Impact of selected yeasts and glycosidase on chemical and volatile composition of mango wine

Chemical and volatile composition of mango wines fermented with different Saccharomyces cerevisiae yeast strains.. Impact of two Williopsis yeast strains on the volatile composition o

IMPACT OF SELECTED YEASTS AND GLYCOSIDASE ON CHEMICAL COMPOSITION

OF MANGO WINE

LI XIAO

NATIONAL UNIVERSITY OF SINGAPORE

2013

IMPACT OF SELECTED YEASTS AND GLYCOSIDASE ON CHEMICAL COMPOSITION

i

THESIS DECLARATION

I hereby declare that this thesis is my original work and it has been written

by me in its entirety, under the supervision of Dr Liu Shao Quan, (in the Food Science and Technology research laboratory, S13-05), Chemistry Department, National University of Singapore, between July 2009 and August 2013

I have duly acknowledged all the sources of information which have been used in the thesis

This thesis has not been submitted for any degree in any university previously

The content of the thesis has been partly published in:

1 Li, X., Yu, B., Curran, P., & Liu, S Q (2011) Chemical and volatile

composition of mango wines fermented with different Saccharomyces

cerevisiae yeast strains South African Journal of Enology and Viticulture, 32, 117-128

2 Li, X., Yu, B., Curran, P., & Liu, S Q (2012) Impact of two Williopsis

yeast strains on the volatile composition of mango wine International

Journal of Food Science and Technology, 47, 808-815

3 Li, X., Chan, L J., Yu, B., Curran, P., & Liu, S Q (2012) Fermentation

of three varieties of mango juices with a mixture of Saccharomyces

cerevisiae and Williopsis saturnus var mrakii International Journal of Food Microbiology, 158, 28-35

4 Lee, P R., Li, X., Yu, B., Curran, P., & Liu, S Q (2012)

Non-Saccharomyces yeasts and wine In A S Peeters (Ed.), Wine: Types, Production and Health, (pp 319-333) Hauppauge, New York, USA:

Nova Science Publishers

5 Chan, L J., Lee, P R., Li, X., Chen, D., Liu, S Q., & Trinh, T T T

(2012) Tropical fruit wine: an untapped opportunity In D Cabel (Ed.),

Food and Beverage Asia Dec/Jan 2011/2012 (pp 48–51) Singapore:

Pablo Publishing Pte Ltd ISSN: 2010-2364

ii

6 Li, X., Chan, L J., Yu, B., Curran, P., & Liu, S Q (2013) Influence of

Saccharomyces cerevisiae and Williopsis saturnus var mrakii on mango

wine characteristics Acta Alimentaria, Accepted (in press)

7 Li, X., Lim, S L., Yu, B., Curran, P., & Liu, S Q (2013) Impact of

pulp on the chemical profile of mango wine South African Journal of

Enology and Viticulture, 34, 119-128

8 Li, X., Lim, S L., Yu, B., Curran, P., & Liu, S Q (2013) Mango wine

aroma enhancement by pulp contact and β-glucosidase International

Journal of Food Science and Technology, 48, 2258-2266

iii

ACKNOWLEDGEMENTS

I would like to express the deepest appreciation to my supervisor Dr Liu Shao Quan for his valuable advice, supervision, guidance and encouragements throughout the entire project I feel honored to work with him and I have learned extensively from his creative ideas, profound knowledge, and in-depth research experience in the four years I am also deeply grateful for his kindness and patience throughout the four years and his countless efforts in guiding me to complete my research project Without his guidance and unwavering help this thesis would not have been possible

Next, I give my warmest appreciation to Dr Yu Bin who guided me extensively on techniques of SPME-GC-FID/MS and relative method development I also thank the flavorists from Firmenich Asia for their generous sharing of their knowledge and experience in flavor science and helpful support on the wine sensory tests

I would also thank other professors of Food Science and Technology (FST) program, such as Prof Zhou Wei Biao, Dr Huang Dejian, Dr Yuk Hyun Gyun and Dr David Popovich, for their teaching and enlightenment to expand

my knowledge in different aspects of food science through my four-year study

In addition, a thank you to my previous honors year students Mr Lim Sien Long and Miss Chew Xue Li for their help and commitment in conducting several experiments I would also thank my fellow postgraduate students, Dr Lee Pin Rou, Ms Sun Jing Can, Ms Cheong Mun Wai, Ms Chen Dai and Ms Chan Li Jie for their suggestions in experiments and moral encouragement I

am also grateful for all the technical support provided by FST staff Ms Lee Chooi Lan, Ms Lew Huey Lee, Ms Jiang Xiaohui, Mr Abdul Rahaman and

Ms Maria Chong

Last but not least, I would like to thank my parents and my adorable fiancée Tan Rui for their persistent love, support and encouragement in my life

iv

TABLE OF CONTENTS

THESIS DECLARATION i

ACKNOWLEDGEMENTS iii

SUMMARY ix

LIST OF TABLES xi

LIST OF FIGURES xiii

LIST OF ABBREVIATIONS xvii

LIST OF SYMBOLS xix

CHAPTER 1 Introduction and Literature Review 1

1.1 Mango 1

1.2 Mango wine 2

1.3 General comments of wine flavor 5

1.4 Flavor modulation by different techniques 8

1.4.1 Selection of Saccharomyces 9

1.4.2 Selection of non-Saccharomyces 10

1.4.3 Mixed-culture fermentation 11

1.4.4 Yeast interactions in mixed-culture fermentation 13

1.4.5 Enzymes 15

1.5 Objectives of project 16

CHAPTER 2 Materials and Methods 18

2.1 Fruits and yeast strains 18

2.2 Preparation of mango juice and pre-culture 19

2.3 Fermentation 20

2.4 Instrumental and sensory analysis of wine samples 23

CHAPTER 3 Chemical composition of mango wines fermented with different Saccharomyces cerevisiae yeasts strains 27

v

3.1 Introduction 27

3.2 Result and discussion 27

3.2.1 Brix, pH and yeast growth 27

3.2.2 Changes of sugars and organic acids 28

3.2.3 Volatile compounds in fresh mango juice 29

3.2.4 Volatile composition of mango wines after 14-day fermentation and kinetic changes of major volatiles 30

3.2.4.1 Alcohols 38

3.2.4.2 Esters 40

3.2.4.3 Volatile fatty acid 42

3.2.4.4 Carbonyl compounds 43

3.3 Conclusion 44

CHAPTER 4 Impact of two Williopsis yeast strains on the volatile composition of mango wine 45

4.1 Introduction 45

4.2 Results and discussion 45

4.2.1 Yeast growth, sugar consumption, and acidity 46

4.2.2 Volatile compounds and their evolution change 47

4.2.3 Quantification of major volatile compounds 52

4.3 Conclusion 55

CHAPTER 5 Influence of mixed-starter of Saccharomyces cerevisiae and Williopsis saturnus var mrakii on mango wine characteristics 56

5.1 Introduction 56

5.2 Results and discussion 56

5.2.1 Yeast ratio screening and population evolution 57

5.2.2 Brix, pH, sugars, ethanol, glycerol and organic acids 58

vi

5.2.3 Evolution of volatiles throughout the fermentation 59

5.2.3.1 Terpenes 59

5.2.3.2 Alcohols 61

5.2.3.3 Fatty acids 64

5.2.3.4 Esters 64

5.2.3.5 Other minor volatiles 66

5.2.4 Sensory test 67

5.3 Conclusion 68

CHAPTER 6 Fermentation of three varieties of mango juices with a mixture of Saccharomyces cerevisiae and Williopsis saturnus var mrakii 69

6.1 Introduction 69

6.2 Results and discussion 69

6.2.1 Changes in yeast population, pH, oBrix, sugars and organic acids 69

6.2.2 Volatiles composition and their kinetic changes in mango juices and wines 73

6.2.2.1 Terpenes 73

6.2.2.2 Alcohols 80

6.2.2.3 Esters 81

6.2.2.4 Carbonyls 81

6.2.2.5 Volatile fatty acids 82

6.2.3 Sensory test 82

6.3 Conclusion 83

CHAPTER 7 Antagonistic mechanism between Saccharomyces cerevisiae and Williopsis mrakii in co- and sequential cultures 85

7.1 Introduction 85

7.2 Results and discussion 85

vii

7.2.1 Early death of W mrkaii in co-culture with S cerevisiae 85

7.2.2 Effect of cell density on growth of two yeasts 87

7.2.3 Death of S cerevisiae in sequential inoculation with W mrakii 90

7.2.4 Presence of toxic compounds in W mrakii culture 90

7.3 Conclusion 94

CHAPTER 8 Impact of pulp contact on development of chemical profile of mango wine 95

8.1 Introduction 95

8.2 Results and discussion 95

8.2.1 Yeast growth, oBrix and pH changes 95

8.2.2 Sugar and organic acid contents before and after fermentation 98

8.2.3 Ethanol and glycerol contents of mango wine 98

8.2.4 Volatile profile of mango juice and wine 99

8.2.4.1 Major volatiles in juice 99

8.2.4.2 Evolution of terpenes hydrocarbons 99

8.2.4.3 Evolution of alcohols 105

8.2.4.4 Evolution of fatty acids 106

8.2.4.5 Evolution of esters 107

8.2.4.6 Evolution of sulfur compounds 108

8.3 Conclusion 109

CHAPTER 9 Enhancement of mango wine aroma by β-glucosidase 110

9.1 Introduction 110

9.2 Results and discussion 110

9.2.1 Physicochemical properties of mango wine with maceration and enzyme treatment 110

9.2.2 Volatiles of mango wine with pulp contact and enzyme treatment 112

viii

9.2.2.1 Terpenols and monoterpene hydrocarbons 112

9.2.2.2 Alcohols 116

9.2.2.3 Esters and fatty acids 117

9.2.2.4 Minor volatile compounds 117

9.2.2.5 Sensory characteristics of mango wine 118

9.2.2.6 Principal component analysis of volatiles in mango wine 120

9.3 Conclusion 121

CHAPTER 10 General conclusions and future work 122

REFERENCES 125

ix

SUMMARY

The project firstly screened the fermentation performance of three strains

of Saccharomyces as well as in two strains of Williopsis saturnus in mango juice for volatile and non-volatile transformation S cerevisiae var cerevisiae MERIT.ferm and W saturnus var mrakii NCYC500 were selected for

subsequent multi-starter fermentation due to their relatively optimal production of ethanol and esters

Generally, S cerevisiae was the main producer of ethanol, fusel alcohols, medium to long chain fatty acids and ethyl esters, while W saturnus was the

main producer of acetate esters and acetic acid but was unable to ferment the mango wine to dryness Volatiles that were initially present in mango juice, such as monoterpene hydrocarbons (C10H16) rapidly decomposed in the single

culture of S cerevisiae Decomposition of the mango varietal aroma was much slower in the fermentation by W saturnus The mixed-culture of the two

yeasts provided the opportunity to not only complete the fermentation but also improve aroma complexity and balance But this needs the winemaker to

apply the right ratio of S cerevisiae and W mrakii and control the duration of

fermentation

Three varieties of mango („R2E2‟, „Harum Manis‟ and „Nam Doc Mai‟)

were evaluated for wine-making using mixed-culture fermentation of S

cerevisiae MERIT.ferm and W mrakii NCYC500 at a ratio of 1:1000 and the

distinctive characters of each mango variety were analyzed „R2E2‟ wine had more fruity, sweet and creamy notes, and retained more of its original character „Harum Manis‟ wine had the lowest aroma intensity with more green and terpenic notes associated with higher levels of residual terpenes

„Nam Doc Mai‟ wine possessed the highest aroma intensity with winey, yeasty, fruity and floral notes attributed to higher amounts of alcohols, acetate esters and ethyl esters

The antagonistic mechanism of between W mrakii and S cerevisiae in mixed-starter fermentation was also studied W mrakii was ethanol sensitive and was likely inhibited by higher levels of ethanol produced by S cerevisiae

S cerevisiae was sensitive to killer toxin produced by W mrakii

x

The project further investigated non-fermentation techniques to improve mango wine aroma quality such as pulp maceration and exogenous β-glucosidase addition Pulp maceration was very efficient in retaining monoterpene hydrocarbons (e.g α-terpinolene, δ-3-carene) varietal aroma compounds of fresh mango Exogenous β-glucosidase facilitated not only release of glycosidically bound aroma compounds in mango wine especially terpenols such as β-citronellol, nerol and geraniol, but also balancing the relative ratio of acetate/ethyl esters as well as acetic acid/medium-chain fatty acids

xi

LIST OF TABLES

Table 3.1 Physicochemical properties, organic acid and sugar concentrations

of mango wines before and after fermentation 29

Table 3.2 Major volatile compounds (GC-FID peak area ×106) and their

relative peak areas (RPA) in fresh „Chok Anan‟ mango juice 31

Table 3.3 Major volatile compounds (GC-FID peak area ×106) and their

relative peak areas (RPA) in mango wine (day 14) fermented by

three S cerevisiae yeasts 34

Table 3.4 Concentrations of selected volatile compounds (mg/L) and the

corresponding odor activity values (OAVs) in mango wines

fermented with culture of three S cerevisiae yeasts on Day 14 37

Table 4.1 Physicochemical properties, organic acid and sugar concentrations

of mango wine 47

Table 4.2 Summary of volatiles in mango juice and wine with 21-day

fermentation 48

Table 4.3 Summary of changes of major terpenoids and their derivatives from

mango juice (day 0) to mango wine after 21-day fermentation 50

Table 4.4 Concentrations of odorants (mg/L) and their corresponding odor

activity values (OAVs) in mango wine fermented with two

Williopsis yeast strains on day 21 54

Table 5.1 Changes of sugars, organic acids, ethanol and glycerol in mango

wines before and after fermentation 59

Table 5.2 Concentrations, odor thresholds, odor activity values (OAVs) and

odor description of typical odorants in mango wines 62

Table 6.1 Changes of sugars, organic acids, ethanol and glycerol in mango

wines before and after fermentation of three varieties [„R2E2‟,

„Harum Manis‟ („HM‟), „Nam Doc Mai‟ („NDM‟)] inoculated with

S cerevisiae and W mrakii 72

Table 6.2 Categorization of volatiles of fresh mango juices (Day 0) and

mango wine (Day 21) from three varieties 76

Table 6.3 Concentration, odor thresholds, odor activity values (OAVs) and

odor description of potent odorants in mango wines 78

Table 8.1 Physicochemical properties, alcohol, organic acid, and sugar

concentrations of mango wines with and without pulp 97

xii

Table 8.2 Volatiles (mg/L) and their odor activity values (OAVs) of

non-macerated and non-macerated mango wines 103

Table 9.1 Physicochemical properties, alcohol, organic acid, and sugar

concentrations of non-macerated and macerated mango wine, with and without enzyme treatment 112

Table 9.2 Major volatiles (mg/L) and their odor activity values (OAVs) for

non-macerated and macerated wine, with and without enzyme treatment 114

xiii

LIST OF FIGURES

Figure 3.1 Changes of alcohols in mango wines during fermentation by S

cerevisiae MERIT.ferm (), S chevalieri CICC-1028 (▲) and S

bayanus EC-1118 (■) 38

Figure 3.2 Changes of acetate esters in mango wines during fermentation by

S cerevisiae MERIT.ferm (), S chevalieri CICC-1028 (▲) and S

bayanus EC-1118 (■) 41

Figure 3.3 Changes of ethyl esters in mango wines during fermentation by S

cerevisiae MERIT.ferm (), S chevalieri CICC-1028 (▲) and S

bayanus EC-1118 (■) 42

Figure 3.4 Changes of fatty acids in mango wines during fermentation by S

cerevisiae MERIT.ferm (), S chevalieri CICC-1028 (▲) and S

bayanus EC-1118 (■) 43

Figure 4.1 Evolution changes of selected terpenoids in mango wine during

fermentation by W mrakii NCYC500 (▲) and W suaveolens

NCYC2586 (■) 49

Figure 4.2 Evolution changes of selected alcohols in mango wine during

fermentation by W mrakii NCYC500 (▲) and W suaveolens

NCYC2586 (■) 50

Figure 4.3 Evolution changes of selected acetate esters in mango wine during

fermentation by W mrakii NCYC500 (▲) and W suaveolens

NCYC2586 (■) 51

Figure 5.1(a) Yeast cell populations during mango wine fermentations

inoculated with different ratios of S cerevisiae MERIT.ferm (■) to

W saturnus var mrakii NCYC500 () = 1:1, 1:10, 1:100, 1:1000 (b) Yeast cell populations in monoculture (■) or mixed-culture ()

at a ratio of 1:1000 of S cerevisiae MERIT.ferm to W saturnus var

mrakii NCYC500 58

Figure 5.2 Evolution trend of terpenoids throughout the fermentation of

mango juices: monoculture of S cerevisiae MERIT.ferm (♦),

monoculture of W saturnus var mrakii NCYC500 (▲) and culture of S cerevisiae MERIT.ferm : W saturnus var mrakii

mixed-NCYC500 = 1:1000 (■) 60

Figure 5.3 Evolution trend of (Z)-3-hexenol and isoamyl alcohol throughout

the fermentation of mango juices: monoculture of S cerevisiae MERIT.ferm (♦), monoculture of W saturnus var mrakii

xiv

NCYC500 (▲) and mixed-culture of S cerevisiae MERIT.ferm :

W saturnus var mrakii NCYC500 = 1:1000 (■) 61

Figure 5.4 Evolution trend of acetate and ethyl esters throughout the

fermentation of mango juices: monoculture of S cerevisiae

MERIT.ferm (♦), monoculture of W saturnus var mrakii

NCYC500 (▲) and mixed-culture of S cerevisiae MERIT.ferm :

W saturnus var mrakii NCYC500 = 1:1000 (■) 64

Figure 5.5 Sensory profile of mango wines: monoculture of S cerevisiae

MERIT.ferm (♦), monoculture of W saturnus var mrakii

NCYC500 (▲) and mixed-culture of S cerevisiae MERIT.ferm :

W saturnus var mrakii NCYC500 = 1:1000 (■) 68

Figure 6.1 The mean yeast cell population in mango wines (triplicates for

each mango juice) made from varieties „R2E2‟, „Harum Manis‟ and

„Nam Doc Mai‟: S cerevisiae MERIT.ferm (■) and W saturnus var mrakii NCYC500 () 70

Figure 6.2 Changes in concentration of fructose, glucose sucrose and ethanol

throughout the fermentation of mango juices of three varieties:

„R2E2‟ (♦), „Harum Manis‟ („HM‟) (▲) and „Nam Doc Mai‟

(„NDM‟) (■) 73

Figure 6.3 (a) Bi-plot of principal component analysis for selected volatiles

of fresh mango juices of three varieties [„R2E2‟, „Harum Manis‟ („HM‟) and „Nam Doc Mai‟ („NDM‟)]: 1) δ-3-Carene 2)α-

Phellandrene 3)Limonene 4)(Z)-β-Ocimene 5)(E)-β-Ocimene,

6)allo-Ocimene, 7)α-terpinolene 8)β-Caryophyllene 9)β-Selinene

10)Germacrene-D 11)1-Hexanal 12)(E)-2-Hexenal

13)2,6-Nonadienal 14)Acetaldehyde 15)Acetoin 16)δ-hexanolactone 17)Geranyl acetone 18)γ-octalactone 19)β-Ionone 20)δ-octalactone 21)γ-decalactone 22)δ-nonalactone 23)Ethyl acetate 24)Methyl 2-butenoate 25)Propyl butyrate 26)Butyl butyrate 27)Ethyl 2-

butenoate 28)3-Hexen-1-ol acetate 29)Ethyl 3-hydroxybutyrate

30)Ethanol 31)1-Penten-3-ol 32)1-Hexanol 33)(Z)-3-Hexenol

34)2-Hexen-1-ol (b) Bi-plot of principal component analysis for selected

volatiles in final mango wines of three varieties [„R2E2‟, „Harum Manis‟ („HM‟) and „Nam Doc Mai‟ („NDM‟)]: 1)ethanol 2)isobutyl

alcohol 3)active amyl alcohol 4)isoamyl alcohol 5)(Z)-3-hexenol 6)2-phenylethyl alcohol 7)ethyl acetate 8)isoamyl acetate 9)(Z)-3-

hexenyl acetate 10)citronellyl acetate 11)2-phenylethyl acetate 12)ethyl hexanoate 13)ethyl octanoate 14)ethyl decanoate 15)ethyl dodecanoate 16)acetic acid 17)octanoic acid 18)decanoic acid 19)linalool 20)α-terpineol 21)β-citronellol 22)β-damascenone 23) γ-octalactone 24) β-Ionone 25) δ-octalactone 26) γ-decalactone 27) δ-nonalacton 75

xv

Figure 6.4 Evolution trend of terpene hydrocarbons and β-citronellol

throughout the fermentation of mango juices of three varieties:

„R2E2‟ (♦),‟Harum Manis‟ („HM‟) (▲) and „Nam Doc Mai‟

(„NDM‟) (■) 77

Figure 6.5 Evolution trend of (Z)-3-hexenol and isoamyl alcohol, and their

corresponding acetate esters throughout the fermentation of mango

juices of three varieties: „R2E2‟ (♦),‟Harum Manis‟ („HM‟) (▲) and „Nam Doc Mai‟ („NDM‟) (■) 80

Figure 6.6 Evolution trend of γ-octalactone throughout the fermentation of

mango juices of three varieties: „R2E2‟ (♦),‟Harum Manis‟ („HM‟) (

▲) and „Nam Doc Mai‟ („NDM‟) (■) 82

Figure 6.7 Aroma profile of mango wines from three varieties: „R2E2‟

(♦),‟Harum Manis‟ („HM‟) (▲) and „Nam Doc Mai‟ („NDM‟) (■)

83

Figure 7.1 Cell population dynamics of W mrakii and S cerevisiae in

co-culture Co-inoculation of (a) W mrakii NCYC500 and (b) S

cerevisiae MERIT.ferm in nutrient broth with different

concentrations of glucose: 3% (♦), 6% (■), 9% (▲), 12% (×) and 15% (●) 86

Figure 7.2 Cell population dynamics of W mrakii NCYC500 in nutrient broth

with different concentrations of ethanol: 0% (♦), 2% (■), 4% (▲), 6% (×), 8% (●) and 10% (Ж) 87

Figure 7.3 Cell population dynamics of W mrakii and S cerevisiae inoculated

at different ratios W mrakii (a) at ~105 CFU/mL was co-inoculated with different concentrations of S cerevisiae (b) when glucose concentration was 2% (w/v), S cerevisiae/W mrakii=1:1(♦), 1:10(

■), 1:100(▲), 1:1000(●) S cerevisiae (c) at ~105

CFU/mL was co-inoculated with different concentrations of W mrakii (d) when glucose concentration was 2% (w/v), S cerevisiae/W

mrakii=1:1(♦), 10:1(■), 100:1(▲), 1000:1(●) 88

Figure 7.4 Cell population dynamics of yeasts in co-cultures inoculated at

different ratios The growth of W mrakii (a) in co-culture with S

cerevisiae (b) at different inoculums when glucose concentration

was 15% (w/v), S cerevisiae/W mrakii=1:1(♦), 1:10(■), 1:100(▲

), 1:1000(●) 89

Figure 7.5 Cell population dynamics of yeasts in sequential cultures

Approximately ~104, ~105, ~106 and ~107 CFU/mL of S cerevisiae (a) was inoculated into the medium seven days after W mrakii (b) was inoculated The starting ratio at day 7 was S cerevisiae/W

mrakii=1:1(♦), 1:10(■), 1:100(▲), 1:1000(●) 90

xvi

Figure 7.6 Cell population dynamics of yeasts in supernatants (a) growth of

S cerevisiae in cell-free supernatant of W mrakii on Day 2, and the

inoculums of S cerevisiae as followed: ~101 (♦), ~102 (■), ~103 (

▲), ~104

(×), ~105 CFU/mL (●); (b) growth of S cerevisiae in cell-free supernatant of W mrakii on Day 7, and the inoculums of S

cerevisiae as followed: ~101 (♦), ~102 (■), ~103 (▲), ~104 (×),

~105 CFU/mL (●); (c) growth of S cerevisiae in cell-free

supernatant of W mrakii on Day 14, and the inoculums of S

cerevisiae as followed: ~101 (♦), ~102 (■), ~103 (▲), ~104 (×),

~105 CFU/mL (●) 92

Figure 8.1 Yeast growth and changes in oBrix for mango wine fermentation

with centrifuged (▲) and uncentrifuged juices (■) 96

Figure 8.2 Evolution trends of terpenes and terpene alcohols during mango

wine fermentation with centrifuged (▲) and uncentrifuged juices (

■) 100

Figure 8.3 Evolution trends of alcohols during mango wine fermentation with

centrifuged (▲) and uncentrifuged juices (■) 106

Figure 8.4 Evolution trends of volatile fatty acids during mango wine

fermentation with centrifuged (▲) and uncentrifuged juices (■) 107

Figure 8.5 Evolution trends of esters during mango wine fermentation with

centrifuged (▲) and uncentrifuged juices (■) 108

Figure 9.1 FID chromatogram of (E)-β-damascenone in non-macerated

control, macerated wine control, non-macerated with enzyme and macerated wine with enzyme 118

Figure 9.2 Aroma profile of mango wines: non-macerated control (▲);

macerated wine control (■); non-macerated with enzyme (♦); and macerated wine with enzyme (×) 119

Figure 9.3 Biplot of principal component analysis of mango wines:

non-macerated control (JCtrl, ▲); non-macerated wine control (PCtrl, ■); enzyme-treated non-macerated wine (JEnz, ♦); and enzyme-treated macerated wine (PEnz, ×): (1) acetic acid; (2) hexanoic acid; (3) octanoic acid; (4) decanoic acid; (5) dodecanoic acid; (6) ethanol; (7) isobutyl alcohol; (8) active amyl alcohol; (9) isoamyl alcohol;

(10) (Z)-3-hexenol; (11) linalool; (12) α-terpineol; (13)

β-citronellol; (14) nerol; (15) geraniol; (16) 2-phenylethyl alcohol; (17) ethyl acetate; (18) isobutyl acetate; (19) isoamyl acetate; (20)

(Z)-3-hexenyl acetate; (21) ethyl hexanoate; (22) ethyl octanoate;

(23) ethyl decanoate; (24) citronellyl acetate; (25) 2-phenylethyl acetate; (26) ethyl dodecanoate; (27) α-terpinolene 121

xvii

LIST OF ABBREVIATIONS

ABTS 2,2‟-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid

ELSD-LT Low temperature evaporative light scattering detector

FRAP Ferric reducing antioxidant power

GC- FID Gas chromatography-Flame ionization detector

HPLC High performance liquid chromatography

HS-SPME Head space-solid phase micronextraction

K2S2O5 Potassium metabisulfite

NCYC National Collection of Yeast Cultures

NIST National Institute of Standards and Technology

CHAPTER 1 Introduction and Literature Review

1.1 Mango

Mango (Mangifera indica L.) is commercially one of the most abundant

tropical fruits in Southeast and South Asia, accounting for its large market share of the total mango produced worldwide (Tharanathan, Yashoda, & Prabha, 2006) Over 30 different varieties of mango are grown and appreciated for its intense peel coloration, attractive fragrance, delicious taste and high nutrition value (Spreer, Ongprasert, Hegele, Wünsche, & Müller, 2009)

The chemical composition of mango varies with location of cultivation, variety, and stage of maturity The major constituents of the pulp are water, carbohydrates, organic acids, fats, tannins, vitamins and flavor compounds (Sagar, Khurdiya, & Balakrishnan, 1999) The rate of starch accumulation is rapid at the beginning of fruit growth and slows down later, but it continues to increase up to maturity Both reducing and non-reducing sugars in mango were found to be increasing towards the end of maturity (Mann, Singh, & Pandey, 1974) The main soluble sugars of mango consisted of glucose, fructose and sucrose Organic acids include oxalic, citric, malic, succinic, pyruvic, adipic, galacturonic, glucuronic and mucic acids were identified in mango, with citric acid being the major one (Tharanathan et al., 2006)

The total sugar content of mangoes varied from 11% to 25% (w/w) During ripening of mango, both sugar and pH would increase (Tandon & Kalra, 1984) The increase in sweetness, as a result of glucogenesis, hydrolysis

of polysaccharides (startch), decrease in acidity and accumulation of sugars and organic acids with an excellent sugar/acid blend are responsible for the taste development Twelve amino acids were reported during mango growth (Tandon & Kalra, 1984) Alanine, arginine, glycine, serine, and leucine, isoleucine are major amino acids detected in mango Triglyceride is the major lipids in mango fruit, but its concentration is generally lower than 1% (w/w) (Gholap & Bandyopadhyay, 1975) Mango is high in vitamin C, but it reaches

a maximum in the early stage of growth instead of ripening stage (Spencer, Morris, & Kennard, 1956) However, the change of β-carotene is in a reversed way β-Carotene remained low in the beginning and increased as the fruits

approached maturity and during ripening (Jungalwala & Cama, 1963; Godoy

& Rodriguez-Amaya, 1987) The increase in β-carotene is accompanied by a decrease in acids and an increase in sugar content

The major textural changes resulting in the softening of fruits are due to enzyme-mediated alterations in the structure and composition of cell wall, partial or complete solubilization of cell wall polysaccharides (pectins and celluloses), and hydrolysis of starch and other polysaccharides (Jha, Kingsly,

& Chopra, 2006) The major classes of cell wall polysaccharides that undergo modifications during ripening are pectins, cellulose and hemicelluloses The process is activated by hydrolases and glycosidase (Fuchs, Pesis, & Zauberman, 1980; Muda, Seymour, Errington, & Tucker, 1995)

The volatile compounds that are involved in mango flavor are produced through metabolic pathways during ripening, harvesting and post-harvest storage, and depend on many factors related to the species, variety and type of technological treatments More than one hundred volatiles were detected Terpene hydrocarbons (monoterpene and sesquiterpene hydrocarbons) are the major volatiles of almost all varieties The dominant terpenes are δ-3-carene, α-pinene, α-phellandrene, α-terpinolene and β-caryophyllene In addition, many varieties have considerable amounts of oxygenated volatile compounds, such as esters (e.g ethyl 2-methylpropanoate, ethyl butanoate, methyl

benzoate), aldehydes [e.g (E)-2-nonenal, (E,Z)-2,6-nonadienal, decanal],

furanones (e.g 2,5-dimethyl-4-methoxy-3(2H)-furanone), lactones (e.g hexalactone), C13-norisoprenoids [e.g (E)-β-ionone], etc (Pino, Mesa, Muñoz,

γ-Martí, & Marbot, 2005; Pino & Mesa, 2006) Maturity stage was found to affect overall flavor of mangoes Mango harvested at the green stage exhibited higher amount of monoterpenes, sesquiterpenes, and aromatic compounds, but fruits harvested at the fully ripe stage had higher concentrations of esters, alkanes, and norisoprenoids (Lalel, Singh, & Tan, 2003a)

1.2 Mango wine

The shelf-life of mango is short and its availability is mostly restricted to one season, therefore processed products are likely to have better prospects

than fresh fruit in the export market Various methods of post-harvest handling have been employed to extend the shelf life of mango fruits and reduce losses, through inhibition of respiration, refrigeration, dehydration, heat treatment and ethylene production, which slows deterioration and senescence Besides, efforts have been made to use the surplus fruits for various purposes and one

of the alternatives is wine production Setting-up of fruit wineries could result

in economic growth, generating employment opportunities and providing better returns of produce to the orchardists (Reddy & Reddy, 2005; Reddy & Reddy, 2009)

Research on mango wine still lacked intensive drive till recently Czyhrinciwk (1966) reported the first study on mango wine production by testing „Hilacha‟ variety for possibility of converting mango into wine The study concluded that mango was an excellent raw material for production of good-quality fruit wines and was suitable to be fermented into white semi-sweet table wine (Czyhrinciwk, 1966) In addition, twenty varieties of mangoes from India were screened for production of dessert and madeira-style wine and tested for influence of madeirization on organoleptic quality

(Onkarayya & Singh, 1984) Obisanya et al (1987) studied the fermentation

of mango juice into wine using locally isolated yeasts Saccharomyces

cerevisiae and Schizosaccharomyces species of palm wine and they concluded

that Schizosaccharomyces yeasts were suitable for the production of sweet, table mango wine and Saccharomyces yeasts were suitable for the production

of dry mango wine with a higher ethanol level (Obisanya, Aina, & Oguntimein, 1987)

Some recent studies about mango wine started in India in 2005 A method

of mango juice extraction with pectinase was developed Ethanol as well as some volatile contents of mango wine was characterized It was concluded that the aromatic compounds of mango wine were comparable in concentration to those of grape wine (Reddy & Reddy, 2005) Further results of characterizing kinetic changes of higher alcohols in mango wine were reported and it was concluded that pectinase treatment could enhance the mango juice yield and increase the synthesis of higher alcohols (within a desirable range) as well as mango wine quality (Reddy & Reddy, 2009) Response surface methodology

(RSM) was used to adjust temperature, pH and inoculum levels of

monocultures of S bayanus to optimize the production of ethanol, glycerol,

and volatile acidity They reported that the predicted value for optimisation process conditions were in good agreement with experimental data (Kumar, Prakasam, & Reddy, 2009) Furthermore, influence of condition factors (temperature, pH, SO2, and aeration) on mango wine fermentation was further assessed based on yeast growth, duration, fermentation rate and volatile composition (Reddy & Reddy, 2011) They concluded that fermentation temperature and pH had more influence on volatile composition and yeast biomass Higher temperature induced a decrease of ethyl acetate and certain higher alcohols as well as an increase of biomass SO2 having a critical value between 100-300ppm, below which, SO2 was able to stimulate yeast growth and fermentation power, but above the critical value SO2 would inhibit yeast growth In addition, initial oxygen in the juice was optimal for yeast growth and better quality of mango wine

Various yeast and bacteria strains were used to optimize the flavor of

mango wine Metschnikowia pulcherrima and Torulaspora delbrueckii were separately mixed with S cerevisiae at a ratio of 10:1 The two non-

Saccharomyces yeasts were unable to complete the fermentation, but the

mixed cultures were able to produce similar levels of ethanol to the single

culture of S cerevisiae, with extra advantages of higher glycerol but lower

volatile and total acidity (Varakumar, Kondapalli, & Obulam, 2012) In addition, malolactic fermentation was also carried out in mango wine

fermentation Oenococcus oeni was inoculated simultaneously and sequentially with S cerevisiae The results indicated that wine inoculated with

O oeni gave higher sensory impact than the control wine (monoculture of S cerevisiae), and simultaneous inoculation showed better sensorial attributes in

flavor, fruity aroma, and overall acceptability than sequential inoculation (Varakumar, Naresh, Variyar, Sharma, & Reddy, 2013)

Furthermore, the complete analysis of volatile compounds of mango wine was finished by two other groups (Reddy, Kumar, & Reddy, 2010; Pino & Queris, 2011) Reddy‟s group used liquid-liquid extraction followed by GC-

MS and they reported the concentration of major alcohols, esters, fatty acids

and ketones of mango wine made from „Banginapalli‟ and „Alphonso‟ varieties Pino and Queris (2011) chose „Haden‟ variety for fermentation They used continuous solvent extraction and analyzed the wine samples by GC-FID and GC-MS Forty esters, 15 alcohols, 12 terpenes, 8 acids, 6 aldehydes and ketones, 4 lactones, 2 phenols, 2 furans and 13 miscellaneous compounds were identified Besides ethanol, some higher alcohols were the major constituents of the volatile profile In addition, they tentatively estimated the contribution of the identified volatile compounds to the aroma of mango wine on the basis of their odor activity values, which indicated potent olfactory compounds such as ethyl butanoate and decanal

The non-volatile studies were focused on carotenoids and phenolic compounds as well as their antioxidants effect The composition and concentration of polyphenols were analyzed using LC-MS in mango wines made from eight different varieties (Kumar, Varakumar, & Reddy, 2012) They compared total antioxidant capacity, radical-scavenging activity, and ferric reducing antioxidant power among these eight varieties of mango wine using ABTS, DPPH and FRAP assay, respectively In addition, major carotenoids were investigated by HPLC in wine from seven mango varieties Generally, concentration of carotenoids decreased throughout fermentation, but the decrease of lutein was more significant than β-carotene (Varakumar, Kumar, & Reddy, 2011) The highest radical-scavenging activity was shown

in varieties of „Alphonso‟, „Sindhura‟ and „Banginapalli‟, respectively, whereas „Alphonso‟, „Banginapalli‟ and „Sindhura‟ had shown higher inhibitory effects on low-density lipoprotein oxidation

1.3 General comments of wine flavor

Among the various factors that contribute to the enjoyment of wine, flavor

is possibly the most important one Flavor is the result of the interaction of chemical constituents with the senses of taste and smell of the consumer Wine flavor is composed of non-volatile chemicals which lead to taste sensations and volatile compounds which are responsible for the odor The non-volatiles include sugars, organic acids, polyphenols, etc., which cause sweetness,

sourness and astringency, respectively The concentration of non-volatiles needs to be at level of ~1% to influence wine taste (Rapp & Mandery, 1986) The volatiles include alcohols, esters, aldehydes, ketones, lactones, fatty acids, benzene derivatives, terpenes, C13-norisoprenoid, thiols, etc The volatiles may originate from fruit before or after viticultural processing (primary aroma), fermentation (secondary aroma) or aging process (tertiary aroma) The concentration of volatiles ranges from 10-1 to 10-10 g/L and wine aroma is formed by the balance of these volatile components (Rapp & Mandery, 1986)

Volatiles such as monoterpene hydrocarbons, sesquiterpene hydrocarbons and C13-norisoprenoid from fruit (e.g Muscat grape, mango) dominate the primary aroma These compounds were formed along with growth of fruit Terpenes were derived from the universal C5 precursor isopentenyl diphosphate (IPP) and its allylic isomer dimethylallyl diphosphate (DMAPP), which in fruit cells were generated from two independent pathways located in cytosol and plastids (El Hadi, Zhang, Wu, Zhou, & Tao, 2013) In addition, C13-norisoprenoid was usually formed by oxidative cleavage of carotenoids via three steps: an initial dioxygenase cleavage yielding apocarotenoids, followed by enzymatic transformations of these apocarotenoids leading to the formation of polar aroma precursors, and finally acid-catalyzed conversions of these precursors to volatile compounds (Winterhalter & Rouseff, 2002)

In the secondary aroma, alcohols and esters played a major role In glycolysis, each glucose molecule is metabolized into two pyruvate molecules Afterwards, pyruvate is decarboxylated into acetaldehyde and carbon dioxide

by pyruvate decarboxylase, followed by reduction of acetaldehyde into ethanol

by alcohol dehydrogenase Higher alcohols can be formed from amino acids (valine, leucine, isoleucine, methionine and phenylalanine) through the Ehrlich Pathway: after the initial transamination reaction, the resultant α-keto acids will be further converted to respective higher alcohols such as active amyl alcohol, 2-phenylethyl alcohol and methionol (Hazelwood, Daran, van Maris, Pronk, & Dickinson, 2008) Some higher alcohols can also be formed through sugar metabolic pathway since α-keto acid can be formed in tricarboxylic acid cycle (TCA cycle) (Pietruszka, Pielech-Przybylska, & Szopa, 2010)

Esters are another important group of volatiles There are two major groups of esters in wine: acetate esters and ethyl esters The two groups are of particular interest, as the presence of these esters determines the fruity aroma

of wine Acetate esters are formed by acetyl-CoA reacting with higher alcohols by alcohol acetyl-transferase (AATase) (Fujii, Nagasawa, Iwamatsu, Bogaki, Tamai, & Hamachi, 1994) Ethyl esters are produced by acyl-transferase reactions in which ethanol react with fatty acyl-CoAs derived from fatty acid metabolism (Saerens, Delvaux, Verstrepen, Van Dijck, Thevelein, & Delvaux, 2008) The final concentration of esters is dependent on the balance between synthesis and hydrolysis of esters (Fukuda et al., 1998) In addition, there is a large variation in ester production among yeast strains Other secondary aroma compounds include volatile fatty acids, aldehydes, ketones, etc

After the major modification in composition during fermentation, chemical constituents generally react slowly during ageing resulting in gradual changes

in aroma, the tertiary aroma (Sumby, Grbin, & Jiranek, 2010) In the first few years of ageing, acetate esters are usually diminished which is due to acid-catalyzed hydrolysis, while a range of straight chain fatty acid ethyl esters could remain constant (Rapp, Guentert, & Ullemeyer, 1985; Rapp & Mandery, 1986) The decrease in the concentration of acetate esters might cause the loss

of freshness and fruitiness especially of white wines This phenomenon could

be explained by the fact that hydrolysis of ethyl esters is slower than that of acetate esters considering the high concentration of the hydrolysis product ethanol which may cause inhibition of ethyl ester hydrolysis (Rapp & Mandery, 1986)

Ethyl esters of diprotic acids such as ethyl tartrate show increases caused

by chemical esterification during ageing Therefore, the ratio of acetate to ethyl esters and build-up of ethyl tartrate could be considered as ageing markers (Rapp & Marais, 1993) During ageing, while free monoterpenes slowly undergo oxidation and sensory extinction, acid-catalyzed or enzyme-catalyzed hydrolysis is able to replenish the lost free monoterpenes from the reservoir of still-bound monoterpenes glycosides (Rapp & Mandery, 1986; Rapp & Marais, 1993) Additionally, compounds derived from carotenoid

degradation are of particular interest with reference to wine ageing such as l,1,6-trimethyl-1,2-dehydro-naphthaline and β-damascenone (Marais, 1992; Pineau, Barbe, Van Leeuwen, & Dubourdieu, 2007) The two compounds may diversify wine bouquet by adding in kerosene, floral and tobacco notes Further, oxygen level should be seriously controlled and its concentration is related to the loss of fruitiness and the concentration of aldehydes, acetic acid and other acetals Use of oak barrel for ageing may lead to extraction of more phenolic compounds from lignin degradation and several fragrant lactones (Rapp & Mandery, 1986)

1.4 Flavor modulation by different techniques

The fermentation of fruit and production of wine is a complex ecological and biochemical process involving the interaction of different microbial species, such as fungi, yeasts, lactic acid bacteria as well as acetic acid bacteria, (Fleet & Heard, 1993) Yeasts are predominant species among these microbes and they are the primary catalysts of the bioconversion of fruit juice into wine (Pretorius, Van der Westhuizen, & Augustyn, 1999) In spontaneous fermentation, the yeasts generally come from the surface of fruit skins or surface of fermentation equipment

Non-Saccharomyces such as Kloeckera, Hanseniaspora and Candida predominate in the early stages, followed by several species of Metschnikowia and Pichia in the middle stages when the ethanol reaches 3-4% The latter

stages of spontaneous fermentation would usually be dominated by the alcohol

tolerant strains of S cerevisiae (Pretorius et al., 1999) Other

non-Saccharomyces which may be involved in wine fermentation include Brettanomyces, Kluyveromyces, Schizosaccharomyces, Torulaspora and Zygosaccharomyces (Jolly, Augustyn, & Pretorius, 2006) However, in

modern wine fermentation, winemakers prefer to produce wine with predictable quality and to be able to control and modulate wine quality Therefore, they use physical or chemical means to remove the microbes in the juice and inoculate selected starter cultures of yeasts to start the fermentation

Below is a list of several techniques which are important for controlling and modulating wine flavor in modern wine fermentation

1.4.1 Selection of Saccharomyces

Although some small wineries still prefer using natural spontaneous fermentations, mass production needs rapid, reliable and controllable fermentations to produce wine with consistent flavor and predictable quality This necessitates the use of selected pure yeasts of known characters Large wineries will be the main beneficiaries of yeast strain development which continually produces new yeast strains that have more reliable performance and therefore facilitate the production of affordable high-quality wines

Besides the primary role of Saccharomyces to efficiently convert sugars

into ethanol without the development of off-flavors, yeast should possess properties such as high tolerance to sulfite, osmotic stress, ethanol and copper; genetic stability; production of glycerol and β-glucosidase; minimal lag phase

on rehydration; complete fermentation of sugars at low temperature; and limited production of foam, sulfur dioxide, hydrogen sulfide, volatile acidity, acetaldehyde, pyruvate, ethyl carbamate precursors and polyphenol oxidase (Pretorius, 2000; Suárez-Lepe & Morata, 2012) The importance of these additional yeast characteristics largely depends on the type and style of wine

to be made and the technical requirements of the winery

Saccharomyces strains with specific characteristics are preferred when

making different types of wine Wine yeast strains of Saccharomyces are classified into several different species and/or strains, including S bayanus, S

beticus, S capensis, S chevaleri, S ellipsoideus, S fermentati, S oviformis, S rosei and S vini (S fermentati and S vini were re-classified as T delbrueckii)

(Lambrechts & Pretorius, 2000; Pretorius, 2000; Suárez-Lepe & Morata, 2012)

S ellipsoideus is widely used for the production of dry wine,

ethanol-tolerant and flocculent strains with autolytic properties Other examples

include S bayanus and S oviformis which are preferred for the production of

bottle-fermented sparkling wine, film-forming strains with strong oxidative

and osmotolerant strains forming little or no volatile acids (e.g S rosei) for sweet wine Further, selection of Saccharomyces yeasts also plays a very

important part in wine flavor modulation and different strains may lead to different aroma bouquet (Pretorius, 2000)

In addition, Saccharomyces yeasts are efficient producers of higher alcohols as well as esters, especially ethyl esters S cerevisiae strains that

produce only small amounts of higher alcohols reduce the intensity of this winey aroma, helping to highlight the varietal aromas valued by consumers Selecting yeast strains with a low propensity to produce higher alcohols would help highlight varietal aromas (Suárez-Lepe & Morata, 2012) Higher alcohols and acetyl-CoA are the main precursors of acetate esters that increase wine flavor complexity Therefore, a good compromise of higher alcohol production is needed to ensure the existence of both varietal and fruity ester aroma in wine

1.4.2 Selection of non-Saccharomyces

Wines produced with pure Saccharomyces yeasts often lack complexity of

flavor, stylistic distinction and vintage variability (Lambrechts & Pretorius,

2000) The use of selected non-Saccharomyces strains with enhanced flavor

compounds provides a potential solution to this issue The role of

non-Saccharomyces in wine production has been extensively debated Some old

points always correlate non-Saccharomyces with low SO2 resistance, low fermentation power and off-odor production (e.g acetic acid, ethyl acetate and acetaldehyde) (Ciani, Comitini, Mannazzu, & Domizio, 2010) Nowadays,

many non-Saccharomyces are isolated and studied individually Results showed that many non-Saccharomyces could positively influence the chemical

and sensory characteristics of wine (Jolly et al., 2006)

Kloeckera apiculate (teleomorphic form Hanseniaspora uvarum), which is

one of the most predominant non-Saccharomyces in grape must, can enhance

the production of acetate esters such as 2-phenylethyl acetate and isoamyl acetate, and these esters provide rose and banana flavor to wine (Moreira, Mendes, Guedes de Pinho, Hogg, & Vasconcelos, 2008) Other efficient ester

producers include H guilliermondii, Pichia anomala, etc (Rojas, Gil, Piñaga,

& Manzanares, 2001, 2003; Moreira et al., 2008) In addition, C stellata is

able to produce higher concentrations of glycerol which can improve wine

viscosity and smoothness (Ciani & Picciotti, 1995) Also, many Candida yeasts such as C pulcherrima produces high levels of extracellular enzymes

such as β-glucosidase to release bound-form aroma compounds (Charoenchai,

Fleet, Henschke, & Todd, 1997) Torulaspora delbrueckii is helpful in controlling acetic acid in wine Unlike S cerevisiae, T delbrueckii does not

respond to the hyper-osmotic (high sugar) medium with increasing acetic acid production (Bely, Stoeckle, Masneuf-Pomarède, & Dubourdieu, 2008)

Kluyveromyces wickerhamii and K phaffi have potential to produce a killer

toxin as an anti-microbial agent against spoilage yeast Dekkera/Brettanomyces

(Ciani & Fatichenti, 2001; Comitini, Ingeniis De, Pepe, Mannazzu, & Ciani, 2004)

In this study, Williopsis saturnus var mrakii (formerly Hansenula

saturnus var mrakii) was chosen because of its competency in producing

acetate ester such as isoamyl acetate and 2-phenylethyl acetate (Erten & Campbell, 2001; Yilmaztekin, Erten, & Cabaroglu, 2009; Lee, Ong, Yu, Curran, & Liu, 2010a) The yeast was firstly used in sake to improve its fruity character (Inoue, Fukuda, Wakai, Sudsai, & Kimura, 1994), and also has been inoculated to improve fruity flavor of grape wine (Erten & Tanguler, 2010) and papaya wine (Lee, Ong, Yu, Curran, & Liu, 2010b)

1.4.3 Mixed-culture fermentation

Spontaneous fermentations might be able to produce wines with an incredible aromatic complexity but the uncertainty and unpredictability is always its Achilles‟ heel Fortunately, controlled mixed-culture fermentation may resolve the uncertainty issue by removal of microbes in juice first and

inoculation of Saccharomyces and non-Saccharomyces simultaneously or

sequentially at a certain ratio Different combinations of yeasts would give particular contributions to wine flavor, which mimics the process of spontaneous fermentation but with much control on the whole process Some

oenological disadvantages of non-Saccharomyces might not be expressed or may be modified by Saccharomyces, and the advantages of having both kinds

of yeast might be properly manifested (Ciani et al., 2010) This may

reintroduce the advantages of the non-Saccharomyces yeasts, avoiding the

usual risks connected with a natural fermentation and the standardization of

aromatic profiles experienced when single strains of S cerevisiae are used to

ferment musts

Commercially, Chr Hansen developed the use of non-Saccharomyces yeast strains (K thermotolerans and T delbrueckii) in blends with

Saccharomyces at different percentages, which is able to increase smoothness

and body to the palate (Ciani et al., 2010), and give elegance and a wild character to the aromatic complexity of wine Three factors are important for mixed-culture fermentation: firstly, selection of suitable yeast strains;

secondly, ratio of Saccharomyces to non-Saccharomyces; and thirdly, suitable

inoculation strategy (simultaneous or sequential)

When selecting suitable yeast strains, performance and characteristics of single cultures of each yeast strain must be well studied beforehand As mentioned in Section 1.3.1 and 1.3.2, advantages and disadvantages vary significantly among different yeast strains Good knowledge about the yeasts performance and their optimal growth conditions are very helpful to correctly select suitable yeasts to cater to different fermentation requirements

In mixed culture fermentation, the ratio of Saccharomyces to

non-Saccharomyces yeasts is an important parameter that determines the quality of

grape wine The effects of different inoculum ratios of W saturnus : S

cerevisiae = 1:1, 2:1, 10:1, 20:1 on fermentation of „Emir‟ grape must were

studied Different inoculum ratios influenced the yeast growth, chemical

composition and volatile compounds Lower W : S ratio led to a wider range

of compounds, higher amounts of volatile compounds, lower levels of acetic

acid and higher levels of esters than monocultures, therefore, W : S =1:1

appeared to be suitable (Tanguler, 2012)

The impact of mixed cultures of H osmophila and S cerevisiae with

different initial yeast ratios on grape wine composition has also been

examined Ratios of H : S =9:1, 3:1, 1:1, 1:3, 1:9, 0:100 were tested In contrast, H : S.=9:1, i.e a higher concentration of non-Saccharomyces, was

selected because it was appropriate to produce wines of desired quality with similar levels of ethanol as the monoculture, enhanced 2-phenyethyl acetate production and decreased levels of ethyl acetate (Viana, Gil, Vallés, & Manzanares, 2009) Several other reports showed similar results as Viana et al

(2009), meaning higher starting concentrations of non-Saccharomyces led to a

wider range, higher intensity and more complexity of wine aroma (Jolly, Augustyn, & Pretorius, 2003; Lee et al., 2010b; Trinh, Woon, Yu, Curran, & Liu, 2011)

The inoculation in mixed-culture fermentation could be simultaneous or sequential However, there was no consistent conclusion indicating which strategy was better for wine flavor Recently, the impact of simultaneous and

sequential inoculation of T delbrueckii-S cerevisiae cultures in high sugar

fermentation of grape wine was compared to determine whether it can improve the quality of wines and reduce the acetic acid content (Bely et al.,

2008) Simultaneous fermentation at a ratio of T.: S =20:1 was better at

controlling acetic acid and acetaldehyde than sequential fermentation In

addition, Ciani et al (2006) pointed out several limitations with sequential fermentations such as the persistence of the non-Saccharomyces yeasts during fermentation resulting in limited sugar residue for sequential cultures of T

delbrueckii and K thermotolerans and excessive increase in ethyl acetate for

sequential trials with H uvarum (Ciani, Beco, & Comitini, 2006) In contrast,

sequential fermentation led to more acetate esters and fruitiness than simultaneous fermentaiton in papaya wine fermentation (Lee, Chong, Yu, Curran, & Liu, 2012a) In addition, the sequence of adding yeasts would affect

final aroma profile The result showed that inoculation of S cerevisiae seven days after W saturnus provide more intense fruitiness than inoculation of W

saturnus two days after S cerevisiae

1.4.4 Yeast interactions in mixed-culture fermentation

In simultaneous mixed-culture fermentation, the growth of many

non-Saccharomyces species is only limited to the first few days (Pretorius, 2000;

Bely et al., 2008; Erten & Tanguler, 2010) The early death phenomenon may

seriously compromise the contribution of non-Saccharomyces to wine flavor

On the other hand, the growth of Saccarhomyces might be influenced by

non-Saccarhomyces in sequential mix-culture fermentation For example, W mrakii would experience an early growth arrest in simultaneous mixed-culture

fermentation with S cerevisiae (Erten & Tanguler, 2010) Interestingly, the growth of S cerevisiae could also be inhibited by W mrakii in sequential mixed-culture fermentation where W mrakii was inoculated into the medium first for 7 days (Lee et al., 2012a; Lee, Kho, Yu, Curran, & Liu, 2012b) W

mrakii is known for its efficiency in producing acetate esters (Inoue et al.,

1994) Researchers used this yeast to improve the fruity and floral characters

in various alcoholic beverages Therefore, study about the interaction

mechanism of Saccharomyces and non-Saccharomyces in mixed-culture

fermentation would be helpful for optimization of inoculation strategy

The two main types of interaction that modulate the development of different yeast populations during alcoholic fermentation are the release of

“toxic” compounds and cell-cell contact inhibition (space competition) Firstly, the “toxic” compounds include ethanol, medium-chain fatty acids, peptides, glycoproteins, etc (Lafon-Lafourcade, Geneix, & Ribereau-Gayon, 1984; Gao & Fleet, 1988; Marquina, Santos, & Peinado, 2002; Pina, Santos, Couto, & Hogg, 2004; Albergaria, Francisco, Gori, Arneborg, & Gírio, 2010)

To prove the presence of “toxic” compounds, cell-free supernatants of the mixed culture at different growth stages were inoculated with the non-

Saccharomyces (Pérez-Nevado, Albergaria, Hogg, & Girio, 2006) and it was

found that toxic compounds produced by S cerevisiae triggered the early cell arrest of the H guilliermondii cells in the mixed cultures with S cerevisiae

Furthermore, it was found that the killing effect of mixed supernatants towards

H guilliermondii was inactivated by protease treatments, which revealed the

proteinaceous nature of the toxic compounds (Albergaria et al., 2010) Further results indicated that the (2–10) kDa protein fraction of those supernatants

seemed to contain antimicrobial peptides active against H guilliermondii Space competition with Saccharomyces in mixed culture fermentation could also lead to early death of non-Saccharomyces A dialysis tubing was

used to separate two yeasts into two compartments and proved that the early

death of K thermotolerans and T delbrueckii was regulated by a cell-cell

contact mechanism, being dependent on the presence of viable S cerevisiae

cells at high concentrations (Nissen & Arneborg, 2003; Nissen, Nielsen, & Arneborg, 2003) It was finally concluded that this cell-cell contact inhibition

was due to a response to space limitation and S cerevisiae was more competent to compete for space than non-Saccharomyces Recently, a double-

compartment fermenter was designed to further confirm the mechanism of

cell-cell contact inhibition between T delbrueckii and S cerevisiae (Renault,

Albertin, & Bely, 2013)

Other possible reasons for inducing early death of non-Saccharomyces in

mixed-cultures could be nutrient depletion (e.g nitrogen, oxygen), quorum sensing, intolerance to sulfite, etc (Hansen, Nissen, Sommer, Nielsen, & Arneborg, 2001; Cocolin & Mills, 2003; Fleet, 2003; Cheraiti, Guezenec, & Salmon, 2005)

1.4.5 Enzymes

It is now well established that, apart from free aroma compounds, a significant part of several aroma compounds is accumulated in fruits as odorless non-volatile glycosides (Adedeji, Hartman, Lech, & Ho, 1992; Drider, Janbon, Chemardin, Arnaud, & Galzy, 1994) The aglycone moieties

of glycosides include monoterpenes, terpenols, benzene derivatives, alcohols, aldehydes, acids, esters, C13-norisoprenoids and other compounds To date, various studies have been conducted on glycosidically-bound volatiles in different mango cultivars, most of which possess pleasant floral and fruity aromas with low perception thresholds (Adedeji et al., 1992; Drider et al., 1994; Ollé, Baumes, Bayonove, Lozano, Sznaper, & Brillouet, 1998; Lalel, Singh, & Tan, 2003b)

Acid and/or enzyme hydrolysis of glycosidically bound volatiles could allow the liberation of free volatiles (aglycones) in different fruit juices or wines such as grape, apple and passion fruits (Sarry & Günata, 2004) Acid hydrolysis of glycosides occurs very slowly and may induce compounds (e.g terpenol) rearrangements Enzymatic hydrolysis can be much more effective and specific in liberating aromatic compounds from glycosides and minimise compounds rearrangements (Gunata, Bitteur, Brillouet, Bayonove, &

Cordonnier, 1988) Mango fruits possess glycosidases on their own and their activities are enhanced during the maturation process (Lalel et al., 2003a) The glycosidically-bound aroma compounds therefore continuously increase in the pulp as maturity progresses, but the activities of glycosidases in the fruit are greatly reduced at typical wine fermentation pHs (Lalel et al., 2003b) Some yeast strains are efficient to produce glycosidases (Charoenchai et al., 1997; Mateo & Di Stefano, 1997; Palmeri & Spagna, 2007) On the other hand,

glycosidases from fungi such as Aspergillus niger have shown very good

stability and activity in wine and nowadays these enzymes are extracted purified and sold commercially (Sarry & Günata, 2004)

In mango wine fermentation, mango pulp was a reservoir of abundant glycosides that can serve as aroma precursors and the aroma compounds can

be released during fermentation by yeast or exogenous glycosidases However, this aspect has not been examined in mango wine fermentation and provides a potential method for enhancing the mango varietal aroma into mango wine by fermentation with the inclusion of pulp (mimicking the maceration process of red grape wine fermentation) and addition of exogenous glycosidase

1.5 Objectives of project

The overall aim of this project was to investigate the chemical profile of

mango wine fermented by monoculture and mixed-culture of S cerevisiae and

W saturnus with the intention to modulate the mango wine flavor and to

develop a new tropical fruit wine “mango wine” Besides, different aspects to improve wine flavor were also studied, including selection of suitable mango varieties, juice maceration, enzyme addition and yeast interaction mechanism

In Chapter 3, to investigate the impact of S cerevisiae on the chemical profile of mango wine, three strains of S cerevisiae (EC-1118, MERIT.ferm,

CICC1028) were tested and one was selected for further study of culture fermentation

mixed-In Chapter 4, to investigate the impact of W saturnus on the chemical profile of mango wine, two strains of W saturnus (NCYC500, NCYC2586)

were tested and one was selected for further study of mixed-culture fermentation

In Chapter 5, to investigate the impact of mixed culture of S cerevisiae (MERIT.ferm) and W saturnus var mrakii (NCYC500) on the chemical,

volatile and sensorial profile of mango wine, four different ratios (1:1, 1:10, 1:100, 1:1000) were pre-screened and one ratio was selected for final study This study was completed in conjunction with Miss Chan Li Jie, FST Master Project 2010/2011

In Chapter 6, to investigate the distinction of chemical, volatile and sensorial profile of different varieties of mango wine, three cultivars of

mangoes were selected Mixed-culture fermentation with S cerevisiae (MERIT.ferm) and W mrakii (NCYC500) at ratio of 1:1000 was used This

study was completed in conjunction with Miss Chan Li Jie, FST Master Project 2010/2011

In Chapter 7, to investigate the antagonistic mechanism between S

cerevisiae (MERIT.ferm) and W mrakii (NCYC 500), several

sub-experiments were conducted

In Chapter 8, to enrich flavor compounds of mango wine, pulp maceration was included during fermentation This study was completed in conjunction with Mr Lim Sien Long, FST Honors Project 2011/2012

In Chapter 9, to release bound-form aroma compounds, exogenous glucosidase was added after fermentation This study was completed in conjunction with Mr Lim Sien Long, FST Honors Project 2011/2012

β-CHAPTER 2 Materials and Methods 2.1 Fruits and yeast strains

The main mango variety chosen in this study was Mangifera indica L cv

„Chok Anan‟ (means „luck of infinity‟) from Malaysia „Chok Anan‟ mango

has the ability to produce off-season flowering without chemical induction

(Spreer et al., 2009) Thus, there might be two or three harvests in a year This

characteristic enables „Chok Anan‟ mangoes to have a large stock each year,

which gives it an advantage to be a raw material for fermentation In addition,

ripe „Chok Anan‟ mangoes have a high content of sugar (16.70o Brix) and very strong terpene aroma Other varieties used in this study included „R2E2‟ (means the mango was derived from Row 2 Experiment 2 from the row and position in the field of the original tree at the DEEDI´s Bowen Research Station) from Australia, „Nam Doc Mai‟ (means „sweet water from flowers‟) from Thailand and „Harum Manis‟ (means „fragrant and sweet‟) from Indonesia „R2E2‟ mango is a large, attractive mango with strong red blush, low fiber and long shelf life, and it is known for its sweet and creamy flavor which is due to high concentration of esters and lactones 'Nam Doc Mai' mango is among the best dessert mangos of Thailand, with an exceptional appearance and eating quality (non-fibrous flesh) When ripe, this mango has smooth, silky texture and extreme sweetness and complex bouquet „Haum manis‟ mango is known for its green skin, golden flesh and terpenic (exotic) flavor These mangoes are abundantly available in local markets in Singapore

The following yeast strains were used in this study: Saccharomyces

cerevisiae var bayanus Lalvin EC1118 (Lallemand Inc, Brooklyn Park,

Australia), Saccharomyces cerevisiae var chevalieri CICC1028 (China Centre

of Industrial Culture Collection, Beijing), Saccharomyces cerevisiae MERIT.ferm (Chr.-Han., Denmark), Williopsis saturnus var mrakii NCYC500 and W staurnus var suaveolens NCYC2586 (National Collection

of Yeast Cultures, Norwich, UK) All strains were obtained in freeze-dried form The freeze fried powder or pellet was reconstituted in peptone water, which was followed by streak plating on potato dextrose agar (PDA, Oxoid, Hampshire, England), and incubated at 25oC After 48 h, pure colonies were picked from PDA plates and re-cultured in sterilized nutrient broth [2% D-

glucose (Sigma-Aldrich, Singapore), 0.25% yeast extract (Oxoid), 0.25% malt extract (Oxoid), 0.25% bacteriological peptone (Oxoid), pH=5] for 48 h at

25oC until the cell count reached about 107 CFU/mL The pure culture with added 25% glycerol was dispensed in 1.5 mL sterilized culture tubes and stored at -80oC until use

2.2 Preparation of mango juice and pre-culture

Mango fruits were washed, peeled, cut and then juiced by using a juice extractor (Sona, Cahaya Electronics, Singapore) and centrifuged at 41415g (Beckman Coulter Allegra 64R Centrifuge, Brea, CA, USA) for 10-15min at

4oC to remove the pulp Polyallomer centrifuge tubes (29×104 mm, 50-mL capacity, Beckman instrument Inc., Palo Alto, CA, USA) were used for the centrifugation process The centrifuged juice without sugar adjustment was stored at -50oC before use The juice treatment process was slightly different

in experiments described in Chapters 8 and 9 The mangoes were juiced and divided into two lots The first lot was centrifuged as described above, and the other lot was not centrifuged Both of the centrifuged and un-centrifuged juices without sugar adjustment were stored at -50oC before use

The original pH values for four varieties of mango were 4.63 for „Chok anan‟, 5.00 for „R2E2‟, 4.45 for „Harum manis‟, 5.45 for „Nam Doc Mai‟ In Chapters 3-6, juice was adjusted to pH=3.5 with 3.7 M food-grade D/L-malic acid (Suntop Ltd, Singapore), and treated with 100 ppm potassium metabisulfite (K2S2O5, Goodlife Homebrew center, Norfolk, England) to kill wild microorganisms prior to fermentation However, in Chapters 8 and 9, pasteurization at 60oC for 15-20 min was done after pH adjustment and

K2S2O5 addition, since pH adjustment plus 100 ppm K2S2O5 were not sufficient to kill wild microorganisms in the un-centrifuged mango pulp To be consistent, the centrifuged mango juice was also pasteurized at 60oC for 15-20 min in Chapter 8 and 9

In Chapters 3-6, a pre-culture was prepared from inoculation of 10% (v/v)

of selected yeasts into mango juice which was aseptically filtered by 0.45 μm polyethersulfone filter membrane (Sartorius Stedium Biotech, Germany) In