I Purple Corn (Zea Mays L.) Cob Anthocyanins

Home , Aurantinidin, Blue corn, Capensinidin, Chicha morada, Copigmentation, Corncob

Purple Corn (Zea mays L.) Cob Anthocyanins: Extraction, Quantification, Spray Drying and Complexation with Proteins

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Fei Lao

Graduate Program in Food Science and Technology

The Ohio State University

2016

Dissertation Committee:

Dr. M. Monica Giusti, Advisor

Dr. Sheryl A. Barringer

Dr. C. Lynn Knipe

Dr. John H. Litchfield

Copyrighted by

Fei Lao

2016

Abstract

Interests of applying natural colorants such as anthocyanins as alternatives to synthetic dyes in food have been a market trend in recent years. This study evaluated purple corn cob (PCC) as economic natural source for high quality food-used anthocyanins colorant production.

PCC is an anthocyanin-rich plant source. Different from anthocyanins in other fruits and vegetables, production and application of PCC anthocyanins is not as simple and straightforward due to the fiber-rich hard texture of the cob and relatively complicated biomatrix of the corn. Some of the PCC pigments undergoing traditional processing have acidic water solubility issues, which limit their application in most aqueous-based foods.

The objective of this study is to optimize preparation of PCC anthocyanins, to produce high quality PCC pigments for more general food application. To achieve this goal, the critical conditions in PCC pigments extraction and spray drying were evaluated, and the key structure (anthocyanin-protein complexation) which was believed to be correlated to water solubility problems was investigated using infrared spectroscopy.

Aqueous ethanol with water and ethanol ratio around 1:1 with slightly acid addition

(0.01% v/v HCl) was able to efficiently recover PCC anthocyanins and phenolics. The yield was comparable to analytical labolatory used solvents under the same extraction conditions. Spray drying PCC pigments with mild inlet/outlet temperatures

(150°C/105°C) and appropriate amount (5%, m/v) of carrier could produce satisfactory ii quality PCC pigment powders with least color degradation and more than 85% of pigment yield. The only hazy reconstituted PCC pigment solution was obtained from spray drying alcoholic PCC extract, which suggested the acidic water insolubility issue was mainly coming from the extraction rather than the heated dehydration process.

Anthocyanin-tannin-protein complex was proposed in previous studies to be the major contributor in the insoluble PCC pigment rich particles. Confirmed by the mid infrared

(MIR) spectroscopy, the PCC anthocyanins and protein could form complexation in the aqueous matrix. Hydrogen bonding was the major driving force to stabilize the complexation. The anthocyanin:protein molar ratio in the complexation depended on the environmental pH, as hydrophobic interaction and ion chelation might also be involved into complexation when the matrix pH was acidic. A rapid MIR prediction model to quantify protein levels in anthocyanin-rich matrix was also developed based on protein unique Amide signals.

In addition, the present study also discussed the advantages and disadvantages of four commonly used anthocyanin quantification methods in PCC pigment analysis. All four spectrophotometric and HPLC approaches outcomes for PCC were linearly correlated

(R2≥0.90) to each other. The total anthocyanins methods produced highest values, followed by the pH differential method, HPLC method with intact pigments, and HPLC method with acid hydrolyzed pigments.

Overall, our study showed PCC could be used to produce high quality anthocyanin powders ready for food application, as well as provided important insightful understandings on anthocyanin-protein complexation. Information in this study may help

iii food companies interested in the transition from synthetic dye to natural colorant using

PCC, for cleaner and more consumer-friendly labels.

Acknowledgments

First and earnest, I would like to express my most sincere gratitude to my advisor, Dr. M.

Monica Giusti, for offering me the opportunity to be a part of her lab and this amazing program. Her expertise in anthocyanins, endless creativity and curiosity, patient advisement, and positive attitude inspired me to explore more and trained me to be a food scientist. This exciting but challenging research project would not have been accomplished without her continuous encouragement and support.

I would like to thank my committee members, Dr. Sheryl A. Barringer, Dr. C. Lynn

Knipe, and Dr. John H. Litchfield, for their invaluable suggestions and support. Tons of thanks to Dr. Luis E. Rodriguez-Saona for providing convenient access to experimental facilities and continually guidance to my infrared study.

I would also like to thank my fellow lab mates and department peers in the past five years, for their considerate assistance and kindness, both in the lab and in general.

Thanks to Chinese Scholarship Council for providing fellowship to my first four years graduate study in the Ohio State University.

Thanks to Alicorp S.A.A. (Lima, Peru), Agroindustrial S.A.C (Lima, Peru), Zanaceutica

E.I.R.L. (Lima, Peru), and Globenatural International S.A. (Chorrillos-Lima, Peru) for providing funding, purple corn materials, and pigments production information to this research.

Finally, I am eternally grateful to my parents, Wenbiao Lao and Yixin Liu, as well as my dearest friends, for always standing by my side, offering me helping hands and mental support whenever I needed them the most throughout my PhD program.

Vita

June 23, 1988 ...... Nanning, China

2007 to 2011 ...... B.E., Food Science and Engineering,

China Agricultural University

2011 to present ...... Ph.D, Food Science and Technology, The

Ohio State University

Publications

Lao, F., Giusti, M.M., 2016. Quantification of Purple Corn (Zea mays L.) Anthocyanins

Using Spectrophotometric and HPLC Approaches: Method Comparison and Correlation.

Food Anal. Methods 9, 1367–1380.

Fields of Study

Major Field: Food Science and Technology

vii

Table of Contents

Abstract ...... ii

Acknowledgments...... v

Vita ...... vii

List of Tables ...... xv

List of Figures ...... xvii

Chapter 1: Introduction ...... 1

Chapter 2: Literature Review ...... 5

2.1 Anthocyanin Chemistry...... 5

2.1.1 Structure...... 6

2.1.2 Stability ...... 12

2.2 Purple Corn Pigments and Proteins ...... 20

2.2.1 Purple corn ...... 20

2.2.2 Purple corn anthocyanins ...... 22

2.2.3 Purple corn phlobaphenes ...... 23

2.2.4 Purple corn proteins ...... 24

viii

2.2.5 Anthocyanin-Protein Complexation in Purple Corn ...... 27

2.3 Anthocyanin-Protein Complexation ...... 29

2.3.1 Anthocyanic vacuolar inclusion (AVI)...... 30

2.3.2 Hordeumin ...... 33

2.3.3 Insoluble deposits in red wine bottle ...... 34

2.3.4 anthocyanin-whey protein aggregate particles ...... 35

2.4 Infrared Technology for Proteins Analysis ...... 36

2.5 Industrial Processing of Food Anthocyanins ...... 39

2.5.1 Extraction...... 41

2.5.2 Spray drying ...... 46

Chapter 3: Health Benefits of Purple Corn Phenolics ...... 50

3.1 Abstract ...... 50

3.2 Introduction ...... 51

3.3 Functional compounds in purple corn ...... 53

3.3.1 Anthocyanins ...... 54

3.3.2 Other phenolics ...... 54

3.4 Health benefits of purple corn phenolics ...... 57

3.4.1 Antioxidant ...... 58

3.4.2 Anti-inflammatory ...... 61

3.4.3 Anti-mutagenic ...... 63

3.4.4 Anti-carcinogenic and anti-cancer ...... 64

3.4.5 Anti-angiogenesis ...... 69

3.4.6 Amelioration of lifestyle diseases: obesity, diabetes, hyperglycemia and their

associated diseases ...... 69

3.4.7 Blood pressure regulation ...... 74

3.4.8 Heart health ...... 75

3.5 Purple Corn Texicology ...... 75

3.6 Recommended Dosage ...... 76

3.7 Conclusion ...... 80

3.8 Acknowlegements ...... 81

Chapter 4: The Effect of Solvent and Acidity Selection on Purple Corn (Zea mays L.)

Cob Pigments and Phenolic Compounds Extraction ...... 82

4.1 Abstract ...... 82

4.2 Introduction ...... 83

4.3 Materials and Methods ...... 85

4.3.1 Materials and reagents ...... 85

4.3.2 Standardization of extraction procedure ...... 86

4.3.3 Effect of solvent selection ...... 88

4.3.4 Effect of acidity ...... 88

4.3.5 Monomeric anthocycanins ...... 89

4.3.6 Polymeric Color ...... 89

4.3.7 Total phenolics ...... 90

4.3.8 Anthocyanins HPLC profile ...... 90

4.3.9 Statistical analysis...... 91

4.4 Results and Discusion ...... 92

4.4.1 Standardization of extraction procedure ...... 92

4.4.2 Effect of solvent selection ...... 95

4.4.3 Effect of acidity ...... 99

4.5 Conclusion ...... 102

4.6 Acknowledgment ...... 102

Chapter 5: Quantification of Purple Corn (Zea mays L.) Anthocyanins Using

Spectrophotometric and HPLC Approaches: Method Comparison and Correlation ...... 103

5.1 Abstract ...... 103

5.2 Introduction ...... 104

5.3 Materials and Methods ...... 107

5.3.1 Materials and reagents ...... 107

5.3.2 The total anthocyanins method ...... 108

5.3.3 Acetone extraction ...... 109

5.3.4 The pH differential method and polymeric color ...... 110

5.3.5 HPLC quantification using intact anthocyanins ...... 111

5.3.6 HPLC quantification using acid hydrolyzed anthocyanins ...... 112

5.3.7 Statistical analysis...... 113

5.4 Results and Discusion ...... 113

5.4.1 Anthocyanin contents measured by four different methods ...... 1133

5.4.2 Correlation between different anthocyanins quantification methods ...... 131

5.5 Conclusion ...... 135

5.6 Acknowledgment ...... 136

Chapter 6: The Effect of Pigment Matrix, Temperature and Amount of Carrier on the

Yield and Final Color Properties of Spray Dried Purple Corn (Zea mays L.) Cob

Anthocyanin Powders ...... 137

6.1 Abstract ...... 137

6.2 Introduction ...... 138

6.3 Materials and Methods ...... 140

6.3.1 Materials and reagents ...... 140

6.3.2 Spray drying conditions ...... 141

6.3.3 Quality evaluation parameters ...... 143

xii

6.3.4 Statistical analysis...... 146

6.4 Results and Discusion ...... 146

6.4.1 Spray-dried pigments yield ...... 146

6.4.2 Polymeric color based pigment quality ...... 150

6.4.3 Color properties changes and solubility ...... 152

6.4.4 Anthocyanins HPLC profiles changes ...... 157

6.5 Conclusion ...... 158

6.6 Acknowledgment ...... 159

Chapter 7: Investigation of Purple Corn (Zea mays L.) Anthocyanins and Proteins

Complexation Using Mid-Infrared Technology ...... 160

7.1 Abstract ...... 160

7.2 Introduction ...... 161

7.3 Materials and Methods ...... 163

7.3.1 Materials and reagents ...... 163

7.3.2 Purified anthocyanins preparation ...... 164

7.3.3 Determination of target region for protein analysis ...... 166

7.3.4 Characterization of anthocyanin-protein complexation using infrared

spectroscopy ...... 169

xiii

7.3.5 Multivariate analysis for infrared quantification model development and

quantification of protein content in purple corn pigments crude extracts ...... 170

7.4 Results and Discusion ...... 172

7.4.1 Preparation of purified purple corn anthocyanins ...... 172

7.4.2 Determination of target region for protein analysis ...... 173

7.4.3 Characterization of anthocyanin-protein complexation using infrared

spectroscopy ...... 180

7.4.4 Multivariate analysis for infrared protein quantification model development

and quantification of protein content in purple corn pigments crude extracts ...... 188

7.5 Conclusion ...... 191

7.6 Acknowledgment ...... 192

Chapter 8: Overall Conculsion...... 193

References ...... 195

xiv

List of Tables

Table 2.1 Currently reported minor anthocyanidins and their substitution pattern ...... 7

Table 2.2 Distribution of protein fractions in corn ...... 25

Table 2.3 Conventional solvent extraction conditions for purple corn anthocyanins ...... 45

Table 3.1 The dosages of anthocyanins and phenolics applied in purple corn animals and humans health benefits studies...... 78

Table 4.1 The anthocyanin properties (monomeric and % polymeric anthocyanins, % acylated pigments) and total phenolics of purple corn cob recovered by different solvents acidified with 0.01% 6N HCl under room temperature ...... 98

Table 4.2 The anthocyanin properties (monomeric and % polymeric anthocyanins, % acylated pigments) and total phenolics of purple corn cob recovered by different acidic

70% aqueous acetone or water solution under room temperature ...... 101

Table 5.2 Peak area values of a purple corn anthocyanins ground powders using same integration parameters for their HPLC 520 nm and 500-530 nm max plot chromatograms and corresponding calculated anthocyanin contents ...... 123

Table 5.3 Example of impact of different HPLC integration parameters: peak area sum values of purple corn extract - same run - using different integration parameters ...... 127

Table 5.4 MS spectral data for typical purple corn anthocyanins...... 130

Table 6.1 Purple corn (Zea mays L.) cob pigments spray-drying yield and percent polymeric color in the pigmented solution before and after spray-drying of varying types of inject solution, inlet temperature and maltodextrin amount ...... 148

Table 6.2 Color and %haze changes of purple corn (Zea mays L.) cob pigments before and after spray-drying of varying types of matrix, inlet temperature and maltodextrin amount...... 153

Table 7.1 Water solubility, anthocyanin and protein contents of ten different commercial purple corn pigment powders...... 191

xvi

List of Figures

Figure 2.1 Six major anthocyanidin aglycone structure ...... 6

Figure 2.2 Typical sugar attachments in anthocyanins ...... 9

Figure 2.3 Typical acid attachments in anthocyanins ...... 11

Figure 2.4 Sketch of acylated anthocyanins intramolecular and intermolecular interaction for stabilization via spatial blocking the chromophore ...... 15

Figure 2.5 pH-dependent structure transformation and degradation of anthocyanins in aqueous solutions ...... 17

Figure 2.6 A scheme of biosynthetic pathway of purple corn phlobaphenes and anthocyanins ...... 22

Figure 2.7 Cross-section of corn kernel ...... 25

Figure 2.8 Solubility of purple corn cob anthocyanin-rich wastes at different pH levels and different concentrations of aqueous ethanol...... 28

Figure 2.9. Anthocyanic vacuolar inclusion in an oblique hand-section of the adaxial epidermal peel of the deep purple inner petal region of lisianthus ...... 31

Figure 2.10 Proposed autophagy mechanism for anthocyanic vacuolar inclusion formation...... 33

Figure 2.11 Flow chart of purple corn cob colorant production...... 40

Figure 2.12 A schematic diagram of lab scale mini spray dryer ...... 46 xvii

Figure 3.1 Chemical structures of typical phenolic acids in purple corn (Zea mays L.) .. 56

Figure 3.2 Chemical structures of typical flavonoids in purple corn (Zea mays L.) ...... 57

Figure 4.1 Extraction curve of percent purple corn cob anthocyanins and phenolics soaked in 0.01% 6N HCl acidified water...... 93

Figure 4.2 Percent extracted purple corn cob anthocyanins after each cake washing, extraction and cake washing were done by 0.01% 6N HCl water ...... 95

Figure 4.3 Quadratic predictive models of purple corn cob anthocyanins and phenolics extracted by various aqueous ethanol solutions ...... 99

Figure 5.1 Example of impact of HPLC integration parameters of max plot at 500-530 nm of purple corn (Zea mays L.) pigment extracts ...... 126

Figure 5.2 Typical anthocyanidins and anthocyanins HPLC profile at of purple corn (Zea mays L.) ...... 129

Figure 5.3 Simple linear regression plots between two of the four anthocyanins quantitative methods for fourteen purple corn (Zea mays L.) pigment-rich samples .... 133

Figure 6.1 Comparison of purple corn cob anthocyanins HPLC profile before and after spray drying ...... 152

Figure 7.1 Purple corn anthocyanins HPLC chromatograms at 520 nm and 280 nm before and after MCX-C18 cartridge purification ...... 172

Figure 7.2 Mid-infrared spectra of purified purple corn anthocyanins, BSA, and zein by mid-IR spectroscopy ...... 175

Figure 7.3 Comparison of protein quantification approaches using ultralight at 280 nm,

Bradford method, and mid infrared spectroscopy...... 178

xviii

Figure 7.4 Partial least squares regression (PLSR) plots and regression vector based on the infrared spectra of BSA and zein using FT-MIR spectroscopy ...... 179

Figure 7.5 Soft independent modeling of class analogy (SIMCA) classification plots, infrared spectra at Amide I region and discriminating power based on mid IR spectra for addition of different levels of purified purple corn anthocyanins into 15 μM BSA...... 181

Figure 7.6 Soft independent modeling of class analogy (SIMCA) classification plots, infrared spectra at 1000-1800 cm-1 region and discriminating power based on mid IR spectra for addition of different levels of BSA into 15 μM purified purple corn anthocyanins...... 185

Figure 7.7 BSA conformational isomers under various pH ...... 186

Figure 7.8 Partial least squares regression (PLSR) model plot for MIR protein quantification in anthocyanin-rich matrix and its regression vector based on the infrared spectra of aqueous BSA-anthocyanin mixture and alcoholic zein- anthocyanin mixture using FT-MIR spectroscopy ...... 189

xix

Chapter 1: Introduction

Anthocyanins are a class of water soluble pigments that could provide red, purple, and blue hues to various flowers, fruits, vegetables, and grains. Current market trend has shown preference of dyes from natural sources over synthetic colorants, because the artificial ones might link to some negative side effects on children (Lofstedt, 2013).

Thanks to the capability of providing a wide variety of colors, anthocyanins have been widely used as one of the alternatives to the synthetic dyes in food industry all around the world. In addition, anthocyanins have been known as potent antioxidants and strong anti- inflammatory agents. They are often advertised as health promoter and are believed to have potential to reduce the risk of cardiovascular disease, obesity, diabetes, cancer and chronic diseases (He and Giusti, 2010; Tsuda, 2008). However, application of anthocyanins in food commodities could be quite challenged as they are not as chemically stable as synthetic dyes. Even the pH, light, and temperature sensitivity of anthocyanins could be resolved, the relatively low tinctorial power and high production cost of anthocyanins compared to the artificial colorants is still a limiting step for broader use of this natural color source.

The dark pigmented variety of maize (Zea mays L.), known as purple corn, is a rich source of anthocyanins. The anthocyanin content in purple corn ranges from 6.8mg/g FW 1 to 82.3mg/g FW depending on the sections, which was much higher than most of the known anthocyanins-rich plants such as blueberries, strawberry, chokeberry, red cabbage, and eggplant (Cevallos-Casals and Cisneros-Zevallos, 2003; Li et al., 2008; Wu et al.,

2006). The anthocyanin levels in the husk and cob regions of purple corn was reported to be roughly 2 to10 times more than that in the kernel (Li et al., 2008). The inedible cob portion of purple corn seems to be an ideal economic starting material for anthocyanins production for food application.

However, due to its fiber-rich hard texture and complicated biomatrix, pigment production, quantification, and application of purple corn cob (PCC) is not as simple and straight forward as those for common fruits and vegetables. It is very hard to thoroughly extract all the reddish pigments from PCC even with organic solvents. The quality properties such as yield, color and water solubility of the final commercial PCC pigment powders might not perform as well as other plant sources. The presence of other reddish pigment in purple corn, phlobaphenes, would bring interference to anthocyanin quantification and decrease the purity of the final anthocyanin extracts. In addition, a large quantity of pigment-rich wastes were generated during traditional economic hot acidified water or aqueous ethanol extraction (Jing and Giusti, 2007). These acidic water insoluble wastes might be anthocyanin-tannin-protein complexes and they significantly decreased the overall pigment application efficiency of the plant material.

Aiming at a high product yield with satisfactory color and solubility quality for food application, the overall objective of this study was to provide optimal processing conditions and valuable anthocyanin-protein complexation understandings for better PCC

2 pigments production. Meanwhile, the potential health benefits of purple corn consumption was summarized and the commonly used anthocyanin quantification approaches were compared.

Chapter 3 symmetrically reviewed purple corn health benefit research performed in 21st century. The functional compounds in purple corn and specific health-promoting properties shown in cell, animal and human studies were summarized. The regarding dosage required to show beneficial effects in animals and humans wre also listed to provide valuable reference to those seeking additional health-promoting effects through purple corn consumption.

Chapter 4 and 6 developed optimal processing conditions to achieve high product yield and satisfactory quality of PCC anthocyanin product for food applications. Chapter 4 investigated the effects of extraction solvent and acidity for optimal PCC pigment recovery. Chapter 6 examined the impacts of input pigment matrix, inlet/out let temperature, and amount of carriers on final spray dried pigments product quality (yield, color, and solubility properties).

Chapter 5 compared and discussed advantages and disadvantages of four analytical approaches in purple corn anthocyanins quantification: the total anthocyanin methods, the pH differential methods, HPLC methods using intact pigments, and HPLC methods using acid hydrolyzed aglycones, The correlation among four different methods was also discussed.

Chapter 7 investigated anthocyanin-protein complexation using mid infrared technology.

The fingerprint IR spectra provided valuable structural information of anthocyanin-

3 protein complexation formation, as well as a promising approach to rapidly quantify protein contents in anthocyanin-rich matrix.

Chapter 2: Literature Review

2.1 ANTHOCYANINS CHEMISTRY

Anthocyanins are a group of water soluble bioactive plant phenolic compounds that provide red, orange, purple and blue colors to leaves, stems, roots and skins of many fruits, vegetables, grains and flowers (Bueno et al., 2012; Cooper-Driver, 2001; Jing et al., 2007; Welch et al., 2008). Interests on anthocyanins as food colorant have a long history, it has been of great interests in recent decades because they are odorless and nearly flavorless, although they might have a moderate astringent sensation (Bueno et al.,

2012; Clifford, 2000; He and Giusti, 2010; Jackman et al., 1987). The use of natural colors with additional health benefits (for example, anthocyanins and carotenoids) as alternative to controversial artificial/synthetic dyes has become increasely desirable in the food market. According to a natural color market report published by Mintel and

Leatherhead Food Research (2013), in 2011, for the first time ever, the value of natural colors market exceeded the artificial/synthetic dyes market. Global sales of natural colors reached $600 million with an annual growth rate of more than 7%. On the other hand, the growth of sales value of synthetic colors was more modest, by less than 4% from 2007 to

2011. A better understanding of the anthocyanin pigments performance in food matrix will provide more detailed guidance to their industrial application as food colorants. 5

2.1.1 Structure

Chemically, anthocyanins belong to flavonoids compounds, as they share the typical C6-

C3-C6 triple ring carbon skeleton with all the other compounds in the phenolics family.

Specifically, as shown in Figure 2.1, anthocyanins are glycosylated polyhydroxy and polymethoxy derivatives of 2-phenylbenzopyrylium cation. They consist two or three units, namely anthocyanidin aglycone, glycosylated sugar groups, and possibly aliphatic acids or aromatic acids acrylated to the sugar moieties, to make the molecule even more complicated. More than 20 anthocyanidins have been currently reported (Basso et al.,

2014; Castañeda-Ovando et al., 2009; Kong et al., 2003), the diversity of aglycones are mainly contributed by the hydroxyl and methoxyl substitution at 3, 5, 6, 7, 3’, 4’, and 5’ position of the carbon skeleton (Table 2.1).

Figure 2.1 Six major anthocyanidin aglycone structures. 6

Substitution position Anthocyanidins 3 5 6 7 8 3’ 4’ 5’ 6-hydroxyanthocyanidins Aurantinidin OH OH OH OH H H OH H 6-Hydroxypelargonidin OH OH OH OH H H OH H 6-Hydroxycyanidin OH OH OH OH H OH OH H 6-Hydroxydelphinidin OH OH OH OH H OH OH OH Ricciniodin A OH H OH OH H H OH H 5-methoxyanthocyanidins

5-Methylcyanidin OH OCH3 H OH H OH OH H Pulchellidin OH OCH3 H OH H OH OH OH Eupinidin OH OCH3 H OH H OCH3 OH OH Capensinidin OH OCH3 H OH H OCH3 OH OCH3 7-methoxyanthocyanidins

Hirsutidin OH OH H OCH3 H OCH3 OH OCH3 Rosinidin OH OH H OCH3 H OCH3 OH H 3-deoxyanthocyanidins Apigeninidin H OH H OH H H OH H (Gesneridin) Columnidin H OH H OH OH H OH OH Diosmetinidin H OH H OH H H OH OCH3 Luteolinidin H OH H OH H OH OH H Tricetinidin H OH H OH H OH OH OH 5-deoxyanthocyanidins Guibourtinidin OH H H OH H H OH H Fisetinidin OH H H OH H H OH OH Robinetinidin OH H H OH H OH OH OH 3,5-dideoxyanthocyanidins

Arrabidin H H OH OH H H OH OCH3 Carajurin H H OH OH H H OCH3 OCH3 3´-Hydroxyarrabidin H H OH OH H OH OH OCH3

Table 2.1 Chemical structures of other minor anthocyanidins found in nature and their substitution pattern. (Summerized from: Basso et al., 2014; Castañeda-Ovando et al., 2009; Jing, 2006; Kovinich et al., 2011; Malan et al., 1996; Rahim et al., 2008; Rein, 2005; Stich and Forkmann, 1988; Wolniak and Wawer, 2008)

Among those, six anthocyanidins are most commonly found. As shown in Figure 2.1, they are cyanidin, pelargonidin, peonidin, delphnidin, petunidin and malvidin (Clifford,

2000).

The distribution of the six common anthocyanidins in fruits and vegetables is: Cy 50%,

Dp 12%, Pg 12%, Pn 12%, Pt 7% and Mv 7% (Castañeda-Ovando et al., 2009; Kong et al., 2003). Being found in 80% of colored leaves, 69% in fruits and 50% in flowers, the non-methylated anthocyanidins (Cy, Dp and Pg) glycoside are mostly seen and Cy-3- glucoside is most abundant anthocyanin in nature (Kong et al., 2003).

2.1.1.1 Sugar attachment in anthocyanins

A single anthocyanidin unit is rarely found in nature because of its poor stability when exposed to light, oxygen and non-acidic condition, whereas glycosylated forms predominate (Wallace, 2011). Sugar moieties can be attached at any of the hydroxyl group at 3, 5, 7, 3’, 5’ or even 4’ of the aglycones (Gould et al., 2009). The more widespread glycoside derivatives in nature are 3-monosides, 3-biosides, 3,5- and 3,7- diglucosides, in which the 3-glucoside derivatives is 2.5 more frequent than the 3,5- diglucosides (Bueno et al., 2012; Kong et al., 2003). The most common encountered sugar attachments to anthocyanidin are glucose, galactose, rhamnose, xylose, arabinose, sophorose, sambubiose, rutinose (Clifford, 2000) (Figure 2.2).

Figure 2.2 Typical sugar attachments in anthocyanins.

2.1.1.2 Acid attachment in anthocyanins

Besides glycosylation, some anthocyanins have additional aromatic and/or aliphatic acid conjugations to the hydroxyl group in sugar attachment for further stability (Fossen et al.,

1998; Wallace, 2011). The structural diversity of anthocyanins is also significantly increased by acylation. Up to 2009, more than 65% of the reported anthocyanins whose structures are adequately characterized are acylated (Gould et al., 2009). Acyl substituents usually occur at the C-6 sugar (Bakowska-Barczak, 2005; Giusti and

Wrolstad, 2003; Otsuki et al., 2002). Attachments to 2-hydroxy (Bakowska-Barczak,

2005; Reiersen et al., 2003; Strack et al., 1992), 3-hydroxy (Andersen and Fossen, 1995;

Bakowska-Barczak, 2005) and 4-hydroxy (Bakowska-Barczak, 2005; Fossen et al., 2003) have also been reported. Acylation may occur in the same molecule and the number of acylating residues may go up to three (Clifford, 2000). As shown in Figure 2.3, common acylating agents include aromatic acids such as caffeic, p-coumaric, ferulic, sinapic and gallic acid, as well as a wide range of aliphatic acids such as acetic, malic, malonic, oxalic, and succinic acid (Bakowska-Barczak, 2005; Clifford, 2000; Reiersen et al., 2003;

Strack, 1992). Being identified in 25% of anthocyanins, malonic acid is the most frequently seen acyl moiety among these pigments (Bueno et al., 2012).

To summarize, the structure of different source of anthocyanins may be differ on (1) the type of aglycone, (2) the type of sugars or other substituents present, (3) the sequence of different sugars, (4) the way interglycosidic linkage, and (5) the attachment point of the substituents to aglycones (Liu et al., 2008). Considering that, the number of anthocyanins

10 would be at least 15-20 times to the number of anthocyanidins, currently, more than 600 kinds of anthocyanins have been reported (Bueno et al., 2012).

Figure 2.3 Typical acid attachments in anthocyanins.

2.1.2 Stability

One of the major challenges of applying anthocyanins as a food colorant is its relatively poor stability in a food matrix. The pigments are instable and very susceptible to degradation , particularly once removed from their native environment (Clifford, 2000;

Giusti and Wrolstad, 2003; Woodward et al., 2009). The chemical structure such as aglycone type, acylation, as well as environmental factors such as pH, temperature, light, presence of oxygen, phenolic compounds, proteins, ascorbic acid, sugars, metal ions and enzymes may have strong impacts on the stability of anthocyanins (Castañeda-Ovando et al., 2009; Cavalcanti et al., 2011; Rein, 2005).

2.1.2.1 Aglycone type

The major structure of anthocyanidins is a high resonance system constructed by conjugated A-C rings. The B ring may also contributes to the overall resonance.

Although the triple ring structure is the main reason for the pigment light reactivity, the various groups (namely hydroxyl, or methoxyl) attached to the ring structure, especially to the B ring, may consequently lead to different stability among anthocyanidins

(Delgado-vargas and Paredes-Lopez, 2002). It is widely believed and shown in some in vitro tests that the increased B-ring hydroxylation is associated with deceased pigments stability, since methylation may help blocking the highly reactive hydroxyl sites (Jing,

2006; Woodward et al., 2009). For the mostly commonly occurring six types of anthocyanidins, the three hydroxyl substituted delphinidin is the most vulnerable to chemical reactions, followed by cyanidin with 2 hydroxyl groups in the B ring.

Pelargonidin, and other three with methoxyl substitution anthocyanidins are more stable

12 than delphinidin and cyanidin. Besides chemical reactivity, the hydroxyl and methoxyl pattern may also affect the color display of the pigments. Usually more hydroxyl substitution on B ring tends to give more blue hue while the methoxyl groups attribute more to reddish color (Francis and Markakis, 1989; Rein, 2005; Tanaka and Ohmiya,

2008).

2.1.2.2 Acylation

Acylation with aliphatic or aromatic acids to sugar substitution may provide additional stability to anthocyanin via intermolecular or intramolecular interactions (Castañeda-

Ovando et al., 2009; Clifford, 2000; Giusti and Wrolstad, 2003; Jing, 2006; Wallace,

2011). It has been shown in many studies that the acylated anthocyanins were more stable than non-acylated forms in most unfavorable conditions such as heat, light and SO2

(Cevallos-Casals and Cisneros-Zevallos, 2004; Delgado-Vargas et al., 2000; Fossen et al., 1998; Giusti and Wrolstad, 2003; Matsufuji et al., 2007; Sadilova et al., 2007). The stability of cyanidin 3-O-6’’-O-malonylglucoside coloration in aqueous solution at pH 5 and 7 was higher than those of cyanidin 3-O-glucoside and cyanidin aglycone (Suzuki et al., 2002). Cyanidin-3-O-2’’-O-glucuronosylglucoside also showed improved color stability in response to light compared to both cyanidin 3-O-glucoside and cyanidin 3-O-

2’’-O-diglucoside (Osmani et al., 2009). Usually, the color of glycosylated anthocyanidins is unstable in aqueous solutions at mildly acidic pH values unless aromatic acyl groups are added (Gould et al., 2009). The aqueous color extracts of red sweet potato and red carrot, which are rich in acylated anthocyanins, showed higher color stability under long time storage in different pH (1-12), heat treatment (98°C, 2h) and

13 light exposure (florescence, 2days) compared to the non-acylated anthocyanins dominant purple corn and red grape (Cevallos-Casals and Cisneros-Zevallos, 2004). In pH 3, 10°Bx soft drink model system, the ferulic acid acylated black carrot anthocyanins showed lower degradation rate than the non-acylated blackberry and acai anthocyanins, in 20 to

50°C storage (Zozio et al., 2011). As for the color aspect, aromatic acylation usually causes a blue shift while the aliphatic acylation doesn’t change the color very much.

However, both kinds of acylation have stabilizing effect on anthocyanins (Bakowska-

Barczak, 2005; Giusti and Wrolstad, 2003; Giusti et al., 1999).

The mechanisms of intermolecular and intramolecular interactions protecting acylated anthocyanins from decoloration in neutral or weakly acidic aqueous solutions are shown in Figure 2.4. This decoloration arises from hydration at the C-2 position of the anthocyanidin flavylium nucleus (Gould et al., 2009). It has been suggested the aromatic acyl groups face-to-face stack on the triple ring anthocyanin backbone through π-π interaction, thus to protect the colored flavylium cation from the nucleophilic attack of the water molecule (Cooper-Driver, 2001; Dangles et al., 1993; Figueiredo, Elhabiri et al., 1996; Malien-Aubert et al., 2001; Quina et al., 2009). Acylation by the nonaromatic acyl groups such as malonic acid may also facilitate pigment stability by H-bonding between the carboxylate group and the core aglycone (Osmani et al., 2009). Glycosidic residues were proposed to act as spacers in this folding model, assuring the acylated groups were in correct position(Dangles et al., 1993; Figueiredo et al., 1996).

Figure 2.4 Sketch of acylated anthocyanins intramolecular and intermolecular interaction for stabilization via spatial blocking the chromophore (Giusti and Wrolstad, 2003; Yoshida et al., 2009, 1992).

2.1.2.3 pH

Anthocyanins may reversibly change their structure and color as their environmental pH changes, such that some have used these plant pigments as pH indicator. Generally, anthocyanins present colorless in slightly acidic condition, they turn to red when it becomes more acidic and turn to blue when alkaline is added.

Different structures of anthocyanins co-exist in aqueous solution in equilibrium, in this way the pigments contribute various color under different pH levels. As shown in Figure

2.5, there are five species that have been reported in anthocyanin equilibrium: flavylium cation, carbinol base, chalcone, quinonoidal base and anionic quinonoidal base (Mateus

15 and de Freitas, 2009). When the pH is very acidic (pH1), anthocyanins are primarily in the form of flavylium cation (Figure 2.5) and present vivid red color. As the pH increases to around 4 to 5, the flavylium cation can be rapidly hydrolyzed by nucleophilic attack of water at the 2-position, resulting in the colorless carbinol pseudobase (hemiketal form). This hemiketal form can be equilibrate to the open-ring yellowish chalcone form

(Figure 2.5). At this point, the conjugated C-ring is destroyed. The color-losing degradation ,ay start here from chalcone to instable intermediaries aldehyde and phenolic acid (Castañeda-Ovando et al., 2009; Fang, 2014). When the pH increases to a neutral condition, the violet quinonoidal base is predominant (Giusti and Wrolstad, 2001). The blue anionic quinonoidal base develops as the pH continually increases to 8 or above

(Jing, 2006; Mateus and de Freitas, 2009). Generally, anthocyanins in the form of flavylium cation are much more chemically stable the other four forms. The half-life of pH1 purified strawberry anthocyanins heated at 95°C was determined to be 3.20 hours

(Sadilova et al., 2006) while the half-life of pH3.5 purified strawberry anthocyanins at same condition was only 1.95 hours (Sadilova et al., 2007). A similar observation was also found in the cases of non-acylated pigment rich elderberry and acylated anthocyanins rich black carrot (Sadilova et al., 2006; Sadilova et al., 2007).

Figure 2.5 pH-dependent structure transformation and degradation of anthocyanins in aqueous solutions (Modified and summerized from: Castañeda-Ovando et al., 2009; Fang, 2014; Mateus and de Freitas, 2009; Welch et al., 2008).

2.1.2.4 Temperature

Like most other chemical reactions, higher temperatures speed up the degradation of anthocyanins. As shown in Figure 2.5, it is widely believed that the degradation of anthocyanins start with the open-ring chalcone structure, where the opening of C-ring splits into two parts, the A-ring thus transforms into aldehyde and the B-ring becomes a 17 phenolic acid (Clifford, 2000; Fang, 2014; Sadilova et al., 2006; Sadilova et al., 2007;

Welch et al., 2008). For non-acylated anthocyanins, the degradation initiates with the opening of the pyrylium ring for the chalcone glycosides, while for the acylated anthocyanins, the acyl-glucoside moieties are split off from the anthocyanidin backbone before further degradation (Sadilova et al., 2006; Sadilova et al., 2007). The formation of chalcone is an endothermic reaction. Increasing temperature tends to encourage other anthocyanins forms transforming into the chalcone form (Clifford, 2000). Additionally, the chalcone degradation rapidly accelerates at elevated temperatures (Clifford, 2000;

Maccarone et al., 1985; Rein, 2005; Romero and Bakker, 2000; Sadilova et al., 2006;

Sadilova et al., 2007).

Generally most of the reported thermal degradation of anthocyanins follow first order kinetics (Ochoa et al., 2001; Romero and Bakker, 2000; Sadilova et al., 2006; Sadilova et al., 2007). The half-life of strawberry preserve color stored at room temperature (20°C) was reported to be around 1300 hours, while the half-life of the color for an identical product stored at 38°C significantly dropped to 240 hours (Meschter, 1953). The half-life might extend to 6000 to 8000 hours (250 to 320 days) if stored at 4°C (Meschter, 1953).

A similar study also showed the degradation reaction rate for sweet cherry, sour cheery and raspberry preserves color were much lower at 4°C (1.28, 1.10, 2.00×103 day-1 respectively) and 20°C (3.88, 2.49, 3.38×103 day-1 respectively) in comparison to that at

40°C (6.95, 5.37, 7.10×103 day-1 respectively) (Ochoa et al., 2001). In addition, cranberry lost 62% of its anthocyanins when stored at 20°C while the loss was only 20% when stored at 7°C for the same period of time (Giusti and Jing, 2007).

However, the thermal degradation of anthocyanins could be hindered by very acidic pH and acylation. The half-life of pH1 purified strawberry anthocyanins heated at 95°C was determined to be 3.20 hours (Sadilova et al., 2006) while the half-life of pH3.5 purified strawberry anthocyanins at same condition was only 1.95 hours (Sadilova et al., 2007).

Similar observation was also found in the cases of elderberry and black carrot (Sadilova et al., 2006; Sadilova et al., 2007). The acylated pigments dominated black carrot had longer half-life under the same experiment condition compared to the non-acyalted pigments dominated strawberry, the half-life of black carrot pigments was 4.1 hours for pH1 (Sadilova et al., 2006) and 2.81 hours for pH3.5 (Sadilova et al., 2007). In a thermal stability study of anthocyanins from different plant sources, the acylated anthocyanins rich red cabbage were the most stable ones, following by black currant, grape skin, and the least stable elderberry (Dyrby et al., 2001). It was suggested that the acylated group could have somewhat protective effect against the anthocyanins thermal degradation (Dyrby et al.,

2001).

2.1.2.5 Light

Although light is very critical for the biosynthesis of anthocyanins, it can also be a significant factor in accelerating anthocyanins degradation. For identical amounts of cranberry anthocyanins stored under the same temperature, the half-lives of those stored in dark were much longer than those stored under the light (Attoe and von Elbe, 1981).

At 40°C, the half-life of Cy-3-Gal stored in the dark was 309 hours while it was only 56 hours when stored under the light (Attoe and von Elbe, 1981). Photon of the light could

19 be absorbed and affect the conjugated structure such as aromatic rings and double bonds in anthocyanins, and thus cause the color loss of the pigments (Koutchma, 2009).

Similar to thermal degradation of anthocyanins, the photodegradation kinetics of anthocyanins also fits the first-order reaction model very well (Attoe and von Elbe, 1981;

Chisté et al., 2010; Ochoa et al., 2001). Different illumination sources may cause different degree of damage to anthocyanins. Florescence light exposure resulted in the longest half-life for anthocyanin extracts from mangosteen peel, which was 597 hours, followed by 306 hours for incandescent light, 177 hours for ultraviolet light and 100 hours for infrared light (Chisté et al., 2010).

Luckily, for the anthocyanins, the damage caused by light is not as severe as that caused by high temperatures, but light does speed up the thermal degradation of anthocyanins.

(Attoe and von Elbe, 1981; Cevallos-Casals and Cisneros-Zevallos, 2004; Pala and

Toklucu, 2011). A study of non-thermal processing of pomegranate juice showed that the monomeric anthocyanins content, polymeric color and color density of the juice exposed to UV-C light treatment were slightly lower but showed no statistical significant difference to the untreated fresh juice. However, the juice treated at 90°C, 2 min had much lower values on these quality measurements (Pala and Toklucu, 2011).

2.2 PURPLE CORN PIGMENTS AND PROTEINS

2.2.1 Purple Corn

20 deepest purple shades among the plant kingdom. Due to its richness in purple pigments, purple corn pigments have long been used to color foods and beverages. In South

America, the purple corn extracts are widely applied in coloring two of the most popular homemade dessert and beverages: chicha morada and mazamorra morada (FAO, 2013).

Moreover, throughout the years, other countries have also shown interest in using these rich sources of color to obtain food colorants. According to United Nations BioTrade

Facilitation Program (2012), the average yearly growth of Peruvian purple corn exports value reached 467% during 1998 to 2002, and the price of purple corn had almost doubled from $0.75 to $1.36/kg in this five year period. In 2002, Peru exported total value of $24,220,360 natural colorants, of which $98,000 was contributed by purple corn anthocyanins products (UNEP, 2012). With purple corn color being recognized by the

European Union with the code E-163 and the same code for the Japanese legislation, the import of purple corns and their color products in Germany, France, Italy, Japan and other countries are growing. The exports of Peruvian purple corn business reached

$187,745,641 in 2010 (UNEP, 2012). Its use as food ingredient and as food colorants has been increasing around the world in the recent decades.

The main class of reddish pigments present in purple corn is the water soluble anthocyanins. Another class of pigments also reported in purple corn is the water insoluble but alcohol soluble pholobaphenes (Grotewold et al., 1994; Lee and Harper,

2002). The two pigments share the same biosynthetic route to the form of flavanones in purple corn plants (Figure 2.6) and undergo very different reactions afterwards.

Figure 2.6 A scheme of biosynthetic pathway of purple corn phlobaphenes and anthocyanins (modified from Grotewold, 2005).

2.2.2 Purple Corn Anthocyanins

Anthocyanins are one of the major sources of color which provide the purple reddish color to purple corn. The anthocyanin content in purple corn ranges from 6.8mg/g FW to

82.3mg/g FW depending on the sections (Cevallos-Casals and Cisneros-Zevallos 2003;

Wu et al. 2006; Li et al. 2008), which was much higher than most of the known anthocyanins-rich plants such as 1.3 to 3.8 mg/g found in blueberries (Cevallos-Casals and Cisneros-Zevallos, 2003; Wu et al., 2006), 0.21±0.03 mg/g in strawberry (Wang and

Lin, 2000; Wu et al., 2006), 3.22±0.41 mg/g in red cabbage (Ahmadiani et al., 2014; Wu

22 et al., 2006), 8.57 mg/g in eggplant (Wu et al., 2006) and 14.80 mg/g in chokeberry

(Kulling and Rawel, 2008; Wu et al., 2006). Previous studies reported that the pigment in purple corn was found in especially high levels in the inedible husk and cob regions.

Anthocyanin levels of cobs ranged from 0.49% to 4.60% of the dry or fresh weight, roughly 2 to10 times more than found in the kernel (Li et al., 2008).

The anthocyanins profile of purple corn has been well studied, six major as well as 17 other minor anthocyanins were identified (Aoki et al., 2002; De Pascual-Teresa et al.,

2002; González-Manzano et al., 2008; González-Paramás et al., 2006; Jing and Giusti,

2005; Jing et al., 2007; Li et al., 2008; Montilla et al., 2011; Pedreschi and Cisneros-

Zevallos, 2007; Zhao et al., 2008; Žilić et al., 2012). The six major purple corn anthocyanins are cyanidin-3-glucoside (C3G), pelargonidin-3-glucoside, peonidin-3- glucoside, and their malonic acid derivatives on the 6’’ position of the glucose molecule.

The minor anthocyanins are diglucoside of the three major anthocyanidins (Žilić et al.,

2012), or dimalonyl-derivatives (Aoki et al., 2002; Jing et al., 2007; Montilla et al.,

2011), or rutinose-derivatives of the same anthocyanidins (Žilić et al., 2012), or even some less common pigments such as the major anthocyanidins linked to succinic acid and catechin (González-Manzano et al., 2008; González-Paramás et al., 2006; Li et al., 2008;

Montilla et al., 2011). Delphinidin-3-glucoside was also reported to be present in purple corn, but it was rarely detected (Žilić et al., 2012).

2.2.3 Purple Corn Phlobaphenes

.Besides anthocyanins, other classes of extractable red pigments, phlobaphenes, have also been reported to be found in purple corn (Grotewold et al., 1994; Selinger and Chandler, 23

1999; Winkel-Shirley, 2001).These pigments are alcohol soluble but water insoluble brick-reddish compounds usually be found in the outer layer of a plant, such as bark, pericarp, cob glume and seed coat (Awika et al., 2004; Matus-Cádiz et al., 2008). The building units of phlobaphenes are 3-deoxyanthocyanins. These pigments are mainly accumulated in the cell wall (Grotewold et al., 1998; Koes et al., 2005). Compared to anthocyanins, the 3-deoxyanthocyanins are not as pH sensitive as normal anthocyanins.

The color change of 3-deoxyanthocyanins due to pH change is not as obvious as that performed by anthocyanins (Awika et al., 2004). To our knowledge, no health-promoting properties for animals/humans have been reported for phlobaphenes. In the case of purple corn, the phlobaphenes are predominantly found in pericarp and cobs (Grotewold et al.,

1994; Lee and Harper, 2002). Partial phlobaphenes could be extracted together with anthocyanins if using organic solvents.

2.2.4 Purple Corn Proteins

Proteins range from 6-12% on a dry basis in different corn varieties. Endosperm tissue

(Figure 2.7) contains about 75% of all the proteins in corn, while the remaining comes from germ (~20%) , bran (~2%), and tip (~5%) (Anderson and Lamsal, 2011; Shukla and

Cheryan, 2001). In the corn cob, where anthocyanins are most concentrated, the reported crude protein content is around 3% to 3.5% of the total corn cob weight (Ansah et al.,

2012; Ososanya et al., 2013). As shown in Table 2.2, there are four major kinds of proteins distributed in different areas of corn, namely zein, glutelin, albumin and globulin

(Rodriguez-Nogales et al., 2006). Albumins and globulins are centralized primarily in germ, while the other two are mostly found in endosperm (Anderson and Lamsal, 2011).

Figure 2.7 Cross-section of corn kernel (Shukla and Cheryan, 2001).

% dry basis proteins in Proteins Solubility Location whole kernel

Albumin water germ 2.3-12.4

Globulin saline germ 2.3-9.0

between endosperm Glutelin alkali 31.2-43.6 and germ

Zein alcohol endosperm 33.9-57.5

Table 2.2 Distribution of protein fractions in corn (summerized from Anderson and Lamsal, 2011; Shukla and Cheryan, 2001).

Most corn protein studies were focused on zeins as they are most abundant. Zein is a class of prolamin type proteins that exists only in the endosperm of corn. Zein found elsewhere in corn is either due to contamination or another kind of prolamin that is not zein (Anderson and Lamsal, 2011). Characterized by high glutamine and proline content and generally soluble only in strong alcohol solutions, prolamins are also reported to be found in other cereals such as wheat, barley, rye, and sorghum (Anderson and Lamsal,

2011). Glutelin received less attention than zeins because of its rather heterogeneous constitution (Ludevid et al., 1984). The classification of these four proteins is mainly based on their solubility in different solvents (Table 2.2).

Located in “zein-bodies”, zein is distributed uniformly through the cytoplasm of corn endosperm cells (Duvick, 1961) with molecular weight ranging from 14-31 KD (Wilson,

1991). Among four kinds of zeins, α-zein is the most abundant, and commercial zeins nowadays are composed primarily α-zein (Anderson and Lamsal, 2011). It is soluble in 50-

90% ethanol but not in anhydrous alcohol except methanol. Zein is also soluble in ketones, amide solvents, esters, glycols as well as high concentration of saline and alkaline aqueous solvent (pH>12). Basically, zein is rich in glutamic acid (21-26%) and other nonpolar amino acids (Leu, Pro, Ala), while the absence of basic and aromatic amino acids is quite notable (Pomes, 1971). Being poor at both nutrition value and water solubility, application of zeins in human food has long been limited. It was reported a well-controlled production of zein microparticles might serve as substitute for dietary fat (Stark and Gross, 1992). Use of zeins in food is mainly for coatings. Zein filming was reported to help preserve the

26 firmness and color for storage of broccoli (Rakotonirainy et al., 2001), as well as help maintain the integrity and oil uptake of turkey during frying (Ilter et al., 2008).

Danzer et al. (1975) suggested zein has globular structure like insulin and other conventional globular proteins in non-aqueous solution based on the helical content (33%-

60%). The circular dichroic spectrum suggested zein has a structure with nine adjacent, topologically antiparallel helices clustered within a distorted cylinder (Argos et al., 1982).

In this proposed structure, polar residues in the helical surfaces favored intra- and intermolecular hydrogen bonding to stabilize the structure. The glutamine rich turns between the helices and at the cylindrical caps further stabilize the structure by facilitating side chain interactions, resulting in stacking of the molecular planes. The distorted cylinder model was later modified based on small-angle X-ray scattering measurement, proposing that reduced alpha zein exist as asymmetric particles (Matsushima et al., 1997).

2.2.5 Anthocyanin-Protein Complexation in Purple Corn

High levels of anthocyanin- rich wastes (ARW) was generated during traditional purple corn pigments industrial acidified water extraction. These ARW reduce the yield of pigments extraction and increase the cost of production because their application in food system is very limited due to their poor solubility in an acidified aqueous matrix. As shown in Figure 2.8A, a previous study constructed the pH-%haze curve for 0.91 mg

ARW/mL (Jing and Giusti, 2005). The characteristics of ARW pH-dependent solubility was similar to the typical performance of tannin-protein complex under the same condition, thus the authors proposed the ARW could be anthocyanin-protein-tannin complexes (Jing and Giusti, 2005). The lowest solubility of the complexes was usually

27 not far away from the isoelectric point of the protein, the pH when proteins tend to precipitate due to limited solubility. Although the ARW presented poor solubility in an acidic aqueous environment, its solubility in 50-62.5% (v/v) aqueous ethanol was quite good (Figure 2.8B).

Figure 2.8 Solubility of purple corn cob anthocyanin-rich wastes (A) at different pH levels, ARW=0.91 mg/mL; (B) at different concentrations of aqueous ethanol, ARW=0.50mg/mL. (Jing and Giusti, 2005).

Similar to a typical protein-tannin complex, the interaction within the anthocyanin- protein-tannin complexes might involve hydrophobic effect and hydrogen bonding

(Charlton et al., 1996). A high ratio of proline to other amino acids in protein plays an important role in the hydrophobic effects in the protein-tannin model (Charlton et al.,

1996; Siebert et al., 1996). The hydroxyl groups in tannins and anthocyanins could be excellent hydrogen donors for the carbonyl groups in proteins in hydrogen bond formation (Jing and Giusti, 2005). However, a more detailed structural analysis is still needed to elucidate the chemical interactions involved in ARW complexes.

Due to the poor acidic water solubility of anthocyanin-protein-tannin complexes, Jing and Giusti (2005) applied these ARW to color the milk with neutral pH (around 6.8).

Their data showed the ARW was effective in providing color to milk matrices with satisfactory color stability. Milk components such as proteins and fats appeared to have protective effect on anthocyanins to prevent them from degradation when exposed to heat

(Jing and Giusti, 2005).

2.3 ANTHOCYANIN-PROTEIN COMPLEXATION

Complexations between anthocyanins and proteins is widely found, and they can occur naturally or be formed artificially. Anthocyanins vacuolar inclusion (AVI) is one of the most commonly seen naturally occurring anthocyanin-protein complexations, others include hordeumin, insoluble deposits in red wine bottles, and others. Usually the complexation has additional compounds (tannins, sugar, polyphenols, etc.) involved, however, the complexation could also be formed solely between these two components.

2.3.1 Anthocyanic Vacuolar Inclusion (AVI)

Anthocyanins in nature are found in two major forms: free anthocyanins that are homogeneously distributed in the vacuole or are trapped in dark-colored anthocyanic vacuolar inclusions (AVI) (Irani and Grotewold, 2005; Zhang et al., 2006). As shown in

Figure 2.9, AVIs are densely packed, dark red to purple pigmented, spherical with a 3 to

10mm diameter, membrane-less, sub-cellular proteinaceous matrix, to which highly concentrated anthocyanins are non-covalently bound (Chanoca et al., 2015; Markham et al., 2000; Mizuno et al., 2006; Silva et al., 2016). The AVI formation is believed primarily to increase pigments stability, and prevent their lytic degradation by certain vacuolar enzymes (Conn et al., 2003). AVI is also critical for dark color accumulation in plant tissues (Kovinich et al., 2011; Mizuno et al., 2006), as formation of this complex may decrease in the soluble anthocyanin concentration in the vacuoles, resulting in transport of more free anthocyanin from cytosol to vacuoles (Nozue et al., 1993). Due to its capability in facilitating color accumulation, this specific structure has been reported to be found in more than 70 dark colored plants such as purple corn, purple sweet potato, black soy bean, mung bean, red cabbage, red reddish, grape, berries, carnation, arabidopsis, and lisianthus (Conn et al., 2010; Kovinich et al., 2011; Markham et al.,

2000; Nozue et al., 1993; Nozzolillo and Ishikura, 1988; Poustka et al., 2007). Besides, the presence of AVI is believed to play a role in producing blueness to plant petal tissue, as AVI is widely found in blue hue flowers (Markham et al., 2000; Pourcel et al., 2010;

Zhang et al., 2006).

Figure 2.9 Anthocyanic vacuolar inclusion in an oblique hand-section of the adaxial epidermal peel of the deep purple inner petal region of lisianthus (Markham et al., 2000). White arrows point to the AVIs in the cell vacuoles.

A composition study of grape AVIs showed that these highly dense bodies were a complex mix of anthocyanins, long-chain tannins, small amounts of protein and other unidentified organic compounds encased by a lipid membrane (Conn et al., 2010).

Further analysis showed that protein and lipid profiles in grape AVIs were similar to those in tonoplast extracts, suggesting that the materials used to trap anthocyanins in AVI was directly coming from the tonoplast (Conn et al., 2010). As shown in Figure 2.10, a more recent study suggested an autophagy mechanism for AVI formation in cells

(Chanoca et al., 2015). No clear pattern of protein structure was found in grape AVI

(Conn et al., 2010), suggesting there was no specific protein structure favored in AVI 31 formation. Up to now, the light-induced expressed VP24 in purple sweet potatoes is one of the rarely reported well-studied vascular protein. It was identified to involve in trapping anthocyanins through hydrophobic interactions and forming AVIs (Nozue et al.,

1997). In the ex vivo study, the barely acid (below pH 5.5) soluble VP24 easily combined anthocyanins further support their participation into AVI formation (Nozue et al., 1997). Another interesting fact about the proteins in AVI is, they have been shown tend to selectively trap acylated types of anthocyanins. Cyanidin and delphenidin acylated 3,5-diglycosides have been found in greater proportion in AVIs isolated from lisianthus cells (Markham et al., 2000). A similar anthocyanins profile pattern was also reported in grape AVIs of different varieties. The percentage of acylated (p- coumaroylated or acetylated) anthocyanins was about 28% higher compared to that in the whole cell extract (Conn et al., 2003; Mizuno et al., 2006). A possible explanation for this selectivity might be that acylation is required for anthocyanins to be transported into vacuole by the protein carriers (Zhao, 2015). A grapevine study demonstrated acylation was required for transporting malvidin or cyanidin glucoside into the vacuole by proteins

AM1 and AM3 through H+-dependent route (Gomez et al., 2009). Once the pigments became structural modified and transported into the vacuole, they could interact with other compounds and form AVIs.

Figure 2.10 Proposed autophagy mechanism for anthocyanic vacuolar inclusion (AVI) formation (Modified from Chanoca et al., 2015).

2.3.2 Hordeumin

Hordeumin is a novel purple pigment that formed during uncooked barley bran ethanolic fermentation, through oxidative polymerization among anthocyanins, tannins and proteins (Deguchi et al., 1999; Ohba et al., 1995, 1993). This complex contains 50% proteins, 10% polyphenols, and 15% sugars (Ohba et al., 1993), and has shown to have beneficial anti-mutagenic (Deguchi et al., 2000c), DPPH and superoxide radical scavenging capabilities (Deguchi et al., 2000a). Similar to anthocyanins, hordeumin

33 showed color change properties with various pH values when distributed in 1% HCl methanol, but its hue was very different from grape anthocyanins under identical pH

(Ohba et al., 1993). Hordeumin presents a bluish-purple color when pH is very acidic, and the red hue becomes obvious as the pH raised to 3 and 5 (Ohba et al., 1993). When the pH is above 7, hordeumin starts to show a vivid blue color. The green hue becomes more and more intense as the pH increases to a more alkaline region (Ohba et al., 1993).

Identified by acid hydrolysis, the major anthocyanidins in hordeumin were cyanidin and delphinidin (Ohba et al., 1993), in agreement with the anthocyanidins in lisianthus AVIs

(Markham et al., 2000). A storage stability study showed hordeumin had 2 to 10 times higher pigment retention than standard anthocyanidins under 21 days, dark, 30°C storage and 15 days, light irritated, 15°C storage (Deguchi et al., 2000b). Different from typical anthocyanins which are highly unstable in low acidic to neutral environment, hordeumin presented excellent color stability at a pH range of 5 to 7 (Deguchi et al., 2000b), suggesting it has great potential to serve as natural food colorant alternative to fill the gap in a slight acidic food matrix.

2.3.3 Insoluble Deposits in Red Wine Bottle

The insoluble deposits in red wine bottle has been a problem for Australia, New Zealand,

French, Italian, and Portuguese red wine producers. In the 1990s, about 5% to 10% of the bottle red wines made from all major black grape varieties had this issue. The visual appearance of the wine was not favorable although the wine quality did not seem to be affected by the formation of deposits (Waters et al., 1994). The composition of the deposit was found by 13C NMR to be a phenolic polymer of anthocyanins, tannins, and

34 proteins (Waters et al., 1994). Hydrolysis showed the major amino acids in the deposit were glycine, aspartic acid/asparagine, serine, and threonine (Waters et al., 1994).

Different from anthocyanin pattern in AVI and hordeumin, the strong NMR methoxyl signal at 56 ppm in the deposit was assignable to methoxyl containing malvidin, peonidin, and petunidin units (Waters et al., 1994). The anthocyanin units were believed to be incorporated into the phenolic polymer through strong chemical association, even covalent bonds. Based on the evidences provided by NMR, Waters et al. (1994) proposed that in the slightly acid environment, the electrophilic anthocyanin and tannin react readily with nucleophilic groups on proteins. Potential nucleophilic groups in proteins for the deposit formation could be the thiol group of cysteine and the hydroxyl groups of serine and threonine (Waters et al., 1994). The insoluble property of this complex might be contributed by the extensive cross-linking between the phenolic polymers and proteins.

2.3.4 Anthocyanin-Whey Protein Aggregate Particles

Apart from naturally occurring anthocyanin-protein complexes reported previously, the anthocyanin-whey protein aggregate particles are typical artificially produced complexes.

After mixing dried whey protein isolate (WPI) with diluted cranberry, blackcurrant, and muscadine grape juice, respectively, to obtain 20% w/w protein mixtures at pH 4.5, the

1–100 μm sized aggregate particles were collected by centrifugation (Schneider et al.,

2016). Approximately 9 to 50 proteins were reacted with per polyphenol, and this ratio increased with increasing phenolic content of the juice used to create the particles

(Schneider et al., 2016). Acids and sugars in the juices were all involved in the particle

35 formation, but the size of the particles was mainly determined by the polyphenols

(Schneider et al., 2016). B-type proanthocyanidins, which are rich in muscadine grape juice, were not detected in significant level in WPI- muscadine grape juice aggregate particles (Schneider et al., 2016). However, anthocyanins (rich in all three juices) and A- type proanthocyanidins (rich in cranberry) were successfully transported into aggregate particles in a quite high level (Schneider et al., 2016). Particles had been shown may improve foam stability. In addition, phenolic extracts from the protein–polyphenol particles reduced the expression of genes associated with chronic inflammation

(Schneider et al., 2016), suggested these protein–polyphenol particles can be made to better modify food processing properties and provide additional health benefits at the same time.

2.4 INFRARED TECHNOLOGY FOR PROTEINS ANALYSIS

Infrared (IR) technology detects electromagnetic energy of vibration between atoms in a molecule and thus provides “fingerprint” information about chemical composition and structural conformation of the testing matrix. It has been widely applied in rapid detection market (Wang, 2014). Since each material has its unique atom combination, none of the two compounds may produce identical signal profiles when scanned by IR. The introduction of Fourier transform (FT) to IR spectroscopy enhanced the signal to noise ratio of the spectrum, allowing for more precise and high throughput analysis. In addition, IR is a simple, rapid technique that can provide reliable analytical results without destructing the samples. Taking advantages of these specific characters, FT-IR has been successfully applied in food components analysis for qualitative and

36 quantitative purpose (Rodriguez-Saona and Allendorf, 2011). It has been reported to be used in identification, adulteration testing, chemical profile analysis, and key component quantification in fruits juices (He et al., 2007; Snyder et al., 2014), tomatoes (De Nardo et al., 2009; Wilkerson et al., 2013), milk (Santos et al., 2013a, 2013b), cheese (Rodriguez-

Saona et al., 2006; Subramanian et al., 2009), cereal (Hassel and Rodriguez-Saona, 2012;

Lin et al., 2014; Rodriguez-Saona et al., 2000), snacks (Shiroma and Rodriguez-Saona,

2009; Wang and Rodriguez-Saona, 2012), oil (Aykas and Rodriguez-Saona, 2016; Koca et al., 2010; Maurer et al., 2012; Wenstrup et al., 2014), and even food-born bacteria

(Baldauf et al., 2007; Rodriguez-Saona et al., 2004, 2001).

The IR region of the electromagnetic spectrum ranges from 14000–50 cm−1, and it is divided into three areas: near IR (14,000–4,000 cm−1), mid IR (4,000–400 cm−1), and far

IR (400–50 cm−1). The MIR absorption bands are caused by fundamental vibrations of a specific functional group (Guillén and Cabo, 1997). Because the band intensities are proportional to the concentration of their respective functional group, the MIR spectra are commonly applied in structural identification and specific compound quantifications

(Rodriguez-Saona and Allendorf, 2011). Examples include 966 cm−1 for lipid trans double bond, 1175 cm−1 for lipid triglyceride C-O ester linkage, 1100-1000 cm−1 for aqueous sugar C-O and C-H stretching, 1640 cm−1 for water OH bending and 3300 cm−1 for water OH stretching (Wang, 2014).

Proteins are one of the three major nutrients in food with great importance. Thanks to their unique peptide bond structure, they can be easily identified and quantified by MIR spectroscopy (Plundrich et al., 2014; Strug et al., 2014). Amide I vibration, which is

37 hardly affected by the nature of side chain, absorbs around 1650 cm−1 and is mainly contributed by C=O stretching (Barth, 2007). It depends on the protein backbone secondary structure and therefore commonly used in secondary structure analysis. For instance, α-helices give rise to a main absorption band close to 1655 cm−1 and a shoulder at lower wavenumbers, antiparallel β-sheets exhibit a strong band near 1630 cm−1 and a weaker band near 1685 cm−1, and parallel β-sheets have their main absorption at higher wavenumber than corresponding antiparallel β-sheets though there might be only 4 cm−1 difference (Barth, 2007). The Amide I signal could also be used in hydrogen bonding investigation. It was reported hydrogen bond to C=O could lower the Amide I frequency by 20 to 30 cm−1 and hydrogen bond to NH by 10 to 20 cm−1 (Mennucci and Martínez,

2005). Amide II vibration, absorbing at around 1550 cm−1, is provided by the combination of NH deformation and CN stretching (Barth, 2007). Similar to Amide I,

Amide II signal is hardly affected by the side chain and may be used for secondary structure prediction (Oberg et al., 2004). However, the correlation between protein secondary structure and Amide II signal change is not as straightforward as it is with

Amide I (Barth, 2007). Different from Amide I and II, side chain may affect the mode of

Amide III. Predominantly contributed by CN stretching and NH bending, Amide III absorbs much wider range from 1400 to 1200 cm−1. Because this is a very information- intense region, the Amide III bands rarely used alone. It could be applied in protein secondary structure prediction together with other signals (Cai and Singh, 2004).

Recent examples of MIR application in protein analysis include quantification and structure quality scan. Strug et al. (2014) developed a membrane-based MIR method for

38 rapid biological protein quantification. In this method, at least 2 μL of the biological sample has to pass a hydrophilic polytetrafluoroethylene membrane to allow rapid water removal before scanning for its MIR spectra (Strug et al., 2014). A standard curve was established using the same approach. Quantification was achieved through different

Amide I signal intensities in the standard protein series, based on Lambert-Beer’s Law

(Strug et al., 2014). One of the advantages of this MIR approach over traditional colorimetric method was, this MIR-based analysis was not affected by the vast protein difference in amino acid composition or size. Because the Amide I signal depends mainly on peptide bond, which is universal for all proteins, the MIR standard curve constructed using different proteins (BSA, protein A, and rabbit IgG) were almost identical (Strug et al., 2014). To qualitatively investigate the complexation of peanut protein and anthocyanins from different plant source, Plundrich et al. (2014) applied MIR and found anthocyanins complexation modified the peanut protein secondary structure. The Amide I and II bands of the complex became narrower after the pigments addition, suggesting the binding between the two compounds restricted the conformational freedom of the proteins (Plundrich et al., 2014).

2.5 INDUSTRIAL PROCESSING OF FOOD ANTHOCYANINS

Anthocyanins can be used as food colorant although the use is restricted to some products varying among countries. According to Codex Alimentarius Commission, the European

Union legislation listed anthocyanins as natural colorant with product number of E163.

The same code is assigned to anthocyanins for food application purpose under Japan legislation. In the US, the FDA (Food and Drug Administration) lists anthocyanin as a

“natural” color does not require certification, but anthocyanins can be only obtained either from “grape color extract”, “grape skin extract”, or “fruit juices or vegetable juices” (Mateus and de Freitas, 2009). To make it easier for the industry to apply anthocyanins in their formulation and production, the pigments are usually prepared into dark-colored concentrate or powders. Figure 2.11 showed a typical flow chart for industrial purple corn cob pigments production (Jing, 2006). This protocol may also apply to food grade anthocyanins production for other plant sources. Pigments extraction from the plant material and spray drying the pigment extracts into powders are the two main key steps for anthocyanin-based food colorant processing.

Figure 2.11 Flow chart of purple corn cob colorant production (Jing, 2006).

2.5.1 Extraction

For food used anthocyanins production, an ideal extraction process would be maximizing the pigments yield while minimizing the degradation and alteration the nature state of the target compounds. Extraction of anthocyanins from plant materials could be affected by

(1) the nature of plant tissue, including the tissue structure and the location of pigments to be extracted, (2) pretreatment of the materials, (3) the polarity of the target compounds,

(4) the selection of extraction methods, (5) the choice of extraction solvent, (6) the length of extraction time, (7) the temperature applied in the extraction process, (8) the solid:liquid ratio in the extraction matrix, (9) the application of new extraction techniques, such as assistance with ultrasound, microwave, enzymes, pulsed electric field, high hydrostatic pressure (Azmir et al., 2013; Castañeda-Ovando et al., 2009;

Corrales et al., 2008; Dai and Mumper, 2010; Routray and Orsat, 2012; Vilkhu et al.,

2008).

The purpose of pretreatment before extraction includes facilitating further extraction process, increasing the storage life, and decreasing the space requirement of the plant materials. Grinding, milling, maceration, homogenization, and drying are the commonly seen steps for pretreatment. Grinding and milling the plant materials into small particles or even powders, as well as maceration and homogenization, may increase the contact surface area between solute and plant materials for proper mixing with the extraction solvent (Azmir et al., 2013; Dai and Mumper, 2010). These pretreatments may breakdown the cellular structure of the plant materials and thus enhance the yield of anthocyanins (Routray and Orsat, 2012). Drying of the plant material may increase the

41 unit yield of plant material, it can be done by air-drying or freeze-drying. Freeze-drying maintained the functional compounds better than air-drying (Asami et al., 2003).

However, drying process could cause undesirable functional components profile change, therefore, the drying condition has to be well-controlled to minimize the potential damage (Abascal et al., 2005).

Conventional solvent extraction is the most widely applied extraction technique to isolate anthocyanins from the fruits, vegetables, and other plant tissues. The efficiency of this type of extraction mainly depends on the choice of solvents (Azmir et al., 2013; Dai and

Mumper, 2010). The polarity of the targeted compound is the primary factor for solvent selection. Since anthocyanins are a class of polar compounds, the solvent chosen for anthocyanins extraction were mainly water, methanol, ethanol, acetone, ethyl acetate, n- butanol, propylene glycol and their combinations (Dai and Mumper, 2010; Mateus and de

Freitas, 2009). Organic solvents and their aqueous mixtures may easily denature the cell membranes, dissolve and stabilize the pigments simultaneously. Things need to consider when choosing solvent for anthocyanins extraction include molecular affinity between solvent and solute, mass transfer, use of co-solvent, environmental safety, human toxicity and financial feasibility (Azmir et al., 2013). Acidification of the solvent is widely applied in anthocyanins extraction, it serves to provide a favorable acidic pH medium for the formation of more stable flavylium (Figure 2.5) form (Jackman et al., 1987). Low concentration of strong acids, such as <3.0% trifluoroacetic acid and < 1.0% of hydrochloric acid, as well as weak organic acids, like formic acid, acetic acid, citric acid, tartaric acid, phosphoric acid, are most commonly seen acids in anthocyanins extraction

(Dai and Mumper, 2010; Jackman et al., 1987; Nicoué et al., 2007). The acid level has to be well-controlled if maintaining the native state of pigments is required. Excess acids may lead to loss of labile acyl and sugar residues during anthocyanins extraction or subsequent concentration (Castañeda-Ovando et al., 2009; Jackman et al., 1987). Mild heating (temperature under 70°C) is recommended for anthocyanins extraction, as it can promote higher pigments yield and improve the extraction rate by increasing both solubility and mass transfer rate, decreasing the viscosity and the surface tension of the solvents to facilitate solvents to reach the sample matrices, although rapid pigments degradation could occur when the extraction temperature went too high (Dai and

Mumper, 2010). Longer extraction time may help increase the pigments yield, but the chance of anthocyanins getting oxidized is also getting higher. To further improve the efficiency of the conventional solvent extraction, constant or occasional mechanical shaking, vortexing, stirring, rotary can be applied to increase the further surface and molecular interaction (Azmir et al., 2013; Routray and Orsat, 2012). Some conditions for purple corn anthocyanins conventional solvent extraction are listed in Table 2.3. The optimum condition is a balance for all possible parameters involving into the extraction process.

In recent decades, although the traditional solvent extraction is still in use, novel technologies that provide higher extraction efficiency with smaller amount of plant materials, less organic solvent usage, less time and labor cost, greater specificity and selectivity, are becoming more and more favorable. The new techniques based on applying compressed fluids as extracting agents include subcritical water extraction,

43 supercritical fluid extraction, pressurized fluid extraction, accelerated solvent extraction

(Azmir et al., 2013; Dai and Mumper, 2010; Liu et al., 2008). Other advanced methods such as enzyme-assisted extraction, pulsed electric field extraction, microwave assisted extraction, and ultrasound-assisted extraction, they provide biochemical or mechanical shear force to disrupt biological membranes, facilitate the release of pigments, enhance penetration of solvent into cellular materials, and improve mass transfer (Dai and

Mumper, 2010; Ignat et al., 2011; Routray and Orsat, 2012; Vilkhu et al., 2008). A study working on anthocyanins extraction from grape by-product compared the pigments recovery of the traditional solvent extraction (50% aqueous ethanol, 70°C, 1h) with the three novel methods assisted with high hydrostatic pressure (HHP), pulsed electric field

(PEF) ultrasonics (Corrales et al., 2008). Their results showed both HPP and PEF recovered significantly higher amount of total anthocyanins than the traditional solvent extraction method (Corrales et al., 2008). The new techniques, HPP, PEF, and ultrasonics, presented high selectivity in anthocyanins extraction, suggested by their recovery of acylated pigments were at least 4-folds higher compared to conventional method (Corrales et al., 2008).

Solid- Part Anthocyanin Temperature Solvent Acidity Time liquid Reference (Cultivar) content (DW) (°C) ratio 0.94±0.03 g/100g water 0% 1h 50 Cob 1:25 (Jing and Giusti, 2007) 0.98±0.08 g/100g 70% acetone 0% 1h RT Cob 1% citric (De Pascual-Teresa et 34%, after drying 60% ethanol 3h 60 N/A (cv. Morado) acid al., 2002) (Pedreschi and Cisneros- Cob 37.03 mg/g 60% ethanol 0% 48h RT 1:7 Zevallos, 2007) Husk 9.1% -11.0% Kernels no accumulation methanol 1% HCl 12h 4 1:10 (Li et al., 2008) Cob 0.49%-4.60% Seed 1% 1N (Ramos-Escudero et al., 2.87 mg/g methanol 1h 40 1:10 HCl 2012)

45 1N HCl-95% ethanol

Cob 7.03 mg/g 73min 70 1:25 (Yang et al., 2009) (15:85 v/v) Whole grain 10.56 mg/100g (cv. Ayzuma) Whole grain 27.64 mg/100g (cv. Tuimura) stirred 3h at 0℃, no water-HCl (19:1, v/v) stirring 8h at RT 1:2 Whole grain 14.71 mg/100g (Montilla et al., 2011) (cv. Oke) Whole grain 51.25 mg/100g (cv. Kulli) Flour 63.73 mg/100g 80% ethanol N/A 2min RT 1:5 (cv. Kulli) Table 2.3 Conventional solvent extraction conditions for purple69 corn anthocyanins.

2.5.2 Spray Drying

Spray drying is a dehydrate operation by which a liquid product is atomized in a hot gas

(air or inert gas like nitrogen) current to instantaneously produce dried powders.

Depending on the nature of feeding materials and drying conditions, the mean size of the final powders could be ranging from 0.2 μm to 5 mm (Balassa et al., 1971; Fang and

Bhandari, 2010; Kandansamy et al., 2012). This technique has been widely used for drying heat-sensitive substances like anthocyanins because of the rapid solvent evaporation from the droplets may protect the substances from thermal degradation.

About 80% to 90% of the industrial encapsulated anthocyanins were achieved through spray drying (Mahdavi et al., 2014). Comparing to another anthocyanins drying technique, freeze-drying, spray drying is at least 30 to 50 times less expensive (Desobry et al., 1997). The instrument applies for spray drying operation is called spray dryer.

Figure 2.12 A schematic diagram of lab scale mini spray dryer (modified from Buchi

Corporation). 46

A schematic diagram of a lab-used spray dryer is shown in Figure 2.12. Pilot plant level and industry processing level spray dryers share the similar structure design to this lab scale equipment. There are three basic steps for spray drying, liquid atomization, gas- droplet mixing and dehydration, and product separation (Balassa et al., 1971). In the beginning of spray drying, the feeding materials are pumped into the drying chamber through the nozzle. The feeding materials can be a solution, an emulsion or a suspension of interested compounds and drying carrier mixture. As shown in Figure 2.12, nozzle is the place where liquid atomization happens. The main purpose of atomization is to facilitate heat and mass transfer through maximizing surface between the feeding materials and the dry gas. The most commonly used atomizers include pressure nozzle, pneumatic valve, homogenizing atomizers, sonic nozzle, two fluid nozzle and spinning disk configuration (Gharsallaoui et al., 2007). Once entering into the drying chamber, the feeding material droplets are well-mixed with the hot drying gas, the droplet dehydration process starts simultaneously. The duration each droplet stay inside the chamber is determined by the air flow rate and size of drying chamber, as the chamber size governs the air contact time (Murugesan and Orsat, 2012). A well-designed drying process usually lasts less than 30 seconds (Fogler and Kleninschmidt, 1938). As the droplets traveling in the chamber, heat transfer is carried out from air towards the droplets due to the temperature difference and water transfer is carried out in the opposite direction due to the vapor pressure difference. The heat and mass transfer proceeds until the temperature of the droplet reaches the air temperature. The separation of the product powders and the wet air is accomplished partially in the drying chamber itself by gravity,

47 the larger dried particles are directly deposited to the bottom of the spray dryer, whereas the fine powders are transported to external cyclones for separation (Cocero et al., 2009).

A filter bag is installed in the end of the air flow to collect the extra powders before letting the wet gas into the air.

Critical parameters of spray drying are inlet and outlet temperatures of air, feed flow rate, feed temperature, viscosity of feed, solid content, wall material selection, surface tension, volatility of solvent, and nozzle conditions (Mahdavi et al., 2014). It is widely accepted that the optimum spray-drying condition is a compromise between high air temperature, high solid concentration of the solution, and easy pulverization and drying without expansion and cracks of final particles (Gharsallaoui et al., 2007). Inlet/outlet temperature, feed flow rate, feed temperature, carrier selection, are the most easy- accessed important factors to manipulate for spray drying optimization (Ferrari et al.,

2012b; Liu et al., 2001; Tonon et al., 2010). The air inlet/outlet temperature determines the drying rate and the final moisture content of the product. A study working on spray drying black mulberry anthocyanins showed the moisture content of the spray-dried pigment powders decreased with the increase in inlet air temperature, the moisture content of the product powder was around 1.5% when drying in 150°C while it was nearly 2% when using the same conditions except the air inlet temperature was set to

110°C (Fazaeli et al., 2012). There are two main factors need to consider when determining air inlet temperatures. First, the temperature should be safe to use without damaging the interested compounds or creating operating hazards (Fogler and

Kleninschmidt, 1938). Secondly, the comparative cost of heat sources should be

48 acceptably low (Fogler and Kleninschmidt, 1938). Adjusting feed flow rate can help ensure each sprayed droplet is dried to the desired level before it comes to the separating cyclone. In a spray drying blueberry study, the feed flow rate 8.5 mL/min produced around 16% moisture powders regardless of inlet temperatures (140°C or 160°C), while the powders moisture produced by feed flow rate 9.6 mL/min was at least 3% higher

(Jiménez-Aguilar et al., 2011). The other factor, feed temperature is important because it is responsible for the viscosity of the emulsion. When the feed temperature is increased, viscosity and droplet size would be decreased (Kandansamy et al., 2012). Finally, the choice of a carrier for spray drying pigments can affect the final yield of the product and microcapsule stability. The mostly widely used carrier maltodextrin was reported to yield more blackberry anthocyanins under the identical spray drying condition than the same amount of gum Arabic (Ferrari et al., 2012a). The drying yield of spray dried black mulberry juice when applied 16% maltodextrin as carrier was more than 80% , which was higher than around 75% for 8% maltodextrin operating under same condition

(Fazaeli et al., 2012). A study spray dried acai pigments with different carriers, their half- lives based on total anthocyanins contents varied from 814 days for tapioca starch carrier,

979 days for gum Arabic carrier, to 1248 days for maltodextrin carrier (Tonon et al.,

2010). Typical carriers reported to be suitable for spray drying include natural gums (gum

Arabic, alginates, carrageenans, etc.), proteins (dairy proteins, soy proteins, gelatin, etc.), carbohydrates (maltodextrins and cellulose derivatives) and/or lipids (waxes, emulsifiers)

(Balassa et al., 1971; De Vos et al., 2010; Gharsallaoui et al., 2007; Kandansamy et al.,

2012).

Chapter 3: Health Benefits of Purple Corn (Zea mays L.) Phenolics

3.1 ABSTRACT

Purple corn (Zea mays L.), one of the grains with deepest shade among the plant kingdom, has caught the attention of the food industry as it could serve as an alternative to synthetic color. Rich in phenolics with potential health promoting properties, purple corn has become a rising star in the novel ingredients market. The blooming purple corn business among the world is an indicator of interests on purple corn application. Being widely advertised as a healthy food, the available information on purple corn health benefits have not yet been well reviewed and summarized. In this study, we present compositional information focused on the potential functional phenolic compounds correlated to health-promoting effects. Health benefit studies including in vitro tests, cell models, animals and human trials were also discussed. This paper emphasizes research using purple corn or its extracts. Some other plant sources with similar phenolic composition to purple corn are also mentioned. Dosage and toxicity of purple corn studies are also reviewed. Purple corn phenolics have been shown in numerous studies to have potent antioxidant, anti-inflammatory, anti-mutagenic, anti-carcinogenic, anticancer, anti- angiogenesis properties and capable to ameliorate lifestyle diseases such as 50 obesity, diabetes, hyperglycemia, hypertension and cardiovascular diseases based on their strong antioxidant power involving biochemical regulation amelioration. With promising evidence in cell and animal studies, this rich source of health promoting compounds deserves additional attention to better understand its potential contributions to human health.

Keywords: bioactive compounds, anthocyanins, antioxidants, chemoprotective, dosage

3.2 INTRODUCTION

Purple corn (Zea mays L.), also known as purple maize, is a crop native to the Andes regions of Peru and has been widely cultivated and consumed throughout the Andean region of South America, mainly Peru, Ecuador, Bolivia and Argentina. It has one of the deepest purple shades among the plant kingdom. Due to its richness in purple pigments, purple corn pigments have long been used to color foods and beverages. In South

America, the purple corn extracts are widely applied in coloring homemade desserts and beverages: chicha morada and mazamorra morada, which are two of the most popular drinks and desserts prepared from purple corn in its original places (FAO, 2013).

Moreover, through the years, other countries have also shown interest on using these rich sources of color to obtain food colorants. According to United Nations BioTrade facilitation program (2012), the average yearly growth of Peruvian purple corn exports value reached 467% during 1998 to 2002, and the price of purple corn had almost doubled from $0.75 to $1.36/kg in this five year period. In 2002, Peru exported

$24,220,360 of natural colorants, of which $98,000 was contributed by purple corn

51 anthocyanins products. With purple corn color being recognized by the European Union with the code E-163 and the same code for the Japanese Legislation, the import of purple corns and their color products to Germany, France, Italy, Japan and other countries is growing. The exports of Peruvian purple corn business reached $187,745,641 in 2010. Its use as food ingredient and as food colorants has been blooming around the world in the recent decades.

For many years Japan was the main market for purple corn or purple corn cob, and this is reflected on the predominance of early research on purple corn that originated from Japan or Japanese researchers (Fukamachi et al., 2008; Sasaki et al., 2007; Tsuda, 2008; Tsuda et al., 2006, 2005, 2004, 2003), that closely linked to an interest on the potential health benefits of purple corn. These researches promoted the international purple corn trade, making purple corn a more popular novel ingredient in food industry, and now researchers around the world are paying more and more attention to this rich source of phytochemicals. Anthocyanins, the pigments in purple corn, have been reported to be associated with the potential to reduce the risks of cardiovascular disease, obesity, diabetes, cancer and chronic diseases (He and Giusti, 2010; Konczak and Zhang, 2004).

Besides the pigments, other functional phenolic compounds in purple corn have also been reported to have attenuating effects on chronic diseases such as hypertension, diabetes and cancer (Kim et al., 2013; Long et al., 2013). In 2013, the phenolic-rich purple corn was proposed by different commercial companies to have the status of a super food for its remarkable potential health benefits.

The objective of this paper is to review and summarize the recently updated knowledge of purple corn phenolics health benefits, providing information for the use of purple corn ingredients that have desirable nutritional values to meet the requirement for healthy foods formulation and development. Information covered in this review includes potential functional phenolic compounds previously reported in purple corn, cell, animals and human health beneficial studies relating to purple corn phenolics and extracts, and dosage studies related to purple corn toxicity.

3.3 FUNCTIONAL COMPOUNDS IN PURPLE CORN

Being in the corn (Zea mays L.) family, purple corn is similar to traditional yellow corn that is also nutritious, for being rich in starch (61%-78% dry basis, db), non-starch polysaccharides (about 10% db) (Sinha et al., 2011), proteins (6%-12% db), lipids (3%-

6% db), minerals, and vitamins (Ai and Jane, 2016). The macronutrients in corn and their beneficial effects on postprandial glycemic/insulinemic responses, lipid metabolism, colon health, and mineral absorption have been well summarized. However, it is anthocyanins and other phenolics, which differentiate purple corn from other regular corn varieties, make it stand out as a health promoting food. The typical anthocyanin and phenolic profiles of purple corn stand out as health promoting food. The total phenolic content of Andean purple corn determined by Folin-Ciocalteu assay was 1756mg /100 g, higher than the well-known phenolic rich blueberries with range of 138 to 672 mg/100 g

(Cevallos-Casals and Cisneros-Zevallos, 2003; Prior et al., 1998). Additionally, research also showed that deeper shade corn tended to have higher total phenolics value (Montilla et al., 2011). Studies related to purple corn health benefits have therefore been conducted 53 mainly with anthocyanin and phenolic rich extracts obtained from purple corn. Studies performed with purple corn juices or powders will also be presented in this document.

Additionally, dose dependent correlation between the amount of these purple corn functional compounds and certain health promoting properties were often found in these studies.

3.3.1 Anthocyanins

The anthocyanins profile of purple corn has been well studied, detailed purple corn anthocyanins profile was reviewed previously in part 2.2.2.

3.3.2 Other phenolics

Phenolics are non-polymeric phytochemical compounds that contain at least one phenol structure. In this chapter, other phenolics in purple corn refer to all phenolic compounds except anthocyanins, phenolic acids and flavonoids are the major two subclasses of phenolics found in purple corn. A variety of other phenolics besides anthocyanins may contribute to the potential health promoting properties reported for purple corn and purple corn extracts in humans.

3.3.2.1 Phenolic acids

Phenolic acids in purple corn are usually recovered by organic solvent such as alcohol and acetone (Moore et al., 2005; Žilić et al., 2012) and identified by HPLC-MS

(Pedreschi and Cisneros-Zevallos 2006, 2007; Ramos-Escudero et al., 2012; Žilić et al.,

2012). The p-coumaric acid and ferulic acid contents for Bolivian dark hue purple corn were 607.5mg/100g DW and 154.2mg/100g DW respectively, much higher than the

54 yellow cultivar which reported to have 366.7 mg/100g DW and 132.9mg/100g DW

(Cuevas Montilla et al., 2011). The phenolic acid contents in different purple corn varied among cultivars. A more recent study showed ferulic acid (39.9 mg/kg DW for native and

33.3 mg/kg DW for hybrid maize) was the most abundant phenolic acid, followed by diferulic and coumaric acid in Mexico purple corn (Urias-Lugo et al., 2015). Most of the phenolic acids are in conjugated or bound form in purple corn. Only a small portion are in free soluble form that can be easily extracted without further hydrolysis treatment

(Cuevas Montilla et al., 2011; Žilić et al., 2012). Currently, there are more than 9 different phenolic acids that have been reported to be found in purple corn. As shown in

Figure 3.1, they were protocatechuic acid, vanillic acid, syringic acid, 2,4,6- trihydroxybenzoic acid, p-coumaric acid (also called “p- hydroxycinnamic acid”), caffeic acid, ferulic acid, chlorogenic acid, and p-hydroxyphenyl acetic acid, as well as their derivatives (Kim et al., 2013; Montilla et al., 2011; Ramos-Escudero et al., 2012; Žilić et al., 2012).

3.3.2.2 Flavonoids

Besides phenolic acids, the flavonoids are another class of functional compounds found in purple corn. The total flavonoids content in purple corn ranged from 307.42mg/kg DW to 337.51mg/kg DW while the yellow hue corn only had 248.64mg/kg DW to

281.20mg/kg DW (Žilić et al., 2012). Similar to phenolic acids, flavonoids in purple corn are recovered by organic solvent mixtures (Moore et al., 2005; Žilić et al., 2012). Total flavonoid content in purple corn was determined by forming flavonoid-aluminum complex and measuring absorbance after alkaline treatment (Jia et al., 1999). Rutin,

55 hirsutrin, morin, kaempferol, quercetin, naringenin, hesperitin and their derivatives are the most commonly reported flavonoids that were recovered from purple maize (Kim et al., 2013; Pedreschi and Cisneros-Zevallos 2007; Ramos-Escudero et al., 2012). The chemical structures of these flavonoids are shown in Figure 3.2.

p-hydroxyphenyl acetic acid Quinic acid Figure 3.1 Chemical structures of typical phenolic acids in purple corn (Zea mays L.). 56

Hirsutrin Rutin

Figure 3.2 Chemical structures of typical flavonoids in purple corn (Zea mays L.).

3.4 HEALTH BENEFITS OF PURPLE CORN PHENOLICS

Interest in the health benefits of purple corn anthocyanins and other phenolics has been on the rise in recent decades. There is a large body of evidence suggesting that these compounds may help reduce the incidence of a variety of chronic diseases. However, it is clear that not all anthocyanins or all phenols are equal in their potential health promoting properties. Here we focus specifically on studies regarding the health promoting properties of purple corn. Studies covered here were either performed with purple corn

57 powders, extracts or juices, or conducted with individual compounds known to be abundant in purple corn and purple corn products.

3.4.1 Antioxidant

Free radical involved chain reaction is the generally accepted mechanism for degenerative oxidation in living tissue (Wang and Stoner, 2008). Antioxidant capability often refers to the scavenging ability to these reactive oxygen radicals: superoxide, singlet oxygen, peroxide, hydrogen peroxide, as well as hydroxyl radicals (Wang and

Stoner, 2008). It is thus believed that antioxidants provide health protection to biological systems. The antioxidant property of purple corn has been well evaluated in free radical scavenging tests, as well as cellular studies in vitro and animal studies in vivo.

3.4.1.1 Free radical scavenging tests

The free radical tests used to determine the antioxidant capabilities of purple corn phenolics including DPPH (2,2-diphenyl-1-picrylhydrazyl), ABTS (2,2’-azino-bis-(3- ethylbenzylthiazoline-6-sulfonic acid), APPH (2,2’-azo-bis-(2-amidinopropane) dihydrochloride), ORAC (oxygen radical absorbance capacity), FRAP (ferric reducing antioxidant power) and deoxyribose assay and nitric oxide scavenging assay (Cevallos-

Casals and Cisneros-Zevallos 2003; Del Pozo-Insfran et al., 2006; Pedreschi and

Cisneros-Zevallos 2006; Lopez-Martinez et al., 2009; Lee et al., 2010; Lopez-Martinez et al., 2011; Ramos-Escudero et al., 2012b; Vayupharp and Laksanalamai 2015;

Duangkhamchan and Siriamornpun 2015). Consistent from study to study, various purple corn extracts presented positive antioxidant capability in all these tests throughout the years. Overall, the antioxidant activity of purple corn, besides impacted by methodology,

58 genotypes and location (Khampas et al., 2015), is highly correlated with the amount of bioactive compounds such as polyphenol, flavonoids, flavanols and anthocyanins

(Ramos-Escudero et al.,, 2012). Pedreschi and Cisneros-Zevallos (2006) separated purple corn anthocyanins from other phenolics, and concluded the antioxidant activities for purple corn anthocyanins (1.019±0.05µg Trolox equivlent/µg) was higher than other phenolic compounds (0.838±0.11µg Trolox equivlent/µg). In another study, these two portions showed no significant difference in ABTS reducing power test (Lopez-Martinez et al., 2009). It could not be denied that both purple corn anthocyanins and other phenolic compounds were great antioxidants though some variation on their antioxidant power was found in different studies. The reason why purple corn had relatively high antioxidant activity compared to other fruits and vegetables was studied using DPPH assay. The study showed purple corn phenolics have higher antioxidant capability and faster reaction kinetics than the same amount of blueberry phenolics (Cevallos-Casals and Cisneros-Zevallos, 2003). These results suggested that comparing to blueberry, purple corn had larger number of active hydroxyl groups, as well as more favorable configuration to allow better interaction with the free radicals (Cevallos-Casals and

Cisneros-Zevallos, 2003).

Additionally, the purple corn free radical scavenging capability remains quite high after industrial processing (Del Pozo-Insfran et al., 2006; Lopez-Martinez et al., 2011). The antioxidant properties of cooked purple corn kernels, tortillas and chips were investigated using ABTS, APPH and ORAC methods: for both Mexico and America genotypes.

Anthocyanin losses were similar with average losses of 37%, 54% and 75% when

59 processed into nixtamal, tortillas, and chips, respectively; their average antioxidant capability loss were 28% after nixtamalization, 37% when processed to tortillas, and 55% if processed to chips (Del Pozo-Insfran et al., 2006). The decreased antioxidant response might be associated with losing bioactive phenolic compounds during the alkaline and thermal processing.

3.4.1.2 Cellular and animal studies

The effect of purple corn extracts on cellular antioxidant response in mouse organs was investigated by inducing H2O2 to isolated mouse kidney, liver and brain (Fernando

Ramos-Escudero et al., 2012). The presence of malondialdehyde (MDA) in these organs served as an indicator of cell membrane oxidative injury after applying H2O2. When the organs were treated with purple corn extracts at low MDA levels, the levels of antioxidant enzymes superoxide dismutase (SOD), catalase (CAT) and total peroxidase

(TPX) in the organs increased (Fernando Ramos-Escudero et al., 2012). These enzymes were believed to have the capability to help remove reactive oxygen species and preventing cells from oxidative damage. The results suggested certain functional compounds in purple corn extract could penetrate membranes, participate in helping stimulate antioxidant enzymes secretion to reduce oxidative damage caused by free radicals (Fernando Ramos-Escudero et al., 2012). A similar study demonstrated the protective effect of purple corn on oxidative damage in rat liver and kidney in vivo

(Zhang et al., 2014). Different from study of Ramos-Escudero et al. (2012), the oxidative stress was induced by putting fluoride in rats drinking water (Zhang et al., 2014). In their experiments, swollen cell and vague cell boundaries, which were signs of high oxidative

60 stress, were observed in the fluoride alone treated rats, and these rats also had high MDA levels in serum and liver tissue (Zhang et al., 2014). However, the cells in rats treated with both fluoride and purple corn rich diet suffered less swollen problem and their MDA levels were significantly lower (Zhang et al., 2014). The antioxidant enzymes (SOD and glutathione peroxidase) levels for purple corn treated rats were significantly higher than the fluoride alone treated group (Zhang et al., 2014). These results suggested that purple corn may alleviate the fluoride-induced oxidative damage by elevating the antioxidant ability in liver and kidney of rats (Zhang et al., 2014).

Another study evaluated purple corn extract antioxidant capability in vivo with healthy rats. The purple corn color fed rats presented no significant difference to the regular diet rats in cell histology (Yokohira et al., 2008). However, serum of rats treated with purple corn color presented strong antioxidant power in potential antioxidant (PAO) test in vivo and microarray analyses showed purple corn color could induce RNA expression of

P450 (cytochrome) oxidoreductase, phosphatidylinositol 3-kinase, and phospholipase A2

(Yokohira et al., 2008). The author thus suggested that purple corn may be effective as antioxidants in vivo and have chemopreventive potential against liver preneoplastic lesion development (Yokohira et al., 2008).

3.4.2 Anti-inflammatory

Inflammation could play a role in promoting some kinds of cancer in animals and humans

(Kwon et al., 2007). Anti-inflammatory refers to the property of reducing inflammation.

A study revealed that purple corn pigments can debilitate high glucose induced mesangianl inflammation, expansion and hyperplasia by disturbing the inflammatory

61 action of IL-8 (Li et al., 2012a). When the renal mesangial cells of db/db mice were exposed to high glucose to induce diabetes, the production of interleukin-8 (IL-8), a chemokine that linked to inflammatory process in glomerulus, was markedly elevated (Li et al., 2012a). In contrast, the cells received purple corn anthocyanins-rich extract treatment showed mitigation of IL-8 secretion in a dose-dependence manner, detected by

ELISA (Li et al., 2012a). Similar study also demonstrated that purple corn pigments antagonized diabetic kidney problems through controlling IL-8-Tyk2-STAT signaling pathway (Kang et al., 2012).

Cyanidin-3-glucoside (C3G) is the most abundant anthocyanin throughout the plant kingdom as well as in in purple corn. Numerous researches have been shown that C3G has strong anti-inflammatory properties although some of their sources of C3G were not purple corn (Min et al., 2010; Reddy et al., 2005; Serra et al., 2013; Tsuda et al., 2002).

Cyclooxygenase (COX), a well-known inflammatory protein, whose abnormal up- regulation is commonly found in many cancers (Martin et al., 2003; Wang and Stoner,

2008). By efficiently inhibiting COX-1 and COX-2 enzyme activity, C3G from Prunus cerasus exhibited strong anti-inflammatory property (Reddy et al., 2005). Tsuda and his co-workers (2002) demonstrated of C3G may play a role in prevention of the NO- mediated inflammatory diseases. C3G were found to suppress zymosan-induced inflammatory response in rats when orally administered: the elevation of the peritoneal exudate NOx, tumor necrosis factor (TNF) and other inflammatory markers were significantly suppressed by the administration of C3G (Tsuda et al. 2002). In a lipopolysaccharide-induced inflammation study using RAW 264.7 cells, C3G from black

62 rice and its metabolites showed suppressing the production of the pro-inflammatory cytokines, TNF-α, interleukin-1β (IL-1β), two inflammatory mediators, NO and prostaglandin E2 (PGE2), as well as the gene expression of nitric oxide synthase (iNOS) and COX-2 (Min et al., 2010). In addition, in the in vivo study with carrageenan-induced inflammation BALB/c mice, C3G from black rice was demonstrated significantly inhibiting the leukocyte number and the levels of TNF-α, PGE2, and protein in the exudates of the air pouch in carrageenan-treated mice, as well as COX-2 expression and nuclear factor-kappa B activation (Min et al., 2010). Another study working on cytokine- induced inflammatory response in HT-29 human intestinal cells showed that C3G presented higher anti-inflammation efficiency than 5-aminosalicylic acid, a well- established anti-inflammatory drug, in reducing NO, PGE2, IL-8 production and iNOS and COX-2 expressions (Serra et al., 2013).

3.4.3 Anti-mutagenic

A mutagen is an agent that introduces change to the genetic material, usually DNA, and thus increases the frequency of mutations. The property of anti-mutagenic refers to the capability of reducing the rate of mutation. Anthocyanins and phenolic acids are well- known anti-mutagenic agents through studies in other fruits and vegetables (Santos-

Cervantes et al., 2007; Yoshimoto et al., 2001). The anti-mutagenic property of purple corn was examined by a research group in Texas A&M University using the Ames test.

Both anthocyanin-rich water fraction (WF) and ethyl acetate phenolics-rich fraction

(EAF) presented dose-dependent anti-mutagenic activity against a food mutagen, Trp-P-

1(Pedreschi and Cisneros-Zevallos, 2006). The lower IC50 value of the EAF (95.2±10.95

µg of chlorogenic acid equivalent/ plate, comparing 321.7±21.36 for WF) indicated the purple corn phenolics were more potent anti-mutagen than the anthocyanins (Pedreschi and Cisneros-Zevallos, 2006). This study also revealed that the mechanism of anti- mutagenic behavior of the purple corn anthocyanins are predominantly a blocking effect on the S-9 Mix activation system of the mutagen; whereas for the purple corn phenolics, a dual mechanism involving both blocking of the S-9 Mix and scavenging action on Trp-

P-1 electrophiles (Pedreschi and Cisneros-Zevallos, 2006). Although in the typical food processing steps, some of the pigments were lost during nixtamalization process (the first step for masa and tortilla production), the processed purple corn still presented anti- mutagenic activity against 2-aminoanthracene induced mutagenicity in Ames test

(Mendoza-Díaz et al., 2012), thus consuming purple corn processed food may still have beneficial effect to humans.

3.4.4 Anti-carcinogenic and anti-cancer

A carcinogen is a substance, radionuclide, or radiation that has ability to damage the genome or to disrupt of cellular metabolic processes, and thus exacerbates cancer or increases its propagation (Poirier, 2004). Anti-carcinogenic refers to the property of inhibiting or preventing the activity of a carcinogen or the development of carcinoma.

3.4.4.1 Mammary and prostate carcinogenesis

An estimated 234,190 (231,840 for women and 2,350 for men) of breast cancer as well as

220,800 new cases of prostate cancer are expected to occur in the US during 2015

(American Cancer Society, 2015). Breast cancer for women and prostate cancer for men are the most frequently diagnosed cancer aside from skin cancer (American Cancer

Society, 2015). The anti-carcinogenic property on mammary and prostate carcinogenesis of purple corn has been evaluated in vitro and in vivo (Fukamachi et al., 2008; Long et al., 2013). The purple corn color (PCC) showed significantly inhibition activity on 7,12- dimethylbenz[α]anthracene induced mammary carcinogenesis development in human c-

Ha-ras proto-oncogene transgenic (Hras128) rats in a dose dependence manner as the incidence of middle sized (0.5-2.0g) mammary tumors were significantly decreased in

PCC fed rats (Fukamachi et al., 2008). Although the number and incidence of small sized

(<0.5g) mammary tumors was not suppressed by PCC, indicating PCC was not able to inhibit the emergence of mammary tumors in these rats (Fukamachi et al., 2008). Of the in vitro study, PCC could inhibit cell viability and induce apoptosis in mammary tumor cells derived from Hras128 rat mammary carcinomas (Fukamachi et al., 2008). At the molecular level, PCC treatment preferred activation of caspase-3, a key protease associated with DNA fragmentation and apoptosis, and reduced Ras protein levels in tumor cells (Fukamachi et al., 2008). A similar study also showed that C3G could inhibit the growth of HS578T, a cell line from human breast carcinoma, in a dose dependent manner (Chen et al., 2005). Long et al. (2013) reported the proliferation androgen- dependent prostate cancer cell line, LNCaP, was inhibited by purple corn color through limiting the expression of Cyclin D1 and inhibiting the G1stage of the cell cycle. In their research, the TRAP male transgenic rats for adenocarcinoma of prostate consuming purple corn pigments had lower percentage of adenocarcinoma and a higher percentage of low-grade prostatic intraepithelial neoplasia, suggesting that purple corn color could retard prostate cancer progression (Long et al., 2013).

3.4.4.2 Liver carcinogenesis

An estimated 35,660 new cases (including intrahepatic bile duct cancers) and 23,000 deaths of liver cancer are expected to occur in the US during 2015 (American Cancer

Society, 2015). In the US, the liver cancer incidence rate in men was reported to be 3 times higher than it were in women, the cases in recent decades were attributed to diabetes and/or obesity and alcohol-related disorders (American Cancer Society, 2015). A

Japanese group investigated the chemopreventative effect of purple corn color on diethylnitrosamine-initiated hepatocarcinogenesis in F344 male rats, the increasing gene expression ratio for cyp 2c, 2e1, 3a1 and 3a2, as well as phospholipase C, delta 4 in rat serum suggested purple corn color might have modifying effects on mRNA markers concerning liver carcinogenesis development (Yokohira et al., 2008).

3.4.4.3 Colon cancer

Colorectal cancer is the third most common cancer in both men and women, it is estimated in 2015 around 93,090 cases of colon cancer will be diagnosed and 49,700 deaths from colorectal cancer to be occurred (American Cancer Society, 2015). The health benefits of consuming purple corn on slowing down colon cancer development has been studied by researchers internationally (Hagiwara et al., 2001; Jing et al., 2008;

Reddy et al., 2005; Zhao et al., 2009). In a male F344/DuCrj rats study, purple corn color

(5% in diet) showed significant reduction to colorectal carcinogenesis induced by 1,2- dimethylhydrazine (DMH) and 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine

(PhIP) (Hagiwara et al. 2001). Fewer incidences of macroscopic colon nodules and lower average aberrant crypt foci number were observed in rats fed with both purple corn color

66 and PhIP, suggested purple corn color exerted protection against PhIP promotion of colon tumor development without evidence of adverse effects (Hagiwara et al. 2001). In vitro cell studies afterwards also supported what was found in the in vivo study. Jing and coworkers (2008) measured the inhibition of human colorectal adenocarcinoma (HT29) cell line growth using different anthocyanins from purple corn, chockberry, bilberry, purple carrot, grape, radish and elderberry. All anthocyanins in their study suppressed the proliferation of HT29 cells, the GI50 for Peruvian purple corn anthocyanins was around

14 µg of C3G equiv/mL, much lower than pigments from other sources, suggesting purple corn pigments were potent anti-colon cancer agent (Jing et al., 2008). Similar study also showed Chinese purple corn had anti-proliferative capacity against HT29 cell in a dose dependence manner (Zhao et al., 2009). A study done with C3G from Prunus cerasus also showed a dose-dependent growth inhibition effect to HCT-116 colon tumor cells (Reddy et al., 2005). Although potently inhibit the proliferation of colon cancer derived HT29 cell lines (cell growth was inhibited around 50% after 48h exposure to 25

µg/mL chokeberry anthocyanins extract), study also showed anthocyanins from natural sources did not present inhibition to the growth of nontumorigenic colonic NCM460 cells

(Zhao et al., 2004).

3.4.4.4 Lung cancer

Lung cancer accounts for more deaths than any other cancer in both genders (American

Cancer Society, 2015). An estimated 158,040 deaths from lung cancer, accounting for about 27% of all cancer deaths, are expected to occur in 2015(American Cancer Society,

2015). C3G, the predominant anthocyanin in purple corn, has been shown capable

67 inhibiting the development of lung tumor cells in mice. This pigment purified from blackberry inhibited proliferation of a human lung carcinoma cell line A549 in vitro, reduced the size of A549 tumor xenograft growth and inhibited metastasis in vivo (Ding et al., 2006). The author also performed mechanistic study to indicate that C3G could inhibit migration and invasion of A549 tumor cells (Ding et al., 2006). Another study also demonstrated the inhibition of C3G on the the migration and invasion of Lewis lung carcinoma cells both in vitro and in C57BL/6 male mice (Chen et al., 2005). Moreover,

C3G from Prunus cerasus also showed a dose-dependent growth inhibition effect to NCl-

H460 lung tumor cells (Reddy et al., 2005).

3.4.4.5 Skin cancer

Skin cancer is the most commonly diagnosed cancer in the United States (American

Cancer Society, 2015). The number of skin cancer patients is difficult to estimate because some of them are not required to be reported to cancer registries (American Cancer

Society, 2015). One study of nonmelanoma skin cancers occurrence in the US estimated that in 2006, 3.5 million cases were diagnosed among 2.2 million people (American

Cancer Society, 2015). Pomegranate fruit extract (PFE), which is rich in pelargonidin-3- glucoside and cyanidin-3-glucoside, had been reported could modulate MAPK and NF-B pathways and inhibits against 12-O-tetradecanoylphorbol-13-acetate-induced skin tumorigenesis in CD-1 mice: the skin tumor incidence was reduced 70% at week 16 and tumor multiplicity was reduced 64% at week 30 for the PFE treated mice (Afaq et al.,

2005). From this study we may infer that purple corn, which is also rich in these two pigments, may have potential protective effect on skin cancer.

3.4.5 Anti-angiogenesis

Angiogenesis is a critical factor in tumor growth and metastasis, it is a process of forming new blood vessels from the existing vascular network (Huang et al., 2006).The anti- angiogenesis effect of anthocyanin- rich purple corn extract (PCE) under hyperglycemia condition has been investigated in human endothelial cells ex vivo and in db/db mice in vivo (Kang et al., 2013). In the in vitro study, PCE decreased endothelial expression of vascular endothelial growth factor (VEGF), which is one of the most potent angiogenesis activating agents (Kang et al., 2013; Wang and Stoner, 2008). Meanwhile, the hypoxia inducible factor (HIF)-1a, endothelial marker of platelet endothelial cell adhesion molecule-1 and integrin b3 induced by high glucose media were all attenuated by PCE

(Kang et al., 2013). In the in vivo study, the mice treated with 10mg/kg PCE for 8 weeks showed the glomerular angiogenesis in kidney was alleviated due to PCE attenuated the induction of VEGF and HIF-1a (Kang et al., 2013). Oral administration of PCE diminished the mesangial and endothelial induction of angiopoietin proteins and dampened the induction and activation of VEGF receptor 2 under hypeglycemic conditions (Kang et al., 2013). Thus the author concluded PCE may be served as a potent therapeutic agent for abnormal angiogenesis (Kang et al., 2013).

3.4.6 Amelioration of lifestyle diseases: obesity, diabetes, hyperglycemia and their associated diseases

Obesity is defined as the accumulation of excess adipose tissue resulting from various metabolic disorders (Tsuda et al., 2003). The obesity is a strong risk factor for type 2 diabetes, hyperglycemia, hypertension, hyperlipidemia as well as various cardiovascular

69 diseases (Kahn et al., 2006; Lew and Garfinkel 1979; Tsuda et al., 2003). Purple corn has been showed to have beneficial effect on ameliorating these lifestyle diseases.

A Japanese group did a series of experiments focusing on purple corn anthocyanins regulation of preventing metabolic syndrome (Tsuda, 2008). Purple corn color (PCC) has shown to prevent the obesity in male C57BL/6 mice feeding by 300g lard/kg high-fat

(HF) diet (Tsuda et al., 2003). The mice in the HF diet group had significantly higher body weight than those in control, PCC alone and HF+PCC groups after 5 weeks (Tsuda et al., 2003). The adipocytes in the adipose tissue of HF group mice were much bigger and looser than other groups, but hypertrophy of adipocytes did not occur in HF+PCC group of mice (Tsuda et al., 2003). Interestingly, this study was designed that the energy intake of the mice was not affected by PCC, thus the suppression of gaining body weight was not due to inhibition of dietary fat digestion and reduction of energy intake (Tsuda et al., 2003; Tsuda 2008). This study further examined the serum, liver and fetal lipids, serum glucose, insulin, leptin concentrations, hepatic and edididymal epididymal white adipose tissue (WAT) lipogenic enzyme and SREBP-1 mRNA and WAT TNF-α mRNA levels in the mice, and concluded that PCC down-regulated the mRNA levels of the enzymes included in the lipid synthesis accompanied by the reduction of the SREBP-1 mRNA level in the WAT (Tsuda et al., 2003). To work on the mechanism how the PCC function on preventing obesity, the Tsuda and Osawa (2004; 2005; 2006) group tested the impact of the major anthocyanin and anthocyanidins of purple corn, C3G and

Cyanidin ( Cy, the aglycon of C3G), on the adipocyte function. Research has been shown

C3G and Cy could increase the secretion of adipocytokines, a group of bioactive

70 molecules that involving in lipid regulation, as well as enhance the adipocyte specific gene (lipoprotein lipase, adipocyte fatty acid binding protein 2, uncoupling protein 2 and peroxisome proliferator-activated receptor γ) expression at least 2 folds in isolated rat adipocytes fed with 2g/kg anthocyanins diet (Tsuda et al., 2004). Then they further examined the detailed gene expression changes due to C3G and Cy treatment using microassay analysis (Tsuda et al., 2005). The isolated rat adipocytes were treated with

100µM C3G or Cy for 24 h showed 633 or 427 lipid metabolism and signal transduction- related genes up-regulated for more than 1.5 folds (Tsuda et al., 2005). Human adipocyte cells were also tested with 100µM C3G or Cy for 24 h to see if the gene expression enhancement would also happened in human cells (Tsuda et al., 2006). Their study demonstrated the significant up-regulation of adiponectin expression and downregulation of plasminogen activator inhibitor-1 and interleukin-6, as well as significant induction of some of lipid metabolism related genes in both the C3G and Cy treatment groups (Tsuda et al., 2006). All these results suggest purple corn anthocyanins have a unique therapeutic advantage responsible for the regulation of the adipocyte function in preventing lifestyle disease such as obesity and diabetes.

A Korean group has been worked on the inhibition of diabetes-associated disease using purple corn anthocyanins in cell lines and animals (Kang et al., 2013; Kang et al., 2012;

Li et al., 2012a; Li et al., 2012b). Purple corn anthocyanins were shown to dampen high- glucose-induced mesanginal fibrosis inflammation in vitro, and these pigments might play renoprotective role in diabetes nephropathy (Li et al., 2012a). This group also demonstrated that purple corn anthocyanins could inhibit the diabetes-associated

71 glomerular monocyte activation and macrophage infiltration as well as retard diabetes- associated glomerulosclerosis through disturbing the mesanginal IL-8-Tyk-STAT signal pathway in db/db mice (Kang et al., 2012; Li et al., 2012b). Moreover, their research showed anthocyanin-rich purple corn extract inhibited diabetes-associated glomerular angiogenesis by attenuating the induction of VEGF and HIF-1α (Kang et al., 2013). The studies of this Korean group suggested purple corn anthocyanins may be a potent therapeutic agent for the diabetes and hypertension related kidney failure (Kang et al.,

2013; Kang et al., 2012; Li et al., 2012a; Li et al., 2012b).

Additional research also provided experimental support to the statement that purple corn anthocyanins could help ameliorate diabetes, hyperglycemia and hypertension through inhibition of the enzymes which play critical roles in carbohydrate and lipid metabolism

(Kim et al., 2013; Ranilla et al., 2009). In an in vitro study, purple corn has been shown to have potent α-glucosidase inhibitory capability in a dose dependence manner due to its richness in protocatechuic acid, suggesting their potential antihypertension, antidiabetes activity (Ranilla et al., 2009). Hirsutrin, one of the phenolic compounds from purple corn, presented potent inhibition to aldose reductase activity and galactitol formation in rat lens and erythrocytes under high galactose osmotic stress (Kim et al., 2013). Moreover, purple corn anthocyanins not only enhance the insulin secretion in hamster pancreatic beta cells

(HIT-T15) and lowered blood glucose level in db/db mice, but also showed efficient protection activity of pancreatic beta cell from cell death in HIT-T15 cell culture and db/db mice (Hong et al., 2013). The protection effect refers to purple corn anthocyanins did not bring negative effect to the pancreatic insulin secreting cells while the commonly

72 used diabetes drug, sulfonylurea, had been reported to accelerate gradual deterioration of beta cells function and inducing cell death (Maedler et al., 2005; Del Guerra et al., 2005;

Hong et al., 2013). A recent study demonstrated purple corn having preventive effect on diabetic cataract by incubated enucleated rat lenses in 55mM glucose with different level of purple corn extracts (Thiraphatthanavong et al., 2014). The results showed purple corn extract could help decreasing lens opacity via lowering oxidative stress or suppression aldose reductase, the rate limiting enzyme in polyol pathway (Thiraphatthanavong et al.,

2014). Purple rice extract, which is similar to purple corn extract in the aspect of being abundant in C3G, was recently reported to present suppression of blood glucose increase in healthy male Wistar rats in humans (Shimoda et al., 2015). In their human study, 5 males and 3 females were given a 200g rice ball containing placebo after 11 hours of fasting, their fingertips blood samples were collected and the blood glucoside level at 0,

30, 60, 90 and 120 min were measured (Shimoda et al., 2015). The same subjects were given was a 200g rice ball containing 25mg purple rice extract one week later (Shimoda et al., 2015). Significant suppression of blood glucose was observed in the purple rice treated group at 30 and 90 min compared with that in the control group (Shimoda et al.,

2015). They also studied the carbohydrate absorption suppression mechanism of purple rice extract in vitro using three carbohydrate digestive enzymes, their results confirmed the suppression was achieve by the inhibitory effects of purple rice extract on α- glucosidase, α-amylase and aldose reductase activities (Shimoda et al., 2015). It is possible the similar phenolic profile purple corn extract may also suppress the acute increase in blood glucose after meals as the purple rice extract.

3.4.7 Blood pressure regulation

The effect of purple corn on blood pressure regulation has been studied in both animals and humans. Intragastric administration of purple corn color (PCC) at a dose of 7.4 mg/kg as anthocyanin to hypertensive rats twice daily for 5 weeks, the finding showed that PCC significantly inhibited the increase in systolic blood pressure after 8 days

(Toyoshi and Kohda 2004). Another study fed the hypertensive rats with purple corn anthocyanins color for 15 weeks and their results also presented the systolic blood pressures of the PCC fed rats were significantly lower than the control groups, suggesting that purple corn pigments had anti-hypertensive effect on the hypertensive animals

(Shindo et al., 2007). A small scale (30 individuals) research investigated the benefits of purple corn extract capsule on the blood pressure of mild-to-moderate hypertension

Peruvians (Finkel et al., 2013). This study showed that taking a purple corn extract capsule for a short period of time (3 weeks) had a beneficial effect on reducing systolic and diastolic blood pressure from139/88 to 132/81 mm Hg for the early stage hypertension patients, regardless of age, gender, body mass index level, or initial average blood pressure reading (Finkel et al., 2013). Although the mechanism how purple corn regulate blood pressure in hypertension individual is still not clear, the beneficial effect of purple corn suggests a great potential of application as food supplements and medication.

3.4.8 Heart health

Anthocyanins have been reported to have beneficial effect on cardiovascular health (He and Giusti, 2010; Wallace and Giusti, 2014). Early study believed the beneficial property of phenolic compounds might due to their capability of inhibiting hexanal formation in in

74 human low-density lipoproteins and inhibiting hydroperoxide formation in lecithin liposomes (Heinonen et al., 1998). A study fed the spontaneously hypertensive rats with purple corn anthocyanins color for 15 weeks, the heart rate of anthocyanin-administered rats (about 365 beats/min in average) were significantly lower than the control rats

(about 385 beats/min in average) (Shindo et al., 2007). In addition, other study showed both heart weight and abdominal fat of purple corn anthocyanins fed broiler were significantly lower up to 0.55±0.01% body weight and 1.47±0.03% body weight than the regular yellow corn fed group, suggesting there might be potential heart health benefit of purple corn (Amnueysit et al. 2010).

3.5 PURPLE CORN TOXICOLOGY

The toxicity of purple corn pigment-rich extracts has been evaluated by different research teams, with two major studies conducted in China and Japan. The Chinese research team led by Zhou ran an acute oral toxicity test on Wistar rats by dissolving purple corn pigment powders in distilled water, followed by gavaging the pigment solutions into rats up to the level of 21.5g/(kg bw) (Zhou et al., 2007). The purple corn pigment powders were prepared by spray drying the extracts obtained from soaking the plants into 50-60°C

60-70% acidic ethanol (Zhou et al., 2007). None of the rats showed observable toxic symptoms nor died during the study and the author concluded that the oral LD50 for purple corn pigment-rich extract in rats was greater than 21.5 g/kg bw (Zhou et al., 2007).

In addition, the purple corn extracts presented no significant difference (p>0.05) to the negative control groups in the mice marrow micronucleus test and sperm aberration test

(Zhou et al., 2007). According to the Chinese toxicity grading standards, the tested purple

75 corn pigments were considered no harm, with great potential to be developed as food additive and healthy food materials (Zhou et al. 2007). Another research team in Japan ran a 90-days subchronic oral toxicity study of purple corn color (containing 26.4% of anthocyanins, 57.7% of polyphenol and glucose and 10% of citric acid) using F344 rats.

The no-observed-adverse-effect-level (NOEL) was determined to be 5.0% purple corn color in diet for both sexes (male: 3542 mg/kg bw/day, female: 3849 mg/kg bw/day) under their experimental conditions (Nabae et al., 2008).

3.6 RECOMMENDED DOSAGE

The dosages needed to obtain some type of health benefit from purple corn powders in a variety of studies was mostly in the level of milligrams of purple corn anthocyanins / kg body weight. The dosages of purple corn applied in health benefits study involving animals and humans are shown in Table 3.1. Most health benefits studies conducted with purple corn in mice used dosage of purple corn anthocyanins around 10 mg/kg bw (Hong et al.,, 2013; Kang et al.,, 2013; Kang et al.,, 2012; Li et al.,, 2012a). Studies conducted in rats used dosages of purple corn color ranging from 0.2% to 1.0% of the total diet (Tsuda et al., 2003; Long et al., 2013), or 7.4 mg/(kg bw) anthocyanins equivalent (Yokohira et al., 2008). In a human study, slightly hypertensive patients took capsule of 300mg purple corn extract containing 6% total anthocyanins (18 mg), 15% phenolic compounds (45 mg), 76.6% carbohydrates and negligible amount of fats and proteins on a daily basis and no adverse reaction was reported (Finkel et al., 2013).

To date, there is no accurate recommended dosage of purple corn anthocyanins consumption for human health benefits. Even though there is compounding evidence that

76 strongly suggest that these compounds may indeed be beneficial to human health, additional studies need to be conducted to confirm these results. Anthocyanins consumption in United States was estimated to be about 12.5mg/day/person although it had been previously reported that anthocyanins consumption could be as high as 200 mg/day/person in earlier studies (Wu et al., 2006). However, the dose 0.01% purple corn color (PCC) in daily diet used in Yokohira et al., (2008) experiment, which approximately corresponded to typical average human daily anthocyanins intake, did not showed significant beneficial effect according to their study. The level of PCC required to achieve beneficial impact was 1% PCC in diet, which corresponded to 168.5 mg/kg bodyweight/d in humans based on FAO/WHO Joint Expert Committee on Food

Additives (Yokohira et al., 2008). Taking these numbers into account, considering the clinical trials on the blood pressure control study using purple corn anthocyanins with consumption of 18 mg/day/person, it could be reasonable to speculate that the typical average levels of purple corn anthocyanin daily consumption would do no harm with additional potential for a beneficial effect (Finkel et al., 2013). However, more research is still needed to be able to make a consumption dosage recommendation.

Function Target disease Dose tested in study Purple corn source Subjects (Ref.) and/or organ showing health benefits Japanese purple corn color Anticancer Cancer; F344/DuCrj rats, (Hagiwara et al., (PCC) powders, 5% PCC in daily diet Colon 6-week old, male 2001) 21.5% anthocyanins Mexican purple corn, processed Anticarcinogenic Cancer; to tortilla powers (PCT), Sprague– Dawley rats, (Reynoso-Camacho 27 % PCT/BW (w/w) et al., 2015) Colon 13.03±0.55 mg/100g ACN, 4-5 weeks, male 138.14±2.95 mg/100g phenolics TRAP rats with a Japanese purple corn color 0.1% and 1% PCC in diet Cancer; Sprague–Dawley genetic Anticarcinogenesis (PCC) powders, (25 and 267 mg PCC⁄rat/d) (Long et al., 2013) Prostate background, 20.9% anthocyanins

6-week old, male 78

Japanese purple corn color c-Ha-ras transgenic and Anticarcinogenesis Cancer; (Fukamachi et al., (PCC) powders, non-transgenic rats, 1% PCC in daily diet Mammary 2008) 33.7% anthocyanins female 1% PCC in daily diet Anticarcinogenesis Japanese purple corn color Cancer; F344 rats, (correspond to 168.5 mg Antioxidant (PCC) powders, (Yokohira et al., Liver 4-week old, male anthoyanins/kg BW/d in 33.7% anthocyanins 2008) humans.) Chinese maize purple plant Cancer; Wistar rats, Antioxidant pigments (MPPP), 1% PCC in daily diet (Zhang et al., 2014) Liver and kidney male and female 57.95% anthocyanins Diabetes db/db mice and their non- Anti-inflammatory Korean purple corn kernel (Li et al., 2012a, glomerulosclerosis; diabetic db/m littermates, 10 mg PCE /kg BW daily phenolic-rich extract (PCE) Kang et al., 2012) Kidney Adult, male Continued

Table 3.1 The dosage of anthocyanins and phenolics applied70 in purple corn animals and humans health benefits studies.

Table 3.1 continued

Function Target disease Dose tested in study Purple corn source Subjects (Ref.) and/or organ showing health benefits db/db mice and their Diabetic Korean purple corn kernel non-diabetic db/m 10 mg PCE/kg BW Anti-angiogenesis nephropathy; (Kang et al., 2013) anthocyanin-rich extract (PCE) littermates, daily Kidney 8-week old, male Diabetes; Peruvian purple corn Antidiabetic C57BL/KsJ db/db mice, pancreatic beta anthocyanin powders (PCA), 10 mg PCA /kg BW daily Antihyperglycemic 6-week old, male (Hong et al., 2013) cell 10% anthocyanins Blood pressure Japanese Purple corn color regulation spontaneously 79 Hypertension (PCC), 56.7 mg PCC/kg BW/day (Toyoshi and Kohda, hypertensive rats 2004) 26% anthocyanin Blood pressure and Japanese Purple corn color spontaneously heart rate regulation Hypertension (PCC), hypertensive rats, 1% PCC in diet (Shindo et al., 2007) 26% anthocyanins 3-week old, male Peruvian purple corn extract Blood pressure (PCE), process to capsule Phase-I hypertension regulation Hypertension which contained 300mg PCE, humans, 21-70 years 1 capsule (Finkel et al., 2013) 18mg anthocyanin, old, male and female 45mg phenolics, Cy and C3G standard, Wistar rats, 2 g C3G /kg diet, rats had Obesity prevention Obesity (Tsuda et al., 2004) purity>99% 7-week old, male free access to the diet C3G purified from purple corn, Wistar rats, Obesity prevention Obesity 0.2% C3G in daily diet (Tsuda et al., 2003) purity>95% 8-week old, male

3.7 CONCLUSION

Studies focused on purple corn functional properties have become very popular in recent decades. Although the use of purple corn could be traced back to hundreds of years ago in South America, it is nowadays a rising star in the worldwide ingredients market due to its deep color and potential health promoting impacts. Up to now, more than 20 bioactive phenolic compounds including phenolic acids, anthocyanins and other flavonoids have been reported to be found in purple corn (Zea mays L.). These phenolic compounds, as presented in this review, have been shown in numerous in vivo and in vitro studies to have potent antioxidant, anti-inflammatory, anti-mutagenic, anti-carcinogenic, anticancer, anti- angiogenesis properties and to be capable to ameliorate lifestyle diseases such as obesity, diabetes, hyperglycemia, hypertension and cardiovascular diseases.

These beneficial properties are in large proportions attributed by the strong antioxidant power of the purple corn phenolics as well as to their anti-inflammatory properties. These health beneficial properties still remain at a satisfactory high level even undergoing food industry process. Research has shown that the no-observed-adverse-effect-level (NOEL) of purple corn color was at least 5.0% of the rat diet (~3.5g/kg bw/day), while for hypertension patients, daily consumption of only 300mg purple corn extracts (contain

18mg anthocyanins) for 3 weeks could present beneficial blood pressure reducing effect.

However, most currently available studies on purple corn health benefits were conducted with anthocyanins or phenolic compounds mixtures. Studies working on individual anthocyanin or other phenolic compound are still in highly demand to reveal the potential contribution achieving by individual components. The most powerful functional phenolic

80 compound in purple corn and its potential synergetic effect with other chemicals are still unknown. Since only a small amount of purple corn functional compound mixture may achieve beneficial effects, this promising rich source of health promoting compounds deserves additional attention to clarify its mechanism of action and to better understand its potential contributions to human health.

3.8 ACKNOWLEDGMENTS

We appreciate Alicorp S.A.A. (Lima, Peru) and Chinese Scholarship Council Fellowship for their financial support.

Chapter 4: The Effect of Solvent and Acidity Selection on Purple Corn (Zea mays L.) Cob Pigments and Phenolic Compounds Extraction

4.1 ABSTRACT

Purple corn cob (PCC) is an economic source of anthocyanins that could serve as food colorant. Maximizing pigment recovery efficiency while optimizing quality is critical for the success of PCC for food applications. Due to toxicity concerns, acidified aqueous organic solvents are not desirable for food applications though they produce high anthocyanin recovery from PCC. In this study a consumer-friendly solvent (aqueous ethanol) was used at different ethanol ratios (0-100%) to determine PCC anthocyanin recovery as compared to methanol and 70% (v/v) aqueous acetone. Different acidity levels (0-2% HCl, v/v) were also added into two extraction matrices, water and 70% aqueous acetone, to investigate if the PCC pigment extraction yield could be improved by adjusting solvent acidity.

The PCC extract obtained by water-ethanol ratio around 1:1 (40%, 50%, 60% ethanol) achieved highest yield of monomeric anthocyanins (more than 13.5 mg/g FW) with low polymeric color, comparable to the lab used extraction solvent 70% acetone (14.3 mg/g

FW). The amount of PCC monomeric anthocyanin and phenolics recovery extracted with

82 different water-ethanol combination could be predicted using a quadratic model. The solvent acidity seemed to play a role in water PCC extraction but not much in extraction with 70% acetone.

In summary, extraction performed by 0.01% (v/v) 6N HCl acidified 50% aqueous ethanol for 45 min, with double cake washing could produce high levels of monomeric anthocyanins and phenolics with relatively low polymeric color. This extract could be a potential candidate for food companies interesting in applying PCC pigments as synthetic red dye alternatives.

Keywords: water-ethanol ratio, anthocyanins, extraction, polymeric color.

4.2 INTRODUCTION

With increasing concerns on potential adverse effects of synthetic dyes (Lofstedt, 2013), pigments derived from natural sources are becoming increasingly market desirable.

Anthocyanins are a class of phenolic compounds that provide red, orange, purple and blue color to various fruits, vegetables, flowers, and cereals. Easy to corporate into aqueous food system, anthocyanins can be an important alternative to synthetic dyes for food coloring purpose. Additionally, anthocyanins from plant sources have been shown to have health-promoting effects such as antioxidant (Yokohira et al., 2008; Zhang et al.,

2014), anti-inflammatory (Reddy et al., 2005; Serra et al., 2013; Tsuda et al., 2005), and anti-carcinogenic (Fukamachi et al., 2008; Jing et al., 2008; Urias-Lugo et al., 2015).

Anthocyanins have been recognized by the European Union and Japan, with the code E-

163 to be used as food colorant.

Anthocyanins for food application are usually directly extracted from plant tissues.

Extraction of anthocyanins is often conducted by grinding, drying or lyophilising fruits, or soaking fresh materials into solvents (Barnes et al., 2009). The most widely used solvents are water, ethanol, methanol, acetone, and others. Organic solvents and their aqueous mixtures are preferred since they could disrupt cell membranes and dissolve the pigments at the same time (Mateus and de Freitas, 2009). To facilitate the extraction process and stabilize pigments during the extraction, the solvents are usually acidified.

Selection of acids could have an impact on anthocyanins recovery. In a wild blueberry study, extraction was done using the same levels (0.02%, v/v) of hydrochloric, citric, tartaric, lactic, and phosphoric acid. Phosphoric acid yielded the most pigments from the frozen berries (Nicoué et al., 2007). Besides the extraction solvent and selection of acid, other factors such as extraction temperature, time, solid:liquid ratio, assistance with microwave, ultrasound, sonication, can also affect the pigments extraction from the plant material (Castañeda-Ovando et al., 2009; Fernando Ramos-Escudero et al., 2012; Yang and Zhai, 2010; Yang et al., 2009).

Purple corn (Zea mays L.), originated from the Andean region, Latin America, is a rich source of anthocyanin. The anthocyanin content in purple corn ranges from 6.8mg/g FW to 82.3mg/g FW depending on the sections, which is much higher compared to other anthocyanin-rich fruits such as blueberries, with anthocyanin content only 1.3-3.8 mg/g

FW (Cevallos-Casals and Cisneros-Zevallos, 2003; Li et al., 2008; Wu et al., 2006). The dark-colored inedible cob area of purple corn is an inexpensive starting material for food colorant production. Optimized combinations of extraction time, temperature, solid:liquid

84 ratio, water-methanol mixtures for anthocyanins extraction from purple corn have been well studied (Jing and Giusti, 2007; Fernando Ramos-Escudero et al., 2012; Yang et al.,

2009). However, extraction efficiency of purple corn cob (PCC) anthocyanins using food-friendly water-ethanol combinations, and how they compared with other organic solvents have not yet been systematically investigated. In addition, although the impact of different pH values on plant anthocyanin extraction have been studied by adding the same amount of various organic acids into the extraction matrix, the effect of adding different levels of identical acid on PCC anthocyanins recovery has not yet been elucidated.

The objective of this study was to increase PCC anthocyanins extraction efficiency through optimizing the water-ethanol ratio in the extraction solvent and extraction matrix acidity. The hypothesis was that the different polarities of various water-ethanol combinations and acidities would achieve different pigment recoveries. The optimal polarity and acidity for the pigment might be different from those for other compounds.

We could have a chance to obtain high anthocyanin product by optimizing solvent polarity and acidity.

4.3 MATERIALS AND METHODS

4.3.1 Materials and Reagents

Dried purple corn cob (Zea mays L.) particles were kindly provided by Agroindustrial

S.A.C (Lima, Peru) and Globenatural International S.A. (Chorrillos-Lima, Peru). Gallic acid was acquired from MP Biomedicals, LLC (Aurora, OH). Folin and Ciocalteau’s

85 phenol reagent were purchased from Sigma-Aldrich (St. Louis, MO). All solvents and other chemicals were purchased from Fisher Scientific (Fair Lawn, NJ, USA).

4.3.2 Standardization of Extraction Procedure

Before extraction, the dried purple corn cob (PCC) particles were blended into powders using a coffee blender to facilitate the extraction. Extractions were done at room temperature (70-75°F) on a magnet stir plate with stir rate of 100 rpm. Briefly, a 1 gram sample of weighted PCC powders was placed into a beaker and 50 mL of extracting solvent was added. After sufficient amount of time soaking (~45 min), the colored solvent was filtered through Whatman No.4 filter paper using a Buchner funnel. The cake was re-extracted several (2) times using 25 mL of the same solvent each time until a faintly colored filtrate was obtained. For the 70% (v/v) acetone extracts, two volumes of chloroform were added after the filtrate was transferred to a separatory funnel, and the colored portion was collected after storing at 4°C overnight (Rodriguez-Saona and

Wrolstad, 2001). All extracts with organic solvents were sent to a rotary evaporator to evaporate the organic solvent at 40°C under vacuum. Finally the remaining aqueous extracts were made up to known volume with 0.01% HCl acidified water.

4.3.2.1 Powders soaking time determination

The soaking time required to extract a sufficient amount (>95%) of anthocyanins was determined using 5 gram samples of PCC powder with 250 mL of 0.01% (v/v) 6N HCl acidified water. Water was selected because all other organic solvents used in this study have better penetration capability into plant tissue than water. It was expected the time required for the water to recover sufficient amount of pigments would be the longest.

Every 3 min after the extraction was initiated, 0.5 mL of the extraction slurry was sampled to analyze its monomeric anthocyanins and total phenolics concentrations. This sampling process was done for slightly over an hour. The remaining slurry was continueally extraction for about another 1 hour before determining its monomeric anthocyanins and total phenolics concentrations. The final monomeric anthocyanins and total phenolics concentrations were considered as “fully extracted”. The percentage of recovered pigment and phenolics at each sampling time point was calculated using the following equation:

푐표푛푐. 푎푡 푠푎푚푝푙𝑖푛푔 푡𝑖푚푒 푝표𝑖푛푡 %Recovery = × 100% 푐표푛푐. 푎푓푡푒푟 푓푢푙푙푦 푒푥푡푟푎푐푡푒푑

The percent extracted monomeric anthocyanins and total phenolics (%recovery) were plotted versus their sampling time point. The time required to extract more than 95% of monomeric anthocyanins was selected as the minimum soaking time in the following extraction experiments.

4.3.2.2 Cake washing times determination

After soaking, the extracted pigment solution was filtered through the funnel and the remaining colored insoluble powders formed a cake on the filter paper. Cake washing may help further recover some more extractable pigments, but it would be a large waste of extraction solvent if the washing recovered pigments were not concentrated enough.

Again, due to the relatively low penetration capability of water, in this part of study, the number of cake washing times was determined using 0.01% (v/v) 6N HCl acidified water.

Around 1 gram of PCC powders was soaked in 50 mL of acidified water for about 45 min. The colored solution was filtered through the funnel with vacuum. The ~50 mL filtrate was considered as “0” cake washing. Then the vacuum was turned off and 25 mL of acidified water was applied for the cake washing. This cake washing filtrate was considered as solution after 1st cake washing. The same amount (25 mL) of acidified water was used in the following washing steps until a very pale color filtrate was obtained. The monomeric anthocyanin content of the filtrates after each cake washing was measured. The percentage of recovered monomeric anthocyanins after each cake washing was determined and plotted versus the washing time. Cake washing numbers were determined when more than 95% of extractable pigments were recovered. This cake washing number would be applied in the following extraction study even using different solvents.

4.3.3 Effect of Solvent Selection

The effects of solvent selection on PCC pigments recovery were investigated using water,

3 different organic solvents and their aqueous mixtures: 70% (v/v) aqueous acetone, water, methanol, ethanol, 20% (v/v) aqueous ethanol, 40% (v/v) aqueous ethanol, 50%

(v/v) aqueous ethanol, 60% (v/v) aqueous ethanol and 80% (v/v) aqueous ethanol. All solvents were acidified with 0.01% 6N HCl (v/v). PCC particles used in this part of study were provided by Globenatural International S.A. (Chorrillos-Lima, Peru). The extraction procedure was described previously. Extraction of each solvent was done in triplicate.

The obtained extracts were analyzed for their monomeric and polymeric anthocyanins, total phenolics, and anthocyanins HPLC profiles.

4.3.4 Effect of Acidity

The effect of acidity on PCC pigments recovery were investigated in this part using 70% aqueous acetone and water. Acidity levels used in this experiment were 0%, 0.01%,

0.1%, 1%, 2% (v/v) 6N HCl. The extraction procedure was the same as described above.

PCC particles used for this part of study were provided by Agroindustrial S.A.C (Lima,

Peru). The obtained extracts were analyzed for their monomeric and polymeric anthocyanins, total phenolics, and anthocyanins HPLC profiles.

4.3.5 Monomeric Anthocyanins

Total monomeric anthocyanins of each extract were determined by pH differential method and the result was reported as mg of cyanidin-3-glucoside equivalent per gram of dried powders or mg of cyanidin-3-glucoside equivalent per mL of solution. Absorbance values of pH1.0 and pH4.5 buffer diluted sample were measured at 520 nm and 700 nm by Shimadzu UV-2450 UV-visible spectrophotometer (Shimadzu Corporation, Tokyo,

Japan). Monomeric anthocyanins present vivid red color at pH1.0 while colorless at pH4.5. The monomeric anthocyanin content was calculated based on the color difference between the two buffer systems using the molecular weight of cyaniding-3-glucoside

449.2 and its molecular absorptivity 26900 in aqueous buffer (Giusti and Wrolstad,

2001). Measurement was done in triplicate.

4.3.6 Polymeric Color

Different from monomeric anthocyanins, polymeric anthocyanins still present red color at pH4.5. Bisulfite solution was used to bleach the monomeric anthocyanin while the polymerized colored anthocyanin-tannin complexes were resistant to bleaching. Briefly,

89 absorbance of the bisulfite-bleached and control (distilled water diluted) samples were measured at 420 nm, 520 nm and 700 nm by spectrophotometer. The percentage of polymeric color was calculated using bleached sample color divided by the control sample color (Giusti and Wrolstad, 2001). Each sample was measured three times.

4.3.7 Total Phenolics

Total phenolics of each extract were determined by Folin-Ciocalteu method (Ainsworth and Gillespie, 2007; Waterhouse, 2003). Briefly, 20 µL of PCC extract, or gallic acid standard, or water blank were added into cuvette. Following by 1.58 mL of distilled water and 100 µL of Folin & Ciocalteu reagent, the mixture was mixed thoroughly by pipetting before incubating under dark at room temperature for 5 min. Finally 300 µL of saturated

Na2CO3 was added, the solution was incubated 2h under dark at room temperature before measuring absorbance at 765 nm using spectrophotometer. Measurement was done in triplicate. Total phenolics content was calculated as gallic acid equivalents based on gallic acid standard curve.

4.3.8 Anthocyanins HPLC profile

The anthocyanins profile of each PCC extract was purified by a C18 Sep-Pak cartridge

(Waters Corporation, Milford, MA, USA) followed the procedure of Rodriguez-Saona and Wrolstad (2001) and filtered through 0.45 μm polypropylene syringe filter

(Phenomenex, Torrance, CA, USA) before injecting into the vial for HPLC analysis.

A reverse-phase high performance liquid chromatograph (HPLC) system (Shimadzu

Corporation, Tokyo, Japan) consisting of LC-20AD prominence liquid chromatograph, equipped with a SIL-20AC prominence auto sampler at 4°C, a SPD-M20A prominence 90 diode array detector, with LCMS solution Ver3.30 software was applied for this analysis.

The column was reverse-phase 3.5 µm Symmetry C18 column (4.6×150mm, Waters

Corp., MA, USA) fitted with a 4.6×22 mm Symmetry 2 micro guard column (Waters

Corporation, MA, USA). The solvents used were A: 4.5% (v/v) formic acid in water, and

B: 100% acetonitrile. Solvents were filtered through 0.45µm poly(tetrafluorothylene) membrane filters (Pall Life Sciences, Ann Arbor, MI, USA). Separation was achieved by using a linear gradient from 5 to 25% solvent B in first 25 min, keeping 25% B from 25 to 30 min, followed by linear increasing B from 25% to 40% during 30 to 35 min. The injection volume varied from 50 to 150µL depended on the pigment concentration, with a

0.8 mL/min flow rate. The purple corn anthocyanins chromatogram was collected at 520 nm and all analysis was repeated three times for each sample.

4.3.9 Statistical Analysis

One-way ANOVA test analysis together with Tukey methods at α= 0.05 was applied to evaluate the mean differences on total phenolics, monomeric and polymeric anthocyanins, anthocyanins HPLC profiles, among different extracting solvents and acidities, p<0.05 was considered to be statistically significant.. Correlation between anthocyanins recovery and % ethanol was determined through general linear model.

Minitab 16.0 software (Minitab Inc. State College, PA) was used to analyze data in this study.

4.4 RESULTS AND DISCUSSION

4.4.1 Standardization of Extraction Procedure

Standardization of the extraction procedure, such as extraction soaking time and cake washing number is critical for this study. Since the extraction efficiency of PCC anthocyanins could be affected by various other factors, to better evaluate the effects of solvent selection and acidity, minimizing the bias coming from other aspects would help lead to a more credible conclusion.

4.4.1.1 Powders soaking time determination

PCC anthocyanins and other phenolic compounds dissolved into acidified water rapidly in the beginning stage of extraction. The extracted monomeric anthocyanins and total phenolics increased in an exponential manner in the first 6 min as shown in Figure 4.1.

The increase rate of % extracted anthocyanins and phenolics dramatically slowed down after 9 min of extraction. More than 90% of total monomeric anthocyanins were extracted into the solvent within half an hour, while it took phenolics more than 45 min to reach the same recovery level (Figure 4.1), suggesting PCC anthocyanins have better acidified water affinity than other phenolic compounds. The observation that anthocyanins coming out faster than other phenolic compounds was in agreement with previous study in our lab, working with PCC extraction time using acidified 70% aqueous acetone (Jing and

Giusti, 2007). They reported the total monomeric anthocyanins contents in PCC extracted by aqueous acetone had no statistical difference between the extraction time of 20 min and 60 min, while the total phenolics content for 60 min extraction was significantly higher than 20 min extraction (Jing and Giusti, 2007). More than 95% of monomeric

92 anthocyanins were extracted after 40 min of soaking, and the monomeric anthocyanins level in the solution became quite stable afterwards (Figure 4.1). However, the trend in

Figure 4.1 suggested even after an hour the amount of total phenolics would continue increasing, even though at 69 min around 97% of total water extractable phenolic compounds had already been transported into the solvent.

100%

80%

60% anthocyanins

40% phenolics %Recovery

20%

0% 0 20 40 60 80 100 120 Soaking time (min)

Figure 4.1 Extraction curve of percent purple corn cob anthocyanins and phenolics soaked in 0.01% 6N HCl acidified water.

Most previous purple corn anthocyanin extraction studies reported extraction times above an hour (Montilla et al., 2011; Pedreschi, 2005; Fernando Ramos-Escudero et al., 2012;

Yang et al., 2009), even when assisted with organic solvent and/or heated conditions. Our study showed that ~45 min recovered >95% of PCC anthocyanins at room temperature without extra extraction aids such as ultrasound or microwave. However, soaking time of at least an hour would still be recommended if working on quantification or targeting at maximizing total phenolic compounds recovery from PCC.

4.4.1.2 Cake washing times determination

In order to extract as much pigments as possible, most laboratory scale studies tended to wash the cake several times until a colorless filtrate was obtained. It worked well for quantification purpose but would be a waste of labor, time and solvent from the industry pigments production point of view. As shown in Figure 4.2, as the cake washing numbers increased, the amount of recovered pigments in each washing step decreased dramatically. Over 90% of the PCC monomeric anthocyanins were extracted after one cake washing. Similar results were also reported in PCC pigments cake re-extraction study using acidified aqueous acetone (Jing and Giusti, 2007). Their research showed that an extraction time of 20 min and 60 min plus 1 re-extraction generated 88.36% and

87.64% total monomeric anthocyanins. Additional re-extractions did not contribute much to increase the final total monomeric anthocyanins and phenolics (Jing and Giusti, 2007).

In our study, the 4th cake washing step only extracted 0.69% of total extractable PCC monomeric anthocyanins, which seemed very labor, solvent and time inefficient (Figure

4.2). To better mimic the industrial pigment processing condition, and taking cost factor

94 into consideration, double cake washing, which recovered almost 97% of total extractable pigments, was applied in the following PCC pigments experiments.

90% 75.25% 75%

60%

45%

30%

17.00% % Extracted anthocyanins Extracted % 15% 3.02% 1.63% 0.69% 0% 0 1 2 3 4 Cake washing times

Figure 4.2 Percent extracted purple corn cob anthocyanins after each cake washing, extraction and cake washing were done by 0.01% 6N HCl water.

4.4.2 Effect of Solvent Selection

We would expect that a good solvent for PCC pigments extraction to be high in monomeric anthocyanins and total phenolics recovery while low in polymeric color. High monomeric anthocyanins and total phenolics suggest good penetration capability of the solvent into the plant tissue as well as good affinity/solubility of the solvent to the target

95 compounds. The low polymeric color is an indicator of high anthocyanins quality in the extract, since the fresher or less processed/degraded anthocyanin extracts usually have lower polymeric color (Brownmiller et al., 2008; Choi et al., 2002; Gao et al., 1997;

Hager et al., 2008a; Hager et al., 2008b).

Overall, aqueous organic solvent mixtures worked better in PCC anthocyanins and phenolics extraction compared to water and pure organic solvents (Table 4.1). This was in agreement with most crop anthocyanin and phenolic compounds studies. The mostly used extraction solvents were aqueous organic solvent mixture such as 60%-80% ethanol

(De Pascual-Teresa et al., 2002; Pedreschi and Cisneros-Zevallos, 2007; Zhao et al.,

2008), and 50-70% acetone (Jing and Giusti, 2007; Moore et al., 2005; Žilić et al., 2012).

Among all the solvents applied in this study, 50% ethanol was the most efficient solvent with high recovery of 14.3±0.6mg/g FW of monomeric anthocyanins and 49.8±0.7mg/g

FW phenolics. Compared to the widely used phenolic extraction solvent, acidified 70% aqueous acetone, the aqueous ethanol mixed in a ratio around 1:1 (40%, 50% and 60%) showed no statistical significance in PCC monomeric and polymeric anthocyanins recovery, but the 50% ethanol presented significantly higher (p<0.05) capability in total phenolics recovery than all of the other three solvents. Pure ethanol was the least favorable extraction solvent for PCC pigment extraction in our experimental condition. It extracted the lowest amount of monomeric anthocyanins (3.4±0.4mg/g FW) and total phenolics (7.0±0.6mg/g FW) from the plant material, and its polymeric color

(41.5±3.5%) in the extract was significantly (p<0.05) higher than all others (Table 4.1).

Considering the quality of anthocyanins, the aqueous organic solvent mixtures tended to

96 produce lower level of polymeric color, ranging between 20% and 25% (Table 4.1).

Although the pure methanol extract produced lowest polymeric color of 20.1±0.6%, no statistical significance at α=0.05 level was found between the polymeric color level of methanol extracts and other aqueous organic solvent mixtures. Selection of solvent does not play a very critical role in the PCC anthocyanins profile, the proportion of acylated pigments in different extraction solvents did not vary much much (Table 4.1). In fact, the

PCC anthocyanins HPLC chromatogram from different solvents could overlap each other well, indicating solvents applied in this study did not change the PCC anthocyanin composition during the extraction process.

Targeting at food application, water and ethanol are preferred solvent options since they are less toxic compared to other organic solvent combinations (Escribano-Bailón et al.,

2004). It has been shown in Table 4.1 that the water-ethanol mixtures could serve as good extraction solvents for PCC color recovery. Interested in the recovery efficiency of other water-ethanol ratio mixtures, As shown in Figure 4.3, predictive models for PCC anthocyanins and phenolics water-ethanol extraction yield were generated through fitting the data in Table 4.1. Both predictive models were in quadratic pattern, the highest PCC anthocyanins recovery was predicted to extract with 48.2% (v/v) ethanol, and the highest

PCC phenolics recovery was predicted to extract with 46.7% (v/v) ethanol (Figure 4.3).

Ramos-Escudero et al. (2012) also evaluated the impact of different water-alcohol combinations on purple corn extraction solvents, but using methanol, though their solvent acidity level (1% 1N HCl) was slightly higher than ours. Very different from what we got there, their highest total phenolics recovery was came from 80% methanol, and highest

97 anthocyanins recovery was produced by 100% methanol (Fernando Ramos-Escudero et al., 2012). Since ethanol has higher polarity than methanol, it was expected that when mixing with the least polar water, it would require less ethanol than methanol to reach the

PCC phenolics-favored polarity.

Monomeric Total Extracting % Polymeric % Acylated anthocyanins phenolics solvent anthocyanins anthocyanins (mg/g) (mg/g) 70% acetone 14.3±0.8a 23.2±1.5bc 43.6±1.4b 31.0 MeOH 5.8±0.1d 20.1±0.6c 27.8±1.4d 35.1 Water 6.1±0.3d 25.3±2.9b 20.9±1.8e 32.4 20% EtOH 10.6±0.7c 21.8±2.7bc 35.1±1.6c 28.7 40% EtOH 13.7±0.8a 21.7±1.5bc 43.7±1.2b 31.7 50% EtOH 14.3±0.6a 20.4±2.2c 49.8±1.4a 32.4 60% EtOH 13.9±0.8a 23.2±2.4bc 44.5±1.5b 32.2 80% EtOH 11.9±0.6b 23.7±1.7bc 36.6±1.4c 28.6 EtOH 3.4±0.4e 41.5±3.5a 7.0±0.6f 36.0 Different letters indicated statistically significance (α=0.05, Tukey method) among different extraction solvents.

y = -130.73x2 + 122.12x + 18.769 anthocyanins 50 R² = 0.9526 phenolics

y = -38.025x2 + 36.663x + 5.5411 20 R² = 0.9432

10 Amount of extracted compound (mg/g) compound extracted of Amount

0 0 0.2 0.4 0.6 0.8 1 Propotion of Ethanol

Figure 4.3 Quadratic predictive models of purple corn cob anthocyanins and phenolics extracted by various aqueous ethanol solutions.

4.4.3 Effect of Acidity

Acid addition for anthocyanins extraction is widely used as acids may help stabilize anthocyanins by keeping them in the flavylium cation form. The commonly added acids in purple corn pigments extractions are hydrochloric acid (Jing and Giusti, 2005; Li et al.,

2008; Montilla et al., 2011; Fernando Ramos-Escudero et al., 2012), citric acid (De

Pascual-Teresa et al., 2002; Zhao et al., 2008), formic acid (De Mejia et al., 2015), and trifluoroacetic acid (González-Manzano et al., 2008). In this present study, the difference

99 in HCl addition in PCC color and phenolics extraction is shown in Table 4.2. The solvent acidity seemed to play a role in water PCC extraction but did not contribute much to PCC phenolics recovery with the aqueous organic solvent mixture, 70% acetone.

In the acidified water extraction experiments, the 1% and 2% 6N HCl acidified water produced significantly (p<0.05) higher amounts of monomeric anthocyanins and total phenolics, and lower levels of polymeric color than all other acidity levels (Table 4.2).

The low levels of polymeric color from 1% and 2% acidified water might suggest the possibility that polymeric anthocyanins in PCC underwent acid hydrolysis during the extraction. Among the acid levels under 1%, 0.01% 6N HCl acidified stood out as it recovered higher amount of pigments and phenolics with similar polymeric color levels to others (Table 4.2). This results suggested that the acidity with 0.01% (v/v) 6N HCl addition facilitated the extraction of PCC phenolic compounds without hydrolyzing the anthocyanins profile. Monomeric anthocyanin and total phenolic quantities of 70% acetone PCC extraction using different acidities were not very distinct. Only 0.01% 6N

HCl acidified sample produced significantly higher amount of monomeric anthocyanins than others (Table 4.2). However, the acid level in aqueous organic solvent mixture could impact the quality of the final extracts. Polymeric color levels of highly acidified solutions (1% and 2%) had only about 10% of polymeric color, the 0.1% acidified one also had polymeric color lower than 20% (Table 4.2). Some studies suggested using low level of acid for anthocyanins extraction because strong acid media might degrade acylated anthocyanins through hydrolysis and destroy 3-monoside anthocyanins the by breaking glycoside bonds (Kapasakalidis et al., 2006). But our HPLC chromatograms for

100 different levels of acidities were still nearly identical. The proportion of acylated pigments in PCC were almost unchanged even with 2% (v/v) 6N HCl addition, regardless of extraction solvent (Table 4.2).

Monomeric Total Extracting % Polymeric % Acylated Acidity anthocyanins phenolics solvent anthocyanins anthocyanins (mg/g) (mg/g) 0% 19.5±1.8AB 26.5±2.5A 65.8±6.1A 34.3 0.01% 21.3±1.2A 21.4±3.1B 71.8±3.8A 32.1 70% 0.1% 18.7±1.8B 16.8±2.0C 66.8±4.3A 32.5 acetone 1% 19.6±0.6AB 10.5±1.9D 70.8±4.5A 31.7 2% 18.6±0.6B 10.2±1.0D 72.2±6.2A 30.4 0% 7.9±1.0bc 23.3±1.1a 32.6±1.5d 35.8 0.01% 8.9±1.1b 23.1±0.5a 38.5±0.8bc 34.6 water 0.1% 6.4±1.2c 23.5±1.5a 34.7±1.3cd 31.3 1% 12.8±1.3a 19.8±0.7b 49.4±2.2a 33.6 2% 12.5±1.4a 18.0±0.7b 43.1±2.0b 31.0 Different upper and lower letters indicated statistically significance (α=0.05, Tukey method) among different acidities in 70% acetone and water.

Table 4.2 The anthocyanin properties (monomeric and % polymeric anthocyanins, % acylated pigments) and total phenolics of purple corn cob recovered by different acidic 70% aqueous acetone or water solution under room temperature. Acidic solutions were prepared with 6N HCl. Values were reported as mean ± standard deviation (n=3).

To sum up, 0.01% (v/v) 6N HCl acidified was recommended for PCC color extraction using both water and 70% acetone. This level of acidity could facilitate recovering considerably high amount of pigments and phenolics, and maintain the original PCC 101 anthocyanin composition without hydrolyzing polymeric, acylated and glycosylated pigments.

4.5 CONCLUSION

Solvent selection and acidity impacted PCC anthocyanins and phenolics extraction in both quantity and quality aspects. Aqueous ethanol mixtures were efficient economic solvents for high quantity and quality PCC pigments and phenolics extraction. Solvent acidity achieved by 0.01% (v/v) 6N HCl could effectively facilitate PCC phenolics compounds extraction without chemically changing the pigment composition. Extraction soaking time and cake washing number may also need to be considered for time, labor and cost efficient PCC color production. Extraction performed by 0.01% (v/v) 6N HCl acidified 50% aqueous ethanol for 45 min with double cake washing could produce high level of monomeric anthocyanins and phenolics with relatively low level of polymeric color.

4.6 ACKNOWLEDGMENTS

This work was partially funded by Chinese Scholarship Council. The authors would also like to acknowledge Agroindustrial S.A.C (Lima, Peru) and Globenatural International

S.A. (Chorrillos-Lima, Peru) for providing purple corn cob samples for this study.

102

Chapter 5: Quantification of Purple Corn (Zea mays L.) Anthocyanins Using Spectrophotometric and HPLC Approaches: Method Comparison and Correlation*

*Published in Food Analytical Methods, 2016, 9, 1367–1380.

5.1 ABSTRACT

Purple corn cob is a rich source of anthocyanins with great potential as food colorant.

Accurate quantification of anthocyanin content in cereals such as purple corn is critical to evaluate their nutritional value and facilitate their application in foods. Our objective was to determine the advantages and disadvantages of four common spectrophotometric and chromatographic methods to quantify cereal anthocyanins. Anthocyanins from fourteen purple corn cob samples were extracted, identified by HPLC-PDA-MS and quantified different methods commonly cited in the literature or used by analytical laboratories: total anthocyanins, pH-differential and HPLC methods with intact or acid hydrolyzed pigments. Polymeric color was also determined as it affects color quality. Among the four methods, the total anthocyanin method produced the highest value, followed by the pH-differential method and HPLC with intact pigments, which produced very similar results, while the HPLC with hydrolyzed pigments produced the lowest values. The

103 quantitative differences among the four methods are likely due to the differences in their specificity. Despite the differences in quantitative results, three of the methods: pH- differential and both HPLC methods, showed good linear correlation (R2≥0.98). In addition, the effect of wavelength selection and the criteria for HPLC integration were also evaluated. In summary, anthocyanin quantification results were dependent on the method chosen to quantify the pigments in the material. So reporting methodology is critical and recommended when interpreting the anthocyanin quantification results for better inter-literature comparison.

Keywords: anthocyanins; quantification; HPLC criteria; comparison

5.2 INTRODUCTION

Anthocyanins are a class of water soluble phenolic compounds that provide red, orange, purple, blue color to most fruits and vegetables. The demand to use natural colors as an alternative to controversial artificial/synthetic dyes has increased significantly in the food market. According to a natural color market report published by Mintel and Leatherhead

Food Research (2013), in 2011, global sales of natural colors reached $600 million with an annual growth rate of more than 7%, and the value of the natural colors market exceeded the artificial/synthetic dyes market for the first time ever. On the other hand, the artificial and synthetic colors market growth in value sales were more modest, by less than 4% from 2007 to 2011.The interest of seeking novel natural color sources and applying these new pigments into food industry is blooming.

104

Purple corn (Zea mays L.) is widely cultivated in low valleys throughout the Andean region of South America, especially in Peru, Ecuador, Bolivia and Argentina, and has one of the deepest purple shades among the plant kingdom. Due to its abundance in anthocyanins, the purple corn extract has long been used in coloring homemade dessert and beverage in its original places: chicha morada and mazamorra morada are among the most popular drinks and desserts prepared from purple corn (FAO, 2013). The mean anthocyanin content of whole, fresh purple corn from Peru was 16.4 mg/g (Cevallos-

Casals and Cisneros-Zevallos, 2003), which was much higher than most of the known anthocyanins-rich plants such as 1.3 to 3.8 mg/g found in blueberries (Cevallos-Casals and Cisneros-Zevallos, 2003; Wu et al., 2006), 0.21±0.03 mg/g in strawberry (Wu et al.,

2006), 3.22±0.41 mg/g in red cabbage (Ahmadiani et al., 2014), 8.57 mg/g in eggplant

(Wu et al., 2006) and 14.80 mg/g in chokeberry (Kulling and Rawel, 2008; Wu et al.,

2006). Previous studies reported that pigment in purple corn was found in especially high levels in the inedible husk and cob regions. Anthocyanin levels of cobs ranged from

0.49% to 4.60% of the dry or fresh weight, much higher than the red brownish kernels with less anthocyanins accumulation (Li et al., 2008). Besides their high concentration, purple corn anthocyanins have been reported to be potent antioxidant and anti- inflammatory agents, and may play a role in the prevention of cancer (Fukamachi et al.,

2008; Jing et al., 2008; Long et al., 2013), amelioration of lifestyle diseases such as obesity, diabetes, hyperglycemia and hypertension (Finkel et al., 2013;

Thiraphatthanavong et al., 2014; Tsuda et al., 2003). The economic cost, high

105 anthocyanin content and potential health benefits makes purple corn cob a perfect starting material for colorant production and competitive candidate as a food dye.

Unlike the well-known anthocyanin rich berries, quantification of anthocyanins in cereals such as purple corn is not as consistent even when comparing similar species and varieties. Several methods have been reported to be applied for the quantitation of purple corn anthocyanins. The total anthocyanins method, which extracts the pigments with very acidic alcoholic solution and quantifies anthocyanins spectrophotometrically without further purification, become very popular in cereal anthocyanins quantification due to its simplicity (Leja et al., 2003). First introduced in 1968, the total anthocyanins method is still used today for rapid anthocyanins determination (Fuleki and Francis, 1968a), especially in quantifying anthocyanins content in cereals such as various maize strains

(Abdel-Aal and Hucl, 1999; Li et al., 2008; Lopez-Martinez et al., 2009; Žilić et al.,

2012). The pH differential method was also widely applied for the quantitation of dark colored maize pigments quantification (Jing and Giusti, 2007; F. Ramos-Escudero et al.,

2012; Zhao et al., 2008). In addition, HPLC has also been used for purple corn anthocyanins quantification as well as for the identification of anthocyanin composition

(Aoki et al., 2002; Jing and Giusti, 2007; Montilla et al., 2011). One of the advantages of using HPLC for anthocyanins quantification is that the concentration of each individual anthocyanin could be obtained, although running HPLC is more time consuming and sophisticated than regular spectrophotometric methods (Wu et al., 2006). Most research worked directly with intact pigments to get an overall picture of detailed anthocyanin profile (Abdel-Aal and Hucl, 1999; Pedreschi and Cisneros-Zevallos, 2007; Zhao et al.,

106

2008). However, another method applied pigments after acid hydrolysis for quantification purpose since it presents a simpler chromatogram (Nyman and Kumpulainen, 2001;

Pinho et al., 2011; Truong et al., 2010; Zhang et al., 2004). This method also caught our attention since there are on one hand the risks of anthocyanins undergoing some degradation during the hydrolysis process, but there is also a possibility of releasing anthocyanins otherwise bound to other corn components. Therefore, we could not anticipate how this method would compare to the others.

Anthocyanins quantification methods have been compared in the past in fruits and vegetables, showing high correlations (R≥0.925) (Lee et al., 2008, 2005), but not much information is available about their effectiveness for cereal anthocyanins quantification.

In the present study, four commonly used anthocyanin analytical methods, including two spectrophotometric methods and two HPLC methods, were applied and tested for quantification of anthocyanins in fourteen different purple corn cob anthocyanin-rich samples. The criteria for anthocyanins HPLC quantification were also discussed. This research provide information on the advantages and disadvantages of these different methods for the quantification of purple corn pigments, as well as provide the potential of methods exchangeability through the evaluation of the correlation coefficients between different methods.

5.3 MATERIALS AND METHODS

5.3.1 Materials and Reagents

Fourteen different purple corn samples were evaluated. Three ground dried dark colored purple corn cob (Zea mays L.) coarse particles (Table 1, samples 1-3) as well as eleven

107 industrially processed dark colored commercial anthocyanin-rich purple corn cob fine powdered extracts (Table 1, samples 4-14) were kindly provided by Zanaceutica E.I.R.L.

(Lima, Peru), Alicorp S.A.A. (Lima, Peru), Agroindustrial S.A.C (Lima, Peru), and

Globenatural International S.A. (Chorrillos-Lima, Peru). All samples had moisture levels between 3-7%. The three samples of grounded purple corn cob particles were starting materials for pigment industrial extraction. The commercial powdered extracts were labeled 4-14 based on when they were received, with older samples being assigned higher numbers.

Cyanidin-3-glucoside chloride standard (≥98.0% , HPLC) was purchased from Sigma

(St. Louis, MO, USA). All high performance liquid chromatography (HPLC) grade solvents and other chemicals were purchased from Fisher Scientific (Fair Lawn, NJ,

USA).

5.3.2 The Total Anthocyanins Method

The total anthocyanins method followed the cranberry anthocyanins quantification developed by Fuleki and Francis (1968). Briefly, anthocyanins were extracted from weighted purple corn powders using a 85:15 (v/v) mixture of 95% ethanol and 1.5N HCl.

Each powder was macerated in the extraction solvent overnight at 4°C. Then the filtered extract was made up to known volume and analyzed with Shimadzu UV-2450 UV-visible spectrophotometer (Shimadzu Corporation, Tokyo, Japan) at 535 nm. Measurement was done in triplicate. The anthocyanins content was calculated using the average extinction coefficient of cranberry anthocyanins at 535 nm in acidic alcoholic system: 48500

L/(cm•mg). The final formula of total anthocyanins was reduced to: 108

100 × 퐴 × 퐷퐹 × 푉 Total anthocyanins (mg/100g) = 98.2 × 푥 Where

100/98.2 = a constant that takes the extinction coefficient and unit conversions into consideration;

A = absorbance of sample at 535 nm;

DF = dilution factor (for example, if 0.5 mL sample was diluted to 3 mL, DF=3/0.5=6);

V = the known volume anthocyanins extract was made up to after extraction (mL); x = the weight of purple corn powder for extraction (g).

5.3.3 Acetone Extraction

Weighted purple corn powders was placed in the blender and 50 mL of 0.01% (v/v) 6N

HCl acidified 70% aqueous acetone was added. The slurry was blended at room temperature for 3 min and filtered through Whatman No.4 filter paper using Buchner funnel. The cake was re-extracted until a faintly colored solution was obtained. Two volumes of chloroform were added after the filtrate was transferred to a separatory funnel. The samples were stored at room temperature overnight before evaporating the upper colored portion in a rotary evaporator at 40°C under vacuum. Finally the remaining extract was made up to known volume with 0.01% HCl acidified water (Rodriguez-Saona and Wrolstad, 2001). These extracts were used for pH differential method and two HPLC quantifications.

109

5.3.4 The pH Differential Method and Polymeric Color

In the pH differential method, the absorbance of pH1.0 and pH4.5 buffer diluted purple corn pigments were measured at their maximum absorbance wavelength (around 520 nm) and 700 nm by spectrophotometer. The monomeric anthocyanin was calculated using the molecular weight of cyanidin-3-glucoside 449.2 and its molecular absorptivity 26900 in aqueous buffer (Giusti and Wrolstad 2001). Measurement was done in triplicate. The anthocyanin content of each sample solution was calculated using the equation below:

A = (퐴푣푖푠 푚푎푥 − 퐴700)pH=1.0 − (퐴푣푖푠 푚푎푥 − 퐴700)pH=4.5

퐴 × 449.2 × 퐷퐹 × 1000 Anthocyanins (mg/L) = 26900 × 1

In addition, the polymeric color of each extract was measured. A sulfite solution could bleach the monomeric anthocyanin as sulfonic acid may link to 4-position of monomeric anthocyanins such to form colorless anthocyanin-sulfonic acid adducts (Giusti and

Wrolstad, 2001). However the polymerized colored anthocyanin-tannin complexes were resistant to bleaching because the 4-positions are taken as the mono units need this site to connect each other (Giusti and Wrolstad, 2001). Briefly, absorbance of the bisulfite bleached and control (distilled water diluted) samples were measured at 420 nm, their maximum absorbance wavelength (around 520 nm) and 700 nm by spectrophotometer.

Measurement was done in triplicate. The percentage polymeric color was calculated as follow:

(퐴 − 퐴 ) + (퐴 − 퐴 ) Percent polymeric color = 420 700 bleach 푣푖푠 푚푎푥 700 bleach × 100% (퐴420 − 퐴700)control + (퐴푣푖푠 푚푎푥 − 퐴700)control

110

5.3.5 HPLC Quantification Using Intact Anthocyanins

For each purple corn powders, 2 mL of the crude extract was passed through a C18 Sep-

Pak cartridge (Waters Corporation, Milford, MA, USA) to remove sugars, acids and other non-phenolic compounds. Then the methanol in purified pigments were evaporated at

40°C under vacuum, made up back to 2 mL with acidified water and filtered through 0.45

μm polypropylene syringe filter (Phenomenex, Torrance, CA, USA) into the vial ready for HPLC analysis.

The standard curve for purple corn anthocyanins quantification was made by injecting different volumes of cyanidin-3-glucoside chloride solution into HPLC, plotting the amount of pigments injected versus their respective peak areas. The peak areas of fourteen purple corn cob pigments samples were calculated and compared with standard curve to obtain the anthocyanin content in each sample.

A reverse-phase high performance liquid chromatograph (HPLC) system (Shimadzu

Corporation, Tokyo, Japan) consisted of LC-20AD prominence liquid chromatograph, equipped with a SIL-20AC prominence auto sampler at 4°C, a SPD-M20A prominence diode array detector and LCMS-2010EV liquid chromatograph mass spectrometer (MS), with LCMS solution Ver3.30 software was used for this analysis. The column was reversed-phase 3.5 µm Symmetry C18 column (4.6×150mm, Waters Corp., MA, USA) fitted with a 4.6×22 mm Symmetry 2 micro guard column (Waters Corporation, MA,

USA). The solvents used were A: 4.5% (v/v) formic acid in water, and B: 100% acetonitrile. Solvents were filtered through 0.45µm poly(tetrafluorothylene) membrane filters (Pall Life Sciences, Ann Arbor, MI, USA). Separation was achieved by using a 111 linear gradient from 5 to 25% solvent B in first 25 min, keeping 25% B from 25 to 30 min, followed by linear increasing B from 25% to 40% during 30 to 35 min.

Approximately 1/10 of the HPLC elute separated by a micro-splitter valve (Analytical

Scientific Instruments, Inc. El Sobrante, CA, U.S.A.) was analyzed using MS. Precursor- ion analysis by scanning the precursors of all six major anthocyanidins, including cyanidin (MW 287), delphinidin (MW 303), malvidin (MW 331), peonidin (MW 301), pelargonidin (MW 271), and petunidin (MW 317). Product ion analyses were also conducted to screen all the anthocyanins present in the sample (MW 200 to 1200).

Typical settings of were as follows: detector voltage, 1.5kV; nebulizing gas flow,

1.5L/min; CDL temperature, 250°C; heat blocker temperature, 200°C. The injection volume varied from 50 to 150µL depended on the pigment concentration, with a 0.8 mL/min flow rate. Purple corn anthocyanins chromatograms were collected at 520 nm, with MS helping to identify anthocyanins presented in purple corn. Analysis was repeated three times for each sample.

5.3.6 HPLC Quantification Using Acid Hydrolyzed Anthocyanins

For each purple corn powder, 2 mL of the crude extract was mixed 10 mL 10% HCl in a tube and heated in boiling water for 90 min. After acid hydrolysis, the pigments were passed through C18 Sep-Pak cartridge (Waters Corporation, Milford, MA, USA) for purification and made up to 2 mL with acidified water and filtered through a 0.45 μm polypropylene syringe filter (Phenomenex, Torrance, CA, USA) into the vial ready for

HPLC analysis. Triplicate analyses were conducted in this quantification. Since the anthocyanidins were highly sensitive and easily degraded, the acid hydrolyzed pigments

112 were injected to HPLC within 5 minutes after their preparation. The HPLC analysis procedure of acid hydrolyzed pigments was the same as the HPLC analysis for intact anthocyanins as described previously.

5.3.7 Statistical Analysis

For anthocyanins quantification of purple corn powders, One-way ANOVA test analysis together with Tukey methods at α= 0.05 to evaluate the mean differences of the four methods and percent polymeric color between different samples. Correlation between any two methods was determined on the anthocyanin values obtained from different methods through simple linear regression. For all statistics, p<0.05 was considered to be statistically significant. Minitab 16.0 software (Minitab Inc. State College, PA) was used to analyze data in this study.

5.4 RESULTS AND DISCUSSION

5.4.1 Anthocyanin Contents Measured by Four Different Methods

Anthocyanin contents of fourteen different purple corn cob samples measured by the total anthocyanins method, the pH differential method and the HPLC methods with intact or acid hydrolyzed anthocyanins were reported as mg of cyanidin-3-glucoside (C3G) equivalents per gram of purple corn powder (Table 5.1). Generally, for each given sample, the results of the anthocyanin content were significantly different (p<0.01) depending on the method used. The total anthocyanins method produced the highest quantitative values among the four different methods, followed by the pH differential method, the HPLC with intact pigments, while the quantitation by HPLC of acid hydrolyzed produced the lowest quantitative values of anthocyanins. 113

Anthocyanin content (mg C3G/g samples) HPLC with Sample Total pH HPLC with Polymeric acid ID Anthocyanins differential intact color (%) hydrolyzed method method anthocyanins anthocyanins 1 13.7±0.0a 12.6±0.2b 12.2±0.1b 4.6±0.3c 23.2±0.3DE 2 4.3±0.0a 3.7±0.1b 3.5±0.1b 1.1±0.2c 21.2±1.6EF 3 13.3±0.0a 3.1±0.0b 3.1±0.0b 1.2±0.2c 15.1±1.6G 4 19.0±0.2a 7.7±0.2b 7.6±0.3b 2.5±0.3c 56.0±0.4A 5 23.9±0.1a 15.3±0.0b 15.6±0.2b 7.0±0.4c 44.2±1.4B 6 18.7±0.1a 13.6±0.1b 13.4±0.2b 4.4±0.7c 18.8±2.6F 7 106.1±0.3a 84.5±1.5b 83.0±0.1b 43.1±2.5c 25.3±1.6CD 8 117.4±0.2a 100.3±0.6b 98.1±2.5b 56.1±1.5c 27.5±0.3C 9 109.0±0.2a 89.1±1.5b 86.0±0.7b 44.0±4.2c 24.2±1.2CDE 10 30.8±0.1a 19.7±0.9b 16.7±0.6c 5.7±0.7d 44.8±0.5B 11 25.7±0.0a 15.0±0.4b 12.4±0.2c 2.3±0.3d 44.1±0.3B 12 70.5±0.7a 24.0±0.6b 22.8±0.4b 10.7±0.7c 43.0±1.2B 13 60.3±0.2a 25.9±0.9b 24.4±0.3b 11.7±0.6c 45.5±0.2B 14 65.8±1.6a 37.6±1.3b 37.3±0.1b 18.3±0.6c 46.5±0.7B Different lower case letters in upper right corner indicated statistically significance (α=0.05, Tukey method) among different quantitative methods for the given sample. Different capital letters in upper right corner of polymeric color indicated statistically significant (α=0.05, Tukey method) among different samples.

Table 5.1 Anthocyanin contents and polymeric color of 14 purple corn powders. Pigment contents were measured by total anthocyanins method, pH differential method and HPLC methods with intact or acid hydrolyzed anthocyanins. Values are represented as mean ± standard deviation (n=3). Sample #1-3 were recovered from grounded corn cob, others were industrially processed pigments.

114

5.4.1.1 The total anthocyanins method

As shown in Table 5.1, the total anthocyanins method produced highest anthocyanins quantification among the four methods in this study for all samples, ranging from 4.3 mg

C3G/g to 117 mg C3G/g, regardless of the sample origin and degree of processing.

Overall, the reported anthocyanins levels in the literature for raw dark colored corn using the total anthocyanins methods were quite high, ranging from 0.93±0.01 mg/g to

8.5±0.06 mg/g (Abdel-Aal and Hucl, 1999; Li et al., 2008; Lopez-Martinez et al., 2009;

Žilić et al., 2012).

The low specificity of the total anthocyanins method may explain why this methodology produces higher quantitation than others. By taking single measurement at 535 nm only, the total anthocyanins method measures all reddish color as contributed by anthocyanins.

Fuleki and Francis (1968b) and later other researchers (Lopez-Martinez et al., 2009; Žilić et al., 2012) modified this methodology by subtracting the absorbance at 700 nm, which was believed to be due to turbidity. However, this total anthocyanins protocol fails to consider the possibility of the presence of other reddish pigments. Besides the monomeric and polymeric anthocyanins, other red pigments, phlobaphenes and its building unit, 3- deoxyanthocyanins, have also been reported to be found in purple corn (Grotewold et al.,

1994; Selinger and Chandler, 1999; Winkel-Shirley, 2001).These pigments are alcohol soluble but water insoluble brick-reddish compounds usually be found in most bark, pericarp, cob glume and seed coat of a plant (Awika et al., 2004; Matus-Cádiz et al.,

2008). Taking polymerized anthocyanins and other non-anthocyanin reddish color into

115 account, the total anthocyanins method may overestimate the anthocyanins in cereals that could finally be available for aqueous food preparation.

5.4.1.2 The pH-differential method

The pH differential method produced quantitation ranging from 3.1 mg C3G/g to 100.3 mg C3G/g, lower than the total anthocyanins method for respective samples (Table 5.1).

Previous research also measured the anthocyanins content in purple corn cob using the pH differential method: the reported value were 9.8±0.8 mg/g DW for Peruvian purple corn cob (Jing and Giusti, 2007), and 3.04±0.16 mg/g DW for Chinese purple corn (Zhao et al., 2008). These reported values are generally lower than the quantitative values produced by the total anthocyanins methods in this study.

The pH differential method, also called the total monomeric anthocyanins method, is designed to measure only single anthocyanin units. In contrast to the total anthocyanins method, the pH differential method takes the advantage of monomeric anthocyanins that can significantly change color under different acidic aqueous condition (vivid red at pH1 and colorless at pH4.5) to eliminate the interference of other reddish pigments. The water insoluble pholobaphenes, the polymeric anthocyanins and 3-deoxyanthocyanins which are still colorful under pH4.5 (Awika et al., 2004; Giusti and Wrolstad, 2001), are thus removed from the quantification. The validation of pH differential method has been well studied and verified with the standard curve method using several of fruit and vegetable juices (Dandena et al., 2012; Lee et al., 2005). Due to its specificity, it was expected and also shown in our results (Table 5.1) that the values produced by the pH differential method were lower than that of the total anthocyanins method. 116

5.4.1.3 Polymeric color

Determining how much the non-monomeric anthocyanins contribute to the reddish color, percent polymeric color is an important characteristic for anthocyanins quality evaluation. The polymeric color of samples in this study ranged from 15.1% to 56.0%

(Table 5.1). Older purple corn pigment powders (samples #11-14) tended to have higher %polymeric color while the polymeric color was much lower for the pigments freshly recovered from dried purple corn cob (samples # 1-3). The fresher (shorter storage time) purple corn pigment extracts (samples# 6-9) also had lower percent polymeric color than the older ones (samples #11-14).All these results suggested polymeric color could be a great indicator of the freshness of pigments. Polymeric color could also serve as an indicator of degree of processing. As shown in Table 5.1, pigments directly recovered from dried ground corn cob had lower %polymeric color (15.1% to

23.2%) than the industrially processed extracts (mostly >25%), proving the pigments directly recovered from dried ground corn cob was less degraded/processed than the industrially processed extracts. Although sample #4 and 5 (Table 5.1) were relatively fresh product, their high %polymeric color suggested the purple corn cob anthocyanins suffered severe over-processing, leading monomeric pigments became polymerized. It should also be noticed that the difference between total anthocyanins measurement and pH differential measurement for the same sample was relatively large when the polymeric color (Table 5.1) was relatively high (>40%), while the difference was smaller when the polymeric color was low (<25%). Regression analysis showed that the ratio of difference between the total anthocyanins method and the pH differential method value

117 versus the total anthocyanins method value were correlated (R2=0.704) with polymeric color measurement for given samples. The only deviation point in this regression was sample #3, whose total anthocyanins value was much higher than its pH differential value with only 15.1% polymeric color. Phlobaphenes in purple corn cob might have been extracted into the methanol solution and took into account in the total anthocyanins quantification may explain this deviation. It was also reported by Abdel-Aal and coworkers (2006), that the deoxyanthocyanins (building units of pholobaphenes) accounted for around 50% of the total anthocyanins. Since deoxyanthocyanins do not follow the typical anthocyanins color change due to pH changes, whether or not to count them as anthocyanins can be questioned. Additional attention has to be paid for this point when quantifying anthocyanins in phlobaphenes containing cereals. The correlation and deviation discussed above further showed that the total anthocyanins method measured more than just monomeric and polymeric anthocyanins, other reddish pigments such as phlobaphenes might interfere with this quantification.

5.4.1.4 The HPLC methods

5.4.1.4.1 The HPLC method with intact pigments

The intact pigments HPLC method produced values ranging from 3.1mg C3G/g to 98.1 mg C3G/g for the purple corn cob pigments samples in this study (Table 5.1). It could be seen that the HPLC with intact pigment quantitative values were lower than their respective total anthocyanins method values for identical samples (Table 5.1). Abdel-Aal and coworkers (2006) also reported similar results for their purple corn anthocyanins quantitation, where the HPLC quantitation was 0.965 mg/g, compared to 1.277 mg/g

118 produced by the total anthocyanins method. However, for each given sample, the pH differential quantification and the intact pigments HPLC method produced similar values.

No statistically significant difference (p=8473, t-test) was found between the two methods overall, although the intact pigments HPLC quantitation were somewhat lower than the pH differential ones. Similarity in quantitative values between these two methods were also reported in 5 different kinds of Chinese purple corn (Zhao et al.,

2008), berries (Lee et al., 2016), and most fruits and vegetable juices (Lee et al., 2008).

The similarities between the quantitative results obtained with the pH differential method and the HPLC method with intact pigments were most likely explained by the fact that both methods focus on the monomeric anthocyanins only. The water-insoluble and large polymers are typically retained by the filter before the HPLC injection to protect the very sophisticated capillary system. In addition, the PDA and MS provided a more detailed picture of the pigments, helped further identify different anthocyanins and confirmed that monomeric anthocyanins were the pigments being monitored. Since both methods considered only monomeric anthocyanins, it was expected, and it was shown in this study that the values of the pH differential method and the HPLC method with intact anthocyanins for the identical sample were very similar.

5.4.1.4.2 The HPLC method with acid hydrolyzed pigments

As shown in Table 5.1, the quantitation values produced by the hydrolyzed pigments

HPLC method were much lower that the values produced by previous three methods, from 1.1 mg C3G/g to 56.1 mg C3G/g. The acid hydrolyzed HPLC values were around or even less than half of the values obtained from other methods. Statistically significant 119 differences (p<0.01) were found between the hydrolyzed pigments HPLC and all other quantifications. Although acid hydrolyzed pigments have been reported to be applied in strawberry (Nyman and Kumpulainen, 2001), bilberry (Nyman and Kumpulainen, 2001;

Zhang et al., 2004) and red wine (Pinho et al., 2011), hydrolysis before quantification was rarely reported for anthocyanins determination in cereals.

Acid hydrolysis, which breaks the glycosidic bond of monomeric anthocyanins, hydrolyzes polymeric pigments into smaller pieces, and finally releases anthocyanidins.

Theoretically, HPLC quantification using acid hydrolyzed pigments could provide an indication of how many potential available pigments are in a sample. It was expected that the quantification from HPLC method with hydrolyzed pigments would be higher than the pH differential method and HPLC with intact pigments readings. However, quantification with acid hydrolyzed pigments was not used as often as other methods due to the poor stability of anthocyanidins (Wallace, 2011). Anthocyanidins are easily degraded when exposed to light, oxygen and non-acidic conditions (Harborne, 1983), so care must be taken after hydrolyzing the pigments. Anthocyanidins obtained from acid hydrolysis were protected from light and oxygen, cooled down and purified quickly before injection into the HPLC. The effect of time interval between sample preparation and HPLC injection could be very significant in this quantification. HPLC injection within 5 min after sample preparation could give a value which was around 1.5 times higher than that injected 30 min after preparation (data not shown). The higher standard deviations (Table 5.1) of the HPLC with acid hydrolyzed pigments method compared to others suggested the degradation and time control issue with this method. This could be

120 the reason not as much quantitative anthocyanin analysis was done using acid hydrolyzed pigments.

5.4.1.4.3 Critical parameters when quantifying by HPLC

HPLC is currently one of the most widely used quantitative methods for determing varieties of compounds in pharmaceutical and food analysis (Abdel-Aal and Hucl, 1999).

The most common procedure is, first preparing a standard curve by plotting the known amount of substance versus its HPLC signal (peak high, peak area, etc.), then integrating the HPLC signal of the unknown tested material, comparing the signal to the standard curve and obtaining the amount in tested material. This procedure should work well for single compound quantification. However, in the case of purple corn, which contains more than one anthocyanin, counting the HPLC signals are no longer as straightforward as that with only single compounds. Thus having reasonable and consistent criteria for integration of the HPLC signal is critical to quantitative HPLC measurement. In the present study, wavelength selection and integration criteria for HPLC quantification of purple corn anthocyanins were discussed to minimize possible errors in anthocyanins

HPLC quantifications.

5.4.1.4.3.1 Wavelength selection

Most published anthocyanins HPLC studies collected absorbance at 520 nm, a typical wavelength for reddish color analysis while other compounds and impurities are not seen

(Pinho et al., 2011). However, the wavelength of anthocyanins maximum absorbance varies around 520 nm depending on the anthocyanin chemical structure as well as the solvent (and pH) used. For instance, under the conditions of our HPLC run, the maximum 121 absorbance for catechin-(4,8)-cyanidin-3,5-diglucoside was 528 nm, while the maximum absorbance for pelargonidin-3-glucoside was 504 nm. Although in our purple corn samples, the maximum absorbance for most cyanidin and peonidin derivatives were between 516 nm and 522 nm, using 520 nm solely for quantitative analysis might not accurately determine the amount of different anthocyanins. Therefore, a max plot, which covers wavelength from 500 nm to 530 nm, can be used instead for multiple anthocyanins quantification.

The HPLC chromatogram at 520 nm and 500-530 nm max plot of each purple corn cob pigments sample were very similar, in fact, most of the peaks could even visually overlap each other. However, when applying the same integration parameters to the same chromatograms of a sample, the calculated peak areas are not identical (Table 5.2). The max plot had higher peak area sum and higher calculated anthocyanin content (4.68±0.05 mg C3G/g compared to 4.42±0.06 mg C3G/g processed with 520 nm only), because the max plot captured the maximum signal within 500-530 nm range while the 520 nm chromatogram only presented the signal at single wavelength. When a peak had its highest absorbance other than 520 nm, the signal for this peak is stronger in max plot than in a 520 nm chromatogram. Statistical analysis suggested that application of max plot produced significant higher (p=0.013) quantitative anthocyanins counts than using

520 nm solely. Therefore, using a single wavelength may underestimate the pigments content of a plant material which contains more than one anthocyanin. Based on the comparison of two wavelength selection, max plot 500-530 nm was applied in the anthocyanins quantification in this study.

122

Peak area Compound 520 nm 500-530 nm catechin-(4,8)-cyanidin-3,5- 95123±2757 99262±1179 diglucoside cyanidin-3-glucoside 538137±116533 5702063±78065 pelargonidin-3-glucoside 434065±27814 486133±13993 cyanidin-3-malonylglucoside 91925±2314 93766±1507 peonidin-3-glucoside 1471677±14714 1602593±67353 cyanidin-3-(6''-malonylglucoside) 2349606±35418 2397641±53907 cyanidin-3-succinylglucoside 25983±2446 52516±497 pelargonidin-3-(6''- 138576±2790 164148±2640 maolonylglucoside) peonidin-3-(6''-malonylglucoside) 570330±8498 581506±10734 peonidin-3-dimalonylglucoside 32992±1619 35784±1813 sum of peak area 10591652±142339a 11215414±131542b anthocyanins content in sample 4.42±0.06A 4.68±0.05B (mg C3G/g powders) Different letters in upper right corner indicated statistically significant (α=0.05, paired t- test, p=0.013) between different wavelength.

Table 5.2 Peak area values of a purple corn anthocyanins ground powders (sample #1) using same integration parameters for their HPLC 520 nm and 500-530 nm max plot chromatograms and corresponding calculated anthocyanin contents.

123

5.4.1.4.3.2 Criteria for multiple peaks HPLC quantitative integration

Interestingly, the anthocyanins HPLC quantitative studies rarely specify how their integration was done. One reason might be that the integration of anthocyanins HPLC chromatograms with a single standard could be somewhat subjective, and this might be the step that may introduce serious error in HPLC quantification. To avoid such subjectivity, many researchers would rely on the sophisticated mathematical models of the HPLC software. However, some basic integration parameters must be set up by the analyst so that the software can carry out those calculations. The integration parameters in the HPLC software used for the present study were the slope and minimal (min) area/height, which directly relate to peak recognition and the definition of the baseline.

Slope determines the sensitivity of peak detection, the start of a peak is recognized when the slope of the peak exceeds the setting value. Min area/height distinguishes peaks from noise, only area/height with higher value than the setting will be recognized as peak for quantitative calculation. In addition to parameters setting, manually integration function offers options to horizontal /vertical move the peak starting and ending points, add or reject certain peaks as well as unify peaks. These functions make it even more flexible with the HPLC integration. Depending on different integration parameters/criteria, the calculated outcome could vary very much.

An examples of the impact of different integration parameters on multiple peaks HPLC quantification is shown in Figure 5.1, where the same purple corn pigments chromatogram was integrated in six possible different ways. Figure 5.1a set high values for both slope and min area, which only took the significant peaks into account; while 124

Figure 5.1b set much lower slope and min area to sum up all the possible anthocyanin signals. It was noticed in the major peaks area (11-20 min), that the bottom of the peaks were a little bit higher than the non-signal flat area (after 25 min), and these minor signals between the two major peaks might also be anthocyanins. That is why Figure 5.1c considered these connecting parts by unifying the major peaks. In addition to Figure

5.1c, Figure 5.1d also added the connecting parts of the minor peaks. Figure 5.1e took all the connecting parts between the starting point of the first peak and the finishing point of the last peak into consideration. In Figure 5.1f, all the absorbance (everything above

0) was considered due to the presence of pigments. Here, it is clear that even given the same sample with the identical HPLC chromatogram, the measured anthocyanins contents (Table 5.3) could be much different due to various integration parameters/criteria. The calculated anthocyanin content using integration in Figure 5.1f was 127.8 mg anthocyanins per gram of powders, nearly twice as much as that using integration in Figure 5.1a (70.0 mg/g).

In the present study, the integration shown in Figure. 5.1d was considered the best, because it considered all possible anthocyanin major and minor signals, as well as selectively integrated the areas between the major peaks in a reasonable way. Integration in Figure. 5.1e and 5.1f overestimated the importance of the connecting areas between major peaks group and minor peak signals, which may cause addition of non- anthocyanins signals during quantification. On the other hand, integration in Figure 5.1a,

5.1b and 5.1c ignored the effects of minor peaks and connecting areas, thus leading to underestimation of the pigments when using HPLC quantification.

125

Figure 5.1 Example of impact of HPLC integration parameters of max plot at 500-530 nm of purple corn (Zea mays L.) pigment extracts (sample #7). Black lines with arrows indicated different integrations for peak area calculation.

126

Total Anthocyanin quantification Chromatogram ID Peak Area (mgC3G/g powders) a 32700656 70.0 b 34230083 73.2 c 37788540 80.8 d 38570990 82.5 e 46940422 100.4 f 59767149 127.8 Table 5.3 Example of impact of different HPLC integration parameters: peak area sum values of purple corn extract (sample #7) – same run - using different integration parameters and their calculated anthocyanin quantification corresponding to Fig. 5.1.

Calculating the peak area may not be a problem with a perfect chromatogram, where all peaks achieve full resolution and the baseline is completely flat, and where peaks are high and clean without noise. However, this perfect situation is rarely seen in real-day research when analyzing plant materials. Thus unifying criteria for non-perfect HPLC chromatogram quantitative integration is important. Based on this experiment, by comparing six different integrations, we chose to follow the following criteria to get continuous reasonable quantitative HPLC integration for purple corn anthocyanins quantification. Firstly, unify the target peaks if they are close to each other and whose bottom part is higher than the baseline, in whihch baseline refers to the non-target compounds parts in a chromatogram, not the “0” absorbance line. Secondly, do not unify two peaks if the distance between them is longer than the peak width.

127

5.4.1.4.4 Purple corn anthocyanins identification

PDA-MS was applied in this study to identify different anthocyanins in purple corn cob.

Consistent with published results, only three anthocyanidins were found in purple corn cob after acid hydrolysis (Figure 5.2a). They were identified by MS as cyanidin, pelargonidin and peonidin (Table 5.4) (De Pascual-Teresa et al., 2002; Jing and Giusti,

2005). As for the intact pigments, six major anthocyanins reported previously were found in most purple corn powders in this study (Figure 5.2b and Table 5.4). Based on comparison with the literature (De Pascual-Teresa et al., 2002; Jing and Giusti, 2005;

Zhao et al., 2008; Žilić et al., 2012), together with the PDA and MS data they were identified as cyanidin-3-glucoside (peak 6), pelargonidin-3-glucoside (peak 7), peonidin-

3-glucoside (peak 10), cyanidin-3-(6”-malonylglucoside) (peak 11), pelargonidin-3-(6”- malonylglucoside) (peak 13) and peonidin-3-(6”-malonylglucoside) (peak 15).

Besides six major anthocyanins, eight more minor anthocyanins were also found in some of the purple corn cob samples (Figure 5.2b and Table 5.1). Tentatively identified as catechin-(4,8)-cyanidin-3,5-diglucoside, peak 4 was found in all fourteen samples in this study though its proportion was only around 1%. This flavonol-anthocyanin pigment in purple corn was first reported by González-Paramás and coworkers (2006) and its structure was then characterized by NMR (Montilla et al., 2011; González-Manzano et al., 2008). Cyanidin-3,5-diglucoside (peak 5) in purple corn was previously reported by

Žilić and coworkers (2012). In this study, this diglucoside pigment was only found in sample 12-14. The presence of this pigment in other purple corn powders were negligible, possibly due to glycoside in 5 position of anthocyanidins being more 128 susceptible than others in the processing conditions. Two different minor cyanidin-3- malonylglucosides (peak 8 and 9) was also identified. They might be isomers of the major anthocyanin cyanidin-3-(6''-malonylglucoside) at peak 11, as the acylated group in

6’’ position of glucose could freely rotate and thus form different isomers. The minor peak 12 was identified to be cyanidin-3-succinylglucoside as previously reported by Li and coworkers (2008). The dimalonyl derivative of cyanidin (peak 14), pelargonidin

(peak 16) and peonidin (peak 17) were also found in our samples, similar results were reported previously (Aoki et al., 2002; Jing et al., 2007; Montilla et al., 2011). None of the ethylmalonyl derivatives of anthocyanidins in purple corn reported by De Pascual-

Teresa and coworkers (2002) was found in the fourteen samples in this study.

Figure 5.2 Typical anthocyanidins and anthocyanins HPLC profile at of purple corn (Zea mays L.). 129

Retention m/z Peak Molecular Found in Compound time Fragment # Weight sample # (min) [M+H]+ [M+H]+ 1 cyanidin 17.8 287 287 all 2 pelargonidin 20.8 271 271 all 3 peonidin 22.0 301 301 all catechin-(4,8)-cyanidin-3,5- 4 4.6 899 287 all diglucoside 5 cyanidin-3,5-diglucoside 10.5 611 287 12-14 6 cyanidin-3-glucoside 11.4 449 287 all 7 pelargonidin-3-glucoside 13.0 433 271 all 8 cyanidin-3-malonylglucoside 13.6 535 287 4,5,6,7 1-3,6- 9 cyanidin-3-malonylglucoside 14.2 535 287 11,14 10 peonidin-3-glucoside 14.4 463 301 all 11 cyanidin-3-(6''-malonylglucoside) 16.0 535 287 all 12 cyanidin-3-succinylglucoside 17.4 549 287 6,8,9 pelargonidin-3-(6''- 13 17.7 519 271 1-11,14 maolonylglucoside) 14 cyanidin-3-dimalonylglucoside 18.6 621 287 1,2,6-10 15 peonidin-3-(6''-malonylglucoside) 18.9 549 301 all 16 pelargonidin-3-dimalonylglucoside 20.7 605 271 1,2,4-8,11 17 peonidin-3-dimalonylglucoside 21.5 635 301 1,2,4-8,11 Table 5.4 MS spectral data for typical purple corn anthocyanins (see Figure 4.2).

5.4.1.5 Effect of the extraction procedure

Two different extraction approaches were applied in this study, acidic methanol extraction and acetone-chloroform extraction. Extraction of anthocyanins was mostly carried out with mild acidified methanol, ethanol, water or other polar organic solvents, under different pH and temperature conditions (Escribano-Bailón et al., 2004; Jing and

130

Giusti, 2007). It has been widely reported that extraction could affect the yield of anthocyanins (Jing and Giusti, 2007; Yang and Zhai, 2010), and thus affect the quantification.

As discussed above, the highly acidic alcoholic solution recovered monomeric, polymeric as well as other alcohol soluble pigments, and all the reddish compounds were measured in the total anthocyanins method without further purification. The relatively high capability for pigments extraction may introduce compounds other than anthocyanins into our quantification. In contrast, in the acetone extraction, after the chloroform separation, only water soluble compounds were recovered. The highly polymerized polymeric anthocyanins and phlobaphenes was eliminated due to their insolubility. The 3- deoxyanthocyanins was also mostly eliminated due to its relative lower polarity because of missing hydroxyl group in 3 position, Even if very limited amount of deoxyanthocyanins entered the final analyzed solution, they could be elimated either by pH changes or the purification step for HPLC. As the pH differential method and both the

HPLC methods started with acetone extraction recovered pigments, it was expected and shown in Table 5.1 that these three measurements gave lower anthocyanins readings than the total anthocyanins method.

5.4.2 Correlation between Different Anthocyanins Quantification Methods

The four methods used in this study provided quantitative results that were different.

However, they all followed a linear correlation (R2>0.88) as shown in Figure 5.3. The total anthocyanins method generally had lower correlation coefficients (R2=0.884, 0.901,

0.902) with other three methods when compared to other individually as pairs (Figure 131

5.3 a, b and c). The differences in polymerization levels throughout the samples might be primarily responsible for the lack of fit in the linear regression of these total anthocyanins method pairs. Points of total anthocyanins values between 60 to 80 mg/g which were far lower than the regression line were all high in polymeric colors (Table 5.1). Those with relatively lower polymeric color were much closer surrounded around the regression line.

As for the hydrolyzed HPLC values, although they were well correlated with the pH differential method and the intact HPLC method (Figure 5.3 e and f) values, the high susceptibility to the experimental conditions of anthocyanidins made the slopes of these regression lines much lower than other pairs.

132

Figure 5.3 Simple linear regression plots between two of the four anthocyanins quantitative methods for fourteen purple corn (Zea mays L.) pigment-rich samples.

133

The pH differential method and the intact HPLC method pair not only had the highest correlation (R2=0.998) among all six pairs (Figure 5.3d). It should be noted that the slope for their regression was also very close to 1 (0.98) and the intercept was close to 0 (-

0.83), indicating that the values produced from the two methods were almost identical.

This results suggested that these two approaches could be exchangeable in cereal anthocyanins quantification when working with the appropriate integration parameters.

Known as mostly applied methods, several researches have been done on comparison and correlation between the pH differential method and intact pigment HPLC method (Lee et al., 2008, 2016). Using two anthocyanin standards (cyanidin-3-glucoside and malvidin-3- glucoside), reading with two instruments (cuvettes and microplates) and running with two

HPLC systems (different columns and mobile phase conditions), Lee and coworkers

(2008) tested seven anthocyanins rich juice and concluded the two methods were highly correlated (R≥0.925, p≤0.05) although HPLC method values were a little bit higher.

Other researchers used five different berries and found the intact pigments HPLC and the pH differential methods in good accordance each other, with HPLC analysis having similar trends but lower values than pH differential analysis (Lee et al., 2016). The high correlation and value closeness of these two methods could be explained by application of the same criteria to selectively measured only monomeric anthocyanins. Although results from the two methods could vary depending on the materials, the data from the pH differential method and the HPLC method with intact pigments were quite similar. For those having limited access to HPLC, the pH differential method could work as a simple, economic substitution for anthocyanins quantification.

134

5.5 CONCLUSION

In conclusion, anthocyanin quantification results were dependent on the method chosen to extract and quantify the pigments in the material. When determining anthocyanin contents in purple corn cob, the total anthocyanins method with single wavelength/pH measurement takes all compounds with reddish color into account, such that it may overestimate anthocyanins in the material. Information about the presence of phlobaphenes or deoxyanthocyanins should be taken into consideration when choosing the method and interpreting the results. The pH differential method and the HPLC method with intact anthocyanins eliminate interference from other compounds and only measure the monomeric anthocyanins. They were considered as relatively appropriate true-to label quantification methods for monomeric anthocyanins. Due to instability of anthocyanidins which could lead to pigment degradation during preparation, the HPLC method with acid hydrolyzed pigments tended to underestimate the anthocyanin content and was not recommended for cereal anthocyanins quantification analysis. The HPLC method with intact pigments produce very similar results to the pH differential method with R2=0.998 if appropriate wavelength and integration criteria were applied, suggesting the pH differential method is a great alternative if individual anthocyanin identification and quantification is not required. In addition, it is critical to apply appropriate and unified criteria for HPLC quantitative integration in order to obtain data that can be compared among different literatures. Polymeric color measurement could serve as supplement to pH differential method to better evaluate the pigments quality, offering true-to-label anthocyanins information.

135

5.6 ACKNOWLEDGEMENTS

We appreciate Zanaceutica E.I.R.L. (Lima, Peru), Alicorp S.A.A. (Lima, Peru),

Agroindustrial S.A.C (Lima, Peru) and Globenatural International S.A. (Chorrillos-Lima,

Peru) for providing the purple corn samples.

136

Chapter 6: The Effect of Pigment Matrix, Temperature and Amount of Carrier on the Yield and Final Color Properties of Spray Dried Purple Corn (Zea may L.) Cob Anthocyanin Powders

6.1 ABSTRACT

Spray drying is an economic technique widely used to extend shelf-life of heat sensitive materials such as anthocyanin-based food colorants. Purple corn cob (PCC) is a rich pigment source of industry relevance but some commercial PPC powders show low water solubility leading to color loss. Our goal was to evaluate the impact of different spray drying conditions on the yields and quality of PCC anthocyanin powders.

Yields, monomeric and polymeric anthocyanins, color properties (CIELch, haze), and composition before and after spray drying of varying PCC anthocyanin matrices (hot water, 40% aqueous ethanol, C18 purified), inlet temperature (130, 150, 170°C) and amount of maltodextrin (2%, 5%, 10%) were investigated.

The yield and final color quality of spray dried PCC anthocyanins were affected (p<0.05) by all parameters evaluated. The amount of carrier had the highest impact on anthocyanins yield, 5-10% maltodextrin recovered >80% of the input pigments while 2% maltodextrin only recovered 57%. Ethanolic PCC pigment matrix produced significantly

137 higher final product haze than the aqueous matrix, but anthocyanin purification before dehydration did not significantly improve final powder color nor solubility. The hue of the reconstituted powders matched the initial hue of the solutions, the main changes were higher lightness and lower chroma due to processing pigments loss.

Quality of feeding materials, and drying conditions affected the yield and quality of the final product. Hot water extracts spray dried with 5% maltodextrin at 150°C gave the highest pigment yield and satisfactory solubility with the least color changes.

Keywords: spray drying, anthocyanins, yield, polymeric color, HPLC.

6.2 INTRODUCTION

Color plays an important role in consumers’ acceptance of foods. Natural food colorants have increased in popularity compared to artificial dyes because they may help enhance visual appeal while adding health-promoting phytochemicals. According to a 2013 natural color market report (Mintel and Leatherhead Food Research 2013), the global sales of synthetic dyes stayed almost unchanged from 2007 ($548 million) to 2011 ($570 million), while the sales of natural colors increased dramatically from $465 million in

2007 to over $600 million in 2011, with an annual growth rate of over 7%.

First introduced in milk powders production, spray drying is a continuous operation widely used to encapsulate functional compounds (Balassa et al., 1971). It is 30-50 times less expensive than freeze-drying (Gharsallaoui et al., 2007) and has been successfully applied in fruits and vegetables source colorants production such as acai, berries, pomegranate, purple sweet potato, black carrot (Fazaeli et al., 2012; Jimenez-Aguilar et

138 al., 2011; Robert et al., 2010; Tonon et al., 2010). Quality of the final spray dried products can be affected by inlet/outlet temperature, choice of carrier, physical properties of the feed, among other factors (De Vos et al., 2010; Kandansamy et al., 2012; Mahdavi et al., 2014).

Purple corn (Zea mays L.), originated from Peru, is abundant in health beneficial anthocyanins that could be a competitive candidate as natural pigment source. Compared to famous anthocyanins rich food blueberry (anthocyanin content 1.3-3.8 mg/g FW), purple corn has higher anthocyanins level of 6.8-82.3 mg/g FW depending on the sections (Cevallos-Casals and Cisneros-Zevallos, 2003; Li et al., 2008; Wu et al., 2006).

The purple corn cob (PCC), being low in sugars and rich in soluble fiber, which had been reported as a drying aid in encapsulation in spray dried cactus pear pigments (Saénz et al., 2009), is considered an ideal starting material for spray dry pigment production compared to other plant anthocyanin sources. However, spray dried PCC pigments has not yet been investigated.

Most previous research discussed physicochemical properties of spray dried anthocyanins, such as moisture content, water activity, bulk density, particle size, glass transition temperature, solubility, antioxidant capability, storage stability (Ahmed et al.,

2010; Ersus and Yurdagel, 2007; Fazaeli et al., 2012; Ferrari et al., 2012b). But the detailed chemical profiles of the pigments changes during spray drying process, like

HPLC chromatogram, %polymeric color were rarely discussed. Polymeric color is a widely used index for anthocyanin degradation in aqueous extract, juice and wine (Choi et al., 2002; Gao et al., 1997; Giusti and Wrolstad, 2001; Wrolstad et al., 2005). Different

139 from monomeric anthocyanins, the C-4 position of polymeric anthocyanin is covalently linked to other phenolic compounds, which makes polymeric anthocyanins resistant to bisulfite bleach (Berké et al., 1998). These polymeric anthocyanins generally have higher hydrophobicity, and larger molecular weight compared to monomeric anthocyanins, which may affect their water solubility in a food matrix and bioavailability to humans

(Gao et al., 1997; Hager et al., 2008b). Polymeric anthocyanins are typically formed during thermal juice/puree processing and long-time storage, during which monomeric anthocyanins and other phenolic compounds condense and form polymeric anthocyanin

(Choi et al., 2002; Hager et al., 2008a). Changes in polymeric anthocyanins have been well studied in thermal processing of anthocyanin-rich juice and puree (Brownmiller et al., 2008; Hager et al., 2008b), but changes in polymeric anthocyanins in spray drying anthocyanins-rich materials have not yet been reported.

The objective of this study was to investigate the PCC pigments yield (monomeric anthocyanins), compare the PCC anthocyanins composition (%polymeric color and

HPLC profiles), color properties (CIELch), water solubility (%haze) before and after spray drying, as impacted by choice of feeding pigment matrix, inlet/outlet spray drying temperature as well as the amount of carrier applied.

6.3 MATERIALS AND METHODS

6.3.1 Materials and Reagents

Dried purple corn cob (Zea mays L.) particles were kindly provided by Agroindustrial

S.A.C (Lima, Peru). Maltodextrin (DE=16.5-19.5) were purchased from Sigma-Aldrich

140

(St. Louis, MO). All solvents and other chemicals were purchased from Fisher Scientific

(Fair Lawn, NJ, USA).

6.3.2 Spray Drying Conditions

All spray dry experiments was performed by laboratory scale mini spray dryer B-290

(BÜCHI Labortechnik AG, Switzerland). Briefly, after adjusting the feeding monomeric anthocyanins concentration to around 0.35 mg/mL, 500 mL of PCC pigment extract was mixed with maltodextrin under magnetic agitation and heated to around 50°C, and pumped through a peristaltic pump with drying nitrogen flow rate of 35 m3/h for spray- drying. The feed pigment flow rate was 1.5 mL/min.

6.3.2.1 Pigment matrix

Anthocyanins extracts obtained from three different solvents with different degree of pigment purity were pumped into spray dryer to investigate the effect of pigment matrix on the final quality of spray dried PCC pigments product. The solvents used in this experiment were 40% (v/v) aqueous ethanol, 0.01% 6N HCl acidified hot water (70°C) and 70% (v/v) aqueous acetone. Extraction was performed at room temperature if not specified. Ground PCC powders were soaked into solvent for 1 hour on a magnetic stir plate with a stir rate of 150 rpm. The extract was then filtered through Whatman No.4 filter paper using a Buchner funnel. For the 40% (v/v) aqueous ethanol extract, the solution was sent to rotary evaporator at 40°C under vacuum to remove ethanol before spray drying. As for the 70% (v/v) aqueous acetone extract, the solution was partitioned with chloroform before sending the aqueous colored portion to rotary evaporator at 40°C under vacuum to remove the remaining acetone. The obtained aqueous pigments solution

141 was then passed through a C18 cartridge to remove non-phenolic compounds (Rodriguez-

Saona and Wrolstad, 2001) to make them ready for spray drying. The total soluble solids of the PCC pigment extracts before adding maltodextrin were all <1%, measured by a handheld refractometer (Fisher Scientific, China). The amount of maltodextrin mixed into pigments was 25g, 5% (w/v) of the pigment solution, inlet temperature was 150°C and outlet temperature was 105±1°C. The obtained spray dried powders was solubilized into distilled water to measure their quality.

6.3.2.2 Inlet/Outlet temperature

In this part, three different inlet temperature settings, 130°C, 150°C, 170°C, were applied to investigate the effect of inlet temperature on the final quality of spray dried PCC pigments product. The outlet temperature depends mainly on the inlet temperature. The

PCC pigments used for this part of the experiment were all recovered from 0.01% 6N

HCl acidified hot water (70°C). Similar to the process mentioned previously, the spray dryer was pre-heated to the set inlet temperature, 5% (w/v, 25g) of the maltodextrin was mixed into 500 mL of pigments solution before pumping to the spray dryer. The obtained spray dried powders was solubilize back into distilled water for quality evaluation.

6.3.2.3 Amount of maltodextrin

In this part, three different maltodextrin amounts, 2%, 5%, 10% (w/v), were mixed with the 500 mL colored solution extracted by hot acidified water to investigate the effect of amount of maltodextrin on the final quality of spray dried PCC pigments product. Similar to the process mentioned previously, 2%, 5%, 10% (w/v, 10g, 25g, 50g, respectively) of the maltodextrin was mixed into 500 mL of PCC pigments solution, then spray dryer was

142 pre-heated to 150°C inlet temperature before pumping the pigments in for drying. The obtained spray dried powders was solubilized into distilled water for quality evaluation.

6.3.2.4 Reconstitution of pigmented solution from dried powders

The spray-dried purple corn anthocyanins powders were recovered from the collection vessel only, any particles deposited on the dryer chamber were discarded. Powders were weighted, and only 1/5 of the powders was solubilized back to 100 mL of distilled water.

The reconstructed colored solution was used for pigment quality evaluation as “after spray-drying” sample.

6.3.3 Quality Evaluation Parameters

Quality evaluation of the PCC pigments includes pigment yield based on monomeric anthocyanins, polymeric anthocyanins, color and %haze in the initial and reconstituted pigment solution. The PCC anthocyanins HPLC profiles before and after spray drying were also analyzed.

6.3.3.1 Yield of the pigments.

The spray-dried yield of PCC pigments was calculated using the following equation.

푀표푛표푚푒푟𝑖푐 푎푛푡ℎ표푐푦푎푛𝑖푛 푐표푛푡푒푛푡 𝑖푛 푠푝푟푎푦 푑푟𝑖푒푑 푝표푤푑푒푟푠 푌𝑖푒푙푑 = × 100% 푀표푛표푚푒푟𝑖푐 푎푛푡ℎ표푐푦푎푛𝑖푛 푐표푛푡푒푛푡 𝑖푛 푓푒푒푑 푠표푙푢푡𝑖표푛

The total monomeric anthocyanins in each PCC solution were determined by the pH differential method. Absorbance of pH1.0 and pH4.5 buffer diluted sample were measured at 520 nm and 700 nm by a Shimadzu UV-2450 UV-visible spectrophotometer

(Shimadzu Corporation, Tokyo, Japan). The monomeric anthocyanin was calculated using the molecular weight of cyaniding-3-glucoside 449.2 and its molecular absorptivity

143

26900 in aqueous buffer (Giusti and Wrolstad, 2001). Measurements were done in triplicate.

6.3.3.2 Polymeric color

A bisulfite solution was used to bleach the monomeric anthocyanin while the polymerized colored anthocyanin-tannin complexes were resistant to bleaching. Briefly, absorbance of the bisulfite bleached and control (distilled water diluted) samples were measured at 420 nm, 520 nm and 700 nm by spectrophotometer. The percentage of polymeric color was calculated using bleached sample color divided by the control sample color (Giusti and Wrolstad, 2001). Each sample was measured for three times.

6.3.3.3 Color property and %haze

Since the anthocyanins color is highly affected by pH, the pH of the initial and the reconstituted pigment solutions were measured by pH meter (Mettler-Toledo, model

S220, Schwerzenbach, Switzerland). Color was measured using ColorQuest XE

(HunterLab, Inc., Reston, VA) to detect total transmittance under illuminant D65, 10° observer angle with 1.000 inch view area, reflectance specular included mode. The instrument was calibrated before every session of use. Briefly, after a 1:1(v/v) dilution with distilled water, 2 mL of diluted extract was pipetted into a 2mm pathlength disposable cuvette, the CIE color parameters L*, c*, h* and % haze, were recorded. Each sample was measured in triplicate and the average was used for analysis. The color difference of the pigments before and after spray-drying was evaluated by ΔE, the square root of the sum of square for L*, c* and h*. Solubility of the spray dried PCC anthocyanins powders were evaluated based on %haze. Usually a haze less than 5%

144 indicates a clear solution without visible suspension, suggesting that the solutes are well- dissolved in the solvent (Jing and Giusti, 2005).

6.3.3.4 Anthocyanins HPLC profiles

The anthocyanins profile of each PCC pigments prior and after spray drying was purified by C18 Sep-Pak cartridge (Waters Corporation, Milford, MA, USA) and filtered through

0.45 μm polypropylene syringe filter (Phenomenex, Torrance, CA, USA) before putting into the vial ready for HPLC analysis. A reverse-phase high performance liquid chromatograph (HPLC) system (Shimadzu Corporation, Tokyo, Japan) consisted of LC-

20AD prominence liquid chromatograph, equipped with a SIL-20AC prominence auto sampler at 4°C, a SPD-M20A prominence diode array detector, with LCMS solution

Ver3.30 software was used for this analysis. The column was reversed-phase 3.5 µm

Symmetry C18 column (4.6×150mm, Waters Corp., MA, USA) fitted with a 4.6×22 mm

Symmetry 2 micro guard column (Waters Corporation, MA, USA). The solvents used were A: 4.5% (v/v) formic acid in water, and B: 100% acetonitrile. Solvents were filtered through 0.45µm poly(tetrafluorothylene) membrane filters (Pall Life Sciences, Ann

Arbor, MI, USA). Modified from a previous purple corn anthocyanins study (Lao and

Giusti, 2016), separation was achieved by using a linear gradient from 7 to 23% solvent

B in first 20 min, keeping 23% B for 2 min, followed by linear increasing B from 23% to

40% for 5 min. The injection volume was 50 µL, with a 0.8 mL/min flow rate. PCC anthocyanins chromatogram was collected at 520 nm.

145

6.3.4 Statistical Analysis

One-way ANOVA test analysis together with Tukey methods at α= 0.05 were applied to evaluate the mean differences on yield and polymeric anthocyanins and anthocyanins

HPLC profiles after spray drying among different extract matrices, inlet temperatures and percent maltodextrin. The difference between before and after spray-drying was evaluated with pairwise t-test. For all statistics, p<0.05 was considered to be statistically significant unless specified. Minitab 17.0 software (Minitab Inc. State College, PA) was used to analyze data in this study.

6.4 RESULTS AND DISCUSSION

6.4.1 Spray-dried Pigments Yield

The spray dried PCC anthocyanins yield using different extract matrices and amount of maltodextrin under three inlet temperatures is shown in Table 6.1. The yield of total monomeric PCC anthocyanins in this study ranged from 57.3±8.9% to 89.4±3.9%.

Consistent with previous studies working on spray dried anthocyanins from various plant sources, pigments retention rates using comparable drying conditions were 69%-80% for blackberry (Ferrari et al., 2012b), 77%-86% for acai (Tonon et al., 2008), 65%-93% for

Garcinia Indica Choisy (Nayak and Rastogi, 2010), ~70% for pomegranate (Yousefi et al., 2011), and 50%-77% for black mulberry (Fazaeli et al., 2012).

The PCC pigments retention after spray-drying was mainly depending on the amount of maltodextrin applied to the inject pigment solution, while the pigment matrices and the inlet/outlet temperature did not show significant effects (Table 6.1). As shown in Table

6.1, PCC anthocyanins mixed with 2% of maltodextrin had the lowest pigments recovery

146 rate of 57.3±8.9%, which was significantly lower (p=0.003) than all other groups with maltodextrin amount at least 5%. This observation suggested that a certain amount of carrier agent is critical to efficiently enhance high spray dried anthocyanins recover yield.

Carrier agents such as maltodextrin with various DE levels, gum Arabic, waxy starch and soybean isolates have been demonstrated in different studies to help increasing the pigments retention in the spray drying process due to their capability to change the hygroscopic and thermoplastic character of the powders (Osorio et al., 2010; Robert et al., 2010; Yousefi et al., 2011; Fazaeli et al., 2012; Ferrari et al., 2012a). Without the aid of carrier, pomegranate juice formed a hard glass film on the spray dryer chamber wall in the dehydrate process (Yousefi et al., 2011). In the case of spray drying black mulberry, enhancing maltodextrin concentration from 8% to 12% and 16% significantly increased the process yield (Fazaeli et al., 2012). It was also shown in our 2% maltodextrin run that pigments that could not have enough interaction to the carrier agent tended to adhere to spray-drying wall chamber and cyclone, resulting to loss of the PCC anthocyanins. Much smaller amount of PCC pigments were observed to adhere to the dryer chamber wall when the maltodextrin level was at least 5%. Spray dry purple sweet potato was an exception, Ahmed et al., (2010) reported no significant difference was found in the anthocyanins content between the encapsulated and non-encapsulated flours produced under their spray-drying condition. However, maltodextrin encapsulation did present significantly better capability in total phenolics and flavonoids retention, thus lead to higher antioxidant capability in the final powders compared to the non-encapsulated ones

(Ahmed et al., 2010).

147

Temperature (°C) Type of Pigment Maltodextrin Anthocyanins Polymeric Color (%) Matrix Amount Yield (%) Inlet Outlet Before After 130 89±1 81.3±7.1a 33.6±0.8 35.0±1.0 150 105±1 Hot water 5% 89.4±3.8a 29.4±2.5 31.4±2.0 170 119±1 extract 85.6±3.5a 32.6±0.6 34.3±1.0*

2% 57.3±8.8b 33.8±0.2 35.5±0.6 150 105±1 Hot water 5% 89.4±3.9a 29.4±2.5 31.4±2.0 extract 10% 88.5±3.5a 32.9±0.5 34.4±0.5

40% EtOH 82.3±1.3a 36.9±3.5 37.3±3.5 150 105±1 Hotextract water 5% 89.4±3.9a 29.4±2.5 31.4±2.0 C18extract purified 84.7±4.9a 29.4±1.4 30.5±0.2 Values are represented as mean ± standard deviation (n=3). Different letters and * indicate significant differences at 0.05 level on yield ANOVA-Tukey comparison and before-after polymeric color paired t-test, respectively.

However, increasing the carrier amount up to 10% did not help significantly increase the

PCC pigments yield under our experiment conditions (Table 6.1). When the carriers was already above its appropriate level, adding more carriers did not facilitate yielding more pigments during the spray drying process. Yousefi et al. (2011) presented in their spray- dry pomegranate juice study that enhancing three different carrier agents (maltodextrin, gum Arabic, waxy starch) level from 8% to 12% did not change anthocyanins yield very much. In agreement to the results in our study, Tonon et al. (2008) showed increasing maltodextrin concentration from 10% to 20% and 30% did not significantly increase the spray-dry yield of acai anthocyanins. 148

In the present study, the inlet/outlet temperature as well as the anthocyanin matrix did not play a significant role in the yield of spray dry PCC anthocyanins (Table 6.1). Similar observation was also found in spray dried pomegranate juice and pomegranate ethanol extracts (Robert et al., 2010). However, some other literature reported that a higher inlet/outlet temperature could significantly reduce the pigments retention in the spray drying process. Simply by adjusting pump rate to control the outlet temperature to 80°C,

Fang and Bhandari (2011) were able to recover more than 94% of the injected bayberry anthocyanins with an inlet temperature 150°C using same instrument as in our study.

Blueberry spray dried under 140/80°C (inlet/outlet temperature) had higher total phenolics and anthocyanins retention than that under 160/95°C (Jimenez-Aguilar et al.,

2011). Similarly, increasing inlet temperature from 160°C to 180°C and 200°C led to more black carrot pigments loss, regardless the carrier type was used (Ersus and

Yurdagel, 2007). It was also shown in Amaranthus betacyanin pigments (Cai and Corke,

2000) and acai anthocyanins (Tonon et al., 2008) that increasing inlet temperature from

150°C up to 210°C caused pigments loss in spray drying due to their high sensitivity to high temperature exposure. The highest inlet temperature was only 170°C in our study, which was not very high compared to other studies. Focusing more on the inlet temperature effect, may explain why temperature did not present a significant effect on our PCC pigments yield. In addition, PCC (Figure 6.1) and pomegranate (Gómez-

Caravaca et al., 2013) anthocyanins have higher proportions of acylated pigments compared to non-acylated anthocyanins dominant berries (Seeram et al., 2006). Acylated anthocyanins had been shown to have higher stability under adverse conditions than the

149 non-acylated anthocyanins (Giusti and Wrolstad 2003; Wrolstad et al., 2005). Our observation may also suggested the acylated anthocyanins rich PCC was less affected by the high spray dry temperature, and thus presented more heat-resistant character compared to the berries.

6.4.2 Polymeric Color Based Pigment Quality

Percent polymeric color of PCC anthocyanins before and after various spray-drying conditions are shown in Table 6.1. The percent polymeric color of most PCC anthocyanins increased slightly but not significantly after undergoing the spray drying process. Only the PCC pigment spray dried with highest inlet temperature of 170°C presented significantly higher (p=0.019) percent polymeric color after drying than before, suggesting high inlet/outlet temperature may speed up the formation of polymeric anthocyanins during the spray drying process. In the present study, inlet temperature was the only factor that could significantly affect the polymeric color change between before and after spray drying, PCC anthocyanins matrix and amount of maltodextrin did not show significantly have impact on polymeric anthocyanins during spray drying (Table

6.1).

Temperature have been shown to play a major role in formation of anthocyanins polymeric color in wine fermentation (Gao et al., 1997), blood orange juice refrigerated storage (Choi et al., 2002), berries juice and puree pasteurization and storage

(Brownmiller et al., 2008; Hager et al., 2008a; Hager et al., 2008b), but not in spray drying. To our knowledge, very limited research data have been published about the percent polymeric color change of anthocyanins undergoing spray drying. Acetaldehyde-

150 mediated condensation, copigmentation, and self-association (Mazza and Francis 1995;

Gao et al., 1997; Wrolstad et al., 2005) are the major reactions considered to be responsible for formation of anthocyanins polymeric color. Gao et al., (1997) observed the new formed polymeric anthocyanins in their grape wine HPLC chromatograph.

Proteins (Ososanya et al., 2013; Rodriguez-Nogales et al., 2006), phenolic acids such as protocatechuic acid, vanillic acid, syringic acid, p-coumaric acid, caffeic acid, ferilic acid and chlorogenic acid (Kim et al., 2013; Montilla et al., 2011; Fernando Ramos-Escudero et al., 2012; Žilić et al., 2012), as well as other flavonoids such as rutin, kaempferol, quercetin, naringenin and hesperitin (Pedreschi and Cisneros-Zevallos 2007; Ramos-

Escudero et al., 2012; Kim et al., 2013) present in PCC may all participate in polymeric color formation during spray drying. However, none of the polymeric anthocyanin was observed in our spray dry PCC HPLC chromatogram (Figure 6.1). The relatively large molecular weight polymeric anthocyanins might have either been filtered before send for

HPLC run or did not elute through the C18 column under our experiment condition.

151

Figure 6.1 Comparison of purple corn cob anthocyanins HPLC profile before and after spray drying (Hot water extracted PCC pigments was mixed with 5% maltodextrin, spray-dried under inlet temperature of 150°C).

6.4.3 Color Properties Changes and Solubility

Most of previous studies were interested in the color of dried anthocyanin powders

(Ahmed et al., 2010; Cai and Corke, 2000; Ersus and Yurdagel, 2007; Ferrari et al.,

2012a; Kha et al., 2010; Nayak and Rastogi, 2010; Osorio et al., 2010). Their products presented vivid reddish color with intensive chroma as the pigments in powders were highly concentrated. However, a nice color in the powder form does not guarantee a satisfactory color when dissloving them into water. In the present study, we were focusing on the color of reconstituted solution, and how the reconstituted color appeared after processing compared to their initial status.

152

Before Spray Drying After Spray Drying ΔE pH L* c* h* %haze pH L* c* h* %haze 56.5* 52.8* 17.4 4.0* 64.7 43.7 16.6 1.7 130/89±1 4.1 4.2 12.3 (1.0) (1.1) (0.4) (0.1) (0.8) (0.8) (0.1) (0.1) Inlet/Outlet 57.2* 51.3* 16.3 3.7* 62.7 44.1 16.3 0.7 Temperature 150/105±1 4.1 4.2 9.1 (0.3) (0.4) (0.1) (0.2) (0.3) (0.8) (0.1) (0.1) (°C) 57.8* 51.7* 17.2 3.9* 63.8 43.9 17.7 1.5 170/119±1 4.1 4.3 9.8 (0.8) (0.6) (0.2) (0.1) (0.3) (0.3) (0.1) (0.1)

57.8* 51.5* 16.7 4.2* 72.6 35.9 16.1 0.9 2% 4.1 4.3 20.8 (0.4) (0.3) (0.3) (0.1) (1.1) (1.5) (0.1) (0.1) Maltodextrin 57.2* 51.3* 16.3 3.7* 62.7 44.1 16.3 0.7 Amount 5% 4.1 4.2 9.1 (0.3) (0.4) (0.1) (0.2) (0.3) (0.8) (0.1) (0.1)

153 (%) 57.5* 51.7* 17.4 4.4* 63.4 44.3 17.5 1.0 10% 4.1 4.2 9.5 (0.1) (0.2) (0.0) (0.2) (0.2) (0.2) (0.0) (0.1)

58.9* 49.1* 14.9* 56.2* 65.6 41.8 13.7 14.9 EtOH extract 4.2 4.1 10.0 (0.6) (0.8) (0.1) (0.4) (0.3) (0.5) (0.1) (0.2) Type of 57.2* 51.3* 16.3 3.7* 62.7 44.1 16.3 0.7 water extract 4.1 4.2 9.1 Matrix (0.3) (0.4) (0.1) (0.2) (0.3) (0.8) (0.1) (0.1) 53.2* 67.4* 24.7* 0.2 59.7 58.5 20.5 0.5 C18 purified 3.1 3.3 11.7 (0.6) (0.3) (0.1) (0.1) (0.3) (0.6) (0.2) (0.0) Color and %haze values are represented as mean (n=3), standard deviation were shown in the bracket below the means. The monomeric anthocyanins contents were between 0.31 and 0.33 mg Cy-3-glucoside equivalence/mL solution for all Before Spray Drying color measurements. * is assigned to statistical significant at 0.01 level on before-after color and %haze paired t-test.

Table 6.2 Color and %haze changes of purple corn (Zea mays L.) cob pigments before and after spray-drying of varying 1 types of matrix, inlet temperature and maltodextrin amount.

The color characteristics changes and %haze of the PCC solution before and after spray drying are shown in Table 6.2. Note that the changes in pH between the initial and reconstituted solutions were very minor (Table 6.2), so it is reasonable to believe that the color change between the initial and reconstituted solutions were mainly due to spray drying. As the hue number fell within the range of 13 to 25, both initial and reconstituted

PCC pigments solutions presented a red tone (Table 6.2). The hue of C18 purified PCC pigments and 40% ethanol PCC extracts decreased slightly after spray drying, indicating their reconstituted solution becoming redder as compared to the initial solutions. The hue of water PCC extract reconstituted solutions stayed almost the same as their initial solutions, the inlet/outlet temperature and the amount of maltodextrin did not have significant impacts on changing the hue of PCC water extract during the process (Table

6.2). After undergoing the spray dehydration, all the reconstituted solutions had significantly (p<0.01) higher lightness and lower chroma compared to their initial solutions (Table 6.2), suggesting the color became a little bit paler and less saturated.

Since all the spray drying trials in this study had at least 10% monomeric anthocyanins loss (Table 6.1), these change in lightness and chroma were expected. These changes on lightness and chroma in the reconstituted solutions were in agreement with spray dried blueberry (Jimenez-Aguilar et al., 2011), although their hue changes varied due to the interaction between blueberry anthocyanins and other compounds in the initial slurry.

The overall color changes before and after spray drying were evaluated with ΔE. When

ΔE is less than 1.5 the color difference between the two is usually undistinguished, while when ΔE falls between 1.5 and 5, the color difference can be identified by the trained

154 eyes (Obon et al., 2009). The difference in color becomes evident for most people when the ΔE is larger than 5 (Obon et al., 2009). As shown in Table 2, the color difference between the initial and reconstituted pigment solutions were all evident as the ΔE ranged from 9.1 to 20.8 among different trials in this study. These distinguishable color change after processing was mainly due to loss of anthocyanins, as the monomeric anthocyanins suffered at least 10% loss while the gain of polymeric color was less than 4% during spray drying (Table 6.1). With the highest yield of monomeric anthocyanins

(89.4±3.9%), the PCC water extract spray dried under 150°C together with 5% (m/v) maltodextrin presented the lowest ΔE of 9.1. On the contrary, the color difference was most evident (ΔE=20.8) when the PCC was spray dried together with 2% maltodextrin and the monomeric pigments retention was only 57.3±8.9%. In a spray dried jaboticaba extract study the final product was reconstituted into water to meet the soluble solids level prior processing. Their color difference was all less than 5 when maltodextrin was the carrier (Silva et al., 2013). Similarly, a ΔE around 5 was reported in spray dried blueberry pigments before and after processing (Jimenez-Aguilar et al., 2011). The ΔE in our study might be much less so that the color difference would not be that strong if we reconstituted the pigments powders based on the level of total soluble solids.

Different from most studies in which directly dissolved powders into solvent were used to evaluate their solubility (Yousefi et al., 2011; Fazaeli et al., 2012), this study used

%haze to rapidly estimate the water solubility of the spray dried PCC pigment powders.

Usually %haze less than 5% would be considered as a clear solution (Jing and Giusti,

2005), %haze larger than 8% would have visible suspended compounds in the system,

155 suggesting limited solubility of the solute. Most of our reconstituted pigment solutions gave good aqueous solubility with less than 2%haze (Table 6.2) except the 40% ethanol extract powders. The 40% ethanol extract reconstituted solution had dark color water insoluble powders suspending with 14.9±0.2% of haze (Table 6.2). A previous study investigated dark colored water insoluble compounds obtained from PCC ethanolic extracts and their pH-%haze curve suggested these anthocyanin-rich by-product could be anthocyanin-protein-tannin complex (Jing and Giusti, 2005). The haze of 0.5 mg/mL of these complexes became less than 5% when distributed in 36.5%-62.5% aqueous ethanol while the haze stayed as high as ~15% when suspended in water (Jing and Giusti, 2005).

The 40% ethanol extract reconstituted solution showed large value of haze was possibly due to presence of anthocyanin-protein-tannin complexes. These complexes were formed during PCC pigments extraction, entered into the spray drier and finally became powders in the PCC product. Since these anthocyanin-protein-tannin complexes have limited water solubility, they were not found in large numbers in water extracts. The water solubility for spray dried PCC water extract products was still satisfactory. Additionally, as shown in Table 6.2, most of the %haze of the reconstituted pigments were significantly (p<0.01) lower than in their initial solutions, suggesting the spray dehydration might improve water solubility of the PCC pigments. One explanation for this observation could be those relatively large molecular weight insoluble compounds tended to adhere to spray dryer wall, as the aids provided by the carrier agents such as maltodextrin in changing their hygroscopic and thermoplastic properties were very limited compared to other water soluble compounds. With insoluble complexes retained

156 in the cylinder, the proportion of soluble compounds in the product container increased, thus the haze of the reconstituted solution decreased. The only exception was thw C18 purified PCC pigments, which did not showe significantly solubility improvement because they already had a very low initial haze level (0.2±0.1%).

6.4.4 Anthocyanins HPLC Profiles Changes

Six major purple corn anthocyanins previously reported in the literature were also found in our PCC samples before and after spray drying (Figure 6.1). Based on comparison with our previous publication (Lao and Giusti, 2016), they were tentatively identified as cyanidin-3-glucoside, pelargonidin-3-glucoside, peonidin-3-glucoside, cyanidin-3-(6”- malonylglucoside), pelargonidin-3-(6”-malonylglucoside) and peonidin-3-(6”- malonylglucoside). The PCC anthocyanins composition did not change very much after spray drying process according to their HPLC chromatograph (Figure 6.1), regardless of pigment matrix, inlet/outlet temperature and amount of maltodextrin applied to the solution. In most of our experimental groups, the proportion of non-acylated anthocyanins in the PCC increased slightly after spray drying. Only at the inlet temperature 170°C dried PCC powders presented a slightly decease on %non-acylated pigments from 69.3±1.3% before dehydration to 68.7±1.5% after drying. But no statistically significance was found between before and after spray drying PCC HPLC chromatogram for %non-acylated pigments or individual anthocyanin. Our observation suggested the spray drying process did not make a major change in the PCC anthocyanins composition. None of the PCC pigment was more liable than others when exposed to the operation conditions applied in this thermal dehydration study.

157

Little research has been reported on the pigments HPLC profile change in the spray drying process. In a encapsulation of corozo fruit anthocyanins study, the authors also confirmed by HPLC-PDA that the pigment composition in the microcapsules was similar to that of the fruit without processing (Osorio et al., 2010). Fang and Bhandari (2011) ran bayberry polyphenol HPLC chromatograms before and after spray drying, as well as the polyphenol HPLC chromatograms after 6 months of storage under various water activity.

Their results showed almost no loss of bayberry anthocyanins right after the spray drying process, but the cy-3-glucoside in bayberry powders suffered higher degradation rate than other polyphenols during the storage test (Fang and Bhandari, 2011).

6.5 CONCLUSION

The choice of feeding matrix, inlet/outlet temperature and amount of maltodextrin applied significantly affected the yield and color quality of spray dried PCC anthocyanins product. The yield was majorly impacted by the amount of maltodextrin. An appropriate amount of carrier could help increase the pigments retention. High inlet/outlet temperature resulted in higher formation of polymeric anthocyanins during spray drying.

The choice of feeding matrix played critical roles in %haze of the reconstituted solutions, ethanol PPC could produce a hazy final product due to presence of anthocyanin-protein- tannin complexes. The PCC pigment composition did not change much during the spray drying. Hot water extracts spray dried with 5% (m/v) maltodextrin at 150°C inlet temperature gave the highest pigment yield and satisfactory water solubility with the least color changes. PCC powders produced using conditions provided in this study could

158 represent high quality natural alternatives to the use of synthetic dyes, helping food companies transition from artificial colorants to more consumer-friendly alternatives.

6.6 ACKNOWLEDGMENTS

This work was partially funded by Chinese Scholarship Council Fellowship. The authors would also like to acknowledge Agroindustrial S.A.C (Lima, Peru) for providing purple corn cob samples for this study.

159

Chapter 7: Investigation of Purple Corn (Zea mays L.) Anthocyanins and Protein Complexation Using Mid-Infrared Technology

7.1 ABSTRACT

During the industrial extraction of purple corn cob (PCC) pigments, dark colored complexes are formed and precipitate. Previous studies suggested these acidic water insoluble compounds might be anthocyanin-tannin-protein complexes. Better understanding this complexation and being able to quantify proteins in the anthocyanin matrix may help to improve the pigment recovery efficiency and colorant quality.

Infrared (IR) technology measures the dipole moment change of the molecule through molecular vibration, provides specific molecule fingerprint information. In this study, IR was applied to investigate the protein conformational alteration introduced by anthocyanin complexation. IR spectra of proteins, PCC anthocyanins, and their mixture was collected using a portable infrared spectrometer by applying a drop of sample onto an

ATR-IR diamond crystal. Spectra were analyzed by soft independent modeling of class analogy to characterize the complexation, and partial least square regression to develop a protein quantification model. Our study showed the primary driven force for anthocyanin- protein complexation formation was hydrogen bonding. Hydrophobic interaction and ion

160 chelation may also be involved into complex formation when the matrix pH was lower than 4.5. Based on the unique protein Amide bands (~1645 cm-1, ~1540 cm-1), the predictive model generated using both BSA and zein in this study had an R² = 0.970 and could be used to quantify the protein content with only 1.5 µL of sample. It could be applied for protein measurement up to 10 mg/mL with only 0.276 mg/mL error. The solubility performance of commercial PCC pigments powders appeared to correlated to the protein/anthocyanin ratio in the product.

Keywords: Mid-infrared, anthocyanin-protein complexation, multivariate analysis.

7.2 INTRODUCTION

Anthocyanins are a class of water-soluble phenolic compounds that provide red, orange, purple and blue color to plants. Known as potent antioxidants and strong anti- inflammatory agents, anthocyanins are believed to have potential to reduce the risk of cardiovascular disease, obesity, diabetes, cancer and chronic diseases (He and Giusti,

2010; Konczak and Zhang, 2004; Wallace, 2011). Purple corn (Zea mays L.), with one of the deepest purple shades among the plant kingdom, is rich in anthocyanins. The mean anthocyanin content of whole, fresh purple corn from Peru was around 16.4 mg/g, compared to 1.3 to 3.8 mg/g found in blueberries (Cevallos-Casals and Cisneros-

Zevallos, 2003). The high anthocyanin content and economic cost makes purple corn an ideal starting material for food colorant production.

However, unlike anthocyanins in other fruits and vegetables, the recovery of purple corn color is generally not efficient. In the industrial processing, large amount of pigment-rich

161 waste was generated during traditional economic hot acidified water or aqueous ethanol extraction (Jing and Giusti, 2007). These dark-colored wastes had low acidic water solubility, which limited their application in most aqueous food matrix (Jing and Giusti,

2005). Previous study suggested the pigment-rich wastes might be protein-tannin- anthocyanin complexation (Jing and Giusti, 2007). Similarly, dark colored precipitates were also observed in aqueous purple corn pigments solution in their storage period.

Since tannin generally has poor water solubility and has a low chance to be directly extracted from the plant tissue by water, the tannin in the complexes might be condensed by the extractable anthocyanins. Investigation of the protein-anthocyanin complexation pattern and quantification of protein content in anthocyanin-rich matrix would provide further understanding to anthocyanin-protein interaction, which may finally lead to solutions to the low efficiency purple corn industrial extraction and storage problems.

However, the straightforward colorimetric quantitation for proteins such as BCA and

Bradford methods could not be applied here because anthocyanins are interference compounds in these analyses (Compton and Jones, 1985; Smith et al., 1985). Infrared technology, on the other hand, measures change in the dipole moment of the molecule through molecular vibration, provides specific fingerprint information of the molecule

(Rodriguez-Saona and Allendorf, 2011). This makes infrared spectroscopy a valuable tool to investigate the proteins, even in the complicated biological matrices. The infrared spectrum has been shown in previous studies to provide information of protein chemical structure vibration (Von Germar et al., 2000), properties of protein neighboring group

(Barth, 2007), redox state of protein in a biological reaction (Nabedryk et al., 1990),

162 change in protein bond strength (Andersson and Barth, 2006), bond angels and conformational freedom (Mizuguchi et al., 1997), as well as hydrogen bonding and electronic field distribution (Dioumaev and Braiman, 1995). In present study, the presence of anthocyanin was taken into consideration by using infrared spectroscopy, as only the protein unique amide signals which come from peptide bond would be used for protein analysis.

The objective of this study was to investigate the purple corn anthocyanin-protein complexation pattern and develop an infrared model for rapid protein quantitation in anthocyanin-rich systems using Attenuated Total Reflectance infrared (ATR-IR) spectroscopy combined with multivariate techniques.

7.3 MATERIALS AND METHODS

7.3.1 Materials and Reagents

Dried purple corn cob (Zea mays L.) powders were kindly provided by Agroindustrial

S.A.C (Lima, Peru). Ten different ready-to-use commercial purple corn pigment powders were provided by Zanaceutica E.I.R.L. (Lima, Peru), Alicorp S.A.A. (Lima, Peru),

Agroindustrial S.A.C (Lima, Peru) and Globenatural International S.A. (Chorrillos-Lima,

Peru). Purified corn protein α-zein was purchased from Acros Organics (Fair Lawn, NJ,

USA). Bovine serum albumin (BSA) was purchased from Bio-Rad (Hercules, CA, USA).

All other solvents and chemicals were purchased from Fisher Scientific (Fair Lawn, NJ,

USA).

163

7.3.2 Purified Anthocyanins Preparation

7.3.2.1 Anthocyanins extraction

Weighted purple corn powders was placed in the blender and 50 mL of 0.01% (v/v) 6N

HCl acidified 70% aqueous acetone was added. The slurry was blended at room temperature for 3 min and filtered through Whatman No.4 filter paper using a Buchner funnel. The cake was re-extracted until a faintly colored solution was obtained. Two volumes of chloroform were added after the filtrate was transferred to a separatory funnel. The samples were stored at room temperature overnight before evaporating the upper colored portion in a rotary evaporator at 40℃ under vacuum. Finally the remaining extract was made up to known volume with 0.01% HCl acidified water (Rodriguez-Saona and Wrolstad, 2001).

7.3.2.2 Anthocyanins purification

The extracted anthocyanins were passed through MCX mixed-mode cation-exchange cartridge (Waters Corporation, Milford, MA, USA) and C18 Sep-Pak cartridge (Waters

Corporation, Milford, MA, USA) for purification. The purpose of using MCX purification was to separate anthocyanins from the matrix of sugars, acids and other phenolic compounds. The purpose of applying C18 purification was to remove salts introduced during MCX purification out form the purified PCC anthocyanins. In MCX cartridge purification, the cartridge was first activated by 2 volumes of methanol, then 2 volumes of 0.1% (v/v) trifluoroacetic acid (TFA) acidified water before loading the sample. Two volumes of 0.1% TFA acidified water was applied to wash sugars, acids, and other non-phenolic compounds away. Afterwards, non-anthocyanin phenolics were

164 removed by 2 volumes of 0.1% TFA acidified methanol. Following these two washing steps, the anthocyanins were eluted by 1 volume of 1% (v/v) NH4OH in methanol followed by 1 volume of 1% NH4OH in water/methanol (40:60, v/v). Finally, formic acid was added to bring the basic eluate to a pH around 1, in order to avoid anthocyanin degradation (He and Giusti, 2011). Methanol in the MCX elute was evaporated in a rotary evaporator at 40°C under vacuum, solubilized into 0.01% (v/v) HCl acidified water before sending to C18 cartridge. In C18 purification, the cartridge was first activated by 2 volumes of methanol, then 2 volumes of 0.01% (v/v) HCl acidified water before loading the sample. Two volumes of 0.01% HCl acidified water was applied to wash the remaining salts in the MCX elute. Then 0.01% (v/v) HCl acidified methanol was applied to elute the anthocyanins out of the cartridge (Rodriguez-Saona and Wrolstad, 2001).

Methanol in the C18 elute was evaporated in a rotary evaporator at 40℃ under vacuum before making to a known volume with 0.01% HCl acidified water. The concentration of purified purple corn pigments was determined by the pH differential method (Giusti and

Wrolstad, 2001).

7.3.2.3 Purified anthocyanins checking

A reverse-phase high performance liquid chromatograph (HPLC) system (Shimadzu

Corporation, Tokyo, Japan) equipped with a SPD-M20A prominence diode array detector and LCMS-2010EV liquid chromatograph mass spectrometer (MS) was applied to check if the purified purple corn anthocyanins were free of proteins. The column was reversed- phase 3.5µm Symmetry C18 column (4.6×150mm, Waters Corp., MA) fitted with a

4.6×22mm Symmetry 2 micro guard column (Waters Corporation, MA). The solvents

165 used were A: 4.5% formic acid in water, and B: acetonitrile. Separation was achieved by using a linear gradient from 5 to 25% solvent B in the first 25 min, keeping 25% B from

25 to 30 min, followed by linearly increasing B from 25% to 40% during 30 to 35 min, with a flow rate of 0.8 mL/min. Purple corn anthocyanins chromatograms before and after MCX and C18 cartridge purification were collected at 280 nm and 520 nm, with MS helping identifying anthocyanins.

7.3.3 Determination of Target Region for Protein Analysis

7.3.3.1 Comparison of mid-infrared spectrum for protein and anthocyanin

The portable 4500a FTIR unit (Agilent Technologies Inc., Santa Clara, CA) was used in all infrared related experiments in this chapter. It was equipped with a Michelson interferometer, Zinc selenide beam splitter, low-powered solid-state laser, wire-wound element infrared source and thermoelectrically cooled deuterated triglycine sulfate detector. The instrument was interfaced with a triple reflection diamond crystal attenuated total reflectance (ATR) accessory with a 2 mm diameter sampling surface and

200 μm active area providing 6 μm effective penetration depth for the energy at mid infrared (MIR) region.

Two different kinds of proteins, 1 mg/mL BSA in water and 1 mg/mL zein in 70% (v/v) aqueous ethanol, as well as 1.125 mg/mL purified purple corn anthocyanins in 0.01%

(v/v) 6N HCl acidified water were used for preliminary MIR scan to explore the MIR pattern difference between the two classes of compounds. Each of the three solutions was vortexed for 30 seconds before applying 1µL onto the diamond ATR crystal. The liquid droplet was vacuum dried till a beautiful thin film was formed. The MIR spectra was

166 collected using MicroLab software (Agilent Technologies, Santa Clara, Calif., U.S.A.) operating in the wavenumber ranges from 4000 to 700 cm−1 with resolution of 4 cm−1, and 64 scans were co-added to improve signal to noise ratio. The collected spectra were compared and the common region for proteins that distinguished them for purple corn anthocyanins was identified.

7.3.3.2 Advantage of applying infrared over traditional colorimetric methods in protein quantification

The objective of this part was to see if the protein standard selection would affect final protein quantification using infrared and other traditional spectrophotometric methods. A series of different concentration (0.1, 0.2, 0.5, 1.0, 2.0, 5.0, 10.0 mg/mL) of BSA in water and the identical series of zein in 70% (v/v) aqueous ethanol were applied. Both proteins were used to construct quantification standard curve via three different approaches: UV at

280 nm quantification, Bradford method, and MIR-ATR.

7.3.3.2.1 Ultraviolet light at 280 nm

Briefly, 200µL of each protein standard was applied in 96 well plate and sent to Spectra

Max 190 plate reader (Molecular Devices, Sunnyvale CA) to collect their absorbance at

280 nm. Reading was done in triplicate for each samples.

7.3.3.2.2 Bradford method

Briefly, 100µL of each protein standard was pipetted into a 15 mL tube and mixed with 5 mL of Bradford dye, pure water or 70% (v/v) aqueous ethanol was served as blank. The tubes were well-vortexed and incubated at room temperature for about 45 minutes.

167

Measurement of each sample was taken using Shimadzu UV-2450 UV-visible spectrophotometer (Shimadzu Corporation, Tokyo, Japan) at 595 nm, in triplicates.

7.3.3.2.3 Mid-infrared spectroscopy

Each of protein standard was vortexed for 30 seconds before applying 3µL to the diamond ATR crystal. The liquid droplet was vacuum dried until a beautiful thin film was formed. The MIR spectrum of each sample was collected and the intensities of

Amide I and Amide II bands were used to build the standard curve. MIR spectra of each sample was collected in triplicates.

7.3.3.3 Multivariate analysis to construct protein quantification prediction curve using the whole Amide region.

The spectral data collected in 7.3.3.2.3 was analyzed by Partial Least Squares Regression

(PLSR) to generate calibration models using Pirouette 4.0 software (Infometrix Inc.,

Woodville, WA). The idea of PLSR is to compress large number of variables into lower dimensions with noise reduction to provide an accurate and reproducible calibration model (Wold et al., 2001). Since the identical concentration of BSA and zein showed similar properties in the Amide region, the final PLSR model was constructed using both data, to provide a more universal approach for protein quantification. Spectra data underwent normalization (100) and smooth (35) transformation, model were cross validated and optimal factors with satisfactory prediction was selected.

168

7.3.4 Characterization of Anthocyanin-Protein Complexation using Infrared

Spectroscopy

7.3.4.1 The effect of anthocyanins addition on the proteins

In this part, different amount of purified purple corn anthocyanins were introduced to constant level of BSA, to explore the protein infrared spectrophotometric change due to the formation of the anthocyanin-protein complexes. To be specific, 1 mg/mL (~15 μM)

BSA was mixed with purified purple corn anthocyanins in molar ratios of 1:0, 1:1, 1:2,

1:3, 1:4, 1:5, 1:6, 1:7, 1:8, 1:9, 1:10. A drop (3 µL) of each mixture was applied onto

ATR crystal after preparation, scan was collected in at least 5 replications.

7.3.4.2 The effect of protein addition on the anthocyanins

In this part, different levels of BSA were added into constant amount of purified purple corn anthocyanins. In the aqueous matrix, 75 mg/L (~15 μM) purified purple corn anthocyanins was mixed with BSA in molar ratios of 1:0, 1:0.01, 1:0.025, 1:0.05, 1:0.1,

1:0.25, 1:0.5, 1:1, 1:2, 1:5. A drop (3 µL) of each mixture was applied onto ATR crystal after preparation, scan was collected in at least 3 replications.

7.3.4.3 FTIR spectroscopy

Each of the anthocyanin-protein mixture was vortexed for 30 seconds before applying a drop to the diamond ATR crystal. The liquid droplet was vacuum dried until a thin film was formed. The infrared spectra of each mixture was collected in the wavenumber ranges from 4000 to 700 cm−1 with resolution of 4 cm−1, and 64 scans were co-added to increase signal to noise ratio.

169

The collected spectra data was normalized (100), smoothed (35) and analyzed by the soft independent modeling of class analogy (SIMCA), using the chemometrics modeling software Pirouette (v4.0, Infometrix Inc., Woodville, Wash., U.S.A.), to investigate the qualitative and possible quantitative characteristics of the anthocyanin-protein complexes. SIMCA is a classification algorithm based on principal component analysis that is applied to each category of interest separately to build the classification model.

7.3.5 Multivariate Analysis for Infrared Quantification Model Development and

Quantification of Protein Content in Purple Corn Pigments Crude Extracts

7.3.5.1 Development of infrared protein quantification model in anthocyanin rich matrix

Two different proteins, BSA and zein, were applied in the protein quantification model development. A series of protein-anthocyanin mixture was prepared under two different solvent matrices for more broad future application. The protein content in mg/mL was the same for identical anthocyanin: protein molar ratio, in both matrices, In the aqueous matrix, 75 mg/L (~15 μM) purified purple corn anthocyanins was mixed with BSA in molar ratios of 1:0, 1:0.01, 1:0.025, 1:0.05, 1:0.1, 1:0.25, 1:0.5, 1:1, 1:2, 1:5. Similarly, in the 70% (v/v) aqueous ethanol matrix, 225 mg/L (~45 μM) purified purple corn anthocyanins was mixed with zein in identical molar ratios. A drop (1.5 μL) of each mixture was sent to IR scan after preparation. The spectra was collected in at least 3 replications. The spectral data collected in were analyzed by Partial Least Squares

Regression (PLSR) to generate calibration models using Pirouette 4.0 software

(Infometrix, Woodville, WA).Second derivative (Savitzky-Golay second-order polynomial filter with a 35-point window), smooth (35) and normalize (100) functions

170 were used to transformed the spectral data to resolve peak overlap and eliminate baseline shifts. Cross-validated (leave-one out approach) PLSR algorithm was used to develop models and the number of optimal factors was determined. Outlier diagnostics, standard error of cross-validation (SEV) and correlation coefficient (R2) were used to evaluate the quantification model.

7.3.5.2 Quantification of protein contents in extracts of commercial purple corn pigment powders

A drop (1.5 µL, after different levels of enrichments) of C18 purified various commercial purple corn pigment extracts in methanol was applied onto the diamond ATR crystal of

4500a FTIR unit (Agilent Technologies Inc., Santa Clara, CA). Spectra of the pigment extracts were collected in the wavenumber ranges from 4000 to 700 cm−1 with resolution of 4 cm−1, and 64 scans were co-added. Model developed in 7.3.5.1 was applied to quantify the protein content in each extract through Pirouette 4.0 software. The obtained

MIR predicted protein contents of the commercial purple corn extracts were compared with their water solubility, to explore potential correlation between anthocyanin-protein complexation and water solubility.

A crude assessment of commercial purple corn pigment powders water solubility was made visually based on the precipitation observed 30 seconds after mixing with acidified water. Around 0.1g of each pigment powders was distributed into 50 mL of pH3.5 HCl acidified water, thoroughly mixing with vortex for 30 seconds before let stand in the rack for precipitates observation. No precipitation = very good solubility, little solubilized color and heavy visible precipitation = very poor solubility.

171

7.4 RESULTS AND DISCUSSION

7.4.1 Preparation of Purified Purple Corn Anthocyanins

The HPLC profiles of purple corn anthocyanins before and after MCX-C18 cartridge purification are shown in Figure 7.1. Anthocyanins have maximum absorbance around

520 nm while protein and phenolic compounds have high absorbance at 280 nm. As shown in Figure 7.1, unpurified purple corn extract may contain proteins or other phenolics besides anthocyanins. However, after MCX-C18 cartridge purification, the chromatograms at 520 nm and 280 nm could nearly overlap each other (Figure 7.1). This suggested that the purple corn anthocyanin extract went through the MCX-C18 purification approach was free of proteins and other phenolic compounds, ready to be applied in the following experiments.

Figure 7.1 Purple corn anthocyanins HPLC chromatograms at 520 nm and 280 nm before and after MCX-C18 cartridge purification. 172

7.4.2 Determination of Target Region for Protein Analysis

7.4.2.1 Comparison of mid-infrared spectrum for protein and anthocyanin

Figure 7.2 shows the MIR spectra of BSA, zein and purified purple corn anthocyanin, in which marked out the most important peaks to discriminate the two compounds. There was a very distinctive pattern difference between anthocyanin and proteins. A small shoulder at around 1720 cm-1 was found in the purified anthocyanin MIR spectra. It was assigned to the carbonyl C=O stretching of protonated carboxylic acids (Coates, 2000; He et al., 2007). The 1600 cm-1 band in anthocyanins was associated with H-bonding

(Coates, 2000; He et al., 2007).The purified anthocyanin had unique small peak pattern at

1450–1510 cm-1, which are assigned to C=C–C aromatic ring stretching (Coates, 2000;

Fernández and Agosin, 2007). The 1400 cm-1 peak of anthocyanins was reported to be corresponded to phenolic hydroxyl bending (Coates, 2000). The band at 1285 cm-1 was also reported to be arising from the pyran-derived ring structure (Edelmann and Lendl,

2002). The several bands at 950-1225 cm-1 and 700-900 cm-1 regions were commonly found in plant phenolics (Fernández and Agosin, 2007; Fragoso et al., 2011a, 2011b; He et al., 2007), they were mainly contributed by aromatic C-H vibration (Coates, 2000;

Fernández and Agosin, 2007). The band at 1086 cm-1 was reported to be associated with

C-O-C stretch in carbohydrates such as glucose (Coates, 2000; Fragoso et al., 2011b).

There was a shared region between the two classes of compounds around 3600 cm-1, which correlated to the O-H group vibration (Coates, 2000). The broad band in anthocyanin suggested that the majority of the hydroxyl groups were in the H-bonded state, while the narrow peaks in both proteins suggested that the O-H groups were mostly

173 in nonbonded status (Coates, 2000). Very different from anthocyanins, a triple bands distribution pattern (~1657, ~1540, ~1340-1415 cm-1) was observed in the IR spectra of the BSA and zein. The infrared spectra of both proteins consisted with previous protein

IR studies (Barth, 2007; Rodriguez-Saona and Allendorf, 2011; Wang et al., 2015). They were mainly contributed by the protein backbone amide vibrations: Amide I (~1650 cm-1) which arises mainly from the C=O stretching, Amide II (~1550 cm-1) which is caused by both NH plane bending and CN stretching, and Amide III (1400-1200 cm-1) corresponding to NH bending and CN stretching. Therefore, the Amide I and II pattern observed in the mid-IR region which was representative for BSA and zein but weak for anthocyanin could be regarded as competitive candidates for protein identification and quantitation in anthocyanin-rich matrix.

174

Figure 7.2 Mid-infrared spectra of purified purple corn anthocyanins, BSA, and zein by mid-IR spectroscopy.

175

7.4.2.2 Advantage of applying infrared over traditional colorimetric methods in protein quantification

Selection of protein standard is critical for traditional colorimetric quantification. Use of known amount of analyte to serve as standard is highly recommended, because otherwise the various protein composition may introduce errors in the quantification. The standard curves constructed by BSA and zein using three different approaches were shown in

Figure 7.3. The traditional UV at 280 nm and the Bradford methods showed very different shapes and slopes for identical amounts of two protein series, while in IR approach, the shapes and slopes for both proteins were very similar (Figure 7.3). These results were in agreement with a previous study that comparing protein quantifications of

IR through Amide I and UV at 280 nm. In that study three different sources of proteins were used, BSA, protein A, and rabbit IgG (Strug et al., 2014). The reason for dramatic slope difference in the traditional colorimetric methods was, only certain kinds of amino acids could serve as chromophores in these quantitative measurement. In the UV measurement, the major chromophores were the tyrosine and tryptophan side chains while in the Bradford method, the major color contributors were arginine and lysine side chains (Lovrien and Matulis, 1995). Different proportions of the key color formation structures in BSA and zein resulted in their slope difference when producing the standard curve. On the contrary, the IR approach based on the universal peptide bond structure, which would lead to a more general pattern in protein quantification regardless of structure or size difference. In addition, since the side chain can hardly affect the Amide I and II signals (Barth, 2007), the variation of protein composition due to side chains could

176 be partially eliminated. In fact, the standard curves constructed by BSA and zein through

IR approach were closed enough that the structure difference between the two proteins could be ignorable (Figure 7.3). These results suggested a promising advantage of infrared total protein quantification that the analysis would not be affected by the vast variation of protein composition and size.

7.4.2.3 Multivariate analysis to construct protein quantification prediction curve using the whole Amide region.

In order to take full advantage of the protein fingerprint information and better identify protein from the complicated biomatrix infrared spectra, multivariate analysis was introduced to construct protein quantification standard curve using the whole Amide region (1100-1700 cm-1). As mentioned previously, the protein source difference did not affect IR quantification, this multivariate model was generated combining both BSA and zein data. As shown in Figure 7.4, the generated predictive model had nice linear shape with R2=0.978 using only 4 factors. Cross validation suggested the model would work well in quantifying proteins ranged from 1.0 mg/mL to 10.0 mg/mL with 0.509 mg/mL prediction error. Examination to the regression vector indicated the bands contributing the most to the predictive models were 1657 cm-1, 1550 cm-1, 1409 and 1269 cm-1, which are associated with Amide I, Amide II, and Amide III vibrations, respectively (Barth,

2007).

177

178

Figure 7.3 Comparison of protein quantification approaches using ultralight at 280nm (A), Bradford method (B), and mid infrared spectroscopy (C and D). Standard curves were constructed using both BSA and zein. Unlike traditional colorimetric methods, MIR presented identical178 slopes for different protein sources.

Figure 7.4 Partial least squares regression (PLSR) plots and regression vector based on the infrared spectra of BSA and zein using FT-MIR spectroscopy.

179

7.4.3 Characterization of Anthocyanin-Protein Complexation using Infrared

Spectroscopy

7.4.3.1 The effect of anthocyanins addition on the proteins

Different levels of purified purple corn anthocyanins were added into 15 μM of BSA to qualitatively investigate how the formation of anthocyanin-protein complexation modified the protein secondary structure. As shown in Figure 7.5 infrared spectra, the

Amide I band was altered as the anthocyanins were added. The changes in Amide I band observed in this study was in agreement with a previous research on the complexation of peanut protein and grape, black currant, and elderberry juice (Plundrich et al., 2014).

Usually, formation of protein-polyphenol complex altered protein FT-IR profiles in the forms of increase/decrease band intensity, band shift, band width change, and/or new band appearance (Plundrich et al., 2014). In our study, the Amide I band shifted to lower wavenumber as the more and more pigments were added (Figure 7.5). The Amide I band shift often indicates the change in H-bonding (Barth, 2007; Plundrich et al., 2014). It was suggested that the major driving force of the BSA-anthocyanins complex formation might be H-bonding. In addition, the Amide I intensity decreased as pigments were introduced to the matrix (Figure 7.5). These changes in band position and intensity suggested the formation anthocyanin-BSA complexation. Other BSA spectra changes due to anthocyanins addition were the band at around 1720 cm-1 and 1610 cm-1. They were assigned to the carbonyl C=O stretching of protonated carboxylic acids in anthocyanins

(Coates, 2000; He et al., 2007), and water or H-bonding (Coates, 2000; Plundrich et al.,

2014), respectively.

180

Figure 7.5 Soft independent modeling of class analogy (SIMCA) classification plots, infrared spectra at Amide I region, and discriminating power based on mid IR spectra for different levels of purple corn anthocyanins addition into 15 μM BSA.

181

Soft independent modeling of class analogy (SIMCA) was applied to explore the possible quantitative pattern in the BSA-anthocyanin complexation. As shown in SIMCA plot of

Figure 7.5, different amount of anthocyanins and their BSA complexations could be classified with only 3 factors. Discriminating power plot (Figure 7.5) showed the predominant class separation contributors were wavenumber 1653 cm-1 and 1610 cm-1, which correlated to Amide I and H-bonding, respectively (Coates, 2000). Information provided in the discriminating power plot also supported the alternation of Amide I signal and formation of H-bonding due to BSA-anthocyanin complexation. As shown in Figure

7.5 SIMCA plot, the class cluster distance became shorter and shorter as the anthocyanins molar ratio increased. The distance between 1:6 (BSA: anthocyanin molar ratio) and 1:7 clusters was much narrower than the lower molar ratio groups. Interestingly, the 1:9 and

1:10 cluster behaved differently to other clusters, suggesting something changed in the complexation. Instead of following the direction all previous clusters moved up, the 1:9 and 1:10 cluster moved in a perpendicular direction (Figure 7.5). This moving pattern change might suggest the BSA-anthocyanin complexes appeared to saturation when the molar ratio was about 1:7. The changes at Amide I and H-bonding signals were very limited as the BSA: anthocyanin molar ratio reached 1:7 (Figure 7.5). The SIMCA output suggested the BSA-anthocyanin complexation through H-bonding became saturated when the BSA: anthocyanin molar ratio approached approximately to 1:7.

7.4.3.2 The effect of protein addition on the anthocyanins

Different levels of BSA were added into 15 μM of purified purple corn anthocyanins to qualitatively investigate how the formation of anthocyanin-protein complexation altered

182 anthocyanins IR spectra. As shown in Figure 7.6 infrared spectra, the shape and intensity of anthocyanins spectra was altered into various degree as BSA was introduced to the matrix. The anthocyanin 1720 cm-1 and 1600 cm-1 bands were swallowed by the BSA signature Amide signals as more and more proteins were added. These teo bands were assigned to the carbonyl C=O stretching of protonated carboxylic acids (Coates, 2000; He et al., 2007) and H-bonding (Coates, 2000; Fernández and Agosin, 2007). Another sharp band for anthocyanin at 1086 cm-1, which was contributed by carbonyl vibrations

(Coates, 2000; Fragoso et al., 2011b), showed decreased in the intensity and band width as BSA was added (Figure 7.6). Again, the changes in band intensity and width suggested formation of BSA-anthocyanin complexation (Plundrich et al., 2014). Similar reducing bandwidth phenomenon was also observed in peanut protein complexation with plant phenolics (Plundrich et al., 2014). The narrower band width suggested the conformational freedom of the carbonyl groups in anthocyanins were restricted due to complexation with BSA (Kong and Yu, 2007).

Soft independent modeling of class analogy (SIMCA) was applied to explore the critical factors for the formation of BSA-anthocyanin complexation. As discussed in 7.4.3.1, shown in Figure 7.6 SIMCA plot, different amount of BSA and their anthocyanin complexations was classified with only 3 factors. Discriminating power plot (Figure 7.6) showed the predominant class separation contributors were wavenumber 1086, 1125,

1322, 1372, 1456, 1491, and 1657 cm-1, which corresponded to the carbonyl (Coates,

2000; Fragoso et al., 2011b), aromatic rings C-H (Coates, 2000; Fernández and Agosin,

2007), C−O stretch from pyran-derived ring structure (Coates, 2000; Fernández and

183

Agosin, 2007), deprotonated carboxylate group (Barth, 2000; Gerwert et al., 1989),

C=C−C aromatic ring stretch (Coates, 2000; Fernández and Agosin, 2007), aromatic rings double bonds (Coates, 2000; Fernández and Agosin, 2007), and Amide I (Barth,

2007), respectively. The carbonyl band suggested the BSA-anthocyanins complexation formation was driven by the carbonyl groups involved H-bonding. Since the bands at

1125, 1322, 1456 and 1491 cm-1 were mainly associate with alternation on functional groups related to anthocyanin aromatic rings, the discriminating power plot suggested the aromatic rings centered hydrophobic interaction could play an important role in the BSA- anthocyanin complexes. The complexation redistributed the electrons in the aromatic ring region, resulted in the change of IR vibration response and the restriction of structural freedom. The band 1372 cm-1 was reported to be associated with ion chelation (Wang,

2014). It was used to investigate Ca2+ binding to proteins previously (Fabian et al., 1996).

The band might suggest the chelation between negatively charged BSA carboxylic acid side chains (aspartate and glutamate) and the protonated anthocyanin flavylium. Other important bands that contributed to class separation included Amide I (1657 cm-1) and

Amide II (1538 cm-1) due to different level of total proteins, as well as 1600 cm-1 which was produced by H-bonding. Similar to part 7.4.3.1, the clusters appeared to behaved differently as the BSA:anthocyanin molar ratio was higher than 1:0.5 (Figure 7.6). The

SIMCA plot (Figure 7.6) appeared to suggest the complexation involving hydrophobic interaction and ion chelation became saturated when the anthocyanin: BSA molar ratio came to 1:0.5 (BSA: anthocyanin molar ratio = 1:20).

184

7.4.3.3 The effect of matrix pH on the BSA-anthocyanin complexation formation

Previous studies showed BSA may present in five conformations when placed under different pH environments (Bhattacharya et al., 2011; Carter and Ho, 1994; Peters, 1985).

They are the E-form (Extended; pH<3), the F-form (Fast; pH 3–4.5), the N-form

(Native/Normal; pH 4.5–7), the B-form (Basic; pH 7–8.5) and the A-form (Aged; pH>8.5). As shown in Figure 7.7, BSA is in its natural state and highly compact into a heart shape N-form when the pH is slightly acidic to neutral. In the N-form, the hydrophobic groups are tightly packed inside while the hydrophilic side chains tend to come to the outside surface (Carter and Ho, 1994; Michnik et al., 2005). As the pH decreased into the range of 3 to 4.5, the partly open F-form (Figure 7.7) becomes predominant, in which some of the hydrophobic groups begin to stretch out and expose to the outer environment (Carter and Ho, 1994). The BSA conformation becomes even more extended to the E-form (Figure 7.7) as the pH continues to decrease.

Figure 7.7 BSA conformational isomers under various pH (Carter and Ho, 1994). 186

The conformation of BSA could affect its complexation with anthocyanins. In the case of adding anthocyanins into constant amount of BSA, since BSA was predominant in the background matrix, the starting environmental pH was around neutral. The system pH dropped slightly as anthocyanins were added. Since BSA was in the tightly packed N- form, the anthocyanins could only interact with the surface hydrophilic side chain through H-bonding to form the complexation. The surface active side chain spots available for H-bond formation were limited, so only 7 anthocyanin molecular might be able to interact with a BSA and the complexation got saturated soon (Figure 7.5). While in the case of adding BSA into constant amount of anthocyanins, because anthocyanins was predominant in the starting point, the initial pH was acidic (pH~4). The matrix pH increased slightly as BSA was introduced. The new added BSA was in the partially open

F-form once entered the matrix, so the anthocyanins aromatic rings quickly located the hydrophobic region in the BSA and formed the complexation through hydrophobic interaction. In addition, the pH~4 environment facilitated the deprotonation of aspartate and glutamate side chains. The negatively charged carboxylic groups in proteins could chelate the positively charged anthocyanin flavylium, to form the complexation. Lastly, the H-bonding available spots were still available for additional pigment binding. Thus in the BSA addition to constant amount of anthocyanins case, the BSA-anthocyanin complexation could be achieved through three different manners: hydrophobic interaction, chelation, and H-bonding. This might explain the complexation became saturated when anthocyanin: BSA molar ratio was up to 1:0.5 (BSA: anthocyanin = 1:20)

(Figure 7.6) in this experimental setting.

187

7.4.4 Multivariate Analysis for Infrared Protein Quantification Model Development and Quantification of Protein Content in Purple Corn Pigments Crude Extracts

7.4.5.1 Development of infrared protein quantification model in anthocyanin rich matrix

Collaborating both BSA and zein, aqueous and alcoholic matrices, the infrared protein quantification model developed through our approach was expected to be able to predict protein levels in various anthocyanin-rich plant extract matrix in a more universal way.

Simply by applying 1.5 µL of protein standard, as shown in Figure 7.8, the generated predictive model for infrared protein quantification in pigment-based environments had nice linear shape with R2=0.970 using only 4 factors. The optimal factor selected for model was able to represent enough information to build the calibration model, and at the same time, was not be too large to include structural information from other sources. Our selected model explained more than 99.9% variables in the wavenumber range of 1700 to 1500 cm-1. Cross validation suggested the model would work well in quantifying proteins in anthocyanin-rich matrix, up to protein level of 10.0 mg/mL, with only 0.276 mg/mL prediction error. In agreement with our quantification model established solely by pure proteins in part 7.4.2.3, examination to the regression vector indicated the bands contributed the most to the predictive models were Amide I at 1658 cm-1, and Amide II at

1562 cm-1 (Barth, 2007). Again, H-bonding related band at 1597 cm-1 which indicated the complexation of anthocyanins and proteins, was also critical in the predictive model.

188

189

7.4.5.2 Quantification of protein contents in extracts of commercial purple corn pigment powders

The MIR protein quantification model developed above was applied to predict protein contents in ten different commercial purple corn pigment powder extracts. A C18 purification was applied to all commercial purple corn pigment extracts to simplify the matrix MIR spectra by removing other interfering carbohydrates. As shown in Table 7.1, the protein/anthocyanins ratio in the powders appeared to correlate to the difference in water solubility. High protein/anthocyanins ratios tended to present poor water solubility

(Table 7.1). The water solubility of the commercial powders was overall good when the protein/anthocyanin ratio stayed in a low level (<10). The water solubility became poor as the protein/anthocyain ratio in the commercial powders reached to around 17 (Table

7.1). The solution haze became heavier as the protein/anthocyanin ratio increased to

34.73 (Table 7.1). The observation of high protein/anthocyanin ratio tended to show poor solubility property in this study appeared to suggest a positive correlation between the anthocyanin-protein complexation and limited water solubility in purple corn pigment commercial product.

190

Anthocyanin Protein Protein/pigment Water Sample content content ratio solubility (mg/mL) (mg/mL) Ag -new 0.061±0.002 0.67±0.01 11.01 good Ag-old 0.028±0.001 0.17±0.0 6.05 fair A-Anthocyanin 0.191±0.001 0.53±0.04 2.78 very good A-Anthocyanin MS2 0.208±0.003 1.23±0.03 5.91 very good Alicorp-11% 0.185±0.002 1.55±0.05 8.36 very good G-WI 0.043±0.001 0.26±0.03 5.98 good G-MS1 0.049±0.001 1.22±0.05 24.77 poor G-MS2 0.066±0.002 0.55±0.02 8.33 very good Z-S1 0.020±0.001 0.71±0.03 34.73 very poor Z-S2 0.040±0.00 0.71±0.00 17.57 poor Table 7.1 Water solubility, anthocyanin and protein contents of ten different commercial purple corn pigment powders. Anthocyanin content was measured by pH differential method, protein content was measured by mid infrared predictive model.

7.5 CONCLUSION

FT-MIR spectroscopy allowed for rapid (about 2 min preparation and analysis time) characterization of protein-anthocyanin complexation and quantification of proteins in pigment-rich matrix. Hydrogen bonding was suggested by the infrared discrimination power to be the major driving force for the anthocyanin-protein complexation formation in neutral environment. When the matrix became more acidic (pH<4.5), hydrophobic interaction and ion chelation might also involve into the anthocyanin-protein complexation. The MIR protein predictive model developed in this study provided a simple, rapid approach to quantify protein levels in an anthocyanin-rich matrix using only

1.5 µL of pigmented solution. The model might help predict the water solubility

191 properties of the pigments products and has great potential to provide an easy quality control method for rapid product quality evaluation.

7.6 ACKNOWLEDGEMENTS

The author would like to thank Dr. Luis E Rodriguez-Saona for providing access to the infrared instruments in his lab, his guidance throughout the IR project, and help in the data analyzing. The author would also like to thank Dr. Marcal Plans Pujolras for his technical advice and assistance in MIR instrument set-up and training.

192

Chapter 8: Overall Conclusion

Rich in anthocyanins and other phenolic compounds, purple corn (Zea mays L.) anthocyanins could be great value-added natural pigments for food coloring.

To produce high quality food application anthocyanins powders from purple corn cob

(PCC), extraction solvent, acidity, spray drying input matrix, inlet/outlet temperature and the amount of carriers applied in the processing have to be carefully considered and well- controlled. In the extraction step, solvent selection played a more significant role in anthocyanins recovery than acidity. Our study suggested extraction performed by 0.01%

(v/v) 6N HCl acidified aqueous ethanol (water: ethanol ratio around 1:1) could produce high level of monomeric anthocyanins and phenolics with relatively low level of polymeric color. In the spray drying step, the choice of feeding matrix played critical roles in %haze of the reconstituted solutions, the yield was majorly impacted by the amount of maltodextrin, and high inlet/outlet temperature resulted in higher formation of polymeric anthocyanins during spray drying. Overall, 70°C hot water extracts spray dried with 5% (m/v) maltodextrin at 150°C/105°C inlet/outlet temperature gave the highest pigment yield and satisfactory water solubility with the least color changes.

PCC anthocyanin quantification results were dependent on the method chosen and the differences among various methods are likely due to the differences in their specificity. 193

Among the four commonly used methods, the total anthocyanin method produced the highest value, followed by the pH-differential method, HPLC with intact pigments, and

HPLC with hydrolyzed pigments. Three of the methods: pH-differential and both HPLC methods, showed good linear correlation between each other (R2≥0.98). The total anthocyanins method had lower correlation to all other quantification because it accounted other pigments other than monomeric anthocyanins.

Proteins and anthocyanin could form complexation in the aqueous matrix was confirmed by FT-MIR spectroscopy. Hydrogen bonding was suggested by the infrared discrimination power to be the major driving force for the anthocyanin-protein complexation formation in neutral environment. When the matrix became more acidic

(pH<4.5), hydrophobic interaction and ion chelation might also involve into the complexation. The MIR protein predictive model developed in this study provided a simple, rapid approach to quantify protein levels in anthocyanin-rich matrix. This model might also help predict the water solubility properties of the pigments products and has great potential to provide an easy quality control tool for fast product quality evaluation.

Overall, our study presented important insightful understandings on production of high quality PCC pigment product for food application, from optimal extraction and spray drying conditions, to exploration of anthocyanin-protein complexation which might lead to poor solubility issue of PCC commercial pigments. Information in this study could provide an option for food companies to smoothly transit from artificial colorants to more consumer-friendly natural alternatives.

194

References

Abascal, K., Ganora, L., Yarnell, E., 2005. The effect of freeze-drying and its implications for botanical medicine: a review. Phyther. Res. 19, 655–660.

Abdel-Aal, E.-S.M., Hucl, P., 1999. A Rapid Method for Quantifying Total Anthocyanins in Blue Aleurone and Purple Pericarp Wheats. Cereal Chem. 76, 350–354.

Abdel-Aal, E.-S.M., Young, J.C., Rabalski, I., 2006. Anthocyanin Composition in Black, Blue, Pink, Purple, and Red Cereal Grains. J. Agric. Food Chem. 54, 4696–4704.

Afaq, F., Saleem, M., Krueger, C.G., Reed, J.D., Mukhtar, H., 2005. Anthocyanin- and hydrolyzable tannin-rich pomegranate fruit extract modulates MAPK and NF-κB pathways and inhibits skin tumorigenesis in CD-1 mice. Int. J. Cancer 113, 423– 433.

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. J. Agric. Food Chem. 62, 7524–7531.

Ahmed, M., Akter, M.S., Lee, J., Eun, J., 2010. Encapsulation by spray drying of bioactive components, physicochemical and morphological properties from purple sweet potato. LWT - Food Sci. Technol. 43, 1307–1312.

Ai, Y., Jane, J., 2016. Macronutrients in Corn and Human Nutrition. Compr. Rev. Food Sci. Food Saf. 15, 581–598.

Ainsworth, E.A., Gillespie, K.M., 2007. Estimation of total phenolic content and other oxidation substrates in plant tissues using Folin–Ciocalteu reagent. Nat. Protoc. 2, 875–877.

American Cancer Society, 2015. Cancer Facts & Figures 2015.

Andersen, Ø. M., Fossen, T., 1995. Anthocyanins with an unusual acylation pattern from stem of Allium victorialis. Phytochemistry, 40(6), 1809-1812.

Anderson, T.J., Lamsal, B.P., 2011. Zein Extraction from Corn, Corn Products, and 195

Coproducts and Modifications for Various Applications: A Review. Cereal Chem. 88, 159–173.

Andersson, J., Barth, A., 2006. FTIR studies on the bond properties of the aspartyl phosphate moiety of the Ca2+–ATPase. Biopolymers 82, 353–357.

Ansah, T., Agbolosu, A.A., Teye, G.A., Akwasi, A., Opoku-agyeman, M., 2012. Evaluation of Corn Cob on the Growth Performance of Grasscutter (Thryonomys swinderianus). Eval. Corn Cob Growth Perform. Grasscutter 45, 7–10.

Aoki, H., Kuze, N., Kato, Y., 2002. Anthocyanins isolated from purple corn (Zea mays L.). Foods Food Ingredients J. Japan 199, 41–45.

Argos, P., Pedersen, K., Marks, M.D., Larkins, B. a., 1982. A structural model for maize zein proteins. J. Biol. Chem. 257, 9984–9990.

Asami, D.K., Hong, Y.-J., Barrett, D.M., Mitchell, A.E., 2003. Comparison of the Total Phenolic and Ascorbic Acid Content of Freeze-Dried and Air-Dried Marionberry, Strawberry, and Corn Grown Using Conventional, Organic, and Sustainable Agricultural Practices. J. Agric. Food Chem. 51, 1237–1241.

Attoe, E.L., von Elbe, J., 1981. Photochemial Degradation of Betanine & Selected Anthocyanins. J. Food Sci. 46, 1934–1937.

Awika, J.M., Rooney, L.W., Waniska, R.D., 2004. Properties of 3-deoxyanthocyanins from sorghum. J. Agric. Food Chem. 52, 4388–4394.

Aykas, D.P., Rodriguez-Saona, L.E., 2016. Analytical Methods Assessing potato chip oil quality using a portable infrared spectrometer combined with pattern recognition analysis. Anal. Methods. 8(4), 731-741.

Azmir, J., Zaidul, I.S.M., Rahman, M.M., Sharif, K.M., Mohamed, A., Sahena, F., Jahurul, M.H.A., Ghafoor, K., Norulaini, N.A.N., Omar, A.K.M., 2013. Techniques for extraction of bioactive compounds from plant materials: A review. J. Food Eng. 117, 426–436.

Bakowska-Barczak, A., 2005. Acylated anthocyanins as stable, natural food colorants–a review. Pol. J. Nutr, Sci. 14, 107–116.

Balassa, L.L., Fanger, G.O., Wurzburg, O.B., 1971. Microencapsulation in the food industry. C R C Crit. Rev. Food Technol. 2, 245–265.

Baldauf, N.A., Rodriguez-Romo, L.A., Männig, A., Yousef, A.E., Rodriguez-Saona, L.E., 2007. Effect of selective growth media on the differentiation of Salmonella enterica serovars by Fourier-Transform Mid-Infrared Spectroscopy. J. Microbiol. Methods 68, 106–114. 196

Barnes, J.S., Nguyen, H.P., Shen, S., Schug, K.A., 2009. General method for extraction of blueberry anthocyanins and identification using high performance liquid chromatography–electrospray ionization-ion trap-time of flight-mass spectrometry. J. Chromatogr. A 1216, 4728–4735.

Barth, A., 2007. Infrared spectroscopy of proteins. Biochim. Biophys. Acta - Bioenerg. 1767, 1073–1101.

Barth, A., 2000. The infrared absorption of amino acid side chains. Prog. Biophys. Mol. Biol. 74, 141–173.

Basso, M.C., Lacoste, C., Pizzi, A., Fredon, E., Delmotte, L., 2014. MALDI-TOF and 13C NMR analysis of flexible films and lacquers derived from tannin. Ind. Crops Prod. 61, 352–360.

Berké, B., Chèze, C., Vercauteren, J., Deffieux, G., 1998. Bisulfite addition to anthocyanins: Revisited structures of colourless adducts. Tetrahedron Lett. 39, 5771–5774.

Bhattacharya, M., Jain, N., Bhasne, K., Kumari, V., Mukhopadhyay, S., 2011. pH- induced Conformational Isomerization of Bovine Serum Albumin Studied by Extrinsic and Intrinsic Protein Fluorescence. J. Fluoresc. 21, 1083–1090.

Brownmiller, C., Howard, L.R., Prior, R.L., 2008. Processing and Storage Effects on Monomeric Anthocyanins, Percent Polymeric Color, and Antioxidant Capacity of Processed Blueberry Products. J. Food Sci. 73, H72–H79.

Bueno, J.M., Sáez-Plaza, P., Ramos-Escudero, F., Jiménez, A.M., Fett, R., Asuero, A.G., 2012. Analysis and Antioxidant Capacity of Anthocyanin Pigments. Part II: Chemical Structure, Color, and Intake of Anthocyanins. Crit. Rev. Anal. Chem. 42, 126–151.

Cai, S., Singh, B.R., 2004. A Distinct Utility of the Amide III Infrared Band for Secondary Structure Estimation of Aqueous Protein Solutions Using Partial Least Squares Methods †. Biochemistry 43, 2541–2549.

Cai, Y.Z., Corke, H., 2000. Production and Properties of Spray-dried Amaranthus Betacyanin Pigments. J. Food Sci. 65, 1248–1252.

Carter, D.C., Ho, J.X., 1994. Structure of Serum Albumin. Adv. Protein Chem. 45, 153– 203.

Castañeda-Ovando, A., Pacheco-Hernández, M.D.L., Páez-Hernández, M.E., Rodríguez, J. a., Galán-Vidal, C.A., 2009. Chemical studies of anthocyanins: A review. Food Chem. 113, 859–871.

197

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 Res. Int. 44, 499–509.

Cevallos-Casals, B. a., Cisneros-Zevallos, L., 2004. Stability of anthocyanin-based aqueous extracts of Andean purple corn and red-fleshed sweet potato compared to synthetic and natural colorants. Food Chem. 86, 69–77.

Cevallos-Casals, B. a., Cisneros-Zevallos, L., 2003. Stoichiometric and Kinetic Studies of Phenolic Antioxidants from Andean Purple Corn and Red-Fleshed Sweetpotato. J. Agric. Food Chem. 51, 3313–3319.

Chanoca, A., Kovinich, N., Burkel, B., Stecha, S., Bohorquez-Restrepo, A., Ueda, T., Eliceiri, K.W., Grotewold, E., Otegui, M.S., 2015. Anthocyanin Vacuolar Inclusions Form by a Microautophagy Mechanism. Plant Cell 27, 2545–2559.

Charlton, A.J., Baxter, N.J., Lilley, T.H., Haslam, E., McDonald, C.J., Williamson, M.P., 1996. Tannin interactions with a full-length human salivary proline-rich protein display a stronger affinity than with single proline-rich repeats. FEBS Lett. 382, 289–292.

Chen, P.-N., Chu, S.-C., Chiou, H.-L., Chiang, C.-L., Yang, S.-F., Hsieh, Y.-S., 2005. Cyanidin 3-glucoside and peonidin 3-glucoside inhibit tumor cell growth and induce apoptosis in vitro and suppress tumor growth in vivo. Nutr. Cancer 53, 232–243.

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.). Int. J. Food Sci. Technol. 45, 1902–1908.

Choi, M.H., Kim, G.H., Lee, H.S., 2002. Effects of ascorbic acid retention on juice color and pigment stability in blood orange (Citrus sinensis) juice during refrigerated storage. Food Res. Int. 35, 753–759.

Clifford, M.N., 2000. Anthocyanins – nature, occurrence and dietary burden. J. Sci. Food Agric. 80, 1118–1125.

Coates, J., 2000. Interpretation of Infrared Spectra, A Practical Approach, in: Meyers, R.A. (Ed.), Encyclopedia of Analytical Chemistry. John Wiley & Sons Ltd., Chichester, NY, pp. 10815–10837.

Cocero, M.J., Martín, Á., Mattea, F., Varona, S., 2009. Encapsulation and co- precipitation processes with supercritical fluids: Fundamentals and applications. J. Supercrit. Fluids 47, 546–555.

Compton, S.J., Jones, C.G., 1985. Mechanism of dye response and interference in the

198

Bradford protein assay. Anal. Biochem. 151, 369–374.

Conn, S., Franco, C., Zhang, W., 2010. Characterization of anthocyanic vacuolar inclusions in Vitis vinifera L. cell suspension cultures. Planta 231, 1343–1360.

Conn, S., Zhang, W., Franco, C., 2003. Anthocyanic vacuolar inclusions (AVI) selectively bind acylated anthocyanins in Vitis vinifera L. (grapevine) suspension culture. Biotechnol. Lett. 25, 835–839.

Cooper-Driver, G.A., 2001. Contributions of Jeffrey Harborne and co-workers to the study of anthocyanins. Phytochemistry 56, 229–236.

Corrales, M., Toepfl, S., Butz, P., Knorr, D., Tauscher, B., 2008. Extraction of anthocyanins from grape by-products assisted by ultrasonics, high hydrostatic pressure or pulsed electric fields: A comparison. Innov. Food Sci. Emerg. Technol. 9, 85–91.

Dai, J., Mumper, R.J., 2010. Plant Phenolics: Extraction, Analysis and Their Antioxidant and Anticancer Properties. Molecules 15, 7313–7352.

Dandena, A., Leimane, I., Kletnieks, U., 2012. Validation of Monomeric Anthocianin Determination Method for Bilberry Juice and Marc Extracts. J. Life Sci. 6, 1378– 1382.

Dangles, O., Saito, N., Brouillard, R., 1993. Anthocyanin intramolecular copigment effect. Phytochemistry 34, 119–124.

Danzer, L.A., Ades, H., Rees, E.D., 1975. The helical content of zein, a water insoluble protein, in non-aqueous solvents. Biochim. Biophys. Acta - Protein Struct. 386, 26– 31.

De Mejia, E.G., Dia, V.P., West, L., West, M., Singh, V., Wang, Z., Allen, C., 2015. Temperature Dependency of Shelf and Thermal Stabilities of Anthocyanins from Corn Distillers’ Dried Grains with Solubles in Different Ethanol Extracts and a Commercially Available Beverage. J. Agric. Food Chem. 63, 10032–10041.

De Nardo, T., Shiroma-Kian, C., Halim, Y., Francis, D., Rodriguez-Saona, L.E., 2009. Rapid and simultaneous determination of lycopene and beta-carotene contents in tomato juice by infrared spectroscopy. J. Agric. Food Chem. 57, 1105–12.

De Pascual-Teresa, S., Santos-Buelga, C., Rivas-Gonzalo, J.C., 2002. LC-MS analysis of anthocyanins from purple corn cob. J. Sci. Food Agric. 82, 1003–1006.

De Vos, P., Faas, M.M., Spasojevic, M., Sikkema, J., 2010. Encapsulation for preservation of functionality and targeted delivery of bioactive food components. Int. Dairy J. 20, 292–302. 199

Deguchi, T., Ohba, R., Ueda, S., 2000a. Radical Scavenging Activity of a Purple Pigment, Hordeumin, from Uncooked Barley Bran-Fermented Broth. J. Agric. Food Chem. 48, 3198–3201.

Deguchi, T., Ohba, R., Ueda, S., 1999. Effects of Reactive Oxygen and Temperature on the Formation of a Purple Pigment, Hordeumin, from Barley Bran-Fermented Broth. Biosci. Biotechnol. Biochem. 63, 1151–1155.

Deguchi, T., Shohara, S., Ohba, R., Ueda, S., 2000b. Effects of pH and Light on the Storage Stability of the Purple Pigment, Hordeumin, from Uncooked Barley Bran Fermented Broth. Biosci. Biotechnol. Biochem. 64, 2236–2239.

Deguchi, T., Yoshimoto, M., Ohba, R., Ueda, S., 2000c. Antimutagenicity of the Purple Pigment, Hordeumin, from Uncooked Barley Bran-Fermented Broth. Biosci. Biotechnol. Biochem. 64, 414–416.

Del Guerra, S., Marselli, L., Lupi, R., Boggi, U., Mosca, F., Benzi, L., Del Prato, S., Marchetti, P., 2005. Effects of prolonged in vitro exposure to sulphonylureas on the function and survival of human islets. J. Diabetes Complications 19, 60–64.

Del Pozo-Insfran, D., Brenes, C.H., Serna Saldivar, S.O., Talcott, S.T., 2006. Polyphenolic and antioxidant content of white and blue corn (Zea mays L.) products. Food Res. Int. 39, 696–703.

Delgado-Vargas, F., Parades-Lopez, O., 2002. Natural Colorants for Food and Nutraceutical Uses. CRC Press, Boca Raton, Florida.

Delgado-Vargas, F., Jiménez, A. R., Paredes-López, O., 2000. Natural pigments: carotenoids, anthocyanins, and betalains--characteristics, biosynthesis, processing, and stability, Critical reviews in food science and nutrition, 40, 173-289.

Desobry, S.A., Netto, F.M., Labuza, T.P., 1997. Comparison of Spray-drying, Drum- drying and Freeze-drying for β-Carotene Encapsulation and Preservation. J. Food Sci. 62, 1158–1162.

Ding, M., Feng, R., Wang, S.Y., Bowman, L., Lu, Y., Qian, Y., Castranova, V., Jiang, B.-H., Shi, X., 2006. Cyanidin-3-glucoside, a Natural Product Derived from Blackberry, Exhibits Chemopreventive and Chemotherapeutic Activity. J. Biol. Chem. 281, 17359–17368.

Dioumaev, A.K., Braiman, M.S., 1995. Modeling Vibrational Spectra of Amino Acid Side Chains in Proteins: The Carbonyl Stretch Frequency of Buried Carboxylic Residues. J. Am. Chem. Soc. 117, 10572–10574.

Duangkhamchan, W., Siriamornpun, S., 2015. Quality attributes and anthocyanin content

200

of rice coated by purple-corn cob extract as affected by coating conditions. Food Bioprod. Process. 96, 171–179.

Duvick, D.N., 1961. Protein Granules of Maize Endosperm Cells. Cereal Chem.

Dyrby, M., Westergaard, N., Stapelfeldt, H., 2001. Light and heat sensitivity of red cabbage extract in soft drink model systems. Food Chem. 72, 431–437.

Edelmann, A., Lendl, B., 2002. Toward the Optical Tongue: Flow-Through Sensing of Tannin−Protein Interactions Based on FTIR Spectroscopy. J. Am. Chem. Soc. 124, 14741–14747.

Ersus, S., Yurdagel, U., 2007. Microencapsulation of anthocyanin pigments of black carrot (Daucus carota L.) by spray drier. J. Food Eng. 80, 805–812.

Escribano-Bailón, M.T., Santos-Buelga, C., Rivas-Gonzalo, J.C., 2004. Anthocyanins in cereals. J. Chromatogr. A 1054, 129–141.

Fabian, H., Yuan, T., Vogel, H., Mantsch, H., 1996. Comparative analysis of the amino- and carboxy-terminal domains of calmodulin by Fourier transform infrared spectroscopy. Eur. Biophys. J. 24, 195–201.

Fang, J., 2014. Bioavailability of anthocyanins. Drug Metab. Rev. 46, 508–520.

Fang, Z., Bhandari, B., 2011. Effect of spray drying and storage on the stability of bayberry polyphenols. Food Chem. 129, 1139–1147.

Fang, Z., Bhandari, B., 2010. Encapsulation of polyphenols – a review. Trends Food Sci. Technol. 21, 510–523.

FAO, 2013. Traditional High Andean Cuisine: Allin Mikuy / Sumak Mikuy, 1st ed. Food and Agriculture Organization of the United Nations.

Fazaeli, M., Emam-djomeh, Z., Ashtari, A.K., Omid, M., 2012. Food and Bioproducts Processing Effect of spray drying conditions and feed composition on the physical properties of black mulberry juice powder. Food Bioprod. Process. 90, 667–675.

Fernández, K., Agosin, E., 2007. Quantitative Analysis of Red Wine Tannins Using Fourier-Transform Mid-Infrared Spectrometry. J. Agric. Food Chem. 55, 7294– 7300.

Ferrari, C.C., Germer, S.P.M., Alvim, I.D., Vissotto, F.Z., de Aguirre, J.M., 2012a. Influence of carrier agents on the physicochemical properties of blackberry powder produced by spray drying. Int. J. Food Sci. Technol. 47, 1237–1245.

201

Ferrari, C.C., Germer, S.P.M., de Aguirre, J.M., 2012b. Effects of Spray-Drying Conditions on the Physicochemical Properties of Blackberry Powder. Dry. Technol. 30, 154–163.

Figueiredo, P., Elhabiri, M., Saito, N., Brouillard, R., 1996. Anthocyanin intramolecular interactions. A new mathematical approach to account for the remarkable colorant properties of the pigments extracted from Matthiola incana. J. Am. Chem. Soc. 118, 4788–4793.

Finkel, M.L., Sanchez, S., Mak, T., Granstein, J., Lefkowitz, A., 2013. Anthocyanin-Rich Purple Corn Extract and Its Effects on the Blood Pressure of Adults. J. Evid. Based. Complementary Altern. Med. 18, 237–242.

Fogler, B.B., Kleninschmidt, R. V., 1938. Spray Drying. Ind. Eng. Chem. 30, 1372–1384.

Fossen, T., Cabrita, L., Andersen, O.M., 1998. Colour and stability of pure anthocyanins influenced by pH including the alkaline region. Food Chem. 63, 435–440.

Fossen, T., Øvstedal, D.O., Slimestad, R., Andersen, Ø.M., 2003. Anthocyanins from a Norwegian potato cultivar. Food Chem. 81, 433–437.

Fragoso, S., Aceña, L., Guasch, J., Busto, O., Mestres, M., 2011a. Application of FT- MIR Spectroscopy for Fast Control of Red Grape Phenolic Ripening. J. Agric. Food Chem. 59, 2175–2183.

Fragoso, S., Aceña, L., Guasch, J., Mestres, M., Busto, O., 2011b. Quantification of Phenolic Compounds during Red Winemaking Using FT-MIR Spectroscopy and PLS-Regression. J. Agric. Food Chem. 59, 10795–10802.

Francis, F. J., Markakis, P. C., 1989. Food colorants: anthocyanins. Critical Reviews in Food Science & Nutrition, 28, 273-314.

Fukamachi, K., Imada, T., Ohshima, Y., Xu, J., Tsuda, H., 2008. Purple corn color suppresses Ras protein level and inhibits 7,12-dimethylbenz a anthracene-induced mammary carcinogenesis in the rat. Cancer Sci. 99, 1841-1846.

Fuleki, T., Francis, F.J., 1968a. Quantitative Methods for Anthocyanins.: 1. Extraction and Determination of Total Anthocyanin in Cranberries. J. Food Sci. 33, 72–77.

Fuleki, T., Francis, F.J., 1968b. Quantitative Methods for Anthocyanins.: 2. Determination of Total Anthocyanin and Degradation Index for Cranberry Juice. J. Food Sci. 33, 78–83.

Gao, L., Girard, B., Mazza, G., Reynolds, A.G., 1997. Changes in Anthocyanins and Color Characteristics of Pinot Noir Wines during Different Vinification Processes. J. Agric. Food Chem. 45, 2003–2008. 202

Gerwert, K., Hess, B., Soppa, J., Oesterhelt, D., 1989. Role of aspartate-96 in proton translocation by bacteriorhodopsin. Proc. Natl. Acad. Sci. 86, 4943–4947.

Gharsallaoui, A., Roudaut, G., Chambin, O., Voilley, A., Saurel, R., 2007. Applications of spray-drying in microencapsulation of food ingredients: An overview. Food Res. Int. 40, 1107–1121.

Giusti, M.M., Jing, P., 2007. Natural pigments of berries: Functionality and application, in: Yanyun Zhao (Ed.), Berry Fruit: Value-Added Products for Health Promotion. CRC Press, pp. 105–146.

Giusti, M.M., Rodríguez-Saona, L.E., Wrolstad, R.E., 1999. Molar absorptivity and color characteristics of acylated and non- acylated pelargonidin-based anthocyanins. J. Agric. Food Chem. 47, 4631–4637.

Giusti, M.M., Wrolstad, R.E., 2003. Acylated anthocyanins from edible sources and their applications in food systems. Biochem. Eng. J. 14, 217–225.

Giusti, M.M., Wrolstad, R.E., 2001. Characterization and Measurement of Anthocyanins by UV-Visible Spectroscopy, in: Wrolstad, R.E., Acree, T.E., An, H., Decker, E.A., Penner, M.H., Reid, D.S., Schwartz, S.J., Shoemaker, C.F., Peter, S. (Eds.), Current Protocols in Food Analytical Chemistry. John Wiley & Sons, Inc., New York, pp. F.1.2.1–F1.2.13.

Gomez, C., Terrier, N., Torregrosa, L., Vialet, S., Fournier-Level, A., Verries, C., Souquet, J.-M., Mazauric, J.-P., Klein, M., Cheynier, V., Ageorges, A., 2009. Grapevine MATE-Type Proteins Act as Vacuolar H+-Dependent Acylated Anthocyanin Transporters. PLANT Physiol. 150, 402–415.

Gómez-Caravaca, A.M., Verardo, V., Toselli, M., Segura-Carretero, A., Fernández- Gutiérrez, A., Caboni, M.F., 2013. Determination of the Major Phenolic Compounds in Pomegranate Juices by HPLC–DAD–ESI-MS. J. Agric. Food Chem. 61, 5328– 5337.

González-Manzano, S., Pérez-Alonso, J.J., Salinas-Moreno, Y., Mateus, N., Silva, A.M.S., de Freitas, V., Santos-Buelga, C., 2008. Flavanol–anthocyanin pigments in corn: NMR characterisation and presence in different purple corn varieties. J. Food Compos. Anal. 21, 521–526.

González-Paramás, A.M., Lopes da Silva, F., Martín-López, P., Macz-Pop, G., González- Manzano, S., Alcalde-Eon, C., Pérez-Alonso, J.J., Escribano-Bailón, M.T., Rivas- Gonzalo, J.C., Santos-Buelga, C., 2006. Flavanol–anthocyanin condensed pigments in plant extracts. Food Chem. 94, 428–436.

Gould, K., Davies, K.M., Winefield, C., 2009. Anthcyanins Biosynthesis, Functions and 203

Applications. Springer Science & Business Media.

Grotewold, E., 2005. Plant metabolic diversity: A regulatory perspective. Trends Plant Sci. 10, 57–62.

Grotewold, E., Chamberlin, M., Snook, M., Siame, B., Butler, L., Swenson, J., Maddock, S., St Clair, G., Bowen, B., 1998. Engineering secondary metabolism in maize cells by ectopic expression of transcription factors. Plant Cell 10, 721–740.

Grotewold, E., Drummond, B.J., Bowen, B., Peterson, T., 1994. The myb-homologous P gene controls phlobaphene pigmentation in maize floral organs by directly activating a flavonoid biosynthetic gene subset. Cell 76, 543–553.

Guillén, M.D., Cabo, N., 1997. Infrared spectroscopy in the study of edible oils and fats. J. Sci. Food Agric. 75, 1–11.

Hager, A., Howard, L.R., Prior, R.L., Brownmiller, C., 2008a. Processing and Storage Effects on Monomeric Anthocyanins, Percent Polymeric Color, and Antioxidant Capacity of Processed Black Raspberry Products. J. Food Sci. 73, H72–H79. x

Hager, T.J., Howard, L.R., Prior, R.L., 2008b. Processing and storage effects on monomeric anthocyanins, percent polymeric color, and antioxidant capacity of processed blackberry products. J. Agric. Food Chem. 56, 689–695.

Hagiwara, A., Miyashita, K., Nakanishi, T., Sano, M., Tamano, S., Kadota, T., Koda, T., Nakamura, M., Imaida, K., Ito, N., Shirai, T., 2001. Pronounced inhibition by a natural anthocyanin, purple corn color, of 2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine (PhIP)-associated colorectal carcinogenesis in male F344 rats pretreated with 1,2-dimethylhydrazine. Cancer Lett. 171, 17–25.

Harborne, J.B., 1983. Anthocyanins as food colours. Phytochemistry 22, 1067–1068.

Hassel, S., Rodriguez-Saona, L.E., 2012. Application of a Handheld Infrared Sensor for Monitoring the Mineral Fortification in Whole Grain Cornmeal. Food Anal. Methods 5, 571–578.

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. J. Chromatogr. A 1218, 7914–7922.

He, J., Giusti, M.M., 2010. Anthocyanins: Natural Colorants with Health-Promoting Properties. Annu. Rev. Food Sci. Technol. - (new 2010) 1, 163–187.

He, J., Rodriguez-Saona, L.E., Giusti, M.M., 2007. Midinfrared Spectroscopy for Juice AuthenticationRapid Differentiation of Commercial Juices. J. Agric. Food Chem. 55, 4443–4452. 204

Heinonen, I.M., Meyer, A.S., Frankel, E.N., 1998. Antioxidant Activity of Berry Phenolics on Human Low-Density Lipoprotein and Liposome Oxidation. J. Agric. Food Chem. 46, 4107–4112.

Hong, S.H., Heo, J., Kim, J., Kwon, S., Yeo, K., Bakowska-Barczak, A.M., Kolodziejczyk, P., Ryu, O.-H., Choi, M.-K., Kang, Y.-H., Lim, S.S., Suh, H.-W., Huh, S.-O., Lee, J.-Y., 2013. Antidiabetic and Beta Cell-Protection Activities of Purple Corn Anthocyanins. Biomol. Ther. 21, 284–289.

Huang, C., Li, J., Song, L., Zhang, D., Tong, Q., Ding, M., Bowman, L., Aziz, R., Stoner, G.D., 2006. Black raspberry extracts inhibit benzo(a)pyrene diol-epoxide-induced activator protein 1 activation and VEGF transcription by targeting the phosphotidylinositol 3-kinase/Akt pathway. Cancer Res. 66, 581–587.

Ignat, I., Volf, I., Popa, V.I., 2011. A critical review of methods for characterisation of polyphenolic compounds in fruits and vegetables. Food Chem. 126, 1821–1835.

Ilter, S., Dogan, I.S., Meral, R., 2008. Application of food grade coatings to Turkey buttocks. Ital. J. Food Sci. 20, 203–212.

Irani, N.G., Grotewold, E., 2005. Light-induced morphological alteration in anthocyanin- accumulating vacuoles of maize cells. BMC Plant Biol. 5, 1–15.

Jackman, R.L., Yada, R., Tung, M.A., 1987. A review: separation and chemical properties of anthocyanins used for their qualitative and quantitative analysis. J. Food Biochem. 11, 279–308.

Jia, Z., Tang, M., Wu, J., 1999. The determination of flavonoid contents in mulberry and their scavenging effects on superoxide radicals. Food Chem. 64, 555–559.

Jimenez-Aguilar, D.M., Ortega-Regules, A.E., Lozada-Ramirez, J.D., Perez-Perez, M.C.I., Vernon-Carter, E.J., Welti-Chanes, J., 2011. Color and chemical stability of spray-dried blueberry extract using mesquite gum as wall material. J. Food Compos. Anal. 24, 889–894.

Jing, P., 2006. Purple Corn Anthocyanins : Chemical Structure, Chemoprotective Activity and Structure / Function Relationships (Dissertation). The Ohio State University. Retrieved from .

Jing, P., Bomser, J. a., Schwartz, S.J., He, J., Magnuson, B. a., Giusti, M.M., 2008. Structure-function relationships of anthocyanins from various anthocyanin-rich extracts on the inhibition of colon cancer cell growth. J. Agric. Food Chem. 56, 9391–9398.

Jing, P., Giusti, M.M., 2007. Effects of extraction conditions on improving the yield and

205

quality of an anthocyanin-rich purple corn (Zea mays L.) color extract. J. Food Sci. 72, C363-C368.

Jing, P., Giusti, M.M., 2005. Characterization of anthocyanin-rich waste from purple corn cobs (Zea mays L.) and its application to color milk. J. Agric. Food Chem. 53, 8775–8781.

Jing, P., Noriega, V., Schwartz, S.J., Giusti, M.M., 2007. Effects of growing conditions on purple corncob (Zea mays L.) anthocyanins. J. Agric. Food Chem. 55, 8625– 8629.

Kahn, S.E., Hull, R.L., Utzschneider, K.M., 2006. Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature 444, 840–846.

Kandansamy, K., College, K.E., Somasundaram, P.D., Engineering, K., 2012. Microencapsulation of Colors by Spray Drying – A Review Microencapsulation of Colors by Spray Drying – A Review 8. doi:10.1515/1556-3758.2647

Kang, M.-K., Li, J., Kim, J.-L., Gong, J.-H., Kwak, S.-N., Park, J.H.Y., Lee, J.-Y., Lim, S.S., Kang, Y.-H., 2012. Purple corn anthocyanins inhibit diabetes-associated glomerular monocyte activation and macrophage infiltration. AJP Ren. Physiol. 303, F1060–F1069.

Kang, M.K., Lim, S.S., Lee, J.Y., Yeo, K.M., Kang, Y.H., 2013. Anthocyanin-rich purple corn extract inhibit diabetes-associated glomerular angiogenesis. PLoS One 8, 1–10.

Kapasakalidis, P.G., Rastall, R.A., Gordon, M.H., 2006. Extraction of Polyphenols from Processed Black Currant (Ribes nigrum L.) Residues. J. Agric. Food Chem. 54, 4016–4021.

Kha, T.C., Nguyen, M.H., Roach, P.D., 2010. Effects of spray drying conditions on the physicochemical and antioxidant properties of the Gac (Momordica cochinchinensis) fruit aril powder. J. Food Eng. 98, 385–392.

Khampas, S., Lertrat, K., Lomthaisong, K., Simla, S., Suriharn, B., 2015. Effect of location, genotype and their interactions for anthocyanins and antioxidant activities of purple waxy corn cob. Turkish J. F. Crop. 20, 15–23.

Kim, T.H., Kim, J.K., Kang, Y.-H., Lee, J.-Y., Kang, I.J., Lim, S.S., 2013. Aldose Reductase Inhibitory Activity of Compounds from Zea mays L. Biomed Res. Int. 2013, 1–8.

Koca, N., Kocaoglu-Vurma, N.A., Harper, W.J., Rodriguez-Saona, L.E., 2010. Application of temperature-controlled attenuated total reflectance-mid-infrared (ATR-MIR) spectroscopy for rapid estimation of butter adulteration. Food Chem.

206

121, 778–782.

Koes, R., Verweij, W., Quattrocchio, F., 2005. Flavonoids: a colorful model for the regulation and evolution of biochemical pathways. Trends Plant Sci. 10, 236–242.

Konczak, I., Zhang, W., 2004. Anthocyanins - More than nature’s colours. J. Biomed. Biotechnol. 2004, 239–240.

Kong, J., Yu, S., 2007. Fourier Transform Infrared Spectroscopic Analysis of Protein Secondary Structures. Acta Biochim. Biophys. Sin. (Shanghai). 39, 549–559.

Kong, J.M., Chia, L.S., Goh, N.K., Chia, T.F., Brouillard, R., 2003. Analysis and biological activities of anthocyanins. Phytochemistry 64, 923–933.

Koutchma, T., 2009. Advances in Ultraviolet Light Technology for Non-thermal Processing of Liquid Foods. Food Bioprocess Technol. 2, 138–155.

Kovinich, N., Arnason, J.T., De Luca, V., Miki, B., 2011. Coloring Soybeans with Anthocyanins. In: Gang, D.R. (Ed.), The Biological Activity of Phytochemicals. Springer New York, New York, NY, pp. 47–57.

Kovinich, N., Saleem, A., Arnason, J.T., Miki, B., 2011. Combined analysis of transcriptome and metabolite data reveals extensive differences between black and brown nearly-isogenic soybean (Glycine max) seed coats enabling the identification of pigment isogenes. BMC Genomics 12, 38, 1. doi: 10.1186/1471-2164-12-381.

Kulling, S.E., Rawel, H.M., 2008. Chokeberry (Aronia melanocarpa) - A review on the characteristic components and potential health effects. Planta Med. 74, 1625–1634.

Kwon, J.Y., Lee, K.W., Hur, H.J., Lee, H.J., 2007. Peonidin inhibits phorbol-ester- induced COX-2 expression and transformation in JB6 P+ cells by blocking phosphorylation of ERK-1 and -2. Ann. N. Y. Acad. Sci. 1095, 513–520.

Lao, F., Giusti, M.M., 2016. Quantification of Purple Corn (Zea mays L.) Anthocyanins Using Spectrophotometric and HPLC Approaches: Method Comparison and Correlation. Food Anal. Methods 9, 1367–1380.

Lee, C.H., Garcia, H.S., Parkin, K.L., 2010. Bioactivities of kernel extracts of 18 strains of maize (Zea mays). J. Food Sci. 75, 667–672.

Lee, E.A., Harper, V., 2002. Suppressor of pericarp pigmentation 1 (spp1), a novel gene involved in phlobaphene accumulation in maize (Zea mays L.) pericarps. Maydica 47, 51–58.

Lee, J., Durst, R.W., Wrolstad, R.E., 2005. Determination of total monomeric anthocyanin pigment content of fruit juices, beverages, natural colorants, and wines 207

by the pH differential method: Collaborative study. J. AOAC Int. 88, 1269–1278.

Lee, J., Rennaker, C., Wrolstad, R.E., 2008. Correlation of two anthocyanin quantification methods: HPLC and spectrophotometric methods. Food Chem. 110, 782–786.

Lee, S.G., Vance, T.M., Nam, T.-G., Kim, D.-O., Koo, S.I., Chun, O.K., 2016. Evaluation of pH differential and HPLC methods expressed as cyanidin-3-glucoside equivalent for measuring the total anthocyanin contents of berries. J. Food Meas. Charact. doi:10.1007/s11694-016-9337-9

Leja, M., Mareczek, A., Ben, J., 2003. Antioxidant properties of two apple cultivars during long-term storage. Food Chem. 80, 303–307.

Lew, E. a, Garfinkel, L., 1979. Variations in mortality by weight among 750,000 men and women. J. Chronic Dis. 32, 563–576.

Li, C.Y., Kim, H.W., Won, S.R., Min, H.K., Park, K.J., Park, J.Y., Ahn, M.S., Rhee, H.I., 2008. Corn husk as a potential source of anthocyanins. J. Agric. Food Chem. 56, 11413–11416.

Li, J., Kang, M.K., Kim, J.K., Kim, J.L., Kang, S.W., Lim, S.S., Kang, Y.H., 2012a. Purple corn anthocyanins retard diabetes-associated glomerulosclerosis in mesangial cells and db/db mice. Eur. J. Nutr. 51, 961–973.

Li, J., Lim, S.S., Lee, J.Y., Kim, J.K., Kang, S.W., Kim, J.L., Kang, Y.H., 2012b. Purple corn anthocyanins dampened high-glucose-induced mesangial fibrosis and inflammation: Possible renoprotective role in diabetic nephropathy. J. Nutr. Biochem. 23, 320–331.

Lin, C.A., Ayvaz, H., Rodriguez-Saona, L.E., 2014. Application of Portable and Handheld Infrared Spectrometers for Determination of Sucrose Levels in Infant Cereals. Food Anal. Methods 7, 1407–1414.

Liu, E.H., Qi, L.W., Cao, J., Li, P., Li, C.Y., Peng, Y.B., 2008. Advances of modern chromatographic and electrophoretic methods in separation and analysis of flavonoids. Molecules 13, 2521–2544.

Liu, X.-D., Atarashi, T., Furuta, T., Yoshii, H., Aishima, S., Ohkawara, M., Linko, P., 2001. Microencapsulation of emulsified hydrophobic flavours by spray drying. Dry. Technol. 19, 1361–1374.

Lofstedt, R., 2013. Communicating Food Risks in an Era of Growing Public Distrust: Three Case Studies. Risk Anal. 33, 192–202.

Long, N., Suzuki, S., Sato, S., Naiki-Ito, A., Sakatani, K., Shirai, T., Takahashi, S., 2013. 208

Purple corn color inhibition of prostate carcinogenesis by targeting cell growth pathways. Cancer Sci. 104, 298–303.

Lopez-Martinez, L.X., Oliart-Ros, R.M., Valerio-Alfaro, G., Lee, C.H., Parkin, K.L., Garcia, H.S., 2009. Antioxidant activity, phenolic compounds and anthocyanins content of eighteen strains of Mexican maize. LWT - Food Sci. Technol. 42, 1187– 1192.

Lopez-Martinez, L.X., Parkin, K.L., Garcia, H.S., 2011. Phase II-Inducing, Polyphenols Content and Antioxidant Capacity of Corn (Zea mays L.) from Phenotypes of White, Blue, Red and Purple Colors Processed into Masa and Tortillas. Plant Foods Hum. Nutr. 66, 41–47.

Lovrien, R., Matulis, D., 1995. Assays for Total Protein, in: Coligan, J.E., Dunn, B.M., Speicher, D.W., Wingfield, P.T. (Eds.), Current Protocols in Protein Science. John Wiley & Sons, Inc., Hoboken, NJ, USA, pp. 1–24.

Ludevid, M. D., Torrent, M., Martinez-Izquierdo, J. A., Puigdomenech, P., Palau, J., 1984. Subcellular localization of glutelin-2 in maize (Zea mays L.) endosperm. Plant molecular biology, 3, 227-234.

Maccarone, E., Maccarone, A, Rapisarda, P., 1985. Stabilization of anthocyanins in blood orange fruit juice. J. Food Sci. 50, 901–904.

Maedler, K., Carr, R.D., Bosco, D., Zuellig, R. a, Berney, T., Donath, M.Y., 2005. Sulfonylurea induced beta-cell apoptosis in cultured human islets. J. Clin. Endocrinol. Metab. 90, 501–506.

Mahdavi, S.A., Jafari, S.M., Ghorbani, M., Assadpoor, E., 2014. Spray-Drying Microencapsulation of Anthocyanins by Natural Biopolymers: A Review. Dry. Technol. 32, 509–518.

Malan, E., Swinny, E., Ferreira, D., Steynberg, P., 1996. The structure and synthesis of proguibourtinidins from cassia abbreviata. Phytochemistry 41, 1209–1213.

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. J. Agric. Food Chem. 49, 170– 176.

Markham, K.R., Gould, K.S., Winefield, C.S., Mitchell, K.A., Bloor, S.J., Boase, M.R., 2000. Anthocyanic vacuolar inclusions - Their nature and significance in flower colouration. Phytochemistry 55, 327–336.

Martin, S., Giannone, G., Andriantsitohaina, R., Martinez, M.C., 2003. Delphinidin, an

209

active compound of red wine, inhibits endothelial cell apoptosis via nitric oxide pathway and regulation of calcium homeostasis. Br. J. Pharmacol. 139, 1095–1102.

Mateus, N., de Freitas, V., 2009. Anthocyanins as Food Colorants, in: Gould, K., Davies, K., Winefield, C. (Eds.), Anthocyanins: Biosynthesis, Functions and Applications. Springer, New York, pp. 283–304.

Matsufuji, H., Kido, H., Misawa, H., Yaguchi, J., Otsuki, T., Chino, M., Takeda, M., Yamagata, K., 2007. Stability to light, heat, and hydrogen peroxide at different pH values and DPPH radical scavenging activity of acylated anthocyanins from red radish extract. J. Agric. Food Chem. 55, 3692–3701.

Matsushima, N., Danno, G., Takezawa, H., Izumi, Y., 1997. Three-dimensional structure of maize α-zein proteins studied by small-angle X-ray scattering. Biochim. Biophys. Acta - Protein Struct. Mol. Enzymol. 1339, 14–22.

Matus-Cádiz, M. a, Daskalchuk, T.E., Verma, B., Puttick, D., Chibbar, R.N., Gray, G.R., Perron, C.E., Tyler, R.T., Hucl, P., 2008. Phenolic compounds contribute to dark bran pigmentation in hard white wheat. J. Agric. Food Chem. 56, 1644–53.

Maurer, N.E., Hatta-Sakoda, B., Pascual-Chagman, G., Rodriguez-Saona, L.E., 2012. Characterization and authentication of a novel vegetable source of omega-3 fatty acids, sacha inchi (Plukenetia volubilis L.) oil. Food Chem. 134, 1173–1180.

Mazza, G., Francis, F.J., 1995. Anthocyanins in grapes and grape products. Crit. Rev. Food Sci. Nutr. 35, 341–371.

Mendoza-Díaz, S., Ortiz-Valerio, M.D.C., Castaño-Tostado, E., Figueroa-Cárdenas, J.D.D., Reynoso-Camacho, R., Ramos-Gómez, M., Campos-Vega, R., Loarca-Piña, G., 2012. Antioxidant Capacity and Antimutagenic Activity of Anthocyanin and Carotenoid Extracts from Nixtamalized Pigmented Creole Maize Races (Zea mays L.). Plant Foods Hum. Nutr. 67, 442–449.

Mennucci, B., Martínez, J.M., 2005. How to Model Solvation of Peptides? Insights from a Quantum-mechanical and Molecular Dynamics Study of N -Methylacetamide. 1. Geometries, Infrared, and Ultraviolet Spectra in Water. J. Phys. Chem. B 109, 9818– 9829.

Meschter, E.E., 1953. Fruit Color Loss, Effects of Carbohydrates and Other Factors on Strawberry Products. J. Agric. Food Chem. 1, 574–579.

Michnik, A., Michalik, K., Drzazga, Z., 2005. Stability of bovine serum albumin at different pH. J. Therm. Anal. Calorim. 80, 399–406.

Min, S.W., Ryu, S.N., Kim, D.H., 2010. Anti-inflammatory effects of black rice,

210

cyanidin-3-O-??-d-glycoside, and its metabolites, cyanidin and protocatechuic acid. Int. Immunopharmacol. 10, 959–966.

Mintel and Leatherhead Food Research, 2013. Food Colours: Market, Technical and Regulatory Insights. Leatherhead Food Research, Leatherhead.

Mizuguchi, M., Nara, M., Kawano, K., Nitta, K., 1997. FT-IR study of the Ca 2+ - binding to bovine α-lactalbumin. Relationships between the type of coordination and characteristics of the bands due to the Asp COO- groups in the Ca2+-binding site. FEBS Lett. 417, 153–156.

Mizuno, H., Hirano, K., Okamoto, G., 2006. Effect of anthocyanin composition in grape skin on anthocyanic vacuolar inclusion development and skin coloration. Vitis - J. Grapevine Res. 45, 173–177.

Montilla, E.C., Hillebrand, S., Antezana, A., Winterhalter, P., 2011. Soluble and Bound Phenolic Compounds in Different Bolivian Purple Corn (Zea mays L.) Cultivars. J. Agric. Food Chem. 59, 7068–7074.

Moore, J., Hao, Z., Zhou, K., Luther, M., Costa, J., Yu, L. (Lucy), 2005. Carotenoid, Tocopherol, Phenolic Acid, and Antioxidant Properties of Maryland-Grown Soft Wheat. J. Agric. Food Chem. 53, 6649–6657.

Murugesan, R., Orsat, V., 2012. Spray Drying for the Production of Nutraceutical Ingredients — A Review. Food and Bioprocess Technology, 5, 3-14.

Nabae, K., Hayashi, S.M., Kawabe, M., Ichihara, T., Hagiwara, A., Tamano, S., Tsushima, Y., Uchida, K., Koda, T., Nakamura, M., Ogawa, K., Shirai, T., 2008. A 90-day oral toxicity study of purple corn color, a natural food colorant, in F344 rats. Food Chem. Toxicol. 46, 774–780.

Nabedryk, E., Leonhard, M., Mäntele, W., Breton, J., 1990. Fourier transform infrared difference spectroscopy shows no evidence for an enolization of chlorophyll a upon cation formation either in vitro or during P700 photooxidation. Biochemistry 29, 3242–3247.

Nayak, C.A., Rastogi, N.K., 2010. Effect of Selected Additives on Microencapsulation of Anthocyanin by Spray Drying. Dry. Technol. 28, 1396–1404.

Nicoué, E.É., Savard, S., Belkacemi, K., 2007. Anthocyanins in Wild Blueberries of Quebec: Extraction and Identification. J. Agric. Food Chem. 55, 5626–5635.

Nozue, M., Kubo, H., Nishimura, M., Katou, A., Hattori, C., Usuda, N., Nagata, T., Yasuda, H., 1993. Characterization of Intravacuolar Pigmented Structures in Anthocyanin-Containing Cells of Sweet Potato Suspension Cultures. Plant Cell

211

Physiol. 34, 803–808.

Nozue, M., Yamada, K., Nakamura, T., Kubo, H., Kondo, M., Nishimura, M., 1997. Expression of a vacuolar protein (VP24) in anthocyanin-producing cells of sweet potato in suspension culture. Plant Physiol. 115, 1065–1072.

Nozzolillo, C., Ishikura, N., 1988. An investigation of the intracellular site of anthocyanoplasts using isolated protoplasts and vacuoles. Plant Cell Rep. 7, 389– 392.

Nyman, N.A., Kumpulainen, J.T., 2001. Determination of Anthocyanidins in Berries and Red Wine by High-Performance Liquid Chromatography. J. Agric. Food Chem. 49, 4183–4187.

Oberg, K.A, Ruysschaert, J.M., Goormaghtigh, E., 2004. The optimization of protein secondary structure determination with infrared and circular dichroism spectra. Eur. J. Biochem. 271, 2937–2948.

Obon, J.M., Castellar, M.R., Alacid, M., Fernandez-Lopez, J.A., 2009. Production of a red-purple food colorant from Opuntia stricta fruits by spray drying and its application in food model systems. J. Food Eng. 90, 471–479.

Ochoa, M.R., Kesseler, A.G., De Michelis, A., Mugridge, A., Chaves, A.R., 2001. Kinetics of colour change of raspberry, sweet (Prunus avium) and sour (Prunus cerasus) cherries preserves packed in glass containers: Light and room temperature effects. J. Food Eng. 49, 55–62.

Ohba, R., Kitaoka, S., Ueda, S., 1993. Properties and precursors of hordeumin produced from uncooked barley bran by ethanol fermentation. J. Ferment. Bioeng. 75, 121– 125.

Ohba, R., Nakayama, I., Suda, I., Ueda, S., 1995. Participation of reactive oxygens in the formation of a purple pigment, hordeumin, from barley bran-fermented broth. J. Ferment. Bioeng. 79, 73–75.

Osmani, S.A., Hansen, E.H., Malien-Aubert, Ć., Olsen, C.E., Bak, S., Møller, B.L., 2009. Effect of glucuronosylation on anthocyanin color stability. J. Agric. Food Chem. 57, 3149–3155.

Osorio, C., Acevedo, B., Hillebrand, S., Carriazo, J., Winterhalter, P., Morales, A.L., 2010. Microencapsulation by Spray-Drying of Anthocyanin Pigments from Corozo (Bactris guineensis) Fruit. J. Agric. Food Chem. 58, 6977–6985.

Ososanya, T.O., Alabi, B.O., Sorunke, A.O., 2013. Performance and Digestibility of Corncob and Cowpea Husk Diets by West African Dwarf Sheep. Pakistan J. Nutr.

212

12, 85–88.

Otsuki, T., Matsufuji, H., Takeda, M., Toyoda, M., Goda, Y., 2002. Acylated anthocyanins from red radish (Raphanus sativus L.). Phytochemistry 60, 79–87.

Pala, Ç.U., Toklucu, A.K., 2011. Effect of UV-C light on anthocyanin content and other quality parameters of pomegranate juice. J. Food Compos. Anal. 24, 790–795.

Pedreschi, R., 2005. Fractionation of phenolic compounds from a purple corn extract and evaluation of antioxidant and antimutagenic activities (Thesis). Texas A&M University. Retrieved from .

Pedreschi, R., Cisneros-Zevallos, L., 2007. Phenolic profiles of Andean purple corn (Zea mays L.). Food Chem. 100, 956–963.

Pedreschi, R., Cisneros-Zevallos, L., 2006. Antimutagenic and antioxidant properties of phenolic fractions from Andean purple corn (Zea mays L.). J. Agric. Food Chem. 54, 4557–4567.

Peters, T., 1985. Serum Albumin. Adv. Protein Chem. 37, 161–245.

Petrussa, E., Braidot, E., Zancani, M., Peresson, C., Bertolini, A., Patui, S., Vianello, A., 2013. Plant Flavonoids—Biosynthesis, Transport and Involvement in Stress Responses. Int. J. Mol. Sci. 14, 14950–14973.

Pinho, C., Melo, A., Mansilha, C., Ferreira, I.M.P.L.V.O., 2011. Optimization of Conditions for Anthocyanin Hydrolysis from Red Wine Using Response Surface Methodology (RSM). J. Agric. Food Chem. 59, 50–55.

Plundrich, N.J., Kulis, M., White, B.L., Grace, M.H., Guo, R., Burks, a. W., Davis, J.P., Lila, M.A., 2014. Novel strategy to create hypoallergenic peanut protein-polyphenol edible matrices for oral immunotherapy. J. Agric. Food Chem. 62, 7010–7021.

Poirier, M.C., 2004. Chemical-induced DNA damage and human cancer risk. Nat. Rev. Cancer 4, 630–637.

Pourcel, L., Irani, N.G., Lu, Y., Riedl, K., Schwartz, S., Grotewold, E., 2010. The Formation of Anthocyanic Vacuolar Inclusions in Arabidopsis thaliana and Implications for the Sequestration of Anthocyanin Pigments. Mol. Plant 3, 78–90.

Poustka, F., Irani, N.G., Feller, A., Lu, Y., Pourcel, L., Frame, K., Grotewold, E., 2007. A Trafficking Pathway for Anthocyanins Overlaps with the Endoplasmic Reticulum- to-Vacuole Protein-Sorting Route in Arabidopsis and Contributes to the Formation of Vacuolar Inclusions. PLANT Physiol. 145, 1323–1335.

Prior, R.L., Cao, G., Martin, A., Sofic, E., Mcewen, J., Brien, C.O., Lischner, N., 213

Ehlenfeldt, M., Kalt, W., Krewer, G., Mainland, C.M., 1998. Antioxidant capacity as influenced by total phenolic and anthocyanin content, maturity, and variety of Vaccinium Species. J. Agric. Food Chem. 8561, 2686–2693.

Quina, F.H., Moreira, P.F., Vautier-Giongo, C., Rettori, D., Rodrigues, R.F., Freitas, A. a., Silva, P.F., Maçanita, A.L., 2009. Photochemistry of anthocyanins and their biological role in plant tissues. Pure Appl. Chem. 81, 1687–1694.

Rahim, A. a., Rocca, E., Steinmetz, J., Jain Kassim, M., Sani Ibrahim, M., Osman, H., 2008. Antioxidant activities of mangrove Rhizophora apiculata bark extracts. Food Chem. 107, 200–207.

Rakotonirainy, A.M., Wang, Q., Padua, G.W., 2001. Evaluation of zein films as modified atmosphere packaging for fresh broccoli. J. Food Sci. 66, 1108–1111.

Ramos-Escudero, F., González-Miret, M.L., García-Asuero, A., 2012. Effect of various extraction systems on the antioxidant activity kinetic and color of extracts from purple corn. Vitae, 19, 41-48.

Ramos-Escudero, F., Muñoz, A.M., Alvarado-Ortíz, C., Alvarado, Á., Yáñez, J. a., 2012. Purple Corn (Zea mays L.) Phenolic Compounds Profile and Its Assessment as an Agent against Oxidative Stress in Isolated Mouse Organs. J. Med. Food 15, 206– 215.

Ranilla, L.G., Apostolidis, E., Genovese, M.I., Lajolo, F.M., Shetty, K., 2009. Evaluation of Indigenous Grains from the Peruvian Andean Region for Antidiabetes and Antihypertension Potential Using In Vitro Methods. J. Med. Food 12, 704–713.

Reddy, M.K., Alexander-Lindo, R.L., Nair, M.G., 2005. Relative inhibition of lipid peroxidation, cyclooxygenase enzymes, and human tumor cell proliferation by natural food colors. J Agric Food Chem 53, 9268–9273.

Reiersen, B., Kiremire, B.T., Byamukama, R., Andersen, Ø.M., 2003. Anthocyanins acylated with gallic acid from chenille plant, Acalypha hispida. Phytochemistry 64, 867–871.

Rein, M., 2005. Copigmentation reactions and color stability of berry anthocyanins (Dissertation). University of Helsinki.

Reynoso-Camacho, R., Guerrero-Villanueva, G., de Dios Figueroa, J., Gallegos-Corona, M.A., Mendoza, S., Loarca-Piña, G., Ramos-Gomez, M., 2015. Anticarcinogenic Effect of Corn Tortilla Against 1,2-Dimethylhydrazine (DMH)-Induced Colon Carcinogenesis in Sprague–Dawley Rats. Plant Foods Hum. Nutr. 70, 146–152.

Robert, P., Gorena, T., Romero, N., Sepulveda, E., Chavez, J., Saenz, C., 2010.

214

Encapsulation of polyphenols and anthocyanins from pomegranate (Punica granatum ) by spray drying. Int. J. Food Sci. Technol. 45, 1386–1394.

Rodriguez-Nogales, J.M., Garcia, M.C., Marina, M.L., 2006. High-performance liquid chromatography and capillary electrophoresis for the analysis of maize proteins. J. Sep. Sci. 29, 197–210.

Rodriguez-Saona, L.E., Allendorf, M.E., 2011. Use of FTIR for Rapid Authentication and Detection of Adulteration of Food. Annu. Rev. Food Sci. Technol. 2, 467–483.

Rodriguez-Saona, L.E., Fry, F.S., Calvey, E.M., 2000. Use of fourier transform near- infrared reflectance spectroscopy for rapid quantification of castor bean meal in a selection of flour-based products. J. Agric. Food Chem. 48, 5169–5177.

Rodriguez-Saona, L.E., Khambaty, F.M., Fry, F.S., Calvey, E.M., 2001. Rapid detection and identification of bacterial strains by Fourier transform near-infrared spectroscopy. J. Agric. Food Chem. 49, 574–579.

Rodriguez-Saona, L.E., Khambaty, F.M., Fry, F.S., Dubois, J., Calvey, E.M., 2004. Detection and identification of bacteria in a juice matrix with Fourier transform-near infrared spectroscopy and multivariiate analysis. J. Food Prot. 67, 2555–2559.

Rodriguez-Saona, L.E., Koca, N., Harper, W.J., Alvarez, V.B., 2006. Rapid determination of Swiss cheese composition by Fourier transform infrared/attenuated total reflectance spectroscopy. J. Dairy Sci. 89, 1407–12.

Rodriguez-Saona, L.E., Wrolstad, R.E., 2001. Extraction, Isolation, and Purification of Anthocyanins, in: Wrolstad, R.E., Acree, T.E., An, H., Decker, E.A., Penner, M.H., Reid, D.S., Schwartz, S.J., Shoemaker, C.F., Peter, S. (Eds.), Current Protocols in Food Analytical Chemistry. John Wiley & Sons, Inc., New York, pp. F1.1.1– F1.1.11.

Romero, C., Bakker, J., 2000. Effect of storage temperature and pyruvate on kinetics of anthocyanin degradation, vitisin A derivative formation, and color characteristics of model solutions. J. Agric. Food Chem. 48, 2135–2141.

Routray, W., Orsat, V., 2012. Microwave-Assisted Extraction of Flavonoids: A Review. Food Bioprocess Technol. 5, 409–424.

Sadilova, E., Carle, R., Stintzing, F.C., 2007. Thermal degradation of anthocyanins and its impact on color and in wfroantioxidant capacity. Mol. Nutr. Food Res. 51, 1461– 1471.

Sadilova, E., Stintzing, F.C., Carle, R., 2006. Thermal Degradation of Acylated and Nonacylated Anthocyanins. J. Food Sci. 71, C504–C512.

215

Saénz, C., Tapia, S., Chávez, J., Robert, P., 2009. Microencapsulation by spray drying of bioactive compounds from cactus pear (Opuntia ficus-indica). Food Chem. 114, 616–622.

Santos, P.M., Pereira-Filho, E.R., Rodriguez-Saona, L.E., 2013a. Rapid detection and quantification of milk adulteration using infrared microspectroscopy and chemometrics analysis. Food Chem. 138, 19–24.

Santos, P.M., Pereira-Filho, E.R., Rodriguez-Saona, L.E., 2013b. Application of Hand- Held and Portable Infrared Spectrometers in Bovine Milk Analysis. J. Agric. Food Chem. 61, 1205–1211.

Santos-Cervantes, M.E., Ibarra-Zazueta, M.E., Loarca-Piña, G., Paredes-López, O., Delgado-Vargas, F., 2007. Antioxidant and antimutagenic activities of Randia echinocarpa fruit. Plant Foods Hum. Nutr. 62, 71–77.

Sasaki, R., Nishimura, N., Hoshino, H., Isa, Y., Kadowaki, M., Ichi, T., Tanaka, A., Nishiumi, S., Fukuda, I., Ashida, H., Horio, F., Tsuda, T., 2007. Cyanidin 3- glucoside ameliorates hyperglycemia and insulin sensitivity due to downregulation of retinol binding protein 4 expression in diabetic mice. Biochem. Pharmacol. 74, 1619–1627.

Schneider, M., Esposito, D., Lila, M.A., Foegeding, E.A., 2016. Formation of whey protein–polyphenol meso-structures as a natural means of creating functional particles. Food Funct. 7, 1306–1318.

Seeram, N.P., Adams, L.S., Zhang, Y., Lee, R., Sand, D., Scheuller, H.S., Heber, D., 2006. Blackberry, Black Raspberry, Blueberry, Cranberry, Red Raspberry, and Strawberry Extracts Inhibit Growth and Stimulate Apoptosis of Human Cancer Cells In Vitro. J. Agric. Food Chem. 54, 9329–9339.

Selinger, D.A., Chandler, V.L., 1999. A mutation in the pale aleurone color1 gene identifies a novel regulator of the maize anthocyanin pathway. Plant Cell 11, 5–14.

Serra, D., Paixão, J., Nunes, C., Dinis, T.C.P., Almeida, L.M., 2013. Cyanidin-3- Glucoside Suppresses Cytokine-Induced Inflammatory Response in Human Intestinal Cells: Comparison with 5-Aminosalicylic Acid. PLoS One 8, 1–9.

Shimoda, H., Aitani, M., Tanaka, J., Hitoe, S., 2015. Purple Rice Extract Exhibits Preventive Activities on Experimental Diabetes Models and Human Subjects. J. Rice Res. 03, 2–5.

Shindo, M., Kasai, T., Abe, A., Kondo, Y., 2007. Effects of dietary administration of plant-derived anthocyanin-rich colors to spontaneously hypertensive rats. J. Nutr. Sci. Vitaminol. (Tokyo). 53, 90–93. 216

Shiroma, C., Rodriguez-Saona, L., 2009. Application of NIR and MIR spectroscopy in quality control of potato chips. J. Food Compos. Anal. 22, 596–605.

Shukla, R., Cheryan, M., 2001. Zein: The industrial protein from corn. Ind. Crops Prod. 13, 171–192.

Siebert, K.J., Troukhanova, N. V, Lynn, P.Y., 1996. Nature of Polyphenol−Protein Interactions. J. Agric. Food Chem. 44, 80–85.

Silva, P.I., Stringheta, P.C., Teofí lo, R.F., De Oliveira, I.R.N., 2013. Parameter optimization for spray-drying microencapsulation of jaboticaba (Myrciaria jaboticaba) peel extracts using simultaneous analysis of responses. J. Food Eng. 117, 538–544.

Silva, V.O., Freitas, A.A., Maçanita, A.L., Quina, F.H., 2016. Chemistry and photochemistry of natural plant pigments: the anthocyanins. J. Phys. Org. Chem. doi:10.1002/poc.3534

Sinha, A.K., Kumar, V., Makkar, H.P.S., De Boeck, G., Becker, K., 2011. Non-starch polysaccharides and their role in fish nutrition – A review. Food Chem. 127, 1409– 1426.

Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K., Gartner, F.H., Provenzano, M.D., Fujimoto, E.K., Goeke, N.M., Olson, B.J., Klenk, D.C., 1985. Measurement of protein using bicinchoninic acid. Anal. Biochem. 150, 76–85.

Snyder, A.B., Sweeney, C.F., Rodriguez-Saona, L.E., Giusti, M.M., 2014. Rapid authentication of concord juice concentration in a grape juice blend using Fourier- Transform infrared spectroscopy and chemometric analysis. Food Chem. 147, 295– 301.

Stark, L.E., Gross, A.T., 1992. Hydrophobic protein microparticles and preparation thereof. U.S. Patent No. 5,021,248. Washington, DC: U.S. Patent and Trademark Office.

Stich, K., Forkmann, G., 1988. Biosynthesis of 3-deoxyanthocyanins with flower extracts from Sinningia cardinalis. Phytochemistry 27, 785–789.

Strack, D., 1992. Two anthocyanins acylated with gallic acid from the leaves of Victoria amazonica. Phytochemistry 31, 989–991.

Strug, I., Utzat, C., Iii, A.C., Gutierrez, S., Amara, R., Lento, J., Capito, F., Skudas, R., Chernokalskaya, E., Nadler, T., 2014. Development of a Univariate Membrane- Based Mid-Infrared Method for Protein Quantitation and Total Lipid Content Analysis of Biological Samples. Journal of Analytical Methods in Chemistry,

217

doi:10.1155/2014/657079.

Subramanian, A., Harper, W.J., Rodriguez-Saona, L.E., 2009. Cheddar cheese classification based on flavor quality using a novel extraction method and Fourier transform infrared spectroscopy. J. Dairy Sci. 92, 87–94.

Suzuki, H., Nakayama, T., Yonekura-sakakibara, K., Fukui, Y., Nakamura, N., Yamaguchi, M., Tanaka, Y., Kusumi, T., Nishino, T., 2002. cDNA Cloning , Heterologous Expressions , and Functional Characterization of Malonyl-Coenzyme A : Anthocyanidin 3-O-Glucoside-6 -O-Malonyltransferase from Dahlia Flowers. Plant Physiol. 130, 2142–2151.

Tanaka, Y., Ohmiya, A., 2008. Seeing is believing: engineering anthocyanin and carotenoid biosynthetic pathways. Curr. Opin. Biotechnol. 19, 190–197.

Thiraphatthanavong, P., Wattanathorn, J., Muchimapura, S., Thukham-mee, W., Wannanon, P., Tong-un, T., Suriharn, B., Lertrat, K., 2014. Preventive Effect of Zea mays L. (Purple Waxy Corn) on Experimental Diabetic Cataract. Biomed Res. Int. 2014, 1–8.

Tonon, R. V, Brabet, C., Hubinger, M.D., 2010. Anthocyanin stability and antioxidant activity of spray-dried açai (Euterpe oleracea Mart.) juice produced with different carrier agents. Food Res. Int. 43, 907–914.

Tonon, R. V, Brabet, C., Hubinger, M.D., 2008. Influence of process conditions on the physicochemical properties of acai (Euterpe oleraceae Mart.) powder produced by spray drying. J. Food Eng. 88, 411–418.

Toyoshi, T., Kohda, T., 2004. Antihypertensive Activity of Purple Corn Color in Spontaneously Hypertensive Rats. Foods Food Ingredients J. Japan 209, 671–679.

Truong, V.D., Deighton, N., Thompson, R.T., McFeeters, R.F., Dean, L.O., Pecota, K. V., Yencho, G.C., 2010. Characterization of Anthocyanins and Anthocyanidins in Purple-Fleshed Sweetpotatoes by HPLC-DAD/ESI-MS/MS. J. Agric. Food Chem. 58, 404–410.

Tsuda, T., 2008. Regulation of adipocyte function by anthocyanins; Possibility of preventing the metabolic syndrome. J. Agric. Food Chem. 56, 642–646.

Tsuda, T., Horio, F., Uchida, K., Aoki, H., Osawa, T., 2003. Dietary cyanidin 3-O-beta- D-glucoside-rich purple corn color prevents obesity and ameliorates hyperglycemia in mice. J. Nutr. 133, 2125–2130.

Tsuda, T., Ueno, Y., Aoki, H., Koda, T., Horio, F., Takahashi, N., Kawada, T., Osawa, T., 2004. Anthocyanin enhances adipocytokine secretion and adipocyte-specific

218

gene expression in isolated rat adipocytes. Biochem. Biophys. Res. Commun. 316, 149–157.

Tsuda, T., Ueno, Y., Kojo, H., Yoshikawa, T., Osawa, T., 2005. Gene expression profile of isolated rat adipocytes treated with anthocyanins. Biochim. Biophys. Acta 1733, 137–147.

Tsuda, T., Ueno, Y., Yoshikawa, T., Kojo, H., Osawa, T., 2006. Microarray profiling of gene expression in human adipocytes in response to anthocyanins. Biochem. Pharmacol. 71, 1184–1197.

UNEP, 2012. Green Economy Sectoral Study: BioTrade – A catalyst for transitioning to a green economy in Peru. Retrieved from

Urias-Lugo, D.A., Heredia, J.B., Muy-Rangel, M.D., Valdez-Torres, J.B., Serna-Saldívar, S.O., Gutiérrez-Uribe, J.A., 2015. Anthocyanins and Phenolic Acids of Hybrid and Native Blue Maize (Zea mays L.) Extracts and Their Antiproliferative Activity in Mammary (MCF7), Liver (HepG2), Colon (Caco2 and HT29) and Prostate (PC3) Cancer Cells. Plant Foods Hum. Nutr. 70, 193–199.

Vayupharp, B., Laksanalamai, V., 2015. Antioxidant Properties and Color Stability of Anthocyanin Purified Extracts from Thai Waxy Purple Corn Cob. J. Food Nutr. Res. 3, 629–636.

Vilkhu, K., Mawson, R., Simons, L., Bates, D., 2008. Applications and opportunities for ultrasound assisted extraction in the food industry — A review. Innov. Food Sci. Emerg. Technol. 9, 161–169.

Von Germar, F., Barth, A., Mäntele, W., 2000. Structural Changes of the Sarcoplasmic Reticulum Ca2+-ATPase upon Nucleotide Binding Studied by Fourier Transform Infrared Spectroscopy. Biophys. J. 78, 1531–1540.

Wallace, T.C., 2011. Anthocyanins in Cardiovascular Disease 1. Adv. Nutr. 2, 1–7.

Wang, Shiow Y., Lin, H.S., 2000. Antioxidant activity in fruits and leaves. J. Agric. Food Chem. 48, 140–146.

Wang, L.S., Stoner, G.D., 2008. Anthocyanins and their role in cancer prevention. Cancer Lett. 269, 281–290.

Wang, T., 2014. The Authentication of Whey Protein Powder Ingredients and Understanding Factors Regulating Astringency in Acidic Whey Protein Beverages to Estimate Astringency by Infrared Spectroscopy ― An Instrumental Approach

219

(Dissertation). The Ohio State University. Retrieved from

Wang, T., Rodriguez-Saona, L.E., 2012. Rapid Determination of Sugar Level in Snack Products Using Infrared Spectroscopy. J. Food Sci. 77, 874–879.

Wang, T., Tan, S.Y., Mutilangi, W., Aykas, D.P., Rodriguez-Saona, L.E., 2015. Authentication of Whey Protein Powders by Portable Mid-Infrared Spectrometers Combined with Pattern Recognition Analysis. J. Food Sci. 80, C2111–C2116.

Waterhouse, A.L., 2003. Determination of Total Phenolics, in: Wrolstad, R., Acree, T., An, H., Decker, E., Penner, M., Reid, D., Schwartz, S., Shoemaker, C., Sporns, P. (Eds.), Current Protocols in Food Analytical Chemistry. John Wiley & Sons, Inc., Hoboken, NJ, USA, pp. 11.1.1–11.1.8.

Waters, E.J., Peng, Z., Pocock, K.F., Jones, G.P., Clarke, P., Williams, P.J., 1994. Solid- State 13C NMR investigation into insoluble deposits adhering to the inner glass surface of bottled red wine. J. Agric. Food Chem. 42, 1761–1766.

Welch, C.R., Wu, Q., Simon, J.E., 2008. Recent Advances in Anthocyanin Analysis and Characterization, Current Analytical Chemistry.

Wenstrup, M.J., Plans, M., Rodriguez-Saona, L.E., 2014. Effect of a novel induction food-processing device in improving frying oil quality. Int. J. Food Sci. Technol. 49, 2223–2229.

Wilkerson, E.D., Anthon, G.E., Barrett, D.M., Sayajon, G.F.G., Santos, A.M., Rodriguez- Saona, L.E., 2013. Rapid assessment of quality parameters in processing tomatoes using hand-held and benchtop infrared spectrometers and multivariate analysis. J. Agric. Food Chem. 61, 2088–2095.

Wilson, C.M., 1991. Multiple zeins from maize endosperms characterized by reversed- phase high performance liquid chromatography. Plant Physiol. 95, 777–786.

Winkel-Shirley, B., 2001. Flavonoid biosynthesis. A colorful model for genetics, biochemistry, cell biology, and biotechnology. Plant Physiol. 126, 485–493.

Wold, S., Sjöström, M., Eriksson, L., 2001. PLS-regression: a basic tool of chemometrics. Chemom. Intell. Lab. Syst. 58, 109–130.

Wolniak, M., Wawer, I., 2008. 13C CPMAS NMR and DFT calculations of anthocyanidins. Solid State Nucl. Magn. Reson. 34, 44–51.

Woodward, G., Kroon, P., Cassidy, A., Kay, C., 2009. Anthocyanin stability and recovery: Implications for the analysis of clinical and experimental samples. J. Agric. Food Chem. 57, 5271–5278. 220

Wrolstad, R.E., Durst, R.W., Lee, J., 2005. Tracking color and pigment changes in anthocyanin products. Trends Food Sci. Technol. 16, 423–428.

Wu, X., Beecher, G.R., Holden, J.M., Haytowitz, D.B., Gebhardt, S.E., Prior, R.L., 2006. Concentrations of Anthocyanins in Common Foods in the United States and Estimation of Normal Consumption. J. Agric. Food Chem. 54, 4069–4075.

Yang, Z., Chen, Z., Yuan, S., Zhai, W., Piao, X., Piao, X., 2009. Extraction and identification of anthocyanin from purple corn (Zea mays L.). Int. J. Food Sci. Technol. 44, 2485–2492.

Yang, Z., Zhai, W., 2010. Optimization of microwave-assisted extraction of anthocyanins from purple corn (Zea mays L.) cob and identification with HPLC–MS. Innov. Food Sci. Emerg. Technol. 11, 470–476.

Yokohira, M., Yamakawa, K., Saoo, K., Matsuda, Y., Hosokawa, K., Hashimoto, N., Kuno, T., Imaida, K., 2008. Antioxidant effects of flavonoids used as food additives (purple corn color, enzymatically modified isoquercitrin, and isoquercitrin) on liver carcinogenesis in a rat medium-term bioassay. J. Food Sci. 73, 561–568.

Yoshida, K., Kondo, T., Goto, T., 1992. Intramolecular stacking conformation of gentiodelphin, a diacylated anthocyanin from Gentiana makinoi. Tetrahedron 48, 4313–4326.

Yoshida, K., Mori, M., Kondo, T., 2009. Blue flower color development by anthocyanins: from chemical structure to cell physiology. Nat. Prod. Rep. 26, 884.

Yoshimoto, M., Okuno, S., Yamaguchi, M., Yamakawa, O., 2001. Antimutagenicity of deacylated anthocyanins in purple-fleshed sweetpotato. Biosci. Biotechnol. Biochem. 65, 1652–1655.

Yousefi, S., Emam-Djomeh, Z., Mousavi, S.M., 2011. Effect of carrier type and spray drying on the physicochemical properties of powdered and reconstituted pomegranate juice (Punica Granatum L.). J. Food Sci. Technol. 48, 677–684.

Zhang, H., Wang, L., Deroles, S., Bennett, R., Davies, K., 2006. New insight into the structures and formation of anthocyanic vacuolar inclusions in flower petals. BMC Plant Biol. 6, 1–14.

Zhang, Z., Kou, X., Fugal, K., McLaughlin, J., 2004. Comparison of HPLC Methods for Determination of Anthocyanins and Anthocyanidins in Bilberry Extracts. J. Agric. Food Chem. 52, 688–691.

Zhang, Z., Zhou, B., Wang, H., Wang, F., Song, Y., Liu, S., Xi, S., 2014. Maize Purple Plant Pigment Protects Against Fluoride-Induced Oxidative Damage of Liver and

221

Kidney in Rats. Int. J. Environ. Res. Public Health 11, 1020–1033.

Zhao, C., Giusti, M.M., Malik, M., Moyer, M.P., Magnuson, B. a., 2004. Effects of commercial anthocyanin-rich on colonic cancer and nontumorigenic colonic cell growth. J. Agric. Food Chem. 52, 6122–6128.

Zhao, J., 2015. Flavonoid transport mechanisms: how to go, and with whom. Trends Plant Sci. 20, 576–585.

Zhao, X., Corrales, M., Zhang, C., Hu, X., Ma, Y., Tauscher, B., 2008. Composition and Thermal Stability of Anthocyanins from Chinese Purple Corn (Zea mays L.). J. Agric. Food Chem. 56, 10761–10766.

Zhao, X., Zhang, C., Guigas, C., Ma, Y., Corrales, M., Tauscher, B., Hu, X., 2009. Composition, antimicrobial activity, and antiproliferative capacity of anthocyanin extracts of purple corn (Zea mays L.) from China. Eur. Food Res. Technol. 228, 759–765.

Žilić, S., Serpen, A., Akıllıoğlu, G., Gökmen, V., Vančetović, J., 2012. Phenolic Compounds, Carotenoids, Anthocyanins, and Antioxidant Capacity of Colored Maize (Zea mays L.) Kernels. J. Agric. Food Chem. 60, 1224–1231.

Zozio, S., Pallet, D., Dornier, M., 2011. Evaluation of anthocyanin stability during storage of a coloured drink made from extracts of the Andean blackberry (Rubus glaucus Benth.), açai (Euterpe oleracea Mart.) and black carrot (Daucus carota L.). Fruits 66, 203–215.

222