. 16(6): 538-542. 173 Chand, P., Aruna, A . , Maqsood, A . M . and Rao, L . V . , 2005. Novel mutation method for increased cellulase production. Journal of Applied Microbiology. 98(2): 318-323. Chandrakant, P. and Bisaria, V . S . , 1998. Simultaneous bioconversion of cellulose and hemicellulose to ethanol. Critical Reviews in Biotechnology. 18(4): 295-331. Chang, V . S . , Kaar, W . E . , Burr, B . and Holtzapple, M . T . , 2001a. Simultaneous saccharification and fermentation of lime-treated biomass. Biotechnology Letters. 23(16): 1327-1333. Chang, V . S . , Nagwani, M . , K i m , C . H . and Holtzapple, M . T . , 2001b. Oxidative lime pretreatment of high-lignin biomass - Poplar wood and newspaper. Applied Biochemistry and Biotechnology. 94(1): 1-28. Chauvet, J . M . and Lamy, F., 1990. Determination of Soluble Lignin and Proteins in the Presence of Each Other. Analytica Chimica Acta. 235(2): 299-306. Chen, S. and Wayman, M . , 1993. Use of Sorbose to Enhance Cellobiase Activi ty in a Trichoderma-Reesei Cellulase System Produced on Wheat Hydrolysate. Biotechnology Techniques. 7(5): 345-350. Cherry, J. , 2003. Cellulase cost reduction. Enzyme sugar platform project review. http://wwwl.eere.energv.gov/biomass/pdfs/novozymes esp review.pdf.. Novozymes Biotech, Inc., Golden, Colorado. Chum, H . L . , Johnson, D . K . and Black, S.K. , 1990. Organosolv Pretreatment for Enzymatic-Hydrolysis of Poplars .2. Catalyst Effects and the Combined Severity Parameter. Industrial & Engineering Chemistry Research. 29(2): 156-162. Conklin-Brittain, N . L . , Dierenfeld, E.S. , Wrangham, R .W. , Norconk, M . and Silver, S.C., 1999. Chemical protein analysis: A comparison of Kjeldahl crude protein and total ninhydrin protein from wild, tropical vegetation. Journal of Chemical Ecology. 25(12): 2601-2622. Converse, A . O . , 1993. Substrate factors limiting enzymatic hydrolysis. In: J. Saddler (Editor), Bioconversion of Forest and agricultural plant residues. C A B International, Oxon, pp. 93-106. Converse, A . O . , Ooshima, H . and Burns, D.S., 1990. Kinetics of Enzymatic-Hydrolysis of Lignocellulosic Materials Based on Surface-Area of Cellulose Accessible to Enzyme and Enzyme Adsorption on Lignin and Cellulose. Applied Biochemistry and Biotechnology. 24-5: 67-73. Creagh, A . L . , Ong, E . , Jervis, E . , Kilburn, D . G . and Haynes, C . A . , 1996. Binding of the cellulose-binding domain of exoglucanase Cex from Cellulomonas fimi to insoluble microcrystalline cellulose is entropically driven. Proceedings of the National Academy of Sciences of the United States of America. 93(22): 12229-12234. 174 De Queiroz, A . A . A . , Vargas, R.R. , Higa, O.Z. , Ribeiro, R.R. and Vitolo, M . , 2002. Lactam-amide graft copolymers as novel support for enzyme immobilization. Journal of Applied Polymer Science. 84(4): 767-777. Demirbas, A . , 2005. Bioethanol from cellulosic materials: A renewable motor fuel from biomass. Energy Sources. 27(4): 327-337. Deshpande, M . V . , 1992. Ethanol-Production from Cellulose by Coupled Saccharification Fermentation Using Saccharomyces-Cerevisiae and Cellulase Complex from Sclerotium-Rolfsii U V - 8 Mutant. Applied Biochemistry and Biotechnology. 36(3): 227-234. Deshpande, M . V . and Eriksson, K . E . , 1984. Reutilization of Enzymes for Saccharification of Lignocellulosic Materials. Enzyme and Microbial Technology. 6(8): 338-340. Dourado, F. , Bastos, M . , Mota, M . and Gama, F . M . , 2002. Studies on the properties of Celluclast/Eudragit L-100 conjugate. Journal of Biotechnology. 99(2): 121-131. Duff, S.J.B., Moritz, J .W. and Andersen, K . L . , 1994. Simultaneous Hydrolysis and Fermentation of Pu lp -Mi l l Primary Clarifier Sludge. Canadian Journal of Chemical Engineering. 72(6): 1013-1020. Duff, S.J.B., Moritz, J .W. and Casavant, T.E. , 1995. Effect of Surfactant and Particle-Size Reduction on Hydrolysis of Deinking Sludge and Nonrecyclable Newsprint. Biotechnology and Bioengineering. 45(3): 239-244. Duff, S.J.B. and Murray, W . D . , 1996. Bioconversion of forest products industry waste cellulosics to fuel ethanol: A review. Bioresource Technology. 55(1): 1-33. Egyhazi, A . , Dienes, D . , Reczey, K . and Simandi, B . , 2004. Examination of cellulase enzyme production by Trichoderma reesei rut C30 using supercritical carbon dioxide cell disruption. Chemical and Biochemical Engineering Quarterly. 18(3): 257-261. Eklund, R. and Zacchi, G . , 1995. Simultaneous Saccharification and Fermentation of Steam-Pretreated Wi l low. Enzyme and Microbial Technology. 17(3): 255-259. Energy Information Administration, 2000. International energy outlook, Energy Information Administration: Available from the Supt. of Docs., U .S . G.P.O. , Washington, D . C . Energy Information Administration, 2005. International energy outlook, Energy Information Administration: Available from the Supt. of Docs., U .S . G.P.O. , Washington, D . C . Eon-Duval, A . , Gumbs, K . and Ellett, C , 2003. Precipitation of R N A impurities with high salt in a plasmid D N A purification process: Use of experimental design to 175 determine reaction conditions. Biotechnology and Bioengineering. 83(5): 544-553. Eriksson, T., Borjesson, J. and Tjerneld, F., 2002. Mechanism of surfactant effect in enzymatic hydrolysis of lignocellulose. Enzyme and Microbial Technology. 31(3): 353-364. Esterbauer, H . , Steiner, W. , Labudova, I., Hermann, A . and Hayn, M . , 1991. Production of Trichoderma Cellulase in Laboratory and Pilot Scale. Bioresource Technology. 36(1): 51-65. Farrell, A . E . et al., 2006. Ethanol can contribute to energy and environmental goals. Science. 311(5760): 506-508. Fogler, H.S. , 1999. Elements of chemical reaction engineering. Prentice-Hall international series in the physical and chemical engineering sciences. Prentice Hal l PTR, Upper Saddle River, N J , 252 pp. Fujishima, S., Yaku , F. and Koshijima, T., 1987. Immobilization of Beta-Glucosidase Using Wood Residue of Enzymatic-Hydrolysis. Mokuzai Gakkaishi. 33(12): 992-993. Fujishima, S., Yaku, F. and Koshijima, T., 1991. Immobilization of Beta-Glucosidase Using Wood Residue for Enzymatic-Hydrolysis. Journal of Carbohydrate Chemistry. 10(4): 671-680. Gadgil , N . J . , Daginawala, H.F. , Chakrabarti, T. and Khanna, P., 1995. Enhanced Cellulase Production by a Mutant of Trichoderma-Reesei. Enzyme and Microbial Technology. 17(10): 942-946. Galbe, M . and Zacchi, G . , 2002. A review of the production of ethanol from softwood. Applied Microbiology and Biotechnology. 59(6): 618-628. Gan, Q., Al len , S.J. and Taylor, G . , 2003. Kinetic dynamics in heterogeneous enzymatic hydrolysis of cellulose: an overview, an experimental study and mathematical modelling. Process Biochemistry. 38(7): 1003-1018. Gan, Q., Al len , S.J. and Taylor, G . , 2005. Analysis of process integration and intensification of enzymatic cellulose hydrolysis in a membrane bioreactor. Journal of Chemical Technology and Biotechnology. 80(6): 688-698. Ghose, T . K . , 1987. Measurement of Cellulase Activities. Pure and Applied Chemistry. 59(2): 257-268. Gilarranz, M . A . , Rodriguez, F. and Oliet, M . , 2000. Lignin behavior during the autocatalyzed methanol pulping of Eucalyptus globulus - Changes in molecular weight and functionality. Holzforschung. 54(4): 373-380. 176 Gilkes, N . R . et al., 1992. The Adsorption of a Bacterial Cellulase and Its 2 Isolated Domains to Crystalline Cellulose. Journal of Biological Chemistry. 267(10): 6743-6749. Girard, D.J . and Converse, A . O . , 1993. Recovery of Cellulase from Lignaceous Hydrolysis Residue. Applied Biochemistry and Biotechnology. 39: 521-533. Gollapalli , L . E . , Dale, B . E . and Rivers, D . M . , 2002. Predicting digestibility of ammonia fiber explosion (AFEX)-treated rice straw. Applied Biochemistry and Biotechnology. 98: 23-35. Gomez, J . M . , Romero, M . D . and Fernandez, T . M . , 2005. Immobilization of beta-Glucosidase on carbon nanotubes. Catalysis Letters. 101(3-4): 275-278. Governors' Ethanol Coalition, 2005. Ethanol from biomass America's 21 Century Transportation Fuel, http:www.ethanol-gec.org. Goyal , A . , Ghosh, B . and Eveleigh, D. , 1991. Characteristics of Fungal Cellulases. Bioresource Technology. 36(1): 37-50. Gregg, D . and Saddler, J .N. , 1995. Bioconversion of lignocellulosic residue to ethanol: Process flowsheet development. Biomass & Bioenergy. 9(1-5): 287-302. Gregg, D.J . , Boussaid, A . and Saddler, J .N. , 1998. Techno-economic evaluations of a generic wood-to-ethanol process: Effect of increased cellulose yields and enzyme recycle. Bioresource Technology. 63(1): 7-12. Grethlein, H . E . and Converse, A . O . , 1991. Common Aspects of A c i d Prehydrolysis and Steam Explosion for Pretreating Wood. Bioresource Technology. 36(1): 77-82. Grous, W.R. , Converse, A . O . and Grethlein, H.E . , 1986. Effect of Steam Explosion Pretreatment on Pore-Size and Enzymatic-Hydrolysis of Poplar. Enzyme and Microbial Technology. 8(5): 274-280. Gruno, M . , Valjamae, P., Pettersson, G . and Johansson, G . , 2004. Inhibition of the Trichoderma reesei cellulases by cellobiose is strongly dependent on the nature of the substrate. Biotechnology and Bioengineering. 86(5): 503-511. Hao, X . C . , Y u , X . B . and Yan , Z . L . , 2006. Optimization of the medium for the production of cellulase by the mutant Trichoderma reesei WX-112 using response surface methodology. Food Technology and Biotechnology. 44(1): 89-94. Haynes, C . A . , Sliwinsky, E . and Norde, W. , 1994. Structural and Electrostatic Properties of Globular-Proteins at a Polystyrene Water Interface. Journal of Col lo id and Interface Science. 164(2): 394-409. 177 Hayward, T . K . , Hamilton, J., Tholudar, A . and M c M i l l a n , J.D., 2000. Improvements in titer, productivity, and yield using Solka-Floc for cellulase production. Applied Biochemistry and Biotechnology. 84-6: 859-874. Helle, S.S., Duff, S.J.B. and Cooper, D . G . , 1993. Effect of Surfactants on Cellulose Hydrolysis. Biotechnology and Bioengineering. 42(5): 611-617. Henrissat, B . , Driguez, H . , Viet, C . and Schulein, M . , 1985. Synergism of Cellulases from Trichoderma-Reesei in the Degradation of Cellulose. Bio-Technology. 3(8): 722-726. Hirs, C .H .W. , 1967. Detection of Peptides by Chemical Methods. Methods in Enzymology. 11: 325-329. Hogan, C M . and Meshartree, M . , 1990. Recycle of Cellulases and the Use of Lignocellulosic Residue for Enzyme-Production after Hydrolysis of Steam-Pretreated Aspenwood. Journal of Industrial Microbiology. 6(4): 253-261. Holtzapple, M . T . and Humphrey, A . E . , 1984. The Effect of Organosolv Pretreatment on the Enzymatic-Hydrolysis of Poplar. Biotechnology and Bioengineering. 26(7): 670-676. Holtzapple, M . T . , Ripley, E.P. and Nikolaou, M . , 1994. Saccharification, Fermentation, and Protein Recovery from Low-Temperature Afex-Treated Coastal Bermudagrass. Biotechnology and Bioengineering. 44(9): 1122-1131. Hsu, T . A . (Editor), 1996. Pretreatment of biomass. Handbook on bioethanok.productiona nd utilization. Taylor & Francis, Washington, 180-212 pp. Hsu, T .A . , Himmel , M . , Schell, D. , Fanner, J. and Berggren, M . , 1996. Design and initial operation of a high-solids, pilot-scale reactor for dilute-acid pretreatment of lignocellulosic biomass. Applied Biochemistry and Biotechnology. 57-8: 3-18. Hui , J . P . M . et al., 2001. Characterization of cellobiohydrolase I (Cel7A) glycoforms from extracts of Trichoderma reesei using capillary isoelectric focusing and electrospray mass spectrometry. Journal of Chromatography B-Analytical Technologies in the Biomedical and Life Sciences. 752(2): 349-368. Ismail, A . , Soultani, S. and Ghoul, M . , 1998. Optimization of the enzymatic synthesis of butyl glucoside using response surface methodology. Biotechnology Progress. 14(6): 874-878. Iyer, P . V . and Lee, Y . Y . , 1999. Product inhibition in simultaneous saccharification and fermentation of cellulose into lactic acid. Biotechnology Letters. 21(5): 371-373. Iyer, P .V . , W u , Z .W. , K i m , S.B. and Lee, Y . Y . , 1996. Ammonia recycled percolation process for pretreatment of herbaceous biomass. Applied Biochemistry and Biotechnology. 57-8: 121-132. 178 Jackson, L .S . , Joyce, T .W. , Heitmann, J .A. and Giesbrecht, F . G . , 1996. Enzyme activity recovery from secondary fiber treated with cellulase and xylanase. Journal of Biotechnology. 45(1): 33-44. Jaycock, M . J . and Parfitt, G .D. , 1981. Chemistry of interfaces. Halsted Press, New York, 279 pp. Jensen, J .K. , 2003. The Bioprocess challenge enzymatic cellulose hydrolysis. http://www 1 •eere.eiiergy.gov/biomass/pdfs/genencor esp review.pdf., Genencor International, Inc. Jia, J., Dyer, P.S., Buswell , J .A. and Peberdy, J.F., 1999. Cloning of the cbhl and cbhl l genes involved in cellulose utilisation by the straw mushroom Volvariel la volvacea. Molecular and General Genetics. 261(6): 985-993. Jorgensen, H . , Eriksson, T., Borjesson, J., Tjerneld, F. and Olsson, L . , 2003. Purification and characterization of five cellulases and one xylanase from Penicil l ium brasilianum IBT 20888. Enzyme and Microbial Technology. 32(7): 851-861. Jorgensen, H . , Morkeberg, A . , Krogh, K . B . R . and Olsson, L . , 2005. Production of cellulases and hemicellulases by three Penicillium species: effect of substrate and evaluation of cellulase adsorption by capillary electrophoresis. Enzyme and Microbial Technology. 36(1): 42-48. Jorgensen, H . and Olsson, L . , 2006. Production of cellulases by Penicillium brasilianum IBT 20888 - Effect of substrate on hydrolytic performance. Enzyme and Microbial Technology. 38(3-4): 381-390. Juhasz, T., Szengyel, Z. , Reczey, K . , Siika-Aho, M . and Vi ikar i , L . , 2005. Characterization of cellulases and hemicellulases produced by Trichoderma reesei on various carbon sources. Process Biochemistry. 40(11): 3519-3525. Juhasz, T., Szengyel, Z . , Szijarto, N . and Reczey, K . , 2004. Effect of p H on cellulase production of Trichoderma reesei R U T C30. Applied Biochemistry and Biotechnology. 113-16: 201-211. Kaar, W . E . and Holtzapple, M . T . , 1998. Benefits from Tween during enzymic hydrolysis of corn stover. Biotechnology and Bioengineering. 59(4): 419-427. Kaar, W . E . and Holtzapple, M . T . , 2000. Using lime pretreatment to facilitate the enzymic hydrolysis of corn stover. Biomass & Bioenergy. 18(3): 189-199. Kadam, K . L . and Keutzer, W.J . , 1995. Enhancement in Cellulase Production by Trichoderma-Reesei Rut-C30 Due to Ci t r ic-Acid. Biotechnology Letters. 17(10): 1111-1114. 179 Kadam, K . L . and Newman, M . M . , 1997. Development of a low-cost fermentation medium for ethanol production from biomass. Applied Microbiology and Biotechnology. 47(6): 625-629. Kadam, K . L . et al., 2000. Softwood forest thinnings as a biomass source for ethanol production: A feasibility study for California. Biotechnology Progress. 16(6): 947-957. Kalman, G . , Varga, E . and Reczey, K . , 2002. Dilute sulphuric acid pretreatment of corn stover at long residence times. Chemical and Biochemical Engineering Quarterly. 16(4): 151-157. Karlsson, J., Siika-aho, M . , Tenkanen, M . and Tjerneld, F. , 2002. Enzymatic properties of the low molecular mass endoglucanases C e l l 2 A ( E G III) and Cel45A ( E G V ) of Trichoderma reesei. Journal of Biotechnology. 99(1): 63-78. Katchalske-Katzir, E . and Kraemer, D . M . , 2000. Eupergit C , a carrier for immobilization of enzymes of industrial potential. J. M o l . Catal. B-Enzym. 10(1-3): 157-76. Katchalski-Katzir, E . and Kraemer, D . M . , 2000. Eupergit (R) C, a carrier for immobilization of enzymes of industrial potential. Journal of Molecular Catalysis B-Enzymatic. 10(1-3): 157-176. K i m , J.S., Lee, Y . Y . and Park, S.C., 2000. Pretreatment of wastepaper and pulp mi l l sludge by aqueous ammonia and hydrogen peroxide. Applied Biochemistry and Biotechnology. 84-6: 129-139. K i m , S. and Holtzapple, M . T . , 2005. Lime pretreatment and enzymatic hydrolysis of corn stover. Bioresource Technology. 96(18): 1994-2006. K i m , S.B. and Chun, J.W., 2004. Enhancement of enzymatic digestibility of recycled newspaper by addition of surfactant in ammonia-hydrogen peroxide pretreatment. Applied Biochemistry and Biotechnology. 113-16: 1023-1031. K i m , T . H . and Lee, Y . Y . , 2005a. Pretreatment and fractionation of corn stover by ammonia recycle percolation process. Bioresource Technology. 96(18): 2007-2013. K i m , T . H . and Lee, Y . Y . , 2005b. Pretreatment of corn stover by soaking in aqueous ammonia. Applied Biochemistry and Biotechnology. 121: 1119-1131. Knauf, M . and Moniruzzaman, M . , 2004. Lignocellulosic biomass processing: A perspective. International Sugar Journal. 106(1263): 147-150. Knutsen, J.S. and Davis, R . H . , 2004. Cellulase retention and sugar removal by membrane ultrafiltration during lignocellulosic biomass hydrolysis. Applied Biochemistry and Biotechnology. 113-16: 585-599. 180 Koullas, D.P. , Christakopoulos, P., Kekos, D . , Maoris, B . J . and Koukios, E . G . , 1992. Correlating the Effect of Pretreatment on the Enzymatic-Hydrolysis of Straw. Biotechnology and Bioengineering. 39(1): 113-116. Krishna, S .H. and Chowdary, G . V . , 2000. Optimization of simultaneous saccharification and fermentation for the production of ethanol from lignocellulosic biomass. Journal of Agricultural and Food Chemistry. 48(5): 1971-1976. Krogh, K . B . R . , Morkeberg, A . , Jorgensen, H . , Frisvad, J .C. and Olsson, L . , 2004. Screening genus Penicillium for producers of cellulolytic and xylanolytic enzymes. Applied Biochemistry and Biotechnology. 113-16: 389-401. Kyriacou, A . , Neufeld, R .J . and Mackenzie, C .R. , 1988. Effect of Physical Parameters on the Adsorption Characteristics of Fractionated Trichoderma-Reesei Cellulase Components. Enzyme and Microbial Technology. 10(11): 675-681. Lamb, S.B. and Stuckey, D . C , 2000. Enzyme immobilization on colloidal liquid aphrons ( C L A s ) : the influence of system parameters on activity. Enzyme and Microbial Technology. 26(8): 574-581. Lee, D . , 1994. The use of cellulase adsorption and recycle as a means of assessing and enhancing htehydrolysis of cellulosic substrates. M S Thesis, Univeristy of British Columbia, Vancouver, 40 pp. Lee, D . , Y u , A . H . C . and Saddler, J .N. , 1995. Evaluation of Cellulase Recycling Strategies for the Hydrolysis of Lignocellulosic Substrates. Biotechnology and Bioengineering. 45(4): 328-336. Lee, D. , Y u , A . H . C , Wong, K . K . Y . and Saddler, J .N. , 1994. Evaluation of the Enzymatic Susceptibility of Cellulosic Substrates Using Specific Hydrolysis Rates and Enzyme Adsorption. Applied Biochemistry and Biotechnology. 45-6: 407-415. Lee, J., 1997. Biological conversion of lignocellulosic biomass to ethanol. Journal of Biotechnology. 56(1): 1-24. Lee, S.B., Shin, H.S. , Ryu, D . D . Y . and Mandels, M . , 1982. Adsorption of Cellulase on Cellulose - Effect of Physicochemical Properties of Cellulose on Adsorption and Rate of Hydrolysis. Biotechnology and Bioengineering. 24(10): 2137-2153. Lee, W . G . et al., 1996. Effect of surfactants on ethanol fermentation using glucose and cellulosic hydrolyzates. Biotechnology Letters. 18(3): 299-304. Lejeune, R. and Baron, G . V . , 1995. Effect of Agitation on Growth and Enzyme-Production of Trichoderma-Reesei in Batch Fermentation. Applied Microbiology and Biotechnology. 43(2): 249-258. Lemos, M . A . , Teixeira, J .A. , Domingues, M . R . M . , Mota, M . and Gama, F . M . , 2003. The enhancement of the cellulolytic activity of cellobiohydrolase I and endoglucanase 181 by the addition of cellulose binding domains derived from Trichoderma reesei. Enzyme and Microbial Technology. 32(1): 35-40. Levitsky, V . Y . , Lozano, P. and Iborra, J .L. , 1999. Kinetic analysis of deactivation of immobilized alpha-chymotrypsin by water-miscible organic solvent in kyotorphin synthesis. Biotechnology and Bioengineering. 65(2): 170-175. L i , X . L . , Dien, B.S. , Cotta, M . A . , W u , Y . V . and Saha, B . C . , 2005. Profile of enzyme production by Trichoderma reesei grown on corn fiber fractions. Applied Biochemistry and Biotechnology. 121: 321-334. Limam, F. , Chaabouni, S.E., Ghrir, R. and Marzouki, N . , 1995. Cellobiohydrolases of Penicillium-Occitanis Mutant Pol-6 - Purification and Properties. Enzyme and Microbial Technology. 17(4): 340-346. L i u , B . L . and Tzeng, Y . M . , 1998. Optimization of growth medium for the production of spores from Bacillus thuringiensis using response surface methodology. Bioprocess Engineering. 18(6): 413-418. L i u , Y . , 2006. Some consideration on the Langmuir isotherm equation. Colloids and Surfaces a-Physicochemical and Engineering Aspects. 274(1-3): 34-36. Lo , C M . , Zhang, Q., Lee, P. and Ju, L . K . , 2005. Cellulase production by Trichoderma reesei using sawdust hydrolysate. Applied Biochemistry and Biotechnology. 121: 561-573. L u , Y . P . , Yang, B . , Gregg, D . , Saddler, J .N . and Mansfield, S.D., 2002. Cellulase adsorption and an evaluation of enzyme recycle during hydrolysis of steam-exploded softwood residues. Applied Biochemistry and Biotechnology. 98: 641-654. Luey, J .K. , Mcguire, J. and Sproull, R .D . , 1991. The Effect of p H and N a C l Concentration on Adsorption of Beta-Lactoglobulin at Hydrophilic and Hydrophobic Silicon Surfaces. Journal of Col loid and Interface Science. 143(2): 489-500. Lynd, L . R . , Elander, R .T . and Wyman, C .E . , 1996. Likely features and costs of mature biomass ethanol technology. Applied Biochemistry and Biotechnology. 57-8: 741-761. Lynd, L . R . , Weimer, P.J., van Z y l , W . H . and Pretorius, I.S., 2002a. Microbial cellulose utilization: Fundamentals and biotechnology. Microbiology and Molecular Biology Reviews. 66(3): 506-+. Lynd, L . R . , Weimer, P.J., van Z y l , W . H . and Pretorius, I.S., 2002b. Microbial cellulose utilization: Fundamentals and biotechnology. Microbiology and Molecular Biology Reviews. 66(3): 506-577. 182 Mabee, W . E . and Saddler, J .N. , 2005. I E A Bioenergy Task 39 Liquid biofuels from biomass-Progress in enzymatic hydrolysis of lignocellulosics. Technology report. http://www.valbiom.be/LiploadPDF/Progress in Enzymatic Hydrolysis.pdf, Dublin, Ireland. Maheswari, D . K . , Jahan, H . , Paul, J. and Varma, A . , 1993. Wheat Straw, a Potential Substrate for Cellulase Production Using Trichoderma-Reesei. World Journal of Microbiology & Biotechnology. 9(1): 120-121. Mais, U . , Esteghlalian, A . R . and Saddler, J .N. , 2002. Influence of mixing regime on enzymatic saccharification of steam-exploded softwood chips. Applied Biochemistry and Biotechnology. 98: 463-472. Makkar, H.P.S. , Dawra, R . K . and Singh, B . , 1987. Protein Precipitation Assay for Quantitation of Tannins - Determination of Protein in Tannin Protein Complex. Analytical Biochemistry. 166(2): 435-439. Mandecki, W. , Shallcross, M . A . , Sowadski, J. and Tomazicallen, S., 1991. Mutagenesis of Conserved Residues within the Active-Site of Escherichia-Coli Alkaline-Phosphatase Yields Enzymes with Increased Kcat. Protein Engineering. 4(7): 801-804. Mansfield, S.D. and Meder, R., 2003. Cellulose hydrolysis - the role of monocomponent cellulases in crystalline cellulose degradation. Cellulose. 10(2): 159-169. Mansfield, S.D., Mooney, C. and Saddler, J .N. , 1999. Substrate and enzyme characteristics that limit cellulose hydrolysis. Biotechnology Progress. 15(5): 804-816. Mao, X . P . et al., 2006. A novel method to prepare chitosan powder and its application in cellulase immobilization. Journal of Chemical Technology and Biotechnology. 81(2): 189-195. Martin, M . T . , Plou, F.J. , Alcalde, M . and Ballesteros, A . , 2003. Immobilization on Eupergit C of cyclodextrin glucosyltransferase (CGTase) and properties of the immobilized biocatalyst. Journal of Molecular Catalysis B-Enzymatic. 21(4-6): 299-308. Martineau, F., 2002. Canada intends to ratify the Kyoto Protocol: Implications for Business, Investors and Lenders. EnviroBulletin. Martinez, A . T . et al., 1991. Solid-State Nmr-Studies of Lignin and Plant Polysaccharide Degradation by Fungi. Holzforschung. 45: 49-54. Martino, A . , Pifferi, P . G . and Spagna, G . , 1996. Immobilization of beta-glucosidase from a commercial preparation.2. Optimization of the immobilization process on chitosan. Process Biochemistry. 31(3): 287-293. 183 Mateo, C , Abian, O., Femandez-Lafuente, R. and Guisan, J . M . , 2000. Increase in conformational stability of enzymes immobilized on epoxy-activated supports by favoring additional multipoint covalent attachment. Enzyme and Microbial Technology. 26(7): 509-515. Mateo, C. et al., 2002a. Epoxy sepabeads: a novel epoxy support for stabilization of industrial enzymes via very intense multipoint covalent attachment. Biotechnol. Prog. 18(3): 629-634. Mateo, C. et al., 2002b. Epoxy sepabeads: A novel epoxy support for stabilization of industrial enzymes via very intense multipoint covalent attachment. Biotechnology Progress. 18(3): 629-634. M c M i l l a n , J.D., 1994. Pretreatment of lignocellulosic biomass. In: M . E . Himmel , J.O. Baker and R.P. Overend (Editors), Enzymatic conversion of biomass for fuels production. American Chemical Society, Washington, D C , pp. 499. Medve, J. , 1997. Cellulose hydrolysis by Trichoderma reesei cellulases: studies on adsorption, sugar production and synergism of cellobiohydrolase I, II and endoglucanase II. Ph.D. Thesis, Lund University, Lund. Medve, J., Lee, D . and Tjerneld, F., 1998. Ion-exchange chromatographic purification and quantitative analysis of Trichoderma reesei cellulases cellobiohydrolase I, II and endoglucanase II by fast protein liquid chromatography. Journal of Chromatography A . 808(1-2): 153-165. Medve, J., Stahlberg, J. and Tjerneld, F., 1994. Adsorption and Synergism of Cellobiohydrolase-I and Cellobiohydrolase-Ii of Trichoderma-Reesei During Hydrolysis of Microcrystalline Cellulose. Biotechnology and Bioengineering. 44(9): 1064-1073. Medve, J. , Stahlberg, J. and Tjerneld, F., 1997. Isotherms for adsorption of cellobiohydrolase I and II from Trichoderma reesei on microcrystalline cellulose. Applied Biochemistry and Biotechnology. 66(1): 39-56. Meshartree, M . , Hogan, C M . and Saddler, J .N. , 1987. Recycle of Enzymes and Substrate Following Enzymatic-Hydrolysis of Steam-Pretreated Aspenwood. Biotechnology and Bioengineering. 30(4): 558-564. Mishra, C . and Rao, M . , 1988. Mode of Action and Synergism of Cellulases from Penicillium-Funiculosum. Applied Biochemistry and Biotechnology. 19(2): 139-150. Mollmann, S.H., Elofsson, U . , Bukrinsky, J.T. and Frokjaer, S., 2005. Displacement of adsorbed insulin by Tween 80 monitored using total internal reflection fluorescence and ellipsometry. Pharmaceutical Research. 22(11): 1931-1941. Montgomery, D. , 2000. Design and analysis of experiments. Wiley, N Y . 184 Mooney, C . A . , Mansfield, S.D., Touhy, M . G . , et al., 1998. The effect of initial pore volume and lignin content on the enzymatic hydrolysis of softwoods. Biores. Technol. 64(2): 113-119. Moreno, E .C . , Kresak, M . and Hay, D.I., 1982. Adsorption Thermodynamics of Acid ic Proline-Rich Human Salivary Proteins onto Calcium Apatites. Journal of Biological Chemistry. 257(6): 2981-2989. Mores, W . D . , Knutsen, J.S. and Davis, R . H . , 2001. Cellulase recovery via membrane filtration. Applied Biochemistry and Biotechnology. 91-3: 297-309. Morlot, J .C., 1999. National climate policies and the Kyoto protocol. Organisation for Economic Co-operation and Development, Paris, 87 pp. Morris, S.C., 1982. Synergism of Geotrichum-Candidum and Penicillium-Digitatum in Infected Citrus-Fruit. Phytopathology. 72(10): 1336-1339. Mozhaev, V . V . , Berezin, I .V. and Martinek, K . , 1988. Structure-Stability Relationship in Proteins - Fundamental Tasks and Strategy for the Development of Stabilized Enzyme Catalysts for Biotechnology. C R C Critical Reviews in Biochemistry. 23(3): 235-281. Murray, P. et al., 2004. Expression in Trichoderma reesei and characterisation of a thermostable family 3 beta-glucosidase from the moderately thermophilic fungus Talaromyces emersonii. Protein Expression and Purification. 38(2): 248-257. Neilson, M . J . , Shafizadeh, F., A z i z , S. and Sarkanen, K . V . , 1983. Evaluation of Organosolv Pulp as a Suitable Substrate for Rapid Enzymatic-Hydrolysis. Biotechnology and Bioengineering. 25(2): 609-612. Nevalainen, K . M . H . , Palva, E.T. and Bailey, M . J . , 1980. High Cellulase-Producing Mutant Strain of Trichoderma-Reesei. Enzyme and Microbial Technology. 2(1): 59-60. Nguyen, Q .A . et al., 1999. Dilute acid hydrolysis of softwoods. Applied Biochemistry and Biotechnology. 77-9: 133-142. Nguyen, Q .A. , Tucker, M . P . , Keller, F . A . and Eddy, F.P., 2000. Two-stage dilute-acid pretreatment of softwoods. Applied Biochemistry and Biotechnology. 84-6: 561-576. Nidetzky, B . , Steiner, W . and Claeyssens, M . , 1994. Cellulose Hydrolysis by the Cellulases from Trichoderma-Reesei - Adsorptions of 2 Cellobiohydrolases, 2 Endocellulases and Their Core Proteins on Filter-Paper and Their Relation to Hydrolysis. Biochemical Journal. 303: 817-823. 185 Nogawa, M . , Goto, M . , Okada, H . and Morikawa, Y . , 2001. L-Sorbose induces cellulase gene transcription in the cellulolytic fungus Trichoderma reesei. Current Genetics. 38(6): 329-334. Norde, W. , 1998. Driving forces for protein adsorption at solid surfaces. In: M . Malmsten (Editor), Biopolymers at interfaces. Marcel Dekker, New York, pp. 27-54. Oh, K . K . , K i m , Y . S . , Yoon, H . H . and Tae, B.S. , 2002. Pretreatment of lignocellulosic biomass using combination of ammonia recycled percolation and dilute-acid process. Journal of Industrial and Engineering Chemistry. 8(1): 64-70. Olsson, L . , Christensen, T .M.I .E . , Hansen, K . P . and Palmqvist, E . A . , 2003: Influence of the carbon source on production of cellulases, hemicellulases and pectinases by Trichoderma reesei Rut C-30. Enzyme and Microbial Technology. 33(5): 612-619. Ong, E . , Gilkes, N .R . , Mil ler , R .C . , Warren, R . A . J , and Kilburn, D . G . , 1993. The Cellulose-Binding Domain (Cbdcex) of an Exoglucanase from Cellulomonas-F imi - Production in Escherichia-Coli and Characterization of the Polypeptide. Biotechnology and Bioengineering. 42(4): 401-409. Ong, E . , Kilburn, D . G . , Mil ler , R . C . and Warren, R . A . J . , 1994. Streptomyces-Lividans Glycosylates the Linker Region of a Beta-1,4-Glycanase from Cellulomonas-F imi . Journal of Bacteriology. 176(4): 999-1008. Ooshima, H . , Burns, D.S . and Converse, A . O . , 1990. Adsorption of Cellulase from Trichoderma-Reesei on Cellulose and Lignacious Residue in Wood Pretreated by Dilute Sulfuric-Acid with Explosive Decompression. Biotechnology and Bioengineering. 36(5): 446-452. Ooshima, H . , Sakata, M . and Harano, Y . , 1986. Enhancement of Enzymatic-Hydrolysis of Cellulose by Surfactant. Biotechnology and Bioengineering. 28(11): 1727-1734. Ortega, N . , Busto, M . D . and Perez-Mateos, M . , 2000. Enzymatic saccharification of pretreated wheat straw by T-reesei cellulases and A-niger beta-glucosidase. Biocatalysis and Biotransformation. 18(4): 311-330. Otter, D .E . , Munro, P .A . , Scott, G . K . and Geddes, R., 1984. Elution of Trichoderma-Reesei Cellulase from Cellulose by Ph Adjustment with Sodium-Hydroxide. Biotechnology Letters. 6(6): 369-373. Otter, D . E . , Munro, P .A . , Scott, G . K . and Geddes, R., 1989. Desorption of Trichoderma-Reesei Cellulase from Cellulose by a Range of Desorbents. Biotechnology and Bioengineering. 34(3): 291-298. Ozawa, T. et al., 2001. Thermostabilization by replacement of specific residues with lysine in a Bacillus alkaline cellulase: building a structural model and 186 implications of newly formed double intrahelical salt bridges. Protein Engineering. 14(7): 501-504. Ozer, A . , Ozer, D . and Ozer, A . , 2004. The adsorption of copper(II) ions on to dehydrated wheat bran ( D W B ) : determination of the equilibrium and thermodynamic parameters. Process Biochemistry. 39(12): 2183-2191. Palonen, H . , 2004. Role of lignin in the enzymatic hydrolysis of lignocellulose. PhD Thesis, V T T , E S P O O , Finland, 21 pp. Palonen, H . , Tjerneld, F., Zacchi, G . and Tenkanen, M . , 2004. Adsorption of Trichoderma reesei C B H I and E G II and their catalytic domains on steam pretreated softwood and isolated lignin. Journal of Biotechnology. 107(1): 65-72. Pan, X . J . et al., 2005a. Biorefining of softwoods using ethanol organosolv pulping: Preliminary evaluation of process streams for manufacture of fuel-grade ethanol and co-products. Biotechnology and Bioengineering. 90(4): 473-481. Pan, X . J . and Sana, Y . , 1999. Atmospheric acetic acid pulping of rice straw IV: Physico-chemical characterization of acetic acid lignins from rice straw and woods. Part 2. Chemical structures. Holtzforschung. 53: 590-596. Pan, X . J . , X i e , D . , Gilkes, N . , Gregg, D.J . and Saddler, J .N. , 2005b. Strategies to enhance the enzymatic hydrolysis of pretreated softwood with high residual lignin content. Applied Biochemistry and Biotechnology. 121: 1069-1079. Pan, X . J . , Zhang, X . , Gregg, D.J . and Saddler, J .N. , 2004. Enhanced enzymatic hydrolysis of steam-exploded Douglas fir wood by alkali-oxygen post-treatment. Applied Biochemistry and Biotechnology. 113-16: 1103-1114. Park, J.W., Takahata, Y . , Kajiuchi, T. and Akehata, T., 1992. Effects of Nonionic Surfactant on Enzymatic-Hydrolysis of Used Newspaper. Biotechnology and Bioengineering. 39(1): 117-120. Petzelbauer, I., Kuhn, B . , Splechtna, B . , Kulbe, K . D . and Nidetzky, B . , 2002. Development of an ultrahigh-temperature process for the enzymatic hydrolysis of lactose. IV. Immobilization of two thermostable beta-glycosidases and optimization of a packed-bed reactor for lactose conversion. Biotechnology and Bioengineering. 77(6): 619-631. Pierce Biotechnology Inc., 2003. B C A Protein Assay Ki t : Instructions, Rockford, IL. Puis, J. and Wood, T . M . , 1991. The Degradation Pattern of Cellulose by Extracellular Cellulases of Aerobic and Anaerobic Microorganisms. Bioresource Technology. 36(1): 15-19. Pye, E . K . and Lora, J .H. , 1991. The Alce l l Process - a Proven Alternative to Kraft Pulping. Tappi Journal. 74(3): 113-118. 187 Ram, K . A . and Venkatasubramanian, K . , 1982. Enhancement of Starch Conversion Efficiency with Free and Immobilized Pullulanase and Alpha-1,4-Glucosidase. Biotechnology and Bioengineering. 24(2): 355-369. Ramos, L .P . , Breuil , C . and Saddler, J .N. , 1993. The Use of Enzyme Recycling and the Influence of Sugar Accumulation on Cellulose Hydrolysis by Trichoderma-Cellulases. Enzyme and Microbial Technology. 15(1): 19-25. Ramos, L .P . and Saddler, J .N. , 1994. Enzyme Recycling During Fed-Batch Hydrolysis of Cellulose Derived from Steam-Exploded Eucalyptus-Viminalis. Applied Biochemistry and Biotechnology. 45-6: 193-207. Ramos, L .P . , Zandona, A . , Deschamps, F .C . and Saddler, J .N. , 1999. The effect of Trichoderma cellulases on the fine structure of a bleached softwood kraft pulp. Enzyme and Microbial Technology. 24(7): 371-380. Rani, A . S . , Das, M . L . M . and Satyanarayana, S., 2000. Preparation and characterization of amyloglucosidase adsorbed on activated charcoal. Journal of Molecular Catalysis B-Enzymatic. 10(5): 471-476. Rao, M . , Seeta, R. and Deshpande, V . , 1983. Effect of Pretreatment on the Hydrolysis of Cellulose by Penicillium-Funiculosum Cellulase and Recovery of Enzyme. Biotechnology and Bioengineering. 25(7): 1863-1871. Ravinder, T., Ramesh, B . , Seenayya, G . and Reddy, G . , 2000. Fermentative production of acetic acid from various pure and natural cellulosic materials by Clostridium lentocellum SG6. World Journal of Microbiology & Biotechnology. 16(6): 507-512. Ray, L . and Majumdar, S.K., 1994. Enzymatic-Hydrolysis of Starch to Glucose, and Immobilization of Amyloglucosidase in Calcium Alginate Gel . Research and Industry. 39(2): 109-112. Reczey, K . , Szengyel, Z . , Eklund, R. and Zacchi, G . , 1996. Cellulase production by T-reesei. Bioresource Technology. 57(1): 25-30. Reese, E.T. , 1976. History of Cellulase Program at US Army Natick Development Center. Biotechnology and Bioengineering(6): 9-20. Reese, E.T. and Mandels, M . , 1980. Stability of the Cellulase of Trichoderma-Reesei under Use Conditions. Biotechnology and Bioengineering. 22(2): 323-335. Renewable Fuels Association, 2005. Annual world ethanol production by country. http://www.ethanolrfa.Org/industry/statistics/#E, Accessed June 2, 2006. Riedel, K . and Bronnenmeier, K . , 1998. Intramolecular synergism in an engineered exo-endo-l,4-beta-glucanase fusion protein. Molecular Microbiology. 28(4): 767-775. 188 Riou, C , Salmon, J . M . , Vallier, M . J . , Gunata, Z . and Barre, P., 1998. Purification, characterization, and substrate specificity of a novel highly glucose-tolerant beta-glucosidase from Aspergillus oryzae. Applied and Environmental Microbiology. 64(10): 3607-3614. Robinson, J., Keating, J.D., Boussaid, A . , Mansfield, S.D. and Saddler, J .N. , 2002. The influence of bark on the fermentation of Douglas-fir whitewood pre-hydrolysates. Applied Microbiology and Biotechnology. 59(4-5): 443-448. Rughani, J. and Mcginnis, G.D. , 1987. Combined Rapid Steam Hydrolysis and Organosolv Pretreatment of Mixed Southern Hardwoods. Abstracts of Papers of the American Chemical Society. 194: 66-Cell . Ryu, D . D . Y . , K i m , C . and Mandels, M . , 1984. Competitive Adsorption of Cellulase Components and Its Significance in a Synergistic Mechanism. Biotechnology and Bioengineering. 26(5): 488-496. Ryu, D . D . Y . and Mandels, M . , 1980. Cellulases - Biosynthesis and Applications. Enzyme and Microbial Technology. 2(2): 91-102. Saddler, J .N. , Meshartree, M . , Y u , E . K . C . and Brownell, H . H . , 1983. Enzymatic-Hydrolysis of Various Pretreated Lignocellulosic Substrates and the Fermentation of the Liberated Sugars to Ethanol and Butanediol. Biotechnology and Bioengineering: 225-238. Saha, B . C . , Iten, L . B . , Cotta, M . A . and W u , Y . V . , 2005a. Dilute acid pretreatment, enzymatic saccharification and fermentation of wheat straw to ethanol. Process Biochemistry. 40(12): 3693-3700. Saha, B . C . , Iten, L . B . , Cotta, M . A . and W u , Y . V . , 2005b. Dilute acid pretreatment, enzymatic saccharification, and fermentation of rice hulls to ethanol. Biotechnology Progress. 21(3): 816-822. Sardar, M . , Agarwal, R., Kumar, A . and Gupta, M . N . , 1997. Noncovalent immobilization of enzymes on an enteric polymer Eudragit S-100. Enzyme and Microbial Technology. 20(5): 361-367. Sarkanen, S., Teller, D .C . , Hal l , J. and McCarthy, J .L. , 1981. Lignin .18. Associative Effects among Organosolv Lignin Components. Macromolecules. 14(2): 426-434. S A S Institute Inc., 2002. J M P 4 Design of Experiments, Cary, N C , U S A , 80-81 pp. Saville, B . A . , Khavkine, M . , Seetharam, G . , Marandi, B . and Zuo, Y . L . , 2004. Characterization and performance of immobilized amylase and cellulase. Applied Biochemistry and Biotechnology. 113-16: 251-259. Sawalha, M . F . , Gardea-Torresdey, J .L. , Parsons, J .G., Saupe, G . and Peralta-Videa, J.R., 2005. Determination of adsorption and speciation of chromium species by 189 saltbush (Atriplex canescens) biomass using a combination of X A S and ICP-OES. Microchemical Journal. 81(1): 122-132. Schell, D. , Nguyen, Q., Tucker, M . and Boynton, B . , 1998. Pretreatment of softwood by acid-catalyzed steam explosion followed by alkali extraction. Applied Biochemistry and Biotechnology. 70-2: 17-24. Seidle, H.F. , Marten, I., Shoseyov, O. and Huber, R .E . , 2004. Physical and kinetic properties of the family 3 beta-glucosidase from Aspergillus niger which is important for cellulose breakdown. Protein Journal. 23(1): 11-23. Seip, J.E., Fager, S.K., Gavagan, J.E., Anton, D . L . and D i Cosimo, R., 1994. Glyoxyl ic acid production using immobilized glycolate oxidase and catalase. Bioorganic & Medicinal Chemistry. 2(6): 371-8. Sellers, T., Lora, J .H. and Okuma, M . , 1994. Organosolv-Lignin-Modified Phenolics as Strandboard Binder. 2. Strandboard Manufacture and Properties. Mokuzai Gakkaishi. 40(10): 1079-1084. Shahbazi, A . , L i , Y . B . and Mims, M . R . , 2005. Application of sequential aqueous steam treatments to the fractionation of softwood. Applied Biochemistry and Biotechnology. 121: 973-987. Shaver, J . M . et al., 1996. Single-amino acid substitutions eliminate lysine inhibition of maize dihydrodipicolinate synthase. Proceedings of the National Academy of Sciences of the United States of America. 93(5): 1962-1966. Sheehan, J.S. and Himmel, M . E . , 2001. Outlook for bioethanol production from lignocellulosic feedstocks: Technology hurtles. Agro Food Industry Hi-Tech. 12(5): 54-57. Shen, X . L . and X i a , L . M . , 2004. Production and immobilization of cellobiase from Aspergillus niger ZU-07. Process Biochemistry. 39(11): 1363-1367. Shevchenko, S . M . , Chang, K . , Dick, D . G . , Gregg, D.J . and Saddler, J .N. , 2001a. Structure and properties of lignin in softwoods after S02-catalyzed steam explosion and enzymatic hydrolysis. Cellulose Chemistry and Technology. 35(5-6): 487-502. Shevchenko, S . M . , Paszner, L . and Saddler, J .N. , 2001b. Treatment of organosolv and steam explosion lignins with molecular hydrogen iodide: A structural probe. Cellulose Chemistry and Technology. 35(3-4): 295-318. Shoemaker, S., Watt, K . , Tsitovsky, G . and Cox, R., 1983. Characterization and Properties of Cellulases Purified from Trichoderma-Reesei Strain-L27. Bio-Technology. 1(8): 687-690. 190 Sidiras, D . and Koukios, E . , 2004. Simulation of acid-catalysed organosolv fractionation of wheat straw. Bioresource Technology. 94(1): 91-98. Sinitsyn, A . P . , Gusakov, A . V . and Vlasenko, E . Y . , 1991. Effect of Structural and Physicochemical Features of Cellulosic Substrates on the Efficiency of Enzymatic-Hydrolysis. Applied Biochemistry and Biotechnology. 30(1): 43-59. Sjostrom, E . , 1993. Wood Chemistry Fundamentals and Applications. Harcourt Brace Jovanovich, Toronto, 59 pp. Sorensen, H.R. , Pedersen, S., Vikso-Nielsen, A . and Meyer, A . S . , 2005. Efficiencies of designed enzyme combinations in releasing arabinose and xylose from wheat arabinoxylan in an industrial ethanol fermentation residue. Enzyme and Microbial Technology. 36(5-6): 773-784. South, C.R. , Hogsett, D . A . and Lynd, L . R . , 1993. Continuous Fermentation of Cellulosic Biomass to Ethanol. Applied Biochemistry and Biotechnology. 39: 587-600. Spagna, G . , Pifferi, P .G . and Martino, A . , 1993. Pectinlyase Immobilization on Epoxy Supports for Application in the Food-Processing Industry. Journal of Chemical Technology and Biotechnology. 57(4): 379-385. Starcher, B . , 2001. A ninhydrin-based assay to quantitate the total protein content of tissue samples. Analytical Biochemistry. 292(1): 125-129. Steele, E . , Raj, S., Nghiem, J. and Stowers, M . , 2005. Enzyme recovery and recycling following hydrolysis of ammonia fiber explosion-treated corn stover. Applied Biochemistry and Biotechnology. 121: 901-910. Stenberg, K . , Bollok, M . , Reczey, K . , Galbe, M . and Zacchi, G . , 2000. Effect of substrate and cellulase concentration on simultaneous saccharification and fermentation of steam-pretreated softwood for ethanol production. Biotechnology and Bioengineering. 68(2): 204-210. Stockburger, P., 1993. A n Overview of near-Commercial and Commercial Solvent-Based Pulping Processes. Tappi Journal. 76(6): 71-74. Sun, Y . and Cheng, J.J., 2005. Dilute acid pretreatment of rye straw and bermudagrass for ethanol production. Bioresource Technology. 96(14): 1599-1606. Sun, Y . and Cheng, J .Y. , 2002. Hydrolysis of lignocellulosic materials for ethanol production: a review. Bioresource Technology. 83(1): 1-11. Sundstrom, D .W. , K le i , H .E . , Coughlin, R .W. , Biederman, G.J . and Brouwer, C . A . , 1981. Enzymatic-Hydrolysis of Cellulose to Glucose Using Immobilized Beta-Glucosidase. Biotechnology and Bioengineering. 23(3): 473-485. 191 Szengyel, Z . and Zacchi, G . , 2000. Effect of acetic acid and furfural on cellulase production of Trichoderma reesei R U T C30. Applied Biochemistry and Biotechnology. 89(1): 31-42. Tan, L . U . L . , Y u , E . K . C . , Campbell, N . and Saddler, J .N. , 1986. Column Cellulose Hydrolysis Reactor - an Efficient Cellulose Hydrolysis Reactor with Continuous Cellulase Recycling. Applied Microbiology and Biotechnology. 25(3): 250-255. Taniguchi, M . et al., 2005. Evaluation of pretreatment with Pleurotus ostreatus for enzymatic hydrolysis of rice straw. Journal of Bioscience and Bioengineering. 100(6): 637-643. Tanyildizi , M . S . , E l ibo l , M . and Ozer, D . , 2006. Optimization of growth medium for the production of alpha-amylase from Bacillus amyloliquefaciens using response surface methodology. Journal of Chemical Technology and Biotechnology. 81(4): 618-622. Tengborg, C , Galbe, M . and Zacchi, G . , 2001. Reduced inhibition of enzymatic hydrolysis of steam-pretreated softwood. Enzyme and Microbial Technology. 28(9-10): 835-844. Thygesen, A . et al., 2003. Production of cellulose and hemicellulose-degrading enzymes by filamentous fungi cultivated on wet-oxidised wheat straw. Enzyme and Microbial Technology. 32(5): 606-615. Tjerneld, F., 1994. Enzyme-Catalyzed Hydrolysis and Recycling in Cellulose Bioconversion. Aqueous Two-Phase Systems. 228: 549-558. Tomme, P. et al., 1998. Characterization and affinity applications of cellulose-binding domains. Journal of Chromatography B . 715(1): 283-296. Tomme, P. et al., 1995. Comparison of a Fungal (Family-I) and Bacterial (Family-Ii) Cellulose-Binding Domain. Journal of Bacteriology. 177(15): 4356-4363. Tomme, P. et al., 1988. Studies of the Cellulolytic System of Trichoderma-Reesei Qm-9414 - Analysis of Domain Function in 2 Cellobiohydrolases by Limited Proteolysis. European Journal of Biochemistry. 170(3): 575-581. Torees-Bacete, J. et al., 2000. Covalent immobilization of penicillin acylase from Streptomyces lavendulae. Biotechno. Lett. 22: 1011-1014. Torres-Bacete, J. et al., 2000. Covalent immobilization of penicillin acylase from Streptomyces lavendulae. Biotechnology Letters. 22(12): 1011-1014. Tu, M . B . et al., 2006. Immobilization of beta-glucosidase on Eupergit C for lignocellulose hydrolysis. Biotechnology Letters. 28(3): 151-156. 192 Tucker, M . P . , K i m , K . H . , Newman, M . M . and Nguyen, Q .A . , 2003. Effects of temperature and moisture on dilute-acid steam explosion pretreatment of corn stover and cellulase enzyme digestibility. Applied Biochemistry and Biotechnology. 105: 165-177. U .S . Department of Energy, 2006. Cellulase Cost-Reduction Contracts, Biomass Program. U.S . Deparment of Energy, Energy Efficiency and Renewable Energy. Valjamae, P., Pettersson, G . and Johansson, G . , 2001. Mechanism of substrate inhibition in cellulose synergistic degradation. European Journal of Biochemistry. 268(16): 4520-4526. Valjamae, P., Si ld, V . , Pettersson, G . and Johansson, G . , 1998. The initial kinetics of hydrolysis by cellobiohydrolases I and II is consistent with a cellulose surface -erosion model. European Journal of Biochemistry. 253(2): 469-475. Vallander, L . and Eriksson, K . E . , 1987. Enzyme Recirculation in Saccharification of Lignocellulosic Materials. Enzyme and Microbial Technology. 9(12): 714-720. van Wyk, J .P.H. , 1998. Paper hydrolysis by cellulase from Penicill ium funiculosum and Trichoderma viride. Bioresource Technology. 63(3): 275-277. Vance, C P . , Kirk , T . K . and Sherwood, R.T. , 1980. Lignification as a Mechanism of Disease Resistance. Annual Review of Phytopathology. 18: 259-288. Varga, E . , Reczey, K . and Zacchi, G . , 2004. Optimization of steam pretreatment of corn stover to enhance enzymatic digestibility. Applied Biochemistry and Biotechnology. 113-16: 509-523. Vinzant, T . B . et al., 1994. Ssf Comparison of Selected Woods from Southern Sawmills. Applied Biochemistry and Biotechnology. 45-6: 611-626. Vitousek, P . M . , Mooney, H . A . , Lubchenco, J. and Mel i l lo , J . M . , 1997. Human domination of Earth's ecosystems. Science. 277(5325): 494-499. Walker, L . P . and Wilson, D . B . , 1991. Enzymatic-Hydrolysis of Cellulose - an Overview. Bioresource Technology. 36(1): 3-14. Watson, D . L . , Wilson, D . B . and Walker, L .P . , 2002. Synergism in binary mixtures of Thermobifida fusca cellulases Cel6B, Cel9A, and Cel5A on B M C C and Avice l . Applied Biochemistry and Biotechnology. 101(2): 97-111. Wi l l i s , J . M . and Herring, F . G . , 1987. Effect of Water in the C-13 Cp Mas Nmr-Spectrum of White Spruce Wood. Macromolecules. 20(7): 1554-1556. Wingren, A . , Galbe, M . and Zacchi, G . , 2003. Techno-economic evaluation of producing ethanol from softwood: Comparison of SSF and S H F and identification of bottlenecks. Biotechnology Progress. 19(4): 1109-1117. 193 Withers, S.G., Rupitz, K . , Trimbur, D . and Warren, R . A . J . , 1992. Mechanistic Consequences of Mutation of the Active-Site Nucleophile Glu-358 in Agrobacterium Beta-Glucosidase. Biochemistry. 31(41): 9979-9985. Wood, T . M . , 1992. Fungal Cellulases. Biochemical Society Transactions. 20(1): 46-53. Wood, T . M . and Bhat, K . M . , 1988. Methods for Measuring Cellulase Activities. Methods in Enzymology. 160: 87-112. Wood, T . M . and Mccrae, S.I., 1986. The Cellulase of Penicillium-Pinophilum -Synergism between Enzyme Components in Solubilizing Cellulose with Special Reference to the Involvement of 2 Immunologically Distinct Cellobiohydrolases. Biochemical Journal. 234(1): 93-99. Wood, T . M . , Mccrae, S.I. and Bhat, K . M . , 1989. The Mechanism of Fungal Cellulase Action - Synergism between Enzyme Components of Penicillium-Pinophilum Cellulase in Solubilizing Hydrogen Bond-Ordered Cellulose. Biochemical Journal. 260(1): 37-43. Woodward, J., 1991. Synergism in Cellulase Systems. Bioresource Technology. 36(1): 67-75. W u , M . M . et al., 1999. Optimization of steam explosion to enhance hemicellulose recovery and enzymatic hydrolysis of cellulose in softwoods. Applied Biochemistry and Biotechnology. 77-9: 47-54. W u , Z . W . and Lee, Y . Y . , 1997. Ammonia recycled percolation as a complementary pretreatment to the dilute-acid process. Applied Biochemistry and Biotechnology. 63-5: 21-34. Wyman, C , 1996. Handbook on bioethanol: production and utilization. Applied energy technology series. Taylor & Francis, Washington, D C , 424 pp. Wyman, C .E . , 1999. Biomass ethanol: Technical progress, opportunities, and commercial challenges. Annual Review of Energy and the Environment. 24: 189-226. Wyman, C . E . et al., 2005. Coordinated development of leading biomass pretreatment technologies. Bioresource Technology. 96(18): 1959-1966. Wyman, C . E . , Spindler, D . D . and Grohmann, K . , 1992. Simultaneous Saccharification and Fermentation of Several Lignocellulosic Feedstocks to Fuel Ethanol. Biomass & Bioenergy. 3(5): 301-307. X i a , L . M . and Shen, X . L . , 2004. High-yield cellulase production by Trichoderma reesei ZU-02 on corn cob residue. Bioresource Technology. 91(3): 259-262. 194 Xiao, Z .Z . , Zhang, X . , Gregg, D.J . and Saddler, J .N. , 2004. Effects of sugar inhibition on cellulases and beta-glucosidase during enzymatic hydrolysis of softwood substrates. Applied Biochemistry and Biotechnology. 113-16: 1115-1126. Xiong, H.R. , von Weymarn, N . , Leisola, M . and Turunen, O., 2004. Influence of p H on the production of xylanases by Trichoderma reesei Rut C-30. Process Biochemistry. 39(6): 729-733. Yang, B . , Boussaid, A . , Mansfield, S.D., Gregg, D.J . and Saddler, J .N. , 2002. Fast and efficient alkaline peroxide treatment to enhance the enzymatic digestibility of steam-exploded softwood substrates. Biotechnology and Bioengineering. 77(6): 678-684. Yang, B . and Wyman, C .E . , 2006. B S A treatment to enhance enzymatic hydrolysis of cellulose in lignin containing substrates. Biotechnology and Bioengineering. 94(4): 611-617. Yasokawa, D . , Shimizu, T., Nakagawa, R., Ikeda, T. and Nagashima, K . , 2003. Cloning, sequencing, and heterologous expression of a cellobiohydrolase c D N A from the basidiomycete Corticium rolfsii. Bioscience Biotechnology and Biochemistry. 67(6): 1319-1326. Yoon, H . H . , Wu , Z . W . and Lee, Y . Y . , 1995. Ammonia-Recycled Percolation Process for Pretreatment of Biomass Feedstock. Applied Biochemistry and Biotechnology. 51-2: 5-19. Yuan, X . Y . , Shen, N . X . , Sheng, J. and Wei , X . , 1999. Immobilization of cellulase using acrylamide grafted acrylonitrile copolymer membranes. Journal of Membrane Science. 155(1): 101-106. Zandona, F . A . , Cancelier, P .D. , Siika-Aho, M . and Ramos, L .P . , 2003. Characterization of the activity profile in cellulases derived from recombinant strains of Trichoderma reesei. Applications of Enzymes to Lignocellulosics. 855: 285-303. Zhang, Y . H . P . and Lynd, L .R . , 2005. Determination of the number-average degree of polymerization of cellodextrins and cellulose with application to enzymatic hydrolysis. Biomacromolecules. 6(3): 1510-1515. Zhao, Y . , W u , B . , Yan, B . X . and Gao, P.J., 2004. Mechanism of cellobiose inhibition in cellulose hydrolysis by cellobiohydrolase. Science in China Series C-Life Sciences. 47(1): 18-24. Zheng, Y . M . , Zhao, Q .B . and Y u , H.Q. , 2005. Adsorption of a cationic dye onto aerobic granules. Process Biochemistry. 40(12): 3777-3782. Zhou, S.D., Davis, F .C . and Ingram, L . O . , 2001. Gene integration and expression and extracellular secretion of Erwinia chrysanthemi endoglucanase C e l Y (celY) and 195 C e l Z (celZ) in ethanologenic Klebsiella oxytoca P2. Applied and Environmental Microbiology. 67(1): 6-14. Zhou, S.D. and Ingram, L .O . , 2001. Simultaneous saccharification and fermentation of amorphous cellulose to ethanol by recombinant Klebsiella oxytoca SZ21 without supplemental cellulase. Biotechnology Letters. 23(18): 1455-1462. Zimbardi, F. et al., 1999. Steam explosion of straw in batch and continuous systems. Applied Biochemistry and Biotechnology. 77-9: 117-125. Zyabreva, N . V . , Isakova, E.P. and Biryukov, V . V . , 2001. Conditions for ethanol production during bioconversion of cellulose-containing raw material. Applied Biochemistry and Microbiology. 37(5): 483-488. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0074969/manifest

Comment

Related Items