Petunidin Derivatives from Black Goji and Purple Potato As Promising Natural Colorants

Petunidin Derivatives from Black Goji and Purple Potato as Promising Natural Colorants,

and Their Co-pigmentation with Metals and Isoflavones

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy

in the Graduate School of The Ohio State University

Peipei Tang

Graduate Program in Food Science and Technology

The Ohio State University

2018

Dissertation Committee

Dr. M. Mónica Giusti, Advisor

Dr. Luis Rodriguez-Saona

Dr. Christopher Simons

Dr. John Litchfield 1

Copyrighted by

Peipei Tang

2018

Abstract

Color is an important factor in consumer perception and quality of food products.

Although extensively used in food industry, artificial colorants have been increasingly questioned among consumers due to potential health concerns. With current “clean label” trends as well as customer demands, manufactures are seeking alternatives for replacement of synthetic dyes. However, there are limited options for stable natural colors. The first overall objective of this dissertation was to explore new sources of natural colorants. We focused on petunidin-derivatives in black goji and purple potato, as they both are rich anthocyanin sources that could be promising candidates for natural food colors; the second overall objective was to explore co-pigmentation between the petunidin-derivatives and metal ions or soybean isoflavones as methods for anthocyanin color enhancement and stabilization.

Currently the most prevalent anthocyanin-based pigments were cyanidin and pelargonidin derivatives (from red cabbage, black carrot, purple sweet potato, and red radish, among others). Focusing on these anthocyanidins would limited the innovation of natural colorant application. Therefore, we started from investigating the pigments in black goji, as it was reported to contain abundant petunidin-derivatives. The black goji extracts produced various vivid hues over wide ranges of pH, with red, purple, and blue colors in acidic, neutral and alkaline conditions, respectively. Cis and trans isomeric

ii petunidin-3-p-coumaroyl-rutinoside-5-glucoside were the major pigments. The colorimetric and spectrophotometric traits of black goji anthocyanins were significantly impacted by purity, pH, acylation, and acyl moiety spatial configuration.

The cis and trans isomeric petunidin-3-p-coumaroyl-rutinoside-5-glucoside displayed different color properties. They could be served in food products depending on diverse color demands. However, only 12 cis acylated anthocyanins were reported in reviewing the current literatures. We then explored the controlled conversion of trans to cis petunidin-derivative by UV-irradiation, aiming to provide an efficient method to obtain cis isomeric anthocyanins. A dose-dependent trans to cis isomerization in petunidin-3-p-cou-rut-5-glu was observed, with trans : cis ratio being dependent on container materials and irradiation energy received.

Next, we studied pigments in purple potato since petunidin-derivatives were also reported in it previously. The anthocyanin profile was similar but simpler compared to that of black goji. The extracts exhibited analogous color properties as black goji, but with higher color saturation.

We further investigated co-pigmentation between petunidin-derivatives and metal ions or isoflavones as methods to enhance and stabilize petunidin-derivatives.

Anthocyanin color and stability were demonstrated to be influenced by metal complexation. Although various types of anthocyanin-Mn+ complexes exhibiting blue colors have been studied, most of these studies mainly focused on cyanidin and delphinidin derivatives at acidic or mildly acidic conditions (pH ≤ 6). We studied the chelation between petunidin-derivatives (extracted from black goji and purple potato) and

iii

Fe3+ or Al3+ at pH 7-9. The petunidin-derivatives experienced varied bluing effects and enhanced stabilities, depending on pH, metal source, and anthocyanin-metal molar ratios.

Isoflavonoids from red clover (Trifolium pratense) was reported to enhance overall color and stability of anthocyanins in muscatine grape juice and wine through intermolecular co-pigmentation. However, the underlying anthocyanin-isoflavone co- pigmentation characteristics remained to be explored. Soybean isoflavones, being more common in food than red clover isoflavones, could be potential co-pigments for anthocyanins. We demonstrated the co-pigmentation between anthocyanin and soybean isoflavones at pH 3 and 7, enhancing and stabilizing the color of acylated anthocyanins.

This study showed petunidin-derivatives from black goji and purple potato as promising sources for natural colorants, producing various vivid hues over a wide range of pH. The co-pigmentation with metal ions and soybean isoflavones could provide food industry with various vivid violet, blue, and green colors with enhanced stability.

Dedication

This dissertation is dedicated to my family.

Acknowledgments

I would like to express my great gratitude and appreciation to my sincere advisor,

Dr. M. Mónica Giusti, for offering me the great opportunity to explore the fantastic anthocyanin world in her lab. For the past five years, she has taught me a lot, and I am definitely motivated by her amazing personality, incredible knowledge and expertise in this field. What I have obtained during this invaluable five years are not only the analytical food chemistry experience, but also, most importantly, the way to enjoy science and research life.

I would also like to thank my committee members, Dr. Luis Rodriguez-Saona

Dr. Christopher Simons, and Dr. John H. Litchfield, for their support, suggestion, and instruction from every aspect regarding the experiment design, data analysis, and dissertation writing.

Many thanks to my fellow lab mates and department peers in the past five years, for their considerate help and kindness. Special gratitude would be given to Neda

Ahamadiani, Gregory Sigurdson, Fei Lao, Jacob Farr, and Kevin Wong.

I am eternally grateful to my parents, Bowei Tang and Xuexia Mao, as well as my dearest wife, Qianying Yao, for always standing by my side, offering me assistance and mental support whenever I needed them the most throughout my Ph.D. program.

Thanks to the Ohio Soybean Council for providing research funding for the anthocyanin-isoflavone co-pigmentation projects.

vii

Vita

Born on April 08, 1989 ...... Wuxi, Jiangsu Province, China

2008 to 2012 ...... B.S., Biotechnology,

Fudan University, Shanghai China

2012 to 2013 ...... Ph.D. Program, Cell, Molecular

Developmental Biology, and Biophysics

Program, The Johns Hopkins University

2013 to present ...... Ph.D., Food Science and Technology, The

Ohio State University

Fields of Study

Major Field: Food Science and Technology

Minor Field: Statistical Data Analysis

Publications

Tang, P. & Giusti, M.M., 2018. Black goji as a Potential Source of Natural Color in a

Wide pH Range. Submitted to Food Chemistry

viii

Tang, P. & Giusti, M.M., 2018. Metal Chelates of Petunidin-derivatives Exhibit

Enhanced Color and Stability. In preparation for Journal of Agricultural and Food

Chemistry

Sigurdson, G.T., Tang, P., and Giusti, M.M., 2018. Cis–Trans Configuration of

Coumaric Acid Acylation Affects the Spectral and Colorimetric Properties of

Anthocyanins. Molecules, 23(3), 598.

Sigurdson, G.T., Tang, P. & Giusti, M.M., 2017. Natural Colorants: Food Colorants from Natural Sources. Annual Review of Food Science and Technology, 8(1), pp.261–

280.

Giusti, M.M., Ahmadiani, N., Tang, P. and Ottinger, M.A., 2015. Isoflavone and flavonoid supplemented eggs in health. In Handbook of eggs in human function (pp. 334-

364). Wageningen Academic Publishers.

Shi, H.X., Liu, X., Wang, Q., Tang, P., Liu, X.Y., Shan, Y.F. and Wang, C., 2011.

Mitochondrial ubiquitin ligase MARCH5 promotes TLR7 signaling by attenuating

TANK action. PLoS pathogens, 7(5), p.e1002057.

Table of Contents

Abstract ...... ii

Dedication ...... v

Acknowledgments...... vi

Vita ...... viii

Table of Contents ...... x

List of Tables ...... xviii

List of Figures ...... xix

Chapter 1 Introduction ...... 1

Chapter 2 Natural Colorants: Food Colorants from Natural Sources ...... 9

2.1 Natural Colorant...... 9

2.2 Classification by Source ...... 12

2.2.1 Plant ...... 13

2.2.2 Animal...... 13

2.2.3 Microbial ...... 14

2.2.4 Mineral ...... 14

2.3 Classification by Chemical Structure...... 15

2.3.1 Flavonoid Derivatives: Anthocyanins...... 15

2.3.2 Isoprenoid Derivatives: Carotenoids ...... 17

2.3.3 Pyrrole Derivatives: Chlorophylls ...... 18

2.3.4 Nitrogen-Heterocyclic Derivatives: Betalains ...... 19

2.4 Classification by Hue and Application ...... 20

2.4.1 Red ...... 20

2.4.2 Orange-Yellow ...... 23

2.4.3 Purple ...... 24

2.4.4 Blue ...... 25

2.4.5 Green ...... 28

2.4.6 Brown ...... 30

2.4.7 Black ...... 32

2.4.8 White ...... 33

Chapter 3 Isoflavone in Human Health ...... 38

3.1 Flavonoids ...... 38

3.2 Isoflavone ...... 39

3.2.1 Isoflavone structure ...... 40

3.2.2 Isoflavone in Foods ...... 43

3.3 Isoflavone and Health ...... 44

3.3.1 Isoflavones Bioavailability and Metabolism ...... 44

3.3.2 Isoflavones as Phytoestrogens ...... 46

3.3.3 Isoflavones and Women’s Health ...... 47

3.3.4 Isoflavones and Breast Cancer ...... 48

3.3.5 Isoflavones and Osteoporosis ...... 49

3.3.6 Isoflavones and Cardiovascular Disease ...... 50

3.3.7 Isoflavones antioxidant activity ...... 51

3.4 Isoflavone and Safety Concerns...... 52

3.5 Analysis of Isoflavones ...... 54

3.5.1 Isoflavone Preparation and Extraction ...... 54

3.5.2 Isoflavone Separation, Identification and Quantification ...... 56

Chapter 4 Black Goji as a Potential Source of Natural Color in a Wide pH Range ...... 58

4.1 Abstract ...... 58

4.2 Introduction ...... 59

4.3 Materials and Methods ...... 62

4.3.1 Materials & Reagent ...... 62

4.3.2 Pigments Extraction ...... 62

xii

4.3.3 Pigments Purification ...... 63

4.3.4 Pigments Saponification ...... 64

4.3.5 Pigments Identification and Isolation ...... 64

4.3.6 Pigments Quantification...... 66

4.3.7 Buffer System and Sample Preparation ...... 66

4.3.8 Spectrophotometric Analysis ...... 67

4.3.9 Colorimetric Analysis ...... 67

4.3.10 Statistical Analysis ...... 67

4.4 Results and Discussion ...... 68

4.4.1 Identification of Anthocyanin Profiles in Black Goji ...... 68

4.4.2 Effect of Purification on Anthocyanin Profiles in Black Goji ...... 72

4.4.3 Colorimetric Properties of Black Goji Anthocyanin Extracts and Isolates ..... 73

4.4.4 Spectrophotometric Properties of Black Goji Extracts and Isolates ...... 79

4.4.5 Comparison of Color Stabilities among the Black Goji Anthocyanin Extracts

and Isolates...... 82

4.5 Conclusion ...... 86

Chapter 5 UV-Induced trans-to-cis Isomerization of Coumaric Acid in Petunidin- derivative...... 87

5.1 Abstract ...... 87

xiii

5.2. Introduction ...... 88

5.3. Materials and Methods ...... 90

5.3.1 Materials & Reagent ...... 90

5.3.2 Pigments Extraction ...... 90

5.3.3 Pigments Purification ...... 91

5.3.4 Pigments Isolation and Identification ...... 92

5.3.5 Pigments Quantification...... 93

5.3.6 Buffer System and Sample Preparation ...... 93

5.3.7 UV Light Exposure ...... 93

5.3.8 Spectrophotometric Analysis ...... 94

5.3.9 Colorimetric Analysis ...... 95

5.3.10 Statistical Analysis ...... 95

5.4. Results and Discussion ...... 95

5.4.1 Influence of UV Irradiation on the Petunidin-Derivative ...... 95

5.4.2 Influence of UV Irradiation on Colorimetric Properties of Petunidin-derivative

...... 99

5.4.3 Influence of UV Irradiation on Spectrophotometric Properties of Petunidin-

derivative...... 100

5.5. Conclusion ...... 107

xiv

Chapter 6 Metal Chelates of Petunidin-derivatives Exhibit Enhanced Color and Stability

...... 108

6.1 Abstract ...... 108

6.2. Introduction ...... 109

6.3 Materials and Methods ...... 111

6.3.1 Materials & Reagent ...... 111

6.3.2 Pigments Extraction ...... 112

6.3.3 Pigments Purification ...... 113

6.3.4 Pigments Identification ...... 113

6.3.5 Pigments Quantification...... 114

6.3.6 Buffer System and Sample Preparation ...... 114

6.3.7 Spectrophotometric Analysis ...... 115

6.3.8 Colorimetric Analysis ...... 116

6.3.9 Statistical Analysis ...... 116

6.4 Results and Discussion ...... 116

6.4.1 Anthocyanin Profiles in Black Goji and Purple Potato ...... 116

6.4.2 Spectrophotometric and Colorimetric Properties of the Purple Potato and Black

Goji Extracts ...... 118

6.4.3 Spectrophotometric Properties of Petunidin-derivatives Chelated with Metals

...... 120

6.4.4 Colorimetric Properties of Petunidin-derivatives Chelated with Metals ...... 126

6.4.5 Stability of Petunidin-derivatives with or without Metal Chelation ...... 132

6.5. Conclusion ...... 137

Chapter 7 Stabilization and Color Enhancement of Anthocyanin by Soybean Isoflavones

...... 139

7.1 Abstract ...... 139

7.2 Introduction ...... 140

7.3 Materials and Methods ...... 142

7.3.1 Materials and Reagents ...... 142

7.3.2 Anthocyanin and Isoflavone Extraction ...... 143

7.3.3 Pigment and Isoflavone Purification ...... 144

7.3.4 Pigments Saponification ...... 145

7.3.5 Pigment Quantification ...... 146

7.3.6 Pigment and Isoflavone Identification ...... 146

7.3.7 Isoflavone Quantification ...... 147

7.3.8 Sample Preparation ...... 147

xvi

7.3.9 Investigation on the Colorimetric, Spectrophotometric properties, and Stability

of Anthocyanins ...... 148

7.4 Results and Discussion ...... 148

7.4.1 Anthocyanin Profiles in Red Cabbage, Red Radish, Black Goji, and Chinese

Eggplant ...... 149

7.4.2 Isoflavone Profiles in SCR...... 151

7.4.3 Spectral Characteristics of Anthocyanins Co-Pigmented with Soybean

Isoflavones ...... 152

7.4.4 Colorimetric Properties of Anthocyanins Co-Pigmented with Soybean

Isoflavones ...... 157

7.4.5 Anthocyanin Color Stabilities When Co-Pigmented with Soybean Isoflavones

...... 160

7.4.6 Anthocyanin Stabilities When Co-Pigmented with Soybean Isoflavones ..... 163

7.5 Conclusion ...... 166

Chapter 8 Overall Conclusion ...... 168

Bibliography ...... 172

xvii

List of Tables

Table 2.1 Major colorants from natural sources (or nature identical): chemical classification sources, colors and regulatory status in the USA ...... 37

Table 3.1 Mean value of total isoflavones concentrations in various food products in a wet weight basis...... 45

Table 3.2 Molecular weight and maximum UV absorption wavelength of selected isoflavones ...... 57

Table 4.1 Colorimetric (CIE-L*, C*ab, hab) and spectrophotometric (λmax) data of black goji anthocyanin extracts and isolates. Values presented are means (n=3) and (SD)...... 77

Table 5.1 Colorimetric data of petunidin-3-rutinoside-(trans-p-coumaroyl)-5-glucoside over UV irradiation in quartz cuvettes...... 103

Table 5.2 Colorimetric data of petunidin-3-rutinoside-(trans-p-coumaroyl)-5-glucoside over UV irradiation in glass cuvettes...... 104

Table 5.3 Spectrophotometric properties of petunidin-3-rutinoside-(trans-p-coumaroyl)-

5-glucoside over UV irradiation in quartz cuvettes...... 105

Table 6.1 Colorimetric (CIE-L*, C*ab, hab) and spectrophotometric (λmax) data of purple potato and black goji anthocyanin extracts over pH 3-10...... 120

xviii

List of Figures

Figure 1.1 Representative chemical structures of anthocyanins ...... 2

Figure 1.2 Intermolecualr co-pigmentation structure illustration ...... 6

Figure 2.1 Chemical structures of selected natural pigments with potential food use, organized according to their chemical classification ...... 35

Figure 2.2 Representative food colorants from natural sources, organized according to their hue ...... 36

Figure 3.1 Structure comparison of different flavonoids ...... 41

Figure 3.2 Chemical structures of soy isoflavones, categorized into four forms: aglycones, glucosides, acetylglucosides and malonylglucosides ...... 42

Figure 3.3 Chemical structure of equol ...... 42

Figure 3.4 The structure relationship between 17β-estradiol and Genistein (left) and the crystal structure of the estradiol-human estrogen receptor-α ligand binding domain complex (right) as proposed by Tanenbaum et al. (1998)...... 47

Figure 4.1 Chromatogram of black goji anthocyanin extracts and isolates at 520 nm and

280-700nm, and their identifications ...... 71

Figure 4.2 Color expression of black goji anthocyanin extracts and isolates at pH 3-10 76

Figure 4.3 Spectrophotometric characteristics of black goji anthocyanin extracts and isolates at pH 3-10...... 80

xix

Figure 4.4 Color changes of black goji extracts and isolates that were stored under refrigerated condition in dark within three weeks testing period...... 84

Figure 4.5 Color changes (described as ΔE) of black goji extracts and isolates at pH

3,4,7,8,9, and 10. Samples were stored under refrigerated condition in dark within three weeks testing period ...... 85

Figure 5.1 HPLC chromatograms of petunidin-3-rutinoside-(trans-p-coumaroyl)-5- glucoside sample over different times of exposure to UV irradiation, expressed as total energy received by the cuvette. Peak 1: petunidin-3-rutinoside-(cis-p-coumaroyl)-5- glucoside; Peak 2: petunidin-3-rutinoside-(trans-p-coumaroyl)-5-glucoside ...... 97

Figure 5.2 Ratio of cis and trans isomers of petunidin-3-rutinoside-(p-coumaroyl)-5- glucoside and total area under the curve (AUC) of pigments in chromatogram over UV irradiation ...... 98

Figure 5.3 Colorimetric properties of petunidin-3-rutinoside-(trans-p-coumaroyl)-5- glucoside over UV irradiation in quartz cuvette ...... 101

Figure 5.4 Colorimetric properties of petunidin-3-rutinoside-(trans-p-coumaroyl)-5- glucoside over UV irradiation in glass cuvette ...... 102

Figure 5.5 Visible absorbance spectra of petunidin-3-rutinoside-(trans-p-coumaroyl)-5- glucoside over UV irradiation in quartz cuvette ...... 106

Figure 6.1 Chromatograms of black goji and purple potato anthocyanin extracts at 520 nm and 280-700nm, and their identifications...... 118

Figure 6.2 Visible light spectra of purple potato anthocyanin metal chelates at alkaline pH...... 122

Figure 6.3 Visible light spectra of black goji anthocyanin metal chelates at alkaline pH.

...... 123

Figure 6.4 Quantification of spectrophotometric changes in purple potato anthocyanin metal chelates ...... 124

Figure 6.5 Quantification of spectrophotometric changes in black goji anthocyanin metal chelates...... 125

Figure 6.6 Colorimetric changes of the purple potato anthocyanin extracts chelated with metal ions at pH 7-9 after 60 min equilibrium. Color changes (ΔE) were calculated based on the samples without metal chelation at certain pH level...... 128

Figure 6.7 Colorimetric changes of the black goji anthocyanin extracts chelated with metal ions at pH 7-9 after 60 min equilibrium. Color changes (ΔE) were calculated based on the samples without metal chelation at certain pH level...... 129

Figure 6.8 Color data of purple potato anthocyanin metal chelates expressed in CIE-LAB color space...... 130

Figure 6.9 Color data of black goji anthocyanin metal chelates expressed in CIE-LAB color space...... 130

Figure 6.10 Color changes (ΔE) of purple potato anthocyanin metal chelates over 28 days in dark refrigerated condition...... 134

Figure 6.11 Color changes (ΔE) of black goji anthocyanin metal chelates over 28 days in dark refrigerated condition...... 135

Figure 6.12 Color stability of purple potato anthocyanin metal chelates over 28 days in dark refrigerated condition ...... 136

xxi

Figure 7.1 HPLC Chromatograms of anthocyanin profiles in red radish, red cabbage,

Chinese eggplant, and black goji...... 150

Figure 7.2 Identification of isoflavone profiles in soybean curd residue by HPLC chromatogram...... 152

Figure 7.3 Spectrophotometric data of acylated anthocyanins co-pigmented with soybean isoflavones ...... 155

Figure 7.4 Spectrophotometric data of non-acylated anthocyanins co-pigmented with soybean isoflavones ...... 156

Figure 7.5 Color expression of anthocyanins when co-pigmented with soybean isoflavone extracts...... 158

Figure 7.6 Colorimetric properties of acylated anthocyanins co-pigmented with soybean isoflavones at pH 3 and 7 ...... 159

Figure 7.7 Color changes (ΔE) of acylated anthocyanins co-pigmented with soybean isoflavones at pH 3 and 7 over 31 days storage test (4°C, dark)...... 161

Figure 7.8 Color changes (ΔE) of non-acylated anthocyanins co-pigmented with soybean isoflavones at pH 3 and 7 over 31 days storage test (4°C, dark)...... 162

Figure 7.9 The color stability of acylated anthocyanins co-pigmented with soybean isoflavones at pH 7 over 31 days storage test (4°C, dark)...... 163

Figure 7.10 Monomeric anthocyanin content (uM) of acylated anthocyanins co- pigmented with soybean isoflavones over 31 days storage test (4°C, dark)...... 165

Figure 7.11 Monomeric anthocyanin content (uM) of non-acylated anthocyanins co- pigmented with soybean isoflavones over 31 days storage test (4°C, dark)...... 166

xxii

Chapter 1 Introduction

Color plays a critical role in food quality and consumer perception. During the last few years, there have been increasing health concerns towards the use of synthetic dyes, which are suspected to cause behavior problems in children with attention deficit hyperactivity disorder (ADHD) (Sharma, McKone, & Markow, 2010). Started in 2010, the European Union (EU) demanded the warning labels “May have an adverse effect on activity and attention in children” for food products that contain certain synthetic food colorants (Ronald E. Wrolstad & Culver, 2012). One year later, the U.S. Food and Drug

Administration (FDA) also conducted the review of scientific evidence on artificial colorants and possible association with ADHD in children (FDA, 2011). Although a causal relationship between color additive and ADHD has not been confirmed in this review, the condition of ADHD might be promoted by synthetic colorants. The increasing consumer demands for natural colorants and the current market trend of “clean label”, have driven the food industry to focus on the replacement of artificial colorants with those natural alternatives.

Among various natural colorants, anthocyanins are a group of important water- soluble natural pigments widely distributed in fruits and vegetables. They impart vivid red to blue colors to plants, and have important application in coloring food products

(Obón, Castellar, Alacid, & Fernández-López, 2009; Sigurdson, Tang, & Giusti, 2017). 1

Differing in the substituent patterns on the different positions on the B rings, there are six major anthocyanin aglycones (also known as “anthocyanidins”) commonly identified in nature: cyanidin (Cy), delphinidin (Dp), pelargonidin (Pg), peonidin (Pn), petunidin (Pt), and malvidin (Mv) (Figure 1.1) (He & Giusti, 2010). It undergoes characteristic pH- dependent structure transformations, exerting diverse color appearance. Given the fact of possessing potent antioxidant activity and being associated with health-promoting effects, anthocyanins have become an increasingly prevalent alternative to the current widely-use synthetic food dyes (He & Giusti, 2010b; Sharma et al., 2010).

Figure 1.1 Representative chemical structures of anthocyanins

However, the transition from synthetic dyes to natural ones has encountered multiple hurdles, as there is limited option for natural blue and purple hues, and food scientists have to consider the factors such as cost, abundancy, availability, stability and 2 tinctorial strength of natural colorants (Sigurdson et al., 2017). Generally speaking, naturally derived colorants are less stable compared to the artificial counterparts, and their stabilities are greatly influenced by food processing and storage conditions (Delia B.

Rodriguez-Amaya, 2016). Pigments sourced from vegetable extracts (for example red cabbage, red radish, and red beet) could also impart undesirable aroma and flavor, resulting in difficulty for their direct food application (Sigurdson, Tang, & Giusti, 2017).

Therefore, in order to keep being competitive and sustainable, it is of great interests for food industry to find sources of affordable and stable natural colorants that could cover a wide range of color hues, especially violet-blue.

The first overall objective of this dissertation was to explore new sources of natural colorants, providing the food industry with more options of plant-based pigments.

Currently, the most prevalent anthocyanin-based pigments are cyanidin and pelargonidin derivatives from red cabbage, black carrot, elderberry, purple sweet potato, chokeberry, and red radish, among others. However, focusing on these anthocyanidins would limit our understanding of the key elements in the application of natural colorants, as well as the innovation of application. We aim to discover other candidates in anthocyanin family, broadening natural colorant markets.

It has been reported previously that Lycium ruthenicum Murr. (black goji) and

Solanum tuberosum L. subsp. andigenum (purple potato) contained abundant petunidin- derivatives (M. Monica Giusti, Polit, Ayvaz, Tay, & Manrique, 2014; Zheng et al., 2011).

Black goji fruit from Tibet contained 550-500mg anthocyanin per 100g FW, most of which (>90%) were acylated ones (Zheng et al., 2011). Five major anthocyanins were

3 identified and petunidin derivative accounted for 95% of the anthocyanin content. Total polyphenol content (~1310 mg GA equivalents/100gFW) and strong antioxidant activity

(~1060mg GA equivalents/100g FW) were also reported (Zheng et al., 2011). Purple potato was also found rich in polyphenols, anthocyanins, and phenolic acids (Shiroma-

Kian, Tay, Manrique, Giusti, & Rodriguez-Saona, 2008). It has been reported that it contained up to 150 mg anthocyanin per 100g DW, and petunidin derivative was the predominant (~63%) pigment (M. Monica Giusti et al., 2014). An average of 4.68 g gallic acid equivalent polyphenols per kg DW were discovered in the purple potato, 3-4 folds compared to that in the white flesh cultivar (M. Monica Giusti et al., 2014;

Lachman, Hamouz, Orsák, Pivec, & Dvořák, 2008).

There was no study conducted to investigate the potential use of these petunidin- derivatives as natural colorants, which inspired us to start from the two plants. We would like to elucidate the colorimetric and spectrophotometric properties of the petunidin- derivative pigments.

We will begin with exploring the profile, color characteristics, and stability of the anthocyanins in black goji, as well as the influence of purity, pH, acylation and acyl moiety spatial configuration on the colorimetric and spectrophotometric properties. We will also study the colorimetric and spectrophotometric properties of petunidin-derivative pigments in the purple potato (Chapter 4-6).

The second overall objective of this dissertation was to explore co-pigmentation between anthocyanins and metal ions or soybean isoflavones as methods for anthocyanin color enhancement and stabilization.

Previous studies have demonstrated that the color of anthocyanins can be stabilized and strengthened by co-pigmentation interactions, which is often manifested as a bathochromic effect and hyperchromic effect when a colorless co-pigment is added to an aqueous solution of the pigment (Malien-Aubert et al., 2001; Pacheco-Palencia &

Talcott, 2010; Fossen et al., 2007). Intermolecular interactions occur between pigment and co-pigment through a weak π-π overlap, dipole-dipole interaction, or possible hydrogen bonding (R. Brouillard, Mazza, Saad, Albrecht-Gary, & Cheminat, 1989), resulting in an overlapping arrangement of the two molecules (Figure 1.2). It could prevent the nucleophilic attack on the anthocyanin by water, promoting anthocyanin stability (O. Dangles, Stoeckel, Wigand, & Brouillard, 1992).

It has been demonstrated that isoflavoniods, Formononetin, Biochanin A and prunetin, from red clover (Trifolium pratense) could enhance overall color and stability of anthocyanin in muscatine grape juice (Vitis rotundifolia) and wine through intermolecular co-pigmentation (Talcott, Peele, & Brenes, 2005). However, the underlying anthocyanin-isoflavone co-pigmentation characteristics remained to be explored. Soybean isoflavones, being more common in food than red clover isoflavones, could be potential co-pigments for anthocyanins.

Figure 1.2 Intermolecualr co-pigmentation structure illustration

Anthocyanin color and stability could be influenced by metal complexation

(Cavalcanti, Santos, & Meireles, 2011). It occurs between anthocyanins with at least two free hydroxyl groups on the B ring, and multivalent metal ions (Cortez, Luna-Vital,

Margulis, & Gonzalez de Mejia, 2017; G. T. Sigurdson, Robbins, Collins, & Giusti,

2016; Yoshida, Kitahara, Ito, & Kondo, 2006; Yoshida, Mori, & Kondo, 2009). The metal ion acts in competition with H+ to attach to the catechol or pyrogallol moieties on the B-ring, triggering anthocyanin transformation from red flavylium cation form to purple-blue quinoidal base anion form. This transformed molecule would then stacks with other flavylium cation molecules to form a stable complex, leading to bathochromic shift in the spectrum and bluing effect (Schreiber, Swink, & Godsey, 2010). Although various types of anthocyanin-Mn+ complexes exhibiting blue colors have been reported; most of these studies mainly focused on cyanidin and delphinidin derivatives at acidic or mildly acidic conditions (pH ≤ 6) (Pyysalo & Kuusi, 1973; Schreiber et al., 2010; G. T.

Sigurdson et al., 2016; G. T. Sigurdson, Robbins, Collins, & Giusti, 2017; Gregory T.

Sigurdson & Giusti, 2014; Tachibana, Kimura, & Ohno, 2014). It remains to be explored the metal chelation in petunidin-derivatives and at neutral or alkaline pHs.

We will study the co-pigmentation between petunidin-derivatives (extracted from black goji and purple potato) and metal ions, as well as its influence on the pigments color expression and stability at neutral and alkaline conditions. Next, our goal will be to elucidate the properties in anthocyanins and soybean isoflavones co-pigmentation, especially the impact of anthocyanin structures (aglycone and acylation) on their interaction (Chapter 6 and 7).

What comes next in the remaining dissertation will be two review chapters regarding the background of food colorants from natural sources and isoflavones in human health. The following four research chapters will present main findings in

7 petunidin-derivatives as natural colorants and their interactions with metal ions and isoflavones.

Chapter 2 Natural Colorants: Food Colorants from Natural Sources

Published. Sigurdson, G. T., Tang, P., & Giusti, M. M. (2017). Natural colorants: food colorants from natural sources. Annual Review of Food Science and Technology, 8, 261- 280.

2.1 Natural Colorant

Color is an important sensory attribute of foods that often plays an important role in the market success of a product. Color is often used by the consumer as an indicator of a variety of qualities of the food product such as flavor, safety, nutritional value, and more. The coloring of foods is believed to have emerged around 1500 BC (Burrows,

2009). Ancient Egyptian writings describe the coloring of drugs, and the coloring of wine has been described by Romans in 400 BC. Color additives can be used to serve several purposes including standardizing raw ingredient colors, providing color identities to otherwise colorless foods, and accounting for loss during processing or storage.

Historically, food coloring substances were derived from nature; some popular sources included saffron, paprika, turmeric, and various flowers (Burrows, 2009). The development of the first synthetic organic colorant mauveine in 1856 ushered in the discovery of several others (Burrows, 2009). These synthetic colorants gained popularity as food colorants due to their low production cost, high tinctorial strength, and chemical stability. Consumer demand has been a driving force for the replacement of artificial

9 colorants with those derived from natural sources. This conversion can be related to consumer demand for more “natural” products, possible links of synthetic colorants to hyperactivity in children and their allergenicity in sensitive populations (McCann et al.,

2007; Potera, 2010). Although currently used synthetic colorants have long records of safety evaluations and strict regulations, the food industry is seeking alternatives to meet changing market demands and regulatory restrictions.

Food colorant regulations vary throughout the world making universal legality of all pigments difficult to discuss. Many countries from Latin America and Europe have adopted the specifications of the Codex Alimentarius or the Joint FAO (Food and

Agricultural Organization)/WHO (World Health Organization) Expert Committee on

Food Additives (JECFA) (Ronald E. Wrolstad & Culver, 2012). While others regulate food coloring additives independently such as the U.S., Korea, and Japan (Ronald E.

Wrolstad & Culver, 2012).

According the U.S. Food and Drug Administration (FDA), a color additive is

“any material, not exempted under section 201(t) of the act, that is a dye, pigment, or other substance made by a process of synthesis or similar artifice, or extracted, isolated, or otherwise derived, with or without intermediate or final change of identity, from a vegetable, animal, mineral, or other source and that, when added or applied to a food, drug, or cosmetic or to the human body or any part thereof, is capable (alone or through reaction with another substance) of imparting a color thereto” (CFR 2016). Food colorants are regulated apart from other food additives and cannot receive status as

Generally Recognized as Safe (GRAS) according to the FDA and the Code of Federal

Regulations (CFR) title 21, parts 70-74, 80, and 82. Current FDA regulations permit 2 major classifications of food colorants: certified and exempt from certification.

The classification of food colorants as certified or exempt from certification loosely distinguishes whether the FDA views the colorants as synthetic or natural

(Burrows, 2009). Current permitted certified pigments are chemically synthesized, and each batch of the colorant produced must be evaluated for identity and sufficient purity levels (Downham & Collins, 2000). Colorants exempt from certification generally include “natural” pigments, but no legal definition for the term “natural” has been adopted yet, leading to consumer and industrial confusion. Colorants exempt from certification include a variety of pigments derived from plant, animal, or mineral natural sources and also synthesized compounds (nature-identical), despite the thought that colorants exempt from certification are natural.

Despite the many types of pigments derived from natural sources, these pigments typically carry more limitations for coloring foods compared to synthetic colorants. In general, naturally derived pigments are typically more expensive than synthetic counterparts associated with large amounts of raw production materials and higher usage levels (Delia B. Rodriguez-Amaya, 2016; Ronald E. Wrolstad & Culver, 2012). Many naturally sourced colorants can also impart undesirable flavors and aromas to products, such as those from red radishes or red beets. The replacement of synthetic dyes may also be limited by inability of natural pigments to match the color characteristics of synthetic colorants, especially blue and green hues. Most are more sensitive to heat, light, oxygen resulting in color loss or alterations in hue (Lakshmi, 2014; Delia B. Rodriguez-Amaya,

2016; Ronald E. Wrolstad & Culver, 2012). Others may be sensitive to environmental matrix conditions such pH, protein, presence of metal ions, or other organic compounds.

Therefore there is generally no universal answer for replacement of a synthetic dye with a naturally derived option. In this review the sources, chemical natures and color applications of a variety of color substances derived from natural sources and their possible applications for coloring of foods are summarized. In addition, some discussion of materials that are chemically synthesized to replicate nature is included. It is important to note that some of the bright colors observed in nature are not produced by pigments, but to a phenomenon called structural coloration. These colors are the result of an optical effect that results in the reflection, refraction or diffraction of certain wavelengths of light, while other wavelengths are absorbed, resulting on a colorful display. This type of phenomena is responsible for the colors of some bird feathers, butterfly wings and beetle shells. Textural and pattern changes can also produce iridescent effects, as seen in peacock feathers, soap bubbles, films of oil, and mother-of-pearl, because the reflected color depends upon the angle of view. It is noted that structural coloration may be another alternative to the use of synthetic colorants in foods; however, they are not part of the scope of this review.

2.2 Classification by Source

The classification of naturally derived colorants can become very complex due to the wide variety of innate properties of the coloring substances. They can be derived from variety of sources in nature, and therefore, natural colorants also exhibit a wide variety of

12 chemical compositions that affect properties, solubilities, and stabilities differently. These properties impact their potential applications in certain matrices and also play a role in regulations of the different pigments. Some of the more commonly used or more prevalent colorants are discussed in this review and are summarized in Table 2.1.

2.2.1 Plant

Plant derived pigments are the result of biochemical pathways within the organism that result in a variety of organic compounds with unique physicochemical properties. These colored compounds are abundant in nature and play important roles in photosynthetic pathways, attracting pollinators, and protection from predators and solar energy. Plant pigments include a variety of chemical classes that can selectively absorb certain wavelengths of light while reflecting others, such as porphyrins, carotenoids, anthocyanins and betalains which are discussed later.

2.2.2 Animal

Similarly, different animals produce a variety of chemical compounds resulting in unique colorations across the animal kingdom. These pigments also serve a variety of purposes including transportation of blood oxygen, protection from predators or UV radiation, mating, and more (Aberoumand, 2011; Delgado-Vargas, Jiménez, & Paredes-

López, 2000). As previously mentioned, the many colors of animals can be result of physical nanoscale structuring causing reflectance of certain light, like butterfly wings,

13 and by pigments including pterins, purines, anthraquinones, melanins, and others

(Delgado-Vargas et al., 2000; Newsome, Culver, & Breemen, 2014).

2.2.3 Microbial

Pigment production by various microbial organisms has been used to help identify certain species. Bacterial and fungal organisms produce a variety of different types of pigments, like carotenoids and monascus pigments, resulting in color expression of almost all hues (Joshi, Attri, Bala, & Bhushan, 2003). The use of microorganisms to produce colorants may be promising and have large economic potential due to growth conditions and renewability (Joshi et al., 2003).

2.2.4 Mineral

Minerals are generally described as elements or chemical compounds that are crystalline and formed by geological processes (Nickel, 1995). They have long historical use as colorants in foods, cosmetics, and art. Minerals present a variety of hues that depends on their chemical composition and/or physical structuring. Many minerals contain metallic cations, and metallic cations containing d orbital electrons often absorb and reflect visible light (Mason, 2013). As an example, viridian, a chromium oxide mineral is a green pigment that has been used in painting but also in cosmetics as colorants exempt from certification (CFR 2016).

2.3 Classification by Chemical Structure

2.3.1 Flavonoid Derivatives: Anthocyanins

Flavonoids are a group of secondary plant metabolites characterized by a C6C3C6 carbon skeletal backbone, the basic structure may be found in Figure 2.1. Of these polyphenolic compounds, anthocyanins are an important subclass of water-soluble pigments that impart vivid red to blue color to plants (Obón et al., 2009). Six major aglycones (anthocyanidins) are found in commonly consumed fruits and vegetables, differing in degree of hydroxylation and methoxylation. These aglycones are bound to sugars in nature, and may also be further acylated with aromatic or aliphatic acids. Both glycosylation and acylation enhance the anthocyanin stability and explaining why anthocyanidins are rarely found in nature (He & Giusti, 2010c). Anthocyanins compose the largest group of water soluble naturally occurring pigments; more than 700 unique structures have been identified (Andersen & Jordheim, 2014). The application of anthocyanins as natural food pigments has encountered difficulties due to poor stability; the color of anthocyanins is sensitive to light, heat, oxygen, and pH condition, which limits its use in different food products (Eiro & Heinonen, 2002). Thus, food industry is seeking methods to improve anthocyanin stability.

Studies have shown the color of anthocyanins can be stabilized and strengthened by co-pigmentation interactions with colorless co-molecules in solution with the pigment.

Co-pigmentation is manifested by shifts in the wavelength maximum absorption or increases in the intensity of absorbance (Malien-Aubert et al., 2001; Pacheco-Palencia &

Talcott, 2010; Fossen et al., 2007). The mechanism of interaction is thought to be result 15 of hydrophobic interactions or Π-Π stacking (Di Meo, Sancho Garcia, Dangles, &

Trouillas, 2012). Acylation of the anthocyanin is also thought to affect the color expression and increase stability of anthocyanins by means of intramolecular co- pigmentation (Malien-Aubert et al., 2001). Acylated anthocyanins are more common to vegetable and floral systems while non-acylated anthocyanins are predominant in fruits.

The superior stability of acylated anthocyanins explains the application of vegetable- based anthocyanin colorants in the food industry (Wallace & Giusti 2008, Giusti &

Wrolstad 2003).

The sensitivity of anthocyanins to pH allows for interconversion between different red and blue structural forms and the development of purple colors. As pH is increased from acidic conditions, the flavylium cations (usually appear red, pH ≤ 3) become deprotonated, losing color (pH 3-6), and then eventually forming quinonoidal bases (appearing purple-blue, pH ≥ 6). Due to these structural changes as response to pH, anthocyanins typically appear red or in low intensity in the pH environments of many food products.

Additional expansion of the color expression of anthocyanins can also be result of metal ion chelation by the pigment, a mechanism often observed in floral systems. Blue shifts can only be induced on anthocyanins with catechol or pyrogallol moieties on the B- ring (Bayer, Egeter, Fink, Nether, & Wegmann, 1966). Metal ions displace hydrogen ions, inducing conversion of flavylium cations (red) to quinonoidal bases (blue) and then coordinate stacking of the pigment with another anthocyanin molecule (Schreiber et al.,

2010).

2.3.2 Isoprenoid Derivatives: Carotenoids

Carotenoids are an important group of lipid-soluble yellow-orange-red pigments widely distributed in nature, including higher plants, bacteria, fungi, yeast, birds and insects (Tanaka, Sasaki, & Ohmiya, 2008). They are 40-carbon tetraterpenoids constituted by isoprenoid units joined in a head-to-tail pattern. They can be classified into two categories: carotenes, containing only polyunsaturated hydrocarbons and xanthophylls, having oxygen functional groups. The carotenoids can be acyclic, monocyclic, or dicyclic (e.g. lycopene, γ-carotene, lutein, and β-carotene, respectively); examples may be found in Figure 2.1.

The long conjugation systems of carotenoids, having π-electrons delocalized along the polyene chain, are responsible for their absorbance in the visible spectrum, resulting in yellow-orange-red hues (Rivera & Canela-Garayoa, 2012). Larger conjugated systems result in display of redder hues; while short double bond chains, as in phytofluene with five double bonds, show achromaticity. Cyclization also influences the color appearance, as lycopene, β-carotene, and γ-carotene, with same number of double bonds, exhibit red, orange, orange-red color respectively.

Due to their electron-rich, highly unsaturated chemical composition, carotenoids are vulnerable to oxidation and isomerization during food processing and storage (D. B.

Rodriguez-Amaya, 1999). Oxidation is a major cause of carotenoids loss and is influenced by many factors including light, moisture, temperature, peroxides, metals, enzymes, lipids, and antioxidants. There are more than 600 different natural carotenoids, with an estimated yield of 100 million tonnes every year (Riggi, 2010). 17

2.3.3 Pyrrole Derivatives: Chlorophylls

Pyrroles are 5-member ring consisting of 4 carbon atoms and 1 nitrogen atom, and in biological systems, they play many useful roles: forming hydrogen bonds, coordinating metals, and stacking interactions (Walsh, Garneau-Tsodikova, & Howard-Jones, 2006).

They are the building blocks of many heteroaromatic rings and linear polypyrroles. Some of the most studied pyrrole derivatives include heme and chlorophyll pigments.

Tetrapyrrole compounds, such as chlorophyll, have been found to create almost all colors of the visible spectrum (Newsome et al., 2014).

The basic structure of chlorophyll is a porphin ring, a symmetrical cyclic tetrapyrrole, found in nature with a phytol attachment and centralized magnesium ion,

Figure 2.1 (Humphrey, 2004). The latter two components are integral to its functionally in foods as a colorant (Ronald E. Wrolstad & Culver, 2012). With alkaline saponification, the phytol attachment is cleaved forming chlorophyllin which changes the pigment from lipophilic to hydrophilic (Humphrey 2004). Treatment of the pigment with weak acid results in displacement of the centralized Mg2+ ion by hydrogen ions. The formation of this pigment, pheophytin, is indicated by expression of drab olive-green colors. Although this reaction is irreversible, strong complexes can be formed when Cu2+ or Zn2+ ions are centralized in this chromophore. These metal coordinations not only add stability to the pigment but also cause expression more desirable green colors. Typically these reactions are conducted on extracted chlorophyll carried commercial applications, but the formation of Cu-chlorophyll complexes also occurs naturally. Many plants contain

18 significant levels of the mineral, and Cu-chlorophyll has been documented in plant extracts obtained with supercritical CO2 (Humphrey, 2004).

2.3.4 Nitrogen-Heterocyclic Derivatives: Betalains

Of the many N-heterocyclic compounds, betalains are yellow and red pigment immonium derivatives of betalamic acid. They are comprised of two major groups: red- violet betacyanins, condensation products of betalamic acid with cyclo-Dopa and yellow- orange betaxanthins, the condensation products of betalamic acid and amines, Figure 2.1

(Stintzing, Schieber, & Carle, 2002). Betacyanins can also be substituted with mono- or di-saccharides, and the resulting glycosides could be linked to acylation groups, leading to diverse betacyanin structures.

Betalains are water-soluble and responsible for the colors of plants of order

Caryophyllales and are commonly found in red beet and cactus pear (Strack, Vogt, &

Schliemann, 2003; Wyler, 1986). Stable at a broad pH range from pH 3-7, betalains are mainly applied in low acid foods like dairy products (yogurt and ice-cream) (Stintzing &

Carle, 2004). In addition, the color of betalains is largely independent on the pH values compared to other natural colorants. Therefore betalains often complement anthocyanins in food applications, especially in low acid and neutral foods.

Betalain stability is influenced by various factors such as temperature, light, oxygen, water activity, and other food components. Hydrolytic cleavage of the aldimine bonds could occur at higher pH condition and during the heat treatment (Schwartz & von

Elbe, 1983), producing yellow betalamic acid. This causes decreased tinctorial strength as 19 well as undesired yellow hues in food product, limiting betalain application to products with short shelf-life, low-temperature storage, and opaque packaging. Several additives have been shown to stabilize betalains. For example, antioxidants (e.g. ascorbic acid), chelating agents (EDTA, citric acid), preservatives, and gums (e.g. pectin, locust bean gum) were shown to increase the betalain stability (Herbach, Stintzing, & Carle, 2006).

2.4 Classification by Hue and Application

2.4.1 Red

Red colorants are one of the most used in the food industry likely due to prevalent distribution in fruits and vegetables and in stimulation of appetite (Singh, 2006). Red colorants are widely in foods including beverages, confectionaries, dairy, meats, snacks and cereals. However, obtaining a vibrant red shade in food is challenging; natural red color shade was considered as the most challenging hue for new product development

(39%) and followed by green (19%) and blue (13%) (DDW The Color House 2013).

Anthocyanins are viable options for reds due to their chemistry, showing red color in acidic pH. In applications such as maraschino cherries, red radish anthocyanins expressed color similar to FD&C Red No. 40 (M. Mónica Giusti & Wrolstad, 2003).

Nonacylated anthocyanins often express red coloration in wider acidic pH ranges than acylated counterparts which appear more pink-purple. Figure 2.2 demonstrates possible hues of anthocyanins and many naturally occurring pigments. Anthocyanins have been applied to beverages, yogurts, and dry mixes and are gaining popularity. They are

20 permitted in foods as extracts from grape color or skin, vegetable juice, and fruit juice, exempt from certification in the U.S. (CFR 2016).

Pyranoanthocyanins are red pigments derived from anthocyanins during the wine fermentation and aging process first isolated in 1996 (Rentzsch, Schwarz, &

Winterhalter, 2007). Anthocyanin cyclisation with pyruvic acid increased their color stability, allowing for greater color expression at higher pH and resistance to bleaching by SO2 (Bakker & Timberlake, 1997; Fulcrand, Benabdeljalil, Rigaud, Cheynier, &

Moutounet, 1998). They exhibit hypsochromic absorbance shifts, showing more red- orange color. Like anthocyanins, color intensity expressed by the pigments is pH- dependent but comparatively much lower (Bakker & Timberlake, 1997).

Being stable in pH 3-7, betalains are suitable for coloring acidic to neutral pH foods to obtain a relative stable red hue. Red beet, exempt from certification in the U.S. and as E-162 in the E.U., has been established in the market as a widely used red food colorant. It has been commercially used for coloring food like yoghurt, confectionary, ice-cream, syrups and sausages. Red beets carry an earthy flavor, which could be imparted to food products. Recently, fruits from Cactaceae family have been introduced as promising betalain sources, which provides an alternative to red beet pigments.

Many carotenoid compounds express variations of red hues. Lycopene is a bright red pigment found in tomatoes, watermelon, guava, and pink fruit. As a typical carotenoid, lycopene is water-insoluble and displays orange hues in nonpolar solvent.

Lycopene, as tomato lycopene extract or concentrate, is exempt from certification in the

U.S. and also allowed in the E.U. Water dispensable and lycopene oleoresin are

21 commercially available for coloring purposes. Paprika (Capsicum annuum) is another good source of carotenoids, with high concentrations of capsanthin, capsorubin, zeaxanthin, capsilutein, and violaxanthin (Minguez-Mosquera & Hornero-Mendez,

1993). The colorant displays yellow-orange-red colors and is dependent on the composition of carotenoids. It is applied in meat products, soups, pickles, and snacks.

Another alternative for synthetic red colorants can be found in cochineal insects from superfamily Coccoidea, native to South America. The predominate pigment is carminic acid, a water-soluble anthraquinone. Its color changes with pH, being orange in acidic environments, red with decreasing acidity, and, violet in alkaline conditions.

Carminic acid has a good stability to heat, light, and oxygen and can also chelate metal ions to form carmine, colored pink to red and violet. Carmine is less dependent on pH and is extremely stable to oxidation and SO2. It has been applied in beverages, bakery, dairy products, jams, and confectionery. Limitations of these pigments include: poor consumer perception on the insect source, potential allergenicity associated with insect protein, and relative high cost (Acero et al., 1998).

Monascus pigments are products from fermentation of rice with Monascus spp yeast. They have long been used in Asia as naturally derived food colorants and food preservatives due to their antibacterial properties (Danuri, 2008). The color of Monascus pigments range yellow-orange-red. They are soluble in ethanol and slightly soluble in water (Figure 2.1). Monascus pigments have been increasingly applied to food products like meat and sausage replacing nitrates/nitrites, enhancing color, and improving stability

(Pattanagul, Pinthong, & Phianmongkhol, 2007). Although not approved in U.S.,

Monascus pigments are promising natural colorants for their various color appearance and other benefits.

2.4.2 Orange-Yellow

Carotenoids also offer potential alternatives for synthetic yellow-orange colorants.

Commonly used pigments involve α-, β-, γ-carotene, lycopene, capsanthin, bixin, and etc.

Although abundant in nature, the bulk of carotenoids are mainly obtained by synthetic methods. Lutein is a common yellow carotenoid found and is commercially available as a food colorant. Aztec marigold (Tagetes erecta) is an important source of lutein

(Breithaupt, Wirt, & Bamedi, 2002), but its application is limited to poultry feed for yellow enhancement of eggs and skins. Annatto is a yellow-orange colorant extracted from the seeds of tropical tree Bixa orellana. The major pigment responsible for the annatto seeds is bixin, a carotenoid type compound of 26-carbons. Bixin is lipid-soluble and is extracted with vegetable oil or organic solvents. When alkali is used for extraction, bixin is saponified forming norbixin, a water-soluble pigment. Thus both water-soluble and lipid-soluble annatto preparations are commercially available. Annatto has been used in cheddar cheese, beverages, flavored milk drinks.

Saffron is the dried stigma of crocus sativus flower which contains the water- soluble pigment crocin, a glycoside of crocetin. It is quite stable to light and heat but is labor intensive and therefore high cost. It is approved as food colorant in the United

States but not in the E.U. where it is considered as spice (CFR 2016). Gardenia fruits

(Gardenia jasminoides Ellis) are source of a similar yellow compound. Extracts of 23 gardenia fruit contain crocetin giving yellow, and also other red and blue compounds. It is widely adopted in Asian countries as a natural colorant though not legal in the U.S.

About 320 tons of gardenia yellow are consumed in Japan annually, and demand has been increasing (Nakamura, 1995).

Turmeric is a yellow spice from the dried rhizomes of the herb curcuma longa.

Three lipid-soluble pigments are found in turmeric: curcumin, demethoxycurcumin, and bisdemethoxycurcumin. The hue of the colorants can range from a pale yellow to a vibrant yellow depending on the concentration, Figure 2.2. Turmeric oleoresin and curcumin are allowed as food colorants in the U.S., while only curcumin is approved colorant in E.U. (CFR 2016). Various extraction methods can lead to different yellow variations depending on the content and ratio of the contained pigments. By particle size reduction and emulsifiers, aqueous dispersible curcumin is produced commercially.

Turmeric is stable to heat but susceptible to light, oxidation, and alkaline conditions. It can be applied in margarines, baked goods, bouillon, compound coatings, dry beverage mixes, ice cream, panned candies, sauces, seasonings, soups, and frostings, dairy, and cereal (DDW The Color House 2014).

2.4.3 Purple

Purple color is considered the range of hues between blue and red and not traditionally defined in the visible light spectrum (Gilbert & Haeberli, 2008). However, violet is described as one of the spectral components of white light and is associated with wavelengths of 380-420 nm (Gilbert & Haeberli, 2008). As a combination of red and blue 24 colors (such as anthocyanins and spirulina), purple hues in food products can be obtained by mixes of pigments expressing these colors.

Acylated anthocyanin derivatives from red cabbage, purple sweet potato, and purple carrot have been shown to express purple color in pH as low as 5 (Ahmadiani,

2012). The uniquely acylated anthocyanins of the butterfly pea flower (Clitoria ternatea) have been found to resulted in purple color expression in pH 3-5 (Abdullah, Lee, & Lee,

2010). Additional expansion of the purple color expression of anthocyanins can also be result of metal ion. Violet hues were found to be produced by a variety of anthocyanin and metal ion combinations in acidic pH but was more common with cyanidin derivatives

(Buchweitz, Brauch, Carle, & Kammerer, 2013b; Gregory T. Sigurdson & Giusti, 2014).

Another recently identified purple pigment includes violacein, a bacteriocidal and antiviral pigment produced by Chromobacterium violaceum (Rettori & Durán, 1998).

The pigment was found to be soluble in water, and its color was stable from pH 1-11 expressing hue angles from 278-295 (Venil et al., 2015). In model systems of yogurt and agar based gels, the pigment’s color was stable >1 month of storage at 4 ºC (Venil et al.,

2015). Being a recently discovered, approved usage in foods will require safety evaluations and regulatory approval.

2.4.4 Blue

Although not necessarily uncommon in nature, blue colors are difficult to reproduce in foods and beverages. Several chemical classifications of pigments can be responsible for blue color observation; those that do typically possess π-bond 25 conjugation, aromatic ring systems, heteroatoms, and ionic charges (Newsome et al.,

2014).

Blue color expression by anthocyanins typically employs pH increase, copigmentation, and/or metal chelation. The aforementioned anthocyanins of Clitoria ternatea were found to express blue hues in pH 5-7, and the anthocyanins of red cabbage and purple sweet potato expressed colors similar to FD&C Blue No. 2 in pH 8 (Abdullah et al., 2010; Ahmadiani, 2012). Recently, a naturally derived colorant has been commercially developed from vegetable sources that shows blue hues in pH ~3-7. The plant source or chemical structure is not public knowledge, but the pigment’s response to pH suggests the pigment to be anthocyanin based (Sensient Colors LLC, 2016).

Anthocyanin macromolecular complexes with flavones and metallic cations have been identified as responsible for the blue colors of multiple floral systems (Yoshida et al., 2009). Commelinin, from Asiatic dayflowers, is a blue pigment stable in concentrated solutions in pH ≥ 2.4. With dilution, the macromolecule easily dissociated losing blue color (Takeda, Fujii, & Iida, 1984; Yoshida et al., 2009). Methods to stabilize the complex, such as encapsulation, may improve the stability of the blue pigment.

Anthocyanin chelates of Al3+ and Fe3+ cations also developed blue colors in pH ≥ 2.5, depending on anthocyanin structure and ratio to cation (Gregory T. Sigurdson & Giusti,

2014). More heavily substituted anthocyanins typically expressed blue colors in widest acidic pH ranges (Buchweitz, Carle, & Kammerer, 2012; Gregory T. Sigurdson & Giusti,

2014).

Dimers of pyranoanthocyanins, or portisins, found to form naturally during the aging of wine also exhibit blue hues in acidic pH (Mateus, Oliveira, Haettich-Motta, & de

Freitas, 2004). One of these pigments isolated from port wine expressed a turquoise blue at pH 2 (Oliveira et al., 2010). The pigments would require regulatory approval for food use, but occurring naturally in a historically consumed beverage, toxicity in low concentrations may prove unlikely.

Blue colorants also approved for broad range food use in the U.S. include phycocyanins as spirulina (Arthrospira platensis) extracts (CFR 2016). Phycocyanin is a protein complex that shows some pH sensitivity, changing λmax between pH 5-7, and is insoluble at pH 3 (Jespersen, Strømdahl, Olsen, & Skibsted, 2005). The use of this pigment is also further limited by its poor light stability and high sensitivity to heat, losing color at temperatures ≥ 45 ºC (Jespersen et al., 2005). It exhibits a bright blue color in some applications, Figure 2.2, and has found use in many confectionary products and ice creams.

Iridoid derivative blue colorants have long historical use and relatively accepted records for safety. Precursors of the blue pigments are found in plants of Rubiaceae family, like gardenias or the huito fruit (S. Wu, Ford, & Horn, 2009). After expressing juice from plant materials, colorless iridoid glycosides are hydrolyzed by β-glucosidase releasing genipin and glucose (Jespersen et al., 2005). Blue colors then result when genipin reacts with amino acids; condensation with different amino acids results in different blue hues (S. Wu et al., 2009). These pigments show color similar to FD&C

Blue No. 2 (Ahmadiani, 2012). The use of blue gardenia pigments in foods is not legal in

27 the U.S. or E.U. but is in Japan and occurs in China and Korea (Jespersen et al., 2005;

Newsome et al., 2014). The blue colorants from huito or combinations with other fruits is offered as a fruit juice for color (Wild Flavors and Specialty Ingredients, 2016).

Other recently investigated blue pigments include mareninne and tetrapyrrole derivatives from processed garlic. Marennine is a blue-green pigment synthesized by the marine diatom Haslea ostrearia; its chemical structure is still unknown (Pouvreau et al.,

2006). The pigment is water soluble and somewhat responsive to pH, being blue in acidic conditions and greener in alkaline pH (Gastineau et al., 2014; Pouvreau et al., 2006). In the case of tetrapyrroles, biliverdins are one of the few blue pigments of the animal kingdom, responsible for the bluing of bruises (Newsome et al., 2014). The blue-green discoloration of processed garlic was also found to be result of pyrrole compounds

(Block, 2010). The formation of the pigments is closely related to decrease in thiosulfinate compounds, which also resulted in decrease of allium flavors (Block, 2010).

As many as 8 compounds have been identified as blue or green, of which those are tri- and tetra-pyrroles (Block, 2010). Further investigations would be required for their suitability as a food colorant.

2.4.5 Green

Green food colorants from natural sources are relatively limited despite the prevalence of green hues in natural settings. Chlorophyll is the most widely distributed pigment in nature and is found in many sources including plants, algae, and bacteria

(Humphrey, 2004; Ronald E. Wrolstad & Culver, 2012). Likely related to its functions in 28 photosynthesis and as a catalyst, the pigment is highly unstable and losses its green color quite easily (Humphrey, 2004). Cu-chlorophyllin and associated complexes are approved for broad range use in foods in the E.U., but in the U.S., use of chlorophyll to color foods is limited to Na-Cu-Chlorophyllin in citrus-based dry beverage mixes (CFR 2016,

Humphrey, 2004). The Cu-chlorophyllin complex is so stable that that the Cu ion cannot be freed by normal metabolism or even concentrated HCl digestion (Humphrey, 2004).

Other green coloring, typically considered a detriment to quality, has been observed in products such as sweet potato and burdock particularly in injured or processed products (Yabuta, Koizumi, Namiki, Hida, & Namiki, 2001). The compounds responsible for this green coloration were benzacridine trihydroxy derivatives, condensation products of 2 molecules of oxidized chlorogenic acid and primary amino compounds under aeration, Figure 2.1 (Yabuta et al., 2001). The water soluble green pigment was found to develop more quickly in alkaline pH. The pigment may show as a food colorant as it showed no mutagenicity by the Ames test, but further evaluation will be required (Yabuta et al., 2001).

Aside from chlorophyll derivatives, current options for natural green food colorants are typically mixtures of blue and yellow pigments resulting green color observation. These mixtures can include a variety of sources and pigments, depending on application.

2.4.6 Brown

The perception of brown color can be considered a human response to variety of chromophores, having no specific λmax observable by a visible absorbance spectrum. The absorbance spectrum of caramel color shows no distinctive peaks and decreasing absorbance at increasing wavelengths (Ronald E. Wrolstad & Culver, 2012). Caramel color accounted for more than 90% by weight of all colors produced and is approved for use the U.S., Europe, and most other countries (Ronald E. Wrolstad & Culver, 2012). In the U.S., caramel coloring has a long history of use with first commercial manufacturing dating to 1863 (Chappel & Howell, 1992); current regulations list caramel colors as exempt from certification (CFR 2016). Caramels produce a variety of brown colors (light yellow to red-brown and black-brown) depending on processing procedures. Referred to as burnt sugars, caramel is produced by thermal treatment of a variety of sugars, sometimes in conjunction with added acids, alkalis, nitrogen, or sulfur sources (CFR

2016). Caramelization is a complex series of reactions characterized by sugar-amino condensation, Amadori or Heyns rearrangement, sugar dehydration, Strecker degradation, and condensation and polymerization reactions (Sepe, Parker, Nixon, & Kamuf, 2008).

These different production procedures resulted in classification of caramels into 4 categories with different functional properties: Class I produced only from sugars, Class

II processed with sulfites, Class III with ammonium, and Class IV with both ammonium and sulfites added (Ronald E. Wrolstad & Culver, 2012). Addition of nitrogen sources results in products similar to those of the Maillard reaction. Recent review of toxicology data has led the EFSA to conclude no evidence of carcinogenic activity (EFSA, 2012). 30

Despite the safety of caramel colors, consumer demand has further instigated the investigation of alternatives for brown food colorants. Polyphenolic compounds are responsible for several colors found in nature including some yellow and brown hued pigments. The phenolic concentrates of apple are being sold as an option for naturally sourced brown colorants in international markets including Europe and Japan (Akazome,

2004; Herbarom Ingredients GmbH, 2013). Applications have been described as similar to that of caramel colors (Herbafood Ingredients GmbH 2013). The extracts appear rather safe, showing a no observed adverse effect level (NOAEL) of 2g/kg in rat models

(Akazome, 2004). As an extract, the colorant does not have legal use in the U.S.; but may as a juice concentrate, exempt from certification (CFR 2016).

Other alternatives for brown colorants include toasted defatted cottonseed flour, which imparts light-dark brown color due to quercetin glycosides and gossypol derivatives (Blouin & Cherry 1980, Wrolstad & Culver 2012). In the U.S., this can be used with good manufacturing practices as a colorant exempt from certification (CFR

2016) but is not legal in Europe (Ronald E. Wrolstad & Culver, 2012).

Recently, an additional alternative for brown colorants was released with ingredients including vegetable juice, β-carotene, annatto, paprika, and/or turmeric

(Watson 2014). The technology may function by employing the idea that no specific λmax correlates to brown color observation and creating dense saturation of multiple pigments to allow for broad absorbance over the visible spectrum.

2.4.7 Black

The observation of black color is the result of a lack of reflectance or emittance of all wavelengths of visible light. In color communication systems, black is considered achromatic; however in application, it is considered a color. Black colors can be achieved through means that will allow for absorbance of light of all wavelengths, such as combinations of many chromophores or high concentrations of that same chromophore

(Ronald E. Wrolstad & Culver, 2012).

Currently in the U.S., no single sources for black colors from natural sources may be used in food products; with the pseudo exception of iron oxides. Certain compositions of black iron oxide occur in nature; however only synthetic varieties are permitted as exempt from certification (CFR 2016). The use of these water insoluble pigments is limited to coloring of sausage casings or confections for human consumption and dog or cat food (CFR 2016).

In the E.U., vegetable carbon (E153) is permitted (Miranda-Bermudez, Belai,

Harp, Yakes, & Barrows, 2012). Carbon black is permitted in the U.S. to color drugs and cosmetics as a certified color (Black D&C No.2); this is to ensure there is a limitation of harmful by-products from manufacturing the colorant. Vegetable carbon (E153) is produced by the combustion (carbonization) of materials such as wood, cellulose, coconut or other shells (Miranda-Bermudez et al., 2012). A long history of use as a medical substance and chemical inertness in vivo also attest to its safety. Carbon is an insoluble, odorless, and tasteless pigment whose coloring power relates to its particle

32 size, being more intense as finer particles (European Food Safety Authority (EFSA),

2012).

Other potential options for black colorants for foods include melanin based pigments. Two major classes express black colors: eumelanins and allomelanins (Swan,

1974). Eumelanins are polymers containing nitrogen typically from oxidized tyrosine

(animal derived) while allomelanins are plant polymers of oxidized catechols (Nicolaus,

Piattelli, & Fattorusso, 1964; Swan, 1974). Squid ink is a traditional ingredient in

Japanese and Mediterranean cuisines used to impart light flavor and dark color to a variety of foods. Sepiamelanin (the ink from cephalopods) is an insoluble eumelanin that incorporates some other aromatic amino acids (phenylalanine and tryptophan) and occurs as a protein-melanin complex suspended in clear fluid (Swan, 1974). Although it has shown many potential health benefiting capabilities, squid ink is not current permitted as a food coloring additive in the U.S. (Ronald E. Wrolstad & Culver, 2012), but the pigment has been in Japan to color buns and other components of burgers for fast food chains (Burger King Japan, 2015).

2.4.8 White

In contrast to black, white is the result of a high degree of reflectance of all wavelengths of light but is also considered achromatic. Current naturally sourced options for white pigments are typically minerals. The only white pigment approved for food use in the U.S. is titanium dioxide (TiO2), exempt from certification; it is also approved in the

E.U. as E171 (CFR 2016; Weir et al. 2012). Although the pigment occurs in nature, U.S. 33 regulations limit its use to synthetized varieties and its amount in food to not exceed 1%

b by weight (Code of Federal Regulations , 2015). TiO2 was found in the highest abundance in candies, sweets, and chewing gums (Weir et al., 2012). Toxicity is considered extremely low to its extremely low solubility. In food formulations, it is therefore generally dispersed (Weir et al., 2012; Ronald E. Wrolstad & Culver, 2012).

An alternative to TiO2 has been created from calcium carbonate (CaCO3), which was traditionally limited due to expression of poor quality white like tones (Koehler,

2014). Recent technological advancements have improved the quality of the white color increasing its potential use in foods, such as water-dispersible composition including a starch hydrocolloid (Koehler, 2014). There are no negative health concerns for CaCO3 and it is not used commonly used in non-food products, helping to retain positive consumer perception of the pigment (Koehler, 2014). The pigment can be naturally sourced from mining of limestone or found in chicken egg shells; however U.S. regulations limit its use to drugs as synthetically prepared forms (CFR 2016). The pigment is permitted in the E.U. as a food colorant, E170, and is not considered to be a health concern (European Food Safety Authority (EFSA), 2011).

Limited information was found available for a newly developed technology offering an additional alternative for a whitening or opacity agent that contains neither

TiO2 nor CaCO3. The technology is considered to be clean label and shown to effectively color in broad pH from 2.5-7.0 (Sensient Colors LLC, 2016).

Figure 2.1 Chemical structures of selected natural pigments with potential food use, organized according to their chemical classification

Figure 2.2 Representative food colorants from natural sources, organized according to their hue

Table 2.1 Major colorants from natural sources (or nature identical): chemical classification sources, colors and regulatory status in the USA

Chemical CFRa classification Additive Source Color section Grape skin Grapes (marc from pressing) 73.170 Grapeextract color Concord grapes (Vitis labrusca) 73.169 Anthocyanin Red - purple - blue Fruitextract juice Berries 73.250 Vegetable juice Carrot, cabbage, and others 73.260 Betalain Beet powder Beet (Beta vulgaris) Pink - red 73.40 Caramel Caramel Heating of sugars Brown 73.85 Carminic Acid Cochineal Dactylopius coccus Orange - red 73.100 extract Annatto Bixa orellana L. 73.30 Synthesized, Bacterium (Paracoccus carotinifaciens), Yeast (Phaffia 73.35 Astaxanthin 73.355 rhodozyma), Algae (Haematococcus 73.185 pluvialis) β-carotene Natural (carrots) or Synthesized 73.95 Canthaxanthin Synthesized 73.75 Carrot oil Carrot (Daucus carota L.) Yellow - orange - 73.300 Carotenoid Corn endosperm Yellow corn (Zea mays) red 73.315 Paprikaoil 73.345 Paprika (Capsicum annum L.) (oleoresin)Paparika 73.340 Saffron Stigma (Crocus sativus L.) 73.500 Tagetes Aztec Marigold (Tagetes erecta L.) 73.295 Tomato Tomato (Solanum lycopersicum) 73.585 ßlycopene-Apo-8’- Synthesized 73.90 Sodiumcarotenal copper Chlorophyll Alfalfa (Medicago sativa) Green 73.125 chlorophyllin Turmeric 73.600 Curcuminoid Rhizome (Curcuma longa L.) Yellow (oleoresin) 73.615 Iron oxide Synthesized Yellow/orange/red/ 73.200 Mica pearlescent Synthesized brown/Pearlescentblack 73.350 Titanium dioxide Synthesized White 73.575 Mineral Ultramarine blue Synthesized Deep blue 73.50 Ferrous Synthesized Yellowish-gray 73.160 gluconate Ferrous lactate Synthesized Greenish-white 73.165 Phycocyanin Spirulina extract Algae (Arthrospira platensis) Blue-green 73.530 Vitamin Riboflavin Microbial fermentation Yellow 73.450 (CFR 2016; Food Standard Agency 2014)

Chapter 3 Isoflavone in Human Health

Published. Giusti, M. M., Ahmadiani, N., Tang, P., & Ottinger, M. A. (2015). Isoflavone and flavonoid supplemented eggs in health. Chapter 20 in Handbook of Eggs In Human Function (pp. 334-364). Wageningen Academic Publishers.

3.1 Flavonoids

Flavonoids are plant secondary metabolites, which are present in almost all the terrestrial plants (Asfaw & Demissew, 1994). They share the similar structures, a phenylbezopyran C15 (C6-C3-C6) heterocyclic nucleus structure, and vary in the aromatic ring positions and different methoxy and hydroxyl substitutions on the phenolic rings. The general chemical structures of these compounds are shown in Figure 3.1.

These compounds are made of two aromatic rings (A and B rings) joined to each other by benzopyran moiety (C-ring).

Flavonoids are widely distributed in most plants. A comprehensive food flavonoid database was established by Kyle and Duthie, in which isoflavones were found in 35 types of fruits, 31 kinds of vegetables, 26 beverages and 12 herbs species (Kyle and

Duthie 2006). It was estimated that, based on the flavonoids analysis in different foods, the average intake of flavonoids was up to 1 gram per day per person in USA (Kühnau,

1976).

Many researches and studies have demonstrated that flavonoids exert protective effect on human health, such as the anti-oxidative effect (Duthie & Crozier, 2000;

Manthey, Guthrie, & Grohmann, 2001; Morton, Abu-Amsha Caccetta, Puddey, & Croft,

2000), coronary heart disease (CHD) prevention (Hertog et al., 1995), and anti- carcinogenic effect (Birt, Hendrich, & Wang, 2001; Ezekiel, Singh, Sharma, & Kaur,

2013). Given the potential health benefits from isoflavones, many researchers have been exploring different ways to incorporate flavonoids into people’s daily diets.

3.2 Isoflavone

Isoflavone, one of the flavonoid, is a plant secondary metabolite. It has drawn wide attentions over the last few decades because of its potent phytoestrogenic effects

(Setchell, 1998). Isoflavones, unlike flavonoids, are believed to be limited to leguminous plants, such as soybeans (Price and Fenwick 1985). The physiological functions of isoflavones are mainly explained by its similar structure compared to the endogenous estrogen: 17β-estradiol (Figure 3.4). Isoflavones can bind to the estradiol receptors (ER), especially the ER-βreceptor subtype to trigger the downstream gene transcription, thus exerting physiological activities (Duncan, Phipps, & Kurzer, 2003; Markiewicz, Garey,

Adlercreutz, & Gurpide, 1993). Numerous papers have investigated the roles of isoflavones in health, and found isoflavones possess multiple health benefits, such as antioxidant (Brenda J Boersma et al., 2003; Wiseman et al., 2000), anti-cardiovascular diseases (E A Kirk, Sutherland, Wang, Chait, & LeBoeuf, 1998), anti-osteoporosis

(Vitale, Piazza, Melilli, Drago, & Salomone, 2013), and prevention of breast cancer (Lee,

Gourley, et al., 1991).

3.2.1 Isoflavone structure

Isoflavones are the most abundant form of naturally occurring isoflavonoids, with

334 different structures (Asfaw and Demissew 1994). Isoflavones differ from flavones in that the benzyl ring B is joined at position 3 instead of 2 (Figure 3.1). Soy isoflavones are in three aglycone types: Daidzein, Genistein, and Glycitein. These compounds, however, can be glycosylated and acylated to form other chemical forms (Figure 3.2).

Therefore, there are total of 12 isomers of isoflavones (Figure 3.2). Attachment of glucose happens through a glycosidic bond to the A ring, which turn the aglycone form to

β-glucoside form: Daidzin, Genistin, and Glycitin (kudou et al. 1991). Acylation of isoflavones with acetic or malonic acids happens through the formation of ester bond with the attached sugar moiety to form Acetylglucoside or Malonylglucoside forms

(Figure 3.2) (Wang and Murphy 1994a).

Isoflavones occur mainly in their glycoside form and this has effect on their absorption and retention within the human body (Izumi et al., 2000). Chemical structure of isoflavone also affects the functionality of these compounds within the human body.

For instance, equol, which is a minor isoflavone made through the metabolism of

Daidzein, (Daidzein metabolite) can be much more potent than the parent compound

(Figure 3.3) (Morito et al., 2001).

Figure 3.1 Structure comparison of different flavonoids 41

Aglycones

Compound R1 R2 Daidzein H H Genistein OH H Glycitein H OCH3

Compund R3 R4 R5 Daidzin H H H Glucosides Genistin OH H H Glycitin H OCH3 H 6”- O-Acetyldaidzin H H COCH3 6”- O-Acetylgenistein OH H COCH3 6”- O-Acetylglycitin H OCH3 COCH3 6”- O-Malonyldaidzin H H COCH2COOH 6”- O-Malonylgenistein OH H COCH2COOH 6”- O-Malonylycitin H OCH3 COCH2COOH

Figure 2 Figure 3.2 Chemical structures of soy isoflavones, categorized into four forms: aglycones, glucosides, acetylglucosides and malonylglucosides

Figure 3.3 ChemicalFigure 3 structure of equol

3.2.2 Isoflavone in Foods

Dietary isoflavones mainly come from soybean or soy foods (Liu, 1997). In Asia, soybean is regarded as the most important cultivated legumes, and a vital source of dietary proteins. The total isoflavone content in soybeans has been reported to be ranging from 1261 – 2343 μg/g (Wang & Murphy, 1994b). But, this value will change based on different cultivars, environment, locations, and the years (Figallo, 2003; Tsukamoto et al.,

1995; Wang & Murphy, 1994b). In fact, different environment stress has great impact on the isoflavone synthesis. Based on previous studies, genistein is more abundant than daidzein, and daidzein is much more than glycitein in soybeans (Wang & Murphy,

1994b). Also, isoflavones malonyl derivatives are more abundant than β-glucosides, followed by acetyl derivatives and aglycones (Figallo, 2003).

Soybean-related products, such as miso, tempeh and tofu, not only vary in isoflavone concentration, but also differ in isoflavone profiles, due to the different food processing. For example, soy flour, which undergoes little food processing, contains an average of 1.06 mg/g isoflavones, with 80% malonyl derivatives, 15%β- glucosides, respectively (Murphy, Barua, & Hauck, 2002; Umphress, Murphy, Franke, Custer, &

Blitz, 2005). Tofu, due to the soaking process, contains 0.53 mg/g isoflavones, with up to

25% β-glucosides, and 37% malonyl derivatives (Wang & Murphy, 1994b). Another example is tempeh, with the stronger enzymatic activity involved in the food processing,

50% of the total isoflavone (0.86mg/g) are aglycones (Murphy et al., 2002; Wang &

Murphy, 1994b). A comprehensive investigation upon the isoflavone content and profiles in 557 foods has been conducted by Nutrient Data Laboratory of Agricultural Research 43

Service, U.S. Department of Agriculture (Bhagwat et al 2008). Table 3.1 shows the total isoflavone content in some of the soy foods on a wet weight basis (Bhagwat et al 2008).

3.3 Isoflavone and Health

3.3.1 Isoflavones Bioavailability and Metabolism

Despite of the potential phytoestrogenic activities, the bioavailability of isoflavones is considered as a critical factor that influences their functions (Hendrich et al

1998). After the consumption, isoflavones are absorbed through the gastrointestinal tract, in which different chemical reaction and flora promote the metabolic process (such as digestion and absorption) of isoflavones (Turner, Thomson, & Shaw, 2003). Most isoflavones in our diet are β-glucoside derivatives. When passing the gastrointestinal tract, those glucoside derivatives will be hydrolyzed by intestinal β-glucosidases or specific microbial flora, resulting the accumulation of aglycones in small intestine (Day et al., 1998; Izumi et al., 2000). Isoflavones in aglycone forms could be further absorbed by intestinal epithelial cells (Setchell & Cassidy, 1999). In the liver, those isoflavones are metabolized by phase ΙΙ enzymes (either become glucuronidated, or sulfated, or just left as aglycones) (Xu, Harris, Wang, Murphy, & Hendrich, 1995); some isoflavones will also be metabolized by the microbial flora into other compound, such as equol in small intestinal (Axelson et al., 1982). Glucuronides, sulfates, and aglycones are the major isoflavone forms found in urinary excretion (Zhang, Hendrich, & Murphy, 2003). It has been reported that biliary excretion is the final fate of the isoflavone (Stakianos et al

1997). 44

Table 3.1 Mean value of total isoflavones concentrations in various food products in a wet weight basis.

Food Item and Description Total Isoflavones Concentration (mg/100g edible portion) Egg, whole, raw, fresh 0.05 Bacon 9.36-118.50 Black bean, sauce 10.26 Breakfast Cereals 0.02-93.90 Clover 0.25-21.00 Instant beverage, soy, powder, not 109.51 reconstituted Mayonnaise, made with tofu 16.80 Miso 41.45 Miso soup 1.52-69.84 Natto 82.29 Oncom 6.50-9.70 Soy cheese 6.02-25.72 Soy drink 7.85 Soy flour 150.94-178.10 Soy infant formula 2.21-28.01 Soy lecithin 15.70 Soy meal, defatted, raw 209.58 Soy noodles, flat 8.50 Soy paste 38.24 Soy protein products 11.49-94.65 Soy sauce 0.10-1.18 Soy yogurt 33.17 Soy-based liquid formula for adults 0.34-0.57 Soybean products 34.68-131.53 Soybean seeds 2.82-178.81 Soybeans 12.50-48.95 Soymilk product 0.70-196.05 Tempeh product 29.00-72.80 Tofu product 3.62-83.20 Data are from USDA Database for the Isoflavone Content of Selected Foods (http://www.ars.usda.gov/SP2UserFiles/Place/12354500/Data/isoflav/Isoflav_R2.pdf)

3.3.2 Isoflavones as Phytoestrogens

Estrogen is a sexual hormone that works as a signaling compound, regulating many biological processes in our body, such as sexual development, sexual reproductions, cholesterol transportation, and bone metabolism (Mechoulam,

Brueggemeier, & Denlinger, 1984). Their target cells are distributed in breast, liver, heart, and bones. When estrogen binds its receptors (ERα or ERβ), the activated ligand- receptor complex will bond to estrogen response elements (Azeke & Ekpo, 2008) on the

DNA sequence. With the help from co-activators, specific genes will be transcribed, then, the protein products from those genes can promote cell proliferation (Wuttke, Jarry, &

Seidlová-Wuttke, 2007).

Isoflavone aglycones have the similar structure when compared to the robust estrogen: 17β-estradiol, especially in the phenol rings and the unique distance (11.5Å) between two –OH groups (Figure 3.4) (Leclerq et al 1979). 17β-estradiol was reported to bind estrogen receptors ERα and ERβ to activate the biological signaling pathway, promoting the cell proliferation (Figure 3.4). The estrogen-targeted tissues include breast, bone, liver, and brain (Setchell, 1998). The similar structure of isoflavones compared to estradiol, therefore, explains their functions as phytoestrogens: anti-cancer, anti-cardiovascular diseases, anti-osteoporosis and other diseases (Anderson, Johnstone,

& Cook-Newell, 1995; Messina, Persky, Setchell, & Barnes, 1994; Potter et al., 1998;

Sirtori, 2001). Although the potency of isoflavones’ binding to ER is relatively low, the high level of isoflavones in serum makes isoflavone display robust physiological effects

(Adlercreutz et al., 1991; Miksicek, 1994). Furthermore, isoflavones are believed to have 46 both estrogenic and anti-estrogenic effects in the low-level and high-level estrogen environments, respectively.

b Figure 3.4 The structure relationship between 17β-estradiol and Genistein (left) and the crystal structure of the estradiol-human estrogen receptor-α ligand binding domain complex (right) as proposed by Tanenbaum et al. (1998).

Figure 4

3.3.3 Isoflavones and Women’s Health

Interests in the effect of isoflavone on women’s health are mostly focused on postmenopausal women. The hormonal change that occurs during menopause can cause a variety of symptoms, such as night sweats, hot flashes, and increase the risk for heart

47 disease and osteoporosis. Hormone replacement therapy (HRT) is usually prescribed to help prevent the negative health effect of menopause. Alternatively, isoflavone could be consumed to lower the incidence of menopausal symptoms by estrogenic effects of the isoflavones (Burke 1996; Harding et al 1996). Isoflavones may, however, act as an estrogen antagonistic in premenopausal women who have a high level of endogenous estrogen. In this case, phytoestrogens may protect women from hormone-dependent cancers by increasing menstrual cycle length and reducing overall lifetime estrogen exposure (Cassidy, Bingham, & Setchell, 1994; Lu, Anderson, Grady, & Nagamani,

1996).

3.3.4 Isoflavones and Breast Cancer

Breast cancer accounts for nearly one fifth of the cancer death in women, especially in Western Countries. However, in Asia like Japan where soy consumption is relatively high, the death rate is significantly low (American Cancer Society 2005).

People tend to assume that those who consume high level of soybean products, which mainly contain isoflavones, will have low risk of breast cancer. However, the relationship between soy isflavones intake and breast cancer risk is controversial.

Both in vitro and in vivo studies have reported that genistein could stimulate the growth of breast cancer cell MCF-7 (Zava & Duwe, 1997). But, two case-control studies from Singapore and Japan showed that the regular consumption of soyfoods is linked to a significant decreased risk of breast cancer in premenopausal but not postmenopausal women (Lee et al. 1991; Hirose et al. 1995). In addition, genistein administration in 48 neonatal rats resulted in tumor reduction when tumors were chemically induced in rats as adults phase of life (Lamartiniere et al. 1995). And genistein intaken in prepubertal rats leads to lower incidence of adenocarcinomas than that in elder rats. These findings imply that early uptake of phytoestrogen may benefit later life in the breast cancer prevention

(Murrill et al., 1996). That is the reason why in Asian countries, where soy products are consumed throughout the lifetime, the breast cancer risk is significantly lower than that in

Western countries (Setchell, 1998).

3.3.5 Isoflavones and Osteoporosis

Phytoestrogens have been considered as the main factor explaining the lower bone loss in Asian women as compared to Western women (Cooper, Campion, & Melton,

1992). It has been hypothesized that isoflavone consumption may increase bone calcium retention; therefore, increasing bone mineral density (Riggs et al. 1990). More than 10 studies have demonstrated that the consumption of soy products or the clover-derived isoflavone supplement for 0.5-2 years could prevent lumbar spine bone mineral loss and increase the bone mineral density in postmenopausal women throughout the world

(Atkinson, Compston, Day, Dowsett, & Bingham, 2004; Chen, Ho, Lam, Ho, & Woo,

2003; Lydeking-Olsen, Beck-Jensen, Setchell, & Holm-Jensen, 2004; Potter et al., 1998;

Uesugi, Fukui, & Yamori, 2002). Nevertheless, two trials (Brink et al., 2008;

Vupadhyayula, Gallagher, Templin, Logsdon, & Smith, 2009) found that long-term consumption of isoflavone-enriched diet resulted in no significant effect on bone mineral density in postmenopausal women. Those results seem to be contradictory. However, 49 when evaluating those trials, other factors should be taken into account, such as different human populations, age, their ability to metabolize daidzein into equol; differences in soy cultivars used, food supplements and other factors.

3.3.6 Isoflavones and Cardiovascular Disease

Cardiovascular diseases are among the major cause of death in most developed and developing countries (Finegold, Asaria, & Francis, 2013). The capabilities of isoflavones to reduce low-density lipoproteins (LDL), which have central role in the atherosclerotic process, have been extensively studied. There is some data that suggest soy isoflavones may be largely responsible for cholesterol reduction and the protective effect for coronary artery atherosclerosis (Anderson et al., 1995; Potter et al., 1998).

Additionally, isoflavone have not only been able to lower LDL cholesterol, but also increased high-density lipoproteins (HDL), which have a protective effect and act to prevent LDL oxidation and to remove cholesterol that accumulates in the blood vessel wall (Anderson, Diwadkar, & Bridges, 1998; E A Kirk et al., 1998; Munro et al., 2003).

Although isoflavones have been proven to have the beneficial effect on lowering the risk of cardiovascular diseases and reducing the lipid oxidation, its effect on the atherosclerosis progression and how their metabolites would affect the cardiovascular diseases still remains to be explored.

3.3.7 Isoflavones antioxidant activity

Numerous studies have observed the antioxidant effect of isoflavones both in vivo and in vitro (Wei et al. 1995; Foti et al. 2005; Boersma et al. 2003; Wiseman et al. 2000).

For example, the consumption of soy-enriched vegetarian burgers (containing 21.2 mg daidzein and 34.8 mg genistein) in healthy adults for 17 days would not only significantly reduce the plasma concentrations of 8-epi-PGF2α, a biomarker of lipid peroxidation in vivo, but also increase the resistance of LDL oxidation (Wiseman et al., 2000). A group of scientists also examined the antioxidative effect of isoflavone supplement on oxidative damage in blood. Six women and six men were provided soy supplementations (50 mg isoflavone once a day for women and twice a day for men) for 3 weeks, and they found the 5-hydroxymethyl- 2’-deoxyuridine level in blood, a biomarker for oxidative DNA damage, was decreased for 47% in women and 61% in men (Djuric, Chen, Doerge,

Heilbrun, & Kucuk, 2001). These antioxidant activities were regarded as the possible mechanisms for cardiovascular disease protective effect. Furthermore, genistein was reported to inhibit the endothelial cell proliferation and in vitro angiogenesis, which is related to tumor development if in a disorder and disease status (Fotsis et al., 1993). Since angiogenesis and inflammation response are often associated with cancer development, those findings could explain the anti-carcinogenesis effect of isoflavones.

While there are many clear and supportive evidences demonstrated that soy- enriched diet have antioxidant effect on human, other studies indicated no effect. For example, in Djuric’s study, though 5-hydroxymethyl- 2’-deoxyuridine level in blood dropped for nearly 50% after the consumption of isoflavone supplements, the plasma 8- 51 isoprostanes level, an indicator for lipid peroxidation and oxidative stress, was unaffected

(Djuric et al., 2001). Same phenomenon was also observed in Hodgson’s work, in which the urinary level of 8-isoprostanes was not influenced by soy isoflavone supplementation

(Hodgson et al., 1999). It seems that the food matrix do influence the properties of isoflavones. Based on the comparison of previous publications, soy-enhanced diet would exhibit more antioxidant activities, while isoflavone supplements made from soy extract do not. In addition, the metabolites of isoflavones also affect the antioxidant properties.

Equol, the metabolite of daidzein, displays more potent estrogenic effect than the other two isoflavones, genistein and daidzein (Morito et al., 2001). Therefore, the difference of isoflavone bio-metabolism in different populations and ages would be another reason for the inconsistence results.

3.4 Isoflavone and Safety Concerns

Despite the beneficial effect of isoflavones on human health, there have been some concerns raised over the potential adverse effects of isoflavone consumption as more biological effects of these compounds are learned. The concerns are mainly towards the soy-based infant formulas, a substitute for cow milk, as the high-exposure to isoflavones were suspected to result in adverse effect, especially due to the estrogen toxicity, in infants (Sheehan, 1998). Those concerns are not exaggerations, since previously animal studies showed the symptom of delayed development in animal infants that exposed to estrogenic compounds. It is also worth noting that infertility was reported in animals fed with high dose of isoflavones (Shutt 1974). The concentrations of

52 isoflavones in infant formula powder have varied. Setchell et al. and Murphy et al. obtained values of 316μg/g and 232 μg/g, respectively (Setchell et al 1998, Murphy et al

1997). Although Fort et al. has found the association between early-life soy-based formula feeding in kids and the autoimmune thyroid disease based on the retrospective analysis, no adverse effects of soy-based formula has been ever observed in humans, and no direct correlation has been found between isoflavone content and negative effects in infants (Miniello et al., 2003). A study developed by Strom et al indicated that there were no main differences in the comparison of the adult results of infants fed breast milk and soy-based formulas during infancy (Strom et al., 2001). To date, no detrimental effects have been confirmed and soy-based infant formulas continue to be used.

Another concern about the isoflavone intake is focused on the estrogenicity and potential harm of isoflavones for breast cancer patients and those people who are more likely to develop breast cancer, because in vitro studies have demonstrated that genistein could stimulate the breast cancer cell growth (Zava & Duwe, 1997). Also isoflavone was reported to influence thyroid function (Villar, Saconato, Valente, & Atallah, 2007), and might be associated with increased risk of thyroid cancer. However, those studies cannot prove that the isoflavone in the food matrix could be a direct harm to human, and they did not specify the physiological effect of isoflavone metabolites during the digestion. Most of the observation studies have found no adverse effect of isoflavone on cancer risk.

Overall, the current researches and data are insufficient to make absolute conclusions on the isoflavone safety issues. Further studies and clinical trials should be undertaken to safely draw the conclusions.

3.5 Analysis of Isoflavones

Analyzing the isoflavones in soy and soy food products involves several steps.

First of all, the isoflavones should be extracted from the food matrix, by either organic or aqueous organic solvents. Next, chromatographic techniques are used to separate different isoflavones, depending on their polarity. Lastly, each individual isoflavone is detected, identified and quantified by various techniques, such as UV-Vis Spectrum,

HPLC and Mass Spectrum (Lin et al., 2004; Saitoh et al., 2001).

3.5.1 Isoflavone Preparation and Extraction

Sample preparation and extraction step is critical to the isoflavone analysis, because sample loss and chemical changes may happen during the extraction, which causes misleading results. Before the extraction, internal standards are required to be added in order to precisely calculate the loss and the recovery rate of isoflavone samples.

The internal standards should be carefully selected, and meets the following requirements

(Oomah, 2002; Song et al., 1998):

1) Not be present in the sample that will be investigated.

2) Should be chromatographically separable from analytes.

3) Be stable, and will not interact with other components in the sample or food matrix

Based on previous findings, 2,4,4’-trihydroxydeoxybenzoin (THB) works well as an internal standards, since it is similar to the isoflavones, chromatographically separable from analytes, and quite resistant to heat and acid (Song, Barua, Buseman, & Murphy,

1998; Sugiyama, Sakurai, & Hirota, 2010). 54

Also, to better calculate the aglycone concentrations, isoflavone samples can be incubated with β-Glucuronidase overnight. The incubation of the enzyme will help to broke the glycoside bond and release the 3 major aglycones in soybeans. Thus, the peaks in the chromatograph will be easier to identify and quantify (Lin, Wu, Abdelnabi,

Ottinger, & Giusti, 2004).

Extraction of isoflavones has been conducted by using methanol, ethanol, acetone, and acetonitrile with water or dilute acid (Hendrich and Murphy. 2001). Naim et al extracted isoflavones from the commercially defatted soybean flakes by using 60% ethanol (Naim, Gestetner, Zilkah, Birk, & Bondi, 1974). Eldridge compared the recovery rate of isoflavones extracted from different solvents, such as 50%, 80% and 100% ethanol, methanol, ethyl acetate and acetonitrile. Maximum extraction efficiency was achieved by 80% methanol (Eldridge, 1982). Murphy also conducted a study on the isoflavone extraction efficiency, and demonstrated that acetonitrile with 0.1N HCl exhibit increased recovery rate of isoflavones (Murphy et al.1981). Therefore, 80% methanol and acidified 83% acetonitrile have become common extraction solvents for isoflavone analysis. However, the method might not work if the isoflavone samples contain acetylglucoside and malonylglucoside isomers.

Temperature is another factor that influence the isoflavone extraction efficiency.

Coward et al has compared the isoflavone extraction from soybeans in different temperatures: 80°C, 25°C and 4°C for 2-72 hours by 80% methanol. The results showed that the extraction conduced at 4°C for 2-4 hours had the highest yield of malonyl isomers of isoflavones (Coward, Smith, Kirk, & Barnes, 1998).

3.5.2 Isoflavone Separation, Identification and Quantification

Isoflavones are separated, determined and quantified by either HPLC or gas chromatography after the extraction from the samples (Ghosh & Fenner, 1999). After different isoflavones are chromatographically separated their identity and purity can be authenticated by several techniques such as UV-Vis Spectrum, HPLC retention time,

Mass Spectral (MS) Analysis, nuclear magnetic resonance (NMR) spectroscopy, and infrared (IR) spectroscopy (Hendrich and Murphy. 2001). Each isoflavone has unique

UV absorption spectra that work as a identity for the HPLC identifications by using diode array detectors (PDA). The spectra usually are composed of two maximum absorptions in the range of 245-275 and 300-330 nm (Oomah, 2002). Table 3.2 listed the prevail isoflavone UV-spectrum parameters. Also, MS analysis provides the isoflavone mass and purity information; NMR and IR spectroscopy offer additional chemical structure data for the isoflavones (Coward et al., 1998).

Analysis of isoflavones in food matrix may need authentic standards for quantification and identification. The commercial standards of isoflavones, either in aglycone form or β-glucosides and acetylglucosides are available now. But malonylglucoside standard products are still under development due to the instability.

Nevertheless, HPLC retention time, UV-spectra absorption patterns, together with the

MS spectroscopy data, could provide a fair conclusion on the isoflavone identification.

Table 3.2 Molecular weight and maximum UV absorption wavelength of selected isoflavones

Maximum UV Molecular Isoflavone absorption Weight References Name wavelength (g/mol) (nm) Genistin 432.3 262 (Lin et al., 2004) Genistein 270.2 260，328 (Rijke et al. 2006) Daidzin 416.4 260 (Nurmi et al., 2002) Daidzein 254.2 250，302 (Rijke et al. 2006) Glycitin 446.4 262 (Nurmi et al., 2002) Glycitein 284.3 262 (Nurmi et al., 2002) Biochanin A 284.3 261，326 (Delmonte et al. 2006) Formononetin 268.3 249，302 (Rijke et al. 2006)

Chapter 4 Black Goji as a Potential Source of Natural Color in a Wide pH Range

A portion of this chapter was submitted to Food Chemistry

4.1 Abstract

Lycium ruthenicum Murr is a traditional Chinese herb widely distributed in Tibet. The fruit, known as black goji, is popular in traditional Chinese medicine for disease- treatment. The objective of this study was to investigate its anthocyanin profile and the colorimetric and spectrophotometric properties. Black goji extracts contained abundant petunidin derivatives, with cis and trans isomers of petunidin-3-p-coumaroyl-rutinoside-

5-glucoside. The colorimetric and spectrophotometric traits of black goji anthocyanin were significantly impacted by solid-phase-extraction, acylation, and acyl moiety spatial configuration. MCX cartridge removed considerable polyphenolics from extracts but attenuated the saturation of color expression. The predominate petunidin-3-trans-p- coumaroyl-rutinoside-5-glucoside contributed most of the color expression and showed superior stability. Acylation strengthened the petunidin derivatives color retention and enhanced the color intensity and stability. The two isomers displayed distinctive color properties, which might be explained by their unique acyl moiety orientation. Black goji anthocyanins produced various vivid hues over wide ranges of pH, making them promising candidates for natural color source.

Keywords: Anthocyanin; Natural colorants; Petunidin; Black goji (Lycium ruthenicum

Murr); Acylation

4.2 Introduction

Union (EU) demanded the warning labels for all the food products that contain certain synthetic food colorants (CSPI, 2010). One year later, the U.S. Food and Drug

Administration (FDA) also conducted the review of scientific evidence on artificial colorants and possible association with ADHD in children (FDA, 2011). Although a causal relationship between color additive and ADHD has not been confirmed in this review, the condition of ADHD might be promoted by synthetic colorants. The increasing consumer demands for natural colorants and the current market trend of “clean label”, therefore, have driven the food industry to focus on the replacement of artificial colorants with those natural alternatives.

Despite their various advantages over synthetic dyes, natural pigments bare several weaknesses that limits their application in food. Generally speaking, naturally derived colorants are less stable compared to the artificial counterparts, and their stabilities are greatly influenced by food processing and storage conditions (Delia B.

Rodriguez-Amaya, 2016). Pigments sourced from vegetable extracts, for example red

59 cabbage, red radish, and red beet, could also impart undesirable aroma and flavor, resulting in difficulties for their direct food application (Sigurdson, Tang, & Giusti,

2017). In addition, limited choices of naturally derived pigments are available to match the color traits of synthetic ones, especially for blue and green hues (Gregory T

Sigurdson et al., 2017; Ronald E. Wrolstad & Culver, 2012). Therefore, it remains to be a huge challenge for food industry to explore, develop, and employ new sources of naturally-existing colorants that possess extraordinary stability, bear strong tinctorial strength, and cover a wide range of vivid hues without unpleasant sensory attributes.

Lycium ruthenicum Murr. is a traditional Chinese herb widely distributed in

Qinghai-Tibet plateau. The fruit, known as black goji, has a pleasant aroma and flavor, and is popular in traditional Chinese medicine for disease treatment, such as heart disease, abnormal menstruation and menopause (Jin et al., 2015; Zheng et al., 2011). Its health benefits have been associated with the antioxidant activities of the anthocyanins, which are responsible for the black-bluish color of the fruit (Potterat, 2010). (Zheng et al.,

2011) black goji fruit from Tibet contained 550-500mg anthocyanin per 100g FW, most of which (> 80%) were acylated ones. Five major anthocyanins were identified and petunidin derivative accounted for 95% of the anthocyanin content. Total polyphenol content (~1310 mg GA equivalents/100gFW) and strong antioxidant activity (~1060mg

GA equivalents/100g FW) were also reported. With plentiful acylated anthocyanins and antioxidant capacity, black goji anthocyanins seem to be good candidates for natural colorants. Although it has been well studied in terms of anthocyanin content, antioxidant

60 activity, and biosynthesis, the color properties of black goji pigments remain to be explored, especially the influence of purity, and acylation on the color expression.

Anthocyanins are an important group of water-soluble natural pigments found in fruits and vegetables. They render vivid red to blue color to plants, and have important application in coloring food products (Obón, Castellar, Alacid, & Fernández-López,

2009; Sigurdson, Tang, & Giusti, 2017). Anthocyanin colors are significantly affected by their chemical structures including chromophore methoxylation, hydroxylation, glycosylation and acylation. Generally, more hydroxyl groups on B ring leads to the bluer shift in spectrum while more methoxyl groups causes the redder shift in spectrum (He &

Giusti, 2010; Heredia, Francia-Aricha, Rivas-Gonzalo, Vicario, & Santos-Buelga, 1998).

The glycosylation at C3 and C5 not only impacts color expression and intensity, but also color retention and stability (Rakic et al., 2015; Stintzing, Stintzing, Carle, Frei, &

Wrolstad, 2002; Zhao et al., 2014). Acylation on anthocyanin structure was generally believed to enhance anthocyanin stability as the acyl group influences the pigment structure configuration and protects the chromophore from hydration, which explains the predominance of acylated anthocyanins in natural food colorants industry (M. Monica

Giusti & Wrolstad, 2003; M. Giusti & Wrolstad, 1996). Acylation substituent patterns, including various acyl moieties and attachment locations, matter in terms of anthocyanin colorimetric and spectrophotometric properties (Ahmadiani, Robbins, Collins, & Giusti,

2016).

The objective of this study is to investigate the profile, color characteristics, and stability of the anthocyanins in black goji, as well as the influence of solid-phase

61 extraction, acylation and acyl group spatial configuration on the colorimetric and spectrophotometric properties, providing a new potential natural color source for the food industry.

4.3 Materials and Methods

4.3.1 Materials & Reagent

Black goji pigments were extracted from dried black goji berries that were purchased from a grocery store (LianHua Supermarket) in Shanghai, China.

The chemicals and reagents (ACS or HPLC grade) were purchased from Fisher

Scientific (Fair Lawn, NJ USA), including acetone, chloroform, methanol, trifluoroacetic acid (TFA), ammonium hydroxide (NH4OH), acetonitrile (HPLC and LC/MS grade), potassium hydroxide (KOH), citric acid, sodium phosphate dibasic (Na2HPO4), sodium phosphate monobasic (NaH2PO4), sodium carbonate (Na2CO3), sodium bicarbonate

(NaHCO3), potassium chloride, and sodium acetate. ACS grade ethyl acetate and formic acid were obtained from Mallinckrodt Chemicals (Bedminster Township, NJ USA) and

Honeywell (Morris Plains, NJ USA), respectively.

4.3.2 Pigments Extraction

The extraction of black goji anthocyanins by acetone-chloroform partitioning method were based on Rodriguez-Saona & Wrolstad, (2005). Dried black goji berries

(~50g) were mixed with liquid nitrogen, and powdered by using a food blender. The

62 powders were then mixed with 100%(v/v) aqueous acetone with 0.01% HCl. The slurry was filtered through filter paper (Whatman No.4 filter paper, Whatman Incorporation.

NJ, US) by water vacuum. After filtration, re-extraction was performed by adding

70%(v/v) aqueous acetone with 0.01% HCl until the slurry becoming faded. Two volumes of chloroform were then added into the aqueous acetone solution, and the whole filtrate was transferred into the separatory funnel. After the gentle mixing, samples were stored in 4°C refrigerator overnight. On the next day, after transferring the upper aqueous portion to flask, the acetone and chloroform were evaporated by rotary evaporator.

4.3.3 Pigments Purification

The C-18 cartridge purification method was adapted from Rodriguez-Saona &

Wrolstad, 2005. Aqueous anthocyanin extracts obtained from the above procedure were pumped to pass through the Sep-Pak® C18 cartridge (Waters Corporation. Milford MA,

USA) after activating the cartridge with methanol. Then the cartridge was washed with two column volumes of acidified water (0.01% v/v HCl), and anthocyanin was eluted into a boiling flask by adding 0.01% v/v HCl acidified methanol. After the removal of the methanol by rotary evaporator, the purified anthocyanin samples were re-dissolved by acidified distilled water (0.01% v/v HCl).

The remaining phenolics in the C-18 purified anthocyanin sample could be further removed by Oasis® MCX extraction cartridge (Waters Corporation. Milford MA, USA), a novel solid-phase extraction method introduced by He & Giusti, 2011. The MCX cartridge was first activated by two columns of 0.1% TFA methanol and water, 63 respectively. Then samples were loaded onto the activated MCX cartridge, and two washes of 0.1% TFA water followed by 0.1% TFA methanol were performed to remove the salts, sugars, and other phenolics. The anthocyanin samples were then recovered with

1% NH4OH in methanol, and collected in a round flask containing formic acid that helps to neutralize the alkaline condition of eluent.

4.3.4 Pigments Saponification

Alkaline hydrolysis (saponification) was performed to cleave the ester bond between acyl group and anthocyanin glycoside (Giusti, Rodríguez-Saona, & Wrolstad,

1999). Purified anthocyanin samples were mixed with 10 ml 10% aqueous KOH in a capped test tube for 12 min. Then 2N HCl was added to the mixture to adjust the pH back to mildly acidic conditions. Samples were then purified by C-18 cartridge to remove acids as described above.

4.3.5 Pigments Identification and Isolation

Individual anthocyanin pigments in extracted and purified samples were then analyzed and identified by high performance liquid chromatography (HPLC) (Shimadzu,

Columbia, MD) coupled to a SP-M20A Photodiode Array Detector (Shimadzu,

Columbia, MD) and a LCMS-2010EV Liquid Chromatograph Mass Spectrometer. A reverse phase Symmetry C-18 (5μm, 4.6*150mm) column (Phenomenex, Torrance, CA

US) was used. All of the extracts were filtered through a 0.22 um syringe filter

(Phenomenex, Torrance, CA US) before injection into the HPLC. Samples were analyzed using a flow rate of 0.8 ml/min. The mobile phase consisted of solvent (A) 4.5% (v/v) formic acid and solvent (B) 100% acetonitrile. The linear gradient used in the analysis was from 8% B 0min-5min, 8%-15% B 5min-35min, 15%-35% B 35min-37min, 35%-

8% B 37-40min, 8% B 40-45min. Anthocyanin and all phenolics elutions were monitored at 500-530nm and 280-700nm, respectively. Total ion scans and selected ion monitoring

(Mass/charge ratios of 271, 287, 303, 301, 317 and 331, corresponding to the most common anthocyanin aglycones) were conducted.

The major pigment, petunidin-3-trans/cis-p-cou-rut-5-glu, was isolated by Luna reverse-phase PFP column (5μm particle size and 100 Å pore size with 250*21.20 nm column size, Phenomenex, Torrance, CA US) and semi-prep reverse-phase HPLC

(Shimadzu, Columbia, MD) composed of pumps (LC-6AD), autosampler (SIL-20A HT), column oven (CTO-20A), Photodiode Array Detector (SPD-M20A), and communication module (CBM-M20A). The flow rate was 10 ml/min, with mobile phase consisted of solvent (A) 4.5% (v/v) formic acid and solvent (B) 100% acetonitrile. The linear gradient used for pigment isolation was from 12% B 0min-2min, 12%-21% B 2min-25min, 21%-

21% B 25min-30min, and 21%-30% B 30-50min. For non-acylated (saponified) petunidin derivative, the gradient was from 10% B 0min-1min, 10%-15% B 1min-31min.

Each pigment was collected manually based on the real-time absorbance over 280-

700nm. The collected pigments were purified by C-18 cartridge and their identification and purity were confirmed by Kinetex reverse-phase PFP column (2.6μm particle size

65 and 100 Å pore size with 100*4.6nm column size, Phenomenex, Torrance, CA US) and the same pigment identification method as described above.

4.3.6 Pigments Quantification

The monomeric anthocyanin content was determined by the pH differential method (Mónica Giusti & Wrolstad, 2005). Buffer solutions were prepared using 0.1M potassium chloride at pH 1.0 and 0.4M sodium acetate at pH 4.5. Absorbance of samples at pH 1.0 and 4.5 were measured at 700nm and its λmax (512nm) by using UV-vis

Spectrophotometer (Shimadzu corporation. Tokyo, Japan). Measurements were done in triplicates, and each black goji anthocyanin extracts or isolates was expressed as cyanindin-3-glucoside equivalence.

4.3.7 Buffer System and Sample Preparation

The buffer systems in this study were citric acid-Na2HPO4 buffer solutions for pH

3-7; Na2HPO4-NaH2PO4 buffer solution for pH8; Na2CO3-NaHCO3 buffer solutions for pH9-10 (Dawson, Elliott, Elliott, & Jones, 1986). All the black goji extracts and isolates were diluted in these buffer solutions (pH3-10) at concentration of 25 μM. The pH condition after mixing was confirmed using pH meter (Mettler Toledo Inc, Columbus,

OH US). The initial spectrophotometric and colorimetric measurement were performed after 1 hour equilibrium. Samples were stored at refrigerated condition in the dark for three weeks for stability tests. Analysis was done in triplicates.

4.3.8 Spectrophotometric Analysis

After the mixture of black goji samples with various buffer systems, 250 μL of each sample was aliquoted to poly-D-lysine coated polystyrene 96 well plates, and the spectrums were analyzed from 380 nm to 700 nm with 1 nm interval by using a

SpectraMax 190 Microplate Reader (Molecular Devices, Sunnyvale CA). The spectrophotometric analysis was conducted 1hour after mixture, as well as on day 1, 2, 3,

7, 14, and 21.

4.3.9 Colorimetric Analysis

The spectral absorbance data for each sample from 380 nm to 700 nm as described above were converted into colorimeteric data using the ColorBySpectra software (according to CIE 1964 standard observer, D65 illuminant spectral distribution, and 10° viewer angle) (Farr, JE; Srivastava, A; Machiraju, R; Giusti, 2017). The samples were kept in refrigerated condition at 4°C in the dark for three weeks. The color changes were measured as ΔE.

4.3.10 Statistical Analysis

The tables and corresponding statistical analysis were presented by using Prism software (GraphPad, La Jolla, CA US). One-way ANOVA (two-tailed, α=0.05) and post hoc Tukey’s test (family-wise α=0.05) were conducted to evaluate the differences in L,

C*ab, h ab values among different black goji anthocyanin extracts/isolates at a certain pH condition.

4.4 Results and Discussion

The anthocyanin profiles, colorimetric & spectrophotometric properties, and stabilities were compared among various black goji anthocyanin extracts or isolates. In the following sections, “crude extract” refers to the black goji anthocyanin extracts without any solid-phase purification procedures; “C-18 purified extract” and “MCX purified extract” indicate the extracts purified by C-18 cartridge and C-18 & MCX cartridges, respectively; “cis isomer” and “trans isomer” represent the two isolated petunidin-derivative isomers: petunidin-3-cis-p-coumaroyl-rutinoside-5-glucoside and petunidin-3-trans-p-coumaroyl-rutinoside-5-glucoside; and “SPO” stands for the petunidin-3-rutinoside-5-glucoside, which was produced from the saponification of the trans isomer.

4.4.1 Identification of Anthocyanin Profiles in Black Goji

Four major anthocyanins accounting for 97% of the total peak area were separated based on the chromatogram prepared by RP-HPLC-PDA-MS (Figure 4.1). The relative peak area of the four pigments (peak 1 to peak 4) was approximately 18%, 4%, 7%, and

71%, respectively. Two aglycones, delphinidin (m/z 303) and petunidin (m/z 317) were found in black goji, between which petunidin was the most abundant (93%). According to the MS data, the majority of the black goji anthocyanins in crude extracts were 68 acylated (80%), and they were identified to be petunidin-3-galactoside-5-glucoside (peak

1), delphinidin-3-trans-p-coumaroyl-rutinoside-5-glucoside (peak 2), petunidin-3-cis-p- coumaroyl-rutinoside-5-glucoside (peak 3), and petunidin-3-trans-p-coumaroyl- rutinoside-5-glucoside (peak 4), in agreement with the identification of predominant pigments in a previous study (Zheng et al., 2011).

Petunidin-3-trans-p-coumaroyl-rutinoside-5-glucoside accounted for almost 71% of the total anthocyanins in black goji (peak 4 in Figure 4.1), which was barely reported in berry fruits (X. Wu & Prior, 2005). Taking into account the abundancy in black goji, it is reasonable to believe that this pigment would be representative in terms of overall black goji color expression and other properties including color stability (as described later). Besides, it is interesting to find both cis and trans petunidin isomers (relative peak area was 7% and 71%) with same structure block but different spatial configuration (peak

3 and peak 4 in Figure 4.1). Acyl groups in acylated anthocyanins are typically trans configured, and there were few cis form identified in nature (George et al., 2001;

Hosokawa, 1995; Ichiyanagi et al., 2005). Thus, the co-existing of both plentiful trans and cis isomers from same source is limited. To this end, black goji may serve as a good subject to study the impact of acyl spatial configuration on anthocyanin color expression.

Acylation on the anthocyanin structure is generally believed to enhance anthocyanin stability since the acyl group could alter the pigment structure configuration and protect the chromophore from hydration, which explains the predominance of acylated anthocyanins in food natural colorants industry (M. Monica Giusti & Wrolstad,

2003; M. Giusti & Wrolstad, 1996). Unlike other acylated anthocyanins which are

69 commonly found in large amounts in vegetables such as red cabbage (containing acylated cyanidin derivatives) and red radish (containing acylated pelargonidin derivatives), black goji as fruit are rich in acylated anthocyanin. Some scholars postulated that this high- level content of acylated pigments in black goji was due to the extreme altitude, ultraviolet rich environment, and harsh weather in Tibet. The abundant acylated anthocyanins would protect plants from the tough conditions (Dixon & Paiva, 1995).

Nevertheless, it is a promising advantage of black goji in that the vegetable extracts usually carry unique aroma and flavors while black goji extracts do not.

The two acyl isomers were isolated (labeled as “cis/trans isomers” in Figure 4.1) to compare the spectrophotometric and colorimetric characteristics. In addition, saponification of the petunidin-3-trans-p-coumaroyl-rutinoside-5-glucoside yielded the petunidin-3-rutinoside-5-glucoside (labeled as “SPO” in Figure 4.1) as a non-acylated anthocyanin comparison.

Figure 4.1 Chromatogram of black goji anthocyanin extracts and isolates at 520 nm and 280-700nm, and their identifications

4.4.2 Effect of Purification on Anthocyanin Profiles in Black Goji

The behavior of pigments can be greatly affected by surrounding compounds, as other phenolic including phenol, phenolic acid, flavone, and tannin could significantly influence colorimetric properties as well as pigment stabilities through intermolecular co- pigmentation. (Fossen et al., 2007; Gómez-Míguez et al., 2006; Kunsági-Máté et al.,

2006; L. A. Pacheco-Palencia & Talcott, 2010). Anthocyanin-rich vegetables and fruits generally contain abundant polyphenols and lipids, which complicated the qualitative and quantitative analysis of anthocyanin (Rodriguez-Saona & Wrolstad, 2001). Thus, purity is a critical factor in pigment analysis. Unfortunately, the commonly used extraction methods are not specific for anthocyanin, and a solid-phase purification procedure would be necessary to exclude the extraneous interfering compounds. In this study, two resins were used to accomplish this goal: C-18 and MCX cartridges. The former one contains

18 carbon-chains that are bonded on a silicon base. Hydrophobic organic compounds such as anthocyanin and phenolic compounds would retain on the cartridge while acids and sugars would be washed out by acidified water. Phenolics other than anthocyanin then could be partially removed using ethyl acetate wash (Oszmianski & Lee, 1990;

Rodriguez-Saona & Wrolstad, 2001). MCX cartridge was implemented for anthocyanin purification by He & Giusti, 2011. Basically, this isolation method used cation-exchange and reversed-phase mechanisms to isolate positive charged anthocyanin. It was found to exhibit higher anthocyanin selectivity than other solid phase extraction methods.

The crude extract, C-18 purified extract, and MCX purified extract were investigated regarding their anthocyanin and polyphenolics profiles, which were 72 monitored at 520 nm and 280-700 nm, respectively. As demonstrated in Figure 4.1, the

C-18 cartridge purification removed some of the polyphenolics from the crude extract, but there were no huge differences in the chromatogram as compared to that of crude extract. A following wash with the MCX cartridge significantly eliminated polyphenolics, demonstrating MCX as a powerful tool for anthocyanin purification. As expected, anthocyanin profiles and content before and after the above purifications steps were intact since the proportions of peaks remained unchanged.

4.4.3 Colorimetric Properties of Black Goji Anthocyanin Extracts and Isolates

Generally, all the black goji anthocyanin extracts (crude extract, C-18 purified extract, and MCX purified extract) and isolates (trans isomers and SPO) exhibited a similar “red-purple-blue” pattern of color expression as pH increased from acidic to alkaline (Figure 4.2). Red hue colors were found in acidic condition, and then they gradually became colorless in mildly acidic condition as the anthocyanin structure undergoes pH-dependent transformation from red flavylium cation form to colorless carbinol pseudobase. These extracts and isolates exhibited various vivid purple, blue, and greenlish-blue colors in neutral to alkaline pH environments, where anthocyanin is predominant in quinoidal base form (Raymond Brouillard & Delaporte, 1977a). At pH

10, all the extracts quickly faded in color, except the isolated trans isomers which still retained an attenuated tinctorial strength. Although the overall pattern “red-purple-blue” was the same, the colorimetric properties of black goji anthocyanins were entirely influenced by purification procedures, and acylation, in that there were pronounced 73 discrepancies among these extracts and isolates throughout the whole pH ranges tested

(Figure 4.2).

Due to the unknown molar absorptivity of the corresponding individual black goji anthocyanin, they were quantified as cyaniding-3-glucoside equivalent. Thus, a divergence in L* (lightness) and C*ab (chroma) among extracts and isolates would be expected due to the different dilution factors for each pigment; but hab (hue angle) would still be comparable since hab was less variable in a small range of concentration. The colorimetric data are presented in Table 4.1 and Figure 4.2.

The differences in L*, C* ab and hab among crude extract, C-18 purified extract,

MCX purified extract, and trans isomer were small in acidic and mildly acid conditions from pH 3 to pH6. But this dissimilarity became very distinct between pH 7 and pH 9.

As black goji passed through MCX cartridges, L* value increased and C*ab value decreased compared to that of crude extract, resulting in more pale and dull hues.

However, there was no significant difference before and after C-18 cartridge purification in terms of the colorimetric property. It could be explained based on the chromatogram, as the polyphenolics and anthocyanin contents remained almost the same between the crude extract and C-18 purified sample, while the MCX cartridge considerably removed the interfering polyphenolics which were believed to impact anthocyanin color properties and stability by means of inter-molecular co-pigmentation (Gómez-Míguez et al., 2006;

Malien-Aubert et al., 2001). The removal of these polyphenolics would predictably alter the pigments color expression and other traits. On the other hand, the trans isomer displayed decreased L* and increased C*ab values compared to that of MCX purified

74 extract, leading to a similar vivid and intensified purple or blue hues as crude extract.

This similarity demonstrated the role of the trans isomer as the major pigment in black goji and its leading contribution for black goji color expression. The trans isomer was even able to express color at pH 10 while other extracts faded away in color quickly after mixture. Nevertheless, the hue angle (h*ab) remained unchanged among all extracts and trans isomer.

A comparison between trans isomer and SPO demonstrated the importance of acylation in anthocyanin color expression (Table 4.1 and Figure 4.2). The two isolates exhibited red hues in acidic and mildly acidic conditions, but acylation helped to strengthen and intensified the red color, as the C*ab values of trans isomer were larger than those of SPO. Compared to SPO, acylated petunidin suffered less color loss when pH arose from 3 to 6. That is because the acyl group could form a folded structure, intramolecular co-pigmentation, to protect the chromophore from hydration, resulting in increased pKh (hydration constant) and thus strengthening the color retention; On the contrary, SPO without acyl moiety protection would be more prone to hydration and structurally transformed into colorless carbinol pseudobase forms (George et al., 2001;

M. Monica Giusti & Wrolstad, 2003). When pH increased to neutral and alkaline conditions, huge divergence in the h*ab value was found. The trans isomer showed purplish blue hues (267°-306°), whereas SPO expressed greenish blue hues (218°-274°).

To this end, anthocyanin colorimetric properties were greatly influenced by the acylation on its structure.

Figure 4.2 Color expression of black goji anthocyanin extracts and isolates at pH 3-10

Table 4.1 Colorimetric (CIE-L*, C*ab, hab) and spectrophotometric (λmax) data of black goji anthocyanin extracts and isolates. Values presented are means (n=3) and (SD)

pH3 pH4 pH5 pH6 pH7 pH8 pH9 pH10 90.7 93.1 93.8 82.2 70.4 68.0 68.9 76.6 Crude (2.1) a (2.1) a (1.4) a (2.1) b (1.4) c (1.1) d (2.3) c (1.3) c

90.4 93.0 93.6 81.9 70.2 68.1 69.7 77.1 C-18 (3.1) a (1.1) a (1.1) a (1.1) b (0.1) c (1.2) d (1.4) c (1.1) c 93.2 95.4 96.0 95.8 88.7 85.9 87.2 94.9 MCX (1.1) a (1.2) a (1.1) a (1.1) a (2.2) a (1.2) a (1.1) a (1.2) a 91.7 94.8 95.5 92.7 82.1 82.8 83.3 83.8 cis (0.7) a (0.2) a (0.2) a (0.5) a (0.7) b (0.8) b (1.1) b (0.1) b 93.7 95.0 95.8 95.2 85.1 82.3 82.2 91.2 L* (lightness) trans (2.3) a (1.2) a (1.2) a (1.3) a (2.1) b (1.1) b (1.1) b (1.1) a 93.9 96.0 96.3 95.7 83.0 78.3 79.9 93.6 SPO (0.3) a (0.1) a (0.1) a (0.4) a (1.7) b (2.0) c (1.9) b (0.5) a 11.1 4.9 3.5 33.5 41.4 38.7 32.4 20.8 Crude (1.1) a (0.5) b (1.1) a (1.5) b (1.9) a (1.9) a (1.3) a (1.7) b

11.4 5.0 3.7 39.7 40.7 36.4 24.8 27.7 C-18 (0.9) a (0.5) b (1.1) a (1.7) a (1.4) a (1.1) a (1.9) b (1.1) a 10.6 8.2 2.0 19.9 28.4 28.5 27.5 13.8 MCX (2.1) a (0.3) a (1.3) ab (1.3) d (1.2) b (1.6) b (2.3) b (0.8) c 10.3 2.4 1.2 1.2 15.7 26.3 26.0 22.7

(Chroma)

cis (1.4) a (0.2) c (0.1) b (0.8) e (0.7) c (1.1) b (2.0) b (0.2) b

ab 11.9 2.7 1.4 27.8 38.2 34.4 36.4 26.1

C* trans (1.1) a (0.1) c (0.9) ab (1.1) c (1.1) a (1.1) a (1.2) a (0.1) a 6.4 2.1 0.9 0.8 13.4 25.8 23.6 10.8 SPO (0.5) b (0.2) c (0.1) b (0.2) e (1.8) c (2.0) b (2.2) b (1.3) c 350.0 16.5 30.0 330.8 307.5 276.8 261.0 279.6 Crude (2.9) a (2.4) c (1.8) c (1.8) b (2.9) a (1.8) a (1.2) b (2.9) a

350.2 16.5 30.1 327.7 306.6 275.6 259.3 282.6 C-18 (2.4) a (1.8) c (2.4) c (1.7) a (1.7) a (1.4) a (1.8) b (1.4) a 348.9 16.5 30.1 330.3 304.8 271.5 260.4 262.2 MCX (3.2) a (1.1) c (2.7) c (1.6) b (1.3) a (1.3) b (2.1) b (1.8) c 344.1 3.3 52.6 336.4 277.9 214.5 211.4 206.6 cis b d b a b d d d

(Hue angle)(Hue (0.9) (3.4) (1.3) (1.2) (0.3) (0.0) (0.3) (0.3)

ab 343.5 123.8 39.3 330.6 306.0 275.8 267.0 266.8

h trans (2.5) b (1.3) a (1.3) c (1.7) b (1.1) a (1.2) a (1.2) a (2.1) b 5.3 60.2 85.2 76.9 274.2 233.5 218.3 96.3 SPO (1.8) c (2.0) b (3.5) a (0.3) c (1.2) b (1.0) c (0.4) c (0.8) e 525 527 536 537 557 575 578 Crude NA (0) b (0.7) b (0.7) a (0.7) a (0) b (0) d (0) c 525 527 536 539 558 578 578 C-18 NA (0.7) b (0) b (0.7) a (1.4) a (0.7) b (0) c (0) c 525 527 536 539 558 578 578 MCX NA (1.4) b (0) b (0) a (0) a (0) b (0) c (0) c

max

λ 531 532 531 534 566 628 630 628 cis (1.0) a (1.6) a (3.2) b (1.0) b (3.2) c (0) a (0.6) a (0.6) a 523 529 535 540 558 578 578 578 trans (0) b (0) b (1.2) a (0.7) a (0) b (0) c (1.2) c (0) b 519 514 511 582 604 604 SPO NA NA (1.2) c (0.6) c (1.9) c (2.1) a (0.6) b (1.2) b 77

A comparison among cis, trans isomers and SPO demonstrated the important roles of acylation and the acyl moiety spatial configuration in anthocyanin color expression (Table 4.1 and Figure 4.2). All of the above three isolates exhibited red hues in acidic and mildly acidic conditions, but acylation helped to strengthen and intensified the red color as the C*ab values for both cis and trans isomers were larger than that of

SPO. In these pH environment, the acylated cis and trans isomers showed pinklish red hues (3.3° - 330°) while SPO exhibited yellowish red hue (5.3° - 85°). Compared to SPO, acylated ones suffered less color loss when pH arises from 3 to 6. That is because the acyl group could form a folded structure, intramolecular co-pigmentation, to protect the chromophore from hydration, resulting in increased pKh and thus strengthen the color retention; while SPO, without acyl moiety protection, would be more prone to hydration and structurally transformed into colorless carbinol pseudobase forms (George et al.,

2001; M. Monica Giusti & Wrolstad, 2003). In addition, trans isomer was superior than cis isomer in terms of color intensity from pH 4 to pH 6. When pH increases to neural and alkaline conditions, huge divergence in the h*ab value was found. The trans isomer showed purplish blue hues (267°-306°), whereas cis isomer and SPO expressed similar greenlish blue hues (211°-278° and 218°-274°, respectively). Based on the above phenomenon, anthocyanin colorimetric properties are not only influenced by the acylation but also by the acyl moieties spatial configuration. The dissimilarity in color expression behaviors of cis and trans isomers (as well as the corresponding difference in visible light absorption spectra that will be discussed the following section) could possibly demonstrate their different impacts on anthocyanin structure including

78 intramolecular co-pigmentation, structure distortion, and electronic charge density along the chromophore (George et al., 2001). But this diversity would empower the two isomers as promising blue and greenlish blue natural color sources for food application.

4.4.4 Spectrophotometric Properties of Black Goji Extracts and Isolates

Corresponding to the colorimetric data as discussed above, the spectrophotometric properties of black goji extracts were dependent on purity and pH (Figure 4.3 and Table

4.1). At pH 3, the visible light absorption spectra of all the extracts and trans isomer displayed a unique single peak with the maximum absorbance found around 525 nm.

Besides, the -3,5-glycosylation of black goji anthocyanin was well recognized by low ratio of A440/Avis-max (Durst & Wrolstad, 2005). In mildly acidic condition from pH 4 to 6, the spectra flattened because of the anthocyanin structure hydration and the formation of colorless carbinol pseudobase. As pH further ascended to neutral and alkaline condition, bathochromic and hyperchromic shifts were observed, with the λmax ranged from 558 nm at pH 7 to 578 nm at pH 9. While the MCX purified extract showed a similar absorption intensity at pH 3 as other extracts, a significant diminution in absorption was found from pH 7 to pH 9, corresponding to the relative pale and dull colors as described in previous section. All the extracts displayed a sharp peak in spectra when pH is greater and equal to 7, matching the vivid and highly intensified hues. Interestingly, a distinctive peak shoulder existed right to the λmax at around 640 nm for all the extracts containing petunidin-3-trans-p-coumaroyl-rutinoside-5-glucoside, while it was absented in SPO.

Figure 4.3 Spectrophotometric characteristics of black goji anthocyanin extracts and isolates at pH 3-10.

There was an increasingly significant incongruity between trans isomer and SPO considering their visible light absorption spectra as pH arose from 3 to 9 (Figure 4.3 and

Table 4.1). The non-acylated petunidin derivative was documented by a single peak with

λmax being around 525 nm at pH 3, and a broad peak with λmax > 600 nm above pH 8. The trans isomer, however, was characterized by its sharp peaks in the spectrum, which was linked to its vivid color expression. While the SPO exhibited greater λmax than the trans isomer in alkaline conditions, it presented less λmax throughout pH 3 to pH 6. Overall, the variations in λmax between the two isolates were small in acidic condition, but became considerable when pH was greater than 6.

It is very interesting that anthocyanin with the same structure block but slight different spatial configuration (cis and trans isomers) would have brought such a significant variance in terms of colorimetric and spectrophotometric properties. A previous research (George et al., 2001) studied trans and cis isomers of delphinidin-3-p- coumaroyl-glucopyranoside-5-malonyl-glucopyranoside found in Triteleia bridgeseii.

They concluded that the cis isomer of this delphinidin derivative had p-coumaroyl residue parallel to the aglycones while the trans isomer was in the shape of a quasi-perpendicular conformation. Moreover, the distance to the aglycones was smaller for the cis acyl group than the trans one, which caused difference in hydration equilibrium constant pKh and the corresponding color retention in mildly acidic conditions. It seems reasonable to assume that this tiny configuration disparity might prominently impact the whole molecular folding (namely the intramolecular co-pigmentation) and electronic charge density distribution, leading to diverse colorimetric and spectrophotometric properties.

4.4.5 Comparison of Color Stabilities among the Black Goji Anthocyanin Extracts and

Isolates

All the extracts and isolates were stored at refrigerated condition in dark for three weeks, and the color changes were expressed as ΔE (Figure 4.4 and Figure 4.5). Black goji anthocyanin color stability was greatly influenced by its composition, acylation, and pH. Typically, black goji anthocyanins are stable at an acidic environment, and became vulnerable to degradation as pH increases. At pH 3 and 4, all of the extracts and isolates remained almost the same color in that ΔE’s were < 5 throughout the testing time frame.

At neutral and alkaline pH, the discrepancy enlarged. As expected, SPO degraded in the fastest pace among all the samples, while acylation significantly boosted stability from pH 7 to 9. Compared to other black goji extracts, the isolated trans isomer was extraordinary stable at alkaline condition as its purple-blue hue endured up to three weeks, illustrating the influential role of plant polyphenolics in anthocyanin color stability. The crude extract, C-18 purified extract, and MCX purified extract shared similar stability pattern, probably due to their similar anthocyanin profiles.

It is noteworthy to find that the stabilities of black goji extracts and isolates were superior at pH 8 than that in any other pH environment, which might be explained by their pH-dependent transformation in structure configuration. As pH arises from acidic to mildly acidic or neutral conditions, the red flavylium cation form either undergoes hydration and turns into colorless carbinol pseudobase, or bears deprotonation and exists in blue-purple quinoidal base form. As the pH further increases, the quinoidal base form 82 could be ionized and becomes one or two negatively charged forms (Raymond Brouillard

& Delaporte, 1977a; He & Giusti, 2010c). A previous study has reported that the pKa2 and pKa3 (dissociation constants for the structure transformation from quinoidal base to one negatively charged form and from one to two charged form, respectively) of petunidin aglycones were pH 6.99 and 8.27 (León-Carmona, Galano, & Alvarez-Idaboy,

2016). Although these two numbers would not be the same as the dissociation constants of petunidin-derivative pigments in black goji, it is still rational to postulate that a larger proportion of two negatively charged quinoidal base forms exists at pH 8, and therefore the isolated pigments were more resistance to degradation in this condition. Nevertheless, the isolated trans isomers showed excellent stability at alkaline condition, demonstrating its potential capability as various stable blue-hues natural pigments in food application.

Figure 4.4 Color changes of black goji extracts and isolates that were stored under refrigerated condition in dark within three weeks testing period.

5 pH 3 5 pH 4

4 4

3 Ra3w Extract Raw Extract C-18

E C-18

Δ Δ C-18 & MCX C-18 & MCX 2 Ci2s isomer Cis isomer Trans isomer Trans isomer 1 SPO1 SPO

0 0 0 5 10 15 20 25 0 5 10 15 20 25 Time (days) Time (days)

30 pH 7 50 pH 8

20 Ra30w Extract Raw Extract C-18 C-18

Δ Δ C-18 & MCX C-18 & MCX Ci20s isomer Cis isomer 10 Trans isomer Trans isomer SPO10 SPO

0 0 0 5 10 15 20 25 0 5 10 15 20 25 Time (days) Time (days)

40 pH 9 50 pH 10

40 30

Ra30w Extract Crude Extract C-18 C-18 E 20 E

Δ Δ C-18 & MCX MCX Ci20s isomer cis isomer Trans isomer trans isomer 10 SPO10 SPO

0 0 0 5 10 15 20 25 0 5 10 15 20 25 Time (days) Time (days)

Figure 4.5 Color changes (described as ΔE) of black goji extracts and isolates at pH 3,4,7,8,9, and 10. Samples were stored under refrigerated condition in dark within three weeks testing period

4.5 Conclusion

Black goji extracts contained abundant petunidin derivatives, with cis and trans isomers of petunidin-3-p-coumaroyl-rutinoside-5-glucoside. The colorimetric and spectrophotometric traits of black goji anthocyanin were greatly influenced by purification procedures, acylation, and acyl moiety orientations. MCX cartridge removed considerable polyphenolics from fruit extracts, and attenuated the saturation of color expression. The predominant petunidin-3-trans-p-coumaroyl-rutinoside-5-glucoside contributed most of the black goji anthocyanin color properties and showed a better color stability compared to other extracts over time. Acylation not only strengthened the color retention in mildly acidic condition, but also enhance the tinctorial strength and stability of pigments. The distinctive difference in the colorimetric and spectrophotometric properties of the two isomers might be explained by their unique acyl moiety spatial configuration. Nevertheless, this dissimilarity shed light on the chemical attributes that could be used to manipulate the color expression in natural pigment application.

This study demonstrated that black goji is a promising source for natural colorants, producing various vivid hues over a wide range of pH. Its cis and trans petunidin isomers, with strong stability and attracting vibrant hues, would broaden the choices of natural pigments in food industry, catering to the current trends of shifting from artificial colorant to natural alternatives.

Chapter 5 UV-Induced trans-to-cis Isomerization of Coumaric Acid in Petunidin- derivative

5.1 Abstract

Acyl spatial configuration plays an important role in anthocyanin colorimetric properties.

Petunidin-3-cis-p-cou-rut-5-glu from black goji was shown to express bluer hues than its trans counterpart. However, only 12 cis acylated anthocyanins were reported in reviewing the current literatures. The objective of this study was to investigate the controlled conversion of trans to cis petunidin-derivative by UV-irradiation, providing an alternative method to obtain cis configured anthocyanins. Petunidin-3-trans-p-cou-rut-5- glu was isolated and purified from black goji, and dissolved in 100% methanol (10-3M) in quartz and glass cuvettes. The pigment identification and its colorimetric and spectrophotometric properties were investigated over UV irradiation (254 nm) up to 180 min (45.7J). UV light triggered the formation of petunidin-3-cis-p-cou-rut-5-glu from its trans counterpart. The trans isolates in quartz and glass cuvettes required 7.64J and

30.48J to reach a plateau phase with cis : trans ratio of 4:6 and 55:45, respectively. The isomerization caused decrease in hue angles, leading to bluer hues at neutral and alkaline conditions.

Keywords: Natural Color; Isomers; cis/trans-acylation; Lycium ruthenicum Murr; UV- irradiation

5.2. Introduction

Anthocyanins are water soluble pigments that impart vivid red-violet-blue colors to various vegetables and fruits. They have becoming increasingly popular among consumers due to its potent antioxidant activities and current ongoing “clean label” trends in food industry. Consumer attitudes and regulatory changes are driving the transition from synthetic colors with no nutritional value to natural pigments (such as anthocyanins) that possess health enhancing properties. Incorporation natural pigments into food matrix will lead to cutting-edge food product development, and ensure a future market share.

Anthocyanin colors are significantly affected by its chemical structures including chromophore methoxylation, hydroxylation, glycosylation and acylation. Generally, more hydroxyl groups on B ring leads to the bluer shift in spectrum while more methoxyl groups causes the redder shift in spectrum (He & Giusti, 2010; Heredia, Francia-Aricha,

Rivas-Gonzalo, Vicario, & Santos-Buelga, 1998). Acylation on anthocyanin structure was generally believed to enhance anthocyanin stability as the acyl group influences the pigment structure configuration and protects the chromophore from hydration, which explains the predominance of acylated anthocyanins in natural food colorants industry

(M. Monica Giusti & Wrolstad, 2003; M. Giusti & Wrolstad, 1996). In addition, acylation substituent patterns, including various acyl moieties and attachment locations, could affect anthocyanin colorimetric and spectrophotometric properties (Ahmadiani,

Robbins, Collins, & Giusti, 2016).

Cis and trans acylated anthocyanins were found together in nature, for example black goji (Lycium ruthenicum Murr.), Chinese eggplant (Solanum melongena L.), and

88 purple bell peppers (Capsicum annuum L.) were reported to contain both cis and trans petunidin or delphinidin derivatives (Ichiyanagi et al., 2005; Inami, Tamura, Kikuzaki, &

Nakatami, 1996; Jin et al., 2015; Sigurdson & Giusti, 2014; Yoshida, Kondo, Kameda, &

Goto, 1990). In previous chapter, it has been shown that the spectrophotometric and colorimetric properties of the petunidin-3-p-cou-rut-5-glu in black goji were significantly influenced by the stereochemistry of the p-coumaric acid. The cis isomer showed greater

λmax in a wide range of pH, expressing bluer hues than the trans isomer. The trans isomer, on the other hand, exhibited improved color stabilities over time. Thus, the cis and trans acylated anthocyanins could be used in food products depending on diverse purposes, either for bluer color or better color stability.

However, the trans isomer always dominated in nature, not only in occurrence but also in quantity. In fact, we only found 12 cis acylated anthocyanin derivatives reported in current studies (George et al., 2001; Hosokawa, 1995; Ichiyanagi et al., 2005; Jin et al.,

2015). Thus, it would benefit the food industry if an efficient way could be provided to obtain cis acylated anthocyanins.

Photo-irradiation was previously reported to induce isomerization from trans to cis isomers in cyanidin derivatives from the purple leaves of Perilla ocimoides (Yoshida,

Kondo, Kameda, & Goto, 1990). In addition, most of the naturally occurring cis acylated anthocyanins are found in the portion of plants receiving large amount of light.

Therefore, these findings led us to study the role of UV irradiation in controlling the trans to cis isomerization in petunidin-derivative.

The objective of this study was to investigate the controlled conversion of trans to cis petunidin-derivative by UV-irradiation, providing an alternative method to obtain cis isomeric anthocyanins.

5.3. Materials and Methods

5.3.1 Materials & Reagent

Black goji pigments were extracted from dried black goji berries that were purchased from a grocery store (LianHua Supermarket) in Shanghai, China.

The chemicals and reagents (ACS or HPLC grade) were purchased from Fisher

(NaHCO3), potassium chloride, and sodium acetate. ACS grade ethyl acetate and formic acid were obtained from Mallinckrodt Chemicals (Bedminster Township, NJ USA) and

Honeywell (Morris Plains, NJ USA), respectively.

5.3.2 Pigments Extraction

The extraction of black goji anthocyanins by acetone-chloroform partitioning method were based on Rodriguez-Saona & Wrolstad, (2005). Dried black goji berries

(~50g) were mixed with liquid nitrogen, and powdered by using a food blender. The

90 powders were then mixed with 100%(v/v) aqueous acetone with 0.01% HCl. The slurry was filtered through filter paper (Whatman No.4 filter paper, Whatman Incorporation.

NJ, US) by water vacuum. After filtration, re-extraction was performed by adding

70%(v/v) aqueous acetone with 0.01% HCl until the slurry becoming faded. Two volumes of chloroform were then added into the aqueous acetone solution, and the whole filtrate was transferred into the separatory funnel. After the gentle mixing, samples were stored in 4°C refrigerator overnight. On the next day, after transferring the upper aqueous portion to a flask, the acetone and chloroform were evaporated by a rotary evaporator.

5.3.3 Pigments Purification

The C-18 cartridge purification method was adapted from Rodriguez-Saona &

Wrolstad, 2005. Aqueous anthocyanin extracts obtained from the above procedure were pumped to pass through the Sep-Pak® C18 cartridge (Waters Corporation. Milford MA,

USA) after activating the cartridge with methanol. Then the cartridge was washed with two column volumes of acidified water (0.01% v/v HCl), and anthocyanin was eluted into a boiling flask by adding 0.01% v/v HCl acidified methanol. After the removal of the methanol by a rotary evaporator, the purified anthocyanin samples were re-dissolved in acidified distilled water (0.01% v/v HCl).

5.3.4 Pigments Isolation and Identification

The major pigment, petunidin-3-trans-p-cou-rut-5-glu, was isolated by Luna reverse-phase PFP column (5μm particle size and 100 Å pore size with 250*21.20 nm column size, Phenomenex, Torrance, CA US) and semi-prep reverse-phase HPLC

21% B 25min-30min, and 21%-30% B 30-50min. The collected pigments were purified by C-18 cartridge and their identification and purity were confirmed by Kinetex reverse- phase PFP column (2.6μm particle size and 100 Å pore size with 100*4.6nm column size, Phenomenex, Torrance, CA US) and the linear gradient used for identity confirmation was from 8% B 0min-5min, 8%-15% B 5min-35min, 15%-35% B 35min-

37min, 35%-8% B 37-40min, 8% B 40-45min. Anthocyanin was monitored at 500-

530nm and 280-700nm, respectively. Total ion scan and selected ion monitoring

(Mass/charge ratios of 271, 287, 303, 301, 317 and 331, corresponding to the most common anthocyanin aglycones) were conducted. The method used for the identification of UV-irradiation samples was the same as described above, but with the following linear gradients: 7% B 0min-2min, 7%-20% B 2min-30min, 20%-50% B 30min-32min, 50%-

7% B 32-38min, 7% B 38-45min.

5.3.5 Pigments Quantification

The monomeric anthocyanin content was determined by the pH differential method (Mónica Giusti & Wrolstad, 2005). Buffer solutions were prepared using 0.1M potassium chloride at pH 1.0 and 0.4M sodium acetate at pH 4.5. Absorbances of samples at pH 1.0 and 4.5 were measured at 700nm and its λmax (512nm) by using UV-vis

Spectrophotometer (Shimadzu Corporation. Tokyo, Japan). Measurements were done in triplicates, and the black goji anthocyanin extracts or isolates was expressed as cyanidin-

3-diglucoside-5-glucoside (Ahmadiani et al., 2016).

5.3.6 Buffer System and Sample Preparation

The buffer systems in this study were 0.025 M KCl for pH 1 and 2; 0.1 M sodium acetate for pH 3-6; 0.25 M TRIS for pH 7and 8; and 0.1 M sodium bicarbonate for pH 9.

All the black goji extracts and isolates were diluted in these buffer solutions (pH1-9) at a concentration of 50 μM. The pH value after mixing was confirmed using a pH meter

(Mettler Toledo Inc, Columbus, OH US). The initial spectrophotometric and colorimetric measurement were performed after 20min equilibrium. Analysis was done in triplicates.

5.3.7 UV Light Exposure

Isolated and purified petunidin-3-trans-p-cou-rut-5-glu samples were dissolved in

100% MeOH (10-3 M) in quartz and glass cuvettes capped with PTFE stoppers

(FireflySci Inc Staten Island, NY), respectively. Samples were then put in a UV light 93 chamber (UV Stratalinker Model 1800, Stratagene, San Diego CA), where the power of

UV irradiation (254 nm) delivered was 3000uWatts/cm2.The span of UV light exposure time for the petunidin-derivative samples were determined based on their plateau stage.

In our preliminary study, the plateau stages in quartz and glass cuvettes were achieved after 30 min and 120 min, respectively. Therefore, in this study samples in quartz cuvettes were kept in UV irradiation for 0, 1, 3, 5, 10, 15, 30 min, corresponding to irradiation energy 0, 0.25, 0.76, 1.27, 2.55, 3.82, 7.64 J, respectively. Samples in glass cuvettes were kept in UV irradiation for 0, 5, 15, 30, 45, 60, 120, 180 min, with corresponding energy of 0, 1.27, 3.82, 7.64, 11.46, 15.28, 30.48, 45.72 J, respectively. At each time point, 100 ul sample was pipetted out from cuvette and mixed with 500 ul 4.5% formic acid for examination of pigment identification (HPLC chromatogram); and 10 ul sample was mixed with 200 ul buffers (pH 1-9) to investigate the spectrophotometric characteristics at different pH levels at time points 0, 5, 10, 15, 30 min from quartz cuvette and 0, 15, 30, 45, 60, 120 180 min from glass cuvette, respectively.

5.3.8 Spectrophotometric Analysis

After the mixture of petunidin-derivative samples with various buffers in the poly-D-lysine coated polystyrene well plates, the spectrum was analyzed from 380 nm to

700 nm with 1 nm intervals by using SpectraMax Microplate Reader (Molecular Devices,

Sunnyvale CA). The spectrophotometric analysis was conducted 20 min after mixing.

5.3.9 Colorimetric Analysis

The spectral absorbance data for each sample from 380 nm to 700 nm as described above were used to be converted into colorimeteric data using the

ColorBySpectra software (according to CIE 1964 standard observer, D65 illuminant spectral distribution, and 10° viewer angle) (Farr, JE; Srivastava, A; Machiraju, R; Giusti,

2017).

5.3.10 Statistical Analysis

Statistical analysis was performed by using Prism software (GraphPad, La Jolla,

CA US). One-way ANOVA (two-tailed, α=0.05) and post hoc Tukey’s test (family-wise

α=0.05) were conducted to evaluate the differences in C*ab, hab values among samples at different time points at certain pH conditions.

5.4. Results and Discussion

5.4.1 Influence of UV Irradiation on the Petunidin-Derivative

The chromatograms of the isolated black goji petunidin-3-trans-p-cou-rut-5-glu over UV-light exposure, and the corresponding quantification of cis : trans ratio are shown in Figure 5.1 and Figure 5.2, respectively. Before the UV irradiation began, there was only one peak (peak 2), representing the petunidin-3-trans-p-cou-rut-5-glu in the chromatograms. Its cis isomeric counterpart, petunidin-3-cis-p-cou-rut-5-glu (peak 1),

95 was gradually generated over the UV irradiation (254nm) process in a dose-dependent manner (Figure 5.1).

In quartz cuvettes, where usable range is 190-2500 nm, a small amount of energy

(0.25 J) was able to trigger the formation of cis isomers from its trans counterpart, with cis : trans ratio around 5 : 95 (≃ 0.05) (Figure 5.2). The formation of cis isomer followed first order kinetics with k=0.08 min-1. The cis : trans ratio finally reached 4 : 6

(≃ 0.67) in the plateau phase after the sample was exposed to 7.64 J UV irradiation.

In glass cuvettes, similar results were observed. The cis isomer was increasingly produced from trans isomer throughout UV light irradiation (Figure 5.1 and 5.2).

However, since part of the UV waves was blocked by the glass cuvette due to its usable range (340-2500 nm), a larger amounts of energy was needed for the formation of cis isomer. For example, it required 7.64 J energy in order for the cis : trans ratio to reach

34 : 66 (≃ 0.52) in glass cuvettes, as compared to 2.55 J in quartz cuvettes. It is no surprise that the first order kinetic parameter k in glass cuvettes was 0.02 min-1, smaller than that in quartz cuvettes. Although extensive energy was needed, petunidin-3-trans-p- cou-rut-5-glu in glass cuvette was able to produce larger proportion of cis isomers, as the final cis : trans ratio in plateau phase was 55 : 45 (≃ 1.2), twice amount of that in quartz cuvettes.

The total area under the curve of the pigments in the chromatograms declined

(~16% in both cuvettes) upon reaching the plateau stage (Figure 5.1 and 5.2), as anthocyanins were vulnerable to degradation when exposed to UV irradiation (Chisté,

Lopes, & de Faria, 2010). One should balance the trans-to-cis conversion and pigment

96 degradation at the same time by controlling the exposure span. Nevertheless, UV irradiation is still a valuable method for obtaining cis isomeric anthocyanins.

Figure 5.2 Ratio of cis and trans isomers of petunidin-3-rutinoside-(p-coumaroyl)-5- glucoside and total area under the curve (AUC) of pigments in chromatogram over UV irradiation

5.4.2 Influence of UV Irradiation on Colorimetric Properties of Petunidin-derivative

The colorimetric properties and data of the samples under UV irradiation are shown in Figure 5.3, 5.4 and Table 5.1, 5.2. In chapter 4, it was demonstrated that petunidin-3-cis-p-cou-rut-5-glu displayed bluer colors compared to trans counterpart at neutral and alkaline pH. Its hue angle values (hab) were 277.9°, 214.5°, 211.4° and 206.6° at pH 7, 8, 9, and 10, respectively (Chapter 4, Table 4.1), in contrast to 306.0°, 275.8°,

267.0°, and 266.8° in petunidin-3-trans-p-cou-rut-5-glu in the same pH conditions. With the formation of cis isomer due to the UV-induced isomerization, the hue angle of the original trans isomer decreased consecutively at pH 7-9 (Table 5.1 and 5.2). The hab dropped most at pH 9 (Δhab = 15.3 and 20.9 in quartz and glass cuvettes, respectively), followed by that at pH 8 (Δhab = 13.5 and 18.3) and pH 7 (Δhab = 8.1 and 4.4). As a result, the color of the petunidin isolate changed from purplish-blue to blue over the exposure under UV irradiation (Figure 5.3 and 5.4).

The above colorimetric differences were not observed at acid and mildly acidic pH, as the colors of the isolated petunidin trans isomer stayed in the same red hue angle throughout the UV treatment, in agreement with previous findings in Chapter 4 that there were no color differences between the petunidin cis and trans isomers at acidic pH

(Chapter 4, Table 4.1).

Although UV light caused trans to cis isomerization in petunidin-derivative, it was also expected to damage the chromophore structure, leading to the color degradation in anthocyanin (Chisté, Lopes, & de Faria, 2010). In fact, the chroma (C*ab) of trans isomer sample decreased over UV irradiation (Table 5.1 and 5.2). Thus, cautions should 99 be used when applying this method to obtain cis isomers, since the cis : trans ratio would remain the same but C*ab would keep decreasing once the isomerization reaches its plateau phase.

5.4.3 Influence of UV Irradiation on Spectrophotometric Properties of Petunidin- derivative

Corresponding to the decline in the chroma of petunidin samples, the absorbance at λmax decreased under UV irradiation, especially at alkaline pH (ΔA = 0.154 and 0.115 for sample in quartz cuvette at pH 9 and pH 8, respectively (Table 5.3). Similar results were observed in glass cuvette. The λmax, however, remained the same over UV light treatment at all pH levels (Table 5.3).

In the previous chapter, it has been shown that the cis isomer displayed unique broad peaks with λmax around 628 and 630 nm at pH 8 and 9, respectively; whereas the

λmax of the trans isomer was 578 nm at these two pHs, and a peak shoulder around 630 nm (Chapter 4, Figure 4.3). Although the λmax of petunidin-derivatives remained the same throughout the UV irradiation process, the shapes of the visible absorbance spectra were changing; there was a growing peak shoulder around 630 nm at pH 8 and 9 (Figure

5.5), which could be explained by the generation of cis isomers that caused the higher absorbance in ~630 nm region.

100

pH 1 2 3 4 5 6 7 8 9

0 J

1.27 J

2.55 J

3.82 J

7.64 J

-a* 0 a* -25 -20 -15 -10 -5 0 5 10 15 20 25

-5

-10 7.64 J

-15 7.64 J 7.64 J 0 J

-20 0 J

0 J -25 -b*

Figure 5.3 Colorimetric properties of petunidin-3-rutinoside-(trans-p-coumaroyl)-5- glucoside over UV irradiation in quartz cuvette

101

1 2 3 4 5 6 7 8 9

0 J

3.82 J

7.64 J

11.46 J

15.28 J

30.48 J

0 -a* -25 -20 -15 -10 -5 0 5 10 15 20 25 a*

-5

-10

30.48 J -15 0 J 30.48 J

-20 30.48 J 0 J 0 J-25

-b*

Figure 5.4 Colorimetric properties of petunidin-3-rutinoside-(trans-p-coumaroyl)-5- glucoside over UV irradiation in glass cuvette

102

Table 5.1 Colorimetric data of petunidin-3-rutinoside-(trans-p-coumaroyl)-5-glucoside over UV irradiation in quartz cuvettes.

pH of the buffer solution 7 8 9 L 0 82.2(0.1) 78.7(0.2) 78.1(0.3) 1.27 82.1(0.3) 79.4(0.2) 78.8(0.1) 2.55 82.6(0.1) 78.9(0.5) 78.9(0.2) 3.82 83.3(0.5) 81.1(1.1) 80.1(0.5) 7.64 83.7(0.4) 81.6(0.4) 81.4(0.2) C*ab 0 17.6(0.2)a 22.5(0.5)a 23.1(0.2)a

1.27 17.2(0.2)a 21.0(0.3)b 22.6(0.3)a 2.55 16.1(0.6)b 21.1(0.5)b 22.9(0.1)a 3.82 15.6(0.7)b 19.6(1.3)b 22.1(0.4)b 7.64 14.6(0.2)c 18.9(0.7)c 20.2(0.1)c hab 0 295.3(0.3)a 273.3(0.1)a 261.1(0.2)a 1.27 292.8(0.5)b 265.5(0.8)b 252.1(0.1)b 2.55 291.0(1.3)b 264.4(0.5)b 248.9(0.5)b 3.82 290.0(0.1)b 260.6(2.2)b 245.8(0.8)c 7.64 287.2(1.3)c 259.8(0.0)c 245.8(0.6)c

Different superscript letters indicate significant differences (p<0.05) in C*ab or hab over increasing UV irradiation energy given a trans isomer sample at a certain pH level.

103

Table 5.2 Colorimetric data of petunidin-3-rutinoside-(trans-p-coumaroyl)-5-glucoside over UV irradiation in glass cuvettes.

pH of the buffer solution 7 8 9 L 0 82.2(0.1) 78.7(0.2) 78.1(0.3) 3.82 83.9(0.3) 79.9(0.1) 80.9(0.2) 7.64 83.0(0.6) 78.7(0.5) 79.9(0.8) 11.46 82.0(0.4) 78.9(0.3) 78.6(0.2)

15.28 82.8(0.7) 78.7(0.5) 78.4(0.2) 30.48 83.2(0.1) 80.2(0.3) 80.5(0.0) C*ab 0 22.1(0.2)a 22.5(0.5)a 23.1(0.2)a 3.82 16.9(0.1)b 22.4(0.2)a 21.2(0.2)b 7.64 16.0(0.9)b 23.1(0.1)a 21.8(0.1)a 11.46 16.4(0.2)b 23.2(0.0)a 21.5(0.2)b 15.28 16.3(0.1)b 23.6(0.3)a 22.5(0.3)a

UV Irradiation Energy Energy (J) Irradiation UV 30.48 16.1(0.0)b 21.5(0.7)a 22.4(0.1)a hab 0 289.9(0.3)a 273.8(0.1)a 261.0(0.2)a 3.82 290.3(0.6)a 267.5(0.3)b 255.9(0.6)b 7.64 289.7(0.5)a 264.9(0.9)b 250.6(0.3)b 11.46 287.9(1.0)a 260.3(0.7)c 243.3(1.0)c 15.28 286.7(0.5)b 258.9(0.2)c 240.9(0.3)c 30.48 285.5(0.0)b 255.5(1.0)d 240.1(1.2)c

Different superscript letters indicate significant differences (p<0.05) in C* ab or hab over increasing UV irradiation energy given a trans isomer sample at a certain pH level.

104

Table 5.3 Spectrophotometric properties of petunidin-3-rutinoside-(trans-p-coumaroyl)- 5-glucoside over UV irradiation in quartz cuvettes.

pH 1 2 3 4 5 6 7 8 9

λmax 525.5 a 527.0 a 525.0 a 539.0 a 560.5 a 575.5 a 578.0 a 0 NM NM (0.6) (0.0) (0.0) (1.2) (0.6) (2.4) (0.0) 525.5 a 527.5 a 526.5 a 539.0 a 560.5 a 577.5 a 579.0 a 1.27 NM NM (0.6) (0.6) (0.6) (1.2) (0.6) (0.6) (0.0) 525.0 a 527 a 527.0 a 539.0 a 560.0 a 577.5 a 579.5 a 2.55 NM NM (1.2) (0.0) (1.2) (0.0) (1.2) (0.6) (0.6) 527.5 a 527.5 a 530.0 a 539.5 a 560.0 a 578.0 a 579.5 a 3.82 NM NM (1.5) (0.6) (1.2) (2.4) (1.2) (0.0) (0.6) 527 a 528.5 a 528.0 a 537.0 a 561.0 a 578.5 a 579.0 a 7.64 NM NM (0.6) (0.6) (0.0) (1.2) (1.2) (0.6) (1.2)

Absorbance 0.432 a 0.325 a 0.139 b 0.086 b 0.316 a 0.496 a 0.611 a 0 NM NM (0.025) (0.011) (0.001) (0.006) (0.005) (0.007) (0.009) a a a b a a b UV IrradiationUV (J) Energy 0.424 0.343 0.155 0.096 0.295 0.454 0.558 1.27 NM NM (0.022) (0.006) (0.004) (0.003) (0.005) (0.001) (0.001) 0.488 a 0.330 a 0.162 a 0.114 a 0.294 a 0.448 a 0.544 b 2.55 NM NM (0.023) (0.009) (0.006) (0.021) (0.005) (0.008) (0.006) 0.397 b 0.321 a 0.155 a 0.127 a 0.286 b 0.402 b 0.504 b 3.82 NM NM (0.047) (0.014) (0.012) (0.006) (0.002) (0.023) (0.015) 0.350 b 0.277 b 0.137 b 0.097 b 0.258 b 0.381 b 0.457 c 7.64 NM NM (0.030) (0.010) (0.004 (0.007) (0.014) (0.014) (0.001) Different superscript letters indicate significant differences (p<0.05) over increasing UV irradiation energy given a trans isomer sample at a certain pH level. NM: not measurable due to low absorbance.

105

Figure 5.5 Visible absorbance spectra of petunidin-3-rutinoside-(trans-p-coumaroyl)-5- glucoside over UV irradiation in quartz cuvette

106

5.5. Conclusion

UV irradiation (254 nm) induced dose-dependent trans to cis isomerization in petunidin-3-p-cou-rut-5-glu isolated from black goji. The trans isolate in quartz cuvette

(usable range 190-2500 nm) reached plateau phase soon after UV irradiation (7.64 J), with cis : trans ratio of 4:6; while in glass cuvettes (usable range 340-2500 nm) more energy was need in order to get the plateau cis : trans ratio of 55 : 45. The isomerization from trans to cis petunidin-3-p-cou-rut-5-glu caused decrease in hue angles, leading to bluer hues at neutral and alkaline conditions. However, UV irradiation also triggered color degradation, resulting in decline in chroma and absorbance in visible spectra. This study demonstrated that controlled application of UV-irradiation could be a potential method to obtain cis isomeric coumaroyl anthocyanins. It could be used for creating different coloration based on the pigments obtained from the same source.

107

Chapter 6 Metal Chelates of Petunidin-derivatives Exhibit Enhanced Color and Stability

Tang, P. & Giusti, M.M., 2018. Metal Chelates of Petunidin-derivatives Exhibit Enhanced Color and Stability. Submitted to Journal of Agricultural and Food Chemistry

6.1 Abstract

Anthocyanins with catechol or pyrogallol moieties on the B-ring are known to chelate with metals, showing bluing effect. However, most studies have focused on cyanidin and delphinidin derivatives in pH ≤ 6. In this study we investigated metal chelation of petunidin-derivatives in a wide pH range (3-10). Acylated petunidin-derivatives (mainly petunidin-3-rutinoside-(p-coumaroyl)-5-glucoside) from purple potato and black goji

(25uM), were combined with Al3+ or Fe3+ at ratios 1:0 to 1:60 in buffers pH 3-10.

Anthocyanins displayed red, purple, violet-blue colors at acidic, neutral and alkaline pH, respectively. Metal ions in small concentrations triggered bathochromic shifts at alkaline

3+ pH, resulting in vivid blue hues (hab 200°-310°). Fe addition caused higher bathochromic shift than Al3+, producing green colors at pH 8-9. Metal ions increased the color stability in a dose-dependent manner over 28 days, particularly at pH 8. Petunidin- derivatives and its metal chelate could serve as potential alternatives to synthetic food colorants.

Keywords: Purple Potato (Solanum tuberosum L. subsp. andigenum); Black Goji

(Lycium ruthenicum Murr.); Anthocyanin; Natural Colorant; Metal Chelation 108

6.2. Introduction

Anthocyanins are water-soluble pigments that render brilliant red, purple and blue colors to vegetables and fruits (He & Giusti, 2010). They undergo characteristic pH- dependent structural transformations, exerting diverse color appearance (Brouillard &

Delaporte, 1977). Given the fact of possessing potent antioxidant activity and being associated with health-promoting effects, anthocyanins have becoming an increasingly prevalent alternative to the current widely-use synthetic food dyes which are suspected to cause activity problems in children (He & Giusti, 2010b; Sharma et al., 2010).

Anthocyanin color and stability could be influenced by metal complexation

(Cavalcanti et al., 2011). It occurs between anthocyanins with at least two free hydroxyl groups on the B ring, and multivalent metal ions, including but not limited to Fe2+, Fe3+,

Ga3+, Al3+, Mg2+, Sn2+, Ti2+, and Cu2+ (Cortez et al., 2017; G. T. Sigurdson et al., 2016;

Yoshida et al., 2006, 2009). The metal ion acts in competition with H+ to attach to the catechol or pyrogallol moieties on the B-ring, triggering anthocyanin transformation from red flavylium cation form to purple-blue quinoidal base anion form. This transformed molecule would then stacks with other flavylium cation molecules to form a stable complex, leading to a bathochromic shift in the spectrum and bluing effect (Schreiber et al., 2010). Although various types of anthocyanin-Mn+ complexes exhibiting blue colors have been reported, most of these studies mainly focused on cyanidin and delphinidin derivatives at acidic or mildly acidic conditions (pH ≤ 6) (Pyysalo & Kuusi, 1973;

Schreiber et al., 2010; G. T. Sigurdson et al., 2016, 2017; Gregory T. Sigurdson & Giusti,

109

2014; Tachibana et al., 2014). It remains to be explored the metal chelation in petunidin- derivatives and at neutral or alkaline pHs.

Various cultivars of pigmented-flesh potatoes including white, yellow, red, and purple ones have been cultured, among which purple potato (Solanum tuberosum L. subsp. andigenum) has been increasingly popular among consumers because of its potential health benefits during the past few years (Giusti, Polit, Ayvaz, Tay, &

Manrique, 2014). Compared to white and yellow flesh potatoes, purple potato is rich in polyphenols, anthocyanins, and phenolic acids (Shiroma-Kian et al., 2008). It has been reported that purple potato contained up to 150 mg anthocyanin per 100g DW, and petunidin derivative was the predominant (~63%) pigment (M. Monica Giusti et al.,

2014). An average of 4.68 g gallic acid equivalent polyphenols per kg DW were found in purple potato, 3-4 folds compared to that in white flesh cultivar (M. Monica Giusti et al.,

2014; Lachman et al., 2008).

Lycium ruthenicum Murr. is a Chinese herb widely found in Qinghai-Tibet plateau. Its fruit, black goji berry, was reported to contain abundant anthocyanins (550-

500mg per 100g FW), most of which (>80%) were acylated ones (Zheng et al., 2011).

Five major anthocyanins were identified and petunidin derivatives accounted for 95% of the total anthocyanin content. High polyphenol content (~1310 mg GA equivalents/100gFW) and strong antioxidant activity (~1060mg GA equivalents/100g

FW) were also reported (Zheng et al., 2011). Our previous study has shown that petunidin-3-trans-p-coumaroyl-rutinoside-5-glucoside isolate contributed most of the

110 color expression of the black goji extract, and showed superior stability compared to other extracts over time (Tang and Giusti, 2017).

Anthocyanin-Mn+ chelation has been studied in food, such as strawberry puree, cranberry juice cocktail, crowberry juice, and in food model system with sugar beet pectin, isolated pectin fraction, and polysaccharides (Buchweitz et al., 2012; Cavalcanti et al., 2011; Tachibana et al., 2014; R. E. Wrolstad & Erlandson, 1973) as a practical method to manipulate and stabilize anthocyanin color. Metal chelation of anthocyanins is known to cause bathochromic shifts and can result on expression of blue colors.

The objective of this study was to investigate the colorimetric and spectrophotometric properties of petunidin-derivative pigments in purple potato and black goji, and to study the impact of their co-pigmentation with metal ions on the color expression and stability at neutral and alkaline conditions, complementing the knowledge of anthocyanin-metal chelation and providing the food industry with alternatives for synthetic dyes.

6.3 Materials and Methods

6.3.1 Materials & Reagent

Purple potato pigments and black goji pigments were extracted from purple potatoes (cultivar Purple Majesty) purchased from a local grocery store (Whole Foods

Market) in Columbus, Ohio, and black goji berry obtained from LianHua Supermarket in

Shanghai, China, respectively.

111

The chemicals and reagents used in this study were obtained from Fisher

Scientific (Fair Lawn, NJ USA) in either ACS or HPLC grade: acetone, chloroform, methanol, citric acid, sodium phosphate dibasic (Na2HPO4), sodium phosphate monobasic (NaH2PO4), sodium carbonate (Na2CO3), sodium bicarbonate (NaHCO3), potassium chloride (KCl), sodium acetate (CH3COONa), lab grade aluminum sulfate hydrate, reagent grade ferric chloride hexahydrate, and LC/MS grade acetonitrile. In addition, ethyl acetate (C4H8O2) and formic acid (HCO2H) in ACS grade were acquired from Mallinckrodt Chemicals (Bedminster Township, NJ USA) and Honeywell (Morris

Plains, NJ USA), respectively.

6.3.2 Pigments Extraction

Purple potato and black goji pigments were extracted by using the acetone- chloroform method reported by Rodriguez-Saona & Wrolstad, 2001. Approximately

100g purple potato and black goji berry were sliced and mingled with liquid nitrogen before being powdered in food blender. Acidified acetone (0.01% HCl) was then used to flush the powders, and then the pigment slurry was filtered through Whatman No.4 filter paper (Whatman Incorporation. NJ, US). Re-extraction was conducted as 70% (v/v) acidified aqueous acetone was used to re-wash the slurry until red color was invisible.

The filtrate was then transferred to a separatory funnel, and gently mixed with chloroform. After that, the separatory funnel was stored in refrigerated dark condition for overnight, so that the pigments will be concentrated in the upper aqueous phase. It was

112 collected on the next day in a round flask to remove the acetone and chloroform by a rotary evaporator.

6.3.3 Pigments Purification

The purification of pigment crude extracts was performed based on the method proposed by Rodriguez-Saona & Wrolstad, 2001. The aqueous anthocyanin extracts were forced to pass the Sep-Pak® C18 cartridge (Waters Corporation. Milford MA, USA) which had been activated with methanol. Two volumes of acidified water (0.01% v/v

HCl) was then pumped to wash the column, followed by one volume of ethyl acetate.

Next, the purple potato or black goji anthocyanin was recovered by acidified methanol

(0.01% v/v HCl) and collected in a round flask. After removal of methanol by using a rotary evaporator, the purified anthocyanin was dissolved in acidified distilled water

(0.01% v/v HCl).

6.3.4 Pigments Identification

The purple potato extracts and black goji extracts were evaluated and their pigments were identified by high performance liquid chromatography (HPLC)

(Shimadzu, Columbia, MD) composed of a SP-M20A Photodiode Array Detector

(Shimadzu, Columbia, MD) and a LCMS-2010EV Liquid Chromatograph Mass

Spectrometer. The separation of the pigments was achieved by a reverse phase Symmetry

C-18 column (5μm particle size and 4.6*150mm column size) (Phenomenex, Torrance,

113

CA US). Before injection into HPLC, all of the extracts were filtered through a 0.22 um syringe filter (Phenomenex, Torrance, CA US). The mobile phase was made of solvent

(A) 4.5% (v/v) formic acid and solvent (B) 100% acetonitrile. Flow rate was set to be 0.8 ml/min. The linear gradient in this study was programed as follows: 7% B 0min-1min,

7%-20% B 1min-30min, 20%-40% B 30min-36min, 40% B 36-40min, 40%-7% B 40-

42min, 7% B 42-52min. Anthocyanin and all phenolics elutions were monitored at 500-

530nm and 280-700nm, respectively. Total ion scan and selected ion monitoring were performed.

6.3.5 Pigments Quantification

The monomeric anthocyanin content was quantified by the pH differential method adopted from Giusti & Wrolstad, 2005. Purified anthocyanin samples were dissolved in

0.025 M potassium chloride buffer (pH 1) and 0.4 M sodium acetate buffer (pH 4.5), respectively. After 15 min equilibrium, the absorbance at λvis-max (524 nm) and at 700 nm were measured by UV-vis Spectrophotometer (Shimadzu Corporation. Tokyo, Japan).

The quantification was accomplished in triplicates, and the monomeric anthocyanin content of the extracts were expressed as cyanindin-3-glucoside equivalence.

6.3.6 Buffer System and Sample Preparation

A series of buffer solution ranging from pH 3 to 10 was prepared (citric acid-

Na2HPO4 buffer solutions for pH 3-7; Na2HPO4-NaH2PO4 buffer solution for pH 8;

114

Na2CO3-NaHCO3 buffer solutions for pH 9-10) (Dawson et al., 1986). The overall principle was to 1) use the buffer that covers a wide pH range; and 2) keep the source of metal ions same among different buffer systems so that the influence of exogenous metal ion was minimum. Purple potato and black goji anthocyanin extracts (25 μM) were diluted in the above buffers and the final mixture pH environment was confirmed by a pH meter (Mettler Toledo Inc, Columbus, OH US). For metal chelation study, the metal ions

(Al3+ or Fe3+) was prepared in concentration of 0.06 M before it was mixed with 25 μM purple potato or black goji anthocyanin extracts based on certain anthocyanin and metal- ion molar ratio in buffers pH 7-9. The final pH condition in each combination was confirmed as well by a pH meter.

The initial spectrophotometric and colorimetric measurement were examined after

60 min equilibrium. Anthocyanin samples chelated with metal ions were stored at refrigerated condition at dark for 4 weeks for stability test. Experiments were done in triplicates.

6.3.7 Spectrophotometric Analysis

After the mixing of anthocyanin samples with or without metal ions in different buffer systems, 250 μL of each sample was transferred to poly-D-lysine coated polystyrene 96 well plates. The spectrum of these samples from 380 to 700 nm (1nm interval) was investigated by SpectraMax 190 Microplate Reader (Molecular Devices,

Sunnyvale CA).

115

6.3.8 Colorimetric Analysis

The color characteristics of each mixture were analyzed by ColorQuest spectrophotometer (data presented in CIE-L*C*abhab with 10° viewer angle and D65 illuminant spectral distribution) over 4 weeks under refrigerated storage condition in the dark (60 min after mixing, as well as on day 1, 2, 3, 4, 5, 7, 14, 21, and 28). The color changes were measured as ΔE.

6.3.9 Statistical Analysis

Statistical analysis was conducted by Prism software (GraphPad, La Jolla, CA

US). One-way ANOVA (two-tailed, α=0.05) and post hoc Tukey’s test (family-wise

α=0.05) were conducted to evaluate the differences among different samples.

6.4 Results and Discussion

6.4.1 Anthocyanin Profiles in Black Goji and Purple Potato

Three major anthocyanins accounting for 93% of the total peak area at 520nm were separated from black goji (Figure 6.1). The relative peak areas of the three pigments (peak 1, 3 and 4) were approximately 18%, 7%, and 71%, respectively.

According to the MS data and acid hydrolysis, all major anthocyanins in black goji were petunidin derivatives and 80% of the total pigments were acylated. A minor peak (peak

2) was identified as a delphinidin derivative and represented less than 5% of the anthocyanins. The pigments were identified to be petunidin-3-galactoside-5-glucoside

116

(peak 1), delphinidin-3-trans-p-coumaroyl-rutinoside-5-glucoside (peak 2), petunidin-3- cis-p-coumaroyl-rutinoside-5-glucoside (peak 3), and petunidin-3-trans-p-coumaroyl- rutinoside-5-glucoside (peak 4), in agreement with the identification of predominant pigments in previous study (Zheng et al., 2011).

One anthocyanin in purple potato – cultivar Purple Majesty accounted for 91% of the total peak area (Figure 6.1) and was identified as petunidin-3-rutinoside-(p- coumaroyl)-5-glucose (peak 4). A second minor peak (peak 2) representing ~4 of the total peak area was delphinidin-3-rutinoside-(p-coumaroyl)-5-glucoside. The same pigments were identified in purple potato varieties by previous researchers (M. Monica

Giusti et al., 2014; Nayak, Berrios, Powers, & Tang, 2011; Stushnoff et al., 2008) although their proportions varied widely. It is interesting that these two anthocyanins match two of the pigments found in black goji extracts as well.

117

Figure 6.1 Chromatograms of black goji and purple potato anthocyanin extracts at 520 nm and 280-700nm, and their identifications.

6.4.2 Spectrophotometric and Colorimetric Properties of the Purple Potato and Black

Goji Extracts

Generally speaking, both extracts displayed vivid red, purple, and violet-blue hues in acidic, neutral, and alkaline pH, respectively (Figure 6.1). Anthocyanins undergo pH- dependent structure transformation (Raymond Brouillard, Delaporte, & Dubois, 1978;

118

Raymond Brouillard & Dubois, 1977), and under acidic pH, the flavylium cation predominates. Purple potato and black goji extracts exhibited hue angles (h) around 355 ° in CIE-L*C*abhab color space at pH 3, corresponding to red hues (Table 6.1). When pH increases to mildly acidic environment, hydration of the flavylium cation is expected with formation of the colorless carbinol pseudobase form. As a result, at pH 4 and 5 the hue angles of both extracts moved towards yellow-hue direction, and the chroma (C*ab) decreased as the color of the extracts became pale and dull (Table 6.1). Starting from neutral pH, the quinoidal base is converted, and it starts to carry negative charges as pH further reaches alkaline conditions, leading to blue colors. Consequently, the hue angles of the purple potato extracts turned from 341.6° at pH 6, to 207.8° at pH 10. Similarly, black goji extracts experienced the same changes in hue angles, but in less magnitude

(327.7° at pH 6 to 282.6° at pH 10). Remarkably, petunidin-derivatives displayed various vivid blue hues in neutral and alkaline conditions, with corresponding hue angles ranging from 242.6° and 306.6° (Table 6.1). The chroma of both extracts also enlarged in these neutral-alkaline pHs.

The increase of pH with the resulting color changes toward the blue region were a result of the bathochromic changes in the visible spectra of the pigments (Table 6.1). The

λmax of both extracts at acidic pH were ~524 nm and it boosted to 578 nm at pH 8 and 9.

The visible spectra of purple potato and black goji extracts at alkaline pH were documented by a distinctive peak shoulder (~630 nm) next to the λmax (Figure 6.2 and

Figure 6.3).

119

Table 6.1 Colorimetric (CIE-L*, C*ab, hab) and spectrophotometric (λmax) data of purple potato and black goji anthocyanin extracts over pH 3-10.

pH3 pH4 pH5 pH6 pH7 pH8 pH9 pH10 Purple Potato L* 91.5(0.6) 93.0(0.9) 94.6(0.3) 90.4(0.4) 80.2(1.6) 75.7(0.5) 74.4(0.9) 80.8(2.2)

C*ab 8.6(1.2) 5.0(1.5) 2.6(0.1) 6.9(0.8) 17.7(2.1) 18.8(0.7) 18.6(1.2) 7.8(1.7)

hab 357.3(3.0) 17.8(1.2) 54.3(1.7) 341.6(4.3) 306.5(0.4) 262.2(1.9) 242.6(3.4) 207.8(5.6)

λmax 524(0.0) 524(0.0) 529(0.6) 539(0.0) 558(0.0) 577(0.0) 578(0.0) 578(0.0)

Black Goji L* 90.4(3.1) 93.0(1.1) 93.6(1.1) 81.9(1.1) 70.2(0.1) 68.1(1.2) 69.7(1.4) 77.1(1.1) C*ab 11.4(0.9) 5.0(0.5) 3.7(1.1) 39.7(1.7) 40.7(1.4) 36.4(1.1) 24.8(1.9) 27.7(1.1) hab 350.2(2.4) 16.5(1.8) 30.1(2.4) 327.7(1.7) 306.6(1.7) 275.6(1.4) 259.3(1.8) 282.6(1.4) λmax 525(0.7) 527(0) 536(0.7) 539(1.4) 558(0.7) 578(0) 578(0) NM NM: not measureable due to color loss.

6.4.3 Spectrophotometric Properties of Petunidin-derivatives Chelated with Metals

The visible absorbance spectra, and quantifications of λmax and maximum absorbance of Petunidin-M3+ chelates (abbreviated as Pt-M3+) are shown in Figure 6.2,

6.3, 6.4, and 6.5. Generally, bathochromic and hyperchromic shifts were perceived for

Pt-M3+ at pH 7-9, similar to the previous studies on cyanidin or delphinidin-M3+ chelates at acidic pHs (Buchweitz, Brauch, Carle, & Kammerer, 2013a; G. T. Sigurdson et al.,

2016, 2017; R. E. Wrolstad & Erlandson, 1973). Pt-M3+ chelation was greatly influenced by pH and metal ion source.

The addition of Al3+ to purple potato petunidin-derivatives introduced the most pronounced bathochromic shift (Δλmax = 42 nm) at pH 7 and maximum hyperchromic

120 response (ΔA = 0.282) at pH 9, while Fe3+ chelates experienced their maximum shifts in larger magnitudes (Δλmax = 79 nm and ΔA = 0.399, respectively) at the same pH as that of Al3+ (Figure 6.4). Black goji petunidin-derivatives exerted the maximum

3+ 3+ bathochromic shifts with Al (Δλmax = 33 nm) and Fe (Δλmax = 66 nm) at pH 9 and 7, respectively (Figure 6.5). Additionally, in purple potato extracts the largest λmax and absorbance achieved by Pt-Al3+ was 609 nm and 0.939 at pH 9, respectively; and Pt-Fe3+ reached its maximum λmax (637 nm) and absorbance (1.046) at pH 7 and pH 9,

3+ 3+ respectively. In black goji extracts, the maximum λmax of Pt-M were 611 nm (Pt-Al ) and 630 nm (Pt-Fe3+). Overall, Pt-Fe3+ showed higher bathochromic and hyperchromic responses than Pt-Al3+, similar to previous reported results in cyanidin-M3+ (G. T.

Sigurdson et al., 2016). It was proposed by Sigurdson et al (2016) that the outer electrons of Fe3+ appeared to be in a high spin configuration (with 1 electron in each 5d orbitals), resulting in an enhanced overlapping with anthocyanin and a stronger association. This unique electron arrangement was expected to be responsible for the pronounced shifts in visible spectra. Thus, Fe3+ in lower concentration was able to produce a large bathochromic response.

121

120

1.5 pH 7 Al3+ 100 1.5 pH 7 Fe3+ 1.2 1.2 0.9 0.9

0.6 0.6 0.3 0.3 80 1:0 0 0 400 450 500 550 600 650 700 400 450 500 550 600 650 700 1:0.1 1.5 pH 8 Al3+ 1.5 pH 8 Fe3+ e 1:0.5 c 1.2 1.2

a 1:1 b 0.9 0.9

o 60

s 0.6 0.6 1:2

A 0.3 0.3 1:5 0 0 400 450 500 550 600 650 700 400 450 500 550 600 650 700 1:10

1.5 3+ 1.5 3+ 1:30 pH 9 Al 40 pH 9 Fe 1.2 1.2 1:60 0.9 0.9

0.6 0.6 0.3 0.3

0 20 0 400 450 500 550 600 650 700 400 450 500 550 600 650 700 Wavelength (nm)

Figure 6.2 Visible light spectra of purple potato anthocyanin metal chelates at alkaline 0 0 pH. 5 10 15 20 25 30

3+ The Pt-M molar ratio that was necessary to achieve maximum λmax and absorbance at

3+ each pH level was closely related to M source. In purple potato extracts, the largest λmax of Pt-Al3+ was accomplished at the molar ratio of 1:60, 1:10, and 1:5 at pH 7, 8 and 9, respectively; while it was 1:1, 1:2, 1:30 in Pt-Fe3+ at these three pHs (Figure 6.4).

Similar trends could be found in black goji extracts (Figure 6.5). Actually, the [Al3+] required for reaching the maximum λmax dropped as pH increased, aligning with the idea that M3+ acts in competition with H+ for B-ring attachment in anthocyanins (Olivier

Dangles, Elhabiri, & Brouillard, 1994). However, the necessary [Fe3+] to reach the largest

122

λmax increased when pH increased, which might be explained by the aggregation and precipitation of anthocyanin pigment triggered by the addition of Fe3+, as well as the poor stability of Fe3+ at alkaline pH (Buchweitz et al., 2013a, 2012; G. T. Sigurdson et al.,

2016). 120

100 0.8 pH 7 Al3+ 0.8 pH 7 Fe3+ 0.6 0.6

0.4 0.4

0.2 0.2 80 1:0 0 0 380 430 480 530 580 630 680 380 430 480 530 580 630 680 1:0.1

0.8 3+ 0.8 3+ e pH 8 Al pH 8 Fe 1:0.5

n 0.6 0.6 a 1:1

o 0.4 60 0.4

s 1:2

A 0.2 0.2 1:5 0 0 380 430 480 530 580 630 680 380 430 480 530 580 630 680 1:10 0.8 2 1:30 pH 9 Al3+ 40 pH 9 Fe3+ 0.6 1.5 1:60

0.4 1

0.2 0.5

0 20 0 380 430 480 530 580 630 680 380 430 480 530 580 630 680 Wavelength (nm)

Figure 6.3 Visible light spectra of black goji anthocyanin metal chelates at alkaline pH. 0 0 5 10 15 20 25 30

123

640

630

620

610

a 600

m 1.2 λ 590

580 1.1

570 1.0 pH7 Al 560 0.9 pH7 Fe 550 0.8 1:0 1:0.1 1:0.5 1:1 1:2 1:5 1:10 1:30 1:60 pH8 Al 0.7 pH8 Fe 3+ ACN:M Ratio pH9 Al 0.6 pH9 Fe 1.2 0.5

1.1 0.4

a l 1 2 5 0 0 0 tr .1 .5 : : : 1 3 6

m 1 1 1 : : : 1.0 C :0 :0 1 1 1

λ 1 1

a 0.9

n 0.8

r 0.7

b 0.6

A 0.5

0.4 1:0 1:0.1 1:0.5 1:1 1:2 1:5 1:10 1:30 1:60

ACN:M3+ Ratio

Figure 6.4 Quantification of spectrophotometric changes in purple potato anthocyanin metal chelates

124

640

630

620

610

x a 600

λ 590

580 1.2

570 1.1 560 1.0 pH7 Al 550 0.9 1:0 1:0.1 1:0.5 1:1 1:2 1:5 1:10 1:30 1:60 pH7 Fe 0.8 ACN:M3+ Ratio pH8 Al 0.7 pH8 Fe pH9 Al 0.6 0.9 pH9 Fe 0.5

a 0.8

m 0.4

l t r .1 .5 :1 :2 :5 0 0 0 t 0 0 1 1 1 :1 :3 :6 a 0.7 C : : 1 1 1 1 1

a 0.6

s 0.5

A 0.4

0.3 1:0 1:0.1 1:0.5 1:1 1:2 1:5 1:10 1:30 60 ACN:M3+ Ratio

Figure 6.5 Quantification of spectrophotometric changes in black goji anthocyanin metal chelates.

125

Another interesting phenomenon was that when [M3+] increased, there would be a hypsochromic effect after the bathochromic effect, indicating a “saturation status” at the turning point (Figure 6.4 and 6.5). However, Pt-Al3+ at pH 7 was an exception, which could probably be explained that the [Al3+] tested in this study was not large enough to reach the “saturation status”. Similarly, hypochromic response appeared before the hyperchromic effect; and the turning point in λmax corresponded to that in absorbance

(Figure 6.4 and 6.5).

Fe3+ in higher concentration usually expresses yellow hue colors in aqueous systems, documented by the higher absorbance in the spectra region 380-450 nm.

Consequently, Pt-Fe3+ gradually developed higher absorbance in a [Fe3+] dose-dependent manner in this region at pH 7-9 (Figure 6.2 and 6.3). Combined with the blue color from petunidin-derivatives and the bluing effect introduced by metal chelation, greenish blue colors were observed (Figure 6.6, 6.7, 6.8, and 6.9) whereas there was still a local maximum peak around 610 nm (as discussed in later section).

6.4.4 Colorimetric Properties of Petunidin-derivatives Chelated with Metals

In alignment with the bathochromic and hyperchromic shifts in spectrophotometric data, bluing responses with vivid colors were found in the Pt-M3+

(Figure 6.6, 6.7, 6.8, and 6.9). The color hues of Pt-M3+ were dependent upon the metal ion source, pH, and Pt-M3+ ratio.

With the increase of [Al3+], purple potato extracts developed a variety of blue colors. At pH 7, brilliant blue with hue angle ~261° appeared starting from Pt:Al3+ =1:30. 126

Darker blue (~240°) and greenish-blue (~200°) were obtained when Pt:Al3+ =1:2 at pH 8 and pH 9, respectively. Fe3+ also triggered bluing shifts, as the increasing [Fe3+] led to the transition of Pt-Fe3+ from quadrant IV to quadrant III in the CIELAB color space (Figure

6.8). As discussed above, Fe3+ in higher concentration expresses yellow hue. Therefore, mixed with blue color, Fe3+ began to develop green colors when its ratio to petunidin- derivatives was greater than approximately 30:1, exhibiting hue angles ~216°, 210°, and

167° at pH 7, 8, and 9, respectively (Figure 6.8). Similar results were found in black goji

Pt-M3+ chelates, where various blue and greenish-blue colors were obtained in neutral and alkaline pH (Figure 6.9).

Corresponding to spectrophotometric properties, the necessary ratio between petunidin-derivatives and M3+ for blue color formation was strictly related to the pH condition and M3+ source. It required less [Al3+] for petunidin to display a significant transition into blue hues when pH increased from 7 to 9, supporting the theory of the competition between metal ion and hydrogen ion for the attachment to B-ring catechol moiety (Olivier Dangles et al., 1994). But it did not apply to the case of Fe3+, where an increasing [Fe3+] was essential for the blue-green hues development. As discussed previously, it could be due to the aggregation and precipitation of anthocyanin pigment by adding Fe3+, as well as its poor solubility in alkaline pH (Buchweitz et al., 2013a,

2012; G. T. Sigurdson et al., 2016).

127

128

129

Figure 6.8 Color data of purple potato anthocyanin metal chelates expressed in CIE-LAB color space.

Figure 6.9 Color data of black goji anthocyanin metal chelates expressed in CIE-LAB color space.

130

In order to measure the minimum [M3+] that is necessary to demonstrate a color change in the purple potato and black goji extracts, ΔE was calculated. Typically, ΔE = 5 is believed to be the threshold for untrained eyes to differentiate the color differences

(CIE, 2004). As shown in Figure 6.6 and 6.7, ΔE reached 5 with less [Al3+] and more

[Fe3+] in alkaline pH compared to that in neutral pH, as explained above. It should be noted that both the metal ions at all pHs tested were able to cause color changes even when Pt:M3+ was less than 1:1. Black goji extracts demanded more metals for blue color formation as compared to purple potato extracts, which might be due to its more complicated anthocyanin profiles. In purple potato extracts, Al3+ in small concentration

(Pt:Al3+ = 1:0.5) could produce blue colors at pH 8 and 9, and it was the same for Fe3+ at pH 7 and 8. These low threshold properties is desirable since there is concern regarding the neurotoxicity associated with metals (Becaria, Campbell, & Bondy, 2002). However, both aluminum sulfate and ferric chloride were considered by FDA as Generally

Regarded as Safe (GRAS), with no limit in usage (Code of Federal Regulations, 2017). In addition, a metallic salt is an important ingredient in FD&C lake pigments, e.g. Red 40

Aluminum Lake, Blue #1/2 Aluminum Lake. The metal ion concentrations used in this study was lower than that in FD&C Lake pigments. Therefore, we proposed a tentative new method to decrease metal usage in food pigments in neutral and alkaline pH.

131

6.4.5 Stability of Petunidin-derivatives with or without Metal Chelation

Pt-M3+ were stored in a dark refrigerated condition for 28 days. The color changes, expressed as ΔE, shown in Figure 6.10, 6.11 and 6.12. Pt-M3+ generally exhibited enhanced color stability in a [M3+] dose-dependent manner. Higher [Al3+] significantly detained the blue colors hue after 7 days at all pH value tested, while the petunidin-derivatives without metal chelation gradually became colorless after 2 days.

Interestingly, brown hues steadily formed in higher [Fe3+] chelates after mixture with petunidin-derivatives and was more severe in alkaline pHs. Fe3+ is known for its reductive-oxidative capabilities, and has been reported to catalyze the oxidative degradation of the flavonoid (Makris & Rossiter, 2000). Therefore, the brown color could be developed due to the degradation of anthocyanin, especially in the alkaline pH range.

In addition, Fe3+ in large concentration would express yellow-brown colors, which was covered by the blue color produced by purple potato anthocyanins soon after their mixture, exhibiting green hues. But when blue color vanished, the yellow hues would outstand and dominate the mixture, which also would partially explain the browning phenomenon.

Among the three pH values in this study, pH 8 seemed to be the best condition for petunidin-derivatives and their metal chelates regarding their color stability. At this pH, a lower concentration of M3+ still showed a promising stabilization capability. At neutral and alkaline pH, anthocyanin is supposed to exist dominantly in quinoidal base forms, and would take on increasing negative charges with a further increase in pH (Raymond

Brouillard & Delaporte, 1977b). This process could be documented by pKa2 and pKa3, the 132 disassociation constant for the transition from quinoidal base configuration to that with one negative charges, and the transformation from one negative charged to two negative charged. Previous study reported that the pKa2 and pKa3 of petunidin aglycone were 6.99 and 8.27, respectively (León-Carmona et al., 2016). Although the true pKa’s for petunidin-3-rutinoside-(p-coumaroyl)-5-glucose would be different, it is reasonable to postulate that the purple potato anthocyanin carries negative charges at pH 8 and thus is more stable, compared to other close neutral and alkaline pHs.

Overall, metal ions helped to stabilize the petunidin-derivatives in a dose dependent behavior, and significantly slowed down the escalation of ΔE. Fe3+ in high concentration, because of its oxidative capability, accelerated the color degradation and promoted browning effect.

133

120

120 120 3+ 100 3+ 100 pH 7 Al 100 pH 7 Fe

80 80

60 60

40 40

20 20 80 0 0 1:0

) 0 5 10 15 20 25 30 0 5 10 15 20 25 30

E 1:0.1

Δ 120 120

(

3+ 3+ s 100 pH 8 Al 100 pH 8 Fe 1:0.5

e g 80 80 1:1

n a 60 60 60 h 1:2

40 40

r 1:5

l 20 20

o 1:10 0 0

C 0 5 10 15 20 25 30 0 5 10 15 20 25 30 1:30

120 120 40 1:60 3+ 3+ 100 pH 9 Al 100 pH 9 Fe

80 80

60 60

40 40 20 20 20

0 0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Time (days)

0 Figure 6.10 Color changes (ΔE) of purple potato0 anthocyanin5 10 1metal5 20 chelates25 3 0over 28 days in dark refrigerated condition.

134

120

120 120 3+ 3+ pH 7 Al p1H00 7 Fe 100 100 80 80 60 60 40 40 20 20 0 0 80 0 5 10 15 20 25 30 0 5 10 15 20 25 30 1:0

)

E 100 100 1:0.1

( 3+ 3+ pH 8 Al pH 8 Fe s 80 80 1:0.5

g 1:1 n 60 60

a 60

h 1:2 40 40

r 1:5 20 20

o 1:10 0 0 C 0 5 10 15 20 25 30 0 5 10 15 20 25 30 1:30

100 100 40 1:60 pH 9 Al3+ pH 9 Fe3+ 80 80

60 60

40 40

20 20 20 0 0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Time (days)

0 Figure 6.11 Color changes (ΔE) of black goji anthocyanin0 5 metal10 1 5chelates20 2over5 3280 days in dark refrigerated condition.

135

pH 7 Al3+ ACN:M+ pH 7 Fe3+ ACN:M+ 1:0 1:0.1 1:0.5 1:1 1:2 1:5 1:10 1:30 1:60 1:0 1:0.1 1:0.5 1:1 1:2 1:5 1:10 1:30 1:60 60min 60min Day1 Day1 Day2 Day2 Day3 Day3 Day4 Day4 Day5 Day5 Day7 Day7 Day14 Day14 Day21 Day21 Day28 Day28

pH 8 Al3+ ACN:M+ pH 8 Fe3+ ACN:M+ 1:0 1:0.1 1:0.5 1:1 1:2 1:5 1:10 1:30 1:60 1:0 1:0.1 1:0.5 1:1 1:2 1:5 1:10 1:30 1:60 60min 60min Day1 Day1 Day2 Day2 Day3 Day3 Day4 Day4 Day5 Day5 Day7 Day7 Day14 Day14 Day21 Day21 Day28 Day28

3+ pH 9 Al3+ ACN:M+ pH 9 Fe ACN:M+ 1:0 1:0.1 1:0.5 1:1 1:2 1:5 1:10 1:30 1:60 1:0 1:0.1 1:0.5 1:1 1:2 1:5 1:10 1:30 1:60 60min 60min Day1 Day1 Day2 Day2 Day3 Day3 Day4 Day4 Day5 Day5 Day7 Day7 Day14 Day14 Day21 Day21 Day28 Day28

Figure 6.12 Color stability of purple potato anthocyanin metal chelates over 28 days in dark refrigerated condition

136

6.5. Conclusion

Purple potato and black goji extracts contained abundant petunidin-derivatives, with petunidin-3-(p-coumaroyl)-rutinoside-5-glucoside as the predominant pigment. The extracts displayed vivid red, purple, violet-blue colors from acidic, neutral, to alkaline pH, respectively. In the presence of metal ions, Al3+ and Fe3+, the petunidin-derivatives experienced bathochromic (up to 79 nm) and hyperchromic shifts (increased up to 0.399 in absorbance at λmax), resulting in various blue to green hues and intensified colors in neutral and alkaline conditions. The magnitudes of the metal chelation effect were dependent on metal ions, pH, and Pt:M3+ ratios. Fe3+ with a higher spin configuration in the outer electrons triggered further bathochromic shift than that of Al3+, and its unique yellow color in aqueous system led to green hues when combined with petunidin- derivative’s blue colors. The oxidative capability of Fe3+ towards flavonoids limited its application in larger concentration (>125 uM). Al3+ in low concentration (~12.5 uM) was able to chelate Pt and produce blue colors at pH 8 and 9. The concentration of [Al3+] needed for blue production decreased upon the increase of pH. The chelation of metal ions significantly enhanced the color stability (up to 4 times) of petunidin-derivatives over neutral and alkaline conditions.

This study demonstrated that petunidin-derivatives could chelate with metal ions

Al3+ and Fe3+ in neutral and alkaline pH, resulting in enhanced color and stabilities.

Purple potato and black goji, containing abundant petunidin-derivatives are promising sources for natural colorants over a wide range of pH. Chelated with metals, these

137 pigments could provide the food industry with various vivid violet, blue, and green colors with enhanced stability.

138

Chapter 7 Stabilization and Color Enhancement of Anthocyanin by Soybean Isoflavones

7.1 Abstract

The color of anthocyanins could be stabilized and enhanced by co-pigmentation in the presence of some flavonoids. The objective of this study was to investigate the potential co-pigmentation between anthocyanins and soybean isoflavones, and the effects of anthocyanin structure on their interaction. Acylated anthocyanins were extracted and purified from various plants. Their non-acylated counterparts were prepared by saponification. These pigments were mixed with soybean curd residue isoflavones at pH

3 and 7. Anthocyanins formed co-pigmentation with soybean isoflavones, resulting in enhanced and stabilized colors. This interaction was influenced by anthocyanin structures. More hydroxyl groups on the anthocyanin B-ring led to more pronounced bathochromic and hyperchromic shifts. Acylated anthocyanins showed more pleasant colors and better favored co-pigmentation with soybean isoflavones than non-acylated ones. The results indicated that soybean isoflavones could serve as potent co-pigments to enhance and stabilize anthocyanin colors, providing an innovative way of broadening the natural pigment application.

Keywords: Co-pigmentation; Acylation; Soybean curd residue; Isoflavonoids;

139

7.2 Introduction

Color is one of the important attributes that consumers consider in food quality.

During the last few years, there have been increasing health concerns towards the use of certain synthetic dyes, which might cause behavior problems in children with attention deficit hyperactivity disorder (ADHD), as well as potential cancer development (Sharma et al., 2010). For the replacement of artificial colorants, natural pigments that are commonly extracted from fruits and vegetables, have been demanded by consumers.

Responsible for red, blue, and purple colors in vegetables and fruits, anthocyanins are a class of natural pigments commonly found in plants. They have important application in coloring food products (Obón et al., 2009). Colorants made of anthocyanins are widely manufactured for food use from fruits and vegetables (Pintea,

2007). According to Code Federal Regulation (CFR), anthocyanins from grape skin & grape color extract, vegetable juice (such as red cabbage, black carrot, purple sweet potato, radish), and fruit juice (raspberry, blueberries and elderberry) are natural colorant additives, which is exempted from certification (21 CFR Part 73, subpart A-Foods).

However, the application of anthocyanins as natural food pigments has encountered difficulties due to their poor stability. In fact, the color of anthocyanins is sensitive to light, heat, oxygen, and pH, which limits their use in different food products

(Eiro & Heinonen, 2002). Thus, the food industry is seeking for methods to improve anthocyanin stability. Previous studies have demonstrated that the color of anthocyanins can be stabilized and strengthened by co-pigmentation interactions, which is often manifested as bathochromic effect (the shift of the maximum absorption wavelength,

140

Δλmax, in the visible range towards higher wavelength) and hyperchromic effect (the increase in the intensity of absorbance) when a colorless co-pigment is added to an aqueous solution of pigments (Malien-Aubert et al., 2001; Pacheco-Palencia & Talcott,

2010; Fossen et al., 2007).

There are two crucial mechanisms for the co-pigmentation phenomenon: 1) intramolecular interactions, in which a covalent linkage of anthocyanin molecules occurs with an organic acid, or an aromatic acyl group; 2) intermolecular interactions, in which the pigment and the co-pigment interact through a weak π-π overlap, dipole-dipole interaction, or possible hydrogen binding (R. Brouillard et al., 1989), resulting in an overlapping arrangement of the two molecules. Intermolecular co-pigmentation will prevent the nucleophilic attack on the anthocyanin by water, promoting anthocyanin stability (O. Dangles et al., 1992). The more hydroxyl groups there are in the co-pigment molecule, the stronger the complex forms (Gómez-Míguez et al, 2006).

Isoflavones are plant secondary metabolites commonly found in leguminous plants (Wang & Murphy, 1994b). They have drawn wide attention over the last few decades because of their potent phytoestrogenic effects. Numerous papers investigated the roles of isoflavones in health, and found isoflavones possess multiple health benefits, such as antioxidant (B J Boersma et al., 2003), anti-cardiovascular diseases (Elizabeth A

Kirk, Sutherland, Wang, Chait, & LeBoeuf, 1998), and the prevention of breast cancer

(Lee, Lee, et al., 1991). With the awareness of the health-promoting effect of isoflavones, scientists have been seeking a way to incorporate the isoflavones into daily diets,

141 especially in the western world due to the low acceptability of soy foods and eating habits

(Akdemir & Sahin, 2009; Lin et al., 2004).

Soybean Curd Residues (SCR) are wastes produced during soy processing

(O’Toole, 1999). About 800,000 tons of SCR (5% of total soy consumed) is generated annually as soybean byproduct in Japan (Li et al., 2013). However, they have potential for value-added utilization because of their high isoflavone contents (355 mg/100 DW)

(Li et al., 2013). Utilizing soybean wastes would add value to soybean farmers’ products, raising commodity prices, and reducing wastes.

A previous study demonstrated that isoflavoniods, Formononetin, Biochanin A and prunetin, from red clover (Trifolium pratense) could enhance overall color and stability of anthocyanins in muscatine grape juice (Vitis rotundifolia) and wine through intermolecular co-pigmentation (Talcott et al., 2005). However, the underlying anthocyanin-isoflavone co-pigmentation characteristics remained to be explored.

Soybean isoflavones, being more common in food than red clover isoflavones, could be potential co-pigments for anthocyanins. The objectives of this study were to elucidate the properties in anthocyanin and soybean isoflavone co-pigmentation, especially the impact of anthocyanin structures (anthocyanidin and acylation) on their interaction.

7.3 Materials and Methods

7.3.1 Materials and Reagents

Red cabbage, eggplant, and red radish were purchased from local grocery in

Columbus, OH. Black goji was bought from LianHua supermarket in Shanghai, China. 142

Chemicals and reagents used in this study were in either ACS or HPLC grade unless specified. Acetone, chloroform, methanol trifluoroacetic acid, citric acid, formic acid, HCl, sodium phosphate dibasic, sodium phosphate monobasic, sodium carbonate, sodium bicarbonate, sodium acetate, potassium hydroxide, ammonium hydroxide, hexane, and HPLC-MS grade acetonitrile were obtained from Fisher Scientific (Fair

Lawn, NJ); Ethyl acetate were purchased from Mallinckrodt Chemicals (Bedminster

Township, NJ).

7.3.2 Anthocyanin and Isoflavone Extraction

Anthocyanins from different food sources were extracted based on (Rodriguez-

Saona & Wrolstad, 2005). Nearly 50g dried fruit or vegetables were mixed with liquid nitrogen, and blended into powders, which were later rinsed with 100%(v/v) acetone acidified with 0.01% HCl. The slurry was filtrated through Whatman No.4 filter paper

(Whatman Incorporation NJ US) and re-washed using 70%(v/v) acetone acidified with

HCl until the colors were totally missing. The filtrate was then transferred to a separatory funnel and mixed with two volumes of chloroform. After gentle mixing, the samples were kept in fridge at 4C for overnight. The upper portion was collected, and the samples were roto-evaporated to remove the acetone and chloroform. For eggplant only, aqueous acetone was acidified with 1.5% trifluroacetic acid (TFA) instead to ensure that the reactive delphinidin derivatives were not oxidized.

Soybean curd residue (SCR) was made in the following procedures. Soybeans were soaked in water for overnight and peeled off. The bulky beans were then mixed with 143 water and pureed in a food processor. The slurry was filtered through cheese cloth and repeated soaking-filtration procedures. The SCR was obtained in the cheese cloth.

Soybean isoflavones were extracted from SCR according to (Lin et al., 2004).

Approximately 50-gram soybean sample was mixed with liquid nitrogen and powdered by a food blender. Methanol was added to the powder and stirred for 2 hours. Then the solutions were filtered by Whatman No.4 filter paper to remove the particles. Precipitated slurry was re-dissolved by 80% methanol and filtered. The supernatants were transferred to a separatory funnel and mixed with hexane. After standing one-night, the lower portion was collected and roto-evaporated to remove the methanol and hexane.

7.3.3 Pigment and Isoflavone Purification

Anthocyanin purifications were performed using C-18 and MCX catridge based on the methods described by Rodriguez-Saona & Wrolstad, 2005 and He & Giusti, 2011, respectively. An aqueous anthocyanin solution was forced to pass through the Sep-Pak®

C18 cartridge (Waters Corporation, Milford MA), which was first activated with two volumes of methanol. One volume of acidified water (0.01% v/v HCl) and one volume of ethyl acetate were used to wash out the sugars and phenolic compounds. Then the anthocyanin was eluted and collected by acidified methanol (0.01% v/v HCl). After roto- evaporation, purified anthocyanin was dissolved in acidified water. In the next step, the

C-18 purified anthocyanin sample was pumped to pass through Oasis® MCX cartridge

(Waters Corporation, Milford MA) activated by acidified methanol (0.1% TFA) and acidified water (0.1% TFA). The loaded sample was then washed with TFA acidified 144 water and methanol to eliminate the salts, sugars and polyphenols. The anthocyanin was recovered by methanol with 1% NH4OH, and collected in beaker containing formic acid.

The elute was then purified using C-18 based on the same method as above to remove the unwanted salts.

Similar to anthocyanin, isoflavone purification was achieved by C-18 cartridge.

After activation with methanol, the soybean isoflavone extracts were loaded to the column, and washed by water and 10% acetonitrile. The isoflavone was recovered by

80% (v/v) aqueous methanol and 80% (v/v) ethyl acetate in methanol. After roto- evaporation, isoflavone extracts were dissolved in 20% methanol.

7.3.4 Pigments Saponification

Red cabbage, Chinese eggplant, red radish and black goji contain abundant acylated anthocyanins. In order to generated non-acylated anthocyanins, alkaline hydrolyses

(saponification) were conducted to eliminate the acylation groups attached to glycosides

(M. Monica Giusti, Rodriguez-Saona, Griffin, & Wrolstad, 1999). Aqueous10% KOH was added to anthocyanin extracts in a capped container for 12 min in the dark. The solution was then neutralized by using HCl and the hydrolysate was purified by C-18 cartridge to remove the salts. After the removal of methanol by a rotary evaporator, the anthocyanins were dissolved in acidified water (0.01% v/v HCl). The completion of saponification was confirmed by HPLC.

145

7.3.5 Pigment Quantification

After extraction and saponification, the monomeric anthocyanin of each sample was quantified by the pH differential method described by Giusti and Wrolstad (2005). Buffer system were prepared using 0.1 M potassium chloride at pH 1 and 0.4 M sodium acetate at pH 4.5. The absorbance of the samples at the two pH conditions were measured at their

λmax and 700 nm respectively using UV-vis Spectrophotometer (Shimadzu corporation,

Tokyo Japan). Pigments were expressed in cyaniding-3-glucoside equivalence.

7.3.6 Pigment and Isoflavone Identification

Purified anthocyanin and isoflavone extracts were analyzed and identified by

HPLC (Shimadzu, Columbia MD) coupled to SP-M20A Photodiode Array Detector,

LCMS-2010 EV Liquid Chromatograph Mass Spectrometer, and a reverse phase

Symmetry C-18 column (5uM, 4.6*150mm) (Phenomenex, Torrance CA). All the samples were run at a flow rate of 0.8ml/min. The mobile phase for anthocyanin were solvent A: 4.5%v/v formic acid in HPLC water, and solvent B: 100% HPLC acetonitrile.

For isoflavone, the two solvents were instead 0.1% v/v acetic acid in HPLC water, and

0.1% v/v acetic acid in acetonitrile, respectively. The linear gradient used in the analysis for anthocyanin was as follow. For red cabbage, red radish and Chinese eggplant extracts:

0-1min 0-10% B, 1-46min 10%-30% B, 46-47min 30%-40% B; for black goji extracts: 0-

2min 7% B, 2-30min 5%-20% B, 30-32min 20%-60% B. While for isoflavone, the gradient was 0-25min 20%-60% B, 25-30min, 60% B. Total ion scan and selected ion

146 monitoring for anthocyanin were performed at m/z ratio of 271, 287, 303, 301, 317, and

331 (6 common anthocyanidins). Similarly, isoflavone was analyzed at m/z ratio of 270,

255, and 284 (genistein, daidzein, and glycitein)

7.3.7 Isoflavone Quantification

A series of solutions of genistein standards (in increasing concentrations) were prepared.

These solutions were investigated by HPLC, and the area under the curve (AUC) in the chromatograms was measured at 254 nm to develop a standard curve (namely the linear relationship between isoflavone concentration and HPLC AUC). The soybean isoflavone content was quantified based on this linear model and expressed as genistein equivalence.

7.3.8 Sample Preparation

Citric acid-Na2HPO4 buffers of pH 3 and 7 were prepared based on Dawson et al.,

1986. Anthocyanin samples at a concentration of 50uM were added to buffers in each pH condition. Then the solution was mixed with soybean isoflavone extracts (or isoflavone standards) based on the anthocyanin:isoflavone mole ratio of 1:0, 1:2, 1:5 and 1:10. Each mixture was stored under refrigerated storage condition (4C) in the dark over a month period.

147

7.3.9 Investigation on the Colorimetric, Spectrophotometric properties, and Stability of

Anthocyanins

After equilibrium, the color characteristics and spectrum (380 nm – 700 nm, 1 nm interval) of each anthocyanin sample were analyzed by ColorQuest XE spectrophotometer (data presented in CIE-L*C*abhab with 10 degrees observer angle, D65 illuminant), and microplate reader SpectraMax 190 (Molecular Devices, Sunnyvale CA), respectively. The monomeric anthocyanin concentration was monitored by pH differential method, and color differences were expressed as ΔE over one-month

(measured on day 0, 1, 3, 10, 17, 31).

7.4 Results and Discussion

In the following sections, we will show that soybean isoflavones interact with anthocyanins through co-pigmentation, resulting in enhanced color and stability. The co- pigmentation is dependent on anthocyanin structures. A few abbreviations will be used:

Cy, Pg, Pt, and Dp indicate cyanidin, pelargonidin, petunidin, and delphinidin, respectively; while A-Cy, A-Pg, A-Pt, and A-Dp stand for each four acylated anthocyanins. SCR represents soybean curd residue.

148

7.4.1 Anthocyanin Profiles in Red Cabbage, Red Radish, Black Goji, and Chinese

Eggplant

In order to investigate the influence of anthocyanin structure on its co- pigmentation with soybean isoflavones, pigments were extracted from red cabbage, red radish, black goji and Chinese eggplant since each of them were known to contain only one major anthocyanidin (M. Monica Giusti & Wrolstad, 2003; G. T. Sigurdson et al.,

2017; Gregory T. Sigurdson & Giusti, 2014; Zheng et al., 2011). Red cabbage, red radish, and Chinese eggplant contain A-Cy, A-Pg, and A-Dp, increasing in numbers of hydroxyl groups on the B ring of anthocyanidin structure; black goji contains A-Pt, with one hydroxyl group and one methoxyl group on the B ring.

The HPLC chromatograms and identification of the pigments are shown in

Figure 7.1. Similar to previous studies, there were primarily 8 pigments found in red cabbage, all of which were various types and numbers of acylated Cy-3-sophphoroside-5- glucoside derivatives (Ahmadiani, Robbins, Collins, & Giusti, 2014; Gregory T.

Sigurdson & Giusti, 2014). Peak 1 was non-acylated Cy-3-sophphoroside-5-glucoside.

Peak 2 - peak 5 were Cy-3-sophphoroside-5-glucoside with one acyl moiety attachment

(sinapic acid, p-coumeric acid, or ferulic acid), while peak 6 -peak 8 were Cy-3- sophphoroside-5-glucoside coupled to two acyl groups. Five major Pg derivatives were found in red radish extracts, with peak 1 being Pg-3-sophphoroside-5-glucoside, followed by two mono-acylated and two di-acylated Pg-3-sophphoroside-5-glucoside derivatives

(peak 2 – peak 5), in agreement with a previous study (M. Giusti & Wrolstad, 1996). Dp-

3-p-coumaroyl-rutinoside-5-glucoside and Pt-3-p-coumaroyl-rutinoside-5-glucoside were 149 the predominant pigments in Chinese eggplant and black goji respectively (Gregory T.

Sigurdson & Giusti, 2014; Zheng et al., 2011).

Figure 7.1 HPLC Chromatograms of anthocyanin profiles in red radish, red cabbage, Chinese eggplant, and black goji.

150

All of the four pigment sources were abundant in acylated anthocyanins (>80%).

Although differing in anthocyanidin structures, their glycosylation and acylation patterns were similar: disaccharide attachment on position 3 and monosaccharide on position 5 in anthocyanidins; additionally, they all contained pigments bearing p-coumaroyl acyl moiety on the structure block.

7.4.2 Isoflavone Profiles in SCR

The chromatogram of SCR isoflavones profile is exhibited in Figure 7.2. Similar to previously reported soybean isoflavone contents, SCR contained genistein, daidzein, glycitein, as well as their respective glycosides and malonate conjugates forms (Lin et al.,

2004; Riedl et al., 2007). The glycosylated isoflavones (≈24%) eluted fastest, as the sugar moiety brings greater polarity. Malonyl isoflavones (≈59%) had longer retention time due to the non-polar acyl groups. Isoflavone aglycones daidzein, glycitein, and genistein

(≈16%) showed greatest non-polarity and eluted last. Genistein derivatives were the most abundant isoflavones in SCR (≈51%), followed by daidzein derivatives (≈44%) and glycitein derivatives (≈5%).

151

Figure 7.2 Identification of isoflavone profiles in soybean curd residue by HPLC chromatogram.

7.4.3 Spectral Characteristics of Anthocyanins Co-Pigmented with Soybean Isoflavones

The spectrophotometric properties of anthocyanin and soybean isoflavones co- pigmentation are shown in Figure 7.3 and 7.4.

At pH 3, acylated Cy, Pg, Pt, and Dp derivatives extracted from red cabbage, red radish, black goji and Chinese eggplant exhibited a single peak in the visible absorption spectra with λmax equal to 531.5 nm, 510 nm, 524 nm, and 520.5 nm, respectively. The -

3,5-glycosylation of these anthocyanins was well documented by the low ratio of

A440/Avis-max in the spectra (Durst & Wrolstad, 2005). Co-pigmented with soybean isoflavones extracts, dose-dependent bathochromic and hyperchromic shifts were 152 observed in anthocyanin-isoflavone mixtures. Acylated Dp derivatives from Chinese eggplant (bearing three hydroxyl groups on B ring) showed the most pronounced shifts, as their λmax increased 6 nm and the absorbance at λmax boosted up to 57%

After saponification, though there was still one peak in the spectra at pH 3, these four anthocyanins underwent hypsochromic shifts. The λmax of Cy, Pg, Pt derivatives decreased to 511 nm, 497.5 nm, and 504 nm, respectively, agreeing with previous reports that acyl group in anthocyanin structure would result in bluing effects, manifested as bathochromic effect (Ahmadiani et al., 2016; M. Mónica Giusti & Wrolstad, 2003).

Saponified Dp derivatives from Chinese eggplant quickly lost their color at pH 3, leading to a flat line in its visible absorption spectra. Its λmax and corresponding absorbance were therefore unidentifiable. When mixed with soybean isoflavones, dose-dependent hypsochromic and hyperchromic effects were observed in these saponified anthocyanins.

The magnitude of hyperchromic shift was found largest in Cy derivatives (two hydroxyl groups on B-ring) from red cabbage, as the absorbance at its λmax increased 109%. To this end, anthocyanins (whether acylated or not) with more hydroxyl groups attached to the

B-ring were more reactive in anthocyanin-isoflavone co-pigmentation at pH 3.

At pH 7, A-Cy derivatives showed a broader peak along 550 - 600 nm in the absorption spectra, while A-Pg derivatives exerted a peak shoulder around 635 nm, next to its λmax. The λmax of both anthocyanins exhibited bathochromic shifts to 586.5 nm and

554.5 nm respectively as pH increased from pH 3 to pH 7. The visible absorption spectra of A-Pt and A-Dp anthocyanins rapidly changed the shapes at pH 7 compared to that at pH 3. The main peak (~532 nm) in A-Pt attenuated in its absorbance. A-Dp not only

153 displayed a bathochromic shift (λmax increased to 570 nm) when pH was raised to 7, but also showed a second peak around 450 nm. Co-pigmented with soybean isoflavones, dose-dependent bathochromic and hyperchromic shifts were found in these acylated anthocyanins. Similar to what have been found at pH 3, A-Dp from Chinese eggplant was more pronounced in bathochromic and hyperchromic effects compared to other anthocyanins tested.

Saponification of A-Pt and A-Dp anthocyanins resulted in a straight line in their visible absorption spectra, and the λmax started to fade away after being dissolved in buffer. On the other hand, saponified Cy and Pg anthocyanins generally retained their shape. Dose-dependent hypsochromic and hyperchromic shifts were observed in saponified Cy derivatives, while saponified Pg derivatives still showed bathochromic and hyperchromic effects. Saponified Cy derivatives exhibited the most hyperchromic shifts

(increased 63%), similar to that at pH 3.

Anthocyanin structure played an important role in the co-pigmentation between anthocyanin and soybean isoflavones, based on the spectrophotometric properties of the mixture at pH 3 and 7. Anthocyanin with more hydroxyl groups on the B-ring co- pigmented more with Isoflavone. It could be explained that hydroxyl groups help to form hydrogen bonds between the two flavonoid structures, favoring the formation of intermolecular co-pigmentation (Gómez-Míguez et al., 2006).

154

Acylated Anthocyanins

0.8 A-Cy 0.6 A-Pg 40 0.6 0.4 0.4

e 0.2 0.2

35 n

3 0

a 0

b 380 480 580 680 380 480 580 680

30 b 0.6 0.6 A A-Pt A-Dp 0.4 0.4 25 0.2 0.2 0 0 1:0 380 480 580 680 380 480 580 680 20 1:2 Wavelength (nm) 1:5 15 0.6 A-Cy 0.6 A-Pg 1:10 0.4 0.4

10 e 0.2 0.2

a 7 0 0

r 380 480 580 680 380 480 580 680

5 o

b 0.6 0.6 A A-Pt A-Dp 0 0.4 0.4 0 5 100.2 15 20 0.2 25 30 35

0 0 380 480 580 680 380 480 580 680 Wavelength (nm)

Figure 7.3 Spectrophotometric data of acylated anthocyanins co-pigmented with soybean isoflavones

155

Non-acylated Anthocyanins

0.4 0.4 40 0.3 Cy 0.3 Pg 0.2 0.2

e 0.1 0.1

35 n

3 0 0

b 380 480 580 680 380 480 580 680

p s 0.6 30 b 0.6

A Dp 0.4 Pt 0.4 25 0.2 0.2 0 0 1:0 380 480 580 680 380 480 580 680 20 1:2 Wavelength (nm) 1:5

15 0.6 0.6 1:10 Cy Pg 0.4 0.4

10 e 0.2 0.2

a 7 0 0

r 380 480 580 680 380 480 580 680

5 o

b 0.6 0.6

A Pt Dp 0 0.4 0.4 0 5 100.2 15 20 0.2 25 30 35

0 0 380 480 580 680 380 480 580 680 Wavelength (nm)

Figure 7.4 Spectrophotometric data of non-acylated anthocyanins co-pigmented with soybean isoflavones

156

7.4.4 Colorimetric Properties of Anthocyanins Co-Pigmented with Soybean Isoflavones

At pH 3 acylated Cy, Pg, Pt, and Dp derivatives showed red (hab = 355°), orange

(hab = 22°), red (hab = 353°), and reddish-pink (hab = 2°) colors at pH 3, respectively

(Figure 7.5 and 7.6). A dose-dependent shift towards pinkish red hue was observed in these acylated anthocyanins co-pigmented with soybean isoflavones extracts, resulting in various enhanced colors (hab = 359°, 26°, 360°, 7°, respectively for the four anthocyanin extracts). Non-acylated anthocyanins lost their vivid hues at pH 3 and showed little color enhancement effect with soybean isoflavones.

At pH 7, acylated anthocyanins exerted diverse colors (Figure 7.5 and 7.6).

While A-Pg displayed pinkish red hue (hab = 333°), A-Cy, A-Pt, and A-Dp showed purple colors (hab = 298°, 317°, 336°, respectively). When mixed with soybean isoflavones, A-Cy, A-Pt, and A-Dp underwent a dose-dependent shift towards blue hue

(hab = 287°, 293°, and 291°, respectively). A-Pg derivative showed a shift to red hue (hab

= 336°). Similar to what was found at pH 3, saponified anthocyanins quickly lost vivid color and there was limited effect introduced by soybean isoflavones.

Acylated anthocyanin showed more pleasant colors as compared to non-acylated ones at both pH 3 and pH 7. The colors were enhanced by soybean isoflavones, resulting in more attracting colors.

157

Acylated Anthocyanin Non-acylated Anthocyanin ACN:ISOF ratio 1:0 1:2 1:5 1:10 1:0 1:2 1:5 1:10

r RR

A BG

r RR

A BG

Figure 7.5 Color expression of anthocyanins when co-pigmented with soybean isoflavone extracts.

158

a (Red) 0 0 10 20 30 40 50 60 RR 1:0 1:20 pH 7 -10 1:0 Acylated CE ACN -20

1:20 -30 1:0

BG 1:0 -40 1:20

RC -50 1:20

-60 -b (Blue)

b (Yellow) 30 1:20 RR pH 3 20 1:0 Acylated ACN 10 1:20 CE -a (Green) 1:0 1:20 a (Red) 0 -80 -60 -40 -20 0 21:0 20 40 60 80 100 BG RC -10 1:0 1:0

-20

-30 -b (Blue)

Figure 7.6 Colorimetric properties of acylated anthocyanins co-pigmented with soybean isoflavones at pH 3 and 7

159

7.4.5 Anthocyanin Color Stabilities When Co-Pigmented with Soybean Isoflavones

Color changes expressed as ΔE were calculated as compared to initial color (1 hr equilibration after mixture) for each sample over 31 days (Figure 7.7, 7.8 and 7.9).

Acylated anthocyanins were quite stable at pH 3 since the ΔE of the four acylated pigments were all smaller than 5, a threshold under which untrained eyes could barely distinguish the color differences. Similarly, saponified anthocyanins exhibited little color changes (ΔE ≤ 5). The adding of soybean isoflavones seemed to have limited effect on color stability of these anthocyanins. The ΔE of A-Cy, A-Pg, and all four non-acylated anthocyanins remained the same throughout 31 days when co-pigmented with soybean isoflavones. Although soybean isoflavones in higher concentration slightly accelerated the color changes of A-Pt and A-Dp derivatives, most of these ΔE’s were still close to 5.

Besides, the trivial increase in ΔE could be due to the initial color enhancement effect by soybean isoflavones, so that the color difference in these samples over 31 days would be more significant compared to those with no color enhancement in the beginning.

At pH 7, acylated anthocyanins suffered a significant color loss over 31 days as compared to the situation at pH 3. The ΔE’s were greater than 60 for A-Pt, and A-Dp derivatives, while A-Cy and A-Pg derivatives showed better color stabilities (ΔE ≤ 30).

Non-acylated anthocyanins displayed considerable less color change with ΔE ≤ 5 (except non-acylated Pt derivatives). It was due to the fact that saponified anthocyanins were vulnerable at pH 7, and they lost their color soon after dissolved in buffers. Mixed with soybean isoflavones, acylated anthocyanins showed substantial decrease in color

160

degradation. Surprisingly, soybean isoflavones drastically stabilized the A-Pg color, as

ΔE’s

Acylated Anthocyanins

10 10 8 A-Cy 8 A-Pg

) 6 6

40 E 4 4

( 2 2

s 0 0

e 35 3

g 0 10 20 30 40 0 10 20 30 40

a p 10 10 h A-Pt A-Dp 30 8 8

r 6 6

l 4 4

25 o 2 2 C 1:0 0 0 20 0 10 20 30 40 0 10 20 30 40 1:2 Time (days) 1:5 15 1:10 40 15

30 10 ) 10

E 20

Δ 5 ( 10 A-Cy 5 s A-Pg

e 7 0 0

g 0 10 20 30 40 0 10 20 30 40

a 0 p h 100 80

C 80 0 5 10 15 2060 25 30 35 r 60

l 40 40

o 20 C 20 A-Pt A-Dp 0 0 0 10 20 30 40 0 10 20 30 40 Time (days)

Figure 7.7 Color changes (ΔE) of acylated anthocyanins co-pigmented with soybean isoflavones at pH 3 and 7 over 31 days storage test (4°C, dark).

161

Non-acylated Anthocyanins

10 10 8 Cy 8 Pg

) 6 6 40 E 4 4

( 2 2

s 3 0 0

35 g 0 10 20 30 40 0 10 20 30 40

p a 10 10

h 8 30 C 8 Dp

Pt r 6 6

o l 4 4

o 25 2 2

C 0 0 1:0 0 10 20 30 40 0 10 20 30 40 20 1:2 Time (days) 1:5 15 1:10 20 10 Cy 8 Pg 15 ) 6

10 E 10 4

( 5 2

5 e 0 0

H 0 10 20 30 40 0 10 20 30 40

0 h 40 15

C Dp 30 0 r 5 10 15 20 10 25 30 35

l 20

o 5 10 C Pt 0 0 0 10 20 30 40 0 10 20 30 40 Time (days)

Figure 7.8 Color changes (ΔE) of non-acylated anthocyanins co-pigmented with soybean isoflavones at pH 3 and 7 over 31 days storage test (4°C, dark).

162

RC RR CE BG ACN:ISOF ratio 1:0 1:2 1:5 1:10 1:0 1:2 1:5 1:10 1:0 1:2 1:5 1:10 1:0 1:2 1:5 1:10

Day0

Day1

Day3

Day10

Day17

Day31

Figure 7.9 The color stability of acylated anthocyanins co-pigmented with soybean isoflavones at pH 7 over 31 days storage test (4°C, dark).

were less than 5 through 31 days. On the other hand, the color loss of non-aclyated anthocyanins seemed to be severe with the existence of soybean isoflavones at pH 7. Due to their completely faded color in the initial mixture, soybean isoflavones enhanced the colors at the beginning, and thus provided the room for later color losses. Overall, soybean isoflavones extracts increased the color stability of acylated anthocyanin at pH 7, particularly and more pronounced with A-Pg derivatives from red radish.

7.4.6 Anthocyanin Stabilities When Co-Pigmented with Soybean Isoflavones

Monomeric anthocyanin concentration (50 uM in the initial mixture) was monitored over 31 days to evaluate anthocyanin stability when co-pigmented with soybean isoflavones (Figure 7.10 and 7.11). At pH 3 soybean isoflavones extracts exhibited limited effects on both acylated and non-acylated anthocyanins, due to the good 163 stability of anthocyanins at this pH condition (except non-acylated Dp derivatives).

Besides, acylated anthocyanins were relatively more resistant to degradation compared to their counterparts; Dp derivatives, whether acylated or not, were most reactive, and showed the worst stability among the four anthocyanin derivatives.

At pH 7, acylated anthocyanins generally were more susceptible to degradation compared to that at pH 3, especially A-Pt and A-Dp derivatives. Chinese eggplant extracts containing A-Dp anthocyanins degraded more than 80% after 24 hours. In contrast, A-Cy derivatives from red cabbage showed an extraordinary stability (20% loss). Soybean isoflavones showed a dose-dependent protective effect on A-Cy and A-Pg for the pigment degradation, as the monomeric anthocyanin concentrations of A-Cy and

A-Pg remained up to 45 uM after 31 days at pH 7 when co-pigmented with soybean isoflavones. It seemed that red radish A-Pg derivatives showed the most significant improvement in pigment stability with help from soybean isoflavones. Unfortunately, non-acylated anthocyanins degraded so fast at pH 7 that their pigment concentration dropped more than 90% soon after dissolving in buffer, and their co-pigmentation effect with soybean isoflavones was also barely noticed.

164

Acylated Anthocyanins

50 50 40 40 40 30 30 20 20 10 A-Cy 10 A-Pg

35 3 0 0

) 0 10 20 30 40 0 10 20 30 40

u 50 50

(

30 t 40 40

t 30 30 n 20 20 25 o

10 A-Pt 10 A-Dp s 1:0

i 0 0

n 0 10 20 30 40 0 10 20 30 40 20 a 1:2

o Time (days) 1:5

15 n 1:10

a 50 50

r 40 40

e 30 30 10 m

o 20 20

o 10 A-Cy 10 A-Pg

5 M 0 0 H 0 10 20 30 40 0 10 20 30 40

p 50 50 40 0 40 30 0 5 10 15 A-Pt20 30 25 30 A-Dp 35 20 20 10 10 0 0 0 10 20 30 40 0 10 20 30 40 Time (days)

Figure 7.10 Monomeric anthocyanin content (uM) of acylated anthocyanins co- pigmented with soybean isoflavones over 31 days storage test (4°C, dark).

165

35 Non-acylated Anthocyanins

)

30 u 50

(

n 40 40

e t 30 25 n 30 o 20 c 20 1:0 s 10 10 n Cy Pg

n 0 20 3 0 1:2

a 0 10 20 30 40 0 10 20 30 40

H c 1:5

h 50 50 t 1:10 15 n 40 40

a Dp

c 30 30

i r 20 20 10 e m 10 Pt 10

n 0 0

o 0 10 20 30 40 0 10 20 30 40

5 M Time (days)

0 0 Figure5 7.11 1Monomeric0 1 anthocyanin5 20 content (uM)25 of non-3acylated0 anthocyanins35 co- pigmented with soybean isoflavones over 31 days storage test (4°C, dark).

7.5 Conclusion

Isoflavones were able to modulate the color and stability of different anthocyanins at pH

3 and pH 7, in phenomena referred to as co-pigmentation. This co-pigmentation was

influenced by the structures of anthocyanins, including anthocyanidin aglycone structure,

and existence of acylation groups. Overall, delphinidin-derivatives with the most

hydroxyl groups on the B-ring exhibited the most pronounced bathochromic and

hyperchromic shifts with soybean isoflavones at both pHs. Acylated anthocyanins

showed a more pleasant color and favored co-pigmentation with isoflavones as compared

to non-acylated ones. Soybean isoflavones extracts enhanced and stabilized the color of

acylated anthocyanins. They particularly exerted more pronounced effects in stabilizing 166

A-Pg derivatives at pH 7. The results indicated that isoflavones could be used as potent co-pigments to enhance and stabilize anthocyanin colors, providing the food industry an innovative way of broadening the natural pigment application. Also, isoflavones could be obtained from soybeans, adding value to an underutilized subproduct from soybean processing. Meanwhile, it suggested a novel method for incorporation of soybean nutrition into the daily diet.

167

Chapter 8 Overall Conclusion

Petunidin-derivatives from black goji and purple potato were demonstrated to be promising candidates for natural colorants. Their co-pigmentation with metal ions and soybean isoflavones could result in various enhanced colorations and stabilities.

Black goji extracts contained abundant petunidin-derivatives, with petunidin-3- trans-p-coumaroyl-rutinoside-5-glucoside being the main pigment (~80% of total pigment). The colorimetric and spectrophotometric traits of black goji anthocyanins were influenced by purity, acylation, and acyl moiety orientations. The MCX cartridge removed a large proportion of polyphenolics from the fruit extracts, and attenuated the saturation of color expression. Petunidin-3-trans-p-coumaroyl-rutinoside-5-glucoside contributed most of the black goji anthocyanin color properties and showed a better color stability compared to other extracts over time. Acylation not only strengthened the color retention in mildly acidic condition, but also enhanced the tinctorial strength and stability of pigments. The spectrophotometric and colorimetric properties of the petunidin-3-p- coumaroyl-rutinoside-5-glucoside in black goji were significantly influenced by the stereochemistry of the coumaric acid acylation. Pigments acylated with the cis isomer showed greater λmax over a wide range of pH, expressing bluer hues than the trans isomer. Pigments acylated with the trans isomer, on the other hand, exhibited improved color stabilities over time. Thus, the cis and trans acylated anthocyanins could be applied

168 to food to obtain different hues depending on the type of product and the target coloration. Nevertheless, this dissimilarity shed light on the chemical attributes that could be used to manipulate the color expression in natural pigment applications.

UV irradiation (254 nm) induced dose-dependent trans to cis isomerization in petunidin-3-p-coumaroyl-rutinoside-5-glucoside isolated from black goji. The trans isolate in quartz cuvettes (usable range 190-2500 nm) reached a plateau phase soon after

UV exposure (7.64 J), with a cis:trans ratio of 4:6; while in glass cuvettes (usable range

340-2500 nm) more energy (30.48 J) was needed in order to get the plateau stage with a cis : trans ratio of 55:45. The isomerization from trans to cis petunidin-3-p-coumaroyl- rutinoside-5-glucoside caused decrease in hue angles, leading to bluer hues at neutral and alkaline conditions. UV-irradiation could be a potential method to obtain cis isomeric coumaroyl anthocyanins.

Similar to black goji, purple potato extracts contained abundant petunidin- derivatives, with petunidin-3-(p-coumaroyl)-rutinoside-5-glucoside as the predominant pigment. The extracts displayed vivid red, purple, violet-blue colors from acidic, neutral, to alkaline pH, respectively. In the presence of metal ions, Al3+ and Fe3+, the petunidin- derivatives from black goji and purple potato experienced bathochromic (up to 79 nm) and hyperchromic shifts (up to 60% increase in absorbance at λmax), resulting in various blue to green hues and intensified colors in neutral and alkaline conditions. The magnitudes of the metal chelation effect were dependent on metal ions, pH, and Pt:M3+ ratios. Fe3+ with a higher spin configuration in the outer electrons triggered further bathochromic shift than that of Al3+, and its unique yellow color in aqueous system led to

169 green hues when combined with petunidin-derivative’s blue colors. The oxidative capability of Fe3+ towards flavonoids limited its application in larger concentration (>125 uM). Al3+ in low concentration (~12.5 uM) was able to chelate Pt (25uM) and produce blue colors at pH 8 and 9. The concentration of [Al3+] needed for blue production decreased upon the increase of pH. The chelation of metal ions significantly enhanced the color stability (up to 4 times) of petunidin-derivatives over neutral and alkaline conditions.

Soybean isoflavones were able to interact with anthocyanins through co- pigmentation at pH 3 and pH 7. This co-pigmentation was influenced by the structures of anthocyanins, including the anthocyanidin basic structure, and existence of acylation group. Overall, delphinidin derivatives with the most hydroxyl groups on the B-ring exhibited the most pronounced bathochromic and hyperchromic shifts with soybean isoflavones at both pHs. Acylated anthocyanin showed a more pleasant color and favored co-pigmentation with soybean isoflavones as compared to non-acylated ones. Soybean isoflavones extracts enhanced and stabilized the color of acylated anthocyanin. They particularly exerted more pronounced effect in stabilizing acylated pelargonidin- derivatives at pH 7.

This study demonstrated petunidin-derivatives from black goji and purple potato as promising sources for natural colorants, producing various vivid hues over a wide range of pH. Chelated with metals, these pigments could provide the food industry with various vivid violet, blue, and green colors. The co-pigmentation with soybean

170 isoflavones could enhance and stabilize anthocyanin colors, revealing an innovative way of broadening the soybean application.

171

Bibliography

Abdullah, R., Lee, P. M., & Lee, K. H. (2010). Multiple color and pH stability of floral anthocyanin extract: Clitoria ternatea. CSSR 2010 - 2010 International Conference on Science and Social Research, (09), 254–258.

Aberoumand, A. (2011). A Review Article on Edible Pigments Properties and Sources as Natural Biocolorants in Foodstuff and Food Industry. World Journal of Dairy and Food Sciences, 6(22), 71–78.

Acero, S., Tabar, a I., Alvarez, M. J., Garcia, B. E., Olaguibel, J. M., & Moneo, I. (1998). Occupational asthma and food allergy due to carmine. Allergy, 53(9), 897– 901. http://www.ncbi.nlm.nih.gov/pubmed/9788693

Adlercreutz, H., Honjo, H., Higashi, A., Fotsis, T., Hämäläinen, E., Hasegawa, T., & Okada, H. (1991). Urinary excretion of lignans and isoflavonoid phytoestrogens in Japanese men and women consuming a traditional Japanese diet. The American Journal of Clinical Nutrition, 54(6), 1093–100. http://www.ncbi.nlm.nih.gov/pubmed/1659780

Ahmadiani, N. (2012). Master of Science Thesis: Anthocyanin Based Blue Colorants. Retrieved from https://etd.ohiolink.edu/pg_10?0::NO:10:P10_ETD_SUBID:76437

Ahmadiani, N., Robbins, R. J., Collins, T. M., & Giusti, M. M. (2014). Anthocyanins contents, profiles, and color characteristics of red cabbage extracts from different cultivars and maturity stages. Journal of Agricultural and Food Chemistry, 62(30), 7524–7531. http://doi.org/10.1021/jf501991q

Ahmadiani, N., Robbins, R. J., Collins, T. M., & Giusti, M. M. (2016). Molar absorptivity (ε) and spectral characteristics of cyanidin-based anthocyanins from red cabbage. Food Chemistry, 197, 900–906. http://doi.org/10.1016/j.foodchem.2015.11.032

Akazome, Y. (2004). Characteristics and physiological functions of polyphenols from apples. Biofactors, 22, 311–314.

Akdemir, F., & Sahin, K. (2009). Genistein supplementation to the quail: effects on egg production and egg yolk genistein, daidzein, and lipid peroxidation levels. Poultry Science, 88(10), 2125–31. http://doi.org/10.3382/ps.2009-00004

Andersen, Ø. M., & Jordheim, M. (2014). Basic Anthocyanin Chemistry and Dietary Sources. In T. C. Wallace & M. M. Giusti (Eds.), Anthocyanins in Health and Disease (1st ed., pp. 13–90). Boca Raton, FL: CRC Press (Taylor & Francis Group). 172

Anderson, J. W., Diwadkar, V. A., & Bridges, S. R. (1998). Selective effects of different antioxidants on oxidation of lipoproteins from rats. Proceedings of the Society for Experimental Biology and Medicine. Society for Experimental Biology and Medicine (New York, N.Y.), 218(4), 376–81. http://www.ncbi.nlm.nih.gov/pubmed/9714083

Anderson, J. W., Johnstone, B. M., & Cook-Newell, M. E. (1995). Meta-analysis of the effects of soy protein intake on serum lipids. The New England Journal of Medicine, 333(5), 276–82. http://doi.org/10.1056/NEJM199508033330502

Asfaw, N., & Demissew, S. (1994). Phytochemical Dictionary: A Handbook of Bioactive Compounds from Plants. Economic Botany, 48(3), 258–258. http://doi.org/10.1007/BF02862326

Atkinson, C., Compston, J. E., Day, N. E., Dowsett, M., & Bingham, S. A. (2004). The effects of phytoestrogen isoflavones on bone density in women: a double-blind, randomized, placebo-controlled trial. Am J Clin Nutr, 79(2), 326–333. Retrieved from http://ajcn.nutrition.org/content/79/2/326.long

Axelson, M., Kirk, D. N., Farrant, R. D., Cooley, G., Lawson, a M., & Setchell, K. D. (1982). The identification of the weak oestrogen equol [7-hydroxy-3-(4’- hydroxyphenyl)chroman] in human urine. The Biochemical Journal, 201(2), 353– 357. http://doi.org/10.1042/bj2010353

Azeke, M. A., & Ekpo, K. E. (2008). Egg yolk cholesterol lowering effects of garlic and tea. Journal of Biological Sciences, 8(2), 456–460. http://doi.org/10.3923/jbs.2008.456.460

Bakker, J., & Timberlake, C. F. (1997). Isolation , Identification , and Characterization of New Color-Stable Anthocyanins Occurring in Some Red Wines. Journal of Agricultural and Food Chemistry, 8561(96), 35–43. http://doi.org/10.1021/jf960252c

Bayer, E., Egeter, H., Fink, A., Nether, K., & Wegmann, K. (1966). Complex Formation and Flower Colors. Angewandte Chemie International Edition in English, 5(9), 791– 798.

Becaria, a, Campbell, a, & Bondy, S. C. (2002). Aluminum as a toxicant. Toxicology and Industrial Health, 18(7), 309–320. http://doi.org/10.1191/0748233702th157oa

Birt, D. F., Hendrich, S., & Wang, W. (2001). Dietary agents in cancer prevention: Flavonoids and isoflavonoids. Pharmacology and Therapeutics. 90(2-3), 157-177. http://doi.org/10.1016/S0163-7258(01)00137-1

173

Block, E. (2010). Garlic and Other Alliums, The Lore and the Science (1st ed.). Cambridge, United Kingdom: The Royal Society of Chemistry

Blouin, F. A., & Cherry, J. P. (1980). Identification of color-causing pigments in biscuits containing cottonseed flour. Journal of Food Science, 45, 953–961.

Boersma, B. J., D’Alessandro, T., Benton, M. R., Kirk, M., Wilson, L. S., Prasain, J., … Patel, R. P. (2003). Neutrophil myeloperoxidase chlorinates and nitrates soy isoflavones and enhances their antioxidant properties. Free Radic Biol Med, 35(11), 1417–1430. http://doi.org/10.1016/j.freeradbiomed.2003.08.009

Boersma, B. J., D’Alessandro, T., Benton, M. R., Kirk, M., Wilson, L. S., Prasain, J., … Patel, R. P. (2003). Neutrophil myeloperoxidase chlorinates and nitrates soy isoflavones and enhances their antioxidant properties. Free Radical Biology & Medicine, 35(11), 1417–30. http://www.ncbi.nlm.nih.gov/pubmed/14642389

Breithaupt, D. E., Wirt, U., & Bamedi, A. (2002). Differentiation between lutein monoester regioisomers and detection of lutein diesters from marigold flowers (Tagetes erecta L.) and several fruits by liquid chromatography-mass spectrometry. Journal of Agricultural and Food Chemistry, 50(1), 66–70. http://doi.org/10.1021/jf010970l

Brink, E., Coxam, V., Robins, S., Wahala, K., Cassidy, A., & Branca, F. (2008). Long- term consumption of isoflavone-enriched foods does not affect bone mineral density, bone metabolism, or hormonal status in early postmenopausal women: a randomized, double-blind, placebo controlled study. The American Journal of Clinical Nutrition, 87(3), 761–70. http://www.ncbi.nlm.nih.gov/pubmed/18326616

Brouillard, R., & Delaporte, B. (1977a). Chemistry of Anthocyanin Pigments 2.1 Kinetic and Thermodynamic Study of Proton Transfer, Hydration, and Tautomeric Reactions of Malvidin 3-Glucoside. Journal of the American Chemical Society, 99, 8461–8468. http://doi.org/0.1021/ja00468a015

Brouillard, R., Delaporte, B., & Dubois, J. E. (1978). Chemistry of Anthocyanin Pigments. 3.1 Relaxation Amplitudes in pH-Jump Experiments. Journal of the American Chemical Society, 100(19), 6202–6205. http://doi.org/10.1021/ja00487a041

Brouillard, R., & Dubois, J. E. (1977). Mechanism of the Structural Transformations of Anthocyanins in Acidic Media. Journal of the American Chemical Society, 99(5), 1359–1364. http://doi.org/10.1021/ja00447a012

Brouillard, R., Mazza, G., Saad, Z., Albrecht-Gary, A. M., & Cheminat, A. (1989). The Copigmentation Reaction of Anthocyanins: A Microprobe for the Structural Study 174

of Aqueous Solutions. Journal of the American Chemical Society, 111(7), 2604– 2610. http://doi.org/10.1021/ja00189a039

Buchweitz, M., Brauch, J., Carle, R., & Kammerer, D. R. (2013a). Application of ferric anthocyanin chelates as natural blue food colorants in polysaccharide and gelatin based gels. Food Research International, 51(1), 274–282. http://doi.org/10.1016/j.foodres.2012.11.030

Buchweitz, M., Brauch, J., Carle, R., & Kammerer, D. R. (2013b). Colour and stability assessment of blue ferric anthocyanin chelates in liquid pectin-stabilised model systems. Food Chemistry, 138(2-3), 2026–2035.

Buchweitz, M., Carle, R., & Kammerer, D. R. (2012). Bathochromic and stabilising effects of sugar beet pectin and an isolated pectic fraction on anthocyanins exhibiting pyrogallol and catechol moieties. Food Chemistry, 135(4), 3010–3019. http://doi.org/10.1016/j.foodchem.2012.06.101

Burrows, A. (2009). Palette of our palates: A brief history of food coloring and its regulation. Comprehensive Reviews in Food Science and Food Safety, 8(4), 394– 408.

Cassidy, A., Bingham, S., & Setchell, K. (1994). Biological effects of a diet of soy protein rich in isoflavones on the menstrual cycle of premenopausal women. American Journal of Clinical Nutrition, 60(3), 333–340. http://ajcn.nutrition.org/content/60/3/333.abstract

Cavalcanti, R. N., Santos, D. T., & Meireles, M. A. A. (2011). Non-thermal stabilization mechanisms of anthocyanins in model and food systems-An overview. Food Research International. http://doi.org/10.1016/j.foodres.2010.12.007

Chappel, C. I., & Howell, J. C. (1992). Caramel colours--a historical introduction. Food and Chemical Toxicology : An International Journal Published for the British Industrial Biological Research Association, 30(5), 351–357.

Chen, Y.-M., Ho, S. C., Lam, S. S. H., Ho, S. S. S., & Woo, J. L. F. (2003). Soy isoflavones have a favorable effect on bone loss in Chinese postmenopausal women with lower bone mass: a double-blind, randomized, controlled trial. The Journal of Clinical Endocrinology and Metabolism, 88(10), 4740–7. http://doi.org/10.1210/jc.2003-030290

Chisté, R. C., Lopes, A. S., & de Faria, L. J. G. (2010). Thermal and light degradation kinetics of anthocyanin extracts from mangosteen peel (Garcinia mangostana L.). International Journal of Food Science and Technology, 45(9), 1902–1908. http://doi.org/10.1111/j.1365-2621.2010.02351.x 175

Cooper, C., Campion, G., & Melton, L. J. (1992). Hip fractures in the elderly: a world- wide projection. Osteoporosis International : A Journal Established as Result of Cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA, 2(6), 285–9. http://www.ncbi.nlm.nih.gov/pubmed/1421796

Cortez, R., Luna-Vital, D. A., Margulis, D., & Gonzalez de Mejia, E. (2017). Natural Pigments: Stabilization Methods of Anthocyanins for Food Applications. Comprehensive Reviews in Food Science and Food Safety, 16(1), 180–198. http://doi.org/10.1111/1541-4337.12244

Coward, L., Smith, M., Kirk, M., & Barnes, S. (1998). Chemical modification of isoflavones in soyfoods during cooking and processing. The American Journal of Clinical Nutrition, 68(6 Suppl), 1486S–1491S. http://www.ncbi.nlm.nih.gov/pubmed/9848521

Dangles, O., Elhabiri, M., & Brouillard, R. (1994). Kinetic and Thermodynamic Investigation of the Aluminium-Anthocyanin Complexation in Aqueous Solution. Journal of the Chemical Society, Perkin Transactions 2, (12), 2587–2596. http://doi.org/10.1039/p29940002587

Dangles, O., Stoeckel, C., Wigand, M. C., & Brouillard, R. (1992). Two very distinct types of anthocyanin complexation: Copigmentation and inclusion. Tetrahedron Letters, 33(36), 5227–5230. http://doi.org/10.1016/S0040-4039(00)79139-8

Danuri, H. (2008). Optimizing Angkak Pigments and Lovastatin Production By Monascus purpureus. Biosciences, 15(2), 61–66.

Dawson, R. M. C., Elliott, D. C., Elliott, W. H., & Jones, K. M. (1986). Data for biochemical research. Data for biochemical research. http://doi.org/10.1016/0307- 4412(87)90110-5

Day, A. J., Dupont, M. S., Ridley, S., Rhodes, M., Rhodes, M. J. c., Morgan, M. R. a., & Williamson, G. (1998). Deglycosylation of flavonoid and isoflavonoid glycosides by human small intestine and liver β-glucosidase activity. FEBS Letters, 436(1), 71–75. http://doi.org/10.1016/S0014-5793(98)01101-6 de Rijke, E., Out, P., Niessen, W. M. A., Ariese, F., Gooijer, C., & Brinkman, U. A. T. (2006). Analytical separation and detection methods for flavonoids. Journal of Chromatography. A, 1112(1-2), 31–63. http://doi.org/10.1016/j.chroma.2006.01.019

Delgado-Vargas, F., Jiménez, A. R., & Paredes-López, O. (2000). Natural Pigments: Carotenoids, Anthocyanins, and Betalains — Characteristics, Biosynthesis, 176

Processing, and Stability. Critical Reviews in Food Science and Nutrition (Vol. 40). 40(3), 173-289.

Delmonte, P., Perry, J., & Rader, J. I. (2006). Determination of isoflavones in dietary supplements containing soy, Red Clover and kudzu: extraction followed by basic or acid hydrolysis. Journal of Chromatography. A, 1107(1-2), 59–69. http://doi.org/10.1016/j.chroma.2005.11.060

Di Meo, F., Sancho Garcia, J. C., Dangles, O., & Trouillas, P. (2012). Highlights on anthocyanin pigmentation and copigmentation: A matter of flavonoid pi-stacking complexation to be described by DFT-D. Journal of Chemical Theory and Computation, 8(6), 2034–2043.

Dixon, R. A., & Paiva, N. L. (1995). Stress-Induced Phenylpropanoid Metabolism. The Plant Cell, 7(7), 1085. http://doi.org/10.2307/3870059

Djuric, Z., Chen, G., Doerge, D. R., Heilbrun, L. K., & Kucuk, O. (2001). Effect of soy isoflavone supplementation on markers of oxidative stress in men and women. Cancer Letters, 172(1), 1–6. http://www.ncbi.nlm.nih.gov/pubmed/11595123

Downham, A., & Collins, P. (2000). Colouring our foods in the last and next millennium. International Journal of Food Science and Technology, 35(1), 5–22.

Duncan, A. M., Phipps, W. R., & Kurzer, M. S. (2003). Phyto-oestrogens. Best Practice & Research Clinical Endocrinology & Metabolism, 17(2), 253–271. http://doi.org/10.1016/S1521-690X(02)00103-3

Durst, R. W., & Wrolstad, R. E. (2005). Separation and Characterization of Anthocyanins by HPLC. In Handbook of Food Analytical Chemistry. Vol. 2–2, pp. 33–45 EDITORS. http://doi.org/10.1002/0471709085.ch18

Duthie, G., & Crozier, A. (2000). Plant-derived phenolic antioxidants. Current Opinion in Lipidology ;3(6):447-51 http://doi.org/10.1097/00041433-200002000-00007

EFSA. (2012). Refined exposure assessment for caramel colours (E 150a, c, d) 1. EFSA Journal, 10(12), 3030.

Eiro, M. J., & Heinonen, M. (2002). Anthocyanin color behavior and stability during storage: effect of intermolecular copigmentation. Journal of Agricultural and Food Chemistry, 50(25), 7461–6. http://doi.org/10.1021/jf0258306

Eldridge, A. C. (1982). Determination of isoflavones in soybean flours, protein concentrates, and isolates. Journal of Agricultural and Food Chemistry, 30(2), 353– 355. http://doi.org/10.1021/jf00110a035 177

European Food Safety Authority (EFSA). (2011). Scientific opinion on re-evaluation of calcium carbonate (E 170) as a food additive. EFSA Journal, 9(7), 1–73.

European Food Safety Authority (EFSA). (2012). Scientific Opinion on the re-evaluation of vegetable carbon ( E 153 ) as a food additive. EFSA Journal, 10(4), 2592–2626.

Ezekiel, R., Singh, N., Sharma, S., & Kaur, A. (2013). Beneficial phytochemicals in potato - a review. Food Research International, 50(2), 487–496. http://doi.org/10.1016/j.foodres.2011.04.025

Farr, JE; Srivastava, A; Machiraju, R; Giusti, M. (2017). ColorBySpectra: an application to automate conversion of spectral data to color spaces. Software X.

Figallo, V. B. (2003). Effect of planting time and location on the isoflavone content and profile of different soybeans cultivars. M.S. Thesis, Nutrition and Food Science Department, Univerisity of Maryland http://drum.lib.umd.edu/handle/1903/106

Finegold, J. A., Asaria, P., & Francis, D. P. (2013). Mortality from ischaemic heart disease by country, region, and age: statistics from World Health Organisation and United Nations. International Journal of Cardiology, 168(2), 934–45. http://doi.org/10.1016/j.ijcard.2012.10.046

Fossen, T., Rayyan, S., Holmberg, M. H., Nimtz, M., & Andersen, O. M. (2007). Covalent anthocyanin-flavone dimer from leaves of Oxalis triangularis. Phytochemistry, 68(5), 652–62. http://doi.org/10.1016/j.phytochem.2006.10.030

Foti, P., Erba, D., Riso, P., Spadafranca, A., Criscuoli, F., & Testolin, G. (2005). Comparison between daidzein and genistein antioxidant activity in primary and cancer lymphocytes. Archives of Biochemistry and Biophysics, 433(2), 421–7. http://doi.org/10.1016/j.abb.2004.10.008

Fotsis, T., Pepper, M., Adlercreutz, H., Fleischmann, G., Hase, T., Montesano, R., & Schweigerer, L. (1993). Genistein, a dietary-derived inhibitor of in vitro angiogenesis. Proceedings of the National Academy of Sciences of the United States of America, 90(7), 2690–4. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=46161&tool=pmcentrez &rendertype=abstract

Fulcrand, H., Benabdeljalil, C., Rigaud, J., Cheynier, V., & Moutounet, M. (1998). A new class of wine pigments generated by reaction between pyruvic acid and grape anthocyanins. Phytochemistry, 47(7), 1401–1407. http://doi.org/10.1016/S0031- 9422(97)00772-3 178

Gastineau, R., Turcotte, F., Pouvreau, J. B., Morancais, M., Fleurence, J., Windarto, E., … Mouget, J. L. (2014). Marennine, promising blue pigments from a widespread Haslea diatom species complex. Marine Drugs, 12(6), 3161–3189.

George, F., Figueiredo, P., Toki, K., Tatsuzawa, F., Saito, N., & Brouillard, R. (2001). Influence of trans-cis isomerisation of coumaric acid substituents on colour variance and stabilisation in anthocyanins. Phytochemistry, 57(5), 791–795. http://doi.org/10.1016/S0031-9422(01)00105-4

Ghosh, P., & Fenner, G. P. (1999). Improved Method for Gas Chromatographic Analysis of Genistein and Daidzein from Soybean ( Glycine max ) Seeds. Journal of Agricultural and Food Chemistry, 47(9), 3455–3456. http://doi.org/10.1021/jf990157e

Gilbert, P. U. P. ., & Haeberli, W. (2008). Physics in the Arts. Burlington, MA: Elsevier Academic Press.

Giusti, M. M., Polit, M. F., Ayvaz, H., Tay, D., & Manrique, I. (2014). Characterization and quantitation of anthocyanins and other phenolics in native andean potatoes. Journal of Agricultural and Food Chemistry, 62(19), 4408–4416. http://doi.org/10.1021/jf500655n

Giusti, M. M., Rodriguez-Saona, L. E., Griffin, D., & Wrolstad, R. E. (1999). Electrospray and tandem mass spectroscopy as tools for anthocyanin characterization. Journal of Agricultural and Food Chemistry, 47(11), 4657–4664. http://doi.org/10.1021/jf981242+

Giusti, M. M., Rodríguez-Saona, L. E., & Wrolstad, R. E. (1999). Molar absorptivity and color characteristics of acylated and non- acylated pelargonidin-based anthocyanins. Journal of Agricultural and Food Chemistry, 47(11), 4631–4637. http://doi.org/10.1021/jf981271k

Giusti, M. M., & Wrolstad, R. E. (2003). Acylated anthocyanins from edible sources and their applications in food systems. In Biochemical Engineering Journal (Vol. 14, pp. 217–225). http://doi.org/10.1016/S1369-703X(02)00221-8

Giusti, M., & Wrolstad, R. (1996). Characterization of red radish anthocyanins. Journal of Food Science, 61(2), 322–326. http://doi.org/10.1111/j.1365- 2621.1996.tb14186.x 179

Gómez-Míguez, M., González-Manzano, S., Teresa Escribano-BailóN, M., Heredia, F. J., & Santos-Buelga, C. (2006). Influence of different phenolic copigments on the color of malvidin 3-glucoside. Journal of Agricultural and Food Chemistry, 54(15), 5422–5429. http://doi.org/10.1021/jf0604586

Handbook of Nutraceuticals and Functional Foods, Second Edition - CRC Press Book. http://www.crcpress.com/product/isbn/9780849364099

He, J., & Giusti, M. M. (2010). Anthocyanins: Natural Colorants with Health-Promoting Properties. Annual Review of Food Science and Technology, 1(1), 163–187. http://doi.org/10.1146/annurev.food.080708.100754

He, J., & Giusti, M. M. (2011). High-purity isolation of anthocyanins mixtures from fruits and vegetables - A novel solid-phase extraction method using mixed mode cation-exchange chromatography. Journal of Chromatography A, 1218(44), 7914– 7922. http://doi.org/10.1016/j.chroma.2011.09.005

Herbach, K. M., Stintzing, F. C., & Carle, R. (2006). Betalain Stability and Degradation- Structural and Chromatic Aspects. Journal of Food Science, 71(4), R41–R50. http://doi.org/10.1111/j.1750-3841.2006.00022.x

Herbarom Ingredients GmbH. (2013). Herbarom - Apple extracts for natural colouring. Anuga Food Tec Quarterly, 3, 1–2.

Heredia, F. J., Francia-Aricha, E. M., Rivas-Gonzalo, J. C., Vicario, I. M., & Santos- Buelga, C. (1998). Chromatic characterization of anthocyanins from red grapes - I. pH effect. Food Chemistry, 63(4), 491–498. http://doi.org/10.1016/S0308- 8146(98)00051-X

Hertog, M. G., Kromhout, D., Aravanis, C., Blackburn, H., Buzina, R., Fidanza, F., … Nedeljkovic, S. (1995). Flavonoid intake and long-term risk of coronary heart disease and cancer in the seven countries study. Archives of Internal Medicine, 155(4), 381–386. http://www.ncbi.nlm.nih.gov/pubmed/7848021

Hirose, K., Tajima, K., Hamajima, N., Inoue, M., Takezaki, T., Kuroishi, T., … Tokudome, S. (1995). A large-scale, hospital-based case-control study of risk factors of breast cancer according to menopausal status. Japanese Journal of Cancer Research : Gann, 86(2), 146–54. http://www.ncbi.nlm.nih.gov/pubmed/7730137

Hodgson, J. M., Puddey, I. B., Croft, K. D., Mori, T. A., Rivera, J., & Beilin, L. J. (1999). Isoflavonoids do not inhibit in vivo lipid peroxidation in subjects with high-normal blood pressure. Atherosclerosis, 145(1), 167–172. http://doi.org/10.1016/S0021- 9150(99)00029-5 180

Hosokawa, K. (1995). Seven acylated anthocyanins in the blue flowers of Hyacinthus orientalis. Phytochemistry, 38(5), 1293–1298. http://doi.org/10.1016/0031- 9422(94)00778-R

Humphrey, A. M. (2004). Chlorophyll as a Color and Functional Ingredient. Journal of Food Science, 69(5), C422–C425.

Ichiyanagi, T., Kashiwada, Y., Shida, Y., Ikeshiro, Y., Kaneyuki, T., & Konishi, T. (2005). Nasunin from Eggplant Consists of Cis − Trans Isomers of Delphinidin 3-[4- ( p- Coumaroyl)- l -rhamnosyl (1→6)glucopyranoside]-5-glucopyranoside. Journal of Agricultural and Food Chemistry, 53(24), 9472–9477. http://doi.org/10.1021/jf051841y

Inami, O., Tamura, I., Kikuzaki, H., & Nakatami, N. (1996). Stability of anthocyanins of Sambucus canadensis and Sambucus nigra. Journal of Agricultural and Food Chemistry, 44(10), 3090–3096. http://doi.org/10.1021/jf9507809

Izumi, T., Piskula, M. K., Osawa, S., Obata, A., Tobe, K., Saito, M., … Kikuchi, M. (2000). Soy isoflavone aglycones are absorbed faster and in higher amounts than their glucosides in humans. The Journal of Nutrition, 130(7), 1695–9. http://www.ncbi.nlm.nih.gov/pubmed/10867038

Jespersen, L., Strømdahl, L. D., Olsen, K., & Skibsted, L. H. (2005). Heat and light stability of three natural blue colorants for use in confectionery and beverages. European Food Research and Technology, 220(3-4), 261–266.

Jin, H., Liu, Y., Guo, Z., Yang, F., Wang, J., Li, X., … Liang, X. (2015). High- performance liquid chromatography separation of cis - Trans anthocyanin isomers from wild Lycium ruthenicum Murr. employing a mixed-mode reversed- phase/strong anion-exchange stationary phase. Journal of Agricultural and Food Chemistry, 63(2), 500–508. http://doi.org/10.1021/jf504525w

Joshi, V. K., Attri, D., Bala, A., & Bhushan, S. (2003). Microbial Pigments. Indian Journal of Biotechnology, 2, 362–369.

Kirk, E. A., Sutherland, P., Wang, S. A., Chait, A., & LeBoeuf, R. C. (1998). Dietary Isoflavones Reduce Plasma Cholesterol and Atherosclerosis in C57BL/6 Mice but not LDL Receptor–Deficient Mice. The Journal of Nutrition, 128(6), 954–959.

Kirk, E. A., Sutherland, P., Wang, S. A., Chait, A., & LeBoeuf, R. C. (1998). Dietary isoflavones reduce plasma cholesterol and atherosclerosis in C57BL/6 mice but not LDL receptor-deficient mice. The Journal of Nutrition, 128(6), 954–9. http://www.ncbi.nlm.nih.gov/pubmed/9614153 181

Koehler, K. (2014). Composition Comprising Calcium Carbonate as a White Pigment. U.S. Patent No. 8,765,180. Washington, DC: U.S. Patent and Trademark Office.

Kudou, S., Fleury, Y., Welti, D., Magnolato, D., Uchida, T., Kitamura, K., & Okubo, K. (1991). Malonyl Isoflavone Glycosides in Soybean Seeds. Agricultural and Biological Chemistry, 55(9), 2227–2233. http://ci.nii.ac.jp/naid/110006325717/en/

Kühnau, J. (1976). The flavonoids. A class of semi-essential food components: their role in human nutrition. World Review of Nutrition and Dietetics, 24, 117–91. http://www.ncbi.nlm.nih.gov/pubmed/790781

Kunsági-Máté, S., Szabó, K., Nikfardjam, M. P., & Kollár, L. (2006). Determination of the thermodynamic parameters of the complex formation between malvidin-3-O- glucoside and polyphenols. Copigmentation effect in red wines. Journal of Biochemical and Biophysical Methods, 69(1-2), 113–119. http://doi.org/10.1016/j.jbbm.2006.03.014

Lachman, J., Hamouz, K., Orsák, M., Pivec, V., & Dvořák, P. (2008). The influence of flesh colour and growing locality on polyphenolic content and antioxidant activity in potatoes. Scientia Horticulturae, 117(2), 109–114. http://doi.org/10.1016/j.scienta.2008.03.030

Lakshmi, G. C. (2014). Food Coloring: The Natural Way. Research Journal of Chemical Sciences, 4(2), 87–96.

Lamartiniere, C. A., Moore, J. B., Brown, N. M., Thompson, R., Hardin, M. J., & Barnes, S. (1995). Genistein suppresses mammary cancer in rats. Carcinogenesis, 16(11), 2833–40. http://www.ncbi.nlm.nih.gov/pubmed/7586206

Lee, H. P., Gourley, L., Duffy, S. W., Estéve, J., Lee, J., & Day, N. E. (1991). Dietary effects on breast-cancer risk in Singapore. Lancet, 337(8751), 1197–200. http://www.ncbi.nlm.nih.gov/pubmed/1673746

León-Carmona, J. R., Galano, A., & Alvarez-Idaboy, J. R. (2016). Deprotonation routes of anthocyanidins in aqueous solution, pK a values, and speciation under physiological conditions. RSC Advances, 6(58), 53421–53429. http://doi.org/10.1039/C6RA10818K

Li, S., Zhu, D., Li, K., Yang, Y., Lei, Z., & Zhang, Z. (2013). Soybean Curd Residue: Composition, Utilization, and Related Limiting Factors. ISRN Industrial Engineering, 2013, 1–8. http://doi.org/10.1155/2013/423590

Lin, F., Wu, J., Abdelnabi, M. a, Ottinger, M. a, & Giusti, M. M. (2004). Effects of dose 182

and glycosylation on the transfer of genistein into the eggs of the Japanese quail (Coturnix japonica). Journal of Agricultural and Food Chemistry, 52(8), 2397–403. http://doi.org/10.1021/jf034921f

Liu, K. (1997). Chemistry and Nutritional Value of Soybean Components. Boston, MA: Springer US. http://doi.org/10.1007/978-1-4615-1763-4

Lu, L. J., Anderson, K. E., Grady, J. J., & Nagamani, M. (1996). Effects of soya consumption for one month on steroid hormones in premenopausal women: implications for breast cancer risk reduction. Cancer Epidemiology, Biomarkers & Prevention : A Publication of the American Association for Cancer Research, Cosponsored by the American Society of Preventive Oncology, 5(1), 63–70. http://www.ncbi.nlm.nih.gov/pubmed/8770469

Lydeking-Olsen, E., Beck-Jensen, J.-E., Setchell, K. D. R., & Holm-Jensen, T. (2004). Soymilk or progesterone for prevention of bone loss--a 2 year randomized, placebo- controlled trial. European Journal of Nutrition, 43(4), 246–57. http://doi.org/10.1007/s00394-004-0497-8

Makris, D. P., & Rossiter, J. T. (2000). Heat-induced, metal-catalyzed oxidative degradation of quercetin and rutin (quercetin 3-O-rhamnosylglucoside) in aqueous model systems. Journal of Agricultural and Food Chemistry, 48(9), 3830–3838. http://doi.org/10.1021/jf0001280

Malien-Aubert, C., Dangles, O., & Amiot, M. J. (2001). Color stability of commercial anthocyanin-based extracts in relation to the phenolic composition. Protective effects by intra- and intermolecular copigmentation. Journal of Agricultural and Food Chemistry, 49(1), 170–176. http://doi.org/10.1021/jf000791o

Manthey, J. A., Guthrie, N., & Grohmann, K. (2001). Biological properties of citrus flavonoids pertaining to cancer and inflammation. Current Medicinal Chemistry. 8(2), 135-153 http://doi.org/10.2174/0929867013373723

Markiewicz, L., Garey, J., Adlercreutz, H., & Gurpide, E. (1993). In vitro bioassays of non-steroidal phytoestrogens. The Journal of Steroid Biochemistry and Molecular Biology, 45(5), 399–405. http://doi.org/10.1016/0960-0760(93)90009-L

Mason, R. P. (2013). Chemical thermodynamics and metal(loid) complexation in natural waters. In Trace Metals in Aquatic Systems (1st ed., pp. 49–123). Chichester, West Sussex, U.K.: Blackwell Publishing (John Wiley & Sons, Ltd.).

Mateus, N., Oliveira, J., Haettich-Motta, M., & de Freitas, V. (2004). New Family of Bluish Pyranoanthocyanins. Journal of Biomedicine and Biotechnology, 2004(5), 299–305. http://doi.org/10.1155/S1110724304404033 183

McCann, D., Barrett, A., Cooper, A., Crumpler, D., Dalen, L., Grimshaw, K., … Stevenson, J. (2007). Food additives and hyperactive behaviour in 3-year-old and 8/9-year-old children in the community: a randomised, double-blinded, placebo- controlled trial. Lancet, 370(9598), 1560–1567.

Mechoulam, R., Brueggemeier, R. W., & Denlinger, D. L. (1984). Estrogens in insects. Experientia, 40(9), 942–944. http://doi.org/10.1007/BF01946450

Messina, M. J., Persky, V., Setchell, K. D., & Barnes, S. (1994). Soy intake and cancer risk: a review of the in vitro and in vivo data. Nutrition and Cancer, 21(2), 113–131. http://doi.org/10.1080/01635589409514310

Miksicek, R. J. (1994). Interaction of naturally occurring nonsteroidal estrogens with expressed recombinant human estrogen receptor. The Journal of Steroid Biochemistry and Molecular Biology, 49(2-3), 153–60. http://www.ncbi.nlm.nih.gov/pubmed/8031711

Minguez-Mosquera, M. I., & Hornero-Mendez, D. (1993). Separation and Quantification of the Carotenoid Pigments in Red Peppers (Capsicum annuum L.), Paprika, and Oleoresin by Reversed-Phase HPLC. J. Agric. Food Chem, 41, 1616–1620.

Miniello, V. L., Moro, G. E., Tarantino, M., Natile, M., Granieri, L., & Armenio, L. (2003). Soy-based formulas and phyto-oestrogens: a safety profile. Acta Paediatrica (Oslo, Norway : 1992) Supplement, 91(441), 93–100. http://doi.org/10.1080/08035320310018727

Miranda-Bermudez, E., Belai, N., Harp, B. P., Yakes, B. J., & Barrows, J. N. (2012). Qualitative determination of carbon black in food products. Food Additives and Contaminants Part a-Chemistry Analysis Control Exposure & Risk Assessment, 29(1), 38–42.

Mitchell, J. H., Gardner, P. T., McPhail, D. B., Morrice, P. C., Collins, A. R., & Duthie, G. G. (1998). Antioxidant efficacy of phytoestrogens in chemical and biological model systems. Archives of Biochemistry and Biophysics, 360(1), 142–8. http://doi.org/10.1006/abbi.1998.0951

Mónica Giusti, M., & Wrolstad, R. E. (2005). Characterization and Measurement of Anthocyanins by UV-visible Spectroscopy. Chapter F 1.2 in Handbook of Food Analytical Chemistry Vol. 2–2, pp. 19–31. http://doi.org/10.1002/0471709085.ch18

Morito, K., Hirose, T., Kinjo, J., Hirakawa, T., Okawa, M., Nohara, T., … Masamune, Y. (2001). Interaction of phytoestrogens with estrogen receptors alpha and beta. Biological & Pharmaceutical Bulletin, 24(4), 351–6. 184

http://www.ncbi.nlm.nih.gov/pubmed/11305594

Morton, L. W., Abu-Amsha Caccetta, R., Puddey, I. B., & Croft, K. D. (2000). Chemistry and biological effects of dietary phenolic compounds: relevance to cardiovascular disease. Clinical and Experimental Pharmacology & Physiology, 27(3), 152–159. http://doi.org/http://dx.doi.org/10.1046/j.1440-1681.2000.03214.x

Munro, I. C., Harwood, M., Hlywka, J. J., Stephen, A. M., Doull, J., Flamm, W. G., & Adlercreutz, H. (2003). Soy isoflavones: a safety review. Nutrition Reviews, 61(1), 1–33. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12638461

Murphy, P. A., Barua, K., & Hauck, C. C. (2002). Solvent extraction selection in the determination of isoflavones in soy foods. Journal of Chromatography. B, Analytical Technologies in the Biomedical and Life Sciences, 777(1-2), 129–38. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12270206

Murrill, W. B., Brown, N. M., Zhang, J. X., Manzolillo, P. A., Barnes, S., & Lamartiniere, C. A. (1996). Prepubertal genistein exposure suppresses mammary cancer and enhances gland differentiation in rats. Carcinogenesis, 17(7), 1451– 1457. http://doi.org/10.1093/carcin/17.7.1451

Naim, M., Gestetner, B., Zilkah, S., Birk, Y., & Bondi, A. (1974). Soybean isoflavones. Characterization, determination, and antifungal activity. Journal of Agricultural and Food Chemistry, 22(5), 806–810. http://doi.org/10.1021/jf60195a031

Nayak, B., Berrios, J. D. J., Powers, J. R., & Tang, J. (2011). Thermal degradation of anthocyanins from purple potato (Cv. Purple Majesty) and impact on antioxidant capacity. Journal of Agricultural and Food Chemistry, 59(20), 11040–11049. http://doi.org/10.1021/jf201923a

Newsome, A. G., Culver, C. a, & Breemen, R. B. Van. (2014). Nature’s Palette : The Search for Natural Blue Colorants. Journal of Agricultural and Food Chemistry, 62, 6498–6511.

Nickel, E. H. (1995). The Definition of a Mineral. Mineralogical Journal, 17(7), 346– 349.

Nicolaus, R. a., Piattelli, M., & Fattorusso, E. (1964). The structure of melanins and melanogenesis—IV. Tetrahedron, 20(5), 1163–1172.

Nurmi, T., Mazur, W., Heinonen, S., Kokkonen, J., & Adlercreutz, H. (2002). Isoflavone content of the soy based supplements. Journal of Pharmaceutical and Biomedical Analysis, 28(1), 1–11. http://doi.org/10.1016/S0731-7085(01)00612-4

185

O’Toole, D. K. (1999). Characteristics and use of okara, the soybean residue from soy milk production - A review. Journal of Agricultural and Food Chemistry. 47(2), 363-371 http://doi.org/10.1021/jf980754l

Obón, J. M., Castellar, M. R., Alacid, M., & Fernández-López, J. a. (2009). Production of a red-purple food colorant from Opuntia stricta fruits by spray drying and its application in food model systems. Journal of Food Engineering, 90(4), 471–479. http://doi.org/10.1016/j.jfoodeng.2008.07.013

Oliveira, J., Azevedo, J., Silva, A. M. S., Teixeira, N., Cruz, L., Mateus, N., & De Freitas, V. (2010). Pyranoanthocyanin dimers: A new family of turquoise blue anthocyanin-derived pigments found in port wine. Journal of Agricultural and Food Chemistry, 58(8), 5154–5159.

Oszmianski, J., & Lee, C. Y. (1990). Isolation and HPLC determination of phenolic compounds in red grapes. American Journal of Enology and Viticulture, 41(3), 204– 206.

Oyvind, A., & Markham, K. (2006). Flavonoids. Chemistry, Biochemistry and Applications. Edited by Øyvind M. Andersen and Kenneth R. Markham. Angewandte Chemie International Edition, 45(41), 6786–6787. http://doi.org/10.1002/anie.200685399

Pacheco-Palencia, L. a., & Talcott, S. T. (2010). Chemical stability of a??ai fruit (Euterpe oleracea Mart.) anthocyanins as influenced by naturally occurring and externally added polyphenolic cofactors in model systems. Food Chemistry, 118(1), 17–25. http://doi.org/10.1016/j.foodchem.2009.02.032

Pacheco-Palencia, L. A., & Talcott, S. T. (2010). Chemical stability of acai fruit (Euterpe oleracea Mart.) anthocyanins as influenced by naturally occurring and externally added polyphenolic cofactors in model systems. Food Chemistry, 118(1), 17–25. http://doi.org/10.1016/j.foodchem.2009.02.032

Pattanagul, P., Pinthong, R., & Phianmongkhol, A. (2007). Review of Angkak Production (Monascus purpureus). Chiang Mai J. Sci., 34(3), 319–328.

Pintea, A. M. (2007). Pigments as Food Colorants. Food Colorants, Chemical and Functional Properties, 551–616.

Potera, C. (2010). The Artificial Food Dye Blues. Environmental Health Perspectives, 118(10), 428.

Potter, S. M., Baum, J. A., Teng, H., Stillman, R. J., Shay, N. F., & Erdman, J. W. (1998). Soy protein and isoflavones: their effects on blood lipids and bone density in 186

postmenopausal women. The American Journal of Clinical Nutrition, 68(6 Suppl), 1375S–1379S. http://www.ncbi.nlm.nih.gov/pubmed/9848502

Potterat, O. (2010). Goji (Lycium barbarum and L. chinense): Phytochemistry, pharmacology and safety in the perspective of traditional uses and recent popularity. Planta Medica. http://doi.org/10.1055/s-0029-1186218

Pouvreau, J. B., Morançais, M., Massé, G., Rosa, P., Robert, J. M., Fleurence, J., & Pondaven, P. (2006). Purification of the blue-green pigment “marennine” from the marine tychopelagic diatom Haslea ostrearia (Gaillon/Bory) Simonsen. Journal of Applied Phycology, 18(6), 769–781.

Price, K. R., & Fenwick, G. R. Naturally occurring oestrogens in foods--a review. Food Additives and Contaminants, 2(2), 73–106. http://doi.org/10.1080/02652038509373531

Pyysalo, H., & Kuusi, T. (1973). The role of iron and tin in discoloration of berry and red beet juices. Zeitschrift Fur Lebensmittel-Untersuchung Und -Forschung, 153(4), 224–233. http://doi.org/10.1007/BF01142093

Rakic, V., Skrt, M., Miljkovic, M., Kostic, D., Sokolovic, D., & Poklar-Ulrih, N. (2015). Effects of pH on the stability of cyanidin and cyanidin 3-O-β-glucopyranoside in aqueous solution. Hemijska Industrija, 69(5), 511–522. http://doi.org/10.2298/HEMIND140711072R

Rentzsch, M., Schwarz, M., & Winterhalter, P. (2007). Pyranoanthocyanins - an overview on structures, occurrence, and pathways of formation. Trends in Food Science and Technology, 18(10), 526–534.

Rettori, D., & Durán, N. (1998). Production, extraction and purification of violacein: An antibiotic pigment produced by Chromobacterium violaceum. World Journal of Microbiology and Biotechnology, 14(5), 685–688.

Riedl, K. M., Lee, J. H., Renita, M., St Martin, S. K., Schwartz, S. J., & Vodovotz, Y. (2007). Isoflavone profiles, phenol content, and antioxidant activity of soybean seeds as influenced by cultivar and growing location in Ohio. Journal of the Science of Food and Agriculture, 87(7), 1197–1206. http://doi.org/10.1002/jsfa.2795

Riggi, E. (2010). Recent patents on the extraction of carotenoids. Recent Patents on Food, Nutrition & Agriculture, 2(1), 75–82. http://www.ncbi.nlm.nih.gov/pubmed/20653552

Rivera, S. M., & Canela-Garayoa, R. (2012). Analytical tools for the analysis of carotenoids in diverse materials. Journal of Chromatography A. 1224, 1-10 187

http://doi.org/10.1016/j.chroma.2011.12.025

Rodriguez-Amaya, D. B. (1999). Changes in carotenoids during processing and storage of foods. Archivos Latinoamericanos de Nutricion, 49(3 Suppl 1), 38S-47

Rodriguez-Amaya, D. B. (2016). Natural food pigments and colorants. Current Opinion in Food Science. (pp. 1-35) http://doi.org/10.1016/j.cofs.2015.08.004

Rodriguez-Saona, L. E., & Wrolstad, R. E. (2005). Extraction, Isolation, and Purification of Anthocyanins. Chapter F 1.11 in Current Protocols in Food Analytical Chemistry. http://doi.org/10.1002/0471142913.faf0101s00

Schreiber, H. D., Swink, A. M., & Godsey, T. D. (2010). The chemical mechanism for Al3+ complexing with delphinidin: A model for the bluing of hydrangea sepals. Journal of Inorganic Biochemistry, 104(7), 732–739. http://doi.org/10.1016/j.jinorgbio.2010.03.006

Schwartz, S. J., & von Elbe, J. H. (1983). Identification of betanin degradation products. Zeitschrift Für Lebensmittel-Untersuchung Und -Forschung, 176(6), 448–453. http://doi.org/10.1007/BF01042560

Sensient Colors LLC. (2016). Natural Colors, Closing the Gap between Nature and Science. St. Louis, Missouri.

Sepe, H. A., Parker, O. D., Nixon, A. R., & Kamuf, W. E. (2008). Global Color Quality of Beverages Utilizing Caramel Color, 226–240.

Setchell, K. D. (1998). Phytoestrogens: the biochemistry, physiology, and implications for human health of soy isoflavones. The American Journal of Clinical Nutrition, 68(6 Suppl), 1333S–1346S. http://www.ncbi.nlm.nih.gov/pubmed/9848496

Setchell, K. D., & Cassidy, A. (1999). Dietary isoflavones: biological effects and relevance to human health. The Journal of Nutrition, 129, 758S–767S.

Sharma, V., McKone, H. T., & Markow, P. G. (2010). A Global Perspective on the History, Use, and Identification of Synthetic Food Dyes. Journal of Chemical Education, 88(1), 24–28. http://doi.org/10.1021/ed100545v

Sheehan, D. M. (1998). Herbal medicines, phytoestrogens and toxicity: Risk:benefit considerations. Proceedings of the Society for Experimental Biology and Medicine, 217(3), 379–385. https://www.scopus.com/inward/record.uri?eid=2-s2.0- 0031912347&partnerID=40&md5=86b20a273ecb15d29c8bd976f4b86ed5

Shiroma-Kian, C., Tay, D., Manrique, I., Giusti, M. M., & Rodriguez-Saona, L. E. 188

(2008). Improving the screening process for the selection of potato breeding lines with enhanced polyphenolics content. Journal of Agricultural and Food Chemistry, 56(21), 9835–9842. http://doi.org/10.1021/jf801716b

Sigurdson, G. T., & Giusti, M. M. (2014). Bathochromic and hyperchromic effects of aluminum salt complexation by anthocyanins from edible sources for blue color development. Journal of Agricultural and Food Chemistry, 62(29), 6955–6965. http://doi.org/10.1021/jf405145r

Sigurdson, G. T., Robbins, R. J., Collins, T. M., & Giusti, M. M. (2016). Evaluating the role of metal ions in the bathochromic and hyperchromic responses of cyanidin derivatives in acidic and alkaline pH. Food Chemistry, 208, 26–34. http://doi.org/10.1016/j.foodchem.2016.03.109

Sigurdson, G. T., Robbins, R. J., Collins, T. M., & Giusti, M. M. (2017). Effects of hydroxycinnamic acids on blue color expression of cyanidin derivatives and their metal chelates. Food Chemistry, 234, 131–138. http://doi.org/10.1016/j.foodchem.2017.04.127

Sigurdson, G. T., Tang, P., & Giusti, M. M. (2017). Natural Colorants: Food Colorants from Natural Sources. Annual Review of Food Science and Technology, 8(1), 261– 280. http://doi.org/10.1146/annurev-food-030216-025923

Singh, S. (2006). Impact of color on marketing. Management Decision, 44(6), 783–789.

Sirtori, C. R. (2001). Risks and Benefits of Soy Phytoestrogens in Cardiovascular Diseases, Cancer, Climacteric Symptoms and Osteoporosis. Drug Safety, 24(9), 665–682. http://doi.org/10.2165/00002018-200124090-00003

Song, T., Barua, K., Buseman, G., & Murphy, P. A. (1998). Soy isoflavone analysis: quality control and a new internal standard. The American Journal of Clinical Nutrition, 68(6 Suppl), 1474S–1479S. http://www.ncbi.nlm.nih.gov/pubmed/9848519

Stintzing, F. C., & Carle, R. (2004). Functional properties of anthocyanins and betalains in plants, food, and in human nutrition. Trends in Food Science and Technology. 15(1), 19-38. http://doi.org/10.1016/j.tifs.2003.07.004

Stintzing, F. C., Schieber, A., & Carle, R. (2002). Identification of betalains from yellow beet (Beta vulgaris L.) and cactus pear [Opuntia ficus-indica (L.) Mill.] by high- performance liquid chromatography-electrospray ionization mass spectrometry. Journal of Agricultural and Food Chemistry, 50(8), 2302–2307. http://doi.org/10.1021/jf011305f

189

Stintzing, F. C., Stintzing, A. S., Carle, R., Frei, B., & Wrolstad, R. E. (2002). Color and antioxidant properties of cyanidin-based anthocyanin pigments. Journal of Agricultural and Food Chemistry, 50(21), 6172–6181. http://doi.org/10.1021/jf0204811

Strack, D., Vogt, T., & Schliemann, W. (2003). Recent advances in betalain research. Phytochemistry. 62(3), 247-269. http://doi.org/10.1016/S0031-9422(02)00564-2

Strom, B. L., Schinnar, R., Ziegler, E. E., Barnhart, K. T., Sammel, M. D., Macones, G. A, Hanson, S. A. (2001). Exposure to soy-based formula in infancy and endocrinological and reproductive outcomes in young adulthood. JAMA : The Journal of the American Medical Association, 286(7), 807–814. http://doi.org/10.1001/jama.286.7.807

Stushnoff, C., Holm, D., Thompson, M. D., Jiang, W., Thompson, H. J., Joyce, N. I., & Wilson, P. (2008). Antioxidant properties of cultivars and selections from the Colorado potato breeding program. In American Journal of Potato Research 85, 267–276. http://doi.org/10.1007/s12230-008-9032-4

Sugiyama, Y., Sakurai, Y., & Hirota, A. (2010). Isolation of 2,4,4’- trihydroxydeoxybenzoin and 3'-hydroxydaidzein from soybean miso. Bioscience, Biotechnology, and Biochemistry, 74(6), 1293–4. http://www.ncbi.nlm.nih.gov/pubmed/20530891

Swan, G. a. (1974). Structure, Chemistry, and biosynthesis of melanin. In Fortschritte der Chemie Organischer Naturstoffe/Progress in the Chemistry of Organic Natural Products (pp. 521-582). Springer, Vienna.

Tachibana, N., Kimura, Y., & Ohno, T. (2014). Examination of molecular mechanism for the enhanced thermal stability of anthocyanins by metal cations and polysaccharides. Food Chemistry, 143, 452–458. http://doi.org/10.1016/j.foodchem.2013.08.017

Takeda, K., Fujii, T., & Iida, M. (1984). Magnesium in the blue pigment complex commelinin. Phytochemistry, 23(4), 879–881.

Talcott, S. T., Peele, J. E., & Brenes, C. H. (2005). Red clover isoflavonoids as anthocyanin color enhancing agents in muscadine wine and juice. Food Research International, 38(10), 1205–1212. http://doi.org/10.1016/j.foodres.2005.05.004

Tanaka, Y., Sasaki, N., & Ohmiya, A. (2008). Biosynthesis of plant pigments: Anthocyanins, betalains and carotenoids. Plant Journal. http://doi.org/10.1111/j.1365-313X.2008.03447.x

190

Tanenbaum, D. M., Wang, Y., Williams, S. P., & Sigler, P. B. (1998). Crystallographic comparison of the estrogen and progesterone receptor’s ligand binding domains. Proceedings of the National Academy of Sciences of the United States of America, 95(11), 5998–6003. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=27574&tool=pmcentrez &rendertype=abstract

Tsukamoto, C., Shimada, S., Igita, K., Kudou, S., Kokubun, M., Okubo, K., & Kitamura, K. (1995). Factors Affecting Isoflavone Content in Soybean Seeds: Changes in Isoflavones, Saponins, and Composition of Fatty Acids at Different Temperatures during Seed Development. Journal of Agricultural and Food Chemistry, 43(5), 1184–1192. http://doi.org/10.1021/jf00053a012

Turner, N. J., Thomson, B. M., & Shaw, I. C. (2003). Bioactive isoflavones in functional foods: The importance of gut microflora on bioavailability. Nutrition Reviews. http://doi.org/10.1301/nr.2003.jun.204-213

Uesugi, T., Fukui, Y., & Yamori, Y. (2002). Beneficial effects of soybean isoflavone supplementation on bone metabolism and serum lipids in postmenopausal japanese women: a four-week study. Journal of the American College of Nutrition, 21(2), 97– 102. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11999549

Umphress, S. T., Murphy, S. P., Franke, A. A., Custer, L. J., & Blitz, C. L. (2005). Isoflavone content of foods with soy additives. Journal of Food Composition and Analysis, 18(6), 533–550. http://doi.org/10.1016/j.jfca.2004.04.008

Venil, C. K., Aruldass, C. A., Abd Halim, M. H., Khasim, A. R., Zakaria, Z. A., & Ahmad, W. A. (2015). Spray drying of violet pigment from Chromobacterium violaceum UTM 5 and its application in food model systems. International Biodeterioration and Biodegradation, 102, 324–329.

Villar, H. C. C. E., Saconato, H., Valente, O., & Atallah, A. N. (2007). Thyroid hormone replacement for subclinical hypothyroidism. The Cochrane Database of Systematic Reviews, (3), CD003419. http://doi.org/10.1002/14651858.CD003419.pub2

Vitale, D. C., Piazza, C., Melilli, B., Drago, F., & Salomone, S. (2013). Isoflavones: estrogenic activity, biological effect and bioavailability. European Journal of Drug Metabolism and Pharmacokinetics, 38(1), 15–25. http://doi.org/10.1007/s13318- 012-0112-y

Vupadhyayula, P. M., Gallagher, J. C., Templin, T., Logsdon, S. M., & Smith, L. M. (n.d.). Effects of soy protein isolate on bone mineral density and physical performance indices in postmenopausal women--a 2-year randomized, double-blind, placebo-controlled trial. Menopause (New York, N.Y.), 16(2), 320–8. 191

http://doi.org/10.1097/gme.0b013e3181844893

Wallace, T. C., & Giusti, M. M. (2008). Determination of color, pigment, and phenolic stability in yogurt systems colored with nonacylated anthocyanins from Berberis boliviana L. as compared to other natural/synthetic colorants. Journal of Food Science, 73(4). http://doi.org/10.1111/j.1750-3841.2008.00706.x

Walsh, C. T., Garneau-Tsodikova, S., & Howard-Jones, A. R. (2006). Biological formation of pyrroles: Nature’s logic and enzymatic machinery. Natural Product Reports, 23(4), 517–531.

Wang, H., & Murphy, P. A. (1994a). Isoflavone Composition of American and Japanese Soybeans in Iowa: Effects of Variety, Crop Year, and Location. Journal of Agricultural and Food Chemistry, 42(8), 1674–1677. http://doi.org/10.1021/jf00044a017

Wang, H., & Murphy, P. A. (1994b). Isoflavone Content in Commercial Soybean Foods. Journal of Agricultural and Food Chemistry, 42(8), 1666–1673. http://doi.org/10.1021/jf00044a016

Wei, H., Bowen, R., Cai, Q., Barnes, S., & Wang, Y. (1995). Antioxidant and antipromotional effects of the soybean isoflavone genistein. Proceedings of the Society for Experimental Biology and Medicine. Society for Experimental Biology and Medicine (New York, N.Y.), 208(1), 124–30. http://www.ncbi.nlm.nih.gov/pubmed/7892286

Weir, A., Westerhoff, P., Fabricius, L., Hristovski, K., & von Goetz, N. (2012). Titanium dioxide nanoparticles in food and personal care products. Environmental Science & Technology, 46(4), 2242–50.

Wild Flavors and Specialty Ingredients. (2016). Acid-Stable Blue. Erlanger, KY.

Wiseman, H., O’Reilly, J. D., Adlercreutz, H., Mallet, A. I., Bowey, E. A., Rowland, I. R., & Sanders, T. A. (2000). Isoflavone phytoestrogens consumed in soy decrease F(2)-isoprostane concentrations and increase resistance of low-density lipoprotein to oxidation in humans. The American Journal of Clinical Nutrition, 72(2), 395–400. http://www.ncbi.nlm.nih.gov/pubmed/10919933

Wrolstad, R. E., & Culver, C. A. (2012). Alternatives to Those Artificial FD&C Food Colorants. Annual Review of Food Science and Technology, 3, 59–77. http://doi.org/10.1146/annurev-food-022811-101118

Wrolstad, R. E., & Erlandson, J. A. (1973). Effect of metal ions on the color of strawberry puree. Journal of Food Science, 38(3), 460–463. 192

http://doi.org/10.1111/j.1365-2621.1973.tb01454.x

Wu, S., Ford, C., & Horn, G. (2013). U.S. Patent No. 8,557,319. Washington, DC: U.S. Patent and Trademark Office.

Wu, X., & Prior, R. L. (2005). Identification and characterization of anthocyanins by high-performance liquid chromatography-electrospray ionization-tandem mass spectrometry in common foods in the United States: vegetables, nuts, and grains. Journal of Agricultural and Food Chemistry, 53(8), 3101–3113. http://doi.org/10.1021/jf0478861

Wuttke, W., Jarry, H., & Seidlová-Wuttke, D. (2007). Isoflavones-Safe food additives or dangerous drugs? Ageing Research Reviews. 6(2), 150-188. http://doi.org/10.1016/j.arr.2007.05.001

Wyler, H. (1986). Neobetanin: A new natural plant constituent? Phytochemistry, 25(9), 2238. http://doi.org/10.1016/0031-9422(86)80100-5

Xu, X., Harris, K. S., Wang, H. J., Murphy, P. a, & Hendrich, S. (1995). Bioavailability of soybean isoflavones depends upon gut microflora in women. The Journal of Nutrition, 125(9), 2307–2315.

Yabuta, G., Koizumi, Y., Namiki, K., Hida, M., & Namiki, M. (2001). Structure of green pigment formed by the reaction of caffeic acid esters (or chlorogenic acid) with a primary amino compound. Bioscience Biotechnology and Biochemistry, 65(10), 2121–2130.

Yoshida, K., Kitahara, S., Ito, D., & Kondo, T. (2006). Ferric ions involved in the flower color development of the Himalayan blue poppy, Meconopsis grandis. Phytochemistry, 67(10), 992–998. http://doi.org/10.1016/j.phytochem.2006.03.013

Yoshida, K., Kondo, T., Kameda, K., & Goto, T. (1990). Structure of anthocyanins isolated from purple leaves of perilla ocimoides l. Var. Crispa benth and their isomerization by irradiation of light. Agricultural and Biological Chemistry, 54(7), 1745–1751. http://doi.org/10.1080/00021369.1990.10870188

Yoshida, K., Mori, M., & Kondo, T. (2009). Blue flower color development by anthocyanins: from chemical structure to cell physiology. Natural Product Reports, 26(7), 884–915.

Zava, D. T., & Duwe, G. (1997). Estrogenic and antiproliferative properties of genistein and other flavonoids in human breast cancer cells in vitro. Nutrition and Cancer, 27(1), 31–40. http://doi.org/10.1080/01635589709514498

193

Zhang, Y., Hendrich, S., & Murphy, P. A. (2003). Glucuronides are the main isoflavone metabolites in women. The Journal of Nutrition, 133(2), 399–404. http://www.ncbi.nlm.nih.gov/pubmed/12566474

Zhao, C. L., Chen, Z. J., Bai, X. S., Ding, C., Long, T. J., Wei, F. G., & Miao, K. R. (2014). Structure-activity relationships of anthocyanidin glycosylation. Molecular Diversity. 18(3), 687-700. http://doi.org/10.1007/s11030-014-9520-z

Zheng, J., Ding, C., Wang, L., Li, G., Shi, J., Li, H.,Suo, Y. (2011). Anthocyanins composition and antioxidant activity of wild Lycium ruthenicum Murr. from Qinghai-Tibet Plateau. Food Chemistry, 126(3), 859–865. http://doi.org/10.1016/j.foodchem.2010.11.052

194