UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Isolation and characterization of the major fraction of flaxseed proteins Chung, Mavis W. Y. 2001

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

Item Metadata

Download

Media
831-ubc_2001-0172.pdf [ 24MB ]
Metadata
JSON: 831-1.0089714.json
JSON-LD: 831-1.0089714-ld.json
RDF/XML (Pretty): 831-1.0089714-rdf.xml
RDF/JSON: 831-1.0089714-rdf.json
Turtle: 831-1.0089714-turtle.txt
N-Triples: 831-1.0089714-rdf-ntriples.txt
Original Record: 831-1.0089714-source.json
Full Text
831-1.0089714-fulltext.txt
Citation
831-1.0089714.ris

Full Text

ISOLATION AND CHARACTERIZATION OF THE MAJOR FRACTION OF FLAXSEED PROTEINS by Mavis W . Y . Chung B.Sc. (Agr), The University of British Columbia, 1998 A thesis submitted in partial fulfillment of the requirement for the degree of M A S T E R OF S C I E N C E in T H E F A C U L T Y OF G R A D U A T E STUDIES Department of Food Science We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y OF BRITISH C O L U M B I A April, 2001 © Mavis W.Y. Chung, 2001 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of r o o The University of British Columbia Vancouver, Canada DE-6 (2/88) ABSTRACT Flaxseed has not been well characterized in terms of the physicochemical and functional properties of its constituent proteins. This information is essential for its utilization. The major fraction of flaxseed proteins was isolated and characterized in this study. Proteins from NorMan flaxseeds were extracted with 0.10 M NaCl in 0.10 M Tris at pH 8.6 and fractionated by DEAE-Sephacel anion-exchange chromatography. The major protein fraction was eluted at 0.25 M NaCl and comprised 63.7 % of the total proteins. The gel electrophoretic profiles, isoelectric points, sulfhydryl and disulfide groups, and amino acid composition of the major fraction were investigated. In addition, the surface hydrophobicity, apparent viscosity, solubility and foaming properties of the major fraction in buffer conditions with two NaCl concentrations (0.01 or 1.0 M) at three pHs (3, 5 or 7) were examined. At least six different size classes of polypeptides, with predominant components having molecular weights of 20 + 1, 26 + 2 and 31 + 1 kDa, were observed for the major fraction in reducing SDS-PAGE. The fraction had a high content of disulfide linkages but low content of sulfhydryl groups. Similar to the storage globulin of other oilseeds, it had high level of arginine, glutamate (and/or glutamine) and aspartate (and/or asparagine). The isoelectric points of three components in the major fraction separated by isoelectric focusing were 4.7 + 0.0, 5.1 + 0.0, and 5.6 + 0.1. Its surface hydrophobicity (S0) measured by the PROD AN fluorescent probe was significantly affected by pH and NaCl concentrations. S0 increased with the addition of NaCl and increased in pH. ii The major fraction showed a typical U-shaped solubility curve as a function of pH at low NaCl concentration while addition of NaCl broadened and shifted the region of minimal solubility to the acidic side. The fraction had foaming properties comparable to that of a commercial egg albumen powder in 0.01 M NaCl at pH 7 and it had the densest and most stable foam in the same buffer at pH 3. Results from this thesis serve as a basis for more in-depth characterization of the major fraction of flaxseed proteins. iii TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iv LIST OF TABLES x LIST OF FIGURES xii ACKNOWLEDGEMENTS xvi CHAPTER 1—INTRODUCTION 1 CHAPTER 2—LITERATURE REVIEW 3 2.1 O V E R V I E W OF F L A X A N D F L A X S E E D S 3 2.1.1 Flax 3 2.1.2 Flaxseed/Linseed 4 2.1.3 Flaxseed/Linseed Meal 5 2.1.4 Flaxseed/linseed proteins 5 2.2 F L A X S E E D / L I N S E E D PROTEINS 6 2.2.1 Chronological review of the extraction studies on flaxseed/linseed proteins 6 2.2.2 Isolation and characterization of flaxseed/linseed proteins 13 2.2.2.1 High molecular weight protein fraction 14 2.2.2.2 Low molecular weight protein fraction 18 2.3 S T R U C T U R A L C H A R A C T E R I Z A T I O N OF PROTEINS 18 2.3.1 Isoelectric focusing 19 2.3.1.1 Isoelectric points of flaxseed/linseed proteins 19 2.3.2 Hydrophobicity 19 2.3.2.1 Hydrophobicity of flaxseed/linseed proteins 20 2.3.3 Sulfhydryl (SH) and Disulfide (SS) 21 2.3.3.1 SH and SS groups for flaxseed/linseed proteins 22 2.4 F U N C T I O N A L PROPERTIES OF PROTEINS 22 2.4.1 Solubility 24 iv 2.4.1.1 Solubility of flaxseed/linseedproteins 25 2.4.2 Foaming 27 2.4.2.1 Methods of foam generations 28 2.4.2.2 Methods of protein foam measurement 31 2.4.2.3 Analysis of foaming data 32 2.4.2.4 Foaming properties of flaxseed/linseed proteins 33 2.4.3 Viscosity 36 2.4.3.1 Viscosity of flaxseedAinseed proteins 37 CHAPTER 3—MATERIAL AND METHODS FOR PHASE 1: ISOLATION OF 39 THE MAJOR FRACTION OF FLAXSEED PROTEINS 3.1 M A T E R I A L S 39 3.1.1 Chemicals 39 3.1.2 Proteins 39 3.1.3 Flaxseeds 39 3.2 A N A L Y S E S 40 3.2.1 Protein determination 40 3.2.2 Amino acid analyses 40 3.2.3 Gel electrophoresis 40 3.3 E X T R A C T I O N OF F L A X S E E D PROTEINS 42 3.3.1 Studies on seed 45 3.3.1.1 Moisture and total solid determination 45 3.3.3.2 Protein determination 45 3.3.2 First and second screening experiments 46 3.3.3 Third Screening Experiment 47 3.3.4 Studies on extraction yield 47 3.3.4.1 First extraction study 47 3.3.4.2 Second extraction study 47 3.4 FRACTIONATION OF F L A X S E E D PROTEINS 48 3.4.1 First and second screening experiments 48 3.4.2 Third screening experiment 50 3.4.3 Scale-up fractionation of flaxseed proteins 51 v CHAPTER 4—MATERIAL AND METHODS FOR PHASE 2: 55 CHARACTERIZATION OF THE MAJOR FRACTION OF FLAXSEED PROTEINS 4.1 P R E P A R A T I O N OF S A M P L E S FOR C H A R A C T E R I Z A T I O N 55 4.2 SOLUBILITY CHARACTERISTICS 5 6 4.2.1 Preparation of buffers 56 4.2.2 Preparation of samples 57 4.2.3 Determination of solubility 57 4.3 ISOELECTRIC POINTS 57 4.4 PROTEIN S U R F A C E HYDROPHOBICITY 58 4.4.1 Preparation of PROD AN stock solution 58 4.4.2 Measurement of relative fluorescence intensities (RFI) of samples 59 4.4.3 Expression of protein surface hydrophobicity and statistical analyses 59 4.5 A M I N O A C I D COMPOSITION 60 4.6 C A L C U L A T I O N OF B I G E L O W ' S T O T A L HYDROPHOBICITY V A L U E (HC» 60 4.7 S U L F H Y D R Y L G R O U P (SH) A N D DISULFIDE G R O U P (SS) D E T E R M I N A T I O N 61 4.7.1 Reactive SH 61 4.7.2 Total SH 6 2 4.7.3 Changes in reactive and total SH values over time 6 2 4.7.4 Total SH and SS determination 63 4.7.5 Statistical analyses 63 4.8 C A L C U L A T I O N OF C Y S T E I N E C O N T E N T 64 4.9 F O A M I N G PROPERTIES 64 4.9.1 Preparation of samples 64 4.9.2 Foaming apparatus 65 4.9.3 Preliminary experiments for experimental conditions and set up 66 4.9.4 Foam generation 66 4.9.5 Measurement and description of foam 67 Set 1—Foam volume and drainage 67 Set 2—Foam volume and conductivity measurement 68 Set 3—Photographic illustration of foams 7 0 v i 4.9.6 Cleaning of foaming apparatus 71 4 . 1 0 VISCOSITY 71 4.10.1 Preparation of samples 71 4.10.2 Determination of apparent viscosity 7 2 4.10.3 Statistical analyses 7 2 4 .11 REGRESSION A N A L Y S E S OF THE F U N C T I O N A L CHARACTERISTICS OF T H E M A J O R 7 2 F R A C T I O N CHAPTER 5—RESULTS AND DISCUSSIONS FOR PHASE 1: ISOLATION OF 73 THE MAJOR FRACTION OF FLAXSEED PROTEINS 5.1 A N A L Y S E S O N F L A X S E E D S 73 5.2 CHARACTERISTICS OF F L A X S E E D PROTEIN E X T R A C T S 75 5.2.1 First screening experiment 75 5.2.2 Second screening experiment 7 9 5.2.3 Studies on extraction yield 83 5.2.3.1 First extraction study 83 5.2.3.2 Second extraction study 85 5.3 CHARACTERISTICS OF F L A X S E E D PROTEIN F R A C T I O N A T E D B Y A N I O N - E X C H A N G E 89 C H R O M A T O G R A P H Y 5.3.1 First screening experiment 89 5.3.1.1 DEAE-Sephacel chromatographic pattern 89 5.3.1.2 Protein content of each fraction 91 5.3.1.3 Gel electrophoretic patterns 9 2 5.3.1.4 Amino acid compositions 96 5.3.2 Second screening experiment 9 9 5.3.2.1 DEAE-Sephacel chromatographic pattern 99 5.3.2.2 Protein content of each fraction 101 5.3.2.3 Gel electrophoretic patterns 102 5.3.3 Third screening experiment 110 5.3.3.1 DEAE-Sephacel chromatographic pattern 110 5.3.3.2 Gel electrophoretic patterns 111 5.3.4 Scale-up fractionation of flaxseed proteins 118 5.4 C O N C L U S I O N FOR P H A S E 1 123 vii CHAPTER 6—RESULTS AND DISCUSSION FOR PHASE 2: 125 CHARACTERIZATION OF THE MAJOR FRACTION OF FLAXSEED PROTEINS 6.1 G E L ELECTROPHORETIC PATTERNS 125 6.2 SOLUBILITY A N D ISOELECTRIC POINTS 132 6.2.1 Effect of pH and low ionic strength (0.01 M NaCl) 134 6.2.2 Effect of pH and high ionic strength (1.0 M NaCl) 137 6.3 S U R F A C E HYDROPHOBICITY 138 6.4 A M I N O A C I D COMPOSITIONS A N D T O T A L HYDROPHOBICITY V A L U E S (He))): 141 6.5 SS A N D SH 144 6.6 C Y S T E I N E C O N T E N T 145 6.7 F O A M I N G 146 6.7.1 Preliminary experiments and experimental set up 146 6.7.2 Description of foams 148 Set 1—Foam volume and drainage 148 Set 2—Foam volume and conductivity measurement 150 (1) Conductivity measurement using the YSI conductivity probe 150 (2) Difference in conductivity readings between two conductivity meters 152 (3) Conductivity measurement using the Radiometer conductivity probe 153 (4) Interference of conductivity probes on foaming properties 157 (5) Foaming Capacity (FC) 158 (6) Relative Foam Conductivity (Cf) 159 (7) Foam Stability Index (FSI) 161 (8) Interpretation of results from the conductivity set 162 Set 3—Photographic illustration of foams 163 6.7.3 Interpretation of results from different foaming tests 175 6.8 VISCOSITY 176 6.9 REGRESSION A N A L Y S E S OF T H E F U N C T I O N A L PROPERTIES OF T H E M A J O R 179 F R A C T I O N 6.9.1 Factors Affecting Solubility 179 6.9.1.1 Hydrophobicity and solubility 180 viii 6.9.1.2 Charge and solubility 181 6.9.2 Factors affecting viscosity 182 6.9.3 Factors affecting foaming properties 183 6.9.3.1 Hydrophobicity and foaming 185 6.9.3.2 SH and SS content and foaming 186 6.9.3.3 Solubility and foaming 187 6.9.3.4 Charge, pi and foaming 188 6.9.3.5 Viscosity and foaming 189 CHAPTER 7—GENERAL CONCLUSIONS AND RECOMMENDATIONS FOR 190 FUTURE RESEARCH REFERENCES 193 APPENDLX A - Minitab™ Printout for Regression Analyses for Solubility of the 200 Major Fraction APPENDIX B - Minitab™ Printout for Regression Analyses for Viscosity of the 203 Major Fraction APPENDIX C - Minitab™ Printout for Regression Analyses for C f of the Major 207 Fraction APPENDIX D - Minitab™ Printout for Regression Analyses for FSI of the Major 210 Fraction ix LIST OF TABLES Table 1. Summary of the different approaches to extract flaxseed/linseed proteins 7 reported in the literature. Table 2. Characteristics of the high- and low- molecular weight flaxseed proteins 13 reported in the literature. Table 3. Solubility studies on flaxseed/linseed meals reported in the literature. 26 Table 4. Foaming properties studies on flaxseed/linseed protein products reported in the 34 literature. Table 5. Additional treatments of the flaxseed and flaxseed extracts investigated in the 46 first two preliminary experiments. Table 6a. Percent solid and moisture content of the NorMan flaxseeds used in the 74 experiment. Table 6b. Protein content (wet basis) of the NorMan flaxseeds used in the experiment. 74 Table 7. Extraction yield of the four batches of NorMan flaxseed protein extracts 75 prepared in the 1st screening experiment. Table 8. Molecular weight calculations of protein bands in lanes 4-7 of Figure 7. 78 Table 9. Extraction yield of the four batches of NorMan flaxseed protein extracts 79 prepared in the 2nd screening experiment. Table 10. Molecular weight calculations of protein bands in lanes 2, 3, 4 and 6 of 82 Figure 8. Table 11. Protein content estimated by the BCA Protein Assay on the 16-hr and 1-hr 85 extraction sets in the 2n d extraction study for NorMan flaxseed proteins. Table 12. Protein concentration of each fraction collected from the 1st screening 91 experiment. Table 13. Comparison of amino acid compositions (g/lOOg) of laboratory prepared 0.20 97 and 0.25 M fractions with two reported values. Table 14. Comparison of amino acid compositions (g/lOOg) of laboratory prepared FT 98 and 0.10 M fractions with three reported values. Table 15. Protein concentration of each fraction collected from the 2nd experiment. 101 x Table 16. Molecular weight calculations of protein bands in lanes 2- 6 and 8 of Figure 105 16. Table 17. Molecular weight calculations of protein bands in lanes 2- 7 of Figure 17. 107 Table 18. Molecular weight calculations of protein bands in lanes 3 -8 of Figure 21. 115 Table 19. Molecular weight calculations of protein bands in lanes 4-8 of Figure 22. 117 Table 20. Molecular weight calculations of protein bands in lanes 2, 6, 7, and 8 of 128 Figure 26. Table 21. Molecular weight calculations of protein bands in lanes 3 and 5 of Figure 27. 129 Table 22. Calculation of pi of the protein bands in lanes 1, 2, 6, 7, 8 of Figure 29. 135 Table 23. Amino acid compositions (g/ lOOg protein) of the major protein fraction from 142 NorMan flaxseed as compared with the reported values. Table 24. Amino acid compositions (g/ lOOg protein) of the whole protein extract from 143 NorMan flaxseed as compared with the reported values. Table 25. Sulfhydryl groups and disulfide bonds of the major fraction and whole 145 protein extract from NorMan flaxseeds. Table 26. Conductivity of buffers used in the foaming test measured by two different 152 conductivity probes and meters. Table 27. Foaming capacity (FC) of the whole extract, egg albumen, BSA, ovalbumin 158 and the major fraction at concentrations of 1 mg/ml. Table 28. Relative foam conductivity (Cf) of the whole extract, egg albumen, BSA, 159 ovalbumin and the major fraction at concentrations of 1 mg/ml. Table 29. Foam stability index (FSI) of the whole extract, egg albumen, BSA, 161 ovalbumin and the major fraction at concentrations of 1 mg/ml. Table 30. Apparent viscosity of the protein samples at concentrations of 1.0 mg/ml. 177 xi LIST OF FIGURES Figure 1. Simplified procedures for isolation of flaxseed 1.6S and 12S proteins reported 16 by Madhusudhan and Singh (1985c,d). Figure 2. Simplified procedures for the isolation of 1 IS flaxseed globulin by Dev and 17 Sienkiewicz (1987). Figure 3. Procedures for the isolation of flaxseed proteins in preliminary experiments. 43 Figure 4. Procedures for the isolation of the major fraction of flaxseed proteins. 53 Figure 5. An "overloaded" DEAE-Sephacel chromatographic pattern of NorMan 54 flaxseed proteins. Figure 6. Foaming apparatus for determination of the foaming properties of the major 65 fraction of flaxseed proteins at a concentration of 1.0 mg/ml. Figure 7. Reducing and non-reducing SDS-PAGE of flaxseed protein extracts extracted 77 with or without 2-mercaptoethanol on 10-15 gradient PhastGel®. Figure 8. Reducing SDS-PAGE of flaxseed protein extracts extracted with or without 81 2-mercaptoethanol from the 1st and 2nd screening experiments on 10-15 gradient PhastGel®. Figure 9. Non-reducing SDS-PAGE of the extracts and residues from extraction of 84 NorMan flaxseed proteins, as well as the extracts from re-extraction of the residues on 8-25 gradient PhastGel®. Figure 10. Reducing SDS-PAGE of NorMan flaxseed protein extracts obtained by 87 repeated extractions on 8-25 gradient PhastGel®. Figure 11. Reducing SDS-PAGE of the residues after three consecutive extractions of 88 NorMan flaxseeds and pellets from centrifugation in the 2nd extraction study on 8-25 gradient PhastGel®. Figure 12. Elution profile for fractions collected from a 50 ml DEAE-Sephacel column 90 in the 1st screening experiment. Figure 13. Reducing and non-reducing SDS-PAGE profiles of the fractions collected 93 from the 1 experiment on 10-15 gradient PhastGel . Figure 14. Native PAGE of fractions collected from the 1st screening experiment on 8- 95 25 gradient PhastGel®. Figure 15. Elution profile for fractions collected from a 50 ml DEAE-Sephacel column 100 in the 2nd screening experiment. xii Figure 16. Reducing SDS-PAGE of the fractions collected from the 2n a screening 104 experiment on 10-15 gradient PhastGel®. Figure 17. Non-reducing SDS-PAGE of the fractions collected from the 2nd screening 106 experiment on 10-15 gradient PhastGel®. Figure 18. Native PAGE of the fractions collected from the 2nd screening experiment on 108 8-25 gradient PhastGel®. Figure 19. Elution profile for fractions collected from a 50 ml DEAE-Sephacel column 113 in the 3rd screening experiment. Figure 20. Elution profile for reloading 35ml of the unbound (FT) fraction, collected 113 from the suspected overloaded 50 ml DEAE-column in the 3rd screening experiment, onto a 225 ml DEAE-column. Figure 21. Reducing and non-reducing SDS-PAGE of low-molecular weight fractions 114 collected from the 1st and 3rd screening experiments on high density PhastGel®. Figure 22. Reducing and non-reducing SDS-PAGE of the major fraction from 3rd 116 screening experiment as compared to the major fraction from the 1st screening experiment on high density PhastGel®. Figure 23. Typical DEAE-Sephacel chromatographic pattern of flaxseed proteins 120 (225ml column). Figure 24a. Ultraviolet absorption spectrum of the peak fraction from the FT fraction of 121 flaxseed protein from a 225 ml DEAE-column at 0.10 M NaCl. Figure 24b. Ultraviolet absorption spectrum of the peak fraction from the 0.25 M 121 fraction eluted from a 225 ml DEAE-column at 0.25 M NaCl. Figure 24c. Ultraviolet absorption spectrum of the peak fraction from the 0.50 M 122 fraction eluted from a 225 ml DEAE-column at 0.50 M NaCl. Figure 25. Native PAGE of the 0.25 M fractions from the 6 scale-up fractionation 126 batches on 8-25 gradient PhastGel®. Figure 26. Reducing SDS-PAGE of the 0.25 M fractions from the 6 scale-up 128 fractionation batches on 8-25 gradient PhastGel®. Figure 27. Non-reducing SDS-PAGE of the FT, 0.25 M fractions and whole extract on 129 8-25 gradient Phastgel®. Figure 28. Solubility profiles of the FT fraction, major fraction and whole protein 133 extract from NorMan flaxseeds at different pHs and two NaCl concentrations. xiii Figure 29. IEF profiles for the 0.25 M fractions and whole extract on IEF 3-9 135 PhastGel®. Figure 30a. Surface Hydrophobicity of the major fraction measured at pH 3, 5, 7 in 139 citrate-phosphate buffer with 0.01 M or 1.0 M NaCl. Figure 30b. Surface Hydrophobicity of the whole extract measured at pH 3, 5, 7 in 139 citrate-phosphate buffer with 0.01 M or 1.0 M NaCl. Figure 31. Foam conductivity decay curves for measured using the YSI conductivity 151 probe. Figure 32. Simplified illustration of the YSI conductivity probe (YSI Model 31, Yellow 151 springs, Ohio). Figure 33. Foam conductivity decay curves measured using the Radiometer 153 conductivity probe. Figure 34. Simplified illustration of the Radiometer conductivity probe (Copenhagen 154 Type CDC 104; Radiometer Analytical, Lyon, France). Figure 35. Foam conductivity decay curve for the major fraction at concentrations of 1 156 mg/ml, in citrate-phosphate buffer with 0.01 M NaCl at pH 3. Figure 36. Foam conductivity decay curve for the major fraction at concentrations of 1 156 mg/ml, in citrate-phosphate buffer with 0.01 M NaCl at pH 5 (a) and pH 7 (•). Figure 37. Foam conductivity decay curve for egg albumen, whole extract, ovalbumin 157 and BSA, at concentrations of 1 mg/ml, in citrate-phosphate buffer with 0.01 M NaCl at pH 7. Figure 38. Comparison of the foam of the whole protein extract (left) with the major 164 protein fraction (right) from NorMan flaxseed after sparging of nitrogen gas at a flow rate of 35 ml/min into 5 ml of samples at 1 mg/ml for 1 minute. Figure 39. Foam of the major fraction of NorMan flaxseed at 1 mg/ml in citrate- 166 phosphate buffer with 0.01 M NaCl at pH 7 in the column at 3 min. Figure 40. Foam of the major fraction of NorMan flaxseed at 1 mg/ml in citrate- 167 phosphate buffer with 0.01 M NaCl at pH 3 in the column at 3 min. Figure 41. Foam of egg albumen at 1 mg/ml in citrate-phosphate buffer with 0.01 M 168 NaCl at pH 7 in the column at 3 min. Figure 42. Foam of the major protein fraction of NorMan flaxseeds at 1 mg/ml in 169 citrate-phosphate buffer with 0.01 M NaCl at pH 7 in the column at 28 min. xiv Figure 43. Foam of the major protein fraction of NorMan flaxseed at 1 mg/ml in 170 citrate-phosphate buffer with 0.01 M NaCl at pH 3 in the column at 26 min. Figure 44. Foam of egg albumen at 1 mg/ml in citrate-phosphate buffer with 0.01 M 171 NaCl at pH 7 in the column at 25 min. Figure 45. Deterioration over time of foams of the major protein fraction from 172 NorMan flaxseed, at 1 mg/ml in citrate-phosphate buffer with 0.0IM NaCl at pH 7 or 3, as compared to that of the egg albumen. Figure 46. Deterioration over time of foam of the major fraction, at 1 mg/ml in citrate- 173 phosphate buffer with 0.0IM NaCl at pH 7, scooped out of the column and placed on a slide. Figure 47. Deterioration over time of foam of the major fraction, at 1 mg/ml in citrate- 174 phosphate buffer with 0.01M NaCl at pH 3, scooped out of the column and placed on a slide. xv ACKNOWLEDGEMENTS I would like to express my deepest and most sincere gratitude to my supervisor, Dr. Eunice Li-Chan, for her guidance and advice throughout my studies. She is not only a respectable and knowledgeable supervisor but also a supportive and understanding friend. It is really a blessing to be her student! I am thankful to Dr. T. Durance and Dr. C. Seaman for being members of my supervisory committee and for their valuable advice. Special thanks to Dr. D. Oomah for being a member of my examining committee and Dr. B. Skura for being my examination chair. I would also like to thank Mr. Sherman Yee, Ms. Val Skura, Ms. Angela Gerber, Ms. Joyce Tom, Ms Jeannette Law, and Ms Brenda Barker for their technical supports throughout my studies. Thanks to Dr. Sultanbawa for passing on important techniques for my project. Thanks to everyone at Food Science who shared their valuable experience with me, learned with me and helped out with me, especially members of Dr. Li-Chan's research group, Andrea, Bo, Eddy, Eugene, Hai Hong, Judy, Karen and Nooshin. I would also like to acknowledge the University of British Columbia for granting the University Graduate fellowship, the Natural Science & Engineering Research Council of Canada (NSERC) and the Faculty of Agricultural Sciences for providing me scholarships for my studies. In addition, many thanks to the Flax Council of Canada, the NSERC, and Agriculture & Agri-Food Canada for funding this research project. Last but not least, I would like to thank my family, friends, and brothers & sisters from SVPGMBC for their continuous support, love and prayers. Mom, Dad and brother, thanks for your endless love and caring; Anthony, thanks for your encouragement and understanding through these years. xvi CHAPTER 1—INTRODUCTION Flaxseed, also called linseed, is the small flat oval seed from flax (Linum usitatissimum). The textile variety of flax has been used by humans for more than two centuries while the oilseed variety had been used mainly for industrial applications until recently, when it has been gaining popularity in the health food market because of its reported health benefits and disease preventive properties. The market for edible flaxseed is currently limited to its oil and the whole intact flaxseed. The defatted meal, which is the residue after crushing of flaxseed oil, is mainly used as livestock feed. Recent research also focuses mainly on the whole intact flaxseed, its oil and their health benefits to humans. Limited attention has been brought to the proteins of flaxseed. With the increasing demand for vegetable sources of proteins, knowledge on flaxseed proteins as a food source is valuable. Isolation and characterization of a new or novel protein are the basic steps toward a better understanding of the protein and the application of the protein into food systems. Since the first isolation of flaxseed proteins by Osborne in 1892, most research on flaxseed proteins has been concerned with the extraction of protein from the oil-crushed meal. Limited research has been done on the characteristics and functionality of the specific proteins fractionated from flaxseed. To date, information on specific components of flaxseed protein shows that flaxseed consists of two major storage proteins, one with high (11-12 S) and the other with low (1.6-2 S) molecular weights. The high molecular weight fraction is the more dominant fraction and is more soluble in salt solution while the low molecular weight fraction is water soluble and 1 basic in nature. As the two major proteins have different solubility and charge properties, they can be separated based on these distinct properties. The higher molecular weight major fraction has received greater interest, primarily because it is the major fraction. Preliminary findings suggest that it has properties similar to the major storage proteins of other important oilseeds such as soy and canola, however, most of the properties of this major fraction of flaxseed proteins are still awaiting investigation. With all the unexplored area on this valuable protein of great potential for food uses, the main objective of this thesis was to isolate and characterize the major fraction of flaxseed proteins. Two phases were carried out to fulfill the objective. Phase 1 focused on the modification and simplification of protocols established for a concurrent project of isolation of cadmium-binding proteins of flaxseed, as well as a basic characterization of the protein fractions fractionated by anion-exchange chromatography. The isolation procedure was repeated to obtain sufficient quantity of the major fraction for further characterization in Phase 2. In Phase 2, the molecular characteristics and structural properties such as surface hydrophobicity, isoelectric points, sulfhydryl groups and disulfide bonds of the major fraction were determined. Amino acid compositions of the fraction as compared to the whole protein extract were also determined. In addition, the functional characteristics of solubility, foaming and viscosity of the major fraction and the relationships of these structural and functional characteristics were also studied. 2 CHAPTER 2—LITERATURE REVIEW 2.1 OVERVIEW OF FLAX AND FLAXSEEDS In this thesis, the word "flax" refers to the whole plant of Linum usitatissimum. The word "flaxseed" or "linseed" refers to the seeds of the flax plant. Depending on the terms used by the reference cited, the words "flaxseed" and "linseed" are used interchangeably. Generally, the term "linseed" is more commonly used to refer to non-edible oilseed, in which the oil is used in industrial applications such as paints, and the meal or fat-cake after oil removal is mainly used as livestock feed. The term "flaxseed" more commonly refers to edible seeds from flax which are consumed as whole seed or are used for production of edible oils and food applications. 2.1.1 Flax Flax {Linum usitatissimum) is one of the oldest crops known to man (Peterson, 1958). It is an erect annual herb, about 60 to 120 cm high, woody-stemmed, with small bluish, bluish violet or whitish flowers that grow on terminal panicles and small strap-like, sessile leaves (Peterson, 1958; Salunkhe et al., 1992). Its flowers are self-fertile and its fruits are capsular with five cells, each containing two seeds (Peterson, 1958; Salunkhe et al., 1992). There are two main categories of varieties of cultivated flax: one is grown for its fibre to make linen and the other one is for oil. Fibre varieties grow tall with few branches while oilseed varieties are comparatively short with more branching (Peterson, 1958). Flax cultivated in Canada is mainly the oilseed varieties (Flax Council of Canada, 2001). Canada contributes up to 40% of the world's total flax production and has become a 3 leader in the production and export of flax since 1994. Flax is the third major oilseed crop in Canada, after canola and soy (Flax Council of Canada, 2001). 2.1.2 FlaxseedVLinseed Flaxseed/linseed is yellowish or blackish brown, small, flattened, and oval with a smooth shiny coat (Salunkhe et al., 1992). It has an average dimension of 2.5 mm x 5 mm x 1 mm and an average weight of 3-13 mg per seed, with its fibrous seed coat accounting for 30-39% of the seed weight (Wanasundara and Shahidi, 1997a). The seed coat consists of five distinct layers: epidermis (mucilage cells), round cells, fibre cells, cross (transverse fibre) cells, and pigment cells (Peterson, 1958; Freeman, 1995). The outermost epidermal layer expands extremely rapidly upon hydration (Freeman, 1995) and accounts for 2-7.5 % of the dry weight of the seed (Madhusudhan and Singh, 1983). The next four layers underneath are also termed the "true hull", which is characterized by its tough, fibrous, low oil and low protein content (Peterson, 1958). Under the seed coat is the endosperm (inner hull), which contains some oil and protein (Peterson, 1958). It is difficult to separate it from the true hull and thus the two are usually analysed together as "hull" (Peterson, 1958). The two cotyledons under the endosperm are where most of the oil and protein are found. Although Canada is the largest producer of flaxseed/linseed, very limited amounts of flaxseed/linseed are crushed and thus a very small amount of flaxseed/linseed meal is available in Canada (Bhatty and Cherdkiatgumchai, 1990). Proximate composition of whole flaxseed/linseed varies considerably, depending on cultivar, growing conditions, seed processing and analytical methods. Bhatty and Cherdkiatgumchai (1990) reported 4 proximate composition of a Canadian-grown flaxseed to be 41 % (dry basis) fat, 26 % protein (% N x 6.25), 4 % ash, 5 % crude fibre and 24 % nitrogen-free extract (by difference). 2.1.3 Flaxseed/Linseed Meal The term "flaxseed/linseed meal" is quite ambiguous. It has been used for ground unextracted seed with about 35 % oil, ground seed cake with about 10 % oil and solvent-extracted cake meal with about 3% oil (Salunkhe et al., 1992). The term "defatted flaxseed/linseed meal" does not specify the oil content or the method of delipidation. Seeds could be pressed with hydraulic expeller or screw press, cold pressed under high pressure with supercritical carbon dioxide, or extracted with different solvents. Some of the anti-nutrients and/or pro-nutrients in flaxseed/linseed meal include non-starch polysaccharides (mucilage), cyanogenic glycosides, phytic acid, phenolics, mammalian trypsin and chymotrypsin inhibitors, linatine (a vitamin B6 antagonist) and lignans (Bhatty and Cherdkiatgumchai, 1990). 2.1.4 Flaxseed/linseed proteins Oomah and Mazza (1993a) reviewed the literature on the protein content, amino acid composition, fractionation, and functional properties of flaxseed proteins. They found that the reported protein content in flaxseeds ranges from 10.5 % to 31 %. The protein content (N x 6.5) of defatted, moisture-free flaxseed in Canadian cultivars varies between 31.3 and 36.7 % (Oomah and Mazza, 1993a). Oomah and Mazza (1995) later reported the protein content (N x 5.41) of 109 flaxseed accessions from the world collection as being normally distributed ranging from 20.9^ 18.1 %, with a mean of 34.5 %. Over two-thirds 5 (69 %) of the population have protein contents ranging from 29-37 % (Oomah and Mazza, 1995). Differences can be attributed to both genetic and environmental factors, as well as the nitrogen-to-protein conversion factor used (N x 6.5 vs N x 6.25 vs N x 5.41). 2.2 FLAXSEED/LINSEED PROTEINS Flaxseed/linseed protein has not been extensively studied. The earliest report on linseed proteins was recorded in 1892 by Osborne, who extracted lipid-free linseed meal with water, 10 % NaCl solution, or dilute alkali (Vassel and Nesbitt, 1945). Osborne found a globulin with 18.6 % nitrogen and an albumin like protein with 17.7 % nitrogen. As a result, he estimated the nitrogen-to-protein conversion factor of 5.5 for the whole protein of the linseed (Smith et al, 1946). Vassel and Nesbitt (1945) were the next group which worked on flaxseed proteins and named the high molecular weight (MW) 12S globulin "linin" and the low molecular weight 2S "conlinin". Thereafter, most studies on flaxseed/linseed proteins have mainly been focusing on the whole protein extract from defatted meals. 2.2.1 Chronological review of the extraction studies on flaxseed/linseed proteins The presence of polysaccharides in the seed coat that swelled in aqueous medium was found to hinder the separation of proteins (Smith et al, 1946; Sosulski and Bakal, 1969; Madhusudhan and Singh, 1983). Table 1 summarizes in chronological order, the different approaches used to reduce the viscosity caused by the mucilage, as well as the different methods of grinding, delipidation, and protein extraction. 6 Table 1. Summary of the different approaches to extract flaxseed/linseed proteins reported in the literature. Researchers Grinding Dehulling / Mucilage reduction Delipidation Extraction condition (or suggested optimum) Painter and Roller mill Grinding defatted meal for Nesbitt, followed by ~ 1.5 hr in ball mill; 1946 ball mill/ sifting through #40 Wiley mill (0.42 mm) sieve. 20 hr Soxhlet 3 g seed to 100 ml of 1 N NaCl / apparatus with M g C l 2 ; pet ether shaken at Rm Temp; centrifuge & supernatant taken Smith et al, Ball mill Sifting through graded Hexane 2 water extractions with 1:20 or 1946 sieves; Air separation with the Raymond Whizzer l:10w/v H 2 0 respectively; alkaline extraction & acid precipitation Sosulski and Waring Hexane @ Rm 5 g defatted meal in 100 ml 0.2% Bakal, blender Temp lhr x 2; NaOH 1 hr; 1969 during Not done 1:10 seed to centrifuged @ 1000 x g 15 min; delipidation solvent Waring blender re-extracted once; isoelectric precipitation @ pH 4.4 Madhusudhan A P V fruit Soaking in H 2 0 (1:3 w/v) Hexane @ Rm 5 g defatted meal in 25 ml 1 M and Singh, pulper; 1 hr x 5; Temp x 5 NaCl, 1 hr; 1983;1985c flaking passing through A P V fruit centrifuged @ 6000 rpm 30 min; machine; pulper 1/32 inch sieve & supernatant dialyzed against 0.05 Satake paddy nylon brushes x 4; M phosphate buffer, pH 7.6 separator flaking machine with roll gap during 0.4 mm; mucilage drying @ 40 °C 18 hr reduction crossflow drier 55 nrVmin; treatment separating hulls with Satake paddy separator Dev et al, 1986a Coffee grinder Not done Hexane @ Rm 1.0 M NaCl; Temp; pH 8.0; vacuum drying Rm Temp @40 °C 6-8 hr Dev and Sienkiewicz, 1987 M i l l (25 mm screen) Not done Repeated alternating extraction with diethyl ether & acetone @ 4°C; desolventized at Rm Temp 1.0 M NaCl in phosphate buffer. 1:20 (w/v); repeated centrifugation @ 3000 x g 15 min @ 20 °C; diluted with 5.5 x vol of H 2 0 ; cryoprecipitated @ 4°C; redissolved pellet upon centrifugation 7 Table 1 (con't). Summary of the different approaches to extract flaxseed/linseed proteins reported in the literature. Researchers Grinding Dehulling / Reducing mucilage Delipidation Extraction condition (or suggested optimum) Oomah et al, Hexane; 1:16 (w/v) of 1.28 M NaCl in 1994 dried in forced-air 50 m M phosphate buffer, Not applicable oven 24 hr pH6.8 (started with commercial cold pressed @ 5 0 ° C centrifuged @ 20 000 x g, flaxseed defatted meal) 30 min; supernatant taken Wanasundara Wire mesh Soaking in N a H C 0 3 Hexane 1:33 (w/v) of 2.75% (w/v) and Shahidi, (2 mm) (0, 0.05 or 0.10 M) (1:5 w/v) Sodium Hexametaphosphate, 1996; 1997a during 1:10 (w/v) for 3, 6 or 12 hr; 1 min; pH8.9 mucilage draining excess H2O; 5 x in Warning reducing dispersing in 0.10 N H C 1 ; blender treatment; passing through 2mm wire mesh; Waring washing 5x with H 2 0 blender (1:10 w/v); during drying in forced-air oven @45°C delipidation Wanasundara Wire mesh Adding seeds to 0.01 M and Shahidi, (2 mm) acetate buffer (1:5 w/v) with 1997a during mucilage reducing different enzymes of Celluclast® 1.5L, Pectinex® Ultra SP & treatment Viscozyme® L ; Not applicable (not the focus of the study) incubating for 1, 3, or 6 hr; stopping reaction by 50 ml O . lONNaOH; washing & drying as above 8 After the pioneering work of Osborn in 1892, followed by that of Vassel and Nesbitt in 1945, Painter and Nesbitt (1946) investigated the nitrogenous constituents of flaxseed and concluded that 85-90 % of the flaxseed nitrogen was dispersed at a pH near 11 and minimal solubility (15-21 % nitrogen) was found at pH 3.5-4.0. They also found that "more nitrogen dispersed from the endosperm than from whole meal or hull and 20% of the nitrogen was not precipitable with trichloroacetic acid (non-protein nitrogen)". Smith et al. (1946) isolated the flaxseed proteins by alkali extraction followed by acid precipitation and they found that about 21-23% of nitrogen was soluble at the point of minimal nitrogen solubility (pH 3.8 in one preparation and pH 5.1 in another). They also observed that the presence of polysaccharide gums from flaxseed hulls interfered with the settling and isolation of the protein. As a result, they tried to improve protein extraction by removing the hull with graded sieves and air separation. They found that removal of hull effectively improved the preparation of protein isolate but they did not successfully develop an efficient and economical method of hull removal. Sosulski and Bakal (1969) extracted protein from hexane-defatted ground flaxseed with 0.2% NaOH (1:20 meal-to-solvent ratio) and then precipitated the protein by adjusting the pH to the isoelectric point of 4.5. They also investigated the extraction rates and characteristics of isolated protein concentrates obtained from rapeseed, flax and sunflower meals as compared with soybean meal. A tan-coloured flaxseed protein meal was prepared and it had a lower protein but higher fibre content than soybean meal. The flaxseed proteins were found to be less water-soluble but were more soluble in salt solutions as compared to soy proteins. 9 Sosulski and Sarwar (1973) continued to explore the amino acid composition of the different oilseed meals and found that flaxseed meal and protein isolate contained more of the nonessential amino acids arginine, aspartic acid as well as glutamic acid and rated poorly in essential amino acid indices and protein scores. Madhusudhan and Singh (1983) developed a method to dehull and demucilage the linseed. They also did an extensive study on the proteins extracted in 1.0 M NaCl. Solubility of this protein fraction in water "exhibited a typical U-shape with minimum solubility in the range pH 3-6, and more or less constant after pH 8.0" (Madhusudhan and Singh, 1983). Addition of NaCl shifted the pH minimum to 0.5-4.5, while addition of sodium hexametaphosphate shifted the minimum to pH 0.6-5.3. The extinction coefficient (E 1 % i c m ) was found to be 10.1 %"1cm"1 in 1.0 M NaCl. The gel filtration pattern consisted of three peaks, which contributed to 3, 67, and 30 % of the total protein. They suggested that the first peak contained mostly nucleic acid and some amount of high molecular weight aggregate while the second peak consisted of more than one protein component when analysed with disc electrophoresis. They also found that the DEAE-Sephadex chromatographic pattern showed four components and they noted that "the first two peaks appeared to be anionic in nature as they eluted without salt gradient". However, fractions unbound to a DEAE-Sephadex anion-exchange column should in fact be cationic because the positively charged protein fractions would be repelled by the positively charged DEAE-Sephadex resins. These authors later investigated the detoxification of linseed meal by water boiling (Madhusudhan and Singh, 1985a). They concluded that the detoxification procedure they developed resulted in an increase in in vitro digestibility by 10 38 % and decrease in available lysine content by 30 %. The high molecular weight fraction from the detoxified meal was also dissociated. Dev et al. (1986a) investigated the nitrogen extractability and buffer capacity of linseed flour prepared by grinding and sieving (0.25mm sieve) hexane-delipidated cold-pressed seed meal. They found that the smallest amount of nitrogen (20 %) was extracted in water in the pH range 4.0-4.6 and the greatest (80 %) at pH 12.0. The extraction time did not affect nitrogen extractability but the seed to water ratio did. More nitrogenous material was extracted with progressively higher amounts of solvents. Addition of NaCl (0.10-1.0 M) broadened the pH range of minimum nitrogen extractability and shifted it toward lower pH region. Addition of NaCl also increased nitrogen extractability, particularly in the pH range 4.0-8.0. A higher buffer capacity of the flour was observed in the acidic medium than in the alkaline medium. They concluded that maximum nitrogen extractability was 76 % at pH 8.0 in 1.0 M NaCl. They also recommended that pH for the extraction of defatted linseed meal protein should not exceed 10.0 at room temperature due to possible denaturation of proteins, loss of lysine and cysteine, and formation of undesirable amino acid derivatives such as lysinoalaine, ornithoalanine, fi-amino-alanine and lanthionine. Oomah et al. and Wanasundara & Shahidi were the two groups that further studied the extraction conditions for the flaxseed proteins. Oomah et al. (1994) used response surface methodology to optimize the protein extraction conditions from commercial flaxseed meal. Meal extracts were prepared by constant magnetic stirring of meal-solvent mixtures for 30 minutes at room temperature. Five solvent-to-meal ratios (v/w) expressed in 1/kg 11 were used: 10, 16, 25, 34, 40. The solvent consisted of 50 mM sodium phosphate buffer (pH 6.0-10.0) to which sodium chloride was added to give concentrations of 0.32, 0.80, 1.28 and 1.60 M. The protein content of the supernatant recovered after centrifugation at 20 000 x g for 30 minutes was determined. They found that solvent pH was not a significant factor affecting protein solubility and recovery but solvent-to-meal ratio and ionic strength were highly significant. At a solvent-to-meal ratio of 16, both protein solubility and protein content were high. Up to 82 % of flaxseed protein could be solubilized at pH 6.8, ionic strength of solvent of 1.28 M NaCl and a solvent-to-meal ratio of 16. Wanasundara and Shahidi (1996) tried a different approach and found different results. They developed a two-phase solvent extraction system consisting of methanol-ammonia-water/hexane which could simultaneously extract the oil from flaxseeds and detoxify the meal by reducing the cyanogenic glycoside, phenolic acid, condensed tannin and soluble sugar contents. Such systems were however unable to remove the phytic acid from the meal (Wanasundara and Shahidi, 1994a). They also used sodium hexametaphosphate to isolate proteins from mucilage-reduced flaxseed meal and employed response surface methodology with a central composite rotatable design to study the nitrogen extractability and protein recovery as a function of medium pH, sodium hexametaphosphate concentration and solvent-to-meal ratio (Wanasundara and Shahidi, 1996). In contrast to Oomah et al. (1994), they found that with incorporation of sodium hexametaphosphate in the extraction system, pH became the most significant factor while solvent-to-meal ratio was the least significant factor affecting nitrogen extractability and protein recovery. Up to 77 % of total nitrogen could be extracted from the low-mucilage flaxseed meal with 12 sodium hexametaphosphate of 2.75 % (w/v) at pH 8.9 and a meal-to-solvent ratio of 1:33 (w/v) (Wanasundara and Shahidi, 1996). The sodium bicarbonate soaking treatment for mucilage reduction might also account for the difference in findings of the Wanasundara and Shahidi group from the Oomah et al. group. 2.2.2 Isolation and characterization of flaxseed/linseed proteins Since the classification of linseed proteins into the high molecular weight 12S globulin linin and the low molecular weight 2S conlinin by Vassel and Nesbitt (1945), there has been limited research on the separate fractions. Table 2 summarizes the characteristics of the high- and low- molecular weight flaxseed proteins reported in the literature. Table 2. Characteristics of the high- and low- molecular weight flaxseed proteins. h i g h - M W fraction l o w - M W fraction Molecular Weight (kDa) 252 -298 1 ; 3202 15 -18 3 Sedimentation velocity US 3 ; 12S 11.45S4 2S 3; 1.6 S 5 % of total protein 58% 3; 66% 1 42% 3; 20% 5 Solubility salt1'2'4'6 water 3 ' 5 ' 6 # of components SDS-PAGE l 5 Urea-PAGE 6(1 acidic, 2 neutral, 3 basic)1 n.a.7 Secondary structure (%) alpha-helix 3-4 1 265 beta-pleat structure 17 1 325 Aperiodic -80 1 42 5 11, 18, 29, 42,61 \ Component M W (kDa) 14.4, 24.6, 30.0, 35.2, 50.9 2 — Total # of a.a per 100 kg protein 871 1 885 5 A.a composition rich in D, E , R 1 , 2 ' 6 K , C , E , G 3' 5' 6 1 Madhusudhan and Singh, 1985c 5 Madhusudhan and Singh, 1985d 2 Marcone et al, 1998 6 Dev et al. 1986b 3 Youle and Huang, 1981 7 Not available 4 Dev and Sienkiewicz, 1987 13 2.2.2.1 High molecular weight protein fraction Vassel and Nesbitt (1945) isolated the high-molecular-weight fraction by differential isoelectric precipitation at pH 5.7. The purified 12S protein, which they named linin, was shown to be homogeneous, with an isoelectric point of 4.75, and contained 17 % nitrogen, 0.6 % sulphur, and 0.54 % carbohydrate. Youle and Huang (1981) extracted the 2S and 1 IS protein from linseed with 0.035 M sodium phosphate buffer containing 1 M NaCl, pH 7.5 and centrifuged the extract in a 5 to 30 % sucrose gradient in the same buffer. They also found that linseed contained a minor protein peak with a sedimentation value higher than 1 IS. The 1 IS protein fraction accounted for 58 % of total proteins and was 41, 61 and 82 % soluble in water, 0.05 M NaCl and 0.5 M NaCl respectively. Madhusudhan and Singh (1985c) did a detailed study on the isolation and characterization of the major fraction of flaxseed proteins. Figure 1 summarizes the procedure used to isolate the two major flaxseed proteins. The fraction was isolated to homogeneity by gel filtration on Sepharose 6B. The major fraction constituted 66 % of the total proteins, contained no phosphorous, had a molecular weight (MW) of 294 000 Dal tons (Da) with a sedimentation value of 12, and was made up of at least five non-identical components. In DEAE-Sephadex chromatography, it was eluted at salt concentration of 0.24 M NaCl. Its ultraviolet absorption maximum and minimum were at 280 nm and 250 nm respectively, with an extinction coefficient of 7.6 %'x cm"1. It contained a high amount of aspartic acid, glutamic acid and arginine, which was typical of most globulins of other oilseed proteins. Its intrinsic viscosity in 14 phosphate buffer was 3.1 ml/g. In PAGE at pH 3.6 with buffers of low ionic strength of fx 0.025, the fraction was found to dissociate into at least 2 components. They commented that the dissociation of oligomeric proteins at low ionic strength and pH had been reported for many oilseeds such as soybean, sunflower and rapeseed. Dev et al. (1986b) compared the electrophoretic patterns and amino-acid profiles of the defatted flour, protein isolate, water- and salt- soluble fractions of flaxseed. The defatted flour was prepared by milling cold-pressed flaxseed meal; the protein isolate was prepared by alkaline extraction at pH 9.5 followed by isoelectric precipitation at pH 4.2; the water-soluble and salt-soluble fractions were prepared by successive extractions with distilled water and 0.5 M NaCl. The salt-soluble fraction consisted of protein of medium mobility while the water-soluble fraction had a high mobility component which was absent in the salt-soluble fraction. Dev and Sienkiewicz (1987) isolated the major fraction using different procedures (Figure 2) and characterized it as an 1 IS globulin. They also reported the oligomeric nature of the linin which would dissociate into smaller components at lower salt concentration and lower pH. Marcone et al. (1998) isolated and characterized the salt-soluble seed globulins of various dicotyledonous plants and found great similarities in the structural and chemical properties among these storage globulins. Flaxseed globulins had a molecular weight of 320 kDa and consisted of basic and acid subunits (Marcone et al, 1998). 15 Dehulling, grinding & delipidation Khategaon flaxseed Water extraction & Centrifugation Supernatant Pellet 1 1 Cation - Exchange Chromatography Repeated extractions Step gradient 0-0.4 M NaCl ^ 1 fNH,YvSn, P ( 4)2 04 recipitation Fraction eluted at 0.09 M NaCl -1.6 S protein " Pellet Supernatant ^ Gel Permeation^Chromatography Major Fraction - 12S protein Figure 1. Simplified procedures for isolation of flaxseed 1.6S and 12S proteins reported by Madhusudhan and Singh (1985c,d)-16 Grinding & delipidation I 1 M NaCl Extraction (1:20 w/v), repeated Centrifugation Pellet Supernatant - total protein I Dilution with 5.5 (v) dd H 2 0 I Cryoprecipitation 4 °C Supernatant Pellet i Dissolved in extraction buffer - crude globulin \ Gel Permeation Chromatography, Ultrafiltration 2x I Purified 11S globulin Figure 2. Simplified procedures for the isolation of 1 IS flaxseed globulin by Dev and Sienkiewicz (1987). 17 2.2.2.2 Low molecular weight protein fraction Vassel and Nesbitt (1945) isolated conlinin from a dioxane-treated glycol extract of flaxseed meal. Youle and Huang (1981) extracted the 2S protein and 1 IS protein from flaxseed in 0.035 M sodium phosphate buffer with 1 M NaCl, pH 7.5 and centrifuged the extract in a sucrose gradient from 5 to 30 % sucrose in the same buffer. The 2S fraction accounted for 42 % of the total proteins and was 93, 97, and 99 % soluble in water, 0.05 M NaCl and 0.5 M NaCl respectively. Madhusudhan and Singh (1985d) noted that the fraction was basic in nature but they did not show details of the study. They isolated the low-molecular weight protein fraction by water extraction and CM-Sephadex C-50 Chromatography (Figure 1). The 1.6S protein they isolated had an ultraviolet absorption maximum at 280 nm. A shoulder was found at 290 nm. Both the Youle & Huang, and Madhusudhan & Singh groups characterized the low-molecular weight fraction of flaxseed meal as having high content of glutamate, glutamine, asparagine, aspartate, arginine and cysteine, and being water-soluble. 2.3 STRUCTURAL CHARACTERIZATION OF PROTEINS Hydrophobic, electrostatic and steric parameters are three important parameters that affect the structure of protein (Nakai, 1983). These three parameters are important in the understanding of the biological properties (such as enzymatic activity) as well as the functional properties of protein molecules (Li-Chan and Haskard, 1998). 18 2.3.1 Isoelectric focusing Isoelectric focusing (IEF) can be regarded as electrophoresis within a pH gradient. Macromolecules migrate through the gradient as long as they retain a net positive or negative charge until they reach the point in the pH gradient which corresponds to their isoelectric point (Andrews, 1986). 2.3.1.1 Isoelectric points of flaxseed/linseed proteins: Vassel and Nesbitt (1945), found the pi of the 12S protein they isolated at 4.75. Sosulski and Bakal (1969) used pH 4.5 as the isoelectric point of the flaxseed meal proteins for isoelectric precipitation. Wanasundara and Shahidi (1994b) used 3.55 + 0.05 as the isoelectric pH for precipitating the flaxseed extracts in sodium hexametaphosphate. 2.3.2 Hydrophobicity The term "hydrophobic" describes a solute that has little or no affinity for an aqueous solution. Hydrophobicity is the tendency of non-polar solutes to adhere to one another in an aqueous environment (Cardamone and Puri, 1992). Such association is referred to as the hydrophobic effect or hydrophobic interactions, rather than hydrophobic bonding (Li-Chan, 1999). The hydrophobic effect is triggered by the tendency of proteins to reduce the entropically unfavourable contact between non-polar groups with water and the need to form enthalpically favourable non-covalent interactions (Li-Chan, 1999). Surface (or effective) hydrophobicity has more influence on functionality than total hydrophobicity (Li-Chan and Haskard, 1998). 19 Fluorescent probe methods have been the most popular means of quantifying protein surface hydrophobicity due to their simplicity, speed, ability to predict functionality and use of small quantities of purified protein for analysis (Li-Chan and Haskard, 1998). The quantum yield of fluorescence and wavelength of maximal emission of the probe depend on the polarity of its environment (Li-Chan, 1999). The fluorescent probe methods measure hydrophobic groups on the protein surface that are able to bind the probe. Under conditions with excess probe, the initial slope (S0) of the fluorescence intensity versus protein concentration plot has been shown to be correlated to the effective hydrophobicity (Li-Chan and Haskard, 1998). The most popular types of fluorescent probes include anionic probes of the aromatic sulfonic class, such as amphiphilic l-anilinonaphthalene-8-sulfonate (ANS), and anionic probes of the fatty acid analogue type, such as ds-parinaric acid (Li-Chan, 1999). The use of anionic probes, however, is more likely to result in an inaccurate estimation of protein hydrophobicity because of the contribution of electrostatic interaction on the binding of probes. As a result, a non-polar non-dissociable fluorescent probe, 6-propionyl-2-(N,N-dimethyl-amino)naphthalene (PRODAN) has been proposed as an effective alternative (Li-Chan, 1999; Alizadeh-Pasdar and Li-Chan, 2000). 2.3.2.1 Hydrophobicity of flaxseed/linseed proteins: The only report of surface hydrophobicity measurement on flaxseed meal proteins was by Wanasundara and Shahidi (1997b) where the effect of acylation on surface hydrophobicity was measured with ANS fluorescent probe. The fluorescence intensity of the unmodified flaxseed protein isolated was reported to be 200 units per mg of protein. Acylation increased the surface hydrophobicity of the flaxseed protein isolates. 20 2.3.3 SULFHYDRYL (SH) AND DISULFIDE (SS) Sulfhydryl (SH) and disulfide (SS) groups have been widely implicated as important functional groups in many food proteins (Beveridge et al, 191 A). SS bonds are formed following attainment of the most thermodynamically stable conformation. They enhance tertiary structure and confer molecular stability of proteins (Kinsella, 1981). Ellman's method (Ellman, 1959) has been the most common method for analyzing the sulfhydryl groups of proteins. The Ellman's reagent, 5,5-dithiobis(2-nitrobenzoic acid, (DTNB) undergoes the thio-disulfide interchange reaction in the presence of a free SH group, and gives rise to a characteristic chromophoric product, 5-thiobis-(2-nitrobenzoic acid) (TNB). The stoichiometry of the SH group to TNB formed is 1:1 (Ming et al, 1995). Thannhauser et al (1983) developed a method for quantitation of SS bonds in polypeptides and proteins. The sample of protein or polypeptides containing disulfide bonds was added to a solution of 2-nitro-5-thio-sulfobenzoate (NTSB) in 0.2 M Tris base containing 2 M guanidine thiocyanate (GuSCN), 3 mM EDTA, and 0.1M sodium sulfite at pH 9.5. Guanidine thiocyanate was added to the protein solutions to denature the proteins, making the disulfide bonds accessible. EDTA was added to complex the metal ions which catalyze reoxidation of thiols to disulfides (Thannhauser et al, 1983). The disulfide bonds of proteins or peptides were cleaved by excess sodium sulfite and forming chromophoric 2-nitro-5-thiobenzoate (NTB). Since 1 mole of disulfide bonds produced 1 mole of NTB quantitatively, the concentration of disulfide bonds could be calculated by 21 use of the extinction coefficient of NTB at 412 nm, 13 600 M'cm"1. Damodaran (1984) later found out that the NTB underwent subsequent photochemical reaction with excess sulfite in the medium, leading to formation of a nonchromophoric derivative of NTB. They suggested that the incubation procedure should be carried out in the dark. 2.3.3.1 SH and SS groups of flaxseed/linseed proteins: To date, no reported value for SH and SS groups of flaxseed/linseed proteins was found. 2.4 FUNCTIONAL PROPERTIES OF PROTEINS Functional properties were defined by Kinsella (1979) as "the intrinsic physicochemical characteristics which affect the behaviour of protein in food systems during processing, manufacturing, storage and preparation". They are controlled by the composition and structure of proteins, and the interactions of protein with one another and with other substances (Jones and Tung, 1983). According to Nakai and Powrie (1981), functional properties include: 1) sensory and kinaesthetic properties such as flavour, colour, odour and texture; 2) hydration, dispersibility, solubility and swelling; 3) surface-active properties such as emulsification, foaming, adsorption; 4) rheological properties such as gelation and texturization; 5) other properties such as adhesive, cohesive, dough making, film and fibre making. Functionality in a broad sense is any property of a protein, other than its nutritional value, that affects its utilization (Pomeranz, 1991). Several researchers have studied the possibility of using flaxseed meal in food products. Dev and Quensel (1986) found that both linseed flour and the protein isolate they prepared exhibited favourable water-, moisture-, oil-absorptions, emulsifying activity and emulsion 22 stability compared with soybean flour and isolate. The foaming capacity and stability of linseed flour were inferior to soybean flour while those of linseed isolates were better than soybean isolates. They also found that the emulsifying properties were pH-dependent and were adversely influenced by NaCl. The same group later investigated the functional properties of linseed protein flour and concentrate containing different levels of mucilage and found that high-mucilage protein concentrate had better water absorption and emulsifying properties, higher foaming capacity but lower nitrogen solubility, oil absorption and foam stability than the low-mucilage flour and concentrates. The high-mucilage protein also had higher viscosity (Dev and Quensel, 1988). Dev and Quensel (1989) continued to assess the functional properties of the low- and high-mucilage linseed products in food systems such as canned fish sauce, meat emulsion and ice cream. All linseed protein products were found to reduce cooking loss and firmness of meat emulsion but the low-mucilage products in general performed better as meat extender. The high-mucilage protein concentrate was an excellent emulsifier-stabilizer ingredient in fish sauce and to have great potential as an ice cream stabilizer. Madhusudhan and Singh (1985b) compared the functional properties of raw and water-boiled linseed meal with those of soybean meal and found that water boiling reduced nitrogen solubility of linseed meal in water, NaCl and sodium hexametaphosphate. Heat processing also diminished the foam capacity and stability, emulsification capacity and fat absorption capacity but increased water absorption capacity of the linseed meal. 23 Oomah and Mazza (1993b) studied the effect of methanol, aqueous methanol, methanol-ammonium hydroxide-water as primary extraction solvents on the physicochemical characteristics of commercially cold pressed NorMan flaxseed meal. All solvent extraction systems used in the study significantly reduced particle size, water hydration capacity and viscosity but improved the emulsifying capacity and foaming characteristics of the flaxseed meals. They suggested that the changes in the physicochemical characteristics of the meals were affected by the degree of polarity of the primary extraction solvent used. Wanasundara and Shahidi (1994b) investigated the functional properties and amino acid composition of flaxseed meal treated with alkanol-ammonia water. Their results were similar to those of Oomah and Mazza (1993b), in that solvent-extracted flaxseed meals had improved emulsifying and foaming properties. However, they found that water-adsorption was increased in the alkanol-ammonia water extracted flaxseed meal. The presence of hull polysaccharides was suggested to be an important factor affecting water adsorption (Wanasundara and Shahidi, 1994b). 2.4.1 Solubility Solubility of proteins relates to surface hydrophobic (protein-protein) and hydrophilic (protein-solvent) interactions with water (Vojdani, 1996). Solubility is a very important functional property of proteins because it has significant influence on other functional properties of proteins (Hailing, 1981). Vojdani (1996) noted that the proportion and distribution of surface hydrophilic and hydrophobic patches are the main factors in determining the degree of protein solubility, rather than the total hydrophobicity and charge density based on amino acid composition. Environmental factors such as pH, ionic 24 strength, temperature, solvent composition and the presence of other food components affect protein solubility. In addition, physical, chemical and thermal treatments during processing, method of isolation, interaction with other food components, and mechanical treatments prior to solubility testing will also influence the degree of protein solubility. 2.4.1.1 Solubility of flaxseed/linseed proteins: A few of the early research studies reported on the nitrogen extractability of flaxseed/linseed proteins. The minimum extractability of nitrogenous matter from oil-free linseed meal was observed to be at pH 3.8-5.1 by Smith et al (1946), and at pH 3.5-4.0 by Painter and Nesbitt (1946). Sosulski and Bakal (1969) also noted the wide pH range of minimum solubility in contrast to the distinct isoelectric point for soybean proteins. More recently, three research groups investigated the solubility characteristics of flaxseed/linseed proteins. Table 3 summarizes the conditions of solubility studies reported by the three recent groups. Madhusudhan and Singh (1983; 1985b) reported a broader pH range of least nitrogen solubility (pH 3.0-6.0) for demucilaged, defatted, and dehulled linseed meal. Dev and Quensel (1986) found that linseed flour had lower nitrogen solubility in water than soy flour in pH range of 6.0-7.0 and at pH below 3 but linseed isolate had better nitrogen solubility than soybean isolate throughout the whole pH range of 2.0-12.0. Dev et al. (1986a) also reported a broad pattern of nitrogen extractability for linseed flour at varying pH and ionic strength. Its region of minimum solubility of the linseed flour in water was pH 4.0-4.6. About 20 % of nitrogenous matter was extracted at the isoelectric pH region. 25 Table 3. Solubility studies on flaxseed/linseed meals reported in the literature. References Sample tested Sample concentration Solvent pH Incubation Centrifuge Protein conditions determination Dev and Quensel, 1986 linseed flour & isolate; soybean flour & isolate 0.2 g /20 ml d d H 2 0 2-12 30 min room temp 3000 x g 20 min % N i n solution; micro-Kjeldahl Dev et ai, 1986a defatted linseed flour 1 g / 20 ml dd H 2 0 , 0.1M, 0.3M, 0.6M, 1.0 M NaCl 2-12 30 min room temp 2500 x g 30 min % N i n solution; micro-Kjeldahl Dev and Quensel, 1988 linseed flours, protein concentrates & isolate at different mucilage levels 0.2 g /20 ml d d H 2 0 2-12 30 min room temp 3000 x g 20 min % N i n solution; micro-Kjeldahl Madhusudhan and Singh, 1983; 1985b. soybean meal, linseed meal, water-boiled linseed meal 2 g/20ml dd H 2 0 , 0.5 M , l .OMNaCl , 2 % S H M P •12 mechanically shaken lhr room temp 6000 rpm 30 min % N i n solution; micro-Kjeldahl Wanasundara and Shahidi, 1994a 5 laboratory prepared flaxseed meals & 2 commercial meal extracted, defatted with different solvent systems 5 g/250ml d d H 2 0 2-11.8 Burrel wrist-action shaker 2hr room temp 1500 x g 10 min % N i n solution; micro-Kjeldahl Wanasundara and Shahidi, 1997b flaxseed protein isolate (79 % protein), 5, 10,20 % acylated isolate & 5, 10, 20% succinylated isolate 0.25g/ 25ml dd H 2 0 , 0.35M, 0.70M NaCl 3-11 not available 20 000 g 15 min Lowry's, B S A std. Solubility index (supernatant / suspension) 26 Dev and Quensel (1988) later also found that the nitrogen solubility of linseed flour and concentrate with higher mucilage contents were lower than their low-mucilage counterparts at pH 6 and above but higher at pH below 4. Wanasundara and Shahidi (1994a) observed the minimum solubility of nitrogenous compounds in hexane-extracted and methanol-ammonia-water/hexane-extracted meals to occur between pH 3.0-3.5. Wanasundara and Shahidi (1997b) found that both acetylation and succinylation increased solubility of flaxseed protein isolates. 2.4.2 Foaming Foam is defined by German and Phillips (1994) as a two-phase colloidal system containing at least initially a continuous liquid phase and a gas phase dispersed as bubbles or air cells. Applications of foams could be found in the food industry, household and cosmetic products, construction sites, firefighting. There are two important phases in protein foaming: the effectiveness of gas encapsulation and the lifetime of the foam. To form foam efficiently, protein needs to adsorb rapidly during the transient stage of foam formation (Wilde and Clark, 1996). The adsorption process can be further described by three sub-processes, namely, transportation from bulk solution to the interface, penetration into the surface layer and reorganization of the protein structure in the adsorbed layer. Protein size, surface hydrophobicity and structural flexibility are the three main factors affecting foamability. Proteins are transported to the interface by diffusion, convection or a combination of the two processes (Wilde and Clark, 1996). Ffigh agitation through 27 whipping and shaking or direct gas introduction through sparging or injecting introduce small gas bubbles into the protein sample yielding a dispersion of gas bubbles entrapped by the protein solution. As more gas is introduced, entrapped bubbles are forced upward carrying with them the continuous aqueous phase (German and Phillips, 1994). Once the foam is formed, the rate of rupture of the bubbles is a function of the lamella thickness and interfacial viscoelasticity (German and Phillips, 1994). 2.4.2.1 Methods of foam generations: Wilde and Clark (1996) classified methods of foam generation into four categories: (1) Sparging, (2) Whipping, (3) Shaking, (4) Pouring. Sparging is the most commonly used method in basic studies on foams because it yields more uniform bubble sizes. It involves forcing gas through the liquid via an aperture or frit to create bubbles and hence a foam (Wilde and Clark, 1996). It requires a lower protein concentration than other methods. The typical range employed is 0.01 - 0.2 %, but up to 2 % has been used in some studies (Patel and Fry, 1987). The volume of foam produced is dependent on the total amount of protein in the solution being bubbled rather than the concentration of the protein (Hailing, 1981). Industrial application of sparging is limited but it is more widely used as a research tool (Wilde and Clark, 1996). The low protein concentration required for sparging is a great advantage for testing novel research grade proteins which are in short supply (Patel and Fry, 1987). Another advantage of the sparging method over other methods is the ability to control the amount of gas and energy input. Hailing (1981), however, commented that foams produced by sparging usually seem to have 28 much lower stability but this might be due to the much lower protein concentrations employed in the studies. Various conditions have been used in the sparging method. Waniska and Kinsella (1979) placed 15 ml of protein solution with concentration range of 0.01-1.00 % in a temperature-controlled glass condenser with a dimension of 85 cm x 1.1 cm and sparged with nitrogen at flow rates ranging from 4 to 36 ml/min. Townsend and Nakai (1983) placed 15 ml of 0.1 % protein solution in a 100 cm x 2 cm glass column fitted with a sintered glass disc with pore size of 40-60 p.m and sparged with air at a flow rate of 35 ml/min for 2 minute. Kato et al. (1983) placed 5 ml of 0.1% protein solution in a 2.4 cm x 30 cm glass column and sparged with air at a flow rate of 90 ml/min for 15 seconds. Balmaceda et al. (1984), from a technical centre of a commercial food corporation, developed a standard test for foaming in which they used a relatively large column with internal diameter of 45-48 mm and sparged nitrogen gas at 100 ml/min into 50 ml of 1 % protein solution for 2 minutes. Wright and Hernmant (1987) did not specify the dimension of the column but used two glass sinters with different porosities for different sample volumes and flow rates. They sparged 25 ml of 0.1 or 0.5 % protein solution with air at 60 cm3/min through a glass sinter (porosity 4, radius 15 mm), or 2 ml of solution with a flow rate of 2.4 cm3/min through an aperture of 15 u,m. Sparging time was not controlled but sparging was terminated when the foam reached a certain mark. Whipping is the most common method reported for producing foams and is still the most popular industrial method (Patel and Fry, 1987; Wilde and Clark, 1996). It 29 involves the high-speed movement of whisks or blades through a liquid, thus beating atmospheric air into the foam. Whipping generally produces the most stable and stiff foams because it progressively decreases the bubble size and increases the liquid fraction of the foam with increasing time. One disadvantage of whipping is that it requires a much higher concentration of protein (3-40 %) for foam production (Waniska and Kinsella, 1979). Various conditions and different types of food mixers are reported in literature but the types of mixer used have a profound effect on the volume and properties of the final foam. It is therefore very difficult to compare foaming properties of samples prepared by whipping method in different laboratories. Shaking is used as a simple research method for foam generation in a closed system (Wilde and Clark, 1996). It generally involves shaking a sealed jar or graduated cylinder mechanically or manually with frequencies varying from 1 to 50 Hz at a variety of amplitudes of agitation (Wilde and Clark, 1996). About 1 % protein is required for foam production (Waniska and Kinsella, 1979). Shaking is not considered a favourable method by industry and its applications are limited to specific studies such as the study of aerosol-generated foams (Wilde and Clark, 1996). Pouring has the most applications in the brewing industry. It is used by the American Society of Brewing Chemists as a standardized method for determination of beer foam stability (Anonymous, 1976). 30 2.4.2.2 Methods of protein foam measurement: There are many different methods for reporting foam measurement. Direct measurement of the increase in foam volume upon the introduction of a gas into protein solution has been the most commonly used method. It is usually expressed as a volume of foam obtained for a given input of work (eg. time of foaming and volume of gas used). The measurement of the decrease of foam volume or the increase of drainage over time is typically used as a measurement of foam stability. The times taken for half of the foam to collapse or half of the liquid to drain are commonly used. However, the degree of changes in the foam volume can vary greatly among proteins. Some protein foams disintegrate with a minimal change in foam volume while others disintegrate with a significant change in volume. German and Phillips (1994) considered the system of gas-liquid protein dispersion highly unstable. They emphasized that the disintegration of foam as characterized by the drainage of continuous liquid phase, and bubble breakage should be considered a part of the process of foam formation. Therefore, measurements of bubble size distributions and lamella dimensions are needed for complete description of foams because measurement of air incorporation and foam stability do not completely describe important differences between protein foams (German and Phillips, 1994). Conductivity of foam is proportional to its density (Clark, 1948). Conductivity also relates to the amount of liquid in the foam (Guillerme et al, 1993). Foam conductivity is thus an indirect method of measuring the foam density (the ability of a foam to take up and hold liquids) (Wilde and Clark, 1996). Kato et al (1983) 31 developed a conductivity measurement method for determination of the properties of protein foams generated by the sparging method. Foam consists of gas bubbles separated by liquid and solid films. By measuring the conductivity of foams which contain conductors, such as the fluid adsorbed to the film of foams, the foaming properties of a protein can be estimated. Kato et al. (1983) also found that the initial conductivity and the changes in the conductivity over time correlated well with the foaming capacity and the foam stability index of proteins, respectively. Monitoring bubble size distribution with time is another approach to understand the bubble rupture and the changes in bubbles. Methods such as photographing a set of bubbles over time and analyzing freeze-dried foams by scanning electron microscopy have been reported (Wilde and Clark, 1996). 2.4.2.3 Analysis of foaming data The simplest and most popular expression of foamability is the foam volume (Vf0) created during a defined period of foam generation. (Wilde and Clark, 1996). Foam expansion (FE%) is Vf0 expressed as a percentage of the initial sample volume (Vii). This parameter is employed for the ease of comparing of results between research groups because the initial volume between different groups varies (Wilde and Clark, 1996). FE% = Vf0 I Vii * 100% 32 The efficiency or foaming capacity (FC) was used for foam generated using the sparging technique (Wilde and Clark, 1996). The foam volume Vf0 is adjusted by the input volume of gas Vg. FC = Vf0/Vg. Foam conductivity when the foam has been created (Ci), has also been used as a measurement of foamability (Kato et al. 1983). On the other hand, since the initial conductivity of foam is dependent on the conductivity of the bulk liquid Q, a more representative expression is the relative foam conductivity C/ (Wilde and Clark, 1996). C/= Ci/Ci * 100% The Cf term provides an estimate of the relative density of the foam but is not a true representative value of the liquid content of the foam since the charge distribution through a foam lamella is not similar to that of the bulk solution as a result of interfacial adsorption. 2.4.2.4 Foaming properties of flaxseed proteins: Only three research groups have studied the foaming properties of flaxseed/linseed meal proteins. Table 4 outlines the different conditions used by these groups. 33 Table 4. Foaming properties studies on flaxseed/linseed protein products reported in the literature. Sample Foam concentration Foam measure- Foam Foam References Sample tested (g/lOOml) Solvent pH generation ment capacity stability Dev and Quensel, 1986 linseed flour & isolate; soybean flour & isolate 1.0 dd H 2 0 2-12 blender 2500 rpm 5 min 250 ml measuring cylinder % Foam Vol increase 0-120 min in foam or volume as half-life Madhusudhan soybean meal, and Singh, linseed meal, 1985b water-boiled linseed meal 2.0 0.10-1.0 NaCl (30 °C-90 °C) 1-12 Braun blender 1600 rpm 5 min 250 ml measuring cylinder read @ 30s % increase in foam volume Foam vol at 5, 10, 20, 30, 45, 60, 90, 120 min Wanasundara 5 laboratory and Shahidi, prepared flaxseed 1994a meals & 2 commercial meal extracted, defatted with different solvent systems 3.0 d d H 2 0 Polytron 250 ml % homogenizer measuring increase 10 000 rpm/min lmin cylinder in foam volume Foam vol at 5, 10, 20,40, 60, 120 min Wanasundara flaxseed protein and Shahidi, isolate 1997b (79 % protein), 5, 10, 20 % acylated isolate & 5, 10,20% succinylated isolate 1.0 dd H 2 0 Polytron homo-genizer 10 000 rpm/min for lmin 250 ml measuring cylinder increase in foam volume 3% solution foam vol after 0.5, 5, 10 15 & 20 min 34 Dev and Quensel (1986) compared the foaming properties of linseed flour and linseed isolate to those of soybean flour and soybean isolate. Although the foaming capacity and stability of linseed flour were inferior to soybean flour, the foaming capacity and stability of linseed isolates were better than soybean isolates. A pronounced dependence of foaming capacity on the proportion of soluble protein was observed. Foaming capacity was lowest at the isoelectric pH and the maximum foaming capacity for linseed isolate was observed at strongly acidic and alkaline pH levels. Foam stability of linseed isolate increased with decreasing value of pH. Incorporation of NaCl caused an enormous increase in the foam stability of linseed isolate, probably due to the fact that NaCl reduced leakage and improved foam stability (Dev and Quensel, 1986). Madhusudhan and Singh (1985b) compared the foaming properties of linseed meal with those of water-boiled linseed meal and soybean meal. Minimum foam capacity was observed at the region of minimal nitrogen solubility (pH 2.0-6.0). Soybean meal had better foam capacity than linseed meal at all pHs tested. The linseed meal had better foaming properties after mild heat treatment which resulted in surface denaturation of the protein to expose the hydrophobic regions of the protein while maintaining it in solution. In addition, they studied the effect of NaCl concentration on the foam capacity of the linseed and soybean meals. An increase in foam capacity was observed for linseed meal from 0.0-0.2 M NaCl and a gradual decrease from 0.2-1.0 M. Soybean meal had higher foam capacity than linseed meal at all NaCl concentrations. The increase in foam capacity up to 0.2 M NaCl and the decrease in foam capacity at higher salt concentration were attributed respectively to salting-in 35 and salting-out of proteins. In terms of foam stability, the linseed meal foam was more stable at acidic and neutral pHs compared to the soybean and heat-treated linseed meal. Addition of NaCl reduced the foam stability probably due to charge repulsions (Madhusudhan and Singh, 1985b). Wanasundara and Shahidi (1994a) investigated the foaming properties of seven laboratory prepared linseed meals defatted with seven different solvent systems and two commercial meals defatted with two solvent systems. Foam produced by all solvent-extracted samples had fine bubble structure and they suggested that could be a unique characteristic of linseed meal proteins. The presence of non-protein nitrogen compounds and carbohydrates helped to stabilize the foams. They later found that acylation of flaxseed protein isolate reduced both its foaming capacity and stability, probably due to the increase in negative charge caused by acylation which hindered protein-protein interactions for film formation around air bubbles (Wanasundara and Shahidi 1997b). 2.4.3 Viscosity Viscosity was defined by Isaac Newton as "the resistance which arises from the lack of slipperiness of the parts of the liquid" (Barnes et al, 1989). Viscosity is synonymous with "internal fraction" and is a measure of "resistance to flow" (Barnes et al, 1989). Quantitatively, viscosity is defined as "the shear stress divided by the rate of shear in stead simple-shear flow" (Barnes et al, 1989). Liquids with viscosity independent of the shear rate applied are termed Newtonian liquids. Most liquids, however, are non-Newtonian, in 36 which their viscosities are shear rate dependent. The terms "apparent viscosity" or "shear-dependent viscosity" are commonly used for non-Newtonian liquids. Viscosity of protein solutions is affected by the concentrations of the solution and inherent physicochemical properties such as molecular weight, polydispersity, hydrophobicity and conformation of each protein species (Schenz and Morr, 1996). The concentration and physicochemical properties of other ionic and non-ionic solutes also exert an important influence on the viscosity of protein solutions by contributing directly to viscosity and by their tendency to interact and modify the physicochemical properties of the proteins. Rha and Pradipasena (1986) noted that the viscosity of proteins in dilute solutions and in the absence of interactions is governed by the shape and size of the molecules. When the molecules are completely isolated, the concentration of the protein approaches zero and the term "intrinsic viscosity" is used to define the viscosity of proteins (Rha and Pradipasena, 1986). Intrinsic viscosity was defined by Rha and Pradipasena (1986) as "the limit of specific viscosity divided by concentration or reduced viscosity when the concentration of protein approaches zero". Intrinsic viscosity is therefore an indication of the hydrodynamic shape and size of protein in solution (Rha and Pradipasena, 1986). 2.4.2.1 Viscosity of flaxseed/linseed proteins: Most viscosity studies on flaxseed/linseed focused on the viscosity contributed by the mucilage. A study by Oomah and Mazza (1993b) investigated the viscosity of the flaxseed, which contained lipid, carbohydrate, and other minor components in addition to proteins. Oomah and Mazza (1993b) measured the apparent viscosity of 37 solvent extracted flaxseed meals and found that flaxseed meal exhibited non-Newtonian shear-thinning properties. They concluded that this shear thinning behaviour was affected by protein concentration, NaCl concentration and solvent extraction. The same group later investigated the effect of commercial processing (cleaning, flaking, pressing and solvent extraction) on viscosities of the processed products and observed that viscosities were not significantly affected but shear-thinning indices increased with processing (Oomah and Mazza, 1998). Madhusudhan and Singh (1985c) was the only group that measured viscosity of the isolated protein of flaxseed. The reported intrinsic viscosity for the 12S linseed protein in phosphate buffer was 3.1 ml/g. 38 CHAPTER 3—MATERIALS AND METHODS FOR PHASE 1; ISOLATION OF THE MAJOR FRACTION OF FLAXSEED PROTEINS 3.1 MATERIALS 3.1.1 Chemicals All Tris buffer was made from Ultra-pure grade Tris buffer salt (ICN Biochemical, Inc. Costa Mesa, CA). 0.02% sodium azide was added to all buffer solutions as an antimicrobial agent. All other chemicals were of analytical reagent grade. 3.1.2 Proteins Bovine serum albumin (BSA) A-4503, ovalbumin A-5503 and pMactoglobulin L-6879 were purchased from Sigma (St. Louis, MO). Bovine IgG 55917 was purchased from ICN Biomedicals, Inc. (Costa Mesa, CA). 3.1.3 Flaxseeds Mechanically dehulled flaxseed (NorMan cultivar) was donated by the Pacific Agri-Food Research Centre (PARC) in Summerland, BC. The flaxseed, obtained from the Agriculture and Agri-Food Diversification Research Centre, Morden, Manitoba, was dehulled at PARC by mechanically grinding the seeds through a Strong Scott barley pearler fitted with a 2 mm screen and fitting the ground seeds into an air separator (The Cuthbert Co. Ltd, Winnipeg, MB) to separate the hulls from the seeds. 39 3.2 ANALYSES 3.2.1 Protein determination Protein contents of the solid samples were determined by the Leco nitrogen combustion method in the Faculty of Agricultural Sciences, UBC (LECO FP-428, LECO Cooperation, Joseph, MI). Protein contents of liquid samples were initially analyzed by the Bio-Rad Protein Assay based on the Bradford assay (Bio-Rad Laboratories, Hercules, CA). Bicinchoninic acid (BCA) Protein Assay (Pierce, Rockford, IL) was later used in place of the Bio-Rad Protein Assay because it was not reliable in estimating the protein content of lower molecular weight proteins. BSA was used as protein standards. 3.2.2 Amino acid analyses Amino acid compositions were determined at the Nucleic Acid Protein Service (NAPS) of the Biotechnology Laboratory of UBC (Vancouver, BC) in which an ABI 420A Amino Acid Hydrolyzer/Derivatizer was used to perform amino acid analysis. Samples were spotted onto a fritted slide where they were hydrolyzed into free amino acids and derivatized into phenythiocarbamyl (PTC) amino acids which were then separated on the ABI 130A separation system via HPLC. 3.2.3 Gel electrophoresis Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) was carried out with PhastGel® gradient 10-15 and 8-25, according to the PhastSystem™ Separation Technique (File No. 110—SDS-PAGE; Amersham Pharmacia Biotech Inc., Quebec). Molecular weight markers (Sigma Marker™ Wide Molecular Weight Range, M-4038; 40 Sigma, St. Louis, MO) were used. BSA, ovalbumin, f3-lactoglobulin and IgG were also prepared as alternative molecular weight standards after the Sigma Marker™ expired. SDS-PAGE on high density gel was carried out according to the PhastSystem™ Separation Technique (File No. 112—SDS-PAGE of low molecular weight protein using PhastGel® high density). BSA, ovalbumin, P-lactoglobulin and IgG, as well as the Pharmacia Peptide Molecular Weight Marker (PMW Marker), MW Range 2512 - 16949 were used as molecular weight standards. Band corresponded to MW of 2512 Da from the PWM Marker was observed. No satisfactory explanation or suggestion for the missing band was found after consultation with the technical support department of the manufacturer. Non-denaturing polyacrylamide gel electrophoresis (Native PAGE) was carried out according to the PhastSystem™ Separation Technique (File No. 120—Native-PAGE with PhastGel® gradient 8-25). All 10-15 and 8-25 gradient PhastGels® were Coomassie stained according to the PhastSystem™ Development Technique (File No. 200—Fast Coomassie staining). High density gels were Coomassie stained according to the PhastSystem™ Development Technique (File No. 201—Coomassie staining of peptides using PhastGel® high density). Ferguson plot for molecular weight estimation were prepared by plotting the logarithms of the protein standard molecular weights against their corresponding relative mobility. The relative mobility of a protein standard was calculated as the distance of migration of the 41 protein band in the resolving gel divided by the distance of migration of the tracking dye. Molecular weights of the samples were determined from their relative mobility using the Ferguson plot (Smith, 1994). 3.3 EXTRACTION OF FLAXSEED PROTEINS Flaxseed proteins were extracted from defatted flaxseed meal by modification of the procedures previously established for the isolation and characterization of cadmium-binding proteins from flaxseed (Sultanbawa, 1998; Li-Chan et al. 1999). A series of preliminary experiments was carried out to determine the most suitable conditions for extraction of flaxseed proteins. The first two experiments were carried out to investigate the effect of 2-mercaptoethanol on the extraction yield of flaxseed protein. The third experiment was performed to explore the feasibility of speeding up the isolation procedure. Figure 3 summarizes the protocol used in the preliminary experiments. 42 Dehulling NorMan flaxseed Grinding Wiley Mill #10 (2mm) sieve I Delipidation, Air drying 1:5:5 (w/v/v) seed: methanol: chloroform 1 hr x 3, magnetically stirred; filter though Whatman® # 2 filter paper; air dry seed in fume hood over night \ Extraction 1:16 (w/v) seed : 0.10 M NaCl, 0.10 M Tris @ pH 8.6; 16 hr @ 4°C, magnetically stirred I Filtration Cheesecloth Residue Supernatant I Centrifugation 20 400 x g, 30 min @ 8-10 °C Pellet A-280 (nm) Conductivity (mS/cm)/' i • Supernatant—protein extract I Dialysis 25mM Tris @ pH 8.6 I Fractionation Anion-exchange Chromatography (2.5 x 20 cm column: 50 ml DEAE); Equilibrated with 25 mM Tris @ pH 8.6; step gradient elution I A <£. 1 s t experiment < 2 n d experiment 3r experiment 40M Figure 3. Procedures for the isolation of flaxseed proteins in preliminary experiments. 43 Dehulled NorMan flaxseeds obtained from the PARC Summerland research centre were used directly in the first screening experiment while the seeds were further manually sorted until only about 10 % of hulls remained (by observation) in the second experiment. The seeds were then ground through a #10 (2mm) sieve with a Wiley Mill (Arthur, H. Thomas Co. Scientific Apparatus, Philadelphia, P.A), followed by delipidation using methanol-chloroform (Sultanbawa, 1998). A 1:10 seed to solvent ratio (w/v) was used and the proportion of methanol to chloroform was 1:1 (v/v). Seeds were delipidated using three one-hour solvent extractions and were magnetically stirred during each extraction. Delipidated seeds were recovered by filtering through a Whatman® filter paper number two (Whatman International Ltd, Maidstone, UK) and air-dried overnight in the fume hood. Extraction of proteins from dehulled delipidated ground seeds were carried out at 4 °C for 16 hours with 0.10 M NaCl in 0.10 M Tris buffer at pH 8.6 as extraction buffer and with 1:16 (w/v) seeds to buffer ratio, as described by Sultanbawa (1998). The seeds were magnetically stirred during extraction. After the 16-hour extraction, the extract was passed through a double layer of cheesecloth to retain all the large particles. The protein extract was then centrifuged at 20 400 x g for 30 minutes at 8-10 °C as described by Sultanbawa (1998). A single-step centrifugation was used as opposed to the two-step centrifugation used for the cadmium-binding protein project (Sultanbawa, 1988; Li-Chan et al, 1999) due to the lower amount of sample handled in this project. 44 3.3.1 Studies on seed The effect of dehulling and delipidation on the moisture and protein contents of the flaxseeds were investigated. 3.3.1.1 Moisture and total solid determination: Moisture and total solid contents of the whole dehulled seeds and the dehulled delipidated ground seed were determined by AO AC method (925.09 Total solids and Moisture in Flour Vacuum Oven Method; AO AC, 1990). One batch of the seeds analyzed was manually sorted so that there were only approximately 10 % hulls remaining (defined as 90 % dehulled); the other batch was manually sorted until no observable hulls remained (defined as 100 % dehulled). 3.3.3.2 Protein determination: Protein contents of the 90 % & 100 % dehulled seeds and dehulled delipidated ground seed were determined by the Leco nitrogen combustion method (UECO FP-428, LECO Cooperation, Joseph, MI). The significance of the difference in moisture and protein contents of the flaxseeds was determined by one-way Analysis of Variance (ANOVA) followed by Tukey's pairwise comparison test with a family error rate of 5 (p = 0.05) using MINITAB™ version 13 (Minitab, Inc., State College, PA). 45 3.3.2 First and second screening experiments The first two experiments were carried out on small-scale because of the limited amount of dehulled flaxseeds available. The 2nd experiment was carried out because the findings of the 1st experiment were inconclusive. The effect of addition of 10 mM 2-mercaptoethanol during extraction on the yield and overall characteristics of the proteins was investigated using Bio-Rad Protein Assay and SDS-PAGE on PhastGel® gradient 10-15 respectively. The difference in the protein concentration in the extract and the overall yield of extract was compared by one-way ANOVA followed by Tukey's pairwise comparison test with a family error rate of 5 (p = 0.05) using MINITAB™ version 13. Feasibility of small-scale extraction and effectiveness of glass wool filtration in clarifying the extract were also investigated. Table 5 summarizes the treatments of the flaxseeds for extraction during the first two screening experiments. Table 5. Additional treatments of the flaxseed and flaxseed extracts investigated in the first two preliminary experiments. Experiments Batches Weight of seeds (g) Manual sorting of hulls 2-Mercaptoethanol during extraction (10 mM) Glasswool filtration after centrifugation 1 1 1.0 - - + 2 1.0 - + + 3 1.0 - - + 4 1.0 - + + 2 1 0.80 + - -2 0.80 + + -3 0.26 + -4 0.26 + + -46 3.3.3 Third Screening Experiment 22.02 g of dehulled NorMan seed was delipidated as described above. 8.91 g dehulled delipidated ground seed was used for protein extraction. Glasswool filtration of the supernatant collected upon centrifugation of protein extract was not employed. Mira-cloth was used instead to remove any large size insoluble materials. 106 ml of crude protein extract was prepared for further fractionation (to be described in Section 3.4.2). 3.3.4 Studies on extraction yield 3.3.4.1 First extraction study: The objective of the first extraction study was to determine whether the extraction process was extracting a representative portion of the proteins. 16 ml of extraction buffer (0.10 M NaCl in 0.10 M Tris at pH 8.6) was added to 0.93 g of the residues from a previous protein extraction, and the re-extraction was carried out for 16 hours under the same conditions outlined in Figure 3. A non-reducing SDS-PAGE was carried out using PhastGel® gradient 8-25 on the extracts from previous extractions, residues and the extracts obtained from re-extraction to compare the PAGE profiles of the three. 3.3.4.2 Second extraction study: The main objectives of the second extraction study were to investigate whether the long 16-hour extraction was necessary to extract the proteins from flaxseed and to explore whether repeated extractions would increase the total extraction yield. A two-step centrifugation as described by Sultanbawa (1998) was also used in place of the one-step centrifugation. 47 One set of sample was extracted for 16 hours as described in the original protocol (Figure 3) while the other set was extracted only for 1 hour. Residues retained by the cheesecloth were then subjected to another 2-hour extraction (thus termed the "2nd extract"). Residues retained by the cheesecloth were further extracted for another 2 hours (thus termed the "3rd extract"). The term "1-hr" and "16-hr" applied only to the first extraction. All subsequent extractions were carried out for 2 hours. Reducing SDS-PAGE was carried out on PhastGel® gradient 8-25 to investigate the molecular compositions of the three consecutive extracts from 16- and 1-hr extractions and the residues after extractions. 3.4 FRACTIONATION OF FLAXSEED PROTEINS Protein extracts were fractionated by anion-exchange chromatography. The extracts from the 1st and 2nd screening experiments were fractionated to explore the elution profile of flaxseed proteins and to select the fraction which constituted the major peak. The extract from the 3rd screening experiment was fractionated to determine the feasibility of scaling up and speeding up the fractionation procedure. The final protocol established was used for repeated isolation of the major fraction of interest. 3.4.1 First and second screening experiments The protein extracts obtained from batch 3 of the 1st screening experiment and batch 1 of the 2n d screening experiment were dialyzed against 25mM Tris buffer at pH 8.6 using Spectra/Por® Molecularporous dialysis membrane (Spectrum®, Laguna Hills, CA) with molecular weight cut off of 6-8 kDa. The dialyzed protein extracts were loaded onto DEAE-Sephacel (Amersham Pharmacia Biotech Inc., Quebec) packed in a Bio-Rad 2.5 48 cm x 20 cm column, equilibrated with 25 mM Tris buffer at pH 8.6. The same buffer was used to wash the column after sample loading and the fraction collected was termed "Flow through" or simply "FT". A step gradient of 0.10-0.30 M NaCl (0.10 M, 0.15 M, 0.20 M, 0.25 M, 0.30 M) in 0.10 M Tris buffer, pH 8.6, was used to elute the samples in the 1st screening experiment. A step gradient of 0.10-0.50 M NaCl (0.10 M, 0.15 M, 0.20 M, 0.25 M, 0.30 M, 0.45 M, 0.50 M) in 0.10 M Tris buffer, pH 8.6, was used to elute the samples in the 2n d screening experiment. The "0.45 M" and "0.50 M" fractions were collected for the interest of the cadmium-binding project (Li-Chan et al., 1999). An Isco Cygnet™ fraction collector (series 2170-001, Isco, Inc. Lincoln, NB) was used to collect 2 ml of samples in each tube. Absorbance at 280 nm measured by a Unicam UV/Vis Spectrophotometer UV2 (Unicam Ltd, Analytical Technology Inc, Cambridge, UK) and conductivity measured by Yellow Springs Instrument Conductivity Bridge (YSI Model 31, Yellow Springs, Ohio) were used to monitor the progress of elution. Samples eluted from the same buffer with their respective ionic strength were pooled together. Sample eluted from 0.10 M NaCl in 0.10 M Tris at pH 8.6 was thus termed "0.10 M fraction"...etc. All fractions were freeze-dried (LABCONCO® model 75018, Labconco Corporation, Kansas City, MO), reconstituted with distilled-deionized water, and dialyzed against 25. mM Tris buffer at pH 8.6 using Spectra/Por® Molecularporous dialysis membrane described above. Dialyzed fractions and the original protein extract were loaded onto PhastGel® gradient 10-15 for SDS-PAGE under both reducing and non-reducing conditions. Native PAGE of all fractions and the original protein extract was also carried out, using a PhastGel® 49 gradient 8-25. Protein recovery from fractionation of the 1st screening experiment was determined by the BCA Protein Assay. Protein recovery from fractionation of the 2nd screening experiment was determined by the Bio-Rad Protein Assay. Amino acid analysis was carried out on the "FT", "0.10 M", "0.20 M" and "0.25 M" fractions of the 1st screening experiment. 3.4.2 Third screening experiment Fractionation procedures were simplified based on the observations made from the first two screening experiments. Protein extract was loaded undialyzed onto the same 2.5 cm x 20 cm column described above. The column was equilibrated with the extraction buffer (0.10 M NaCl, 0.10 M Tris, pH 8.6). The "FT" collected thus contained both the FT and 0.10 M fractions collected in the first two screening experiments. The two fractions were combined to speed up the elution process. The major fraction was obtained by replacing the three-step elution with 0.15 M, 0.20 M and 0.25 M NaCl in 0.10 M Tris at pH 8.6 with a single elution with 0.25 M NaCl in 0.10 M Tris at pH 8.6. Combined elution of the three steps was carried out because the proteins in the three fractions have similar compositions observed under SDS-PAGE and similar amino acid compositions. Madhusudhan and Singh (1985a) also reported a major peak eluting at 0.24 M NaCl. Samples were eluted with a step gradient of 0.25, 0.40, 0.45 and 0.50 M NaCl. The later three steps of the gradient elution were carried out for the interest of the cadmium-binding project (Li-Chan etal, 1999). 10 ml of sample was collected in each tube using the fraction collector described above. Progress of elution was monitored as described above except absorbance at 254 nm was 50 also monitored for the interest of the cadmium-binding project (Li-Chan et al. 1999). Samples of the same ionic strength were pooled, freeze-dried, reconstituted, dialyzed and analyzed as described above. High density SDS-PAGE under both reducing and non-reducing conditions were carried out on the "combined FT" and "combined 0.25 M" fractions collected in the 3rd experiment. The names "combined FT" and "combined 0.25 M" fractions were used to distinguish these fractions from the fractions eluted separately from 25 mM Tris and 0.10 M NaCl, and 0.15, 0.20 and 0.25 M NaCl respectively. The FT, 0.10 and 0.20 M fractions from the 1st experiment were used as comparison. 3.4.3 Scale-up fractionation of flaxseed proteins As the isolation using ion-exchange chromatography was time consuming and the capacity of the original column used was too small to obtain workable amount of protein per run, a larger column and fewer steps for the gradient elution were used. Figure 4 shows the finalized extraction and fractionation procedure for the isolation of the major fraction of flaxseed proteins. A larger Bio-Rad 5 cm x 20 cm column was packed with 225 mL DEAE-Sephacel. Protein extract was prepared as described previously. Protein extract was loaded directly onto the column equilibrated with 0.10 M NaCl in 0.10 M Tris at pH 8.6 as described in the 3rd experiment. The ultraviolet absorption spectra of the fractions collected were recorded using a Unicam UV/Vis Spectrophotometer UV2. Besides preparing protein extracts from dehulled delipidated seed as shown in Figure 4, an "overloaded 0.25 M fraction" from a concurrent project in our laboratory on isolation of 51 cadmium-binding flaxseed proteins (Li-Chan et al, 1999) was also used as the starting material for large scale isolation of the major fraction. For the latter, 250 ml of flaxseed protein extracts was prepared from 17.74 grams of dehulled delipidated seed as described above. The extract was adjusted to 0.25 M NaCl in 0.10 M Tris at pH 8.6 and loaded onto a 50 ml DEAE-Sephacel column equilibrated with 0.25 M NaCl in 0.10 M Tris at pH 8.6. The total amount of material loaded exceeded the capacity of the 50 ml DEAE-Sephacel and the fraction thus collected was all the unbound material (a combination of all fractions that eluted before 0.25 M NaCl, 19-22 mS/cm, and the whole extract which did not bind to DEAE because of the limited sites for binding). Figure 5 shows the DEAE-Sephacel profile performed under the "overloaded" conditions obtained from Lei (1999). The fraction was therefore termed "overloaded" because the column was purposely overloaded to obtain sufficient amount of the later fractions (Sultanbawa, 1998). The "overloaded 0.25 M fraction" obtained in a frozen form was thawed, dialyzed to reduce conductivity back to 10-11 mS/cm and applied to the DEAE-Sephacel column as described for the freshly prepared whole extract (Figure 4). The overloaded fraction from the cadmium-binding flaxseed protein project was used because of the limited supply of NorMan flaxseeds from the same batch. A total of six fractionations were carried out. The major fraction eluted at 0.25 M NaCl in 0.10 M Tris at pH 8.6 from each set was desalted by dialyzing against 5 mM Tris at pH 8.6 and was kept at -35 °C until used. A total of 147 ml of the "0.25 M fraction" was collected for further characterization in phase 2. 52 Dehulling NorMan flaxseed I Grinding Wiley Mill #10 (2mm) sieve I Delipidation, Air drying 1:5:5 (w/v/v) seed: methanol: chloroform 1 hr x 3, magnetically stirred; filter though Whatman® # 2 filter paper; air dry seed in fume hood over I Extraction 1:16 (w/v) seed : 0.10 M NaCl in 0.10 M Tris @ pH 8.6; 16 hr @ 4°C; magnetically stirred Filtration Cheesecloth Residue Pellet Pellet Supernatant Centrifugation 10 400 x g ,30 min @ 8-10 °C Supernatant Centrifugation 20 400 x g , 30 min @ 8-10 °C Overloaded 0.25 M fraction 1 i Dialysis <^Supernatant--protein extracP> °- 1 M N a C 1 i n 0.1 M Tris @ pH8.6 Filtration <• Miracloth A-280 (nm) Conductivity (mS/cm)/' Fractionation — — — — — — — — ——— — — —I • ! / ' Anion-exchange Chromatography (5.0 x 20 cm column: 255 ml DEAE); Equilibrated with 0.1 M NaCl in 0.1 M Tris @ pH 8.6; step gradient elution C^FJfO.lOMp 0.50 M Figure 4. Procedures for the isolation of the major fraction of flaxseed proteins. 1 Overloaded 0.25 M fraction was defined and described in the text. 0.25 M 0.50 M 0 200 400 600 800 1000 1200 1400 1600 Elution volume (ml) Figure 5. An "overloaded" DEAE-Sephacel chromatographic pattern of NorMan flaxseed proteins1. Profile obtained from Lei (1999), used with permission. 1250 ml flaxseed protein extract was prepared from 17.7 g dehulled delipidated NorMan flaxseeds and loaded onto a 50 ml DEAE-Sephacel column equilibrated with 0.25 M NaCl in 0.10 M Tris at pH 8.6. 54 CHAPTER 4—MATERIAL AND METHODS FOR PHASE 2; CHARACTERIZATION OF THE MAJOR FRACTION OF FLAXSEED PROTEINS 4.1 PREPARATION OF SAMPLES FOR CHARACTERIZATION The "0.25 M fractions" collected from the 6 scale-up batches were thawed. To check for uniformity of the samples from the 6 batches, Native PAGE and SDS-PAGE were carried out on aliquots from 6 individual batches and also a 1:1:1:1:1:1 (by volume) mixture of the 6 batches. Procedures for running native- and SDS-PAGE were described in Section 3.2.3. PhastGel® gradient 8-25 was used in both cases. The relative intensities of the bands in each lane in the Native PAGE were determined by the Phastlmage™ Gel Analyser and the Phastlmage™ software complemented with the PhastSystem™ (Amersham Pharmacia Biotech Inc., Quebec). The "0.25 M fractions" from the 6 batches were then pooled together and referred to as the "major fraction" throughout this thesis. The protein concentration of the pooled fraction was determined to be 8.7 mg/ml by the BCA Protein Assay (described in Section 3.2.2). This major fraction was used as the stock solution for all subsequent characterization tests and it was divided into smaller portions in 15 or 45 ml disposable centrifuge tubes. The smaller portions were kept at -35 °C until used. The "FT fraction" collected from the first scale-up batch was also thawed and made into smaller portions for use as stock solution for subsequent characterization tests which required 55 the "FT fraction" for comparison purposes. Its protein concentration was determined to be 1.9 mg/ml by the BCA Protein Assay. Protein extracts obtained by extraction as described in Section 3.3 at a concentration 7.7 mg/ml (determined by the BCA Protein Assay) was used as stock solution for subsequent characterization tests for comparison purposes. The protein extract was referred to as the "whole extract" throughout this thesis. 4.2 SOLUBILITY CHARACTERISTICS Solubility of the major fraction, whole extract and FT fraction were evaluated over the pH range of 2.0-8.5 in buffers of low (0.01 M NaCl) and high (1.0 M NaCl) ionic strength. Conditions used were based on the methods and conditions most commonly used for solubility tests as reviewed by Vojdani (1996). 4.2.1 Preparation of buffers Citrate-phosphate buffers at pH 3.0, 4.0, 5.0, 6.0 and 7.0 were prepared according to Mohan (1997). The concentration (in M) of citrate: phosphate of the five buffers were 0.040: 0.020, 0.031: 0.038, 0.024: 0.051,0.018: 0.063, and 0.007: 0.087 respectively. Two sets of buffers were prepared, one with low (0.01 M NaCl) and the other with high salt (1.0 M NaCl). Citrate-phosphate buffer at pH 2.0 and 2.5 were prepared by the addition of 1.0 N HC1 to the pH 2.5 buffer. Citrate-phosphate buffer at pH 8.5 was prepared by the addition of 1.0 N NaOH to the pH 7.0 buffer. Citrate-phosphate buffer at pH 3.5, 4.5, 5.5 and 6.5 were respectively prepared by mixing the buffers at pH 3.0 and 4.0, pH 4.0 and 5.0, 5.0 and 6.0, 6.0 and 7.0 in one to one ratio and adjusted pH to yield the final pHs. 56 4.2.2 Preparation of samples Protein stock solutions frozen in 5 mM Tris at pH 8.6 were thawed and diluted to 1.0 mg/ml protein by citrate-phosphate buffer solution at different pHs in the range of 2.0-8.5. Samples were allowed to equilibrate in a water bath held at 24°C for 1 hour. pH of the samples were checked again. Instead of replicating analysis at the same pH, solubility profiles of proteins were determined in two sets of pHs: 3.0, 4.0, 5.0, 6.0, 7.0 and 3.5, 4.5, 5.5, 6.5, 8.5. Solubility of the proteins in buffers of high ionic strength at lower pHs of 2.0 and 2.5 were also investigated because of the reported minimal solubility shifting and broadening effect in the presence of NaCl (Madhusudhan and Singh, 1983; Dev et al., 1986a). 4.2.3 Determination of solubility Samples were centrifuged at 10 000 x g for 30 min. Absorbance at 280 nm of the samples before and after centrifugation was measured with a Unicam UV/Vis Spectrophotometer UV2 and plotted against pH. Protein contents of the samples before centrifugation and the supernatants after centrifugation were also determined using the BCA Protein Assay. Solubility of protein was expressed as the percent ratio of the protein content of the supernatant to that of the sample before centrifugation. 4.3 ISOELECTRIC POINTS Isoelectric points of the major fraction were determined by isoelectric focusing following the procedure outlined by the PhastSystem™ Separation Technique (File No. 100—IEF and electrophoretic titration curve analysis). PhastGel® pH 3 -9 was used. Approximately 2 fig of sample was loaded on each lane. Samples were applied in the middle position of the gel as 57 described in the PhastSystem™ user manual. Duplicate gels were prepared. The Sigma IEF-Mix 3.6 - 9.3, 1-3018 (Sigma, St. Louis, MO) was used as marker. The standard curve was prepared by plotting the pis of the standards against the relative mobility of each standard protein. Calculation of relative mobility was described in Section 3.2.3. The pis of the samples were calculated from the standard curve. 4.4 PROTEIN SURFACE HYDROPHOBICITY PRODAN (6-propionyl-2-(dimethylamino) naphthalene) (Molecular Probes, Inc., Eugene, OR) was used according to Alizadeh-Pasdar and Li-Chan (2000) and Cheng (2001) for measuring the surface hydrophobicity of the major fraction and whole extract at pH 3.0, 5.0, 7.0 in buffers of low (0.01 M NaCl) and high (1.0 M NaCl) ionic strength. 4.4.1 Preparation of PRODAN stock solution The PRODAN stock solution was prepared by dissolving approximately 3.2 mg of PRODAN in 10 ml of methanol. The concentration of the PRODAN stock solution was later determined to be 1.55 x 10"3 M by measuring A 3 6 i and using e =1.8 x 104 M^cm"1 (Haugland, 1996). The PRODAN stock solutions were stored in smaller aliquots in small micro-centrifuge vials wrapped with aluminium foil, in a -18 °C freezer. Because of the light sensitive nature of the fluorescent probe, all processes for the preparation of the probe were done in the dark (Cheng, 2001). On the day of the experiment, a vial of PRODAN stock solution was removed from the freezer and placed in an ice bath to minimize methanol evaporation during use. 58 4.4.2 Measurement of relative fluorescence intensities (RFI) of samples Protein stock solutions were diluted in citric-phosphate buffers of pH 3.0, 5.0 or 7.0 and NaCl concentrations of 0.01 M or 1.0 M, each yielding five final concentrations from 0.015 to 0.15 mg/ml protein. The dilution series were prepared in duplicate. 5 ul of PROD AN solution was then added to 2 ml of each diluted sample and mixed well by vortexing. After 15 minutes incubation at ambient temperature in the dark, relative fluorescence intensities (RFI) of the samples were measured with a Shimadzu RF-540 (Shimadzu Corporation, Kyoto, Japan) spectrofluorophotometer using a 1-cm path length quartz cuvette (Starna Ltd., Romford, Essex). Excitation and emission wavelengths of 365 nm and 465 nm were used respectively. RFI values of the buffers and diluted samples without PRODAN were referred to as "blanks" (Alizadeh-Pasdar and Li-Chan, 2000). The net RFI was obtained by subtracting the RFI of each "blank" at a particular dilution from that of the corresponding protein solution with PRODAN. In order to minimize the effect of day-to-day instrumental fluctuations in the RFI values, standardization was performed by measuring the RFI of 2 ml of methanol with 5 ul of PRODAN and corrected to a standard value of 50 (Alizadeh-Pasdar and Li-Chan, 2000). The net RFI values of each sample were standardized using the ratio of the standard value of 50 and the measured RFI value of PRODAN in methanol. 4.4.3 Expression of protein surface hydrophobicity and statistical analyses Protein surface hydrophobicity of the samples was calculated by linear regression analysis using Microsoft Excel (Microsoft Corporation, Roselle, IL) and was expressed as the 59 initial slope (S0) of the plot of net RFI versus the protein concentration (%) as described by Alizadeh-Pasdar and Li-Chan (2000). Statistical analyses were conducted using MINITAB™ version 13. Analysis of variance was conducted on the S0 values of the whole extract and the major fraction as a function of pH and salt concentration. Differences between samples were compared using one-way ANOVA followed by Tukey's pairwise comparison test with a family error rate of 5 (p = 0.05). 4.5 AMINO ACID COMPOSITION Amino acid compositions of the major fraction and the whole extract were determined at the UVic Protein Chemistry Centre (University of Victoria, Victoria, BC). Samples were hydrolysed and applied to an Applied Biosystems Model 420 derivatizer/analyser system. Analyses of PTC-residues were done by RP-HPLC and amino acid compositions were expressed in picomoles and mole %. 4.6 CALCULATION OF BIGELOW'S TOTAL HYDROPHOBICITY V A L U E (H§) The total hydrophobicity (H(|)) values of the whole extract and the major fraction were calculated according to Bigelow (1967) based on the amino acid compositions determined. Analysis of variance and Tukey's pairwise comparison test (at p = 0.05) were conducted on the (H(|)) values of the whole extract and the major fraction using MINITAB™ version 13. The total hydrophobicity (H<|)) values of the "FT" and "0.10 M" fractions from the 1st screening experiment were also calculated for comparison. 60 4.7 SULFHYDRYL GROUP (SH) AND DISULFIDE GROUP (SS) DETERMINATION The SH content of the major fraction and the whole extract was determined using Ellman's reagent (5,5-dithiobis(2-nitrobenzoic acid), DTNB) (Ellman, 1959) with modification of the methods described by Beveridge et al (1974) and Mine (1997). The total content of SH and SS groups was determined using the method of Thannhauser et al. (1983) with modification of the method described by Damodaran (1984). 4.7.1 Reactive S H 0.50 ml of protein sample (at a concentration of 8.7 mg/ml for the major fraction and 7.7 mg/ml for the whole fraction) was added to 0.50 ml Tris-Glycine-EDTA buffer at pH 8.0. The final concentration of the buffer was 85 mM Tris, 100 mM Glycine and 4 mM EDTA, pH 8.0. After incubation at 40 °C for 30 min, 30 ul of 4 mg/ml freshly made Ellman's reagent solution (in Tris-Glycine-EDTA buffer, pH 8.0) was added. The samples were allowed to stand for 15 minutes at room temperature and the colour changes were measured by the absorbance at 412 nm with a Unicam UV/Vis Spectrophotometer UV2. The colour changes of the samples after 30, 60 and 120 minutes were also recorded. Duplicate analyses were conducted. The Tris-Glycine-EDTA buffer was used as blank for adjusting zero on the spectrophotometer while the same buffer with Ellman's reagent was used as reagent blank. A412 of the protein samples without Ellman's reagent was measured as protein blank. The sulfhydryl residues were calculated as: umole SH/g protein = 73.53 * net A412 (D/C) 61 where C is the sample concentration in mg/ml; D is a dilution factor net A412 = A412 Sample with Ellman's reagent - reagent blank - protein blank 73.53 is derived from 106/1.36 * 104 in which 1.36 * 104 is the molar absorptivity (Ellman, 1959), and 106 is for conversion from molar to |xmole/ml basis and from mg to g protein 4.7.2 Total SH 0.50 ml protein sample was added to 0.50 ml of Tris-Glycine-EDTA buffer at pH 8.0 with denaturants. 2.5 M and 2.0 M guanidine thiocyanate (GuSCN) were first used as denaturant as described by Cheng (2001). GuSCN was later replaced by SDS at a concentration of 5.0 mg/ml as suggested by Beveridge et al (1974) because of the observed interference of GuSCN on the assay. Duplicate analyses were carried out. Samples were incubated, treated with Ellman's reagent, incubated and the absorbances of samples were read as described above for reactive SH. 4.7.3 Changes in reactive and total SH values over time Samples were initially incubated at ambient temperature as described by Sultanbawa (1998) instead of 40 °C. The net A412 readings were recorded after 30, 60 and 120 min and an increase in the net A412 readings over time was observed. On the other hand, when the samples were incubated at 40 °C as described by Mine (1997), no trend of increasing net A412 were observed. Incubation temperature of 40 °C was thus used. 62 4.7.4 Total SH and SS determination The NTSB (2-nitro-5-thiosulfobenzoate) stock solution was prepared by dissolving 100 mg Ellman's reagent in 10 ml of 1.0 M Na2S03 and adjusting the pH to 7.5 to produce a bright red coloured solution. The solution was then transferred into a 38 °C water bath and oxygen bubbled for approximately 2 hours until the colour of the solution changed to pale-yellow. The NTSB stock solution was divided into small aliquots in small micro-centrifuge vials and stored at -18 °C. NTSB working solution was prepared by diluting the stock solution with freshly made solution containing 2.0 M guanidine thiocyanate, 0.20 M Tris, 100 mM sodium sulfite and 3 mM EDTA at pH 9.5 in a ratio of 1 to 100. 100 ul of protein stock solution was added to 1.5 ml NTSB working solution. The mixtures were incubated in the dark for 25 min. Absorbance at 412 nm was read against a blank of 1.5 ml of NTSB working solution mixed with 100 ul water. Triplicate analyses were carried out. Total SH and SS residues were calculated as: umole (total SH + SS)/g = 75.53 netA4i2 (D/C), where C is the sample concentration in mg/ml and D is a dilution factor. Calculation of netA4i2 was described above. Disulfide groups were determined by subtracting the content of total SH group (Beveridge et al, 197r4; Mine, 1997) from the total SH and SS groups (Thannhauser et al, 1983). 4.7.5 Statistical analyses Analysis of variance and Tukey's pairwise comparison test (p = 0.05) was conducted on the reactive SH, total SH, total SH + SS and SS content of the whole extract and the major fraction using MINITAB™ version 13. 63 4.8 CALCULATION OF CYSTEINE CONTENT Results of the SH and SS determination were used to estimate the content of cysteine in the whole flaxseed protein extract and the major fraction with the assumption that one unit of SH was equivalent to one unit of cysteine and one unit of SS was equivalent to two units of cysteine. 4.9 FOAMING PROPERTIES Foaming properties of the major fraction at pH 3.0, 5.0, 7.0 in low (0.01 M NaCl) and high (1.0 M NaCl) ionic strength buffer were investigated. In addition, foaming properties of BSA, ovalbumin, P-lactoglobulin, commercial spray dried egg albumen and the whole extract at pH 7.0 in low ionic strength (0.01 M NaCl) buffer were also studied as comparisons. 4.9.1 Preparation of samples Protein stock solutions of the major fraction (with protein concentration of 8.7 mg/ml) and whole extract (7.7 mg/ml) in 5 mM Tris at pH 8.6 were thawed and diluted with citric-phosphate buffer to a concentration of 1.0 mg/ml. Stock solutions of BSA, |3-lactoglobulin, ovalbumin and commercial spray dried egg albumen (Canadian Inovatech Inc., Abbotsford, BC) at 8.5 mg/ml (by weight) were prepared in 5 mM Tris at pH 8.6. Fresh working solutions were prepared in duplicates on the day of the experiment by diluting the stock solution with citric-phosphate buffer to a concentration of 1.0 mg/ml (determined by the BCA Protein Assay). 64 4.9.2 Foaming apparatus The foaming apparatus was set up as described by Kato et al. (1983) except that conductivity cells were not included. A 150 ml graduated glass column with a diameter of 24 mm was used. The bottom of the column was fitted with a glass frit and a G-4 glass filter circle (flow rate of 75 ml/min, porosity of 5.5 s per 100 cm3, thickness of 0.28 mm; Fisher Scientific, Pittsburgh, PA). The glass frit and G-4 filter were wetted with distilled deionized water. The G-4 glass filter was placed underneath the glass frit support with the smooth side of the filter in contact with the glass frit and the porous side facing down. Figure 6 shows a picture of the foaming apparatus used in this experiment. Flow meter Nitrogen gas in Figure 6. Foaming apparatus for determination of the foaming properties of the major fraction of flaxseed proteins at a concentration of 1.0 mg/ml. 65 4 . 9 . 3 Preliminary experiments for experimental conditions and set up Preliminary experiments were carried out using commercial spray dried egg albumen known to be a good foamer (Cheng, 2001) and BSA. Higher flow rates of 80, 90, and 110 ml/min and shorter sparging time of 15, 20, 30 s were initially used based on the condition of 90 ml/min for 15 seconds reported by Kato et al. (1983). Slower flow rates of 20-80 ml/min and longer sparge times of 40-60 s were later adopted due to the difficulty of maintaining consistent conditions at higher flow rate and shorter time. The slow flow rate of 20 ml/min was tried because it was the optimum flow rate reported by Waniska and Kinsella (1979) The same condition was repeated three times. The ease of control, repeatability and reproducibility were main criteria in the selection of experimental conditions. The set of conditions yielding a more reproducible, repeatable and stable foam and which were easily repeatable was selected. The higher volume of foam produced and the longer time for the liquid to drain at different conditions were used as indicators of a set of good condition. The flow rate of 35 ml/min and a sparging time of 1 minute were finally chosen because of the ease of control of experimental conditions at this flow rate, the ease of calculation and standardization at 1 minute and the relatively reproducible foam produced. 4 . 9 . 4 Foam generation Nitrogen gas was introduced through the flow control knob at the bottom of the column and the flow rate was adjusted through a flow meter until a constant rate of 35 ml/min was 66 obtained. 5 ml of protein solution at a concentration of 1 mg/ml was then introduced from the top of the column. At the end of 1 minute, the flow control knob was closed and the tubing connecting the flow control knob of the column and the flow meter was removed to stop sparging. 4.9.5 Measurement and description of foam Because of the complexity of the foaming mechanism of different proteins, different methods and combinations of the methods were used to measure and describe the foams objectively: Set 1— Foam volume and drainage: Foamability and foam stability were initially measured solely by the volume of the foam right after sparging and the change in the foam volume over time respectively. The foam volume was denoted by the height of the foam in the graduated column with 1 ml intervals. The volume of the foam was measured every minute. Measurement of the liquid drained from foam was later used in combination with measurement of foam volume in describing the foaming properties of the protein studied. Because of the small amount of samples (5 ml) and the construction of the column, it was impossible to record the volume of liquid that did not go into foam and volume of liquid that drained out over time as described in the foam collapse rate method by the American Society of Brewing Chemists (Anonymous, 1976). Drainage of protein foams were thus measured by opening the flow control knob and allowing the protein solution to drip through the glass frit and glass filter along a thin 67 glass opening into a 10 ml graduated cylinder. Drainage of liquid was measured every minute. Set 2— Foam volume and conductivity: Foams were generated in the same way as described above except that a conductivity probe (YSI Model 31, Yellow springs, Ohio) was lowered into the column until the end of the probe reached the 30 ml mark of the column. Foams of the samples foamed around the probe as well as into the probe. Readings were first taken frequently at 45, 50, 70 and 75 s after sparging started and then in 15 and 30 s intervals thereafter. The 30 ml mark was chosen so that the conductivity probe was consistently measuring the same section of foam. The volume of the foam obtained in the presence of the probe 1 minute after foaming was also noted. Because of the failure of the YSI conductivity probe to measure the conductivity of the foam of the major fraction, another conductivity probe by Radiometer Copenhagen Type CDC 104 (Radiometer Analytical, Lyon, France) was used. The probe was lowered into the column so that it was situated between the 20-70 ml region of the column. Readings were first taken frequently at 45, 50, 70 and 75 s after sparging started and then in 15 and 30 s intervals thereafter. The volume of the foam with the probe at the end of the 1-minute sparging period was also noted and the foam volume was adjusted for the volume displaced by the probe. The adjusted foam volume was used to calculate the foaming capacity (FC), which was defined by the ratio of the volume of gas in foam to the volume of gas sparged, as described by Waniska and Kinsella (1979) and Wilde and Clark (1996). 68 ¥ C = Vf0/Vg Vf0 is the volume of foam at the end of the 1-minute sparging period Vg is the total volume of gas inputted The relative foam conductivity (Cf ) was used as an index of foaming power as described by (Wilde and Clark, 1996 ) except that Cmax was used in place of C,. (Kato et al, 1983 ) . Because of the construction of the conductivity probe, there was a delay for the foam conductivity of some samples to reach the maximum. While the length of the delay period varied among samples (denser foams had longer delay), the C, measured right after sparging stopped did not representatively reflect the density of the foams of different samples. C f = Cn^l Q * 1 0 0 % Cmax is the maximum conductivity recorded Ci is the conductivity of the protein solution Foam stability of the protein foams was expressed as the foam stability index (FSI) as described by Kato et al. ( 1 9 8 3 ) . FSI = C o * At / AC t At / AC t is the slope of the second phase of the conductivity decay curve C 0 is the y-intercept obtained by extrapolating the line for the slope of the second phase of the conductivity decay curve Analysis of variance of the FC, C f , FSI values of the whole extract, the major fraction, BSA, ovalbumin, pMactoglobulin and egg albumen was conducted as a function of pH and salt concentration using MINITAB™ version 13. One-way 6 9 ANOVA and Tukey's pairwise comparison tests (p = 0.05) were also conducted to compare the difference across all samples tested, samples in 0.01 M NaCl at pH 7 only, and only the major fraction at three pHs and two NaCl concentrations, using the same statistical programme. Set 3— Photographic illustration of foams: A Wild M3 stereomicroscope fitted with a Leitz Wetzler camera (Wild Leitz Canada Ltd, Willowdale, ON) was used to capture the change in the sizes and matrices of the observed "good foamers". An Olympus® D-620L digital camera (Olympus America Inc., Melville, NY) was used to take pictures of the overall change in foam height in the foam of the same set of samples. Pictures of the foam of the-observed "poor foamers" were also taken for comparison. Foams were generated as described above. The flow control knob remained closed at the end of the foam generation. The column was undamped, gently rotated for 90° and placed horizontally under the stereomicroscope. Photos were taken using the Wild M3 camera on three different sections of the column: 50-60 ml, 30-40 ml and 10-20 ml at different time intervals. The column was clamped back in vertical position after taking the pictures to allow draining by gravity. Photos of the foam in the column when the column was clamped in vertical positions were taken using the Olympus® digital camera. 70 4.9.6 Cleaning of foaming apparatus The foaming apparatus was disassembled and cleaned after each run. The G-4 glass filter was discarded after each use and the glass frit was soaked with hot water with regular household detergent (Palmolive dishwashing liquid, Colgate-palmolive Canada Inc., Toronto, ON) in between replicates of the same sample and with Terg-a-zyme® enzymatic cleaner (Alconx Inc., New York, NY) in between different protein samples. The glass frit was then rinse with hot running water and soaked in distilled deionized water. The glass column was cleaned with hot running water. Household detergent was then added and the column was rinsed with hot running water, followed by distilled deionized water. The (R) column was then wipe dried with Kimwipe EX-L Delicate Task Wipers (Kimberly-Clark® Corporation, Roswell, CA) with the aid of a long glass-stirring rod. 4.10 VISCOSITY Viscosities of protein samples at the same concentrations of 1.0 mg/ml used for foaming properties determination were measured to investigate whether viscosities of samples contributed to the differences in foaming properties observed. A Brookfield Synchro-Lectric LVT series viscometer attached with a small sample adapter and spindle number 18 (Brookfield Engineering Laboratories, Inc., Stoughton, MA) was used. 4.10.1 Preparation of samples Fresh working solutions of whole extract, egg albumen, BSA, ovalbumin in 0.01 M NaCl at pH 7 and major fraction in buffers of low (0.01 M) and high (1.0 M) NaCl at pH 3, 5 and 7, at concentrations of 1.0 mg/ml, were prepared in duplicates as described above for 71 foaming (Section 4.9.1). Samples were allowed to equilibrate at room temperature for 1 hour before measurements. 4.10.2 Determination of apparent viscosity 9 ml of sample was placed into the small sample adapter and the viscometer was turned to a speed of 0.6 RPM. Readings were taken after 5 complete rotations of the dial. At least three dial readings were obtained for each sample replicate. Viscosity of distilled-deionized water was also measured. The apparent viscosity in was calculated by multiplying the dial reading by a factor, which corresponded to the viscometer spindle and speed combination utilized, from a conversion table supplied by the manufacturer. For spindle 18 and speed of 0.6, the factor was 50. 4.10.3 Statistical analyses Analysis of variance and Tukey's pairwise comparison test was conducted on the apparent viscosity of the whole extract, the major fraction, egg albumen, BSA and ovalbumin using MINITAB™ version 13. 4.11 REGRESSION ANALYSES OF THE FUNCTIONAL PROPERTIES OF THE MAJOR FRACTION The relationships between the different structural and functional characteristics of the major fraction were analyzed by multiple regression model and stepwise (forward and backward) regression model using MINITAB™ version 13. 72 CHAPTER 5—RESULTS AND DISCUSSION FOR PHASE 1; ISOLATION OF THE MAJOR FRACTION OF FLAXSEED PROTEINS 5.1 ANALYSES ON FLAXSEEDS The percent solid, moisture, nitrogen and the calculated protein content of the dehulled flaxseed as well as the dehulled delipidated seeds are shown in Table 6a and 6b. Table 6a shows that the 90 % dehulled seeds did not differ significantly from the 100 % dehulled seeds in terms of the % solid and % moisture contents. The 90% dehulled delipidated ground seeds showed lower solid and higher moisture contents (p < 0.05). This was mainly due to the fact that the ground seeds had greater surface areas for more effective removal of moisture during drying and thus lower % solid determined and higher % moisture calculated by difference. Table 6b shows that the extent of hull removal significantly affect the protein contents of the seeds (p < 0.05). The higher the content of hulls remaining on the seeds, the lower the protein content. Flaxseed has been reported to contain 41 % (dry basis) fat (Bhatty and Cherdkiatgumchai, 1990). The delipidation procedure removed the fat from the seeds and therefore, higher % of nitrogen and protein were determined after delipidation of the 90 % dehulled seed. Baseds on the results from these analyses, the mechanically dehulled seed were further manually dehulled until no observable hulls remained (100 % dehulled) before subjecting to delipidation. 73 Table 6a. Percent solid and moisture content of the NorMan flaxseeds used in the experiment1. Seed description 2 % solid3 % moisture 3 90% dehulled delipidated 43.84 + 0.14a 56.16 ± 0.14 b 90% dehulled 46.82 ± 0.08 b 53.18 ± 0.08 a 100% dehulled 46.48 + 0.37 b 53.52 ± 0.37 a 1 Values are reported as means + standard deviations of 3 replicates. 2 90% dehulled flaxseeds were prepared by manually sorting the mechanically dehulled NorMan seeds until only about 10% hulls were left. The percentage was only a rough estimation made by observation. 100% dehulled flaxseeds were prepared by manually sorting the dehulled seeds until there was no hull remaining. 3 Values in the same column with different superscripts are significantly different (p < 0.05). Table 6b. Protein content (wet basis) of the NorMan flaxseeds used in the experiment. Protein [%] (Nitrogen x 5.5) Seed description (wet basis) 90% dehulled delipidated 1 45.7 + 1.3 c 90% dehulled 2 24.3 + 0.6 a 100% dehulled3 25.3 ± 0.3 b 1 90% dehulled delipidated flaxseeds were prepared by delipidating seeds which were further dehulled by manually sorting the mechanically dehulled NorMan seeds until only about 10% hulls were left. The delipidation procedures were described in Section 3.3. The percentage was only rough estimation made by observations. Values are reported as means + standard deviations of 5 replicates. 2 90% dehulled flaxseeds were prepared by manually sorting the mechanically dehulled NorMan seeds until only about 10% hulls were left. They were used as starting materials for preparation of 90 % dehulled delipidated seeds. Values are reported as means + standard deviations of 10 replicates. 3 100% dehulled flaxseeds prepared by manually sorting the mechanically dehulled NorMan seeds until there were no hulls remaining. Values are reported as means + standard deviations of 2 replicates. 4 Values with different superscripts are significantly different (p < 0.05). 74 5.2 CHARACTERISTICS OF FLAXSEED PROTEIN EXTRACTS The extraction yields and gel electrophoretic profiles of the protein extracts obtained from the first two screening experiments, as well as the findings of the two extraction studies are described in the following sections. 5.2.1 First screening experiment After extraction and centrifugation, supernatants of batches 1, 2 and 4 were cloudy. Filtering through glass wool after centrifugation did not help to clear up the extracts. Re-centrifugation of the cloudy supernatants at 20 400 x g for 30 minutes did not yield any pellet or clarify the extracts. Table 7 shows the protein yield of the four batches extracted in the 1st screening experiment. Table 7. Extraction yield of the four batches of NorMan flaxseed protein extracts prepared in the 1st screening experiment. 2-Mercapto- Protein Total Weight ethanol during content volume of Extraction yield of seeds extraction (mg/ml) of extract Batch (g) (10 mM) extract12 (ml) 1* 1.0 - 17.4 ± 2.8 d 10.4 39.2 + 6.4b 2* 1.0 + 13.1 ± 1.3° 12.5 35.6 ± 3.4b 3 1.0 - 7.3 ± 2.0a 9.8 12.9 ± 3.6a 4* 1.0 + 10.1 ± 0.9b 13.7 37.4 ± 3.5 b Cloudy extracts 1 Extraction yield refers to the protein recovered in extract expressed as a percentage of the protein content of the dehulled delipidated seeds used for extraction. The dehulled delipidated seeds used in this screening experiment had a protein content of 46.3 %. Protein contents of protein extracts were determined using Bio-Rad Protein Assay. Values are reported as the means + standard deviations of 6 assay replicates (3 dilution replicates each for 2 different dilutions). 2 Values in the same column with different superscripts are significantly different (p < 0.05). 75 Results from ANOVA and Tukey's Pairwise comparison test showed that the protein yield of batch 3 (extracted without 2-mercapthoethanol) was significantly lower than those of the other three batches (p < 0.001). On the other hand, the yield of batch 1 (also extracted without 2-mercapthoethanol) was the highest (although not significantly different) among the four batches. The cloudiness of the extracts might contribute to the overestimation of the cloudy batch by the Bio-Rad Protein Assay leading to the contradictory result. The cloudiness might have been caused by the presence of mucilage from the seed hulls because the seeds were not manually sorted to remove hulls that remained on the mechanically dehulled seeds. An additional problem was observed in the protein assay results, in which higher dilutions of the sample resulted in higher estimated protein content. This might be again caused by the cloudiness of the sample. As a result, the 1st screening experiment was inconclusive in determining the effect of 2-mercapthoethanol on the extraction yield. Figure 7 shows the SDS-PAGE profiles of the protein extracts obtained from the 1st screening experiments and Table 8 shows the calculations for the corresponding bands in lanes 4-7. Bands 1, 3 and 6 (corresponding to MW of 12, 21-22 and 31-32 kDa) were the major bands in lane 4 and 5 under reducing conditions. Bands 9 and 11 (corresponding to MW of 45 and 52 kDa) were the major bands in lanes 6 and 1 under non-reducing conditions. 76 M W (kDa) 205. 116-97 84 66 45 55 36 29 24 20 14.2 6.5 4 9 Lane 1 2 3 4 5 6 7 8. Band # 1 (lanes 4-7) y 9,io,n > 6 , 7 I 2,3,4,5 Lane 1. Sigma Marker™ (wide-range; M W range 6.5 - 205 kDa) 2. M W marker (BSA 66 kDa, ovalbumin 45 kDa, (3-lactoglobulin 18kDa) 3. M W marker (IgG 25 kDa & 55 kDa) 4. Reducing S D S - P A G E of Protein extract from 1 s t experiment; extracted without 2 - M E 2 5. Reducing S D S - P A G E of Protein extract from 1 s t experiment; extracted with 2 - M E 6. Non-reducing S D S - P A G E of Protein extract from 1 s t experiment; extracted without 2-ME 7. Non-reducing S D S - P A G E of Protein extract from 1st experiment; extracted with 2 - M E 5. Sigma Marker™ (wide-range; M W range 6.5 - 205 kDa) Figure 7. Reducing and non-reducing SDS-PAGE of flaxseed protein extracts extracted with or without 2-mercaptoethanol on 10-15 gradient PhastGel®. 1 Refer to Table 8 for details in the MW of the bands. 2 - M E = 2-mercaptoethanol. 7 7 Table 8. Molecular weight calculations of protein bands in lanes 4-7 of Figure 7. Reducing SDS Non-reducing SDS Lane 4 Lane 5 Lane 6 Lane 7 Band MW (kDa) Band MW (kDa) Band MW (kDa) Band MW (kDa) 1* 12 x* 12 1* 12 1* 13 2 21 2 21 2 21 2 22 3 22 3 22 3 22 3 23 4 23 4 24 4 24 4 25 5 26 5 27 6 31 6 31 6 32 6 32 8 36 8 38 7 35 7 35 9 45 9 45 10 49 10 50 10 48 10 49 11 52 11 52 Band numbers in bold represent darker bands observed from the gel. * MW of the bands were calculated from the extrapolation of the standard curve. The change in SDS-PAGE profiles of the protein extracts under reducing (lanes 4 and 5) and non-reducing (lanes 6 and 7) conditions indicated the presence of disulfide linkages in flaxseed proteins. Bands 8, 9 and 11 in lanes 6 and 7 (corresponding to MW of 36-38, 45 and 52 kDa, respectively) were absented in lanes 4 and 5. More bands of lower molecular weights were found under reducing conditions (bands 5 and 7 in lanes 4 and 5). Fligher intensities of lower molecular weight bands of 12 kDa and in the range of 21-32 kDa were also found in the reducing SDS lanes (lanes 4 and 5). The electrophoretic profiles of the samples extracted with (lanes 5 and 7) and without 2-mercaptoethanol (lanes 4 and 6) were similar. The addition of 2-mercaptoethanol during protein extraction did not result in any observable difference in protein composition of the extracts. 78 5.2.2 Second screening experiment None of the four extracts prepared in the 2nd screening experiment was as cloudy as compared to the extracts of the 1st screening experiment. Manual removal of hulls helped to reduce the cloudiness of the protein extract prepared and the cloudy substances in the extracts were most likely the mucilage from the hulls. Table 9 shows the protein yields in the extracts of the four batches extracted in the 2n d screening experiment. Table 9. Extraction yield of the four batches of NorMan flaxseed protein extracts prepared in the 2n d screening experiment. 2-Mercapto- Total ethanol Protein volume Batch Weight during content of of seeds extraction (mg/ml) of extract Extraction (g) (10 mM) extract12 (ml) yield (%)12 1 0.80 - 13.5 ± 2.6 ab 6.9 25.1 ± 4.8 a 2 0.80 + 13.1 ± 1.7 a 8.9 31.4 ± 4 . 1 b 3 0.26 - 18.9 ± 3.2 c 1.0 31.5 ± 5.4 b 4 0.26 + 16.3 ± 4.2 be 2.1 28.5 ± 7.3 ab 1 Extraction yield refers to the protein recovered in extract expressed as a percentage of the protein content of the dehulled delipidated seeds used for extraction. The dehulled delipidated seeds used for this screening experiment had a protein content of 46.3 %. Protein contents of protein extracts were determined using Bio-Rad Protein Assay. Values are reported as the means + standard deviations of 6 assay replicates (3 dilution replicates each for 2 different dilutions). 2 Values in the same column with different superscripts are significantly different (p < 0.001). 79 For batch size of 0.80 g (batches 1 and 2), the yield of sample extracted in the presence of 2-mercaptoethanol was significantly higher than the one without (p < 0.001). On the other hand, for the 0.26 g batches (batches 3 and 4), the batch extracted without 2-mercaptoethanol had higher (but not significant) yield. The difference in protein yield was mainly caused by the difference in volume of supernatants recovered after extraction and centrifugation. Comparing batches with the same starting materials (batch 1 vs 2 and batch 3 vs 4), the batches extracted with 2-mercaptoethanol (batches 2 and 4) had higher supernatant volume recovered. An increase in the protein content (mg/ml) estimated by the Bio-Rad Protein Assay with increasing sample dilution was still observed. The Bio-Rad Protein Assay was experiencing some interference but the source was unidentified. Figure 8 shows the reducing SDS-PAGE profiles of the flaxseed protein extracts obtained from the 1st and 2nd screening experiments and Table 10 shows the molecular weight calculation of the corresponding bands in different lanes on the gel. The electrophoretic patterns observed in Figure 8 were similar to lanes 4 and 5 of Figure 7, in which the flaxseed protein extracts had at least one band of low-molecular weight (< 12 kDa), a group of very close bands at molecular weight range of 19-27 kDa, another group at 30-35 kDa, and another band of higher molecular weight at about 49 or 50 kDa. Bands 1,4 and 7 with corresponding MW of 11, 20-21 and 30-31 kDa were again observed to be the major bands as in Figure 7. 80 M W (kDa) 1 Band #1 (lanes 2-7) 6,7 A j 3,4,5] 2 1 :M|H££ I^HHP l l i d f r ^Br ^Pr - « » 4 n » *•* Lage _ 1 2 1 - 6. 5 205 116 97 84 66 55 -45 36 29 24 -14.2 20 6.5 Lane 1. Sigma Marker™ (wide-range; M W range 6.5 - 205 kDa) 2. Protein extract from the 1 s t experiment; extracted without 2 - M E 2 (5-fold diluted) 3. Protein extract from the 1 s t experiment; extracted with 2 - M E (5-fold diluted) 4. Protein extract from the 2 n d experiment; extracted without 2 - M E (5-fold diluted) 5. Protein extract from the 2 n d experiment; extracted without 2 - M E (10-fold diluted) 6. Protein extract from the 2 n d experiment; extracted with 2 - M E (5-fold diluted) 7. Protein extract from the 2 n d experiment, extracted with 2 - M E (10 fold diluted) 8. Sigma Marker™ (wide-range; M W range 6.5 - 205 kDa) Figure 8. Reducing SDS-PAGE of flaxseed protein extracts extracted with or without 2-mercaptoethanol from the 1st and 2nd screening experiments on 10-15 gradient PhastGel®. 1 Refer to Table 10 for details in the M W of the bands. 2 - M E = 2-mercaptoethanol. 81 Table 10. Molecular weight calculations of protein bands in lanes 2, 3, 4 and 6 of Figure 8. Lane 2 Lane 3 Lane 4 Lane 6 Band MW (kDa) Band MW (kDa) Band MW (kDa) Band MW (kDa) 1* 11 1* 11 1* 11 1* 11 2 15 2 15 2 15 2 15 3 19 3 19 3 19 3 19 4 20 4 20 4 20 4 21 5 22 5 23 5 23 5 23 6 24 6 24 6 24 6 25 7 30 7 30 7 30 7 31 8 33 8 33 8 33 8 34 9 48 9 48 9 49 9 49 Band numbers in bold represent darker bands observed from the gel. * MW of the bands were calculated from the extrapolation of the standard curve Findings similar to those of Figure 7 were observed in Figure 8, in which the electrophoretic profiles of the flaxseed protein extracts extracted with 2-mercaptoethanol (lanes 3, 6 and 7) were similar to those without (lanes 2, 4 and 5). The addition of 2-mercaptoethanol during protein extraction did not make any observable difference in compositions of the extracts. The addition of 2-mercaptoethanol increased the extraction yield by improving the yield of supernatant recovered upon centrifugation of the extracts but did not change the relative compositions of proteins being extracted. The addition of 10 mM 2-mercaptoethanol in the extraction media for globulins has been reported to increase the amount of protein extracted (Marcone et al, 1998). 2-mercaptoethanol prevented aggregation of 1 IS protein by cleaving the intermolecular SS bonds (Peng et al, 1984). The higher supernatant 82 recovery from centrifugation might therefore be contributed by the preventive action of 2-mercaptoethanol against aggregation of the flaxseed proteins. Results from the first two screening experiments showed that the presence of hull in the flaxseed seriously affects the clarity of the protein extracts. The use of 2-mercaptoethanol increased the yield but did not affect the profile of the protein extracted. In spite of the higher yield obtained with the extraction with 2-mercaptoethanol, it was not used for subsequent extraction because of the concern over the possible change in the properties of the protein in the major fraction which might result from reduction of disulfide linkages by 2-mercaptoethanol. 5.2.3 Studies on extraction yield 5.2.3.1 First extraction study: Figure 9 shows the non-reducing SDS-PAGE profiles of the residues and the extracts obtained from re-extraction of residual ground flaxseeds as compared to the extract obtained in the first extraction. The residues from extraction and the extracts from re-extraction had very similar PAGE profiles with the extracts from the first extraction. Despite the low extraction yield, the extract obtained contained representative proteins from the seeds. 83 M W (KDa) 66 m m 55 W 45 Lane ,1 2 3 4 25 18 6 7 8 Lane 1. Residue from extraction (7.14 mg/ml) 2. Residue from extraction (3.57 mg/ml) 3. Residue from extraction (15.32 mg/ml) 4. M W marker (BSA 66 kDa, ovalbumin 45 kDa, |3-lactoglobulin 18kDa, IgG 25 kDa & 55 kDa) 5. Protein extract from the 1 s t extraction (~ 1.3 mg/ml) 6. Protein extract from re-extraction of residue 7. Protein extract from re-extraction of residue (2x diluted) 8. Protein extract from re-extraction of residue (5x diluted) Figure 9. Non-reducing SDS-PAGE of the extracts and residues from extraction of NorMan flaxseed proteins, as well as the extracts from re-extraction of the residues on 8-25 gradient PhastGel®. 5.2.3.2 Second extraction study: Table 11 shows the results of the 2n d extraction study. The 1st extractions at 16-hr or 1-hr were only able to extract 22.7 and 21.3 % of proteins from flaxseeds, respectively. The 2nd extractions were only able to extract the remaining 4.8 and 4.3 % proteins from the residues of the 16 hr and 1 hr sets, respectively. In the 3rd re-extractions, only 2.3 and 4.3 % proteins were extracted from the residues from the 16-hr and 1-hr sets, respectively. The two-step centrifugation at 10 400 and 20 400 x g in place of the one-step at 24 000 x g did not result in a higher yield. The total yield after three consecutive extractions was around 30 %. Extraction for 16 hours and 1 hour did not differ greatly in the total amount of protein extracted except that the extraction for 16 hours appeared to be able to extract slightly more proteins in the 1st and 2nd extraction than extraction for 1 hour. Table 11. Protein content estimated by the BCA Protein Assay on the 16-hr and 1-hr extraction sets in the 2nd extraction study for NorMan flaxseed proteins. 16-hr extraction set 1-hr extraction set g protein % extracted g protein % extracted 1st extract 0.520 22/7 0.889 213 2nd extract (2hr) 0.100 4.8 0.170 4.3 3rd extract (2hr) 0.046 2.3 0.170 4.3 Total extracted 6.666 29.8 1.23 29.9 Original protein1 2.29 4.17 (5.00g dd seed)2 (9.02 g dd seed) 1 Based on 45.7% protein in 90 % dehulled delipidated flaxseeds determined by Leco nitrogen analysis. 2 dd seed = dehulled delipidated NorMan flaxseeds 85 Figures 10 and 11 show the SDS-PAGE profiles of the extracts, residues and pellets from protein extraction. The PAGE profiles of the 16-hr and 1-hr extracts in Figure 10 were very similar except that the 16-hr extract samples (lanes 2-4) showed more distinguishable higher molecular weight bands at around 45-66 kDa than the 1-hr samples (lane 5-7). The extraction for 16 hours appeared to be more effective in extracting some of the high-molecular weight protein components. The PAGE profiles of the residue collected after three consecutive extractions (lanes 1, 2, 8) and the pellet from centrifugation (lanes 3, 7) in Figure 11 were very similar to that of the first extract (lane 6). Repeated extractions of the residues repeatedly extracted a small but representative portion of the total proteins of flaxseeds. No further studies were carried out to investigate the method of increasing extraction yield because this was not the main objective of the thesis. The 16-hr and single extraction procedure was used for all subsequent extractions. Studies by Sultanbawa (1998) suggested that re-grinding of dehulled ground seeds which had passed once through the Wiley Mill # 10 (2 mm) sieve with a smaller sieve size # 20 (1 mm), and grinding the delipidated ground seed with the same # 20 (1 mm) sieve would increase extraction yield. The finer ground flaxseeds gave greater surface area for more effective extraction and would probably give higher extraction yield. 86 jfc MW (KDa) 66 55 45 25 18 Lane%Jl 8 .A Lane 1. M W marker (BSA 66 kDa, ovalbumin 45 kDa, (3-lactoglobulin 18kDa, IgG 25 kDa & 55 kDa) 2. 1 s t extract from 16 hr extraction 2 n d extract from 16 hr extraction 3 r d extraction from 16 hr extraction 1 s t extract f roml hr extraction 2 n d extract from 1 hr extraction 3. 4. 5. 6. 7. 3 r d extraction from 1 hr extraction 8. M W marker (BSA 66 kDa, ovalbumin 45 kDa, P-lactoglobulin 18 kDa, oc4actalbumin 14 kDa) Figure 10. Reducing SDS-PAGE of NorMan flaxseed protein extracts obtained by repeated extractions on 8-25 gradient PhastGel®. (Samples at concentration of approximately 2 mg/ml were applied to each lane) 87 Lane 1. Residue from the 16 hr extraction set (8.64 mg/ml) 2. Residue from the 1 hr extraction set (9.00 mg/ml) 3. Pellet from centrifugation (pellet pooled from all centrifugations; 9.06 mg/ml) 4. MW marker (BSA 66 kDa, ovalbumin 45 kDa, (3-lactoglobulin 18kDa, IgG 25 kDa & 55 kDa) 5. MW marker (BSA 66 kDa, Ovalbumin 45 kDa, 3-lactoglobulin 18kDa, a-lactalbumin 14kDa) 6. 1st extraction from 16 hr extraction 7. Pellet from centrifugation (pellet pooled from all centrifugations; 9.06 mg/ml) 8. Residue from the 16 hr extraction set (10.06 mg/ml) Figure 11. Reducing SDS-PAGE of the residues after three consecutive extractions of NorMan flaxseeds PhastGel0 and pellets from centrifugation in the 2nd extraction study on 8-25 gradient 88 5.3 CHARACTERISTICS OF FLAXSEED PROTEIN FRACTIONATED BY ANION-EXCHANGE CHROMATOGRAPHY Fractions collected from the DEAE anion-exchange chromatography from the first three preliminary experiments and from the scale-up isolations are characterized in the following sections. 5.3.1 First screening experiment Protein extract prepared without the addition of 2-mercaptoethanol from batch 3 of the 1st screening experiment was selected for further fractionation because it was the only batch of non-cloudy protein extract among the four. Loading a cloudy extract onto a DEAE-Sephacel column would result in undesirable clogging of the column and interference with anion-exchange process. 5.3.1.1 DEAE-Sephacel chromatographic pattern: Figure 12 shows the elution profile for the 1st screening experiment. The fraction eluted at 0.20 M NaCl (at conductivity of 13.5-17.0 mS/cm; also called the "0.20 M fraction" for simplicity) accounted for the largest peak. The fraction eluted at 0.10 M NaCl (at conductivity of 10.0-10.5 mS/cm; also called the "0.10 M fraction" for simplicity) accounted for the second highest peak. A distinctive peak was not observed for the 0.15 M NaCl fraction (at conductivity of 11.0-13.0 mS/cm; also called the "0.15 M fraction" for simplicity) but the A2sofor the fraction stayed in the region of 0.2-0.4. 89 0.20M 0.25M A 0.30M A AA A •A-280 • A- • - Conductivity 26.0 24.0 22.0 20.0 18.0 16.0 14.0 12.0 + 10.0 8.0 6.0 4.0 2.0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 Elution volume (ml) Figure 12. Elution profile for fractions collected from a 50 ml DEAE-Sephacel column in the 1st screening experiment. The column was equilibrated with 25 mM Tris at pH 8.6. The numbers "0.10 M". . . refer to the concentration of NaCl used for step gradient elution. 9.0 ml of protein extract, extracted without 2-mercaptoethanol, from 1.00 g of dehulled delipidated NorMan flaxseed was loaded onto the column. 90 5.3.1.2 Protein content of each fraction Table 12 shows the protein concentrations of the fractions determined by the BCA Protein Assay. The total percentage of protein found in the 0.15 M fraction eluting as a shoulder next to the 0.20 M peak was higher than that of the 0.10 M fraction, which appeared from the A280 to be the 2nd largest peak. Table 12. Protein concentration of each fraction collected from the 1st screening experiment. Fraction Protein concentration of fraction 1 (mg/ml) % of total protein fractionated 2 FT 0.313 2.65 0.10 M 0.923 12.2 0.15 M 0.783 15.3 0.20 M 2.010 58.9 0.25 M 0.464 9.73 0.30 M 0.071 1.33 'Protein concentration determined by the BCA Protein Assay. 2 % total protein calculated based on the sum of protein content of fractions FT-0.30 M fraction. 91 5.3.1.3 Gel electrophoretic patterns Figure 13 shows the reducing and non-reducing SDS-PAGE profiles of fractions eluted in the 1st screening experiment. Fractions were loaded onto the gel without any dilution or concentration. The protein concentration of each fraction loaded into each lane can be found in Table 12. From the results of BCA assay, it appeared that the FT fraction had lower protein concentration than the 0.10 M fraction. On the other hand, the intensities of the SDS-PAGE bands of the FT fraction were higher than those of the 0.10 M fraction. The BCA assay might be underestimating the protein concentration of the FT fraction or the FT fraction was simply more sensitive to Coomassie staining. Fractions that was not bound to the DEAE column and that eluted at 0.10 M NaCl both consisted of low-molecular weight components, which migrated close to the position of the tracking dye on the PhastGel® (lanes 2 and 3, Figure 13). The 0.15, 0.20 and 0.25 M fractions had more high-molecular weight components as compared to the FT and 0.10 M fractions. Comparison of the electrophoretic profiles of reducing and non-reducing SDS-PAGE showed the presence of disulfide linkages in the 0.15 M, 0.20 M, 0.25 M fractions and the whole protein extract. The FT and 0.1 M fractions had bands that were too close together and too close to the bottom of the gel and therefore, it was hard to compare the changes in the electrophoretic profiles between reducing and non-reducing conditions. 92 Reducing SDS Non-reducing SDS • Lane 1 2 3 4 5 6 7 8 Lane 1 Lane 7. Sigma Marker (wide-range; MW range 6.5-205 kDa) 2. Flow through (FT) fraction 3. 0.10 M fraction 4. 0.15 M fraction 5. 0.20 M fraction 6. 0.25 M fraction 7. 0.30 M fraction 8. Whole protein extract (-1.3 mg/ml) Figure 13. Reducing and non-reducing SDS-PAGE of the fractions collected from the 1st screening experiment on 10-15 gradient PhastGel . The same sequence of sample applications was used for the two gels. Protein concentrations of each fraction (listed in Table 12) were not adjusted before loading onto the gel. 93 Figure 14 shows the Native PAGE profiles of the fractions collected in the 1st screening experiment. Fractions were loaded onto the gel without any dilution or concentration as described for the SDS-PAGE. The proteins in the 0.15 M, 0.20 M and 0.25 M fractions (lanes 5, 6 and 7) appeared to have very similar charges and sizes. These fractions accounted for the majority of the protein from the whole flaxseed protein extract. There were two minor bands on either side of the major bands for 0.15 M, 0.20 M and 0.25 M fractions and the whole protein extract (lanes 5, 6,7 and 8). There was a noticeable thick band at the point of sample applications in lane 6 (0.20 M fraction) and very weak ones in lanes 5, 7, and 8 (0.15, 0.25 M fractions and whole protein extract). These might be contributed by larger aggregates that did not enter the gel. The FT fraction did not migrate along the gel but migrated in opposite directions toward the anode suggesting that it was cationic at the running conditions of the Native PAGE (lane 3). According to the PhastSystem™ Separation Technique File No. 120, "the pH of the buffer system for Native PAGE was 8.8 and proteins with pis less than 8.5 would take on a net negative charge and migrate through the homogenous stacking gel zone". This suggested that the FT fraction had pi higher than 8.5. Madhusudhan and Singh (1985d) also reported the basic nature of the low-molecular weight fraction they isolated. The weak band positioned close to the anode in lane 8 for the whole extract was probably also contributed by the FT fraction. The 0.10 M (lane 4) fraction appeared to consist of many small peptides of different sizes and charges. 94 Interface between stacking and resolving gel + Lane 1. BSA 66 kDa, ovalbumin 45 kDa, p-lactoglobulin 18kDa 2. IgG 160 kDa 3. FT fraction (4x diluted) 4. 0.10 M fraction 5. 0.15M fraction 6. 0.20M fraction 7. 0.25M fraction 8. Whole protein extract (lOx diluted) Figure 14. Native PAGE of fractions collected from the 1st screening experiment on 8-25 gradient PhastGel®. Protein concentrations of each fraction (listed in Table 12) were not adjusted (unless otherwise stated) before loading onto the gel. 95 5.3.1.4 Amino acid compositions: Results for amino acid analyses of the 0.20 and 0.25 M fractions compared to the literature values are shown in Table 13. The 0.20 and 0.25 M fraction had very similar amino acid compositions to the flaxseed 12S protein reported by Madhusudhan and Singh (1985c) and Marcone et al. (1998). The two fractions differed from the Marcone et al: (1998) globulin with lower E and I and higher S. The two fractions and the 12S isolated by Madhusudhan and Singh (1985c) differed from the "salt-soluble fraction" reported by Dev et al. (1986), with lower E and H, as well as higher D, S, A, L, and K. This could be explained by the fact that the "salt-soluble fraction" by Dev et al. (1986) was prepared from extraction in 0.50 M salt solution without further purification while the 0.20 and 0.25 M, 12S protein fractionated by Madhusudhan and Singh (1985c), and flaxseed globulin prepared by Marcone et al. (1998) were more purified. Results for amino acid analysis of the unbound (FT) and 0.10 M fractions compared with literature values are shown in Table 14. Variations in the amino acid compositions of the three literature values were found, especially for E, P, G, A, M, Y, F, H and R. The FT fraction had similar composition as the "small molecular weight protein fraction" isolated by Madhusudhan and Singh (1985d) except that it had higher G and lower H, A, P, V and I. The 0.10 M fraction had similar composition as the "water-soluble fraction" reported by Dev et al. (1986) except that it had higher D, G and L and lower Ff. Both FT and 0.10 M fractions had higher G, K, C, and E but lower D, H and F contents compared to the 0.20 and 0.25 M fractions. 96 Table 13. Comparison of amino acid compositions (g/lOOg) 1 of laboratory prepared 0.20 and 0.25 M fractions with the reported values. Amino Acid Madhusudhan & Singh, 1985c 12S protein Dev& Sienkiewicz, 1987 Salt-soluble fraction Marcone et al, 1998 Flaxseed globulin Laboratory prepared 0.20 M 0.25 M fraction fraction Aspartic acid2 D 11.3 8.28 12.4 11.7 12.7 Glutamic acid2 E 19.8 24.5 24.3 21.1 20.8 Serine S 5.10 3.98 3.1 5.90 5.27 Glycine G 4.80 5.55 5.4 5.54 5.19 Histidine H 2.50 3.53 2.4 2.49 2.42 Arginine R 12.5 10.2 12.6 12.3 12.7 Threonine T 3.90 3.08 3.6 3.84 3.40 Alanine A 4.80 3.83 5.5 5.00 5.45 Proline P 4.50 4.09 — 3.76 4.00 Tyrosine Y 2.30 3.16 2.4 2.39 2.67 Valine V 5.60 4.32 5.1 4.89 5.20 Methionine M 1.70 1.89 1.3 1.34 1.25 Cysteine C 1.40 n.a3 0.9 0.58 0.54 Isoleucine I 4.60 4.15 5.6 4.47 3.97 Leucine L 5.80 4.54 5.9 5.48 5.78 Phenylalanine F 5.90 4.30 6.3 5.75 5.34 Lysine K 3.10 4.37 3.1 2.86 3.01 1 Cysteine was not derivatized before analysis; Tryptophan was not analyzed. 2Asparagine and glutamine were quantitatively converted to aspartic and glutamic acids respectively. 3 n.a = not available. 97 Table 14. Comparison of amino acid compositions (g/lOOg)1 of laboratory prepared FT and 0.10 M fractions with the reported values. Madhusudhan Dev& Youle & Huang, 1981 & Singh, 1985d Sienkiewicz, 1987 Laboratory prepared Amino Acid Small MW protein Water-soluble fraction 2S protein Unbound (FT) fraction 0.10 M fraction Aspartic acid 2 D 5.50 8.62 6.36 5.16 8.51 Glutamic acid E 35.0 28.2 23.7 37.3 25.7 Serine S 3.90 4.24 6.05 4.13 4.78 Glycine G 4.80 5.55 2.42 8.47 7.15 Histidine H 1.60 3.65 1.23 0.86 1.65 Arginine R 13.1 10.3 6.03 12.2 9.38 Threonine T 2.10 4.44 3.58 2.46 4.09 Alanine A 4.80 3.83 4.00 2.58 4.74 Proline P 3.00 3.63 1.57 1.65 3.38 Tyrosine Y 1.40 3.35 1.47 1.32 2.34 Valine V 5.60 4.32 1.25 2.92 4.54 Methionine M 1.70 1.89 0.54 1.13 1.55 Cysteine C 3.50 n.a.3 4.09 3.23 1.50 Isoleucine I 4.60 4.15 5.78 3.08 3.50 Leucine L 5.80 4.54 5.34 5.71 6.03 Phenylalanine F 2.40 4.75 2.16 2.93 4.04 Lysine K 4.90 5.10 5.96 4.78 5.27 1 Cysteine was not derivatized before analysis; Tryptophan was not analyzed. 2 Asparagine and glutamine were quantitatively converted to aspartic and glutamic acids respectively. 3 n.a. = not available. 98 5.3.2 Second screening experiment To be consistent with the 1st screening experiment which used a protein extract prepared without the addition of 2-mercaptoethanol for fractionation, protein extract prepared without the addition of 2-mercaptoethanol from batch 1 of the 2nd screening experiment was used for further fractionation by anion-exchange chromatography. 5.3.2.1 DEAE-Sephacel chromatographic pattern: Elution profile for the 2n d screening experiment (Figure 15) was similar to that of the 1st screening experiment (Figure 12). The fractionation procedure was very reproducible. The fraction eluted at 0.20 M NaCl (at conductivity of 15.0-19.0 mS/cm) again accounted for the highest peak, followed by the 0.10 M fraction (10.0-11.0 mS/cm) and the flow through fraction. The 0.15 M fraction (13.0-15.0 mS/cm) again appeared as a shoulder next to the 0.20 M peak. The 0.25 M fraction (19.0-22.0) again showed a small peak while the 0.30 M fraction (23.0-25.0 mS/cm) had almost no peak. Fractions of 0.45 M (26-36 mS/cm) and 0.50 M (37^ 11 mS/cm) were collected for the interest of the cadmium-binding protein project. 99 Figure 15. Elution profile for fractions collected from a 50 ml DEAE-Sephacel column in the 2 screening experiment. The column was equilibrated with 25 mM Tris at pH 8.6. The numbers "0.10 M". . . refer to the concentration of NaCl used for step gradient elution. 6.9 ml of protein extract, extracted without 2-mercaptoethanol, from 0.80 g of dehulled delipidated NorMan flaxseed was loaded onto the column. 100 5.3.2.2 Protein content of each fraction Table 15 shows the protein contents of the fractions from the 2n d screening experiment determined by the Bio-Rad Protein Assay. The Coomassie dye in the Bio-Rad Protein Assay was apparently ineffective in detecting the proteins in the FT fraction even after concentration of the fraction and the A595 readings remained very low. This might be due to the fact that the molecular weight of the FT fraction was too low to be detectable by the protein assay. The detection limit of the Bio-Rad Protein Assay suggested by the manufacturer was 3-5 kDa while FT fraction consisted mainly of low-molecular weight components. The Bio-Rad Protein Assay was therefore not used for any subsequent protein determination. Table 15. Protein concentration of each fraction collected from the 2n screening experiment. Fraction Protein concentration of fraction (mg/ml)1 % of total protein fractionated2 FT 0.098 1.62 0.10 M 0.269 5.59 0.15 M 0.434 9.22 0.20 M 1.824 65.3 0.25 M 0.489 14.6 0.30 M 0.123 3.75 'Protein concentration determined by the Bio-Rad Protein Assay. 2 % total protein calculated based on the sum of protein content of fractions FT-0.30 M fraction. 101 5.3.2.3 Gel electrophoretic patterns Figures 16 and 17 respectively show the reducing and non-reducing SDS-PAGE profiles of the fractions collected from the 2nd screening experiment, while Tables 16 and 17 respectively show the corresponding molecular weight calculations for the bands in the two Figures. Fractions were again applied onto the gel without further dilution or concentration. The concentration of each fraction loaded onto the gel was listed in Table 15. The intensities of the bands from the FT fraction again appeared darker than those of the 0.10 M fraction even though the protein assay results showed that its protein concentration was lower than that of the 0.10 M fraction. Fractions eluted with buffer of the same ionic strength from the 1st and 2nd screening experiments had very similar compositions observed under SDS-PAGE suggesting reproducibility of the fractionating procedure. (Figures 13, 16, 17). The FT and 0.10 M fractions had a few low-molecular weight components (12-18 kDa) while the 0.15 M, 0.20 M and 0.25 M fractions had very similar compositions with molecular weight ranging from 12-51 kDa. Comparing the PAGE profiles and molecular weight of the bands in reducing and non-reducing SDS-PAGE (Figures 16 and 17), higher molecular weight bands of the 0.20 M fraction and the whole extract at about 44, 68, 86 and 97 kDa (bands 6, 7, 9, 10, 11) were only found in the gel under non-reducing conditions. The absence of a few higher molecular weight bands accompanied by the increase in intensities in bands of 21-25 kDa molecular weight range and around 30 kDa in the gel under reducing conditions indicated that some higher molecular weight components were 102 disulfide linked and were broken down to smaller components as the bonds were broken by 2-mercaptoethanol. The molecular weights of the lower molecular weight bands in the fractions and in the extract might not have been accurately estimated because they appeared so close to the solvent front. Their molecular weights might have exceeded the separation capacity of the PhastGel 10-15 gradient, which was designed for a molecular separation range of 10-250 kDa. Using a PhastGel® gradient with a better separation power for low-molecular weight components might show a better separation of the low-molecular weight component and better estimation of the molecular weights. 103 M W (kDa) 97 66 45 ad 1.6 •>A 6.5 14.2 Lane Band #1 (lanes 2-8) 1MB 2 ,3 ,4 ,5 ,6 III, IV 1/1, II 3 4 5 6 7 8 V,.. 1. Sigma Marker™ (wide-range; MW range 6.5-205 kDa) 2. Flow through (FT) fraction 3. 0.10 M fraction 4. 0.15 M fraction 5. 0.20 M fraction 6. 0.25 M fraction 7. 0.30 M fraction 8. Whole protein extract (-2.6 mg/ml) Figure 16. Reducing SDS-PAGE of the fractions collected from the 2n d screening experiment on 10-15 gradient PhastGel®. Protein concentrations of each fraction (listed in Table 15) were not adjusted before loading onto the gel. 1 Refer to Table 16 for details in the MW of the bands. 104 Table 16. Molecular weight calculations of protein bands in lanes 2- 6 and 8 of Figure 16. Lane Fraction Band# MW (kDa) Lane Fraction Band# MW (kDa) 2 FT I* 12 3 0.10M I* 12 II* 14 II* 14 III 16 III 15 IV 18 IV 17 4 0.15 M 1* 12 5 0.20 M 1* 12 2 21 2 21 3 22 3 22 4 24 4 24 5 25 5 27 6 30 6 32 7 47 7 49 6 0.25 M 1* 12 8 Whole 1/1* 12 2 22 extract II* 13 3 23 III 15 4 25 IV 17 5 27 3 22 6 33 4 24 5 27 6 28 7 51 * MW of the bands were calculated from the extrapolation of the standard curve Band numbers in bold represent darker bands observed from the gel. 105 1. Sigma Marker™ (wide-range; MW range 6.5 - 205 kDa) 2. Unbound (How through, FT) 3. 0.10 M fraction 4. 0.15 M fraction 5. 0.20 M fraction 6. 0.25 M fraction 7. Whole protein extract 8. Sigma Marker™ (wide-range; MW range 6.5 - 205 kDa) Figure 17. Non-reducing SDS-PAGE of the fractions collected from the 2n screening (It) experiment on 10-15 gradient PhastGel . Protein concentrations of each fraction (listed in Table 15) were not adjusted before loading onto the gel. 1 Refer to Table 17 for details in the MW of the bands. 106 Table 17. Molecular weight calculations of protein bands in lanes 2- 7 of Figure 17. Lane Fraction Band# MW (kDa) Lane Fraction Band# MW (kDa) 2 FT I* 12 3 0.10M I* 12 II* 14 II* 14 III 18 III 16 IV 21 IV 22 V 28 4 0.15M 1* 12 5 0.20M 1* 12 2 18 2 18 3 22 3 22 4 24 4 25 5 31 5 31 6 36 6 36 7 45 7 44 8 51 8 51 9 68 10 86 11 97 6 0.25M 1* 13 7 Whole 1/1* 12 2 18 extract II* 14 3 22 III 16 4 24 2 17 5 31 3 22 7 45 4 24 8 51 V 28 5 33 6 37 7 45 8 52 9 68 10 86 11 97 * MW of the bands were calculated from the extrapolation of the standard curve Band numbers in bold represent darker bands observed from the gel. 107 Figure 18 shows the Native-PAGE profiles of the fractions collected from the 2na screening experiment. The Native-PAGE profiles of fractions with protein concentrations unadjusted from the 2nd screening experiment (the left gel on Figure 18) were very similar to that of the 1st screening experiment (Figure 14), with FT being basic, 0.10 M fraction consisting of multiple bands, 0.15, 0.20 and 0.25 M fraction having similar profiles. Sample concentrations unadjusted Sample concentrated or diluted to 2 mg/ml « Lane 1 . 2 3 4 5 6 7 8 Lane: 1. BSA 66 kDa, ovalbumin 45 kDa, P-lactoglobulin 18kDa 2. Flow through (FT) fraction 3. 0.10 M Fraction 4. 0.15 M Fraction 5. 0.20 M Fraction 6. 0.25 M Fraction 7. Whole protein extract 8. BSA 66 kDa, ovalbumin 45 kDa, P-lactoglobulin 18kDa Lane^^ 2 3 4 5 6 7 8 ^. Lane: 1. Flow through (FT) fraction 2. 0.10 M Fraction 3. 0.15 M Fraction 4. 0.20 M Fraction 5. 0.25 M Fraction 6. 0.30 M Fraction 7. Whole protein extract 8. Flow through fraction (10 x diluted) Figure 18. Native PAGE of the fractions collected from the 2nd screening experiment on 8-25 gradient PhastGel®. 108 Fractions from the 2n screening experiment were also concentrated by freeze-drying a calculated amount of the fractions and reconstituting them in a pre-determined amount of distilled deionized water to obtain final concentrations of 2 mg/ml before loading onto the gel (the right gel on Figure 18). Calculations were made based on the protein concentrations estimated by the Bio-Rad Protein Assay. The PAGE profiles on the right gel of Figure 18 again suggested that the Bio-Rad Protein Assay was not accurate in estimating the protein concentrations of the flaxseed protein fractions. Although all fractions from lanes 1-8 were estimated by the assay to have the same concentrations, they showed distinctively different intensities on the gel. For instance, the extremely thick and dark band remaining in area of sample application in lane 1 indicated that the lane had been heavily overloaded. The Native PAGE profiles of the fractions on the right gel of Figure 18 also showed that there was a slight difference in the distance travelled by the protein bands of 0.15, 0.20 and 0.25 M fractions, with the 0.25 M fraction being further down, followed by the 0.20 and 0.15 M fractions. The slight difference in the distance travelled might be attributed by the difference in charge of the protein component as indicated by their affinity to the DEAE-Sephacel anion-exchange column. On the other hand, the observed difference in the distance the protein band migrated might also have been caused by an uncontrolled inconsistency in the running condition as indicated by the slanted solvent front. 109 Results from the first two screening experiments showed that the method established successfully isolated the low-molecular weight protein fraction from the high-molecular weight protein fraction. Since the 0.20 M fraction was found to be the major fraction of the flaxseed protein and had similar amino acid composition as the major storage protein isolated by other groups (Madhusudhan and Singh, 1985c; Dev and Sienkiewicz, 1987), this fraction was selected for detailed study. 5.3.3 Third screening experiment The third screening experiment was carried out to explore the feasibility of speeding up the fractionation procedure. Although the FT and 0.10 M fractions had different characteristics, they were eluted together in a single elution step using 0.10 M NaCl because they were not the fractions of interest. Since the 0.15 M, 0.20 M, and 0.25 M fractions were found to have very similar electrophoretic and amino acid profiles, the three-step elution was combined to a single one of 0.25 M NaCl. 5.3.3.1 DEAE-Sephacel chromatographic pattern: The elution profile of 100 ml of flaxseed protein extract on 50 ml DEAE-Sephacel column is shown in Figure 19. More than 10 bed volumes were required to wash out the first FT peak using 0.10 M NaCl, 0.10 M Tris at pH 8.6. The area of this peak 110 was greater than that of the 0.25 M peak. The broader and larger peak of the FT fraction suggested that the column might have been overloaded. This was later confirmed by reloading the suspected "overloaded FT fraction" onto a 225 ml DEAE-Sephacel column. Figure 20 shows that the suspected overloaded FT fraction from the 3rd screening experiment could be further fractionated into two separate fractions, one unbound at 0.10 M NaCl and the other eluted at 0.25 M NaCl. The capacity of the 50 ml DEAE-Sephacel column used in the 3rd screening experiment was not large enough to bind all the protein which should be bound (ie. All the resins had been saturated and there was no more binding site). As a result, any unbound proteins were collected as FT. 5.3.3.2 Gel electrophoretic patterns As the 10-15 gradient PhastGel® was unable to separate the low-molecular weight components, high density PhastGel® was used. Figure 21 shows a high density SDS-PAGE of the low-molecular weight fraction (FT and 0.10 M fractions in the 1st screening experiment versus the "combined FT" in the 3rd screening experiment). The "combined FT fraction" was the fraction not bound to the 225 ml DEAE-column equilibrated with 0.10 M NaCl in 0.10 M Tris at pH 8.6 upon reloading the "overloaded FT fraction" collected from the 3rd screening experiment. It was called "combined FT" to distinguish from the FT and 0.10 M fractions from the 1st screening experiment, which were eluted separately at 25 mM Tris and 0.10 M NaCl respectively. Table 18 shows the molecular weight calculations for the 111 corresponding bands in each lane of the gel. The FT and 0.10 M fractions consisted of at least 4 bands with molecular weight ranging from 8-29 kDa. The change in the intensities of the major bands of the FT and 0.1 M fractions under non-reducing (lanes 3 and 5) and reducing (lanes 4 and 6) conditions suggested that these fractions were disulfide linked. The combined FT fraction from the 3rd screening experiment was over diluted before loading onto the gel diluted and as a result, only very weak bands were observed. Figure 22 shows the high density PhastGel® SDS-PAGE profiles of the "combined 0.25 M" fraction from the 3rd screening experiment as compared to the 0.20 M fraction from the 1st screening experiment. The fraction eluted at a single step of 0.25 M NaCl in 0.10 M Tris at pH 8.6 in the 3rd screening experiment was termed the "combined 0.25 M" to distinguish from the 0.15, 0.20 and 0.25 M fractions which were respectively eluted at 0.15, 0.20 and 0.25 M NaCl from the 1st screening experiment. Table 19 shows the molecular weight calculations for the corresponding bands in each lane of the gel. 112 Figure 19. Elution profile for fractions collected from a 50 ml DEAE-Sephacel column in the 3r screening experiment. The column was equilibrated with 0.10 M NaCl in 0.10 M Tris at pH 8.6. The numbers "0.10 M".. . refer to the concentration of NaCl used for step gradient elution. 100 ml of protein extract prepared from 8.91 g dehulled delipidated NorMan flaxseed was loaded onto the column. Elution volume (ml) Figure 20. Elution profile for reloading 35ml of the unbound (FT) fraction, collected from the suspected overloaded 50 ml DEAE-Sephacel column in the 3rd screening experiment, onto a 225 ml DEAE-Sephacel column. The column was equilibrated with 0.10 M NaCl in 0.10 M Tris at pH 8.6. The numbers "0.10 M".. . refer to the concentration of NaCl used for step gradient elution. 113 Lank 6 7 8 •r Lane 1. Pharmacia PMW marker (M.W. range 2.5- 17 kDa)1 2. MW marker (BSA 66 kDa, ovalbumin 45 kDa, P-lactoglobulin 18kDa, IgG 25 kDa & 55 kDa) 3. Non-reducing SDS-PAGE of the FT fraction from the 1st screening experiment 4. Reducing SDS-PAGE of the FT fraction from the 1st screening experiment 5. Non-reducing SDS-PAGE of the 0.10 M fraction from the 1st screening experiment 6. Reducing SDS-PAGE of the 0.10 M fraction from the 1st screening experiment 7. Non-reducing SDS-PAGE of the combined FT from the 3rd screening experiment (5x diluted) 8. Reducing SDS-PAGE of the combined FT from 3rd screening experiment (5x diluted) Figure 21. Reducing and non-reducing SDS-PAGE of low-molecular weight fractions collected from the 1st and 3rd screening experiments on high density PhastGel®. Protein concentrations of the fractions from the 1st screening experiment were not adjusted before loading onto the gel. 1 Band corresponded to MW of 2.5 kDa was not observed. 114 Table 18. Molecular weight calculations of protein bands in lanes 3 - 8 of Figure 21. Non-reducing SDS-PAGE MW Lane Fraction Band# (kDa) Lane Reducing SDS-PAGE Fraction Band # MW (kDa) 3 FT 1 8.8 4 FT 1 9.1 (1st 2 11 (1st 2 12 screening 3 18 screening 3 17 experiment) 4 23 experiment) 4 24 5 28 5 29 5 0.10 M 2' 13 6 0.10 M r 8.2 (1st 3' 16 (1st 2' 13 screening 4' 20 screening 3' 16 experiment) 5' 24 experiment) 4' 18 6' 26 5' 24 7' 29 6' 26 7' 29 7 Combined I 12 8 Combined II 18 FT II 18 FT (3rd III 23 (3rd screening screening experiment) experiment) Band numbers in bold represent darker bands observed from the gel. Lane 1. Pharmacia PMW marker (M.W. range 2.5- 17 kDa)1 2. MW Marker (BSA 66 kDa, ovalbumin 45 kDa, 3-lactoglobulin 18 kDa) 3. MW marker (IgG 25 kDa & 55 kDa) 4. Non-reducing SDS-PAGE of the combined 0.25 M fraction from the 3rd screening experiment (2x diluted) 5. Non-reducing SDS-PAGE of the combined 0.25 M fraction from the 3rd screening experiment (5x diluted) 6. Non-reducing SDS-PAGE of the 0.20 M fraction from the 1st screening experiment 7. Reducing SDS-PAGE of the combined 0.25 M fraction from the 3rd screening experiment (5x diluted) 5. Reducing SDS-PAGE of the 0.20 M fraction from the 1st screening experiment Figure 22. Reducing and non-reducing SDS-PAGE of the major fraction from 3rd screening experiment as compared to the major fraction from the 1st screening experiment on high density PhastGel®. Protein concentrations of the fractions from the 1st screening experiment were not adjusted before loading onto the gel. 1 Band corresponded to MW of 2.5 kDa was not observed. 116 Table 19. Molecular weight calculations of protein bands in lanes 4-8 of Figure 22. Non-reducing SDS-PAGE Reducing SDS-PAGE Lane Fraction Band # MW (kDa) Lane Fraction Band # MW (kDa) 4 Combined 1 10 7 Combined 1 11 0.25 M 2 12 0.25 M 2 12 (3rd 3 24 (3rd 3 25 screening 5 35 screening 4 28 experiment) 6 44 experiment) 5 38 7 53 7 53 8 88 5 Combined 1 11 0.25 M 2 12 (3rd 3 25 screening 5 36 experiment) 6 50 7 58 8 88 6 0.20 M 1 11 5 0.20 M 1 11 (1st 2 12 (1st 2 13 screening 3 27 screening 3 26 experiment) 4 30 experiment) 4 29 5 39 5 41 6 49 7 54 7 59 8 89 Band numbers in bold represent darker bands observed from the gel. Molecular weights of the different components of the 0.20 M fraction calculated from the high density PhastGel® (Figure 22 and Table 19) were slightly different from those calculated from the 10-15 gradient PhastGel® (Figures 16 and 17 and Tables 16 and 17). In general, similar profiles were obtained using the two types of gels, but better resolution of the lower versus higher molecular weight components were observed for the higher density versus 10-15 gradient PhastGel®. The lower molecular weight components of the 0.20 M fraction from the 1st screening experiment and the combined 0.25 M fraction in the 3rd screening experiment appeared to be composed of at least two bands with molecular weight of around 10-11 kDa and 12-13 kDa. Results from the 3rd screening experiment showed that elution with fewer step gradient was feasible. The fraction eluted in a single step of 0.25 M NaCl had comparable electrophoretic profiles as the major 0.20 M fraction eluted with the three-step gradient elution of 0.15, 0.20 and 0.25 M NaCl. The fractions unbound to the column equilibrated with 25 mM Tris and the fraction eluted at 0.10 M NaCl in 0.10 M Tris at pH 8.6 had different electrophoretic profiles but were eluted together in one step to speed up the fractionation procedure. The DEAE-chromatographic pattern from the 50 ml column from the 3rd screening experiment showed that the capacity of the DEAE-column packed was not large enough for a larger-scale isolation of the major fraction of flaxseed proteins. As a result, a larger column was packed. 5.3.4 Scale-up fractionation of flaxseed proteins Figure 23 shows a typical elution profile of the fractions eluted from a 225 ml DEAE-column. The fraction eluted at 0.25 M NaCl in 0.10 M Tris at pH 8.6 (also called the 118 "0.25 M fraction" for simplicity) contributed 63.7% total protein while the fraction not bound to the column equilibrated with 0.10 M NaCl in 0.10 M Tris at pH 8.6 (also called the "FT fraction" for simplicity) accounted for 35.9% protein. Figures 24a, b and c show the ultraviolet absorption (UV) spectra of the three fractions collected. The FT fraction had a broader absorption maximum than the 0.25 M fraction. The maximum for the FT fraction was between 275-280 nm while the maximum for the 0.25 M fraction was at 280 nm. Both fractions had absorption minima at around 255 nm. A small shoulder was observed for the UV spectrum of the FT fraction at 290 nm. Madhusudhan and Singh (1985d) also reported a maximum at 280 nm and a shoulder at 290 nm for the small molecular weight 1.6S protein they isolated. In addition, they reported a maximum at 280 nm and a minimum at 250 nm for the major 12 S fraction (Madhusudhan and Singh, 1985c). Both the first two fractions (FT and 0.25 M) had UV absorption spectra typical of proteins, with maxima around 280 nm contributed by the tryptophan and tyrosine in the proteins and minima around 255 nm. However, the later fraction (0.50 M) was distinctively different since it had no observable maximum or minimum in the 250-350 nm range. A total of six scale-up fractionations were carried out and 147 ml of the major fraction at a concentration of 8.7 mg/ml was collected. 119 0.25M - 45.0 40.0 35.0 30.0 - 25.0 - 20.0 15.0 10.0 5.0 0.0 500 1000 1500 Elution volume (ml) 2000 E o E, o 3 TJ C o o Figure 23. Typical DEAE-Sephacel chromatographic pattern of flaxseed proteins (225ml column). This chromatogram was obtained from the 2nd scale-up fractionation batch. The column was equilibrated with 0.10 M NaCl in 0.10 M Tris at pH 8.6. The numbers "0.10 M".. . refer to the concentration of NaCl used for step gradient elution. 56 ml of protein extract prepared from 8.69 g dehulled delipidated NorMan flaxseed was loaded onto the column. 120 2.000 1.500 1.000 A B S 0.500 0.000 250.0 280.0 300.0 320.0 WAVELENGTH 350.0 Figure 24a. Ultraviolet absorption spectrum of the peak fraction from the FT fraction of flaxseed protein from a 225 ml DEAE-column at 0.10 M NaCl. 2.762 2.500 2.000 A B S 1.500 1.000 0.500 0.060 250.0 280.0 300.0 320.0 WAVELENGTH 350.0 Figure 24b. Ultraviolet absorption spectrum of the peak fraction from the 0.25 M fraction eluted from a 225 ml DEAE-column at 0.25 M NaCl. 121 3.905 3.000 2.000 A B S 1.000 -0.074 200 250.0 300.0 WAVELENGTH 350.0 400.0 Figure 24c. Ultraviolet absorption spectrum of the peak fraction from the 0.50 M fraction eluted from a 225 ml DEAE-column at 0.50 M NaCl. 122 5.4 CONCLUSIONS FOR PHASE 1 Protocols established for isolation of cadmium-binding proteins have been modified and simplified to isolate the major fraction of flaxseed protein. Flaxseed proteins were extracted from dehulled delipidated NorMan flaxseed with a 16-hour extraction with 0.10 M NaCl in 0.10 M Tris at pH 8.6. The residual hulls in the mechanically dehulled seeds significantly affected the clarity of the protein extracts and therefore, the hulls that remained were manually removed until no observable hulls were found in all subsequent extractions. Although the addition of 2-mercaptoethanol in the extraction buffer increased the extraction yield by increasing the volume of supernatant recovered upon centrifugation, it was not used in subsequent extractions to reduce the possibility of changes in the properties of the proteins in the major fraction that might result from reduction of disulfide linkages by 2-mercaptoethanol. Flaxseed proteins were fractionated by DEAE anion-exchange chromatography. Fractions not bound to the column equilibrated at 25 mM Tris and eluted at 0.10 M NaCl in 0.10 M Tris at pH 8.6 consisted of lower molecular weight components with respective major bands of 18 and 20 kDa observed by SDS-PAGE. Fractions eluted at 0.15, 0.20 and 0.25 M NaCl in 0.10 M Tris at pH 8.6 had very similar electrophoretic profiles, with major protein bands of 23-24, 25-27 and 30-33 kDa observed under reducing SDS-PAGE and 22, 31, 44-45 and 51 kDa observed under non-reducing SDS-PAGE. The decrease in the intensities of higher molecular weight protein bands accompanied by the increase in the intensities of the lower molecular weight protein bands indicated the presence of disulfide linkages in the protein fractions. The amino acid compositions of the two low-molecular weight fractions were 123 comparable to those reported for the low-molecular weight flaxseed/linseed proteins in the literature. The 0.20 and 0.25 M fractions had very similar amino acid compositions and their values were comparable to those reported for the high-molecular weight flaxseed/linseed proteins in the literature. The finalized scale-up fractionation procedure involved a step gradient elution with three different NaCl concentrations: 0.10, 0.25 and 0.50 M NaCl in 0.10 M Tris at pH 8.6. The low-molecular weight fraction remained unbound to the column equilibrated with 0.10 M NaCl in 0.10 M Tris at pH 8.6 while the high-molecular weight major fraction eluted at 0.25 M NaCl in 0.10 M Tris at pH 8.6. The major fraction accounted for 63.7 % of the total proteins fractionated. A total of 147 ml of the major fraction at a concentration of 8.7 mg/ml was isolated from 6 fractionation batches for further characterization in Phase 2. 124 CHAPTER 6—RESULTS AND DISCUSSION FOR PHASE 2: CHARACTERIZATION OF THE MAJOR FRACTION OF FLAXSEED PROTEINS The characteristics of the major fractions eluted at 0.25 M NaCl in 0.10 M Tris at pH 8.6 (also called the "0.25 M fraction") collected from 6 scale-up fractionation batches described in Chapter 5 are described in this chapter. The terms "0.25 M fraction" and the "major fraction" are used interchangeably. 6.1 G E L E L E C T R O P H O R E T I C P A T T E R N S The uniformity of the major fractions collected from the 6 scale-up fractionation batches was assessed by Native-and SDS-PAGE. Figure 25 shows the Native PAGE profiles of the six 0.25 M fractions. The major band of the major fraction accounted for an average of 93.4 + 3.8 percent of the total intensities of the bands per lane (n = 6). There were again some materials that did not enter the gel and remained at the point of sample application and in the boundary between the stacking and the resolving gel. The relative intensities of the materials that remained at the point of sample application and the boundary between the stacking gel and resolving gel were 2.0 +1.3 and 0.8 + 0.4 percents respectively (n = 6). The bands in lane 6 were not included in the calculation of the relative intensities because it had a remarkably high amount of materials that did not enter the resolving gel and had unusual slanted bands when compared with protein bands of the other batches. All six batches showed a major component and two minor components. Native PAGE profiles of the separate 0.15 M, 0.20 M and 0.25 M fractions as well as the whole extract also showed 125 the presence of the two minor bands (Figures 14 and 18 in Chapter 5). The two minor components might be impurities or could be either dissociated or aggregated forms of the subunits comprising the major band. The relative intensities of the minor bands above and below the major band were 1.0 + 0.3 and 3.9 + 1.1 percents respectively (n = 6). Marcone et al. (1998) also reported the presence of a minor protein band of slightly higher relative electrophoretic mobility in peanut, buckwheat, alfalfa, caraway, cumin, amaranth, and soybean globulins, probably due to electrostatic repulsion between subunits of non-covalently linked protein subunits. Lane 1. I s scale-up batch from fresh extract (F25-1) 2. 2 n d scale-up batch from fresh extract (F25-2) 3. 3 r d scale-up batch from overload fraction (R25-1) m $ 4. 4 t h scale-up batch from overload fraction (R25-2) 5. 5 t h scale-up batch from overload fraction (R25-3) 6. 6 t h scale-up batch from fresh extract (F25-3)) Lane \L 2 3 4 5 6 7 8y 7. Mixture of the 6 batches 8. Whole protein fraction Figure 25. Native PAGE of the 0.25 M fractions from the 6 scale-up fractionation batches on 8-25 gradient PhastGel®. 1 Code "F" stands for "Fresh" and "R" stands for "Reload". "25" stands for the 0.25 M fraction. Number after the hyphen designates the batch number. 126 Figure 26 shows the reducing SDS-PAGE profiles of the six 0.25 M fractions and Table 20 shows the corresponding molecular weight calculations for the bands in the fractions. The estimated molecular weights of the components of the 0.25 M fractions were similar to those from the preliminary experiments (Figures 16 and 17; Tables 16 and 17 in Chapter 5). A total of 12 bands were observed, with predominant bands corresponding to molecular weights of 20 + 1, 26 + 2 and 31 + 1 kDa. Oomah and Mazza (1998) also observed the presence of 14, 24, 25, 34 and 51 kDa polypeptides in flaxseed meal products, with the bands at 14, 24 and 34 kDa being predominant. Marcone (1999) also observed 5 components from reducing SDS-PAGE with molecular weights of 14.4, 24.6, 30.0, 35.2 and 50.9 kDa, with the components at 24.6 and 35.3 kDa corresponded to the acidic and basic subunits, respectively. Figure 27 shows the non-reducing SDS-PAGE profiles of the 0.25 M fractions prepared from fresh flaxseed protein extract as compared to the one prepared from the overloaded 0.25 M fraction (described in Section 3.4.3) and Table 21 shows the corresponding molecular weight calculations for the bands in the fractions. A total of 15 bands were observed and the bands corresponding to approximate molecular weights of 40 and 47-48 kDa were the most intense. The intensities of some bands were weak, making it difficult to distinguish between separate component or simply part of an adjacent band. 127 Lane M W (kDa) _ 66 55 45 25 18 s s * m m • a 1. M W markers (BSA 66k, Ovalbumin 45k, P-lactoglobulin 18k, IgG 25k & 55k) 2. 1st scale-up batch from fresh extract (F25-1) B a n d * J. 2n d scale-up batch from fresh extract (F25-2) 12,13, 4. 3r d scale-up batch from overload fraction (R25-1) 14 5. 4 t h scale-up batch from overload fraction (R25-2) 5 ,6,7, 6. 5 t h scale-up batch from overload fraction (R25-3) 8,9 7. 6 t h scale-up batch from fresh extract (F25-3) 1,2,3, 8. Mixture of the 6 batches Lane 1 7 8 Figure 26. Reducing SDS-PAGE of the 0.25 M fractions from the 6 scale-up fractionation batches on 8-25 gradient PhastGel®. Sample codes were defined in Figure 25. Table 20. Molecular weight calculations of protein bands in lanes 2, 6, 7, and 8 of Figure 26. Band1 M W 2 (kDa) 1* 9 + 0.5 2* 10 ± 0 3* 14 + 2 4* 17 ± 1 5 20 ± 1 6 23 ± 2 7 26 ± 2 8 31 ± 1 9 34 ± 1 12 47 ± 6 13 61 ± 2 14 69 ± 2 * MW of the bands were calculated from the extrapolation of the standard curve Band numbers in bold represent darker bands observed from the gel. 1 Band numbers are common for Figures 26 & 27, Tables 20 & 21. 2 MW are reported as means + standard deviations of lanes 2, 6, 7, 8. 128 *"5" — N Lane; MW Band # (kDa) 7. M W marker ( B S A 66k, Ovalbumin 45k, P-lactoglobulin 18k, IgG 25k & 55k ) 2. 0 . 1 0 M fraction (FT) from fresh extract 3. 0 . 2 5 M fraction (R25-1) (2x diluted) 4. 0 . 2 5 M fraction (R25-1) (4x diluted) 66 \ 13,14,15,16 5. 0 . 2 5 M fraction (F25-1) (2x diluted) 45 T 4fe - S <fg S S 5 2 j - 10,11,12,13 6. 0 . 2 5 M fraction (F25-1) (4x diluted) 2 5 4 ? 5^  6 ; 7 ; g ; 9 7. Whole protein fraction ( lOx diluted) 1 8 - 1 5. M W marker ( B S A 66k, Ovalbumin 45k, K 1,2,3 **—• • —«* J P-lactoglobulin 18k, IgG 25k & 55k) La\e 1 2 3 4 5 6 7 8 Figure 27. Non-reducing SDS-PAGE of the FT, 0.25 M fractions and whole extract on 8-25 gradient Phastgel®. Sample codes were defined in Figure 25. Table 21. Molecular weight calculations of protein bands in lanes 3 and 5 of Figure 27. Lane 3 Lane 5 Band 1 MW(kDa) Band 1 MW(kDa) 1* 10 1* 10 2* 11 2* 11 3* 14 3* 15 4 18 4 18 5 21 5 21 6 23 6 24 8 30 8 30 9 34 9 35 10 40 10 40 11 43 11 45 12 47 12 48 13 60 13 62 14* 68 14* 70 15* 84 15* 88 16* 95 16* 99 * MW of the bands were calculated from the extrapolation of the standard curve. Band numbers in bold represent darker bands observed from the gel. 1 Band numbers are common for Figures 26 & 27, Tables 20 & 21. 129 A change in the relative intensity of different bands was observed upon comparison of the PAGE profiles of the 0.25 M fraction under reducing (Figure 26) and non-reducing (Figure 27) conditions. For instance, bands 10, 11, 15, 16 corresponding to estimated molecular weight ranges of 40, 43-45, 84-88 and 95-99 kDa under non-reducing conditions (Figure 27) were missing under reducing conditions (Figure 26). Bands 9 and 12, which corresponded to the estimated molecular weight ranges of 34-35 and 47-48 kDa, became less intense under reducing condition (Figure 26). On the other hand, bands 5 and 8 with molecular weights of 20 + 1 and 31 + 1 kDa became more intense under reducing conditions. Band 7 with an estimated molecular weight of 26 + 2 was not found under non-reducing conditions but appeared as a very intense band under reducing conditions. The shifts in the intensities of bands from relatively higher molecular weight components in the non-reducing conditions to lower ones in the reducing conditions suggested the disruption of disulfide-linked higher molecular weight components of the fraction into smaller ones by reducing agents. Similar decreased in intensities or disappearance of higher molecular components with the increase in intensities in lower molecular components was also reported for rapeseed globulin and was common in the storage proteins of various seeds (Dalgalarrondo et al., 1986). Based on the molecular weights calculated for the proteins bands of the major fraction under reducing- and non-reducing- SDS-PAGE, components that were disulfide linked could be speculated. For instance, bands 9 and 10 (MW of 34 + 1 and 40 kDa) observed under non-reducing conditions might be formed by a dimer of bands 4 or 5 (MW of 17 + 1 or 20 + 1 kDa) observed under reducing conditions. Bands 11 and 12 (MW of 43-45 and 47-48 kDa) might be respectively formed by linkage of heterogeneous components of bands 5 and 6 (MW of 20 ± 1 and 23 + 2 kDa), and bands 6 and 7 (MW of 23 ± 2 or 26 + 2 kDa). The higher molecular 130 weight components might also be formed by a combination of the smaller components. For instance, band 15 (MW around 86 kDa) could be formed by the combination of the two dimers of bands 4 and 7 (MW 17 ± 1 and 26 ± 2 kDa). Band 16 (MW around 97 kDa) might be formed by the combination of a dimer of 23 + 2 kDa (band 6) and a monomer of 47 + 6 kDa (band 7), or a combination of a dimer of 17 + 1 kDa (band 4), a monomer of 14 + 2 kDa (band 3) and a monomer of 47 + 6 kDa (band 12). Madhusudhan and Singh (1985c) also ran SDS-PAGE of the 12S protein that they isolated (a brief summary of the isolation procedure was outlined in Figure 1 in Section 2.2.2.1). They found at least five non-identical subunits with molecular weights of 11, 18, 29, 42 and 61 kDa, with the 18 and 42 kDa bands being more intense. It was however, very difficult to directly compare their results with those obtained in this experiment. 10 % homogenous gel was used in their study while 8-25 % gradient pre-made gel was used in this experiment. Also, conditions for electrophoretic separation were not specified. Dev and Sienkiewicz (1987) carried out SDS-PAGE on the total protein, the crude and purified 1 IS globulin (a brief summary on the production of the three was found in the literature review, Figure 2 Section 2.2.2.1), with 8.7 % polyacrylamide using 0.5 % SDS-0.05 M Tris-Glycine (pH 8.3) buffer system. They found nine bands in the total protein samples, of which four were distinctly predominant over the others. The total number of bands obtained in the crude and purified globulin was not reported, however, two and three predominant bands were found respectively. They did not report the molecular weight of the bands on the gel but they noted that "the crude and purified globulin was an oligomeric protein with the one most predominant component of low molecular mass". They did not define what was considered 131 "low" and "high" molecular mass and therefore, it was difficult to compare their results with the findings in this experiment. The high number of protein bands found in the electrophoretic patterns of the major fraction is not uncommon among oilseeds. Kapoor and Gupta (1977) found 15 bands for the whole protein of soybean and 11 bands for the 1 IS glycinin. They also noted that Larsen observed 21 bands for the whole protein of soybeans in 1967 and Smith et al. observed 14 bands in soybean globulins in 1955 (Kapoor and Gupta, 1977). Kapoor and Gupta did not specify whether their electrophoretic patterns were obtained under reducing or non-reducing conditions. In the SDS-PAGE analysis of the crude US soybean glycinin, Wolf and Nelsen (1996) found at least 12 bands. Dalgalarrondo et al. (1986) studied the subunit compositions of rapeseed 12S globulin and observed three groups of bands in the MW range of 22-24, 29-31 and 35-37 and 52-60 kDa under non-reducing SDS. They named the bands with 29-31 and 35-37 kDa the same group but in two sub-classes. Each group consisted of at least two to three bands. 6.2 SOLUBILITY AND ISOELECTRIC POINTS Solubility of the whole extract, major fraction and FT fraction as a function of pH was first expressed in terms of A28o- However, because of the turbidity of most of the samples of the major fraction and whole extract at different pHs and the presence of insoluble particulates in some samples, the plot of A2so against pH was not very reliable in describing the true solubility. In addition, different absorptivities at 280 nm between the whole extract ( E 1 % i c m = 10.1 in 1 M NaCl; Madhusudhan and Singh, 1983), major fraction ( E 1 % i c m = 7.6 in phosphate 132 buffer; Madhusudhan and Singh, 1985c) and FT fraction (not reported) made it difficult to directly compare the solubility of the three based on A280. Figure 28 shows the solubility profiles for the whole extract, major fraction and FT fraction as a function of pH, expressed as percentage protein soluble in supernatant, with protein concentration determined by the BCA Protein Assay. 0 -I , 1 1 1 1 1 1 1 — 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 p H Figure 28. Solubility profiles of the FT fraction, major fraction and whole protein extract from NorMan flaxseeds at different pHs and two NaCl concentrations. 133 6.2.1 Effect of p H and low ionic strength (0.01 M NaCl) The region of minimal solubility for the major fraction occurred at pH 4.0-5.5. Early researchers already observed such a wide pH range of minimum solubility for linseed meal (Sosulski and Bakal, 1969). Smith et al. (1946) and Painter and Nesbitt (1946) respectively reported the minimum nitrogen extractability of linseed meal in water as 3.8-5.1 and 3.5-4.0. Madhusudhan and Singh (1983) also found that the nitrogen solubility profile of linseed meal proteins in water exhibited a typical U-shape with minimum solubility in the pH range of 3-6 and the solubility remained more or less constant at pH above 8.0. Dev et al. (1986a) found similar results in that the nitrogen solubility of linseed proteins increased on both sides of the isoelectric range of protein at 4.0-4.6. Wanasundara and Shahidi (1994a) observed that the minimum solubility of nitrogenous compounds in hexane-extracted and methanol-ammonia-water/hexane-extracted meals occurred between pH 3.0-3.5. The pH range of minimal solubility also corresponded with the results of the isolectric focusing of the major fraction. The isoelectric points of the major fraction in 5 mM Tris at pH 8.6 were 4.7 ± 0.0, 5.1 ± 0.0 and 5.6 ± 0.1 (n =4; Figure 29 and Table 22). Vassel and Nesbitt (1945) also reported a pi at 4.75 for the 12S protein they isolated. The thick and dark band at the place of sample application might have been caused by the protein aggregates which could not migrate through the gel (Figure 29). 134 ; — — 7. 0.25 M fraction (F25-1) 2. 0.25 M fraction (R25-2) 3. FT fraction from the I s screening experiment Site of 4. Sigma IEF Marker (pi 3.6-9.3) sample application *? ~ ~ 5. Sigma IEF Marker (pi 3.6-9.3) 3 " 3 6. 0.25 M fraction (F25-3) ^ 2 i i 7. Combined 0.25 M fraction from the 6 batches 8. Whole protein extract m *" * Lane ^ 2 3 4 5 6 7 8 J Figure 29. IEF profiles for the 0.25 M fractions and whole extract on IEF 3-9 PhastGel . Code "F" stands for "Fresh" and "R" stands for "Reload". "25" stands for the 0.25 M fraction. Number after the hyphen designates the batch number. Table 22. Calculation of pi of the protein bands in lanes 1, 2, 6, 7, 8 of Figure 29. Sample Band Pi 0.25 M fraction1 1 4.7 ± 0.0 2 5.1 ± 0.0 3 5.6 ± 0.1 Whole extract 1 4.7 2 5.0 3 5.6 'pis for 0.25 M fraction are reported as the means + standard deviations of lanes 1, 2, 6 and 7. 135 No band was observed within the separation range of pH 3-9 in the IEF gel for the FT fraction from the 1st screening experiment. Results from the Native PAGE of the FT fraction (Figure 14 and 18) also suggested that the fraction had pi greater than 8.5. Based on the results from Native PAGE and IEF, the FT fraction should have pi higher than 9. Madhusudhan and Singh (1985d) noted that the low-molecular weight fraction of linseed was basic in nature but they did not discuss any detailed findings on it. The pi of the whole extract was similar to that of the major fraction as expected because the major fraction accounted for at least 60 % of the proteins of the whole extract and the pi for the FT fraction which accounted for at least 30 % of the total protein was out of the determinable range. At pH near the isoelectric point, the net charge was minimal and thus electrostatic repulsive forces between protein molecules preventing the hydrophobic regions of proteins from interacting with each other was also minimal. The hydrophobic interactions of proteins resulted in insolubility. Solubility increased as the pH increased or decreased from the pis. About 90 % of protein was soluble at pH 3.0 and 7.0. About 20 % of protein in the major fraction was still soluble at the region of minimal solubility. Madhusudhan and Singh (1983) also found 22-24 % of the total nitrogen was soluble at the point of minimum solubility irrespective of the solvents used. Smith et al. (1946) also found that 21-23 % of the total nitrogen was soluble at the region of low nitrogen extractability. Painter and Nesbitt (1946) reported approximately 20 % of the total non-protein nitrogen was not precipitable with trichloroacetic acid. Madhusudhan 136 and Singh (1983) suggested that the presence of non-protein nitrogen or low-molecular weight proteins might be the reason for the soluble nitrogen at region of minimum solubility. The 20 % soluble protein at regions of minimal solubility in this experiment was not contributed by the non-protein nitrogen because the percentage protein was determined by the BCA Protein Assay instead of total nitrogen combustion method. The BCA Protein Assay is based on the reaction of bicinchoninic acid with cuprous ion, where the cuprous ions are formed upon reduction of cupric ions under alkaline conditions by peptide bonds of proteins and peptides containing at least two peptide bonds. The whole extract had solubility profile similar to that of the major fraction and exhibited minimal solubility between pH range of 4.0-5.5. Close to 50 % of protein remained soluble in this pH range. The difference of the % soluble protein between the whole extract and the major fraction at regions of minimal solubility could be explained by the presence of soluble FT fraction in the whole extract but not in the major fraction. The FT fraction was found to be soluble under low salt concentrations throughout the pH range studied in this experiment. 6.2.2 Effect of p H and high ionic strength (1.0 M NaCl) Addition of NaCl shifted and broadened the regions of minimum solubility of both the whole extract and the major fraction to pH 4 and below. This agrees with the findings of Madhusudhan and Singh (1983) where the pH of minimum solubility of the linseed meal 137 proteins shifted to the acidic side (pH 0.5-4.5) in the presence of 0.5 M and 1.0 M NaCl. They suggested that binding of anions by the proteins might be the cause of the shift. Dev et al. (1986a) also observed the widening of the isoelectric range of linseed meal proteins and shifting towards the more acidic side in the presence of NaCl; but solubility increased sharply with increasing pH beyond 4. Solubility of both the whole extract and major fraction in this experiment also increased at pH higher than 4.0. Solubility of the major fraction was less than 10 % below pH 3.0 while that of the whole extract was around 30^ 10 % at or below pH 3.0. The FT fraction was highly soluble (around 90 % soluble) in 1.0 M NaCl solution irrespective of the pH. 6.3 SURFACE HYDROPHOBICITY Figure 30 shows the surface hydrophobicity (S0), measured by the PRODAN fluorescence probe method, of the major fraction and whole extract at three pHs and two NaCl concentrations. The major fraction was significantly more hydrophobic than the whole fraction at pH 5 and 7 in both 0.01 M and 1.0 M NaCl (p < 0.05). 138 450 400 350 300 I 250 o 200 (/) 150 100 50 0 • Ave So 0.01 M NaCl • Ave So 1.0 M NaCl 5 PH Figure 30a. Surface Hydrophobicity 1 of the major fraction measured at pH 3, 5, 7 in citrate-phosphate buffer with 0.01 M or 1.0 M NaCl. Bars with different letters (a-f) represent significant (p < 0.05) difference in SQ across the two figures (30a and b). Figure 30b. Surface Hydrophobicity 1 of the whole extract measured at pH 3, 5, 7 in citrate-phosphate buffer with 0.01 M or 1.0 M NaCl. Bars with different letters (a-f) represent significant (p < 0.05) difference in Sp across the two figures (30 a and b). 1 So values of all samples reported are the average of two replicate analyses except for the major fraction in 0.01 M NaCl at pH 7.0, which involved the average of four replicates. 139 NaCl concentration and pH significantly affected surface hydrophobicity of the major fraction (p < 0.05). At the same pH, the surface hydrophobicity of the major fraction in 1.0 M NaCl was significantly higher than that in 0.01 M NaCl (p < 0.05, except for pH 3 at p < 0.10). Surface hydrophobicity of the major fraction was highest at pH 7 in 1.0 M NaCl and lowest at pH 3 in 0.01 M NaCl. Surface hydrophobicity of the whole extract in 1.0 M NaCl at all three pHs was not significantly different from each other (p < 0.05). Surface hydrophobicity at the same pH but different NaCl concentration was not significantly different except for pH 3, which showed a significantly lower surface hydrophobicity in 0.01 M than 1.0 M NaCl (p < 0.05). Cheng (2001) found that RFI values of PRODAN were significantly lower at acidic than at neutral or alkaline pH and suggested that the dampening effect of acidic pH on the RFI values obtained with PRODAN probe should be considered when comparing S0 values and interpreting the effects of pH on protein surface hydrophobicity. On the other hand, Cheng (2001) also noted that the relative magnitude of the solvent acidity effect on PRODAN emission was small in comparison to the changes in S0 in proteins as a function of pHs. For example, the lactoferrin studied by Cheng (2001) had significantly higher S0 at pH 3 as compared to pH 7 and 9. In this context, the low surface hydrophobicity of the major fraction at pH 3 might be slightly underestimated by PRODAN but the fraction itself exhibited a lower surface hydrophobicity at that pH. The whole extract also showed no difference in surface hydrophobicity between pH 3, 5 and 7 in 1.0 M NaCl, indicating that the magnitude of solvent acidic effect on PRODAN was small. 140 6 .4 AMINO ACID COMPOSITIONS AND TOTAL HYDROPHOBICITY VALUES (H<))): Table 23 shows the amino acid composition of the major fraction as compared to the reported values from literature. Similar to other storage proteins reported in the literature, the major fraction of flaxseed proteins had high content of arginine, glutamate (and/or glutamine) and aspartate (and/or asparagine). Such an amino acid composition with high nitrogen content was important for supplying nitrogen for germination (Youle and Huang, 1981). Protein sources rich in arginine have recently gained popularity because of its potential preventative functions against heart disease (Pszczola, 2000). Food sources rich in glutamine are also receiving great interest because of its potential functions of supporting the immune system and improving athletic performance (O'Carrol, 1995; Blenford, 1996) Table 24 shows the amino acid composition of the whole extract as compared to the reported literature values. The amino acid composition of the flaxseed/linseed varied greatly across research groups. The difference in the variety of seeds used and the method of preparation might be partly account for the different reported amino acid compositions. The whole extract had higher contents of lysine and possibly cysteine (although cysteine was not derviatized before analysis, the values for cysteine detected was higher in the whole extract than in the major fraction). The higher contents of lysine and cysteine were most likely contributed by the lower molecular weight fractions in the whole extract (as indicated by the amino acid compositions of the low molecular weight fractions in Table 14, Section 5.3.1.4). The lysine/arginine ratio, a determinant of the cholesterolemic and artherogenic effects of a protein (Oomah and Mazza, 2000), were 0.27 for the major fraction and 0.34 for the whole extract, compared to 0.37 for flaxseed protein, and 0.88 for either soybean or canola proteins 141 reported by Oomah and Mazza (2000). Both the major fraction and the whole extract were less lipidemic and artherogenic than soybean and canola proteins. Table 23. Amino acid compositions (g/ lOOg protein)1 of the major protein fraction from NorMan flaxseed as compared with the reported values. Amino Acid D e v & Madhusudhan Sienkiewicz, & Singh, 1985 1987 Flaxseed 12S Salt-soluble protein protein fraction Marcone et al., 1998 Flaxseed globulin Garcia et al, 1997 Soybean U S glycinin Wolf & Nelsen, 1996 Soybean U S glycinin Gruener & Ismond, 1997 Canola 12S globulin Major fraction' Aspartic acid 3 D 11.30 8.28 12.4 13.9 12.65 8.48 12 34 + 0 09 Glutamic acid 3 E 19.80 24.50 24.3 25.1 15.49 18.99 21 82 + 0 94 Serine S 5.10 3.98 3.1 6.5 5.30 5.04 4 58 + 1 64 Glycine G 4.80 5.55 5.4 5.0 7.73 6.01 5 61 + 0 69 Histidine H 2.50 3.53 2.4 2.6 1.80 3.27 2 52 + 0 12 Arginine R 12.50 10.2 12.6 8.9 5.51. 8.31 11 88 + 0 16 Threonine T 3.90 3.08 3.6 4.1 3.71 3.43 3 07 + 0 86 Alanine A 4.80 3.83 5.5 4.0 5.55 4.24 5 67 + 0 66 Proline P 4.50 4.09 0.0 6.9 6.17 8.29 4 17 + 0 15 Tyrosine Y 2.30 3.16 2.4 4.5 2.77 2.91 2 44 + 0 24 Valine V 5.60 4.32 5.1 4.9 5.72 3.87 4 65 + 0 99 Methionine M 1.70 1.89 1.3 1.3 1.56 1.76 1 25 + 0 33 Cysteine C 1.40 n .a 4 0.9 1.7 0.70 1.78 0 58 + 0 14 Isoleucine I 4.60 4.15 5.6 4.9 4.61 3.42 4 55 + 0 38 Leucine L 5.80 4.54 5.9 8.1 7.04 8.18 5 83 + 0 44 Phenylalanine F 5.90 4.30 6.3 5.5 4.30 5.22 5 81 + 0 41 Lysine K 3.10 4.37 3.1 5.7 4.23 4.66 3 24 + 0 12 1 Tryptophan was not analyzed. Cysteine was not derivatized before analysis. 2 Values are reported as means + standard deviations of three analyses. 3 Asparagine and glutamine were quantitatively converted to aspartic and glutamic acids respectively. n.a = not available. 142 Table 24. Amino acid compositions (g/ lOOg protein)1 of the whole protein extract from NorMan flaxseed as compared with the reported values. Amino Acid Mad-, . husudhan Cherdkiatgumchai, „ ~. , * & Singh, Bhatty & 1990 Lab Corn-prepared mercial meal meals 1985a Linseed meal Wanasundara & Shahidi, 1994b Hexane-extracted Methanol ammonia extracted Com-mercial meal Klockeman et al, 1997 Whole extract2 soy bean rapeseed meal Aspartic acid 3 D 12.50 12.40 9.40 9.18 9.16 8.03 12.8 7.82 10.32± 0.12 Glutamatic acid 3 E 26.30 26.40 18.00 16.70 16.36 14.45 20.47 17.75 21.50± 0.59 Serine S 5.80 5.90 6.40. 4.94 4.99 4.48 5.60 4.81 6.36± 0.54 Glycine G 7.00 7.10 11.50 6.44 6.26 5.64 4.57 5.25 10.85 + 0.98 Histidine H 2.90 3.10 1.90 2.69 2.46 2.36 2.77 2.57 1.78 + 0.04 Arginine R 11.80 11.10 8.30 11.50 11.20 9.78 7.91 5.38 8.38± 0.37 Threonine T 4.90 5.10 4.30 3.40 3.30 3.00 4.40 4.54 3.93 + 0.14 Alanine A 5.40 5.50 6.90 4.81 4.64 4.61 4.66 4.71 6.86± 0.22 Proline P 5.20 5.50 4.90 3.64 3.65 3.32 6.01 5.82 3.94± 0.34 Tyrosine Y 2.90 3.10 1.90 2.21 2.12 1.98 3.43 2.88 1.70 + 0.25 Valine V 5.60 5.60 6.90 5.75 5.64 5.02 5.25 3.66 4.71 ± 0.80 Methionine M 2.20 2.50 1.30 1.45 1.41 1.24 1.38 1.74 1.31 + 0.51 Cysteine C 3.80 4.30 1.80 3.29 3.39 3.16 1.45 0.77 1.37 + 0.05 Isoleucine I 5.20 5.00 4.40 4.78 4.54 4.19 4.97 2.82 4.16± 0.40 Leucine L 6.80 7.10 6.30 6.70 6.39 5.96 7.81 7.22 5.86± 0.43 Phenylalanine F 5.30 5.30 4.00 5.13 4.91 4.63 5.41 3.84 4.07 ± 0.48 Lysine K 4.10 4.30 3.80 4.38 4.14 3.92 6.98 5.62 3.44± 0.39 Tryptophan was not analyzed. Cysteine was not derivatized before analysis. Values are reported as means + standard deviations of three analyses. Asparagine and glutamine were quantitatively converted to aspartic and glutamic acids respectively. 143 The total hydrophobicity values (H<])) calculated based on the amino acid compositions for the major fraction and whole extract were 652.3 + 54.8 and 613.8 + 33.3 kcal/ 100 g protein respectively (n = 3). These calculated values might underestimate the true H(|) values because cysteine was not derivatized before analysis and tryptophan was not included in the analysis. As a result, these values were not applicable for comparison with other literature values but could be used for comparison of the H<|) values of different samples (consistently analyzed without derivatization of cysteine and consistently not including tryptophan in the analysis) in this thesis. The H(|) for the major fractions was slightly higher but was not significantly different from that of the whole extract (p > 0.10). The H(|) of the major fraction and the whole extract were both higher than the "FT" and "0.10 M fraction" in the 1st screening experiment (491.31 and 609.1 respectively). 6.5 SS and SH Table 25 shows the reactive SH, total SH, total SH + SS and SS values for the major fraction and the whole extract. The number of reactive and total SH groups determined by this assay should be viewed with caution due to the high protein blank readings arising from the turbidity of samples in the testing buffer. Although these blank readings were taken into consideration in the calibration of netA4i2 and SH content, greater variability or error in the results might be anticipated. No significant difference were found in the number of reactive and total SH of the major fraction and the whole extract (p > 0.10). The whole extract, however, had significantly higher content of SS as compared with the major fraction (p < 0.001). 144 Table 25. Sulfhydryl groups and disulfide bonds of the major fraction and whole protein extract from NorMan flaxseeds. Sample Reactive SH 1 Total SH 1 Total SH + SS 2 SS 2 ' 3 umole/ pmole/ umole/ pinole/ g protein g protein g protein g protein 0.25 M fraction 0.618±0.09 a 3.40± 0.06 a' 64.8± 2.1 x 61.4± 2.1 x' whole extract 0.708±0.08 a 3.25± 0.10 a' 124± 1 y 121 ± 1 y' 1 Values reported are means + standard deviations of two replicates. 2 Values reported are means + standard deviations of three replicates. 3 SS = Total SH + SS - Total SH. a, a' , x, y, x', y ' y a i u e s j n m e s a m e column with different superscripts are significantly different at p < 0.05. The flaxseed protein extract and major fraction were characterized by high number of SS and low number of SH. The high SS content compared to SH is common among storage proteins of many dicotyledonous seeds, which have common structures referred to as "legumin-like" consisting of heavy and light disulfide-linked polypeptides (Inquello et al. 1993). The total SH and SS contents of soy US protein reviewed by Peng et al. (1984) were 15.84 and 104.64 pmole/g protein, respectively. The major fraction of flaxseed proteins had lower SH and SS contents than those of the soy 1 IS protein. 6.6 CYSTEINE CONTENT The cysteine contents of the major fraction and whole extract calculated based on the SS and SH content were 1.53 + 0.05 and 2.93 + 0.02 g/100 g proteins respectively. The cysteine content of the major fraction was close to that reported by Madhusudhan & Singh (1985c). 145 6.7 F O A M I N G Foaming properties of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 7 were compared with that of the whole extract to see if the isolated fraction would have superior or inferior functional properties compared to the original protein. The foaming properties of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 7 were also compared with those of egg albumen, BSA, P-lactoglobulin and ovalbumin in the same buffer conditions to determine how well would the major fraction serve as a foaming agent. Foaming properties of the major fraction in citrate-phosphate buffer in low (0.01 M) and high (1.0 M) NaCl concentrations at pH 3, 5, 7 were also studied to investigate the effect of NaCl concentrations and pHs on the foaming properties of the major fraction. Because of the lack of standardized conditions used for foam generation, a series of preliminary experiments was carried out to obtain repeatable and reproducible conditions for foam generation. In order to objectively describe the dynamic nature of foams, three sets of experiments were carried out, namely the foam volume-drainage set, foam volume-conductivity set and the photographic illustration set. 6.7.1 Preliminary experiments and experimental set up Slower flow rate and a longer sparge time were chosen over the initial higher ones at 90, 100, 110 cm3/min and shorter sparge times of 15, 20, 30 seconds because of the difficulty in controlling the experimental conditions at higher gas flow rates and shorter sparging times. Fluctuation of flow rate as indicated by the flow meter was higher at higher flow rate. The short sparging time made it very difficult to adjust the flow rate back to the proper range because the sparging time might have been over before corrections were 146 made. The combination of high fluctuation and short sparging time resulted in higher batch-to-batch errors and thus lower reproducibility and repeatability. Waniska and Kinsella (1979) found that at higher gas flow rates, more liquid was initially entrapped in the foam and less gas was lost during sparging. The large amount of liquid initially in the foam due to the inadequate time for the foam lamellae to reach equilibrium with the liquid would lead to an increase in drainage. On the other hand, at lower gas flow rate, foam had longer periods to drain and thus had lower amount of liquid initially trapped in the foam and lower drainage. However, the gas lost during sparging was higher. They suggested a flow rate of 20 ml/min as a compromise between the two effects. The selection of optimum foam generating conditions could be an interesting study that would require more in-depth investigation and optimization. The final conditions used in this experiment were selected based on results and observations from experiments with different combinations of flow rate and sparging time. Fluctuation of flow rate was lower at low flow rate. The flow rate of 35 ml/min for 1 minute was chosen because of the ease of control of experimental conditions, and the ease of calculation and standardization at 1 minute. By observations, sparging of 35 ml/min for 1 minute was sufficient for most of the sample to go into foam. In the preliminary trials, samples were first placed in the column before gas was sparged. The liquid sample placed on top of the glass frit created pressure and thus reduced the flow rate of the nitrogen being sparged. This undesirable fluctuation in flow rate was later 147 overcome by first applying nitrogen gas to obtain a steady flow rate before introducing the samples from above. 6.7.2 Description of foams Set 1—Foam volume and drainage: The measurement of foam volume as the sole means of describing the foaming power and foam stability was inadequate. The within-sample variability of the foam volume was high compared to sample-to-sample variability. Foams of different proteins disintegrated differently. Foams of ovalbumin and whole fraction disintegrated with very noticeable change in bubble sizes and foam heights, in which the bubbles were completely destroyed in a short time. The foam of BSA was more stable than those of ovalbumin and whole fraction, but the foam collapsed with a noticeable change in foam height. Foams of egg albumen and the major fraction at pH 7 showed a less noticeable change in height. It was very difficult to record the exact volume of foam that disintegrated over time. For instance, foams of egg albumen and the major fraction generally disintegrated partially while retaining the overall structure and height of the foam. In some cases, foams close to the bottom of the column collapsed while those at the top or middle sections of the column were retained. The change in foam volume over time was more noticeable for all foams when the drainage was collected in a graduated cylinder rather than leaving the drained liquid 148 in the column as in the set measuring solely foam volume. The continuous removal of liquid created a driving force to drain the liquid trapped in the foam and dragged the whole foam matrix downwards. Nevertheless, the variation of foam volume of replicates of the same sample was still substantial. When drainage was measured by the volume of liquid that dripped from the column, through the glass frit and filter, past the opened flow control knob, down the narrow glass opening and into the graduated cylinder, there was about 1 ml of dead-volume of drainage for all the protein samples. This 1 ml accounted for a considerable proportion of the total 5 ml, limiting this method in distinguishing the stability of different foams. In addition, the drainage rate was also seriously affected by the presence of tiny air bubbles which blocked the narrow glass opening, causing an overestimation of stability. The integrity of the G-4 glass filter also affected the drainage rate. A slight crack on the thin glass filter might result in increased drainage and thus an underestimation of stability. With all the complications and uncontrolled factors, the drainage method was not very reproducible. The variation of the foam capacity and drainage within samples was so great that it was difficult to compare differences across samples. Kato et al. (1983) also noted that the foaming properties determined from foam volume do not necessarily reflect the quality factors of foams, such as the thickness and size, and it was difficult to accurately measure a decrease in foam volume and the rate of fluid leakage from foam. 149 Set 2—Foam volume and conductivity measurement: (1) Conductivity measurement using the YSI conductivity probe: Figure 31 shows the foam conductivity decay curve of the whole fraction, egg albumen, BSA, p-lactoglobulin and ovalbumin in citrate-phosphate buffer with 0.01 M NaCl at pH 7. Conductivity of the protein foams decreased as the foam disintegrated over time. The observed "poor foamers" of the whole extract and ovalbumin showed sharper decrease in conductivity compared to the better foamers. The stability of the foams based on the conductivity decay curves was in the following order: egg albumen > P-lactoglobulin > BSA > ovalbumin and whole extract. This trend was similar to the foam stability index reported by Kato et al. (1983), in which the stability of P-lactoglobulin > BSA > ovalbumin. The conductivity method was generally a more effective way of representing foam disintegration than foam volume and drainage. Unfortunately, no conductivity reading was obtained for the dense foam of the major fraction because the foam accumulated at the outer end of the probe and was unable to enter the bridge. Figure 32 is a simplified illustration of the YSI conductivity probe. The narrow distance between the outer glass protective cover and the conductivity bridge, and the long distance between the opening of the probe and the bridge might be the main hindrance to the foam reaching the bridge. 150 •*— Whole extract • Egg albumen A BSA X Ovalbumin H— beta-lactoglobulin • • •Pf ' i 1 1 1 1 1 r "1 1 1 r 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 Time (min) Figure 31. Foam conductivity decay curves for measured using the YSI conductivity probe. Samples of the whole flaxseed protein extract and egg albumen, BSA, ovalbumin and |3-lactoglobulin were at concentration of 1 mg/ml in citrate-phosphate buffer with 0.01 M NaCl at pH 7. Conductivity of each sample is an average of three replicates. 1.0 cm Conductivity probe Fixed glass cover protecting the probe 0.2 cm t 2.5 cm ^—1.3 cm—• Figure 32. Simplified illustration of the YSI conductivity probe (YSI Model 31, Yellow springs, Ohio). The two darker rings are the concentric electrodes. Samples must cover both rings of electrode for readings. 151 (2) Difference in conductivity readings between two conductivity meters: Table 26 shows the conductivity of the buffers used in the foaming experiment measured with two probes and meters. Conductivity values of the same sample differed when measured with two different probes and meters. The range of conductivity measured by the YSI probe and meter was lower. The Radiometer conductivity probe and meter was more sensitive and accurate in measuring the conductivity of samples (as shown by the readings of the KC1 standards in Table 26). The pH effect on conductivity contributed by the different concentrations of buffer salts was more apparent in buffers with 0.01 M NaCl but was minimal in buffers with 1.0 M NaCl. Although the conductivity of the foams measured by the two conductivity meters were different, they showed the same trends in the foam conductivity decay curves for good and poor foamers. Table 26. Conductivity of buffers used in the foaming test measured by two different conductivity probes and meters. Conductivity measured (mS/cm) Buffer (citrate-phosphate) YSI conductivity probe , ^^iometer f r . / • jr conductivity probe and conditions and meter ; r2 meter 0.01 M NaCl at pH 3 3.64 3.72 0.01 M NaCl at pH 5 6.40 6.90 0.01 M NaCl at pH 7 9.50 10.9 1.0 M NaCl at pH 3 62.0 76.0 1.0 M NaCl at pH 5 62.0 77.0 1.0 M NaCl at pH 7 62.0 76.0 0.01 M K C 1 3 1.34 1.41 1 YSI Model 31, Yellow springs, Ohio. 2 Copenhagen Type CDC 104; Radiometer Analytical, Lyon, France. 3 0.01 M KC1 standards should have a conductivity reading of 1.41 mS/cm (Anonymous, 1992). 152 (3) Conductivity measurement using the Radiometer conductivity probe: Figure 33 shows the foam conductivity decay curve for protein samples in citrate-phosphate buffer with 0.01 M NaCl at pH 7, measured by the Radiometer conductivity probe. The whole extract and ovalbumin were again found to have a steeper decay curve compared to other proteins, similar to the results obtained using the YSI conductivity probe, while the major fraction had slightly higher conductivity and a similar decay curve compared to egg albumen and BSA. Figure 33. Foam conductivity decay curves measured using the Radiometer conductivity probe. Samples of whole flaxseed protein extract, the major fraction, egg albumen, BSA and ovalbumin were at concentrations of 1 mg/ml in citrate-phosphate buffer with 0.01 M NaCl at pH 7. Conductivity of each sample is an average of two replicates except for the major fraction with five replicates. 153 Although the design of the Radiometer conductivity probe made it a better candidate for measuring the conductivity of most samples, the three narrow constrictions on the probe still limited some foam from entering the electrodes. Figure 34 is a simplified illustration of the Radiometer showing the three constrictions which hindered some protein foams from entering the probe. Fixed glass cover protecting the probe Conductivity probe Narrow constrictions Figure 34. Simplified illustration of the Radiometer conductivity probe (Copenhagen Type CDC 104; Radiometer Analytical, Lyon, France). The three darker rings are the concentric electrodes. Samples must cover all three rings of electrode for accurate readings. 154 The fine and stiff foam of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 3 did not move up to the readable region of the probe. In a few replicates, the foam of this sample slowly moved up to the readable region of the probe, bringing an increase in conductivity reading over time (Figure 35). Once in the probe, the foam of this fraction was caught in the narrow constrictions and stably stayed inside for a long time while the foam in other regions of the column disintegrated. This led to an overestimation of foam stability. In another replicate, the foam of this fraction did not reach the readable region of the probe and resulted in an underestimation of foam stability. There was no obvious trend in the conductivity values measured in the replicate trials, due to the variability in how the foam of this fraction moved up to the readable region and whether it was immobilized inside the probe (Figure 35). Foams of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 7 also had minor problems moving up to the electrode as well as being immobilized inside the probe, as indicated by the greater variability in the conductivity decay curve (Figure 36) as compared to other proteins at the same pH (Figure 37). Foams of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 5 were comparable to those of the whole extract and ovalbumin as indicated by its steep foam conductivity decay curves (Figure 36). 155 600 500 S «» 400 *f 3 0 0 '•S u -§ 200 c 5 100 A A • • A A Mil > A A 4A o y*-0.00 1 A --A A A A A A 4A M ! ! ! ! 1 1 1 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 Time (min) Figure 35. Foam conductivity decay curve for the major fraction at concentrations of 1 mg/ml, in citrate-phosphate buffer with 0.01 M NaCl at pH 3. Conductivity was measured using the Radiometer conductivity probe. Replicate analyses are denoted by the same symbol in different shading intensities. 320 280 40 0. I o°o °c# • • 6 ^ 1 0 6 GO 00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 T i m e (min) Figure 36. Foam conductivity decay curve for the major fraction at concentrations of 1 mg/ml, in citrate-phosphate buffer with 0.01 M NaCl at pH 5 (•) and pH 7 (•). Conductivity of was measured using the Radiometer conductivity probe. Replicate analyses are denoted by the same symbol in different shading intensities. 156 320 280 1240 3 200 = 160 f l 2 0 10 O 80 U 40 0 H > - E g g l - • " E g g 2 -B— whole 1 - •— whole 2 O Ovalbumin 1 6 Ovalbumin 2 A B S A 1 A BSA 2 2 $ 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 Time (min) 16.00 18.00 20.00 22.00 24.00 Figure 37. Foam conductivity decay curve for egg albumen, whole extract, ovalbumin and BSA, at concentrations of 1 mg/ml, in citrate-phosphate buffer with 0.01 M NaCl at pH 7. Conductivity was measured using the Radiometer conductivity probe. Results of duplicate analyses of each sample are denoted by "1" and "2". (4) Interference of conductivity probes on foaming properties: In general, the presence of the conductivity probe in the centre of the column increased the mechanical resistance as the foam moved up the column. The presence of the probe also hindered the formation of a continuous uniform matrix of foam in the column. Part of the foam was destroyed as it hit the probe while another part of the foam was forced to foam around the probe. The thicker the probe, the more interference during foam formation. On the other hand, the thick conductivity probe placed at the centre of the column appeared to act as a support preventing the foam of some proteins from collapsing. 157 (5) Foaming Capacity (FC): The measurement of foam height and the calculation of foam capacity by including the volume of gas input broadly distinguished poor foamers from the better ones. Table 27 summarizes the foaming capacity of different protein samples. The foam of the major fraction in citrate-phosphate buffer with 1.0 M NaCl at pH 3 collapsed during its formation and hence foaming capacity was not measured. The foaming capacity of the whole fraction and ovalbumin were significantly lower than those of other samples (p < 0.001). Foaming capacity of BSA, egg albumin and major fraction at all the pHs and NaCl concentrations showed no significant difference. Table 27. Foaming capacity (FC) of the whole extract, egg albumen, BSA, ovalbumin and the major fraction at concentrations of 1 mg/ml. Sample Buffer (citrate-phosphate) Foaming capacity (FC)1 Whole 0.01 M NaCl, pH 7 1.20 ± 0.00 a Egg Albumen 0.01 M NaCl, pH 7 1.64 ± 0.18 " BSA 0.01 M NaCl, pH 7 1.70 ± 0.01 b Ovalbumin 0.01 M NaCl, pH 7 1.29 ± 0.04 a 0.25M fraction 0.01 M NaCl, pH7 1.70 ± 0.03 " 0.25M fraction 0.01 M NaCl, pH5 1.65 ± 0.08 " 0.25M fraction 0.01 M NaCl, pH 3 1.78 ± 0.06 b 0.25M fraction 1.0MNaCl,pH7 1.90 ± 0.04 b 0.25M fraction 1.0MNaCl,pH5 1.80 ± 0.02 b 0.25M fraction 1.0MNaCl,pH3 not measurable 1 FC values are reported as means + standard deviations of 2 replicates for all samples except for the major fraction in 0.01 M NaCl at pH 3 and 7 with 5 replicates. a " b Values in the same column with different superscripts are significantly different at p < 0.05. 158 (6) Relative Foam Conductivity (Cf): Table 28 shows the relative foam conductivity (Cf) of all the samples. The relative foam conductivity is an index of the foaming power of different samples (Wilde and Clark, 1996). Table 28. Relative foam conductivity (Cf) 1 of the whole extract, egg albumen, BSA, ovalbumin and the major fraction at concentrations of 1 mg/ml. Sample Buffer (citrate-phosphate) Cf Whole 0.01 M NaCl pH7 1.11 ± 0.40 % a x a' Egg Albumen 0.01 M NaCl pH 7 1.60 ± 0.07 % a xy a' BSA 0.01 M NaCl pH 7 1.94 ± 0.00 % a xyz a' Ovalbumin 0.01 M NaCl pH 7 1.46 ± 0.00 % a xy a' Major fraction 0.01 M NaCl pH 7 2.42 ± 0.59 % a yz b' Major fraction 0.01 M NaCl pH 5 3.06 ± 0.10 % b z Major fraction 0.01 M NaCl pH 3 9.39 ± 4.64 % b N.A. Major fraction 1.0 M NaCl pH7 1.94 ± 0.18 % a xyz Major fraction 1.0 M NaCl pH5 2.15 ± 0.10 % a xyz Major fraction 1.0 M NaCl pH3 2.18 ± 0.10 % a xyz 1 Cf values are reported as the means + standard deviations of 2 replicates for all samples except for the major fraction in 0.01 M NaCl at pH 3 and 7 with 5 replicates. a-b; a'-b' Values in the same column with different superscripts are significantly different at p < 0.05. x " z Values in the same column with different superscripts are significantly different at p < 0.05. N.A.= not applicable. The results for the sample of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 3 were not included in ANOVA and Tukey's comparison test. 159 The major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 3 had the highest relative foam conductivity among the samples studied, followed by the same fraction at pH 5. These two samples had significantly higher Cf when compared with other samples (p < 0.05). The rest of the samples were not significantly different (column coded a, b in Table 28) but the Cf of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 7 was the third highest, followed by the same fraction in 1.0 M NaCl pH 3, 5, 7, BSA, ovalbumin, egg albumen and whole extract. In a separate statistical analysis comparing only the Cf of the proteins in citrate-phosphate buffer with 0.01 M NaCl at pH 7 (the first five rows, column with superscripts a'-b' in Table 28), the major fraction had significantly higher (p < 0.05) Cf than BSA, ovalbumin, egg albumen and whole extract. In another separate statistical analysis, the Cf of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 3 was not considered due to the high standard deviation of this fraction, which might reduce power of the analysis for detecting the differences among other samples. This ANOVA were still unable to detect a significant difference of the Cf among samples except that the Cf of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 5 was significantly higher (p < 0.05) than those of BSA, egg albumen and whole extract. The Cf of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 7 was significantly higher (p < 0.05) than that of the whole extract (column coded x, y, z in Table 28). 160 (7) Foam Stability Index (FSI): Table 29 shows the foam stability indices of protein samples. Foam stability index for the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 3 was not determined because of the low reproducibility and repeatability (described in section (3) above). Table 29. Foam stability index (FSI) of the whole extract, egg albumen, BSA, ovalbumin and the major fraction at concentrations of 1 mg/ml. Sample Buffer (citrate-phosphate) FSI = C 0 * AC/At (uS/cm)1 Whole extract 0.01 M NaCl pH 7 3.7 + 1.0 a Egg Albumen 0.01 M NaCl pH7 35.2 + 1.1 d BSA 0.01 M NaCl pH7 19.8 + 2.1 b Ovalbumin 0.01 M NaCl pH 7 3.9 + 0.1 a Major fraction 0.01 M NaCl pH 7 25.2 + 2.9 be Major fraction 0.01 M NaCl pH 5 4.9 + 0.9 a Major fraction 0.01 M NaCl pH 3 N. D . 2 Major fraction 1.0MNaClpH7 32.4 + 5.7 cd Major fraction 1.0MNaClpH5 6.7 + 0.1 a Major fraction 1.0 M NaCl pH3 5.8 + 1.1 a 1 FSI values reported are the means + standard deviations of 2 replicates except for the major fraction in 0.01 M NaCl at pH 7 with 5 replicates. 2 N.D.= not determined (See text for explanation). a " d Values in the same column with different superscripts are significantly different at p < 0.001. 161 From Table 29, it is observed that egg albumen in 0.01 M NaCl at pH 7 and the major fraction in 1.0 M NaCl at pH 7 had significantly higher foam stability indices than the rest of the samples (p < 0.001). The major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 7 was also relatively stable compared to other samples. Foams of the whole extract and ovalbumin were the least stable. Comparing the FSI of all the proteins in citrate-phosphate buffer with 0.01 M NaCl at pH 7 (the first five rows in Table 29), the foam of egg albumen was significantly more stable than those of the other samples at the same pH (p < 0.001). Stabilities of the foams of major fraction and BSA were not significantly different but were significantly higher than those of the whole extract and the ovalbumin (p < 0.001). Comparing the FSI of the major fraction at the three pHs and two NaCl concentrations (the last six rows in Table 29), the foam of major fraction at pH 7 in both high and low NaCl concentrations were significantly more stable than the foam of the same sample at other pHs and NaCl concentrations. (8) Interpretation of results from the conductivity set: Each of the parameters FC, Cf or FSI on their own could not fully describe the foam properties. For instance, the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 5 had very good foaming power as indicated by the high relative foam conductivity and yet its foam disintegrated in a very short period of time as indicated by the poor foam stability index. The major fraction in citrate-phosphate buffer with 1.0 M NaCl at pH 3 and 5 had relatively higher Cf than at pH 7 and yet those foams 162 were not as stable as that at pH 7, as indicated by the lower FSI values. The Cf of the major fraction in citrate-phosphate buffer with 1.0 M NaCl at pH 3 was higher (but not significantly) than the whole extract and ovalbumin but the FC of the fraction in 1.0 M NaCl at pH 3 was low to a point that it was not measurable because the foam disintegrated as it was being formed. Based on the results from the three parameters FC, Cf or FSI and the findings noted by observations, the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 3 and 7 and in citrate-phosphate buffer with 1.0 M NaCl at pH 7, and egg albumen had good foaming power as well as foaming stability. Set 3—Photographic illustration of foams: Figure 38 shows a comparison of the foam of a "good foamer", the major fraction, with a "poor foamer", the whole extract, in 0.01 M NaCl at pH 7. The "good" and "poor" foamers had distinctive differences observed in the foam matrices. 163 Figure 38. Comparison of the foam of the whole protein extract (left) with the major protein fraction (right) from NorMan flaxseed after sparging of nitrogen gas at a flow rate of 35 ml/min into 5 ml of samples at 1 mg/ml for 1 minute. 164 Figures 39, 40, 41 respectively show the foams of the major fraction at pH 7, pH 3 and egg albumen in citrate-phosphate buffer with 0.01 M NaCl at pH 7 in the column at 3 min. The three "good foamers" were characterized by very fine foams. The sizes of the bubbles were generally larger toward the bottom of the column. Figures 42, 43, 44 show the same foams in the column after 25-28 min. The foams tended to disintegrate from the top and bottom edges into the middle section of the column. Larger bubble sizes were observed as the foam deteriorated. Figure 45 compares the change in the same foams in the column after longer periods of time (55-128 min). The foam of the major fraction at pH 3 was the most stable among the three. The foam of egg albumen started out with very fine foams but the foam was not as stable as the major fraction at pH 3. Figures 46 and 47 respectively show the disintegration of the foams of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 3 and 7 placed on slides. The bubbles of the foam of the major fraction of flaxseed in citrate-phosphate buffer with 0.01 M NaCl at pH 7 merged together and formed large bubbles as they disintegrated (Figure 46). On the other hand, foam of the major fraction in the same buffer at pH 3 did not grow in size but simply broke apart as they disintegrated (Figure 47). 165 Figure 39. Foam of the major fraction of NorMan flaxseed at 1 mg/ml in citrate-phosphate buffer with 0.01 M NaCl at pH 7 in the column at 3 min (Sparging of gas started at 0 min and stopped at 1 min). 166 Figure 40. Foam of the major fraction of NorMan flaxseed at 1 mg/ml in citrate-phosphate buffer with 0.01 M NaCl at pH 3 in the column at 3 min (Sparging of gas started at 0 min and stopped at 1 min). Figure 41. Foam of egg albumen at 1 mg/ml in citrate-phosphate buffer with 0.01 M NaCl at pH 7 in the column at 3 min (Sparging of gas started at 0 min and stopped at 1 min). 168 Figure 42. Foam of the major protein fraction of NorMan flaxseeds at 1 mg/ml in citrate-phosphate buffer with 0.01 M NaCl at pH 7 in the column at 28 min (Sparging of gas started at 0 min and stopped at 1 min). 169 Figure 43. Foam of the major protein fraction of NorMan flaxseed at 1 mg/ml in citrate-phosphate buffer with 0.01 M NaCl at pH 3 in the column at 26 min (Sparging of gas started at 0 min and stopped at 1 min). 170 Figure 44. Foam of egg albumen at 1 mg/ml in citrate-phosphate buffer with 0.01 M NaCl at pH 7 in the column at 25 min (Sparging of gas started at 0 min and stopped at 1 min). 171 Flaxseed major fraction pH 7; 65 min Flaxseed major fraction pH 3; 55 min Egg albumen pH 7; 65 min Flaxseed major fraction pH 7; 128 min Flaxseed major fraction pH 3 ; 128 min Egg albumen pH 7; 100 min Figure 45. Deterioration over time of foams of the major protein fraction from NorMan flaxseed, at 1 mg/ml in citrate-phosphate buffer with 0.01M NaCl at pH 7 or 3 , as compared to that of the egg albumen (Sparging of gas started at 0 min and stopped at 1 min). 7.0 min 12.5 min Figure 46. Deterioration over time of foam of the major fraction, at 1 mg/ml in citrate-phosphate buffer with 0.01M NaCl at pH 7, scooped out of the column and placed on a slide (Time when sparging of gas started was used as time 0; Sparging ended at 1 min). 173 9.0 min 15.0 min 12.5 min Figure 47. Deterioration over time of foam of the major fraction, at 1 mg/ml in citrate-phosphate buffer with 0.01M NaCl at pH 3, scooped out of the column and placed on a slide (Time when sparging of gas started was used as time 0; Sparging ended at 1 min). 174 6.7.3 Interpretation of results from different foaming tests The observed properties of foams were different depending on the way in which they were formed and the methods of analysis. For set 1 (foam volume and drainage), the foaming power and stability was lower when the foam was drained with the flow control knob opened (drain liquid collected in a separate measuring cylinder) than when the foam was drained with the flow control knob closed (drain liquid remained in the column). Removal of liquid forced the liquid trapped in the foam to drain continuously. In addition, the foam in the drainage set with drained liquid collected in a separate measuring cylinder was drier compared to those of the other two sets because the liquid in the foam was drained out. For set 2 (foam volume and conductivity measurement), the conductivity probe placed in the centre of the column hindered foam formation and the foam was forced around the probe. The resistance from the wall and the conductivity probe made the formation of uniform and stable foam matrix difficult. As a result, the integrity and stability of the foam formed and measured in this set was the worst among the three. For set 3 (photographic illustration of foams), with pictures of the foam taken inside the column, the foaming power and foam stability was highest because the foam generated in the column was allowed to stand undisturbed as the foam gradually disintegrated. For the foam that was scooped out from the column and placed onto a slide, the foam matrix was disturbed or even destroyed by the mere action of 175 scooping and isolating a small part of the foam from the whole foam matrix. In addition, the small amount of foam placed on a slide was vulnerable to evaporation in a short time. The difference in the degree of physical and mechanical disturbance applied to the foam in different methods resulted in apparent differences in the measured foaming properties of the protein foam of the same sample. For instance, the foam of a sample at 8 min in the photographic set might still be in good shape while the foam of the same sample in the drainage set and conductivity set might have been totally or partially destroyed. It is therefore important to interpret results from each set of foaming tests individually and compare the general trends in each test to give an overall picture of the foaming properties of the foams of different samples. Based on the results from the three sets, the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 3 had the best foaming capacity and stability. 6.8 V ISCOSITY Table 30 shows the apparent viscosities of the whole extract, the major fraction, egg albumen, BSA and ovalbumin in citrate-phosphate buffer with 0.01 M NaCl or 1.0 M NaCl at pH 7. The apparent viscosities of the major fraction in 0.01 M NaCl or 1.0 M NaCl at pH 3 or 5 were also measured. Viscosity of this set of samples was measured to investigate whether viscosity of the sample solution contributed to the difference in foaming properties observed in the previous section. 176 Table 30. Apparent viscosity of the protein samples at concentrations of 1.0 mg/ml compared to water. Sample Buffer Apparent Viscosity (mPa .s) Water Whole extract1 Egg albumen 3 B S A 4 Ovalbumin 4 Major fraction 4 Major fraction 5 Major fraction 4 Major fraction 3 Major fraction 6 Major fraction 1 citrate-phosphate citrate-phosphate citrate-phosphate citrate-phosphate citrate-phosphate citrate-phosphate citrate-phosphate citrate-phosphate citrate-phosphate citrate-phosphate buffer with buffer with buffer with buffer with buffer with buffer with buffer with buffer with buffer with buffer with 0.01 M NaCl, pH 7 0.01 M NaCl, pH 7 0.01 M NaCl, pH7 0.01 M NaCl, pH 7 0.01 M NaCl, pH 7 0.01 M NaCl, pH 5 0.01 M NaCl, pH3 1.0 M NaCl, pH7 1.0 M NaCl, pH5 1.0 M NaCl, pH3 480 + 20 870 + 150 370 + 90 1100 + 140 620 + 70 600 + 110 730 + 56 600 + 60 820 + 40 720 + 80 700 + 30 b b vx be yz b v cd z be y be xy 1 Values are means + standard deviations of 9 replicates. 2 Values are means + standard deviations of 13 replicates. 3 Values are means + standard deviations of 11 replicates. 4 Values are means + standard deviations of 7 replicates. 5 Values are means + standard deviations of 10 replicates. 6 Values are means + standard deviations of 8 replicates. a - e ; v-z y a j u e s m t n e s a m e c o i u m n with different superscripts are significantly different at p < 0.001. The major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 5 and also in buffer with 1.0 M NaCl at pH 3 and 5 were not soluble. The presence of insoluble particulates might affect the viscosity measured. Shen (1981) commented that viscosity of proteins is highly sensitive to the history of the protein and the techniques used to measure it. BSA had the highest apparent viscosity among all the samples, followed by the whole extract and the major fraction in 1.0 M NaCl at pH 7. The apparent viscosities of the major fractions 177 at the three pHs and two NaCl concentrations were not significantly different from one another except it was higher in 1.0 M NaCl at pH 7 (p < 0.001). The apparent viscosities of ovalbumin and the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 7 were not significantly different. Apparent viscosity of egg albumen was not significantly different from water. In a separate statistical analysis comparing only the apparent viscosities of the major fraction at the three pHs and two NaCl concentrations, the major fraction in 0.01 M NaCl at pH 3 and 7 had significantly lower apparent viscosity than those of the same fraction at other conditions (column with superscript v-z in Table 30). The apparent viscosity of the major fraction in 1.0 M NaCl at pH 3 was not significantly higher than that of the same fraction in 0.01 M NaCl at pH 7 but was significantly higher than the one in 0.01 M NaCl at pH 3. The observed apparent viscosities of the samples were extremely high. For instance, distilled deionized water, which should have a viscosity of 1 mPa .s (Barnes et al, 1989), had a viscosity measured to be 480 + 20 mPa .s. This great discrepancy between the theoretical and measured viscosity readings suggested that the presence of wall-effect on the viscometer might be seriously affecting the measurement, in which the viscosity measured was caused mainly by the friction between the solution and the wall of the container, rather then the internal fraction between liquid. Viscosity of a Newtonian viscosity standard supplied by the manufacturer, which should have a viscosity of 4.93 mPa -s, was therefore measured using the same combination of spindle (18) and rotational speed (0.6 rpm) to investigate the prevalence of wall-effect. The 4.93 mPa .s standard had apparent viscosity measurements of 275 ± 6 mPa .s (n=3). Upon increasing the rotational speed, the apparent viscosity obtained after adjustment 178 with the factor corresponding to the spindle and speed factor decreased. The apparent shear-thinning effect observed for a supposingly Newtonian viscosity standard showed that the wall-effect was substantially affecting the viscosity measurement. The viscosities determined in this study should be viewed with caution due to an overestimation of viscosity caused by the wall-effect. On the other hand, as viscosity of each sample was consistently determined using the same spindle and rotational speed, with dial readings taken consistently after five complete revolutions, the difference in apparent viscosities determined should reflect the trends in apparent viscosities among samples. Such a general trend showing the difference among samples were adequate for the purpose of investigating how solution viscosities correlate with foaming properties. 6.9 REGRESSION ANALYSES OF THE FUNCTIONAL PROPERTIES OF THE MAJOR FRACTION Regression analyses were employed to investigate the relationships among different characteristics of the major fraction studied in this thesis and how they were related to the functional properties observed. 6.9.1 Factors Affecting Solubility Environmental factors such as pH, ionic strength, temperature, solvent composition and the presence of other food components may affect protein solubility. In addition, physical, chemical and thermal treatments during processing, method of isolation, interaction with other food components, and mechanical treatments prior to solubility testing will also influence the degree of protein solubility (Vojdani, 1996). 179 A quadratic regression model for solubility of the major fraction with pH, NaCl, S0, reactive SH, total SH and SS as parameters showed that the 2nd order term of pH and the interaction of pH with NaCl, as well as S0 and its interactions with pH, NaCl, reactive SH and total SH were significant factors affecting solubility (P < 0.001 for regression model; R2adj = 99.7 %; p < 0.05 for 3 predictors; p < 0.10 for 4 predictors;). Details of the regression models of solubility of the major fraction can be found in Appendix A. A stepwise (forward and backward) regression on solubility showed that the 2nd order term of pH was the most significant factor affecting solubility (p < 0.05; R adj = 18.04 %). A separate stepwise regression on solubility with pH and NaCl in every step showed that the interaction of S0 with NaCl and pH, together with the 2nd order term of total SH and SS, were the significant factors affecting solubility after pH and NaCl (p < 0.05; R 2 a d j = 94.49 %). pH and NaCl were included in each step because they were the main treatments on the samples. 6.9.1.1 Hydrophobicity and solubility: The quadratic regression model for solubility of the major fraction (Appendix A) showed that surface hydrophobicity was an important factor determining solubility through its direct effect on solubility and its interaction with pH, NaCl, reactive and total SH (p < 0.1 except for its interaction with pH and NaCl at p < 0.05). On the other hand, the total hydrophobicity (H<j)) calculated based on amino acid composition had no significant effect on solubility. Vojdani (1996) also noted that the proportion and distribution of surface hydrophilic and hydrophobic patches are the main factors in determining the degree of solubility of protein, rather than the total hydrophobicity and charge density based on amino acid composition. 180 Nakai and Powrie (1981) noted that protein solubility generally decrease as hydrophobicity increase. The negative coefficient of S0 in the regression equation of the present study also suggested that solubility decreased as surface hydrophobicity increased. The major fraction had significantly higher surface hydrophobicity than the whole extract for most of the NaCl and pH conditions used in this study while it also had lower solubility throughout the pH range study. The trend of higher hydrophobicity with lower solubility, however, was not as simple because surface hydrophobicity also combined with NaCl and pH effects in its effect on solubility (Appendix A). The presence of NaCl generally increases surface hydrophobicity and leads to decrease of solubility due to the facilitation of hydrophobic interactions by NaCl (Damodaran and Kinsella, 1982). On the other hand, the presence of NaCl also increases solubility through its salting-in action. The changes in pH alter the net charge of the protein and the overall balance of electrostatic and hydrophobic interactions which affect solubility. Hayakawa and Nakai (1985) suggested that the exposed aromatic amino acids might play a more important role than aliphatic amino acids in the protein insolubility. Nakai (1983) commented that hydrophobicity is not the only structural factor that determines the functional behaviour of a protein. 6.9.1.2 Charge and solubility: The net charge effect can be usually shown as pH effect (protein solubility as a function of pH) (Townsend and Nakai, 1983). Based on the results from the quadratic 181 regression model for solubility of the major fraction (Appendix A), pH was an important factor determining solubility through its interactions with S0 and NaCl (p < 0.05). The 1st stepwise regression model for solubility (Appendix A) also showed that the pH had a 2n d order effect on solubility (p < 0.05). The net charge of protein varies as pH changes. At the isoelectric pH of proteins, net charge preventing hydrophobic aggregates from forming was minimal and thus solubility decreased due to the formation of aggregates. In the present study, solubility of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 5 was also lowest among the same sample at the three pHs studied. 6.9.2 Factors affecting viscosity Sample concentration, composition of samples, dispersion conditions, pH, ionic strength, and processing conditions all affect viscosity of a sample (Schenz and Morr, 1996). Quadratic regression analysis on the viscosity of the major fraction showed that no single factor had a significant effect on viscosity and yet the regression model was significant (P = 0.002; R 2 adj = 80.8 %). Details of regression analysis on viscosity of the major fraction can be found in Appendix B. All the factors contributed to the viscosity of the major fraction but no single factor dominated in their effects on viscosity. A stepwise (forward and backward) regression analysis for viscosity of the major fraction showed that the interaction of pH and NaCl was the most significant factor affecting viscosity, followed by the 2nd order term of solubility, and NaCl (p < 0.001; R 2 a dj = 84.34 %). Results from a separate stepwise regression analysis with NaCl and pH in every step showed that the interaction between solubility and pH, as well as the interaction between pH and NaCl, 182 were significantly affecting viscosity of the major fraction (p < 0.001 except for NaCl p < 0.1; R 2 a d j = 84.17%). Shen (1981) investigated the relationship of solubility and viscosity in soy protein isolates and found a general trend of viscosity being inversely correlated with solubility. Schenz and Morr (1996), however, commented that the concentrations and physicochemical properties of other ionic and non-ionic solutes also exerted an important influence on the viscosity of protein solutions and tended to confound the underlying inverse relationship of protein solubility and viscosity. Shen (1981) also noted that viscosity was not necessarily correlated to the solubility because many factors such as conformation, hydration, exposure of hydrophobic groups, charge distribution, would all contribute to the intermolecular interactions that affect viscosity. In this study, viscosity was measured mainly to investigate its correlation with the foaming properties of samples and therefore, viscosity was not measured in different sample concentrations and shear rates as in other more in-depth studies on viscosity. 6.9.3 Factors affecting foaming properties The structure, surface hydrophobicity, charges, pi and flexibility all contribute to the properties of a protein-stabilized foam. Knowledge of these properties and the effect of environmental conditions or treatments can often explain changes in foaming behaviour (Wilde and Clark, 1995). 183 A multiple regression model for Cf of the major fraction with NaCl, pH, S0, reactive SH, total SH, SS, H0, solubility and viscosity as parameters showed that all the factors (except H(()) and most of their interactions were significantly affecting the Cf (P < 0.001 for regression model; R2adj = 100%; p < 0.001 for most factors). Details of the regression models for Cf can be found in Appendix C. This shows that the foaming capacity of the major fraction was affected by its charge, surface hydrophobicity, solubility and viscosity, and the synergistic effects of all these factors. A stepwise regression model of Cf showed that the interaction between S0 and total SH was the most significant factor affecting Cf (p < 0.001; R2adj = 32.61 %). However, as the interaction between solubility and reactive SH, as well as the interaction between solubility and S0 entered the regression model, the interaction between S0 and total SH became insignificant. A separate stepwise regression model including NaCl and pH in every step resulted in the interactions between pH and NaCl, H(J) and viscosity as the most significant factors affecting Cf (p < 0.001; R 2 adj = 67.88 %). The interaction between reactive SH and solubility also significantly affected Cf before the interaction of H(J) and viscosity entered the model (p < 0.05; R adj = 64.57 %). NaCl and pH were included in every step because they were the two main sample treatments on the major fraction. A full quadratic regression model for FSI with Cf with NaCl, pH, So, H<|), SH, SS and solubility could not be carried out because of the small sample size (FSI for the major fraction in 0.01 M NaCl at pH 3 was not measured) which limited the power of the regression model. A stepwise regression model for FSI of the major fraction showed that the 2n d order term of pH, pH, the interactions between viscosity and S0, reactive SH and S0 were the most significant factors affecting FSI (p < 0.05; R 2 a d j = 96.58 %). Details of the 184 regression models on FSI can be found in Appendix D. A separate stepwise regression model including pH and NaCl in every step showed that up to 16 factors (including the interactions between parameters) were significantly affecting FSI (p < 0.001 for 15 factors; p < 0.05 for H(j)* S0; R 2 adj = 100%). This again shows that the stability of foam was affected by different factors and their interactions among factors. 6.9.3.1 Hydrophobicity and foaming: In the stepwise regression model of Cf, the interactions of S0 with total SH and solubility were the two most significant factors affecting Cf. Surface hydrophobicity plays a governing role as the trigger of foaming because amphiphilic proteins possessing high surface hydrophobicity are forcefully adsorbed at the air and water interface to cause a pronounced reduction of surface tension that readily facilitates foaming (Kato et al, 1983). Surface hydrophobicity affects the initial surface activity of a protein and its capability for rapid formation of an initial stable interface (Wilde and Clark, 1996). Townsend and Nakai (1983), on the other hand, found a significant correlation between the foaming capacity of 11 model proteins and the average hydrophobicity (H(])ave) but found no significant relationship between the surface hydrophobicity and foaming capacity. They suggested that proteins were extensively uncoiled at the air/water interface and a measure of total hydrophobicity would give a better correlation with foaming capacity. In the present study, however, the calculated total hydrophobicity (H(j)) did not correlate with Cf except for its interaction with viscosity. 185 Wilde and Clark (1996) suggested that the correlation between surface hydrophobicity and foaming properties of proteins only applied to foaming power (the ease of foaming) but not foam stability. Kato et al. (1983) also observed no significant correlation between foam stability and surface hydrophobicity of proteins and they suggested that the ability to associate and form a film due to denaturation might be essential for the foam stability of proteins. In this study, surface hydrophobicity significantly affected foam stability through its interaction with reactive SH and viscosity (in the first stepwise regression model for FSI). When NaCl and pH effects were considered in every step in the 2n d stepwise regression model, S0, its 2nd order term and its interactions with NaCl and reactive SH were all significantly affecting FSI. 6.9.3.2 SH and SS content and foaming: Townsend and Nakai (1983) used the ratio of number of SS linkage/mol weight as an index of molecular flexibility. They found a significant negative correlation between the SS linkage/mol weight and foaming capacity. They suggested that the flexible protein molecules have higher tendency to spread out at the air/water interface to stabilize fresh air cells and thus preventing the collapsing of foams. The whole extract had almost double the amount of SS than the major fraction. Since the major fraction comprised approximately 64 % of the total protein in the whole extract, and was composed of mainly higher molecular weight component while the remaining 36 % of proteins in the whole extract consisted mainly of lower molecular weight components, the SS linkage/ molecular weight ratio for the whole extract would be much higher than the major fraction. The major fraction was comparatively more 186 flexible than the whole extract which may be related to its better foaming capacity. From the regression model for Cf of the major fraction, reactive, total SH and SS, and their interactions with S0 significantly affected Cf (p < 0.001; Appendix C). The 2nd order term of SS, total SH, and the interactions between reactive SH with S0 and solubility were all significantly affecting FSI of the major fraction (stepwise regression model with NaCl and pH in each step; p < 0.001; Appendix D). 6.9.3.3 Solubility and foaming: Comparing the foaming properties of the major fraction with their solubility profiles, the observed better foamers (the major fraction in 0.01 M NaCl at pH 3 and 7 and in 1.0 M NaCl at pH 7) were also the more soluble ones. The regression model for Cf showed that solubility of the major fraction had a significant positive correlation with Cf (p < 0.001; Appendix C). The 2nd stepwise regression model for FSI also showed that the interactions between solubility and reactive SH, viscosity and H<J) were important factors determining the stability of the major fraction (p < 0.001 except p < 0.05 for the interaction between solubility and H(|); Appendix D). Both Madhusudhan and Singh (1985a) and Dev and Quensel (1986) reported a pronounced dependence of foaming capacity on solubility of protein of linseed meal or linseed isolates. Both research groups found that the foaming capacity of the linseed meal and isolated in 0.1 M NaCl or distilled deionized water showed a U-shaped curve similar to the solubility profile. 187 6.9.3.4 Charge, pi and foaming: pH was a significant factor affecting both the Cf and FSIof the major fraction (p < 0.001; Appendix C and D respectively). At the isoelectric pH of proteins, net charge is minimal. Wilde and Clark (1996) commented that the foam of a protein is usually more stable at pH close to the isoelectic point (pi) of a protein because electrostatic repulsion reducing the protein-protein interactions involved in forming a viscoelastic film in the foam is minimised. However, results of the present study, based on observations and the foam volume-conductivity set (Set 2), did not show the trend observed by Wilde and Clark (1996). In a comparison of the foam stability of the major fraction in citrate-phosphate buffer with 0.01 M NaCl as a function of pH (3, 5 or 7), it was observed that foam stability was lowest at pH 5, which was the pH closest to the isoelectric points of the fraction. The decrease in electrostatic repulsion at the isoelectric points may not only reduce the interference of the protein-protein interactions from forming a stable foam, but may also lead to formation of insoluble aggregates because the repulsions preventing protein-protein interactions from forming aggregates are also minimized. The insolubility of protein may have led to poor foaming properties in this study. Dev and Quensel (1986) also found that foaming capacity was lowest at the isoelectric pH. Dev and Quensel (1986) found that maximum foaming capacity for linseed flour and isolate in distilled deionized water were observed at acid pH 2 below the isoelectric region. The foaming capacity (FC) and the relative foam conductivity (Cf) of the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 3 were also highest among the same sample at the three pHs. 188 Results from the photographic set show that the foams from the major fraction in citrate-phosphate buffer with 0.01 M NaCl at pH 3 were more stable than those at pH 7. This agreed with the findings of Madhusudhan and Singh (1985) and Dev and Quensel (1986) which showed an increase in foam stability with decreasing pH. Dev and Quensel (1986) also commented that in general, foam stability was maximum at a pH somewhat below the isoelectric range. 6.9.3.5 Viscosity and foaming: From the regression model for Cf, viscosity and its interaction with SS were significant factors affecting Cf. (p < 0.1; Appendix C). The interactions between viscosity and solubility, and viscosity and NaCl had significant effect on Cf. (p < 0.001; Appendix C). A negative coefficient of viscosity in the Cf regression model for the major fraction suggested that a lower viscosity would give arise to better foaming power. On the other hand, Townsend and Nakai (1983) suggested that high viscosity of the protein solution enhance foaming capacity. Bikerman (1973) studied the relationship between foaming and solution viscosity by the addition of inert materials to alter the viscosity and found conflicting results. The insolubility of some samples might also contribute to the difference in the trends found in this study and the literature. Viscosity and its various interactions with pH, S0, SS and solubility also significantly affected FSI (p < 0.001; the 2nd stepwise regression model with pH and NaCl in every step; Appendix D). 189 CHAPTER 7—GENERAL CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH The major fraction of flaxseed proteins was isolated and characterized in this study. Flaxseed proteins were extracted with 0.10 M NaCl in 0.10 M Tris at pH 8.6, and the protein components were separated based on their charges using DEAE-Sephacel anion-exchange chromatography. The major protein fraction, which eluted at a NaCl concentration of 0.25 M and conductivity of 19-22 mS/cm, comprised of 63.7 % of the total flaxseed proteins. At least six different size classes of polypeptides with estimated molecular weights of 8.5-10, 12-28, 30-35, 41-53, 59-64 and 67-71 kDa were observed when the fraction was analyzed by reducing SDS-PAGE. In particular, polypeptides of 20 + 1, 26 + 2 and 31 ± 1 kDa were the major components within the six size classes, similar to those reported in the literature for 1 IS seed globulins. In non-reducing SDS-PAGE, protein components with molecular weights of 21, 30, 40 and 47-48 kDa were predominant. The high content of disulfide linkages (SS) and low content of sulfhydryl groups (SH) determined by colorimetric reaction with Ellman's reagent and NTSB indicated that most of the cysteine residues in this fraction existed in a disulfide linked cystine form. This could also be observed by comparison of the non-reducing-and reducing-SDS-PAGE profiles of the fraction. The disappearance or decrease in intensities of the higher molecular weight bands at 34-35, 40, 43-45, 84-88 and 95-99 kDa, accompanied by the increase in intensities or appearance of lower molecular weight bands at 12-18, 24-28 and 30-32 kDa, revealed the contribution of the disulfide linkage of cystine in stabilization of the overall structure of the major fraction. 190 The major fraction exhibited a typical U-shaped solubility curve as a function of pH at low NaCl concentration (0.01 M), with a broad region of minimal solubility between pH 4.0 and 5.5. The presence of 1.0 M NaCl broadened and shifted the region of minimal solubility to the acidic side below pH 4.0. The isoelectric points of three components in the major fraction separated by isoelectric focusing were 4.7 + 0.0, 5.1 + 0.0, and 5.6 + 0.1 (n = 4). Surface hydrophobicity measured by the PRODAN fluorescent probe was significantly affected by pH and NaCl concentration (p < 0.05). The addition of NaCl increased the surface hydrophobicity of the major fraction at all three pHs studied (3, 5 and 7). Surface hydrophobicity of the major fraction decreased as pH decreased from 7 to 3. Similar to the storage globulin of other oilseeds, the major fraction had high levels of arginine, glutamate (and/or glutamine) and aspartate (and/or asparagine). Its apparent viscosities at the three pHs and two NaCl concentrations were not significantly different from one another except in 1.0 M NaCl at pH 7, which was significantly higher. The foaming properties of the major fraction were affected by pH, NaCl, solubility, viscosity, surface hydrophobicity, the number of SH and SS, and the interactions among these factors. The fraction had the finest, densest, and most stable foam in buffer with low NaCl concentration (0.01 M) and low pH (3). The foaming properties of the major fraction in 0.01 M NaCl at pH 7 were comparable to that of a commercial egg albumen powder. The major fraction had higher surface and total hydrophobicity but lower number of SH and SS groups, as well as lower apparent viscosity than the unfractionated whole protein extract. Its foaming properties were much better but its overall solubility was lower than that of the whole extract. 191 This thesis outlined a procedure for isolation of the major fraction of flaxseed proteins. The separation of the high- and low-molecular weight components in one chromatographic step can serve as a starting point for the investigation of the low-molecular weight components of flaxseed proteins in the future. Results from the gel electrophoresis and isoelectric focusing of the major fraction laid an important foundation toward more in-depth molecular characterization (such as to investigate its subunit compositions, isoelectric points of the individual subunits, association-dissociation of subunits, and the overall structure) of the protein. Results from foaming studies served as a basis for further investigation of the foaming properties of the fraction in different pHs and NaCl concentrations, as well as in combinations with other ingredients in model food systems or non-food systems (such as cosmetics and household products). Whipping method can be used in placed of the sparging method for application to real industrial systems. Functional characteristics of the major fraction other than solubility and foaming are still awaiting exploration for its food and non-food application potentials. The findings of the present study on surface hydrophobicity, SH and SS contents, apparent viscosity and solubility of the major fraction* can serve as a basis to explain the various functional properties to be studied in future research. 192 REFERENCE Alizadeh-Pasdar, N. and Li-Chan, E. C.Y. 2000. Comparison of protein surface hydrophobicity measured at various pH values using three different fluorescent probes. J. Agri. Food Chem. 48:328-334. Andrews, A.T. 1986. Electrophoresis theory, techniques, and Biochemical and Clinical Applications. 2nd ed. Clarendon Press, Oxford University Press, Toronto. ON. Association of Official Analytical Chemists (AOAC). 1990. Cereal Foods. Chp. 32 In Official Methods of Analysis. 15th ed. p. 777. Anonymous. 1976. Foam collapse rate: A—Sigma value method. BEER-22. In Methods of Analysis of the American Society of Brewing Chemists. 7th ed. American Society of Brewing Chemists, St. Paul, MN. p. 1-6. Anonymous. 1992. Omega Complete pH & Conductivity Measurement Handbook and Encyclopedia0. Volume 28. Omega Engineering, Inc., Stamford, CT. Balmaceda, EA., Kim, M.K., Franzen, R., Mardones, B. and Lugay, J.C. 1984. Protein functionality methodology—standard tests. Appendix 27-1 in Food Protein Chemistry— An Introduction for Food Scientists. J.M. Regenstein and C E . Regenstein (Ed). Academic Press Inc., Toronto, ON. p. 284-286. Barnes, H.A., Hutton, J.F. and Walters, K. 1989. An Introduction to Rheology. Elsevier Science Publishing Company Inc., New York, NY. Beveridge, T., Toma, S.J. and Nakai, S. 1974. Determination of SH- and SS- groups in some food proteins using Ellman's reagent. /. FoodSci. 39: 49-51. Bhatty, R.S. and Cherdkiatgumchai. 1990. Compositional analysis of laboratory-prepared and commercial samples of linseed meal and of hull isolate from Flax. fAOCS. 67: 79-83. Bigelow, C C . 1967. On the average hydrophobicity of proteins and the relation between it and protein structure. J. Theoret. Biol. 16: 187-211. Bikerman, J.J. 1973. Foams. Springer-Verlag. New York, N.Y. Blenford, D.E. 1996. Use of amino acids and peptides in sports nutrition. International Food Ingredients 3: 20-23. Cardamone, M. and Puri, N. 1992. Spectrofluorimetric assessment of the surface hydrophobicity of proteins. Biochem. J. 282: 589-593. Chakrabarti, A. 1996. Fluorescence of spectrin-bound PRODAN. Biochem. Biophys. Res. Commun. 226:495-497. 193 Clark, N.O. 1948. The electrical conductivity of foam. Trans. Faraday Soc. 44: 13-15. Cheng, E.S.Y. 2001. A study of protein surface hydrophobicity and structure-function analysis of proteins in spray dried egg albumen. M.Sc. Thesis. The University of British Columbia, Vancouver, BC. Dalgalarrondo, M., Robin, J.M., and Azanza, J.L. 1986. Subunit composition of the globulin fraction of rapeseed (Brassica Napus. L.). Plant Sci. 43: 115-124. Damodaran, S. and Kinsella, J.E. 1982. Effects of ions on protein conformation and functionality. Chp. 13. In Food Protein Deterioration Mechanisms and Functionality. LP. Cherry. (Ed). American Chemical Society, Washington, D.C. p. 327-357. Damodaran, S. 1984. Estimation of disulfide bonds using 2-nitro-5-thiosulfobenzoic acid: limitations. Anal. Biochem. 145: 200-204. Dev, D.K. and Quensel, E. 1986. Functional and microstructural characteristics of linseed (Linum usitatissimum L.) flour and a protein Isolate. Lebensm.-Wiss. U.-Technol. 19: 331-337. Dev, D.K., Quensel, E. and Hansen, R. 1986a. Nitrogen extractability and buffer capacity of defatted linseed (Linum usitatissimum L.) flour. J. Sci. Food Agric. 37: 199-205. Dev, D.K., Sienkiewicz, T., Quensel.E. and Hansen, R. 1986b. Isolation and partial characterization of flaxseed (Linum usitatissimum L.) proteins. Die Nahrung 30: 391-393. Dev, D.K. and Sienkiewicz, T. 1987. Isolation and subunit composition of 1 IS globuline of linseed (Linum usitatissimum L.) Die Nahrung 31:767-769. Dev, D.K. and Quensel, E. 1988. Preparation and functional properties of linseed protein products containing different levels of mucilage. J. Food Sci. 53: 1834-1857. Dev, D.K. and Quensel, E. 1989. Functional properties of linseed protein products containing different levels of mucilage in selected food systems. J. Food Sci. 54: 183-186. Ellman, G.L. 1959. Tissue sulfhydryl groups. Arch. Biochem. Biophys. 82: 70-72. Flax Council of Canada. 2001. www.flaxcouncil.ca Freedman, T.P. 1995. Structure of Flaxseed. Chp. 1. In Flaxseed in Human Nutrition. S.C. Cunnane andL.U. Thompson (Ed). AOCS Press, Champaign, Illinois, p. 11-21. Garcia. M.C., Torre, M., Marina, M.L., and Laborad, F. 1997. Composition and characterization of soybean and related products. Crit. Rev. Food Sci. Nutr. 37: 361-391. 194 German, B.J. and Philips, L. 1994. Protein interactions in foams: Protein-gas phase interactions. Chp. 6. In Protein Functionality in Food Systems. N.S. Hettiarachchy and G.R. Ziegler (Ed). Marcel Dekker, Inc., New York, p.181-208. Gruener. L. and Ismond, M.A.H. 1997. Effect of acetylation and succinylation on the physiochemical properties of the canola 12S globulin. Part I. Food Chem. 60: 357-363. Guillerme, C., Loisel, W., Bertrand, D. and Popineau, Y. 1993. Study of foam stability by video image analysis. Relationship with the quantity of liquid in the foams. J. Texture Studies 24: 287-302. Hailing, P.J. 1981. Protein-stabilized foams and emulsions. Crit. Rev. Food. Sci. Nutr. 15: 155-201. Haughland, R.P. 1996. Fluorescent phospholipids, fatty acids and sterols. In Handbook of Fluorescent Probes and Research Chemicals. Molecular Probes, Eugene, OR. P. 299-300. Hayakawa, S. and Nakai, S. 1985. Relationships of hydrophobicity and net charge to the solubility of milk and soy proteins. J. Food Sci. 50: 486-491. Inquello, V., Raymond, J. and Azanza, J.L. 1993. Disulfide interchange reactions in 1 IS globulin subunits of Cruciferae seeds. Relationships to gene families. Eur. J.Biochem. 211: 891-895. Jones, L.J. and Tung, M.A. 1983. Functional properties of modified oilseed protein concentrates and isolates. Can. Inst. Food Sci. Technol. f. 16: 57-62. Kapoor. A.C. and Gupta, Y.P. 1977. Chemical evaluation and electrophoretic pattern of soy proteins. J. Food Sci. 42: 1558-1561. Kato, A., Takahashi, A., Matsudomi, N., and Kobayashi, K. 1983. Determination of foaming properties of proteins by conductivity measurements, f. Food Sci. 48: 62-65. Klockman, D.M, Toledo, R. and Sims, K.A. 1997. Isolation and characterization of defatted canola meal protein, f. Agri Food Chem. 45: 3867-3870. Kinsella, J.E. 1979. Functional properties of soy proteins. J. Am. Oil Chem. Soc. 56: 242-248. Kinsella, J.E. 1981. Relationship between structure and functional properties of food proteins. Chp 3. In Food Proteins. P.F. Fox and J..J. Condon (Ed). Applied Science Publishers, New York, p.51-103. Lei, B. 1999. Unpublished data. The University of British Columbia, Vancouver, BC. 195 Li-Chan, E.C.Y. and Haskard, C A . 1998. Hydrophobicity of bovine serum albumin and ovalbumin determined using uncharged (PRODAN) and anionic (ANS) fluorescent probes. J. Agric Food Chem. 46: 2671-2677'. Li-Chan, E.C.Y. 1999. Hydrophobicity in food protein systems. Chp. 144. In Encyclopedia of Food Science and Technology. 2nd ed. F.J. Francis. (Ed). Wiley. New York. p. 1321-1331. Li-Chan, E. C. Y., Sultanbawa, F., Losso, J. N., Oomah, D. and Mazza, G. 1999. Isolation and characterization of cadmium binding proteins from flax (Linum usitatisisiumum) seed. Poster P067. CJFST Annual Conference, Kelowna, BC, June 6-9, 1999. Marcone, M.F., Kakuda, Y. and Yada, R.Y. 1998. Salt-soluble seed globulins of various dicotyledonous and monocotyledonous plants—I. Isolation/purification and characterization. Food Chem. 62: 27-47. Marcone, M.F. 1999. Biochemical and biophysical properties of plant storage proteins: a current understanding with emphasis on 1 IS seed globulins. Food Res. Int. 32: 79-92. Madhusudhan, K.T. & Singh, N. 1983. Studies on linseed proteins. J. Agric. Food Chem. 31: 959-963. Madhusudhan, K.T. Singh, N. 1985a. Effect of detoxification treatment on the physiochemical properties of linseed proteins. J. Agri. Food Chem. 33: 1219-1222. Madhusudhan, K.T. Singh, N. 1985b. Effect of heat treatment on the functional properties of linseed meal. J. Agri. Food Chem. 33: 1222-1226. Madhusudhan, K.T. Singh, N. 1985c. Isolation and characterization of the major fraction (12S) of linseed proteins. J. Agri. Food Chem. 33: 673-677. Madhusudhan, K.T. Singh, N. 1985d. Isolation and characterization of a small molecular weigh protein of linseed meal. Phytochemistry 24: 2507-2509. Mine, Y. 1997. Effect of dry heat and mild alkaline treatment on functional properties of egg white proteins. J. Agric. Food Chem. 45: 2924-2928. Ming, D., Markley, J.L, and Hellekant, G. 1995. Quantification of cysteinyl sulfhydryl residues in peptides and proteins by ESI-MS or MALDI-MS. BioTechniques 18: 808, 810. Mohan, C. 1997. Buffers—A guide for the preparation and use of buffers in biological systems. Calbiochem-Novabiochem Corporation, San Diego, CA. Nakai, S. 1983. Structure-function relationships of food proteins with an emphasis on the importance of protein hydrophobicity. J. Agric. Food Chem. 31: 676-683. 196 Nakai, S. and Powrie, W.D. 1981. Modification of proteins for functional and nutritional improvements. Chp 11. In Cereals—A Renewable Resource, Theory and Practice. Y. Pomeranz and L. Munck (Ed). American Association of Cereal Chemists, St. Paul, MN. p. 217-242. O'Carrol. P. 1995. A bio-active future. World of Ingredients Nov./Dec, 28-30. Oomah, B.D. and Mazza, G. 1993a. Flaxseed proteins—a review. Food Chem. 48:109-114. Oomah, B.D. and Mazza, G. 1993b. Processing of flaxseed meal: effect of solvent extraction on physiochemical characteristics. Lebensm.-Wiss. U.-Technol. 26: 312-317. Oomah, B.D., Mazza, G. and Cui, W. 1994. Optimization of protein extraction from flaxseed meal. Food Res. Int. 27: 355-361. Oomah, B.D. and Mazza, G. 1995. Functional properties, uses of flaxseed protein. INFORM. 6: 1246-1252. Oomah, B.D. and Mazza, G. 1998. Compositional changes during commercial processing of flaxseed. Industr. Crops Products. 9: 29-37. Oomah, B.D and Mazza, G. 2000. Bioactive components of flaxseed: occurrence and health benefits. Chp 9. In Phytochemicals and Phytopharmaceuticals. F. Shahidi and C.T. Ho (Ed). AOCS Press, Champaign, Illinois. pl06-121. Painter, E.P. and Nesbitt, L.L. 1946. Nitrogenous constituents of flaxseed. Industr. Engng. Chem. 38: 95-98. Patel, P.D. and Fry, J.C. 1987. The search for standardised methods for assessing protein functionality. Chp 7. In Developments in Food Proteins. 5th ed. B.J.F. Hudson (Ed). Elsevier Applied Science Publishers, London, p. 299-333. Peng, I.C., Quass, D.W., Dayton, W.R. and Allen, C.E. 1984. The physicochemical and functional properties of soybean 1 IS globulin—a review. Cereal Chem. 61: 480-490. Peterson, S.M. 1958. Linseed oil meal. Chp 21. In Processed Plant Protein Foodstuffs. A.M. Altschul (Ed). Academic Press, Inc., New York. p. 593-617. Pomeranz, Y. 1991. Functional Properties of Food Components. 2nd ed. Academic Press, Inc. Toronto. Pszczola. D E . 2000. Genes and diet: the specialized role ingredients may play. Food Technol. 54 (10): 82, 84, 96. 197 Rha, C K . and Pradipasena, P. 1986. Viscosity of Proteins. Chp. 2. In Functional properties of Food Macromolecules. J.R. Mitchell and D.A. Ledward (Ed). Elsevier Applied Science Publishers, London, p. 79-120. Salunkhe, D.K., Chavan, J.K., and Adsule, R.N., and Kadam, S.S. 1992. World Oilseeds: Chemistry, Technology, and Utilization. Van Nostrand Reinhold, New York. Schenz, T.W. and Morr, C.V. 1996. Viscosity. Chp. 5. In Methods of Testing Protein Functionality. 1st ed. G.M. Hall (Ed). Blackie Academic & Professional, New York. p. 61-75. Shen. J.L. 1981. Solubility and viscosity. Chp. 5. In Protein Functionality in Foods. American Chemical Society, Washington, D.C p.89-109. Smith, A. K., Johnson, V.L. and Beckel, A . C 1946. Linseed proteins: alkali dispersion and acid precipitation. Inustr. Engng. Chem. 38: 353-356. Smith, D.M. 1994. Protein separation and characterization procedures. Chp. 15. In Introduction to the Chemical Analysis of Foods. S.S. Nielsen (Ed). Jones and Bartlett Publishers, Boston, p. 221-232. Sosulski, F.W. and Bakal., A. 1969. Isolated proteins from rapeseed, flax, and sunflower meals. Can. Inst. Food Sci Technol. J. 2: 28-32. Sosulski, F.W. and Sarwar, G. 1973. Amino-acid composition of oilseed meals and protein isolates. Can. Inst. Food Sci Technol. J. 6: 1-5. Sultanbawa, F. 1998. Unpublished data. The University of British Columbia, Vancouver, BC. Thannhauser, T.W., Konishi, Y., and Scheraga, H.A. 1983. Sensitive quantitative analysis of disulfide bonds in polypeptides and proteins. Anal. Biochem. 138: 181-188. Townsend, A-A. and Nakai, S. 1983. Relationships between hydrophobicity and foaming characteristics of food proteins. J. Food Sci. 48: 588-594. Vassel, B. and Nesbitt, L.L. 1945. The nitrogenous constituents of flaxseed. II. The isolation of a purified protein fraction. J. Biol. Chem. 159: 571-584. Vojdani, F. 1996. Solubility. Chp. 2. In Methods of Testing Protein Functionality. 1st ed. G.M. Hall (Ed). Blackie Academic & Professional, New York. p.11-60. Wanasundara, P.KJ.P.D. and Shahidi, F. 1994a. Alkanol-arnmonia-water/hexane extraction of flaxseed. Food Chem. 49: 39-44. Wanasundara, P.KJ.P.D. and Shahidi, F. 1994b. Functional properties and amino-acid composition of solvent-extracted flaxseed meal. Food Chem. 49: 41-51. 198 Wanasundara, P.KJ.P.D. and Shahidi, F. 1996. Optimization of hexametaphosphate-assisted extraction of flaxseed proteins using response surface methodology. J. Food Sci. 61: 604-607. Wanasundara, P.KJ.P.D. and Shahidi, F. 1997a. Removal of flaxseed mucilage by chemical and enzymatic treatments. Food Chem. 59: 47-55. Wanasundara, P.KJ.P.D. and Shahidi, F. 1997b. Functional properties of acylated flax protein isolates. J. Agric. Food Chem. 45: 2431-2441. Waniska, R.D and Kinsella. 1979. Foaming properties of proteins: evaluation of a column aeration apparatus using ovalbumin. J. Food Sci. 44: 1398-1411. Wilde, P.J. and Clark, D.C. 1996. Foam formation and stability. Chp. 5. In Methods of Testing Protein Functionality. 1st ed. G.M. Hall (Ed). Blackie Academic & Professional, New York. p. 110-152. Wolf, W.J. and Nelsen, T.C. 1996. Partial purification and characterization of the 15S globulin of soybeans, a dimer or glycinin. J. Agric. Food Chem. 44: 785-791. Wright, D.J. and Hemmant, J.W. 1987. Foaming properties of protein solutions: comparison of large-scale whipping and conductimetric methods. J. Sci. Food Agric. 41: 371-371. Youle, R.J. and Huang, H.C. 1981. Occurrence of low molecular weight and high cysteine containing albumin storage proteins in oilseeds of diverse species. Amer. J.Bot. 68: 44-48. 199 APPENDIX A - Minitab™ Printout for Regression Analyses for Solubility of the Major Fraction R e g r e s s i o n A n a l y s i s for % s o l u b l e ( Q u a d r a t i c m o d e l : p H , N a C l , S Q , R S H , T S H , S S , H(|)) * NaCl * NaCl i s h i g h l y c o r r e l a t e d with other X v a r i a b l e s * NaCl * NaCl has been removed from the equation * T-SH i s h i g h l y c o r r e l a t e d with other X v a r i a b l e s * T-SH has been removed from the equation * T-SH*2 i s hi g h l y c o r r e l a t e d with other X v a r i a b l e s * T-SH^2 has been removed from the equation The regression equation is % soluble = - 1078 - 88 pH + 19.0 pH.* pH - 26.2 pH * NaCl - 5.2 NaCl + 0.482 NaCl * So - 311 So + 0.00266 So * So - 0.360 So * pH + 1517 R-SH - 1202 R-SH~2 +5.40 R-SH*So +90.3 T-SH*So + 24.1 SS - 0.182 SS* SS + 0.00405 SS*So Predictor Coef SE Coef T P Constant -1078 1810 -0. 60 0 565 PH -88.2 148.5 -0. 59 0 566 pH * pH 19.03 10.06 1. 89 0 088 * pH * NaCl -26.17 11.54 -2 . 27 0 047 ** NaCl -5.15 42.08 -0. 12 0 905 NaCl * So 0.4818 0.1953 2 . 47 0 033 ** So -310.7 158 .7 -1. 96 0 079 * So * So 0.002659 0.001993 1. 33 0 212 So * pH -0.3603 0.1440 -2 . 50 0 031 ** R-SH 1517 2333 0. 65 0 530 R-SHA2 -1202 1855 -0. 65 0 531 R-SH*So 5.400 2.761 1. 96 0 079 * T-SH*So 90.28 45.96 1. 96 0 078 * s s 24.10 48.44 0. 50 0 630 s s * s s -0.1817 0.3763 -0. 48 0 640 SS*So 0.004047 0.002504 1. 62 0 137 Ho 1.5860 0.9420 1. 68 0 131 Ho"2 -0.0012114 0.0007152 -1. 69 0 129 S = 1.833 R - S q = 99.9% R -• S q ( a d j ) = 99.7% Analysis of Variance Source DF SS MS F Regression 15 24692.8 1646 .2 489.73 Residual Error 10 33.6 3 .4 Total 25 24726.5 *** S i g n i f i c a n c e at p < 0.001 ** S i g n i f i c a n c e at p < 0.05 * S i g n i f i c a n c e at p < 0.1 200 Stepwise Regression for solubility A l p h a - t o - E n t e r : 0.2 Alpha-to-Remove: 0.2 Response i s % s o l u b l on 18 p r e d i c t o r s , w i t h N = 26 S t e p 1 2 C o n s t a n t 36.84 130.53 pH * pH 0.85 4.95 T-Value 2.55 1.67 P-Value 0.018 0.109 pH -42 T-Va l u e -1.39 P-Value 0.178 S 28.5 27.9 R-Sq 21.32 27.42 R - S g ( a d j ) 1 8 . 0 4 2 1 . 1 1 PRESS 22678.3 22903.7 R-Sq(pred) 8.28 7.37 b e s t a l t . V a r i a b l e pH So T-Value 2.34 -1.25 P-Value 0.028 0.225 V a r i a b l e So * pH T-SH*So T-Value 2.08 -1.25 P-Value 0.048 0.225 201 Stepwise Regression for % soluble (pH and NaCl in every step) Alpha-to -Enter: 0. 2 Alpha-to-Remove: 0.2 Response i s % s o l u b l on 18 pre d i c t o r s , with N 26 Step 1 2 3 4 5 Constant 27 . 00 69.09 -228 .76 1089 .14 1168 .26 PH 8 .0 -0.4 107 .7 113.9 114.0 T-Value 2 . 37 -0 .11 10 .59 11 .88 13 .35 P-Value 0 . 027 0.913 0. 000 0. 000 0. 000 NaCl -14 .0 -72.3 -85.6 -86.8 -91.9 T-Value -1. 23 -4.07 -11 .81 -13 .18 -14 .78 P-Value 0.231 0 . 001 0. 000 0. 000 0. 000 NaCl * So 0.207 0. 600 0. 625 0. 640 T-Value 3.81 14 .07 15 .60 17 .68 P-Value 0.001 0. 000 0. 000 0. 000 SO * pH -0. 206 -0. 219 -0. 220 T-Value -10 .73 -12 .02 -13 .58 P-Value 0. 000 0. 000 0. 000 TSH A2 -115 -128 T-Value -2 .36 -2 .93 P-Value 0 . 029 0. 009 ss*ss 0.0185 T-Value 2 .49 P-Value 0. 022 S 28 .6 22 .7 9 .14 8 .28 7 .38 R-Sq 23.65 54.03 92.91 94.45 95.81 R-Sq(adj) 17. 01 47.76 91 .56 93 .06 94 .49 PRESS 23551 .3 14093 .4 2548 .71 2313 .86 1911 .42 R-Sq(pred) 4. 75 43 .00 89 .69 90 .64 92 .27 best a l t . Vari a b l e pH * NaC So * So R-SH~2 SS T-Value 3.11 -7 .71 2 .35 2 .48 P-Value 0.005 0. 000 0. 029 0. 023 Va r i a b l e pH * pH SS *So R -SH pH * PH T-Value 1.62 -5 .49 2 .35 1 .52 P-Value 0.119 0. 000 0. 029 0. 145 APPENDIX B - Minitab™ Printout for Regression Analyses for Viscosity of the Major Fraction Regression Analysis: Viscosity (Quadratic model: pH, NaCl, So, Solubility, H<|>) * NaCl * NaCl is highly correlated with other X variables * NaCl * NaCl has been removed from the equation The regression equation is Vis = 100668 - 46376 pH + 4314 pH * pH - 12661 pH * NaCl + 10421 NaCl - 20.9 NaCl * So + 47.8 So + 0.070 So * So - 1.98 So * pH + 1066 %sol*NaCl + 54 %sol*pH - 0.86 %sol*So - 712 % soluble + 3.35 % sol^2 + 68.8 Ho - 0.0527 Ho 2^ + 0.00672 Ho*%Sol Predictor Coef SE Coef T P Constant 100668 65848 1. 53 0 161 pH -46376 27974 -1. 66 0 132 pH * pH 4314 2563 1. 68 0 127 pH * NaCl -12661 9875 -1. 28 0 232 NaCl 10421 12490 0. 83 0 426 NaCl * So -20.94 24.40 -0. 86 0 413 So 47 .77 27.33 1. 75 0 114 So * So 0.0704 0.1689 0. 42 0 687 So * pH -1.976 5.751 -0. 34 0 739 %sol*NaC 1065.7 749.2 1. 42 0 189 %sol*pH 53.8 113 .9 0. 47 0 648 %sol*So -0.860 1.483 -0. 58 0 576 % solubl -712.0 585.0 -1. 22 0 254 % sol"2 3.349 1.992 1. 68 0 127 Ho 68.75 39.69 1. 73 0 117 Ho~2 -0.0526J9 0.03018 -1. 75 0 115 Ho*%Sol 0.00672'3 0.006137 1. 10 0 302 S = 34.84 R-Sq = 93.1% R- Sq(adj) = 80.8% Analysis of Variance Source DF SS MS Regression 16 147317 9207 7 . 5 Residual Error 9 10925 1214 Total 25 158242 *** Significance at p < 0.001 ** Significance at p < 0.05 * Significance at p < 0.1 203 Stepwise Regression: Viscosity (pH, NaCl, So, solubility, TSH, RSH, SS, H<|>) Alpha-to-Enter: 0.2 Alpha-to-Remove: 0.2 Response i s Vis on 29 pr e d i c t o r s , with N = 26 S t e p 1 2 3 Constant 639.3 695.3 743 .3 p H * N a C l 20.7 21.0 50.4 T-Value 5 . 04 6 . 81 7 .22 P-Value 0 . 000 0.000 0 . 000 % s o l A 2 -0.0123 -0.0198 T-Value -4.43 -7.45 P-Value 0.000 0.000 N a C l -176 T-Value -4.46 P-Value 0.000 S 56.6 42.5 31.5 R-Sq 51.40 73 .77 86.22 R - S q ( a d j ) 4 9 . 3 8 7 1 . 4 9 8 4 . 3 4 PRESS 88661.3 53973.0 30050.6 R-Sq(pred) 43.97 65.89 81. 01 best a l t . Va r i a b l e NaCl * N % s o l u b l NaCl * N T-Value 4.36 -3 .82 -4.46 P-Value 0.000 0 . 001 0.000 Var i a b l e NaCl %sol*T-S So * So T-Value 4.36 -3.82 3.22 P-Value 0.000 0.001 0.004 Regression Analysis: Viscosity (based on significant parameters from the 1 s t stepwise regression; NaCl, pH * NaCl, pH, % solA2) The regression equation is Vis = 723 - 155 NaCl + 46.3 pH * NaCl +4.08 PH - 0.0198 Predictor Coef SE Coef T P Constant 722.81 30.35 23.81 0.000 *** NaCl -154.88 47 .38 -3 .27 0.004 ** pH * NaC 46.251 8.727 5.30 0.000 *** pH 4.083 5.070 0.81 0.430 % sol"2 -0.019821 0 . 002685 -7.38 0.000 *** S = 31.74 R - S q = 8 6 . 6 % R - S q ( a d j ) = 8 4 . 1% Analysis of Variance Source DF SS MS F Regression 4 137087 34272 34. 02 Residual Error 21 21155 1007 To t a l 25 158242 Source DF Seq SS NaCl 1 69904 pH * NaC 1 11542 pH 1 726 % sol^2 1 54915 Unusual Observations Obs NaCl V i s F i t SE F i t Residual St Resid 14 0.01 650.00 591.69 14.16 58.31 2.05R R denotes an observation with a large standardized r e s i d u a l Stepwise Regression for Viscosity (NaCl and pH in every step) Alpha-to-Enter: 0.2 Alpha-to-Remove: 0.2 Response i s Vis on 29 pr e d i c t o r s , with N = 26 Step 1 2 3 Constant 584.1 516.7 579.7 p H 10.9 47.1 40.8 T-Value 1.57 5.00 5.78 P-Value 0.130 0.000 0.000 N a C l 105 98 -86 T-Value 4.49 5.67 -1.99 P-Value 0.000 0.000 0.059 % s o l * p H -0.340 -0.434 T-Value -4.56 -7.41 P-Value 0.000 0.000 p H * N a C l 36.3 T-Value 4.46 P-Value 0.000 S 58.9 43.2 31.7 R-Sq 49.57 74.08 86.70 R - S q ( a d j ) 4 5 . 1 8 7 0 . 5 5 8 4 . 1 7 PRESS 98770.7 58055.5 33077.6 R-Sq(pred) 37.58 63.31 79.10 best a l t . Vari a b l e % sol~2 NaCl * S T-Value -3.67 3.61 P-Value 0.001 0.002 Varia b l e %sol*SS % sol^2 T-Value -3.11 3.60 P-Value 0.005 0.002 205 Regression Analysis: Viscosity (Based on significant parameters from the 2n d stepwise regression with pH and NaCl in every step; NaCl, pH * NaCl, %sol*pH, pH) The regression equation is V i s = 580 - 85.7 NaCl + 36.3 pH * NaCl - 0.434 %sol*pH + 40.8 pH Predictor Constant NaCl pH * NaC %sol*pH P H S = 31.65 Coef 579.69 -85.66 36.280 -0.43380 40.786 SE Coef 27.15 42.98 8.126 0.05853 7.056 T 21.35 -1.99 4.46 -7.41 5.78 P 0.000 0.059 0.000 0 .000 0.000 R - S q = 8 6 . 7 % R - S q ( a d j ) = 84 .23 Analysis of Variance Source DF SS Regression 4 137202 Residual Error 21 21040 Total 25 158242 MS 34300 1002 F 34.23 0.000 *** Source NaCl pH * NaC %sol*pH P H DF 1 1 1 1 Seq SS 69904 11542 22283 33473 Unusual Observations Obs NaCl Vis F i t SE F i t Residual St Resid 14 0.01 650.00 591.67 14.12 58.33 2.06R R denotes an observation with a large standardized r e s i d u a l 206 APPENDIX C - Minitab™ Printout for Regression Analyses for C f of the Major Fraction Regression Analysis for C-f * T-SH i s h i g h l y c o r r e l a t e d w i t h o t h e r X v a r i a b l e s * T-SH has been removed f r o m t h e e q u a t i o n * % s o l * T - S H i s h i g h l y c o r r e l a t e d w i t h o t h e r X v a r i a b l e s * % s o l * T - S H has been removed f r o m t h e e q u a t i o n The regression equation is C-f = 265 + 3378 pH + 4793 pH * N a C l - 7652 N a C l - 54.6 N a C l * So + 280 So - 4.48 So * pH + 1730 R-SH - 4.42 R-SH*So - 85.3 T-SH*So - 197 SS + 0.0578 SS*So - 110 % s o l * N a C l - 88.1 % s o l * p H + 1.16 % s o l * S o + 127 % s o l u b l e - 22.6 % s o l * R - S H + 2 . 3 4 % s o l * S S - 0.222 V i s . -0.000570 V i s * s o l +0.000032 V i s * S o + 0.0260 V i s * N a C l - 0.00330 V i s * p H + 0.00384 V i s * S S P r e d i c t o r Coef SE Coef T p C o n s t a n t 264.622 8.691 30 45 0 001 PH 3377.52 22 .42 150 64 0 000 * * * pH * NaC 4793.48 22.55 212 57 0 000 * * * N a C l -7652.16 49.37 -154 98 0 000 * * * N a C l * So -54.6285 0.2321 -235 42 0 000 * * * So 280.205 2 .272 123 32 0 000 * * * So * pH -4.48411 0.03435 -130 54 0 000 * * * R-SH 1729.69 10.73 161 13 0 000 * * * R-SH*So -4.42169 0.04785 -92 42 0 000 * * * T-SH*So -85.3353 0.6614 -129 02 0 000 * * * SS -196.984 0.955 -206 26 0 000 * * * SS*So 0.0577907 0.0002831 204 11 0 000 * * * % s o l * N a C l -109.746 0.471 -233 06 0 000 * * * % s o l * p H -88.1297 0.4287 -205 59 0 000 * * * % s o l * S o 1.15880 0.00624 185 73 0 000 * * * % s o l u b l i t y 126.608 0.749 169 06 0 000 * * * % s o l * R S H -22.5847 0.2070 -109 13 0 000 * * * % s o l * S S 2.33933 0.00858 272 51 0 000 * * * V i s -0.22213 0.06917 -3 21 0 085 * V i s * s o l -0.00057038 0.00002015 -28 31 0 001 * * * V i s * S o 0.00003155 0.00003103 1 02 0 416 V i s * N a C l 0.025981 0.002402 10 82 0 008 V i s * p H -0.003300 0.002343 -1 41 0 294 V i s * S S 0.003845 0.001160 3 31 0 080 * V i s * T - S H -0.011539 0.008392 -1 37 0 303 V i s * R - S H 0.0336 0.2399 0 14 0 901 Ho -0.04172 0.04448 -0 94 .0 447 H o * % S o l -0.00025413 0.00008699 -2 92 0 100 H o * v i s 0.00008091 0.00005398 1 50 0 273 S = 0.008859 R - S q = 100.0% R - S q ( a d j ) = 100.0% *** s i g n i f i c a n c e a t p < 0.001 ** s i g n i f i c a n c e a t p < 0.05 * s i g n i f i c a n c e a t p < 0.1 207 Analysis of Variance Source DF SS MS F P Regression 23 290.749 12.641 161068.95 0.000 *** Residual Error 2 . 0.000 0.000 Total 25 290.750 1 s t Stepwise Regression: C-f versus pH, pH * pH, Alpha-to -Enter : 0. 2 Alpha-to-Remove: 0.2 Response i s C - f on 35 pr e d i c t o r s , with N Step 1 2 3 4 Constant 6. 807 4 . 407 1. 787 2 . 126 T S H * S o -0.00389 -0.00487 0.00072 T-Value -3 .62 -5 .23 0 .29 P-Value 0. 001 0. 000 0. 772 % S o l * R S H 0. 085 0. 147 0. 141 T-Value 3 .49 4 .36 5 .56 P-Value 0. 002 0. 000 0. 000 % s o l * S o -0.00027 -0.00024 T-Value -2 .44 -6 .38 P-Value 0. 023 0. 000 S 2 .80 2 .31 2 . 10 2 .05 R-Sq 35 .30 57 .73 66 .72 66 .59 R - S q ( a d j ) 3 2 . 6 1 54 . 0 6 62 . 1 9 6 3 . 6 9 PRESS 23 6. 535 171. 630 137. 812 136. 758 R-Sq(pred) 18 .65 40 .97 52 .60 52 .96 best a l t . Va r i a b l e So % s o l u b l % s o l *pH T-Value -3 .62 3 .31 -1 .74 P-Value 0. 001 0. 003 0. 096 Varia b l e SS *So % s o l *SS Vis *So T-Value -3 .60 3 .31 1 .27 P-Value 0. 001 0. 003 0. 218 26 208 2n d Stepwise Regression for C-f (pH, NaCl in very step) Alpha-to-Enter: 0.15 Alpha-to-Remove: 0.15 Response i s C-f on 39 pr e d i c t o r s , with N = 26 Step 1 2 3 4 5 Constant 10.099 8.743 11.336 21.023 24.752 pH -0.99 -1.48 -1.79 -1.71 -1.68 T-Value -3.19 -5.24 -5.53 -5.42 -5.41 P-Value 0.004 0.000 0.000 0.000 0.000 NaCl -3.04 -2.26 -7.42 -9.34 -10.96 T-Value -2.89 -2.60 -2.41 -2.95 -4.48 P-Value 0.008 0.017 0.025 0.008 0.000 %Sol*RSH 0.091 0.067 0.029 T-Value 3.71 2.43 0.82 P-Value 0.001 0.024 0.423 pH * NaCl 0.99 1.59 1.95 T-Value 1.74 2.44 4.15 P-Value 0.096 0.024 0.000 Ho*vis -0.00002 -0.00003 T-Value -1.67 -2.94 P-Value 0.110 0.008 S 2.65 2.12 2.03 1.95 1.93 R-Sq 44.57 65.92 70.24 73.89 73.02 R-Sq(adj) 39.75 61.27 64.57 67.36 67.88 PRESS 209.947 144.192 132.033 129.858 126.959 R-Sq(pred) 27.79 50.41 54.59 55.34 56.33 best a l t . V a r i a b l e %sol*SS Vis*NaCl Ho^2 T-Value 3.59 1.58 -1.43 P-Value 0.002 0.129 0.167 Va r i a b l e V i s * s o l Ho"2 Ho T-Value 3.56 -1.38 -1.43 P-Value 0 . 002 0.183 0.167 APPENDIX D - Minitab™ Printout for Regression Analyses for FSI of the Major Fraction 1 Stepwise Regression for FSI Alpha-to-Enter: 0.2 Alpha-to-Remove: 0.2 Response i s FSI on 35 predi c t o r s , with N 21 S t e p 1 2 3 4 5 Constant -6.091 48. 314 71.491 65.555 87 . 175 p H * p H 0.668 2 . 833 3.706 3.506 3 . 707 T-Value 8.67 6 .95 10.14 10.82 11 .02 P-Value 0.000 0 . 000 0.000 0.000 0 . 000 P H -22 .7 -33 .7 -30.9 -33.0 T-Value -5 .35 -8.23 -8.39 -8 .72 P-Value 0. 000 0.000 0.000 0. 000 V i s * S o 0.00004 0.00007 0.00007 T-Value 4.10 5.21 5 .63 P-Value 0 . 001 0.000 0. 000 R - S H * S o -0.045 -0 . 047 T-Value -2.63 -2 .89 P-Value 0.018 0. 011 SS*SS -0.0039 T-Value -1 .56 P-Value 0. 141 S 5.51 3 .52 2.56 2.21 2 .12 R-Sq 79.81 92 .20 96.08 97 .27 97 .65 R - S q ( a d j ) 7 8 . 7 5 9 1 . 3 4 9 5 . 3 9 9 6 . 5 8 9 6 . 8 6 PRESS 713.925 283 . 099 178.012 136.636 140. 579 R-Sq(pred) 74.98 90 .08 93 .76 95.21 95 .07 best a l t . Va r i a b l e %sol*pH Vis*NaCl pH * NaC T-SH"2 SS T-Value 7.23 3 .20 3 .92 1.62 -1 .54 P-Value 0.000 0. 005 0.001 0.125 0. 143 Vari a b l e PH NaCl * N Vis*pH R-SH~2 Vis *SS T-Value 6.87 3 .12 3.91 -1.61 -1 .26 P-Value 0.000 0. 006 0.001 0.126 0. 226 Vari a b l e Vis*pH NaCl Vis~2 T-SH R -SH T-Value 6.07 3 .12 3.81 1.61 1 .09 P-Value 0.000 0. 006 0.001 0.127 0. 293 Vari a b l e So * pH pH * NaC Vis*T-SH R-SH T -SH T-Value 5.84 2 .42 3 .73 -1.61 -1 .09 P-Value 0.000 0. 026 0.002 0.127 0. 293 210 R e g r e s s i o n A n a l y s i s for FSI (s ignif icant p a r a m e t e r s + N a C l ) The regression equation is FSI = 85.0 - 32.2 pH + 3.64 pH * pH + 0.31 NaCl +0.000070 Vis*So - 0.0475 R-SH*So - 0.00392 SS*SS P r e d i c t o r Constant PH pH * pH NaCl Vis*So R-SH*So SS*SS Coef 85.05 -32.208 3.6417 0.307 0.00007023 -0.04750 -0.003915 SE Coef 24.12 7.772 0.6371 2.518 0.00002308 0.01699 0.002617 T ,53 .14 .72 0.12 04 80 50 0.003 ** 0.001 *** 0.000 **i 0.905 0.009 ** 0.014 ** 0.157 S = 2.189 R-Sq = 97.6% R-Sq(adj) = 96.6% A n a l y s i s of Variance Source DF SS MS Regression 6 2786.40 464.40 Residual E r r o r 14 67.11 4.79 To t a l 20 2853.51 F 96.88 0.000 *** 2n d S t e p w i s e R e g r e s s i o n for FSI ( N a C l , p H in e v e r y s tep) Alpha-to-Enter: 0.15 Alpha-to-Remove: 0.15 Response i s FSI on 39 p r e d i c t o r s , w i t h N 21 S t e p Constant 1 -29.23 2 36.31 3 24.12 4 45.10 5 -17.86 6 -52 .23 -116.13 PH T-Value P-Value 7 .41 8.21 0.000 -19.47 -5.58 0.000 -15.38 -5.18 0.000 -16.80 -5.76 0.000 4.80 0.35 0.734 16.44 18.27 0.000 16.28 29.41 0.000 NaCl T-Value P-Value 7 .1 2.56 0.019 4.7 3 .41 0.003 7 . 0 5.54 0.000 7.4 6.09 0.000 27 .0 2 .18 0.047 37.1 12.94 0.000 126 . 8 7.14 0.000 pH * pH T-Value P-Value 2.58 7.76 0.000 2 .42 9.19 0.000 2.58 9.73 0.000 0.91 0.84 0.414 %Sol*RSH T-Value P-Value -0.187 -3 .46 0.003 -0.203 -3 .91 0.001 -0.262 -4.23 0.001 -0.294 -6.02 0.000 -0.570 -9.18 0.000 ss*ss T-Value P-Value -0.0042 -1.72 0.105 -0.0049 -2.07 0.057 -0.0051 -2.19 0.045 -0.0112 -5.98 0.000 NaCl *So T-Value P-Value -0.0481 -1.59 0.134 -0 . 0728 -10.41 0.000 -0.3807 -6.26 0.000 So T-Value P-Value 0.344 5.08 0.000 211 S 5.77 2.79 2.17 2.05 1.95 1.93 1.19 R-Sq 78.98 95.37 97.35 97.79 98.13 98.03 99.31 R - S q ( a d j ) 7 6 . 6 4 9 4 . 5 6 9 6 . 6 9 9 7 . 0 5 9 7 . 3 3 9 7 . 3 8 9 9 . 0 1 PRESS 795.897 195.399 130.575 134.277 132.124 126.304 41.6178 R-Sq(pred) 72.11 93.15 95.42 95.29 95.37 95.57 98.54 b e s t a l t . V a r i a b l e T - Value P-Value V a r i a b l e T - Value P-Value So % s o l * S o -6.86 -2.65 0.000 0.018 T-SH*So V i s * N a C l -6.83 0.000 2.56 0.021 SS V i s * s o l -1.71 1.31 0.108 0.213 SS*So % s o l u b l -1.38 0.189 1.10 0.289 SS*So 4.78 0.000 T-SH*So 4.25 0.001 S t e p C o n s t a n t 8 -101.7 9 -100.1 10 188.9 11 383.1 12 417.0 13 393 .7 14 405.8 P H T-Value P-Value 11.06 18.75 0.000 11.41 21.46 0.000 13 . 67 14.76 0.000 9.94 8.81 0.000 11.63 9.42 0.000 13.03 9.30 0.000 14.70 10.09 0.000 N a C l T-Value P-Value 114.8 17.12 0.000 112.9 19.28 0.000 116.3 23.92 0.000 101.5 20.63 0.000 102.8 24.28 0.000 103 .4 26.66 0.000 101.6 29.52 0.000 pH * pH T-Value P-Value %SOl*RSH T-Value P-Value -0.366 -11.60 0.000 -0.371 -13.56 0.000 -0.466 -11.45 0.000 -0.282 -5.24 0.000 -0.307 -6.49 0.000 -0.316 -7 .27 0.000 -0.304 -8.11 0.000 ss*ss T-Value P-Value -0.01247 -17.62 0.000 -0.01263 -20.51 0.000 -0 . 01221 -23 .59 0.000 -0.01225 -36.00 0.000 -0 .01223 -42.14 0.000 -0.01221 -46.08 0.000 -0 . 01483 -11.14 0.000 N a C l * S o T-Value P-Value -0.477 -19.26 0.000 -0.484 -22 .38 0.000 -0.445 -20.08 0.000 -0.477 -28.61 0.000 -0.471 -32.36 0.000 -0.467 -34.73 0.000 -0.467 -40.72 0.000 S o T-Value P-Value 0.397 15.42 0.000 0.402 17.96 0.000 0.378 19.05 0.000 0.406 27 .43 0.000 0.401 31.34 0.000 0.371 17.50 0.000 0.350 16.63 0.000 p H * N a C l T-Value P-Value 6.93 9.42 0.000 7.70 10.72 0.000 5.28 5.03 0.000 9.31 7.53 0.000 8.80 8.15 0.000 8.48 8.47 0.000 8.81 10.13 0.000 V i s * p H T-Value P-Value -0.00108 -2.32 0.039 -0.00123 -0.00107 -3.25 -4.23 0.008 0.002 -0.00276 -3 .43 0.007 -0.00447 -3.56 0.007 -0.00714 -4.17 0.004 T - S H T - V a l u e P-Value -86 -2.76 0.019 -139 -5.66 0.000 -151 -6.98 0.000 -144 -7.16 0.000 -148 -8.56 0.000 R - S H * S O T - V a l u e P-Value -0.0370 -3.93 0.003 -0.0345 -4.27 0.002 -0.0315 -4.13 0.003 -0.0346 -5.18 0.001 V i s ~ 2 T -Value 0.00001 2 .18 0.00001 2.35 -0.00001 -1.06 2 1 2 P-Value 0.057 0.047 0.325 Vis*So T-Value P-Value 0.00004 1.68 0.132 0.00007 2.78 0.027 Vis*SS T-Value P-Value 0.00048 2.00 0.086 S R-Sq R-Sq(adj) PRESS R-Sq(pred) best a l t . Va r i a b l e T-Value P-Value Vari a b l e T-Value P-Value 0.441 99.91 9 9 . 8 6 5 .17860 99.82 V i s * s o l -7.61 0.000 TSHA2 -7 .15 0.000 0.381 99.94 9 9 . 9 0 4.71471 99.83 Vis*TSH -2 .23 0.046 Vis*SS -2.20 0.048 0.306 0.201 99.96 99.99 9 9 . 9 3 9 9 . 9 7 5.95332 73.8266 99.79 97.41 RSH %sol*T-S 2.76 2.93 0.019 0.015 RSH~2 % so l u b l 0.171 99.99 9 9 . 9 8 57 . 0253 98.00 0.156 99.99 9 9 . 9 8 47.7523 98.33 Vis*SS So * pH 0.133 100.00 9 9 . 9 9 39.7420 98.61 Vi s 2.75 0.019 2 . 93 0.015 2.09 1.28 1.19 0.066 0.235 0.274 Vis Ho Vis*TSH 2.01 -0.94 1.11 0.076 0.373 0.302 S t e p 15 16 17 18 19 Constant 403.7 423.9 554.9 680.7 687.4 PH T-Value P-Value 14.05 10.55 0.000 14.70 11.68 0.000 12 .43 9.11 0.000 10.74 50.91 0.000 10.51 86.27 0.000 NaCl T-Value P-Value 102.64 31.03 0.000 101.33 32.94 0.000 97 .31 33 .31 0.000 109.51 136.79 0.000 109.82 264.54 0.000 pH * pH T-Value P-Value %SOl*RSH T-Value P-Value -0.3132 -8.49 0.000 -0.3133 -9.43 0.000 -0.2333 -5.51 0.001 -0.0934 -9.49 0.000 -0.0856 -15.85 0.000 ss*ss T-Value P-Value -0.01350 -29.47 0.000 -0.01746 -7.37 0.000 -0.02048 -9.18 0.000 -0.02259 -68.43 0.000 -0.02248 -131.63 0.000 NaCl *So T-Value P-Value -0.4662 -40.41 0.000 -0.4580 -39.94 0.000 -0.4684 -47.28 0.000 -0.4940 -247.74 0.000 -0.4952 -467.72 0.000 So T-Value P-Value 0.3600 19.21 0.000 0.3428 17 .40 0.000 0.3430 22 .48 0.000 0.3747 135.33 0.000 0.3775 238.82 0.000 pH *NaCl T-Value P-Value 8.590 10.10 0.000 8.384 10.82 0.00.0 9.147 13.44 0.000 7.877 66.59 0.000 7.856 129.88 0.000 Vis*pH T-Value P-Value -0.00598 -4.51 0.002 -0.00676 -5.28 0.001 -0.00593 -5.63 0.001 -0.00380 -20.12 0.000 -0.00355 -30.77 0.000 213 T - S H T-Value P-Value R - S H * S O T-Value P-Value Vis^2 T-Value P-Value V i s * S o T-Value P-Value V i s * S S T-Value P-Value V i s T-Value P-Value S o * S o T-Value P-Value V i s * s o l T-Value P-Value H o * % S o l T-Value P-Value S R-Sq R - S q ( a d j ) PRESS R-Sq(pred) best a l t . Vari a b l e T-Value P-Value Vari a b l e T-Value P-Value -147.0 -8.47 0.000 -147.7 -9.45 0.000 -179.6 -9.94 0.000 -214.5 -67.44 0.000 -216.4 -127.60 0.000 -0.0327 -0.0327 -0.0583 -0.1066 -0.1090 -5.04 -5.59 -4.99 -33.55 -62.85 0.001 0.001 0.002 0.000 0.000 0.00005 0.00006 0.00006 0.00008 0.00007 2.66 3.40 4.12 34.38 60.85 0.029 0.011 0.006 0.000 0.000 0.00024 0.00099 0.00153 0.00195 0.00193 3.21 2.21 3.69 31.44 60.55 0.012 0.063 0.010 0.000 0.000 -0.0482 -1.69 0.134 -0.0861 -3 .17 0.019 -0.1206 -28.52 0.000 -0.1202 -55.75 0.000 0.00006 2.38 0.055 0.00013 24.63 0.000 0.00013 46.68 0.000 -0.00016 -0.00016 -17.63 -34.80 0.000 0.000 -0.00000 -3.91 0 .017 0.134 0.121 0.0937 99.99 100.00 100.00 9 9 . 9 9 9 9 . 9 9 9 9 . 9 9 39.0956 31.7787 88.3559 98.63 98.89 96.90 0.0129 0.00658 100.00 100.00 1 0 0 . 0 0 1 0 0 . 0 0 2.50616 0.097187 99.91 100.00 Vis*T-SH -1.65 0.143 Vis*NaCl -1.10 0.306 % s o l u b l 2.29 0.062 %sol*T-S 2.29 0.062 Vis*R-SH -2.66 0.045 Vis*NaCl -2.28 0.071 Ho*vis -3 .67 0.021 Ho -3 .47 0.026 Regression Analysis for FSI (using significant parameters in above stepwise) T h e regression equation is F S I = 1005 + 65.2 pH - 4.73 pH * pH + 129 N a C l - 0.173 % s o l * R - S H - 0.0146 SS*SS - 0.586 N a C l * So + 0.218 So + 9.73 pH * N a C l - 0.00188 V i s * p H - 357 T-SH - 0.109 R-SH*So +0.000012 V i s ~ 2 +0.000023 V i s * S o +0.000266 V i s * S S - 0.0303 V i s +0.000467 So * So -0.000035 V i s * s o l +0.000003 H o * % S o l P r e d i c t o r Coef SE Coef T P C o n s t a n t 1004.8 120.2 8 36 0 014 PH 65.19 20.72 3 15 0 088 * pH * pH -4.734 1.793 -2 64 0 119 N a C l 129.486 7 .494 17 28 0 003 * * % s o l * R S H -0.17313 0.03296 -5 25 0 034 * * SS*SS -0.014599 0.003003 -4 86 0 040 * * N a C l * So -0 .58570 0.03447 -16 99 0 003 * * So 0.21797 0.06042 3 61 0 069 * pH * N a C l 9.7286 0.7135 13 64 0 005 * * V i s * p H -0.0018810 0.0006207 -3 03 0 094 * T-SH -356.54 53.11 -6 71 0 021 * R-SH*So -0.109409 0.001200 -91 16 0 000 * * * V i s * 2 0.00001209 0.00000440 2 75 0 111 V i s * S o 0 .00002275 0.00001943 1 17 0 362 V i s * S S 0.0002655 0.0006348 0 42 0 716 V i s -0.03033 0.03470 -0 87 0 474 So * So 0.0004669 0.0001277 3 66 0 067 * V i s * s o l -0.00003493 0.00004646 -0 75 0 531 H o * % S o l 0.00000256 0 . 00000244 1 05 0 404 S = 0.004216 R - S q = 1 0 0 . 0 % R - S q ( a d j ) = 1 0 0 . 0 % A n a l y s i s o f V a r i a n c e S o u r c e DF SS R e g r e s s i o n 18 2853.51 R e s i d u a l E r r o r 2 0.00 T o t a l 20 2853.51 MS F 158.53 8.918E+06 0.00 0 . 0 0 0 *** 215 

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.831.1-0089714/manifest

Comment

Related Items