IIIIIIIIIIIIE ‘‘‘‘‘

 

MSU is an Affirmative Action/Eq ual Opportunity Institution

MICHIGAN STATE LIBRARIES

llllllllllllll‘ll ll \‘lll‘l l. “Will

3 1293 01826 4758

LIBRARY
Michigan State
University

 

 

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

COMPOSITIONAL, FUNCTIONAL AND NUTRITIONAL CHARACTERIZATION
OF ULTRAFILTRATION PROCESSED COWPEA (Vigna unguiculata)
AND NAVY BEAN (Phaseolus vulgaris) PROTEIN FRACTIONS

 

presented by

JOSE CANDACE JACKSON

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Food Science

flax/(yaw

Major professor7

Date May 14, 1292

012771

6

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1!” WWW“

 

COMPOSITIONAL, FUNCTIONAL AND NUTRITIONAL CHARACTERIZATION
OF ULTRAFILTRATION PROCESSED COWPEA (Vigna unguiculata)
AND NAVY BEAN (Phaseolus vulgaris) PROTEIN FRACTIONS

Jose Candace Jackson

A DISSERTATION

Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of

DOCTOR OF PHILOSOPHY
Department of Food Science & Human Nutrition

1999

CW?

 

 

 

ABSTRACT
COMPOSITIONAL, FUNCTIONAL AND NUTRITIONAL CHARACTERIZATION
OF ULTRAFILTRATION PROCESSED COWPEA (Vigna unguiculata)

AND NAVY BEAN (Phaseolus vulgaris) PROTEIN FRACTIONS

By

Jose Candace Jackson

Ultrafiltration technology is increasingly being used to simultaneously purify,
concentrate and fractionate macromolecules without the application of heat or the use of
either extreme chemical or physical conditions. This is particularly important in isolating
proteins since it results in very little modification of structure and functionality.
Ultrafiltration is now a unit operation in the dairy industry. In the vegetable protein
industry, most of the research has been limited to soy protein. There is thus tremendous
potential to evaluate the effects of ultrafiltration processing on the physico-chemical
properties of proteins fi'om under-utilized leguminous species.

The optimum conditions for aqueous extraction of protein from milled cowpea
and navy bean seeds was determined using particle sizes 0.79mm - 6.35mm, pH values 2-
12 for 60 minutes each at 25 or 50°C. The legume flours were extracted three times and
the extracts combined and pumped through a plate and flame ultrafiltration system. The
flow rate (L/hr), flux (L/mthr), protein content, recovery, molecular weight
characterization and functional properties of freeze dried fractions were compared to a
commercial soy protein isolate (SP1). The protein quality of the residue remaining after

aqueous alkali extraction was determined. Legume and wheat flour diet blends were

made up to 10% protein and contained 30%, 70% and 100% by weight of the COWpea and
navy bean residue supplemented with whole-wheat flour to 100%. Arrowroot starch was
added as the major carbohydrate source. A 2% albumin diet was used to determine

metabolic nitrogen, and a modified AIN-93G diet used as a control.

Alkaline pH was more effective than acidic pH in extracting protein. Particle size
had a highly significant inverse relationship and temperature had no effect. Results
indicate that the optimum protein extraction conditions were particle size of 1.59mm, at
pH 10 and 25°C. Permeation flux ranged fi'om 0.18 - 4.03 L/mZ/hr and 3.12 — 24.80
L/mZ/hr for cowpea and navy bean, respectively. Protein content of the fractions ranged
fi'om l8 - 53%. About 84% of the protein in the extract was recovered in the cowpea
protein isolate (CPI), and 80% in navy bean isolate (NBI). Sodium dodecyl sulfate
polyacrylamide gel electrophoresis (SDS PAGE) indicated similar discrete band patterns
for all protein fiactions. The physico-chemical properties of cowpea and navy bean
proteins were significantly affected by ultrafiltration (UF) processing, and there were
significant differences between the experimental fiactions and soy protein isolate (SP1).
The protein quality of all experimental diets was significantly lower than that of the
modified AIN-93G diet except the 30% cowpea diet. Diets with cowpea or navy bean as
the primary source of protein were considered poor quality protein sources. The
extraction and ultrafiltration system employed, provided discrete protein fractions that
can be used as ingredients in the food industry; the 30%CP wheat flour diet blend could

be recommended for use as food for pre-school children, ages 2 — 5 yr.

Copyfightby
JOSE CANDACE JACKSON

1 999

To my Mother - Evelyn

For your strength, courage and guidance

For always being there

Thank you

ACKNOWLEDGEMENTS

I would like to express my sincere gratitude and appreciation to my Graduate
Committee members: Dr. Mark Uebersax, Dr. Maurice Bennink, Dr. Perry Ng and Dr.
George Hosfield, for their guidance and encouragement throughout my graduate
program In particular, I am especially grateful to my Advisor “Dr. U” for encouraging
me to pursue my interests in both Food Science and International Development. Thank
you for providing the support, opportunity and independence for my professional growth.

I am forever grateful to my Mother. She has been a constant source of
encouragement and support throughout my life and is now also my fiiend. My other
family members, Karen, Uncle Cardon and Auntie Wyn — thanks for those weekly phone
calls, they were a great help. I am fortunate to have made truly great fiiends at MSU -
Leps, Jackie, Beverley, Dale, Melise, Carole, Rochelle, Chantalle - it was great meeting
you all, and I know that our fi'iendships will last through the end of time. Leps, in
addition to your fi'iendship, thanks for your help with interpreting the statistics of my
studies. My fiiends in the Department — Alicia, Sharon, Yong Soc, John and Kay, thanks
for being available to lend a listening car when things got rough with both school and life
in general, and also for lending a helping hand in the Processing Plant. 1 was fortunate
also to have received support fi'om the Organization of American States (OAS), and the
Office of the Ombudsman for my graduate education, thanks Julie, Stan and Joy.

Finally, but most importantly, my thanks for the spiritual guidance that I have

received fi'om my father, the Almighty God.

vi

 

 

[151

[\IF

Llli

Iltl

 

TABLE OF CONTENTS

LIST OF TABLES ...................................................................................... . x
LIST OF FIGURES ..................................................................................... xv

LITERATURE REVIEW ............................................................... 3

Food Legumes ..................................................................... 3

Production, Distribution and Consumption.............................. 3

Seed Structure .............................................................. 5

Chemical Composrtron 6

Protein ............................................................ 8
Carbohydrates ................................................... 10
Lipids, Mineral and Vitamin ................................... 12
Antinutn'tional Factors ........................................ l3
Legume Proteins .................................................................. l7
ProteinExtractionu 18
ProteinFunctionality 20
Nutntional Attributes of Legume Proteins ..... . . . . . . . . . 26
Membrane Separations ........................................................... 38
Membrane Module and System Structures.............................. 42
Ultrafiltration Processing ............................................... 46
MATERIALS AND METHODS ..................................................... 49
ProxrmateAnalysrs 49
Milling and Laboratory Scale Protein Extraction .................................... 49
Pilot Scale Protein Extraction_ ...................................................... 50
Ultrafiltration Processing of Aqueous Extracts ................................ 53
SDS Polyacrylamide Gel Electrophoresis.......................................... 56
Thermal Stability ................................................................. 58
Amino Acid Analysis ............................................................. 58
HCl Hydrolysis ........................................................... 58
Derivatization ............................................................ 59
HPLC Analysis ........................................................... 61
Methanesulfonic Acid (MSA) Hydrolysis - Tryptophan Analysis. 62
Forrnic Acid Oxidation - Analysis of Cysteine ....................... 63
Functional Properties ............................................................. 63
Oil and Water Absorption Capacity ................................... 63
Protein Solubility .......................................................... 64
Foaming Capacity and Stability .......................................... 64

vii

 

 

STI-I
PR1
It 1'

I)
.2

 

PRC
Cm

Emulsion Capacity and Stability ........................ . ................ 65
Gelation Characteristics ................................................... 65
Colorimetric Analysis ..................................................... 66
Nutritional Characterization ....................................................... 66
Screening for Phytohemagglutinins ...................................... 66
In-vitro protein digestiblity ................................................ 67
pH-Drop Assay .................................................... 67
pH-Stat Assay ...................................................... 69
In-vivo protein digestibility ............................................... 71
Preparation of Cowpea and Navy Bean Residue ............... 71
Preparation of diets for evaluation .............................. 71
Feeding Study ...................................................... 73

Food Consumption and Rat weight .............................. 73
Calculation ofProtein Quality 77
PDCAAS Calculation and interpretation ....................... 77
Statistical Analysis ......................................................... 78

STUDY 1. OPTIMIZATION OF THE AQUEOUS EXTRACTION OF

PROTEINS FROM COWPEAS (Vigna unguiculata) AND

NAVY BEANS (Phasedus vulgaris) ................................................... 79
Abstract .............................................................................. 79
Introduction............. 80
MatenalsandMethods 81
ResultsandDrscussron 83
Conclusron 92
FutureResearch ...................................................................... 93

STUDY 2. ULTRAFILTRATION PROCESSING, CHARACTERIZATION

AND FUNCTIONAL PROPERTIES OF COWPEA (Vigna unguiculata)

AND NAVY BEAN (Phaseolus vulgaris) PROTEIN FRACTIONS .............. 94
Abstract ............................................................................... 95
Introduction 96
Materials and Methods98
Results and DrscussronlOO
Conclusion 129
Future Research ...................................................................... 131

STUDY 3. NUTRITIONAL QUALITY OF COWPEA AND NAVY BEAN

PROTEIN DIETS BY EV VITRO AND IN VIVO PROTEIN DIGESTIBILITY

CORRECTED AMINO ACID SCORE (PDCAAS) ................................. 132
Abstract ............................................................................... 132
Introduction 134
Materials and Methods 136
Results and Discussion 138

viii

 

5U

AP?

US"

 

 

Conclusron156

Future Research ...................................................................... 156
SUMMARY AND RECOMMENDATIONS .................................................... . 157
APPENDICES ............................................................................... 160

APPENDIX 1 ........................................................................ 161

APPENDIX 2 ........................................................................ 173

APPENDIX 3 ........................................................................ 188
LIST OF REFERENCES ............................................................................... . 199

ix

.g‘
u

 

-\ .\ i .v
. . ..A - . ... . . in - 1

.1m .1‘~ We...» .. ya!“ .4‘ a.“ .4‘ ..Il
. . .. . r . . T r I T; T1 Tr

 

. p .h.
Ti...

Table 1

Table 2

Table 3
Table 4
Table 5
Table 6

Table 7

Table 8
Table 9
Table 10

Table 11

Table 12

Table 13

Table 14
Table 15
Table 16
Table 17

Table 18

LIST OF TABLES

Production of Legumes in Developing and Developed Nations of the
World

Proxirnate Analyses (%) and Nutritional Values of selected
Leguminous seeds

Protein Contents (%) of Selected Food Legumes

Lectin and Trypsin inhibitor activities of selected legume flours
Composition of proteins in selected leguminous seeds

Available methods for testing protein nutritional quality

Comparison of Suggested Patterns of Amino Acid Requirement for
Humans with that of the rat

Preparation of Running and Stacking Gels in SDS PAGE Analysis
Preparation of Blakesley Solution in SDS PAGE Analysis
Preparation of Reagents used for Amino Acid Analysis

Nitrogen content and Sample weight of a 10 mg Nitrogen sample of
Cowpea and Navy Bean Fractions after Ultrafiltration Processing

Preparation of Enzymes A and B used for pH Drop In-vitro
Digestibility of Cowpea and Navy Bean Protein Fractions

Preparation of Enzyme Solution used in the pH Stat In-vitro
Digestibility of Cowpea and Navy Bean Protein Fractions

Calculation of 10% Legume-Wheat Flour Diet Blends

Rat weight Ranking in the Feeding Study

Diet Assignment of Rats in the Feeding Study

Proximate Composition of Cowpea and Navy Bean Flour (%db)

Multiple Regression Summary of Fit Statistics of Protein Extraction
fi'om Cowpea and Navy Bean Flour

Page

l3
17
29

33

57
57

6O

69

7O

74
75
76
83

88

 

 

1 .

 

 

 

1"

 

Table 19

Table 20

Table 21

Table 22

Table 23

Table 24

Table 25

Table 26

Table 27

Table 28

Table 29

Table 30
Table 31

Table 32

Table 33

Analysis of Variance of the Whole Model Test of Protein Extraction
in COWpea and Navy Bean Flour

Correlation Matrix of Process Variables for Protein Extraction of
Cowpea and Navy Bean Flour

Protein content in Cowpea and Navy Bean Fractions (flour particle
size 1.59 mm) after three sequential extractions at pH 10 and 25°C

Protein Yield fiom Cowpea and Navy Bean Flour (particle size 1.59
mm) after three sequential extractions at pH 10 and 25°C

Protein Content (%db) and % Solids of Cowpea and Navy Bean
Protein Fractions after Aqueous Extraction and Utrafiltration
Processing

Analysis of Variance of Cowpea and Navy Bean Fractions afier UF
Processing

Transition Temperatures and Enthalpies of Cowpea and Navy Bean
Fractions afier Ultrafiltration Processing

Amino Acid Composition (mg/ 100g protein) of Cowpea and Navy
Bean Fractions

Protein Solubility of Cowpea and Navy Bean Protein Fractions at
pH 7

Water Absorption Capacity of Cowpea and Navy Bean Protein
Fractions

Oil Absorption Capacity of Cowpea and Navy Bean Protein
Fractions

Foam Capacity of Cowpea and Navy Bean Protein Fractions
Emulsion Capacity of Cowpea and Navy Bean Protein Fractions

Gelation Characteristics of Cowpea and Navy Bean Protein
Fractions alter Ultrafiltration Processing '

Factor Analysis Loading of 15 Functional Properties on the first four
factors

xi

Page

88

90

91

91

103

104

111

114

115

116

116

117
118

121

126

 

uh.
ad:
5.1

...w.

Table 34

Table 35

Table 36

Table 37

Table 38

Table 39
Table 40
Table 41

Table 42

Table 43

Table 44

Table 45

Table 46

Table 47

Table 48

Screening Cowpea and Navy Bean Protein Fractions for
Phytohaemaglutinins (Lectins)

In-virro Protein Digestibility of Cowpea and Navy Bean Fractions
after Ultrafiltration Processing

Cowpea and Navy Bean Protein Diet Composition

Actual Protein Content and standard deviation of the Experimental
and Modified AIN-93G diets

Amino Acid Composition of Cowpea and Navy Bean Diets (mg /
100 g protein)

Essential Amino Acid Score for Cowpea and Navy Bean Diets
Relative Lectin Activity of Experimental Diets
In-vitro Digestibility and Statistical Analyses of Diets

Average Food Intake and Standard Deviation of the Experimental
Diets and Rat Growth over the Feeding Period

Protein content and weight of fecal matter during feeding study

Average in-vivo Apparent Protein Digestibility (AD) and True
Protein Digestibility (TD) and Standard Deviation of the
Experimental Diets

Estimates of Protein Nutritional Quality of Cowpea and Navy Bean
Diets fi'om Relative Protein Efficiency Ratio (RPER), Relative Net
Protein Ratio (RNPR), Relative True Protein Digestibility (RTD),
and Protein Digestibility Corrected Amino Acid Score (PDCAAS)

Protein Content, Predicted Protein Content, Soluble Solids, and
Protein Yield from Cowpea Flour after Extraction under various
conditions of pH, particle size, and temperature

Protein Content, Predicted Protein Content, Soluble Solids, and
Protein Yield from Navy Bean Flour afier Extraction under various
conditions of pH, particle size, and temperature

Contribution of pH, Particle Size, Extraction Temperature and Bean
Type to Protein Content in Cowpea and Navy Bean Extract

xii

Page

127

128

139

140

141

143
144
146

148

150

151

152

162

165

168

Table 49

Table 50

Table 51

Table 52

Table 53

Table 54

Table 55

Table 56

Table 57

Table 58

Table 59

Table 60

Table 61

Table 62

Table 63

Permeate and Retentate Flux during Ultrafrltration Processing of
Aqueous Alkali Cowpea Extracts

Permeate and Retentate Flux during Ultrafiltration Processing of
Aqueous Alkali Navy Bean Extracts

Volume of Aqueous Alkali Cowpea Extracts during Ultrafiltration
Processing

F lowrate of Aqueous Alkali Cowpea Extracts during Ultrafiltration
Processing

Volume of Aqueous Alkali Navy Bean Extracts during
Ultrafiltration Processing
Flowrate of Aqueous Alkali Navy Bean Extracts during

Ultrafiltration Processing

Protein Recovered (% of Aqueous Extract) afier Extraction at pH 10
and 25°C

Retention times and Area under the peak (uVolt-Sec) of Amino
Acids of Cowpea and Navy Bean Fractions

Hunter Color Characteristics of Cowpea and Navy Bean Protein
Fractions after Ultrafiltration Processing

F -va1ues and Statistical Significance of Functional Properties of
Cowpea and Navy Bean Protein Fractions after UF Processing

Correlation Matrix of the Functional Properties of Cowpea and Navy
Bean Protein Fractions

Correlation Matrix of the Functional Properties of Cowpea Protein
Fractions

Correlation Matrix of the Functional Properties of Navy Bean
Protein Fractions

Peak Area and Retention times of Amino Acids of Cowpea and
Navy Bean Diets

Correlation of Protein Nutritional Quality Tests

xiii

Page

174

175

176

177

178

179

180

181

183

184

185

186

187

189

198

Figure 1
Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11
Figure 12

Figure 13

Figure 14

Figure 15

LIST OF FIGURES

Membrane Concepts
Schematic of Protein Extraction of Cowpea and Navy Bean Flour

Schematic of Pilot Plant Scale Extraction of Cowpea and Navy
Bean Flour

Flow diagram of Ultrafrltration Processing of Cowpea and Navy
Bean Aqueous Alkali Protein Extracts

Ultrafiltration Processing Cleaning Program recommended for
Protein Products

Processing scheme for Production of Cowpea and Navy Bean
Residue Ingredients for m-vivo protein digestibility assay

Cq)wpea Protein Extraction Curve, pH and Particle size Effects at
25 C

Cq)wpea Protein Extraction Curve, pH and Particle size Effects at
50 C

Navy Bean Protein Extraction Curve, pH and Particle size Effects at
25°C

Navy Bean Protein Extraction Curve, pH and Particle size Effects at
50°C

Flux of Cowpea Proteins during Ultrafiltration Processing
Flux of Navy Bean Proteins during Ultrafiltration Processing

Mass Balance of Cowpea Protein Fractions after Ultrafiltration
Processing

Mass Balance of Navy Bean Protein Fractions afier Ultrafiltration
Processing

SDS PAGE Pattern of Cowpea and Navy Bean Protein Fractions
afier Ultrafiltration Processing

xiv

Page

51

52

54

55

72

85

85

86

86

101
102

105

105

108

 

-11

’ "'4

' [’11

.7.
'1

'T‘
l'v1

 

 

Figure 16

Figure 17
Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Figure 23

Figure 24

Figure 25

Figure 26

Figure 27

Figure 28

Figure 29
Figure 30

Figure 31

SDS PAGE Pattern of Cowpea and Navy Bean Permeate Fractions
after Ultrafiltration Processing

Foam Stability of Cowpea and Navy Bean Protein Fractions
Emulsion Stability of Cowpea and Navy Bean Protein Fractions

Hunter Color L-Value (Lightness) of Cowpea and Navy Bean
Protein Fractions

Hunter Color a-Value (Redness-Greenness Characteristics) of
Cowpea and Navy Bean Protein Fractions

Hunter Color b-Value (Yellowness—Blueness Characteristics) of
Cowpea and Navy Bean Protein Fractions

Average Food Intake of Experimental Diets and Rat Growth over
the Feeding Period

Leverage Plot for the Whole-Model Test of pH, Temperature,
Particle Size, and Bean Type Effects on % Protein Extracted from
Cowpea and Navy Bean Flour

Residual Plot for the Whole-Model Test of pH, Temperature,
Particle Size, and Bean Type Effects on % Protein Extracted from
Cowpea and Navy Bean Flour

Leverage Plot for the Significant Effect pH on % Protein Extracted
fi'om Cowpea and Navy Bean Flour

Leverage Plot for the Significant Effect Particle Size on % Protein
Extracted from Cowpea and Navy Bean Flour

Leverage Plot for the Significant Effect Bean Type on % Protein
Extracted fi'om Cowpea and Navy Bean Flour

Texture Profile Analysis (TPA) Characteristics of 18% Cowpea and
Navy Bean Protein Gels

Food Intake of rats during feeding period 1
Food Intake of rats during feeding period 2

Food Intake of rats during feeding period 3

XV

Page

109

119
119

122

123

149

168

169

170

171

172

182

190
191

192

Figure 32
Figure 33
Figure 34
Figure 35

Figure 36

Food Intake of rats during feeding period 4

Change in Rat Weight during feeding period 1
Change in Rat Weight during feeding period 2
Change in Rat Weight during feeding period 3

Change in Rat Weight dun'ng feeding period 4

xvi

Page

194
195
196

197

‘11

 

 

 

 

1R;

‘4:

 

INTRODUCTION

The word legume is derived from the latin “legumen” which means seeds harvested
in pods. Alternative terminology for edible seeds of leguminous plants is “pulse” from the
latin word “puls” meaning pottage. The term food legumes is used to cover both the
immature pods and seeds as well as mature dry seeds used for human foods. According to
the Food and Agriculture Organization (FAO, 1979), the word legume can be used for all
leguminous plants. However for those containing small amounts of fat such as cowpeas

and navy beans, the term “pulse” is used, and for those containing a high proportion of fat
such as soybeans and peanuts, the term “leguminous oilseed” can be used (Salunkhe &
Kadam, 1989).

Legume seeds have been an important component of human diet and an
economical source of supplementary protein for many populations lacking animal protein
for centuries, particularly so in developing countries. The average protein content of
legumes (peas, chickpeas and dry beans) is about 22% compared with cereal crops such as
rice and wheat with 7 and 12% respectively. Food legumes are however still under-
utilized, primarily because of their prolonged cooking time requirements; deficiency of
sulfur-amino acids; antinutritional components such as hemagglutinins (lectins), enzyme
inhibitors, phytates, flatus factors, and tannins; as well as a low protein digestibility
(Salunkhe & Kadam, 1989).

Sulfur amino acid deficiency, in particular methionine, has been well reported in
the literature (Bressani & Elias, 1988; Deshpande & Damodaran, 1990, 1991; Bliss &

Hall, 1977). On average, legumes contain 1 g methionine / 100 g protein, the FAO

 

 

.
'3';
..11.

i

‘I
r

 

..." ‘?

 

 

reference pattern suggests an intake of at least 2.2 g/100 g protein for optimal nutrition
(Deshpande, 1992). Supplementation with the limiting amino acid has been recommended.
Antinutritional components in legumes are of great importance since they can limit the
nutritional potential for both human and animal consumption (Gatehouse, 1991). The two
main types of antinutritional components in legumes include the proteinase inhibitors
(primarily trypsin) and the lectins, and these have also been proven to be toxic in animal

studies.

The use of ultrafiltration (UF) technology to separate macromolecules is a
relatively new process in the food industry. It is an established unit operation in the
chemical, environmentaL and petroleum industries. In the food industry, the most
important application is for concentrating proteins in dilute solutions, and most of the
commercial applications involve the concentration of milk proteins and oil-seed proteins
such as soybeans. Thus, there is tremendous potential for commercial concentration of

proteins fi'om under-utilized legumes such as cowpeas and navy beans.
The dissertation research has been divided into three studies as listed below:
1 Optimization of protein extraction from selected legumes

2 UP processing of aqueous alkali protein extracts into cowpea and navy bean
protein fi'actions, and characterization and functional properties of the fi'actions

compared to a commercial soy protein isolate

3 Evaluation of the nutritional properties of legume-wheat protein diet blends
compared to a modified American Institute of Nutrition (AIN-93G) diet as the

control

 

 

fit

 

LITERATURE REVIEW

FOOD LEGUMES

Production. Distribution and Consumption

The major food legumes grown in all continents of the world include soybeans,
groundnuts, dry beans, peas, broad beans, chickpeas and lentils. Others such as pigeon
peas are grown only in some countries depending on the climatic conditions needed to
support the growth and food habits of the population (Kadam & Salunkhe, 1989; FAO,
1990). The production of major legumes in developing and developed countries is
indicated in Table 1. With the exception of soybeans and dry peas, production of legumes
was significantly higher overall in developing countries than in developed countries.

World legume production, particularly of pulses, declined during the late 1970‘s to
mid 1980's. This was due to the traditional caution of farmers which prevents them from
allocating too high a proportion of their land to pulses because of low yields, uncertain
harvests, slow maturation, and sensitivity of legumes to growing conditions at all periods
of development and severe losses caused by pests. In addition, methods of preparation and
cooking necessary to ensure a digestible product are often lengthy and costly in terms of
fuel consumption. However, as research on legumes increased, production has steadily
increased (Doughty & Walker, 1982; FAO Production Yearbook, 1996).

Food legumes are distributed throughout the world, probably because of their
unique capacity to fix atmospheric nitrogen. Peas, broad beans, and lentils are in general

more popular in the Middle East. Soybeans are consumed in large quantities in China,

I.

am

 

 

 

Japan, Indonesia and Central America; while chickpeas and pigeon peas are popular in
India, Bangladesh and Pakistan. Preference for one type of legume over another is
determined by availability in that region, which is influenced by the environmental
conditions that favor higher yields of certain species over another (Bressani, 1973; Kadam

& Salunkhe, 1989).

 

 

 

 

 

 

Table 1 Production of Legumes in Metric Tons (MT x 103) in Deve10ping and
Developed Nations of the World '
Legume Developed Developing Total (T)
Nations Nations

MTx103 %r M'rxlo3 %r M'rxlo3
Dry Beans 2,206 12.5 15,376 87.5 17,582
Dry Broad Beans 415 12.0 3,031 88.0 3,458
Chick Peas 395 4.9 7,607 95.1 8,007
Cowpeas 56 2.2 2,502 97.8 2,560
Ground nuts 1,980 6.2 29,722 93.8 31,708
Lentils 545 18.6 2,390 81.4 2,954
Dry Peas 8,516 77.6 2,454 22.4 1 1,048
Pigeon Peas --- --- 2,787 100.0 2,787
Soybeans 69,054 52.9 61,551 47.1 130,658
‘ FAO, 1996

 

.N
5'...
‘w

I! j

r ,‘1

4'1

 

 

 

Dietary surveys indicate that bean consumption is very high in some countries, and
that significant amounts of protein, calories and other nutrients are provided. However.
there is a limitation to the amount of legume foods that are consumed by humans because
they cause gastrointestinal disorders due to their bulk and low digestibility (Stanton et a1.

1966;1(adam & Salunkhe, 1989).

Seed Structure

Food legumes are classified into two categories: those in which energy is stored as
fat such as peanuts and soybeans; and those in which energy is stored as starch or gum
such as cowpeas and navy beans (Doughty & Walker, 1982). However the seed structures
in both types are similar. Generally, mature leguminous seeds have three major
components: the seed coat, cotyledons and embryo axis.

The outermost layer of the seed is the testa or seed coat, and accounts for about
7.7% of the total dry weight in the mature seed (Powrie et a1, 1960). The presence of
polyphenolic compounds, primarily tannins, in varied amounts determines the seed coat
color. The least amounts of tannins are shown in white beans like navy beans, and the
amounts increase in the colored beans (black, read and brown beans) like cowpeas. In
most legumes, the endosperm is only found at an early stage of development; in the
mature seed, it is reduced to a thin layer surrounding the cotyledons or embryo.
Characteristic external features include the hilium, micropyle', and raphe. The hilium is a
large oval scar near the middle of the edge, where the seed breaks away fi'om the stalk.

The micropyle is a small opening in the seed coat beside the hilium and is the original site

 

 

1‘.

N"

'9?

131.

of

 

 

where the pollen tube enters the valve. The raphe is a ridge at the side of the hilium
opposite to the micrOpyle and represents the base of the stalk, which by maturity has fused
with the seed coat (Doughty & Walker, 1982).

Bean seeds generally have a thick seed coat. The pltunule or embryonic stem is
fairly well developed in the resting seed and lies between two cotyledons or seed leaves.
The radicle or embryonic root has almost no protection except for that provided by the
seed coat. Thus, the seed is unusually vulnerable to breakage especially when dry and
roughly treated (Doughty & Walker, 1982).

Studies have shown that the seed coat of many beans is made up of thick
interwoven fibrous bundles, which is known as the cuticle. The hilium shows a great deal
of variation in shape and size ranging from round to oblong and to oval and elliptical. The
smaller seeds usually have a greater total surface area covered by the hilium. The
micropyle varies from circular and triangular to fork shaped. The inside surface of the seed
coat and the cotyledon surface have numerous hills and valleys, and appear to be

complementary structures (Deshpande, 1985).

Chemical Commsition

A wide variety of compositions exists for difl‘erent legumes (Table 2), and
this is governed by the cultivar, geographic location and growth conditions (Krober,
1968). They are characterized by a relatively large content of carbohydrates, ranging fi‘om
24-68%. The carbohydrates are either water-soluble components such as sugars and

pectins; or insoluble fi’actions such as starch and cellulose. Protein content is generally

about 20 - 45%. Lipid content ranges from 1-7% except in the oilseeds such as soybeans
and peanuts, which contain in some cases up to 50% lipids. Legumes are also good
sources of dietary fiber and minerals, being a particular rich source of calcium, iron and

water soluble vitamins such thiamin, riboflavin and nicotinic acid (Salunkhe & Kadam,

1989).

Table 2 Proximate Analyses (%)l and Nutritional Values of selected Leguminous
seeds

 

Legume H20 Ash Fiber Fat CHO Protein Nutritional Values

 

PER 2 _§\_/_3
Blackgram 9.7 4.8 3.8 1.0 57.3 23.4 60-64
Chick Pea 9.8 2.7 3.9 5.3 61.2 17.1 1.7 52-78
Cowpea 11.0 3.6 3.9 1.3 56.8 23.4 45-72
Groundnut 5.0 3.7 2.4 48.2 15.9 24.8
Mungbean 9.7 4.0 3.3 1.2 58.2 23.6 2.1 39-66

NavyBeans‘ 12.6 3.2 9.6 1.2 63.3 23.4
PigeonPea 10.1 3.8 8.1 1.5 57.3 19.2 1.5 46-74

‘ Siegel & Fawcett, 1976

2 Engel, 1978; and Luse & Rachie, 1979
3 Bressani, 1973

4 Uebersax and Occefia, 1991

 

 

Y’

or

(.I)

 

 

man

The proteins of legume seeds are either metabolic, structural or storage in nature.
The enzymatic (metabolic) and structural proteins are responsible for normal cellular
activities including the synthesis of structural proteins and storage proteins. The storage
proteins which are smaller in number, accounts for about 70% of the seed nitrogen
(Moose & Pernollet, 1982), and occur within the cell in discrete protein bodies (Pemollet,
1978). The genotype and the environmental conditions under which the legumes were
grown govern protein content. In most cases, variation between cultivars may be as much

as 10-15%. The range of protein contents in various legumes is indicated in Table 3.

 

 

Table 3 Protein Contents (%) of Selected Food Legumes l
Legume Protein Range

Chickpeas (Cicer arietinum) 14.9-29.6
Peas (Pisum sativum) 21.2—32.9
Faba beans (Vicia faba) 22.9-38.5
Cowpeas (V igna unguiculata) 20.9-34.6
W'mged beans (Psophocarpus tetragonolobus) 29.8-37.4
Pigeonpeas (Cajanus cajun) 18.8-28.5
Soybeans (Glycine max) 33.2-45.2
Lentils (Lens culinaris) 20.4-30.5

 

'Salunkhe& Kadam, 1985

 

 

9F

11.»

 

3‘5,

(.4-
p

.‘1.
'4

 

 

The crude protein content based on the nitrogen determination of legumes involves
the mixture of difi‘erent nitrogen compounds. Along with proteins, there are free amino
acids, amines, complex lipids, purine and pyrimidine bases, nucleic acids and alkaloids.
These compounds are classified as non-protein nitrogen (NPN) compounds, and have been
analyzed extensively in many legumes. The ratio of NPN to total seed nitrogen is in the
range of 10 - 15%, and it is this which influences the factor that is used in calculating
protein content for various protein sources (Earle & Jones, 1962; Gupta, 1982).

Proteins are located in the cotyledons and embryonic axis of beans with only a
small amount being present in the seed coat (Singh et a1, 1968). In navy bean seeds, the
seed coat contains about 4.8% while the cotyledon and embryonic axis contains about
27.5% and 47.6% crude protein. The cotyledons because of their greater weight
contribute the major amount of protein to the whole seeds. The outer cotyledon layer,
which is about 60% by weight, was found to be richer in protein (Zimmerman et a1, 1967).

Proteins have traditionally been classified as globulins, prolamins, albumins and
glutelins based on their solubility (Osborne, 1907) properties. Globulins are soluble in
dilute salt solutions, prolamins are soluble in 70% in ethyl alcohol, albumins are water
soluble, and glutelins are soluble in diluted acids or bases. Storage proteins in most legume
seeds are mainly composed of globulins, which acts as the carbon and nitrogen source
during germination. These can be further sub-divided into phaseolin (vicilin), phaselin and
conphaseolin (Bressani, 1975 and Kay, 1979). Phaseolin has been reported to have
between three to five subunits ranging in size fiom 23KD to 56KD (Pusztai and Watt,

1970 and Derbyshire et al, 1976).

Studies on the amino acid composition of leguminous seeds, in particular S-amino
acid deficiency (methionine), has been well reported in the literature (Bliss & Hall, 1977;
Bressani & Elias, 1988; Deshpande & Damodaran, 1990, 1991). Legume proteins are
mainly deficient in sulfur-containing amino acids (methionine) and tryptophan, but are rich
in lysine, in which cereals are relatively deficient. In many agricultural systems throughout
the world, legumes and cereals have been linked since they complement each other
nutritionally. On average, legumes contain 1 g methionine/ 100 g protein, the FAO
reference pattern suggests an intake of at least 2.2 g/ 100 g protein for optimal nutrition
(Deshpande, 1992). Supplementation with the limiting amino acid has generally been
recommended. The cotyledon as the major component of the seed accounts for 93% of
methionine and tryptophan of the whole seed, while the seed coat is poorest in these
amino acids. The embryo is rich in methionine and tryptophan but it contributes only about

2.5% of their total quantity in the seed (Kapoor & Gupta, 1977).

Carbohydrates

The total carbohydrates of dry legumes range fi‘om 24% in winged beans to about
68% in coma, including mono- and oligosaccharides, starch and other polysaccharides.
Starch is the primary carbohydrate, and varies from 24% in wrinkled peas to 56.5% in
pinto beans. Soybean, lupine and winged bean are reported to have the lowest starch
content from 0.2 - 6.5%. The starch is believed to be embedded in a dense proteinaceous
matrix, and the average size of the native bean starch granule ranges fiom 2S - 28 um

depending on the variety of bean (Kawamura et a1, 1955).

10

Legume seeds are also reported to contain oligosaccharides such as raffinose.
stachyose, and verbascose, the predominance of each type depends on the type of legume.
Verbascose is the major oligosaccharide in broad beans and mung beans; whereas
stachyose is found in peas, red kidney beans, cowpeas, soybeans and lentils. These
oligosaccharides have reported to be involved in flatulence production in man and animals,
although this is not harmfiil, it is a social discomfort to many people (Rockland et a1,
1969; Levine, 1979; and Fleming, 1981). These sugars cannot be absorbed through the
intestinal wall and cannot be digested by humans because the intestinal tract does not
contain the enzyme - ct-l,6-ga1actosidase that is required to split these oligosaccharides
into simple sugars. These oligosaccharides pass through the gut and small bowel and
enters the colon where bacteria readily utilize them as fermentation substrates and produce
large amounts of carbon dioxide and hydrogen, and a small amount of methane (Levine,
1979)

The degree of flatulence appears to be related to the level of oligosaccharides in
beans, as well as other non-oligosaccharide substances such as fiber meddy et a1, 1984;
Fleming et aL 1980). Legumes contain an appreciable amount of crude fiber, which
consists of cellulose and hemicellulose. There is generally a large variation in fiber content
between different legumes (Table 2). Cellulose is the main component of fiber in red
kidney beans, navy beans, and cowpeas. In others such as lentils, broad beans and

mungbeans, hemicellulose is the major component (Ali et a1 1981).

11

 

 

 

Lipid
06380151111.
from 13“..
$11610sz
1.111 31710112:
liiltilmjc ac.
Milfmnce

F000
1013331411 a:
30t°0f1he 1
late diwetit
191-183th Il’l.’
mm] mm},
:3 h0573110015
lint alifim '1’
5315 In the 511;

Food:
in M W
Mitchell 1?

lb508116,, 3
IGngOr}. & 1.;

f“12811

 

Lipids, Minerals and Vitamins

Lipid content is only of significance in the oilseed legumes such as soybeans and
peanuts which contain about 20 and 40% lipids respectively; in general, the content ranges
from 1-7%. In mature legumes, a major portion of lipids are stored in oil bodies or
spherosomes or lipid containing vesicles in the cotyledons. Most legume lipids contain
high amounts of essential fatty acids, the most important of which include linoleic and
linolenic acids, which are required for growth, physiological functions and body
maintenance (Doughty & Walker, 1982).

Food legumes are good sources of minerals such as calcium, iron, copper, zinc,
potassium, and magnesium (Doughty & Walker, 1982). Potassium contributes about 25-
30% of the total mineral content, and is useful in particular for the diets of people who
take diuretics to control hypertension and who suffer from excessive excretion of
potassium through body fluid. Calcium values vary widely, depending on variety, climate,
cultural methods and mineral content of the soil. Legumes contain a significant amount of
phosphorus which is largely present in the form of phytic acid, an antinutritional factor
that affects the absorption and utilization of calcium through its precipitation as insoluble
salts in the stomach and duodenum (Makower, 1969; Deshpande, 1992).

Food legumes are good sources of Vitamin B (thiamine and riboflavin), and niacin,
but poor sources of Vitamin C (ascorbic acid) and Vitamin A (retinol) in the diet (Tapper
& Ritchey, 1981). Studies have shown that the availability of Vitamin Bo for intestinal
absorption is reduced due to the presence of nondigestible polysaccharides and lignin
(Gregory & Kirk, 1981). Legumes are also good sources of vitamin E (tocopherol) and

folic acid.

12

 

 

Shot ..
. It n
the 11111.71:th
animatrttion

lIilf‘lC 41, .1

1:1ka
\

Li‘éume Flo.
.\1
1011116} Bea;
CONN:
Pea:
Iaba Bean :
501188113
R3“ 301;:
Deg-{It‘d :
Indium;
2 “3613an

3 Udnet an.
30111161: lSi

Phflc.

 

Antinutritional Factors

 

Antinutritional components in legumes are of great importance since they can limit
the nutritional potential for both human and animal consumption. The two main types of
antinutritional components in legumes include the lectins and the proteinase inhibitors

(Table 4), and these have also been proven to be toxic in animal studies (Gatehouse,

 

 

1991).
Table 4 Lectin and Trypsin inhibitor activities of selected legume flours
Legume Flour Lectin activity Trypsin inhibitor activity
(units / mg dry matter) (T.U.I. / mg dry matter)
Kidney Bean ‘ 13.2
Cowpea 2 < 0.05 21.1
Pea 2 100 - 400 4.5 - 9.3
Faba Bean2 25 - 100 5.6- 11.8
Soy bean 3
Raw flour --- 70
Defatted flour 1600 - 3200 85
Industrial meal (toasted) 25 - 200 0.63 - 5.5

 

rValdebaouze et a1, 1980
2 Rouanet and Besancon, 1979
3 Boulter, 1981

Phytohemagglutinins (PHA) or lectins were identified as one of the first factors
involved in the toxicity of raw legtunes to laboratory animals (Jafl‘e and Vega Lette, 1968;
Pusztai and Palmer, 1977 and Wilson et a1, 1980). Phytohemagglutinins (PHA) are

tetrameric carbohydrate-binding proteins that exists as a mixture of five hybrids with

13

similar chemical properties but slightly different biological activities (Leavitt et a1, 1977).
The molecular weight of PHA of dry beans ranges from llSKD to lSOKD, and the two
subunits have molecular weights of 34KD and 36KD. Phytohemagglutinins (PHA) has
potent biological activity because of its ability to bind complex carbohydrates and other
glycoproteins. The adverse effects of PHA are the agglutination of erythrocytes and
bacteria, binding to intestinal epithelium, and the stimulation of lymphocyte formation

(Coffey, 1985).

It is believed that PHA releases intestinal amylase and lipase from binding sites on
the glycocalyx of the intestinal epithelium, inhibits activity of intestinal saccaharase and
brush border dipeptidases (Sandholrn and Scott, 1979; Rouanet and Besancon, 1979; and
Kim et a1, 1976). Additionally, it has been shown that feeding pure PHA to rats’ binds to
and disrupts the intestinal epithelium leading to the formation of abnormal microvilli and
damaged epithelial surfaces. This interference with intestinal surfaces and enzyme activity
results in reduced protein digestibility and inhibits growth due to the depressant effect on
appetite. The toxic effects, manifested as gastrointestinal discomfort, have also been
reported in humans from a review of food poisonings in Britain. Seven outbreaks were
attributed to food poisoning from kidney beans. In each case observed, the onset of
symptoms of nausea, vomiting, diarrhea, and abdominal pain was rapid (1 - 3 hrs). Eating

cooked and under-cooked beans were reportedly responsible (King et a1, 1980).

Cooking beans has been reported to inactivate much of the lectin activity in dry
beans. For complete elimination of the antinutritional efi‘ects, pre-soaking for at least 4 - 5

hrs, followed by heating either for 4 hr at 90°C, 90 min. at 95°C or 10 min. at 100°C is

14

required (RRI, 1982). Coffey et al (1985) reported on the lectin activity of low-
temperature cooked kidney beans, and found that activity was still present in beans
exposed to low temperatures for up to 12 hrs. This supported the work of Honavar et a1,
(1962) and Liener, (1958, 1962, 1976). They reported that lectin activity could be almost
wholly eliminated by conventional heat treatments, however fully cooked beans can still

contain a significant amount of the original lectin activity.

The proteinase inhibitors are substances that inhibit proteolytic enzyme activity and
are specific in their interactions with proteinases such as serine proteases, sulphydrl
proteases, metallo-carboxypeptidases and acid proteases, leading to pancreas hypertrophy.
The growth depression caused by this inhibitor may be the consequence of an endogenous
loss of essential amino acids being secreted by a hyperactive pancreas. Since pancreatic
enzymes such as trypsin and chymotrypsin are rich in sulfur containing amino acids, this
pancreatic hypertrophy causes the drain of body tissue with particular amino acids in order

to meet an increased need for the synthesis of these enzymes (Gatehouse, 1991 ).

There have been a number of papers which described the heat stability/lability of
these trypsin inhibitors (Estevez & Luh, 1985; Eicher & Satterlee, 1988; Esaka et a1,
1987; Dhurandhar & Chang, 1990). It was concluded that the activity could be destroyed
easily by 90% if the legumes were processed properly. This process involves at least 30-60
minutes in boiling water or autoclaving at 15 psi for 15-20 minutes (Deshpande, 1985).

Trypsin inlu'bitors fall into two main groups: those that have a molecular weight of
about 20,000-25,000 D with relatively few disulfide bonds; and those with only 6,000-

8,000 D and a high proportion of disulfide bonds (Liener, 1982; and Gatehouse, 1991). A

15

 

{:13 011 1r;
1113?: 1101 EL“
atrium] \;

P1111-
Berries that
2:211:11 cer'.
19851. P1113..-
reduction in
be M a a
3113811 IO 10.

30933311181 10 r

 

 

study on trypsin inhibitors in pigeon peas by Godbole et a1 (1994) suggested that they
were not associated with protein bodies. This suggested that the trypsin inhibitor could be
removed from the legume without affecting storage protein structure and thus improves

nutritional value of legumes.

Phytates, polyphenols, dietary fiber, and oxalates are dietary components in
legumes that influence mineral bioavailability. Polyphenols such as the tannins have been
ascribed certain beneficial effects such as lowering of blood-related disorders (Deshpande,
1985). Phytates interact with proteins, resulting in reduced protein solubility, and thus a
reduction in solubility dependent functional properties of the protein, particularly if it is to
be used as a food ingredient. Soaking of groundnut, pigeon peas, and chickpeas have been
shown to lower phytate levels by leaching out the ions into soak water as a result of a

concentration gradient (Igbedioh et a1, 1994).

A number of studies have been conducted using purification and separation
techniques such as ultrafiltration to remove antinutritional factors from legumes. In the
study reported by Berot et a1 (1987), the authors found that or-galactosides, antitrypsic
and hemagglutinating activity were lowered by submitting the protein extract to
ultrafiltration. Air classification of fababean flour resulted in protein fractions that
contained the larger proportion of trypsin inhibitors and lectins (Elkovvicz & Sosulski,
1982); and alkaline extraction and isoelectric precipitation reduced the activity to about
1/3. The protein micellar mass (PMM) procedure described by Murray et a1 (1978, 1981)
was reported to reduce the levels of antinutritional factors in the precipitated protein to

less than about 5% of the original levels.

16

 

 

11
II mic-3.1:
uhich are
The globe
coeficien:
ratertnel'
proportion
51 al 1981

“131111 10 161

 

LEGUME PROTEINS

The composition of proteins in selected leguminous seeds is presented in Table 5.
It indicates that storage proteins in most legume seeds are mainly composed of globulins,
which are about 90% in soybeans, and 60% and 66% for fababean and pea respectively.
The globulins are made up of two subunits, which are characterized by their sedimentation
coefiicients. They are the 7S and 11S globulins and are called vicilin and legumin
respectively. In soybeans and lupines, the oilseeds, the 7S-like protein is found in a larger
proportion than the 118-like protein; their ratios are 1.6:1 and 1.3:! respectively (Duranti
et a1, 1981). In fababeans and peas, the pulses, legumin is the major protein, with the

vicilin to legumin ratio being close to 1:2.

 

 

 

Table 5 Composition of proteins in selected leguminous seeds
Legume Albumin ‘ Globulin ‘ Glutelin ' Vicilin:

g /100g protein (g/100g protein) g/ 100g protein Legumin
Fababean 20 60 15 1:1.6 - 1:37 ’3
Pea 21 66 12 1:13 -1:4.2 3
Soybean 10 90 0 1.6:1 ‘
Phaseolus 15 75 10
Lupinus 10 - 20 80 - 90 0 13:13
‘ Boulter, 1977

2 Gatehouse et a1, 1980

3 Martensson, 1980

‘ Thanh and Shibasaki, 1976
3 Duranti et a1, 1981

17

 

 

commercial
from a 100;
hthericai
lieiogical it
Dr}
mind. '15?
17} 111mg :.
190.8011 f
process in

68°11 and

53131211112 ’_

 

 

Protein Extraction

There are a number of protein extraction techniques that are available for use
commercially for the production of food or food ingredients; or for purifying a protein
from a food for further study in a laboratory. Generally, extraction techniques utilize the
biochemical differences in protein solubility, size, charge, adsorption characteristics, and

biological affinities for other molecules (Smith, 1994).

Dry processes for extracting proteins includes mainly air classification. In this
method, light or fine protein fractions are separated from heavy or coarse starch fractions
by milling to a fine powder, followed by several runs through an air classifier (Vose et a1,
1976; and Tyler et a1, 1981). Colonna et a1 (1980) reported on a two-run air classification
process in which the protein fraction had a yield of 41% and 35% and a protein content of
68% and 56% for fababean and pea, respectively. Additionally, they found that legumes
with higher lipid contents in the initial flour and a broad distribution of starch granule sizes
such as wrinkled peas, tends to be less efficient when air classification is used for

separating the protein fiom the starch.

Wet processes for protein extraction have been extensively reported in the
literature (Anson and Pader, 1957; Flink and Christiansen, 1973; Murray, 1978; Murray et
a1, 1978, 1981; Berét et a1, 1987; and Suelter, 1985). They are generally based on the
differential solubility characteristics of proteins in solutions, which depend on the type and
charge of amino acids in the molecule. Proteins can be precipitated or solubilized by

changing buffer pH, ionic strength, dielectric constant or temperature (Smith, 1994).

18

 

The
|
Amen 3; i’
10. 10 3013
precipitatig
to produce

mih ext:
around pH

Pn-
PIOIe‘II 61.“
CW Sis of
man f;

Ialiphatic

 

The most common is the isoelectric precipitation method that has been patented by
Anson & Pader (1957). This involves alkali solubilization of the legume flour at pH = 7-
10, followed by centrifugation to remove insoluble components, then isoelectric
precipitation of the proteins at acid pH (3.5-4.5). Fan & Sosulki (1974) used this method
to produce isolates from nine legume flours, and was able to show that the proteins were
easily extractable in an alkaline solution and are less solubilized in the acid pH range

around pH 4.

Proteins have been extracted with salt solutions. Satterlee et a1 (1975) produced a
protein extract using 2% NaCl solution, followed by centrifugation at 9000 x g and
dialysis of the supernatant for 48 hrs. Chang and Satterlee (1979) reported on a protein
extract from 0.2% salt solution followed by precipitation at pH 4.0 at various
temperatures. Sathe and Salunkhe (1981) fiactionated bean flour using salt solubilization,
dialysis, and finally freeze-drying. They were able to obtain a protein isolate and the
isolated albumins and globulins from the bean protein. The protein content of the isolate

was 92.43%.

Proteins can also be separated on the basis of their size due to the wide range of
molecular weights (IOKD to over 1000KD). Separation actually occurs based on the
Stokes radius of the protein, and not on the molecular weight. Stokes radius is the average
radius of the protein in solution and is determined by protein conformation. Actual
extraction procedures that utilize separation by size include dialysis and ultrafiltration.
Dialysis is used to separate molecules in solution by the use of semi-permeable membranes
that permit passage of small molecules but not larger molecules. It is a simple process, but

relatively slow that requires at least 12 hr. Ultrafiltration is similar to dialysis, in that it

19

uses a semi-perrneable membrane, in this case however, separation occurs under an
applied pressure and is much faster. Molecules larger than the membrane cut-off are
retained and are called the retentate, while smaller molecules pass through the membrane

and are called permeate (Kosikowski, 1986; and Smith, 1994).

Protein Functionality

Functional properties of proteins have been defined as those physico-chemical
properties which give information on how a protein will behave in a food system
(Hermansson, 1979). Functional proteins are important ingredients in the food industry;
which is evidenced by the range of specialist ingredients traded and transported either as
food protein groups or as individual proteins. For instance milk proteins are available as
whole dried milk, casein or whey protein; egg white and wheat gluten are also available as
individual proteins. Currently, there is now increased interest in new fimctional proteins,
which has led to research on the fimctional properties of under-utilized protein sources

such as legumes.

According to Sathe et a1, (1984), protein functional properties can be classified

into three major groups from a food application standpoint:

a) hydration properties are dependent on protein-water interactions, and encompass
water absorption and retention, wettability, swelling, adhesion, dispersibility, solubility
and viscosity;

b) properties related to protein-protein interactions including precipitation, gelation, and

the formation of various other structures such as doughs; and

20

 

 

:1 surface

K:-
organoiep:
complex
chemical r
enummnc'
functional 1
mimic 1-.;
processing
filtering;

The
5181:. h 0
binding ca
fomiaticn ;

Polar 3mm
Ahholl’s'll 1

 

Wei] 35 int

c) surface properties such as surface tension, emulsification and foaming characteristics

Kinsella (1979) summarized the important functional properties of food proteins as
organoleptic, hydration, surface, and structural/rheological. These properties reflect
complex interactions between the composition, structure, conformation, and physico-
chemical properties of the proteins, other food components, and the nature of the
environment in which these are associated or measured. The factors that affect protein
functional properties can generally be divided into three areas of influence: those that are
intrinsic to the individual protein such as composition and conformation; and those
processing treatments such as heating and pH; and environmental factors such as water,

and temperature which afl'ect the application of the protein (Kinsella, 1982).

The ability of proteins to bind and immobilize food components like water and
lipids, is one of the most important fimctional properties in many food systems. This
binding capacity is influenced by pH and ionic strength, and affects adhesion, film
formation and viscosity (Kinsella, 1976). The amino acid composition, in particular, the
polar amino acids, of the protein have a significant influence on protein-water interaction.
Although they are the primary sites for protein-water interactions, electrostatic effects, as
well as interaction between the amino acid residues themselves may also afi‘ect water

binding by proteins (Perutz, 1978).

Protein solubility data is very useful for determining optimum conditions for the
extraction and purification of proteins from natural sources, and for the separation of
protein fi'actions. It has been reported that pH, temperature, processing conditions, and

ionic strength affect solubility (Sathe & Salunkhe, 1981). Solubility behavior provides a

21

good index of the potential applications of proteins, because the degree of insolubility is
the most practical measure of protein denaturation and aggregation. Solubility is also an
important attribute of proteins selected for use in food beverages such as infant food

formulas.

The viscosity of a fluid reflects its resistance to flow, and is expressed as the
viscosity coeflicient 1.1, which is the ratio of the shear stress (1:) to the relative site of shear
or rate of flow (y). That is, r = u y. The main factor influencing the viscosity behavior of
protein fluids is the apparent diameter of the dispersed molecules or particles. This
diameter depends on the intrinsic characteristics of the protein molecule such as molar
mass, size, structure, electric charges; protein-solvent interactions which influence
solubility and swelling; and protein-protein interactions which determine the size of
aggregates. Viscosity generally increases exponentially with protein concentration because
of the high protein-protein interactions, and is an important fimctional property in fluid

foods such as beverages, soups, sauces and creams (Damodaran, 1994).

Protein gels are defined as three-dimensional matrixes or networks of intertwined,
partially associated polypeptides in which water is entrapped. The formation of gels is
important in many foods including coagulated egg white, soybean tofu, and milk casein
curd (Kinsella, 1976 and Schmidt, 1981). Damodaran (1988) has reported the stages that
occur during heat-induced gelation. The initial step requires prior heating of the protein,
which results in modification fiom the native state to a progel state. This involves
dissociation and denaturation of the protein, which allows the functional groups involved
in intra-molecular bonding to become available for inter-molecular bonding resulting in a

gel network. Gel networks can be of two types. Those that contain high levels of non-

22

 

 

p111 1:51;
1113;111:11-
Reiersibic
slut. $1.:
11,-.

1111 1101i

meant

5013:1011 is

 

 

polar residues and undergo random aggregation via hydrophobic interactions are opaque
coagulum-type gels with low elasticity and water-holding capacity (i.e. irreversible gels).
Reversible gels contain low levels of non-polar residues and is ordered, translucent.

elastic, and has high water-holding capacity (Damodaran, 1988).

The protocol for testing gelling ability of proteins has been reported by Matsumura
and Mori (1996). The first step involves solubilization or dispersion of various protein
concentrations in neutral or weakly acidic conditions with gentle stirring. The protein
solution is then heated in a boiling water bath for several hours and determination of the
gelling point determined using dynamic rheology apparatus. The evaluation of gel
properties is an important step in the process and this can be evaluated using scanning
electron microscopy (SEM), spectroscopic analysis, gel solubility experiments, rheological
measurements or measurements of fat-binding or water-holding capacity (Matsumura and

Mori, 1996).

Rheological properties have been reported as the most important factors
determining gel properties (Matsumura and Mori, 1996). The difficulty often lies in which
rheological measurement to choose for the analysis. Deformation mechanical tests have
been recommended such as texture profile, compression tests and tensile tests since they
give rise to many parameters fi'om which the required property can then be assessed.
These parameters include hardness, cohesiveness, adhesiveness, stringiness, gumminess,

chewiness, springiness and fi'acturability.

The behavior of proteins at interfaces influences the formation of food emulsions

and foams. Food products that contain both water and fat form thermodynamically

23

 

unstable n
emulsifier
10 be emt
mama
Many foo.
enuliion

protein 11‘;
unie-id an.
11.1110121113-
Mime

att’ibute :1

Si '.
(Kim and

 

 

 

 

 

Sudden d:
Melted
“16351365

111331. em 1

Sepl‘ar i0 !

Qe.‘ ‘ .
mwhx

unstable mixtures or emulsions, which can then be stabilized with amphiphilic molecules or
emulsifiers that are soluble in both water and non-polar solvents. Proteins are considered
to be emulsifiers, and are capable of coating lipid droplets and providing an energy barrier
to both particle association and phase separation during the formation of the emulsion.
Many food emulsions are expected to remain shelf-stable for months, thus the process of
emulsion stability or breakdown is important in the food industry. To be effective, the
protein must have adequate solubility, must be flexible in order to reach the interface,
unfold and orient itself with the hydrophilic groups towards the aqueous phase and the
hydrophobic groups towards the lipid phase. The ability of the protein to lose its tertiary
structure (denaturation) without loss in solubility is normally considered a positive

attribute for an emulsifier (Mangino, 1994).

Studies on protein stabilized emulsions have been well reported in the literature
(Kim and Kinsella, 1987; Halling, 1981; Swift et a1, 1961; Regenstein, 1988; and Pearce
and Kinsella, 1978). Measurement of emulsion capacity involves oil addition at a given
rate to a defined quantity of protein dispersion until there is a decrease in viscosity or
inversion. The change in viscosity is measured i) subjectively by visual appearance, ii)
sudden drop in viscosity or iii) sudden increase in electrical resistance. This method has
received considerable criticism since it requires an experienced operator, and also it
measures the ability of proteins to form emulsrons at protein-to-lipid ratios that are very

different fi'om what will be encountered in the finished product (Regenstein, 1988).

Measurements of emulsion stability mainly use the principle of oil and/or cream
separation over a specified time period at a stated temperature. The causes of emulsion

instability are coalescence, flocculation, gravitational creaming and Ostwald ripening, and

24

 

 

most stud.
and Afton
scattering
1333:1151. '

A
phase and
Protein 1r.‘
Philips. ‘1‘
protent to
5831611 (1".
charged 1
Infraction
rBorient 1c

df\610p (I.

 

 

most studies generally measure destabilization by coalescence and creaming (Petruccelli
and Anon, 1994). It is possible to follow the formation of layers by the eye, by light
scattering techniques, by confocal scanning light microscopy, and ultrasound (Blonk and

Vanaalst, 1993 and Dagom-Scaviner et a1, 1987).

A foam is a complex two-phase colloidal systems containing a continuous liquid
phase and a gas phase dispersed as bubbles or air cells (German and Phillips, 1994).
Protein interactions in foams have been widely reported in the literature (Graham and
Phillips, 1976; MacRitchie, 1978; Kinsella, 1981 and Kinsella and Phillips, 1989). For a
protein to be a successful foaming agent it must be able to stabilize the surface area being
created during foaming. It must be soluble and rapidly diffilse to the interface with the
charged, polar and non-polar residues correctly distributed for enhanced interfacial
interactions. The protein must be flexible to facilitate unfolding at the interface and
reorient to form a viscous film to maintain discrete bubbles until stabilizing interactions

develop (Prins, 1988; and Kinsella and Phillips, 1989).

Foaming capacity can be determined by three dynamic procedures: whipping,
shaking or sparging (Waniska and Kinsella, 1979; and Yasumatsu et al, 1972). The major
difference between these methods is the protein required; for whipping the amount ranges
from 3 - 40%, 1% for shaking and from 0.01 — 2% for gas sparging (Yasumatsu et a1,
1972). Once a protein foam is formed, its lifetime is dependent largely on the maintenance
of the viscoelastic adsorbed layer. If surface denaturation occurs, it results in an insoluble

coagulum and foam instability. Liquid drainage from the foam also contributes to foam

25

Due to the transient stability of protein foams, they are often diflicult structures to
study. Direct measurement of the foam volume, bubble size distribution and lifetime
provides a physical description of the foam. Indirect methods at the microscopic and
molecular levels provides information about why one protein has better foaming properties
than another does (Halling, 1981). Measuring the change in foam volume with time or the
volume of liquid which drains from the foam are the most popular direct methods for

assessing foam stability (Wang and Kinsella, 1976; and Graham and Phillips, 1976).

Nutritional Attributes of Legume Proteins

The primary function of dietary protein is to supply amino acids for the synthesis
of body proteins and other nitrogen-containing substances. The body metabolism of
proteins can be expressed by the difference between nitrogen intake and nitrogen
elimination - the nitrogen balance. If this difference is positive, as occurs during growth,
then nitrogen retention occurs through tissue deposition and protein synthesis. If it is
negative as occurs with malnutrition, injury or infection however, then nitrogen is lost. In
normal adults, the nitrogen balance is zero, because excess protein fi'om large intakes are

generally converted to energy and urea (Guthrie and Picciano, 1995).

Protein requirements in adults are assessed by measuring the minimal protein
intake that will maintain nitrogen equilibrium, and in infants or children, by measuring the
minimal protein intake that will provide an optimal rate of growth The Food and

Agriculture Organization and the World Health Organization of the United Nations

26

11.1011":
. 3111} :1

countries.

1
|

5311 p101:
person to
111‘. the pi

29791.

 

 

 

 

 

(FAG/WHO, 1973) and the NAS (1980), have recommended “safe protein intakes” for
healthy people which are in agreement with the results of N-balance experiments. In many
countries, particularly those in the developing world however, the value recommended for
safe protein intake is often overestimated to that which has been found necessary for a
person to remain in nitrogen balance. In these countries, protein intake is habitually low,

and the protein requirement is often lower, due to metabolic adaptation (UN University,

1979).

The protein nutritive value of a food corresponds to its ability to meet nitrogen and
amino acid requirements of the consumer, and to ensure proper growth and maintenance.
This is affected by the protein content, protein quality and the amino acid availability.
Foods with protein contents below 3% such as cassava and arrowroot do not meet the
protein requirements of hurmns even when ingested in large amounts more than the

caloric requirements (Guthrie and Picciano, 1995).

The quality of a protein depends on the kinds and amounts of amino acids it
contains, and represents a measure of the efficiency with which the body can utilize the
protein. A balanced or high quality protein contains essential amino acids in sufficient
ratios for human needs. Proteins of animal origin generally tend to be of higher quality
than those of plant origin. Amino acids present in dietary proteins are not necessarily fully
“available” since digestion of the protein or absorption of the amino acids may be
incomplete. Amino acids fiom animal foods are absorbed to an extent of 90%, while those
from plant foods are digested and absorbed to an extent of only 60-70%. This has been
attributed to the protein conformation; binding to metals, lipids, cellulose and other

polysaccharides; the presence of antinutritional factors such as trypsin inhibitors and

27

lectins; a

different;

that “'01.?
addition.

allow the

   

1931:1311
?TOtein. e

external 1

 

lectins; as well as surface area and size of the protein, processing effects and biological

differences amongst individuals (FAO, 1970, 1973).

Evaluation of proteins is useful to predict the amount or mixture of food proteins
that would be necessary to meet amino acid requirements for grth and maintenance. In
addition, it is also useful to rank the protein in terms of its potential nutritive value, and to
allow the detection of changes in proteins after processing and storage (Pellett and Young,
1981; Bodwell et a1, 1980; and Walker, 1983). Pellett (1978) reported that diet type (total
protein, energy, amino acids etc), consumers (age, sex and their physiological status) and
external factors such as food frequency, social, economic and hygienic conditions can

affect protein utilization and hence protein quality.

The available methods for testing protein nutritional quality are shown in Table 6.
Biological (in-viva) assays measure growth or nitrogen balance as indicators of protein
utilization and metabolism. These tests reflect the essential amino acid content,
bioavailability of the amino acids present in the protein, and protein digestibility of the
food being tested. Due to the time (30 - 50 days) and expense of biological assays,
chemical or biochemical (in-vitro) assays have been developed to predict how a protein
will meet nutritional and growth requirements. In-vitro assays include enzyme assays that
model mammalian digestion for estirmtion of protein digestibility. Amino acid
composition data is compared to reference proteins then corrected for digestibility (in-viva

or in-virro) to obtain a protein quality estimate.

28

I I I 441‘. o .- ‘ t...- . ensfi ‘5‘ tug-uni f1.‘ 11..” :Al-Z A-.~A~\ aiwhnuih N’CFII‘mciHh‘.‘ EWQ‘FiLfih

.4.;.... ...q ... head-u. . .. ‘ a..l~..;x~.....p. hoe.
.733:- 3. 9:3 3.. o:::.7.e..3:.: see 1...: 32:22:30 .....oEaa \ 2.5% S; a. IE1 0925.43 2613.2..2 32% £3.0ng

A57: h._.¢=o aioacba‘ :3 5:05:23» =°-¢~3Q=~6mv KQQerkb 9==~Z
Adi-27 .5.C.:...:3= 2.7.33.1 32.2.4.1: ...: MUUCEVZ. U?~...~..=>< 0 02:55

80.8, E
88:. 2.8208 :8 888:8 :8 .288
88 6288: :oon: :o 88: E a...

82:28 .8823 :o 8888:

8 2.: 86:20.86 :8 8:88 8: 283
a: 86:89:: .836: a: 38:88 .0:
28.80 8 .8: E8: .2 8 .88: E 2<m
808 .8 888:8 .90 088:

8 3:28:22 2:88: Bo. E Ea: 0E8:
3:888 o8 :2: 208 8:3 808 :8
.88: E 853 .86: 0E8: 3888 =: .8
82:8 2: 8288 yin—-09 :8 awn—-0
2:8: E28: .8 22:88 80m 828::

8 58:88 a: 88:8 88::

938: 8:28

Z 88.: 8:888: 8: 3:2 :8 >m
8:82:88 :8 28> mm: 2888 8:2
8:82:88

E 88 828: :8 8:88 3288::
8: 88 8.: Mm: .28 :28: 8:8 :

8 38m 823 8:83: 8 8:888 :28
:88: .8828 E88 :8 2:: .8 :8:
:23: 2 26 0: 22:88:88 :8 8m:
A05 5:80 E28: 8 288.80

_ . 553358583. .

8. a Z .2 \ :2 -..: 1 i. u E 8.3 3:588 8:88 - 2.3-5
:8 x _ _z \ a: 1 :7: n 8 3:38:85 526::

2:: + .2 8:88 : =a
:8 22 x 828: 8:282

w _ E 089: m8 \ E28:
:8: w. E 08?: wEV 7 L n H<<m

8282.: 0:3 \ 9:": A735 x28— 22... oEE< 388$—

8E8:88 << 3823””.— u $5-09 95:69 awn—-0 28888.5

8:888 x em: 1 :28 28-2 em: 8538

D: x E28: 8:888 8:259: 88m Eo<

w _ E << .8 m8 \ E28: oE8< 88880 8:53:85 E28:
:8: w _ 5 << :6 2: ... 88:: 88%: 282 23, 8:5...

D... x >m u 888:8
Z \ 72: E28:-8: :8 8m .8

Z 80: 1 8m :8: .8 Z 88: u DmZ

:2 88:
868888 1 Z SE8:
1 88 88: Z 888:: .1. >m 5:2: 888:8 E28: 82
883:8 E28: A>mv 28> 828.05
\ :2: E28:-8: :8 8m 2% 88: E28: :oZ
.8 m8. :3 + 8m 82 .8 Eam S; 1 8:2 ENB 88 88650 E28:
888:8 E28: \ Eam :3 .1. mm: 888m 8on Z :5: £380
, . 8.88:8.5 . .933 .8 282

...-l: 3:1! .

__ hams: .80E:8.:.E28: 882.3: 80.8.0.8 032824

:a 83.:

29

  
   

 

 
 

II . .u . . . . 1.1:! ... quflw ..I ...... . .. Ca» A ~.»I:~
2:1) .7. ... 13...; S: tow-8. t. n:3.:A—-— _Q .AX» Cr. NN 9.x QMN flu nw\o QOLQU¥~Q QrFrU >~m~mn-~rfidvfl~fi~ «5.08281
ANVA-v hum-:uufiv anuav-Avhan an: trunnion-uuunsz unav-u-nuI-ahEOMU . A‘llfsso nun-howl

NU. -~Aqu o 3‘33. h.

8:5on 8: 82 mo £58 co~§x0

£92063 2 033383

82: mo>cm>ton mmZh .Swsm macaw“:
.32 .23 805 L5 3665 38:8??—
EEES 5a 032:; 32 fl mmzo

mcamoooa ho :32 a mm

2685 355: .8 .38: m_ 3.2 22>-..—
.$>:§§ non: Eatoo

EB 3928 of mo $588378 <<

2.: 3 328: 8a @933 EBwoBEBB—z
258::

owfi 2: E 5:858qu 3595:: £325
52 2a << 335mm» chE: Emtoo
8:83 On— 0885830 :8 8523 <<

>38 noun—+3 .85 28:08 202

.Ia 8828 3 883:8 2 >83 §mia
£3383 529a i:

can Bo. 53> $388 5222. 32.2056
23:50 32838 .0: 800 $33
wctsu 285:8 8: In 29:96»: £205
53> In 5 new so woman a 35in
A9: @530 589... .8 3:25—80

Amv 38% E 38.88
<< - E 8.95 << n 85.3 <<

88a :02
z 3% E. u m 22?
.Ae 3:. + 2 .2 u o .x. 533
ES om 3 mm H X 20:3
dc 8.8- $.va n a .x. 85%
5:53.55

1 . ,1 .111}.

.sz a. E - 5
Oman: can 520 - $2
33 822 5:3 2§§<

mmZH

PEG
0593 oE~=m><

32333832

0.53-:—

Aoocfian <<V 03?:—
§3~=§< 22 SEE

3min
@0543 .. 05?:—
Eo £3505 522..
manna‘ ..o 25.2

€53 an:

30

 

prote'm
19841“
ind lo“
concern:
effects 6
source.
the met

m ex;

dUe t0

 

There have been numerous reviews on the nutritional quality of legume derived
proteins and on the measurement of food protein digestibility (Bressani, 1975; Carpenter,
1984; and Satterlee et a1, 1981). Bressani, 1973, showed that most of the legume proteins
had low biological value, ranging from 32-78%. This has been attributed to the low
concentration of sulfur amino acids in legume protein. Kelly (1973) showed the beneficial
effects of the addition of methionine in the diet when legumes are used as the protein
source. Additionally, Kelly (1973) found that both the protein efficiency ratio (PER), and
the average weight increased. However, this was not evident for all legume species, and
was explained on the basis that methionine is not the most or the only limiting amino acid
in legume species. When both methionine and tryptophan were added, protein quality

increased which indicated that both are equally limiting.

The protein efficiency ratio (PER) is the most common method that has been used
to determine protein nutritive value. It is the weight in grams gained by rats per gram of
protein consumed compared to that fed a control diet that contains casein as the sole
source of protein. The method is based on the principle that the better the nutritional
quality of the test protein, the more rapidly the animals will grow, and is generally

reported relative to the casein control (Pellett, 1978).

This ratio depends on the protein consumed by rats, and protein quality can be
overestimated for animal proteins and underestimated for some vegetable proteins. This is
due to the higher need of rats for certain dietary essential amino acids compared to
humans, which grow at a much slower rate. In particular, the PER method tends to

overestimate the requirement for histidine, isoleucine, threonine, valine and sulfur-

31

€031“if

{ML-“11':

and ‘3
gig-35.16
of diets
. parcel

the per:

arrino 2
quahtv

I994). '

 

 

containing amino acids. PER also does not adequately account for proteins used in

maintenance, and assigns a PER = 0 if the protein does not support growth (Rasco, 1994).

The net protein ratio (NPR) takes protein used for maintenance into consideration,
and can be calculated as NPR = (Wt gain + wt loss of protein free group)/ protein
ingested. Another approach calculates the net protein utilization (NPU) or the percentage
of dietary nitrogen or protein that is retained. It is the product of the biological value (BV)
- percentage of absorbed nitrogen retained in the body, and the coefficient of digestibility -

the percentage of ingested nitrogen absorbed (N AS, 1980).

A number of protein quality tests utilize amino acid content data. Generally, the
amino acid content of the test protein is compared to a reference protein, and protein
quality is based upon the first limiting amino acid or all essential amino acids (Rasco,
1994). This amino acid score may be corrected for protein digestibility determined either
by an in-vitro or in-vivo assay, which results in the protein digestibility-corrected amino
acid score (PDCAAS) method, i.e. PDCAAS = amino acid score x % true digestibility.
The amino acid score for each of the nine essential amino acids is the ratio of essential
amino acid in l g of test protein to essential amino acid in l g of reference protein.
Reference proteins rat growth and human pattern of amino acid requirements are indicated

in Table 7.

When the amino score takes digestibility into consideration, it is believed to be a
better estimate of protein quality since the body's utilization of dietary protein is affected
by factors that are not reflected in an amino acid score. The PDCAAS method has been

recommended by the FAQ/WHO (1990) for measuring protein quality and has been

32

Tdk?

Comm

 

adopted by the Food and Drug Administration (FDA) in their new nutritional labeling
regulations for measuring protein quality of all foods except those intended for infants

(F AO/WHO, 1990).

 

 

 

 

 

Table 7 Comparison of Suggested Patterns of Amino Acid Requirements for
Humans' with that ofthe Rat2
Amino Acid Amino Acids as % of Protein
Rat Human
Pre-school Child (2-5 yr.) Adult
Histidine 1.9 1.9 1.4
Isoleucine 4.1 2.8 4.0
Leucine 7.1 6.6 7.0
Lysine 6.1 5.8 5.5
Methionine + Cystine 6.5 2.5 3.5
Phenylalanine + Tyrosine 6.8 6.3 6.0
Threonine 4.1 3.4 4.0
Tryptophan 1.3 1.1 1.0
Valine 4.9 3.5 5.0
Arginine 2.9 --- «-
‘ Rasco, 1994
2 NRC, 1995

The calculated PER (C-PER) is a PER calculated fiom amino acid composition
data of the test protein and in-vitro protein digestibility. The discriminate calculated PER
(DC-PER) is calculated only fi'om the essential amino acid composition of the food and
compared to the FAO / WHO standard. C-PER and DC-PER methods consider the

content of all essential amino acids unlike the amino acid score method (PDCAAS). This

33

may be
essenti;
referent

It is ex;

 

 

may be useful for foods that contain more than one limiting essential amino acid. The
essential amino acid index (EAAI) is calculated using the ratio of the test protein to the
reference protein for each of the eight essential amino acids plus histidine (Rasco, 1994).
It is expressed as EAAI = [9 ‘1 (mg lysine in lg test protein / mg lysine in l g reference

protein) x (etc for all 8 essential amino acids + His)].

Protein digestibility assays provide important information about protein quality
since all proteins are digested, absorbed and utilized to different extents. Differences in
digestibility is directly related to the structure of the protein, the presence of non-protein
dietary constituents (anti-nutritional components), processing and storage conditions, and
modifications of the protein. Digestibility of the protein is the proportion of protein
nitrogen that is absorbed, and is calculated based on the nitrogen ingested (N i), feed
intake, fecal nitrogen (F..) and fecal metabolic losses of nitrogen (Mn). In-vivo assays
determine the apparent (AD) and true digestibility (TD). AD =[ ( Ni - F“) / Ni x 100 ] and
TD = { [Ni - (Fn - Mn) INi] x 100}. In—vitro assays are calculated by either the pH-drop or
pH-Stat methods using mammalian gastric and / or pancreatic and intestinal enzymes. The
combinations of enzymes include pepsin, pepsin-pancreatin, papain, papain—trypsin,
trypsin, trypsin-chymtrypsin—peptidase, and trypsin-chymotrypsin—peptidase-bacterial

protease.

In-vitro assays measure the extent of hydrolysis or the initial rate of hydrolysis, and
the constituents, pH, and temperature of the incubation medium are often fixed according
to requirements of the enzyme reaction (Rasco, 1994). These assays however, do not take
the amino acid balance of the food being test or fermentation of food proteins in the lower

bowel into consideration.

34

 

llsuet.

OCCUIS .

topsin .
lie pE
home?

10“ an; '

COllS’tazi',

 

Emit;
allow

“litter

 

The pH-Drop method described in AOAC Method 982.30 (1984) from studies by
Hsu et a1 (1977) and Satterlee et al (1981) is based on the principle that a drop in pH
occurs during protein hydrolysis. As the proteolytic enzymes break peptide bonds of the
test protein, the fi'eed carboxyl groups liberate a IF ion, which causes a decrease in pH. It
has been reported that this method is sensitive enough to detect the presence of soybean
trypsin inhibitors and to detect changes in digestibility due to processing (Rasco, 1994).
The pH-Drop in-vitro method generally correlates well with data for in-vivo assays;
however, they do not accurately estimate quantitative differences between samples with
low and high protein digestibility. The major limitation of this method is that pH is not

constant during the reaction and can be affected by other components in the food.

Pedersen and Eggum (1983) developed the pH-Stat method to overcome the
problem in the pH-Drop method that pH is not constant. The two methods use similar
enzymes, however in the pH-Stat method, digestibility is estimated from the volume of
standard alkali (0.1 N NaOH) that is used to maintain a constant pH of 8.0 during
hydrolysis. It has been reported that the pH-Stat is more accurate and gives better
correlation with in-vivo assays. Pedersen and Eggum (1983) and Dimes et al (1994)
reported correlation coefficients of 0.86 - 0.90 for pH-Stat vs. in-vivo digestibilities and

0.71 - 0.78 for pH-Drop vs. in-vivo digestibilities.

An in-vitro digestibility (Immobilized Digestive Enzyme Assay - IDEA) method
has been reported by Chang et al (1990). Digestive tract enzymes are covalently
immobilized on large-pore diameter (2000 A) glass beads via an amide linkage. The beads
allow access of the test protein to the enzymes, and the method has been shown to

COrrelate well with rat bioassays for a wide variety of food proteins.

35

Amino acid analysis is an important chemical technique for predicting protein
quality. Spackman et a1 (1958) and Hill et al (1979) reported on the ion-exchange
chromatography (IEC) and pre-column derivatization of acid hydrolyzates respectively.
Both techniques resulted in complete destruction of tryptophan and partial destruction of
the sulfiJr-amino acids during acid hydrolysis. These amino acids are often the limiting
amino acids in the human diet, and separate hydrolyses of the protein are often
recommended to determine their content. Thus for a complete analysis of all amino acids.
at least three separate hydrolyses would be necessary: acid hydrolysis, performic acid
oxidation followed by acid hydrolysis for sulfiir amino acids, and methane sulfonic acid

(MSA) hydroysis for tryptophan.

In-vitro and in-vivo amino acid availability methods measure the relative
digestibilities of the individual amino acids. The in-vivo assay referred to as the amino acid
balance is calculated as the difference in weight between the amino acid intake and the
amino acid excreted in fecal matter. This method can overestimate protein quality because
it does not account for the loss of some limiting essential amino acids by microbial
fermentation in the large intestine. The in-vitro assay is particularly useful in studies that

evaluate the effect of processing treatments on protein quality.

The available lysine is usually tested in protein quality assays because the fi'ee
amino group on the side chain of lysine can react chemically with other food components
during processing and storage to give complexes that are biologically unavailable.
Reactive lysine (free e-arnino group) can be measured directly using reagents such as 1-

fluoro-2, 4-dinitrobenzene (DNFB), trinitrobenzenesulfonic acid (TNBS), (o-

36

methylisourea, or o-phthalaldehyde. Indirect methods include DNFB difference method,

dye-binding procedure, furosine method, or reduction by NaBI-I4 (Rasco, 1994).

Available sulfur-containing amino acids should also be determined in protein
quality analyses since they are often the limiting amino acids in foods and can be readily
oxidized to forms that are no longer bioavailable. Rasco (1994) reported several methods
to determine available sulfur-containing amino acids. Available cysteine / cystine can be
measured by converting cystine to cysteine with dithiothreitol, reacting cysteine with 5,5’-
dithiobis-Z-nitrobenzoic acid (DTNB) and measuring the quantity of the derivatized form.
Methionine can reduce dirnethyl sulfoxide (Me2S), which can be quantified by headspace
gas chromatography, or it can also react with cyanogen bromide (CNBr) and the reaction
product methylthiocyanate (MeSCN) measured by gas chromatography as an indication of

available methionine.

37

 

’7')
.6

:hm ei‘...
tCher;
heme:

allows

 

llaltsiu

MEMBRANE SEPARATIONS

The primary role of a membrane is to act as a selective barrier, permitting the
passage of certain components and retaining other components of a mixture. This implies
that either the permeate or retentate phases should be enriched in one or more components
(Cheryan, 1986). A membrane has been defined as “a region of discontinuity interposed
between two phases”, or as “a phase that acts as a barrier to prevent mass movement but
allows restricted and/or regulated passage of one or more species through it”
(Lakshminarayanaiah, 1969).

The basic concept of a membrane separation is shown in Figure 1. A feed stream
generally enters the membrane system and a driving force, such as pressure or
concentration, is applied across the membrane effecting the separation. Certain
components such as solutes, gases or solvents, can pass through the membrane, while
others do not pass through, or only very slowly. This selective transport is the basis of
membrane separations. The stream that is able to traverse the membrane is referred to as
the permeate or filtrate, while that which is retained is called the retentate or concentrate
(Mohr et a1, 1989).

The flux is the rate that the permeate passes through the membrane, and is usually
expressed in units of volume of permeate per unit time per unit membrane area (flux units
gal/fizlday, GFD; or L/mz/hr LMH). The rejection is a measure of the membrane’s ability
to separate or retain solution components. It is the fiaction of the components that is

retained by the membrane, and is generally expressed as a percentage.

38

htnhe‘

“hen:

rate a:
unnSc

OfdfiK

 

Rejection = Percent of rejected solute(s)
Z retained

 

Feed ——' Retentate or isolate

 

Flux = Total quantity passed
1 through membrane
Membrane area x Time

 

 

 

Membrane

 

 

 

Figure 1 Membrane Concepts (Mohr et a1, 1989)

Mathematically, it can be expressed as:

Rejection = 100 (Fi - P i)

F1
where i = specific component or group of components
F = concentration of i in the feed stream
P = concentration of i in the permeate stream
Recovery is the fraction of the feed that is recovered as permeate, and permeability is the
rate at which components permeate through the membrane. It is generally expressed in
units of quantity times membrane thickness per unit time per unit membrane area per unit

of driving force (Cheryan, 1986; NFPA, 1993).

39

ultra}
ox‘er
requt
This
separ.

V
memr

by Ch

mazeri

 

 

 

The five major membrane separation processes include reverse osmosis.
ultrafiltration, microfiltration, dialysis and electrodialysis; and are capable of separations
over a wide range of particle sizes. Centrifugal separations are also versatile, however it
requires that there be a density difference between the two phases that are to be separated.
This is not so for membrane separations, and their usefiilness lies in that they permit
separation of dissolved molecules down to the ionic range provided an appropriate
membrane is used (Cheryan, 1986).

Membranes, their components and characteristics have been extensively described
by Cheryan, 1986; Mohr et a1, 1989; and NFPA, 1993. They are made up of various
materials, either inorganic or organic; and may be isotropic or anisotropic. Isotropic
membranes have a similar texture and pore structure fiom one side of the membrane to the
other, while anisotropic membranes change structure from the surface to the interior of the
membrane.

The factors that are important considerations for membrane selection include
separation capabilities, i.e. retention or selectivity; separation rate or flux; chemical and
mechanical stability of the membrane; and the membrane material cost. Desirable
membrane properties are high flux and selectivity, chemical resistance, and durability. Flux
and selectivity depend on the membrane material, while chemical resistance and durability
are affected by the entire membrane system including membrane, housing, adhesive
materials etc.

The organic membranes are the most common commercially, and include those
made from cellulose acetate (CA), polysulfones (PS), and polyamides (PA). CA is utilized

mostly for R0 and some UF membranes, their use in food processing is restricted because

40

they can only tolerate a narrow pH range (3-8) and a maximum temperature range of 35-
40°C. They are also subject to degradation from organic solvents, and microbial activity;
and are sensitive to chlorine. Aromatic PS membranes are used in UF systems, and are
more suited for food processing. They can tolerate temperatures up to 150°C; and are
resistant to low levels of chorine and other oxidizing agents and solvents; and can tolerate
a fairly wide pH range (~ 2-14). The aromatic PA’s are common in R0 systems, and to a
lesser extent UP and MF. They are generally resistant to chemicals, higher temperatures,
and pH; but are sensitive to chlorine and other oxidizing agents

The inorganic membranes are more suited to food applications than the organic
membranes because they can tolerate harsher conditions in terms of temperature (up to
300°C), chemicals, pH (1 - l4), and chlorine tolerance. They are used only in UP and MF
applications. Materials for inorganic membranes include glass, sintered metaL ceramics,

and inorganic polymers.

Membrane Module a_nd System Structures

Membrane separation equipment or modules have been reviewed by (Cheryan,
1986; Mohr et al, 1989; and Ho & Sirkar, 1992). They comprise of four basic types: flat
plate, spiral-wound, tubular and hollow-fiber. The type of module being utilized depends
on the type of separation, throughput/cost, process flexibility required, ease of
maintenance, and ease of operation. The flat plate or plate-and-frame module was one of
the earliest systems, and can be used with any membrane that is available in a flat sheet. It

consists of a series of membranes sandwiched between spacers that act as flow channels,

41

3020.]

that

10 165

andt

high 2

 

 

and which results in a high packing density. The membranes are bonded to an inert
support material, which is porous and offers little resistance to flow. This module can
accommodate low levels of suspended solids and viscous fluids due to the thin feed flow
channels, which provide high fluid turbulence at lower pumping rates. It can also be used
to test, and monitor different membranes. The membrane packing density is somewhat low
and thus requires more space, initial capital cost is high, membrane replacement labor is
high and the modules are very susceptible to fouling due to the thin channels.

Spiral wound modules are an extension of the flat plate systems. They comprise of
two layers of membrane, generally sandwiched around a porous, woven permeate carrier
or spacer. Three sides of this sandwich are sealed, and the fourth is open and attached to a
perforated tube. A feed channel carrier or spacer is layered over the membrane envelope,
and all is wound spirally around the central perforated tube or collection pipe. The feed
flows axially into the channels, and permeate penetrates the membrane and travels up the
permeate carrier spirally to the central tube (Cheryan, 1986; NFPA, 1993).

The spiral wound system has a higher membrane packing density, and this
minimizes the space needed for the system. It generally requires a lower capital because of
the less expensive materials used in manufacturing. Its disadvantages includes its inability
to handle suspended solids, the solids tend to plug the mesh of the feed channel spacer;
this makes cleaning difficult if particles get caught up in the feed channel spacer. In
addition, the use of plastics in the permeate carrier and feed channel spacers limits their
use at higher temperatures due to deformation which can occur at high temperature and

pressure (Cheryan, 1986; NFPA, 1993).

42

Tut
the tiles r
g‘il- “it?
mrodueed
troughth
home:
minut
EWM.

Th
ofwht.
whams

amend

 

 

Tubular modules resemble the structure used in heat exchangers. The diameter of
the tubes range fi'om 1/8” up to about 1”; and the modules are able to handle suspended
solids with little or no pretreatment, and higher viscosity fluids. Feed solution is
introduced at one end of the tube and it flows through the tube, while the permeate passes
through the membrane. Permeate is collected in an outer shell, and the retentate exits at
the other end of the tube. The larger the tube, the higher the pumping rate needed to
minimize buildup of the boundary layer on the interior of the tube (Cheryan, 1986; NFPA,
1993)

These tubular systems are operated under turbulent flow, which minimizes buildup
of solids on the membrane. They are relatively easy to clean by CIP (clean-in-place)
techniques; the diameter allows dirtier water with suspended solids and fluid foods to be
processed; and if one tube fails it can be plugged OE and the rest of the system allowed to
run. The disadvantages of these modules are their lower membrane packing density,
which requires more floor space, the high capital cost, and their higher pumping energy
costs due to large feed channels, large pressure drops and high flow rates. In addition,
there is usually a high hold up volume per unit of membrane area (Cheryan, 1986; NFPA,
1993).

Hollow fiber modules consist of hollow, hair-like fibers with an outside diameter of
50m to about 1mm, which are bundled together into either a U-shaped or a straight-
through configuration. In the U shaped or closed end modules, a loop of fibers is inside a
pressure vessel, feed enters the shell, is pressurized, and permeate passes through to the
center of the hollow fibers. The permeate exits the open fiber ends at the same end of the

pressure vessel that the feed entered. The remaining fluid exits the module at the end

43

oppos
SllSpt.’
open ;
passr.
accor-r

ultrafi.

andtit

Space 7

mustb

a decli
desert
Panicu
1977;

1989

 

pimp;

 

opposite the feed entrance. These fibers tend to be smaller in diameter, and cannot handle
suspended solids (Cheryan, 1986; NFPA, 1993).

In the straight-through modules, fibers are placed inside a pressure vessel and are
’open at both ends. The feed fluid is circulated on the inside of the fibers with permeate
passing through to the outside of the fibers. To minimize pressure drops, these fibers have
somewhat larger diameters than those used in the U-shaped module, which allows them to
accommodate low levels of suspended solids. This type of module is used for
ultrafiltration.

Hollow fiber modules have very high packing density; this requires less floor space
and thus has lower capital cost. They can be back-flushed for cleaning, and low-pressure
drops and flow rates contribute to lower energy costs. However, the modules are delicate,
and can be easily damaged during use. The hollow fine fibers are highly susceptible to
fouling by suspended solids build-up, and cleaning can be diflicult because there is little
space between the fibers. Further to this, if a capillary fiber is damaged, the entire module
must be replaced (Cheryan, 1986; NFPA, 1993).

During membrane processing, fouling of the membrane can occur, which results in
a decline in flux with time of operation. Considerable research has been conducted to
describe the fouling process, and design conditions to reduce fouling of membranes,
particularly whey protein fouling of membranes (Blatt et al, 1970; Parkin, 1975; Lee,
1977; Smith and MacBean, 1978; Merin and Cheryan, 1980; Luss, 1985a, 1985b; Kim,
1989 and Rarnachandra et al 1994).

Fouling occurs primarily because of a deposition of matter (either by adsorption,

precipitation, pore plugging, particulate adhesion, chemical reaction or other interactions)

 

 

C3565

com:

fouls

 

SCp
Weig

and

flou
031T
hex:

Dre:

within the filtration device or within the internal pore structure of the membrane or onto
the membrane surface. Foulants can include dissolved organic matter such as proteins or
carbohydrates; microorganisms; inorganic compounds such as carbonates, or sulfates; and
colloidal or particulate matter such as suspended solids or metal oxides.

Cleaning can generally restore fouled membranes completely or partially. In some
cases however, when the fouling leads to changes in the membrane structure such as with
compaction or creep (change in membrane which decreases its permeability and flux) the

fouling leads to permanent change and is irreversible (NFPA, 1993).

Ultrafiltration Processing

Ultrafiltration (UF) is a relatively new process utilized in the Food Industry to
separate molecules through a porous polymeric membrane on the basis of their molecular
weight. It results in concentration of higher molecular weight solutes in the concentrate,
and the production of a lower molecular weight permeate stream. UF membranes
generally rejects solutes in the size range of 1.0 to 100 nanometers or molecular weight of
300-500 to 300,000-500,000 Dalton depending on the molecular weight cutofl' (MWCO)
of the membrane (NFPA, 1993).

Under normal conditions of transport through a semi-permeable membrane, water
flow occurs from the dilute to the more concentrated solution, a process referred to as
osmosis. The pressure that causes this type of flow is called the osmotic pressure. If
however a pressure is applied to the concentrated solution that exceeds the osmotic

pressure, then the flow direction is reversed, and water migrates from the more

45

concentrated t
concentration 6
flow direction
overcome the

IT is
lice proteins.
pigments. la1
process are
equipment a
1101 require
electricai e,
afnbient 1e
lemperamr

COmPQnem
transfer. L

“thh is d

andthe hig

concentrated to the dilute solution (Lewis, 1982; Cheryan, 1986). This results in further
concentration of the more concentrated solution. Thus the pressure to cause this opposing
flow direction must be high enough to exceed the osmotic pressure of the solution and to
overcome the hydraulic resistance of the membrane.

UP is used for the separation of fairly large molecules, such as natural polymers
like proteins, starch, and gums, and colloidally dispersed compounds such as clays, paints,
pigments, latex particles etc.; hence the osmotic pressures involved in this separation
process are relatively low. This means that UF has the advantage of having lower
equipment and operating (pumping) costs by a considerable margin. In addition, it does
not require any complicated heat transfer or heat-generating equipment since only
electrical energy is needed to drive the pump motor. UF processes can be operated at
ambient temperatures even though at times it may be necessary to operate at low
temperatures to prevent microbial growth problems or denaturation of heat sensitive
components; or high temperatures to lower viscosity of the retentate or to improve mass
transfer. UP is however limited in the sense that it cannot take the solutes to dryness
which is due to the low mass transfer rates obtained with concentrated macromolecules,
and the high viscosity that makes pumping of the retentate diflicult (Cheryan, 1986).

Due to the fact that UF membrane systems retains macromolecules or particles
larger than about 0.001-0.02 um, they have found use as a technique for simultaneously
purifying, concentrating and fractionating macromolecules without the application of heat
or the use of either extreme chemical or physical conditions. This is extremely important in

protein work since it results in very little modification of structure and fimctionality

(Lewis, 1982).

46

LT is n
the fractionati.
and is well rep
N00}. 1980: B
Chiang. 1984; t

\‘egetr
countries for c
their food val:
Etched COHSh
rePorted on th
Obtain a Single
defatted SO}'be..
Skim Milk to p
1b” removal ot‘
filnctional pro
Omega}, & C

More n
[hill desirabflj: -
“he” Protein-
“to Name;

and fOund that

 

which Weasel

]

UF is now a well accepted processing operation in the dairy industry. It is used in
the fractionation of cheese whey and pre-concentration of milk for cheese manufacture.
and is well reported in the literature (Chapman et al, 1974; Abbot et al, 1979; DeBoer and
Nooy, 1980; Bush et al, 1983; Garoutte, 1983; Barbano and Bynum, 1984; Cheryan and
Chiang, 1984; Cheryan and Kuo 1984; Emstrom and Anis, 1985; Emstrom, 1986).

Vegetable protein products have been a part of the diet of many Asian and African
countries for centuries, but it was not until recently that the western world recognized
their food value. Soybeans, now the major source of edible oil in the United States, has
received considerable research as a potential protein source. Lawhon et al (1981, 1982)
reported on the use of a combined aqueous extraction and membrane isolation process to
obtain a single isolation procedure to produce protein and oil food products fiom un-
defatted soybeans. They also used UF membranes to co-process soy protein extracts and
skim milk to produce a soymilk food ingredient. A number of authors have also studied
the removal of antinutritional factors from soybeans using UF membranes, and improving
functional properties of soy protein isolates (Cheryan, 1979, 1988; Hartman, 1979;
Omosaiye & Cheryan, 1979; and Lab & Cheryan 1980).

More recently, research has begun to be expended on other leguminous species, as
their desirability as less expensive protein sources becomes obvious, particularly to nations
where protein-energy malnutrition was prevalent. Ko et al, (1994) reported on the use of
UP to recover protein fi'om simulated waste water during mung bean starch preparation,
and found that protein recovery was about 87.8% using a 30,000 MWCO membrane,
which increased the utilization of mung bean, and reduced pollution problems. Ulloa et a1

(1988) obtained a protein concentrate fiom chickpea (Cicer arietinum) by UF membranes

47

and used it in
extracts and

functional on

 

and used it in infant formulas. Berot et al, (1987) studied ultrafiltration of fababean protein
extracts and found that yields were higher at alkaline pH, and the isolate had better

functional properties such as whipping and thickening than standard isolates.

48

o‘ttair
walli-

Hum;-

Coll

p110:

C011
COD

lat

MATERIALS AND METHODS

Materials

Mature dry seeds of cowpea (Vigna unguiculata cv. California Blackeye No. 5)
and navy beans (Phaseolus vulgaris cv. Huron) grown in the 1995 crop year were
obtained from Bayside Best Beans LLC, Sebewaing, MI. These seeds were stored in a
walk-in cooler in the Food Processing Laboratory in the Department of Food Science and

Human Nutrition at Michigan State University at 4°C (39°F) prior to use.

Proximate Analysis

Cowpea and navy bean seeds were milled in a Udy Cyclone Mill (Udy Co., Fort
Collins, CO), and all other samples (extracts and ultrafiltration fractions) fi'eeze dried
prior to proximate analyses.

The moisture content was determined using AACC Method 44-40 (1984). Protein
content was determined using the micro-kjeldahl Method 46-13 (AACC, 1984), and the
conversion factor of 5.7 used for legume proteins, and 6.25 used for casein and albumin.
Fat was determined using AACC Method 30-25 (1984), and ash by AACC Method 08-01
(1984). Carbohydrate was calculated by difference, that is, subtracting the components

determined above fi'om the total solids (TS) as determined from moisture content.

Milling and Laboratory Scale Protein Extraction
An overview of the protein extraction procedure is shown in Figure 2. The seeds
were milled in a Fitzpatrick mill (Model D Comminuting Machine, Fitzpatrick Co.,

Chicago, IL) to pass through five different screen sizes (0.79-6.35mm). The bean flour (3

49

g) was dispersed in 24 ml of distilled water and the pH adjusted to between 2 and 12 with
1N acetic acid or 1N sodium hydroxide (NaOH). The mixture was shaken in a 25°C or
50°C water bath for 60 minutes. The extracts were separated from the residue by
centrifugation (8800 g, 20 min), then nitrogen content determined on the extract by the
micro-Kjeldahl method (AACC, 1984). The protein content was calculated using the
nitrogen to protein factor of 5.7. Soluble solids (%) of the extracts and protein yield
expressed as g protein / 100 g flour were also determined. The flours were then extracted
sequentially three times using the optimal conditions and protein content and yield

determined.

Pilot Scale Protein Extraction

A schematic process for the pilot plant extraction of cowpea and navy bean flour
is indicated in Figure 3. The seeds were milled in a Fitzpatrick mill (Model D
Comminuting Machine, Fitzpatrick Co., Chicago, IL) to sieve size 1.59 mm. A series of
extractor baskets (200 mesh) was used in the extraction process to separate the residue
fi'om extract. 20 L of distilled water and 2 Kg of bean flour was added to the extractor
basket, and the pH adjusted to 10 with 400 ml 1N sodium hydroxide (NaOH). The
mixture was re-circulated for 60 minutes at 25°C, to ensure complete extraction, then
passed through a series of cheesecloth to separate any fibrous matter that passed through
the basket. The flour was re-extracted twice under the same conditions until there were
three extracts produced. The three extracts were combined, then insoluble components

removed after overnight sedimentation (4°C).

50

 

Whole Seeds

 

 

 

 

 

Milling Conditions
sieve size:
0.79, 1.59, 2.36, 3.17, 6.35mm

 

 

 

 

 

Extraction Conditions:
Flour : water slurry (1:8 ratio)
pH=2,4,6,8,10 or 12
Temp = 25C or 50C for 60 min.

 

 

 

 

 

[Centrifuge at 8800 x g for 20 min...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l
L T
Residue I Liquid Extract l
I Freeze Dry l
Lyophilized Protein Extract

 

 

 

Figure 2 Schematic of Laboratory Scale Protein Extraction from Cowpea and Navy
Bean Flour

51

Cowpea Flour (CPF)
Navy Bean Flour (NBF)

Distilled H20 l NaOH

/

 

 

 

 

 

 

 

 

 

 

 

 

Cowpea Residue
Extractor (CPR)
Basket Navy Bean Residue
(200 mesh) (NBR)
l
l
j 3 sequential
: extractions of flour
1
I
V
Combined Extracts
(l, 2 and 3)
/ \Wernight Sedimentation (4°C)
Insoluble Components Soluble Components
Cowpea Extract (CPPE)
Navy Bean Extract (NBPE)

 

 

 

Figure 3 Schematic of Pilot Plant Scale Extraction of Cowpea and Navy Bean Flour

52

dist
Tit
thrc
purr
HG
off
"51‘

pres

1m;

fract

calcr

[in

were

in,

Ultrafiltration Processing of Aqueous Extrfls

The soluble protein extracts were poured into a 114-L holding vat. A positive
displacement pump was used to transfer the extract to the Ultrafiltration Lab Module,
Type 20 system (Dow Danmark A/S Separation Systems, Denmark) where it passed
through a pre-filter (80 micron) and then into a large feed stock vat. The extract was
pumped from the feed stock vat and passed through Desal (Osmonics Inc., California)
HG19, HNZ8 and HZ20 flat membranes in series (Figure 4). The molecular weight cut
off (MWCO) points for the membranes were 5-10KD, 15-30KD and 50-100KD
respectively, and the total surface area was 0.863 ml. The average transmembrane
pressure was set at 3.4 atm, and pump speed 3, to maximize flow.

The volume of permeates and retentate collected in timed intervals from 0 — 7 hr
using a graduated cylinder and stopwatch was measured, and the total weight of each
fraction determined at the end of the run. The flow rates [volume / time = L/hr] per batch
and permeation rates or flux [volume (L) / membrane area (m2) / time (hr) = LMH] were
calculated.

Afier each run, the membranes were cleaned using a caustic (Ultrasil 10 &
Ultrasil 25) and acid (Ultrasil 75) UF cleaning program for proteins (Figure 5). They

were then soaked in a 0.25% metabisulfite solution until the next use (EcoLab, St. Paul,

MN).

53

Soluble Protein Extracts

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(CPPE and NBPE)
l UF Flat Plate
................ Membranes
Pre-filter -I-I-I-I-I-I-I-I
~I-I-I-I-Z-I-I-I 5_] 0 KB 1212:2212;21:23::221:2: _’ Permeate l
\\\\\\\\\\\\\ N
\\\\\\\\\\\\\:::::V (CPPI ’ NBP 1)
agggggggg (CPP2, NBP2)
50—100 KD *::::::::::::::::E + Permeate 3
{-Z-Z-Z-l-Z-Z-Z-I (CPP3, NBP3)
v
Retentate
Freeze Dry
Lyophilized Isolates
(CPI and NBI)

Figure 4 Flow diagram of Ultrafiltration Processing of Cowpea and Navy Bean
Aqueous Alkali Extracts

54

Figure 5

 

 

Rinse UF System

 

 

 

 

 

Caustic Wash
0.2 % Ultrasil 10
pH 11.0 - 11.5, Temp 120 -125F
Time 20 min

 

 

 

 

 

Rinse UF System

 

 

 

 

 

Acid Wash
0.3% Ultrasil 75
pH 1.8 - 2.2, Temp 120 - 125F
Time 20 min

 

 

 

 

 

Rinse UF System

 

 

 

 

 

Caustic Wash
0.25% Ultrasil 25
pH 10.5 -11.0, Temp 120 - 125F
Time 20 min

 

 

 

 

 

 

Rinse UF System

Add Ultrasi125 or
chlorine to maintain
150 — 180 ppm

 

 

 

 

 

 

Check Warer Flux Readings
(repeat Uhrasil 25 mustic wash iflow)

 

 

Soak in 0.25% Ultrasil MBS
No heat
Circulate 10 min

Shut Down

 

 

 

 

 

 

 

 

Ultrafiltration Processing Cleaning Program recommended for Protein

Products, (EcoLab, 1992)

55

 

SDS Polyacrylamide Gel Electrophoresis

A modification of the SDS PAGE method of Ng (1970) was used. Solutions for
10% running gel and 7% stacking gel were prepared as shown in Table 8. The running
gel solutions were mixed in a beaker and the reagents for polymerization (APS and
Temed) added. The resulting solution was mixed briefly and the gel poured. The gel was
covered with a thin layer of water or alcohol and allowed to stand for at least one hour.
After polymerization, the top of the gel was rinsed with water and the plates turned
upside down to remove excess water. The reagents for the stacking gel were mixed and
stirred similarly. The gel was poured to completely fill the plate and the comb inserted; it
was then allowed to stand for at least one hour. Afier polymerization of the stacking gel,
the combs were removed, and the sample slots rinsed with water then filled with
electrode buffer.

Sigma wide molecular weight range (M-4038) protein marker (15p!) was loaded
at each end of the gel, and sample (10—25 ul) loaded in the slots between. Electrophoresis
was carried out with a Hoeffer Vertical Electrophoresis unit Model SE 600 (Hoeffer
Scientific Instruments, San Francisco, CA) using a constant voltage power supply (Fisher
Biotech Electrophoresis System, Model FB 458, Pittsburgh, PA). A constant current of
25mA was applied for 24 hrs until the dye front reached the bottom of the gel. After the
run, the plates were opened and the gels placed in staining solution for 24 hours. The gels
were destained in several washes of deionized water, then a second staining was done in
Blakesley solution (Table 9) for 24 hours (Ng, 1970). The gel was then rinsed in
deionized water, and photographed. Sub-unit molecular weights of the protein fractions

were estimated using the Sigma protein markers.

56

3i.

‘11

r...“

til

Table 8 Preparation of Running and Stacking Gels for SDS PAGE Analysis

 

 

 

 

 

 

 

Reagents Volume (ml)
10% Runnin' g Gel 7% Stacking Gel

(30 ml) (10 ml)
30% acrylamide 10 2.50
1.5% bis-acrylamide 2.60 1.10
Tris-HCl 12 4
Water 4.35 2.29
10% sodium dodecyl sulfate (SDS) 0.30 0.10
1% ammonium persulfate (APS) 0.75 0.25
N,N,N’,N’tetramethyl- 0.02 0.0067
ethylenediamine(Temed)
Table 9 Preparation of Blakesley Solution for SDS PAGE Analysis
Reagent Quantity
Coomassie Blue G90 2 g in 1 L H20
2 N H2SO4 55 ml in 945 ml H20
10 N KOH 123.4 g up to 220 ml with distilled H20
100% TCA 300 g in 300 ml H2O

 

57

Thermal Stability

The values of onset (Ti) and peak (Tp) temperatures (in °C) and enthalpy change
(AH in W/g) for the protein test samples were studied using differential scanning
calorimetry (DSC). The magnitudes of Ti and Tp were determined as the intersection
between the baseline and the tangent to the descending segment of the endotherm and the
minimum point on the curve, respectively. AH was determined as the area under the
curve. A DSC (Model 2200, TA Instruments, New Castle, DE) was used to examine
thermal transitions. A 10 - 15 mg solid protein sample was placed in a pre-weighed DSC
aluminum pan, hermetically sealed and scanned over 30 - 150°C in the DSC at a heating

rate of 5°C/min. A sealed empty pan was used as a reference.

Amino Acid Analysis
HCl Hydrolysis

A modification of the Waters Pico-Tag® method (Bidlingmeyer et al, 1984) was
used for amino acid analysis. Samples were ground to pass through a 1 mm screen. About
50 - 100 mg (0.050 - 0.100 g) was weighed, depending on the protein content of the
sample. Approximately 25 ml of 6N HCl was added to each sample under the hood.
Nitrogen (N2) gas was added to each tube to displace the air. A slow stream of N2 was
allowed to flow into tube for 10 seconds or more without lid, then for another 10 seconds
or more with lid almost on. The lid was fastened quickly. The tubes were autoclaved for
24hrs to hydrolyze the proteins to amino acids. Afier hydrolysis, the pressure was

allowed to come down slowly, and the samples cooled to room temperature.

58

The bottom of the tubes were rinsed, and dipped into water in case of acid
residue. Each tube was mixed carefully on a Vortex with the lid still on. 50-ml graduated
cylinders were lined up and the contents of each tube carefully transferred to the labeled
graduated cylinders. 2.5 ml of 10-mM norleucine was added to each sample in the
cylinder. The volume was brought up to 40 ml with HPLC water, and then mixed. Using
parafilm fastened securely on top of each cylinder, they were inverted and shaken 6-8

times.

Screw-cap tubes (16mm x 125m) in a plastic rack (1 tube per sample) were
lined up and labeled. A small funnel was set into each tube, and a 7cm #2 Whatman filter
paper folded and placed into each funnel. The solution was filtered carefully (at least 1
funnel firll for each sample) and the unfiltered portion poured into a waste beaker under
the hood so that the cylinder can be washed and used again. The filtrates were saved for

derivatization.

Derivat ization

Mini-tubes (6mmx50mm) were lined up in racks and labeled using a fine tip black
pen only. 25 1.11 each of sample hydrolyzate and norleucine as internal standard was added
to each tube using a new tip for each sample. 25 ul of amino acid standard, the soy
protein isolate and casein controls and norleucine was added to separate tubes. The
samples were dried in the speed vacuum oven set to medium for 1 hour. After drying, 20
ul of Redry solution (Table 10) was added to each dried tube and mixed. The samples

were dried again in the vacuum oven for 1/2 hour.

59

Table 10 Preparation of Reagents for Amino Acid Analysis

 

 

Reagent Quantity
Redry Solution

0.5N NaAc 200 pl
MeOH 200 pl
TEA 100 1.11
Derivatization Solution

H20 100 pl
TEA 100 pl
MeOH 700 pl
PITC 100 ill
Stock EDTA (1000 ppm)

EDTA 10 mg
H20 100 ml
Eluent A l' 2

NaAc 19 g
H20 1 L
TEA 0.5 ml
EDTA 200 pl
Eluent B 2

Acetonitrile 300 ml
H20 200 ml
EDTA 100 pl
Diluent Solution

Eluent A 100 ml
MeOH 100 ml
10 mM Norleucine

Norleucine mg
H20 up to 500 ml
AA Standard H

1:2 (0.5 STD) lmlSTDH:lmlH20
1:4 (0.25 STD) 1 ml STD H : 3 ml H20

_¥

‘ Titrate solution to pH 6.4 with acetic acid, filter, use 940 ml and add 60 ml acetonitrile
2 Degas under vacuum for 60 sec

60

The dried samples were stored in a dessicator in the freezer. Samples were
equilibrated to room temperature prior to opening the dessicator to avoid water
condensation on the inside and outside of the tubes. The derivatizing solution with PITC
(Table 10) was made up just prior to use, and each chemical added in the order listed,

mixing after each addition.

20 pl of the derivatizing solution was added to each tube quickly and without
stopping between samples except to mix. Each tube was mixed carefully but quickly on
vortex ensuring that no spillage occurred. The timer was set for 10 minutes after mixing
was complete. The tubes were placed in a vacuum oven and the samples dried. It is
recommended that only 24 tubes at a time be derivatized to keep reaction times constant.
The samples were then stored desiccated in fieezer until ready to run on the HPLC. When
run is ready, 200 pl of the diluent solution (Table 10) was added to each tube. This was

mixed on a vortex until dissolved (ideally) ensuring that no spillage occurs.

HPLC Analysis

The PITC amino acids were analyzed by direct injection with a Waters HPLC
system using the following conditions: Column: YMC, C18, 5 pm, 120A°, .46x25cm;
Carrier: Helium; peak sensitivity = 95; base sensitivity = 60; minimum area = 30; dilution
= 1. The volume of eluent during the run was % B: 0min=2, 12min=10, l8min=25,
40min=42, 41min=100 @ 1.5ml/min, hold to 46min, 47min=2, 60min=2, 61min=2 @

1ml/min, equilibrate @ 61min. The sample volume was 60 pl, cycle time 59.5min and

61

one injection per sample was used. The column heater was maintained at 46°C
throughout the run. The amino acids were detected using an ultraviolet (UV) detector at a
wavelength (2.) of 254 nm, and a chromatogram showing amino acid and retention time,
was generated using the software package PeakSirnple II, Version 3.85 (SRI Instruments,
Torrance, CA). Amino acids from the experimental samples were determined using a

Sigma Chemicals amino acid standard solution (A9656).

Methanesulfonic Acid (MSA) Hydrolysis - Tryptopm Anlalysis

The analysis of tryptophan (Trp) in proteins and peptides is complicated by the
instability of this amino acid under the normal hydrolysis conditions with 6N HCI. A
modification of the Waters procedure using MSA was used to generate intact Trp. About
1 mg protein for each fraction was weighed and added to an appropriate size tube. 2 ml of
4M MSA containing 0.2% (w/v) tryptamine HCl and 10 ml H2O were added to each tube.
The reaction vial was sealed for hydrolysis using the usual procedure. The mixture was
hydrolyzed at 110°C (15 psi) for 24 hrs. Alter hydrolysis, 2.2 ml of 4M KOH (sufficient
to neutralize) was added to each sample. The contents were filtered carefully to remove
any suspended material. The usual derivatization procedure was then followed. A Trp

solution was made up (2.5 pmol/ml) in H20 and derivatized.

62

Formic Acid Oxi_d_ation - Analysis of Cysteine

Cysteine (Cys) analysis in peptide and protein samples is also complicated by the
instability of the amino acid under acid hydrolysis conditions. Unfortunately, unlike Trp,
acid alternatives to HCI as well as base hydrolysis are unsatisfactory. The procedure used
for Cys analysis involves conversion to more stable derivatives via oxidation to the acid

stable sulfonic acid, cysteic acid (Cya).

The sample (about 1 mg protein) was weighed and added to an appropriate size
tube. 19 volumes of 97% formic acid was mixed with 1 volume H202 and the mixture
allowed to stand covered for 24 hours. 1 ml of this reagent was added to the sample, and
allowed to stand for 30 min at 22°C. The resulting mixture was dried in a Speed-Vac for
30 min. then sealed for hydrolysis using the usual nitrogen procedure. The derivatization

procedure was then followed, and the samples analyzed using a Waters HPLC.

Functional Promies

Oil and Water Absorption Capacity

Flour containing 1g protein and lO-ml distilled water or oil were mixed for 30 sec
in a mixer (V ari-whirl, mixing control - “fast”). The samples were allowed to stand at RT
for 30 min, then centrifuged at 5000 x g for 30 min. The volume of the supernatant was
recorded and the results expressed on a dry wt basis as g of water or oil absorbed / g of

flour or protein (Okezie and Bello, 1988). The density of water = 1 g/ml, and oil = 0.88

g/ml.

63

Protein Solubility

A modification of the AACC Method 46-23 (1984) was used to determine protein
solubility. A 500-mg protein sample was weighed into a ISO-ml beaker. Deionized
distilled water was added with stirring to form a smooth paste, then more added until the
total volume of the dispersion was 40 ml. The beaker was placed on a magnetic stirrer
insulated with plastic sink matting to prevent heating during the subsequent stirring
period. The mixture was stirred using a 2.5 cm smooth plastic coated stir bar at a rate that
just failed to form a vortex. The pH of the dispersion was measured, then adjusted to 7.0
with 0.1N HCl or NaOH. It was stirred for 1 hr under these conditions, and checked and
maintained intermittently. The mixture was transferred to a 50-ml volumetric flask, and
diluted to the mark with water, then mixed by inverting and swirling.

The dispersion was centrifirged at 20,000 x g for 30 min., and the supernatant
filtered through a Whatman No. 1 filter paper. The protein content of the filtrate was
determined by the micro-kjeldahl method, and protein solubility in distilled water
calculated as - Protein Solubility (%) = {[Supernatant protein conc. (mg/ml) x 50] /

[sample wt (mg) x sample protein content (%)/100]} x 100

Foaming Capacity and Stability
A modification of the method reported by Sathe and Salunkhe (1981a) was used.

A 1-g protein sample and 100 ml distilled water was whipped for 5 min at high speed in a
Waring Blender at RT. The mixture was poured into a 250 ml measuring cylinder, and

the total volume measured at time intervals of 0.0, 0.5, 1.0, 2.0, 3.0, 8.0, and 36 hr. The

effect of concentration using 1, 2, 3, 5, 7 and 10 % (w/v) aqueous protein solutions was
evaluated. Foam capacity was calculated using the following equations:

Volume increase (%) = vol. after whip - vol. before whip / vol. before whip

Emulsion Capacity and Stability

A modification of the method reported by Sathe and Salunkhe (1981a) was used.
A 2-g protein sample was blended in a Waring Blender with 100-ml distilled water for 30
sec at “HI” speed. Oil in was added in 5-ml portions (fi'om burette) and blending
continued at RT until a drop in consistency was observed. The drop in consistency,
judged by a decrease in resistance to blending, was taken as the point of discontinuation
of oil addition. The oil added to this pt = emulsifying capacity of the sample. The
emulsion was allowed to stand in a graduated cylinder, and the volume of water separated

at time intervals of 10, 20, 35, 60, 100, 120, 140, 160, 180, 200, and 780 hr recorded.

QgLation Characteristic;

A modification of the method reported by Sathe and Salunkhe (1981a) was used.
A 2, 4, 6, 8, 10, 12, 14, l6, l8 and 20% (w/v) protein sample was prepared in 5-ml of
distilled water in testubes. The mixture was heated for 1 hr in a boiling water bath
followed by rapid cooling under running cold tap water. It was cooled further for 2 hr at
4°C. The least gelation concentration (Gel P) = concentration when sample fi'om inverted

tube did not fall down or slip. A texture profile analysis (TPA) of an 18% gel for each

65

fraction was generated from the Kramer Shear Press, and the hardness (HARD), work
done on the gel (WORK), cohesiveness (COHES), adhesiveness (ADI-LES), gumminess

(GUM), and chewiness (CHEW) determined.

Colorimetric Analysis

The color of 15-ml samples of each of the liquid fiactions was measured using a
Hunter Color Meter (Model D25D2) with a yellow reference standard. Color was
expressed in terms of L, a and b, where L = lightness, + a = red, - a = green, + b = yellow

and - b = blue.

Nutritionyal Characterization
Screeningfor Phytohemagglutinin

Phytohemagglutinin (PHA) was determined using the method reported by Occena
(1994) with anti-Phaseolus vulgaris lectin (Sigma P0526) as irnmunogen and Phaseolus
vulgaris lectin (Sigma P8629) as reference. 0.01 g/ml slurry of the protein test sample in
deionized distilled water was prepared, and 40-pl aliquots pipetted into round bottom
microtiter wells. Each test sample and a purified Phaseolus vulgaris lectin used as a
reference were reacted with 10 pl of antibody. The same quantities of protein slurry were
pipetted into adjacent rows of wells, but no antibody added to distinguish sample settling

fi'om agglutination. The formation of the precipitate was determined subjectively after an

66

hour. The results were recorded using the following notation: + indicates the presence of
a precipitate; number of + indicates the degree of precipitation (++—++ strong positive, +
slight positive); ND indicates that precipitation was not detected; --- indicates that the

sample was not analyzed.

In-vitro Protein Digestibility

The in-vilro digestibility of experimental samples and reference casein were
measured using the four enzyme pH-Drop method outlined by AOAC (1990) and the

three enzyme pH-Stat method described by McDonough et a1 (1990).
pH Drop Method

Cowpea and navy bean fractions, soy protein isolate and a sodium caseinate
standard (Sigma C-8654) were weighed to contain approximately lO-mg nitrogen (Table
11). Ten ml of deionized distilled water was added to each test tube containing a stirring
bar. Protein suspensions were held in a water bath maintained at 37°C for at least an
hour.

One ml of "Enzyme A" solution (Table 12) which had previously been
equilibrated to pH 8.0+0.03, was added to the protein suspension after it had also been
equilibrated to pH 8.0+0.03. After incubation with Enzyme A for 10 minutes at 37°C, 1.0
ml of "Enzyme B" solution (Table 12) was added to the protein suspension. This was
further incubated for 9 minutes in a water bath maintained at 55°C. At precisely 20

minutes from the addition of "Enzyme A", the pH was read at 37°C. The extent to which

67

pH of the protein suspension dropped was used as a measure of protein digestibility.
Percent digestibility was calculated as follows:

% Protein Digestibility = 234.84 - 22.56 (X), where X = pH at 20 minute.

Table 11 Nitrogen Content and Sample Weight of a 10 mg Nitrogen sample of
Cowpea and Navy Bean Fractions afier Ultrafiltration Processing

 

 

Fraction Nitrogen Content Sample Wt (g)
(% N) 10 mg Nitrogen

CPF 4.68 0.214

CPPE 6.16 0.162

CPI 9.08 0.110

CPR 3.42 0.292

NBF 4.27 0.234

NBPE 6.54 0.153

NBI 8.41 0.119

NBR 1.60 0.625

SF 6.16 0.162

SP1 14.01 0.071

 

68

Table 12 Preparation of Enzymes A and B used for pH-Drop in- Vitro
Digestibility of Cowpea and Navy Bean Protein Fractions

 

 

 

 

 

Enzyme Quantity of Enzyme
(mg / 50 ml)
Enzyme A Enzyme B

Porcine Pancreatic Trypsin 68 n/a
(Type IX, Sigma T-0134

Bovine Pancreatic Chymotrypsin 166 n/a
(Type H, Sigma C-4129)

Porcine Intestinal Peptidase 26 n/a
(Grade K, Sigma P-7500)

Bacterial Protease n/a 397.5
(Pronase E)

pH-Stat Method

A solution containing all 3 enzymes (Table 13) was adjusted to pH 8.0 with 0.1N
HC1 and / or NaOH, maintained for exactly 2 min. then transferred to an ice bath and
kept at 0°C. The multienzyme solution was prepared daily and its activity checked using
an aqueous suspension of non-alkali-treated sodium caseinate (1 mg N / ml H20). Ten ml
of the sodium caseinate suspension was placed in a reaction vessel at 37°C, the pH was
adjusted to 8.0 and maintained for 5 —— 10 min. before adding 1 ml enzyme solution.
Enzyme activity was determined from the amount of 0. IN NaOH required to maintain the
pH at 7.98 for exactly 10 min. Percent true digestibility (TD) was calculated using the
following equation: TD = 76.14 + 47.77B, where B = ml 0.1N NaOH added

(McDonough et al, 1990).

69

 

 

 

la

Table 13 Preparation of Enzyme used for pH-Stat in- Vitro Digestibility
of Cowpea and Navy Bean Protein Fractions

 

 

Enzyme Quantity of Enzyme
(mg / 50 ml)

Porcine Pancreatic Trypsin 68

(Type IX, Sigma T-0134

Bovine Pancreatic Chymotrypsin 166

(Type II, Sigma C-4129)

Porcine Intestinal Peptidase 26

(Grade K, Sigma P-7500)

 

An amount of sample containing 10-mg nitrogen (Table 11) was dissolved in 2.5

ml water, then, 2.5 ml 0.2N NaOH was added and the solutions incubated at 37°C. After

30 min., 5.0 ml 0.075N HCl was added to the solutions and the pH adjusted to 8.0.

Digestibility of triplicate samples of the fractions was determined by adding 1 ml of the

multienzyme solution. The amount of 0.1N NaOH required to maintain the pH at 7.98 for

exactly 5 min. was measured.

An uncorrected protein digestibility value (UT D) was calculated as follows: UTD

= 79.28 + 40.74B, where B = ml 0.1N NaOH used during 5 min. True digestibility of the

fractions was determined by multiplying the UTD value by a correction factor (1) derived

fi'om sodium caseinate, f = 100 / UTD of sodium caseinate.

70

ingredie
remainet

ant inutri

are indie.
3.5% mi.
hl'droxyrr
Source. A
SiOWegt s;
for an ad:
piaSIlC CO.
diet AIN

metabolic

 

In-vivo True Protein Digestibility
Preparation of C ampea and Navy Bean Residue
The processing scheme for production of coma and navy bean residue
ingredients is illustrated in Figure 6. The raw cowpea and navy bean residue, which
remained after alkali extraction, was cooked in kettles for 15 min at 100°C to inactivate

antinutritional factors. The cooked residues were then air-dried for 48 hours at 45°C.

Preparation of diets for evaluation

The quantities used in preparation of the 10% protein (wt/wt) experimental diets
are indicated in Table 14. Each diet contained 10% protein, 13.43% fiber, 10.75% fat,
3.5% mineral mix, 1% vitamin mix, 0.25% of choline bitartrate, and 0.0014% butylated
hydroxytoluene. The fat source was corn oil, and arrowroot flour, the major starch
source. All of the ingredients except oil were mixed for 10-min in a mixing bowl at the
slowest speed to reduce dust. The oil component was added slowly and the diets mixed
for an additional 25-min at slow speed. The diets were stored at 4°C in tightly sealed
plastic containers prior to use. A modified American Institute of Nutrition (AIN) rodent
diet AIN-93G was used as the control, and a 2% albumin diet provided an estimate of

metabolic fecal nitrogen.

71

Legume Flour

 

 

 

 

 

 

 

 

 

 

 

(CPF and NBF)
Distilled Water
NaOH
Extractor basket 9 Residue
(200 mesh)
A Y
Cook at 100°C
for 10 min
i Air-dry for
48 hrs at 45°C
CPR and NBR
For diet

 

 

 

Figure 6 Processing scheme for production of cowpea and navy bean residue
ingredients for in-vivo digestibility study

72

ranked 1:
randoml)
received
protocol
care and
T
experime
addition (
cage to at
7. a Intro;

air dried.

10 a mwd

 

 

 

 

Feeding Study

Twenty male Sprague Dawley (Harlan Sprague Dawley Inc.) rats were initially
ranked by weight, 1 being the heaviest and 20 the lightest (Table 15). Each rat was
randomly assigned a diet over the four-week period (Table 16), and none of them
received the same diet twice. This was repeated until each diet (1 — 8) had ten rats. The
protocol of the All-University Committee on Animal Use and Care (AUCAUC) on the
care and use of laboratory animals was followed.

The rats were housed in individual cages and given free access to the
experimental diets and water for 7 days. The jars containing diet were weighed upon the
addition of flesh food. A cardboard collection box was constructed and placed under each
cage to account for all eaten and spilled food, as well as fecal matter. Between days 3 and
7, a nitrogen balance study was conducted. Spilled food and fecal matter were separated,
air dried, and weighed. Fecal matter collected during the nitrogen balance was pulverized

to a powder with a mortar and pestle, and the protein content determined in triplicate.

Food Consumption and Rat weight
The food consumed was calculated using the initial (Wt), spilled (W5) and final
weights (Wf), i.e. food consumed = W, - (Wf + W,). The rats were weighed on a daily

basis, and the increase or decrease in weight determined over the feeding period.

73

Table 14

Diets are
100 g 0ft
be = 10 g
The expe
following
WWI
NBR
CPR
Consider
30 co
X (1 l
X = 3
-0 973
X (l 1
X = 7
Consider

 

Table 14

PROTEIN CONTENT
Diets are based on 10% (w/w) protein. In
100 g of diet, the total protein will then
be = 10 g.
The experimental samples have the
following protein content:
WWF = 11.03 %
NBR = 15.25 %
CPR = 14.27 %
Consider protein in WWF (X):
30 % of 10 g = 3 g
X(11.03 %)= 3 g
X=3/11.03%=27.19g
70 % of 10 g = 7 g
X(ll.03 %)=7g
X=7/11.03%=63.46g
Consider protein in NBR:
30 % = 19.67 g
70 % = 45.90 g
100 % = 65.57 g
Consider protein in CPR:
30 % = 21.02 g
70 % = 49.05 g
100 % = 70.07 g

FIBER CONTENT

The AIN-93G diet normally has 5 g of
fiber, however the experimental diets
will have significantly more because of
the high fiber content of the samples.
The fiber content of all diets will have to
be based on the diet with the highest
quantity of fiber. The experimental
samples have the following fiber
content: WWF = 15.2%

NBR = 18%

CPR = 18%

Consider the 70:30 NBR‘WWF diet

15.2 %x 27.19 gWWF = 4.13 g fiber

18 % x 45.90 g NBR = 8.26 g fiber
Total = 12.39 g fiber

Consider 70:30 CPR:WWF diet

15.2 % x 27.19 g WWF = 4.13 g fiber

74

Calculation of 10% Legume-Wheat Flour Diet Blends

18 % x 49.05 g CPR = 12.96 g fiber
Total = 12.96 g fiber
Consider the 30:70 NBR:WWF diet
15.2 % x 63.46 g W = 9.64 fiber

18 % x 19.67 g NB = 3.54 g fiber
Total = 13.18 g fiber

Consider 30:70 CPR:WWF diet

15.2 % x 63.46 g WWF = 9.64 g fiber

18 % x 21.02 g CPR = 3.78 g fiber
Total = 13.43 g fiber

Consider 100 % NBR diet

18 % x 65.57g NBR = 11.80 g fiber
Total = 11.80 g fiber

Consider 100% CP diet

18 % x 65.57 g NB = 12.61 g fiber
Total = 12.61 g fiber

Thus all experimental diets must be
based on 13.43 g fiber, which represents
8.43 g extra fiber than the 5 g AIN-93G
diet. This displaces 8.43 g of digestible
carbohydrate, and hence 8.43 g x 4 ‘
Kcal/g = 33.72 Kcal of energy.

This is equivalent to 33.72 Kcal / 9 2
1(0an = 3.74 g of fat. The fat content of
the AIN-93G diet is 7%, thus in order to
maintain the nutrient : energy ratio, it is
necessary to increase the total fat. Thus
the total fat in the diets = 3.74 g + 7 g =
10.74 g.

' 1g digestible carbohydrate = 4 Kcal
2 l g fat = 9 Kcal

CODE:

W = whole wheat flour
NBR = navy bean residue
CPR = cowpea residue

Table 15 Ranking of rats according to weightl

 

 

Weight of rat (g) Assigned #
1 12 1
l 10 2
1 10 3
107 4
104 5
100 6
100 7
99 8
98 9
98 10
97 1 1
97 12
97 13
97 14
97 15
96 16
96 17
95 18
95 19
94 20

 

lheaviest = l and lightest = 20

75

Table 16

 

.AJ

 

 

 

 

 

 

 

 

Table 16 Rat and Diet Assignment for Feeding Study
Rat # Diet Assignment

Group 1 Group 2 Group 3 Group 4
1 2%ALB 100%CP 70%NB 100%NB
2 100%NB 30%NB 70%CP 100%CP
3 100%CP 30%CP AIN 30%NB
4 30%NB 70%NB 2%ALB 30%CP
5 30%CP 70%CP 100%NB 70%NB
6 70%NB AIN 100%CP 70%CP
7 70%CP 2%ALB 30%NB AIN
8 AIN 100%NB 30%CP 2%ALB
9 2%ALB 100%CP 70%NB 100%NB
10 100%NB 30%NB 70%CP 100%CP
11 100%CP 30%CP AIN 30%NB
12 2%ALB 70%NB 2%ALB 30%CP
13 30%CP 70%CP 100%NB 70%NB
14 70%NB AIN 100%CP 30%CP
15 70%CP 30%CP 30%NB 70%NB
16 AIN 70%NB 2%ALB 70%CP
17 2%ALB 70%CP 100%NB AIN
18 100%NB AIN 100%CP 70%CP
19 100%CP 2%ALB 30%NB AIN
20 30%NB 1 00%NB 30%CP 2%ALB

 

76

Ar

ingested a
Tr

[Ni - (Fn
R‘

values w
RPER = (
diet l NF
consurm.

rat i [7101:

 

 

 

Calculation of Protein Quality

Apparent digestibility (AD) was calculated based on the amount of nitrogen (N)
ingested and the feed intake, i.e. AD = (N t — Fn)/ N x 100.

True digestibility (TD) was corrected for metabolic losses in the feces, i.e. TD =
[N — (Fn —Mn)] / Ni x 100. N; = N intake, Fn = fecal N and Mn = fecal metabolic N loss.

Relative protein efficiency ratio (RPER) and relative net protein ratio (RNPR)
values were calculated, using the following equations from Sarwar and Peace, (1994).
RPER = (PER of test diet / PER of modified AIN-93G) x 100. The RNPR = (NPR of test
diet / NPR of modified AIN-93G) x 100. PER is the weight gain of test rat / protein
consumed by test rat, and NPR is the weight gain of the test rat + weight loss of 2%ALB

rat / protein consumed by test rat.

PDCAAS Calculation and interpretation

Protein digestibility corrected amino acid score (PDCAAS) is calculated as the
TD x amino acid score, where amino acid score is the mg of essential amino acid in 1 g
of test protein / mg of the same amino acid in l g of reference protein. Scores for the nine
essential amino acids were calculated using a human growth pattern of amino acids
requirements for pre-school children (2 — 5 yr.) (FAO/WHO/UNU, 1985) as the reference
proteins. Any PDCAAS above 100% is rounded down to 100% for further calculations.
There is no nutritional advantage to consuming proteins with PDCAAS greater than
100% since excess amino acids are not utilized by the body, instead the nitrogen is

excreted as urea, and the carbon skeleton is used for energy or stored in the body as fat.

77

All pr

compl

Starisi

des'iati
Insiiiur
produc
models
means '

extracti.

 

All proteins with a PDCAAS of 100% are high quality proteins, and are considered

complete in being able to meet the essential amino acid requirements of humans.

Statistical Analysis

 

A Microsofi Excel 97 spreadsheet was used to compute the averages and standard
deviations of all the data The statistical analysis programs Stat View (Version 5, SAS
Institute Inc., 1998) and IMP IN (SAS Institute Inc., 1996) were used to analyze the data,
producing analysis of variance, factor analysis, correlations and multiple regression linear
models. The post-hoe test Fisher's LSD was used for multiple comparisons of treatment
means when a significant F value was obtained to determine the specific differences in

extraction conditions which existed between the various treatments.

78

STI’I

 

sizes ((
Values 1
used to
h'uphili
for extn
enmlic
e.‘ttraeii
uhrafiltr
conditit

 

STUDY 1 OPTIMIZATION OF THE AQUEOUS EXTRACTION 0F
PROTEINS FROM COWPEAS (Vigna unguiculata) AND NAVY
BEAN S (Phaseolus vulgaris)

Aha—mg

This study was conducted to evaluate the effect of pH, particle size and
temperature on the efficiency of protein extraction from cowpeas (Vigna unguiculata)
and navy bean (Phaseolus vulgaris) flours, and to assess optimal extraction conditions for
filrther ultrafiltration processing. The seeds were milled to produce five different particle
sizes (0.79mm - 6.35mm screens), and extracted in distilled water at six different pH
values (2-12) for 60 minutes each at 25 or 50°C. Sodium hydroxide and acetic acid were
used to adjust to alkaline and acidic conditions, respectively. The aqueous extracts were
lyophilized and protein yield determined. Alkaline pH was more effective than acidic pH
for extracting protein. Particle size had a highly significant inverse relationship to protein
extraction efficiency. Temperature did not have any significant effect on protein
extraction. Results indicate that the optimum protein extraction conditions for further
ultrafiltration processing were particle size of 1.59 mm, at pH 10 and 25°C. Under these
conditions, 87% and 91% of the total protein was recovered fiom cowpea and navy bean

flour respectively, after three sequential extractions.

79

literatu
1980:!
Stanle)
various
methoc

reporte

depend
applica
Sosulsl

Physic:

4L199t

 

 

Introduction

Protein extraction of various legume seeds has been extensively reported in the
literature (Evans and Kerr, 1963; Hang et al, 1970; Fan and Sosulski, 1974; Khan et a1,
1980; Sumner et a1, 1981; Sauvaire et al, 1984; and Tzeng et al, 1990). Sefa-Dedeh &
Stanley (1979) reported optimum conditions for protein extraction fi'om cowpeas using
various extraction solvents. Kohnhorst et al, (1990) reported on an air classification
method for the production of high protein flours. However, very little work has been
reported on extraction conditions for further ultrafiltration processing for these legumes.

According to Sathe et a1 (1984), the procedures for extracting proteins differ
depending on the purpose and end use. They reported that a convenient method for food
applications is the wet extraction-isoelectric precipitation method described by Fan and
Sosulski (1974). Other authors have reported on dry methods such as air classification to
physically separate the protein rich fiaction (Sosulski and Young, 1979; and Kohnhorst et
al, 1990).

Fractionation according to molecular weight for'desirable functional properties
was reported by Sathe et a1 (1984). They recommended extraction in aqueous alkaline
solution followed by ultrafiltration separation of the protein. Ultrafiltration has
traditionally been used in the dairy industry to recover and purify proteins fiom skim
milk (Peri et al, 1973). It is being used increasingly with legume seeds, and is generally
recommended as a low energy processing method that produces high quality protein
isolates with desirable functional properties (Lawhon et a1, 1979, 1981). There has been
very little work reported on optimized legume protein extraction for ultrafiltration

processing.

80

ertractic
W"
1981; ar

1
variable:

on the y

 

 

 

Studies focused on optimization of protein extraction indicate that type of legume,
extraction conditions (pH, time, solvent, temperature), and particle size are important
parameters that affect the protein recovered (Fan & Sosulki, 1974; Sathe and Salunkhe,
1981; and Farias et a1, 1995).

The null hypothesis (Ho) being tested in this study is the selected process
variables (pH, particle size, bean type, and extraction temperature) do not have any effect
on the yield of protein extracted from cowpea and navy bean flour.

Thus, the objective of this research was to study the effect of particle size,
extraction solvent pH and extraction temperature on the efficiency of protein extraction

from cowpeas and navy beans using a completely randomized experimental design.

Materials and Methods

Mature dry seeds of cowpea (Vigna unguiculata cv. California Blackeye No. 5)
and navy beans (Phaseolus vulgaris cv. Huron) were purchased fi'om Bayside Best Beans
LLC, Sebewaing, MI. The seeds were milled in a standard fitzmill (0.79 - 6.35 mm
screens) and the flour stored at 4°C until use.

Proteins were extracted fi'om the flours (Figure 2), the extracts freeze dried and
nitrogen content determined on a dry basis by the micro-Kjeldahl method (AACC, 1984).
Protein content was calculated using the nitrogen to protein factor of 5.7, soluble solids
(%) of the extracts and protein yield expressed as g protein / 100 g flour, were calculated.
The optimal conditions of protein extraction were determined based on protein yield,

milling time, and requirements for further processing steps.

81

C011

COIT

Micrr
devia
extrat
factor
lnstitu
fitted.

anal ys

intErac-
Variabl
treat

differe

 

The experiment was repeated using three sequential extractions of the optimal
conditions, and the three supernatants produced were combined. A mass balance of the

combined extracts was calculated.

Statistical Analysis

In this experiment, a 6x5x2x2 factorial arrangement of treatments was used. A
Microsoft Excel 97 spreadsheet was used to compute the averages and standard
deviations of the data. The independent variables in the experiments included pH,
extraction temperature, legume type, and particle size, while the dependent (response)
factor was protein extracted. Stat View (SAS Institute Inc., 1998) and IMP IN (SAS
Institute Inc., 1996) were used to analyze the data, and a multiple regression linear model
fitted. If any of the independent variables were insignificant, they were removed from the
analysis, and the model re-examined.

Predicted protein values were computed and leverage plots for main and
interaction treatment effects generated to show the relationships between the independent
variables and protein extracted. Fisher’s LSD was used for multiple comparisons of
treatment means when a significant F value was obtained to determine the specific

differences in extraction conditions which existed between the treatments.

82

COWI

indicate
The pit

in may

Tablel7

Compot

7%
lipid
Ash

MOlStun

Carbon}

D
H
w

 

Results and Discussion

The particle sizes of cowpea and navy bean flours ranged from a fine texture
(0.79mm screen) to very coarse particles (6.35mm screen). Visual examination of the
finest fiaction showed that it had the texture of wheat flour, while the coarsest fraction
had much larger particles, and tended to look and feel like split beans. Regrinding the
coarser fiactions through all screen sizes produced the smaller particle sizes.

Table 17 shows the proximate composition of cowpea and navy bean flours. It
indicates that carbohydrate was the most significant component in both legume flours.
The protein and moisture contents were higher in cowpea, while lipid and ash was higher

innavy bean.

Tab1e17 Proximate Composition of Cowpea and Navy Bean Flour (%db)

 

 

Component Cowpea Navy Bean
Protein 26.7 i 0.84 24.4 :1: 0.59
Lipid 2.1 i 0.23 3.3 :t: 0.35
Ash 3.5 i 0.56 3.8 i: 0.15
Moisture 11.9: 1.1 8.1i0.79
Carbohydrate 55.8 i 0.68 60.4 :1: 0.47

 

n=3

regress
conten
solids

respect
beans.

proport
100 g 1
highest

particle

extract‘u
IESldUes

”14.17

 

& Sailin

 

The data on protein content of the extracts, the predicted protein fiom the
regression model, soluble solids, and their protein recovery yield indicated that protein
content ranged fi'om 11 - 38% for cowpeas, and 11 - 41% for navy beans. The soluble
solids ranged from 0.75 - 5.07% and 0.16 - 4.86% for cowpea and navy bean
respectively. The protein yield ranged from 0.7 — 41% for cowpeas and 4 — 55% for navy
beans. The yields of protein from cowpea and navy bean flour were generally
proportional to the protein contents of the extracts. The lowest yield of 0.68 g protein /
100 g flour was reported for navy bean at pH 10, particle size 6.35 mm and 50°C. The
highest yield of 55.06 g protein / 100 g flour was reported for navy bean at pH 12,
particle size 0.79 mm and 25°C.

The results indicated that alkaline pH was more effective than acidic pH in
extracting protein, which was due to the greater concentration of acidic amino acid
residues found in legumes. Protein content increased with increasing pH except at about
pH 4. The cowpea extraction curves (Figures 7 and 8) showed a well defined lowering in
protein content at pH 4, however for navy bean (Figures 9 and 10), this ranged between
pH 4 - 6. Although the isoelectric points were not determined, it was assumed that the pH
or pH range of minimum protein content in the extract was the isoelectric point for the
legume protein. The steep portion of the nitrogen extraction curves for cowpea (Figures 7
and 8) were again more pronounced than those for navy bean (Figures 9 and 10), and
occurred between pH 4 - 8. These results appear to agree with the general protein
extraction pattern for other legume seeds as reported by Fan & Sosulski (1974) and Sathe

& Salunkhe (1981).

84

as; .53.... :5... agentnw E39..—

\ -

 

3. Elli] l

4 3
..\.L Lao: 39: target-.3 532$

 

43.0)

 

38(1)
33(1) 4

28(1) .
23.00 . //
18(1) . v1]

\ /

/,

Protein Extracted from Flour (%

 

 

 

prr

 

10

12

14

_+O.79
+1.59
+2.36
+3.17

eat—6.35

 

 

 

Figure 7 Cowpea Protein Extraction Curve, pH and Particle Size Effects at 25°C

 

 

4&00
43.00 t
3800 .
33.00 ~
2800 «
23.00 ~

18.“) .

Protein Extracted from Flour (%

13.1!) ‘

 

aoo

 

 

 

 

14

 

 

Figure 8 Cowpea Protein Extraction Curve, pH and Particle Size Effects at 50°C

85

Figur

9A.» 5:0:— EPC to.

2]

in

98.3 n H 5.3.9.:—

Film!

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.,\° 44(1)
'5 39.(l)f
O
E . _.
a 34.00 ~ r—o—0.79:
E 29.“) / + 1.59i
3 ' ' ’+2.36‘
:5 24m 1 v/ ‘ ‘
,3 +3.17
M 19-00 ‘ +6.35
§ 14.00 t
2
“e 9.00 .
4.11)
O 2 4 6 8 10 12 14
on
Figure 9 Navy Bean Protein Extraction Curve, pH and Particle Size Effects at 25°C

2: 44.00 .
g 39.00 «
o , __.
g 34.00 « ;-c—0.79r
é 29.m T ’ 1+ 1.59

[+2.36t

\ I /
g 24.00 i V i—x—3J7l
iii 19-00 ‘ i—r&—6351
g 14.1!) ~
3" 9.“) r
4.1!) .
0 2 4 6 8 10 12 14
pH

 

 

 

Figure 10 Navy Bean Protein Extraction Curve, pH and Particle Size Effects at 50°C

86

 

decrease.

solvent

extractio

tempera:

 

 

squirt

Additionally, as particle size increased, the protein content of the extract
decreased. The greater surface area available for smaller particle sizes with the extraction
solvent can explain this trend. There was no definitive relationship observed for
extraction temperature and protein content of the bean extracts. Overall, the higher
temperature used in this experiment, did not appear to be an advantage in increasing the
protein recovered.

A visual examination of the protein extracts at alkali conditions (pH 10 and 12)
indicated that those for cowpea were darker in appearance than those for navy bean. This
could be explained by the higher tannins content of cowpea compared to navy bean
(Salunkhe, 1985). .

Multiple regression was used to investigate pH, particle size, and extraction
temperature, and bean type as predictors of protein content in cowpea and navy bean
extract. Table 18 shows the summary of fit, and indicates that the multiple correlation
coefficient (R) is 0.89, the multiple coefficient of determination (R2) is 0.80, and the
adjusted R2 is 0.79. These are statistically significant at well under the alplm level of p <
0.05. R = 0.89 indicates that there is a good fit between the linear combination of pH,
particle size, temperature, bean type and protein content. Furthermore, R2 indicated that
80% of the variance in protein content was predictable from this linear combination. The
remaining residual erfor is estimated to have a standard deviation of 3.14 (root mean

square error).

87

'Taide 1

degree:
\euiatk
rauo ot
ovenah
fit usei
W'suall

Signifi

 

 

 

Table 18 Multiple Regression Summary of Fit Statistics

 

R 0.89
R2 0.80
R2 Adjusted 0.79
Root Mean Square Error 3.14
Mean of response 27.04

 

Table 19 shows the analysis of variance data that lists the sums of squares and
degrees of freedom used to form the whole model test'. The error mean squares or the
variation around the regression line is 9.88 and the model mean squares is 402. The F
ratio of 40.68, is highly significant (P < 0.0001), and suggests that the process variables
overall had a significant effect on protein extraction. A leverage plot of the whole-model
fit uses a scatter plot of actual response values against the predicted values to show
visually the significance of the variables. It indicated that the whole-model F test was

significant and also that the model fitted the data.

Table 19 Analysis of Variance of the Whole Model Test of protein content vs. the
predicted protein content in cowpea and navy bean extract

 

 

Source of variation DF Sum of Squares Mean Square F Ratio
Model 1 1 4420.68 401.88 40.68
Error 108 1066.87 9.88 P < 0.0001
C Total 119 5487.55

 

88

l
extractit
the ind:
highly s
was Sig

<0.18i|

ind ical t
horizor

Conclu

 

 

Particl.
cram
Sigma
0.70 a

and b

on p
aHal]
eVal
hoot.

(9H

R and R2 provide valuable information on the contribution of pH, particle size,
extraction temperature, and bean type taken as a group. However, it is useful to determine
the independent contributions of each of the four variables. Particle size and pH had a
highly significant effect (p < 0.0001) on protein content in the bean extracts. Bean type
was significant at p < 0.02, but extraction temperature did not have a significant effect (p
< 0.18) on protein extraction.

Leverage plots of the contribution of the main effects to the whole model fit
indicate that the confidence curves for pH, particle size and bean type crosses the
horizontal line, while that for extraction temperature does not. This further supports the
conclusion that pH, particle size and bean type were significant at p < 0.05 level.

Table 20 shows the correlation matrix of each individual process variable, i.e. pH,
particle size, extraction temperature and bean type correlated with protein content in the
extracts, and each other. pH and particle size were the only process variables that showed
significant correlations with protein content. pH had the highest positive correlation of
0.70 and particle size had a negative correlation of - 0.46. Both extraction temperature
and bean type were poorly correlated with protein content. None of the variables showed
any significant correlations with each other.

Since the previous statistical analysis reported no significant effect of temperature
on protein content in the extracts, temperature was removed fi'om the model, and the
analysis recalculated. A model that included all main and interaction effects was
evaluated. In this case, pH and particle size variables showed statistical significance
however, there was also a significant lack of fit. The three-factor interaction effect

(pH‘particle size‘bean) was removed fi'om the model and then recalculated. Significant

89

Table 2

EH

Proteir
Particlt

Tempe
Bean 1

 

‘ Absc

unde

“do

COWpe

TCCOVe

 

beane

in con

 

Table 20 Correlation Matrix of Variables in Protein Extraction of Cowpea and Navy

 

 

Bean Flour
pH Protein Particle Size Temperature Bean Type
pH 1.00
Protein Gym 1.00
Particle Size 0 -0.46 1.00
Temperature 0 0.06 ' 0 1 .00
Beanlype 0 0.10 0 0 1.00

 

‘ Absolute values of correlation coefficients 2 0.40 were significant at 5% levels and are
underlined

main and interaction effects were reported for pH, particle size, bean type, pH’bean type
and pH‘particle size. Lack of fit was not significant at 0.05.

Table 21 shows the protein content after three sequential extractions of 3g of
cowpea and navy bean flour at pH 10 and 25°C. These data were used to calculate protein
recovered from flour. Table 22 indicates that 87% of the total protein was recovered in
the cowpea extract and 8% in the residue, additionally, 91% was recovered in the navy
bean extract, and 5% in the residue. The remaining 5% and 4% protein unaccounted for

in cowpea and navy bean respectively were presumably lost due to experimental error.

90

 

 

Flour

Extract
Extract
Extract
Residu

‘

n = 3
1lnitia

Table

EXtra

RESid

 

 

 

 

 

Table 21 Protein content in Cowpea and Navy Bean Fractions (flour particle size
1.59 mm) afier three sequential extractions at pH 10 and 25°C

Sample Cowpea Navy Bean

% Protein Wt. Protein (g) % Protein Wt. Protein (g)
Flour 26.39 i 0.06 0.74 i 0.04 24.37 3: 0.09 0.68 i 0.14
Extract 1 48.44 i: 0.31 0.34 i 0.04 38.66 i 0.43 0.38 i 0.04
Extract 2 29.04 i 0.16 0.19 i 0.02 34.91 i 0.06 0.15 :t 0.09
Extract 3 18.48 i 0.02 0.12 :1: 0.10 34.61 i 0.22 0.09 i 0.14
Residue 2.15 i 0.30 0.06 i 0.02 11.98 i 0.09 0.04 :t 0.02
n = 3

‘ Initial Flour Wt = 3 g

 

 

 

Table 22 Protein Yield fiom Cowpea and Navy Bean Flour (particle size 1.59 mm)
after three sequential extractions at pH 10 and 25°C
Sample Protein Yield (g protein/ 100 g flour)
Cowpea Navy Bean
Extract 87.84 ‘ 90.86
Residue 8.11

 

91

 

panic
acidic
point
signif
partic

onthe

Ho wet
require:
CXIractit
on the li
Propertie
h‘

heating u
the emac
denalw'ed,
additionallj
Processing
SignificanCe
Particle 5,28

Optr'mUm eXtr.

Conclusions

The quantity of protein extracted from beans was significantly affected by pH,
particle size and bean type. Alkaline pH had the highest protein extraction yield, while
acidic pH had the lowest protein extraction yield, particularly at the apparent isoelectric
point of the legume proteins. The interaction effect of pH and bean type was highly
significant on percent protein extraction (p < 0.01), while the interaction of pH and
particle size was significant at p < 0.05. Temperature did not have any significant effect
on the protein content in the bean extracts.

The highest protein content was obtained at pH 12 and screen size 0.79 mm.
However, in determining optimized conditions for protein extraction, the energy
requirements involved during milling of cowpea flour, as well as heating during
extraction, had to be taken into consideration. The optimal pH conditions also depended
on the limits permissible for further ultrafiltration processing, and the resulting functional
properties of the protein extracted.

More energy was required for production of the smallest particle size as well as
heating to the higher temperature. In addition, screen size of 0.79-mm was too small for
the extraction processing equipment used in this study. At pH 12, the protein would be
denatured, which would affect the functional properties of the protein isolate;
additionally, this is also higher than the upper limit pH for further ultrafiltration
processing. The main efl‘ects of pH and particle size when investigated to determine
significance of the levels, showed no significant difference between pH 10 and 12, and
particle size 0.79 mm and 1.59 mm on protein content in the bean extracts. Thus, the

optimum extraction conditions selected for further ultrafiltration processing was pH 10,

92

part
fiorr

extra

donc

flours.

signific

CX
2' Cc

C01

particle size 1.59 mm and temperature 25°C. Under these conditions, 87% of the protein
from cowpea flour and 91% fi'om navy bean flour was extracted after three sequential
extraction experiments.

Ho: the process variables, pH, extraction temperature, particle size and bean type,
do not lmve any significant effect on protein extraction from cowpea and navy bean

flours.

Reject the H0 as stated and conclude that particle size, pH and bean type had

significant effects on protein extraction fiom cowpea and navy bean flours.

Future Research
The goals for future study include the following:

1. Conduct pilot scale optimization studies to obtain expanded data for Optimal
extraction conditions for further ultrafiltration processing
2. Conduct complete cost analysis to determine if the optimum protein extraction

conditions are cost efi‘eetive

93

STI

 

extre
since
ultraf‘;
range:
used tc
and mat
the cov
(CPPl, t
6'11! bait
fimctiona
isolate (sj
respecrjye)
for nary be
based on the

“’92 and Cl

 

STUDYZ COMPOSITIONAL AND FUNCTIONAL CHARACTERIZATION
OF ULTRAFILTRATION PROCESSED COWPEA (Vigna
unguiculata) AND NAVY BEAN (Phaseolus vulgaris) PROTEIN
FRACTIONS

Am

Ultrafiltration technology is increasingly being used to simultaneously purify, concentrate
and fi'actionate macromolecules without the application of heat or the use of either
extreme chemical or physical conditions. This is extremely important in isolating proteins
since it results in very little modification of structure and functionality. A plate and frame
ultrafiltration system with three flat sequential membranes (molecular weight cutoff
range: 5-10 KD, 15-30 KD and 50-100 KD, membrane area approximately 0.863 m2) was
used to evaluate the separation and functional characteristics of aqueous alkaline cowpea
and navy bean protein extracts (CPPE and NBPE). The flow rate (L/h) and flux (LMH) of
the cowpea and navy bean retentates (CPI or NBI respectively) and three permeates
(CPPl, CPP2, CPP3; and NBPl, NBP2, and NBP3 respectively) were measured during a
6-hr batch process. Protein content, recovery, molecular weight characterization and
functional properties of freeze-dried samples were compared to a commercial soy protein
isolate (SP1). Flow rate and flux ranged from 0.19 — 2.22 L/h and 0.52 - 4.10 LMH
respectively for cowpea; and from 2.45 — 9.40 L/hr and 2.83 - 10.89 LMH respectively
for navy beans. Protein content of the fiactions ranged from 9% - 52%. Protein recovery
based on the aqueous extract, was 75.6% in CPI, and 7.49%, 3.63% and 0.67% in CPPl,
CPP2 and CPP3 respectively; and 72.6% in NBI, and 4.70%, 2.14% and 1.98% in NBPl,
NBP2 and NBP3 respectively. Sodium dodecyl sulfate polyacrylamide gel

electrophoresis (SDS-PAGE) of the protein fractions indicated band patterns in the range

94

of 6

C PP
Ware
comp.
absorr
(EC)a
FC lou
CPI and
SP1. Am
UF proce

discrete p

 

of 6.5 - 66 KD for CPI, l4 - 20 KD for CPPl, 14.5 - 29 KD for CPP2, 14 - 55 KD for
CPP3, 14 — 66KD for NBI, 7 KD for NBPl, lSKD for NBP2 and 16 —24 KD for NBP3.
Water absorption capacity (WAC) and protein solubility (PS) of CPI were higher
compared to SP1; however for NBI, WAC was lower and PS higher than SP1. Oil
absorption capacity (OAC) for both CPI and NBI were higher than SPI. Emulsification
(EC) and foaming capacity (FC) of NBI was higher than SP1. EC for CPI was higher, but
FC lower than SP1. The critical protein concentration for gelation and gel strengths of
CPI and NBI were lower than SPI. Thermal stability was greatest for CPI, then NBI and
SP1. Amino acid profiles for the fiactions were similar and increased afier extraction and
UF processing. The plate and frame ultrafiltration membrane system employed, provided

discrete protein fi'actions with functional properties.

95

11391122

I
fractiona
chemicai
it results
utilized i
on the b
weight 5
permeate
(NFPA, l

L'
the fractir

and is We

Introduction

Ultrafiltration technology (UF) is used to simultaneously purify, concentrate and
fractionate macromolecules without the application of heat or the use of either extreme
chemical or physical conditions. This is particularly important in isolating proteins since
it results in very little modification of structure and firnctionality. It is being increasingly
utilized in the food industry to separate molecules through a porous polymeric membrane
on the basis of their molecular weight. It results in concentration of higher molecular
weight solutes in the concentrate, and the production of a lower molecular weight
permeate stream, depending on the molecular weight cutoff (MWCO) of the membrane
(NFPA, 1993).

UP is now a well accepted processing operation in the dairy industry. It is used in
the fiactionation of cheese whey and pre-concentration of milk for cheese manufacture,
and is well reported in the literature (Tirnmer, 1997; Sirnbuerger et aL 1997; Cheryan &
Chiang, 1984; Emstrom, 1985, 1986). In the fruit industry, Yu et al, (1997) and Gao et al,
(1997) reported on the use of UP for West Indian cherry juice and apple juice processing,
respectively. In vegetable protein processing, UF has been increasingly utilized as a
replacement for conventional isolation methods. Lawhon et al (1981, 1982) reported on
the use of a combined aqueous extraction procedure to produce protein and oil food
products from un-defatted soybeans. They also used UF membranes to co-process soy
protein extracts and skim milk to produce a soymilk food ingredient. Lui et a1 (1989), and
Kwon et a1(1996), reported on the UP conditions for processing soy and coconut protein

isolates, respectively, to improve protein functionality.

96

soybear
it was c
successf
use of ‘
preparat
bean inc
(1987) 5
yields of
LT procc

lr
processin
attracts \
the native

Tl
Chemical 1

50y preten

More recently, research has been expended on leguminous species other than
soybeans. Bolles (1997), reported on the quality of a navy bean retentate, and found that
it was of similar quality and functionality as a soy protein isolate and could be utilized
successfully in an aseptic processed infant food drink. Ko et al, (1994) reported on the
use of UF to recover protein from simulated waste water during mung bean starch
preparation. They found that protein recovery improved significantly, utilization of mung
bean increased, and pollution problems reduced. Juarez and Lopez (1994) and Berot et al
(1987) showed that the stability of juice from the tropical legume jicama improved, and
yields of fababean protein isolates were higher with improved functional properties after
UF processing, respectively.

In this study, the separation performance of flat plate membranes used in UP
processing of cowpea (Vigna unguiculata) and navy bean (Phaseolus vulgaris) protein
extracts was evaluated. The physico-chemical properties were then determined to assess
the native status of the protein isolates, and their suitability as food protein ingredients.

The null hypothesis (Ho) being tested is there is no difference in the physico-
chemical properties of ultrafiltered cowpea and navy bean protein isolates compared to a

soy protein isolate.

97

Co
lllachinei
at 4°C prior

The
alkali solve
extracts we
California) .
lOOKD resp
flint calculat
schedule rect

The f

using the ”ill

from this. SD

 

 

Materials and Methods

Cowpea and navy bean flour milled in a Fitzpatrick mill (Model D Comminuting
Machine, Fitzpatrick Co., Chicago, IL) with screen size 1.59 mm was prepared and stored
at 4°C prior to use.

The bean flours were extracted three times in extractor baskets in an aqueous
alkali solvent mixture, combined then sedimented overnight at 4°C. The combined
extracts were pumped through three series of flat plate membranes (Osmonics Inc.,
California) with molecular weight cut off (MWCO) points of 5-10KD, 15-30KD and 50-
100KD respectively, and total surface area 0.863m2. The flow rates were measured and
flux calculated. The UF unit was cleaned after each run using the acid/alkali cleaning
schedule recommended for protein products (EcoLab, 1992).

The fiactions were fieeze-dried, protein content of the samples were determined
using the micro-kjeldahl method (AACC, 1984), and a protein mass balance calculated
fiom this. SDS-PAGE was used to determine the molecular weight range of the fi‘actions,
and thermal properties were determined using a DSC Thermal Analyzer Model 2200 (TA
Instruments, New Castle, DE).

Functional properties including water and oil absorption capacity, protein
solubility, foaming capacity and stability, emulsion capacity and stability, gelation and
colorimetric characteristics were determined for all fi'actions (Sathe & Salunkhe, 1981;
Satire et al, 1982; Coffinan & Garcia, 1977; Beuchat, 1977; AACC, 1983).

The amino acid profile of each fraction was determined using acid hydrolysis,
methanesulfonic acid (MSA) hydrolysis for tryptophan and formic acid oxidation prior to

acid hydrolysis for cysteine. Phytohemagglutinin screening was measured using the

98

method 1
digestibil

three enz

ME

M
LT proce
content d
samples L'
Was used
most affe

mttritional

method reported by Occena (1994) with Sigma Chemical kits P0526 and P8629. In-vitro
digestibility was determined by the four enzyme pH-Drop method (AOAC, 1990) and the

three enzyme pH-Stat method described by McDonough et a1 (1990).

Statistical Analysis

Microsoft Excel was used to compute the averages and standard deviations of the
UP processing data. Stat View (SAS Institute Inc., 1998) was used to analyze the protein
content data, functional and nutritional properties and determine differences between
samples using Fisher's LSD. Factor analysis from Stat View (SAS Institute Inc., 1998)
was used to determine which of the firnctional properties were the principal components
most affected by ultrafiltration processing. Correlation matrices of the functional,

nutritional and thermal properties of the fractions were generated.

99

Results.‘

1
and 12 It
experime
permeate
reported

mineral r

Results & Discussion

The flux data for cowpea and navy bean is illustrated graphically in Figures 11
and 12 respectively. In both cases, flux decreased particularly during the initial part of the
experiment. Within about 1/2 hr the flux was significantly reduced for all three
permeates, then remained almost constant during final concentration. This was also
reported for ultrafiltration of fababean protein extracts by Berot et a1 (1987) using a

mineral membrane and for peanut protein products by Lawhon et a1 (1981).

This rapid decrease in flux suggests possible plugging of the membrane as the
CPI and NBI concentrations increased, and occurred as soluble protein in the aqueous
extract formed a gel during UF processing which lead to fouling of the membrane. Berot
et al, (1987) described this phenomenon in terms of the structure of this gel or
polarization layer and suggested that it was dependent on the conformation of the
proteins.

The lower molecular weight components all had higher flux rates. Similarly, the
mean flow rates and flux for NBP3 and NBI were significantly higher than the
corresponding values for CPP3 and CPI. This can be explained by the molecular weight
differences between NB and CP. The molecular weight distribution from SDS-PAGE
indicates that components in the navy bean fractions have lower molecular weight
distribution than cowpeas. This supported the study by Lawhon et al, (1981). They found
that during membrane processing of soybean proteins, the mean flux was higher with
lower molecular weight extracts. Kwon et a1 (1996) also reported on higher initial flux

rates using a 10KD membrane compared to a 5KD, during hollow-fiber UF processing of

100

3...—
...5. «5.: a

Figure l

«2Q Ext: 3.:
~

Figure l

 

 

 

10

Flux (L/m 21hr)

 

 

 

 

0 l 2 3 4 5 6

Time (hr)

 

 

 

Figure l 1

Flux of Cowpea Protein Fractions during Ultrafiltration Processing

 

 

 

 

 

 

 

i-O—NBPI t

 

 

 

+NBP2 ' *
+NBP3 t -

+NBI ll

 

r
"rm (hr) ‘
Figure 12 Flux of Navy Bean Protein Fractions during Ultrafiltration Processing

101

COCO I'll.

more 3

(CPI) a
3 (CPF
for CPI
extract
process
was 10.
.\PP3 a
COWpea
and CPI

LMH. 3.

coconut proteins. As UF operation time increased, the 5KD membrane yielded higher and
more stable flux rates.

Cowpea protein extract (CPPE) was concentrated from 45L to 14L of retentate
(CPI) after a 7-hr batch process. The volume of CPPE separated into permeates l, 2, and
3 (CPPl, CPP2, CPP3) was 15L, 12L and 1.32L, respectively. The mean flow rate (L/hr)
for CPPl, CPP2, CPP3 and CPI were 2.22, 1.78, 0.19, 2.03 L/hr. The navy bean protein
extract (NBPE) concentrated from 32L to 11L of retentate (NPI) afler a 6-hr batch
process. The volume of NBPE separated into permeates 1, 2, and 3 (NPPl, NPP2, NPP3)
was 10.5L, 5.7L and 4.3L, respectively. The mean flow rates (L/hr) for NPPl, NPP2,
NPP3 and NPI were 3.54, 2.45, 2.94 and 9.40 L/hr respectively. The mean flux rates for
cowpea were 4.10 LMH, 3.03 LMH, 0.52 LMH and 2.25 LMH for CPPl, CPP2, CPP3
and CPI respectively. The corresponding values for navy bean were 4.07 LMH, 2.83

LMH, 3.41 LMH and 10.89 LMH for NPPl, NPP2, NPP3 and NPI respectively.

Table 23 shows the protein content (dry basis) and % solids of cowpea and navy
bean protein fractions after pilot plant protein extraction at pH 10 and 25°C and
ultrafiltration processing respectively. The protein content of CP fi'actions was greater
than NB. CPI had a protein content of 52% and NB had 48%, compared to 89% for the
commercial soy protein isolate. The values for the experimental samples were lower than
what was reported for other legume products. Berot et a1 (1987) reported protein contents
of 85% - 91% for fababean protein isolates, Lawhon et al (1979) 80% - 92% for soy
isolate fi'om toasted and untoasted soy flour and Vose (1980) 89% and 94% for field pea

and horsebean isolate respectively.

102

.22.--- VCZQZCAEZCL I

......ESSEED ccw :ozuwbxm £5937< u

NFEmOUOLQ
ace $520.2... Sexual teem x>nZ pea aoQfiOUko £50m

mm Uta 233200 Segment MN 23.3.

mm.oHo_.m

N_.OHov.o

mud H 3.0

ONO H 3.0

co; H 5.0—

9: H8.—

modH 5.—

vvd H oNN

Cod H awe

ood H 2.3

£50m o\c

:mom >>mZ

mvd H case

mm._ HN_.NN

MN.NHw_.:

v.0.— H www—

om; Hm;

mmd. Homsm

mwh H mde

mud H 2.0m

_o._ Hmm.mm

8.0 H Slum

5085 X.

mm; H 5.—

N_.0H0m.o

and H mmd

0N6 H Omd

cog H SNN

00; H ONN

m_.onv.o

oOoHEA

two H owN

owd H NOmw

mezom e\o

cuckoo

mmd HER—m
mm.m H522
Om.m H 3.:
ma H N90.
vmg H913

mm.o_ HEdm

3% H 3.3
mud H wmhm
5; thav

mod H boom

£82m c\c

. m u. a
2856M

m 08083;

N 032:0;

_ 888:8

0:281

flombxm 859500

mfimmoooi

constants :5 :ocombxm 3833c coca Ectoenm £88m :wom >32 :5 monaoo .«o mum—om o\.. :5 33:00 530:—

m Sabxm

m guacam—

_ Sahxm

50E

3:88:80

mm 033.

103

A
highly sir
indicated

and CPR

Table 2

 

V at law

 

Fractic
Residu

\

Prote
Prote
and .
Were
37.9

I‘esp

170/c

An analysis of variance (Table 24) indicates that ultrafiltration processing had a

highly significant effect (p < 0.0001) on the protein content of the fractions. Fishers LSD

indicated that all of the samples were significantly different fi'om each other except CPPl

 

 

and CPR.

Table 24 Analysis of Variance of Cowpea and Navy Bean Fractions after UF
Processing

Variance DF SS MS F Value P Value

Fraction 14 17603.76 1257.41 14801.19 < 0.0001

Residual 30 2.54 0.08

 

Figures 13 and 14 illustrate the protein mass balance of cowpea and navy bean

protein fractions after UF processing. It indicated that about 63% and 52% of the total

protein in cowpea and navy bean flour respectively was recovered in the extract, and 33%

and 41% in the residue respectively. Recoveries based on the total protein in the flour

were 47.2%, 4.7%, 2.3% and 0.4% for CPI, CPPl, CPP2 and CPP3 respectively; while

37.9%, 2.4%, 1.1% and 1.0% were received for NBI, NBPl, NBP2 and NBP3

respectively. The total loss of protein during extraction and UP processing was 13% and

17% for cowpea and navy bean respectively.

104

Figur

Figure 13 Mass Balance of Cowpea Protein Fractions after Ultrafiltration Processing
NBF
100 parts
1
T l l
NBPE NBR Loss
52.1 parts 41.2 parts 6. 7 parts
1 l l l
NBI NBPl NBP2 NBP3 Loss
37.9 parts 2.4 parts 1.1 parts 1.0 parts 9. 7 parts
Figure 14 Mass Balance of Navy Bean Protein Fractions after
Ultrafiltration Processing

 

CF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100 parts
1
l l l
CPPE CPR Loss
62.8 parts 33.1 parts 4. 9 parts
1
l l l L l
CPI CPPl CPP2 CPP3 Loss
47.2 parts 4.7 parts 2.3 parts 0.4 parts 8.5 parts

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

105

res

frac
NB

(Ro

bean
cow;
triple
corre:
Obser
much
flour ;
l'SOIate

more:

The total protein recovered from CPPE and NBPE was 87% and 81%
respectively. About 75% of the total protein was recovered in CPI, and 7.5%, 3.6% and
0.7% recovered in CPPl, CPP2, CPP3 respectively. This was similar for the navy bean
fi‘actions, about 73% was recovered in NBI, and 4.7%, 2.1%, and 2% recovered in NBPl,
NBP2, NBP3 respectively. These values agreed with experiments reported for chickpeas

(Romero-Baranzini et al, 1995).

Figure 15 shows a photograph of the SDS-PAGE pattern from cowpea and navy
bean protein fractions in the presence of 2-ME (2-mercaptoethanol). The pattern for
cowpea flour, extract, residue and isolate showed distinct bands at about 97KD, a major
triplet band at 62KD, 56KD, 52KD, and a doublet band at 36KD and 32KD which
corresponds mainly to the globulin and albumin fi'actions. Nine other protein bands were
observed in the cowpea flour and ranged from 156KD — 38KD. However these were
much less prominent than the subunits previously described. Two of these bands in the
flour fraction were greater than the 205KD protein standard. The extract, residue and
isolate had four other bands that were similar to the flour fraction and corresponded to

lO6KD - 38KD.

The SDS PAGE pattern for the navy bean fiactions showed distinct doublet bands
at 98KD and 96KD and again 50KD and 46KD. A single band was observed at 33KD.
Other bands observed ranged from 156KD — 39KD. The residue fraction produced bands
only at SOKD, 46KD and 39KD. A similar molecular weight profile was reported for
navy bean meals, extracts and concentrates by Komhorst et a1 (1990) after air

classification and isoelectric precipitation of navy bean meal.

106

bands
NBF 1
were 5

similar

- 101<
reportt
concer
actual
memb

{\11 Cl
memb

Obsen

that W

ProfilE

Figure 16 shows the pattern for the cowpea and navy bean permeates. No protein
bands were observed on the gel for permeates 1 and 2 for both legumes (CPPl, CPP2,
NBPl and NBP2). CPP3 showed distinct bands at 62KD, 56KD, 52KD and 36KD that
were similar to bands in CPF; and NBP3 showed bands at 50KD and 33KD that were

similar to the major bands in NBF.

The MWCO points of the three series of flat plate membranes used included 5KD
— 10KD, 15KD - 30KD and 50KD - 100KD respectively. The SDS-PAGE profile
reported above, reflects the MWCO ranges with some changes in selectivity due to the
concentration polarized layer as reported by Blatt et al (1970), as well differences in
actual molecular weight during manufacture of the membrane. This suggests that the UP
membranes were able to fractionate the protein extracts according to their molecular size.
All components smaller than 10KD, 30KD and 100KD should pass through the
membranes and produce permeates 1, 2 and 3. However, the lowest molecular weight
observed on the gel was about 36KD which explains why no bands were observed for
permeates l and 2. The isolates for both legumes showed a wide molecular weight profile

that was similar to the corresponding flour fractions.

These data also indicate that there is a difference in the protein molecular weight
profile of navy bean and cowpea proteins. This was expected since the legumes are two
distinct species. The molecular weight of the major subunits in navy beans was generally
lower compared to the major bands in the cowpea fractions. Kohnhorst et al (1990) and

Khan et a1 (1980) also reported these molecular weight ranges for navy beans and

cowpeas respectively.

107

8V 0:202 £085 :03 .90: .33 0:28: :82 >20: AC “08:8 5055 :002 >>0= A8 So:
:32 in: AC 20.02 E053 «0958 A3 03200: 00958 AC 80.58 2085 009500 AS Soc «0958 A: uaucflm
M3312 2.3% 36802; 5:8:ch Eta 228:5 E085 :00m x>mZ 0:: 009500 .8 E033 mostwmm 2 05mm.“

Hammofi mwmmfl

108

I
I

|'\iI
I!!!

 

8V m 0.00853 :00: :90: .va N 0.00.53 :00: >20: AC _ 030503 :00: >>m= 48

So: :00: >20: AC M 06085: 03300 Avv N 0000::03 03300 AC _ 0:008:03 03300 AC Soc 0338 A: Bagsm
3912 «SEE 3600083 gag—@055 Sta Ectofim 0:005:03 :00m >202 :5 03300 30 E00003 $83-QO

I
I

’
I

II”
III!

I

 

 

2 050a

109

The thermograms of lO-mg solid samples of cowpea, navy bean and soybean
protein fractions were similar to the generalized thermal curve for legume globulin
reported by Petruccelli and Anon (1996). The peaks for all samples did not demonstrate
the characteristic broadening expected with denaturation, these data suggests that the
samples were not completely denatured. Nagano et al (1994) and Wang and Damodaran
(1991) reported that broadening peaks indicated a shift in conformation of the protein to
the unfolded state, which is associated with hydrogen bond disruption in protein

denaturation.

The thermal properties of the fractions are presented in Table 25. They indicate
that initial transition temperatures (Ti) ranged from 82.9°C - 105°C, peak transition
temperatures (Tp) ranged from 935°C — 125°C, and enthalpies ranged fi'om 0.33 — 0.55
W/g. The transition temperatures were higher and the AH values lower than what has
been reported for other legume proteins. This was due to the moisture content of the
samples in this study, which ranged from 3-5%. This was significantly lower compared to
samples in other studies in which slurries of up to 20% moisture content were used
(Gorinstein et al, 1996; Martinez and Anon, 1996; Liao et al, 1996; Erdogdu et al, 1995;
and Okechukwu and Rao, 1997).

It has been reported that lowering moisture content increases the thermal stability
of proteins (Amtfield et al, 1985) resulting in higher transition temperatures. They
reported that as water is removed fi'om a protein system, there is insufficient water in the
vicinity of the protein molecule to bring about the thermal transition. As protein
denaturation occurs and the protein unfolds, there is a transfer of hydrogen bonds from

protein-protein to protein-water. If there is limited water in the system, then this reaction

110

is only possible to a limited extent resulting in lower AH values. Further, at low moisture
levels, water is held tightly due to its association with the protein. This suggests that the
energy that is required for mobilizing water is increased, resulting in elevated transition

temperatures.

Table 25 Transition Temperatures and Enthalpies of Cowpea and Navy Bean
Fractions after Ultrafiltration Processing ‘

 

 

 

Samples Ti (°C) Tp (°C) AH (W/g)
CPPE 82.9 i 1.27 a 94.15 i 1.34 a 0.37 i 0.01 a
CPI 97.0 i 2.82 b 111.5 i 3.53 b 0.34 i 0.03 8,1)
NBPE 91.5 i 1.20 b,c 115.7 :1: 0.99 b,c 0.33 i 0.01 C
NBI 104.5 :1: 0.71 d 125 i 2.83 d 0.52 :1: 0.02 (1
SP1 105 :t 4.24 d,e 1 16 i 2.83 b,c,e 0.55 :t 0.03 d,e
In=2

Means within the same column followed by different letters are significantly
difl‘erent (p < 0.05)

UF processing had a significant effect on the thermal stability of the fractions (p
< 0.05). Both transition temperatures and AH values increased for all samples after
extraction and UP processing. Navy bean fractions had higher transition temperatures and
enthalpies overall than the cowpea and soy fiactions, which indicates that these samples
unfolded more during extraction and UP processing. There were no significant
differences in Ti between NBI and SP1 and NBPE and CPI. For Tp, there were no

significant differences between NBPE and SP1; and CPI, SPI and NBPE. AH for CPI and

111

CPPE was significantly different from both NBI and SPI; however NBI and SP1 were not
significantly different from each other. Additionally, AH for CPPE and CPI were not
significantly different. These results indicated that denaturation after processing was

more significant with navy bean and soybean fi'actions than cowpea fractions.

Arntfield and Murray (1981) observed an increase in transition temperatures
after heat processing. They suggested that the increase observed was due to the removal
of low temperature denaturing proteins during the heat process, which resulted in higher
average transition temperatures for the remaining proteins. In this present study, although
there was no heating, low temperature proteins may have separated into permeates after

extraction and UP processing resulting in higher temperature proteins in the isolates.

The results on AH support the study by Murray et a1 (1985) on thermal behavior
of fababean and field pea proteins during purification. They reported an increase in the
AH values as protein content increased during the purification process, and found a
significant correlation between protein content and AH. In this study, a significant
correlation of 0.77 (p < 0.05) was found between AH and protein content of the fractions.
This could be interpreted as an increase in the proportion of denatured protein as protein

content increases, which is reflected by an increase in AH as measured by DSC.

The amino acid chromatograms of cowpea and navy bean fiactions were well
resolved and allowed identification and good quantification of the amino acids present in
the legume fractions. Due to instability to acid hydrolysis, cysteine and tryptophan were
either absent or present at very low levels in all fractions, and in addition, proline was

absent in both NBPE and NBI, while histidine was absent in NBI. The amino acid

112

composition for standard hydrolyzate, cowpea fractions, navy bean fractions, soy isolate
and casein, expressed as mg amino acid / 100g protein is presented in Table 26. The
amino acid profile of casein, soy isolate and the flour fractions corresponded to what has
been reported in the literature (Mnenbuka and Eggum, 1995; Patel et al, 1980; and Elias

et al, 1976).

The amino acid profile of CPF, CPPE and CPI reflected similar types of amino
acids, which increased after extraction and ultrafiltration processing. This was not
observed overall for navy bean fractions. The profile of NBPE and NBI did not reflect
similar amino acids compared to NBF. Proline was not detectable in either NBPE or NBI.
Alanine and methionine was either not detected or reported in very low levels for all
fi'actions. Changes in retention and resolution for navy bean may have been due to poor
resolution of alanine with proline, or by sample interference from components such as
lipids, minerals and other non-proteinaceous materials. Both cowpea and navy bean
fractions were richest in aspartic acid, glutamic acid, lysine and valine, and contained

lowest levels in the sulfur amino acids and tryptophan.

These results suggest that UF processing concentrated the quantity of each amino
acid significantly fi'om flour to isolate. However, the amino acid content of these
fractions was still considerably lower compared to those for SP1 and casein. The amino
acid profile observed for the flour fractions was similar to what has been reported in the
literature for cowpea and navy bean flours (Maneepun et al, 1974; Hang et al, 1980;

Kanamori et al, 1982).

The data for protein solubility (PS) is shown in Table 27. PS indicates how the

protein will perform when applied in particular to liquid food systems and is also a

113

 

 

00.0 8.0 00.0 0_.0 0h.~ 00.0 N00 N06 _0.h N00 0.3
0.0 000— 00.0 0N0 _0.0 00.0 2.0 00.5 00.0 00.0 0.:
N0.— 00.0 00.0 00.0 00.0 00.0 20.0 00.0 00.0 0:. :04
0N0 NNN 00.0 00.0 00.0 _0.0 _0.0 00.0 00.0 00.0 0:
00d 02 _0.0 00. _ 92 0.2 DZ 2 92 0m; 02
N50 00.0 No.0 00 .0 N00 0_.0 00.0 00.0 00.0 N00 _0>
00.0 00.0 RN 00.0 00; 002 000 0N0 00.~ 00.0 EC.
00.0 000 :N 0.0 2 Q2 0.0 _0.0 00.0 00.0 8n—
00.0 2 8.0 00.0 0N._ 00.N OZ DZ QZ ~02 0?
00.0 00.~ 0N0 _0.0 000 00.0 00.0 3.0 0N0 «0.0 .5...
00.0 3.0 00.0 00.0 00.0 00.: 0.0 00.0 00.0 00.0 0.3.
0N0 00.0 00.0 00.0 00; 00; 00.0 00.0 06 00.0 $2
00.0 N00 00.0 9.0 00.0 3.0 00.0 _0.0 00.0 2.0 05
No.0 00.0 00.0 2.0 :4 00.0 0N0 00.0 00.0 5.0 00m
00.: 00.2 00.0. 00._N 00.0 0N.0N 000— 00. _N 00. 3 0~.0_ 20
00.0. 00.: 0.: 000— 02.0 00.0_ 00.02 000— 0.3 0_.: 02
250400 Em ~52 MED 52 03002 .02 EU man—O ”EU 20< 055

 

0:00.000 :00m .902 0:0 03260 .00 25083 000. \00 5203800 200. 05:10.

00 030,—.

114

practical measure of the extent of protein denaturation due to processing. The higher
solubility of the experimental fractions compared to SPI at neutral pH indicates that CPI
and NBI were less denatured than SP1, and would be useful in liquid food formulations.
This data agreed with the results from the thermal study, which indicated that NBI and
SP1 were more denatured than CPI. These findings were also reported by Kohnhorst et al

(1990) afier air classification of navy and kidney bean flours.

Table 27 Protein Solubility of Cowpea and Navy Bean Protein Fractions at pH 7 '

 

Fractions Protein Solubility (%)

CPF 39.39 i 0.37 a

CPPE 25.04 i 0.531»

CPI 52.09 i 0.88 c

NBF 22.44 i 0.35 d

NBPE 16.25 i 0.40 e

NBI 36.26 i 0.63 f

SPI 16.61 i 0.21g, e
..l n = 3 ..............

Means followed by different letters are significantly different (p < 0.05)

Tables 28 and 29 present the data for water and oil absorption capacity (WAC and
OAC) respectively, and indicates that WAC of the cowpea isolate (CPI) was significantly
different fi'om the soy isolate (SPI) and the navy bean isolate (NBI). This is related to
protein denaturation of the SP1 and NBI fi'actions as a result of isoelectric precipitation
production for SP1 and increased shearing during UF processing for NBI. The OAC of

CPI and NBI were significantly higher than SPI. This agreed with other work reported by

115

Berot et al, (1987) and Sathe et al, (1982) for Fababean and the Great Northern Bean
protein isolates respectively. These values are good indicators that the test proteins could

be incorporated into food formulations like doughs and meat extenders.

Table 28 Water Absorption Capacity of Cowpea and Navy Bean Protein Fractions '

 

Fractions 9 i WAC
(g H20 / g protein)
Cowpea Navy Bean Soybean
Flour 3.75 :t 0.01 a 8.21 i 0.66 d —--
Extract 2.36 i 0.25 b 6.70 i 0.69 e --
Isolate 9.61 i 0.15 c 6.26 i 0.40 f 7.72 i 0.45 g

Means followed by different letters are significantly different (p < 0.05)

Table 29 Oil Absorption Capacity of Cowpea and Navy Bean Protein Fractions '

.A‘u-‘M _ “.4...” ,...-.. . ....

 

. ...-“W” 0.. -0. -..—--- .

Fractions . OAC “—
(g H20 / g protein)
Cowpea Navy Bean Soybean
Flour 19.80 :t 0.29 a 26.67 i O. 14 d «-
Extract 16.58 i 0.07 b,e 18.77 i 0.61 c,d ---
Isolate 18.61 i- 005 c 16.68 i 0.10 e 7.96 i 0.11 f
1__a,:3_,._,,,_._-., “......“ .-- ----,_...--...- .----2- -...

Means followed by different letters are significantly different (p < 0.05)

116

Foam capacity (FC) calculated as the percentage volume increase of 5%
suspensions of the protein samples is shown in Table 30. It was highest for the navy bean
fractions (NBI, NBPE), then SPI, and much lower for CPI. The foam stability over a 6-
hour period is illustrated in Figure 17. The decrease in volume was greatest for CF.
CHPE and NBF compared to SPI, NBPE and NBI. However, after 6 hrs, the only
samples with foams remaining were CF and CPI.

These results can be explained by the properties of a protein for foamability and
stability. In NBI, the protein structure is more denatured than in the flour or extract
because of shearing during UF processing, and in SPI, unfolding occurred during
isoelectric precipitation. This unfolded structure facilitates migration to the interface, and
thus the formation of foams. The presence and amount of native rather than denatured
proteins have been shown to be related to higher foam stability (Yasumatsu et al, 1972;

and Lin et al, 1974). This supports the result that the CF and CPI fi'actions were the only

Table 30 Foam Capacity of Cowpea and Navy Bean Protein Fractions '

" .fiv ix ail—.027.“ ‘0 a_fi. $13..- .‘ La-Efifi‘“.hn“ ‘.Z;T*'I1_ .g’ 2“; _‘

 

 

Ftaétiotis ’ " '" Foétit C0563?
(% Volume increase)
Cowpea Navy Bean Soybean
Flour 10.67 i 0.16 a 19.33 i 0.58 d ---
Extract 11.67 i 0.58 a,b 56.67 i 3.05 e ---
Isolate 29.00 i: 1.00 c 58.67 i- 1.15 e,f 52.00 i 2.00 g
-1 n = 3 -

Means followed by different letters are significantly different (p < 0.05)

117

samples with foams remaining afier 6 hrs. The high foam capacity and stability of NBI
makes it desirable as an ingredient in whipped food products.

Emulsion capacity (BC) was determined subjectively, and calculated as the ml
oil/mg protein (McWatters & Cherry, 1981). EC of all the fractions was higher than the
SPI, and decreased with each processing step (Table 31). This supported the results of
Okezie & Bello (1988) and McWatters & Cherry (1981) for winged bean and field pea
respectively, indicating that as protein content increased, the emulsion capacity
decreased. The higher values for the NB fractions and SPI was due to increased
denaturation of the proteins in those samples. The emulsion stability of the isolates and
SP1 (Figure 18) was high, after 38 hrs there was only about a 10-ml separation. This high
emulsion stability may be due to the globular nature of the major proteins in legumes, and

suggest that the fiactions could be used in meat, ice cream, and textured protein products.

Table 31 Emulsion Capacity of Cowpea and Navy Bean Protein Fractions

Fractions if Emulsion
(g oil / mg protein)
Cowpea Navy Bean Soybean
Flour 0.14 i 0.03 a 0.29 i 0.02 d ---
Extract 0.13 i 0.02 b 0.21 i 0.01 e ---
Isolate 0.11 i 0.02 c 0.19 i 0.04 f 0.06 i 0.03 g
1.3: _3._,,.,.,,__-_,--___._.. ...-..W._-_--. W..- .... -__..----.W-.- .-. _- ..W

Means followed by different letters are significantly difl‘erent (p < 0.05)

118

 

 

160
a 150 \ Efl-O-"SPT—WI
g I40 +CF
L2 .+CPPE l
“5 130 3+Cn
2 120 .+.\BF ,
.5. no \ '+‘\BPEi
g \ K , g—I—NBI 1

lm ‘ :Tf’. .22, :pi. _

90 . 0

o 5 60 120 150 180 300
T1me(minutes)

 

 

 

Figure 17 Foam Stability of Cowpea and Navy Bean Protein Fractions

 

  

O+t I—r r—Ibtl0 i+m3

‘ .
10 Z) 35 60 III) 14) 84) 2230 L-l—mil
l

 

'fa‘ 6)

G 50

B 4) 1—O—81 l
E. 30 1+0: 1 1
{3 20 1+0! 1 i
.2 10 3
O

2

.5

O

>

Tm: (nints)

 

Figure 18 Emulsion Stability of Cowpea and Navy Bean Protein Fractions

119

Table 32 shows the gelation characteristics of the protein fractions. The critical
gelation protein concentrations (% wt/vol.) were 2.66, 6.32 and 9.32 for CF, CPPE and
CPI respectively. These were significantly different from the corresponding navy bean
fractions which were 1.95, 6.71 and 7.67 for NBF, NBPE and NBI respectively. The
critical gelation protein concentration for SP1 was 12.27.

Findings of this study indicate that the flour and extract fractions gelled at lower
protein concentrations than the isolates, and that the critical protein concentrations (%
wt/ml) for gelation increased with increasing protein content. This supports the work of
Schmidt (1981), who reported that a high protein concentration is required for the
gelation of globular proteins, and that solutions containing both protein and

polysaccharides will form gels at relatively low concentrations of the gelling material.

The textural characteristics from an instrumental TPA of 18% gels are also
indicated in Table 32. The gel strength determined by 1) the work required to penetrate
the gel and 2) gel hardness was highest for the flour fractions than extract or isolate,
however, SPI had the highest gel strength compared to the experimental isolate fractions.
This supports the least gelation concentration data reported above. In general, the CPI
and NBI gels were more adhesive and cohesive, and less chewy and gummy than the SP1

gels.

Color reflectance values for the protein fractions are illustrated graphically in
Figures 19 - 21. The data indicates that the CP and NB fractions were darker, and had
more yellowness and greenness characteristics afier extraction and UF processing. SPI
was lighter, and had more redness and blueness characteristics than the experimental

samples. Analysis of variance indicates that UF processing had a highly significant (p <

120

00.0 v 3 80000.00 000000000? 000 8000— 2—05:00 00 0032—00 08200 0800 000 .0003 8002
0 u a _

 

w 00.0H00.0_ 0 00.0H00.00 000.0H00.0_ 0 :.0H0_.0 00_.0H00.: 000.0H 00.0 000.0H00.0 Em
.0 0_ .0 H 00.0 0.0 00.0 H 00.0. 0.0 0. .0 H 00.0 0.0.0 00.0 H 00.0 0.0 2.0 H 00.0 0.0 00.0 H 00.0 0.0 .00 H 00.0 52
0 00.0 H 00.0 0.0.0 5.0 H 00. .0 0 00.0 H 00.00 0.0 00.0 H 00.0 0 00.0 H 00.0 0 00.0 H 00.0 0 00.0 H 00.0 mmmz
0 00.0 H 00.. 0 00.0 H _0.00 0 00.0 H 2.00 0 00.0 H 00.0 0.0 2.0 H 00.0_ 0 00.0 H 0: 0.0 00.0 H 00.0 ”52
0 00.0 H 00.0 0 00.0 H 00.00 0 00.0 H 00.0 0 00.0 H 00.0 0 2.0 H 00.0 o 00.0 H 00.0 0 00.0 H 00.0 EU
0 2.0 H 00.0 0 00.0 H 00.00 0 00.0 H 000— 0 00.0 H 00.0 0 00.0 H 00.0 0 00.0 H 00.. 0 3.0 H 00.0 man—U

0.00.0 H 00.0 0 _0.0 H 0000 0 00.0 H 00.00 0 00.0 H 2.0 0 00.0 H .00. 0 00.0 H 0K..— 0 _0.0 H 00.0 “EU

 

25:30.0 2 55.2 .552 00 88.2 z
802.5 .0 com 0.0 0 cow :0: cow 0.»: com .000 com .00: com 0.00
.250 508.0 .0055 0000000: 0:00 0.83 00053000 00058800 0005300000. 00000300000 00280..“—

 

_ 05000080 0000000005 0000 000000..”— 5085 50m .902 0:0 000300 00 00000000000000 0030—00 00 200.0

121

0.0001) effect on the color of the fractions. In addition, all of the fractions were
significantly different fiom SPI, the control. Xu and Diosady (1994) and Tjahjadi et al
(1988) observed this darkening color for chinese rapeseed and adzuki bean proteins,
respectively after extraction and isolation.

The F-values of all fianctional properties and their statistical significance indicate
that extraction and UF processing had a significant effect on the functional properties of
the fractions. There was also a significant difference in the fimctional properties of the

experimental fractions and SP1.

 

 

w
C
.L.

I.Vdue

‘

CPEII
%fifigfififififiigfi?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

M

Figure 19 Hunter Lab L-value (Lightness) of Cowpea and Navy Bean Protein
Fractions afier UF Processing

122

 

 

                               

...0
.22
0....2
000.2
.....2
0.52
05.2
0.00.7.
2.0.2
.....Z
...U
0....U
0.7.0
.....U
”......U

 

002.5 0

 

Hunter Lab a-Value (redness-greeness characteristics) of Cowpea and

Navy Bean Protein Fractions alter UF Processing

Figure 20

 

 

 

mmmz
0mmZ
.mmz
umz

 

l
00.00.30?

I
000.00%. ... .

| m “luv

..0a.§.§.

 

 

0000
0mmb
.000

Protein Mons

0mmb
0mmo
.mmo
0&0

 

 

-104
-151
-204
-253

 

Hunter Lab b-Values (yellowness-blueness characteristics) of Cowpea and

Navy Bean Protein Fractions afier UF Processing

Figure 21

123

The correlation matrix of the functional properties of cowpea, navy bean and
soybean protein fractions indicated that EC, OAC, GEL P, and ADHES were highly
correlated to %P of the fractions, while FC and WORK had mid-range correlations. All
had negative correlations to %P except FC. HARD and GUM were highly correlated to
all color values of the fractions, while chewiness had mid-range correlations. High to
mid—range correlations was found between WORK and the other TPA textural
characteristics, and mid to low-range correlations found with PC, EC and OAC. All of
these correlations were positive except FC, GEL P and COHES. OAC had high positive
and negative correlations respectively to EC and GEL P, while WAC had only a mid-
range negative correlation with ADHES. PS showed positive correlations with COHES
and B color value.

The correlation matrices of the fimctional properties of cowpea and navy bean
protein fractions grouped by bean type indicated that %P in CP fi'actions showed high to
mid-range correlations with all properties except OAC, while %P for NB had high
correlations with all properties. These results indicate that the protein content and hence
UF processing was important in functionality of the cowpea and navy bean protein
fractions.

Factor Analysis in Stat View (SAS Institute Inc.) was used to determine which of
the functional properties was most affected by ultrafiltration processing, that is, which
were the principal components. The factors were then transformed by a varimax
orthogonal rotation to achieve a more meaningfiil interpretation. The number of factors
retained was determined by those factors having eigenvalues equal to or greater than one,

and fimctional properties with loadings between 0.70 and 0.99 were considered major.

124

The number of variables were fifteen. and an estimated seven factors were
extracted. Of these, four had eigenvalues greater than one, and they accounted for about
95% of the variance observed. Table 33 indicated that CHEW, HARD, WORK and GUM
exhibited the highest loading in factor I, which accounted for 44% of the total variance
observed. These properties were all characteristics of the legume protein gels, thus factor
I can be called a "gelation" factor. The correlation matrix of the fiinctional properties of
cowpea and navy bean protein fractions indicated high positive correlations between
CHEW, HARD, WORK and GUM.

OAC, ADHES, GEL P, and Hunter "b" had the highest loading on factor II, which
accounted for 27% of the total variance observed. These variables represent a mixture of
several properties including hydration, gelation and color. However, the correlation
matrix indicated low correlations between hunter color characteristics and OAC or GEL
P. Thus Factor 11 can be characterized as ”hydration-gelation." Table 33 indicated that
the loading coefficients for OAC and ADHES were positive, while that for GEL P was
negative. This supported the high negative correlation observed between OAC and GEL
P, and ADHES and GEL P; and the high positive correlation observed for OAC and
ADHES.

WAC was the only variable that expressed a high loading coefficient in factor III,
which accounted for 12% of the variance observed. Since WAC is a hydration fimctional
property, factor [II can thus be characterized as a "hydration" factor. EC, FC and PS
showed the highest loading coefficients in Factor IV, which accounted for 11% of the
total variance observed. These variables have both surface property and hydration in

common, and could be characterized as a "surface property + hydration" factor.

125

In conclusion, the results of the factor analysis suggested that the four factors
extracted from the fiinctionality study were "gelation", "hydration + gelation",
"hydration" and "surface property + hydration.” Thus, gelation, hydration and surface
property characteristics can be considered as the principle fiinctional properties that are

affected by ultrafiltration processing of legume proteins.

 

Table 33 Factor Analysis Loading of 15 Functional Properties on the first four
f 5 ‘ _ g factors (eigenvalues > 1) after Varirnax rotation ‘_ .

Functionality Factor 1 Factor 2 Factor 3 Factor 4
FC - 0.593 -0.451 0.127 0.512
EC 0.290 0.595 0.305 mg
WAC -0.202 -0.366 9._80_9 0.266
OAC 0.386 9.171 0.400 0.246
PS 0396 0.370 0.504 M
GEL P -0.637 -_0_._7_5(_) -0.068 -0.108
WORK m 0.375 -0.069 0.151
HARD M9 -0.275 0.095 -0. 177
ADHES 0.506 m -0.285 -0.208
CHEW 9_.9_44_ -0.040 0.300 -0.086
GUM 9.1m -0.314 0.321 -0.245
COHES -9j_6_8 0.057 0.475 -0.268
Hunter ”L” 9.184 -0.442 0.186 -0.326
Hunter "a” 0.606 0131 0.069 0.231
Hunter ”b" -0.674 0.681 0.163 ~0.126
Eigenvalue 6.67 4. 10 l .76 l .66
Variance (%) 44.5 27.3 11.7 11.1
Cumulative 44.5 71.8 83.5 94.6

Yaflance (%) ._ .m m ...... ......
1 Loading values > 0.70 are underlined and considered significant

 

 

126

Table 34 summarizes the results of an assay for screening phytohemagglutinin

(Lectin) activity. It indicates that significant lectin activity was observed in flour and

extract fractions for all samples, but less so for isolates. High activities were also

observed in the residue and permeate 3 fractions. These results are consistent with UF

processing reducing lectin activity in the isolates. This was supported by the data

available on the doublet band size of phytohemagglutinin (32KD and 31KD) by Coffey,

(1983). In this study, permeates 2 and 3 corresponding to the UF membranes with

MWCO of 15-30KD and 50-100KD recorded the highest lectin activities, which suggests

that lectin permeated those UF membranes and accumulated in the permeates.

 

Table 34 Screening Cowpea and Navy Bean Protein Fractions for
Phytohaemaglutinins (Lectins) ‘
"Fractions Activity observed afler 1 hour
2 * Cowpea T T Navy Bea} ”' Soybean

Flour +++~+ ++++ +—+++
Extract +++ 44+ ---
Residue +-+-+ ++—+ ---
Permeate 1 + + ---
Permeate 2 + + ---
Permeate 3 ++ ++ ---
Isolate + H +
1 n= 3;

+ indicates the presence of a precipitate
number of + indicates degree of precipitation
--- indicates sample not analyzed

127

The data on in-vitro digestibility is shown in Table 35. It indicates that for both
cowpeas and navy beans, digestibility increased afier aqueous extraction of flour and UP
processing of the extract. Digestibility of cowpea fractions was higher than navy beans
for the two in-vitro methods tested, however digestibility of CPI and NBI were both
lower than that for SP1. This was expected since the unfolded state of SPI permitted
easier digestion. A highly significant difference in treatments was observed for both

methods (p < 0.001).

Table 35 [n-vitro Protein Digestibility of Cowpea and Navy Bean Fractions after
Ultrafiltration Processing

 

Fractions ' " 1511:5131” ‘ ’ " f ’7 f ’7 "611-15166

CPF * l 81.29 i 0.38 a,d,e,f 68.49 :- 1.49 a

CPPE 82.87 i 0.49 a,b 77.23 i 1.86 b,e

CPI 82.68 i 0.62 a,b,c 75.37 i 1.43 b,c

NBF 79.38 i 0.82 d 75.68 i 1.07 b,d
NBPE 79.37 i 0.98 d,e 79.03 i 0.89 e

NBI 81.69 i 0.40 b,e,f 80.99 i 0.82 e,f

SPI 9011:021 g 94.32: 1.14g

I n = 3 n

Means within the same column followed by different letters are significantly
different (p < 0.05)

Interpretation of the in-vitro digestibility for fractions that had extremely low
protein content (permeates) and possibly low digestibility was difficult by the pH-Stat
assay. The digestibilities indicated for cowpea and navy bean permeate and residue

fractions may have been overestimated using this method. None of these fractions

128

required any additional NaOH to maintain the pH at 7.98, hence based on the equation
used, digestibility was calculated as 79.28%. The equation however assumed that the

minimum digestibility would be 79.28%, but this may not be case with those samples.

There are limitations with using derived equations in determining in-vitro
digestibility. The 3 enzyme in-vitro method described by Hsu et al (1977) was reported to
underestimate protein digestibility particularly from animal sources (Hsu et al, 1978).
This was modified by Satterlee et al (1979) who introduced a four enzyme method.
however it was concluded that only approximate estimates of protein digestibility was
possible with these procedures (Bodwell et al, 1980). It was reported that the three
enzyme pH-Stat method provided more accurate estimates of digestibility because the
activity of the enzymes used was pH dependent. Since pH was kept constant during
digestion, this ensured more uniform enzyme activity and thus more accuracy than with

the pH Drop method (McDonough et al, 1990).

However, in a collaborative test by McDonough et al (1990) in which the pH-Stat
method was used, they reported average digestibilities of 99% for soy isolate, 97.6% for
pea concentrate, 93.8% for canned chick peas and 93.7% for canned pinto beans. These
digestibility values are extremely high for legume products. They are significantly higher
than what is reported in this study for cowpeas and navy beans; and in other studies of
legume protein digestibility (Rubio et al, 1991; Eicher and Satterlee, 1988; Rozan et al,
1997; Kohnhorst et al, 1990 and Hernandez et al, 1997). Despite the problems with using
both methods, they produced comparable results for estimates of protein digestibility.

The two in-vitro digestibility assays showed a high positive correlation (0.74), and mid-

129

range negative correlations with lectin activity, that is, pH Stat vs. lectin. r = — 0.66 and

pH Drop vs. lectin, r = — 0.64.

Conclusion

Results of this study indicated that UF processing was effective at separating
proteins on the basis of their size. The physico-chemical properties of cowpea and navy
bean proteins were significantly affected by UF processing, and there were significant
differences between the experimental fractions and SP1. Protein and amino acid
composition of both legume fi'actions increased significantly. The limiting amino acids in

the fractions were the sulfur amino acids, methionine and cysteine.

The fractions underwent denaturation from shearing during UF processing;
cowpea fractions appeared to be less denatured than navy bean, while SPI was more
denatured than both cowpea and navy bean fiactions. The denatured state of the protein
fractions affected protein functionality, particularly protein solubility of the fractions.
SPI, with the greatest unfolded structure, had the lowest protein solubility, which would
affect its use in commercial hydration applications. Protein digestibility determined by in-
vitro assays, significamly improved after UF processing, which was associated with

reduced lectin activity, and increased protein and amino acid content.

The high PS of CPI and NBI indicate that these fractions would have better
hydration properties than SP1, and would be usefirl in liquid applications. Additionally,
the high WAC of CPI compared to SPI makes this fi'action better than SP1 for liquid

products. The high FC and FS of NBI makes it better than SPI in whipped products such

130

as cakes, desserts and ice cream. OAC and EC of CPl and NBI were higher than SP1. and
ES highest in NBI, these make the experimental isolates, particularly NBI, better in
simulated meat products, cakes and dressings than SPI. Critical protein required for
gelation was highest for SP1, which results in the experimental isolates being better in

protein gels such as meats, cakes and cheese products.

Therefore, the null hypothesis (Ho), stated as there are no significant differences
between the physico-chemical properties of cowpea and navy bean protein fractions
compared to a commercial SPI after UF processing, must be rejected. The conclusion
from this study is that UF processing influences the physico-chemical properties of

proteins.

Future Resear__c_l_1

The goals for future study include the following:

1. Optimize protein extraction and improve membrane performance by introducing a
continuous slurry centrifiigation operation and examine its effect on protein yield

2. Further research on the structure of individual proteins in cowpeas and navy beans
is necessary to explain the differences observed in flux measurements between the
two during UF processing

3. Expand characterization studies on permeates to determine its composition and

utilization in commercial applications

131

STUDY3 NUTRITIONAL QUALITY OF COWPEA AND NAVY BEAN
PROTEIN DIETS BY IN VITRO AND IN VIVO PROTEIN
DIGESTIBILITY CORRECTED AMINO ACID SCORE (PDCAAS)

m

The nutritional value of the insoluble residue that remains after aqueous alkali
extraction of legume flours requires additional research effort to assess its potential as a
food ingredient, particularly in countries where protein-energy malnutrition persists. The
protein digestibility of cowpea (CP) and navy bean (NB) residue diets were studied by in
vivo and in vitro assays. The legume residues were cooked then air-dried prior to diet
preparation. All diets contained 10% protein (w/w). Legume proteins supplied 30%, 70%
and 100% of the protein; the remainder was derived from wheat flour. A modified
American Institute of Nutrition (AIN) rodent diet AIN-93G was used as the control and a
2% albumin diet was used to determine metabolic fecal nitrogen. Food intake and body
weight of the rats was measured and apparent and true protein digestibilities calculated for
the in vivo study. The pH stat and pH drop in vitro assays were also used to determine
protein digestibility. Amino acid composition for all diets were measured and in-vivo and
in-vitro protein digestibility corrected amino acid scores (PDCAAS) calculated. The food
intake of all CP diets was greater than that observed for the control, while that for all NB
diets was less. Intake of the 30% and 70% CP and NB diets was not significantly difi‘erent
fiom each other. The 100% CP and NB diets were consumed the least. All diets supported
rat growth except 100% NB. All rats fed CP diets had significantly higher weights than the

controL unlike rats fed NB diets which all had lower weights. In-vivo digestrbility ranged

132

from 73.7% - 87.5% and 62.6% - 78.2% for CP and NB diets respectively, compared to
98.1% for the control. All diets had significantly difl'erent digestibilities than the control
except 30% CP. The 70% CP, 100% CP and NB diets were not significantly diflerent
fiom each other. Significant correlation’s (p < 0.05) were 0.73 for pH stat vs pH drop,
0.86 for pH stat vs in viva and 0.89 for pH drop vs in viva. Amino acids that are generally
limiting in legumes increased in concentration after supplementation with wheat flour. The
100%CP, 100%NB and 30%NB diets were limiting in sulfur amino acids and lysine
respectively, and were the only diets that did not meet the suggested pattern of
requirement for pre-school children (2-5yrs). Lysine (Lys) was lowest in the 30%CP and
70%NB diets; methionine + cysteine (Met + Cys) was lowest in 70%CP; and threonine
(Thr) was lowest in the modified AIN-93G. In-vivo and in-vitra assays for PDCAAS was
based on these limiting amino acids, and were highly correlated (r = 0.94, r = 0.97 and r =
0.98, p < 0.01). These results suggest that in-vr'va and in-vitra PDCAAS could accurately
predict protein nutritional quality. Additionally, the insoluble CP and NB residues could
not be recommended nutritionally as food for pre-school children, but could be

recommended if they are supplemented with wheat flour.

133

Introduction

The quality of a protein depends on the kinds and amounts of amino acids it
contains, and represents a measure of the efficiency with which the body can utilize the
protein. A balanced or high quality protein contains essential amino acids in sufficient
ratios for human needs. Proteins of animal origin generally tend to be of higher quality
than those of plant origin. Amino acids present in dietary proteins are not necessarily fully
“available” since digestion of the protein or absorption of the amino acids may be
incomplete. Amino acids from animal foods are absorbed to an extent of 90%, while those
from plant foods are digested and absorbed to an extent of only 60-70%. This has been
attributed to the protein conformation; binding to metals, lipids, cellulose and other
polysaccharides; the presence of antinutritional factors such as trypsin inhibitors and
lectins; as well as surface area and size of the protein, processing efiects and biological

differences amongst individuals (FAO, 1970, 1973).

The nutritional quality of legume proteins has been extensively reported in the
literature. Bressani (1973) showed that most legume proteins had low biological value,
ranging fi'om 32-78%, because of the low concentration of sulfur amino acids in legume
protein. Kelly (1973) showed the beneficial efiects of the addition of methionine in the diet
when legumes are used as the protein source; he found that both the protein efficiency
ratio (PER), and the average weight increased. However this was not evident for all
legume species, and was explained on the basis that methionine is not the most or the only
limiting amino acid in legume species. When both methionine and tryptophan were added,

protein quality increased which indicated that both are limiting.

134

More recently, significant research has begun to be expended on other leguminous
species, as their desirability as less expensive protein sources becomes obvious,
particularly to nations where protein-energy malnutrition was prevalent. Ulloa et a1 (1988)
obtained a protein concentrate from chickpea (Cicer arietinum) by UF membranes and
used it in infant formulas. Bolles (1997), reported on the nutritional quality of a navy bean
retentate used in an aseptically processed infant product and found that it was of similar

quality and functionality as a soy protein isolate.

However, there has been very little work reported on the nutritional quality of the
residue that remained after aqueous alkali extraction, and its use as a food ingredient.
Preliminary results from this work have indicated that the protein content of the residue
ranges between 18 - 25% after extraction, and about 35% of the total available protein
remained in the residue (Jackson et al, 1997). Additionally, sodium dodecyl sulfate
polyacrylamide gel electrophoresis (SDS PAGE) analysis fi'om an earlier study indicated
similarity in protein bands between the residue and flour. Both flour and residue also
showed similar amino acid profiles. Opportunities therefore exist to evaluate the
applicability of the residue as a food ingredient. Six diets that contained only cowpea
residue and navy bean residue, 3 mixture of wheat and the cowpea or navy bean residue,
and two other diets containing 2% albumin and a control modified AIN-93G diet were
used. All diets contained about 10% protein and arrowroot starch as their carbohydrate

source.

The objectives of this study were 1) to determine the efi‘ect of diet type on food

consumption and rat growth, and 2) to estimate the protein quality of the diet blends.

135

The null hypotheses (Ho) being tested is there are no significant differences in

protein quality of the experimental diets compared to the control modified AIN-93G diet.

MLeriais & Methods

The raw residue that remained after aqueous alkali extraction of cowpea and navy
bean flour was cooked in kettles to inactivate antinutritional factors, then air-dried for 48
hours at 45°C. They were stored at 4°C prior to use.

Experimental diets were prepared based on a 10% protein diet, using a modified
AIN-93G diet as the control. Actual protein content of each diet was determined by the
micro-kjeldahl method (AACC, 1984), using a nitrogen to protein factor of 5.7 for the
legume diets and 6.25 for the modified AIN-93G and 2%ALB diets. Proteins were
hydrolyzed with HCl to determine total amino acid composition of each diet, except
tryptophan and cysteine. Methane sulfonic acid (MSA) hydrolysis was used for tryptophan
analysis and formic acid oxidation followed by acid hydrolysis for cysteine analysis.

In-vitra digestibilities of the experimental diets were measured using the three-
enzyme pH Stat assay (McDonough et al, 1990), and the four-enzyme pH Drop assay
(AOAC 1990). Digestflailities of the diets were calculated using the following equations:
pH Drop % Protein Digestibility = 234.84 - 22.56 (X), where X = pH at 20 min
pH Stat TD = 76.14 + 47.77B, where B = ml 0.1N NaOH added

Twenty weanling male Sprague-Dawley rats (Harlan Sprague Dawley Inc., MI)
were used in the in-viva feeding study over a period of 4 weeks. The rats were ranked by

weight and assigned diets so that none of them received the same diet twice. Each diet

136

was fed to groups of ten rats in a completely randomized design. Rats were housed in
individual cages and given free access to diets and water for 7 days. The animals were
weighed daily, food intake was measured, and spilled food and fecal matter collected, air-
dried and weighed between days 3 and 7. Fecal samples were ground in a mortar and
stored at 4°C prior to analyses. Apparent and true digestibilities were calculated.

Protein quality of the experimental diets was estimated using relative protein
efficiency ratio (RPER), relative net protein ratio (RNPR), relative true protein
digestibility (RTD), and in-viva and in-vitra protein digestibility corrected amino acid

score (PDCAAS).

Statistical Analfls

Microsoft Excel was used to compute the averages and standard deviations of the
variables. Stat View (SAS Institute Inc., 1998) was used to analyze the food intake, rat
growth and digestibility data, and determine difl‘erences between diets using F isher's LSD.

Correlation matrices of protein nutritional quality tests for the diets were generated.

137

Results and Discussion

The composition of a two-kg sample of each experimental diet is shown in Table
36. The protein contents of the whole-wheat flour, navy bean and cowpea residues were
11.03%, 15.25% and 14.27% respectively. Diets 1 - 6 contained a combination of 30%,
70%, 100% navy bean or cowpea protein supplemented with 70%, 30%, 0% wheat flour
respectively. Diet 7 was 2% albumin to estimate metabolic fecal nitrogen and diet 8 was
the control modified AIN-93G diet with casein as the protein source. The major
carbohydrate source for all diets was arrowroot starch, which ranged in weight from 0.01
- 71% of the total diet weight.

Table 37 shows the actual protein content and standard deviation of each diet.
Protein content ranged fi'om 9.08 —- 11.78% indicating that none of the samples contained
10% protein. The protein contents were within the range of values reported by Sarwar
(1997) for use in in-vr'va assays. CP diets had higher protein contents than the NB diets,
100% CP had the highest protein content of 11.78% while 30% NB had the lowest
protein content of 9.08%. The actual protein content of the diets were highly significantly
different (p < 0.0001) from each other. All of the NB diets were not significantly different
fiom the each other, however, only 30%CP and 100%CP diets were not different. The
70%CP, 30%NB, 70%NB, and 100%NB were not significantly different fi'om the control,

modified AIN-93G diet.

138

ooov:

vwdem
00.0
00.0.
00.004
0003
«N53
wwdmv

05.0-5-2

..:000N.:

94.03; 0_.~5N
«.00 no.0
00.m 00.m
00.0w 00.0w
00.05 00.05
$.00N Muno—

vo.3~ v0.3m
00.N

Vm. 53

micsm 3°88.

-_oso~ .

5.0va
5.0.0
00.m
00.0w
00.05
mmdm
v0.3m

5.0.22

mzxoo _. _.

30

8.0

006
00.0w
00.05

«.065

0v.0~v

5N.00N_

x 0.98

ova
8.0
006
00.0N
00.05
50.04
v0.3m

Simon

2002

mZSOm ,

.--;-sm¢m

newn—
no.0
00.m
00.0w
00.05
mmd
v0.3a

00.50

50. gm

. .w 608.05

mm.50~
no.0
00.m
00.0.5
00.05
00.0N
v0.3m

5.0.05
5%an

mzxe. .

.J ITO? Al‘ 111......1

5280800 85 5205 50m .062 05 80300

8238?

.530

mo

58 £833
is 0:05.:
0050

:0 88
£83
EBB—m
MED

052

”:53

25:60qu

0m 033.

gawk.

139

Table 37 Actual protein content and standard deviation of the experimental and
modified AIN-93G diets '

 

 

 

Diet Protein content (%)
30% CP 11.66 i: 0.35 a
70% CP 9.63 i 0.45 b,h
100% CP 11.78 i 0.47 a,c
30% NB 9.08 d: 0.58 b,d,h
70% NB 9.76 i 0.24 b,d,e,h
100% NB 9.42 i 0.29 b,d,e,f,h
2% Albumin 2.35 i 0.37 g
AIN-93G 9.26 i 0.27 h

ln = 3

means with the same letters are not significantly different (p < 0.0001)

Chrornatograms of the acid hydrolyzed amino acid profile of each diet indicated
that the peaks of all diets were well resolved and showed a similar profile as CPF and NBF
fi'actions. The average peak areas and retention times for amino acids in the experimental

diets indicated similar retention times to the amino acid standard solution (Sigma A9781).

The amino acid composition of the residues and diets is shown in Table 38. The
cooked residues had lower concentrations of amino acids than the uncooked residue. The
lower concentration was due to the processing treatment that the cowpea and navy bean
residues received during cooking and drying prior to diet preparation. The amino acid

content of the experimental diets were significantly different fi’om each other (p < 0.05).

140

mix—80.3 <m2 .3 $00.98 9.5 N

mam—80.3 Eon 0:0 500008 00:08.0 .3 mix—«:0 who _

 

 

 

2.2 a; 8.0 80 8.2 and 2.0 m3
004 _ 0m.~ 05.0 0.0 94.— 3.0 00.0 NEH
50.0. 504.0 00.5 00.0 00$ 00.3 3.» BE
3.: 5.0— m _ .0 3.2 00.2 00.0. _5.0 :04
00.0 3.0 0.0 «0.0 v0.» 000— 00.0 0:
_~.5_ 50.N 0.0 50.0. 50.. 50.N 00.04 .900
86 av; 00." 05.0 and 0nd 00.0. 82
3.: 0N5 00.0 00.5 050— 00.: 00.0 10>
3.0 mw.~ 00d 00.0 0~.v 0m.0 ~00 EC.
5.3 5.04 2.0 00.0 00.0 00.0. 50.0 0.5

m 0 .0 00.0 00.0 OZ 0N0 2V.— 000 «2
5.5 00.0 00.04 and 3.0 50.2 05.5 :3.
NE 0: 2.0 ~00 MS: 2:: 3.0 mi
3.0 00.0 000 0 _ .0 0044 00.: V0 .0 m5
:4 a; 8.0 m2 m3 3.: o: :0
3.0 00.5 N05 0 0.5 0.0 050— 05.0 :om
2.00 0%.: 00.3 3.00 mwdm 0~.0v 50.nm :5

1 00d: 00.2 00.0 00.3 ~5.0~ 05.2 00.2 0.64
00.0-22-2 m200_ m205 9400 0000— 0005 0000 20¢. 80:2
A5380 0 \ Eon 05:8 0:0 805 :95 .902 05 89569 00 5280830 20¢. 65:3 00 030.5

141

All diets were richest in glutamic and aspartic acids. The concentrations of
cysteine, methionine and tryptophan, the limiting amino acids in legumes, increased after
complementation with wheat protein in the diets. As expected, diets that had the highest
wheat protein, that is, 30% and 70% diets had the lowest levels of lysine than the other
diets. Similarly, diets that had the highest legume protein, that is, 100%CP and 100%NB
diets had lower levels of sulfur amino acids and tryptophan than those with greater wheat
protein. The modified AIN-93G diet was significantly difl‘erent from all experimental diets

and had high amounts of the amino acids that were limiting in the legume diets.

Table 39 shows the amino acid score of the experimental diets based on the
reference amino acid pattern for pre-school children (2 —- 5 yr.) (FAQ/WHO, 1990). It
indicates that only the 100%diets and 30%NB did not satisfy the reference amino acid
pattern for pre-school children. Met + Cys were the limiting amino acids in the 100% diets
and Lys in the 30%NB diet. All of the other diets satisfied the amino acid requirements for
pre-school children. This data is consistent with what has been reported in the literature
for lysine and the sulfur amino acids, as the limiting amino acids in wheat and legume

protein sources respectively (FAQ/WHO, 1973).

Phytohenmgglutinin (Lectin) activity of the cooked and uncooked residues and
each diet is shown in Table 40. It indicates that lectin activity was only observed in the
raw cowpea and navy bean residue samples, CPR and NBR respectively. There was no
detectable lectin activity in the cooked samples after receiving heat treatments during

cooking and air-drying. Similarly, none of the experimental diets or the control diet

142

208 Ron 05:8 0832 2: 88:00 30523:: N
E m I 3 8.220 69.8.»... 8. 3882.32 ...8 ages... 83.2. 8.30 32050355. 2.. .5 Baum.

 

 

Sm 3.. 0: SN .3 N. .m 2.. .a>
....N a... on. s... 8.. 03 an. .3
...m Ba ...... :.~ ”3. Sam 3.. 0..
..X ..Z 8.. :2 SN $0 $0 m...
9.0 .2 mm... 3.. aé SN R. 5 + 2...
fl 3... 3.. ..Q N: .2 m..~ 5
SN 8.. 8|. ad 3.. SN 3.. m3
«3 «3 ..3 $0 3.. SN 02 a:
,, ..3 Mad 3.. 8... fl... Nami. 3.. as. + m8
03.22.: 028. sz mzom 88. “.85 82 23.. 8.5.

 

.305 :mom 052 05 30260 80 980m Eo< oEE< 33:85

00 030.0

143

displayed any lectin activity. Occena (1994), Coffey et al (1992) and Dhurandar and
Chang (1990) also reported on lectin activity. They found that wet heat was most effective
in inactivating lectins, and a heat treatment of 100°C for 10 min was sufficient to inactivate

all lectin activity in navy beans.

Table 40 Relative Lectin Activity of Residues and Experimental Diets '

 

Diets Relative Lectin Activity

 

CP

 

 

30% ND
70% ND
100% ND
Raw Residue +++

algae a

 

' ND - not detectable

Cooked Residue ND
Activity in modified AIN-93G diet had

ND

In-vitro protein digestibilities determined by pH Stat and pH Drop methods are
shown in Table 41. The average values for CP diets ranged from 80.30% - 84.54% and
77.82% - 79.93% for the pH stat and pH drop methods respectively. The average values
for NB diets ranged from 79.28% - 79.59% and 66.51% - 74.89% for the pH stat and pH
drop methods respectively. All of the CP diets had higher digestibilities than the NB diets
using both in-vitro assays. A similar trend was observed for the digestibilities obtained by

the two methods in all samples. As legume protein concentration in the diet increased,

144

digestibility decreased for all diets. That is, the 30% CP and 30% NB diets had the highest

digestibility of each of their groups with both assays.

Statistical data (Table 41) indicated that there was a highly significant diflerence (p
< 0.0001) in digestibility between the diets using both assays. Fisher's LSD indicated that
all of the experimental diets were significantly different fi'om the control AIN-93G diet.
The digestibility of the navy bean diets did not appear to be significantly different fi'om
each other, however the digestibility of the 30% and 70%CP diets appeared to be more
similar than the 100%CP diet. This suggests that supplementation with wheat flour had a
significant effect on in-vitro digestibility, and is related to the increase in essential amino

acid composition, particularly the sulfur amino acids, that were limiting in the residue.

The improvement in in-vitro protein digestibility of cereal supplemented legume
based diets was also reported by Dominguez et al (1993) for 70% press-dried millet and
30% press-dried cowpea diet blend. Wolzak et al, (1981) reported on the in-vitro protein
digestibility of various cereal-legume blends. They found that a 70:30 ratio of cereal to
legume, and rice rather than maize supplemented with black beans gave higher in-vitro
protein digestibilities. This was possibly due to the range of amino acids present with these

combination of foods.

Table 41 also shows that R = 0.98 for both pH Stat and pH Drop, which indicates
that there was a good fit between diet type and in-vitro protein digestibility for both
assays. The values of 0.97 and 0.96 for R2 of pH Stat and pH Drop respectively, indicated
that 97% and 96% of the variance in protein digestibility was predictable from the model.

Although lower values were observed for protein digestibility using the pH Drop method,

145

there was a highly significant correlation, r = 0.78 (p < 0.0001), between the two methods.
This significant correlation for pH Drop and pH Stat methods was also reported by El and

Kavas (1996) also observed for digestibility of rainbow trout.

Table 41 In-vitro Digestibility ‘ and Statistical Analyses 2 of Diets

 

 

 

 

 

Diets In-vitro Digestibility
pH-Stat pH-Drop
30%CP 84.54 :t 0.83 a 79.93 i 0.42 a
70%CP 82.54 i 0.87 a,b 77.97 i 0.84 a,b
100%CP 80.30 :1: 0.28 c 77.82 i 0.04 a,b,c
30%NB 79.59 i 0.32 c,d 74.89 i: 0.94 b,c,d
70%NB 79.55 i 0.38 c,d,e 68.77 i 0.31 e
100%NB 79.28 i 0.04 c,d,e,f 66.51 :1: 0.11 e,f
Modified AIN-93G 93.72 i 0.71 g 84.92 i 0.74 g
Statistics
F Ratio 51.70 28.02
R 0.98 0.98
R2 0.97 0.96
‘ n = 3
2 Means within a column followed by the same letters are not significantly
different (p < 0.0001)

pH Stat vs. pH Drop, r = 0.78

146

The raw data on food intake for each rat during the four feeding indicates that
there were fluctuations in the quantity of food ingested by each rat, which is related to the
type of diet ingested. The average food intake and standard deviation for each diet over
the feeding period is indicated in Table 42. It ranged from 27.1 g for 100%NB diet to 75.6
g for 30%CP diet. The food intake of all CP diets was greater than that observed for all
NB diets and the modified AIN-93G diet. However, food intake for all NB diets was less
than the modified AIN-93G diet. The food intake of the 2% Albumin (2%Alb) diet was

less than all diets except 100% NB.

Regression was used to investigate the effect of diet on food intake. Diet had a
highly significant efl‘ect (p < 0.0001) on food intake, and both diet and food intake were
highly correlated, R = 0.86 and R2 = 0.74. Fisher's LSD indicated that the food intake of
all experimental diets were significantly different from the modified AIN-93G diet except
100%CP and 70%NB. Food intake of the 30% and 70% diets were not significantly
different from each other for both CP and NB diets. Food intake of 100%NB and 2%Alb
were significantly different fi'om all other diets.

The data on rat weight for each group indicates that the final weight range for
group 1 rats is 87 -142 g, group 2 is 94 - 174.8 g, group 3 is 87 — 175 g and group 4 is
118.4 — 189.3 g. The rat growth according to diet is shown in Table 42. It indicates that
all CP diets, 30%NB, 70%NB and modified AIN-93G supported rat growth. As expected,
the rats fed 2%Alb decreased in weight, as well as those fed 100%NB diet.

The data suggests that when NB or CF residue was the major source of protein,
the growth of rats was impaired. This was due primarily to the reduced quantity of

essential amino acids in these diets. A regression analysis indicates that diet type had a

147

highly significant effect (p < 0.0001) on rat growth, and diet and rat growth were highly
correlated, R = 0.89 and R2 = 0.79. The change in rat grth and food intake with diet is
presented in Figure 22 and shows that when food intake decreased, rat growth also
decreased.

Regression analysis of the effect of diet type, food intake and their interaction on
rat growth indicates that both diet and food intake had a significant effect (p < 0.02 and p
< 0.0001 respectively), while the interaction had no effect on rat growth. The multiple
correlation coefficient R = 0.93, R2 = 0.86 and the lack of fit is not significant. This

suggests that the model is able to adequately predict rat growth.

Table 42 Average Food Intake1 and Standard Deviation of the Experimental Diets
and Rat Growth2 over the Feeding Period

 

 

Diets Food intake (g) Rat Growth (g)

30 CP 75.65 1 12.37 a 42.07

70 CP 69.44 i 8.86 a,b 39.09

100 CP 50.11 :1: 9.00 c 18.95

30 NB 46.21 i 11.06 c,d 8.46

70 NB 45.01 0 11.27 c,d,e 3.81

100 NB 27.06 :1: 7.60 f - 11.76

2 ALB 36.7 i 4.24 g - 10.77
M-AlN-93G 48.03 1: 10.85 c,d,e,h 11.91

 

‘ F = 29.07 p < 0.0001 R = 0.86 R2 = 0.74
2 F = 28.78 p < 0.0001 R = 0.89 R2 = 0.79

148

02.50 @030 20 .83 5380 SM 05 305 03:258qu 00 8.8:— 0oom 00803.:

a. 05w...

 

 

 

a. 23:. .38... n.
a. 5380 3. n

 

 

 

 

305 035.5395

1 00%NB
70%CP
30%CP

 

 

 

 

 

 

 

 

(3) 1m

 

 

149

The data on fecal matter for each diet is shown in Table 43, and indicates that fecal
weight and fecal protein content was higher for all experimental diets compared to the
modified AIN-93G diet. Fecal protein content was highest in the 100% diets, while fecal
weight was highest in the 70% diets. The higher excretion of nitrogen is likely to originate
in part fiom microbial growth in the large intestine and reflects lower protein digestibility
of that diet. The effect of food intake, diet and their interaction on fecal protein content
indicates that only food intake had a highly significant effect on fecal protein (p < 0.0001),

R = 0.81 and R2 = 0.65.

Table 43 Protein content and weight of fecal matter during feeding study

 

 

 

Diet Protein content (% db) Wt (g)

30CP 21.54 i 2.24 a 8.76 i 1.67 a

70 CP 22.992t0.85 a,b 11.15i2.12b

100 CP 26.16 i 1.49 b,c 8.89 :1: 1.69 a,c

30 NB 20.33 i 1.14 a,b,d 8.25 :t 0.87 a,c,d

70 NB 18.64 i 2.51 a,d,e 10.48 i 1.42 a,b,c,e
100 NB 24.12 i: 0.98 a,b,c,d,f 7.59 :l: 2.85 a,c,d,f
2% ALB 10.40 i 1.22 g 7.69 i 1.25 a,c,d,f,g
Modified AIN-93G 10.78 :1: 2.46 g,h 7.95 :1: 2.02 a,c,d,f,g,h
' n = 3

2 Means within a colurrm followed by the same letters are not significantly
different (p < 0.0001)

150

The average apparent (AD) and true digestibility (TD) are shown in Table 44 and
indicates that TD ranges fi'om 62.6% for 100%NB to 98.1% for the modified AIN-93G
diet. The average values for CP diets ranged from 73.7 % - 87.5% and 62.6% - 78.2% for
NB diets. All of the CP diets had higher digestibilities than the NB diets, and the 30%CP
and 30%NB diets contained the highest TD fi'om each of their groups. TD of 30%CP diet
was not significantly different from the modified AIN-93G diet; all others were
significantly different. The effect of diet and food intake on TD indicates that diet and the

interaction of food intake’diet had a highly significant efl‘ect (p < 0.0001).

Table 44 Average in-vivo Apparent Protein Digestibility (AD) and True Protein
Digestibility (TD) and Standard Deviation of the experimental Diets

 

 

Diets AD TD

30 CP 76.56 i: 5.03 a 87.49 i 5.32 a

70 CP 61.42 :t 8.41 b 73.74 i: 8.79 b

100 CP 64.13 i: 8.53 c 74.61 i 9.28 c

30 NB 49.48 1: 12.09 d 78.23 :1: 9.30 d

70 NB 64.09 i 8.57 e 69.39 1 19.56 c

100NB 17.21 :1: 15.12f 62.64i21.61 f
Modified AIN-93G 75.65 i 9.03 g 98.15 i: 9.41a,g

 

'n = 8
2 Means within a column followed by the same letters are not significantly
difi‘erent (p < 0.0001)

151

Estimates of the protein nutritional quality (RPER, RNPR, RTD and PDCAAS) of
the experimental diets are shown in Table 45. It indicates that the 30% diets generally had
the highest protein quality while the 100% diets had the lowest quality. The 100%NB diet
was the only diet that had zero protein quality using PER and NPR because this diet did
not support rat growth. The PDCAAS values (based on human requirements) were
highest for the 70% diets, values over 100% were observed for the 70%CP diet and
modified AIN-93G. In the methods that assess protein quality based on growth, the RNPR
values of experimental diets were higher than the RPER values because the RNPR method

also credits protein used for both growth and maintenance.

Table 45 Estimates of Protein Nutritional Quality of Cowpea and Navy Bean Diets
fiom Relative Protein Efficiency Ratio (RPER) ', Relative Net Protein
Ratio (RNPR) 2, Relative True Protein Digestibility (RTD)3, and Protein
Digestibility Corrected Amino Acid Score (PDCAAS)‘

 

Diet RPER RNPR RTD PDCAAS PDCAAS PDCAAS
in-vivo pH-Stat pH-Drop

 

30CP 138 339 89.14 89.24 84.19 81.22
70CP 144 329 75.13 95.12 106.48 99.81
lOOCP 81 1 13 76.01 58.20 62.79 60.70
30NB 103 121 79.70 74.32 75.32 71.15
7ONB 73 39 70.69 74.25 86.57 71.44
100NB 0 O 63.82 38.84 49.15 40.00

 

' RPER = PER oftest diet / PER of modified AIN-93G

2 RNPR = NPR oftest diet / NPR of modified AIN-930

3 RTD = TD of test protein / TD of modified AIN-93G

4 PDCAAS of test protein fiom in-vivo and in-vitro assays

152

The growth tests, RPER and RNPR, showed high significant correlations with
each other (r = 0.90). The scoring methods correlated highest with each other, in-vivo
PDCAAS vs. pH-Stat PDCAAS, r = 0.94; in-vivo PDCAAS vs. pH-Drop PDCAAS, r =
0.97; and pH-Stat PDCAAS vs. pH-Drop PDCAAS, r = 0.97). The lowest correlations
were found between the scoring methods and RTD, pH-Stat PDCAAS and RTD, r = 0.41;
pH-Drop PDCAAS and RTD, r = 0.57; and in-vivo PDCAAS vs. RTD, r = 0.69. The
growth tests showed medium to high correlations with the scoring tests.

Harris et al (1988) reported significant correlations (r = 0.90) between NPR and
PER, and Pellett and Young (1981) and the Codex Alimentarius Commission (1984)
reported that RNPR was the most appropriate rat test for routine assessment of protein
quality of foods for humans. They found that NPR method gave similar results to net
protein utilization (NPU) and biological value (BV).

Sarwar (1997) reported that PDCAAS overestimated the protein quality of protein
sources that contained antinutritional factors, particularly when the scores were calculated
based on human requirements. He found that mustard flour and black beans had a
PDCAAS of 84% and 45% respectively compared to RPER and RNPR of 0. This was
also observed for 100%NB diet in this study, with PDCAAS of 49%, 40% and 39% by
pH-Stat, pH-Drop and in-vivo assays respectively, compared to 0 for RPER and RNPR.
He suggested that vegetable protein sources contain growth-depressing factors such as
isothiocyanates in mustard flour, trypsin inhibitors and lectins in black beans and
lysinoalanine in alkaline-treated soy isolate. Sarwar (1997), Eggum et al (1989) and
Sarwar and Peace (1986) reported that the PDCAAS method also does not take into

account the bioavailability of individual amino acids, which may be up to 40% lower than

153

the overall digestibility of protein in the same food. They concluded that the PDCAAS
method may give misleading results about the quality of proteins co-limiting in more than

one essential amino acid.

Rubio et al (1998) reported on the nutritional utilization of chickpea and its
isolated globulin proteins by rats. As expected, they found that growth of rats was
impaired when chickpea meal (uncooked) was included as the sole source of protein in the
diet, and was due to interference with systemic protein metabolism. They also added that
the lower protein efficiency observed was not due to the presence of insoluble residue
(starch and fiber) in the diet, since the presence of fiber in similar quantities in the control

diet had no detrimental effect on performance of rats or the NPU values.

Supplementation of cooked bean diets with limiting amino acids has been
extensively reported to improve protein quality. Kakade and Evans, (1965) found that
autoclaved navy beans supplemented with either methionine alone or with all the limiting
essential amino acids had a similar PER as casein. Bressani et al, (1963) reported an
improved an improved PER and BV for black beans when supplemented with 0.2%
methionine. Sarwar (1997) reported on the beneficial effects of supplementation with
limiting amino acids. He suggested that the PDCAAS method assumes complete biological
efiiciency of the supplemented amino acids. In his study on amino acid supplemented zein,
a protein of low digestibility and poor quality, he found a marked difference between the
PDCAAS and RPER or RNPR of amino acid supplemented zein. Supplementation of

cooked beans with other protein sources including corn, wheat, rice and oats, has also

154

been reported in the literature. Yadav and Liener (1977) reported similar PER values for
navy bean supplemented with various cereals compared to the PER for casein.

There is much interest in determining the optimum nutritional ratio of consumption
between the cereal and legume protein sources. Bressani and Elias (1974) have established
this ratio as 2.6:], which corresponds to a diet constituted by 72 parts corn and 28 parts
bean. Arroyave (1973) reported that children fed a corn : bean diet, in a 70:30 ratio by
weight were able to meet their protein and calorie needs. However, Murillo et a1 (1974)
suggested that this ratio was too bulky for the weight and size of laboratory animals due
to their gastric capacity. Bressani et al (1980) however found that a 70:30 corn : bean diet
increases the consumption of the diet mixture in 5 week old pigs fed ad libitum. This
supported the results of this study that the 30% legume diets had greater digestibilities

compared to the other experimental diets.

155

Conclusions

Results of this study indicated that the amino acid content of cowpea and navy
bean residues decreased after heat processing. The protein quality of cowpea and navy
bean diets was significantly lower than the modified AIN-93G control diet. All cowpea
diets had higher protein quality than the navy bean diets. Cowpea or navy bean as the sole
source of protein in the diet did not provide the pattern of essential amino acid
requirements for pre-school children. Supplementation with wheat flour improved the
protein quality of both legume diets. Lysine was the limiting amino acid for 30%CP,
30%NB and 70%NB diets, while the sulfur-amino acids were limiting in 70%CP,

100%CP, and 100%NB diets.

Ho: there are no significant difierences between the protein quality of cowpea and

navy bean protein diets compared to a modified AIN-93G diet.

Reject the H0 as stated and conclude that diet composition significantly affected

the protein quality.

Future Research

The goals for fiiture study include the following:

1. To determine the bioavailability of the limiting amino acids lysine and methionine
in each diet so that a more accurate prediction of protein quality could be

determined.

2. Research on the proportion of the total sulfitr amino acid requirement which can

be met by cystine for protein sources that have low cystine levels

156

SUMMARY AND RECOMMENDATIONS

There is considerable interest for increasing utilization of legumes as a protein
source, particularly in countries where protein-energy malnutrition persists. The legumes
chosen for study in this dissertation are of economic importance in the United States and
a number of developing countries in Afi'ica. Cowpeas are one of the major legumes
consumed in the developing countries of Africa, particularly in West Africa, and navy
beans are a major crop in the state of Michigan. Ultrafiltration processing has been
increasingly utilized in the food industry to separate macromolecules such as proteins
without adverse effects to chemical structure. Most of the research has been focused on
dairy proteins, and increasingly on soybean proteins. There has been very little published
on ultrafiltration processing of legumes other than soybeans. This research attempted to
provide information on ultrafiltration processing of cowpea and navy beans and
compositional, functional and nutritional characterization of the proteins fractions.

The results indicated the feasibility and potential for a number of value-added
products from ultrafiltration processing of legumes. Aqueous extraction produced a high
protein aqueous extract (protein concentrate) which was used for further ultrafiltration
processing to produce legume protein isolates. These isolates, although different from a
commercial soy protein isolate, had favorable functional and nutritional properties and
could be utilized as ingredients in the food industry.

The high protein solubility of the isolates would be useful properties in liquid
applications. Additionally, the high water absorption capacity of cowpea protein isolate

also makes this fraction acceptable for liquid products. The high foam capacity and

157

stability of the navy bean isolate makes it acceptable for whipped products such as cakes,
desserts and ice cream. Oil absorption capacity and emulsion capacity of the isolates,
particularly navy bean isolate, suggests that they can be utilized in simulated meat
products, cakes and dressings. The critical protein required for gelation was highest for
soy protein isolate, which results in the experimental isolates being more acceptable for
protein gels such as meats, cakes and cheese products. Protein digestibility of the isolates,
determined by in-vitro assays, improved after ultrafiltration processing. This was

associated with reduced lectin activity and increased amino acid content of the isolates.

The residue that remained after aqueous extraction is generally considered a waste
by-product of the ultrafiltration process, and is either fed to animals or used as a soil
amendment in organic farming. However, the results reported indicated that the residue
had up to 40% of the total protein remaining after aqueous alkali extraction, and therefore
had potential as human food. When the legume residue was combined in diets
supplemented with wheat flour, it was found that diets with legume as the primary source
of protein did not meet the nutritional requirements for pre-school children. However, a
30% cowpea-wheat flour diet blend was found to meet their nutritional requirements.

The results of these studies indicated that ultrafiltration processing has the
potential to improve the utilization of legumes due to the fi'actionation of high quality,
high value components while enabling tremendous savings in waste of residual
components. The additional advantages of low operation costs, and simplicity in
operation, make ultrafiltration processing an important technology for increased
utilization in the food industry. Typical annual operating costs including maintenance,

replacement membranes and electricity are generally about 10% of the initial capital

158

investment. This is significantly lower than the average operating cost of other types of
technologies currently in use by the food industry, which may range fiom 25 — 40% of
the initial investment.

The success of ultrafiltration processing into emerging economies must be based
on the existence of a demand for the processed product, a demand that can be satisfied
profitably. Current statistics on global hunger and poverty, and protein malnutrition
suggests that this processing technology could play a major role in improving global food
security. This would be possible mainly through the production of safe and nutritious
food for the consumer, as well as providing a source of livelihood, and therefore money
to access food. The flexibility of ultrafiltration processing in terms of the production of a
wide range of products ensures full usage of equipment. The choice of membrane
systems can make it accessible to small, rural communities or large-scale commercial
applications in urban areas. Due to the high initial capital cost, large-scale commercial
operations would be recommended for technology transfer only to middle and high-
income emerging economies.

The research conducted in this dissertation provides a fiamework for further
research using other types of membrane modules such as mineral membranes, to assess
the properties of the separated components, the reduction in fouling and hence
improvement in yields. Additionally, an economic assessment is needed to evaluate if this

technology is an “appropriate” technology for the developing world.

159

APPENDICES

160

 

APPENDIX 1

SUPPLEMENTARY MATERIAL

STUDY 1 OPTIMIZATION OF THE AQUEOUS EXTRACTION OF PROTEINS
FROM COWPEAS (Vigna unguiculata) AND NAVY BEANS (Phaseolus
vulgaris)

161

I62

05... .55 55.5. 55... 55 55.0 0 55
55.5 55.5 55.5. 00.... 55 55.0 5 .6
55.05 55.5 55.55 .5..5 55 5. .5 5. :5
55.5. 55.. :55 55.55 55 5. .5 5. 8
50.0. 50.5 .555 55.55 55 5. .5 5 :o
05... 50.5 55.55 55.5. 55 5.5 0 .6
55.5 55.5 55.0. 5.0. 55 5. .5 0 :5
3.5 50.5 55.5. 55.5. 55 5.5 5 :5
50.05 50... 55..5 $.05 55 05.5 5. .6
05.55 55.5 55.55 55.55 55 05.5 5. :0
55.5. 00.5 55.55 55..5 55 05.5 5 :55
5. .5. 05.5 55.05 0555 55 05.5 0 55
55... .55 50.5. 50.5. 55 05.5 .0 ..u
55... 05.5 55.55 55.5. 55 05.5 5 :5
55.55 55.5 55.55 55.55 55 55.. 5. :0
53.5 05... 55.55 55.55 55 55.. 5. ,5
55.05 55.5 05.5 50.55 55 55.. 5 .6
05.55 55.5 0. .55 05.55 55 55.. 0 :5
55.5 .55 ....5 5.... 55 55.. 0 ..o
55.5. .55 55.05 55.5. 55 55.. 5 .6
50.5.. 55... 55.55 05.55 55 55.5 5. .6
50.55 50.5 55.55 0....5 55 55.5 5. :0
50.55 .55 55.55 55.55 55 55.5 5 5:
50.55 55.5 55.55 55.55 55 55.5 0 :5
55.5. .5.5 50. .5 50.0. 55 55.5 0 :55
50.5. 55.5 55.55 55.55 55 55.5 5 :55
:8: 5 55. 2.225 50 .850 5.50 050 6.0 :55
03; £820 50=om 20200 5890 082020 85:08 :_ 5280 2:55:00th 0505 22:50 0.00 20:30

 

8380083 0:5 5505. 250:5: .30 0: 5520.55 5.5:? 30::
5505808 5505 5000 55.0 550300 5:50 0.5; 5280 0:5 50:00 53250 .5:55:55 5280 082080 4:55:05 5220 00 5350.

163

0. .0. 55.5 05.55 5.05 05 5.5 5 5.0
50.5. 05.5 50.05 55.55 05 5. .5 0 5.0
5.5. .05 55.5. .0..5 05 5.5 0 5.0
55.5. 50.5 55.05 50.55 05 5.5 5 5.0
55.05 55.0 00.55 0.55 05 05.5 5. 5.0
.005 00.5 0...5 50.5 05 05.5 0. 5.0
00.0 55.0 05.05 05.05 05 05.5 5 5.0
00.0. 50.5 55.55 50.05 05 05.5 0 5.0
50.5 05.5 50.5. 05.0. 05 05.5 0 5.0
05.0. 05.0 50.5 0.5. 05 05.5 5 5.0
0.55 50.0 50.05 05.05 05 05.. 5. 00
00.55 00.5 .505 .055 05 05.. 0. 5.0
5.05 50.5 50.55 50.5 05 05.. 5 5.0
5. .55 5.5 00.55 50.05 05 05.. 0 5.0
5.5 50.. 05.5 55.5. 05 05.. 0 5.0
05.5. 05.5 0.55 0.55 05 05.. 5 5.0
.005 50.0 50.05 00.55 05 05.0 5. 5.0
50.05 50.5 50.55 0.05 05 05.0 0. 5.0
55.55 00.5 5.55 50.05 05 05.0 5 5.0
.0.5 0... 05.05 50.05 05 05.0 0 5.0
0. .0. 50.5 50.55 05.05 05 05.0 0 5.0
50.5. 55.5 50.55 05.05 05 05.0 5 5.0
5...5 05.5 00.05 50.05 55 55.0 5. 5.0
55.5 50.. 00.55 55.55 55 55.0 0. 5.0
00.5. 55.5 05.55 50.55 55 55.0 5 5.0
0. .5 55.. 55.0. 05.55 55 55.0 0 5.0
000.. 5 55. 2.0.25 5. 05. 05. 05. 605 05.5
0.5; 50.25. 50:05 0.02.05 50.20 008.02.. .5808 ... 50.85. 235.0080... 05.5 0.0.5.5. 5.5. 0.0855

5.58. 00 0.00..

0. .5.
55...
05.0
50.5
55.0
00.0
50.55
55.0.
.50.. 5 00. 2.0.8.. 5.
0.0; 0.0.25.

15:;

05..
.5..
05..
5.5
05.5
.50
55.5
000.

50505 0.0205

5.05
55.05
.005
05.5.
05.0.
55.5
0.05
.05.

.555
5.55
55.55
05.5.
55.5.
55.5
50.55
05.

0.89.5. 00.200... 855.58 ... 0.0.85.

05
05
05
05
05
05
05

G...

55.0
55.0
55.0
55.0
55.0
5.5
5.5
00.5

95.82.80... 05.5 0.0035.

0. 5.0

5 5.0

0 5.0

0 5.0

5 5.0
5. 5.0
0. 5.0
.... 0.0.055

5.000. 00 0.005

164

165

55.. .05 5.0. 0.... 55 55.0 0 m2
55.5 05.5 05.5. 05.5 55 55.0 5 5.2
55.05 55.5 55.55 55.05 55 5.5 5. 5.2
55.5. 55.. 05.05 55.55 55 5.5 0. m7.
55.5. 5.5 55.55 50.55 55 5.5 5 m2
00.55 .05 0. .55 05.55 55 5. .5 0 m2
00.5. .05 0. .5. 55.55 55 5.5 0 m2
05.5 50.5 55.5 55.55 55 5.5 5 m7.
.055 55.0 55.55 00.55 55 055 5. 5.2
05.5 00.5 50.5 05.55 55 05.5 0. 5.7.
50.0. 55.0 55.05 50.55 55 05.5 5 5.7.
00.0. 50.5 0. .05 55.55 55 05.5 0 m2
50.0. 05.5 50.0. 50.55 55 05.5 0 5.2
.0.5 05.0 55.55 .505 55 05.5 5 m2
00.55 50.0 50.05 55.55 55 05.. 5. 5.7.
5.0. 00.5 00.55 0.55 55 05.. 0. m2
55.55 50.5 0.55 50.05 55 05.. 5 m2
.055 5. .5 55.05 0.55 55 05.. 0 m7.
05.05 50.. 00.55 05.55 55 05.. 0 5.7.
.00 05.5 50.55 50.55 55 05.. 5 .07.
00.50 50.0 .555 55.00 55 05.0 5. 5.7.
50.0. 50.5 00.55 .055 55 05.0 0. m2
05.55 00.5 55.55 ...55 55 05.0 5 m7.
...55 0... 50.05 55.55 55 05.0 0 5.2
55.0. 50.5 50.55 0.55 55 05.0 0 m7.
05.0 55.5 55.05 05.05 55 05.0 5 m7.
000.. 5 55. 2.0.0.5 5. 05. 05. 05. 6... 00.5
0.0; 50.0.0 50:05 0.00.05 50.80 008.080 5055.50 :. 50.0.0 0535:0080... 05.5 0.0.5.50 5.0 0.0855

 

uln:.ol‘0.. .

05.5.0080. 0:5 05.5 0.0.5.50 £0 50 5:020:00 500...? 500:: 30055.50

5005 50.... 52. :500 5>5Z 80:5 0.0; 50.0.0 0:5 50:05 0.00.05 0:05:00 50.050 00.0.0050 ..:0.:00 50.0.... 50 0.05...

166

55. .5 55.5 05.55 55.55 55 05.5 5 5.2
00.5. 55.5 55.05 55.55 55 05.5 0 02
55.5. .5.. 55.5. 55.55 55 05.5 0 02
50.5. 5... 55.55 .505 55 05.5 5 5.2
55.05 50.0 50.55 50.55 55 55.. 5. 5.2
55.55 55.5 55.55 55.55 55 55.. 5. 5.2
55.5. 50.5 55.55 55.55 55 55.. 5 5.2
5. .55 5. .5 55.55 5.55 55 55.. 0 02
.05. 50.. 05.55 50.55 55 55.. 0 5.7.
55.5. 55.5 50.05 55.55 55 55.. 5 5.2
55.50 50.0 55.55 55.55 55 55.5 5. 5.2
55.55 55.5 50.05 00.55 55 55.5 5. 02
55.5 55.5 55.05 50.55 55 55.5 5 5.2
55.55 0... 55.55 55.55 55 55.5 0 02
55.5. 55.5 55.55 55.05 55 55.5 0 02
55.5. 55.5 .555 05.05 55 55.5 5 5.2
55.0 55.5 55.55 50.55 55 55.0 5. 5.7.
50.5 50.. 55.05 50.55 55 55.0 5. 5.2
50.0 55.5 05.05 55.5 55 55.0 5 5.2
50.0 55.. 5...5 55.5. 55 55.0 0 5.7.
55.. .55 5. .0. 0.... 55 55.0 0 02
55.5 55.5 05.5. 05.5 55 55.0 5 5.2
55.55 55.5 55.55 55.55 55 5.5 5. 5.2
55.5. 55.. 05.55 55.55 55 5.5 5. 5.2
55.5. 5.5 55.55 55.55 55 5.5 5 5.2
00.55 .55 5.55 55.55 55 5.5 0 02
000.. 5 55. 2.0.0.5 5. .00. .05. .05. .0... 00.0.

0.0; :.0.0.0 50:05 0.00.05 50.0.0 030.080 .05....0 :. 50.0.0 0.0.5.0080... 05.5 0.0.5.0 500 0.0855

05.000. 50 0.005

167

05.5. 50.. 0. .05 50.55 05 55.0 5. .07.
50.0. 0. .0 55.55 50.55 05 55.0 0. 5.7.
55.5 55.0 00.55 50.5 05 55.0 5 5.7.
55... 00.. 00.5 55.5. 05 55.0 0 07.
50.5 05.0 .00. 0.... 05 55.0 0 5.7.
.55 55.0 5.5. 05.5 05 55.0 5 5.2
5.55 50.5 ...55 55.05 05 5.5 5. 02
.505 .55 05.5 55.55 05 5.5 0. 5.7.
55.5. 5. .5 50.05 50.55 05 5.5 5 5.2
0.0 .05 50.55 05.55 05 5. .5 0 5.7.
05.0. 0. .5 50.5. 55.55 05 5.5 0 m2
55.5 05.0 5. .55 55.55 05 5.5 5 5.2
55.55 55.0 50.05 00.55 05 05.5 5. 5.2
50.55 55.5 50.55 05.55 05 05.5 0. m2
000.. 5 55. 2.0.0.5 5. .05. .05. 05. .0... 000..
0.0. > :.0.0.0 50005 0.00.05 80.0.0 008.080 88.50 :. 80.0.0 0.0.5.0080... 05.5 0.0...50 20 0.0855

.0....00. 50 0.005

Table 48 Contribution of pH, particle size, extraction temperature and bean type to
protein content in cowpea and navy bean extract

-.a-‘-

 

 

 

"€3.10 Nparm [513""""§1;;a of éauares FRatio Pr'iSB'f>F'"

 

 

 

pH 5 5 3099.87 62.76 < 0.0001
Particle size 4 4 1247.51 31.57 < 0.0001
Temperature 1 1 18.27 1.85 O. 18
Bean 1 l 55.03 5.57 0.02
40 4 '
35 -
3O - ._
25 - . .- I . .
-‘ . .I i...
. ' .r "-
20 _ I I .
15 4 .' . '
.1
10 " r r r r l ' r ' l r r

 

 

 

Predicted Protein Extracted (%)

Figure 23 Leverage Plot for the Whole-Model Test of pH, temperature, particle size
and bean type efl'ects on % protein extracted from Cowpea and Navy Bean
flour

168

 

 

 

 

 

lO
5" . - .
J ...- . .U ..I' .... I. .
. I ..- . ..fi. I. a .'o .
Residual 0 ' -t _ :' :' -: t: .':'
. - " -:' I '- '
-5- : '
'10 r I i If' I ' I r l t r

Predicted Protein Extracted (%)

Figure 24 Residual Plot for the Whole-Model Test of pH, temperature, particle size
and bean type effects on % protein extracted from Cowpea and Navy Bean
flour

l69

 

40-+

35‘

d

30‘ '

 

 

 

Protein
Extracted (%) 25 1

 

 

 

20- .
15-1 E _
10- ,
T r I ' T V T
20 25 30 35

pH Leverage

Figure 25 Leverage Plot for the significant effect pH on % protein extracted from
Cowpea and Navy Bean flour

170

 

40‘

 

 

35W :=
.1 ' I I
30 “ E /'
Protein ' ; r' u 3
Extracted (%) 25 T ' l
. . I a
20 ‘ : I 3
.J I
15 fi .
10 I I I I r f I

 

 

 

16 18 20 22 24 26 28 30 32

Particle Size (Screen Size) Leverage

Figure 26 Leverage Plot for the significant effect particle size on % protein extracted
from Cowpea and Navy Bean flour

171

 

 

Protein
Extracted (%)

 

 

 

 

 

 

10 Trlrf'l’l‘l

25.0 25.5 26.0 26.5 27.0 27.5

Bean Leverage

Figure 27 Leverage Plot for the significant effect bean type on % protein extracted
from Cowpea and Navy Bean flour

172

APPENDIX 2

SUPPLEMENTARY MATERIAL

STUDY 2 ULTRAFILTRATION PROCESSING, CHARACTERIZATION AND
FUNCTIONAL PROPERTIES OF COWPEA (Vigna unguiculata) AND
NAVY BEAN (Phaseolus vulgaris) PROTEIN FRACTIONS

173

55.5 55.0 00.5 55.5 55.0

.00.. .50 .05 .55 00.0
5.0.. 05.0 00.5 50.5 00.5
0v. w. .0 50.5 00.5 00...
5.... 5. .0 05.. .055 00.5
0... 0.0 55.. 05.5 00.5
05.. 0. .0 0.5.. 00.5 0.0..
55.. 0.0 0.5.. 55.5 55..
50.. 05.0 05.. 50.5 00..
00.5 55.0 55.. 00.5 55.0
0V5 05.0 50.5 0. .5 0.0.0
0V5 .50 05.5 05.5 05.0
.05 55.0 05.5 0.0.5 05.0
50.5 05.0 05.5 55.5 5.0.0
0.0.5 5v.0 00.5 ...v 0.5.0
505 0v.0 00.5 h...» 55.0
50.5 05.0 .55 55... 05.0
5. .5 55.0 05.5 05.... 0.5.0
05.5 50.0 005 50.». 55.0
00.5 :0 00..V 05.0 05.0
0.0.5 55.0 00.5 05.0 0.. .0
.0..5 55.. 05.0 55.0 0.0
v0... .05.. 00.0 00.0 50.0
.0.. .00... 50.0 055 50.0
3.5 05.. v0.0 V5.0 50.0
00.0 0 0 0 0
0.080.050 5 0.008.00 5 0.008.00 . 0.008.00
...55. .5 . E ......

tingsatvo . ti. Igbn

0.00.00 000300 ..0...< 0:00:30 00 5.5000080 =0..0......0...D 05.0.. .3... 0.0.80.3. 080 8008.00 av 0.000.

174

55.5 00.0 50.. 50.. 00.5
50.5 05.. 55.. 05.5 . 55.5
00.5 5.5 .5.. v0.5 0.0..
555 .5.5 00.. 50.5 50..
55... 55.5 00.5 0.5 55.0
05... 50.5 55.5 05.5 05.0
55.5. 50.5 55.5 .05 50.0
.05. 00.5 50.5 50.5 0.5.0
.55. 50.5 50.5 .0...V 55.0
00.5. 55.5 00.5 3.... 05.0
00.5. .0..5 .05 50.0 0.5.0
.05. 55.5 50.5 50..V 55.0
55.5. 55.5 0. .5 00.5 05.0
00.5. 50.0 0.5.5 50.5 0...0
00.5. 55..V .0..5 .00 0.0
55.5. 00.0 05.0 00.0 0.0.0
05.05 .05 05.5 05.0 50.0
00.5. 05.0. 505 00.0 50.0
00.0 00.0 00.0 00.0 00.0
0.0.80.3. 5 2008.00 5 8008.00 . 0.008.00
, 1.83. x2. . . .2. mm...

.lttu: .....

0.00.0.5. 0005. .902 ..0...< 8000...... 00 5530080 85.08.0085 55.00 .3... 0.0.80.0”. 0:0 0.88.00 05 0.000.

175

176

00.5. 55.. 00.5. 00.5. 50.0
05.5 .... 50.0. 00.5. 00.0
00.0 55.0 00.5 50... 00.5
50.5 50.0 00.0 50.0 00.0
50.5 55.0 0.... 0.0. 00.5
00.5 05.0 55.5 50.0 00.5
55.. 55.0 55.. 00.5 0.0..
55.. 55.0 55.. 00.5 55..
55.. 0. .0 00.. 05.5 50..
50.. 5.0 55.. 5.5 55.0
55.. 0.0 5... 55.. 00.0
05.. 0.0 .... 00.. 55.0
v... 5.0 50.. 05.. 05.0
50.0 v. .0 50.0 .5.. 50.0
00.0 5. .0 05.0 05.. 0.5.0
550 5.0 55.0 05.. 55.0
00.0 5.0 55.0 0... 05.0
50.0 5.0 50.0 50.. 05.0
50.0 5.0 55.0 00.. 55.0
05.0 5. .0 00.0 50.0 05.0
05.0 5. .0 05.0 00.0 :0
55.0 ...0 550 .0..0 0.0
0. .0 0. .0 55.0 05.0 00.0
00.0 50.0 05.0 .50 50.0
50.0 50.0 5.0 5. .0 50.0
00.0 00.0 00.0 00.0 0

0.0.00.3. 5 8008.00 5 8008.00 . 0.008.00
, -..... .0520.» :i .05 00......
0500000... =0..0..__..0...D 55.00 0.00.8.0 00030.0 ..0...< 0:00:30 ..0 080.0> .m 0.00...

8.0 A A 0.5.0 00.. _ ~05 . 00.0

0.... a. .o 3.. t .0 8.0
00.. :0 0... 000 8.0
00.. 0.0 m... 0.0 8...
3.. :0 2.. 2.5 8.5
8.. 2.0 2.. .00 8.5
N... 3.0 o... 2.0 .0..
2.. :0 a... .00 50.
o... 0:. 00. £0 00..
0... mm... 8.. 03 50.0
8.0 3.0 2.. 9.0 $0
2 .0 00.0 8.. 3.0 00.0
00.5 05.0 3.0 8.5 8.0
2.0 X... 8.0 05.5 00.0
$0 .00 .00 2.... .00
8.0 00.0 8.0 00.5 55.0
$0 a... 2.5 $5 000
8.5 02. 8.0 00.5 0.0.0
5.5 a... 8.5 on... 50.0
ohm .00 2.0 02. 05.0
8.. ~00 N}. o...“ :0
on... o... 80 o... 2.0
on. 02 00.0 on. 8.0
8.. on. 8.0 000 8.0
8.5 8.. 00... 000 8.0
o o o o o
0.0.00.3. 5 8008.00 5 0.00800 . 8008.00
0.3. 0.0. 32... 00 0...;

5500000... 05.00.0005 05.00 0.00.05. 000300 00...... 00000.5. ..0 0.0. 30.0 55 0.000.

177

178

5.: 5N0 «Nam 000— $6
000— 50.0» 3.0. 900 00.m
090— mad 0m.m 0N6 mad
00.x >00“ mm._ 005 he.—
mhd >0.N 0m; 00d 00..
00.m 00.. 9: on; and
00.m 0m; 0N4 3; 000
00% NM._ 00; 002- Ned
00.v 0N._ 00.0 mm; hmd
09m no; 2.0 9; mmd
mm.m 00.0 3.0 2; 0m.0
00d 00.0 No.0 50; had
00d :0 and 00.0 mNd
0m.~ no.0 mmd no.0 0N0
00d and bvd 00.0 2.0
mm; Nvd and $0 2.0
m0; 5N0 mad and 0.0.0
00.0 2.0 2.0 3.0 8.0
mad _ _ .0 2.0 0H .0 No.0
00.0 00.0 00.0 00.0 00.0
0030003— m 0000.50.— N 0308000 0 0808000
.. ..m.00.0e=_o>....- .‘ £0...me ., _

.O'OOO.lal. ... ....o t-l.lo

05000005 0002:0805 mats—0 308wa 000m .902 =8=< 0:00:30 .00 0€:_0> mm 030%

00.. - 000 000 00.. . 00.0

3N 5.0 N00 he; 00.m
2.0. 0: m0._ 0.0 mm.m
00¢ vw; m_._ wmd no.—
mmd _0._ mm; ovd mo;
:.0_ 09m mmN nod wmd
00.2 00.m 00d N0.m 0m.0
@90— Ca 00d N_.m N10
5.0. has” OWN Sim two
0m.0~ m3” _m.m 0.0m mmd
2.: 0.6m 0m.~ 0.00 0&0
00.2 80. SN 5v 5.0
3.: 0m.m mmN vmé mad
0m.: mm.m mud «he 0N0
00.N~ wed NmN Ed :0
009 0N."g 0N.m 0_.m 2.0
3.2 8+ 2am 05m 00.0
00.w~ 0_.m 0w.v ova 8.0
09: N50 00.0 00.0 No.0
00.0 00.0 00.0 00.0 00.0
000000000 0 0.008000 N 0000800 _ 0.008000
.. £9 22 30E... . SEEP

Iliq...li.. u.v.v... A.vnllund¢lnu|II-Il.

$500005 000050005 macaw 080be 0000 >>0Z :00=< 0:00:30 .00 €030 0000 30E Vm 0.00%

179

Table 55

Samples

Isolate
Permeate 1
Permeate 2

Permeate 3

Total

Protein Recovered (% of Aqueous Extract) after Extraction

at pH 10 and 25°C

 

 

 

Protein Recovered (%‘bf Extract)

COWpea 7
75.6
7.49
3.63

0.67

87.48

180

Ndvy Bearr
72.6
4.70
2.14

1.98

81. 43

00.0000
00.000
.0000
.0000
00.000

00.000.
00.000

0000 . 0
00.00.
0. .000
00.000
.0.0.0

00.000.

00. .00.

00.000

00.000.

Z.m.0<U

0000000... 0000. >>0Z 000 000300 ..0 0.0.000 0:.E<00.000-..0>10 0.000 0... 000:: 002 .000 008.... 00.000000

00.000
00.000.
00.0.0
00.000
02
.0000
00.00.
00.000
02
00.000
00.00.
00.00
00.000
00.000
...000.
00.0.0

.00

00.000
.0000
00.00
00.000
00.00
00. .00
00.00.
DZ
0000.0
00.000
00. .00
0.00.
00.000
00.00.
0. .000
00.000

.02

00.00. ._
00.000
00.000
00.0.0
00.00.
00.000
0 . .00.
Q7.
00.000
00.000
00.000
0000.
00.000
00.000
. . .000.
00.000

00.0.7.

00.000
00.000
00.000
0.000
00.00
00.000
00.00.
00.000
00.00
00.000
00.000
00.000
00.000
00.000
0 . .000
00.000

..Dllnnn‘ .

00.000
00.000
00.00
00.000
.000
00.000
.000.
00.000
00.00
.0000
00.000
00.000
.0000
00.000
0. . .0..
00.000

:0.

00.0.0
00.000
00.000
00.000
00.00
00.000
0.00.
00. .00
00.00
0.000
.0000
00.000
00.000
00.000
00.000
00.000

. 000.0.

n v'leO‘. 4:.

00.000
.00.0
00.00
00.. .0
00.00

00.000
00.00.
00.0.0
00.00.
000.0
00.000
00000
00000
0.000
00.000
00.00

".00

0300000030 “0: I OZ

0.00
0.2
00.00
00.00
...0N
00.00
00.0
N00.
0. .0.
00.0.
00...
00.0.
00.0
00.0
00.0
00.0

00..
0....
00..

0..
.02
.0>
00...
9.0—
0.<
E...
02
00
00
..00

0.0
00<

302:. ..mm.....0£0£0<

0m 0.00...

181

 

 

 

 

 

 

 

   

 

 

 

   

 

 

 

 

 

 

 

 

 

 

AD!” ADHES
CPF (,0 . N BF 60 -
mar P. 40 T crux CRlT P ~ 40 r CHEN
20 0 20 f
mm a ‘ 0 GL‘M HARDvk \‘K‘ ~ GUM
WORK? cows WORK“ 'COHFS
Arms
CPPE NBPE ‘ :
GUT P- may (1 my
HARD “-0-. (IN HARD?“ ‘* ~(1M
- Arm-s
30" l
l .
l '5 l .
CRIT P ( our P . ‘ l on
. . l0
HARD?" '
,4” . ‘
W(RK‘ ’Cm mils; _ m
Amns
SPI so l
our P 40 T CHEW
‘ 201
HARD 0‘ " A :\ \ cur
.'/ \
r' “\r
would] \‘(UHI-‘S
Figure 28 Texture Profile Analysis of Cowpea and Navy Bean Protein Gels

(adhesiveness (ADHES), chewiness (CHEW), gumminess (GUM),
cohesiveness (COHES), work done on gel (WORK), hardness (HARD),
critical protein for gelation (CRIT P)

182

Table 57 Hunter Color Characteristics of Cowpea and Navy Bean Protein Fractions
after Ultrafiltration Processing

‘--.....-----.-...----..-_.......-.. ..-.- .-. ”-....._.............A.A.....-‘- -.u....-._-.--....._. -. -.. . .....w...-u ,7 ..H...

mess"; L a b
CPF 35.5 ' 1.8 -14.5
CPEl 31.9 0.9 -10.4
CPE2 21.2 -1 -51
CPE3 19 0.2 -7
CPPE 23.4 0.6 -10.4
CF?! 18.7 0.9 -6.8
CPP2 19.1 0.4 -6.5
CPP3 18.6 0.7 -6.6
CPI 26.9 0.4 -34
NBF 35.9 2.2 -172
NBEl 16.9 1.6 -9.8
mm 17.5 0.7 -8.8
NBE3 14.5 1 -73
NBPE 16.7 1 -8.8
NBPl 12.1 -o.5 -3.1
NBP2 13.1 -0.3 -4
NBP3 12.5 -0.5 -34
NBI 15.6 0.9 -8.2
SP1 37.9 3.5 -231

183

Table 58

F-Values and Statistical Significance of Functional Properties of C owpea

and Navy Bean Protein Fractions after UF Processing

Variable =

Hydration

WAC

OAC

Protein Solubility

Structural Rheological
Critical Protein for Gelation
Work done on Gel
Hardness of Gel
Adhesiveness
Chewiness
Gumminess
Cohesiveness

Tp

Ti

AH

Surface
Foam Capacity
Emulsion Capacity

Color

L-Value
a-Value
b-Value

F-Value ’ 1’

7401.30
4681.69
1121.72

1710.42
266.58
352.38
319.76
846.42
318.98
40.21
50.56
32.38
22.86

540.76
3311.02

2648.50
155.88
1030.45

‘p‘.vaiue

< 0.0001
< 0.0001
< 0.0001

< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001

0.0003

< 0.0001
< 0.0001

'< 0.0001

< 0.0001
< 0.0001

 

184

03:50:: Em 000 N 83? 8203< .205. few 3 585.36 803 0m.0 N mafia—.008 5:28.50 .8 mos—Q, 2280< _

 

Il|ll

'IIII'

 

 

 

 

3.0- 5.0- $0- vloio. 60- 3.0- 2.0 .310.” 20. 00.0- .340. 3.0 v0.0 30 ~00 n
- -_ld 3.0 $0. Mumd 30 000- de wmo 2.0 .004? R0- 30 50.0- 2.0 a
- 2.0 3.0- mlolo S0 v0.0 wlwlo 9.0 2.0- 3.0- _0.0- 2.0 5.0- 0Y0- 4
- 3.0 00.0 0m.0- a 50.0- @019 $10. 20- a 000 NWO” a 0.x.
- 3.0- $0- 5.0- 3.0. w00- 3.0 a :0- 20 0N0- 3.0 3:00
- W0l0. 2.0 mg 30 :0- :0- 3.0 :0 _0.0 $0- .839
- 000 91d a mind“ 20- ~10 ~00 000 Mmdu .325
- 0m.0 le afidl- 00.0 000 00.0- 3.0 mid” .853
- ole-i 0&0- mm0- :0 00.0- No.0 3.0- 033
- ad” 30- ad 0m.0- .de olmdn x33
- 8.0- a 000 “a Mad 0.00
- 2.0 9.0 30- 20- mm
- 000 3d 3.0- U<O
- 0.0 3.0 055
- 80- um
- Um
w w .18 .u-MEOL .XSU ENE-u ..nmthv-Qfiwt EOE ANTED mm UVQ 03: um UK
_ 82880 5805 swam >32 can 80300 00 mos-500:— Bcozocsm 00 5.32 coca—830 0m 2an

185

005.300: 03 000 N 32? 80.0%? .232 $0 a 008E006 203 000 N 30205000 20:20:00 .00 329 820mm?~ _

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00.. 3.0- 000- 00._ ~de m0.0- 2.0- 00._- 5.0- v0.0- 000 ~00 :0- 00.0 00.0- 00.0 0
00._ .00 00.0- 3.0- Mmd N010. Nam-0 0.0-.0. NWO mg.” 20- mod 0V0- 00.0 00.0- m
00._ 0.de no.0- mwd Maud dad mmd- M040 midl- mmd almd >00- ml0d 000- A
00.. fl N00- :0- 00._- 5.0. 30. 000 N00 000- 00.0 00.0- 00.0 0.x.
00.. 2.0. 3.0. a and- 00.0- 000 N00 30 50.0 05.0- 00.0 3:00
00._ wmlo. 01040 wmd 00d a «N0 «001.0 2.0- a 9:20- 0.50
00._ Nd mol- mlofi a 50.0 a 0N0- 050 00.0- .3000
00._ a de 00.0- 00.0- 0N.0 8.0- 00.0 3.0- .35?
00.. 01010 Mm-dl- 000- mmd. 00.0- 00.0 00.0- 03$
00._ a 3.0- 000 N00- 000 E0- 0:03
00.0 3.0 500- N50 000- 00.0 “:00
00.0 50.0 00.0 flu fl mm
00.. 5.0 vm0 2.0 0<O
00._ mN-oi- a 035
00._ a 0m
00._ 0"—
0 0 - «ER. .3000 .530 .330 303- 30: «83 & EC 0% 0V0 .03: 0M 0%
_ 0:530th 000820055 300 300an 580.0 30300 00 303005 3:000:3— 2000 x502 00:20:00 00 030,—.

186

 

000.0000: 000 00.0 N 30.? 05.030. .m.0>0. $0 00 «080.0% 0003 00.0 N 33.2.0000 00.00.0000 00 30.? 05.0000. _

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00.. 010d“ 00..- 0101.0 0%0 00.0- 00.0- 00.0- 00.- add 00.. .v.0 00.0- 00.0- 00.0- 00.0 0
00.. 00.0 000- 0w0- 00.10 010.0 “040 @010 N40 00.0- 00.0- 00.0 00.0 00.0 00.0- 0
00.. N00- :0- w0.0 00.. H0 00.. 0Nl0 .0010... :0- 00Jl0 0.0.10 001.0 00..- A

00.. 0.0.10 5.0- 00.0- m0.0- 000- 00.0- 00.0 .50 00.0- 00.0- 00.0- .00 0.x.

00.. .00. 3.0- 000- 00 0- 00.0- $0 «0.0 mwd- 0m.0- 5.0- 00.0 .3000

00.. 000 m1010 NWO 30 m1? :0- 0010 00.0 .00 w0.0- .050

00.. 0.010 01.0 0N0 0W0” N00- .wl0J0 00.0 00.0 00.0- .3000

00.. %0 03 $0. 3.0. 00.0 5.0 5.0 :0- .3...0<

00.. H.010 00..- .v.0- .01010 0.01.0 § 00..- 0001

00.. mw.0- 3.0- 00.0 00.0 00.0 000- 0.003

 

00.. 00.0 00..- 000- 00.0- 00.0 0.00

 

00.. 00.0- 50.0- 00.0- 00.0 mm
00.. mg 00.. 00.10- 0<O

00.. 00.. NWO” 0<3

00.. ~001- 0m.

00.. 00

 

m 0 N 00 .3000 .050 0:050 3.0% 0.8: fox: & ~00 0% DVD 03: UM 0k

_ 3.030000 00.80.0005 0000 000.008... 0.0005 000m >37. .00 3.000000 300.600”. 05.00 5003. 00.00.0000 .0 0.02.

187

APPENDIX 3

SUPPLEMENTARY MATERIAL

STUDY 3 NUTRITIONAL QUALITY OF COWPEA AND NAVY BEAN
PROTEIN DIETS BY IN VIT R0 AND IN VIVO PROTEIN
DIGESTIBILITY CORRECTED AMINO ACID SCORE (PDCAAS)

188

0.00.0.6»: <m2 E 0.3.000 0:. a
$00.80.... 0.00 05 00.30.08 0.8.0.. 00 £00.05 000 _

189

vm.m0.. N. .03 0v. . m0 mmde 3.03 mméON. m0.000 m . .0m v.01.
0N000 30.30. 00.m.N $0.3 3.0V. 0000. v0.0Nm 0.0m NE...
0V0?“ 3.0mm 00.00m .0.m.m 0.0.0 0v.N0m mvéNv .0.... 0?.
0w0.0 mN.0.0 m0..~0m 0000m 00000. 00.0.0. m0.m00 000N :04
00.0mm N000.“ ..03. .002 00.000 0..m0m Nv. .30 000.0N 0..
.vam 3...; N0.Nm 0m.0m. 3.0m .00.” 0v.00. N0.0N 100
N000. mm. .0 N023. 00. .0 N000. 0m.0m. 8.me ....N .0.).
0m .00 0vN0v NO0Nv 00. mmv w. 000 0N0m0 00. Sum 00. N .0>
0. 00N N. .0N. .wNm. 3.0m. voNvN 0m0¢N $00. NN..N .00
mm00N. .0. .wmv 0m.N0m 0.0.00.0 mN.mm0 00.00 00.000. N0.m. 0:.
300 00.00 0&me 00.00. 00000 m00N. 00.m0 0.0. 0.<
000$. 000$ 0N.w0N 00.00.v N003. 00.0mm 0m.0.0 v0.2 .2...
00.Nwm ww. .Nm 0?me 0m.NON 000% 3.: ..0mm 00... m2
00..VN mm.m0N Nw.m0. 3.va 000mN ....0m 30..» 00.0. 0.:
$.va 2.30 001me 0.0.. .0 .0000 000mg 00.0m0 000 0.0
.N.vwm 200v N000.v N212. 0N.0w0 N...w0 mN0mm m0.0 Sm
0m.0.m. 0m000 02mm w000m. 0m00v. N003. 00.000. 000 0.0
N030 8.00m 00..Nm 003v 203 NN.0v0 0V0; 0m.V am<
0.8-22-2 . 028. 022,. 028 .4 3.. --...008. . ..-...UE 00% 08.53.,,......2.o.-m.,o.._.£<

 

0.0.0 000m 0>0Z .000 000300 ..0 0.0.3.~ 0:.E< ..0 08.... 00.2.0.0”. 0:0 002 0.000 N0 0300

_ __—.—._ .._—-——._T__—- _._._ ._ __1
‘

 

.

0N 0. 0. 0. 0. m.

 

 

 

30:57. :3. .

v. m. N. .. 0. 0

. [0 .

 

 

 

P

. 00:2. 0503.. 0550 38 ..o 8.8:. .50... .. 0N 830.”.

000m¢mN.
pbpberON

00

 

0v

on

 

00

(3) paumsuog poo J

 

00

£8

 

 

 

 

 

190

 

 

 

0N
\

 

N 02.80 00.03.. 0550 3.2 ..o 8.3:. 08... - on 230:.

 

0. 0. 0. 0. n. v. m.

.fb

 

 

 

 

 

..0.. :57. .3.

N...0.0 0

p .

 

 

0 0

Pl.

 

 

nVMN.

 

 

com

L0N
0m
00
.2
00
00
00

 

 

 

(3) pomnsuog poo J

 

191

 

0N 0.

\. 'LlltL‘IJLII

 

 

0. 0.

 

0. n. v. m.

 

 

 

I?

m 00:2. 0503.. 0::30 £8 ..0 3.3:. 000 H.

:2. :57. .3.

N...0.0000mv

P r p 0 .

MN.
..0“

b

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(8) pawnsuoa pong

.m 0.30....

 

192

 

 

0N 0. 0. 0.

 

 

 

 

F

0 00:2. 0508.. 0:000 38 ..0 8.8:. 000 u.

..0.—8:7. .3.

0 0

0.0.0.2220. 0000N.
m.
nN
mm
00
mm
00
00
00
n0

 

 

 

 

  

 

 

 

 

 

 

 

(3) mums 900:1

 

 

 

 

mm 2%:

193

. 00:2. 0508.. 0::00 £0.03 3: :. 0052.0

 

.305: z .3.

0N0.0.0.0.n.v.n.N...0.0000nvnN.

  

0N
00
00

1 .3 3:... n.

.3 .::.0:0.u

00

00.
0N.
00.

00.

(31M!!!

mm 2%:

|I.|1IJ

 

194

0N 0. 0. 0.

0. n. v. m. N.

,, iamflms
.3 0.50:0 .u

 

N 00:2. 0508,. 0::30 2.0.25 .2 :_ 00:25

 

.3502 .3.

.. 0.

0

 

000m

vnN.

0N
0..
00
00
00.
0N.

(3) 1M 3'1!

00.
00.

 

 

3 ea...

195

 

m 00:2. 0:.000. 0::30 2.0.03 .0: :. 00:20

 

0N 0. 0. 0.

.12. :3: n

.3 .0.—.0..5 n.

0.

n.

v.

m.

N.

 

..0.—:52 .0..

..0.0000nva.

0N
00
00
00
00.
0N.
00.
00.
00.
00N

(3) MA MI

 

 

2 050.:

196

 

0 00:2. 0:.000. 0::00 2.0.03 .0. :. 00:25

2 050:

 

.00—5.7. .0..

0N0.0.0.0.m.v.m.N...0.0000nVMN.

0
0N
0..
00
i‘ i , . i: 00
a .2... n .2
.3 0.50:0 n.

0N.
00.
00.
00.
00N

 

(3) IM Wu

 

197

 

 

Table 63 Correlation of Protein Nutritional Quality Tests
RPER RNPR RTD PDCAAS PDCAAS PDCAAS
(in-viva) (pH-Stat) (pH-Drop)
RPER 1.00
RNPR 0.90 1.00
RTD 0.81 0.75 1.00
PDCAAS 0.95 0.84 0.69 1.00
(in-vivo)
PDCAAS 0.84 0.71 0.41 0.94 1.00
(pH-Stat)
PDCAAS 0.93 0.84 0.57 0.97 0.97 1.00
(pH-Drop)

 

198

BIBLIOGRAPHY

A.A.C.C. American Association of Cereal Chemists. 1984. Approved Methods. (7th Edit)
St. Paul; MN. Method 46-13

A.O.A.C. 1990. Official methods of analysis of the Association of Oficial Analytical
Chemists (A.O.A.C). Washington, DC.

Abbot J., Glover F .A., Muir DD. and Skudder PJ 1979. Application of reverse osmosis
to the manufacture of dried whole milk and skim milk. J. Dairy Research, 46, 663

Aimar P., Taddei C., Lafaille J.P., and Sanchez V. 1988. Mass transfer limitations during
ultrafiltration of cheese whey with inorganic membranes. J. of Membrane Science;
38 (3) 203-221

Ali R., Staub H., Coccodrilli G., Schanbacher L and Jr. 1981. Nutritional significance of
dietary fiber: effect on nutrient bioavailability and selected gastrointestinal
functions. J. Agricultural and Food Chemistry; 29 (3) 448-484

Anson ML. and Pader M. 1957. Extraction of soy protein. US Patent 2 785 155

Arntfield S.D., Murray ED. and Ismond M.A.H. 1985. The influence of processing
parameters on food protein functionality. III. Efl‘ect of moisture content on the
thermal stability of fababean protein. J. Canadian Institute of Food Science and
Technology. 18 (3) 226-232

Barbano D.M. and Bynum DO. 1984. Whole milk reverse osmosis retentates for cheddar
cheese manufacture: cheese composition and yield. J. of Dairy Science. 67, 2839-
2849

Bérot S., Gue'guen J., and Berthaud C. 1987. Ultrafiltration of fababean (Viciafaba L.)
protein extracts: process parameters and functional properties of the isolates.
Lebensmittel Wissenschaft und Technologie. 20(3) 143-150

Beuchat L.R. 1977. Emotional property modification of defatted peanut flour as a result
of proteolysis. Lebensmittel Wissenschaft und Technologie. 10(2) 78-83

Blatt W.F., Dravid A., Michaels AS, and Nelson L. 1970. Solute polarization and cake
formation in membrane ultrafiltration: causes, consequences, and control
techniques. In Membrane Science & Technology, Edited by J .E. Flinn. p 47

Bliss F .A. and Hall TC. 1977. Food Legumes: Compositional and nutritional changes
induced by breeding, Cereal Foods World, 22(3), 106

Blonk J.C.G. and Vanaalst H. 1993. Confocal scanning light microscopy in food research.
Food Research International, 26, 297-311

199

Bodwell C.E., Satterlee L.D. and Hackler L.R. 1980. Protein digestibility of the same
protein preparations by human and rat assays and by in-vitro enzymatic digestion
methods. American J. Clinical Nutrition, 33, 677

Bolles AD. 1997. Ultrafiltration processing of dry bean (Phaseolus vulgaris L.) protein
fractions formulated into a beverage suitable for small children. Doctoral
Dissertation, Michigan State University, East Lansing

Boulter D. 1981. Protein composition of grains of the leguminosae. In Plant Proteins for
Human Food, Kluwer Academic Publishers, Boston, MA

Boulter DA. 1977. Quality problems in Protein Plants with special attention paid to the
proteins of legumes. In Protein quality fiom leguminous crops. CEE EUR 5686
EN:42

Bressani R. 1973. Legumes in human diets and how they might be improved. In
Nutritional Improvement of Food Legumes by Breeding. Edited by J. Milner. UN
Protein Advisory Group, New York, 15-42

Bressani R, 1975. A new assessment of needed research. Nutritional Improvement of
Food Legumes by Breeding. Edited by M. Milner. UN Protein Advisory Group,
New York, 381-3 89

Bressani R. and Elias LG. 1974. Whole soybeans as a means of increasing protein and
calories in maize-based diets. J. Food Science. 39 (3) 577-580

Bressani R. and Elias LG. 1988. Seed quath and nutritional goals in pea, lentil, faba bean
and chickpea breeding. In World Crops: Cool Season Food Legumes, Edited by
R] . Summerfield. Kluwer Academic, Dordrecht, The Netherlands, 381

Bush C.S., Caroutte CA, Amundson CH. and Olson NP. 1983. Manufacture of colby
and brick cheeses from ultrafiltered milk. J. of Dairy Science. 66, 415—421

Carpenter KJ. 1984. The protein nutritional quality of meat and poultry products:
scientific basis for regulation. American J. Clinical Nutrition; 40 (3) 671-742

Chang H.I., Catagnani G.L. and Swaisgood HE. 1990. Protein digestibility of alkali- and
fi'uctose-treated protein by rat true digestibility and by the immobilized digestive
enzyme assay system. J. Agricultural and Food Chemistry. 38, 1016-1018

Chang KC. and Satterlee L.D. 1979. Chemical, nutritional and microbiological quality of
protein concentrate from culled dry beans. J. Food Science, 44, 1589

Chang KC. and Satterlee L.D. 1981. Isolation and characterization of major protein fi'om
Great Northern bean. J. Food Science. 46, 1368

200

Chapman H11, Bines V.E., Glover RA. and Skudder R]. 1974. Use of milk concentrated
by ultrafiltration for making hard cheese, sofi cheese, and yogurt. J. of Dairy
Science & Technology, 27, (3) 151-155

Cheryan M. 1979. Improvement of fimctional properties of protein. Cereal Foods World.
24 (7) 272-273

Cheryan M. 1986. Ultrafiltration Handbook. Technomic Publishing Co., Inc., Lancaster

Cheryan M. and Chiang EH. 1984. performance and fouling behavior of hollow fibre and
spiral wound ultrafiltration modules processing skimmilk. In Engineering Sciences
in the Food Industry Vol 1: Engineering and Food. Edited by B.M McKenna,
Elsevier applied Science Publishers, New York

Cheryan M. and Kuo KR 1984. Hollow fibers and spiral wound modules for
ultrafiltration of whey: energy consumption and performance. J. of Dairy Science.
67, 1406-1413

Cheryan M., and Merin U. 1979. A study of the fouling phenomenon during ultrafiltration
of Cottage cheese whey. Abstracts of Papers, American Chemical Society; 178 (1)
COLL 128

Cofi‘ey D.G. 1985. Studies of Phytohemagglutinin. The Lectin of Phaseolus Vulgaris. PhD
Dissertation, Michigan State University, East Lansing

Coffey D.G., Uebersax M.A., Hosfield G.L. and Bennink MR. 1992. Stability of red
kidney bean lectin. J. Food Biochemistry. 16 (1) 43-57

Cofi‘man CW. and Garcia V.V. 1977. Isolation and functional characterization of a
protein isolate fi'om mungbean flour. In 1" International Mungbean Symposium,
69-73.

Colonna P., Gallant D., and Mercier C. 1980. Pisum sativum and Viciafaba
carbohydrates: studies of fi'actions obtained after dry and wet protein extraction
processes. J. Food Science. 45, 1629-1636

Dagom-Scaviner C., Gueguen J. and Laefebvre J. 1987. Emulsifying properties of pea
globulins as related to their adsorption behaviors. J. Food Science, 52, 335-341

Damodaran S. 1988. Refolding of thermally unfolded soy proteins during the cooling
regime of the gelatin process: Efl‘ect on gelation. J. of Agricultural and Food
Chemistry, 36, 262

DeBoer R. and Nooy P.F.C. 1980. Concentration of raw whole milk by reverse osmosis
and its influence on fat globules. Desalination. 35, 201-211

201

Derbyshire E. and Boulter D. 1976. Isolation of legumin like protein fiom Phaseolus
aureus and Phaseolus vulgaris L. Phytochemistry, 15, 411

Deshpande SS. 1985. Investigations on Dry Bean (Phaseolus vulgaris L.):
Microstructure, Processing and Antinutrients, PhD thesis, University of Illinois,
Urbana-Champaign

Deshpande SS. and Damodaran S. 1990. Food Legumes: Chemistry and Technology,
Advanced Cereal Science Technology, 10, 147

Dhurandhar NV. and Chang KC. 1990. Effect of cooking on firmness, trypsin inhibitors,
lectins and cystine/cysteine content of navy and red kidney beans (Phaseolus
vulgaris), J. Food Science, 55, 470

Dimes L.E., Haard N.F., Dong F.M., Rasco B.A., Forster I.P., Fairgrieve W.T., Amdt R.,
Hardy R.W., Barrows F.T. and Higgs D.A. Estimation of protein digestibility. II.
In vitro assay of protein in salmonid feeds. Comparative biochemistry and
physiology. A, Comparative physiology. 108A (2/3) 363-370. Pergamon Press
Ltd. Oxford

Dominguez H.D., Sema-Saldivar S.O., Gomez M.H. and Rooney L.W. 1993. Production
and nutritional value of weaning foods fiom mixtures of pearl millet and cowpeas.
Cereal Chemistry. 70(1) 14-18

Doughty J. and Walker A. 1982. Legumes in Human Nutrition. Food & Nutrition Paper
20, Food and Agriculture Organization (F AO), Rome, Italy, 27

Duranti M., Restani P., Poniatowska M. and Cerletti P. 1981. The seed globulins of
Lupinus albus Lupine. Phytochemistry, 20 (9) 2071-2075

Earle PR. and Jones 0. 1962. Analysis of seed sample fi'om 113 plant families. Econ Bot,
16, 221

Eggum, B. O., Hansen, 1. 81 Larsen, T. 1989. Protein quality and digestible energy of
selected foods determined in balance trials with rats. Plant Foods Human
Nutrition. 39, 13-21

Eicher NJ. and Satterlee L.D. 1988. Nutritional quality of Great Northern bean proteins
processed at varying pH, J. Food Science, 53, 1139

E1 SN. and Kavas A. 1996. Determination of protein quality of rainbow trout (Salmo
irideus) by in vitro protein digestibility-corrected amino acid score (PDCAAS).
Food-Chemistry. 55 (3) 221-223

Elkowicz K. and Sosulski F.W. 1982. Antinutritive factors in eleven legumes and their air-
classified protein and starch fiactions - Trypsin inhibitors, hemagglutinating

202

activity, saponin, phytic acid, quantitative composition. J. Food Science, 47 (4) p.
1301-1304.

Engel RW. 1978. The importance of legumes as a protein source in Asian diets. Proc.
First International Mungbean Symposium, 35-39, AVRDC, Shanhua, Taiwan

Erdogdu N., Czuchajowska Z. and Pomeranz Y. 1995. Wheat flour and defatted milk
fi‘actions characterized by differential scanning calorimetry. II. DSC of interaction
products. Cereal Chemistry. 72 (1) 76-79

Emstrom CA. and Anis SK. 1985. Properties of products fi‘om ultrafiltered whole milk.
In Proceedings , IDF Seminar, Atlanta, GA. International Dairy Federation, 21-
30

Emstrom. CA. 1986. Uses of UF/RO membranes in the Dairy Foods Industry. National
Workshop on Research Opportunities for the Dairy Industry, Berkley, CA

Esaka M., Suzuki K. and Kubota K. 1987. Effects of microwave heating on lipoxygenase
and trypsin inhibitor activities, and water absorption of winged bean seeds, J.
Food Science, 52, 1738

Estevez A.M. and Luh BS. 1985. Chemical and physical characteristics of ready-to-eat
dry beans, J. Food Science, 50, 777

Evans RJ. and Kerr M.H. 1963. Extraction and precipitation of nitrogenous constituents
of dry mvy beans. J. Agricultural Food Chemistry, 11:26

Eykamp W. 1978. Fouling of membranes in food processing. AIChE Symposium Series;
74 (172) 233-235

Fan T.Y and Sosulski F.W. 1974. Dispersibility and isolation of protein fi'om legume
flours. J. Canadian Institute of Food Science and Technology, 7:256

FAO. 1970. Amino Acid content of foods and biological data on proteins. FAO, Rome

FAO. 1979. Grain legumes: processing and storage problems. Food and Nutrition
Bulletin; 1 (2) 1-7

FAO. 1990. Production Yearbook, Food and Agricultural Organization (F AO), Rome
FAO. 1996. Production Yearbook, Food and Agricultural Organization (F AO), Rome

FAQ/WHO. 1973. Energy and Protein Requirements, Report of a joint FAO/WHO ad hoc
expert Committee. World Health Organization Technical Report. Ser. 522, WHO,
Geneva

203

FAO/WHO. 1990. Protein quality evaluation : report of the Joint FAO/WHO Expert
Consultation on Protein Quality Evaluation. Food and Agriculture Organization of
the United Nations, Rome

F arias Y G.A., Baranzini R AL. and Hoyos B J.M. 1995. Particle size, salt concentration,
and temperature effects on nitrogen extraction of chickpea (Cicer arietinum) by
response surface methodology. J. Food Processing and Preservation, 19:331

Fleming SE. 1980. Measurement of hydrogen production in the rat as an indicator of
flatulence activity. J. Food Science., 45, 1012

Fleming SE. 1981. Flatulence activity of the smooth-seeded field pea as indicated by

hydrogen production in the rat Distribution of flatulence-causing components
within the seeds. J. Food Science, 47 (1) 12-15

Flink J. and Christiansen I. 1973. The production of a protein isolate from Vicia faba.
[Broadbeans]. Lebensmittel Wissenschaft und Technologie, 6(3) 102-106

Gao L., Beveridge T., and Reid CA. 1997. Effects of processing and packaging
conditions on haze formation in apple juices. Lebensmittel Wissenschaft und
Technologie; 30(1) 23-29

Garoutte C. 1983. Whey ultrafiltration - efl‘ect of pore size on flux and yield Process
Biochemistry. 18, 4, 2-4, 12

Gatehouse A.M.R 1984. Antinutritional proteins in plants. Developments in food
proteins. 3. Edited by B. Hudson. Elsevier Applied Science Publishers, London

Gatehouse J .A., Croy R.R.D., McIntosh R., Paul C. and Boulter D. 1980. Quantitative
and Qualitative variation in the storage proteins of material fiom the EEC point
field bean test. In Vicia Faba: Feeding value, processing and viruses, Edited by
DA. Bond, 173-190

German J .B. and Phillips LG. 1989. Structures of proteins important in foaming. In Food
Proteins: Structure and Functional Relationships, Edited by J .E. Kinsella and W.
Soucie. AOCS, Charnpaign, IL

Godbole S.A., Krishna T.G. and Bhatia,-C.R. 1994. Changes in protease inhibitory
activity from pigeon pea (Cajanus cajan (L) Mill sp) during seed development and
germination. J. Science of Food and Agriculture, 66 (4) 497-501

Gorinstein S., Zemser M. and Paredes-Lopez O. 1996. Structural stability of globulins. J.
Agricultural and Food Chemistry. 44 (1) 100-105

Graham DE and Phillips MC. 1976. The conformations of proteins at the air-water
interface and their role in stabilizing foams. In Foams. Edited by RJ. Akers,
Academic Press, New York

204

Gregory J .F.III. and Kirk J .R. 1981. The Bioavailability of Vitamin B6 in Foods. Nutrition
Reviews, 39 (1) 1-8

Gupta Y.P. 1982. Nutritive value of food legumes. In Chemistry and Biochemistry of
Food Legumes, Edited by S.K. Arora. Chp 7, Oxford and IBH, New Delhi, India

Guthrie HA. and Picciano M.F. 1995. Protein. In Human Nutrition, Mosby Year Book,
Inc., St. Louis, MO

Halling. P.J. 1981. Protein stabilized foams and emulsions. CRC Critical Review of Food
Science and Nutrition, 15, 155

Hang Y.D., Steinkraus KH. and Hacker L.R. 1970. Comparative studies on the nitrogen
solubility of mung beans, pea beans and red kidney beans. J. Food Science, 35:318

Harris D.A., Burns RA. and Ali R. 1988. Evaluation of infant formula protein quality:
comparison of in vitro with in vivo methods. J. Association Official Analytical
Chemists. 71 (2) 353-357

Hartman-GH Jr. 1979. Removal of phytate from soy protein J. of the American Oil
Chemists' Society; 56 (8) 731-735

Hermansson A.M. 1979. Methods of studying functional characteristics of vegetable
proteins. J. American Oil Chemists Society. 56, 272

Hill D.W., Walters F.H., Wilson TD. and Stuart JD. 1979. HPLC dtermination of amino
acids in the picomole range. Analytical Chemistry. 51, 1338

Ho W.S.W. and Sirkar KK. 1992. Membrane Handbook. Van Nostrand Reinhold, New
York, NY

Honavar P.M., Shih C.V. and Liener LE. 1962. Inhibition of the growth of rats by purified
haemagglutinin fiactions isolated fi'om (Phaseolus vulgaris). J. Nutrition, 77, 109

Hsu H.W, Vavak D.L, Satterlee L.D. and Miller GA. 1977. A multienzyme technique for
estimating protein digestibility. J. Food Science. 42 (5) 1269-1273

Igbedioh S.O., Olugbemi K.T. and Akpapunam M.A. Effects of processing methods on
phytic acid level and some constituents in barnbara groundnut (Vigna subterranea)
and pigeon pea (Cajanus cajan). Food Chemistry, 50 (2) 147-151

Jackson J.C., Ng P.K.W., Uebersax M.A., Bennink M.R., and Hosfield G.L. 1997.
Optimization of the aqueous extraction of proteins fi'om cowpeas (Vigna
unguiculata). Presented at the American Association of Cereal Chemists (AACC)
Annual Meeting, San Diego, October.

205

Jaffe W.G. and Vega Lette CL. 1968. Heat-labile growth inhibiting factors in beans
(Phaseolus vulgaris). J. Nutrition. 94, 203-210

Juarez MS, and Paredes-Lopez O. 1994. Studies on jicama juice processing. Plant Foods
for Human Nutrition; 46 (2) 127-131

Kadarn SS. and Salunkhe D.K. 1989. Production, Distribution and Consumption. In CRC
Handbook of World Legumes: Nutritional Chemistry, Processing Technology and
Utilization. Vol 1. Boca Raton, F1

Kaiser J .M. and Glatz CE. 1988. Use of precipitation to alter flux and fouling
performance in cheese whey ultrafiltration. Biotechnology Progress; 4 (4) 242-247

Kanamori M., Ikeuchi T., Ibuki F., Kotaru M. and Kan KK. 1982. Amino acid
composition of protein fiactions extracted from Phaseolus beans and the field bean
(V icia faba L.). J. Food Science 47 (6) 1991-1994

Kapoor A.C. and Gupta Y.P. 1977. Distribution of nutrients in the anatomical parts of
soybean and difierent phosphorus compounds in seed and its protein fi'actions.
Indian J. Nutrition Diet., 41, 100

Kay EB. 1979. Haricot Bean. In Food Legumes, Edited by DR. Kay, Tropical Products
Institute, London, UK

Kelly J .F . 1973. Increasing protein quantity and quality. In Nutritional Improvement of
Legumes by Breeding. Edited by M. Milner. Protein Advisory Group, New York

Khan ILL, Gatehouse J.A., and Boulter D. 1980. The seed proteins of cowpea. J.
Experimental Botany, 31 (125) 1599-1611

Kim K]. 1989. The effect of surface pretreatment on the performance of ultrafiltration
membranes. Dissertation Abstracts International, B; 50 (1) 271

Kim SH. and Kinsella J.B. 1987. Surface active properties of food proteins. Effects of
reduction of disulfide bonds on film properties and foam stability of glycerin. J.
Food Science, 52, 128

Kim Y.S., Brophy El, and Nicholson J .A. 1976. Rat Intestinal brush border peptidases.
J. Biological Chemistry, 251 3206 — 3212

King T.P., Pusztai A., and Clarke E.M.W. 1980. Kidney bean (Phaseolus vulgaris) lectin
induced lesions in rat small intestine: 3. Ultrasound studies. J. Comp Path. 92, 357
— 373

Kinsella J .E. 1976. Functional properties of proteins in foods: a survey. Critical Reviews
in Food Science and Nutrition, 12, 219

206

Kinsella J.E. 1979. Functional properties of soy proteins. . J. American Oil Chemists
Society. 56, 242

Kinsella J .E. 1982. Relationships between structure and functional properties of Food
Proteins. In Food Proteins, Edited by PF. Fox and J .J London. Elsevier Applied
Sciences, London, 51-103

Kinsella J .E. and Phillips LG. 1989. Structure function relationships in food proteins: Film
and foaming behavior. In Food Proteins: Structure and Functional Relationships.
Edited by J .E. Kinsella and W. Soucie, AOCS, Champaign, IL

Kinsella J .E. and Srinivasan D. 1981. Nutritioml, Chemical and Physical criteria affecting
the use and acceptability of proteins in foods. In Criteria of Food Acceptance.
Edited by J. Solms and R.L Hall. Forster Verlag, Switzerland

Ko W.C., Chen W.J. and Lai TH. 1994. Recovery and functional properties of protein
fi'om the wastewater of mungbean starch processing by ultrafiltration. J. Japanese
Society of Food Science and Technology [Nippon-Shokuhin-Kogyo-Gakkaishi];
41 (9) 633-638

Ko W.C., Chen W.J., and Lai TH. 1994. Recovery and functional properties of protein
fiom the wastewater of mung bean starch processing by ultrafiltration. J. Japanese
Society of Food Science and Technology, 41 (9) 633-638

Kohnhorst A.L., Smith D.M., Uebersax M.A.; Bennink M.R. 1990a. Compositional,
nutritional and functional properties of meals, flours and concentrates fi'om navy
and kidney beans (Phaseolus vulgaris). J. Food Quality. 13 (6) 43 5-446

Kohnhorst A.L., Uebersax M.A. and Zabik M.E. 1990b. Production and functional
characteristics of protein concentrates. J. American Oil Chemists Society. 67 (5)
285-292

Kosikowski F. 1986. New cheese-making procedures utilizing ultrafiltration. Food
Technology, 40(6) 71-77, 156

Koutake M., Uchida Y., Sato T., Shirnoda K., Kirnura T., Sagara Y., Watanabe A., Nakao
S. 1981. Fundamentals and applications of surface phenomena associated with
fouling and cleaning in food processing. Patent. Alnarp, Sweden; Lund University

Koutake M.; Uchida Y.; Sato T.; Shirnoda K.; Kirnura T.; Sagara Y.; Watanabe A.; and
Nakao S. 1987. Study of membrane fouling in ultrafiltration of skimmilk. Reports
of Research Laboratory, Snow Brand Milk Products C o.; No. 85, 232-234

Krober CA. 1968. Nutritional quality of pulses. J. Post Graduate SchooL Indian
Agricultural Research Institute, 6, 157

207

Kwon K.S., Bae D., Park K.H., and Rhee KC. 1996. Aqueous extraction and membrane

techniques improve coconut protein concentrate functionality. J. Food Science; 61
(4) 753-756

Laemmli U.K. 1970. Cleavage of structural proteins during the assembly of the head of the
bacteriophage T4. Nature (London), 227, 680-685

Lah CL. and Cheryan M. 1980. Protein solubility characteristics of an ultrafiltered full-fat
soybean product. J. of A gricultural and Food Chemistry; 28 (5) 911-916

Lakshminarayanaiah N. 1969. Transport phenomena in membranes. Academic Press, New
York

Lawhon J.T., Hensley D.W., Mizukoshi M., and Mulsow D. 1979. Alternate processes for
use in soy protein isolation by industrial ultrafiltration membranes. J. Food
Science, 44:213

Lawhon J.T., Manak LJ. and Lusas E.W. 1982. Co-processing soy proteins and milk
proteins with commercial ultrafiltration membranes. J. Food Science; 47 (3) 800-
805

Lawhon J.T., Manak L.J., Rhee K.C., Rhee KS. and Lusas E.W. 1981. Combining
aqueous extraction and membrane isolation techniques to recover protein and oil
from soybeans. J. Food Science; 46 (3) 912-916, 919

Lawhon-1T. 1983. Method for processing protein fiom nonbinding oilseed by
ultrafiltration and solubilization. United-States-Patent. US 4 420 425 PA: USA,
Texas A&M University System

Leavitt RD, Felsted R.L. and Bachur NR. 1977. Biological and biochemical properties
of Phaseolus vulgaris isolectins. J. Biological Chemistry 252 (9) 2961 - 2966

Lee CR. 1977. Some factors afi‘ecting permeation and fouling of selected ultrafiltration
membranes. Dissertation Abstracts International, B; 38 (2) 574: Order No. 77-
17107, 225 pp.

Levine AS. 1979. The relationship of diet to intestinal gas. Contemporary Nutrition, 4, 7

Lewis MJ. 1982. Concentration of proteins by ultrafiltration. Development of Food
Proteins. 1, 91-130. Applied Science Publishers, Ltd. Barking, England

Liao H.J., Okechukwu P.E., Damodaran S. and Rao M.A. Rheological and calorimetric
properties of heated corn starch-soybean protein isolate dispersions. J. Texture
Studies. 27 (4) 403-418.

Liener LE. 1958. Inactivation studies on the soybean hemagglutinin. J. Biological
Chemistry, 233, 401

208

Liener I .E. 1962. Toxic factors in legumes and their elimination. American J. Clinical
Nutrition, 11, 281

Liener LE. 1976. Legume toxins in relation to protein digestibility: a review, J. Food
Science, 41, 1076 '

Liener LE. 1982. Toxic constitutents in legumes, In Chemistry and Biochemistry of
Legumes. Edited by S.K. Arora. Oxford and IBH, p217

Lin M.J.Y., Humbert ES. and Sosulski F. 1974. Certain firnctional properties of
sunflower meal products. J. Food Science, 39, 368

Liu F.K., Nie Y.H., and Shen B.Y. 1989. Manufacturing soy protein isolate by
ultrafiltration. In Proceedings of the world congress on vegetable protein
utilization in human foods and anirml feedstufl‘s. Conference. Singapore, October,
84-90

Luse RA. and Rachie KC. 1979. Seed protein improvement in tropical food legumes.
Seed Protein Improvement in Cereals and Grain Legumes 11, 87-104; IAEA,
Vienna

Luss G. 1985a. Membrane Fouling - Field Experience and Observations. The Second
International Conference on Fouling and Cleaning in Food Processing. The
University of Wisconsin, Madison, WI

Luss G. 1985b. Cleaning up those fouled membranes. Water Technology. 11, 28-31
MacRitchie F. 1978. Proteins at interfaces. Advanced Protein Chemistry, 32, 283

Makower R.U. 1969. Changes in phytic acid and acid-soluble phosphorus in maturing
pinto beans. J. Science of Food and Agriculture, 20 (2) 82-84

Maneepun S., Luh BS. and Rucker R.D. Amino acid composition and biological quality
of lima bean protein. J. Food Science. 39 (1) 171-174

Martensson P. 1980. Variation in legumin: vicilin ratio between and within cultivars of
Vicia faba L. var. minor broadbeans. Vicia Faba Feed Value Process Virus Proc.
The Hague, Martinus Nijhofl; (3) 159-172

Martinez E.N. and Anon MC. 1996. Composition and structural characterization of
amaranth protein isolates. An electrophoretic and calorimetric study. J.
Agricultural and Food Chemistry. 44 (9) 2523-2530

Matsumura Y. and Mori T. 1996. Gelation. In Methods of Testing Protein Functionality.
Edited by GM. Hall. Blackie Academic & Professional, London

209

McDonough F .E., Sarwar G., Steinke F.H., Slump P., Garcia S. and Boisen S. 1990. In
vitro assay for protein digestibility: interlaboratory study. J. Association Official
Analytical Chemists. 73 (4) 622-625

McWatters K.H. and Cherry J.P. 1981. Emulsification: Vegetable proteins. In Protein
Functionality in Foods. Edited by J .P. Cherry. ACS Symposium Series 147, 217-
242, American Chemical Society (ACS), Washington, DC

Merin U. and Cheryan M. 1980. Factors affecting the mechanism of flux decline during
ultrafiltration of Cottage cheese whey. J. Food Processing and Preservation; 4 (3)
183-198

Mohr C.M., Leeper S.A., Engelgau DE. and Charboneau BL. 1989. Membrane
Applications and Research in Food Processing. Noyes Data Corporation, Park
Ridge, NJ

Mosse J. and Pemollet J.C. 1982. Storage proteins of legume seeds. In Chemistry and
Biochemistry of Food Legumes, Edited by S.K. Arora. Chp 3 Oxford and IBH,
New Delhi, India

Murillo B., Cabezas T. and Bressani R. 1974. Influence of calorie density on the protein
utilization of diets based on maize and beans. Archives LatinoAmerican Nutrition,
24 (2) 223-241

Murray E.D., Arntfield SD. and Ismond M.A.H. 1985. The influence of processing
parameters on food protein functionality. 11. Factors affecting thermal properties as
analyzed by differential scanning calorimetry. J. Canadian Institute of Food
Science Technology. 18 (2) 158-162

Murray E.D., Myers CD. and Barker L.D. 1978. Protein product and process for
preparing same. Canadian Patent 1,028,552

Murray E.D., Myers C.D., Barker L.D. and Maurice T.J. 1981. Functional attributes of
proteins - a noncovalent approach to processing and utilizing plant proteins. In
Utilization of Protein Resources, Food and Nutrition Press, Westport, CT

N.A.S. (National Academy of Sciences) of the US 1980. Recommended Dietary
Allowances. 9th Ed., Food & Nutrition Board, National Research Council,

Washington, DC

Nagano T., Mori H. and Nishinari K. 1994. Rheological properties and conformational
states of beta-conglycinin gels at acidic pH. Biopolym'ers. 34 (2) 293-298

NFPA (National Food Processors Association). 1993. Membrane Filtration Handbook /
Selection Guide: A guide on membrane filtration technology for the food
processing industry, Dublin, CA

210

Ng P.K.W. 1996. SDS PAGE Laboratory Methods — Preparation of Blakesley solution.
Department of Food Science and Human Nutrition, Michigan State University,
East Lansing, MI

Nilsson IL 1988. Fouling of an ultrafiltation membrane by a dissolved whey protein
concentrate and some whey proteins. J. of Membrane Science; 36, 147-160

Nisbet T.J., Thorn T.M. and Wood P.W. 1981. Observations on the fouling of
polysulphone ultrafiltration membranes by acid whey. New Zealand J. of Dairy
Science and Technology; 16(2) 113-120

NRC (National Research Council). 1995. Nutrient requirements of Laboratory Animals.
4th Edition, 11-79. National Academy of Sciences, Washington, DC.

Occena LG. 1994. Processing strategies to improve dry bean (Phaseolus vulgaris)
digestibility and food utilization. Ph.D. Dissertation. Michigan State University,
East Lansing, MI

Okezie B0. and Bello AB. 1988. Physicochemical and functional properties of winged
bean flour and isolate compared with soy isolate. J. Food Science 53 (2) 450-454

Omosaiye O. and Cheryan M. 1979. Low-phytate, full-fat soy protein product by
ultrafiltration of aqueous extracts of whole soybeans. Cereal Chemistry; 56 (2) 58-
62

Osborne T.B. and Campbell GP. 1989. Proteins of the Pea. J. American Chemical
Society, 20 348

Parkin-MF. 1975. The cleaning of tubular ultrafiltration membranes. Food Technology in
New Zealand; 10 (10) 12-13, 15

Patel K.M., Bedford CL. and Youngs CW. 1980. Amino acid and mineral profile of air-
classified navy bean flour fiactions. Cereal Chemistry. 57 (2) 123-125.

Pearce K.N. and Kinsella J.B. 1978. Emulsifying properties of proteins: Evaluation of a
turbidirnetric technique. J. of Agricultural and Food Chemistry, 26, 716

Pedersen B. and Eggum 8.0. 1983. Prediction of protein digestibility by an in vitro
enzymatic pH-stat procedure. J. Animal Physiology and Animal Nutrition. 49
(4/5) 265-277

Pellett PL. 1978. Protein quality evaluation revisited. Food Technology. 32 (5) 60, 62,
64-66, 68, 70-72, 74, 76

Pellett PL. and Young V.R. 1981. Nutritional Evaluation of Protein Foods. Report of a
working group representing the International Union of Nutritional Sciences and
the United Nations University. United Nations University Press, Tokyo, Japan

211

Peri C., Pompei C. and Roddi F. 1973. Process optimization in skim milk protein
recovery and purification by ultrafiltration. J. Food Science, 38: 135

Pemollet J .G. 1978. Protein bodies of seeds, ultrastructure, biochemistry, biosynthesis and
degradation. Phytochemistry, 17, 147 3

Perutz M.F. 1978. Electrostatic effects in proteins. Science. 201, 1187

Petersen R.J. 1985. Dealing successfirlly with fouling of reverse osmosis membranes.
Membrane Technology / Planning Conference. Business Communications Co.,
Inc., Stamford, CT

Petruccelli S. and Anon M.C. 1994. Relationship between the method of obtention of the
structure and functional properties of soy protein isolates. 2. Surface properties. J.
Agricultural and Food Chemistry, 42, 2170-2176

Pomeranz Y. 1991. Functional Properties of Food Components. 2"" Edition. Academic
Press Inc, CA

Powrie W.D., Adams M.W. and Pflug U. 1960. Chemical anatomical and histochemical
studies on the navy bean seed. J. Agronomy. 52, 163

Prins A. 1988. Principles of foam stability. In Advances in Food Emulsions and Foams.
Edited by E. Dickinson and G. Stainsby. Elsevier Applied Science, NY

Pusztai A. and Palmer R. Nutritional evaluation of kidney beans (Phaseolus vulgaris): the
toxic principle [Rats]. J. Science of Food and Agriculture; 28 (7) 620-623

Pusztai A. and Watt W.B. 1970. The isolation and characterization of a major antigenic
and non-haemagglutinating glycoprotein from Phaseolus vulgaris. Biochim
Biophys Acta, June, 207 (3) 413-431

Ramachandra Rao H.G., Lewis. M.J., and Grandison AS. 1994. Effect of soluble calcium
of milk on fouling of ultrafiltration membranes. J. Science of Food and
Agriculture; 65 (2) 249-256

Rasco B.A. 1994. Protein quality tests. In Introduction to the Chemical Analysis of Foods.
Edited by SS. Nielsen. Chapman & Hall, New York

Redaelli R; Ng PKW and Ward RW. 1997. Electrophoretic characterization of storage
proteins of 37 Chinese landraces of wheat. J. of Genetics & Breeding, 51 (3) 239-
249, 26 ref.

Reddy N.R., Pierson M.D., Sathe SK, and Salunkhe D.K. 1984. Chemical, nutritional
and physiological aspects of dry bean carbohydrates - A Review. Food Chemistry,
13, 25

212

Regenstein J.M. 1988. Are comminuted meat products emulsions or a gel matrix?
Presented at the American Oil Chemists Society Meeting, Phoenix AZ

Reichert RD. and Youngs CG. 1978. Nature of the residual protein associated with
starch fractions from air classified field peas. Cereal Chemistry. 55, 469-480

Rockland L.B., Gardiner BL. and Pieczarka D. 1969. Stimulation of gas production and
growth of Clostridium perfiingens type A (No 3624) by legumes. J. Food Science,
34, 411

Romero-Baranzini A.L., Yanez-Farias GA. and Barron-Hoyos,-J.M. 1995. A high
protein product fiom chickpeas (Cicer arietinum L.) by ultrafiltration. Preparation
and fimctional properties. J. Food Processing and Preservation. 19 (5) 319-329.

Rouanet J.M. and Besancon P. 1979. Effects of an extract of phytohemagglutinins on the
growth, digestibility of nitrogen and invertase activity and (sodium+-potassium+
ions)-ATPase of the intestinal mucosa in the rat. Ann Nutr Aliment Ann Nutr
Food. 33 (3) 405-416

Rubio L.A., Grant G., Dewey P., Brown D. Armand M., Bardocz S. and Pusztai A 1998.
Nutritional utilization by rats of chickpea (Cicer arietinum) meal and its isolated
globulin proteins is poorer than that of defatted soybean or lactalbumin. J.
Nutrition. 128 (6) 1042-1047

Salunkhe D.K. and Kadarn SS. 1989. Introduction. In CRC Handbook of World
Legumes: Nutritional Chemistry, Processing Technology and Utilization. Vol 1.
Boca Raton, F1

Sandholrn M. and Scott ML. 1979. Binding of lipase, amylase and protease to intestinal
epithelium as affected by lectins in-vitro. Acta. Vet. Scand. 20, 329 - 342

Sarwar G. 1997. The protein digestibility-corrected amino acid score method
overestimates quality of proteins containing antinutritional factors and of poorly
digestible proteins supplemented with limiting amino acids in rats. J. Nutrition. 127
(5) 758-764

Sarwar G. and Peace R. W. 1986. Comparisons between true digestibility of total nitrogen
and limiting amino acids in vegetable proteins. J. Nutrition. 116, 1172-1184

Sathe SK. and Sahmkhe D.K. 1981. Functional properties of the Great Northern bean
(Phaseoulus vulgaris L.) proteins: emulsion, foaming, viscosity and gelatin
properties. J. Food Science. 46, 71

Sathe SK. and Salunkhe D.K. 1981. Preparation and utilization of protein concentrates
and isolates for nutritional and functional improvement of foods. J. Food Quality,
4:233

213

Sathe S.K., Deshpande SS, and Salunkhe D.K. 1982. Functional properties of winged
bean (Psophocarpus tetragonolobus L.) proteins. J. Food Science 47, 503-509

Sathe S.K., Deshpande SS, and Salunkhe D.K. 1984. Dry beans of Phaseolus. A review.
Part 1. Chemical Composition: Proteins. Critical Review Food Science and
Nutrition. 20, 1

Satterlee L.D., Bembers M. and Kendrick JG. 1975. Functional properties of Great
Northern bean (Phaseoulus vulgaris) protein isolates. J. Food Science. 40, 81

Satterlee L.D., Kendrick J.G., Jewell D.K. and Brown W.D. 1981. Estimating apparent
protein digestibility from in vitro assays. In Protein Quality in Humans. AVI
Publishing Co. 316-339. Westport, Conn,

Satterlee L.D., Marshall HF. and Tennyson J.M. 1979. Measuring protein quality. .1.
American Oil Chemists Society. 56, 103-109

Sauvaire Y.D., Baccou J.C.F. and Kobrehel,-K. 1984. Solubilization and characterization
of fenugreek seed proteins (Trigonella foenum-graecum). J. Agricultural and
Food Chemistry. 32 (1)41-47

Schmidt RH. 1981. Gelation and coagulation. In Protein Functionality in Foods. Edited
by J.P. Cherry. ACS Symposium Series 147, 131-148, American Chemical Society
(AC S), Washington, DC

Sefa-Dedeh S. and Stanley D. 1979. Cowpea proteins. 1. Use of Response Surface
Methodology in predicting Cowpea (Vigna unguiculata) protein extractability. J.
Agricultural and Food Chemistry, 27: 1238

Siegel A. and Fawcett B. 1976. Food legume processing and utilization. International
Development Research Centre, IDRC-TS; 1f, 60 - 63. Ottawa, Canada.

Simbuerger S., Wolfschoon A.F., Kempter K., Rose M., and Marder U. 1997. Acidified
ultrafiltration concentrates as raw material for the production of processed cheese.
European-Patent-Application (EP075563 0A1)

Singh S., Singh H.D. and Sikka KC. 1968. Distribution of nutrients in anatomical parts of
common Indian pulses. Cereal Chemistry, 45, 13

Singh U. and Jambunathan R. 1982. Distribution of seed protein fi'actions and amino acids
in difl‘erent anatomical parts of chickpea and pigeonpea. Plant Foods in Human
Nutrition, 31, 347

Smith BR. and MacBean RD. 1978. Fouling in reverse osmosis of whey. Australian .l. of
Dairy Technology, 33 (2) 57-62

214

Smith D.M. 1994. Protein Separation and Characterization Procedures. In Introduction to
the Chemical Analysis of Foods. Edited by S. S. Nielsen, Chapman & Hall, NY

Sosulski F.W. and Young CG. 1979. Yield and firnctional properties of air-classified
protein and starch fiactions fiom eight legume flours. J. American Oil Chemists
Society (AOAC), 56:292-295

Spackman D.H., Stein W.H. and Moore S. 1958. Automatic recording apparatus for use
in the chromatography of amino acids. Analytical Chemistry. 30, 1190

Stanton W.R., Doughty J., Orraca-Tettech R., and Steele W. 1966. Grain Legumes in
Afiica, Food and Agriculture Organization (FAO), Rome, Italy, 167

Suelter CH. 1985. A practical guide to enzymology. Biochemistry, John Wiley & Sons,
New York

Sumner A.K., Nielsen M.A. and Youngs CG. 1981. Production and evaluation of pea
protein isolate. J. Food Science. 46(2) 364-366, 372

Swift C.E., Lockett C. and Frayar AJ. 1961. Comrninuted meat emulsions - capacity of
meats for emulsifying fat. Food Technology, 15, 486

Tapper LI. and Ritchey SJ. 1981. Nutritional value of dry beans. Presented at the 41"
Annual Meeting ofIFI‘, Atlanta, June

Timrner J .M.K. 1997. Whey protein concentrates with non-traditional compositions.
European Dairy Magazine; No. 3, 47-49

Tjahjadi C., Lin S. and Breene W.M. 1988. Isolation and characterization of adzuki bean
(V igna angularis cv. Takara) proteins. J. Food-Science. 53 (5) 1438-1443

Tong P.S., Barbano D.M., and Rudan M.A 1988. Characterization of proteinaceous
membrane foulants and flux decline during the early stages of whole milk
ultrafiltration. J. of Dairy Science; 71 (3) 604-612

Tyler R.T., Youngs C.G., and Sosulski F.W. 1981. Air classification of legumes. I.
Separation efliciency, yield, and composition of the starch and protein fiactions.
Cereal Chemistry. 58, 144

Tzeng Y., Diosady LL. and Rubin LJ. 1990. Production of Canola protein materials by
alkaline extraction, precipitation, and membrane processing. J. Food Science,
:1147 .

Tzeng Y.M., Diosady LL, and Rubin L.J. 1988. Preparation of rapeseed protein isolates
using ultrafiltration, precipitation and diafiltration. J Canadian Institute Food
Science & Technology 21, (4) 419-424

215

Tzeng Y.M., Diosady L.L., and Rubin L.J. 1988. Preparation of rapeseed protein isolates
using ultrafiltration, precipitation and diafiltration. J Canadian Institute Food
Science Technology. 21, 4, 419-424

Uebersax M.A. and Occena LG. 1991. Composition and nutritive value of dry edible
beans: commercial and world food relief applications. Michigan Dry Bean Digest,
15 (5) 3-12, 28 ’

Ulloa J .A.,Valencia ME, and Garcia Z.H. 1988. Protein concentrate fiom chickpea:
nutritive value of a protein concentrate from chickpea (Cicer arietinum) obtained
by ultrafiltration and its potential use in an infant formula .1. Food Science; 53 (5)
1 3 96-1 398

UN University, World Hunger Programme. 1979. protein-energy requirements under -
conditions prevailing in developing countries: Current knowledges and research
needs. Food Nutrition Bulletin. Suppl. 1. UNU, Tokyo

Valdebouze P., Bergeron E., Gaborit T. and Delort-Laval J. 1980. Content and
distribution of trypsin inhibitors and hemagglutinins antinutritional factors.
Canadian J. Plant Science, 60(2) 695-701

Velicangil O. and Howell J.A. 1979. Protein ultrafiltration: theory of membrane fouling
and its treatment with immobilized proteases. Abstracts of Papers, American
Chemical Society; 178 (1) COLL 76

Vetier C., Bennasar M., and Tarodo de la Fuente B. 1988. Study of the fouling of a
mineral microfiltration membrane using scanning electron micro scopy and
physicochemical analyses in the processing of milk. J. of Dairy Research; 55 (3)
381-400

Vose J.R., Basterrechea M.J., Gorin P.A.J., F inlayson A.J. and Youngs CG. 1976. Air
classification of field peas and horsebean flours: chemical studies of starch and
protein fi'actions. Cereal Chemistry. 53, 928-936

Vu-Huu Thanh and Shrbasaki K. 1976. Major proteins of soybean seeds. A
straightforward fi'actionation and their characterization. J. Agricultural Food
Chemistry, 24(6) 1117-1121

Walker AF. 1983. The estimation of protein quality. In Developments in Food Proteins 2.
Edited by B.J.F. Hudson. Applied Science Publishers, New York

Wang CH. and Damodaran S. 1991. Thermal gelation of globular proteins: influence of
protein conformation on gel strength. J. Agricultural and Food Chemistry. 39 (3)
433-438

Wang J. and Kinsella J.B. 1976. Functional properties of novel proteins: Alfalfa leaf
proteins. J. Food Science, 41, 286

216

Waniska RD. and Kinsella J .E. 1979. Foaming properties of proteins: Evaluation of a
column aeration apparatus using ovalbumin. J. Food Science, 44, 1398

Wilson A.B., King T.P., Clarke E.M.W. and Pusztai A. 1980. Kidney Bean (Phaseolus
vulgaris) lectin induced lesions in rat small intestine: 2. Microbiological Studies. J.
Comp. Pathol, 90, 597 — 602

Wolmk A., Elias LG. and Bressani R. 1981. Protein quality of vegetable proteins as
determined by traditional biological methods and rapid chemical assays. J.
Agricultural and Food Chemistry. 29 (5) 1063-1068

Wu D., Howell J.A., and Turner NM. 1991. A new method for modelling the tirne-
dependence of permeation flux in ultrafiltration. Food and Bioproducts
Processing; 69 (C2) 77-82

Xu L. and Diosady LL. 1994. Functional properties of chinese rapeseed protein isolates.
J. Food Science, 59(5) 1127-1130

Yadav NR. and Liener LE. 1977. Nutritional evaluation of dry-roasted navy bean flour
and mixtures with cereal proteins. In Nutritional improvement of food and feed
proteins. Edited by M. Friedman. Plenum Press, New York

Yasamatsu K., Sawada K, Moritaka S., Toda J., Wada T. and Ishi K. 1972. Whipping and
emulsifying properties of soybean products. J. Agricultural Biological Chemistry
36, 719-727

Yu Z.R, Tseng C.Y., and Tzeng WC. 1997. Modeling study on component distribution of
West Indian cherry juice in a microfiltration and ultrafiltration system. J. Chinese
Agricultural Chemical Society; 35 (1) 52-60, (Abstract only)

Zimmerman G., Weissman S., and Yannai S. 1967. the distribution of protein, lysine and

methionine and antitryptic activity in cotyledons of some leguminous seeds. J.
Food Science. 32, 129

217

         

   

   

      

     

HICH

RM

16 STATE UNIV. LIBRARIES
ll" ll".llll"lllllllllllllllllWWHI
1 9301 8264758

32