Evaluation of Breadfruit (Artocarpus Altilis and A. Altilis X A. Mariannensis) As A

Home , Artocarpus mariannensis, Breadfruit

Evaluation of Breadfruit (Artocarpus altilis and A. altilis X A. mariannensis) as a

Dietary Protein Source

Ying Liu

B.Sc. (Hons.), Dalhousie University, 2012

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF

THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

The College of Graduate Studies

(Biochemistry and Molecular Biology)

THE UNIVERSITY OF BRITISH COLUMBIA

(Okanagan Campus)

April 2016

The undersigned certify that they have read, and recommend to the College of Graduate Studies for acceptance, a thesis entitled:

Evaluation of Breadfruit (Artocarpus altilis and A. altilis x A. mariannensis) as a Dietary Protein Source

Submitted by Ying Liu in partial fulfillment of the requirements of

The degree of Doctor of Philosophy .

Dr. Susan Murch, Chemistry Supervisor, Professor (please print name and faculty/school above the line)

Dr. Deanna Gibson, Biology Supervisory Committee Member, Professor (please print name and faculty/school in the line above)

Dr. Paula Brown, Biology Supervisory Committee Member, Professor (please print name and faculty/school in the line above)

Dr. Ian Walker, Biology University Examiner, Professor (please print name and faculty/school in the line above)

Dr. Nyree Zerega, Northwestern University External Examiner, Professor (please print name and university in the line above)

April 20, 2016 (Date submitted to Grad Studies)

Additional Committee Members include:

Dr. Fred Menard, Chemistry Please print name and faculty/school in the line above

Dr. Miranda Hart, Biology Please print name and faculty/school in the line above

Abstract

Protein malnutrition is a leading cause of child mortality and chronic development disorders in the tropics. This study evaluated the protein quality of breadfruit using three standards measures: (1) yield of fruit and production of protein,

(2) the protein content and amino acid profile, and (3) the protein digestibility/amino acid availability. I analyzed field data collected from 2006 to 2012 from 24 breadfruit cultivars in National Tropical Botanical Garden in Kauai, Hawaii. Based on a plant density of 100 trees/ha, breadfruit can yield 10-14 tons of fruit/ha. The commercial cultivar Ma’afala can produce up to 0.52 t/ha of protein. This is higher than wheat, rice, cassava, and potato. To determine the quality of breadfruit protein, I conducted an amino acid analysis of 49 breadfruit cultivars using high performance liquid chromatography with fluorescence detection and determined that the full spectrum of the essential amino acids were present in all cultivars. Breadfruit is especially rich in phenylalanine, leucine, isoleucine, and valine. Ma’afala contained a higher total essential amino acid content than many other staples including wheat, corn, rice, soybean, and yellow pea. To determine breadfruit protein digestibility, I developed a human digestion model, and found that Ma’afala protein was 10-25% more digestible than wheat protein. Ma’afala and wheat protein were compared using a human intestinal epithelial cell model. The results showed that Ma’afala induced similar impact on cell morphology and cytokine production in the model. To determine the overall effect on health, I compared breadfruit and wheat diet in a standard rodent model. The results show that the breadfruit diet did not induce any adverse effects on the overall health or growth of the mice. Breadfruit-fed mice gained more weight

iii overall than wheat-fed mice but the crumbly nature of the diet made accurate food intake measures impossible. Chemical and histological analysis did not reveal any significant difference between the two diet groups suggesting that breadfruit is not toxic. Together, these data demonstrate that breadfruit has potential as a staple protein source for reducing malnutrition in tropical countries and as an ingredient in nutritional products, in particular gluten-free food products, for world markets.

iv Preface

A version of the thesis has been prepared for a GRAS application to the U.S. Food and Drug Administration for the regulation of breadfruit as a safe food ingredient.

Ying Liu and Susan Murch (2015). Breadfruit (Artocarpus altilis and Artocarpus altilis x Artocarpus mariannensis) Generally Recognized As Safe (GRAS) notice.

USDA.

Some of the information in Chapter 1 has been published in the review article:

Christina Turi, Ying Liu, Diane Ragone, and Susan Murch. (2014). Breadfruit

(Artocarpus spp.): A traditional crop with the potential to prevent hunger and mitigate diabetes in the Tropics. Trends in Food Science and Technology, 45, 267-272.

I critically reviewed the literature, and shared this information with the author groups. I also summarized and wrote some significant sections for the paper.

A version of Chapter 2 has been published.

Ying Liu, Maxwell P. Jones, Diane Ragone, and Susan Murch. (2014). Crop productivity, yield and seasonality of breadfruit (Artocarpus spp., Moraceae). Fruits

69,345-361.

I conducted the data analysis, and wrote most of the original draft of the paper.

A version of Chapter 3 has been published.

Ying Liu, Diane Ragone, and Susan Murch. (2015). Breadfruit (Artocarpus altilis): A source of high quality protein for food security and novel food products.

Amino Acids, Online, DOI 10.1007/s00726-015-1914-4.

v I modified and developed the methods, conducted the majority of the experiment and analyzed all of the data. I also critically reviewed the literature, summarized and graphed all of the data and wrote the original draft of the paper.

A version of Chapter 4 is being prepared for publication.

A version of Chapter 5 is being prepared for publication.

Presentations:

Data from Chapter 2, 3, and 4 were presented at:

Liu Y., Ragone D., Gibson D., Murch S. Breadfruit: a source of high quality protein. Annual Biology Graduate Symposium. University of British Columbia,

Kelowna, BC. September 11th 2015.

Liu Y. and Murch S. Breadfruit: Breadfruit (Artocarpus altilis, Parkinson (Fos)) for Food Security and Novel Food Products. Botany 2015, Edmonton, AB, Canada,

July 25th – 29th 2015.

Liu Y. and Murch S. Love it or hate it, you gotta eat it. Three Minutes Thesis,

University of British Columbia, Kelowna, February 26th 2014

vi Table of Contents

Abstract ...... iii

Preface ...... v

Table of Contents ...... vii

List of Tables ...... xii

List of Figures ...... xv

List of Symbols and Abbreviations ...... xix

Acknowledgements ...... xxi

Dedication ...... xxiii

Chapter 1. Introduction ...... 1

1.1. History of breadfruit consumption ...... 1

1.2. Current uses and significance of breadfruit in the world ...... 4

1.3. Breadfruit botanical description ...... 5

1.4. Current breadfruit nutritional analysis ...... 6

1.4.1. Carbohydrates ...... 14

1.4.2. Carotenoids ...... 16

1.4.3. Minerals ...... 18

1.4.4. Vitamins ...... 21

1.4.5. Anti-nutrients ...... 21

1.5. Study rationale and objectives ...... 24

Chapter 2. Seasonality, Productivity, and Yield of Breadfruit ...... 25

2.1. Synopsis ...... 25

2.1.1. Objectives ...... 26

2.2. Materials and methods ...... 26

vii 2.2.1. Germplasm repository ...... 26

2.2.2. Data collection ...... 27

2.3. Results ...... 29

2.3.1. Time from planting to fruit production ...... 29

2.3.2. Breadfruit seasonality ...... 33

2.3.3. Breadfruit productivity/fruit number ...... 41

2.3.4. Breadfruit fruit weight...... 44

2.3.5. Breadfruit yield ...... 45

2.3.6. Tree size and growth ...... 46

2.3.7. Commercial breadfruit Ma’afala ...... 47

2.4. Discussion ...... 48

2.4.1. Breadfruit seasonality and production in Kauai ...... 48

2.4.2. Comparison of breadfruit productivity across locations ...... 49

2.4.3. Comparison of breadfruit seasonality across locations...... 54

2.4.4. Commercial breadfruit Ma’afala ...... 54

2.5. Summary ...... 56

Chapter 3. Amino Acid Analysis of Breadfruit Protein ...... 57

3.1. Synopsis ...... 57

3.1.1. Objectives ...... 58

3.2. Materials and method ...... 58

3.2.1. Breadfruit sample collection and preparation ...... 58

3.2.2. Reagents ...... 61

3.2.3. Preparation of amino acid samples ...... 61

3.2.4. Derivatization of amino acids ...... 62

3.2.5. HPLC analysis ...... 62

viii 3.2.6. Data analysis ...... 63

3.3. Results ...... 64

3.3.1. Quantification of essential amino acids ...... 64

3.3.2. Comparison of essential amino acids between commercial breadfruit and other staples ...... 67

3.4. Discussion ...... 72

3.5. Summary ...... 76

Chapter 4. Breadfruit Protein Digestibility Analysis Using an In Vitro Human Digestion Model ...... 77

4.1. Synopsis ...... 77

4.1.1. Objectives ...... 77

4.2. Materials and Method ...... 78

4.2.1. Reagents ...... 78

4.2.2. Sample preparation: ...... 79

4.2.3. In vitro digestion model to mimic human digestion ...... 79

4.2.4. Protein determination in the flours and digestion extracts ...... 81

4.3. Results ...... 83

4.4. Discussion ...... 84

4.5. Summary ...... 85

Chapter 5. Effects of Breadfruit Digestion on Human Colon Epithelial Cells Using an In Vitro Cell Model ...... 86

5.1. Synopsis ...... 86

5.1.1. Objectives ...... 86

5.2. Materials and Method ...... 87

5.2.1. Reagents ...... 87

5.2.2. Caco-2 human epithelial colorectal adenocarcinoma cells ...... 88

ix 5.2.3. Digestion extracts preparation ...... 89

5.2.4. Effects of digestion extracts on Caco-2 cell viability ...... 90

5.2.5. Effects of LPS and/or IL-1β stimulation on Caco-2 cells ...... 90

5.2.6. Effects of digestion extracts combined with LPS and/or IL-1β stimulations on Caco-2 cells ...... 92

5.2.7. Primers design and efficiency testing ...... 93

5.2.8. Statistical analysis ...... 95

5.3. Results ...... 95

5.3.1. Effects of digestion extracts on Caco-2 cell viability ...... 95

5.3.2. Effects of LPS and/or IL-1β stimulations on Caco-2 cells ...... 99

5.3.3. Effects of digestion extracts combined with LPS and/or IL-1β stimulations on Caco-2 cells ...... 104

5.4. Discussion ...... 123

5.4.1. Cytotoxicity of wheat and breadfruit digestion to Caco-2 cells ...... 123

5.4.2. Cytokine response of breadfruit- and wheat-treated Caco-2 cells .... 123

5.5. Summary ...... 126

Chapter 6. Effect of the Breadfruit Diet on Mice Health ...... 127

6.1. Synopsis ...... 127

6.1.1. Objectives ...... 128

6.2. Methods ...... 128

6.2.1. Reagents ...... 128

6.2.2. Diet design ...... 129

6.2.3. Experimental animals ...... 134

6.2.4. Overall growth evaluation and hematology analysis ...... 135

6.2.5. Tissue collection ...... 135

6.2.6. Feces protein, total mineral analysis, and fecal occult blood

x detection ...... 136

6.2.7. Ileum morphology examination ...... 136

6.2.8. Major cytokine response on ileum ...... 137

6.2.9. Quantification of bacterial groups in the colon ...... 138

6.2.10. Statistical analysis ...... 138

6.3. Results ...... 139

6.3.1. Overall health and growth ...... 139

6.3.2. Hematology analysis ...... 146

6.3.3. Fecal protein, total mineral analysis, and fecal blood detection ...... 149

6.3.4. Ileum morphology examination ...... 153

6.3.5. Major cytokine response on ileum ...... 158

6.3.6. Quantification of bacterial groups in the colon ...... 161

6.4. Discussion ...... 164

6.4.1. Overall heath and growth ...... 164

6.4.2. Intestinal health evaluation ...... 167

6.5. Summary ...... 169

Chapter 7. Conclusion ...... 170

References ...... 174

Appendices ...... 194

Appendices A: Breadfruit field data sheet in Kauai ...... 194

Appendices B: JR lab reports of the diets used in the mice study ...... 1945

xi List of Tables

Table 1-1 Common dishes made from breadfruit in different areas ...... 2

Table 1-2 Breadfruit production in 2007 Agricultural Censuses Report Conducted by the U.S. Department of Agriculture ...... 4

Table 1-3 The classification of Breadfruit (Artocarpus altilis and Artocarpus altilis× A. mariannensis) ...... 6

Table 1-4 Summary of Literature Search Using Web of ScienceTM and Google ScholarTM ...... 8

Table 1-5 The minimum and maximum reported values for breadfruit (Artocarpus altilis and A. altilis× A. mariannensis) proximate analyses ...... 12

Table 1-6 The minimum and maximum starch and sugar contents of breadfruit (Artocarpus altilis and A. altilis× A. mariannensis) ...... 15

Table 1-7 The minimum and maximum carotenoid and vitamin content for breadfruit (Artocarpus altilis and A. altilis× A. mariannensis) ...... 17

Table 1-8 Reported minimum and maximum mineral content for breadfruit (Artocarpus altilis and A. altilis× A. mariannensis) ...... 19

Table 1-9 Percent daily of recommended daily intake of essential nutrients provided by consumption of breadfruit ...... 20

Table 1-10 Reported maximum and minimum anti-nutrients for breadfruit (Artocarpus altilis and A. altilis × A. mariannensis)...... 23

Table 2-1 The protocol for the collection of seasonality data for breadfruit accessions by visual estimates ...... 28

Table 2-2 The number of months between planting and fruit production for Artocarpus species and cultivars grown in the McBryde Garden, Kauai ...... 30

Table 2-3 The summary of the total fruit number, average fruit weight, and yield

xii in Artocarpus species after being planted 4 years and 7 years. Different plant number exists due to the difference in plant age ...... 43

Table 2-4 Comparison of fruit growth between Rotuma and Ma’afala in Kauai and New Caledonia (Lebegin et al., 2007) ...... 50

Table 2-5 Comparison of the environmental and climate factors for breadfruit plantings in Kauai and Maui, Hawaii, and New Caledonia ...... 51

Table 2-6 Comparison of Artocarpus camansi (breadnut) production in McBryde Garden in Kauai with the University of the West Indies Field Station in Valsayn, Trinidad, and Tobago (Roberts-Nkrumah 2005) ...... 53

Table 2-7 Comparison of the protein yields between Ma’afala and other staples .... 55

Table 3-1 The cultivar list of breadfruit (Artocarpus altilis and A. altilis × A. mariannesis) used in the amino acid analysis ...... 59

Table 3-2 HPLC gradient elution for the breadfruit amino acid study ...... 63

Table 3-3 The summary of amino acid content in commercially available breadfruit cultivars (Artocarpus altilis and A. altilis x A. mariannensis): Ma’afala, Yellow, White, Piipiia, Puaa, and Ulu fiti ...... 68

Table 3-4 Nutritional value of commercially available breadfruit cultivars (Artocarpus altilis and A. altilis × A. mariannesis): Ma’afala, Yellow, White, Piipiia, Puaa, and Ulu fiti ...... 75

Table 4-1 Composition of saliva, gastric solution, duodenal solution, and bile solution used in the multi-stage enzymedigestion model ...... 81

Table 5-1 Experimental design for testing effects of digestion extracts combined with LPS and/or IL1 beta stimulation on Caco-2 cells (n=3) after 24 hours ...... 93

Table 5-2 RNA sequence, best annealing temperature and primer efficiency for the primers used in the breadfruit in vitro study ...... 94

Table 5-3 Cytokine response of Caco-2 after 24 hours of breadfruit/wheat

xiii digestion treatments under various stimulations ...... 106

Table 6-1 Comparison of nutritional value between the breadfruit (BF) diet and 5LG4 diet ...... 129

Table 6-2 Comparison of average body composition and tissue weight between the BF-fed (breadfruit-fed) mice and 5LG4-fed mice ...... 143

Table 6-3 Comparison of blood hematology between mice fed on the breadfruit (BF) diet and 5LG4 diet ...... 147

Table 6-4 Histological examination of ileum morphology ...... 154

xiv List of Figures

Figure 1-1 The summary of literature cited describing breadfruit (Artocarpus altilis and A. altilis ×A. mariannensis) nutritional Composition. a) Proximate analysis, b) carotenoid and vitamin analysis, c) mineral analysis ...... 14

Figure 2-1 Average fruit number per tree of breadfruit cultivars planted from 2008 to 2012 (a). Average fruit number per month of Artocarpus altilis (n=10), A. camansi (n=4), A. mariannensis (n=1), and A. altilis × A. mariannensis (n=10) from 2008 to 2012 (b) ...... 34

Figure 2-2 The average percentage of years (a) and average percentage of breadfruit trees (b) that breadfruit trees bore fruit, edible fruit and male flowers from 2007 to 2012 in National Tropical Botanical Garden in Kauai, Hawaii edible/harvestable fruit refers to full-sized fruit, mature fruit, and ripe fruit...... 35

Figure 2-3 Comparison of individual edible fruit seasonality profiles of breadfruit cultivars planted in the National Tropical Botanical Garden in Kauai and Maui, Hawaii...... 38

Figure 2-4 Comparison of individual male flower seasonality profiles of breadfruit cultivars planted in the National Tropical Botanical Garden in Kauai and Maui, Hawaii ...... 40

Figure 2-5 Average fruit number per year of breadfruit cultivars planted in the National Tropical Botanical Garden in Kauai, Hawaii from 2 years to 8 years old...... 42

Figure 2-6 Average fruit weight for breadfruit cultivars from 2 years old to 8 years old...... 45

Figure 2-7 Yield for breadfruit cultivars from 2 years to 8 years old...... 46 xv Figure 2-8 Total fruit number (a), and average fruit weight (b) for breadfruit cultivar Ma’afala in the National Tropical Botanical Garden at on Kauai...... 47

Figure 3-1 Breadfruit (Artocarpus altilis) and hybrids (A. altilis × A. mariannensis) essential amino acid content (mg/g protein)...... 65

Figure 3-2 Comparison of essential amino acid content between breadfruit (Artocarpus altilis) and hybrids (A. altilis × A. mariannensis) based on protein weight (a) and dry tissue weight (b) ...... 66

Figure 4-1 A Schematic representation of the in vitro digestion model ...... 80

Figure 4-2 Protein content in wheat and breadfruit flour before and after digestion (a) and digestion extracts (b) by BCA and Modified Lowry Assays (n=3)...... 84

Figure 5-1 Effects of digestion extracts of wheat and breadfruit on cell viability. a) 1% digestion extracts. b) 5% digestion extracts. c) 10% digestion extract. d) 50 % digestion extracts...... 98

Figure 5-2 Estimated total number of Caco-2 cells per ml after treatment with digestion extracts for 4 hrs...... 99

xvi Figure 5-3 The mRNA expression of cytokines from Caco-2 cells with no stimulation, IL-1β, or LPS+IL-1β stimulation for 4 hrs, 8 hrs, and 24 hrs. (a) MCP-1. (b) iNOS. (c) TNF-α. (d) IL-4. (e) IL-6. (f) IL-8. (g) IL- 10. (h) IFN-γ...... 104

Figure 5-4 Effect of digestive enzyme solution, wheat digestion, and breadfruit digestion on cytokine expression in Caco-2 cells without LPS or IL- 1β stimulation for 24 hours. a) IL-4, b) IL 10, c) iNOS, d) TNF-α, e) IFN-γ, f) IL-8,g) IL- 6, h) MCP-1...... 110

Figure 5-5 Effect of digestive enzyme solution, wheat digestion, and breadfruit digestion on cytokine expression in Caco-2 cells stimulated with LPS (1000 ng/mL) for 24 hours. a) IL-4, b) IL 10, c) TNF-α, d) IFN-γ, e) MCP-1, f) IL-8, g) IL- 6 h) iNOS...... 114

Figure 5-6 Effect of digestions on cytokine expression on Caco-2 cells stimulated with IL-1β (100 ng/mL) for 24 hours a) IL-4, b) IL 10, c) iNOS, d) TNF-α, e) IFN-γ, f) MCP-1, g) IL-8, h) IL- 6...... 118

Figure 5-7 Effect of digestive enzyme solution, wheat digestion and breadfruit digestion on cytokine mRNA expression in Caco-2 cells stimulated with LPS (1000ng/mL) and IL-1β(100 ng/mL) for 24 hours. a) IL-4, b) IL 10, c) iNOS, d) TNF-α, e) IFN-γ, f) MCP-1, g) IL-8, h) IL- 6...... 122

Figure 6-1 Comparison of scavenging activity percentage at 15 minutes between breadfruit diet and 5LG4 diet using DPPH methods...... 134

Figure 6-2 Comparison of daily body weight (a) and growth rate (b) of mice fed on breadfruit (BF) and 5LG4 diets...... 140

Figure 6-3 Comparison of food usage of mice fed on the breadfruit (BF) and 5LG4 diets ...... 141

Figure 6-4 Comparison of daily water consumption (a) and average daily water

xvii consumption (b) of mice fed on breadfruit (BF) and 5LG4 diets...... 142

Figure 6-5 Comparison of fecal protein content from mice fed with the breadfruit (BF) and 5LG4 diets at Day 4 (a), Day 7 (b), Day 14 (c), and Day 21 (d) ...... 151

Figure 6-6 Comparison of total mineral content from mice fed with the breadfruit (BF) and 5LG4 diets at Day 4 (a), Day 7 (b), Day 14 (c), and Day 21 (d) ...... 153

Figure 6-7 Representative histopathology of the ileum of 5LG4-fed male mice (a), BF-fed male mice (b), 5LG4-fed female mice (c), and BF-fed female mice (d) at 3 weeks of age stained with haematoxylin and eosin (H&E)...... 157

Figure 6-8 Comparison of cytokine expression between ileums of breadfruit (BF)-and 5LG4-fed mice, a) TNF-α, b) IFN-γ, c) IL-10, d) IL-6, e) iNOS ...... 160

xviii List of Symbols and Abbreviations

Ala Alanine

Arg Arginine

Asn Asparagine

Asp Aspartic acid

BF Breadfruit

BCA Bicinchoninic acid

BSA Bovine serum albumin

Caco-2 cells Human epithelial colorectal adenocarcinoma cells

Glu Glutamic acid

Gly Gluycine

His Histidine

H&E staining Haematoxylin and eosin staining

HPLC High performance liquid chromatography

IFN γ Interferon gamma

IL 1β Interleukin 1 beta

IL Interleukin

Ile Isoleucine iNOS Inducible nitric oxide synthase

Leu Leucine

LPS Lipopolysaccharide

Lys Lysine

MCH Mean corpuscular hemoglobin

xix MCHC Mean corpuscular hemoglobin concentration

MCP-1 Monocyte chemoattractant protein-1

MCV Mean corpuscular volume

Met Methionine

MPV Mean platelet volume

Phe Phenylalanine

Pro Proline

RDW Red cell distribution width

RT-qPCR Real time quantitative polymerase chain reaction

Ser Serine

Thr Threonine

TNF-α Tumor necrosis factor alpha

Trp Tryptophan

Tyr Tyrosine

Val Valine

xx Acknowledgements

I would like to take this opportunity to thank everyone that supported and helped me during my graduate studies. First and foremost, I would like to thank my advisors, Dr. Susan Murch and Dr. Deanna Gibson. Dr. Murch provided the opportunities and environment for the growth of this project. I thank her for always being supportive, wise, and understanding. I thank her for being there for me whenever I need, giving me hope and courage, and making it all possible. She is also a great cook and her food always cheers me up (Yummy!). I also want to express my sincere thanks to Dr. Gibson for providing guidance and necessary training for the project development. I am also indebted to the rest of my advisory committee, Dr. Miranda Hart, Dr. Frederic Menard and Dr. Paula Brown, for their continued support and advice throughout the graduate program. Thank you to Dr.

Diane Ragone for sharing the breadfruit germplasm collection at the National

Tropical Botanical Garden. I would like to acknowledge the contribution of Dr. Paul

Shipley, who generously served as part of my comprehensive exam. I also want to thank Dr. Max Jones, who collected the breadfruit samples used in this project and provided guidance in data analysis.

It is important to give my gratitude to other people at the University of British

Columbia for all the assistance and company they provided during this graduate program. This project would be impossible without the help from everyone in the

Murch and Gibson labs. My deepest gratitude goes to Dr. Christina Turi, Broc

Glover, Katie Axwik, Tiah Lee, Daniella DeCoffe, Sandeep Gill, and Nishat Tasnim for their assistance in conducting the project and/or analyzing the data. While there

xxi are too many names to include here, I also want to thank Alex Lane, Christina

Livingston, and Kathy Baker for providing an efficient lab environment. I want to thank the Center for Scholarly Communication, especially Lori Walter, Amanda

Brobbel, and Ellen Campbell, for their time and support. Additionally, I would like to acknowledge the contribution of the Trustees and Fellows of the National Tropical

Botanical Garden and the Breadfruit Institute. Na Lima Kokua volunteers were crucial to the success of this project. In particular, I would like to acknowledge Billie

Dawson who led a team of volunteers including: Loren Johnson, Joyce Packard,

Linda Spade, Noreen Daniels, Mary Stone, Neil Brosnahan, and Cherisse Kent.

Finally, I would like to thank my parents and friends for their patience and support over these years. I thank my parents for encouraging me and helping me in whichever way they could. Thank you to my friends, Alessandro Errico, Christ

Heckel, and Edward Regier for listening to my frustrations and giving me food and candies when I was down. A special thank you must be given to my husband and colleague, Jensen Lund, for making my life much easier and giving me immeasurable help in the lab. Thank you for all the dishes and proofreading you did for me.

xxii Dedication

I dedicate this work to my parents, who love me more than anything else in the world.

献给我的父亲刘柏生,母亲王月桂.

xxiii Chapter 1. Introduction

1.1. History of breadfruit consumption

The history of human consumption of breadfruit as a staple crop of the Pacific extends at least 3,000 years. Traditionally cultivated varieties were bred by individuals, families, and communities and carried by the Lapita from Papua New Guinea throughout Oceania (Ragone, 1997). The traditional cultivars of breadfruit in Hawaii were brought with the Lapita to Hawaii sometime between 1000 and 1200 A.D.

European collections of breadfruit began with the Spanish and the Dutch East

India Company during the 17th century. In the late 1800s, several governments and individuals engaged in the widespread dissemination of breadfruit to regions outside of Oceania (Barrau, 1976; Ragone, 1997; Smith et al., 1992). In 1769, Joseph Banks traveling with Captain James Cook to Tahiti recognized the potential of breadfruit and observed the following (Banks, 1962):

“In the article of food these happy people may almost be said to be exempt from the curse of our forefather; scarcely can it be said that they earn their bread with the sweat of their brow when their cheifest sustenance Bread fruit is procurd with no more trouble than that of climbing a tree and pulling it down. Not that the trees grow here spontaneously but if a man should in the course of his life time plant 10 such trees, which if well done might take the labor of an hour or thereabouts, he would as compleatly fulfull his duty to his own as well as future generations as we natives of less temperate climates can do by toiling in the cold of winter to sew and in the heat of summer to reap the annual produce of our soil, which when once gatherd into the barn must be again resowd and re-reapd as often as the Colds of winter or the heats of Summer return to make such labor disagreable.” (The Endeavour Journal of Joseph Banks, August 14th 1769)

1 In August 1787, Banks successfully persuaded King George III to invest in an expedition to collect breadfruit and transplant it to the Caribbean as a reliable food source for sugar plantations. Responsibility for the expedition was placed under the leadership of Lieutenant William Bligh RN and the HMS Bounty sailed from England in December 1787. The crew of the Bounty collected 1,015 breadfruit trees but all were lost in the famous mutiny at sea in 1789. Determined to transport breadfruit from

East to West, Sir Joseph Banks ordered Bligh to commence two additional voyages that successfully transplanted 668 breadfruit trees from Tahiti to Jamaica and St.

Vincent in 1793 and 1796 (Powell, 1977).

Breadfruit is commonly used as a staple food in Polynesia and Micronesia, and as a supplementary staple in most of Melanesia (Fosberg, 1960; Lim, 2012; Ragone and

Raynor, 2009). Edible portions of breadfruit for human consumption include immature and ripe fruit, seeds, young leaves, and ripe blossoms. Breadfruit is consumed raw, boiled, roasted, baked, or fried. Preservation of the fruit can be achieved through sun- drying or in pits (Jones et al., 2013a; Lim, 2012; Ragone and Raynor, 2009). Breadfruit is consumed in various ways in the tropical regions (Lim, 2012; Morton, 1987; Meilleur et al., 2004; Ragone and Raynor, 2009) (Table 1-1).

Table 1-1 Common dishes made from breadfruit in different areas (Lim, 2012; Morton, 1987; Meilleur et al., 2004; Ragone and Raynor, 2009)

Region Common dishes made from breadfruit

Bahamas Breadfruit soup made from under-ripe fruit

Breadfruit chips made from overripe or soft fruit Barbados Combined with wheat for bread making

2 Region Common dishes made from breadfruit

Brazil Combined with wheat for bread making

Sir Lanka Breadfruit dipped in a salt solution Dominican Breadfruit bread "buen pan" Republic Philippines Cooked fruit in coconut and sugar “oil down” by cooking breadfruit, meat, coconut milk, and dasheen Grenada leaves Guam Breadfruit paste soaked in water

Boiled under-ripe fruits with butter and sugar, or salt and pepper Breadfruit chowder made from cooking breadfruit with vegetables, Hawaii bacon, and milk Substitution for taro in poi, "poi ulu"

Jamaica Breadfruit flour-based porridge

Malaysia Fried sliced firm-ripe breadfruit in syrup or palm sugar Made into a paste by fermenting breadfruit in banana leaf-lined boxes Micronesia Breadfruit soaked and beaten in the sea New Fermented breadfruit Hebrides Polynesia Fermented or baked breadfruit in a native oven "pana" or "panen," cooked breadfruit with olive oil, onions, and Puerto Rico saturated bacalao (salted cod fish) “masi,” fermented breadfruit with banana and Heliconia leaves to Samoa make a paste, which is cooked with coconut cream Solomon Fruit roasted in an underground oven (imu) Islands Breadfruit chips made from overripe or soft fruit Trinidad “oil down,” cooked breadfruit, meat, coconut milk and dasheen leaves “oil down,” cooked breadfruit, meat, coconut milk and dasheen Tobago leaves

3 1.2. Current uses and significance of breadfruit in the world

Breadfruit is one of 35 crop species identified in the International Treaty on Plant

Genetic Resources for Food and Agriculture as an underutilized crop with the potential to improve food security and interdependence (FAO, 2009). The reported number of breadfruit farms in the United States were around 5000 in 2007 (USDA, 2007) (Table

1-2).

Table 1-2 Breadfruit production in 2007 Agricultural Censuses Report Conducted by the U.S. Department of Agriculture (USDA, 2007)

Northern Virgin Guam American Regions Mariana Islands Island Samoa Islands

Year 2007 2007 2007 2008

Number of farms 54 42 14 4,828

Number of non-bearing 35 46 61 NA trees

Number of bearing trees 280 216 258 NA

Pounds harvested for 10,713 4,774 9,650 252,375 sale Pounds harvested for NA NA NA 3,140,728 consumption

In the past, breadfruit was mainly produced in Hawaii, Puerto Rico, the Marianas

Islands and American Samoa for local use and the commercialization of breadfruit was impossible due to difficulties in propagation (Murch et al., 2008; Ragone and

Raynor, 2009). In 2008, we reported the development of in vitro propagation methods

4 for clonal propagation of breadfruit trees in a sterile, controlled environment (Murch et al., 2008; Shi et al., 2007). This advancement leads to the development of a horticultural industry for mass propagation and distribution of the trees to tropical countries in projects designed to increase food security (www.globalbreadfruit.com).

1.3. Breadfruit botanical description

The common name ‘breadfruit’ refers to A. altilis, which can be triploid

(2n=3x=~84) and producing no seeds or diploid (2n=2x=~56) and producing few to several seeds (Ragone, 2001). Some of the cultivated varieties of breadfruit are interspecific hybrids of A. altilis × A. mariannensis (Fosberg, 1960; Zerega et al., 2004,

2005) (Table 1-3; USDA, 2015a; Stevens, 2012). Early generation hybrids produce fruits that most closely resemble its A. mariannensis parent while later generation hybrids more closely resemble A. altilis and are seedless (Jones et al., 2013 a).

Several thousand cultivated varieties of breadfruit are known across the Pacific tropical islands and the fruit has been used as a staple food source for about 3,000 years. Breadfruit is a moderately large evergreen tree generally growing 15 to 20 m, but sometimes reaching over 30 m tall (Niering, 1963; Ragone, 1997; 2006). Breadfruit trees are monoecious, with both male and female inflorescences in the same tree.

The inflorescences are comprised of about 1500-2000 individual florets connected to the receptacle (Ragone 1997; 2006).

5 Table 1-3 The classification of breadfruit (Artocarpus altilis and Artocarpus altilis× A. mariannensis)

Rank Scientific name and common name

Kingdom Plantae – Plants Subkingdom Embryophyta Unranked Angiosperms Unranked Eudicot Unranked Rosids Order Rosales Family Moraceae – Mulberry family Genus Artocarpus J.R. Forst. & G. Forst. – Breadfruit Artocarpus altilis (Parkinson) Fosberg – Breadfruit Species Artocarpus mariannensis Trécul– Artocarpus

The shape of the fruit is irregular, but the texture is generally flattened or rounded pebble. The advanced hybrids have a similar texture and shape to the early generation hybrids, but the size of the fruit is relatively larger and the fruit is seedless

(Jones et al., 2013a). The general fruit size is 12 cm×16 cm and the fruit weight ranges from 1 to 2 kg. Some cultivars can produce fruit weighing up to 6 kg (Ragone, 1997,

2006; Zerega et al., 2005). During the maturation process, the outside fruit skin turns from light green to yellow and the inside flesh becomes creamy white to yellow.

1.4. Current breadfruit nutritional analysis

Using Web of Science, Google Scholar search engines, and government databases and regional reports between July and Oct. 2014, I identified 41 reports researching on the nutritional profile of breadfruit cultivars (Table 1-4). Proximate data

6 for breadfruit included protein, total carbohydrate, crude fiber, starch, and lipid content as well as total calories, insoluble and soluble fiber content (Figure 1-1a). Studies by

Golden and Williams (2001), Holloway et al. (1985), Jones et al. (2011a&b), Ragone and Cavaletto (2006), Tumaalii and Wootton (1988), and Wootton and Tumalii (1984) included cultivar name and variability between cultivars with respect to protein, carbohydrate and caloric content (Table 1-4). Overall, the literature describes a wide range of values. For example, lipid contents varied between reports (0.08 - 4.90 g per

100g of fresh or cooked breadfruit), but most studies of fats generally reported values of less than 2 g per 100 g (Table 1-5). Fiber content varied widely from 0.90–4.88 g/100g fresh weight, 2.13-7.37 g/ 100g cooked weight, and 1.66 –15.29 g/100g in flour

(Table 1-5). Ragone and Cavaletto (2006) found no significant difference in fiber content of different cultivars. It is difficult to assess the total fiber content of generic breadfruit since some of the variability may be due to different methods for determination of fiber and criteria for fiber definitions that have been revised over the years to include various types of carbohydrates such as cellulose, hemicelluloses, pectins, hydrocolloids, beta-glucans, resistant starch, and non-digestible oligosaccharides (McClearly et al., 2012; Mudgil and Barak, 2013). Furthermore, the fiber content of flours made from selected breadfruit tissues, i.e. peel, core, or pulp, was found to significantly differ (Graham and Negron de Bravo, 1981; Mayaki et al.,

2003).

7 Table 1-4 Summary of Literature Search Using Web of ScienceTM and Google ScholarTM

Source Species1 Cultivar Number Preparation Method

Adegunwa et al. 2014 Aa Instant breadfruit flour Adeniran et al. 2012 Aa Fermented flour Fresh fruit that was picked after immediately Amusa et al. 2002 Ac a dropping from trees Firm and mature fruit made into fermented or Appiah et al. 2011 Aa unfermented pulp flour Aregheore 2000 Aa Flour Aa, Fresh, boiled or baked fruit collected at the mature Bahado-Singh et al. 2006 Aa × Am stage Raw or boiled fruit collected at the ripe mature stage Englberger et al. 2003a Am, Aa 8 (creamy to white flesh), and sampled with or without rind Raw, baked or boiled fruit collected at the mature or Englberger et al. 2003b Am, Aa 10 ripe stage Fresh, boiled or dried fruit, picked when flesh was Englberger et al. 2007 Am, Aa 16 yellow, brown, or cream coloured. Raw, fruit collected at mature or ripe stage, with and Englberger et al. 2014 Aa, Am 7 without skin

8 Source Species1 Cultivar Number Preparation Method

Esuoso and Bamiro 1995 Aa Breadfruit flour Graham and Negron de Aca Breadfruit flour Bravo 1981 Golden and Williams 2001 Aa 1 Fresh fruit Holloway et al. 1985 Aa 1 Fresh fruit suitable for consumption Haydersah et al. 2012 Ac a Flour Huang et al. 2000 Aa (?) Fresh fruit collected when still firm and not overripe Ijarotimi and Aroge 2005 Aa Flour made from freshly harvested fruit Aa, Am, Raw fruit collected at mature stage, not fully ripe; Jones et al. 2011a Aa × Am 94 Flour

Aa, Am, Raw, baked, or boiled fruit collected at mature stage, Jones et al. 2013a Aa × Am, 94 not fully ripe.

Leterme et al. 2005 Ac a Fresh fruit Loos et al. 1981 Ac Starch isolate / flour Aa, A. Flour made from freshly harvested mature breadfruit Malomo et al. 2011 integrifolia, and breadnut Mayaki et al. 2003 Ac a Fruit harvested at mature but unripe stage and made

9 Source Species1 Cultivar Number Preparation Method

into a flour

Nelson-Quartey et al. 2007 Aa Flour made from mature, firm fruit

Nochera and Caldwell Ac a Breadfruit flour 1992 Nwokocha and Williams Aa Starch isolate 2010 Oduro et al. 2007a Aa Flour made from mature unripe fruit

Oduro et al. 2007b Aa Flour made from mature unripe fruit

Oladunjoye et al. 2010 Aa Breadfruit meal Oladunjoye and Ojebiyi Aa Breadfruit meal 2011 Oso et al. 2010 Aa Sun dried breadfruit meal Oulaï et al. 2013 Aa Flour Ragone and Cavaletto Aa, Boiled fruit that were collected at the mature, unripe 20 2006 Aa × Am stage. Ravindran and Ac a Breadfruit meal Sivakanesan 2006 Reeve 1974 Aa Fresh dehydrated fruit

10 Source Species1 Cultivar Number Preparation Method

Boiled (immature green colour) and roasted (mature, Samuda et al. 1998 Aa yellow-brown colour) fruit Tumaalii and Wootton Aa 7 Starch isolate 1988 Wang et al. 2011 Aa Defatted flour Cooked or raw breadfruit that was collected at the Wenkam, 1990 Ac a ripe or green stage. Widanagamage et al. 2009 Aa/Ac a Boiled breadfruit Wootton and Tumaalii, Breadfruit Flour made from immature, mature or very Aa 7 1984 mature fruits. 1Aa = Artocarpus altilis; Ac = Artocarpus communis; Am = Artocarpus mariannensis; Aa × Am = Artocarpus altilis × Artocarpus mariannensis ? Indicates that the species name is not entirely clear a A. communis is a synonym for A. altilis

11 Table 1-5 The minimum and maximum reported values for breadfruit (Artocarpus altilis and A. altilis× A. mariannensis) proximate analyses

Fresh (100g) Cooked (100g) Flour (100g) Nutrient Min Max Min Max Min Max

Ash (%) 0.8 4.6 NA NA 0.8 6.7

Moisture (%) 19 83 53.2 83.6 2.5 21

Dry matter (%) 17 80.9 16.4 46.8 79 97.5

Energy (Kcal) 102 310 80 160.9 279.8 378

Total Carbohydrates (g) 14.3 70.1 18.1 37 50 88

Lipid (g) 0.1 4.5 0.1 4.9 0.5 11.8

Protein (g) 0.07 5.2 0.6 11.4 1.9 18.7

Crude Fiber (g) 0.9 4.9 1.8 7.4 0.8 15.3

Insoluble Fiber (g) 3.1* 25.6* 2.4 20 7.5* 62.3*

Soluble Fiber (g) 0.20 0.2 NA 7.2 0.2* 11.4*

* Values marked with * were extrapolated based on the calculated average dry weight for breadfruit (fresh = 37.55%, baked = 29.35%, flour = 91.40%). A value of NA indicates that a specific group was either not detected or below the limits of quantification.

12 Figure 1-1 a

35 30 25 20 15 10

Number ofpublicationsNumber 5 0

Fresh Cooked Flour

Figure 1-1 b

20 18 16 14 12 10 8 6

4 Number ofpublicationsNumber 2 0 B Ca Cl Co Cu Fe K Mg Mn Na Ni P S Zn

Fresh Cooked Flour

13 Figure 1-1 c

Number ofpublicationsNumber 0

Fresh Cooked Flour

Figure 1-1 The summary of literature cited describing breadfruit (Artocarpus altilis and A. altilis ×A. mariannensis) nutritional composition. a) Proximate analysis, b) carotenoid and vitamin analysis, c) mineral analysis

1.4.1. Carbohydrates

Loos et al. (1981) were the first researchers to focus on the characterization of starch from breadfruit. Their work explored basic physicochemical and structural properties of breadfruit starch, but, the study did not identify the cultivars used. In 1984 and 1988, Wootton and Tumaalii published similar research on starch isolates from breadfruit with clear identification and comparison between cultivars. In general, breadfruit starch has an amylose content ranging from 16.6% to 20.4% and a granular size from 0.5 µm to 37.8 µm (Wootton and Tumaalii, 1984). Breadfruit starch composition changes with the maturity of the fruit, with sugars released from the

14 starches as the fruit matures (Graham and Negron de Bravo, 1981). This difference in nutrient composition with maturity also contributes to reported variation in carbohydrate and sugar content (Golden and Williams, 2001; Graham and Negron de

Bravo 1981; Wootton and Tumaalii, 1984; Table 1-6), which increases the difficulty of cross comparison between studies. Glucose and sucrose were found to be the dominant sugar in fully mature breadfruit across the 8 cultivars studied (Golden and

Williams, 2001; Wootton and Tumaalii 1984). When compared to other staple foods such as corn, sweet potatoes, yam and wheat (Esuoso and Bamiro, 1995; Nwokocha and Williams, 2010), the starch properties of breadfruit are suitable for a range of diets and applications (Adebowale et al., 2004; Haydersah et al., 2012).

Table 1-6 The minimum and maximum starch and sugar contents of breadfruit (Artocarpus altilis and A. altilis× A. mariannensis)

Fresh (g/100g) Cooked (g/100g) Flour (g/100g) Nutrient Min Max Min Max Min Max

Total Starch 15.5 28.4 NA NA 42.6 75.7

Fructose NA 0.16 NA NA 4.3 13.6

Glucose 0.18 0.44 NA NA 6.5 11.3

Sucrose 0.25 0.62 NA NA NA 16.4

Total Sugars NA NA NA NA 2.8 26.8

Total Reducing Sugars NA NA NA NA 2.7 22.5

15 1.4.2. Carotenoids

Quantification of carotenoids included seven nutrients: alpha carotene, beta carotene, beta cryptoxanthin, lutein, lycopene, zeaxanthin and total carotenoids

(Figure 1-1b). For carotenoid analysis, all references included species and cultivar names (Table 1-4). Both Englberger et al. (2003 a&b) and Jones et al. (2013b) observed changes in the carotenoid content of breadfruit as a result of cooking, baking or boiling. In addition to the mode of preparation, cultivar or varietal type, tissue sampled (i.e. with rind or without rind), and stage of maturity (starchy mature vs. ripe) have all been found to impact the quantified carotenoids (Englberger et al., 2014,

2003a&b, Jones et al., 2013b; Ragone and Cavaletto, 2006). It is also important to note that the presence or absence of specific carotenoids were either undetected or below the limits of quantification for several cultivars in these studies. Overall, carotenoid content varied widely in our review of the literature (Table 1-7) and is unlikely to be consistent except in a single orchard planted with clonally propagated trees of a single cultivar with standardized harvest and processing protocols.

16 Table 1-7 The minimum and maximum carotenoid and vitamin content for breadfruit (Artocarpus altilis and A. altilis× A. mariannensis)

Fresh (100g) Cooked (100g) Flour (100g) Nutrient Min Max Min Max Min Max

Total carotenoids (µg) NA 3769 NA 1260 NA 6549

alpha carotene (µg) NA 260 NA 142 NA 537.5

beta carotene (µg) NA 3410 NA 868 NA 5501.7

beta cryptoxanthin (µg) NA 3.3 NA 10.6 NA 20.5

Lutein (µg) NA 690 NA 759 NA 2021.6

Lycopene (µg) NA 48.7 NA 25.9 NA 99.6

Zeaxanthin (µg) NA 60 NA 70 NA 182

Folic acid (µg) NA 1.3 NA 1.0 NA 3.1

Vitamin B1 (mg) 0.12 0.28 0.09 0.14 0.29 0.6

Vitamin B2 (mg) 0.05 0.1 0.02 0.06 0.16 0.4

Vitamin B3 (mg) 0.84 1.7 0.64 1.4 2.30 4.4

Vitamin C (mg) 16.20 21 1.60 12.1 NA 22.7

* Values marked with * were extrapolated based on the calculated average dry weight for breadfruit (fresh = 37.55%, baked = 29.35%, flour = 91.40%). NA indicates that a specific group was either not detected or below the limits of quantification.

17 1.4.3. Minerals

Quantification of minerals in breadfruit was relatively common in the literature; however, a majority of references pertained to flours (Figure 1-1c). Similar to proximate analysis, Jones et al. (2011b) and Ragone and Cavaletto (2006) were the only references that included cultivar names (Table 1-4). Both references report significant differences between cultivars with respect to Cu, Fe, K, Mg, Mn, and Zn concentrations. Furthermore, Graham and Negron de Bravo (1981) found that mineral content in breadfruit will differ, not only between tissues types but also between stages of maturity. Consequently, data collected on the mineral nutrition of breadfruit varied considerably (Table 1-8). Comparison of reported values to recommended intake values provided by the FDA indicates that 5.78 ounce (163.86 g) breadfruit flour can provide up to 131% (Ca), 406% (Cu), 109% (Fe), 82% (Mg), 215% (Mn), 315% (P),

133% (K), 41% (Na), and 32% (Zn) of one’s daily mineral requirements (Table 1-9;

FDA, 2013). Based on the above, breadfruit is a good source of both magnesium and potassium. Still, it is possible that mineral nutrition of breadfruit is greater than what is currently known, given that no studies have investigated how soil type or environment can influence the overall nutrient content of breadfruit.

18 Table 1-8 Reported minimum and maximum mineral content for breadfruit (Artocarpus altilis and A. altilis× A. mariannensis)

Fresh (100g) Cooked (100g) Flour (100g) Nutrient Min Max Min Max Min Max

Boron (mg) 0.5 0.5 0.4 0.4 1.3 1.3

Calcium (mg) 18 54 10 30 5 800

Chlorine (mg) NA 2 NA 1.6 NA 4.9

Cobalt (µg) NA 1.1 NA 0.9 NA 2.70

Copper (mg) 0.08 0.3 0.45 0.5 0.1 4.95

Iron (mg) 0.26 52 NA 1.1 0.5 12

Magnesium (mg) 20 70 14 30 9.9 200

Manganese (mg) 0.04 0.3 0.1 0.3 0.1 2.

Nickel (mg) NA 0.08 NA 0.06 NA 0.19

Phosphorus (mg) 7 116 18 41 73.7 1920

Potassium (mg) 289 2390 240 522 66.9 2830

Sodium (mg) 3 27 2 70 1.90 597

Sulfur 20 31 15.6 24.2 48.7 75.5

Zinc (mg) 0.09 0.53 NA 0.13 0.13 3

19 Table 1-9 Percent daily of recommended daily intake of essential nutrients provided by consumption of breadfruit

Recommended Daily Values Provided by 5.78 Ounce (163.86g) of Breadfruit

Daily Fresh Cooked Flour Nutrient Value Min Max Min Max Min Max

Energy (Kcal) 2000 8% 25% 7% 13% 23% 31%

Total Carbohydrates (g) 300 8% 38% 10% 20% 27% 48%

Lipid (g) 65 0% 11% 0% 12% 1% 30%

Protein (g) 50 0% 17% 2% 37% 6% 61%

Crude Fiber (g) 25 6% 32% 12% 48% 5% 100%

Folate 400 0% 0% 0% 0% 0% 1%

Vitamin B1 (mg) 1.5 13% 31% 10% 15% 31% 61%

Vitamin B2 (mg) 1.7 5% 10% 2% 5% 15% 38%

Vitamin B3 (mg) 20 7% 14% 5% 11% 19% 36%

Vitamin C (mg) 60 44% 57% 4% 33% 0% 62%

Calcium (mg) 1000 3% 9% 2% 5% 1% 131%

Copper (mg) 2 7% 20% 37% 37% 8% 406%

Iron (mg) 18 2% 473% 0% 10% 5% 109%

Magnesium (mg) 400 8% 29% 6% 12% 4% 82%

Manganese (mg) 2 3% 27% 7% 25% 6% 215%

Phosphorus (mg) 1000 1% 19% 3% 7% 12% 315%

Potassium (mg) 3500 14% 112% 11% 24% 3% 133%

Sodium (mg) 2400 0% 2% 0% 5% 0% 41%

Zinc (mg) 15 1% 6% 0% 1% 1% 32% 1 Percent daily intake values are based on a caloric intake of 2,000 calories, for adults and children four or more years of age as described by the FDA (2013)

20 1.4.4. Vitamins

The vitamin content of breadfruit has been reported for folic acid, vitamins B1,

B2, B3, and vitamin C (Figure 1-1c). Vitamin C content ranged between 1.6–12.1 mg per 100g of cooked breadfruit for 20 cultivars. In the United States, daily average consumption of refined grains was 5.78 ounce (163.86 g) for male and female over the age of 2 from 2009 to 2010 (USDA, 2015b). According to recommended intake values provided by the FDA, 5.78 ounce serving of breadfruit flour can provide up to

61, 38, 36, and 62 percent of one’s daily requirements for vitamin B1, B2, B3 and C respectively (Table 1-9). A study by Graham and Negron de Bravo (1981) found that the concentration of vitamin C increased from the very immature to mature fruit, but was undetected at the very latest stage of ripening measures, potentially indicating enzymatic conversion of vitamin C at later stages of development. Similar information is needed but has not been generated, for other essential vitamins such as vitamins

B5, E, H, and K.

1.4.5. Antinutrients

Breadfruit, both raw and cooked, showed limited contents of anti-nutrients summarized in Table 1-10. α-amylase inhibitor was found in breadfruit but trypsin inhibitors was not (Ijarotimi and Aroge, 2005). These inhibitors are antinutrients because they inhibit digestive enzymes but they may also have some beneficial health impacts such as lowering blood glucose, reducing plasma cholesterol and triacylglycerol and treating breast cancer (Slavin et al., 1999).

Phenolic antinutrients, including tannins, catechin, gallic acid, and benzoic

21 acid, are commonly found in bran layers of cereal grains, especially the bran layers of the whole grain (Sidhu and Kabir, 2007; Thompson, 1993) and have been found in breadfruit (Appiah et al., 2012; Famurewa et al., 2015; Oulaï et al., 2014 ) (Table

1-10). Phenolic compounds can cause depression in food/feed intake, the formation of less digestible protein complexes, and inhibition of digestive enzymes, but can also prevent damage caused by reactive oxygen species (Slavin, 2004). Breadfruit has maximum total phenolic compounds of 408 mg/100g (Oulaï et al., 2014), which is much lower than whole grain, and similar to wheat germ (349 mg/100g) (Velioglu et al., 1998).

Oxalate is the major component of kidney stones in humans. In grain flours, the total oxalate content can range from 37-269 mg/100g, and in nuts, the number increases to 42-469 mg/100g (Chai and Liebman, 2005; Siener et al., 2006).

Breadfruit has an oxalate content between 100-192 mg/100g in the flour (Famurewa et al., 2015). The cooking process typically results in a dramatic reduction of oxalate content (Siener et al., 2006).

Phytic acid, and its salt phytate can be found in the cotyledon of legumes, oilseeds, and the bran of cereal grains, at a range from 10 to 6000 mg/100g (Sidhu and Kabir 2007; Thompson, 1993). These are considered as anti-nutrients because they can chelate mineral elements in the body, but many researchers have found that phytate can reduce blood glucose, and reduce plasma cholesterol and triacylglycerols (Jenab and Thompson, 2002; Kumar et al., 2010; Schlemmer et al.,

2009). The maximum phytate content found in breadfruit is 1269 mg/100g

(Famurewa et al., 2015; Oulaï et al., 2014). As with many other anti-nutrients, the

22 cooking process can reduce the phytate content. Cooked breadfruit has a maximum phytate content of 58.33 mg/100g (Oulaï et al., 2014).

Other anti-nutrients such as saponin and lectin were not detected in breadfruit flour or cooked breadfruit (Oulaï et al., 2014).

Table 1-10 Reported maximum and minimum anti-nutrients for breadfruit (Artocarpus altilis and A. altilis × A. mariannensis).

Flour (100g) Cooked (100g) Anti-nutrients Min Max Min Max α-amylase inhibitor ND 482 118 333 activity (AIU) Trypsin inhibitor (TIU) ND ND NA NA Total phenolic ND 408.73 223.61 321.12 compounds (mg) -Tannin (mg) 1 4.3 2.2 4.02

-Catechin (mg) ND 51.07 33.83 40.96

-Gallic acid (mg) ND 102.99 47.02 89.78

-Benzoic acid (mg) ND 66.97 43.15 60.92

Total oxalate (mg) 100 192 NA NA

Phytate (mg) 63.4 1269 36.2 58.33

Lignin (mg) 0 4320 3100 4050 ND: Not detected; NA: No data available

23 1.5. Study rationale and objectives

There are relatively few nutritional studies on breadfruit in the scientific literature but breadfruit flour is poised to become an international staple food crop in the next decade. The overall hypothesis of this thesis was:

“Breadfruit is a good source of protein for human use.”

There are a small number of studies on breadfruit protein quantity but breadfruit protein quality was not previously assessed. Likewise the digestion and impacts of breadfruit flour on mammals have not been previously investigated. Therefore, the specific objectives of this study were:

1. To determine the yield of breadfruit per tree and per hectare for 24 cultivars

2. To determine the quality of breadfruit protein by quantifying the essential

amino acid content

3. To determine the digestibility of breadfruit flour using an enzymatic model

of human digestion

4. To quantify the effects of digested breadfruit flour on human intestinal

epithelial cells in an in vitro cell model

5. To quantify whole body responses to consumption of breadfruit flour in a

rodent model

6. To make recommendations for the use of breadfruit fruit and flour in human

nutrition.

24 Chapter 2. Seasonality, Productivity, and Yield of Breadfruit

2.1. Synopsis

Increased commercialization and utilization of breadfruit will require basic research and data to select suitable germplasm and develop optimized production systems adapted to new locations (Jones et al., 2010). The wild seeded progenitor species of breadfruit, breadnut (A. camansi), is propagated by seed. Domesticated breadfruit, both seeded and seedless forms, is traditionally propagated by adventitious root shoots, root cuttings, or air layering. Recently mass propagation of breadfruit has become possible through in vitro micropropagation techniques (Murch et al., 2008;

Shi et al., 2007).It is believed that breadfruit is one of the most productive crops in the world with estimated yields of 6 t/ha on a dry weight basis in an orchard production system (Bowers, 1981). However, investigations on breadfruit yield are few in number and even less information is available comparing the yield of different cultivars

(Fownes and Raynor, 1993; Lebegin et al., 2007). Additionally, most cultivars are highly seasonal (Fownes and Raynor, 1993; Jones et al., 2010; Morton, 1987), but seasonality studies have been limited to investigations within single locations and it is difficult to predict how they may perform in new regions (Fownes and Raynor, 1993;

Lebegin et al., 2007). Since breadfruit is seasonal and highly perishable, unexpected changes in yield and/or seasonality could cause significant economic losses, wasted resources, and disrupt local food supplies. These losses may be mitigated through careful selection of cultivars to extend the season or to help plan for processing fruit into more stable products such as flour, chips, or frozen fruit (Ragone, 2011).

25 2.1.1. Objectives

The objective of the study was to investigate and quantify differences in yield, seasonality, and production of protein of breadfruit in a common garden. The study also determined the yield of the breadfruit protein. Each cultivar is represented by only one or two individual trees in the germplasm repository since each tree requires significant space and cultivar conservation is a priority. Therefore, true cultivar-to- cultivar comparisons are not possible and only comparisons of the performance of individuals over 7 years are made. Several factors, including propagation methods, the length of juvenile stage, yield, and fruit weight and size, are investigated to determine the degree of variability and identify suitable germplasm for international distribution. Breadfruit yield and seasonality are compared to previous reports (Jones et al., 2010; Lebegin et al., 2007) that used some of the same genetic materials planted in different locations.

2.2. Materials and methods

2.2.1. Germplasm repository

The Breadfruit Institute at the National Tropical Botanical Garden (NTBG) holds the world’s largest curated germplasm collection of breadfruit and its associated relatives including Artocarpus altilis, A. camansi, A. mariannensis, and A. altilis × A. mariannensis hybrids. The collection includes 325 well-documented trees collected from 34 Pacific Islands, the Philippines, Indonesia, Honduras and the Seychelles between 1978 and 2004 (Jones et al., 2011a; 2011b; 2013a). The majority of the trees are conserved in a single 12 acre (4.86 ha) site at Kahanu Garden in Maui

26 (20°47’57.07” N, 156°02’18.42” W) and have been described in detail previously

(Jones et al., 2013a). A subset of 26 accessions was established at the NTBG’s

McBryde Garden on Kauai, Hawaii (21°88’79.43” N, 159°49’23.15” W). Kauai has a subtropical climate and the major rainy season usually starts in October and ends in

April. The McBryde Garden receives rainfall of 939 mm annually with a mean temperature of 24.4 °C, mean maximum temperature of 29.0 °C, and mean minimum temperature of 19.7 °C. The field in land that had been in fallow since it was intensively cultivated with sugarcane until the early 1970s and the soil is compacted from this heavy use. Trees were planted 15 m apart giving a planting density of about 50 trees/ha, irrigated as needed, and mulched and fertilized yearly with a standard N:P:K fertilizer. In 2011, a cover crop of Lablab purpureus was planted beneath the A. camansi trees. Trees were pruned and shaped as they grew beginning with some cultivars in April 2008 and again in November 2011, with the exception of cultivars:

White and Rare autia.

2.2.2. Data collection

A standard methodology was used for all data collection (Table 2-1). Data were collected weekly from 2006 to 2012. Each tree was divided into four quadrants, and data were collected for each quadrant from southwest to southeast going clockwise around the tree. Using a visual assessment, male flowers and fruit in five stages of development were counted in each quadrant. Mature fruits were then harvested, weighed and measured. Fruits with any disease symptoms were counted, harvested and discard. Aborted fruit on the ground were counted. Total fruit number includes

27 new fruit, less than full-sized fruit, full-sized fruit, mature fruit, and ripe fruit.

Edible/harvestable fruit refers to full-sized fruit, mature fruit, and ripe fruit. Plant canopy area, percent leaf area, and plant height were analyzed using the ImageJ

1.47v software (Wayne Rasband, National Institutes of Health, Maryland, USA). Tree circumference was measured at knee height (0.35 m) rather than breast height due to the branching habit of some trees.

Table 2-1 The protocol for the collection of seasonality data for breadfruit accessions by visual estimates

Category Description

Male flowers Inflorescences of any size, small and just emerging from the sheath to full size.

New fruit Small fruits that have recently emerged from the leaf sheath; they are often prickly and the stigmas still protrude and remain green.

Less than full- A wide range of fruit sizes from bigger than the new fruit, 1/2- sized fruit size, up to almost full-sized.

Full-sized fruit Fruits that have reached their maximum size, but have not yet started to mature.

Mature fruit Fruits with distinctive characteristics of maturing, such as latex exudate on the skin and slight changes in skin color and texture.

Ripe fruit Fruits that are soft and ripe on the tree.

Fruits with disease Fruits with black spots or other disease symptoms evident on symptoms the peel.

28 2.3. Results

2.3.1. Time from planting to fruit production

Individual germplasm required different time to yield fruit (Table 2-2).

Artocarpus altilis grown from root cuttings started to bear fruit after an average of 29 months (Table 2-2) while hybrids (A. altilis × A. mariannensis) took significantly longer with an average of 37 months. Fast fruiting cultivars from root cuttings are Rare autia

(23 months), White (25 months), and Meitehid (27 months). Breadnut (A. camansi) trees grown from seeds produced fruit after an average of 39 months with a range from 32 to 57 months. Most of the cultivars produced fruit 4 years after planting (Table

2-2). In vitro-propagated trees of the cultivar, Ma’afala, began fruiting within 23 months

(Table 2-2). Since the repository contains only one tree of several varieties, these data are preliminary and more studies are need.

29 Table 2-2 The number of months between planting and fruit production for Artocarpus species and cultivars grown in the McBryde Garden, Kauai

Month First fruit NTBG ID Species Cultivar Planted date before first Source appear fruit Artocarpus Seed, 000523002 Dugdug April-21-04 Sep.-18-07 41 mariannensis Mariana Islands Seed, 000390002 A. camansi Kapiak April-21-04 Dec.-08-06 32 Papua New Guinea Seed, 000394002 A. camansi Kapiak April-21-04 Dec.-08-06 32 Papua New Guinea Seed, 000395002 A. camansi Kapiak April-21-04 Mar.-19-07 35 Papua New Guinea Seed, 000398005 A. camansi Kapiak April-21-04 Apr.-09-07 36 Papua New Guinea Seed, 000499003 A. camansi Kapiak April-21-04 Feb.-19-07 34 Papua New Guinea Seed, 000500002 A. camansi Kapiak April-21-04 Mar.-12-07 35 Papua New Guinea Seed, 000501004 A. camansi Kapiak April-21-04 May-05-07 37 Papua New Guinea

30 Month First fruit NTBG ID Species Cultivar Planted date before first Source appear fruit Seed, 000501005 A. camansi Kapiak April-21-04 Feb.-12-07 34 Papua New Guinea Seed, 000502002 A. camansi Kapiak April-21-04 Nov.-25-08 55 Papua New Guinea Seed, 000503004 A. camansi Kapiak April-21-04 Jan.-07-09 57 Papua New Guinea A. altilis x A. Roots, 030034001 Meinpadahk April-28-04 Jun.-25-07 38 mariannensis Kahanu Garden A. altilis x A. Roots, 030034002 Meinpadahk April-28-04 Jun.-18-07 38 mariannensis Kahanu Garden A. altilis x A. Roots, 030041001 Rotuma April-28-04 May-13-07 37 mariannensis Kahanu Garden A. altilis x A. Roots, 030045001 Yap April-28-04 Apr.-16-07 36 mariannensis Kahanu Garden Roots, 040051001 A. altilis Afara January-03-06 May-21-08 29 Kahanu Garden Roots, 030033001 A. altilis Meitehid June-20-05 Sep.-18-07 27 Kahanu Garden

31 Month First fruit NTBG ID Species Cultivar Planted date before first Source appear fruit Roots, 030035001 A. altilis Otea January-03-06 May-21-08 29 Kahanu Garden Roots, 030037001 A. altilis Porohiti April-28-04 Jun.-18-07 38 Kahanu Garden Roots, 040518001 A. altilis Rare autia January-03-06 Nov.-02-07 23 Kahanu Garden Roots, 030042001 A. altilis Toneno April-28-04 Apr.-16-07 36 Kahanu Garden Roots, 040063001 A. altilis Ulu fiti June-20-05 Oct.-04-07 28 Kahanu Garden Roots, 030028001 A. altilis White June-20-05 Jul.-24-07 25 Kahanu Garden November-21- in vitro, 070659021 A. altilis Ma’afala Sep.-02-10 22 08 Murch in vitro, 070659022 A. altilis Ma’afala October-20-08 Sep.-03-10 23 Murch

Average 34

32 2.3.2. Breadfruit seasonality

Individual breadfruit cultivars (Artocarpus spp.) produced fruit at different times throughout the year with the major fruiting season from July to November (Figure 2-1 a&b). Monthly fruit production during the peak season was as high as 145 fruit per tree for Meinpadahk, and as low as 16 fruit for Meitehid. Some cultivars, such as

Rotuma (hybrid), White (A. altilis), Toneno (A. altilis), and Mos en Samoa (A. altilis) had another small production peak during the spring (146 fruit for the entire year).

Highly productive cultivars including Meinpadahk (hybrid), Rotuma (hybrid), and Mos en Samoa (A. altilis) maintained twice the average fruit production throughout the year

(Figure 2-1a). Seasonal differences were less obvious for breadnut (A. camansi) and fruit production was lower (Figure 2-1b). About 70% of the trees in our study were in fruit for > 80% of the years measured (Figure 2-2).

Figure 2-1 a

250

200

150

100

0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Average fruit number per tree fruit pernumber Average Month

Afara Meinpadahk Meitehid Mos en Samoa Otea Porohiti Puou Rare autla Rotuma Toneno Ulu fiti White Yap A. camansi A. mariannensis Average

Figure 2-1 b 100 90 80 70 60 50 40 30 20 Average fruit number Average 10 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month Artocarpus altilis A. altilis × A. mariannensis A. camansi A. mariannensis

Figure 2-1 Average fruit number per tree of breadfruit cultivars planted from 2008 to 2012 (a). Each cultivar has one tree except for A. camansi (n=10) and Meinpadahk (n=2). Average fruit number per month of Artocarpus altilis (n=10), A. camansi (n=4), A. mariannensis (n=1), and A. altilis × A. mariannensis (n=10) from 2008 to 2012 (b). Bars represent the standard error of the mean of each month over the 4-year data collection period

Figure 2-2 a

100%

80%

60%

40% fruit/fruit 20%

0% Years in maleflowers/edible in Years Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month Male Flowers Edible Fruit Fruit

34 Figure 2-2 b

100%

80%

60%

fruit/fruit 40%

20%

Trees in maleflowers/edible in Trees 0% Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month Male Flowers Edible Fruit Fruit

Figure 2-2 The average percentage of years (a) and average percentage of breadfruit trees (b) that breadfruit trees bore fruit, edible fruit and male flowers from 2007 to 2012 in National Tropical Botanical Garden in Kauai, Hawaii. Edible/harvestable fruit refers to full-sized fruit, mature fruit, and ripe fruit. Bars represent the standard error of the mean

Seasonality in Kauai and Maui

In terms of seasonality difference between Kauai and Maui, most cultivars, including Afara, Otea, Meinpadahk, Porohiti, Meitehid, Ulu fiti, and Yap, bore edible fruit (full size, mature and/or ripe fruit) at similar times (Figure 2-3). Meitehid had a longer low season in Kauai that lasted from April to August. Ulu fiti maintained the same seasonality as well as the same likelihood of fruiting in Kauai and Maui. Yap tended to be less seasonal than most cultivars and it kept this attribute when it was planted in Kauai. Puou, and White shared similar seasonality in Kauai and Maui, with a 2-month delay in the onset of the low season in Kauai. Rotuma and Toneno displayed different seasonality profiles in Kauai than in Maui. In Maui, Rotuma has a

35 distinct season between November and February and a low season between April and

August. However, when it was planted in Kauai, Rotuma produced fruit much more sporadically and was less seasonal. In Maui, Toneno has a peak season from October to January and a low season from March to July. In Kauai, the peak season shifted to

May to July and the low season moved to July to September. Overall, edible fruit season remained relatively consistent between Maui and Kauai.

100% 100%

80% 80%

60% 60%

40% 40% Years Years in edible fruit 20% Years in edible fruit 20%

0% 0%

Jul

Apr Oct

Jan Jun

Feb Mar

Apr Oct

Aug Sep Nov Dec

Jan Jun

May

Mar Feb

Nov Dec

Aug Sep May Month Month Afara in Kauai Afara in Maui Otea in Kauai Otea in Maui

100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Years Years in edible fruit Years Years in edible fruit

0% 0%

Jul

Apr Oct

Jul

Jan Jun

Feb Mar

Aug Sep Nov Dec

Apr Oct

May

Jan Jun

Feb Mar

Nov Aug Sep Dec May Month Month

Meinpadahk in Kauai A. mariannensis in Kauai Meinpadahk in Maui A. mariannensis in Maui

36 100% 100%

80% 80%

60% 60%

40% 40%

Years Years in edible fruit 20%

0% 0%

Jul

Apr Oct

Jan Jun

Apr Oct

Feb Mar

Jan Jun

Nov Dec

Aug Sep

Feb Mar

May

Nov Dec

Aug Sep May Month Month Porohiti in Kauai Meitehid in Kauai Porohiti in Maui Meitehid in Maui

100% 100% 80% 80%

60% 60%

40% 40%

Years Years in edbile fruit 20%

Years Years in edible fruit 20%

0% 0%

Jul

Apr Oct

Jan Jun

Mar

Feb

Aug Sep Nov Dec

Apr Oct

Jan Jun

May

Mar

Feb

Nov Dec Aug Sep May Month Month Yap in Kauai Yap in Maui Ulu fiti in Kauai Ulu fiti in Maui

100% 100% 80% 80% 60% 60% 40% 40%

Years Years in edible fruit 20%

Years Years in edible fruit 20% 0%

Jul

Apr Oct

Jan Jun

Mar

Feb

Aug Sep Nov Dec

May

Jul

Apr Oct

Jan Jun

Feb Mar

Nov Dec

Aug Sep May Month Month Rotuma in Kauai Rotuma in Maui Puou in Kauai Puou in Maui

37 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Years Years in edible fruit Years Years in edible fruit

0% 0%

Jul

Apr Oct

Jan Jun

Apr Oct

Jan Jun

Mar

Feb

Aug Sep Nov Dec

Mar Feb

Nov Aug Sep Dec

May May Month Month White in Kauai White in Maui Toneno in Kauai Toneno in Maui

Figure 2-3 Comparison of individual edible fruit seasonality profiles of breadfruit cultivars planted in the National Tropical Botanical Garden in Kauai and Maui, Hawaii. Edible/harvestable fruit refers to full-sized fruit, mature fruit, and ripe fruit

Similar trends were seen for the male flower season between trees in Kauai and Maui (Figure 2-4). Afara and Yap had almost the same seasonality and number of years with trees producing male flowers. Cultivars such as White, Porohiti and

Meitehid shared similar seasonality in Kauai and Maui, but had a greater number of years with male flowers. Rotuma, Ulu fiti, and Otea had similar seasonality but a 2 to

3 months longer peak season. Puou, Meinpadahk, and Toneno experienced a 2- month delay in the flowering season in Kauai. Overall, breadfruit kept the same male flower seasonality between Kauai and Maui.

38 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Years Years malein flowers Years Years in maleflowers

0% 0%

Jul

Apr Oct

Jan Jun

Feb Mar

Mar

Feb

Nov Dec

Aug Sep

Aug Sep Nov Dec

May May Month Month A. mariannensis in Kauai A. mariannensis in Maui White in Kauai White in Maui

100% 100% 80% 80%

60% 60%

40% 40%

20% 20%

Years Years malein flowers Years Years in maleflowers

0% 0%

Jul

Apr Oct

Jan Jun

Feb Mar

Nov Dec

Aug Sep

Nov Dec

Aug Sep

May May Month Meitehid in Kauai Month Porohiti in Kauai Meitehid in Maui Porohiti in Maui

100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Years Years in maleflowers Years Years in maleflowers

0% 0%

Jul

Apr Oct

Jan Jun

Feb Mar

Jul

Aug Sep Nov Dec

May

Apr Oct

Jan Jun

Feb Mar

Nov Dec

Aug Sep May Month Month

Afara in Kauai Afara in Maui Yap in Kauai Yap in Maui

39 100% 100%

80% 80%

60% 60%

40% 40%

20% 20% Years Years in maleflowers Years Years in maleflowers 0%

Jul

Apr Oct

Jan Jun

Mar

Feb

Aug Sep Nov Dec

Jul

May

Apr Oct

Jan Jun

Mar

Feb

Aug Sep Nov Dec Month May Month Rotuma in Kauai Rotuma in Maui Ulu fiti in Kauai Ulu fiti in Maui

100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

Years Years in maleflowers Years Years in maleflowers

0% 0%

Jul

Apr Oct

Jan Jun

Mar Feb

Jul

Nov Dec

Aug Sep

May

Apr Oct

Jan Jun

Feb Mar

Nov Dec

Aug Sep May Month Month Otea in Kauai Otea in Maui Puou in Kauai Puou in Maui

100% 100%

80% 80%

60% 60%

40% 40%

20% 20% Years Years in maleflowers 0% Years in maleflowers

Jul

Apr Oct

Jan Jun

Mar

Feb

Sep Aug Nov Dec

Jul

May

Apr Oct

Jan Jun

Feb Mar

Dec Aug Sep Nov Month May Month Meinpadahk in Kauai Meinpadahk in Maui Toneno in Kauai Toneno in Maui

Figure 2-4 Comparison of individual male flower seasonality profiles of breadfruit cultivars planted in the National Tropical Botanical Garden in Kauai and Maui, Hawaii

40 2.3.3. Breadfruit productivity/fruit number

On average 64% of the fruit was produced during the peak season for A. altilis and 58% for hybrids (Figure 2-5). The average fruit number (A. altilis and hybrid) during the entire year was approximately 40 fruit per month; 59 fruit per month during the peak season; and 26 fruit per month during the low season. Dugdug (A. mariannensis) had the most distinct peak season from August to November with a fruit production of 160 (70%). A. camansi trees produced an average of 47 ± 8 fruit produced by the juvenile trees. As the trees aged, the number of fruits increased to

130 ± 22 fruit per tree after 7 years and continued to increase throughout the study

(Figure 2-5). After 7 years, hybrids (A. altilis × A. mariannensis) produced an average of 131 ± 36 fruit per tree, while A. altilis produced an average of 129 ± 28 fruit per tree

(Table 2-3). For 7-year-old breadnut (A. camansi) trees, the total fruit production was lower at about 52 ± 9 fruit per tree. Hybrids produced higher yields than the other species at both 4 and 7 years after planting. Based on these 7 years of field data, most cultivars showed a linear increase in total fruit number per year after initiating fruit production. As a group, the total fruit number increased from 3 to 7 years after planting following a linear equation,푦 = 27.18푥 − 21.43, R2=0.9684 (Figure 2-5). The average increase in fruit number per year was about 25.0, but it varied among cultivars. Porohiti had the highest growth rate at 42.9 (푦 = 42.9푥 − 37.8, R2=0.9258), followed by Meinpadahk at 42.3 (푦 = 42.3 − 40.1, R2=0.9537), and White at 38.7 (푦 =

38.7푥 − 30.1, R2=0.8131).

41 350 y = 27.181x - 21.428 R² = 0.9684 300

250

200

150

100 Total fruit number pernumberyear fruit Total 50

0 2 Years 3 Years 4 Years 5 Years 6 Years 7 Years 8 Years Years after being planted A. camansi Afara A. mariannensis Meinpadahk Meitehid Mos en Samoa Otea Porohiti Rare autla Rotuma Toneno Ulu fiti White Yap Average Linear (Average)

Figure 2-5 Average fruit number per year of breadfruit cultivars planted in the National Tropical Botanical Garden in Kauai, Hawaii from 2 years to 8 years old. Bars represent the standard error of the mean over the individual tree. A. camansi (n=10).

42 Table 2-3 The summary of the total fruit number, average fruit weight, and yield in Artocarpus species after being planted 4 years and 7 years. Different plant number exists due to the difference in plant age

Years after Artocarpus Artocarpus A. altilis x A. Category Artocarpus altilis being planted camansi mariannensis mariannensis

Plant number 10 1 7 4

Total fruit number 11±4 6.00 45±10 49±20 4 Years Average fruit weight 344.00 - 1181.12 735.47 (g) Yield (t/hectare) 0.38 - 3.20 3.80

Plant number 10 1 5 4

Total fruit number 52±9 106.00 129±28 131±36 7 Years Average fruit weight 1612.26±71.70 563.5±52.95 1170.01 2038.02 (g) Yield (t/hectare) 6.7 5.7 10.46 17.12

Plant number 10 1 4 4

Canopy area (m2) 41.73 32.41 23.26 37.90

9 Years % Leaf area 84.30 96.43 90.33 95.16

Perimeter (m) 0.91 0.89 0.94 1.07

Height (m) 7.61 5.51 5.14 5.91

43 2.3.4. Breadfruit fruit weight

The harvested fruit generally weighed between 1.0 kg to 2.0 kg during the study years depending on the cultivar (Figure 2-6). The average fruit weight was 1.2 kg across cultivars. However, differences in fruit size were found among cultivars and the extent of these differences increased as the trees grew older. For 4-year-old trees,

Toneno, Rotuma, Rare autia, and Otea cultivars had average fruit weights over 1.0 kg; Afara and Mos en Samoa weighed over 0.5 kg; and Meitehid weighed less than

0.5 kg. At 7 years after planting, Yap, Toneno, Rotuma, and Mos en Samoa grew heavier, with fruit over 1.5 kg; Meitehid, Meinpadahk, and Rare autia weighed over

1.0 kg, while Porohiti weighed over 0.5 kg. Over this 3-year-period, Rotuma and

Porohiti produced increasingly heavier fruit with an average increase around 0.58 kg.

Most cultivars, including Meitehid, Meinpadahk, Rare autia, Mos en Samoa, Ulu fiti and White, produced fruits that increased in weight by 0.98 kg over the 3 years of production. Seven years after planting, the average fruit weight for hybrids (A. altilis ×

A. mariannensis) was 2.0 ± 0.63 kg, while the average weight for A. altilis was 1.2 ±

0.15 kg (Table 2-3).

44 3500

3000

2500

2000

1500

1000

Average fruit weight fruit (g) weight Average 500

0 2 Years 3 Years 4 Years 5 Years 6 Years 7 Years 8 Years Years after Being Planted A. camansi Afara A. mariannensis Meitehid Meinpadahk Mos en Samoa Otea Porohiti Rare autia Rotuma Toneno Ulu fiti White Yap Figure 2-6 Average fruit weight for breadfruit cultivars from 2 years old to 8 years old. Bars represent the standard error of the mean for fruit produced in each year. Please note: Data was unavailable for certain years in study and a different number of fruits was measured for each tree of each variety in each year

2.3.5. Breadfruit yield

Based on the planting density of 100 trees/ha used in our study, the average projected yield for 4-year-old trees was (3.92 ± 0.27) t/ha, with the highest yield of (5.5

± 0.02) t/ha for Otea and the lowest yield of (1.52 ± 0.02) t/ha (Figure 2-7). At 7 years after planting, the projected average yield reached (12.68 ± 1.01) t/ha but was as high as (18.74 ± 0.05) t/ha for Meinpadahk and as low as (4.56 ± 0.04) t/ha for Rare autia.

The average annual yield for the hybrid cultivars (A. altilis × A. mariannensis) included in our study was 17.12 t/ha after 7 years, while for A. altilis, the number was lower at

10.46 t/ha.

45 30 y = 2.4461x - 2.591 R² = 0.9642 25

10 Annual yield yield (t/hectar)Annual

0 2 Years 3 Years 4 Years 5 Years 6 Years 7 Years 8 Years Years after being planted A. camansi Afara A. mariannensis Mei tehid Meinpadahk Mos en Samoa Otea Porohiti Rare autia Rotuma Toneno Ulu fiti White Yap Average Linear (Average)

Figure 2-7 Yield for breadfruit cultivars from 2 years to 8 years old. Data is unavailable for certain years in study. Bars represent the standard error of the mean over the individual tree. A. camansi (n=10), Meinpadahk (n=2). The yield was calculated as follow: Average fruit weight × (Average total fruit number per year −Aborted fruit − Fruit with rots) × 100 trees

2.3.6. Tree size and growth

According to our results, hybrids had the largest trunk circumference (1.07 m) while A. mariannensis had the smallest (0.89 m) (Table 2-3). Hybrids had a larger canopy area (37.90 m2), higher percent leaf area (95.16%) and were taller (5.91m) than A. altilis (23.26 m2, 90.33%, and 5.14 m, respectively). Breadnut (A. camansi) had the highest tree height at 7.61 m and the largest canopy area 41.73 m2; the percent leaf area was the lowest.

46 2.3.7. Commercial breadfruit Ma’afala

The fruiting season for commercial breadfruit cultivar, Ma’afala, was from Oct. to Dec. after being planted 3 years and July to Oct. after being planted 4 years

(Figure 2-8a). The average fruit weight of Ma’afala grown in Kauai was 576 g, with the smallest fruit measured at 425 g in May 2011 and the largest fruit measured at

886 g in September 2011 (Figure 2-8b).

Figure 2-8 a 500

400

300

200 number

100 Ma'alafa total fruit total Ma'alafa 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

2011 2012 Figure 2-8 b

1200 1000 800 600

400 weight (g) weight 200

Ma'afala average fruit average Ma'afala 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2011 2012

Figure 2-8 Total fruit number (a), and average fruit weight (b) for breadfruit cultivar Ma’afala in the National Tropical Botanical Garden at on Kauai. The number is an average of two trees planted in the garden in 2011 and 2012.

47 2.4. Discussion

2.4.1. Breadfruit seasonality and production in Kauai

The major season for breadfruit production in our study of Kauai is from July to

November. However, not all cultivars produced edible fruit at the same time and there were definite seasonal patterns. In our study, breadfruit trees produced some edible fruit each month every year, therefore careful selection of cultivars can ensure yearlong production of breadfruit. The fruit matured 3 to 4 months later than the male flowers emerged, which is consistent with previous studies (Jones et al., 2010).

Breadfruit produced around 260 to 280 fruit per tree during the peak season, which is higher than previous estimates of about 150–250 fruit per tree per year

(Morton 1987; Ragone 2011). Based on an average fruit moisture content of about

68% (Jones et al., 2011b), hybrids yield 5.4 t/ha dry fruit and A. altilis yields 3.4 t/ha dry fruit by 7 years after planting. This is lower than the estimated breadfruit production of 6 t/ha on a dry weight basis in an orchard production system in Bowers’s study in

1981. The fruit number and fruit yield presented in this study showed a linear increase from 3 to 7 years after plant. The linear relationship and high R2 value suggest that the production will continue to increase in the future and the full potential fruit production of a mature orchard is not reached. Nevertheless, our reported yield still compares favorably to the average global yields of rice wheat, or corn at (4.43, 3.25, and 5.64) t/ha, respectively (USDA, 2015b). The protein production for breadfruit

(including both A. altilis and hybrids) is 0.16 t/ha on average, based on 1.2% protein content reported by Jones et al. (2010).

48 Hybrids showed an advantage in fruit number, weight, and yield. The examination of growth tree size and growth also showed that hybrids had larger canopy area, higher percent leaf area and taller tree. Based on these data, it could be concluded that heterosis may exist in Artocarpus hybrids. Heterosis shows the superiority of growth rate, reproductive success, and yield (Lippman and Zamir,

2007). Hybridization and introgression have been observed and applied to increase crop yields since ancient times (de Ribou et al., 2013). An increase in yield ranging from 15% to 50% has been reported for maize, sorghum, rice, and sunflower.

Hybrids also can have advantages for pathogen-resistance traits or better adaptation to extreme climate or new light regimes (de Ribou et al., 2013).

In spite of high fruit production, fruit dropping was observed in the study.

Given this observation, it may be beneficial to thin the fruit and/or prune the tree to encourage fruit maturation, as is commonly practiced with other tree fruits such as apple, pear, peach, and cherry (Robinson 2007; Sane et al., 2012).

2.4.2. Comparison of breadfruit productivity across locations

The yield data obtained in Kauai was higher than what has been reported using the same genetic stock (Colones of Ma’afla and Rotuma) in New Caledonia (Lebegin et al., 2007) (Table 2-4). In New Caledonia, Rotuma had an average fruit weight of

679 ± 123.7 g, which is less than half of that observed in Kauai 1443 ± 65.4 g.

Likewise, fruit length was 11.7 ± 0.3 cm in New Caledonia compared to 16.4 ± 0.3 cm in Kauai. The average weight and size in Kauai were closer to what were observed in

Maui with a range of 12–16 cm × 12–15 cm and an average weight at 1.1 kg (Ragone,

49 2011). Ma’afala fruits were also about twice as large in the Kauai collection with an average fruit weight of 609 ± 21 g compared to 300 ± 72.1 g in New Caledonia (Table

2-4). Ma’afala fruit size was also significantly larger for trees grown in Kauai with fruit measuring 12 ± 0.2 cm compared to 9.0 ± 0.9 cm in New Caledonia (Lebegin et al.,

2007). The average weight and size of Ma’afala fruits in Kauai were closer to previous reports in the existing literature (12–16 cm × 10–13 cm, 0.8 kg) (Ragone, 2011). While

New Caledonia and Kauai are similar distances from the equator (Table 2-5), the trees planted in New Caledonia suffered from water logging issues and a suspected problem with soil-borne diseases (Lebegin et al., 2007). As such, the lower fruit sizes and productivity observed in New Caledonia are likely a result of micro-climatic and biotic variables rather than the overall environment.

Table 2-4 Comparison of fruit growth between Rotuma and Ma’afala in Kauai and New Caledonia (Lebegin et al., 2007)

Rotuma Ma'afala New New Location Kauai Kauai Caledonia Caledonia

Average fruit weight (g) 679±123.7 1143±65.4 300±72.1 609±21

Average fruit size (cm) 11.7±0.3 16.4±0.3 9.0±0.9 12±0.2

50 Table 2-5 Comparison of the environmental and climate factors for breadfruit plantings in Kauai and Maui, Hawaii, and New Caledonia (WRCC, 2016)

Kauai Maui New Caledonia

21°88’79.43”N, 20°47′57.07″N, 19-22°S, Location 159°49’23.15”W 156°02′18.42″W 158-162°E

Mean temp. (°C) 24.4 24.3 23.1

Mean max temp. (°C) 29 27.1 26

Mean min temp. (°C) 24.4 19.7 20.2

Precipitation (mm) 939 2051 1530

Compared to previous studies with breadnut (A. camansi) trees, the data obtained in Kauai showed higher variation (Table 2-6). In Roberts-Nkrumah’s experiment (Roberts-Nkrumah, 2005), three breadnut trees were grown at the

University of the West Indies Field Station (UWI-FS) in Valsayn, Trinidad and Tobago

(lat. 10°2’ to 11°12’ N, long. 60°30’ to 61°56’ W), which has a mean temperature of

26.8 °C and mean precipitation of 1575 mm. In her study, there were positive correlations between seed mass and seed number (r = 0.87), seed mass and fruit mass (r = 0.83), and seed number and fruit mass (r = 0.77). There was no difference between years or plant age. The fruit size and weight observed in Kauai were larger than they were in UWI-FS in all ages. However, seed number and seed mass were lower in Kauai after 6 and 8 years than they were in UWI-FS. These comparisons indicate that the positive correlations found in Roberts-Nkrumah’s study may not apply when comparing data between her study and our study (Table 2-7). In Kauai, breadnut had a higher seed weight than at UWI-FS. Breadnut in Kauai had a lower percentage 51 of seed mass per fruit at all ages. Seed masses from both locations were lower than

50% of the fruit. Fruit and seed production per tree varied dramatically from year to year (30 kg to 140 kg, 10 kg to 60 kg, respectively). Trees in Kauai were grown from seeds collected from the wild and cultivated trees in Papua New Guinea, the center of origin for this species, while the ones at UWI-FS were likely from a single seed source. Variation in fruit and seed production between individuals was found in Kauai.

52 Table 2-6 Comparison of Artocarpus camansi (breadnut) production in McBryde Garden in Kauai with the University of the West Indies Field Station in Valsayn, Trinidad, and Tobago (Roberts-Nkrumah 2005)

Seed Seed Years Average Average Average Number Seed Fruit Average Seed mass mass/ After fruit fruit fruit of production production Location number weight per fruit being weight length width seeds per tree per tree of fruit (g) fruit mass planted (kg) (cm) (cm) per fruit (kg) (kg) (g) (%) UWI Field 126 1.14 16.16 13.18 60 8.14 471.28 59.38 143.64 41.00 5 years Station McBryde 19 1.51 16.50 15.00 44 11.41 502.00 9.54 28.61 33.33 Garden UWI Field 68 1.06 16.54 13.11 61 8.34 502.79 34.36 72.43 46.00 6 years Station McBryde 36 1.92 17.77 16.77 53 8.42 443.33 15.92 69.10 23.03 Garden UWI Field 25 0.89 14.41 12.40 44 7.56 327.78 8.30 22.54 36.00 7 years Station McBryde 52 1.61 18.27 15.37 49 8.73 428.26 22.23 83.68 26.56 Garden UWI Field 30 0.92 14.52 12.21 59 8.17 475.00 14.09 27.30 42.00 8 Years Station McBryde 23 1.33 18.66 14.55 31 9.18 284.46 6.63 31.06 21.34 Garden

53 2.4.3. Comparison of breadfruit seasonality across locations

As previously reported there does not appear to be an influence of precipitation patterns on seasonality in breadfruit (Quartermain, 2007). The McBryde Garden on

Kauai shares a similar rainy season as the Kahanu Garden on Maui. However, the

Kauai location receives about half of the annual precipitation reported for the Kahanu

Garden (Table 2-5) (Shi et al., 2007). The seasonality of male flowers did not vary between the two locations, indicating that the amount of precipitation is not the dominant factor in determining edible fruit/male flower seasonality. Jones et al. proposed a hypothesis that seasonality of breadfruit is closely related to the distance from the equator, potentially due to differences in the light quality/spectrum throughout the season (Jones et al., 2010). In this hypothesis, breadfruit flowering is induced when the sun reaches its zenith before the summer months and subsequent fruiting extends throughout the summer months. Kauai and Maui share a similar longitude and latitude (Table 2-5); the similarity of the breadfruit seasonality in these datasets provides evidence in support of this hypothesis, but further studies in more widely different locations are required.

2.4.4. Commercial breadfruit Ma’afala

Ma’afala has been the most widely distributed breadfruit cultivar. In the Kahanu

Garden germplasm collection on Maui, mature Ma’afala trees produced fruits most reliably from July to December (Jones et al., 2010). Juvenile clones of this tree planted on Kauai displayed a slight delay in the first two fruiting seasons. However, as the trees matured, the season became more similar to that of more established trees in

54 the Kahanu Garden on Maui, with the most fruits being produced in July of 2012.

Based on these data, a slight shift in seasonality of juvenile breadfruit should be considered for agricultural applications. The flowering periods occurred 3 to 4 months earlier than the fruiting season, which is consistent with previous observations (Jones et al., 2010). The recommended planting density for Ma’afala agricultural practice is

120 trees/ha, which means Ma’afala could be expected to yield approximately 20 t/ha by 8 years after planting. Based on a protein content of 2.6% (Jones et al., 2011a),

20 t/ha production of Ma’afala translated to a protein yield of 0.52 t/ha, which is higher than rice, wheat, and potato, but lower than corn and soybean (Table 2-7). Another factor that needs to be taken into account is that other staples, soybean or corn have a long history of cultivation and commercialization, while breadfruit agriculture has very limited studies. The potential yields of a mature breadfruit orchard can be greater than what is reported here.

Table 2-7 Comparison of the protein yields between Ma’afala and other staples

Protein Content Yield Staple Protein yield t/ha (%) t/ha

Cassava 1.11 13.52 0.15

Potato 1.31 19.42 0.25

Rice 6.71 4.463 0.29

Wheat 12.21 3.253 0.40

Ma’afala 2.6 20 0.52

Corn 9.51 5.643 0.54

Soybean 381 2.73 1.03 1 (FAO, 1981), 2 (USDA, 2015c), 3 (FAOSTAT, 2013)

55 2.5. Summary

The current data describes breadfruit growth habits with a comparison of yield and seasonality, demonstrating the range of diversity among cultivars. Generally speaking, breadfruit, including hybrids, requires around 30 months to fruit from root cuttings. Once established, a breadfruit tree can produce over 250 fruit a year with an average weight of 1.2 kg. Around 60% of fruits are produced during the season from

July to November. Our study demonstrated that the expected yield for breadfruit after

7 years is 10-15 t/ha. Differences exist in cultivars. As expected, hybrids have many advantages, including yield, average larger fruits for mature trees, and denser tree canopies. Even though some cultivars (A. altilis: Toneno and White; A. altilis × A. mariannensis: Rotuma and Meinpadahk) have certain agricultural advantages, selecting cultivars for international distribution also depends upon the flavor, texture, taste, etc. of the cultivar.

To date, Ma’afala is the first breadfruit cultivar available for widespread distribution. In vitro propagated Ma’afala trees produced fruits within 23 months and these data indicate that an orchard of established Ma’afala trees will produce 20 t/ha of fruit by 8 years. This will translate to 0.52 t/ha protein yield which is higher than wheat, rice, and potato, and close to corn and soybean.

Seasonality for male flower and fruit remains similar between the two locations compared in this project, supporting the hypothesis that seasonality of breadfruit is closely related to the distance from the equator and incidence of light (Jones et al.,

2010). Further studies should determine if fundamental characteristics of breadfruit cultivars are conserved across disparate geographic regions and distant climates.

56 Chapter 3. Amino Acid Analysis of Breadfruit Protein

3.1. Synopsis

The fundamental requirement for maintaining good health is access to sufficient food of adequate quality (WHO, 2007), but adequate food is limited in many parts of the world and climate change seems likely to disrupt food supplies by 2050 (Porter and Xie, 2014). In comparison to other tropical crops, breadfruit contains an average of 3.9 % protein on a dry weight basis, which is 1.15 % higher than cassava, 1.1 % higher than banana, and 0.3 % higher than sweet potato (Jones et al., 2011a). One breadfruit cultivar, Ma’afala, has an average of 7.6 % protein based on dry weight, which is comparable to rice (Jones et al., 2011a). Previous work has shown that breadfruit contains all of the essential amino acids (Golden and Williams, 2001;

Morton, 1987). Leucine (0.61 ± 0.01 g/100 g) and lysine (0.8 ± 0.22 g/100 g) made up to 30 % of the total amino acid content during the ripe developmental stage of breadfruit (Golden and Williams, 2001). However, this research was based on only one (Golden and Williams, 2001) or few (Morton, 1987) cultivars and protein quality was not assessed. Protein quality is determined by the essential amino acid content and ratio of amino acids (WHO, 2007). Essential amino acids are indispensable and cannot be synthesised by the human body including lysine (lys), leucine (leu), threonine (thr), tryptophan (trp), histidine (his), isoleucine (ile), valine (val), phenylalanine (phe), tyrosine (tyr), methionine (met), and cysteine (cys) (WHO 2007 ).

Since protein content varied widely depending on the cultivar (Jones et al., 2011a), an assessment of the variability of proteins between cultivars and of the protein quality of individual cultivars was warranted.

57 3.1.1. Objectives

The objective was to evaluate and describe the variation in the amino acid composition of a genetically diverse group of breadfruit cultivars growing in a single location to determine how the essential amino acids vary among cultivars.

3.2. Materials and method

3.2.1. Breadfruit sample collection and preparation

A detailed description of the sample collection and experimental design has been published previously (Jones et al., 2011a, 2013b). Three mature breadfruits were collected from 49 breadfruit cultivars including 41 Artocarpus altilis and 8 hybrids

(A. altilis × A. mariannensis) (Table 3-1). Since different breadfruit cultivars produce fruit in different months, the collection process lasted several months during

November 2008 to December 2010 (Jones et al., 2010, 2013b). After removing the stem, fruits were inverted for about 1 h to drain latex. The skin was removed with a good quality household vegetable peeler, the fruit was cut into quarters, and the core was removed. The edible portion of the fruit was cut into 2 cm slices, frozen, and shipped to the University of British Columbia Okanagan. All the breadfruit samples were stored at −86 °C before analysis.

58 Table 3-1 The cultivar list of breadfruit (Artocarpus altilis and A. altilis × A. mariannesis) used in the amino acid analysis

Island of Grid NTBG ID Cultivar Plant date Species origin 8 900233002 Pulupulu 28/03/1986 A. altilis Samoa Society 9 790492001 Porohiti 17/09/1975 A. altilis Islands Society 12 790488001 Toneno 11/09/1975 A. altilis Islands Society 13 790491001 Tuutou 11/09/1975 A. altilis Islands Society 16 790485001 Puou 16/09/1975 A. altilis Islands Society 20 790486001 Roihaa 20/09/1975 A. altilis Islands Society 27 790487001 Huehue 17/09/1975 A. altilis Islands Society 30 780333001 Ahani 15/09/1975 A. altilis Islands Society 32 780325001 Afara 13/09/1975 A. altilis Islands 35 890258001 Ulu fiti 08/02/1985 A. altilis Samoa Society 36 800269001 Mahani 15/09/1975 A. altilis Islands 38 810289002 Yellow 06/01/1978 A. altilis Seychelles Society 39 780327001 Otea 11/09/1975 A. altilis Islands Society 40 780330002 Fafai 17/09/1975 A. altilis Islands 43 810290001 White 05/01/1978 A. altilis Seychelles Havana Society 47 780291001 17/09/1975 A. altilis pataitai Islands 49 790497002 Meinuwe 05/01/1978 A. altilis Micronesia Society 51 780338001 Tapehaa 03/01/1975 A. altilis Islands 55 770517001 Ma’afala 01/01/1975 A. altilis Samoa A7 900232001 Atu 02/04/1986 A. altilis Cook Islands A8 900264001 Uto ni viti 30/03/1986 A. altilis Fiji Marquesas B6 900237001 Mei puou 01/04/1986 A. altilis Islands Marquesas B8 900242001 Mei kopumoko 26/03/1986 A. altilis Islands D9 910271001 Mei uhpw 12/07/1987 A. altilis Micronesia

59 Island of Grid NTBG ID Cultivar Plant date Species origin E5 900266002 Meiarephe 29/03/1986 A. altilis Micronesia Marquesas F6 900241001 Mei aueka 29/03/1986 A. altilis Islands Society H7 900246001 Tuutou auena 01/04/1986 A. altilis Islands Society I6 900247001 Tuutou ooa 29/03/1986 A. altilis Islands Society I8 900249002 Anahonaho 01/04/1986 A. altilis Islands K7 900260001 Samoan 26/03/1986 A. altilis Samoa M6 900262001 Manua 01/04/1986 A. altilis Samoa Society P7 890464001 Ouo 30/07/1985 A. altilis Islands P8 880690001 Kea 01/04/1986 A. altilis Tonga R4 890477001 Uto samoa 01/08/1985 A. altilis Samoa Society S7 890152002 Puurea 01/08/1985 A. altilis Islands Society T6 890460001 Puaa 29/07/1985 A. altilis Islands Society V3 890463001 Patara 29/07/1985 A. altilis Islands V4 890159002 Meriaur 26/07/1985 A. altilis Palau V6 890470001 Furau 26/07/1985 A. altilis Samoa W3 890471001 Uto dina 29/07/1985 A. altilis Samoa Hamoa Society Y1 890154001 11/02/1985 A. altilis (Maopo) Islands A. altilis x. A. Society 1 790489001 Piipiia 13/09/1975 mariannensis Islands A. altilis x. A. 19 890161001 Yuley 10/02/1985 Micronesia mariannensis A. altilis x. A. A9 910269001 Faine 15/07/1987 Micronesia mariannensis A. altilis x. A. B5 900255001 Meinpwahr 29/03/1986 Micronesia mariannensis A. altilis x. A. E6 910272002 Meinpohnsakar 09/07/1987 Micronesia mariannensis A. altilis x. A. J9 910268001 Meion 09/07/1987 Micronesia mariannensis A. altilis x. A. L8 890183002 Midolab 08/02/1985 Palau mariannensis A. altilis x. A. Z9 790494001 Meinpadahk 05/01/1978 Micronesia mariannensis

60 3.2.2. Reagents

AccQFluor kits, comprising borate buffer (AccQFluor 1), reagent powder

(AccQFluor 2A) and reagent diluent (AccQFluor 2B), and AccQ Tag (6-aminoquinolyl-

N-hydroxysuccinimidyl carbamate, AQC) were purchased from Waters (Milford, USA).

Both reagents were prepared and stored according to the Instruction Manual from

Waters (Waters AccQ Tag Chemistry Package Instruction Manual, Revision 1).

Acetonitrile (HPLC grade) and HCl (1 N) were purchased from Fisher Chemical (Fair

Lawn, New Jersey). KOH and HCl (37 %) were obtained from Sigma–Aldrich (Louis,

USA). Norleucine (Sigma Chemical Company, Louis, USA) stock solution (2.5 mM) was made in 0.1 N HCl. The mixed amino acid standard solution (each at 2.5 ± 0.1

μmol/mL, except cystine at 1.25 ± 0.1 μmol/mL) was Pierce™ Amino Acid Standard

H (comprising 18 amino acids) from Thermo Scientific (Rockford, USA). Water used in this study was purified with DirectQ 3 Water purification System (EMD Millipore

Corporation, Billerica, USA).

3.2.3. Preparation of amino acid samples

200 mg of each frozen sample (49 cultivars × 3 fruits × 3 replications) was weighed and transferred to a 1.5 mL MCT graduated natural tube (Fisherbrand,

Mississauga, ON). The sample was placed in a CentriVap DNA Vacuum Concentrator

(Labconco, Kansas City, USA) for an 18 h drying process, including 999 min centrifugation at 1725 RPM. The dried sample was weighed to determine the dry matter, sliced, and transferred to a screw cap Pyrex vial (17 × 60 mm) (Fisherbrand).

An aliquot of 20 μL of the 2.5 mM norleucine internal standard and 2000 μL 6.0 N HCl

61 were added, the vial was flushed with nitrogen, and then quickly capped. The hydrolysis was performed at 110°C for 19 h in a standard heating block (VWR, Radnor,

USA). 500 μL hydrolyzed sample was filtered through an Ultrafree MC Centrifugation

Unit (Fisherbrand). An aliquot of 1 μL of the hydrolyzed protein sample was diluted to a total volume of 40 μL with borate buffer (AccQFluor Reagent 1).

3.2.4. Derivatization of amino acids

The derivatization was achieved by reacting 20 μL diluted sample or protein hydrolysis standard solution with 20 μL AccQTag derivatizing reagent AQC (6- aminoquinolyl-N-hydroxysuccinimidyl carbamate) in 60 μL borate buffer (AccQFluor

1). All the reagents were pipetted into a 250 μL glass conical insert bottom spring (6

× 29 mm) (Canadian Life Science, Peterborough, Canada), agitated moderately, and kept inside a screw cap injection vial. The vial was heated in the heat block at 55°C for 10 min to complete the reaction. After regaining room temperature, 10 μL derivatized sample was injected into the high performance liquid chromatography

(HPLC) instrument.

3.2.5. HPLC analysis

The analysis of breadfruit amino acid was performed on a Waters 2659 HPLC

Separations Module comprising of a multi λ fluorescent detector (Waters 2475) and a photodiode array detector (Waters 2998), controlled by Empower 3 software (Waters,

Milford, USA) and data system. The amino acids were separated on a 4 μm NavaPak

C column (3.9 × 300 mm) (Waters, Milford, USA). Chromatographic analysis was

62 performed at room temperature (32 ± 5°C) using gradient elution shown in Table 3-2: eluent A was ACCQ Tag/water (10/90, % v/v) and eluent B was acetonitrile/water

(50/50, % v/v). The peaks were detected by fluorescence over a period of 75 min with an excitation wavelength of 250 nm and an emission wavelength of 395 nm.

Table 3-2 HPLC gradient elution for the breadfruit amino acid study

Time (min) Flow %A % B Curve

1 100 0 2 1 93 7 6 4 1 90 10 6 28 1 80 20 3 32 1 75 25 8 38 1 61 33 4 50 1 35 65 6 52 1 0 100 6 55 1 100 0 6 60 1 100 0 2 Eluent A: ACCQ·Tag/water (10/90, v/v) and Eluent B: acetonitrile/water (50/50, v/v)

3.2.6. Data analysis

Genesis 1.7.5 (Institute for Genomics and Bioinformatics; Graz, Austria) was used to build the essential amino acid profile. All statistical analyses were conducted using RStudio Version 0.98.1062 (RStudio Inc. USA) with a type 1 error rate of 0.05.

Essential amino acid data for other staples listed in this article are reported by the

Food and Agriculture Organization of the United Nations (FAO, 1981).

63 3.3. Results

3.3.1. Quantification of essential amino acids

All essential amino acids were found in breadfruit (Artocarpus altilis) and hybrids (A. altilis × A. mariannensis). Significant differences were observed in the essential amino acid content of the protein among breadfruit cultivars and hybrids

(Figure 3-1). Phe, leu, ile, and val were found to be the most abundant essential amino acid in breadfruit. When comparing individual essential amino acids, Ma’afala

(Samoa), Afara, and Hamoa (Maopo) (Society Islands) had the highest levels of all of the essential amino acids, while Ahani and Mahani (Society Islands) and Uto ni viti

(Fiji) had significantly lower protein quality. When the data were expressed as a percentage of total protein, hybrids had a significantly higher content of tyr than the A. altilis cultivars evaluated (Figure 3-2a). When the data were expressed on a dry weight basis, met and tyr were significantly higher in the hybrids than A. altilis cultivars. The hybrids also tended to contain more of other amino acids, but these differences were not statistically significant in our sample size (Figure 3-2b).

Figure 3-1 Breadfruit (Artocarpus altilis) and hybrids (A. altilis × A. mariannensis) essential amino acid content (mg/g protein). N=9 for each cultivar

65 Figure 3-2 a

90 80 70 60 50 40

mg/g Protein mg/g 30 20 10 * * 0 His Thr Tyr Val Met Lys Ile Leu Phe Trp Essential Amino Acid Artocarpus altilis Artocarpus altilis x. A. mariannensis

Figure 3-2 b 4 3.5 3 2.5 2 1.5

mg/g Dry tissueDry mg/g 1 0.5 * * * * 0 His Thr Tyr Val Met Lys Ile Leu Phe Trp Essential Amino Acid Artocarpus altilis Artocarpus altilis x. A. mariannensis

Figure 3-2 Comparisons of essential amino acid content between breadfruit (Artocarpus altilis) and hybrids (A. altilis × A. mariannensis) based on protein weight (a) (N=39×9 for breadfruit, N=8×9 for hybrids) and dry tissue weight (b) (N=41×9 for breadfruit, N=8×9 for hybrids). Comparisons significant at the 0.05 level are indicated by *. Bars represent the standard error of the mean over individual species

66 3.3.2. Comparison of essential amino acids between commercial breadfruit and

other staples

Detailed amino acids profiles for commercial breadfruit cultivars, Ma’afla, Ulu fiti, Yellow, Piipiia, white, and Puua were summarized in Table 3-3. When compared to other staples, Ma’afala had the highest total essential amino acid content as a proportion of the total protein (>568 mg/g protein; Figure 3-3). Ma’afala was especially rich in ile (79.3 mg/g protein) and phe+tyr (149.9 mg/g protein), which were both ~2 to

~4 times higher than other staple crops, and leu (168.6 mg/g protein), which was 1/3–

4 times higher than other staples. Ulu fiti had a total essential amino acid (325.1 mg/g protein) that was most similar to potato and higher than cassava and sweet potato.

Based on fresh weight, the total essential amino acid contents in Ma’afala (1071.5 mg/100 g fresh tissue) and Piipiia (889.2 mg/g fresh tissue) were higher than in non- grain tropical staples, such as potato, sweet potato, cassava, banana, taro, and yam

(Figure 3-4). Ma’afala (1071.5 mg/100 g fresh tissue) contained total essential amino acid that was 1.6 times higher than that of potatoes and ~2.6 times higher than that of sweet potatoes, cassava, or banana. Yellow (538.5 mg/100 g fresh tissue) and Ulu fiti

(489.4 mg/100 g fresh tissue) had higher total essential amino acid contents than cassava, sweet potato, and banana.

67 Table 3-3 The summary of amino acid content in commercially available breadfruit cultivars (Artocarpus altilis and A. altilis x A. mariannensis): Ma’afala, Yellow, White, Piipiia, Puaa, and Ulu fiti

Amino acid Mg/g based on Ma’afala Yellow White Piipiia Puaa Ulu fiti

Dry weight 1.20a,b 0.90b,c 0.33d 0.89b,c 0.65c,d 1.42a Average Ala Protein weight 36a 17b,c 8c 22b 17b,c 25b SE 0.18 0.29 0.11 0.15 0.10 0.10 Dry weight 1.22a 0.70b 0.43b 0.72b 0.42b 1.29a Average Arg Protein weight 37a 13c 11c 18b,c 11c 23b SE 0.18 0.16 0.08 0.10 0.05 0.10 Dry weight 3.05b 2.54b 1.25c 1.99b,c 1.12c 5.00a Average Asn Protein weight 92a 48b 31b 50b 29b 89a SE 0.41 0.64 0.26 0.23 0.20 0.54 Dry weight 2.49b 2.09b,c 1.06d 1.46c,d 0.67d 4.05a Average Asp Protein weight 76a 39b 26b,c 36b,c 18c 72a SE 0.38 0.40 0.22 0.18 0.07 0.46 Dry weight 2.34a 1.35b,c 0.76c 2.02a,b 0.97c 1.89a,b Average Glu Protein weight 71a 25c 19c 50a,b 26c 34b,c SE 0.54 0.30 0.14 0.37 0.13 0.14 Dry weight 1.03a 0.49b 0.35b 0.53b 0.41b 0.99a Average Gly Protein weight 31a 9c 9c 13b,c 11c 18b SE 0.13 0.12 0.06 0.10 0.06 0.11

68 Amino acid Mg/g based on Ma’afala Yellow White Piipiia Puaa Ulu fiti

Dry weight 0.70a 0.42b 0.25b 0.41b 0.25b 0.80a Average His Protein weight 21a 8c 6c 10b,c 7c 14b SE 0.09 0.10 0.05 0.06 0.03 0.06 Dry weight 2.62a 1.58b 0.98b 1.56b 0.88b 2.72a Average Ile Protein weight 79a 30b,c 24c 39b,c 23c 49b SE 0.40 0.38 0.17 0.23 0.10 0.28 Dry weight 5.56a 2.68b 1.72b 2.74b 1.73b 4.78a Average Leu Protein weight 169a 51c 42c 69b,c 46c 85b SE 0.68 0.65 0.29 0.38 0.20 0.43 Dry weight 1.30a 0.92a,b 0.53b,c 0.83b,c 0.42c 1.28a Average Lys Protein weight 39a 17b,c 13b,c 21b,c 11c 23b SE 0.24 0.22 0.10 0.12 0.05 0.13 Dry weight 0.25a 0.11c 0.07c 0.13b,c 0.10c 0.20a,b Average Met Protein weight 8a 2b 2b 3b 3b 4b SE 0.04 0.04 0.02 0.02 0.01 0.02 Dry weight 4.77a 2.60b 1.57b 2.49b 1.56b 4.46a Average Phe Protein weight 145a 49c 38c 62b,c 41c 80b SE 0.60 0.64 0.27 0.34 0.20 0.40 Dry weight 0.41a 0.26b 0.13b 0.26b 0.12b 0.47a Pro Average Protein weight 13a 5b,c 3c 7b,c 3c 8b

69 Amino acid Mg/g based on Ma’afala Yellow White Piipiia Puaa Ulu fiti

SE 0.10 0.07 0.02 0.03 0.02 0.03 Dry weight 1.12a 0.69b 0.29c 0.62b,c 0.43b,c 1.13a Average Ser Protein weight 34a 13b,c 7c 15b,c 11c 20b SE 0.17 0.16 0.07 0.12 0.05 0.08 Dry weight 1.22a 0.70b 0.35b 0.68b 0.39b 1.14a Average Thr Protein weight 37a 13b,c 9c 17b,c 10c 20b SE 0.20 0.16 0.08 0.10 0.04 0.08 Dry weight 0.11a,b 0.07b,c 0.06c 0.08a,b,c 0.04c 0.12a Average Trp Protein weight 3a 1b 1b 2b 1b 2a,b SE 0.03 0.02 0.01 0.02 0.01 0.01 Dry weight 0.17b 0.13b,c 0.04c 0.13b,c 0.04c 0.34a Average Tyr Protein weight 5a,b 2c 1c 3b,c 1c 6a SE 0.05 0.04 0.01 0.02 0.01 0.06 Dry weight 2.04a 1.19b,c 0.59c 1.29b 0.76b,c 2.35a Average Val Protein weight 62a 22c,d 14d 32b,c 20c,d 42b SE 0.30 0.32 0.16 0.19 0.09 0.20 Means with the same letter are not significantly different. N=9. SE represents the standard error of the mean over individual cultivar.

600

500

400

300

200

100 Essential amino acid content (mg/g protein) (mg/g contentacid Essentialamino

Staples

His Ile Leu Lys Met Phe+Tyr Val Thr Trp

Figure 3-3 Comparison of essential amino acid content between commercially available breadfruit cultivars (Artocarpus altilis and A. altilis x A. mariannensis): Ma’afala, Yellow, White, Piipiia, Puaa, and Ulu fiti and other staples based on protein weight. Essential amino acids data for other staples listed is from the Food and Agriculture Organization of the United Nations (FAO, 1981)

71 1200

1000

800

600

400

200

0 Essential amino acid content (mg/100g fresh tissue)fresh(mg/100g contentacid Essential amino

Tropical Staples His Ile Leu Lys Met Phe+Tyr Thr Trp Val

Figure 3-4 Comparison of essential amino acid content between commercially available breadfruit cultivars (Artocarpus altilis and A. altilis x A. mariannensis): Ma’afala, Yellow, White, Piipiia, Puaa, and Ulu fiti and other tropical staples based on fresh tissue weight. Essential amino acids data for other staples listed is from the Food and Agriculture Organization of the United Nations (FAO, 1981)

3.4. Discussion

Children and infants usually have a higher demand for essential amino acids

(WHO, 2007). Children who do not have access to adequate protein sources or essential amino acids in their diets are underweight leading to serious health consequences (WHO, 2009). In 2004, about 20% (112 million) of preschool children

72 were underweight, resulting in 2.2 million child deaths and many other long-term developmental problems. Developing countries experienced most of the underweight cases with almost half of the deaths from this cause occurring in Africa and over

800,000 deaths in Southeast Asia (WHO, 2009).

In the daily diet, staples are consumed in such quantities that they constitute a dominant proportion of nutrient intake, including protein intake. Corn, wheat, and rice are the most common staples that comprise two-thirds of the world’s energy intake

(FAO et al., 2012). However, most cereals have poor protein quality and are low in tryptophan (especially maize) and/or lysine (especially wheat) (WHO, 2007). People usually consume meat, fruit, vegetables, and other dairy products as a supplementary intake for essential amino acids, but these options are not always affordable especially for undernourished populations (WHO, 2007; 2009). In addition, 349 million undernourished people (nearly 50% of the world’s starving population) are found near the equator (FAO et al., 2012). Production of typical staple crops, such as wheat and corn, may be limited by the warm, wet climate and absence of winter freeze that can allow the proliferation of pathogenic microbes. Therefore, a suitable affordable and nutritious alternative staple is critical for improving food security in the tropical regions.

Grain crops, such as maize, rice and wheat, are the most common staples in the world (Lancaster and Coursey, 1984). However, in tropical countries, non-grain staples such as cassava, potato, sweet potato, banana, taro, yam, and breadfruit provide the dietary base for 500-700 million people (Lancaster and Coursey, 1984).

The overall goal of our research was to determine whether a diet rich in breadfruit could provide adequate protein quality for food security in these regions.

73 Previous research has shown that 60% breadfruit with 40% soy flour in a weaning diet for children from 1 to 3 years old can supply the recommended overall dietary nutritional allowances within the tolerably low anti-nutrient levels (Ijarotimi and

Aroge, 2005 ), but protein quality was not assessed. In our study, a full spectrum of essential amino acids was found in breadfruit (both A. altilis and hybrids). This finding is consistent with the previous study (Golden and Williams 2001). Significant difference was observed between cultivars. Breadfruit is especially rich in phe, leu, ile, and val. Quantification of met and trp may have been underestimated due to the destruction of the structure during the hydrolysis process (Pickering and Newton,

1990). Cysteine was not quantified in this study because of limitations of the analytical technique.

The cultivar Ma’afala had the highest essential amino acid content based on both dry weight and protein weight (Table 3-3 ) and was the highest protein variety found in the earlier study (Jones et al., 2011a). Ma’afala contains a higher percentage of the essential amino acids than normally found in soybean, indicating that breadfruit cultivars can have better protein quality than soy. A typical breadfruit based main dish usually uses one mature fruit of an average weight of 1.2 kg (HHFN and NTBG, 2012).

I estimate that a 500 g Ma’afala or Ulu Fiti could provide a significant contribution toward the essential amino acid requirement of a preschool child weighing 20 kg

(Table 3-4). Similarly, these cultivars contribute significantly toward meeting the nutritional requirements of a 60 kg adult (Table 3-4).

74 Table 3-4 Nutritional value of commercially available breadfruit cultivars (Artocarpus altilis and A. altilis × A. mariannesis): Ma’afala, Yellow, White, Piipiia, Puaa, and Ulu fiti

mg/500 g fresh breadfruit Requirement per day

mg/20kg mg/kg per per mg/kg per mg/60kg Ma’afala Yellow White Piipiia Puaa Ulu fiti preschool preschool adult adult kid kid

His 200 108 64 195 67 100 - - 10 600

Ile 749 404 226 668 259 373 27 540 20 1200

Leu 1593 714 450 1167 459 631 54 1080 39 2340

Lys 372 215 109 313 141 217 45 900 30 1800

Met 71 35 25 46 18 24 22 440 10 600

Phe+Tyr 1410 681 409 1173 428 641 40 800 25 1500

Thr 349 178 101 279 100 166 23 460 15 900

Trp 32 20 10 30 15 16 6 128 4 240

Val 581 338 196 575 157 279 36 720 26 1560 Daily requirement are from the World Health Organization (WHO 2007)

75 3.5. Summary

The current study analyzed 49 breadfruit cultivars for their amino acid profile using HPLC based amino acid analysis. The results showed that breadfruit contains all the essential amino acids and is especially rich in phenylalanine, leucine, isoleucine, and valine. Individual essential amino acid profiles were generated for each cultivar, and cultivar differences were compared. Ma’afala contained significantly higher total essential amino acid content (568 mg/100g protein) than other varieties and higher-quality protein than staples such as wheat (336 mg/100g protein), corn

(413 mg/100g protein), rice (374 mg/100g protein), potato (326 mg/100g protein), soybean (443 mg/100g protein), and yellow pea (366 mg/100g protein). Consumption of 500 g Ma’afala can contribute significantly toward meeting the daily essential amino acid requirement.

76 Chapter 4. Breadfruit Protein Digestibility Analysis Using an in Vitro Human Digestion Model

4.1. Synopsis

Protein digestibility is an important part of protein quality evaluation

(Blackburn and Southgate, 1981). The essential amino acid profile can provide a good prediction of protein quality, but in practice, the digestibility determines whether a protein is high quality or not (Blackburn and Southgate, 1981). Human digestion is a complex process and in the past two decades there have been many attempts to model this process. In vivo feeding methods, using animals or humans, usually provide the most accurate results, however, they are expensive and time consuming—sometimes even impossible to conduct. In principle, in vitro digestion models provide a useful alternative (Hur et al., 2011). One stage in vitro digestion models usually gave lower protein digestibility than in vivo studies, but two or more stage in vitro digestion, generally, is found to be highly correlated (r>0.8) with in vivo rat true fecal nitrogen digestibility, especially when analyzing plant protein sources

(Butts et al., 2012). To our best knowledge, there is no study related to the determination of breadfruit protein digestibility using an in vitro digestion model.

4.1.1. Objectives

The objective of this study was to determine the protein digestibility of

Ma’afala flour in comparison to commercial wheat flour using a multi stage enzyme digestion model.

77 4.2. Materials and Method

4.2.1. Reagents

Multi-stage enzyme digestion model

Reagents used for digestion model were: urea (BioReagent Grade for molecular biology, suitable for cell culture; Sigma-Aldrich, St. Louis, MO), α-amylase from human saliva (Type XIII-A, lyophilized powder, 300-1500 units/mg protein, 1 KU;

Sigma-Aldrich), pancreatin from porcine pancreas (4 x USP; Sigma-Aldrich), bile bovine (dried, unfractionated; Sigma-Aldrich), BSA (bovine serum albumin) (heat shock fraction, protease free, fatty acid free, essentially globulin free, pH=7, ≥98%;

Sigma-Aldrich), pepsin from porcine gastric mucosa (lyophilized powder, 3200-4500 units/mg protein; Sigma-Aldrich), KSCN (ReagentPlus, ≥99.0%; Sigma-Aldrich),

KH2PO4 (reagent, Caledon Laboratories Ltd., Georgetown, Ont), Na2SO4 (≥99.0%, plant cell culture tested; Sigma-Aldrich), NH4Cl (cell culture tested; Sigma-Aldrich),

NaOH (reagent grade, ≥98%; Sigma-Aldrich), HCl (37%; Sigma-Aldrich), NaCl (small, white crystal; Fisher Scientific, Ontario, Canada), KCl (certified ACS, crystalline;

Fisher Scientific), NaHCO3 (Fisher Scientific), CaCl2 • 2H2O (certified ACS, reagent grade, Fisher Scientific), MgCl2 (bioreagents; Fisher Scientific), and NaH2PO4 • H2O

(reagent grade, AnaChemia Chemicals, New York, USA).

Bioassays

HEPES (fine white crystals, bioreagents; Fisher Scientific), Tween 20

(PhytoTechnology Laboratories®, Shawnee Mission, KS), BCA (bicinchoninc acid) protein assay kit (Thermo Fisher Scientific), Modified Lowry Protein Assay Kit (Thermo

78 Scientific), bovine serum albumin protein standard solution (Thermo Scientific).

4.2.2. Sample preparation

Breadfruit (cultivar: Ma’afala, Artocarpus altilis) fruits (19 fruits) were harvested from a single 4-year-old-tree grown from commercial micro-propagation at the

McBryde Garden, Kauai, HI (21°54’29.47°N; 159°30’41.45°W). Fruit were harvested when mature, peeled, cored and sliced with a mandolin to 6.35 mm slices. Breadfruit slices were dried overnight at 115°C in an Excalibur Dehydrator to complete dryness and shipped to University of British Columbia Okanagan campus. Flour was prepared by grinding dried breadfruit with a burr-style coffee grinder (Bunn, Illinois, USA). Wheat flour (all purpose, Robin Hood™) was purchased from a local supermarket.

4.2.3. In vitro digestion model to mimic human digestion

The in vitro digestion model for digesting breadfruit and wheat was modified based on the previous methods published by Kiers et al., 2000; Tenore et al., 2013;

Raiola et al., 2012. The digestion process was divided by mouth digestion, stomach digestion, and intestinal digestion (Table 4-1; Figure 4-1). Mouth digestion was mimicked by mixing 6 g of cooked (water boiled) flour with 6 mL artificial saliva in a 50 mL centrifuge tube (Corning Incorporated, New York, USA). The pH of the solution was adjusted to 6-7 with HCl (1N), using a pH meter (VWR International,

Pennsylvania, USA). The solution was incubated at 37°C for 5 min in an incubator

(Innova®44, Eppendorf, Connecticut, USA). The solution was then mixed with 12 mL pepsin solution. The pH of the mixture was adjusted to 2-4 with HCl (6N). The mixture was incubated at 37°C and shaken consistently at 300 RPM for 2 hours. The intestinal

79 digestion was performed by adjusting the solution pH to 7.5 with NaOH (1 N) and then mixing the solution with 12 mL of pancreatic solution and 6 mL of bile solution. The solution was incubated at 37°C and shaken consistently at 300 RPM for 2 hours. The compositions of saliva, pepsin solution, pancreatic solution and bile solution were listed in supplementary material. The digestion samples were adjusted to pH 7.5. The digestion extracts were stored at -80°C before analysis. A set of digestion was conducted following the same digestion process (Figure 4-1) in absent of flour samples to obtain a digestive enzyme solution that contains all the enzymes and buffers used in the process.

Measured total protein content before digestion

6 g of cooked breadfruit/wheat flour

Added 6 mL saliva solution Mouth digestion pH=6-7 5 min shaking at 300 RPM at 37 °C

Added 12 mL pepsin solution Stomach digestion pH=2-4 2 hrs shaking at 300 RPM at 37 °C

Added 12 mL of pancreatic solution and 6 mL bile solution Intestinal digestion pH=7.5 2 hrs shaking at 300 RPM at 37 °C

Measured total protein content after digestion

Breadfruit/wheat digestion extracts stored at -80 °C

Figure 4-1 A Schematic representation of the in vitro digestion model

80 Table 4-1 The composition of saliva, gastric solution, duodenal solution, and bile solution used in the multi-stage enzyme digestion model

Saliva Gastric solution Duodenal solution Bile solution

Item g/l Item g/l Item g/l Item g/l

KCl 1.792 NaCl 5.500 NaCl 0.168 NaCl 10.518

KSCN 0.400 NaH2PO4 0.533 NaHCO3 0.081 NaHCO3 11.570

NaH2PO4 1.776 KCl 1.649 KH2PO4 0.002 KCl 0.753

Na2SO4 1.140 CaCl2·2H2O 0.799 KCl 0.014 Urea 0.500

NaCl 0.600 NH4Cl 0.612 MgCl2 0.001 CaCl2·2H2O 0.444

NaHCO3 3.388 Urea 0.170 Urea 0.200 BSA 3.600

Urea 0.400 BSA 2.000 CaCl2·2H2O 0.400 Bile 60.000 α- 0.002 Pepsin 5.000 BSA 2.000 amylase Pancreatin 18.000 BSA=bovine serum albumin

4.2.4. Protein determination in the flours and digestion extracts

Protein was extracted from the flour following the method published by Jones et al. (2011a). First, 10 mg flour was weighed into a 1.5 mL microcentrifuge tube

(Fisher Scientific, Ottawa, ON) with 0.5 mL protein extraction buffer (20 mM HEPES,

150 mM NaCl, 0.3% Tween 20). The mixture was vortexed and sonicated for 15 min at room temperature (FS20, Fisher Scientific). The tubes were centrifuged for 10 min at 13,000 RPM (Galaxy 16DH; VWR, Westchester, PA). The supernatant was saved for the protein determination of flours.

The protein content of flours (wheat and breadfruit, 3 replicates each) and

81 digestion extracts (digestive enzyme solution, wheat digestion, and breadfruit digestion, 3 replicates each) were first determined by BCA (bicinchoninic acid) protein assay (Pierce) following standard kit instructions. 25 µL of the protein extracts from the flour and each bovine serum albumin (BSA) protein standard solution (0, 0.125,

0.25, 0.5, 1.0, 1.5, 2.0 mg/mL) were plated into a 96 microplate well in duplicates (BD

Biosciences, Mississauga, Canada) with 200 µL BCA working reagent (1:50, v/v,

Reagent A/Reagent B). The plate was gently shaken for 30 seconds and incubated at

37°C for 30 min. Absorbance was measured at 562 nm using a Synergy HT microplate reader (Biotek, Winooski, VT). The protein content of digestion extracts was calculated by comparing to a standard curve of varying concentrations of BSA standard solution.

The protein content of digestion extracts was measured using the same protocol except a 1:5 dilution of the extract in 0.9% NaCl was performed before measurement.

This experiment was repeated twice independently.

An orthogonal method for protein quantification was used. A modified Lowry

Protein Assay Kit (Fisher) using standard protocols quantified protein in 40 µL aliquots of each of the samples by comparison to the same BSA protein standard solutions.

Samples and standards were mixed with 200 µL modified Lowry reagents and plated in a 96-well microplate in duplicates at room temperature for 10 min followed by another 30 min incubation with 20 µL 1X Folin-Ciocalteu Reagent at room temperature. The absorbance was measured at 750 nm using the same microplate reader. The protein content was calculated by direct comparison of samples to the standard curve of varying concentrations of BSA standard solution. This experiment was repeated twice independently.

82 4.3. Results

Before digestion, the protein content of wheat flour was 5.3% (modified Lowry assay), or 3.8% (BCA assay) (Figure 4-2a). Breadfruit flour had a protein content of

6.2 % (modified Lowry assay) or 4.5% (BCA assay). About 87% (modified Lowry assay) or 89% (BCA assay) of the breadfruit protein was fully digested in the in vitro digestion model while 79% (modified Lowry assay) or 71% (BCA assay) of the wheat protein was fully digested (Figs. 4-2a & b). After the digestion reactions, there was significantly more intact protein in the wheat digestion extract than the breadfruit confirming that more of the breadfruit protein could be digested (Figure 4-2b).

Fig 4-2 a

) *

w 60

o ***

( 40

i 20

0 n n n n n n n n tio tio io io tio tio io io s s st st s s st st e e e e e e e e ig ig ig ig ig ig ig ig d d d d d d d d r r r r re e re e re e re e o ft o ft o ft o ft f a f a f a f a e t e t e t e t b a b i b a b i t e it ru t e it ru a h u f a h u f e fr d e fr d h W d a h W d a W a re W a re re B re B B B

83 Fig 4-2 b

)

o 10

(

t **

t *** ***

o 5

0 n n n n n n io tio io io tio io t s st t s st lu e e lu e e o g o g s i ig s i ig d d d d e t it e t it m a u m a u zy e fr zy e fr n h d n h d e W a e W a e re e re iv B iv B st st e e ig ig D D BCA assay Modified Lowry assay

Figure 4-2 Protein content in wheat and breadfruit flour before and after digestion (a) and digestion extracts (b) by BCA and Modified Lowry Assays (n=3). * represents significant difference at α=0.05. ** represents significant difference at α=0.01. *** represents significant difference at α=0.001, using two sample t-test or one-way ANOVA with Tukey- Kramer Honest Significant Difference (HSD) test. Bar represents the standard error of 6 replicates within each treatment

4.4. Discussion

In order to meet protein requirements in the diet and provide adequate consumption of essential amino acids, staple foods must be digested and absorbed through the gut (Latham 1997). The evaluation of protein quality is primarily focused

84 on two aspects: the overall protein content/amino acid profile and the protein digestibility/amino acid availability. Our previous research has established the protein and amino acid content of breadfruit flours.

The differences in overall protein content of wheat and breadfruit flours in the literature and in our assays were not substantial. Wheat flour has a protein content of about 10% based on the USDA index (USDA 2014) but analysis of commercial wheat flours varies depending on analysis methods (Knight and Chambers 2003; Moore et al., 2010). The breadfruit cultivar, Ma’afala, used in this study has a reported protein content of 7.6% (Jones et al., 2011a) but varies slightly between batches and years.

In this study, protein content for both flours was less than the value reported in the literature, and Ma’afala flour had higher protein content than wheat flour. This project showed that Ma’afala had higher total essential amino acid content (568 mg/g protein) than other varieties or wheat flour (336 mg/g protein).

Our current objective was to establish breadfruit flour digestibility using in vitro assays in comparison to wheat. Breadfruit showed a higher protein digestibility than wheat protein in the in vitro model. Wheat digestibility can be problematic as the storage proteins gluten and lectin may be undigested in the gut but these proteins are not found in breadfruit flours (de Mejia and Prisecaru, 2007; de Punder and

Pruimboom, 2013; Simonato et al., 2001; Ijarotimi and Aroge, 2005).

4.5. Summary

Breadfruit showed a higher protein digestibility than commercial wheat flour using an in vitro multi-stage enzyme digestion model.

85 Chapter 5. Effects of Breadfruit Digestion on Human Colon Epithelial Cells Using an In Vitro Cell Model

5.1. Synopsis

Human intestinal cells, especially the epithelium, shoulder the responsibility to absorb nutrients and serve as a barrier to bacteria, pathogens, and other antigens.

Damage to the epithelium, which is usually caused by imbalanced apoptotic cell death, is thought to be responsible for diseases of the gastrointestinal tract (Lee, 1993), therefore cytotoxicity/cell viability is a critical indicator of epithelial health. Cytokines secreted by human epithelial cells play a critical role in mucosal and systemic immunity. Although food is known to be a factor that alters intestinal immune response

(Lied et al., 2011), the extent to which digested food regulates cytokine formation through epithelial cells is poorly understood. Caco-2 cells are a human intestinal epithelial cell line that are widely used to investigate the interaction between food substances and the intestinal epithelium (Shimizu 2010). Cell models can provide larger screening capacity at lower cost than animal models (Langerholc et al., 2011).

To our best knowledge, no study has ever been done to evaluate the impact of breadfruit digestion on intestinal cells.

5.1.1. Objectives

The first objective of this study was to determine the cytotoxicity of breadfruit digestion to the human colon epithelial cells in comparison to wheat digestion. By examining the viability of the cells after digestion treatment, impact of the breadfruit digestion to colon epithelial morphology was compared to the impact caused by the wheat digestion. The second objective of this study was to determine the immune

86 response of the human intestinal cells to breadfruit and wheat digestions quantified as changes in specific cytokines including: (a) the anti-inflammatory cytokines: interleukin

4 (IL-4), interleukin 10 (IL-10); (b) the pro-inflammatory cytokines: tumor necrosis factor alpha (TNF-α), interferon gamma (IFN-γ), inducible nitric oxide synthase

(iNOS); (c) pleiotropic cytokine: interleukin 6 (IL-6); and (d) chemokines: monocyte chemoattractant protein-1 (MCP-1) and interleukin 8 (IL-8). Overall, our results will quantify the effects of breadfruit digestion on human colon epithelial cells.

5.2. Materials and Method

5.2.1. Reagents

Caco-2 cell Culture

Materials used for Caco-2 cell culture were: HEPES buffered Dulbecco’s modified Eagle’s medium (DMEM) (high glucose, 4500 mg/L glucose, no L-glutamine, no sodium pyruvate; Thermo Fisher Scientific, Manassas, USA), sodium pyruvate

(100 mM solution; Corning Cellgro, Mediatech Inc., Manassas, VA), penicillin- streptomycin (100X, 10000 µg/mL; Corning cellgro, Mediatech Inc) , L-alanyl-L- glutamine were purchased (100X, liquid, Mediatech Inc.), FBS (fetal bovine serum)

(Standard; Gibco® Life technologies, Thermo Fisher Scientific), phosphate buffered saline (PBS) (without calcium or magnesium; Lonza, Walkersville, MD) nonessential amino acid solution (100X, 0.1 µm sterile filtered; Thermo Fisher Scientific), trypsin

(1X, 0.25% trypsin in HBSS without calcium or magnesium, Porcine Parvovirus

Tested; Corning Cellgro, Mediatech Inc.)

87 Bioassays

Reagents employed in the assays were: anhydrous ethyl alcohol (molecular grade; Commercial Alcohols, Ontario, Canada), chloroform and isopropyl alcohol

(HPLC graded; OmniPur®, EMD Millipore Corporation, Billerica). RiboZol

(biotechnology grade; Amresco, Ohio, USA), trypan blue solution (0.4% w/v in PBS;

Corning Cellgro, Mediatech Inc.), iScript cDNA synthesis kits (Bio-Rad Laboratories,

Ontario), and SsofastTM EvaGreen Supermix (Bio-Rad Laboratories), and nuclease- free water (Integrated DNA technologies, Coralville, USA).

Treatment applications

Other materials used in this study were: lipopolysaccharide (LPS) from

Escherichia coli 0111:B4 (γ-irradiated, BioXtra, suitable for cell culture; Sigma-

Aldrich), interleukin 1 beta (IL-1β) human (animal-component free, recombinant, expressed in E. coli, ≥98% (SDS-PAGE), ≥98% (HPLC); Sigma-Aldrich). Human primers including: tumor necrosis factor alpha (TNF-α), (interferon gamma) IFN-γ, monocyte chemoattractant protein-1 (MCP-1), inducible nitric oxide synthase (iNOS), interleukin 4 (IL-4), interleukin 6 (IL-6), interleukin 8 (IL-8), interleukin 10 (IL-10) and

18S rRNA (18S ribosomal RNA) were purchased from Integrated DNA Technologies

(Coralville, USA).

5.2.2. Caco-2 human epithelial colorectal adenocarcinoma cells

Cells were obtained from the American Type Culture (ATCC® HTB-37TM) and cultured using established protocols. Cells were cultured in a T-75 flask (Greiner Bio- one, Canada) with standard growth media containing HEPES buffered Dulbecco’s

88 modified Eagle’s medium (DMEM) with 4.5 g/L glucose, 1% nonessential amino acid,

20% fetal bovine serum (FBS), 2% penicillin-streptomycin, 2% L-alanyl-L-glutamine, and 1% sodium pyruvate. Cells were grown in a controlled environment with a temperature of 37°C and a humidified atmosphere of CO2 /air (5/95, %) in an Air-

Jacketed CO2 Symphony incubator (VWR International, Pennsylvania, USA). Cells were passaged every 7 days by trypsinisation. Trypsinisation was performed by incubating PBS washed cells with 3 mL trypsin for 4 minutes at 37°C. The trypsin solution containing cells was deactivated by 9 mL growth media. The solution was centrifuged at 234 g for 2 min. The cells were re-suspended in 6 mL growth media and counted by standard trypan blue staining assay (1:1, v/v, trypan blue/re- suspended cell solution) using a hemocytometer. For the generation, 800,000 cells were sub-cultured into a new T-75 flask with 20 mL growth media. Cells used for stimulation were within generation 15 and 25. Cells were plated into 24 transwells

(Greiner Bio-one, Canada) (40,000 cells per well) with the growth medium and they grew until reached 80% confluence. The entire experiment was performed in triplicate.

5.2.3. Digestion extracts preparation

The preparation of digestive enzyme solution, wheat digestion and breadfruit digestion was described in full details in Chapter 4.2.1-4.2.3. The samples were centrifuged and filtered using an Ultrafree® MC centrifugal filter (Millipore Corporation,

USA) at 13,000 RPM for 10 min before applying to Caco-2 cells.

89 5.2.4. Effects of digestion extracts on Caco-2 cell viability

Caco-2 cells were incubated for 4 hours with four different concentrations (1%,

5%, 10%, and 50%) of three different digested extracts (digestive enzyme solution, wheat digestion, and breadfruit digestion) in the standard growth media. After 4 hours,

200 µL trypsin was added into each well and the cells were incubated with trypsin for

2 min at 37°C in a SymphonyTM incubator (VWR International; Pennsylvania, USA). A gentle shake was applied to the well to ensure complete lifting of the cells. 200 µL standard growth medium was added into each well to deactivate trypsin. An aliquot of

400 µL of cells was aseptically transferred to a 1.5 mL clear Eppendorf tube and mixed with an equal volume of 0.4% (w/v) trypan blue solution in PBS. Cell counts were recorded using a dual-chamber hemocytometer (QiuJing®, Guangzhou, China) and a light microscope (Olympus, Ontario, Canada). Viable and nonviable cells were counted separately. The mean of four independent cell counts were pooled for analysis.

5.2.5. Effects of LPS and/or IL-1β stimulation on Caco-2 cells

Caco-2 cells were stimulated with LPS, IL-1β, and LPS plus IL-1β for 4, 8 and

24 hours. LPS used for the stimulation was made in the following way: 1 mg LPS was re-suspended with 1 mL nuclease-free water. 0.4 µL suspended LPS was mixed with

9.6 µL growth medium and used for the stimulation. 10 µg IL-1β was prepared by re- suspended 100 µL nuclease-free water. 0.4 µL suspended IL-1β was mixed with 9.6

µL PBS and used for the stimulation.

RNA extraction was performed on the Caco-2 cells using RiboZol, following the

90 manufacturer’s instruction. 400 µL RiboZol was added into each well and incubated with the cells at room temperature for 5 min. The RiboZol solution containing cells was transferred into a 1.5 mL tube (Fisher Scientific, Ontario, Canada) with 50 µL chloroform. The mixture was shaken vigorously by hand and incubated at room temperature for 2-3 min. After incubation, the samples were centrifuged at 12,000 g for 15 min at 4°C. The aqueous phase was transferred into a new tube 1.5 mL with

200 µL 100% isopropanol. The samples were incubated at -80°C for 20 min and then centrifuged at 12,000 g for 10 min at 4°C. The supernatant was removed and the RNA pellet was washed by 400 µL 75% ethanol. The samples were centrifuged at 7,500 g for 5 min at 4°C. The supernatant was removed and the RNA pellet was air dried at room temperature for 10 min. The RNA pellets were re-suspended in 50 µL nuclease free water and incubated in a hot water bath (55°C) for 15 min. The concentration of the RNA samples was measured by a NanoDrop 2000c UV-Vis Spectrophotometer

(Thermo Fisher Scientific, Manassas, USA).

Based on RNA concentration, 1 to 2 µg of RNA was reversed transcribed using an iScript cDNA synthesis kit by following the manufacturer’s instructions. The mixture contained RNA samples, nuclease free water, 4 µL 5x iScript reaction mix and 1 µL iScript reverse-transcriptase. The mixture was run through a standard cDNA synthesis cycle, including 25°C for 5 min, 42°C for 30 min, 85°C for 5 min and 4°C in a C1000

TouchTM Thermal Cycler (Bio-Rad Laboratories, Ontario, Canada).

cDNA samples were diluted in 1:10 in nuclease free water. An SsofastTM

EvaGreen Supermix was used to prepare qPCR mixture, following the manufacturer’s instruction. The qPCR mixture contained 5 µL Ssofast 2x Mix, 0.2 µL 5 µM forward

91 primer, 0.2 µL 5 µM reverse primer, 1 µL diluted cDNA and 3.6 µL nuclease free water.

The qPCR mixture was plated in duplicates into a 96 MultiplateTM PCR plates (Bio-

Rad Laboratories, Ontario, Canada) and detected by CFX96 Touch real time detection system (Bio-Rad Laboratories, Ontario, Canada). The qPCR run cycle contained enzyme activation at 95°C for 30 s (1 cycle), denaturation at 95°C for 5 s (55 cycle), annealing at 58°C for 5 s (55 cycle) and melt curve at 95°C for 10 s. The reference gene was 18S rRNA. Gene expression of TNF-α, IFN-γ, MCP-1, IL-10, IL-6, iNOS, IL-

8 and IL-4 were examined. Calculations of the expression value were performed in the Bio-Rad CFX manager 3.1 (Bio-Rad Laboratories, Ontario, Canada) using the

ΔΔCt method.

5.2.6. Effects of digestion extracts combined with LPS and/or IL-1β stimulations

on Caco-2 cells

Caco-2 cells were incubated with digestion extracts (digestive enzyme solution, wheat digestion, and breadfruit digestion) combined with different stimulations (non- stimulated, LPS, IL-1β, LPS, and IL-1β) for 24 hours (Table 5-1). RNA extraction, cDNA synthesis, and qPCR analysis process were performed as above.

92 Table 5-1 Experimental design for testing effects of digestion extracts combined with LPS and/or IL1β stimulation on Caco-2 cells (n=3) after 24 hours

Conc. of Conc. of Block Stimulation Treatment stimulations treatments

0% No treatment

Non Digestive enzyme solution 1 - stimulated 1% Wheat digestion Breadfruit digestion 0% LPS only Digestive enzyme solution 2 1000 ng/mL LPS 1% Wheat digestion Breadfruit digestion 0% IL 1β only Digestive enzyme solution 3 100 ng/mL IL 1β 1% Wheat digestion Breadfruit digestion 0% LPS+IL 1β only LPS: 1000 Digestive enzyme solution 4 ng/mL, IL 1β: LPS+ IL 1β 100 ng/mL 1% Wheat digestion Breadfruit digestion

5.2.7. Primers design and efficiency testing

All the primers (Table 5-2) used in this experiment have been designed using

Primer-BLAST software (National Center for Biotechnology Information, USA). For each cytokine, the primer set includes a reverse primer and a forward primer. The primers were designed to have a PCR product size range from 70 to 150 bps. The primer melting temperatures were selected between 57°C and 63°C. To ensure primer

93 specificity, all primers must have no more than at least 2 total mismatches to unintended targets, including at least 2 mismatches within the last 5 bps at the 3’ end.

Primers with a minimum occurrence of hairpin, self-dimer, hetero dimer were selected using OligoAnalyzer 3.1 software (Integrated DNA technologies, Coralville, USA).

Gradient qPCR runs at an annealing temperature range of 50-63°C to test for the best annealing temperature. The primer efficiency was tested by using LinregPCR

(Analysis of quantitative RT-PCR data) (Dr. J.M. Ruijter, Academic Medical Centre, and Australia).

Table 5-2 RNA sequence, best annealing temperature and primer efficiency for the primers used in the breadfruit in vitro cell model study

Best Primer Forward sequence Reverse sequence annealing Efficiency temp. GTAACCCGTTGAAC CCATCCAATCGGTA 18S 58 °C 100% CCCATT GTAGCG TCTCGAACCCCGA TATCTCTCAGCTCC TNF-α 58°C 100% GTGACAA ACGCCA TTCCATTCCAAGCC CCAAGCCCAGAGA IL-10 58°C 91% TGACC CAAGATAAA GAGAGTAGTGAGG GGTCAGGGGTGGT IL-6 58°C 92% AACAAGCC TATTGC TTCCCCCTCTGTTC GTCTGTTACGGTCA IL-4 58°C 95% TTCCT ACTCGG AGCCCTTTACTTGA TCCATCTTTCACCC iNOS 58°C 100% CCTCCT ACTTGC GGCTGAGACTAAC GAATGAAGGTGGC MCP-1 58°C 95% CCAGAAAC TGCTATGA GGGTTCTCTTGGCT GAGTTCCATTATCC IFN-γ 58°C 92% GTTACT GCTACATCT GAGACAGCAGAGC ACACACAGTGAGAT IL-8 58°C 100% ACACAAG GGTTCC

94 5.2.8. Statistical analysis

All statistical analysis was conducted using JMP® 10.0.0 (SAS Institute, Cary,

NC). A series of independent two sample t-test or ANOVA followed by Tukey-Kramer

Honest Significant Difference (HSD) tests were conducted to determine if there were significant differences between treatments with a type 1 error rate of 0.05. Graphs were created using GraphPad Prism 5.0 software (GraphPad Software Inc., La Jolla,

CA).

5.3. Results

5.3.1. Effects of digestion extracts on Caco-2 cell viability

There was no significant difference in the viability of untreated cells and cells treated for 4 hours with digestion extracts (1% of either digestive enzyme solution, wheat digestion or breadfruit digestion) (Figure 5-1a). At higher concentrations (5%) of the digestive enzyme solution without digestions, a significant reduction in the percentage of live cells was observed (Figure 5-1b). Interestingly, the digestions had a mediating effect with 21% increased cell viability with wheat in the digestive enzyme solution and 30% increased cell viability with breadfruit digestions (Figure 5-1a). When the digestion extracts were included at 5% in the cell cultures, breadfruit digestion treated cells had significantly increased viability relative to both the control digestion and wheat (Figure 5-1b). When the digestion extract was included at 10% in the cell culture medium, cell viability was significantly reduced in all treatments (Figure 5-1c).

At 50% of digestion extracts in the cell cultures, the breadfruit digestion extract had a significantly higher cell viability (42% higher) than the digestive enzyme solution

95 without flours (Figure 5-1d). Overall, Caco-2 cells exposed to breadfruit digestion extracts had a significantly higher cell viability than cells exposed to the digestive enzyme solution alone. Increasing the amount of any of the dilution extracts significantly reduced the absolute number of viable cells and the 1% digestion extracts was most suitable for comparative studies (Figure 5-2).

Figure 5-1 a: Cell viability under 1% treatment

100

L 80

e 60

% 40

E 20

0 n n n n io io io io st t st st e lu e e ig o ig ig d s d d e t it % m a u 0 zy e fr n h d e w a re e % b tiv 1 s % e 1 ig d % 1

96 Figure 5-1 b: Cell viability under 5% treatment

* 100

L 80

e 60

% 40

s 20

0 n o n n n ti io tio io s t s st e lu e e ig o ig ig d s d e t d % a it 0 m e ru zy h f n d e w a re e % b tiv 5 s % e 5 ig d % 5

Figure 5-1 c: Cell viability under 10% treatment

** *** ** 100

L 80

e 60

s 20

0 n o n n n ti io tio io s t s st e lu e e ig o ig g d s d i e t d % a it 0 m e u zy h fr n d e w a re e % b tiv 0 s 1 % e 0 ig 1 d % 0 1

97 Figure 5-1 d: Cell viability under 50% treatment

*** *** *** 100

c 60

% 40

s 20 E *

0 n n o n n ti io tio io s t s st e lu e e ig so ig ig d d d e t t % a i 0 ym e ru z h f n d e w a re e % b tiv 0 s 5 % e 0 ig 5 d % 0 5

Figure 5-1 Effects of digestion extracts of wheat and breadfruit on cell viability. a) 1% digestion extracts. b) 5% digestion extracts. c) 10% digestion extract. d) 50 % digestion extracts. Error bars represent the standard error of 3 replicates within each treatment. * represents significant difference at α=0.05, ** represents significant difference at α=0.01, *** represents significant difference at α=0.001, using one-way ANOVA with Tukey-Kramer Honest Significant Difference (HSD) test

98 L 1,000,000

e **

s **

l l 800,000

c a 600,000

m 400,000

d 200,000

E 0 % % % % 1 5 0 0 1 5

Treatment

Digestive Enzyme solution Wheat Digestion Breadfruit Digestion

Figure 5-2 Estimated total number of Caco-2 cells per ml after treatment with digestion extracts for 4 hrs. * represents significant difference at α=0.05,** represents significant difference at α=0.01,*** represents significant difference at α=0.001, using one-way ANOVA with Tukey-Kramer Honest Significant Difference (HSD) test. Bar represents the standard error of 3 replicates within each treatment

5.3.2. Effects of LPS and/or IL-1β stimulations on Caco-2 cells

To determine the best methods of stimulating Caco-2 cells for immune responses, LPS and IL-1β were compared to unstimulated cells (Figure 5-3). Caco-2 cells stimulated with LPS had significantly increased MCP-1 and iNOS expression after 8 hours as compared to unstimulated cells (Figure 5-3 a & b). There were no significant differences in the remaining cytokines (IL-4, IL-6, IL-8, IL-10, TNF-α, and

IFN-γ) between LPS stimulated cells and unstimulated cells regardless of the

99 stimulation duration (Figure 5-3 c-g). When IL-1β was used to stimulate the Caco-2 cells, significantly higher MCP-1, TNF-α, and IL-8 expression was observed after 8 and 24 hours (Figure 5-3 a, c, and f) and significantly higher IL-6 expression was observed after 4 and 8 hours (Figure 5-3 e). When the LPS and IL-1β stimulation was applied in unison, a significant increase in iNOS and TNF-α was observed after 4 hours (Figure 5-3 b & c). A significant increase in IL-10 was observed after 8 hours

(Figure 5-3 g). Similarly, MCP-1 expression was significantly higher after 4 and 8 hours (Figure 5-3 a), and IL-8 expression was significantly higher at 4, 8, and 24 hour sampling points (Figure 5-3 f). Significant increases in the expression of IL-4 and IL-6 were observed after 24 hours (Figure 5-3 d and e). For the stimulation period studied, mRNA expression of IFN-γ did not show any significant difference among the four different stimulation types (Figure 5-3 h).

Figure 5-3 a: MCP-1 production

0.8 ** ** *

- 0.6 P *

M *

s 0.4

A * N * **

m ** 0.2 ***

0.0 4 Hours 8 Hours 24 Hours Stimulation period (h)

nstimulated Cells ells+LPS ells+IL1 beta ells+LPS+IL1 beta U C C C

100 Figure 5-3 b: iNOS production

S 3

s s 2

e r *

x *

N *

R *

m 1

0 4 Hours 8 Hours 24 Hours Stimulation period (h)

Unstimulated Cells Cells+LPS Cells+IL1 beta Cells+LPS+IL1 beta

Figure 5-3 c: TNF-α production

* 0.5 * * 0.4 0.3 0.2

a 0.08

F *

T ** ***

o **

***

n 0.06

e 0.04

R m 0.02

0.00 4 Hours 8 Hours 24 Hours Stimulation period (h)

Unstimulated Cells Cells+LPS Cells+IL1 beta Cells+LPS+IL1 beta

101 Figure 5-3 d: IL-4 production

1.5 * * 1.0

0.5

- 0.4

s 0.3

A 0.2

0.1

0.0 4 Hours 8 Hours 24 Hours Stimulation period (h)

Unstimulated Cells Cells+LPS Cells+IL1 beta Cells+LPS+IL1 beta

Figure 5-3 e: IL-6 production

4 ** *

e 2 r *

x *

R *

m 1 *

0 4 Hours 8 Hours 24 Hours Stimulation period (h)

Unstimulated Cells Cells+LPS Cells+IL1 beta Cells+LPS+IL1 beta

102 Figure 5-3 f: IL-8 production

1.5 * * ** 1.0 * ** 0.5 *** *** **

8 - 0.08 *

f *** **

o ***

s 0.06

A 0.04

0.02

0.00 4 Hours 8 Hours 24 Hours Stimluation period (h)

Unstimulated Cells Cells+LPS Cells+IL1 beta Cells+LPS+IL1 beta

Figure 5-3 g: IL-10 production

0.5

0.4

n 0.3

r **

x **

0.2 **

m 0.1

0.0 4 Hours 8 Hours 24 Hours Stimulation period (h)

Unstimulated Cells Cells+LPS Cells+IL1 beta Cells+LPS+IL1 beta

103 Figure 5-3 h: IFN-γ production

10 8 6 4

m 2

a 0.4

o 0.3

x 0.2

m 0.1

0.0 4 Hours 8 Hours 24 Hours Stimulation period (h)

Unstimulated Cells Cells+LPS Cells+IL1 beta Cells+LPS+IL1 beta

Figure 5-3 The mRNA expression of cytokines from Caco-2 cells with no stimulation, IL-1β, or LPS+IL-1β stimulation for 4 hrs, 8 hrs, and 24 hrs. (a) MCP-1. (b) iNOS. (c) TNF- α. (d) IL-4. (e) IL-6. (f) IL-8. (g) IL-10. (h) IFN-γ. Error bar represents the standard error of 3 replicates within each treatment.*, **, *** represents significant difference at α=0.05, at α=0.01, and α=0.001 respectively, using one-way ANOVA Tukey-Kramer Honest Significant Difference (HSD) test

5.3.3. Effects of digestion extracts combined with LPS and/or IL-1β stimulations

on Caco-2 cells

Breadfruit digestion triggered a similar cytokine response on Caco-2 cells as wheat digestion. All eight cytokines were examined, but no significant differences were observed for IL-4, IL-10, IL-8, TNF-α or IFN-γ between wheat digestion treated group and breadfruit digestion treated group (Table 5-3; Figure 5-4, 5-5, 5-6 and 5-7). For digestion extract treated unstimulated cells, 1% breadfruit treated cells had

104 significantly higher MCP-1 expression than wheat-treated cells and digestive enzyme solution treated cells (Figure 5-4 h). When LPS stimulation was applied with 1% digestion extracts, cells exposed to the breadfruit digestive extract had a higher iNOS expression than other groups (Figure 5-5 h). When IL-1β stimulation was applied, the

1% wheat digestion extract stimulated a significant IL-6 response but the breadfruit reduced the IL-6 stimulation (Figure 5-6 h). When the LPS and IL-1β stimulations were combined, the expression of IL-6 production was significantly higher in wheat treated cells than breadfruit treated cells (Figure 5-7 h).

105 Table 5-3 Cytokine response of Caco-2 after 24 hours of breadfruit/wheat digestion treatments under various stimulations

Major function Cytokine Non-stimulated LPS stimulated IL 1β stimulated IL 1β+LPS stimulated

IL 4 NSD NSD NSD NSD

Anti-inflammatory IL 10 NSD NSD NSD NSD

iNOS NSD Breadfruit > Wheat NSD NSD

TNF α NSD NSD NSD NSD

Pro-inflammatory IFN γ NSD NSD NSD NSD

MCP-1 Breadfruit > Wheat NSD NSD NSD

Chemokines IL 8 NSD NSD NSD NSD

Pleiotropic IL 6 NSD NSD Wheat > Breadfruit Wheat > Breadfruit

> represents significant higher based on alpha level of 0.05

106 Figure 5-4 a: IL-10 under no stimulation

0.6

o 0.4

A 0.2

0.0 n n n io tio io t s st lu e e o ig ig S D D e t a it ym e u z h fr n W d E a e re iv B st e ig D

Figure 5-4 b: IL-4 under no stimulation

2.0

f 1.5

r 1.0

R 0.5

0.0 n n n io tio io t s st lu e e o ig ig S D D e t a it ym e u z h fr n W d E a e re iv B st e ig D

107 Figure 5-4 c: iNOS under no stimulation

0.8

0.6

e 0.4

R 0.2

0.0 n n n io tio io t s st lu e e o ig ig S D D e t a it ym e u z h fr n W d E a e re iv B st e ig D

Figure 5-4 d: TNF-α under no stimulation

0.15

0.10

0.05

0.00 n n n io tio io t s st lu e e o ig ig S D D e t a it ym e u z h fr n W d E a e re iv B st e ig D

108 Figure 5-4 e: IFN-γ under no stimulation

a 1.5

n 1.0

N 0.5

A 0.0 n n n io tio io t s st lu e e o ig ig S D D e t a it ym e u z h fr n W d E a e re iv B st e ig D

Figure 5-4 f: IL-8 under no stimulation

0.15

s 0.10

e 0.05

0.00 n n n io tio io t s st lu e e o ig ig S D D e t a it ym e u z h fr n W d E a e re iv B st e ig D

109 Figure 5-4 g: IL-6 under no stimulation 4

o 3

e 2

u l 1

0 n n o n io ti io t s st lu e e o ig ig S D D e t a it ym e u z h fr n W d E a e re iv B st e ig D Figure 5-4 h: MCP-1 under no stimulation

2.5 *

1 - **

f 2.0

e 1.5

N 1.0

o 0.5

0.0 n n n io tio io t s st lu e e o ig ig S D D e t a it ym e u z h fr n W d E a e re iv B st e ig D Figure 5-4 Effect of digestive enzyme solution, wheat digestion, and breadfruit digestion on cytokine expression in Caco-2 cells without LPS or IL-1β stimulation for 24 hours. a) IL-4, b) IL 10, c) iNOS, d) TNF-α, e) IFN-γ, f) IL-8,g) IL- 6, h) MCP-1. Error bar represents the standard error of 3 replicates within each treatment. *, ** represent significant difference at α=0.05, α=0.01, respectively, using one-way ANOVA with HSD Test 110 Figure 5-5 a: IL-4 production under LPS stimulation

1.5

o 1.0

A 0.5

0.0

n S S io P P S t L L P la + + L u n n + o o n tim ti ti io S lu s st o e e S ig g P S D i L e t D m a it y e u z h fr n d E W a e % % r 1 1 B % 1

Figure 5-5 b: IL-10 production under LPS stimulation

e 1

n S S io P P S t L L P la + + L u n n + o o n tim ti ti io S lu s st o e e S ig g P S D i L e t D m a it y e u z h fr n d E W a e % % r 1 1 B % 1

111 Figure 5-5 c: TNF-α production under LPS stimulation

T 2

x 1

n S S io P P S t L L P la + + L u n n + o o n tim ti ti io S lu s st o e e S ig g P S D i L e t D m a it y e u z h fr n d E W a e % % r 1 1 B % 1

Figure 5-5 d: IFN-γ production under LPS stimulation

0.4

0.3

o 0.2

e 0.1

0.0

n S S io P P S t L L P la + + L u n n + o o n tim ti ti io S lu s st o e e S ig g P S D i L e t D m a it y e u z h fr n d E W a e % % r 1 1 B % 1

112 Figure 5-5 e: MCP-1 production under LPS stimulation

1.5

f 1.0

0.5

0.0

n S S io P P S t L L P la + + L u n n + o o n tim ti ti io S lu s st o e e S ig g P S D i L e t D m a it y e u z h fr n d E W a e % % r 1 1 B % 1

Figure 5-5 f: IL-8 production under LPS stimulation

e 2

R 1

n S S io P P S t L L P la + + L u n n + o o n tim ti ti io S lu s st o e e S ig g P S D i L e t D m a it y e u z h fr n d E W a e % % r 1 1 B % 1

113 Figure 5-5 g: IL-6 production under LPS stimulation

1.5

1.0

A 0.5

0.0

n S S io P P S t L L P la + + L u n n + o o n tim ti ti io S lu s st o e e S ig g P S D i L e t D m a it y e u z h fr n d E W a e % % r 1 1 B % 1 Figure 5-5 h: iNOS production under LPS stimulation 2.0 * * *

N 1.5

s 1.0

N 0.5

0.0

n S S io P P S t L L P la + + L u n n + o o n tim ti ti io S lu s st o e e S ig g P S D i L e t D m a it y e u z h fr n d E W a e % % r 1 1 B % 1 Figure 5-5 Effect of digestive enzyme solution, wheat digestion, and breadfruit digestion on cytokine expression in Caco-2 cells stimulated with LPS (1000 ng/mL) for 24 hours. a) IL-4, b) IL 10, c) TNF-α, d) IFN-γ, e) MCP-1, f) IL-8, g) IL- 6 h) iNOS. Error bar represents the standard error of 3 replicates within each treatment. * represents significant difference at α=0.05, using one-way ANOVA with HSD Test.

114 Figure 5-6 a: IL-4 production under IL-1β stimulation 0.4

L 0.3

r 0.2

m 0.1

0.0

n ta ta ta tio e e e la b b b u 1 1 1 L L IL tim I I + S + t+ it n a ru ta tio e f e u h d b l W a 1 o re S % B IL e 1 % m 1 zy n E % 1

Figure 5-6 b: IL-10 production under IL-1β stimulation

0.15

o 0.10

A 0.05

0.00

n ta ta ta tio e e e la b b b u 1 1 1 L L IL im I I + t + t+ it S n a u a io e fr t t h e lu d b o W a 1 re S % B IL e 1 m % y 1 z n E % 1

115 Figure 5-6 c: iNOS production under IL-1β stimulation

0.10

S 0.08

n 0.06

x 0.04

m 0.02

0.00

n ta ta ta tio e e e a b b b l 1 1 u 1 L im IL IL I t + t+ it+ S n a u a o e r t ti h f e lu d b o W a 1 re S % B IL e 1 % m 1 zy n E % 1

Figure 5-6 d: TNF-α production under IL-1β stimulation

0.8

0.6

o i 0.4

A 0.2

0.0

n ta ta ta tio e e e la b b b u 1 1 1 L L IL tim I I + S + t+ it n a ru ta tio e f e u h d b l W a 1 o re S % B IL e 1 % m 1 zy n E % 1

116 Figure 5-6 e: IFN-γ production under IL-1β stimulation

0.6

N 0.4

x 0.2

0.0

n ta ta ta tio e e e la b b b u 1 1 1 L L IL tim I I + S + t+ it n a ru ta tio e f e u h d b l W a 1 o re S % B IL e 1 % m 1 zy n E % 1

Figure 5-6 f: MCP-1 production under IL-1β stimulation

0.08

P 0.06

s 0.04

N 0.02

0.00

n ta ta ta tio e e e la b b b u 1 1 1 L L IL im I I + t + t+ it S n a u a io e fr t t h e lu d b o W a 1 re S % B IL e 1 m % y 1 z n E % 1

117 Figure 5-6 g: IL-8 production under IL-1β stimulation

1.5

1.0

A 0.5

0.0

n ta ta ta tio e e e la b b b u 1 1 1 L L IL tim I I + S + t+ it n a ru ta tio e f e u h d b l W a 1 o re S % B IL e 1 % m 1 zy n E % 1

Figure 5-6 h: IL-6 production under IL-1β stimulation 2.0 * **

6 - 1.5

e 1.0

R 0.5

0.0 n ta ta ta tio e e e a b b b l 1 1 1 u L m IL IL I ti + + t+ S n t i a ru ta tio e f e lu h d b o W a 1 re S % B IL e 1 m % y 1 z n E % 1 Figure 5-6 Effect of digestions on cytokine expression on Caco-2 cells stimulated with IL-1β (100 ng/mL) for 24 hours a) IL-4, b) IL 10, c) iNOS, d) TNF-α, e) IFN-γ, f) MCP-1, g) IL-8, h) IL- 6. Error bar represents the standard error of 3 replicates within each treatment. *, **represent significant difference at α=0.05, α=0.01, respectively, using one-way ANOVA with Tukey-Kramer Honest Significant Difference (HSD) test.

118 Figure 5-7 a: IL-4 production under LPS plus IL-1β stimulation 1.0

0.8

i 0.6

0.4

R m 0.2

0.0 n ta ta ta tio e e e la b b b u 1 1 1 IL IL IL tim + + + S S S S P P ta P L L e L + b + t it+ n a u 1 tio e fr IL u h d + l W a S o e P S % r L e 1 B m % zy 1 n E % 1

Figure 5-7 b: IL-10 production under LPS plus IL-1β stimulation

0.20

L 0.15

e 0.10

R 0.05

0.00 n ta ta ta tio e e e la b b b u 1 1 1 IL IL IL tim + + + S S S S P P ta P L L e L + b + t it+ n a u 1 tio e fr IL u h d + l W a S o e P S % r L e 1 B m % zy 1 n E % 1

119 Figure 5-7 c: iNOS production under LPS plus IL-1β stimulation

0.15

o 0.10

A 0.05

0.00 n ta ta ta tio e e e la b b b u 1 1 1 IL IL IL tim + + + S S S S P P ta P L L e L + b + t it+ n a u 1 tio e fr IL u h d + l W a S o e P S % r L e 1 B m % zy 1 n E % 1

Figure 5-7 d: TNF-α production under LPS plus IL-1β stimulation

4 ** ** **

a 3

o i 2

A 1

0 n ta ta ta tio e e e la b b b u 1 1 1 m IL IL IL ti + + + s S S S a P P P t L L L e + + + b n t it 1 o a u L ti e fr I u h d + l w a S o e P S % r L e 1 b m % zy 1 n E % 1

120

Figure 5-7 e: IFN-γ production under LPS plus IL-1β stimulation

1.5

N 1.0

x 0.5

0.0 n ta ta ta tio e e e la b b b u 1 1 1 IL IL IL tim + + + S S S S P P ta P L L e L + b + t it+ n a u 1 tio e fr IL u h d + l W a S o e P S % r L e 1 B m % zy 1 n E % 1

Figure 5-7 f: MCP-1 production under LPS plus IL-1β stimulation

0.3

f 0.2

e 0.1

0.0 n ta ta ta tio e e e la b b b u 1 1 1 IL IL IL tim + + + S S S S P P ta P L L e L + b + t it+ n a u 1 tio e fr IL u h d + l W a S o e P S % r L e 1 B m % zy 1 n E % 1

121 Figure 5-7 g: IL-8 production under LPS plus IL-1β stimulation 0.8

I 0.6

r 0.4

m 0.2

0.0 n ta ta ta tio e e e la b b b u 1 1 1 IL IL IL tim + + + S S S S P P ta P L L e L + b + t it+ n a u 1 tio e fr IL u h d + l W a S o e P S % r L e 1 B m % zy 1 n E % 1 Figure 5-7 h: IL-6 production under LPS plus IL-1β stimulation 8

* 6 *

L 6

e 4

R 2

0 n ta ta ta io e e t e b b la b u 1 1 1 IL IL IL tim + + + S S S S P P P ta L L L e + + + b n t it o a u 1 ti e fr IL u h d + l W a S o e P S % r L e 1 B m % zy 1 n E % 1 Figure 5-7 Effect of digestive enzyme solution, wheat and breadfruit digestions on mRNA cytokine expression in Caco-2 cells stimulated with LPS (1000ng/mL) and IL-1β (100 ng/mL) for 24 hours. a) IL-4, b) IL 10, c) iNOS, d) TNF-α, e) IFN-γ, f) MCP-1, g) IL-8, h) IL- 6. Error bar represents the standard error of 3 replicates within each treatment. *, **, *** represent significant difference at α=0.05, α=0.01, α=0.001, using one-way ANOVA with HSD

122 5.4. Discussion

5.4.1. Cytotoxicity of wheat and breadfruit digestion to Caco-2 cells

There are no previous studies of the cytotoxicity of breadfruit and few studies of wheat cytotoxicity that focused mainly on isolated and/or purified gliadin or gliadin- derived peptides, since they were identified as being important in wheat digestion issues, such as celiac disease (Ciclitira and Ellis, 1987; Lindfors et al., 2012). Wheat gliadin was found to trigger significant apoptosis in human cell lines including Caco-2 cells and this toxicity is considered to be responsible for the intestinal damage in celiac disease patients (Elli et al., 1987; Giovannini et al., 2000; 2003; Dolfini et al., 2002).

Normal healthy epithelial cells have a life span of 3-5 days (Kim et al., 1998). The epithelial cells maintain their morphology by controlling programmed cell death or apoptosis. Acute apoptosis in epithelial layer results in the deletion of epithelial cells and loss of epithelial function (Kim et al., 1998). My study attempted to closely mimic intestinal epithelium response to wheat and breadfruit digestion by using non-isolated wheat and breadfruit flour digestions containing proteins but also starches, minerals and other nutrients. I found that 1% wheat or breadfruit digestion extracts did not significantly alter Caco-2 cell viability. However, as the concentration increases, digestion treatments reduced cell viability. Interestingly, breadfruit digestion treated cells also have a higher cell viability than wheat digestion treated cells, indicating that breadfruit might have a more positive impact on cell viability than wheat.

5.4.2. Cytokine response of breadfruit- and wheat-treated Caco-2 cells

To fully understand the cytokine interaction of wheat and breadfruit digestion on Caco-2 cells, three stimulations were applied to Caco-2 cells in comparison to non-

123 stimulated cells. Caco-2 cells are hyposensitive to stimulation with LPS (Cario et al.,

2000; Eckmann et al., 1993; Savidge et al., 2006) while IL-1β was shown to induce inflammatory responses in Caco-2 cells through nuclear factor-κB pathway in a process that can be reduced by LPS (Savidge et al., 2006; Suzuki et al., 2003).

Together, these stimulants were used to model the responses of cells incubated with breadfruit and wheat digestion through the expression of cytokines (Jung et al., 1995;

Savidge et al., 2006; Suzuki et al., 2003). The results from our study confirmed the earlier reports of hyposensitivity of Caco-2 cells to LPS, and IL-1β induced a significant increase in expression of pro-inflammatory cytokines.

My study for the first time demonstrated the cytokine response of epithelial cells to breadfruit digestion in both immune stimulated conditions. In general, breadfruit digestion induced a very similar cytokine response on Caco-2 cells as wheat digestion with a few significant differences.

The first significant finding was the induction of a higher expression of MCP-1 in non-stimulated cells by breadfruit digestion extracts (Fig 5-4 h). MCP-1 is a chemokine that regulates the migration and infiltration of monocytes, memory T lymphocytes, and natural killer cells (Deshmane et al., 2009). The second significant finding was the induction of iNOS when Caco-2 cells were challenged by LPS and breadfruit digestion extracts (Fig 5-5 h). iNOS can produce nitric oxide (NO), which acts as a cytotoxic agent in pathological processes and is usually expressed in response to immune-stimulatory cytokines, bacterial products, or infection in the intestine (Aktan, 2004). Previous studies described an early response to LPS in epithelium that included the upregulated expression and production of iNOS and NO

124 (Kim et al., 1998; Klampfer, 2011; Witthöft et al., 1998). The induced iNOS expression of breadfruit may be a result of productive immune defensive responses of the epithelial cells that would be beneficial in the context of infection; but, more research using infection models are required to draw the conclusion.

Caco-2 cells stimulated with wheat digestion extract had a significantly higher

IL-6 production. IL-6 can act as pro- and anti-inflammatory cytokine in the immune response depending on its pathway, either “classic-signaling “or “trans-signaling”

(Scheller et al., 2011). Classic-signaling requires IL-6 to bind to the membrane-bound

IL 6 receptor alpha (membrane-bound IL-6 receptor), while in trans-signaling, IL-6 will bind to a soluble IL-6 receptor (Rose-John et al., 2006). Both classic- and trans- signaling are associated with membrane-bound glycoprotein 130 (gp130) and activate intracellular signaling pathways including JAK/STAT (Janus kinase/signal transducer and activator of transcription), EPK/MAPK (extracellular signal-regulated kinases/mitogen-activated protein kinase), and PI3K/AKT (phosphoinositide 3- kinase/akt) (Waldner and Neurath, 2014). Classic-signalling of IL-6 shows anti- inflammatory activities including intestinal epithelial cell proliferation and inhibition of epithelial cell apoptosis (Scheller et al., 2011). In contrast, trans-signaling leads to pro- inflammatory activities, such as inhibition of T cell apoptosis and inhibition of Treg differentiation (Becker et al., 2004a; Scheller et al., 2011). Very few cells express membrane-bound IL-6 receptor including intestinal epithelial cells, neutrophils, monocytes/macrophages and some B and T cells, while soluble IL-6 receptor can be found and activated in a wide range of cell types (Becker et al., 2004b Scheller et al.,

2011; Waldner and Neurath, 2014;). Recent studies found that IL-6/STAT 3 cascade

125 is implicated in proliferation and survival of tumor-initiating epithelial cells, both in vitro and in vivo (Lahm et al., 1992; Klampfer, 2011; Becker et al., 2004a; b; Grivennikov et al., 2009; Waldner and Neurath, 2014). These tumor-promoting effects are believed to be more relevant to trans-signalling since dysplastic epithelial cells exhibit greatly diminished surface expression of the membrane-bound IL-6 receptor (Becker et al.,

2004a). However, during late colitis-associated cancer development, enhancing IL-6 trans-signaling or both classic- and trans-signaling can accelerate tumor growth

(Grivennikov et al., 2009). Blocking of IL-6-dependent pathways has been developed in various therapeutics for gastrointestinal disease (Waldner and Neurath, 2014). The cause of the elevated IL-6 levels in response to wheat digestion extracts requires further study but this response was not seen with the breadfruit digestion extracts.

5.5. Summary

My study showed that breadfruit digestion was not toxic to intestinal cells.

Breadfruit digestion and wheat digestion induced similar cytokine responses in Caco-

2 cells with some important differences. Breadfruit digestion induced a higher production of MCP-1 in unstimulated Caco-2 cells and a higher iNOS production in

LPS stimulated cells. These data may provide an early indication of a possible protective immune response against bacterial infection that requires further study.

The induction of IL-6 expression by wheat flour and not by breadfruit flour under IL 1β and/or LPS stimulation provides an early indication that breadfruit flour may be better tolerated and may also be a topic for future research.

126 Chapter 6. Effect of the Breadfruit Diet on Mice Health

6.1. Synopsis

Model organisms are widely used to research human diseases, drug development and diet evaluation (Fields and Johnston, 2005; Hedges, 2002). Mice are one of the representative model organisms used by scientists because of the similarity in genome between mice and humans (over 90%) (Chinwalla et al., 2002;

Hedges, 2002). The evaluation of animal health and biochemical analysis of animal tissues are considered as an indirect analysis of food consumed (O’Connell and

Hedges, 1999). Although there has been a long history of breadfruit consumption by humans, the health impact of the breadfruit diet remains understudied. For example, the Mauritian population had a long-standing tradition of eating breadfruit; however,

Mahomoodally and Ramalingum (2015) showed that the Mauritian population had very limited knowledge of breadfruit nutritional value and were not aware of breadfruit medicinal value at all. About 50% of the people in the survey did not even recognize breadfruit as a healthy food (Mahomoodally and Ramalingum, 2015).

There are a limited amount of animal studies on breadfruit in the literature. These studies often had confusing experimental designs; therefore no definitive conclusions can be made (Aka et al., 2009; Adepeju et al., 2014; Grant, 1995). A properly conducted animal study of the breadfruit diet is significant not only in the exploration of breadfruit nutritional value but also as safety guidance for the breadfruit industry.

127 6.1.1. Objectives

The objective of this study was to scientifically investigate and evaluate the impact of the breadfruit diet on CB57/6J mice performance as a case study. The first objective of the study was to evaluate the overall growth progress of the mice, including weight and hematological parameters, which are fundamental in food safety examination. The second objective of the study was to examine the intestinal health of the mice, by investigating ileum morphology, cytokine response and colon bacteria, as a continuous study of the in vitro cell model experiment. The outcome of these studies provides evidence to assess whether that the breadfruit diet has any adverse impact on mice overall growth or intestinal health.

6.2. Methods

6.2.1. Reagents

Reagents used in this experiments included: DPPH (2,2-diphenyl-1- picrylhydrazyl; free radical; Sigma, USA), methanol (HPLC grade; Fisher Scientific,

USA), trolox ((±)-6-hydroxy-2, 5, 7, 8-tetramethylchromane-2-carboxylic acid; 97%;

Sigma), Rneasy® fibrous tissue mini kits (Qiagen, Germany), β-mercaptoethanol

(>99%, Sigma), iScript cDNA synthesis kits (Bio-Rad Laboratories, Ontario,

Canada), SsofastTM EvaGreen supermix (Bio-Rad Laboratories), nuclease- free water (Integrated DNA technologies, Coralville, USA), anhydrous ethyl alcohol

(molecular grade; Commercial Alcohols, Ontario, Canada), QIAamp DNA stool mini kits (Qiagen), BCA (bicinchoninc acid) protein assay kits (Thermo Fisher Scientific,

Manassas, USA), HEPES (fine white crystals, bioreagents; Fisher Scientific),

128 Tween 20 (PhytoTechnology Laboratories®, Shawnee Mission, USA), Hemoccult

Sensa (Beckman Coulter, USA).

6.2.2. Diet design

A breadfruit (BF) diet was formulated based on the standard 5LG4 diet (Purina,

Gary Summit, MO, USA) by replacing all the ground wheat and wheat components

(~45.5% of 5LG4) with breadfruit flour. Breadfruit flour was prepared from a single 6- year-old tree of the variety ‘Ma’afala’ as described previously (Chapter 4.2.2.) and provided to the Purina facility. Analysis of the two finished diets was conducted by a commercial lab to industry standard. The diets were isocaloric and nutritionally equivalent (Table 6-1).

Table 6-1 Comparison of nutritional value between the breadfruit (BF) diet and 5LG4 diet

Title Unit 5LG4 diet BF diet

Calories USA Cal/100g 363 357

Calories Canada Cal/100g 328 330

Total fat as triglycerides by GC g/100g 6.53 5.93

Saturated fatty acid g/100g 1.41 1.42

cis-Monounsaturated fatty acid g/100g 1.49 1.45

cis-Polyunsaturated fatty acid g/100g 3.26 2.71

Omega-6 fatty acids g/100g 2.85 2.3

Omega-3 fatty acids g/100g 0.41 0.41

Trans fatty acids g/100g 0.03 0.05

Conjugated linoleic acid g/100g <0.01 <0.01

129 Title Unit 5LG4 diet BF diet

Cholesterol mg/100g 28.6 23.6

Sodium mg/100g 291 223

Carbohydrates g/100g 56.7 60.1

Total dietary fibre g/100g 17.6 13.4

Total sugars g/100g 1.9 4.3

Fructose g/100g 0.24 1

Glucose g/100g <0.2 0.79

Sucrose g/100g 1.6 2.5

Maltose g/100g <0.5 <0.5

Lactose g/100g <0.5 <0.5

Protein g/100g 19.39 15.8

Protein factor 6.25 6.25

Total vitamin A RE/100g 71 <20

Retinol IU/100g 169 <50

Beta carotene IU/100g 200 <50

Calcium mg/100g 1190 1260

Iron mg/100g 39.4 40.4

Ash g/100g 6.94 7.18

Moisture g/100g 10.43 10.95

Vitamin C mg/g <1.0 <1.0

Vanadium ppm (w/w) 1.8 2.65

130 Title Unit 5LG4 diet BF diet

Zinc ppm (w/w) 65.7 54

Zirconium ppm (w/w) 0.5 0.8

Iodine ppm (w/w) 2.13 1.56

Niacin mg/100g 11.5 6.9

Vitamin B1.HCl mg/100g 3.2 1.6

Vitamin B12 mcg/100g 5.5

Vitamin B2 mg/100g 0.91 0.78

Vitamin B6 mg/100g 0.79 0.67

IU/100g 3 3.3 Vitamin E mg/100g 2 2.2

Biotin ppm (w/w) 35.6 32.8

Boron ppm (w/w) 2.6 5.9

Aluminum ppm (w/w) 147 203

Antimony ppm (w/w) 0.02 0.04

Arsenic ppm (w/w) 0.7 0.68

Barium ppm (w/w) 7.71 6.4

Beryllium ppm (w/w) 0.04 0.07

Bismuth ppm (w/w) <0.02 <0.02

Cadmium ppm (w/w) 0.076 0.098

Calcium ppm (w/w) 13400 17200

Chromium ppm (w/w) 1.54 2.23

131 Title Unit 5LG4 diet BF diet

Cobalt ppm (w/w) 0.56 0.66

Copper ppm (w/w) 7.49 7.88

Iron ppm (w/w) 345 486

Lead ppm (w/w) 0.11 0.24

Lithium ppm (w/w) 0.2 0.3

Magnesium ppm (w/w) 2190 1650

Manganese ppm (w/w) 120 143

Mercury ppm (w/w) 0.005 0.007

Molybdenum ppm (w/w) 1.57 0.94

Nickel ppm (w/w) 1.01 1.41

Phosphorus ppm (w/w) 9390 10400

Potassium ppm (w/w) 7160 11000

Selenium ppm (w/w) 0.4 0.5

Silver ppm (w/w) <0.02 <0.02

Sodium ppm (w/w) 2810 2060

Strontium ppm (w/w) 12.9 16.8

Thallium ppm (w/w) <0.01 0.01

Thorium ppm (w/w) 0.2 0.2

Tin ppm (w/w) 0.02 0.02

Titanium ppm (w/w) 19.2 24.6

Uranium ppm (w/w) 0.37 0.67

132 Non-nutrient chemistry of the diets

To determine whether the breadfruit flour included any non-nutrient phytochemicals that may have impacted the animal health, the antioxidant content of the diets was determined by standardized DPPH protocols (Turi and Murch, 2013).

Briefly, 3 g of the ground diets (3 replicates each) was mixed with 10 mL of methanol

(0.3 g/ml), centrifuged and syringe filtered (0.22 µm MCE filter, sterile; Fisher

Scientific). A serial dilution of the sample (0.165 g/mL, 0.091 g/mL, 0.050 g/mL, 0.027 g/mL, 0.015 g/mL, 0.0083 g/mL, and 0.0046 g/mL) and trolox standard solutions (1

µM, 8 µM, 16 µM, 24 µM, 32, µM, 40 µM, and 50 µM) were prepared for analysis. 100

µL diet samples or trolox standard solutions were pipetted (duplicate) into Costar 96- well EIA/RIA clear flat bottom plates (Bio-Rad Laboratories, Ontario) with 100 µL of the DPPH solution (0.1 mM in methanol). The absorbance was read by a Synergy HT

Multi-Mode Microplate Reader (BioTek Instruments, Vermont) using Gen5: Microplate

Data Collection and Analysis Software (BioTek). Data was collected each minute for a total of 20 minutes at 562 nm. Blank samples with either methanol and/or DPPH was analyzed to access the contribution of reagents. The BF diet and 5LG4 diet were not significantly different in their scavenging activity at 15 minutes, indicating that significant quantities of reactive phytochemicals were not present (Figure 6-1).

133 100 90 80 70 60 50 40 5LG4 diet: y = 27.654ln(x) + 79.667

30 R² = 0.958 15 min 15 20 Breadfruit diet: y = 28.092ln(x) + 79.902 10 R² = 0.954 0 0.000 0.005 0.010 0.015 0.020

Diet weight (g) Scavenging activity percentage (AA%) at at (AA%) percentage activity Scavenging Breadfruit diet 5LG4 diet Log. (Breadfruit diet ) Log. (5LG4 diet)

Figure 6-1 Comparison of scavenging activity percentage at 15 minutes between breadfruit diet and 5LG4 diet using DPPH methods. Bars represents standard error calculated from the three replicates

6.2.3. Experimental animals

The animal experiments were performed by the staff at Jackson Laboratories,

California, USA according to standard operating protocols. The animal protocol was approved by the University of California San Diego Institutional Animal Care and

Use Committee (IACUC) to ensure the highest quality animal use and care. Mice

(C57BL/6, Jackson Labs) were 7 week old mice at the beginning of the experiment and were housed for 3 weeks in positively ventilated polycarbonate cages with

HEPA filtered air at a density of 4 mice per cage. For identification, the mice were ear-notched. The light source was provided by using artificial fluorescent lighting, creating a 12 h light/dark cycle. The room temperature was controlled at 22±4°C with

50±15% humidity. A total of 32 mice were randomly assigned into 2 groups. Each

134 group had an equal number of female (8) and male (8) mice. One group received the

5LG4 diet and the other group received the breadfruit diet (BF diet). Problems with the breadfruit diet crumbling in the cages meant that it was not possible to accurately determine food intake or absolute amount of food consumed.

6.2.4. Overall growth evaluation and hematology analysis

Body weight, and food and water intake were measured daily at the same time. Clinical observations were made daily, examining the general health condition and any skin lesions. At the terminal point, the mice were sacrificed by CO2 asphyxiation at the end of the third week. The terminal blood was collected by cardiocentesis and the whole blood was submitted for clinical chemistry analysis in

Comparative Pathology Laboratory (CPL) (Davis, CA, USA). The whole body composition measurements of fat, lean, free water, and total water mass was conducted using EchoMRI-100 Analyzer (EchoMRITM, USA) in the Jackson

Laboratory.

6.2.5. Tissue collection

The organs were weighed individually and sent as snap-frozen tissue to the

UBC Okanagan. Feces were collected on days 3, 7, 14, and 21 from each cage, frozen in liquid nitrogen and then sent to the UBC Okanagan campus. One-third of the ileum sections of the mice were stored in RNAlater®, an RNA stabilization reagent (Qiagen®), and shipped frozen to the UBC Okanagan campus. A 3- to 5-mm section of the ileum was fixed in 10% buffered neutral formalin with paraffin,

135 sectioned following standard histological methods and then stained by Hematoxylin and Eosin (H&E) method. The frozen tissues, feces and RNA later ileum section were stored in -80° before analysis.

6.2.6. Feces protein, total mineral analysis, and fecal occult blood detection

Feces from the same cage and the same day were homogenized and mixed using a CoorsTM porcelain mortar and pestle (Sigma, USA). The protein content was extracted from the feces using a protein extraction buffer as described in Chapter

4.2.4. In summary, 10 mg of the feces were homogenized with 1 mL of protein extraction buffer (20 mM HEPES, 150 mM NaCl) and sonicated for 20 minutes. The protein extracts were diluted 1:1 in the protein extraction buffer before BCA analysis.

The BCA analysis was conducted following the standard protocol as described in

Chapter 4.2.4. The mineral content of the feces was calculated by the difference of weight of the feces before and after incineration at 600°C for 2 h in an Isotemp programmable muffle furnace (Fisher Scientific, Ottawa, ON), according to AACC

08-03.The fecal occult blood was detected using a Hemoccult Sensa® kit (Beckman

Coulter, USA) following the manufacturer’s protocol.

6.2.7. Ileum morphology examination

The morphological analysis of the H&E stained ileum slides was conducted, following the method published by Generoso et al. (2015). Villi pictures (10 pictures × 4 mice) were taken from each group using a camera (QImaging, Surrey,

BC, Canada) attached to a microscope (Olympus America Inc., PA, USA) and each

136 parameter was measured blindly using ImageJ 1.47v software (Wayne Rasband,

National Institutes of Health, Maryland, USA). The following parameters were measured: distance between villi; villus height and thickness; the crypt depth; the thickness of lamina propria and epithelium: the length of mucosa, submucosa, and muscularis externa; and the number of goblet cells and red blood cells.

6.2.8. Major cytokine response on ileum

RNA extraction was performed on ileum tissue of the mice, using a RNeasy® fibrous tissue mini kit, following the manufacturer’s instructions. Briefly, the ileum tissue was lysed in Buffer RLT with β-mercaptoethanol using the Mixer Mill MM 400

(Retsch®, Germany). The sample was incubated with proteinase K at 55°C to remove protein content. Debris content was removed by centrifugation, and the supernatant was mixed with ethanol and filtered through an RNeasy mini column.

The RNA of the sample bound to the silica membrane of the column and the impurities and DNA in the sample were washed away by the DNase solution and

Buffer RDD, Buffer RW1, and Buffer RPE. Finally, the RNA was collected in RNase- free water and the concentration of the RNA was determined by the NanoDrop

2000c UV-Vis Spectrophotometer (Thermo Fisher Scientific, Manassas, USA).

Based on the RNA concentration, complementary DNA (cDNA) copies were synthesized using an iScript cDNA synthesis kit by following the manufacturer’s instructions (as described in Chapter 4.2.7.). cDNA samples were diluted 10 times in

RNase-free water.

The real-time polymerase chain reaction (RT-qPCR) analysis was conducted

137 on the cDNA samples using a SsofastTM EvaGreen Supermix, following the manufacturer’s instructions (as described in Chapter 4.2.7.). The production of TNF-

α, IFN-γ, IL-10, IL-6, and iNOS was examined, using 18S rRNA as a reference gene.

The design and testing of the primers used in the study were conducted by Ghosh et al. (2015).

6.2.9. Quantification of bacterial groups in the colon

DNA was extracted from the frozen colon tissue of the mice using a QIAamp

DNA stool mini kit following the manufacturer’s protocol. In summary, one fifth of the frozen colon tissue was collected and homogenized in Buffer ASL. The protein in the sample was digested by proteinase K in a 70°C water bath. The DNA was bound to the QIAamp membrane and the impurities were washed away by Buffer AW1, AW2, and PRE. Eventually, the DNA was collected in Buffer AE. The concentration of the

DNA was determined by the NanoDrop 2000c UV-Vis Spectrophotometer. The production of Firmicutes members, Bacteroidetes members, Enterobacteriacae family, Lactobacillus spp., and Bifidobacterium spp. in the colon was quantified by

RT-qPCR, using Eubacteria production as a reference in the analysis. The sequence specificity of the primers was verified by Baker et al. in 2012. The qPCR analysis was conducted in the same manner as described in Chapter 4.2.7.

6.2.10. Statistical analysis

All statistical analysis were conducted using JMP® 10.0.0 (SAS Institute, Cary,

NC) and GraphPad Prism 5.0 software (GraphPad Software Inc., La Jolla, CA). A

138 series of independent two sample t-tests or two-way ANOVA analyses were conducted to determine if there were significant differences between treatments with a type 1 error rate of 0.05. Graphs were created using GraphPad Prism 5.0 software and Microsoft Excel 2013 (Microsoft Corporation, Santa Rosa, CA).

6.3. Results

6.3.1. Overall health and growth

During the three-week trial, there was no report of death or any abnormal clinical symptoms for both male and female BF-fed (breadfruit-fed) mice. Both BF- and 5LG4-fed mice showed similar growth patterns (Figure 6-2a). Three weeks of the breadfruit diet resulted in a slightly higher growth rate (7.32% overall, 5.82% male, 8.82% female) in mice than the 5LG4 diet (Figure 6-2b), with an equal effect on both male and female mice (two-way ANOVA, Diet effect: p<0.0001<α=0.05,

Gender effect: p=0.3020> α =0.05, Interaction Effect: p=0.1232> α =0.05). Since the breadfruit diet was very crumbly, the food loss during the trial was significant, and more breadfruit diet was used in the feeding, therefore the usage cannot be interpreted as the food intake (Figure 6-3). There was a significant higher daily water consumption (12.7% overall, 14.4% male, 10.6% female) in mice fed on the breadfruit diet compared to mice fed on the 5LG4 diet (Figure 6-4a & b), but there was no interaction between diet and gender (two-way ANOVA, Diet effect: p<0.0001< α =0.0001, Gender effect: p=0.0011< α =0.05, Interaction effect: p=0.2059> α =0.05).

139 Figure 6-2 a

10 Body weight weight (g) Body

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Days 5LG4-fed mice Female 5LG4-fed mice Male 5LG4-fed mice BF-fed mice Male BF-fed mice Female BF-fed mice

Figure 6-2 b

14% 12% *** *** ** 10% 8% 6%

Growth rate Growth 4% 2% 0% 5LG4 diet BF diet 5LG4 diet BF diet 5LG4 diet BF diet Mice Male mice Female mice

Figure 6-2 Comparison of daily body weight (a) and growth rate (b) of mice fed on breadfruit (BF) and 5LG4 diets. Bars represents the standard error of the measurement among eight mice from the same group. * represents significant difference at α=0.05, ** represents significant difference at α=0.01, *** represents significant difference at α=0.001

140 25

5 Food consumption (g) consumptionFood 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Experimental Days

Female 5LG4-fed mice Male 5LG4-fed mice Male BF-fed mice Female BF-fed mice

Figure 6-3 Comparison of food usage of mice fed on the breadfruit (BF) and 5LG4 diets

Figure 6-4 a

Water consumption (mL)consumption Water 1

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Experimental Days 5LG4-fed mice Female 5LG4-fed mice Male 5LG4-fed mice BF-fed mice

141 Figure 6-4 b

4.5 *** *** ** 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5

Daily (ml/animal/day)intake water Daily 0.0 5LG4 diet BF diet 5LG4 diet BF diet 5LG4 diet BF diet Mice Male mice Female mice

Figure 6-4 Comparison of daily water consumption (a) and average daily water consumption (b) of mice fed on breadfruit (BF) and 5LG4 diets. Bars represents the standard error of the measurement among eight mice from the same group. * represents significant difference at α=0.05, ** represents significant difference at α=0.01, *** represents significant difference at α=0.001

At the end of the three-week trial, the body composition was similar between

BF and 5LG4 diet-fed mice (Table 6-2). Male mice consuming the BF diet had significantly higher body weight (4% higher) and lower fat (2.2 % less) than male mice consuming the 5LG4 diet. There were few significant differences in the weights of various tissues and organs in response to the diet. Female BF-fed mice had significantly larger duodena and male BF-fed mice had significantly heavier femurs, while all the other organs, including the brain, heart, liver, and kidney, were not significantly different in weight.

142 Table 6-2 Comparison of average body composition and tissue weight between the BF-fed (breadfruit-fed) mice and 5LG4-fed mice

Sex Male mice Female mice 5LG4 diet BF diet Diet 5LG4 diet BF diet 5LG4 Diet BF Diet

Name Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE

Body weight (g) 21.0 0.7 22.2 0.7 23.7a 0.4 24.7b 0.3 18.4 0.3 19.7 0.6

Fat 17.1% 0.7% 15.7% 1.0% 15.0%a 0.5% 12.8%b 0.6% 19.1% 0.7% 18.5% 1.2%

Lean 73.4% 0.9% 72.9% 0.9% 75.4% 1.1% 74.7% 0.7% 71.5% 0.9% 71.1% 1.5%

Free water 0.5% 0.1% 0.5% 0.1% 0.5% 0.1% 0.4% 0.0% 0.4% 0.1% 0.6% 0.1%

Total water 60.7% 0.7% 60.4% 0.8% 62.3% 0.9% 62.0% 0.5% 59.0% 0.8% 58.9% 1.3%

Brain (g) 0.45 0.004 0.44 0.007 0.46 0.003 0.45 0.006 0.44 0.006 0.43 0.013

Heart (g) 0.12 0.003 0.12 0.005 0.13 0.004 0.13 0.006 0.11 0.004 0.11 0.005

Kidney (g) 0.25 0.009 0.26 0.008 0.28 0.009 0.29 0.006 0.22 0.008 0.24 0.008

Bladder (g) 0.02 0.002 0.02 0.002 0.02 0.001 0.03 0.003 0.02 0.003 0.01 0.002

143 Sex Male mice Female mice 5LG4 diet BF diet Diet 5LG4 diet BF diet 5LG4 Diet BF Diet

Name Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE

Lungs (g) 0.23 0.010 0.22 0.010 0.24 0.011 0.22 0.010 0.21 0.014 0.21 0.018

Thymus (g) 0.03 0.004 0.03 0.005 0.03 0.006 0.03 0.007 0.03 0.006 0.04 0.005

Spleen (g) 0.06 0.002 0.06 0.003 0.06 0.002 0.07 0.003 0.06 0.004 0.06 0.005

Lymph node (g) 0.01 0.003 0.01 0.002 0.02 0.005 0.01 0.003 0.01 0.004 0.01 0.002

Salivary gland (g) 0.12 0.011 0.12 0.013 0.14 0.017 0.14 0.016 0.11 0.015 0.10 0.019

Liver (g) 0.94 0.052 0.88 0.066 1.12 0.033 0.93 0.126 0.76 0.033 0.83 0.047

Pancreas (g) 0.11 0.007 0.11 0.006 0.12 0.011 0.13 0.006 0.10 0.010 0.10 0.008

Duodenum (g) 0.28 0.024 0.22 0.021 0.28 0.037 0.22 0.025 0.20a 0.027 0.31b 0.028

Jejunum (g) 0.25 0.042 0.21 0.012 0.23 0.024 0.22 0.018 0.27 0.084 0.19 0.013

Ileum (g) 0.20 0.037 0.18 0.008 0.22 0.071 0.19 0.011 0.18 0.028 0.17 0.010

Colon (g) 0.15 0.009 0.16 0.012 0.16 0.012 0.16 0.018 0.13 0.013 0.15 0.017

144 Sex Male mice Female mice 5LG4 diet BF diet Diet 5LG4 diet BF diet 5LG4 Diet BF Diet

Name Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE

Testes (g) 0.18 0.024 0.18 0.024 0.18 0.004 0.18 0.005 / / / /

Epididymis (g) 0.05 0.007 0.05 0.007 0.05 0.007 0.05 0.005 / / / /

Uterus (g) 0.09 0.014 0.08 0.011 / / / / 0.09 0.016 0.08 0.010

Mammary gland (g) 0.22 0.012 0.20 0.015 0.22 0.015 0.18 0.016 0.23 0.021 0.23 0.023

Pituitary gland (g) 0.002 0.001 0.002 0.000 0.002 0.001 0.002 0.001 0.002 0.001 0.001 0.000

Adrenal gland (g) 0.01 0.001 0.01 0.001 0.01 0.001 0.01 0.001 0.01 0.001 0.01 0.000

Ears (g) 0.04 0.005 0.04 0.004 0.06 0.006 0.04 0.007 0.03 0.004 0.04 0.005

Tail tip (g) 0.06 0.006 0.13 0.047 0.06 0.007 0.12 0.058 0.06 0.010 0.13 0.078

Skin (g) 0.009 0.003 0.007 0.001 0.013 0.006 0.008 0.001 0.005 0.000 0.006 0.001

Femur (g) 0.06 0.003 0.09 0.009 0.07a 0.005 0.08b 0.002 0.06 0.003 0.09 0.019

Numbers followed by different letter in each section are significant difference at alpha=0.05, using 2 sample t test

145 6.3.2. Hematology analysis

BF-fed mice established a slightly different white blood cell profile, but very similar red blood cell and platelets (thrombocytes) profiles as the 5LG4-fed mice

(Table 6-3). For the total white blood cells as well as the two major components of white blood cells, neutrophils and lymphocytes, gender played a role in the mice’s responses to the BF diet (two-way ANOVA, Interaction effect between gender and diet for white blood cell: p=0.0006<α=0.05, for neutrophils: p=0.0077<α=0.05, and for lymphocytes: p=0.0002<α=0.05). Female BF-fed mice had significantly lower white blood cell, neutrophil, and lymphocyte count than female 5LG4-fed mice while male BF-fed mice had significant higher lymphocyte than male 5LG4-fed mice. No significant difference was found in the monocyte cell number between the two diets in both genders. For eosinophil and basophil counts, there was no significant difference between the two diets in the female mice, but male BF-fed mice had significantly lower eosinophil and basophil count than male 5LG4-fed mice. When the cell counts were converted into the percentage content in white blood cells, the differences between the two diet groups became smaller. Male 5LG4-fed mice had significantly lower lymphocyte and eosinophil percentage than male BF-fed mice, while female 5LG4-fed mice had significantly higher neutrophil percentage than female BF-fed mice. Differences found in the red blood cell and platelet profiles were not statistically significant with the exception of female BF-fed mice that had a significantly higher MCV (mean corpuscular volume) content than female 5LG4 fed mice.

146 Table 6-3 Comparison of blood hematology between mice fed on the breadfruit (BF) diet and 5LG4 diet

Sex Male mice Female mice 5LG4 diet BF diet Diet 5LG4 diet BF diet 5LG4 Diet BF diet

Title Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE

White blood cell (K/µl) 7.1 0.7 5.7 0.5 5.6 0.3 7.0 0.7 8.5a 1.1 4.4b 0.5

Absolute neutrophil cells (K/µl) 1.7a 0.2 1.2b 0.1 1.5 0.1 1.6 0.2 1.9a 0.3 0.8b 0.1

Absolute lymphocyte cells (K/µl) 4.9 0.5 4.2 0.4 3.7a 0.3 5.2b 0.5 6.1a 0.8 3.2b 0.4

Absolute monocyte cells (K/µl) 0.23 0.03 0.21 0.02 0.19 0.03 0.19 0.03 0.28 0.05 0.23 0.03

Absolute eosinophil cells (K/µl) 0.19 0.03 0.12 0.02 0.16a 0.02 0.09b 0.02 0.22 0.06 0.15 0.03

Absolute basophil cells (K/µl) 0.06a 0.01 0.03b 0.01 0.06a 0.01 0.03b 0.01 0.06 0.02 0.04 0.01

Neutrophil % 24.5a 1.4 19.7b 1.2 27.4 2.0 21.8 2.0 21.5a 1.2 17.7b 0.8

Lymphocyte % 68.6a 1.8 73.6b 1.2 65.1a 2.7 74.1b 2.1 72.1 1.8 73.1 1.3

Monocyte % 3.3 0.3 3.9 0.4 3.4 0.5 2.6 0.2 3.2 0.5 5.2 0.3

Eosinophil % 2.7 0.3 2.2 0.4 3.0a 0.4 1.2b 0.3 2.5 0.6 3.2 0.5

Basophil % 0.9 0.1 0.6 0.1 1.1 0.2 0.4 0.1 0.7 0.2 0.8 0.1

147 Sex Male mice Female mice 5LG4 diet BF diet Diet 5LG4 diet BF diet 5LG4 Diet BF diet

Title Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE

Red blood cell (M/µl) 10.7a 0.1 10.1b 0.2 10.8 0.2 10.0 0.4 10.7 0.1 10.3 0.2

Hemoglobin (g/dL) 14.2a 0.2 13.5b 0.3 14.4 0.3 13.4 0.5 14.1 0.3 13.6 0.1

Hematocrit % 43.3 0.6 41.1 0.9 44.4 0.9 40.9 1.7 42.1 0.7 41.3 0.7

MCV (fL) 40.3 0.2 40.6 0.2 41.1 0.2 40.9 0.1 39.5a 0.2 40.2b 0.2

MCH (pg) 13.3 0.1 13.3 0.1 13.3 0.1 13.4 0.1 13.2 0.1 13.3 0.1

MCHC (g/dL) 32.9 0.2 32.9 0.1 32.5 0.3 32.8 0.2 33.4 0.3 33.0 0.2

RDW % 17.9 0.1 17.6 0.1 17.7 0.1 17.3 0.1 18.1 0.2 17.9 0.2

Platelets (K/µL) 1260 53 1087 81 1327 84 1105 118 1193 61 1069 120

MPV (fL) 5.5 0.0 5.7 0.1 5.6 0.0 5.6 0.1 5.5 0.1 5.7 0.1

Numbers followed by different letter in each section are significant difference at alpha=0.05, using 2 sample t test. MCV (mean corpuscular volume), MCH (mean corpuscular hemoglobin), MCHC (mean corpuscular hemoglobin concentration), RDW (red cell distribution width), MPV (mean platelet volume)

148 6.3.3. Fecal protein, total mineral analysis, and fecal blood detection

Feces collected from the BF-fed mice tended to have a higher (1-7%) protein content than the 5LG4-fed mice (Figure 6-5). Both male and female BF-fed mice showed a significantly higher fecal protein than 5LG4-fed mice at Day 7 (Figure 6-

5b). However, in terms of total mineral content, the breadfruit diet fed mice showed a trend of lower percentage (1-6%) than the 5LG4 diet fed mice (Figure 6-6), especially at Day 7 and 14, where both male and female BF-fed mice had significantly lower fecal mineral content than 5LG4 fed mice (Figure 6-6 b & c). By the end of the three-week trial (Day 21), there was no significant difference in the fecal protein or mineral content between the two diet groups (Figure 6-5d & Figure 6-

6d). In spite of the collection date, gender, and diet, the fecal blood detection showed that there was no blood in all of the feces collected.

Figure 6-5 a

20%

16%

12%

4% Fecal protein content at Day Day protein3 atcontent Fecal

0% 5LG4 Diet BF Diet 5LG4 Diet BF Diet 5LG4 Diet BF Diet Mice Male mice Female mice

149 Figure 6-5 b

25% ** *** * 20%

15%

10%

5% Fecal protein content at Day Day protein7 atcontent Fecal

0% 5LG4 Diet BF Diet 5LG4 Diet BF Diet 5LG4 Diet BF Diet Mice Male mice Female mice

Figure 6-5 c 25% * * 20%

15%

10%

5% Fecal protein content at Day Day protein14atcontent Fecal 0% 5LG4 Diet BF Diet 5LG4 Diet BF Diet 5LG4 Diet BF Diet Mice Male mice Female mice

150 Figure 6-5 d

20%

16%

12%

4% Fecal protein content at Day Day protein21atcontent Fecal 0% 5LG4 Diet BF Diet 5LG4 Diet BF Diet 5LG4 Diet BF Diet Mice Male mice Female mice

Figure 6-5 Comparison of fecal protein content from mice fed with the breadfruit (BF) and 5LG4 diets at Day 4 (a), Day 7 (b), Day 14 (c), and Day 21 (d). *, **, *** represents a significant difference between the two groups using 2 sample t-test at alpha level 0.05, 0.01, 0.001, respectively. Bars represents the standard error of the measurement among eight mice from the same group

Figure 6-6 a 16%

14% ** ***

12%

10%

2% Fecal total minerial content at Day Day 3 atcontent minerialtotalFecal 0% 5LG4 Diet BF Diet 5LG4 Diet BF Diet 5LG4 Diet BF Diet

Mice Male mice Female mice 151 Figure 6-6 b

16%

14% *** ** ** 12%

10%

Fecal total mineral content at Day Day 7 atmineraltotalcontent Fecal 0% 5LG4 Diet BF Diet 5LG4 Diet BF Diet 5LG4 Diet BF Diet Mice Male mice Female mice

Figure 6-6 c 16% *** * ** 14%

12%

10%

Fecal total mineral content at Day Day 14atmineraltotalcontent Fecal 0% 5LG4 Diet BF Diet 5LG4 Diet BF Diet 5LG4 Diet BF Diet Mice Male mice Female mice

152 Figure 6-6 d 18%

16%

14%

12%

10%

Fecal total mineral content at Day Day 21atmineraltotalcontent Fecal 0% 5LG4 Diet BF Diet 5LG4 Diet BF Diet 5LG4 Diet BF Diet Mice Male mice Female mice

Figure 6-6 Comparison of total mineral content from mice fed with the breadfruit (BF) and 5LG4 diets at Day 4 (a), Day 7 (b), Day 14 (c), and Day 21 (d). *, **, *** represents a significant difference between the two group using 2 sample t-test at alpha level 0.05, 0.01, 0.001, respectively. Bars represents the standard error of the measurement among eight mice from the same group

6.3.4. Ileum morphology examination

Mice from both diet groups had no significant difference in the villus height and thickness; distance between the villi; crypt depth; thickness of lamina propria and epithelium; length of mucosa, submucosa, and muscularis externa; and number of goblet cells and red blood cells (Table 6-4). There was no difference in the ileum morphology between 5LG4-fed and BF-fed mice (Figure 6-7).

153 Table 6-4: Histological examination of ileum morphology

Sex Male mice Female mice 5LG4 diet BF diet Diet 5LG4 diet BF diet 5LG4 diet BF diet Title Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE

Distance between villi 6.43 0.95 6.28 0.72 6.26 0.94 5.92 0.49 6.60 1.50 6.63 1.23 (µm)

Height of villus 110.2 9.23 117.3 6.66 120.4 10.41 125.5 6.46 99.98 11.74 109.0 8.90 (µm)

Thickness of 51.04 2.83 52.41 2.50 53.36 1.49 55.59 3.51 48.73 4.81 49.23 2.14 villus (µm)

Crypt depth 68.75 5.00 56.66 2.31 70.38 7.93 61.92 2.07 67.12 4.85 51.40 0.78 (µm)

Thickness of lamina propria 20.71 1.98 18.01 1.93 21.07 1.30 19.05 2.91 20.36 3.46 16.97 2.00 (µm)

Epithelium 15.48 1.09 17.26 0.60 16.45 1.63 18.40 0.52 14.50 1.02 16.13 0.60 thickness (µm)

154 Sex Male mice Female mice 5LG4 diet BF diet Diet 5LG4 diet BF diet 5LG4 diet BF diet Title Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE

Length of 171.3 13.05 170.2 8.00 183.3 16.03 182.8 7.33 159.3 16.33 157.7 9.57 mucosa (µm)

Length of submucosa 9.90 0.68 9.26 0.52 8.86 0.41 9.44 0.60 10.95 0.95 9.08 0.76 (µm)

Length of muscularis 21.24 2.07 21.25 1.50 22.08 3.40 23.50 2.28 20.39 1.77 19.01 0.40 externa (µm)

Number of 6 0 6 0 7 0 7 0 5 0 6 0 goblet cells

Number of red 5 2 5 2 5 1 3 2 6 3 7 2 blood cells

155 Figure 6-7 a: 5LG4-fed male mice

Figure 6-7 b: BF-fed male mice

156 Figure 6-7 c: 5LG4-fed female mice

Figure 6-7 d: BF-fed female mice

Figure 6-7 Representative histopathology of the ileum of 5LG4-fed male mice (a), BF- fed male mice (b), 5LG4-fed female mice (c), and BF-fed female mice (d) at 3 weeks of age stained with haematoxylin and eosin (H&E).Scale bar is 50 µm. 157 6.3.5. Major cytokine response on ileum

Cytokines, including TNF-α, IL-10, IL 6, and IFN-γ, were examined in the ileum tissue of the mice and no significant difference was found as a result of consumption of the 5LG4 or breadfruit diet (Figure 6-8 a-d). Female BF-fed mice had significantly higher iNOS production than female 5LG4-fed mice (Figure 6-8e); however, the difference was less than 0.5 fold. BF-fed mice had very similar cytokine responses to 5LG4-fed mice in their ileums.

Figure 6-8 a: TNF-α production

p 8

T 6

s 4

e 2

N R 0 m Male mice ileum Female mice ileum 5LG4 Diet BF Diet

158 Figure 6-8 b: IFN-γ production

m 10

N 8

o 6

s 4

R 0

m Male mice ileum Female mice ileum 5LG4 Diet BF Diet

Figure 6-8 c: IL-10 production

0 10

f 8

s 6

x 4

N 2

m 0 Male mice ileum Female mice ileum 5LG4 Diet BF Diet

159 Figure 6-8 d: IL-6 production

f 8

s 6

x 4

N 2

m 0 Male mice ileum Female mice ileum 5LG4 Diet BF Diet

Figure 6-8 e: iNOS production

o 6

e 4

A 2

N *

m 0 Male mice ileum Female mice ileum 5LG4 Diet BF Diet

Figure 6-8 Comparison of cytokine expression between ileums of breadfruit (BF)-and 5LG4-fed mice, a) TNF-α, b) IFN-γ, c) IL-10, d) IL-6, e) iNOS. * represents a significant difference between the two groups using 2 sample t-test at alpha level 0.05. Middle line denotes media, whiskers are determined by Tukey method, which represent largest value below 75th percentile plus 1.5 times interquartile distance IQR or lowest value above 25th percentile minus 1.5 times IQR.

160 6.3.6. Quantification of bacterial groups in the colon

Gene expression of bacterial groups, Bifidobacterium spp., Lactobacillus spp., and Enterobacteriacae in the colon were compared and no significant differences were found between 5LG4-fed and BF-fed mice (Figure 6-9 a & b). Both male and female mice fed on the breadfruit diet showed a higher gene expression of

Bacteroidetes than mice fed on the 5LG4 diet. Female BF-fed mice showed a higher gene expression of Firmicutes. However, the ratio between Firmicutes and

Bacteroidetes was not significantly different between the two diets (male 5LG4-fed mice: 0.61±0.17, male BF-fed mice: 0.20±0.05, female 5LG4-fed mice: 0.45±0.12, female BF-fed mice: 0.30±0.13) (Figure 6-9).

Figure 6-9 a: Bacteroidetes

o *

e 60

e *

e 40

t 20

B 0 Male mice Female mice 5LG4 Diet BF diet

161 Figure 6-9 b: Firmicutes

10 ***

s 8

x 6

e t 4

r i 2

0 Male mice Female mice 5LG4 Diet BF diet

Figure 6-9 c: Bifidobacterium spp.

i 40

x 30

s 20

a 10

B 0 Male mice Female mice 5LG4 Diet BF diet

162 Figure 6-9 d: Lactobacillus spp.

n 80

p 60

a 20

a L 0 Male mice Female mice 5LG4 Diet BF diet

Figure 6-9 e: Enterobacteriacae

i 400

x 300

e c 200

b 100

E 0 Male mice Female mice 5LG4 Diet BF diet

Figure 6-9 Comparison of bacteria expression between ileums of breadfruit diet and 5LG4 diet fed mice, a) Bacterioidetes, b) Firmicutes, c) Bifidobacterium spp., d) Lactobacillus spp., e) Enterobacteriacae. *, **, *** represents significant difference between the two groups using 2 sample t-test at alpha level 0.05, 0.01, 0.001, respectively. Middle line denotes media, whiskers are determined by Tukey method, which represent largest value below 75th percentile plus 1.5 times interquartile distance (IQR) or lowest value above 25th percentile minus 1.5 times IQR.

163 6.4. Discussion

6.4.1. Overall heath and growth

Of the few animal studies, especially rodent studies, found related to breadfruit, definitive conclusions are not possible (Adepeju et al., 2014; Grant, 1995;

Aka et al., 2009). Grant (1995) reported the death of 4 male Hooded Lister rats after consumption of breadfruit seed extracts. The small sample size in the experimental design makes the data difficult to interpret, the preparation of the diet is unclear and it is possible that the species was misidentified. The paper identified the seeds as representing the species A. altilis (breadfruit) seeds obtained from a tree at the

Mayaguez Institute of Tropical Agriculture (MITA, Puerto Rico). However, the GRIN

(Germplasm Resources Information Network) database provided by the U.S.

Department of Agriculture, shows that the only A. altilis cultivar that might possibly produce seeds was not received until 2005. In the 1990s, Puerto Rico had only seedless A. altilis, derived from the British and French introductions in late 18th century. Therefore, the plant material used in this paper is suspect. Aka et al. (2009) described the impact of the breadfruit diet on male albino rats. They reported that a raw and partially raw breadfruit diet induced severe weight loss in male rats, but cooked breadfruit resulted in weight gain (Aka et al., 2009). However, it was not clear whether breadfruit was used as a supplement along with the control diet or was processed into a new diet, making it impossible to know what exactly was fed to the mice (Aka et al., 2009). It was also not clear which part of the breadfruit was used.

No information about the composition of either the breadfruit or control diets was presented in the paper (Aka et al., 2009). Similar problems can be found in Adepeju

164 et al.’s (2014) weanling albino rats (both sex) study, in which they reported weight loss in weanling rats fed on a breadfruit diet. In the paper, the diet was simply referred to as various ratios of breadfruit, soybeans, and groundnut without any nutritional analysis or method of preparation to establish the baseline of the study

(Adepeju et al., 2014).

My study is the first complete, fully-designed experiment, and clearly showed that the breadfruit diet did not cause death, clinical discomfort or weight loss. Both male and female mice fed on the breadfruit diet in this study showed a significant increase in body weight/growth rate as compared to the mice fed on the 5LG4 diet

(control diet). This finding is interesting for many reasons. BF-fed (breadfruit-fed) mice had higher body weight than 5LG4 fed mice, but it was not reflected in their fat percentages. Male BF-fed mice actually had a significantly lower fat content than

5LG4-fed mice. Although BF-fed mice had higher body weight, their significantly lower fat content excludes the possibility of obesity. The difference in the whole body weight was not associated with weight gain in any particular organ either. The majority of the organs in the BF-fed mice remained the same as the 5LG4-fed mice.

BF-fed mice showed a larger body mass than 5LG4 fed mice without significant increase in fat or a particular organ after three weeks trial.

Hematology analysis

My data showing effects of the BF diet on rodent hematology contradicts previous reports. Adepeju et al. (2014) reported a decrease in percentage hematocrit (red blood cell related parameter) and an increase in the number of white blood cells and platelets in rats fed breadfruit. Aka et al. (2009) reported a significant

165 decrease in the number of white blood cells and red blood cells, and in the percentage of red blood cell related parameters (hemoglobin, hematocrit, MCV,

MCH, and MCHC), in male rats fed on raw and partially raw breadfruit diets. Male rats fed on the cooked breadfruit diet had no difference in white blood cell numbers but significantly higher MCV and hemoglobin (Aka et al., 2009).

In the current study, differences in white blood cell numbers, red blood cells numbers, red blood cell related parameters or platelets between the male BF- and

5LG4-fed mice were not observed. Significant decreases in major leukocyte cell counts in female BF-fed mice were observed but there is insufficient evidence to determine the cause of this decrease. Mice, as an animal model, experience more biological variations in hematological data than larger animals such as dogs or pigs

(Hedrich, 2012). The current data showed that female BF-fed mice had a white blood cell count between 2.10 to 5.90 k/µL and the female 5LG4-fed mice had a range between 4.94 and 15 k/µL. The reported reference range for white blood cells in female CB57/6J mice can be as low as 1.75 k/µL and as high as 14.9 k/µL for 8-

12 week old female mice according to Jackson’s laboratory database (Jackson’s

Laboratory, 2015). The lowest white blood cell counts found in the BF-fed mice is higher than the reference value and the value from 5LG4-fed mice is higher than the normal range. In addition, the blood samples were collected from the heart through cardiocentesis, which was reported to have lower values than blood samples collected from tails or eye (Nemzek et al., 2001). Further study is required to replicate and validate these results.

166 6.4.2. Intestinal health evaluation

Ileum morphology and cytokine response

This study is the first study to investigate the impact of the breadfruit diet on the intestinal health of an animal. I found that there was no significant morphology difference between the ileum tissues collected from BF- and 5LG4-fed mice. Mice fed on the breadfruit diet had very similar cytokine responses as mice fed on the

5LG4 diet, except that there was a slightly higher expression of iNOS (<0.5 fold) in the breadfruit diet group.

This finding was consistent with our previous in vitro cell model study, in which the breadfruit digestion induced a higher (~1 fold) iNOS production on Caco-2 cells (human epithelial colorectal adenocarcinoma cells) than the wheat digestion under LPS stimulation. LPS is an endotoxin found in Gram-negative bacteria, which are found to be resident in the intestinal tracts of both humans and rodents (van der

Waaij et al., 2005). The consistency between the in vitro cell model and in vivo animal study shows the high validity and reliability of the experiment.

iNOS is a pro-inflammatory cytokine that can produce nitric oxide, which acts as a cytotoxic agent in the pathological process, and can regulate mucosal barrier function (Aktan, 2004; Hoffman et al., 1997). The production of iNOS in normal mice ileum mucosa has been well demonstrated by Hoffman et al. in 1997, in which they explained that normal mice ileum mucosa express iNOS due to the numbers of bacteria present and this expression can vary depending on the bacterial state. It is possible that BF-fed mice had different bacterial groups resident in the ileum than

5LG4-fed mice. This possibility should be tested further, using DNA sequencing.

167 Colon bacteria

The current study examined the major bacterial groups in the colon, which are closely correlated with the bacterial groups in the ileum. The significant differences in the colon bacteria between the diets were found in the two major phyla groups,

Bacteroidetes and Firmicutes. The significant increased Bacteroidetes and

Firmicutes expressions in BF-fed mice indicate that there is a higher bacteria population present in the BF-fed mice colon. Bacteria diversity is very important in intestinal immunity and it is related to issues like obesity, diabetes, and intestinal diseases (Maslowski and Mackay, 2011; Ley et al., 2005). Bacteroidetes and

Firmicutes populations were found to be correlated with obesity in both humans and mice (Ley et al., 2005; 2006). In my study, only very weak correlations (R2<0.1) between fat/lean percentages and Bacteroidetes/Firmicutes expressions were found.

The correlations between growth rate and Bacterioidetes and Firmicutes expressions were weak as well (R2<0.2). Future study should further identify the bacterial groups present using DNA sequencing and testing the health impact of breadfruit using an obesity model.

168 6.5. Summary

The current data showed that breadfruit diet did not induce adverse impact on mice overall health and growth. The mice fed on the breadfruit diet, both male and female, had a significantly higher body weight/growth rate than the mice fed on the

5LG4 diet without significant increase in fat or any organ after the three week trial.

The consistency in the morphology and overall similarity of the cytokine responses in mice fed on the breadfruit and 5LG4 diet both the in vitro cell model and in vivo animal study indicate that the breadfruit diet did not induce any adverse effects or inflammatory responses in the ileum of the mice and it had a similar impact to mice ileum as the 5LG4 diet. Colon bacteria gene analysis indicates that BF-fed mice had different bacterial flora than 5LG4-fed mice. Future research is needed to characterize the effects of a breadfruit diet on the microbiome.

169 Chapter 7. Conclusion

The most recent Food and Agriculture Organization (FAO) estimates indicate that about 805 million people in the world were chronically undernourished in 2012-

2014. About 791 million of the undernourished population live in developing countries and 349 million undernourished people are found near the equator (FAO et al., 2012;

2014). Although food insecurity, hunger, and malnutrition are ubiquitous and complex problems, when analyzing individual countries, we can find that the lack of local food production is one of the determining factors. For example, in Haiti, half of the population is undernourished. One third of the calorie intake of the population depends on rice and wheat (FAO et al., 2014); however, about 80% of rice and 100% of wheat were imported from the international food market, making the local food system sensitive to international commodities markets. The food import cost to the government is 50% higher than the total merchandise exports of the country in 2008-

2010. The common staple foods are unaffordable for much of the population and addressing nutritional deficiencies is crucial to health improvement (FAO et al., 2014).

The motivation of this project was to contribute to the development of breadfruit as a sustainable, environmentally-friendly, and high-production agricultural industry in developing countries and tropical areas so that people can gain the ability to feed and support themselves. To date, there have been more than 75,000 trees planted in over

37 countries through this project and the potential of the project for food security has been recognized (Zielinski, 2013). Quality assurance is a prerequisite for the success of breadfruit introduction in impoverished areas. Protein, as one of the most fundamental elements in our diet, plays an important role in human health. The first

170 requirement of a healthy alternative staple crop is the capacity to provide protein in a digestible form (de Onis et al., 1993; FAO, 2013). Evaluation of breadfruit protein quality is necessary for breadfruit product development.

To answer the basic question of how much protein breadfruit can produce, I quantified the growth, development, yield, and seasonality of 24 breadfruit cultivars

(26 trees) established in Kauai, Hawaii, over a seven-year period from 2006–2012.

Individual production profiles were generated for each accessioned cultivar based on major agricultural factors. Across all cultivars of breadfruit (A. altilis), an average of over 200 fruits per year was produced by each tree with an average fruit weight of 1.2 kg. Based on the planting density of 100 trees/ha, this translates to an average projected yield of 13.79 t/ha after 7 years. Hybrids (A. altilis × A. mariannensis) had a higher yield than breadfruit. I found that Ma’afala (commercial cultivar) trees produced fruit within 23 months and in an established orchard, Ma’afala trees will produce 20 t/ha of fruit by 8 years. This will translate to 0.52 t/ha protein yield which is higher than wheat, rice, and potato. The data presented provided a practical forecast for breadfruit agricultural production.

To determine the quantity and quality of breadfruit protein, I evaluated the amino acid profile of 49 important breadfruit cultivars (41 Artocarpus altilis and 8 hybrids of A. altilis × A. mariannensis). While significant differences were found between cultivars, all varieties contained a full spectrum of the essential amino acids and are especially rich in phenylalanine, leucine, isoleucine, and valine. Ma’afala contained significantly higher total essential amino acid content than other varieties and higher-quality protein than staples such as wheat, corn, rice, potato, soybean,

171 and yellow pea.

Proximate analysis showed that breadfruit can provide a high amount of protein/essential amino acid, but the utilization of the breadfruit protein remained unknown. To understand the digestibility of the breadfruit protein and the possible impacts of breadfruit digestion on intestinal epithelial cells, I developed a multiple- stage enzymatic digestion model, mimicking the human digestion process. The protein analysis before and after the digestion showed that protein from the Ma’afala breadfruit flour was more easily digested than protein from the wheat flour.

To quantify the effects of breadfruit digestion on human intestinal health, I applied breadfruit digestion and wheat digestion to Caco-2 cells, and found that breadfruit digestion more positively impacted cell viability than wheat digestion. Both digestions induced similar cytokine responses in Caco-2 cells with some important differences. The higher production of iNOS in breadfruit digestion treated cells under

LPS stimulation could be a possible immune defense against bacterial infection. The induction of IL-6 expression by wheat flour and not by breadfruit flour under IL-1β and/or LPS stimulation provides an early indication that breadfruit flour may be better tolerated.

To quantify whole body response to the consumption of breadfruit flour, an in vivo study of breadfruit was conducted using C57BL/6 mice as a model. The breadfruit diet did not result in any adverse impact on the overall mice growth and it is applicable for both female and male mice. After the three-week trial, mice fed on the breadfruit diet showed a significantly higher body weight/growth rate than mice fed on the standard diet. I also examined the blood samples from the mice. Hematology analysis

172 between the two diet groups showed that the BF-fed mice (breadfruit diet) had similar red blood cell and platelet profiles as the 5LG4-fed mice (standard diet), supporting evidence that the breadfruit diet did not induce any adverse impact on mice physiology. To test the results obtained from the in vitro cell model study, a close examination of mice intestinal health was conducted. Consistent with the in vitro study results, mice fed on the breadfruit diet showed a very similar cytokine response as mice fed on the standard diet. The higher production of iNOS from the breadfruit treated group was consistent between the in vitro cell model and in vivo animal study.

In conclusion, the breadfruit diet did not induce adverse impact on the overall growth of the mice as well as the intestinal health of the mice.

Together, this set of studies represents the most detailed and inclusive assessment of breadfruit protein quality to date and provides fundamental evidence and information for the development of breadfruit products and industries. There are many avenues for future research including cultivar variability, metabolomics, and microbiome studies.

173 References

Abdel-Aal E.-S. M., P.J. Hucl, C.A. Patterson, D. Gray, 2011. Phytochemicals and heavy metals content of hairless canary seed: a variety developed for food use. L\ATT: Food Science and Technology, 44: 904-910.

Adebowale, O.K., Olu-Owolabi, I.B., Olawumi, K.E., & Lawal, S.O. (2004). Functional properties of native, physically and chemically modified breadfruit (Artocarpus altilis) starch. Industrial Crops and Products, 343-351.

Adegunwa, M.O., Adebowale, A.A., Bakare, H.A., & Ovie, S.G. (2014). Compositional characteristics and functional properties of instant plantain-breadfruit flour. International Journal of Food Research, 1-7.

Adeniran, Adekanmi, H., Gbadamosi, S.O., & Omobuwajo, T.O. (2012). Microbiological and physico-chemical characteristics of fufu analogue from breadfruit (Artocarpus altilis F). International Journal of Food Science and Technology, 47, 332-340.

Adepeju, B. A., Gbadamosi, S. O., Omobuwajo, T. O., & Abiodum, O. A. (2014). Effect of breadfruit based complementary food on weight and haematological parameters in albino rats. Food Science and Quality Management, 28, 78-82.

Aka, L. O., Obidike, L. O. Igokwe, I. R., & Ezema, W. S. (2009). The effect of feeding differently prepared breadfruit (Artocarpus altilis) on the hematology, serum biochemistry, live and relative organ weight in albino rats. Nigerian Veterinary Journal, 30, 26-34.

Aktan, F. (2004). iNOS-mediated nitric oxide production and its regulation. Life Science, 75, 639-653.

Amusa, N.A., Kehinde, I.A., & Ashaye, O.A. (2002). Bio-deterioration of breadfruit (Artocarpus communis) in storage and its effects on the nutrient composition. African Journal of Biotechnology, 1, 57-60.

174 Appiah, F., Oduro, I., & Ellis, W.O. (2011). Proximate and mineral composition of Artocarpus altilis pulp flour as affected by fermentation. Pakistan Journal of Nutrition, 10, 653-657.

Aregheore, E.M. (2000). Nutritive value of breadfruit (Artocarpus altilis, Parkinson) and cassava (Manihot dulcis) in concentrate supplements of goats fed a basal diet of batiki grass. Journal of Animal and Feed Science, 9, 615623.

Bahado-Singh, P.S., Wheatley, A.O., Ahmad, M.H., Morrisson, E.Y.St.A., & Asemota, H.N. (2006). Food processing methods influence the glycaemic indices of some commonly eaten West Indian carbohydrate-rich foods. British Journal of Nutrition, 96, 476-481.

Baker, J., Brown, K., Rajendiran, E., Yip, A. DeCoffe, D., Dai, C., Molcan, E., Chittick A., et al. (2012). Medicinal lavender modulates the enteric microbiota to protect against Citrobacter rodentium-induced colitis. American Journal of Physiology and Gastrointest Liver Physiology, 303, G825-G836.

Banks, J., (1962). The endeavour journal of Joseph Banks 1768-1771. Trustees of the Public Library of New South Wales in association with Angus and Robertson, Sydney, Australia.

Barrau, J., (1976). Breadfruit and relatives. In Simmonds N. W. [ed.], Evolution of Crop Plants, 201-202. Longman, London, England. Document: http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInform ation/LabelingN utrition/ucm064928.htm.

Becker, C., Fantini, M.C., Schramm, C., Lehr, A.H., Wirtz, S., Nikolaev, A., et al. (2004a). TGF-β suppresses tumor progression in colon cancer by inhibition of IL-6 trns-signaling. Immunity, 21(4), 491-501.

Becker, C., Fantini, M.C., Wirtz, S., Nikolaev, A., Lehr, H.A., Galle, P.R., John, R., & Neurath, M.F. (2004b). IL-6 signaling promotes tumor growth in colorectal cancer.

175 Cell Cycle, 4(2), 220-223.

Blackburn, N. A., & Southgate, D.A.T. (1981). Protein digestibility and absorption: effects of fiber, and the extent of individual variation. FAO, Rome.

Bowers, R.D. (1981). Breadfruit – a low energy requirement source of carbohydrate for the wet tropics, Entwickl. Laendlicher Raum, 2, 11–13.

Butts, A.C., Monro, A.J., & Moughan, J.P. (2012). In vitro determination of dietary protein and amino acid digestibility for humans. British Journal of Nutrition, 108, S282-S287.

Cario, E., Rosenberg, M.I., Brandwein, L.S., Beck, L.P., Reinecker, C.H., & Podolsky, K.D. (2000). Lipopolysaccharide activates distinct signaling pathways in intestinal epithelial cell lines expressing toll-like receptors. Journal of Immunology, 164, 966-972.

Chai, W., & Liebman, M. (2005). Oxalate content of legumes, nuts, and grain-based flours. Journal of Food Composition and Analysis, 18, 723-729.

Chinwalla, A. T., Waterston, L. L., Lindblad-Toh, K. D., Birney, G. A., Rogers, L. A., Abril, R. S., Agarwal, T. A., et al.(2002). Initial sequencing and comparative analysis of the mouse genome. Nature, 420 (6915), 520–562.

Ciclitira, P.J., & Ellis, J.H. (1987). Investigation of cereal toxicity in coeliac disease. Postgraduate Medical Journal, 63, 767-775. de Mejia, G.E., & Prisecaru, I.V. (2007). Lectins as bioactive plant proteins: a potential in cancer treatment. Critical Reviews in Food Science and Nutrition, 456, 425-445. de Onis, M., Monteiro, C., Akré, J., & Clugston, G. (1993). The worldwide magnitude of protein-energy malnutrition a overview from the WHO global database on child growth. Bulletin of the World Health Organization, 71, 703-712.

176 de Punder, K., & Pruimboom, L. (2013). The dietary intake of wheat and other grains and their role in inflammation. Nutrients, 5(3), 771-787. de Ribou, S.D., Douam, F., Hamant, O., Frohich, M.W., & Negrutiu, J. (2013). Plant science and agricultural productivity: Why are we hitting the yield ceiling? Plant Science, 210, 159-176.

Deshmane, S, S., Kremlev, S., & AminiSawaya, E.B. (2009). Monocyte chemoattractant protein-1 (MCP-1): an overview. Journal of Interferon and Cytokine Research, 39, 313-326.

Dolfini, E., Elli, L., Dasdia, T., Bufardeci, B., Colleoni, M. P., Costa, B., et al. (2002). In vitro cytotoxic effect of bread wheat gliadin on the LoVo human adenocarcinoma cell line. Toxicol In vitro, 16, 331-337.

Eckmann, L, Jung, C.H., Schürer-Maly, C., Panja, A., Morzycka-Wroblewska, E., & Kagnoff, F.M. (1993). Differential cytokine expression by human intestinal cell lines: regulated expression of interleukin 8. Gastroenterology, 105, 1689-1697.

Elli, L., Dolfini, E., & Bardella, T.M. (1987). Gliadin cytotoxicity and in vitro cell cultures. Toxicol Letters, 146, 1-8.

Englberger, L., Aalbersberg, W., Ravi, P., Bonnin, E., Marks, G.C., Fitzgerald, M.H., & Elymore, J. (2003a). Further analyses on Micronesian banana, taro, breadfruit, and other foods for provitamin A carotenoids and minerals. Journal of Food Composition and Analysis, 16, 219-236.

Englberger, L., Schierle, J., Marks, G.C., & Fitzgerald, M.H. (2003b). Micronesian banana, taro, and other foods: newly recognized sources of provitamin A and other carotenoids. Journal of Food Composition and Analysis, 16, 3-19.

Englberger, L., Alfred, J.M., Lorens, A.S., & Luta, T. (2007). Screening selected breadfruit cultivars for carotenoids and related health benefits in Micronesia. Acta Horticulturae, 757, 193-199.

177 Englberger, L., Lorennij, R., Taylor, M., Tuia, V.S., Aalbersberg, W., Dolodolotawake, U., Tibon, L., Tibon, J., & Alfred, J. (2014). Carotenoid content and traditional knowledge of breadfruit cultivars of the Republic of the Marshall Islands. Journal of Food Composition and Analysis, 34, 192-199.

Esuoso, K.O. & Bamiro, F.O. (1995). Studies on the baking properties of non-wheat flours-I. Breadfruit (Artocarpus altilis). International Journal of food Sciences and Nutrition, 46, 267-273.

Famurewa, J. A. V., Esan, Y. O., Pele, G. I., & Arewa, O. A., 2015. Effect of maturity and drying methods on rheological and physic-chemical properties of reconstituted breadfruit (Artocarpus altilis) flour. IOSR Journal of Engineering, 5(2), ISSN (e): 2250-3021, ISSN (p): 2278-8719.

FAO (1981). Amino-acid content of foods and biological data on proteins. Food and Agriculture Organization of the United Nations, Rome.

FAO (2009). The International Treaty on Plant Genetic Resources for Food and Agriculture. Food and Agricultural Organization of the United Nations, Rome.

FAO (2013). Dietary protein quality evaluation in human nutrition. Food and Agriculture Organization of the United Nations, Rome.

FAO, IFAD, & WFP. (2014).The state of food insecurity in the world 2014.

Strengthening the enabling environment for food security and nutrition. Food and

Agriculture Organization of the United Nations, Rome.

FAO, WFP, & IFAD. (2012). The state of food insecurity in the world 2012: Economic growth is necessary but not sufficient to accelerate reduction of hunger and malnutrition. Food and Agriculture Organization of the United Nations, United Nations World Food Programme, and International Fund for Agricultural Development, Rome.

178 FAOSTAT. (2013). Food and agriculture organization of the United Nations statistics division. Online: http://faostat3.fao.org/home/E

Fields, S. & Johnston, M. (2005). Whither model organism research? Science, 307, 1885-1886.

Food and Drug Administration (FDA). (2013). Guidance for industry: A food labeling

Fosberg, F. R., (1960). Introgression in Artocarpus (Moraceae) in Micronesia. Brittonia, 12, 101-113.

Fownes, J.H., & Raynor, W.C. (1993). Seasonality and yield of breadfruit cultivars in the indigenous agroforestry system of Pohnpei, Federated States of Micronesia, Tropical Agriculture (Trinidad), 70, 103–109.

Generoso, V. S., Rodrigues, M. N., Trindade, M. L., Paiva, C. N., Cardoso, N. V., Carneiro, M. C. A., Ferreira, M. A., Faria, C. M. A., & Maioli, U. T. (2015). Dietary supplementation with omega-3 fatty acid attenuates 5-fluorouracil induced mucositis in mice. Lipids in Health and Disease, 14, 54.

Ghosh S, Decoffe D, Brown K, Rajendiran E, Estaki M, et al. (2013) Fish oil attenuates omega-6 polyunsaturated Fatty Acid-induced dysbiosis and infectious colitis but impairs LPS-dephosphorylation activity causing sepsis. PLoS One 8, e55468.

Giovannini, C., Matarrese, P., Scazzocchio, B., Vari, R., D’Archivio, M. Straface, E., et al. (2003), Wheat gliadin induces apoptosis of intestinal cells via an autocrine mechanism involving Fas-Fas ligand pathway. FEBS Letter, 540, 117-124.

Giovannini, C., Sanchez, M., Straface, E., Scazzocchio, B., Silano, M., & de Vincenzi, M. (2000). Induction of apoptosis in Caco-2 cells by wheat gliadin peptides. Toxicology, 145, 63-71.

Golden, D.K. & Williams, J.O. (2001). Amino acid, fatty acid and carbohydrate

179 content of Artocarpus altilis (Breadfruit). Journal of Chromatographic Science, 39, 243-250.

Graham, H.D. & Negron de Bravo, E. (1981). Composition of the breadfruit. Journal of Food Science, 46, 535539.

Grant, G., More, L. J., Mckenzie, N. H., Dorward, P. M., Buchan, W. C., Telek, L., & Pusztai, A. (1995). Nutritional and haemagglutination properties of several tropical seeds. Journal of Agricultural Science, 124, 437-445.

Grivennikov, S., Karin, E., Terzic, J., Muclda, D., Yu, G.Y., Vallabhapurapu, S., et al. (2009). IL-6 and Stat3 are required for survival of intestinal epithelial cells and development of colitis-associated cancer. Cancer Cell, 15, 103-113.

Guide. Online: http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInform ation/LabelingNutrition/ucm064928.Htm

Haydersah, J., Chevallier, I., Rochette, I., Mouquet-River, C., Picq, C., Marianne- Pepin, T., Icard-Verniere,C., & Guvot, J.P. (2012). Fermentation by amylolytic lactic acid bacteria and consequences for starch digestibility of plantain, breadfruit, and sweet potato flours. Journal of Food Science, 77, M466-M472.

Hedges, S. B. (2002). The origin and evolution of model organisms. Nature Reviews Genetics, 3, 11, 838–849.

Hedrich, H. (2012). The laboratory mouse. Edited by Hedrich J. H Academic Press

HHFN & NTBG (2012). Ho’oulu ka ‘Ulu Cookbook: Breadfruit tips, techniques, and Hawaii’s favorite home recipes. Hawaii Homegrown Food Network and National Tropical Botanical Garden, Halualoa.

Hoffman, A. R., Zhang, G., Nussler, C. N., Cleixner, L. S., Ford, R. H., Simmons, L. R., & Watkins, C. S. (1997). Constitutive expression of inducible nitric oxide

180 synthase in the mouse ileal mucosa. American Journal of Physiology, 272, 383-392.

Holloway, W.D., Monro, J.A., Gurnsey, J.C., Pomare, E.W., & Stace, N.H. (1985). Dietary fiber and other constituents of some Tongan foods. The Journal of Food Science, 50, 1756-1757.

Huang, A.S., Titchenal, C.A., & Meilleur, B.A. (2000). Nutrient composition of taro corms and breadfruit. Journal of Food Composition and Analysis, 13, 859-864. International Diabetes Federation (IDF). (2013). International Diabetes Atlas Sixth Edition. ISBN: 2-930229-85-3.

Hur, J. S., Lim, Q. B., Decker, E. A., & McClements J. D. (2011). In vitro human digestion models for food applications. Food Chemistry, 125, 1-12.

Ijarotimi, S.O. & Aroge, F. (2005). Evaluation of the nutritional composition, sensory, and physical properties of a potential weaning food from locally available food materials – breadfruit (Artocarpus altilis) and soybean (Glycine max). Polish Journal of food and Nutrition Sciences, 14, 411-415.

Jackson Laboratory. (2012). Mouse phenome database at the Jackson Laboratory. Online: http://phenome.jax.org/db/q?rtn=meas/catlister&req=Cblood-- hematology&reqstrainid=7

Jenab, M., & Thompsonk, L.U., (2002). Role of phytic acid in cancer and other disease. In: In: Food Phytates. Reddy. N.R and Sathe, S.K. (Eds) Boca Raton, FL CRC Press, pp, 225-248.

Jones, A.M.P., Murch, S.J., & Ragone, D. (2010). Diversity of breadfruit (Artocarpus altilis, Moraceae) seasonality: A resource for year-round nutrition, Economic Botany, 64, 340–351.

Jones, A.M.P., Murch, S.J., Wiseman, J., & Ragone, D. (2013a) Morphological diversity in breadfruit (Artocarpus, Moraceae): Insights into domestication, conservation, and cultivar identification. Genetic Resources and Crop Evolution, 60,

181 175–192.

Jones, A.M.P., Baker, R., Ragone, D., & Murch, S.J. (2013b). Identification of provitamin A carotenoid-rich cultivars of breadfruit (Artocarpus, Moraceae). Journal of Food Composition and Analysis, 31, 51–61.

Jones, A.M.P., Ragone, D., Aiona, K., Lane, A.W., & Murch, S.J. (2011a). Nutritional and morphological diversity of breadfruit (Artocarpus, Moraceae): Identification of elite cultivars for food security. Journal of Food Composition and Analysis, 24, 1091- 1102.

Jones, A.M.P., Ragone, D., Tavana, N.G., Bernotas, D.W., & Murch, S.J. (2011b). Beyond the Bounty: Breadfruit (Artocarpus altilis) for food security and novel foods in the 21st century. Ethnobotany Research and Applications, 9, 129–149.

Jung, C.H., Eckman, L., Yang, S.K., Panja, A., Fierer, J., & Morzycka-Wroblewska, E. (1995). A distinct array of proinflammatory cytokines is expressed in human colon epithelial cells in response to bacterial invasion. Journal of Clinical Investigation, 95, 55-65.

Kiers, L.J., Nout, M.J.R., & Rombouts, M.F. (2000). In vitro digestibility of processed and fermented soya bean, cowpea and maize. Journal of Science and Food Agriculture, 80, 1325-1331.

Kim, M.J., Eckmann, L., Savidge, C.T., Lowe, C.D., Wittöft, T., & Kagnoff, F.M. (1998). Apoptosis of human intestinal epithelial cells after bacterial invasion. Journal of Clinical Investigation, 1815-1823.

Klampfer, L. (2011). Cytokine, inflammation and colon cancer. Current Cancer Drug Target, 11, 451-464.

Knight, I. M., & Chambers, J. P. (2003). Problems associated with determing protein concentration. Molecular Biotechnology, 23, 19-28.

182 Kumar, V., Sinha, A.K., Makkar, H.P.S., & Becker, K. (2010). Dietary roles of phytate and phytase in human nutrition: a review. Food Chemisty, 120, 945-959.

Lahm, H., Petral-Malec, D., Yilmaz-Ceyhan, A., Fischer, J.R., Lorenzoni, M., Givel, J.C., et al. (1992). Growth stimulation of a human colorectal carcinoma cell line by interleukin-1 and -6 and antagonistic effects of transforming growth factor β1. European Journal of Cancer, 11, 1894–1899.

Lancaster, P.A.P., & Coursey, D.G. (1984), Traditional post-harvest technology of perishable tropical staples. Food and Agriculture Organization of the United Nations, Rome

Langerholc, T., Maragkoudakis, P., Wollgast, J., Gradisnik, L., & Cencic, A. (2011). Novel and established intestinal cell line modes-an indispensable tool in food science and nutrition. Trends in Food Science and Technology, 22, S11-S20.

Lans, C.A., (2006). Enthnomedicines used in Trinidad and Tobago for urinary problems and diabetes mellitus. Journal of Ethnobiology and Ethnomedicine, 13, 2- 45

Latham. C.M. (1997). Protein-energy malnutrition, in Human Nutrition in the Developing World. Rome: Food and Agriculture Organization of United States.

Lebegin, S., Lemerre, Desprez, Z., & MadembaSy, F. (2007). Horticultural evaluation of five introduced and one local breadfruit cultivar in New Caledonia. Acta Horticulturae, 757, 89–92.

Lee, D.F. (1993). Importance of apoptosis in the histopathology of drug related lesions in the large intestine. Journal of Clinic Pathology, 46: 118-122.

Leterme, P., Buldgen, A., Estrada, F., & Londono, A.M. (2005). Mineral content of tropical fruits and unconventional foods of the Andes and the rain forest of Columbia. Food Chemistry, 95, 644-652.

183 Ley, E. R., Bäckhed, F., Turnbaugh, P., Lozupone, A. C., Knight, R. D., & Gordon, I. J. (2005). Obesity alters gut microbial ecology. Proceedings of the National Academy of Sciencs of the United States of America, 102, 11070-11075.

Ley, E. R., Turnbaugh, J. P., Klein S., & Gordon I. J. (2006). Microbial ecology: Human gut microbes associated with obesity. Nature, 444, 1022-1023.

Lied, A.G., LillestØl, K., Lind, R., Caleur, J., Morken, H.M., Vaali, K., et al. (2011) Perceived food hypersensitivity: a review of ten years of interdisciplinary research at a reference center. Scand J Gastroentero, 46, 1169-1178.

Lim, T.K. (2012). Artocarpus altilis. in Edible Medicinal and Non Medicinal Plants, Volume 3, Fruits, Springer Science and Business Media.

Lindfors, K., Rauhavirta, T., Stonman, S., Mäki, M., & Kaukinen, K. (2012). In vitro models for gluten toxicity: relevance for celiac disease pathogenesis and development of novel treatment options. Experimental Biology and Medicine, 237, 119-125.

Lippman, Z.B., & Zamir, D. (2007). Heterosis: revisiting the magic, Trends in Genetics, 23, 60-66.

Loos, J.P., Hood, F.L., & Graham, D.H. (1981). Isolation and characterization of starch from breadfruit. Cereal Chemistry, 282-286.

Mahomoodally, M.F., Ramalingum, N. (2015). An investigation into the consumption patterns, attitude, and perception of Mauritians towards common medicinal food plants. Journal of Herbal Medicine, 5, 99-112.

Malomo, S.A., Eleyinmi, A.F., & Fashakin, J.B. (2011). Chemical composition, rheological properties, and bread making potentials of composite flours from breadfruit, breadnut, and wheat. African Journal of Food Science, 5, 400-410.

Maslowski, M. K. & Mackay R. C. (2011). Diet, gut microbiota and immune

184 responses. Nature Immunology, 12, 1.

Mayaki, O.M., Akingbala, J.O., Baccus-Taylor, G.S.H., & Thomas, S. (2003). Evaluation of breadfruit (Artocarpus communis) in traditional stiff porridge foods. Food, Agriculture & Environment, 1, 54-59.

McClearly, B.V., DeVries, J.W., Rader, J.I., Cohen, G., Prosky, L., Mugford, D.C., Champ, M., & Okuma, K. (2012). Determination of insoluble, soluble, and total dietary fiber (CODEX Definition) by enzymatic-gravimetric method and liquid chromatography: collaborative study. Journal of the AOAC International, 95, 824- 844.

Meilleur, B.A., Jones, R.R., Titchenal, C.A., & Huang, A.S. (2004). Hawaiian breadfruit: Ethnobotany, nutrition, and human ecology, Univ. Hawaii Press, Honolulu, U.S.A.

Moore, C. J., DeVries, W. J., Lipp, M. Griffths, C. J., & Abernethy, R. D. (2010). Total protein methods and their potential utility to reduce the risk of food protein adulteration. Comprehensive Reviews in Food Sciences and Food Safety, 9, 330- 357.

Morton, J.F. (1987) Breadfruit. In: Morton JF (ed) Fruits of Warm Climates. Miami, USA, pp 50–58

Mudgil, D. & Barak, S. (2013). Composition, properties and health benefits of indigestible carbohydrate polymers as dietary fiber: A review. International Journal of Biological Macromolecules, 61, 1-6.

Murch, S.J., Ragone, D., Shi, W.L., Alan, A.R., & Saxena, P.K. (2008). In vitro conservation and sustained production of breadfruit (Artocarpus altilis, Moraceae): modern technologies for a traditional tropical crop, Naturwiss. 95 99–107.

Nelson-Quartey, F.C., Oduro, N.I., Ellis, W.O., & Amagloh, F.K. (2007). Formulation of an infant food based on breadfruit (Artocarpus altilis) and breadnut (Artocarpus

185 camansi). Acta Horticulturae, 757, 215-224.

Nemzek, A. J., Bolgos, L. G., Williams, B. A., & Remick, G. D. (2001). Differences in normal values for murine white blood cell counts and other hematological parameters based on sampling site. Inflammation Research, 50, 523-527.

Niering, W. A. (1963). Terrestrial ecology of Kapingamarangi Atoll, Caroline Islands. Ecological Monographs, 33, 131-160.

Nochera, C., & Caldwell, M. (1992). Nutritional evaluation of breadfruit-containing composite flour products. Journal of Food Science, 57, 1420-1422.

Nwokocha, M.L., & Williams, A. P. (2010). Comparative study of physicochemical properties of breadfruit (Artocarpus altilis) and white yam starches. Carbohydrate Polymers, 294-302.

O’connell, T.C., & Hedges, T.E.M. (1999). Investigations into the effect of diet on modern human hair isotopic values. American Journal of Physical Anthropology, 108, 409-425.

Oduro, I., Ellis, W.O. & Narh, S.T. (2007b). Expanding breadfruit utilization and its potential for pasta production.

Oduro, I., Ellis, W.O., Sulemana, A., & Oti-Boateng, P. (2007a). Breakfast meal from breadfruit and soybean composite. The African Forest Research Network: A Special Edition on Ghana Workshop. Academy Science Publishers, vol 19., S238-S242.

Oladunjoye, I.O., & Ojebiyi, O.O. (2011). Feeding value of processed breadfruit (Artocarpus altilis Fosberg) meal for grower rabbits. African Journal of Basic & Applied Sciences, 3, 98-104.

Oladunjoye, I.O., Ologhobo, A.D. & Olaniy, C.O. (2010). Nutrient composition, energy value and residual antinutritional factors in differently processed breadfruit (Artocarpus altilis) meal. African Journal of Biotechnology, 9, 4259-4263.

186 Oso, A.O., Faboro, R., Isah, O., Oni, A., Bamgbose, A., & Dele, P. (2010). Potential of breadfruit (Artocarpus altilis) an ecologically forest based feed resource in rabbit nutrition. Tropical and Subtropical Agroecosystems, 12, 99-108.

Oulaï, F. S., Koné, F. M. T., Amedée, A. P., Gonnety, J. T., Faulet, B. M., & Kouamé, L. P. (2014). Impact of cooking times on some nutritional and antinutritional factors of Ivorian breadfruit (Artocarpus altilis) flour. International Journal of Recent Biotechnology, 2(3), 34-46.

Oulaï, S.F., Gonnety, J.T., Faulet, B.M., Dje, K.M., Kouassi-Koffi, A.P.A., & Kouame, L.P. (2013). Effect of cooking time on the proximate and mineral composition of breadfruit (Artocarpus altilis) grown in Abidjan, Cote D‟ivoire. Pakistan Journal of Nutrition, 12, 768-774.

Pickering, M.V., & Newton, P. (1990). Amino acid hydrolysis-old problems, new solutions. LC*GC-Magazine of Separation Science, 8, 778.

Porter, J.R., & Xie, L. (2014). Chapter 7: Food security and production systems. In: Climate change 2014: impacts, adaptation and vulnerabilities. Intergovernmental panel on climate change, 5th Assessment Report. WHO/UNEP. http://www.ipcc.ch/report/ar5/ wg2/

Powell, D. (1977). Voyage of the plant nursery, H.M.S Providence. Economic Botany 31, 387–431.

Quartermain A. (2007). Breadfruit in Papua New Guinea, Acta Horticulturae. 757, 109–113.

Ragone, D. & Cavaletto, C.G. (2006). Sensory evaluation of fruit quality and nutritional composition of 20 breadfruit (Artocarpus, Moraceae) cultivars. Economic Botany, 60, 335-346.

Ragone, D. (1997). Breadfruit, Artocarpus altilis (Parkinson) Fosberg. International

187 Plant Genetic Resources Institute, Rome.

Ragone, D. (2001). Chromosome numbers and pollen stainability of three species of Pacific Island breadfruit (Artocarpus, Moraceae). American Journal of Botany, 88, 693-696.

Ragone, D. (2006). Artocarpus altilis (breadfruit). In Elevitch C. R. [ed.], Traditional Trees of Pacific Islands: Their Culture Environment and Use, 85-100. Permanent Agriculture Resources, Holualoa, USA.

Ragone, D. (2011). Farm and forestry production and marketing profile for breadfruit (Artocarpus altilis), in: Elevitch C.R. (Ed.), Specialty Crops for Pacific Island Agroforestry, Perm. Agric. Res. (PAR), Hulualoa, U.S.A.

Ragone, D., & Raynor, B. (2009). Breadfruit and its traditional cultivation and use on Pohnpei. In: Balick MB (ed) Ethnobotany of Pohnpei: Plants, people, and island culture. University of Hawai’i Press and New York Botanical Garden Press, Hulualoa, pp, 63–88.

Raiola, A., Meca, G., Mãnes, J., & Ritieni, A. (2012). Bioaccessiblity of deoxynivalenol and its natural co-occurrence with ochratoxin A and aflatoxin B in Italian commercial pasta. Food and Chemiscal Toxiology, 50, 280-287.

Ravindran, V. & Sivakanesan, R. (2006). Breadfruit (Artocarpus communis) meal: Nutrient composition and feeding value for broilers. Journal of the Science of Food and Agriculture, 69, 379-383.

Reeve, R.M. (1974). Histological structure and commercial dehydration potential of breadfruit. Economic Botany, 28, 82-96.

Roberts-Nkrumah, B.L. (2005). Fruit and seed yields in chataigne (Artocarpus camansi Blanco) in Trinidad and Tobago, Fruits, 60, 387-393.

Robinson, T.L. (2007). Recent advances and future directions in orchard planting

188 systems. Acta Horticulturea. 732, 367-381.

Rose-John, S., Scheller, J., Elson, G., & Jones, A.S. (2006). Interleukin-6 biology is coordinated by membrane-bound and solube receptors: role in inflammation and cancer. Journal of Leukocyte Biology, 80, 227-236.

Samuda, P.M., Bushway, A.A., Beecher, G.R., Cook, R.A., Work, R., Cook, C.M., & Bushway, R.J. (1998). Nutrient content of five commonly consumed Jamaican foods, 11, 262-273.

Sane, F., Guillermin, P., Mauget, J.C., & Delaire, M. (2012). Effects of fruit load and intra-inflorescence competition of fruits on apple growth during fruit development, Acta Horticulturea, 932, 179-186.

Savidge, C.T, Newman, G.P., Pan, H.W., Weng, Q.M., Shi, N.H., Mccormick, A.B., Quaroni, A., & Walker, A.W. (2006). Lipopolysaccharide-induced human enterocyte tolerance to cytokine-mediated interleukin-8 production may occur independtly of TLR/MD-2 signaling. Pediatric Research, 59, 89-95.

Scheller, J., Chalaris, A., Schmidt-Arras, D., & Rose-John, S. (2011). The pro- and anti-inflammatory properties of the cytokine interleukin 6. BBA-Mol Cell Research, 1813(5), 878-888.

Schlemmer, U., Frolich, W., Prieto, R., & Grases, F. (2009). Phytate in foods and significance for humans: food sources, intake, processing, bioavailability, protective role and analysis. Molecular Nutrition and Food Research, 53, 5330-S375.

Shi, W.L., Saxena, P.K., Ragone, D., & Murch, S.J. (2007). Mass-propagation and bioreactorbased technologies for germplasm conservation, evaluation and international distribution of breadfruit. Acta Horticulturea, 757, 169–176.

Shimizu M. (2010). Interaction between food substances and the intestinal epithelium. Bioscience, Biotechnology, and Biochemistry, 74, 232-241.

189 Sidhu, J. S., & Kabir, Y. (2007). Functional foods from cereal grains, International Journal of Food Properties, 10, 231-244

Siener, R., Honow, R., Voss, S., Seidler, A., & Hesse, A. (2006). Oxalate content of cereals and cereal products, Journal of Agricultural and Food Chemistry, 54, 3008- 3011.

Simonato, B., Pasini, G., Giannattasio, M., Peruffo, B.D.A., de Lazzari, F., & Curioni, A. (2001). Food allergy to wheat products: the effect of bread baking and in vitro digestion on wheat allergenic proteins. A study with bread dough, crumb, and crust. Journal of Agriculture and Food Chemistry, 49, 5668-5673.

Slavin, J. (2004). Whole Grains and Human Health. Nutrition Research Reviews, 17, 99–110

Slavin, J.L., Martini, M.C., Jacobs, D.R., & Marquart, L., (1999). Plausible Mechanisms for the Protectiveness of Whole Grains. American Journal of Clinical Nutrition, 70, 459S-463S.

Smith, N. J. H., Williams, J. T., Plucknett, D. L., & Talbot, J. P. (1992). Tropical Forests and Their Crops. Comstock Publishing Associates, Ithaca, NY.

Stevens, P. F. (2012). Angiosperm Phylogeny Website. Online: http://www.mobot.org/MOBOT/research/APweb/welcome.html

Suzuki, M., Hisamatsu, T., & Podolsky, K.D. (2003). Gamma interferon augments the intracellular pathway for lipopolysaccharide (LPS) recognition in human intestinal epithelial cells through coordinated up-regulation of LPS uptake and expression of the intracellular toll-like receptor 4-MD-2 complex. Infect Immunology, 71(6), 3503.

Tenore, C.G., Campiglia, P., Ritieni, A., & Novellino, E. (2013). In vitro bioaccessibility, bioavailability and plasma protein interaction of polyphenols from Annurca apple (M. pumila Miller cv Annurca). Food Chemistry, 141, 3519-3524.

190 Thompson, L. U. (1993). Potential health benefits and problems associated with antinutrients in foods. Food Research International, 26, 131-149.

Tumaalii, F. & Wootton, M. (1988). Properties of starches isolated from Western Samoan breadfruit using a traditional method. Starch, 40, S7-S10.

Turi, C.E., & Murch, S.J. (2013). Targeted and untargeted phytochemistry of Ligusticum canbyi: indoleamines, phthalides, antioxidant potential, and use of metabolomics as a hypothesis-generating technique for compound discovery. Planta Medica, 79, 1370–1379.

United States Department of Agriculture (USDA), 2007. 2007 Census of agriculture. Online, http://www.agcensus.usda.gov/

USDA. (2014). USDA national nutrient database for standard reference. Electronic document, http://ndb.nal.usda.gov/.

USDA. (2015a). USDA fact sheets & plant guides. Database: http://plants.usda.gov/java/factSheet

USDA. (2015b). What we eat in America. Online: http://www.ars.usda.gov/SP2UserFiles/Place/80400530/pdf/fped/Table_1_FPED_G EN_0910.pdf

USDA. (2015c). World Agriculture production, Government Printing Office, Washington, US. van der Waaij, A. L., Harmsen, H. J. M., madjipour, M., Kroese, F. G.M., Zwiers, M., van Dullemen, H. M., de Boer, N. K., Welling, G. W., & Jansen, P. L. M. (2005). Bacterial population analysis of human colon and terminal ileum biopsies with 16S rRNA-based fluorescent probes: commensal bacteria live in suspension and have no direct contact with epithelial cells. Inflammatory Bowel Disease, 11, 865-871.

Velioglu, S.Y., Mazza, G., Gao, L., & Oomah, B.D. (1998). Antioxidant activity and

191 total phenolic in selected fruits, vegetables and grain products. Journal of Agriculture and Food Chemistry, 46, 4113-4117.

Waldner, J.M., & Neurath, F.M. (2014). Master regulator of intestinal disease: IL-6 in chronic inflammation and cancer development. Seminars in Immunology, 26, 75-79.

Wang, X., Chen, L. Li, X., Xie, F., Liu, H., & Yu, L. (2011). Thermal and rheological properties of breadfruit starch. Journal of Food Science, 76, E55-E61.

Wenkam, N. S. (1990). Foods of Hawaii and the Pacific Basin fruits and fruit products: raw, processed, and prepared. In Composition (Vol. 4). Honolulu, Hawaii: Hawaii Institute of Tropical and Agriculture and Human Resources. ISSN: 0271e9916.

WHO (2007) Protein and amino acid requirements in human nutrition. World Health Organization, Geneva

WHO (2009) Global health risks: Mortality and burden of disease attributable to selected major risks. World Health Organization, Geneva

Widanagamage, R.D., Ekanayake, S., & Welihinda, J. (2009). Carbohydrate-rich foods: glycaemic indices and the effect of constituent macronutrients. International Journal of Food Sciences and Nutrition, 60, 215-223.

Witthöft, T., Eckmann, L., Kim, J.M., & Kagnoff, M.F. (1998). Enteroinvasive bacteria directly activate expression of iNOS and NO production in human colon epithelial cells. American Journal of Physiology, 275, G564-G571.

Wootton, M., & Tumaalii, F. (1984). Composition of flours from Samoan breadfruit. Journal of Food Science, 49, 1396-1397.

WRCC (West regional Climate Center). (2016). Online: http://www.wrcc.dri.edu/

Zerega, N.J.C., Ragone, D., & Motley, T. (2004). Complex origins of breadfruit

192 (Artocarpus altilis, Moraceae): implications for human migrations in Oceania. American Journal of Botany, 91, 760-766.

Zerega, N.J.C., Ragone, D., & Motley, T. (2005). Systematics and species limits of breadfruit (Artocarpus, Moraceae). Systematic Botany, 30, 603-615.

Zielinski, S. (2013). Botanists spread the gospel that breadfruit can be manna. Science, 342, 303.

193 Appendices

Appendices A: Breadfruit field data sheet in Kauai

194 Appendices B: JR lab reports of the diets used in mice study

1). Breadfruit (BF) diet

195

196

197

198 B) 5LG4 diet (Standard diet):

199

200

201

202