Effects of sugar concentration and yeast inoculation strategy on mango wine fermentation

... project was to study the effects of sugar concentration and inoculation strategies on mango wine fermentation Aim 1: Effect of sugar concentration on mango wine fermentation with S cerevisiae MERIT.ferm... investigated the effects of initial sugar concentration on volatile and glycerol production by Saccharomyces cerevisiae MERIT.ferm in mango wine fermentation Generally, high sugar concentration had a... 3.2.5 Effects of initial sugar concentration on volatile production 39 3.2.6 Effects of redox potential on overall wine quality 52 3.3 Conclusion 53 Chapter Effects of Co-fermentation

EFFECTS OF SUGAR CONCENTRATION AND YEAST INOCULATION STRATEGY ON MANGO WINE FERMENTATION CHAN LI JIE (B. Sc. National University of Singapore) A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF SCIENCE DEPARTMENT OF CHEMISTRY NATIONAL UNIVERSITY OF SINGAPORE 2014 DECLARATION I hereby declare that the 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 2010 and June 2014. 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: Chan, L.J., Lee, P.R., Li, X., Chen, D., Liu, S.Q., and 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. http://www.foodbeverageasia.com/ebook/FBA_DecJan2012/index.html. Li, X., Chan, L.J., Yu, B., Curran, P., and Liu, S.-Q. (2014) Influence of Saccharomyces cerevisiae and Williopsis saturnus var. mrakii on mango wine characteristics. Acta Alimentaria. 43(3), 473-481. Chan Li Jie Name Signature i 20 June 2014 Date ACKNOWLEDGEMENTS I would like to express my heartfelt gratitude to my supervisor, Dr Liu Shao Quan for his unrelenting patience, support and assistance in this project. I have been very, very fortunate to have such an understanding supervisor. Next, I would like to thank the Food Science and Technology program for all the assistance they have rendered. I would also like to thank the flavorists at Firmenich Asia for assisting with the sensory evaluation despite their busy schedule. In addition, I would like to thank my research group members, Dr Cheong Mun Wai, Dr Lee Pin Rou, Dr Li Xiao, Dr Sun Jingcan and Ms Chen Dai for their support, concern and presence in my academic and personal life. They truly have made my days in the lab brighter. Last but not least, I would like to thank my family for their unwavering support and faith in me. Without them, I wouldn’t have been the person I am. ii Table of Contents DECLARATION……………………………………………………………………….. i ACKNOWLEDGEMENTS……………………………………………………………. ii Table of Contents……………………………………………………………………….iii List of Tables…………………………………………………………………...………vi List of Figures…………………………………………………………………..………ix Chapter 1 Introduction and Literature Review………………………………………....1 1.1 Mango fruit ......................................................................................................... 1 1.1.1 Nutritional content ........................................................................................ 1 1.1.2 Volatile compounds and mango flavour ....................................................... 2 1.2 Mango wine ........................................................................................................ 3 1.3 Biochemistry of fermentation, wine flavour and quality ................................... 6 1.4 Influence of fermentation conditions and yeast strains ...................................... 8 1.4.1 Amelioration of must .................................................................................... 8 1.4.2 Saccharomyces in wine production .............................................................. 9 1.4.3 Non-Saccharomyces species in wine production ....................................... 10 1.4.4 Inoculation strategies in wine fermentation................................................ 10 1.5 Research aims and objectives ........................................................................... 13 Chapter 2 Materials and Methods…………………………………………………….15 2.1 Mango fruits and preparation of mango juice .................................................. 15 2.2 Yeast and culture media ................................................................................... 16 2.3 Preparation of yeast starter culture ................................................................... 16 iii 2.4 Fermentation of mango juice with different initial sugar concentrations — Chapter 3 .................................................................................................... 17 2.5 Fermentation of mango juice with co-inoculated S. cerevisiae and W. saturnus — Chapter 4 .................................................................................................... 17 2.6 Fermentation of mango juice with different sequential inoculation strategies — Chapter 5 .................................................................................................... 18 2.7 Analytical methods ........................................................................................... 19 2.7.1 pH, oBrix and yeast enumeration ................................................................ 19 2.7.2 Analysis of sugars and organic acids ......................................................... 19 2.7.3 Analysis of volatile compounds ................................................................. 20 2.8 Sensory analysis ............................................................................................... 22 2.9 Statistical analysis ............................................................................................ 23 Chapter 3 Effects of Sugar Concentration on Volatile Production by Saccharomyces cerevisiae MERIT.ferm………………………………...…………….…….24 3.1 Introduction ........................................................................................................... 24 3. 2 Results and discussion ...................................................................................... 25 3.2.1 Mango juice volatile composition .............................................................. 25 3.2.2 Changes in pH and organic acids ............................................................... 31 3.2.3 Yeast growth, total soluble solids and sugar concentration ........................ 32 3.2.4 Glycerol production ................................................................................... 36 3.2.5 Effects of initial sugar concentration on volatile production ..................... 39 3.2.6 Effects of redox potential on overall wine quality ..................................... 52 3.3 Conclusion ............................................................................................................ 53 Chapter 4 Effects of Co-fermentation of Saccharomyces cerevisiae and Williopsis saturnus Yeasts on Volatile Production…………………………………...55 4.1 Introduction ...................................................................................................... 55 4.2 Results and discussion ...................................................................................... 56 iv 4.2.1 Total soluble solids, sugar concentrations and yeast cell count ................. 56 4.2.2 pH and organic acids .................................................................................. 59 4.2.3 Volatile composition of mango wines ........................................................ 59 4.2.4 Miscellaneous compounds........................................................................ 77 4.3 Conclusion ........................................................................................................ 78 Chapter 5 Effects of Different Sequential Inoculation Strategies of Saccharomyces cerevisiae and Williopsis saturnus on Volatile Production……………….79 5.1 Introduction ..................................................................................................... 79 5.2 Results and discussion ...................................................................................... 80 5.3 5.2.1 Physicochemical properties of mango wine ............................................... 80 5.2.2 Yeast biomass evolution ............................................................................ 82 5.2.3 Volatile composition................................................................................... 86 5.2.4 Sensory evaluation .................................................................................... 108 Conclusion ............................................................................................... 109 Chapter 6 General conclusions and recommendations……………………………...110 Bibliography………………………………………………………………………….115 v SUMMARY The project first investigated the effects of initial sugar concentration on volatile and glycerol production by Saccharomyces cerevisiae MERIT.ferm in mango wine fermentation. Generally, high sugar concentration had a negative impact on volatile production but enhanced glycerol production. Significantly lower amounts of esters and higher alcohols and more acetic acid and acetaldehyde were produced in high sugar fermentation. The effects of a non-Saccharomyces yeast, Williopsis saturnus var. mrakii NCYC 500 and a mixed culture fermentation consisting of S. cerevisiae and W. mrakii in the ratio of 1:1000 were then studied in mango wine fermentation. The volatile profile of the mango wine produced by the mixed culture fermentation resembled that of the one fermented with a monoculture of S. cerevisiae, while the mango wine produced by fermentation with W. mrakii differed significantly from the wine fermented with a monoculture of S. cerevisiae and the mixed culture fermentation. W. mrakii produced higher amounts of acetate esters but significantly lower amounts of ethyl esters and fusel alcohols. This is highly likely due to the metabolic activities of the different yeast strains. The effects of varying the sequence of inoculation of S. cerevisiae and W. mrakii were subsequently investigated. Simultaneous mixed culture fermentation (MCF) was conducted by co-inoculation of S. cerevisiae and W. mrakii in the ratio of 1:1000. Negative sequential fermentation (NSF) was conducted by first inoculating S. cerevisiae and allowing the fermentation to proceed for 7 days before inactivating the S. cerevisiae by ultrasonication; then inoculating W. mrakii and allowing the second phase of the fermentation to proceed for 14 days. For positive sequential fermentation (PSF), the sequence of inoculation was reversed. vi The volatile profiles of the mango wines that resulted from the three different sequential inoculation strategies varied significantly. NSF generally produced the least amount of volatile, especially the higher alcohols and esters while PSF produced more of the desirable volatile ester compounds. MCF typically produced levels of volatiles between NSF and PSF. The findings obtained in this study potentially could have an impact on fruit wine production and allow wine makers to design an inoculation strategy that would cater to the desired wine style. vii List of Tables Table 3.1 Volatile compounds in fresh Chok Anan mango juice analysed using HS-SPMEGC-MS/FID .......................................................................................................... 28 Table 3.2 Physicochemical properties, organic acid and sugar concentrations of mango wine before and after fermentation ................................................................................ 32 Table 3.3 Volatiles in mango juice catabolised during fermentation ...................................... 39 Table 3.4 Volatile composition of mango wines with different initial sugar concentrations . 40 Table 3.5 Selected volatiles quantified in mango wine and their odour activity values (OAV) in low (unfortified mango juice 16.6oB), medium (initial TSS of 23°B) and high (initial TSS of 30°B) sugar fermentation .............................................................. 44 Table 4.1 Physicochemical properties, reducing sugar and organic acid content before and after fermentation .................................................................................................. 57 Table 4.2 Complete volatiles for mango wine fermented with yeasts S. cerevesiae MERIT.ferm, W. mrakii NCYC 500 and mixed culture ....................................... 61 Table 4.3 Major volatile compounds of alcoholic fermentation by S. cerevisiae, W. mrakii and a mixed culture of S. cerevisiae and W. mrakii .............................................. 65 Table 5.1 Physicochemical properties, yeast cell count, organic acid and sugar concentrations of mango wines ..................................................................................................... 81 Table 5.2 Complete volatiles for mango wine produced from different inoculation strategies ............................................................................................................................... 87 Table 5.3 Major volatiles quantified for mango wines with different inoculation strategies.. ............................................................................................................................... 92 viii List of Figures Figure 3.1 Changes in yeast cell count and TSS (oBrix) for low ( ), medium ( ) and high ( ) sugar fermentation................................................................................................. 34 Figure 3.2 Fructose ( ), glucose ( ) and sucrose ( ) consumption kinetics of S.cerevisiae during fermentation ............................................................................................... 36 Figure 3.3 Concentrations of glycerol in mango wines for low (16.6oBrix), medium (23oBrix) and high sugar (30oBrix) fermentation.................................................................. 38 Figure 3.4 Metabolic pathways governing the production of ethanol, glycerol and acetic acid………………………………………………………………………………49 Figure 4.1 (a) Changes TSS (°B) during fermentation for S. cerevisiae MERIT.ferm monoculture ( ), W. mrakii NCYC 500 monoculture ( )and mixed culture fermentation ( ) (S. cerevisiae:W. mrakii at a ratio 1:1000) (b) Changes in yeast population during fermentation for S. cerevisiae MERIT.ferm monoculture ( ), W. mrakii NCYC 500 monoculture ( ), S. cerevisiae in mixed culture fermentation ( )and W. mrakii in mixed culture fermentation ( ); S. cerevisiae:W. mrakii at a ratio 1:1000 in mixed culture fermentation ........................................ 58 Figure.4.2 Changes in β-myrcene in the fermentation of mango juice with a monoculture of S. cerevisiae MERIT.ferm ( ), monoculture of W. mrakii NCYC 500 ( ) and a mixed culture of S. cerevisiae and W. mrakii ( ) ................................................. 67 Figure 4.3 Changes in ethanol content in the fermentation of mango juice with a monoculture of S. cerevisiae MERIT.ferm ( ), monoculture of W. mrakii NCYC 500 ( ) and a mixed culture of S. cerevisiae and W. mrakii ( ) ................................................. 68 Figure 4.4 Changes in isobutyl alcohol and 2-phenylethyl alcohol content in the fermentation of mango juice with a monoculture of S. cerevisiae MERIT.ferm ( ), monoculture of W. mrakii NCYC 500 ( ) and a mixed culture of S. cerevisiae and W. mrakii ( ) ........................................................................................................ 70 Figure 4.5 Changes in ethyl octanoate in the fermentation of mango juice with a monoculture of S. cerevisiae MERIT.ferm ( ), monoculture of W. mrakii NCYC 500 ( ) and a mixed culture of S. cerevisiae and W. mrakii ( ) ................................................. 72 Figure 4.6 Changes in 2-phenylethyl acetate and ethyl acetate content in the fermentation of mango juice with a monoculture of S. cerevisiae MERIT.ferm ( ), monoculture of W. mrakii NCYC 500 ( ) and a mixed culture of S. cerevisiae and W. mrakii ( ) ............................................................................................................................... 74 Figure 4.7 Changes in acetic acid content in the fermentation of mango juice with a monoculture of S. cerevisiae MERIT.ferm ( ), monoculture of W. mrakii NCYC 500 ( ) and a mixed culture of S. cerevisiae and W. mrakii ( ) ........................... 76 Figure 5.1 Changes in oBrix values during fermentation for PSF1 ( ), NSF2 ( ) and MCF3 ( ) ............................................................................................................................... 82 ix Figure 5.2 Changes in (a) S. cerevisiae MERIT.ferm for PSF1 ( ), NSF2 ( ) and MCF3 ( ), (b) W. mrakii NCYC 500 cell population during fermentation for PSF1 ( ), NSF2 ( ) and MCF3 ( ) (c) Changes in S. cerevisiae MERIT.ferm ( ) and W. mrakii NCYC 500 ( ) in MCF......................................................................................... 84 Figure 5.3 Changes in ethanol concentration during PSF1 ( ), NSF2 ( ) and MCF3 ( ) ........ 95 Figure 5.4 Changes in isoamyl alcohol, 2-phenylethyl alcohol, isobutyl alcohol during PSF1 ( ), NSF2 ( ) and MCF3 ( ) .................................................................................. 97 Figure 5.5 Changes in linalool during PSF1 ( ), NSF2 ( ) and MCF3 ( ) .............................. 99 Figure 5.6 Changes in (a) ethyl octanoate, (b) ethyl decanoate, (c) ethyl acetate during PSF1 ( ), NSF2 ( ) and MCF3 ( ) ............................................................................... 102 Figure 5.7 Changes in 2-phenylethyl acetate and isoamyl acetate during during PSF1 ( ), NSF2 ( ) and MCF3 ( )....................................................................................... 103 Figure 5.8 Changes in acetic acid and hexanoic acid during PSF1 ( ), NSF2 ( ) and MCF3 ( ) ............................................................................................................................. 106 Figure 5.9 Sensory profile of PSF1 ( ), NSF2 ( ) and MCF3 ( ), mango wine .................... 108 x Chapter 1 Introduction and Literature Review 1.1 Mango fruit 1.1.1 Nutritional content The mango fruit (Mangifera indica L.) is one of the most popular and economically important tropical fruits due to its exotic and appealing flavour and taste. Typically, the pulp of the fruit is consumed, and the skin and seed discarded. The major constituents of the pulp are water, carbohydrates, organic acids, fats, tannins, vitamins and flavor compounds (Sagar et al. 1999); the chemical composition varies with location of cultivation, variety, and stage of maturity (Chauhan et al. 2010). In the Southeast Asian region, the Chok Anan mango is one of the most popular cultivars consumed for its mild and pleasant flavour. Reported to have a total soluble solids content of about 14 to16 °Brix (Vásquez-Caicedo et al. 2002), with sucrose as the predominant sugar (approximately 7.5 g/100 g), followed by fructose and glucose at 5 g/100 g and 1.5 g/100 g respectively (Vásquez-Caicedo et al. 2002). Citric acid (0.32 g/100 g) and malic acid (0.25 g/100 g) are the main organic acids; other organic acids include succinic, oxalic, pyruvic, adipic, mucic, galacturonic and glucuronic acids (Tharanathan et al. 2006). During the ripening of mango, both sugars and pH increase, due to glucogenesis and hydrolysis of polysaccharides (starch), resulting in an increase in sweetness. This accumulation of sugars and organic acids results in an excellent sugar/acid ratio that is responsible for taste development (Chauhan et al. 2010). The major amino acids reported in mangoes are alanine, arginine, glycine, serine, leucine and isoleucine, with a protein content of approximately 0.8 g/100 g (Tandon and Kalra 1984). Lipids constitute less than 1% (w/w), with triglyceride being the major 1 component (Gholap and Bandyopadhyay 1975). Although mango is rich in vitamin C, the maximum level occurs in the early stage of growth instead of ripening stage (Spencer et al. 1956). β-Carotene increases as the mango fruit matures and ripens (Jungalwala and Cama 1963); the increase in β-carotene correlates with a decrease in acids and an increase in sugar content (Godoy and Rodriguez-Amaya 1987). 1.1.2 Volatile compounds and mango flavour Although more than a hundred volatiles have been identified in the mango flavour profile, terpene hydrocarbons (monoterpene and sesquiterpene) make up the predominant class of volatiles in mango flavour across all varieties. Amongst the terpene hydrocarbons, the dominant ones include δ-3-carene, α-pinene, α-phellandrene, α-terpinolene and βcaryophyllene (Chauhan et al. 2010). In addition to these terpene hydrocarbons, many varieties also have considerable amounts of oxygenated volatile compounds, including esters (e.g. ethyl 2-methylpropanoate, ethyl butanoate, methyl benzoate), lactones (e.g. γhexalactone), aldehydes (e.g. cis-2-nonenal, 2,6-nonadienal, decanal), furanones (e.g. 2,5dimethyl-4-methoxy-3(2H)-furanone), C13-norisoprenoids (e.g. β-ionone), etc (Pino et al. 2005; Pino and Mesa 2006). These compounds 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 (Léchaudel and Joas 2007). In addition, the maturity of the fruit at harvest was also found to affect the overall flavour of mangoes. Mangoes harvested at the green stage exhibited a higher amount of monoterpenes, sesquiterpenes and aromatic compounds while fruits harvested at the fully ripe stage had higher concentrations of esters, alkanes, and norisoprenoids (Lalele et al. 2003a). Generally regarded as the terpinolene type, the Chok Anan mango is known for its pine needle-like terpene note (Chauhan et al. 2010), likely due to the major volatiles present being 2 δ-thujene, α-pinene, δ-3-carene, mycrene, α-phellandrene and α-terpinolene. α-Terpinolene, the major compound, has been described as sweet, floral with pine-like aroma notes while δ3-carene, the second major monoterpene in Chok Anan mangoes has been described as sweet, floral and mango leaf-like (Laohakunjit et al. 2005). These terpene notes are further enhanced by the presence of compounds such as 3-hexanol, 2-hexanol, γ- and δ-lactones, and furan compounds to give the overall flavour perception (Vásquez-Caicedo et al. 2002). In addition, esters give rise to the overall fruity character while the overall floral top notes of the mango flavour is derived from the alcoholic compounds such as linalool, 2-phenylethyl alcohol, nerol and citronellol (Chauhan et al. 2010). 1.2 Mango wine Despite its popularity, mango juice/pulp is not an optimum substrate for fruit wine fermentation due to its relatively low sugar content, and organic acid and amino acid composition. Wine grapes, the optimum substrate for wine fermentation, typically have a sugar content of approximately 150 to 250 g/100 mL [17 to 26 % (w/v)], with tartaric and malic acids as the main organic acids (Alexandre et al. 1994); glutamate, glutamine, alanine, arginine and proline make up 90% of all amino acids present in wine grapes. Furthermore, certain cultivars of mangoes also contain high levels of pectins which may lead to pectin haze, resulting in undesirable cloudiness in wine (Vásquez-Caicedo et al. 2002). To overcome these differences and limitations, various groups have conducted research to optimise the production of a mango wine with quality comparable to conventional grape wine After the first anonymous report in 1963, the technology for mango wine production was tested on the Hilacha variety in 1966. The study concluded that mango was an excellent raw material for production of good quality white, semi-sweet table wine (Czyhrinciwk 1966). 3 Following which, several varieties of mangoes were screened for their suitability in wine production with the conclusion that mango wine had rather high acceptability (Kulkarni, Singh and Chadha 1980; Onkarayya and Singh 1984; Onkarayya 1986). Another study utilised Saccharomyces cerevisiae (S. cerevisiae) and Schizosaccharomyces species isolated from local palm wine and concluded that Schizosaccharomyces yeasts were suitable for the production of sweet, table mango wine while S. cerevisiae yeasts were suitable for the production of dry mango wine with a higher ethanol level (Obisanya et al. 1987). However, detailed vinification techniques and the chemical composition of the wine produced were not reported in these studies. The gap in knowledge was addressed by Reddy and colleagues with extensive investigations on mango wine with special focus on several aspects (Reddy et al. 2009; Reddy and Reddy 2009; Sudheer et al. 2009; Reddy and Reddy, 2011; Varakumar et al. 2011). After an initial study that reported on the production and characterization of mango wine from popular Indian cultivars (Reddy and Reddy 2005, 2007), studies on the optimisation of fermentation conditions by employing response surface methodology (Sudheer et al. 2009), the effects of enzymatic maceration on synthesis of higher alcohols (Reddy and Reddy, 2009), the analysis of volatile composition of wine fermented from Indian mango cultivars (Reddy et al. 2010), the effects of fermentation conditions on yeast growth and volatile production (Reddy and Reddy, 2011), and the antioxidant potential of mango wine (Sudheer et al. 2012) were reported subsequently. These studies concluded that 25°C, pH 5, with 100 ppm SO2 and initial aeration were optimum fermentation conditions for the production of mango wine. In addition, Reddy et al. (2010) also concluded that the aromatic compounds of mango wine were comparable in concentration to those of grape wine. Pectinase enzyme treatment was also found to be effective in increasing the yield of juice, the production of ethanol, increasing the yield of higher alcohols (Reddy and Reddy 2009) and 4 improving mango wine (Li et al. 2013). Pulp maceration was also discovered to have a positive effect on the chemical profile of mango wine (Li et al. 2013). In addition, it was reported that the cultivar of mango used had an effect on the volatile composition of the mango wine produced, indicating that it might be possible to produce mango wines with ‘varietal’ character, just like grape wines (Li et al. 2012). The effects of mixed yeast culture fermentation (S. cerevisiae and Metschnikowia pulcherrima or Torulaspora delbrueckii; and S. cerevisiae and Williopsis saturnus) on the aroma and sensory properties of mango wine were also investigated (Li et al. 2012; Varakumar et al. 2012). Differences in volatiles produced were observed in wines produced by different yeast strains and/or mixed culture. The two non-Saccharomyces yeasts (M. pulcherrima and W. saturnus ) were unable to complete the fermentation, but the mixed cultures were able to produce similar levels of ethanol relative to the monoculture of S. cerevisiae, coupled with a higher glycerol content but lower volatile and total acidity (Varakumar et al. 2012). Malolactic fermentation with lactic acid bacteria is often conducted to convert the sharp and tart-tasting malic acid into the mellow and softer tasting lactic acid. Oenococus oeni (O. oeni) Lactobacillus and Pediococcus are also some of the common microorganisms used. The simultaneous inoculation of O. oeni and S. cerevisiae resulted in higher amounts of ethyl acetate and a bigger decrease in acetaldehyde content than with S. cerevisiae alone (Varakumar et al. 2013). In addition, mango wine inoculated with O. oeni gave higher sensory impacts than the control wine (monoculture of S. cerevisiae), and simultaneous inoculation showed better sensorial attributes in flavour, fruity aroma, and overall acceptability than sequential inoculation (Varakumar et al. 2012). 5 The analysis of volatile compounds in mango wine has been conducted by a few groups (Reddy and Reddy 2010; Li et al. 2011; Pino and Queris 2011). Reddy and colleagues reported on the concentration of major alcohols, esters, fatty acids and ketones of mango wines made from the Banginapalli and Alphonso varieties. Pino and Queris (2011) and Li et al. (2011) assessed the contribution of the identified volatile compounds to the aroma of mango wine on the basis of their odour activity values (OAV). Pino and Queris (2011) identified 40 esters, 15 alcohols, 12 terpenes, 8 acids, 6 aldehydes and ketones, 4 lactones, 2 phenols, 2 furans and 13 miscellaneous in the mango wine. Li et al (2012) identified 4 acids, 7 alcohols, 25 esters, 8 carbonyl compounds and 1 miscellaneous compound as being important to the flavour of mango wine. 1.3 Biochemistry of fermentation, wine flavour and quality Wine making is a complex process that involves interactions between the fermentative yeasts and the numerous compounds present in the must through a series of biochemical reactions. A wide range of compounds (fermentative products) are produced during fermentation, affecting appearance, aroma, flavour and mouth-feel and overall organoleptic properties of the wine (Plata et al. 2003; Francis and Newton 2005; Jones et al. 2008; SáenzNavajas et al. 2010). The two main biochemical reactions that affect wine quality are primary alcoholic and glyceropyruvic fermentation. The former converts sugars into ethanol anaerobically while the latter produces glycerol which affects the mouthfeel and texture of the wine (Nieuwoudt et al. 2002; Jones et al. 2008). Pyruvate is generated as an intermediate product in both pathways; the by-products of pyruvate metabolism are also precursors for other flavour compounds (Swiegers et al. 2005; Ardö 2006). The major volatile products are typically 6 ethanol, fusel alcohols, esters and other compounds which make up the fermentation bouquet that impacts the overall flavour profile of the wine. In addition, fermentative flavour can also be influenced by other compounds such as long-chain fatty acids, nitrogenous and sulphur containing compounds. These compounds are not directly fermented, but can diffuse into the yeast cells and undergo biochemical reactions producing numerous volatile by-products (Salmon et al. 1993); the production levels and metabolism of these compounds are variable and yeast strain specific (Pretorius et al. 1999). Wine quality is heavily influenced by the chemical composition of the wine at the point of consumption due to their interaction with the sense of taste and smell. Wine flavour is due to the interaction of non-volatile chemicals (which lead to taste sensations) and volatile compounds (which are responsible for the odour — one of the most important factors affecting wine quality). Some major non-volatiles include glycerol, sugars, organic acids and polyphenols which affect the mouthfeel, sweetness, sourness and astringency, respectively. The concentration of non-volatiles needs to be at least approximately 1% to have an effect on the wine taste (Rapp and Mandery 1986). The volatile composition is especially important as many consumers associate complexity in aroma with a higher quality wine. These volatile compounds may originate from the original fruit (primary aroma), or from oenological processes such as fermentation (secondary aroma) or aging process (tertiary aroma), and include alcohols, esters, aldehydes, ketones, lactones, fatty acids, benzene derivatives, terpenes, C13-norisoprenoids, thiols, etc. (Rapp and Mandery 1986). The primary aroma is dominated by the major volatiles found in the fruit such as terpene hydrocarbons and norisoprenoids (El Hadri, et al. 2010), while fermentative products such as alcohols and esters dominate the secondary aroma. During alcoholic fermentation, each glucose molecule is broken down into two pyruvate molecules that are subsequently 7 decarboxylated into acetaldehyde and carbon dioxide by pyruvate decarboxylase. This is followed by the reduction of acetaldehyde to ethanol by alcohol dehydrogenase. In addition, higher alcohols can be formed from amino acids through the Ehrlich pathway. After the initial transamination reaction, the resultant α-keto acids can be further converted into respective higher alcohols (Ardö 2006). Some higher alcohols can also be formed through sugar metabolic pathways since α-keto acids can be formed in the tricarboxylic acid cycle (TCA cycle) (Pietruszka et al. 2010). Furthermore, hydrolytic enzymes produced by yeasts also release a wide range of secondary metabolites initially present in the must as non-volatile flavour precursors, contributing to varietal character (Lambrechts and Pretorius 2000; Swiegers et al. 2005; Dubourdieu et al. 2006; Jolly et al. 2006; Li et al. 2013). Tertiary flavour development during the aging of wine is a slow and gradual process (Sumby et al. 2010). Acetate esters usually decrease in the first few years of aging due to acid-catalysed hydrolysis resulting in a loss of freshness and fruitiness, especially in white wines. On the other hand, straight-chain fatty acid ethyl esters remain relatively constant (Rapp and Mandery 1986) as hydrolysis of ethyl esters is slower than that of acetate esters considering the high concentration of the hydrolytic product ethanol which may cause inhibition of ethyl ester hydrolysis (Rapp and Mandery 1986). 1.4 Influence of fermentation conditions and yeast strains 1.4.1 Amelioration of must Conventionally, wine grapes are carefully cultivated to contain the optimum amounts of nutrients for the production of a balanced wine. However, many factors beyond the 8 control of the viticulturist and winemaker (for instance, weather and climate change) may affect the composition of the grape (Jackson and Lombard 1993), resulting in the need for additions of adjuncts to produce a balanced wine. Some of the common adjuncts include acids (for optimum pH), yeast nutrients (nitrogenous source) and sugars (This et al. 2006; Mendoza et al. 2009). Initial sugar content is one of the most important parameter monitored; insufficient sugars would result in a wine with insufficient ethanol content. However, high sugar concentration may have an effect on final wine quality due to the effects of hyperosmotic stress on the yeasts which includes rapid reduction in internal cell volume, efflux of water from the cell, lowering turgor pressure, reducing water availability and causing cell shrinkage (Hohmann 1997). Yeasts accumulate compatible solutes and osmoprotectants under hyperosmotic conditions (Thomas et al. 1994); and osmotolerant yeasts are able to retain synthesized glycerol as osmoregulator, some species even have active glycerol uptake pumps (van Zyl et al. 1990). However, osmosensitive species, such as S. cerevisiae, tend to leak glycerol significantly, except under hyperosmotic conditions when retention is improved (Bauer and Pretorius 2000). In addition, hyperosmotic conditions also impact volatile production due to the imbalance of redox potential (Jain et al. 2011; Styger et al. 2013). The most significant effects on wine quality are likely to be the consequence of excess acetic acid and acetaldhyde production (Swiegers et al. 2005). 1.4.2 Saccharomyces in wine production The Saccharomyces species, especially S. cerevisiae, is the dominant yeast utilised in commercial wine production today. It is highly adapted to fermenting grape must in monoculture and has the ability to modulate most of the major constituents of wine, including residual sugars, ethanol, polyols, acids and phenolics (Swiegers et al. 2005; Pretorius 2007). 9 Commercially, strain selection places an emphasis on the ability to produce wine with low residual sugar and high ethanol (Jackson and Lombard 1993). However, different strains exhibit different traits and have different nutrient requirements; hence, the suitability of each strain is dependent on both the desired wine style as well as the initial physicochemical properties of the must (Heard and Fleet 1986). S. cerevisiae typically produces fruity/estery wines that contain higher concentrations of ethyl esters of fatty acids with lower concentrations of higher alcohols, which could otherwise mask aroma intensity (van der Merwe and van Wyk, 1981); or wines with enhanced varietal character from the release and/or modification of native flavour compounds to yield varietal aroma compounds such as the fruity, long-chain, polyfunctional thiols 4-mercapto-4-methylpentan-2-one, 3- mercaptohexan-1-ol and 3- mercaptohexyl acetate (Dubourdieu et al. 2006; Swiegers et al. 2008). 1.4.3 Non-Saccharomyces species in wine production The aroma profiles of non-Saccharomyces fermented wines are distinctly different from S. cerevisiae wines. Some species such as Williopsis saturnus and Kloeckera apiculate produce significantly higher amounts of desirable acetate esters which may impart floral and fruity notes to the wine bouquet (Li et al. 2012); while other species such as Torulaspora delbrueckii produce low amounts of acetic acid (Ciani and Maccarelli 1999; Rojas, et al. 2001; Jolly, et al. 2006; Fernández, et al. 2000). In addition, some other high ester producing non-Saccharomyces species include the Hanseniaspora family, Pichia anomala, etc. These non-conventional yeast strains yield novel aroma profiles, many of which are perceived as positive, indicate potential application in wine production by providing aroma diversity (Dubourdieu et al., 2006). 10 Glycosidases can also potentially enhance the aroma and flavour properties of wine. However, most Saccharomyces species have low glycosidase activities (Heard and Fleet 1986; Li et al. 2013), therefore, non-Saccharomyces species with glycosidase activities have been reported to contribute positively to enzymatic reactions during the early stages of vinification. Some of these species include the Candida, Debaryomyces, Hanseniaspora, Kloeckera, Kluyveromyces, Metschnikowia, Pichia, Saccharomycodes, Schizosaccharomyces, and Zygosaccharomyces genera (Rosi, Vinella and Domizio 1994). The liberation of aromatic terpenols from their odourless precursors has also been linked to the enzymatic activities of non-Saccharomyces species (Lagace and Bisson 1990). In addition, some strains such as Kloeckera apiculata has the potential to reduce protein haze due to their significant protease content (Lagace and Bisson 1990). 1.4.4 Inoculation strategies in wine fermentation Due to its weak weak fermentative capability and low ethanol tolerance, complete fermentation with a non-Saccharomyces monoculture is not possible, therefore resulting in the limited use of non-Saccharomyces species in commercial wine fermentation. Consequently, multistarter cultures have been explored and utilised to harness the advantages of non-Saccharomyces yeast species. Two strategies have evolved to harvest the desirable aroma profile and to enable complete fermentation with non-Saccharomyces yeasts — cofermentation with a robust Saccharomyces strain; and sequential fermentation, in which the non-Saccharomyces yeast and Saccharomyces strain are inoculated successively, in order to complete fermentation. Wines produced by these methods have proven to have distinct and desirable traits (Soden et al. 2000; Holzapfel 2002;, Jolly et al. 2006; Li et al. 2012; Maturano et al. 2012) 11 A successful co-fermentation depends on the physiological properties of the individual yeasts – its compatibility with other yeasts and the effects on growth rate and biomass development. Suppression of one yeast by the other can result in its reduced metabolic activity and hence decreased impact on the wine characteristics. One strategy may be the coinoculation of a weakly fermentative yeast at high ratio to a strongly fermentative yeast to also achieve a greater impact of the former yeast and produce a more balanced wine during co-fermentation. Diffusion of various metabolites between yeasts with different ‘metabolic tuning’ can result in metabolite concentrations different from those that would be achieved by blending wines (Ciani and Maccarelli 1999; Fernández et al. 2000; Clemente-Jimenez et al. 2005; Jolly et al., 2005; Augustyn and Pretorius 2006; Lee et al. 2012a; Lee et al. 2012b; Li et al. 2012). A sequential fermentation of Pichia fermentans and S. cerevisiae conferred greater complexity to wine through the enhancement of desirable flavour compounds production and glycerol content (Clemente-Jimenez et al. 2005). In addition, the use of multistarter fermentations to reduce the negative sensorial characteristics and for biological acidification of wines has also been reported. Sequential fermentations can also be used to favour weak fermentative strains by delaying inoculation of S. cerevisiae so that the desirable traits conferred by the weak fermentative strain can be developed first, for instance, good acidity, low volatile acidity, intense fruity ester production, and in some wines, a desirable ‘wild yeast’ fermentation character (Moreno et al. 1991; Kapsopoulou et al. 2007; Bely et al. 2008). 12 1.5 Research aims and objectives The overall objective of this research project was to study the effects of sugar concentration and inoculation strategies on mango wine fermentation. Aim 1: Effect of sugar concentration on mango wine fermentation with S. cerevisiae MERIT.ferm — Chapter 3 High sugar concentration or hyperosmotic pressure creates redox imbalance and the efforts of the yeasts to combat this stress to ensure its survival have implications on wine quality. The effects of initial sugar concentration on yeast biomass accumulation, glycerol and volatile compounds production with special emphasis on acetaldehyde and acetic acid were investigated. While it has been hypothesised that high initial sugar concentration will impact wine quality negatively, the specific effects are unknown. Hence, the specific effects of high initial sugar concentration on volatile production and its relation to redox equilibrium are presented in Chapter 3. Aim 2: Effect of co-inoculation of S. cerevisiae and W. saturnus on mango wine fermentation — Chapter 4 As yeast strain is one of the most crucial factors determining the quality of wine, effects were investigated by analysing the differences in volatile composition and growth kinetics between the two different yeasts, S. cerevisiae MERIT.ferm and W. saturnus NCYC 500 and simultaneous mixed culture fermentation with MERIT.ferm and W. saturnus NCYC 500 in the ratio of 1:1000. The results are presented in Chapter 4. Aim 3: Effect of different sequential inoculation strategies on mango wine fermentation — Chapter 5 Due to the effects that different inoculation strategies have on the flavour and quality of wine, and the lack of studies investigating the effects of sequential inoculation strategies on the production of mango wine involving the use of S. cerevisiae MERIT.ferm and W. 13 saturnus NCYC 500, Chapter 5 presents information on the differences in yeast growth kinetics, volatile composition and sensory perception in three wines with different inoculation strategies. 14 Chapter 2 Materials and Methods 2.1 Mango fruits and preparation of mango juice Mango fruits (Mangifera indica L.) of the Chok Anan variety imported from Malaysia were purchased from a local fresh produce wholesale market. Whole, healthy looking fruits were selected and stored at room temperature until fully ripe and soft. After washing with tap water to remove dirt and being allowed to air dry naturally at room temperature, the skin and flesh were removed manually and separated. The resulting flesh was then juiced in a commercial juicer, Sona juice extractor (Cahaya Electronics, Singapore) with the resulting puree being centrifuged at 4°C, 41 415 x g (Beckman Centrifuge, Brea, CA, USA) for 15 min and the supernatant removed and stored at -50°C before use. The mango juice (initial pH 4.5 to 4.6 and 15 – 18 °Brix) was then adjusted to a pH of 3.5 with 50% (w/v) DL-malic acid solution before the addition of 100 ppm potassium metabisulfite (K2S2O5, Goodlife Homebrew center, Norfolk, England). The mixture was left to stand for 24 h at 25°C for sterilisation. This sterilisation process aimed to kill any wild yeast that may be present in the juice. Handling of all materials was conducted in a bio-fumehood to maintain sterility. The effectiveness of sterilisation by 100 ppm SO2 sterilisation was verified by by streak plating on potato dextrose agar medium plate (39 g/L, Oxoid, Basingstoke, Hampshire, England). 15 2.2 Yeast and culture media Williopsis saturnus var. mrakii NCYC 500 from the National Collection of Yeast Culture (Norwich, UK) and Saccharomyces cerevisiae MERIT.ferm from Chr.-Han. (Denmark) were received in the active freeze dried form used in this study. Active dried yeast was propagated in a sterilised nutrient broth consisting of (on a w/v basis), 2% glucose, 0.25% bacteriological peptone, 0.25% yeast extract and 0.25% malt extract in deionised water, pH 5.0. This solution was first autoclaved for 15 min at 121°C before inoculation. These yeast cultures were then sub-cultured on potato dextrose agar plates (0.4% w/v potato extract, 2% w/v dextrose and 1.5% w/v agar, pH 5.6 at 25°C) for 2 days at 25°C. An isolated single colony was suspended into 10 mL of above-mentioned nutrient broth; following which, yeast strains were maintained in the nutrient broth and incubated at 25°C for 48 – 72 h without aeration. Finally, 20% glycerol was added to the culture before being stored at –80°C until further use. All media and equipment were sterilised at 121°C for 15 min before use and purity checks on stock cultures were conducted prior to all fermentations. 2.3 Preparation of yeast starter culture Each pre-culture was prepared with sterilised mango juice inoculated with 10% (v/v) of with S. cerevisiae MERIT.ferm or W. saturnus NCYC 500. The pre-cultures were then incubated at 25°C for 48 to 72 h for the yeasts to reach a concentration of 107 CFU/mL. Assessment of yeast cell growth was conducted via the spread plating method on PDA plates. This method of pre culture fermentation was used for the single culture fermentations (Chapter 3). 16 2.4 Fermentation of mango juice with different initial sugar concentrations — Chapter 3 Three different sets of mango juice (200 mL of sanitised mango juice in sterilised conical flasks) were prepared for fermentation. Two sets were supplemented with glucose to attain medium and high sugar contents with readings of 23°Brix and 30°Brix, respectively. A set of control fermentation with no sugar supplementation (low sugar fermentation) was also conducted. Each conical flask was inoculated with 105 CFU/mL of S. cerevisiae MERIT.ferm then plugged with cotton wool and wrapped with aluminium foil. Static fermentation was carried out for 35 days at 20°C. The fermentations were conducted in triplicate. 2.5 Fermentation of mango juice with co-inoculated S. cerevisiae and W. saturnus — Chapter 4 Two different starter cultures were prepared as describe in section 2.3 – one S. cerevisiae and one W. saturnus NCYC 500. Triplicates of 200 mL of sanitised mango juice in sterilised conical flasks plugged with cotton wool and wrapped with aluminium foil were fermented for 21 days at 20°C. S. cerevisiae MERIT.ferm and W. saturnus NCYC 500 were each inoculated into a flask of sterilised mango juice (200 mL) at a concentration 105 CFU/mL for the single culture fermentations. For the mixed culture fermentation, S. cerevisiae MERIT.ferm and W. saturnus NCYC 500 from the two single cultures were inoculated in a ratio of 1:1000 into the mango juice simultaneously. 17 2.6 Fermentation of mango juice with different sequential inoculation strategies — Chapter 5 Briefly, three different sequential inoculation strategies were studied. Triplicates of 200 mL of sanitised mango juice in sterilised conical flasks plugged with cotton wool and wrapped with aluminium foil were fermented for 21 days at 20°C. In the simultaneous mixed culture fermentation (MCF), 105 CFU/mL of W. saturnus NCYC 500 and 102 CFU/mL of S. cerevisiae MERIT.ferm, in a ratio of 1000:1 were simultaneously added. For the positive sequential fermentation (PSF), 105 CFU/mL W. saturnus NCYC 500 was added and fermentation was carried out for 14 days before the yeast cells were deactivated by being subjected to ultrasonication. S. cerevisiae MERIT.ferm was then added at 102 CFU/mL and the fermentation continued for another 7 days. For the negative sequential fermentation (NSF), 102 CFU/mL of S. cerevisiae MERIT.ferm was inoculated and fermentation was carried out for 7 days before the yeast cells were deactivated by ultrasonication. 105 CFU/mL of W. saturnus NCYC 500 was then added and fermentation continued for further 14 days. S. cerevisiae MERIT.ferm fermentation was halted at Day 7 because it had been shown that S. cerevisiae was able to complete the fermentation within that timeframe. Ultrasonication was conducted with the probe of the ultrasonicator (Hielscher – Ultrasound Technology, UIP 1000, 1000W).being sterilised with 70% ethanol solution prior to each run. Each conical flask was partially immersed in ice water to prevent the sample from overheating and affecting the flavour profile. The sample was then subjected to 15 min of treatment at 20 kHz. Plating on potato dextrose agar (PDA) was done to check for sterility of mango juice after ultrasonication treatment by drawing samples and doing streak plating. 18 2.7 Analytical methods 2.7.1 pH, oBrix and yeast enumeration Sampling was done at regular intervals throughout the fermentation process. Aliquots of approximately 10 mL were drawn under aseptic conditions after swirling to obtain a homogenous sample. The pH and total soluble solids were measured using a refractomer (Atago, Tokyo, Japan) and pH meter (Metrohm, Herisau, Switzerland), respectively. Cell counts for yeasts were carried out via the spread plating method on potato dextrose agar (PDA). Suitable dilutions of up to 108 were carried out with 1% peptone water to obtain suitable cell counts. The plates were then incubated at 25°C for 48 h before yeast colonies were counted. Lysine agar is unable to support the growth of Saccharomyces yeast (Erten and Tanguler 2010) and hence was used to differentiate Saccharomyces and nonSaccharomyces yeast growth pattern for the mixed culture experiments. Other than the selective media used to differentiate between the MERIT.ferm yeast and NCYC 500 yeast, the appearance of the colonies was used as a means of identification. S. cerevisiae yeasts appeared as small, off-white colonies with a glossy surface while NCYC 500 colonies had a dull, white appearance with a slightly bigger size. All yeast enumeration analyses were done in duplicate. 2.7.2 Analysis of sugars and organic acids The instrumental analysis and quantification of sugars and organic acids were conducted using a Shimadzu modular chromatographic system (LC solution software version 1.25) equipped with LC-20AD XR pumps and coupled to a SPD-M20A photodiode array detector, a low temperature evaporative light, scattering detector (ELSD-LT), a SIL-20AC XR autoinjector. 19 Prior to HPLC analysis, samples were centrifuged at 4 248 g at 4°C for 25 min (Sigma 3-18K centrifuge, Osterode am Harz, Germany), filtered with a 0.20-μm RC membrane (Sartorius, Gottingen, Germany) and stored at –50°C before analysis. Analysis was conducted in triplicate. Compounds were identified by comparing retention time, spectrum and concentration with external reference standards. Analysis of organic acids was conducted with a Supelcogel C-610 H column (300 × 7.8 mm, Supelco, Bellefonte, PA, USA). The mobile phase was 0.1% (v/v) sulphuric acid at a flow rate of 0.4 mL/min at 40°C and detection was done by photodiode array at 210 nm wavelength. Sugars were analysed by using the ELSD-LT (gain: 5; 40°C; 350 kPa) coupled with a Zorbax carbohydrate column (150 x 4.6 mm, Agilent, Santa Clara, CA, USA) using a mixture of acetonitrile and water (80:20 v/v) as the mobile phase with a flow rate of 1.4 mL/min at 40°C. 2.7.3 Analysis of volatile compounds Analysis of volatiles in both mango juice and wine was carried out using optimised headspace-solid phase microextraction combined with gas chromatography-flame ionization detector/mass spectrometry (HS-SPME GCMS/FID). Although traditionally used as a qualitative and semi-quantitative method (Sánchez-Palomo et al. 2005; Trinh et al. 2011), it has been demonstrated that HS-SPME GCMS/FID can be applied quantitatively if extraction and analytical conditions were optimised and consistently employed (Baptista et al. 2001; Lee et al. 2010). The analytical method was optimised by varying desorption temperature and time, flow rate, temperature profile gradient and evaluated based on the peak areas obtained for the compounds of interest. The nature of the matrix, the amount of sample, 20 desorbing conditions, fiber coating, extraction temperature and time can have an effect on the analytical results. A SPME fused silica fiber coated with 85 μm carboxen/polydimethylsiloxane (CAR/PDMS) (Supelco, Sigma-Aldrich, Barcelona, Spain) was used for extraction. A 5-mL sample was placed in a 20-mL glass vial tightly capped with a PTFE/silicone septum and extracted by HS-SPME at 60°C for 40 min with 250 rpm agitation; after which, the fibre was desorbed at 250°C for 3 min and injected into Agilent 7890A GC (Santa Clara, CA, USA), coupled to FID and Agilent 5975C triple-axis MS. Chromatographic separation was achieved via a capillary column (Agilent DB-FFAP) of 60 m × 0.25 mm I.D. coated with 0.25 µm film thickness of polyethylene glycol modified with nitroterephthalic acid. Helium was used as the carrier gas with a liner velocity of 1.2 mL/min, transfer line temperature 280°C. Mass detector conditions were set at 70eV electron impact (EI) mode, 230°C source temperature. The mass scanning parameters were: 3 min → 22 min: m/z 25– 280 (5.36 scan/s); 22 min → 71 min: m/z 25–550 (2.78 scan/s) under full-scan acquisition mode. Volatiles were identified by matching their mass spectra with the Wiley mass spectrum library and confirmed with the linear retention index (LRI) values, which were determined on the FFAP column against a series of alkanes (C5-C25) separated under identical operating conditions. The linear retention index (LRI) was used to identify the compound, and the calculation for the LRI is given as : LRI=100×[(ti-tz)/(tz+1-tz)+z], where ti = retention time of compound, tz = retention time of preceding n-alkane, tz +1 = retention time of subsequent n-alkane 21 Several volatiles were selected to be quantified using external standards (supplied by Firmenich Asia Pte Ltd, Tuas, Singapore) dissolved in 10% v/v micro-filtered (0.45 μm) mango juice diluted with water, except for ethanol that was dissolved in 100% micro-filtered mango juice. The analyses were carried out in triplicate. Following the analyses, a standard curve was constructed for the linear range of each compound. The results shown represent the means of three independent fermentations. Concentrations of volatile compounds were determined by using the linear regression equations of the corresponding standards. Odour activity values (OAVs) of quantified volatiles were calculated according to their known thresholds from the literature (Bartowsky and Pretorius, 2009; Ferreira, Lopez, and Cacho, 2000). The formula for the calculation of OAV used was as follows: OAV = 2.8 . Sensory analysis Sensory analysis was conducted for the mango wines produced by the simultaneous mixed-culture and sequential fermentations by a panel of six experienced flavourists (females and males) from Firmenich Asia Pte Ltd (Singapore) using quantitative descriptive analysis (QDA) methodology. A constant volume of wine was presented in wine-testing glasses and was arbitrarily coded. The samples were sniffed and the aroma intensity of each sensory descriptor was rated on a 5-point hedonic scale, where 0 indicated that the descriptor was not perceivable and 5 indicated that the descriptor had extremely high intensity. The eight mutually agreed sensory descriptors to describe the wine aroma were acidic, alcoholic, buttery, cocoa, fruity, fusel, sweet and yeasty notes. 22 2.9 Statistical analysis The ANOVA test using Microsoft Excel (ver 2007) was used to analyse the data obtained. Results were considered statistically significant if the value of P was less than 0.05. Each set of experiment was conducted in triplicate; hence, mean values and standard deviations were calculated from the triplicate. 23 Chapter 3 Effects of Sugar Concentration on Volatile Production by Saccharomyces cerevisiae MERIT.ferm in Mango Juice Fermentation 3.1 Introduction Oenological conditions could be stressful for most yeast strains, with many of these stresses encountered simultaneously or rapidly one after another. For instance, the high sugar content (hyperosmotic stress), nutrient deficiency (lack of nitrogen sources, oxygen, vitamins and other minerals), low pH, extreme temperatures all present as stressful conditions for the propagation and growth of most yeast strains during the onset of fermentation while ethanol toxicity occurs later in the fermentation process. Although commercial wine yeasts are selected for their inherent ability to cope with such environmental stresses, yeast metabolism is still significantly affected by fermentation conditions. High gravity fermentation or fermentation with high sugar content has been shown to have an effect on volatile production. The level of sugar substrate directly affects yeast metabolism which regulates the biochemical assimilation (energy consumption) and dissimilation (energy generation) of nutrients by yeast cells (Walker 1998); this pathway is intrinsically linked to the production of both volatile compounds and non-volatile compounds. Most studies reported increases in acetaldehyde and acetic acid and glycerol production in high gravity fermentation by yeasts (Bely et al. 2003; Chaney et al. 2006). The plasma membrane and actin skeletion are damaged due to the sudden loss in turgor pressure in response to hyperosmotic stress, potentially leading to the cessation of growth (Arrizon and Gschaedler 2002). To compensate for the efflux of water, water from the vacuole is released to provide a buffer for a short adjustment period (Bauer and Pretorius 2000). Concurrently, glycerol production is stimulated and the export channel closes; 24 accumulation of glycerol occurs until the influx of water restores the critical cell size for cell growth to occur (Chaney et al. 2006). This chapter investigated the response of S. cerevisiae MERIT.ferm to hyperosmotic pressure in mango juice fermentation due to high sugar content with a focus on yeast growth, sugar consumption, glycerol production and key volatile production in low sugar fermentation (unfortified mango juice with initial TSS of 16.6°B), medium sugar concentration (fortified mango juice with initial TSS of 23°B) and high sugar fermentaiton (fortified mango juice with initial TSS of 30°B). The sugar concentrations were selected for this study based on the typical total soluble solid (TSS) in °Brix for wine production. For still table wines, the recommended °Brix typically ranged from 20°B to 23°B (medium sugar fermentation) while German regulations for the production of ice wine (or Eiswein) had to be between 26 to 30°B (high sugar fermentation) (Boulton, et al. 1996). 3. 2 Results and discussion 3.2.1 Mango juice volatile composition In order to ascertain the effects that fermentation has on the volatile compounds, it is necessary to gain knowledge of the original volatile profile of the fresh Chok Anan mango juice. From the literature research conducted, there is only one detailed analysis of the flavour profile of Chok Anan mango (Li et al. 2011). Other information sources mentioned major character impact volatile compounds but did not provide a detailed list. Most studies focused on sensory evaluation (Vásquez-Caicedo et al., 2002) but did not report on the full list of volatiles and only reported on the key compounds believed to have an effect on consumers’ acceptability (Laohakunjit et al. 2005). 25 In this study, more than a hundred volatiles were detected in the fresh mango juice; a total of 59 major volatile compounds deemed to have an impact on the flavour profile were identified in the fresh Chok Anan mango juice (Table 3.1). These major volatile compounds identified (by relative peak area or RPA) were mainly monoterpenes (12), sesquiterpenes (6), alcohols (11) and esters (12). Monoterpenes made up the largest group of volatile compounds and constitute approximately 58% of the volatiles by RPA. The major volatile compound by RPA was α-terpinolene, followed by p-cymenene, pcymene and δ-3-carene. α-Terpinolene has been described as sweet, floral with pine-like aroma notes, while δ-3-carene has been described as sweet, floral and mango leaf-like. This finding differs slightly from data reported by Laohakunjit et al. (2005) where δ-3-carene as the second most abundant monoterpene. Some of the other terpene hydrocarbons identified in the same study were δ-thujene, α-pinene, mycrene, and α-phellandrene. Most of these volatile compounds were also identified in this current study. Terpenic notes were thought to be enhanced by the presence of other volatile compounds such as 3-hexanol, 2-hexanol, γ and δ-lactones, and furan compounds to give the overall flavour perception (Chauhan et al. 2010). However, some of these compounds such as 2-hexanol, 3-hexanol and δ-lactone were not found in this current study. This discrepancy could be due to the differences between the mangoes used. It has been shown previously that different climatic conditions and maturity of the fruit when harvested could lead to differences in the organoleptic quality (Lalel et al. 2003). Many of the volatiles that are regarded as important to the aroma profile of the mango flavour were identified in this study. Some of these include butyl butanoate, hexyl formate, 3-hexenyl acetate, cis-rose oxide, cis-3-hexenol, β-damascenone and trans-2-hexanal (Pino et al. 2005). The esters give rise to the overall fruity character while the overall floral top note 26 of the mango flavour is derived from the alcoholic compounds such as linalool, 2phenylethanol, nerol and citronellol (Chauhan et al. 2010). 27 Table 3.1 Volatile compounds in fresh Chok Anan mango juice analysed using HS-SPME-GC-MS/FID Terpenes CAS No1 007785-70-8 013466-78-9 002867-05-2 018172-67-3 000099-86-5 095327-98-3 000555-10-2 027400-71-1 000099-85-4 000099-87-6 000535-77-3 000586-62-9 000673-84-7 001195-32-0 001879-84-6 000087-44-5 000473-13-2 004630-07-3 017066-67-0 000515-17-3 LRI2 1088 1206 1219 1226 1235 1254 1265 1290 1305 1339 1343 1352 1450 1529 1628 1695 1696 1826 1829 Compound α-Pinene δ-3-Carene α-Thujene β-Pinene α-Terpinene Limonene β- Phellandrene β-Ocimene γ-Terpinene p-Cymene m-Cymene α-Terpinolene allo-Ocimene p-Cymenene β-Farnesene trans-Caryophyllene α-Selinene Valencene β- Selinene γ-Selinene 000064-17-5 000928-95-0 000111-27-3 000928-96-1 000104-76-7 1032 1445 1448 1476 1527 Ethanol trans-2-Hexen-1-ol 1-Hexanol cis-3-Hexanol 2-Ethyl-1-hexanol Subtotal Alcohol Peak Area (×106) 3.78 ± 0.05 52.8 ± 1.02 11.98 ± 0.09 1.14 ± 0.08 35.2 ± 0.61 35.6 ± 0.55 3.04 ± 0.28 1.1 ± 0.25 19.13 ± 0.41 12.7 ± 2.36 73.97 ± 5.06 348.97 ± 5.67 0.73 ± 0.47 60.7 ± 2.1 0.34 ± 0.03 0.39 ± 0.02 0.35 ± 0.03 0.19 ± 0.02 1.12 ± 0.05 0.56 ± 0.06 663.77 7.42± 0.84± 2.73± 65.45± 1.66± 28 0.59 0.06 0.3 1.91 0.08 RPA (%) 0.46 6.49 1.47 0.14 4.33 4.38 0.37 0.13 2.35 1.56 9.09 42.91 0.09 7.46 0.04 0.05 0.04 0.02 0.14 0.07 81.61 0.91 0.1 0.34 8.05 0.2 Odour descriptors3 Fresh, sweet, pine Sweet citrus Woody, green, herb-like Sharp, terpenic, conifer Sharp, terpenic, lemon Citrius, terpenic, orange note Mint terpentine Citrus, green, lime Fatty, terpenic, lime Citrus, terpenic, woody Citrus, terpenic, woody Citrus, lime, pine Floral, nutty, peppery Citrus, pine-like Woody, citrus, sweet Woody, clove note Amber Orange, citrus, woody Herbal Woody Alcoholic Fruity, green, leafy Alcoholic Green, leafy Oily, rose, sweet Table 3.1 (Continued) 000470-08-6 000464-43-7 000106-22-9 000078-70-6 000060-12-8 000128-37-0 Subtotal 1794 1808 1867 1999 2035 2090 β-Fenchol Endo-borneol Citronellol Linalool 2-Phenylethanol p−Cresol 000141-78-6 000105-54-4 000123-92-2 000109-21-7 003681-71-8 002497-18-9 000629-33-4 033467-74-2 016491-36-4 065405-80-3 000103-45-7 000110-38-3 1009 1095 1112 1284 1396 1410 1440 1466 1546 1700 1862 1948 Ethyl acetate Ethyl butanoate Isoamyl acetate Butyl butanoate 3-Hexenyl acetate trans-2-Hexenyl acetate Hexyl formate cis-3-Hexenyl propionate cis-3-Hexenyl isobutyrate cis-3-Hexenyl trans-3-butenoate 2-Phenylethyl acetate Ethyl dodecanote 000064-19-7 000067-43-6 000124-07-2 1549 1728 2171 000066-25-1 006728-26-3 1152 1310 Ester 0.11± 0.21± 0.4± 0.2± 0.24± 0.25± 79.51 0.01 0.08 0.02 0 0.02 0.03 0.01 0.03 0.05 0.02 0.03 0.03 9.78 Camphor-like, woody Camphor-like, woody Floral, rose, sweet, green, citrus Fresh, floral, herbal, rosewood Rose, honey, floral Cresol, medicinal, leather 9.67± 0.76± 11.7± 0.03± 25.07± 3.82± 0.82± 0.11± 0.19± 0.07± 1.19± 0.62± 54.05 0.85 0.07 1.04 0 4.3 0.3 0.02 0.01 0.01 0 0.09 0.05 1.19 0.09 1.44 0 3.08 0.47 0.1 0.01 0.02 0.01 0.15 0.08 6.65 Ethereal, fruity, sweet Sweet, fruity Sweet fruity, banana-like Fruity, pineapple, sweet Sharp, fruity-green, sweet Fruity, green, leafy Green, ethereal, fruity Fresh, fruity, green Apple, fruity, green Green, sweet, fruity Sweet, honey, floral, rosy Sweet, wine, brandy Acetic acid Butanoic acid Octanoic acid 0.23± 0.02 0.72± 0.07 0.6± 0.04 1.55 0.03 0.09 0.07 0.19 Vinegar-like Cheesy, rancid butter Acidic, fatty, soapy Hexanal trans-2-Hexenal 0.17± 0.01 4.59± 0.28 0.02 0.56 Fatty, green, grassy Green, leafy, fruity Subtotal Acid Subtotal Aldehyde 29 Table 3.1 (Continued) 000100-52-7 000620-23-5 000104-87-0 Subtotal 1633 1731 1771 Benzaldehyde 3-Methyl-benzaldehyde p-Toloualdehyde 0.4± 0.03 0.96± 0.08 2.76± 0.23 8.88 0.05 0.12 0.34 1.09 Bitter almond Cherry-like Sweet aromatic, bitter almond 000096-48-0 023696-85-7 000104-50-7 1758 1938 2051 5-Methyl dihydro-2(3H)-furanone β-Damascenone γ—octalactone 0.83± 0.08 2.6± 0.23 0.53± 0.05 3.95 0.1 0.32 0.06 0.49 Herbaceous, waxy, creamy note Sweet, floral, fruity, coconut Coconut, lactonic Furan Subtotal 003208-16-0 1044 2-Ethylfuran 2.52± 0.09 2.52 0.31 0.31 Ethereal run, cocoa note Ether 068780-91-6 016409-43-1 001786-08-9 016409-43-1 1537 1434 1563 1434 trans-Linalool oxide cis-Linalool oxide Nerol oxide cis-Rose oxide 0.58± 0.58± 0.31± 0.17± 0.15 0.15 0.08 0.04 Sweet, lemon Earthy, floral, sweet, woody Floral, orange blossom, green, sweet Rose, geranium Ketone Subtotal Total 813.35 1 0.03 0.04 0.01 0.01 100 CAS numbers were obtained from Wiley database library LRI of all the tables was determined on the DB-FFAP column, relative to C5-C40 alkanes. 3 Descriptors were retrieved from http://www.thegoodscentscompany.com 2 30 3.2.2 Changes in pH and organic acids There were no significant changes in pH throughout fermentation with pH values for all three treatments maintained at about 3.50 to 3.52. This is likely due to the insignificant changes in most organic acids before and after fermentation (Table 3.2). Fermentation had no significant effects on citric, succinic and tartaric acids. Citric acid remained at approximately 0.25 g/100 mL, succinic acid stayed at around 0.08 g/100 mL and tartaric acid varied in the range of 0.12 to 0.15 g/100 mL before and after fermentation. Malic acid decreased by approximately 50% after fermentation from 0.79 to 0.9 g/100 mL before fermentation to 0.33 to 0.41 g/100 mL after fermentation. The relatively high malic acid content before fermentation was due to the addition of D/L-malic acid to a attain a pH of 3.5 for the sanistised mango juice before fermentation. The decrease in malic acid after fermentation was likely due to the passive diffusion of D-malic acid into the yeast cells as reported previously (Coloretti et al. 2002). It was unlikely that malolactic fermentation occurred due to the absence of lactic acid in the mango wine and the absence of lactic acid bacteria in the microbial analyses. Furthermore, the addition of 100 ppm of potassium metabisulfite would have inhibited the growth of wild yeasts and bacteria. 31 Table 3.2 Physicochemical properties, organic acid and sugar concentrations of mango wine before and after fermentation 16.6°B pH o Brix (°B) Day 0 23°B 30°B 3.51± 0.01 a 3.51± 0.01 a 3.52± 0.01 a 16.6 ± 0.01 a 23.1± 0.02 a 30.1± 0.04 a Plate count 5.65±0.58 a (x105 CFU/mL) Sugar (g/100 mL) Fructose 4.96±0.08 a Glucose 0.63±0.02 a Sucrose 12.25±0.31 a 5.85±0.58 a 6.35±0.58a 5.58±0.05 b 5.96±0.09 c 7.46±0.03 b 12.39±0.06 c 13.86±0.12 b 14.19±0.36 c 16.6°B Day 35 23°B 30°B 3.51± 0.01 a 3.50± 0.01 a 3.50± 0.01 a 5.6 ± 0.06 b 7.2 ± 0.01 c 10.6 ± 0.17 d 8567±750 c 923±60 d N.D.* 1.11±0.02 d N.D.* 0.38±0.05 d d 0.013±0.00 2.53±0.12 e 935±58 d 2.46±0.15 e 0.22±0.06 e 3.84±0.36 f Organic acid (g/100 mL) Citric acid 0.26±0.04a 0.26±0.03a 0.24±0.04a 0.23±0.03 a 0.24±0.03 a 0.25±0.03 a a b b Malic acid 0.79±0.05 0.90±0.09 0.88±0.09 0.36±0.05 c 0.33±0.02 c 0.41±0.05 d a a a Succinic acid 0.083±0.012 0.079±0.012 0.081±0.012 0.086±0.011 a 0.079±0.012 a 0.081±0.012 a Tartaric acid 0.12±0.03 a 0.14±0.03 a 0.14±0.03 a 0.14±0.02 a 0.11±0.02 a 0.14±0.02 a abcd ANOVA (n=4) at 95% confidence level with same letters in the same row indicating no significant difference * N.D. Not Detected 3.2.3 Yeast growth, total soluble solids and sugar concentration The total soluble solids (TSS) declined throughout fermentation with similar trends. The oBrix value decreased from 16.6°B to 5.6°B for the low sugar fermentation, from 23.1°B to 7.2°B for the medium sugar fermentation and 30.1°B to 10.6°B for the high sugar fermentation. The decrease in oBrix value was more rapid initially before the rate of decline decreased (Figure 3.1). The decrease in °Brix also correlated to the increase in yeast growth (Figure 3.1) and overall sugar concentrations (Figure 3.2). Different trends for yeast biomass were observed for the three treatments (Figure 3.1). From Day 0 to around Day 10, the increases in yeast biomass and maximum cell populations were inversely correlated with initial sugar concentrations. Similar trends were observed for the low and medium sugar fermentation where viable cell population peaked between Days 9 and 11 before decreasing slowly until Day 16 when viable cell population decreased sharply. 32 However, the maximum cell population for the medium sugar fermentation was about 0.5 log lower than that in the low sugar fermentation (16.6°B). For the high sugar fermentation (30°B), the increase in yeast biomass was slower than the other two fermentations. The maximum viable cell population for the high sugar fermentation peaked between Days 9 to 11 but only attained a level of 7.7×107 CFU/mL, lower than that in the other two fermentations. This could be attributed to the osmotic stress from the high sugar concentration (Hohmann et al. 2003; Devantier, Pedersen and Olsson 2005). However, surprisingly, there was no significant decline in cell numbers after Day 11. Viable cell population was maintained at about 7×107 CFU/mL until the fermentation was terminated at Day 35. This could be due to the presence of sugars and other nutrients, while nutrients were depleted in the other two treatments resulting in the starvation of the yeast cells which ultimately led to the decline in the number of viable cells. Furthermore, S. cerevisiae MERIT.ferm is known to have a high ethanol tolerance according to the manufacturer’s information sheet, hence, in the presence of sufficient nutrients, ethanol toxicity is not expected to have a detrimental effect on biomass increase with increasing sugar concentrations. 33 Log CFU/mL 10 9 8 7 6 5 4 3 2 1 0 10 20 30 Time (Day) TSS (°B) 30 25 20 15 10 5 0 0 10 20 30 Time (Days) Figure 3.1 Changes in yeast cell count and TSS (oBrix) for low ( ), medium ( ) and high ( ) sugar fermentation Fermentation was effectively completed at Day 9 and Day 17 for the low and medium sugar fermentation respectively, when the oBrix value remained constant (Figure 3.1) and the concentration of sugars no longer declined (Figure 3.2). Fermentation proceeded to completion for both low and medium sugar fermentation successfully. On the other hand, fa sluggish fermentation was obseved for high sugar fermentation. The rate of decrease reduced dramatically after the first 11 days of fermentation. This is consistent with many studies that have indicated the likelihood of sluggish/stuck fermentations occurring in high density musts (juices). This has implications for the wine quality as the high residual sugar content could 34 suppport the growth of undesirable spoilage microorganisms that may result in detrimental wine quality or result in a sweet wine. From Day 0 to Day 3, an increase in fructose was observed in the low and high sugar fermentations, with an increase in glucose also observed in the low sugar fermentation. After which, all sugars decreased throughout the fermentation. On the other hand, there was no increase in any of the sugars quantified for the medium sugar fermentation. The rate of decline in sugars was generally inversely proportional to initial sugar concentration. It was also observed that glucose was utilised more rapidly than fructose and sucrose. This is consistent with the available data claiming that S. cerevisiae strains are mostly glucophillic in nature (Berthels et al. 2004). Available literature data also suggests that fructose utilisation was inhibited more than glucose in the presence of high ethanol content (Berthels et al. 2004). This was also observed in this study where there was lesser glucose and more fructose present after the high sugar fermentation than the medium sugar fermentation (Table 3.2). The increase in fructose and glucose, together with the decrease in sucrose, is likely due to the enzymatic activity of S. cerevisiae which produces invertase that converts sucrose into glucose and fructose (Berthels et al. 2004). It might be possible that similar invertase activity occurred in the medium sugar fermentation but the higher sugar concentration (relative to low sugar fermentation) could have led to invertase suppression (Myers et al. 1997), leading to lesser amounts of fructose and glucose production. Although this appears to contradict the data for high sugar fermentation, it can be easily explained that the high sugar concentration affected the normal metabolic functioning of the yeast cells, leading to less efficient sugar consumption (Arrizon and Gschaedler 2002). 35 Sugar content (g/100 mL) Low sugar fermentation 12 8 4 0 0 10 20 Time (Days) 30 Sugar content (g/100 mL) Medium sugar fermentation 20 15 10 5 0 0 10 20 Time (Days) 30 Sugar content (g/100 mL) High sugar fermentation 25 20 15 10 5 0 0 10 20 Time (Days) 30 Figure 3.2 Fructose ( ), glucose ( ) and sucrose ( ) consumption kinetics of S.cerevisiae during fermentation 3.2.4 Glycerol production The amount of glycerol produced at the end of each fermentation increased with increasing sugar content (Figure 3.3). The final glycerol concentrations in the low, medium and high sugar fermentations were 1.34 g/100 mL, 1.45 g/100 mL and 1.59 g/100 mL 36 respectively. Glycerol can affect perceived wine quality positively by improving the mouthfeel of the wine and may have minor contributions to the sweetness of the wine, this effect is concentration dependent. The usual amount of glycerol in wines ranges from 4 to 9 g/L (or 0.4 to 0.9 g/100 mL) (Remize et al. 1999). Since yeast strain is the strongest influencing factor on the amount of glycerol produced, it is likely that MERIT.ferm is a high glycerol producing strain and has a potential to improve wine quality. Other factors that influence glycerol production include sulfite concentration, pH, fermentation temperature, aeration, inoculation level, sugar concentration and nitrogen content (Radle and Schülz 1982). Glycerol is an osmoprotectant naturally synthesized by yeasts in response to the high concentration of osmotically active substances (especially glucose and fructose) present in the must or juice. The hypertonic conditions experienced by the yeast cells lead to an efflux of water from the cell, diminished turgor pressure and reduced water availabilty (Scanes et al. 1998). The induction of glycerol biosynthesis results in the accumulation of glycerol inside the cell for the equilibration of the osmotic pressure between the intracellular and extracellular environment o re-establish cell turgor pressure (Hohmann, 1997). 37 1.80 b 1.60 Glycerol (g/100mL) 1.40 a, b a 1.20 1.00 0.80 0.60 0.40 0.20 0.00 Low Medium High Figure 3.3 Concentrations of glycerol in mango wines for low (16.6oBrix), medium (23oBrix) and high sugar (30oBrix) fermentation ab ANOVA (n=4) at 95% confidence level with same letters in the same row indicating no significant difference 38 3.2.5 Effects of initial sugar concentration on volatile production The volatile profile of mango wine was significantly different from the fresh mango juice. A number of volatiles detected in the fresh mango juice were not detected in the final mango wine. These catabolised compounds include a number of esters, aldehydes, ketones, furans and terpene hydrocarbons. The complete list of catabolised compounds can be found in Table 3.3 below. The absence of some significant character impact odourants suggests that the mango wine produced had a different flavour profile from the mango juice. Table 3.3 Volatiles in mango juice catabolised during fermentation Class Alcohol Compounds trans-2-Hexenol, 2-ethylhexanol, para−cresol, βfenchol, endo-borneol Ester Hexyl formate, cis-3-hexenyl isobutyrate, cis-3hexenyl propionate, cis-3-hexenyl-3-butenoate, butyl butanoate trans-2-Hexenal, 3-methylbenzaldehyde, hexanal Aldehyde Ketone Furan Terpene hydrocarbon 5-Methyldihydro-2(3H)-furanone, γ-hexalactone, 4-methyl-2-heptanone, mesifurane 2-ethylfuran, 2,3-dihydro-2-methyl-benzofuan α-Terpinene, cis-ocimene, β-phellandrene, αthujene, α-pinene, allo-ocimene, γ-selinene, αselinene, trans-farnesene, valencene A total of 11 alcohols, 35 esters, 3 aldehydes, 3 ketones, 8 terpenes and 6 organic acids were detected in the final mango wines (Table 3.4). The FID peak area was used to compare volatile compounds that were not quantified, since FID peak area may semi-quantitatively represent the amount of different volatiles (Alves et al. 2010; Lee et al. 2010; Trinh et al. 2010). The relative peak area (RPA) is the area under each peak expressed as a percentage of the total peak area of all the compounds detected 39 Table 3.4 Volatile composition of mango wines with different initial sugar concentrations CAS1 LRI2 Compound Low Peak Area 000064-17-5 1028 Ethanol 000078-83-1 000123-51-3 000111-27-3 1172 1237 1448 Isobutyl alcohol Isoamyl alcohol Hexanol 000928-96-1 000505-10-2 000470-08-6 000106-24-1 1475 1706 1794 1861 cis-3-Hexenol Methionol β-Fenchyl alcohol Geraniol 000106-22-9 000060-12-8 1867 1964 Citronellol 2-Phenylethyl alcohol 000078-70-6 1999 Linalool Subtotal Alcohol Subtotal Ethyl ester 820.9± 11.55 RPA (%) a 820.9 10.63 ± 0.69 a 48.55 ± 2.75 a 0.65 ± 0.03 a 2.76 ± 0.78 ± 0.55 ± 0.23 ± 0.13 a 0.03 a 0.01 a 0.002 a 1.1 ± 3.32 a 54.82 ± 3.52 a 0.71 ± 0.01 a 120 000141-78-6 000105-54-4 000123-66-0 1009 1034 1297 Ethyl acetate Ethyl butyrate Ethyl hexanoate 64187-83-3 1358 Ethyl cis-3-hexenoate 000106-30-9 054653-25-7 000106-32-1 35194-38-8 000123-29-5 000110-38-3 067233-91-4 1369 1392 1453 1486 1624 1746 1795 Ethyl heptanoate Ethyl 5-hexenoate Ethyl octanoate Ethyl 7-octenoate Ethyl nonanoate Ethyl decanoate Ethyl 9-decenoate Medium Peak Area 67.89 1115.94± 81.93 RPA b RPA 87.43 1515.94± 81.93 91.24 Alcoholic 67.89 1115.94 87.43 1515.94 91.24 0.8792 5.03 ± 0.08 b 0.416 6.11 ± 0.12 c 0.3677 Citrus, orange, floral, rose 4.0154 46.79 ± 0.25 a 3.8699 51.52 ± 0.57 b 3.1009 Ethereal, winey 0.0538 0.31 ± 0.01 b 0.0243 0.34 ± 0.02 b 0.0205 Pungent, ethereal, fruity and alcoholic 0.2283 0.16 ± 0.01 b 0.0132 0.09 ± 0.01 c 0.0054 Green, grassy 0.0645 0.29 ± 0.01 b 0.024 0.2 ± 0.01 c 0.0125 Cooked potato-like 0.0455 0.04 ± 0.003 b 0.0033 0.036 ± 0.006 b 0.0024 Camphor, pine, woody 0.019 0.03 ± 0.002 b 0.0025 0.02 ± 0002 b 0.0012 Fusel, etherial, cognac, fruity 0.091 0.57 ± 0.03 b 0.0471 0.56 ± 0.01 b 0.0337 Floral, rosy, sweet, citrus 4.534 0.66 ± 0.04 b 0.0546 0.78 ± 0.02 b 0.0482 Sweet, floral, rose-like, honey 0.0587 0.21 ± 0.02 b 0.0174 0.2 ± 0.01 b 0.0126 Mild floral 9.99 53.5 4.47 0.1017 1.2439 0.2084 0.91 ± 0.02 b 0.0713 1.43 ± 0.06 b 0.112 7.46 ± 0.18 b 0.5845 0.13 ± 0.02 a 0.0108 0.15 ± 0.01 a 0.0118 0.0074 0.2713 1.8452 0.7229 0.148 10.9447 0.7303 0.11 ± 0.48 ± 1.59 ± 1.66 ± 0.15 ± 0.31 ± 0.74 ± 0.001 a 0.12 a 1.97 a 0.48 a 0.1 a 2.82 a 0.22 a Peak Area c 1.23 ± 0.09 a 15.04 ± 0.56 a 2.52 ± 0.17 a 0.09 ± 3.28 ± 22.31 ± 8.74 ± 1.79± 132.33 ± 8.83 ± Odour descriptors3 High 40 0.01 a 0.02 b 0.11 b 2.16 b 0.01b 0.02 b 0.04 b 0.0086 0.0376 0.1246 0.1301 0.0118 0.0243 0.058 59.36 3.59 0.15 ± 0.01 c 0.009 Ethereal, fruity, sweet 1.02 ± 0.08 c 0.0614 Sweet, fruity, lifting 3.23 ± 0.13 c 0.1944 Sweet, pineapple, fruity, waxy 0.12 ± 0.01 a 0.0072 Fruity, sweet, apple, ethereal 0.09 ± 0.01 a 0.0054 Sweet, waxy, soapy 0.44 ± 0.02 b 0.0265 Fruity, pineapple 0.93 ± 0.04 c 0.056 Waxy, sweet, musty, fruity 0.32 ± 0.01 c 0.0193 Tropical, fruity 0.09 ± 0.01c 0.0054 Waxy, soapy, cognac 0.23 ± 0.01 c 0.0138 Sweet, fruity, pineapple 0.61 ± 0.02 c 0.0367 Fruity, fatty Table 3.4 (Continued) 000106-33-2 000124-06-1 000628-97-7 054546-22-4 Subtotal 1887 2201 2373 2402 000109-60-4 000110-19-0 1002 1020 Propyl acetate Isobutyl acetate 0.75 ± 0.05 a 1.18 ± 0.09 a 0.062 0.0976 0.61 ± 0.04 b 0.0478 0.84 ± 0.06 b 0.0658 000123-86-4 1061 Butyl acetate 0.65 ± 0.04 a 0.0538 0.52 ± 0.04 b 0.0407 000628-63-7 000123-92-2 1097 1112 Active amyl acetate Isoamyl acetate 0.18 ± 0.02 a 0.48 ± 0.03 a 0.0149 0.0397 0.13 ± 0.01 b 0.0102 1.25 ± 0.06 b 0.0979 003681-71-8 2497-18-9 000142-92-7 1324 1367 1411 cis-3-Hexenyl acetate 2-Hexenyl acetate n-Hexyl acetate 0.07 ± 0.01 a 0.11 ± 0.01 a 0.33 ± 0.02 a 0.0058 0.0091 0.0273 0.29 ± 0.02 b 0.0227 0.08 ± 0.01b 0.0063 0.24 ± 0.02 b 0.0188 000150-84-5 16630-55-0 000141-12-8 000105-87-3 000103-45-7 1659 1676 1753 1790 1862 Citronellyl acetate Methionyl acetate Neryl acetate Geranyl acetate 2-Phenylethyl acetate 0.69 ± 0.21 ± 3.69 ± 2.56 ± 2.33 ± 0.07 a 0.01 a 0.09 a 0.12 a 0.15 a 0.0571 0.0174 0.3052 0.2117 0.1927 0.52 ± 0.17 ± 2.81 ± 2.07 ± 1.42 ± Acetate ester Ethyl dodecanoate Ethyl tertradecanoate Ethyl hexadecanoate Ethyl 9-hexadecenoate 1.57 a 0.01 a 0.2 a 0.08 a 218.83 Subtotal Higher ester 16.12 ± 0.51 ± 2.51 ± 3.52 ± 18.1 13.23 005461-06-3 002035-99-6 30673-60-0 030673-38-2 1642 1762 1803 1859 000103-52-6 1907 Isobutyl octanoate Isoamyl octanoate Propyl decanoate Isobutyl decanoate 2-Phenylethyl butanoate 1.3233 0.0422 0.2076 0.2911 1.0942 1.02 ± 0.15 ± 0.28 ± 13.25 ± 0.05 b 0.02 b 0.01 b 1.09 b 29.69 2.33 0.04 b 0.01 b 0.07 b 0.1 b 0.08 b 10.95 0.1844 0.1174 0.0604 0.0438 1.68 ± 2.05 ± 0.58 ± 0.45 ± 1.64± 0.12 a 0.1356 1.33± 0.1 b 41 0.0407 0.0133 0.2202 0.1622 0.1113 0.84 0.1 0.19 10.26 ± ± ± ± 0.04 c 0.0506 Sweet, waxy, fruity 0.01 c 0.006 Sweet, waxy, creamy 0.03 c 0.0114 Waxy, fruity, creamy 0.65 c 0.6175 - 18.62 1.12 0.44 ± 0.03 c 0.0265 Fusel, sweet, fruity 0.6 ± 0.04 c 0.0361 Sweet, fruity, banana, tropical 0.42 ± 0.01 c 0.0253 Ethereal, solvent, fruity, banana 0.1 ± 0.02 c 0.006 Ethereal, fruity, banana 0.91 ± 0.09 c 0.0548 Sweet, fruity, banana, solvent 0.26 ± 0.03 b 0.0156 Fresh, green, sweet, fruity 0.06 ± 0.01 b 0.0036 Sweet, leafy green 0.19 ± 0.01 c 0.0114 Fruity, green, apple, banana, sweet 0.39 ± 0.03 c 0.0235 Floral, green rose, fruity 0.13 ± 0.01 c 0.0078 Fatty, ester 2.23 ± 0.06 c 0.1342 Floral, rose, soapy, citrus 1.68 ± 0.08 c 0.1011 1.03 ± 0.04 c 0.062 Sweet, honey, floral, rosy 0.8579 8.44 0.28 a 0.05 a 0.03 a 0.02 a 2.23 ± 1.42 ± 0.73 ± 0.53 ± 0.0799 0.0118 0.0219 1.0381 0.21 b 0.06 b 0.02 b 0.02 b 0.508 0.1316 0.1606 0.0454 0.0353 1.27 1.45 0.46 0.38 0.16 c 0.26 c 0.02 c 0.02 c 0.0764 0.0873 0.0277 0.0229 0.1042 1.08± 0.08 c 0.065 ± ± ± ± Fruity, green, oily, floral Sweet, fruity, waxy, green Waxy, fruity, fatty, green Oily, sweet, brandy, apricot, cognac Fruity, floral, green, winey Table 3.4 (Continued) 002306-91-4 1973 Isoamyl decanoate 2.44 ± 0.06 a 0.2018 1.25 ± 0.17 b 0.0979 006309-51-9 2180 Isoamyl dodecanoate 0.36 ± 0.02 a 0.0298 3.98 ± 0.12 b 0.3118 Subtotal Aldehyde 9.35 000075-07-0 000100-52-7 000104-87-0 939 1637 1773 Acetaldehyde Benzaldehyde p-Tolualdehyde Subtotal Ketone 0.11 0.0091 0.0215 0.1406 000821-55-6 1398 2-Nonanone 0.23 ± 0.02 a 0.019 000513-86-0 023696-85-7 1401 1938 Acetoin β-Damascenone 0.01 ± 0.001 a 0.25 ± 0.02 a 0.001 0.0207 0.48 0.0397 11.32 0.8869 8.21 3.29 ± 0.17 b 0.2578 0.14 ± 0.01 b 0.011 0.22 ± 0.036 b 0.0172 3.65 0.286 0.16 ± 0.01 c 0.0096 Citrus, terpenic, orange note 0.19 ± 0.09b 0.0114 Sweet, citrus 1.01 ± 0.03 b 0.0608 Citrus, terpenic, woody 0.05 ± 0.01 c 0.003 Citrus, lime, pine 0.12 ± 0.01 b 0.0072 Sweet, woody, spice, clove 1.7 ± 0.1 c 0.1023 Woody 0.92 ± 0.03 b 0.0554 Mint, turpentine 000099-85-4 000099-87-6 000586-62-9 000087-44-5 1305 1339 1352 1695 γ-Terpinene p-Cymene α-Terpinolene β-Caryophyllene 0.1 ± 1.37 ± 0.31 ± 0.4 ± 0.02 a 0.31 a 0.02 a 0.01 a 0.0083 0.1133 0.0256 0.0331 0.15 ± 1.02 ± 0.09 ± 0.11 ± 006753-98-6 017066-67-0 1725 1829 α-Caryophyllene β-Selinene 3.83 ± 4.91 a 0.49 ± 0.01 a 0.3168 0.0405 1.83 ± 0.18 b 0.1434 0.98 ± 0.05 b 0.0768 4.4 0.11 b 0.03 b 0.001 b 0.02 b 0.0118 0.0799 0.0071 0.0086 0.3447 4.15 000064-19-7 1549 Acetic acid 6.54 ± 0.3 a 0.5409 7.53 ± 0.01 b 0.5899 000142-62-1 000124-07-2 1890 2170 Hexanoic acid Octanoic acid 0.73 ± 0.06 a 4.92 ± 0.15 a 0.0604 0.4069 0.69 ± 0.11 b 0.0541 0.71 ± 0.02 b 0.0556 42 0.2702 0.22 ± 0.04 b 0.0172 0.01 ± 0.003 a 0.5376 4.49 Fruity, sweet, waxy, soapy, cheese 0.01 ± 0.001 a 0.001 0.01 ± 0.001 a 0.001 Sweet, buttery, creamy 33.88 ± 3.16 b 2.6544 28.57 ± 0.88 c 1.7196 Woody, sweet, fruity, earthy, floral 33.93 2.6583 28.62 1.7226 Limonene 6.5 4.18 ± 0.25 c 0.2516 Green, cherry 0.12 ± 0.01 a 0.0072 Almond, fruity, nutty 0.19 ± 0.03 b 0.0114 Pungent, fresh, green 0.05 ± 0.003b 0.003 1254 0 0.4941 0.05 ± 0.006 b 0.0039 095327-98-3 Subtotal Organic acid 1.33 ± 0.0012a 0.11 ± 0.01 a 0.26 ± 0.0011a 1.7 Subtotal Terpene 0.7733 0.95 ± 0.06 c 0.0572 Waxy, fruity, sweet, cognac 2.62 ± 0.08 c 0.1577 Cognac, green, waxy 0.2498 8.26 ± 0.03 c 0.4972 Sharp, pungent, sour, vinegar 0.66 ± 0.11 b 0.0397 Sour, fatty, sweat, cheese 0.6 ± 0.01 b 0.0361 Fatty, waxy, rancid Table 3.4 (Continued) 000334-48-5 004436-32-9 000143-07-7 Subtotal Grand total 2390 2474 2544 Decanoic acid 9-Decenoic acid Dodecanoic acid 1.63 ± 0.7 a 1.59 ± 0.06 a 2.68 ± 0.19 a 18.09 1209.08 0.1348 0.1315 0.2217 0.19 ± 0.01 b 0.0149 1.08 ± 0.08 b 0.0846 2.8 ± 0.17a 0.2194 1.4962 13 100 1276.38 1 1.0185 13.64 100 1661.47 CAS numbers were obtained from Wiley library LRI of all the relative tables was determined on the DB-FFAP column, relative to C5-C40 hydrocarbons. 3 Descriptors were retrieved from http://www.thegoodscentscompany.com abc ANOVA at 95% confidence level, same letters in the same row indicate no significant difference 2 43 0.18 ± 0.01 b 0.0108 Rancid, sour, fatty 1.03 ± 0.06 c 0.062 Waxy, green, fatty, soapy 2.91 ± 0.12 a 0.1751 Fatty, waxy 0.821 100 Table 3.5 Selected volatiles quantified in mango wine and their odour activity values (OAV) in low (unfortified mango juice 16.6oB), medium (initial TSS of 23°B) and high (initial TSS of 30°B) sugar fermentation CAS1 LRI2 Alcohol 000078-83-1 1172 000123-51-3 1237 000060-12-8 1964 Compounds Isobutyl alcohol Isoamyl alcohol 2-Phenylethyl alcohol Low (mg /L) 62.3±8.3a 107.2±10.3 a 56.0±7.2 a OAV 1.56 3.57 5.60 Medium (mg /L) 31.9±6.2b 387.1±12.2b 2.26±0.15b OAV 0.80 12.90 0.23 High (mg /L) 39.1±7.3c 414.4±15.6c 2.49±0.18c OAV 0.98 13.81 0.25 Ester 000141-78-6 000110-19-0 000123-92-2 000103-45-7 000123-66-0 000106-32-1 000110-38-3 000106-33-2 1009 1020 1112 1862 1297 1453 1746 1887 Ethyl acetate Isobutyl acetate Isoamyl acetate 2-Phenylethyl acetate Ethyl hexanoate Ethyl octanoate Ethyl decanoate Ethyl dodecanoate 3.93±1.19 a 0.52 2.34±0.82b 0.31 0.56±0.15c 0.07 0.042±0.012 a 0.03 0.025±0.092b 0.02 0.013±0.006c 0.01 0.258±0.083 a 8.60 0.7847±0.11b 26.16 0.5529±0.12c 18.43 0.764±0.16a 3.06 0.423±0.051b 1.69 0.341±0.043c 1.36 0.301±0.025 a 21.50 0.792±0.062b 56.57 0.375±0.028c 26.79 1.79±0.25a 895.00 0.4481±0.052b 224.05 0.3865±0.042b 183.25 1.09±0.23 a 5.45 0.0615±0.008b 0.31 0.0605±0.005b 0.30 a 1.75±0.45 0.30 0.706±0.056b 0.12 0.692±0.052b 0.12 Acid 000064-19-7 000142-62-1 000124-07-2 000334-48-5 000143-07-7 1549 1890 2170 2390 2544 Acetic acid Hexanoic acid Octanoic acid Decanoic acid Dodecanoic acid 442.25±59.1 a 1.81±0.21 a 4.64±0.76 a 2.16±0.25 a 0.6339±0.05 a 2.21 0.60 0.53 0.22 0.06 487.69±56.2b 1.67±0.21b 1.18±0.14b 1.81±0.14b 0.6402±0.081b 1 2.44 0.56 0.13 0.18 0.06 538.63±66.5c 1.57±0.16b 1.08±0.09b 1.79±0.12b 0.6453±0.093b CAS numbers were obtained from Wiley library LRI of all the relative tables was determined on the DB-FFAP column, relative to C5-C40 hydrocarbons. 3 Descriptors were retrieved from http://www.thegoodscentscompany.com abc ANOVA at 95% confidence level, same letters in the same row indicate no significant difference 2 44 2.69 0.52 0.12 0.18 0.06 Organoleptics3 Odour threshold 40 Fruity, wine-like 30 Alcoholic, fruity, banana-like 10 Rose, floral 7.5 1.6 0.03 0.25 0.014 0.002 0.2 5.9 Pineapple, fruity, varnish Fruity, sweet, apple Fruity, banana, sweet Floral, rose, sweet Sweet, pineapple, waxy Floral, fruity, brandy Waxy, sweet, apple Soapy, waxy, floral 200 3 8.8 10 10 Vinegar, pungent Fatty, soapy, sour Fatty, soapy, sour, fruity Fatty, rancid, sour Fatty, waxy 3.2.5.1 Ethanol Ethanol accounted for 67.89%, 87.43% and 91.24% (by RPA) of all volatiles produced for the low (unfortified, 16.6oB), medium (23°B) and high (30°B) sugar fermentations respectively (Table 3.4). At low levels, ethanol may enhance the sensory perception of aroma compounds. However, in excess, ethanol has a masking effect and can directly lead to a burning sensation (Swiegers et al., 2005). This may be especially prominent in the high sugar fermentation resulting in a detrimental effect on wine quality by masking the other volatiles present with ethanol accounting for more than 90% of all volatiles present. At a cellular level, ethanol toxicity can lead to the inhibiton of volatile production by affecting the transport of sugar and nitrogen into the cell due to its effects on membrane permeability and may reduce proton motive force (Bisson and Karpel 2010; Fleet and Heard, 1993). This appears to be the case where volatile production was significantly lower in high sugar fermentation. It has been suggested that oxygenation and increasing assimilable nitrogen content can be used to reduce the effects of ethanol toxicity for the production of a high ethanol wine (Casey and Ingledew 1986) since oxygenation allows for the formation of unsaturated fatty acids which decreases membrane fluidity caused by high ethanol content (Kyung et al. 2003). 3.2.5.2 Fusel alcohols Fusel alcohols are important components of the wine bouquet due to their organoleptic properties. The major fusel alcohols produced were 2-phenylethyl alcohol, isoamyl alcohol and isobutyl alcohol. Most fusel alcohols detected were present at the highest concentrations in the low sugar fermentation (Table 3.4). 2-Phenylethyl alcohol and isobutyl alcohol were found at the highest concentrations in the low sugar fermentation. Both higher alcohols were present at levels above their odour 45 thresholds in the low sugar fermentation; therefore, they are likely to have a positive impact on the wine flavour. However, these two alcohols were present at concentrations below their odour thresholds in the medium and high sugar fermentations. Isoamyl alcohol was found at levels above its odour threshold in all three mango wines, with the highest concentration in the high sugar fermentation and the lowest in the low sugar fermentation. Although isoamyl alcohol has a pleasant fruity and sweet odour, the excessive amounts (387.1 and 414.1 mg/L respectively) detected in the medium and high sugar fermentations may have an undesirable effect on wine aroma. It has been reported that fusel alcohols can be detrimental to wine aroma at levels above 350 mg/L (Swiegers et al. 2005). The reason for this abnormality is not clear, but it may be due to the presence of leucine (one of the major amino acids in mango juice) and the redox imbalance presented due to the hypersomotic stress. Furthermore, isoamyl alcohol can also be synthesized from sugar metabolism via pyruvate (which can be converted to leucine ) (Kohlhaw 2003). The higher amounts of isoamyl alcohol present in the high sugar fermentation is highly likely due to the increased sugar metabolism producing large amounts of pyruvate which were consequently converted to isoamyl alcohol. Previous studies have shown that high sugar concentrations favour the formation of fusel alcohols rather than the corresponding carboxylic acid (Jain et al. 2011; Styger et al. 2013) due to the regeneration of NADH from NAD +. The production of leucine from pyruvate also regenerates NADH from NAD+ (Kohlhaw 2003) Other than their contribution to the wine bouquet, higher alcohols also serve as important precursors for the formation of branch-chained esters which are beneficial to wine flavour. In this study, the amounts of higher alcohols were within the desirable range and likely to make positive contributions to the overall mango wine flavour profile. 46 3.2.5.3 Terpene alcohols A number of terpene alcohols were also detected in all three mango wines (Table 3.4). Citronellol, geraniol and linalool were detected but at levels below their quantification limits. Hence, their contribution to the overall wine flavour profile would be likely due to their synergistic effects. Citronellol also increased after fermentation, indicating its biosynthesis of S. cerevisiae (Carrau et al. 2005) or release from the bound terpenols (glycosides) due to enzymatic activity (Gunata et al. 1986). S. cerevisiae is able to convert geraniol to citronellol (Carrau, et al. 2005). This may explain the low amounts of geraniol present in the mango wines. 3.2.5.4 Esters Ethyl esters were the major esters present in mango wine. In low sugar fermentation, , ethyl esters made up 18.8% of all volatiles (by RPA). The amount of ethyl esters detected in the medium sugar and high sugar fermentation was significantly lower, only 2.23% and 1.13% (by RPA) of total volatile production respectively. In terms of potential impact on the wine flavour profile, ethyl octanoate is likely to have the most significant influence on wine flavour due to its low odour threshold, despite not being produced at an exceptionally high level (Table 3.5). Ethyl hexanoate was also present at levels above its odour threshold in all three mango wines and detected in the highest concentration in the medium sugar fermentation at 0.792 mg/L (Table 3.5). Ethyl esters are likely to be produced enzymatically during the synthesis or degradation of fatty acids (Alves et al. 2010) and/or chemical reactions that occur in the presence of alcohol with their prevalence attributed to the high production of medium-chain fatty acids in S. cerevisiae yeasts (Saerens et al. 2008). Ethyl esters are produced by transferase reactions in which alcohols react with fatty acyl-CoAs derived from the metabolism of fatty acids. It 47 has been shown that S. cerevisiae exhibited enzymatic activity required for the synthesis of medium-chain fatty acid ethyl esters (Saerens et al. 2008). The major acetate esters produced were 2-phenylethyl acetate, isoamyl acetate, geranyl acetate and neryl acetate (Table 3.4). Acetate esters confer fruity and floral notes and are important to wine flavour. Acetate esters are produced from the reaction of acetyl-CoA with alcohols (Perestrelo et al. 2006) and thus, the higher production of acetates in the low sugar fermentation may be due to the higher quantities of branched-chain fusel alcohols present. With the exception of isoamyl acetate, all major acetate esters were found in the highest concentration in the low sugar fermentation. Isoamyl acetate and 2-phenylethyl acetate were both detected at levels above their odour thresholds and may impart fruity and floral notes to the wine flavour. Isoamyl acetate was found in the highest amount in the high sugar fermentation, similar to isoamyl alcohol. This is likely due to the higher amounts of isoamyl alcohol as a precursor present in the high sugar fermentation. 3.2.5.5 Organic acids The main acid produced in all three treatments was acetic acid which made up more than 90% of all acids formed, making it main determining factor of volatile acidity. Acetic acid exerts a detrimental effect on wine quality if present at excessive levels. The amount of acetic acid produced in the low, medium and high sugar fermentations were 442, 487 and 538 mg/L respectively; this translates to OAV of 2.21, 2.44 and 2.69 respectively. Studies have shown that S. cerevisiae produces relatively small quantities of acetic acid in must with a moderate sugar concentration (less than 220 g/L) (Bely et al. 2003), therefore, the relatively high levels of acetic acid produced was likely due to the high sugar concentrations. 48 Other acids that were detected include hexanoic acid, octanoic acid, decanoic acid, 9decenoic acid and dodecanoic acid. These short- and medium-chain fatty acids may contribute to undesirable fatty, rancid and soapy off-odours. The amounts of short- and medium-chain fatty acids produced were all well below their odour thresholds (Table 3.5) and were unlikely to impart negative flavour notes but may impart some desired complexity to the wine flavour. With the exception of acetic acid, most of the other organic acids were found in the highest amount in the low sugar fermentation. During normal alcoholic fermentation, the pyruvate produced during glycolysis is decarboxylated to form acetaldehyde, which is then reduced to ethanol, with the production of NAD+. However, NAD+ is generated when glycerol is produced by yeast cells under hyperosmotic conditions, resulting in an excess of NAD+ (with a deficit of NADH). To rectify this redox imbalance, acetaldehyde is converted into acetic acid and this biochemical reaction reduces NAD+ to NADH instead (Figure 3.4). Consequently, more acetic acid was produced in the mango wine with a higher starting sugar content. This is supported by the fact that glycerol was produced in higher amounts in the medium and high sugar fermentations. 49 Glucose ATP ADP Dihydroxyacetone phosphate Fructose-1,6-Bisphosphate NADH NAD+ Glycerol-3-phosphate Glyceraldehyde-3-phosphate NAD+ NADH Glycerol-3-phosphate Pyruvate NAD+ NADH Acetic acid Acetaldehyde Glycerol NADH NAD+ Ethanol Figure 3.4 Metabolic pathways governing the production of ethanol, glycerol and acetic acid 3.2.5.5 Carbonyl compounds The major aldehydes detected in mango wine were acetaldehyde, p-tolualdehyde and benzaldehyde with acetaldehyde being the main aldehyde. Acetaldehyde production was directly correlated to the initial sugar concentration. This can be explained by the redox imbalance due to the high sugar concentration. As mentioned previously, acetaldehyde is oxidised to acetic acid (instead of reduced to ethanol) for NADH regeneration when excess glycerol is produced under hyperosmotic conditions (Figure 3.4). However, due to the excessive amounts of glycerol produced, a shortage of NADH occurs, causing part of the acetaldehyde to be excreted by S.cerevisiae into the wine. This is further supported by the fact that acetaldehyde and glycerol concentents were directly correlated. 50 Acetaldehyde imparts a fresh, fruity aroma to wines at low levels. However, at higher concentrations, acetaldehyde results in green apple like, even a pungent irritating odour (Miyake and Shibamoto 1993). Other than its impact on wine quality due to its organoleptic properties, acetaldehyde is a precursor for acetoin (Romano et al. 1994). The ketones detected in mango wine were β-damascenone, 2-nonanone and acetoin, with β-damascenone. β-Damascenone was the major ketone and one of the few compounds detected in both fresh mango juice and mango wine; and it decreased during fermentation. Although these ketones had low concentrations, they may contribute to “floral” or “fruity” aroma to mango wines synergistically. In addition, it was found at greater concentrations in the medium and high sugar fermentation than in low sugar fermentation. This appears to suggest that the increased hyperosmotic stress may have disrupted the normal metabolic activities of the yeast, resulting in decreased degradation of certain compounds. 3.2.5.6 Terpene hydrocarbons Despite being the predominant class of volatiles in fresh mango juice, the terpenes made up less than 1% (by RPA) of all volatiles present in mango wine. Most of these compounds were catabolised during the fermentation process (Table 3.1). This suggests that the characteristic flavour of the fresh mango juice may not be present in the mango wine. The major terpene hydrocarbon detected in the mango wine was α-caryophyllene, an isomer of β-caryophyllene (El Hadri et al. 2010), a terpene found in fresh mango juice; therefore, this suggests that some form of biotransformation occurred during the fermentation process. In addition, although not quantified, the peak area of α-caryophyllene was about 10 times that of β-caryophyllene. This suggests that some form of biosynthesis of α- caryophyllene occurred during the fermentation process. Although this compound is a quantitatively minor compound, it is of interest to note that studies have shown that many of 51 the monoterpenes and sequisterpenes, i.e. the hydrocarbon terpenes, have been linked to potential health benefits (El Hadri et al. 2010; Buchbauer and Ilic 2013). Some of the other terpene hydrocarbons linked to possible health benefits include limonene, β-caryophyllene and myrcene, all of which are detected in mango wine (Buchbauer and Ilic 2013). Therefore, reducing the degradation of these terpene hydrocarbons may not only improve the flavour profile of the mango wine but also potentially harness some of the health benefits of mango wine if consumed in adequate amounts. 3.2.6 Effects of redox potential on overall wine quality Mouthfeel and flavour complexity are two of the most important factors affecting perceived wine quality, and they are influenced by glycerol and volatile composition respectively. In turn, hyperosmotic stress influences wine quality due to the corrective mechanisms undertaken by the yeast cells for survival by producing glycerol. Therefore, the balance of NAD+ and NADH can influence the volatile and glycerol production during fermentation and consequently affect wine quality. However, as each yeast strain may have different metabolic capabilities, it may be necessary for more in-depth studies to be conducted on S. cerevisiae MERIT.ferm to fully maximise its potential for the fermentation of a mango wine. Metabolically, glycerol is primarily produced for the maintenance of intracellular redox balance, by converting the excess NADH generated during biomass formation to NAD+ (Hohmann 1997; Remize et al. 1999). Hence, when excess glycerol is produced under hyperosmotic conditions to prevent the exodus of water from the cell, an excess of NAD+ arises, leading to imbalance. excreted. Acetaldehyde is very toxic to cells and excess of it must be While an increase in glycerol concentration may improve wine quality by enhancing the mouthfeel of the wine, this is coupled to an increase in acetic acid and 52 acetaldehyde production and hence may result in an undesirable vinegar-like character if acetic acid is produced at excessive levels (Swiegers and Pretorius, 2005). On the other hand, excess NADH promotes production of fusel alcohol and hinders the production of esters due to the need for precursor fusel acids which requires NAD+ Therefore, it has been hypothesized that the imbalance in NAD+/NADH levels, due to hyperosmotic conditions generally drove the production of higher alcohols in an attempt to reduce NADH to NAD+ (Jain et al. 2011; Styger et al. 2013). Furthermore, the production of a fusel alcohol versus fusel acids is therefore also dependent on the redox requirements of the yeast (Bisson and Karpel, 2010). In addition, despite the stuck/laggish fermentation in the high sugar juice samples, the final ethanol content obtained was still higher than that of the other two fermentations. This indicates that, with the right conditions, the production of mango wines with high alcohol content is possible. Several studies have suggested that supplementing a high sugar fermention with amino acids and nitrogenous compounds reduces the risks of a stuck fermentation (Arrizon and Gschaedler, 2002; Alexandre, Rousseauz and Charpentier, 1994; Chaney et al., 2006). 3.3 Conclusion In conclusion, the initial sugar content in the mango juice had a dramatic effect on the volatile production and fermentation kinetics: Generally, the higher the initial sugar concentration, the more significant the hyperosmotic stress the starting material presented to the yeast. This hyperosmotic stress affected the normal metabolism of the cell and consequently also has a negative effect on the production of volatile compounds. It could be partly due to the redox imbalance hyperosmotic stress presented to the yeast and/or the 53 effects of ethanol toxicity affecting the functioning of the cellular membrane and resulting in a disruption to normal cellular mechanisms. Although it was expected that initial sugar concentration would have a negative impact on wine quality, the results obtained in this chapter demonstrated the specific effects of hyperosmotic stress on the production of volatiles, and how the balance of NAD+ and NADH affects the metabolic pathways and consequently affects the flavour profile of the wine. for instance, higher amounts of ketones and terpene hydrocarbons were detected in wines with higher sugar concentrations. Potentially, the data obtained from this study can be further investigated to use initial sugar concentration as a mean of modulating wine flavour. Furthermore, the data obtained from this study can be used as an indicator of the flavour profile of the wine that will be produced from batches of mangoes with different maturities, and consequently different sugar content. 54 Chapter 4 Effects of Co-fermentation of Saccharomyces cerevisiae and Williopsis saturnus Yeasts on Volatile Production 4.1 Introduction With its ability to modulate most major constituents of wine (including residual sugar, ethanol, polyols, acids and phenolics) (Pretorius, 2000), S. cerevisiae is widely utilised commercially, specifically for producing wines with high ethanol and low residual sugar content. However, non-Saccharomyces yeasts are able to produce a diversity of novel aromatic compounds (such as acetate esters and other branch-chained esters) different from S. cerevisiae wines and has potential application in wine production to increase aroma diversity (Dubourdieu et al. 2006) due to the host of enzymes such as esterases, glycosidases, amongst others. In addition, Williopsis saturnus var. saturnus strains has a high acetate ester producing ability and are known to convert higher alcohols into the corresponding acetate esters (Yilmaztekin et al. 2009) and was chosen for its potential for flavour enhancement by contributing to the fruity and floral note in wine (Iwase et al. 1995; Yilmaztekin et al. 2009). S. cerevisiae MERIT.ferrm was selected due to its vigorous fermenting ability and tolerance toward high sugar and ethanol concentrations. To harness the benefits of both Saccharomyces and non-Saccharomyces yeasts, mixed culture (also known as co-fermentation or co-inoculation) is a potential fermentation strategy that can be utilised. Mixed culture fermentations involving Saccharomyces and non- Saccharomyces yeasts for grape wines have been extensively studied and shown to allow for differentiation of wines due to increased flavour complexity and mouthfeel enhancement (Fleet 2008; Viana et al. 2009). Some commercial non-Saccharomyces yeasts in use now include Kluyveromyces thermotolerans and Torulapora delbrueckii (Pretorius et al. 1999). 55 This chapter studied the chemical composition of three different mango wines – one produced from a S. cerevisiae MERIT.ferm, one produced from W. saturnus NCYC 500, and a mango wine produced from a simultaneous fermentation of yeasts S. cerevisiae MERIT.ferm and W. mrakii NCYC 500. 4.2 Results and discussion 4.2.1 Total soluble solids, sugar concentrations and yeast cell count The growth kinetics, changes in TSS and sugar concentration is illustrated in Table 4.1. The yeast cell populations increased from the initial 5.50×105 CFU/mL for S. cerevisiae, 6.33×105 CFU/mL for W. mrakii to 9.01×108 CFU/mL and 9.65×107 CFU/mL respectively after a 21-day fermentation period. For the mixed culture fermentation, S. cerevisiae and W. mrakii yeasts were inoculated in a 1:1000 ratio. S. cerevisiae was initially present at a level of 2×102 CFU/mL and W. mrakii was present at a level of 1.67 ×105 CFU/mL. At the end of 21 days, no W. mrakii was detected and S. cerevisiae reached a population size of 7.05×108 CFU/mL (Table 4.1). Lysine agar was used to enumerate W. mrakii towards the end of the fermentation when numbers were too low to count on PDA. In the mixed culture fermentation, where S. cerevisiae and W. mrakii were simultaneously inoculated, S. cerevisiae dominated the fermentation. This was expected as S. cerevisiae is known to be a vigorous fermenter with the ability to rapidly adapt to the harsh oenological conditions. S. cerevisiae grew to about 108 CFU/mL and remained stationary until the end of the fermentation. W. mrakii, on the other hand, declined rapidly and steadily (Figure 4.1) These could either be due to the inhibitory effects from the toxic metabolites of S. cerevisiae fermentation (such as ethanol), and/or the lack of nutrients for growth due to its weaker capability to compete for nutrients with the more robust Saccharomyces species. The 56 toxic effects of ethanol coupled with the lack of nutrients might have a detrimental effect on the survival of the W. mrakii. Other potential factors for the early death of W. mrakii yeasts include oxygen unavailability (Holm Hansen, et al. 2001), formation of other toxic compounds, quorum sensing and cell-to-cell contact mechanism (Fleet and Heard 1993; Nissen and Arneborg 2003; Nissen et al. 2003; Farkas et al. 2005). Table 4.1 Physicochemical properties, reducing sugar and organic acid content before and after fermentation Yeast strains Physiochemical properties pH o Brix MERIT Day 0 NCYC 500 Mixed Culture MERIT Day 21 NCYC 500 Mixed Culture 3.51± 0.01a 16.7 ± 0.01a 3.50± 0.01a 16.7 ± 0.01a 3.51± 0.01a 16.7 ± 0.01a 3.52± 0.02a 5.3 ± 0.01b 3.50± 0.01a 11.46 ± 0.5c 3.51± 0.01a 5.5 ± 0.01b Plate count (105 CFU/mL) S. cerevisiae MERIT.ferm W. mrakii NCYC 500** 6.33 ± 1.10a - 6.33 ± 1.10a 965 ± 85b 7051 ± 993d N.D*. 0.26 ± 0.03a 0.90 ± 0.09a 0.079 ± 0.012a 0.15 ± 0.03a 0.25 ± 0.02a 0.91 ± 0.08a 0.079 ± 0.011a 0.14 ± 0.03 a 0.24 ± 0.02 a 0.43 ± 0.02b 0.079 ± 0.012 a 0.13 ± 0.02 a 0.25 ± 0.02 a 0.48 ± 0.05b 0.084 ± 0.014b 0.13 ± 0.01 a 1.11 ± 0.02 b 0.38 ± 0.05 b 2.53 ± 0.12 c N.D.* N.D.* 0.013 ± 0.00 b Organic acid (g/100 mL) Citric acid Malic acid Succinic acid Tartaric acid 0.002 ±0.001b 9012 ± 1291c 6.67±1.10a 0.26 ± 0.02a 0.90 ± 0.07a 0.081 ± 0.012 a 0.14 ± 0.04 a 0.25 ± 0.03a 0.46 ± 0.05b 0.085 ± 0.011b 0.14 ± 0.02 a Reducing sugars (g/100 mL) Fructose 4.99 ± 0.05a 4.97 ± 0.04 a 4.98 ± 0.06 a N.D.* Glucose 0.68 ± 0.03a 0.66 ± 0.05 a 0.68 ± 0.02 a N.D.* Sucrose 12.45 ± 0.12a 12.43 ± 0.16 a 12.47± 0.17 a 0.013 ± 0.00 b *N.D.: not detected ** W. mrakii was enumerated on both lysine and potato dextrose agar abcd ANOVA (n=4) confidence level at 95%, same letters in the same row indicate no significant differences The °Brix values decreased from 16.7 °B (in fresh mango juice) to 5.3 ° B for s monoculture and to 5.5 ° B for the mixed culture after 21 days. The sugar contents of the mango wine fermented with S. cerevisiae monoculture and mixed culture showed no significant differences. Only trace levels of sucrose detected at 0.013 and 0.015g/100 mL 57 respectively (Table 4.1) while both fructose and glucose were completely metabolised. On the other hand, significant amounts of all three sugars were detected in the W. mrakii wine, with 1.11 g/100 mL of fructose, 0.38 g/100 mL of glucose and 2.53 g/100 mL of sucrose; the total soluble solids content was about 11.46 °B in the NCYC 500-fermented wine. Sugar consumption was much lower by Williopsis yeast than Saccharomyces. Interestingly, in the NCYC 500-fermented wine, the largest decrease amongst the sugars was for sucrose. This indicates that W. mrakii utilised sugars in a different manner from the glucophillic S. cerevisiae (Gonzalez et al. 2013). (a) TSS (°Brix) 20 15 10 5 0 0 5 10 15 Time (Days) 20 (b) Log CFU / mL 10 8 6 4 2 0 -2 0 5 10 15 20 Time (Days) Figure 4.1 (a) Changes TSS (°B) during fermentation for S. cerevisiae MERIT.ferm monoculture ( ), W. mrakii NCYC 500 monoculture ( )and mixed culture fermentation ( ) (S. cerevisiae:W. mrakii at a ratio 1:1000) (b) Changes in yeast population during fermentation for S. cerevisiae MERIT.ferm monoculture ( ), W. mrakii NCYC 500 monoculture ( ), S. cerevisiae in mixed culture fermentation ( )and W. mrakii in mixed culture 58 fermentation ( ); S. cerevisiae:W. mrakii at a ratio 1:1000 in mixed culture fermentation 4.2.2 pH and organic acids The changes in pH and organic acid contents before and after fermentation can be found in Table 4.1. The pH remained relatively constant throughout the fermentation period between 3.4 to 3.6. With the exception of malic acid, there were no significant changes in the organic acids (Table 4.1). Malic acid decreased by approximately 50% after fermentation. As D,L-malic acid was added to the fresh mango juice to achieve a pH of 3.5, the total malic acid increased from initial 0.3 g/100 mL to about 0.88 g/100 mL after pH adjustment (data not shown). The absence of malolactic fermentation (as indicated by the absence of lactic acid bacteria and lactic acid), and the inability of the selected yeast strains in this study to metabolise malic acid suggest that the observed decrease in malic acid content is most likely due to the passive diffusion of D-malic acid into the cells (Coloretti et al. 2002). Furthermore, the addition of 100 ppm of potassium metabisulphite to the juice at the beginning of the fermentation would inhibit the growth of unwanted microorganisms. There was a minor increase in the amount of succinic acid after fermentation which could be formed from the reduction of oxaloacetate and malate from the Krebs cycle. Succinic acid is commonly present in alcoholic fermentation (Swiegers et al. 2005) and undesirable at high concentrations due to the “salty” or “bitter” taste imparted (Whiting 1976). 4.2.3 Volatile composition of mango wines A total of 10 alcohols, 49 esters, 3 aldehydes, 1 furan, 4 ketones, 2 lactones, 20 hydrocarbon terpenes, 2 oxides of terpene alcohols, 10 organic acids and 5 other miscellaneous compounds (such as furans and lactones) were identified. The complete list of volatiles produced can be found in Table 4.2 while some of the potentially impactful odourants which were quantified can be found in Table 4.3. 59 While a large number of volatiles found in mango juice were metabolised, a number of volatile compounds were also produced from both chemical and enzymatic reactions during fermentation. Consequently, there were significant differences between the S. cerevisiae, W. mrakii and mixed culture fermented mango wines. The volatile profile for the mixed culture wine was similar to that of the S. cerevisae wine, while the W. mrakii wine exhibited a significantly different volatile profile. This was expected as it was apparent that S. cerevisiae dominated the fermentation in the mixed culture fermentation. 60 Table 4.2 Complete volatiles for mango wine fermented with yeasts S. cerevisiae MERIT.ferm, W. mrakii NCYC 500 and mixed culture CAS1 LRI2 Compounds Peak Area 000064-17-5 1028 Ethanol 2025±50.33a 2025 71.21 71.21 575.5±26.48 b 575.5 53.69 53.69 2188±63.57a 2188 70.25 70.25 Strong, alcoholic s 000078-83-1 1172 Isobutyl alcohol 000137-32-6 1220 Active amyl alcohol 000123-51-3 1237 Isoamyl alcohol 14.19±0.57 a 6.23±0.18 a 28.28±1.36 a 0.499 0.219 0.995 9.06±0.21 b 5.67±0.19 b 15.63±1.22 b 0.845 0.529 1.458 15.5±0.440 c 6.57±0.18 a 43.29±0.650 c 0.498 0.211 1.39 Ethereal, winey Fusel, winey and solvent-like Fusel, alcoholic, cognac, fruity, banana Ethereal, fruity and alcoholic, fusel oil Green, grassy Cooked vegetable Camphor, pine , sweet Subtotal Alcohol RPA (%) Mixed Peak Area RPA (%) 000111-27-3 1448 Hexanol 0.63±0.0288a 0.022 1.03±0.11 b 0.096 0.67±0.031 a 0.021 000928-96-1 1475 cis-3-Hexenol 000505-10-2 1706 Methionol 000470-08-6 1794 β-Fenchol 1.04±0.05 a 1.24±0.029 a 0.63±0.03 a 0.037 0.044 0.022 1.71±0.08 b 0.94±0.033 b 0.12±0.01 b 0.16 0.088 0.011 0.38±0.01 c 1.37±0.010 c 0.15±0.01 c 0.012 0.044 0.005 0.06 1.105 0.04 0.012 4.405 1.31±0.07 a 87.52±1.560 c 1.43±0.070 c 0.22±0.02 a 158.5 0.042 2.811 0.046 0.007 5.087 Floral, rosy, sweet, citrus Floral, rose and honey-like Citrus, orange, floral and rose Mild floral 000106-22-9 000060-12-8 000078-70-6 000128-37-0 1867 1964 1999 2090 Citronellol 2-Phenylethyl alcohol Linalool p−Cresol 1.3±0.07 a 56.97±1.60 a 0.49±0.01 a 0.25±0.02 a 111.3 0.046 2.004 0.017 0.009 3.994 000105-54-4 000123-66-0 64187-83-3 054653-25-7 000106-32-1 1098 1297 1358 1392 1453 Ethyl butanoate Ethyl hexanoate Ethyl cis-3-hexenoate Ethyl 5-hexenoate Ethyl octanoate 2.3±0.10 a 0.28±0.01 a 0.1±0.003 a 3.01±0.030 a 83.84±2.02 a 0.081 0.01 0.004 0.106 2.949 1.21±0.03 b 0.71±0.52 b 0.03±0.001 b 0.02±0.001 b 3.14±0.06 b 0.113 0.066 0.003 0.001 0.293 2.54±0.14 a 6.28±0.220 c 0.1±0.005 a 2.51±0.03 c 138.8±2.140 c 0.082 0.202 0.003 0.081 4.456 35194-38-8 1486 Ethyl 7-octenoate 000123-29-5 1624 Ethyl nonanoate 000110-38-3 1746 Ethyl decanoate 0.41±0.005 a 1.81±2.32 a 305.3±5.12 a 0.014 0.064 10.738 0.01±0.001 b 1.04±0.062 b 1.05±0.05 b 0.001 0.097 0.098 0.36±0.04 c 1.37±0.110 c 246.2±5.470 c 0.012 0.044 7.907 Fruity, sweet, apple, fresh Sweet, pineapple, fruity Sweet, fruity, pineapple, green Fruity, pineapple Waxy, sweet, pineapple and fruity Tropical, fruity Waxy, soapy, cognac Sweet, waxy, fruity 067233-91-4 1795 Ethyl 9-decenoate 87.45±1.39 a 3.076 11.3±0.565 b 1.054 63.07±0.600 c 2.025 Fruity, fatty 000106-33-2 1887 Ethyl dodecanoate a 0.077 c 1.186 Sweet, waxy, soapy Subtotal Ethyl ester 500 Odour descriptor3 Merit Peak Area RPA (%) 71.41±0.53 2.512 0.64±0.02 b 11.84±1.49 b 0.43±0.03 b 0.13±0.01 b 47.2 0.82±0.70 61 b 36.94±0.260 Table 4.2 (Continued) 000124-06-1 2201 Ethyl tertradecanoate 2.42±0.0434a 000628-97-7 2373 Ethyl hexadecanoate 24.88±0.62 a 054546-22-4 2402 Ethyl 9-hexadecenoate 1.22±0.0427 a 598.3 Subtotal Acetate ester 000109-60-4 1002 Propyl acetate 000141-78-6 1009 Ethyl acetate 000110-19-0 1020 Isobutyl acetate 000123-86-4 1061 Butyl acetate 21.044 0.75±0.01 a 4.41±0.40 a 1.17±0.08 a 0.65±0.036 34.32 3.98±5.18 b 0.155 148.6±1.45 b 2.18±0.02 b 0.88±0.00 b 9.63±0.21 b 86.67±1.87 b 13.85±0.30 b 0.023 0.05 0.166 0.005 3.202 0.026 0.041 a 0.54±0.031 b 1.77±0.014 b 0.05±0.001 b 0.371 1.36±0.066 c 1.97±0.041 c 0.8±0.01 c 516.6 0.044 0.063 0.026 Sweet, waxy, creamy Waxy, fruity, creamy 16.588 1.27±0.03 c 0.041 Fusel, sweet, fruity 94.35±0.38 c 3.03 Ethereal, fruity, sweet, green 2.07±0.08 b 0.066 Sweet, fruity, banana, tropical 0.65±0.03 a 0.021 Fruity, banana 3.83±0.16 c 0.123 Ethereal, fruity, banana 0.42±0.01 c 0.013 Sweet, fruity, banana 1.292 2.91±0.14 c 0.093 0.03±0.001 c 0.43±0.01 c Fresh, green, sweet, fruity, apple Sweet, leafy green, fruity Fruity, green apple banana sweet Floral, green, rose, fruit Floral, rose, citrus, pear Sulphurous Floral, green, herbal Sweet, honey, floral rosy 13.869 0.203 0.082 0.21±0.01 a 1.97±0.08 a 003681-71-8 1324 cis-3-Hexenyl acetate 0.66±0.04 a 0.023 002497-18-9 1367 2-Hexenyl acetate 000142-92-7 1413 n-Hexyl acetate 0.11±0.01 a 0.33±0.02 a 0.004 0.012 0.45±0.01 b 2.22±0.138 b 0.042 0.207 0.69±0.01 a 3.69±0.09 a 0.21±0.01 a 2.56±0.12 a 10.56±0.52 a 27.97 0.024 0.13 0.007 0.09 0.371 0.984 2.82±0.04 b 0.49±0.06 b 0.34±0.01 b 3.56±0.151 b 41.93±0.20 b 317.6 0.263 0.046 0.032 0.332 3.913 29.638 0.003 0.12±0.003 b 0.011 0.16±0.001 c 0.005 Waxy, fruity, fatty, green, fruity 0.078 0.026 0.238 0.079 0.058 0.192 3.21±0.10 b 0.38±0.01 b 7±0.320 a 0.37±0.018 b 0.02±0.001 b 0.11±0.004 b 0.3 0.035 0.653 0.034 0.002 0.01 2.92±0.03 c 0.8±0.01 c 9.54±0.110 c 1±0.045 c 0.2±0.001 c 1.96±0.120 c 0.094 0.026 0.306 0.032 0.006 0.063 Fruity, green, oily, floral Waxy, fruity, fatty, fruity Sweet, fruity, waxy Oily, sweet, brandy, apricot Fruity, floral, green, winey Waxy, banana, cognac, green 000628-63-7 1097 Active amyl acetate 000123-92-2 1112 Isoamyl acetate 000150-84-5 000141-12-8 16630-55-0 000105-87-3 000103-45-7 1659 1753 1764 1790 1862 Citronellyl acetate Neryl acetate Methionyl acetate Geranyl acetate 2-Phenylethyl acetate Subtotal Higher esters 0.131 0.875 0.043 000624-13-5 1506 Propyl octanoate 0.09±0.0059 0.007 0.069 0.899 8.087 4.15±0.10 c 0.11±0.01 c 0.33±0.03 b 2.64±0.04 a 40.99±1.10 c 154.2 0.001 0.014 0.133 0.004 0.011 0.085 1.316 4.951 a 005461-06-3 030673-60-0 002035-99-6 030673-38-2 000103-52-6 002306-91-4 1642 1721 1762 1859 1914 1973 Isobutyl octanoate Propyl decanoate Isoamyl octanoate Isobutyl decanoate 2-Phenylethyl butanoate Isoamyl decanoate 2.22±0.03 a 0.73±0.02 a 6.76±0.04 a 2.25±0.094 a 1.64±0.12 a 5.45±0.28 a 62 Table 4.2 (Continued) 006309-51-9 2180 Isoamyl dodecanoate 0.55±0.01 a 005457-70-5 2397 2-Phenylethyl octanoate 0.36±0.01 a 20.05 Subtotal Aldehydes 0.019 0.013 0.705 0.07±0.002 b 0.07±0.000 b 11.35 0.007 0.007 1.059 0.15±0.000 c 0.02±0.001 c 16.75 0.005 0.001 0.538 Cognac, green, waxy Sweet, fruity, creamy, floral 000075-07-0 939 Acetaldehyde 000100-52-7 1637 Benzaldehyde 5.27±0.03 a 0.19±0.0065 0.185 0.007 1.1±0.11 b 0.07±0.003 b 0.103 0.007 10.16±0.530 c 2.11±0.050 c 0.326 0.068 Pungent, fresh, green Almond, fruity, nutty 000104-87-0 1773 p-Tolualdehyde 3.05±0.05 a 8.51 0.107 0.299 1.78±0.03 b 2.95 0.166 0.276 1.79±0.01 b 14.06 0.057 0.452 Cherry 000821-55-6 1331 2-Nonanone 1.03±0.0366 0.036 1.85±0.006 b 0.172 1.32±0.020 c 0.042 Fruity, sweet, waxy 000513-86-0 1401 Acetoin 0.03±0.0012 0.001 0.75±0.027 b 0.07 0.25±0.330 c 0.008 Sweet, buttery creamy, dairy 0.002 0.12±0.007 b 0.011 0.07±0.0035 0.002 Fruity, peach, creamy Subtotal a a Ketones a 002305-05-7 1418 γ-Dodecalactone 0.06±0.0030 a 023696-85-7 1938 β-Damascenone 0.57±0.01 a 1.69 0.02 0.059 1.86±0.044 b 4.58 0.174 0.428 1.05±0.05 c 2.69 0.034 0.086 Woody, sweet, fruity, floral 000539-52-6 1331 3-(4-Methyl-3pentenyl)-furan 0.04±0.002 a 0.001 0.13±0.010 b 0.012 0.04±0.001 a 0.001 Woody 0.04 0.001 0.13 0.012 0.04 0.001 003387-41-5 1173 Sabinene 013466-78-9 1206 δ-3-Carene 000123-35-3 1246 β-Myrcene 0.46±0.01 a 0.09±0.001 a 0.68±0.008 a 0.016 0.003 0.024 3.97±0.07 b 1.55±0.06 b 4.43±0.09 b 0.37 0.145 0.404 0.39±0.014 c 0.19±0.01 c 0.38±0.031 c 0.013 0.006 0.012 095327-98-3 1254 Limonene 000555-10-2 1265 β-Phellandrene 027400-71-1 1290 β-Ocimene 0.07±0.00 a 0.11±0.01 a 0.49±0.0191 0.002 0.004 0.017 0.22±0.01 b 0.6±0.02 b 1.95±0.05 b 0.021 0.056 0.182 0.22±0.009 b 0.18±0.001 c 0.1±0.001 c 0.007 0.006 0.003 Woody, spicy, citrus Sweet, citrus Woody, citrus, fruity, mango, minty Sweet, citrus, peely Mint, turpentine Green, tropical, woody, floral 000099-85-4 1305 γ-Terpinene 000099-87-6 1339 p-Cymene 000586-62-9 1352 α-Terpinolene 0.71±0.03 a 1.59±0.25 a 1.72±0.12 a 0.025 0.056 0.061 1.74±0.01 b 6.1±0.09 b 20.97±1.01 b 0.162 0.569 1.957 0.69±0.032 a 1.99±0.010 c 1.86±0.020 c 0.022 0.064 0.06 Citrus, lime-Iike, oily Woody, lemon-like Sweet, fresh, piney citrus Subtotal Furan Subtotal Terpenes a a 63 043124-56-7 1382 Isoterpinolene 016409-43-1 1434 cis-Rose oxide 0.13±0.00 a 0.32±0.02 a 0.005 0.011 0.57±0.03 b 3.62±2.69 b b 0.053 0.338 0.09±0.004 c 0.31±0.0101 0.003 0.01 Green, vegetative, floral, herbal a a 001195-32-0 1529 p-Cymenene 000087-44-5 1695 β-Caryophyllene 9.76±0.23 0.22±0.0075 0.343 0.008 19.26±1.14 0.64±0.023 b 1.797 0.06 9.66±0.451 a 0.2±0.006 a 0.31 0.006 Phenolic, spicy, musty Sweet, woody, spice, clove 006753-98-6 1725 α- Humulene 017066-67-0 1829 β-Selinene 0.57±0.03 a 0.24±0.0083 0.02 0.009 1.1±0.038 b 0.44±0.02 b 0.103 0.041 0.92±0.010 b 0.21±0.003 a 0.03 0.007 Woody Herbal 6.258 17.4 0.559 a a Subtotal 17.17 0.604 016409-43-1 1434 cis-Linalool oxide Misc compounds 001786-08-9 1563 Nerol oxide Subtotal 0.023±0.0005a 0.71±0.01 a 0.73 0.001 0.025 0.026 0.29±0.22 b 0.95±0.02 b 1.24 0.027 0.089 0.116 0.024±0.001 a 1.23±0.020 c 1.25 0.001 0.039 0.04 Woody, floral, slightly green Green, vegetative, floral, mint 2.96±0.134 a 0.51±0.23 a 0.24±0.011a 0.8±0.036 a 19.41±0.64 a 0.53±0.03 a 0.54±0.02 a 4.46±0.09 a 1.75±0.084 a 31.2 2843 0.104 0.018 0.008 0.028 0.683 0.019 0.019 0.157 0.061 1.097 100 1.03±0.047 b 0.35±0.012 b 0.53±0.02 b 1.09±0.11 b 1.74±0.08 b 0.06±0.002 b 0.55±0.040 a 1.25±0.067 b 0.95±0.052 b 7.55 1072 0.097 0.033 0.049 0.102 0.162 0.006 0.051 0.116 0.088 0.704 100 0.72±0.040 c 0.55±0.025 c 0.23±0.01 a 0.72±0.002 c 19.8±0.360 a 16.33±0.060 c 0.87±0.020 b 2.36±0.050 c 1.92±0.070 c 43.5 3114 0.023 0.018 0.007 0.023 0.636 0.524 0.028 0.076 0.062 1.397 100 Sharp, pungent, sour, vinegar Acidic note Sharp, dairy-like, cheesy Sour, fatty, sweat, cheese Fatty, waxy, rancid Rancid, sour, fatty Faint, balsam Waxy, green, fatty, soapy Mild fatty, coconut, bay oil 000064-19-7 000079-31-2 000067-43-6 000142-62-1 000124-07-2 000334-48-5 000065-85-0 14436-32-9 000143-07-7 Acids Subtotal Grand Total 1549 1575 1728 1890 2170 2390 2455 2474 2544 Acetic acid Isobutyric acid Butanoic acid Hexanoic acid Octanoaic acid Decanoic acid Benzoic acid 9-Decenoic acid Dodecanoic acid 67.07 1 CAS numbers were obtained from Wiley database library LRI of all the relative tables was determined on the DB-FFAP column, relative to C5-C40 hydrocarbons. 3 Descriptors were retrieved from http://www.thegoodscentscompany.com abc ANOVA (n=4)at 95% confidence level with the same letters in the same row indicating no significant difference 2 64 Table 4.3 1 Major volatile compounds of alcoholic fermentation by S. cerevisiae, W. mrakii and a mixed culture of S. cerevisiae and W. mrakii CAS N o LRI2 000078-83-1 000123-51-3 1172 1237 000060-12-8 1964 S. cerevisiae MERIT OAV W. mrakii NCYC 500 OAV Mixed-culture OAV Odour Organoleptics3 (mg /L) (mg /L) (mg /L) threshold Isobutyl alcohol 89.1±11.9a 2.2 20.3±0.98 b 0.5 78.8±13.6 a 2 40 Fruity, wine-like Isoamyl alcohol 101.5±5.7 a 3.4 21.7±5.0 b 0.7 93.8±3.4 c 3.1 30 Alcoholic, fruity, banana, a b a 2-Phenylethyl alcohol 54.0±5.9 5.4 6.7±0.6 0.7 58.2±10.0 5.8 10 Sweet, rose, floral Terpenol 000098-55-5 000106-22-9 1352 1867 α-Terpinolene β-Citronellol 0.01±0.00 a 0.013±0.00 a 0.05 0.13 0.1±0.02 b 0.006±0.000 b 0.5 0.06 0.02±0.00 a 0.017±0.003 c 0.1 0.17 N.A. 0.1 000141-78-6 1009 Ethyl acetate 3.93±1.19 a 0.5 349.5±33.7b 46.6 30.3±4.3 c 4 7.5 000110-19-0 000123-92-2 000103-45-7 000123-66-0 1020 1112 1862 1297 Isobutyl acetate Isoamyl acetate 2-Phenylethyl acetate Ethyl hexanoate 0.026±0.012 a 0.27±0.07 a 0.59±0.15 a 0.27±0.05 a 0.016 9 2.4 19.3 0.22±0.04 b 3.54±0.19 b 3.30±0.18 b 0.006±0.001 b 0.14 118 13.2 0.4 000106-32-1 000110-38-3 000106-33-2 1453 1746 1887 Ethyl octanoate Ethyl decanoate Ethyl dodecanoate 1.13±0.30 a 1.26±0.25 a 1.21±0.45 a 565 6.3 0.2 0.053±0.002 b 0.059±0.002 b 0.083±0.005 b 26.5 0.3 0.01 0.95±0.39 a 0.84±0.23 c 0.89±0.05 c 475 4.2 0.15 0.002 0.2 5.9 000064-19-7 000142-62-1 000124-07-2 000334-48-5 000143-07-7 1549 1890 2170 2390 2544 Acetic acid Hexanoic acid Octanoic acid Decanoic acid Dodecanoic acid 452±49 a 1.39±0.15 a 4.64±0.76 a 1.57±0.27 a 0.65±0.05 a 2.3 0.46 0.53 0.16 0.065 680±14 b 0.32±0.02 b 0.24±0.03 b 0.35±0.01 b 0.48±0.02 b 3.4 0.11 0.03 0.04 0.048 647±105 c 2.17±0.56 c 3.87±0.59 c 1.45±0.03 a 0.66±0.04 a 3.2 0.72 0.44 0.15 0.066 200 3 8.8 10 10 Alcohol Ester Acid Compounds 1 0.027±0.019 a 0.017 0.51±0.05 c 17 a 0.51±0.10 2 0.25±0.08 a 17.9 CAS numbers were obtained from Wiley database library LRI of all the relative tables was determined on the DB-FFAP column, relative to C5-C40 hydrocarbons. 3 Descriptors were retrieved from http://www.thegoodscentscompany.com abc ANOVA (n=4)at 95% confidence level with the same letters in the same row indicating no significant difference 2 65 1.6 0.03 0.25 0.014 Woody, sweet, pine Rose, citrus, green Pineapple, fruity, varnish Fruity, sweet, apple Fruity, banana, sweet Floral, rose, sweet Sweet, pineapple, waxy Floral, fruity, brandy Waxy, sweet, apple Soapy, waxy, floral Vinegar, pungent Fatty, soapy, sour Fatty, soapy, sour Fatty, rancid, sour Coconut, fatty 4.2.3.1 Terpenes The major terpenes in mango wines were α-caryophyllene, β-caryophyllene, β-myrcene, β-ocimene and β-phellandrene (Table 4.3). Terpenes constituted the largest group of volatiles in fresh mango juice (Table 3.2) but most of the terpene hydrocarbons in fresh mango juice decreased substantially after fermentation. The decrease in terpenes was significantly more pronounced in the S. cerevisiae predominant fermentations. This is in contrast with some previous reports which claimed that fermentation would not affect the concentration of terpenes (Rapp 1998; Ong and Acree 1999; Alves et al. 2010). However, it has also been reported that some Saccharomyces strains would cause the decrease of terpenes (Zoecklein et al. 1997). This could be due to the inherent variation in the genome makeup for the different strains and should be further investigated for the production of wine with more varietal character, since studies have indicated that varietal character of mango is often distinguished by the presence of terpene hydrocarbons (Chauhan et al. 2010). This could be an area of W. mrakii fermentation that can be further investigated and utilised to produce wine where the original mango aroma is better preserved, because W. mrakii can better retain the characteristic mango aroma. Some new terpene hydrocarbons were formed. For instance, β-myrcene was not detected in fresh mango juice but detected after fermentation. This could be due to the metabolic activities of the yeasts. Similar to the general trend observed for terpene hydrocarbons, β- myrcene was found in higher amounts in the W. mrakii –fermented wine. The concentration of β- myrcene decreased throughout the S. cerevisiae fermentation with the largest decrease observed from Day 2 to Day 4, before the decrease became more gradual. On the other hand, β- myrcene increased from Day 2 to Day 4 for the W. mrakii and mixed culture fermentation before a significant decrease was obsesrved in mixed culture fermentation and a less drastic decrease was observed in W. mrakii fermentation. β- myrcene 66 steadily increased again from Day 10 in W. mrakii fermentation while β- myrcene decreased gradually in the mixed culture fermentation. The trend in evolution for the mixed culture fermentation could be largely due to the domination of S. cerevisiae after the initial domination of W. mrakii resulting in similar trends as the S. cereisiae monoculture fermentation and final content of β- myrcene. β-myrcene is known to be the product of linalool degradation, hence it is likely that the linalool produced during fermentation simultaneously undergoes degradation due to other metabolic reactions, resulting in the formation of β- myrcene (Förster-Fromme and Jendrossek 2010). β-Myrcene 6.0E+06 Peak area 5.0E+06 4.0E+06 3.0E+06 2.0E+06 1.0E+06 0.0E+00 0 5 10 Time (Days) 15 20 Figure.4.2 Changes in β-myrcene in the fermentation of mango juice with a monoculture of S. cerevisiae MERIT.ferm ( ), monoculture of W. mrakii NCYC 500 ( ) and a mixed culture of S. cerevisiae and W. mrakii ( ) 4.2.3.2 Alcohols Alcohols make up the largest group of volatiles (by RPA) in all three mango wines with ethanol being the most dominant alcohol formed with an RPA of 71.2% (v/v) for strain S. cerevisiae, 26.4% v/v for strains NCYC 500 and 70.2% (v/v) for mixed culture fermentation (Table 4.3,). The major alcohols in fresh mango juice were cis-3-hexanol, ethanol, hexanol and trans-2-hexenol (Table 3.1) while the major alcohols in mango wines were isobutyl alcohol, isoamyl alcohol and 2-phenylethyl alcohol (Table 4.3). The amounts of all higher 67 alcohols found in mango juice decreased after fermentation and significant amounts of new higher alcohols were formed. Generally, the S.cerevisiae-dominated fermentations produced higher amounts of alcohols relative to the W. mrakii-fermented wine. There were significant differences in the amounts of ethanol produced from the three different treatments. The monoculture of S.cerevisiae produced the most ethanol, followed by the mixed culture and lastly the monoculture of W.mrakii. Ethanol production peaked at Day 7 and 10 for S.cerevisiae and mixed culture fermentation respectively, before a gradual decline (Figure 4.3). The reason for this decline is unclear as there were no corresponding increases in the ethyl esters (Figure 4.5) which might have utilised ethanol as a precursor. Peak area Ethanol 4.0E+09 3.5E+09 3.0E+09 2.5E+09 2.0E+09 1.5E+09 1.0E+09 5.0E+08 0.0E+00 0 5 10 15 Time (Days) 20 Figure 4.3 Changes in ethanol content in the fermentation of mango juice with a monoculture of S. cerevisiae MERIT.ferm ( ), monoculture of W. mrakii NCYC 500 ( ) and a mixed culture of S. cerevisiae and W. mrakii ( ) The major fusel alcohols produced were isoamyl alcohol, isobutyl alcohol and 2phenylethyl alcohol. All three alcohols were found at levels higher than their odour thresholds (OAV>1) in the mango wine fermented by the S. cerevisiae yeasts and mixed culture (Table 4.2). The mixed culture fermentation produced slightly lesser amounts of isoamyl and isobutyl alcohols than the S. cerevisiae monoculture. 2-Phenylethyl alcohol was 68 found in the highest concentration in the mixed culture fermentation. W. mrakii fermentation generally produced the lowest amounts of fusel alcohols. Isoamyl and isobutyl alcohol exhibited similar changes during the course of fermentation (data for isoamyl alcohol not shown). Production increased drastically from Day 2 to 4 for S. cerevisiae fermentation and from Day 4 to 7 for mixed culture fermentation (Figure 4.4). After which, both alcohols generally decreased as the fermentation proceeded. On the other hand, W. mrakii, produced isoamyl and isobutyl alcohol a lot more slowly (Figure 4.4). However, the final concentration of isobutyl alcohol did not differ significantly between the mixed culture and W. mrakii fermentation due to the decrease in isobutyl alcohol during the later part of the mixed culture fermentation. The production of 2-phenylethyl alcohol was more gradual for all three treatments. Maximum concentration was attained at Day 10 and Day 14 for mixed culture and S. cerevisiae fermentation respectively while production increased steadily throughout the fermentation period for W. mrakii. Slight decreases were observed for both S. cerevisiae and mixed culture fermentations. Fusel alcohols have a positive effect on wine flavour at concentrations below 350 mg/L; typically, wines have been reported to contain 80–540 mg/L (Rapp and Mandery, 1986). In this study, the major higher alcohol content falls within the desirable range. The production of such moderate levels of alcohols may be beneficial to the flavour profile. S. cerevisiae yeast may be harnessed for its production of desirable higher alcohols. On the other hand, the W. mrakii yeast produced significantly lower amounts of these higher alcohols. It is likely that the individual higher alcohol was not able to contribute much to the flavour profile, but there may be a synergistic effect and the total sum of these alcohols may contribute positively to the overall flavour profile. 69 Isobutyl alcohol 2.5E+07 Peak area 2.0E+07 1.5E+07 1.0E+07 5.0E+06 0.0E+00 0 5 10 15 Time (Days) 20 2-Phenylethyl alcohol 6.0E+07 Peak area 5.0E+07 4.0E+07 3.0E+07 2.0E+07 1.0E+07 0.0E+00 0 5 10 15 Time (Days) 20 Figure 4.4 Changes in isobutyl alcohol and 2-phenylethyl alcohol content in the fermentation of mango juice with a monoculture of S. cerevisiae MERIT.ferm ( ), monoculture of W. mrakii NCYC 500 ( ) and a mixed culture of S. cerevisiae and W. mrakii ( ) Terpene alcohols or terpenols such as citronellol and linalool were also detected in the mango wine. These compounds may impart attributes such as “fruity” or “floral” flavour (Iwase et al. 1995). However, both terpenols were produced at levels lower than their odour thresholds for all three fermentations. Although these compounds may have synergistic effects with other higher alcohols, they may contribute little to the overall flavour profile of the mango wine individually. Increases in linalool were also after fermentation were observed and were inversely proportional to the myrcene content. As β-myrcene is the product of linalool degradation (Förster-Fromme and Jendrossek 2010), moderation of this mechanism might have a positive effect on the flavour profile of mango wine 70 4.2.3.3 Esters Esters made up the second largest group (over 25% RPA). Ethyl esters were the major esters detected in the S. cerevisiae-dominated fermented wines while acetate esters dominated the W. mrakii-fermented wine. The ethyl esters that were quantified include ethyl hexanoate, ethyl octanoate, ethyl decanoate, ethyl dodecanoate while the acetate esters that were quantified were ethyl acetate, isobutyl acetate, isoamyl acetate and 2-phenylethyl acetate. While only two ethyl esters of significance were detected in fresh mango juice, both the number and amount of ethyl esters increased after fermentation. The major ethyl esters present in the S. cerevisiae and mixed culture fermentations were ethyl octanoate, ethyl decanoate and ethyl dodecanoate. In terms of potential contribution to wine flavour profile, ethyl octanoate (described as floral, fruity, brandy) may have a more significant impact. Production of ethyl octanoate peaked at Day 4 and Day 7 for S.cerevisiae and mixed culture fermentation respectively (Figure 4.5). After which, ethyl octanoate decreased steadily for the S.cerevisiae fermentation while some a small increase was observed from Day 10 to 14 in the mixed culture fermentation. On the other hand, ethyl octanoate production was significantly lower in W.mrakii fermentation. A noticeable increase was observed from Day 10 to Day 14 before ethyl octanoate content remained relatively constant. Ethyl hexanoate and ethyl decanoate were also found at levels above their odour thresholds, while ethyl dodecanoate was only present at levels above detection thresholds in the mixed culture fermentation. With the exception of ethyl octanoate, all the other ethyl esters quantified were below their odour thresholds. 71 Ethyl octanoate 1.4E+08 Peak area 1.2E+08 1.0E+08 8.0E+07 6.0E+07 4.0E+07 2.0E+07 0.0E+00 0 5 10 15 Time (Days) 20 Figure 4.5 Changes in ethyl octanoate in the fermentation of mango juice with a monoculture of S. cerevisiae MERIT.ferm ( ), monoculture of W. mrakii NCYC 500 ( ) and a mixed culture of S. cerevisiae and W. mrakii ( ) At appropriate levels, these medium-chain esters add moderate notes of ripe fruits to fermented wine (Alves et al. 2010). In this study, ethyl hexanoate, ethyl octanoate and ethyl decanoate were all found at levels above their odour thresholds in the Saccharomycesfermented wine. In the mixed culture fermentation, ethyl dodecanoate was produced at high levels and exceeded its odour threshold with an OAV of 5.9. The presence of much higher ethyl esters content in the S. cerevisiae fermentation may imply that the efficiency of fatty acid metabolism is much lower in W. mrakii than S. cerevisiae. However, these esters may impart rancid and soapy flavours at high concentrations; as such, careful moderation of these esters may make a positive contribution to the wine aroma. In this respect, mixed culture fermentation resulted in a lower concentration of ethyl decanoate and ethyl dodecanoate. Hence, such a fermentation strategy may prove to be beneficial in moderating wine flavour profile. W. mrakii fermentation produced higher amounts of acetate esters. Acetate esters may have positive contributions to the overall quality of the wine and most impart moderate “floral” or “fruity” flavour notes. All the acetate esters quantified in this study were present 72 in amounts higher than their threshold values (Table 4.3), indicating the likelihood of their contribution to fruity, floral and sweet flavours to the final wine bouquet. The timepoint evolution of ethyl acetate and 2-phenylethyl acetate can be found in Figure 4.6. There was no significant differences between 2-phenylethyl acetate between W. mrakii and mixed culture fermentations after 21 days. 2-Phenylethyl acetate was initially produced at a much faster rate in the mixed culture fermentation and exceeded that in the W. mrakii fermentation until a significant decline occurred during Day 7. The reason for this decline would be due to the increased acetate ester hydrolysing activity of the S. cerevisiae being more dominant (Fukuda et al. 1998). Furthermore, the presence of rather significant amounts of citronellyl acetate and neryl acetate further demonstrated the diverse range of volatiles in the W. markii mango wine. However, one of the possible areas of concern is the high ethyl acetate levels in the NCYC 500-fermented wine. Ethyl acetate increased consistently in the W. mrakii fermentation (Figure 4.6) and is considered beneficial at levels of 150–200 mg/L (Jackson 2008) and possibly detrimental to wine flavour by introducing “nail varnish”-like flavours. Therefore, the high ethyl acetate content in the W. mrakii wine may prove detrimental to wine quality. Acetate esters are produced from the reaction of acetyl CoA with alcohols (Perestrelo et al. 2006). The significantly higher acetate ester contents in W. mrakii fermentations could be due to the higher stability of the alcohol-acetyl transferases (AATase) in W. mrakii (Inoue et al. 1997) relative to labile nature of the same enzymes in S. cerevisiae (Minetoki 1992). The high concentrations of acetate esters could enhance the activity of hydrolyzing esterase of S. cerevisiae (Kurita 2008). This could explain why acetate esters increased and then decreased in mixed-culture fermentation in similar studies. 73 S. cerevisiae consistently produced higher amounts of other higher branched-chain esters than W. mrakii. This could be due to the higher amounts of fusel alcohols present in the S. cerevisiae fermentation than in the W. mrakii fermentation. However, the significantly lower amounts of these esters in mixed culture fermentation was unexpected. A possible explanation would include the potential killer toxins secreted by the W. mrakii yeasts (Magliani et al. 2008) inhibiting the normal functioning of the cellular metabolism of the S. cerevisiae yeast. 2-Phenylethyl acetate 5.0E+07 Peak area 4.0E+07 3.0E+07 2.0E+07 1.0E+07 0.0E+00 0 5 10 15 Time (Days) 20 Peak area Ethyl acetate 1.8E+08 1.6E+08 1.4E+08 1.2E+08 1.0E+08 8.0E+07 6.0E+07 4.0E+07 2.0E+07 0.0E+00 0 5 10 15 Time (Days) 20 Figure 4.6 Changes in 2-phenylethyl acetate and ethyl acetate content in the fermentation of mango juice with a monoculture of S. cerevisiae MERIT.ferm ( ), monoculture of W. mrakii NCYC 500 ( ) and a mixed culture of S. cerevisiae and W. mrakii ( ) 74 4.2.3.4 Acids Acetic, octanoic, decanoic and dodecanoic acids were the major fatty acids detected in all three mango wines and quantified in this study. Acetic acid was the only acid detected at levels above its odour threshold. As discussed in the Chapter 3, acetic acid is produced when yeasts are exposed to hyperosmotic stress and may impart an unpleasant, vinegar off-odour when present at excessive levels. Similar trends in acetic acid production were observed Day 0 to Day 7 for both monoculture fermentations while the mixed culture produced significantly more acetic acid. From Day 10 onwards, acetic acid production for W. mrakii fermentation increased significantly until the end of the fermentation. On the other hand, acetic acid production for S. cerevisiae fermentation remained relatively constant while a decrease in acetic acid was observed from Day 14 until the end of the fermentation in the mixed culture fermentation. It is likely that S. cerevisiae is better adapted to such osmotic stress and therefore, produced lesser amounts of acetic acid relative to the W. mrakii yeast. However, it was interesting to note that acetic acid was present at significantly higher levels in the mixed culture–fermented wine compared to the S. cerevisiae fermentation despite S. cerevisiae dominating the fermentation. This could be due to the inhibitory effects that W. mrakii may have on S. cerevisiae. Previous studies have shown that W. mrakii produce killer toxins that S. cerevisiae are sensitive to. This will be discussed in greater detail in the next chapter. Although acetic acid concentration was higher than the threshold level in all three mango wines, it may not be detrimental to wine quality as a concentration between 0.02 to 0.07 g/100 mL was considered optimal depending on the style of wine (Lambrechts and Pretorius, 2000); the amount of acetic acid produced in this study ranged from 0.045 to 0.068 g/100 mL. 75 Acetic acid 1.2E+07 Peak area 1.0E+07 8.0E+06 6.0E+06 4.0E+06 2.0E+06 0.0E+00 0 5 10 Time (Days) 15 20 Figure 4.7 Changes in acetic acid content in the fermentation of mango juice with a monoculture of S. cerevisiae MERIT.ferm ( ), monoculture of W. mrakii NCYC 500 ( ) and a mixed culture of S. cerevisiae and W. mrakii ( ) The C6 to C10 fatty acids impart mild and pleasant aroma to wine at concentrations of 4 to 10 mg/L; however, at levels beyond 20 mg/L, their impact on wine becomes negative with fatty, rancid and soapy off-odours. Hence, they must be controlled at low levels or at least not higher than their threshold levels. These fatty acids may arise from the auto-oxidation of saturated lipids present as by-products of yeast fatty acid metabolism; and are extremely important to wine flavour not only for their own odour characters but also for their contribution to the synthesis of volatile esters (Saerens et al. 2008; Pino and Queris 2011). In this study, the amounts of these fatty acids produced were well below 20 mg/L and may therefore contribute to some complexity in the mango wine flavour profile. 4.2.3.5 Aldehydes and Ketones Acetaldehyde, benzaldehyde, o-tolualdehyde, and p-tolualdehyde were identified in all three mango wines. Acetaldehyde was the major aldehyde in the Saccharomyces dominated wines while o-tolualdehyde was the major aldehyde in NCYC 500-fermented wines (Table 4.2). Compared with other volatiles, aldehydes were only a minor group with less than 1% RPA. 76 Acetaldehyde is an important compound in terms of influence on wine quality. At low levels, acetaldehyde imparts a fresh, fruity note to wine. However, at excessive levels, it can impart an objectionable, green odour. Acetaldehyde was not quantified in this study due to its instability. However, by comparing RPA, it was present in the highest amount in the mixed culture fermentation, followed by the S.cerevisiae fermentation then W. mrakii fermentation. As mentioned in Chapter 3, acetaldehyde is very toxic to the yeast cells and has to be excreted, the data obtained in this chapter appears to suggest that W. mrakii is more prone to converting acetaldehyde to acetic acid when compared to S. cerevisiae which appears to secrete the acetaldehyde. Acetaldehyde and acetic acid are both undesirable at excessive levels, hence the effects of this discrepancy in acetaldehyde utilisation and secretion on wine flavour have to be further evaluated. 4.2.3.6 Miscellaneous compounds Nerol oxide and cis-linalool oxide were still detected after fermentation. However, trans-linalool oxide (which was present in fresh mango juice) was not detected in any of the mango wines. Generally, S. cerevisiae fermentation produced the least amounts of these compounds. Although cis-linalool oxide decreased after fermentation, it was found in significantly higher concentration in W. mrakii fermentation. On the other hand, nerol oxide was found in the highest concentration in the mixed culture fermentation, followed by the W. mrakii fermentation. Nerol oxide can be degraded to form nerol before being converted into neryl acetate. Although the S. cerevisiae fermentation produced the lowest amount of nerol oxide, it contained the highest amount of neryl acetate. This is interesting because it would be expected that W. mrakii with its high acetate synthesising activity would result in the W. mrakii producing the highest amounts of neryl acetate. 77 One plausible reason for this phenomenon could be due to the possible de novo production of nerol by S. cerevisiae yeast (Herrero et al. 2008). Methionol was identified in and it increased after the fermentation which was probably produced by yeasts through L-methionine metabolism (Seow et al. 2010; Tan et al. 2012). The amount of methionol produced was higher in the S. cerevisiae-dominated fermentations than the W. mrakii fermentation. 4.3 Conclusion The final flavour profile of the wine and consequently wine quality is very heavily influenced by the choice of starter cultures, which includes choice of yeast strains and/or monoculture or mixed culture. This chapter has illustrated the differences in the metabolism and modulation of volatile compounds by the different starter cultures. Although the volatile profile of the mixed culture fermentation is similar to that of the Saccharomyces monoculture, there were still significant differences. The possibility of utilising a mixed culture fermentation to improve the flavour profile of mango wine with similar ethanol content to a S. cerevisiae–fermented is worth investigating. In addition, it points to the possibility of harnessing the benefits of aroma enhancement by W. mrakii yeast and the ability for a complete fermentation by S. cerevisiae. The possibility of using sequential inoculation can be further explored to maximise the benefits of both strains and for the production of a mango wine with superior wine quality. 78 Chapter 5 Effects of Different Sequential Inoculation Strategies of Saccharomyces cerevisiae and Williopsis saturnus on Volatile Production 5.1 Introduction Mixed starter cultures have been demonstrated to enhance flavour complexities in wine production in some studies (Ciani et al. 2006; Moreira et al. 2008; Lee et al. 2010), while others have reported that non-Saccharomyces yeasts in mixed starter cultures haves limited influence due to their early growth arrest (Clemente-Jimenez et al. 2005; Farkas, et al. 2005; Bely, Stoeckle, et al. 2008; Varakumar et al. 2012). The early demise of the non- Saccharomyces yeasts could be attributed to their lower resistance to stresses under oenological conditions. Sequential fermentation has been suggested to be the most adequate strategy in strain combination; the kinetic behaviour is similar to a successful spontaneous fermentation with he wine produced different from that of a simultaneous fermentation (Clemente-Jimenez et al. 2005; Rodríguez et al. 2010). This is likely due to the prolonged action of non- Saccharomyces species in controlled sequential fermentation due to the lack of inhibition by the strongly fermentative Saccharomyces species. To date, there has been no published study on the effects of a controlled sequential fermentation on the quality of mango wine. This chapter investigated the effects of simultaneous mixed culture fermentation (MCF) where S. cerevisiae MERIT.ferm and W. mrakii NCYC 500 were co-inoculated against a positive sequential fermentation (PSF) and negative sequential fermentation (NSF). In PSF, W. mrakii NCYC 500 was first inoculated and fermentation was allowed to proceed for 14 days before killing the W. mrakii NCYC 500 yeast and inoculating the S. cerevisiae MERIT.ferm yeast and fermentation was conducted for further 7 days. For the NSF, the 79 sequence of yeast inoculation was reversed. S. cerevisiae MERIT.ferm was inoculated and fermentation carried out for 7 days before its forced demise and then, W. mrakii NCYC 500 was inoculated and fermentation conducted for further 14 days. 5.2 Results and discussion PSF was meant to stimulate the conditions of natural spontaneous fermentation under controlled conditions. The absence of strongly competitive S. cerevisiae MERIT.ferm yeast and other competitive microorganisms in the sterile mango juice was meant to maximise the action of W. mrakii NCYC 500 yeasts on the mango juice while the strongly fermentative S. cerevisiae MERIT.ferm yeast was inoculated later to produce the necessary ethanol content. On the other hand, NSF reversed the natural sequence of microbial growth. Previous studies have shown that some odour active volatile compounds increased briefly after fermentation, then decreased (Li et al. 2012). Reversing the order of propagation of yeasts would theoretically allow the important volatile compounds to be retained in the final mango wine with the high ethanol content associated with Saccharomyces fermented wines, imparting positive aroma attributes of Williopsis fermented wines (Ciani et al. 1999). 5.2.1 Physicochemical properties of mango wine There were no significant changes in the pH and most organic acids between the three mango wines and fresh mango juice (Table 5.1). The pH of all three mango wines produced ranged from 3.50 to 3.52 while only malic acid showed significant decreases as discussed in Chapter 4. 80 Table 5.1 Physicochemical properties, yeast cell count, organic acid and sugar concentrations of mango wines 1 Day 0 NSF2 Yeast strains PSF Physicochemical properties pH 3.50±0.01 a 3.50±0.01 a Brix (°B) 17.2±0.01a 17.3±0.01a 3.50±0.01 a 17.3±0.01a Plate count (105 CFU/mL) MERIT.ferm 6.67±1.20 a a NCYC 500 5.00±0.95 - 0.002 ±0.001b 6.67±1.10b Organic acid (g/100 mL) Citric acid 0.30±0.05 a Malic acid 0.90±0.08 a Succinic acid 0.083±0.018 a Tartaric acid 0.15±0.04 a Sugars (g/100 mL) Fructose 4.98±0.08 a Glucose 0.63±0.02 a Sucrose 12.25±0.31 a MCF PSF Day 21 NSF 3.51±0.02 a 3.52±0.01 a 5. 61±0.01 b 7.8±0.11 c MCF 3.50±0.01 a 5.58±0.01 b 3.50±0.64 c 7051 ± 993d c 53.4±7.85 N.D*. 0.29±0.06 a 0.92±0.09 a 0.088±0.011 a 0.16±0.04 a 0.29±0.06 a 0.92±0.09 a 0.088±0.011 a 0.16±0.04 a 0.27±0.03 a 0.28±0.03 a 0.36±0.05 b 0.41±0.05 c 0.086±0.011 a 0.081±0.012 0.14±0.02 a 0.14±0.02 a 0.27±0.03 a 0.36±0.05 b 0.086±0.011 a 0.14±0.02 a 5.03±0.09 a 0.62±0.06 a 12.56±0.36 a 4.98 ± 0.06 a 0.68 ± 0.02 a 12.47± 0.17 a N.D.* 2.46±0.15 b N.D.* N.D.* 0.22±0.06 b N.D.* 0.013±0.00 b 3.84±0.36 c 0.013±0.00 b *N.D.: not detected abcd ANOVA (n=4) confidence level at 95%, same letters in the same row indicate no significant difference 1 PSF W.mrakii was inoculated on Day 0, inactivated at Day 14 by ultrasonication. cerevisiae was then inoculated at D14. Fermentation was ceased at Day 21. 2 NSF S. cerevisiae on on Day 0, inactivated at Day 7 by ultrasonication. W.mrakii was then inoculated at D14. Fermentation was ceased at Day 21. 3 MCF S. cerevisiae and W. mrakii were both simultaneously inoculated on Day 0 in the ratio of 1:1000. Fermentation was ceased at Day 21. S. The evolution of oBrix values for NSF and MCF was almost identical for the first 7 days of the fermentation (Figure 5.1). This is likely due to the fact that S. cerevisiae was the dominant yeast species in MCF. Although W. mrakii was present in MCF, due to its low consumption of sugar, the main determining factor of sugar consumption was still due to Saccharomyces growth and cell population. The sudden increase in °Brix at Day 7 for NSF was due to the supplementation of glucose for the next phase of fermentation; glucose was added to ensure that there were sufficient nutrients for the subsequently inoculated NCYC 500. The trend observed for the decrease in oBrix values correlated with the trend for yeast 81 cell growth, especially the growth of S. cerevisiae (Figure 5.2). This is easily explained by TSS (°B) the fact that S. cerevisiae consumed sugars at a much faster rate than W. MRAKII. 20 18 16 14 12 10 8 6 4 2 0 Addition of glucose 0 5 10 Time (Days) 15 20 Figure 5.1 Changes in oBrix values during fermentation for PSF1 ( ), NSF2 ( ) and MCF3 ( ) 1 PSF W.mrakii was inoculated on Day 0, inactivated at Day 14 by ultrasonication. S. cerevisiae was then inoculated at D14. Fermentation was ceased at Day 21. 2 NSF S. cerevisiae on on Day 0, inactivated at Day 7 by ultrasonication. W.mrakii was then inoculated at D14. Fermentation was ceased at Day 21 3 MCF S. cerevisiae and W. mrakii were both simultaneously inoculated on Day 0 in the ratio of 1:1000. Fermentation was ceased at Day 21 5.2.2 Yeast biomass evolution The evolution of S. cerevisiae and W. mrakii for the three types of fermentations is shown in Figure 5.2. In PSF, W. mrakii reached a level of 108 CFU/mL in the sterile mango juice before being killed off via ultrasonication. The absence of the Saccharomyces yeast allowed proliferation of W. mrakii, giving W. mrakii longer persistence to produce esters. Interestingly, the growth kinetics of S. cerevisiae when inoculated after the demise of W. mrakii was different from that of both NSF and MCF. Despite its suitability for alcoholic 82 fermentation and the presence of sufficient nutrients in the juice, S. cerevisiae cell population declined almost immediately after inoculation. It is unlikely that the primary cause of the death of S. cerevisiae in sequential culture was ethanol because ethanol production and sugar consumption by W. mrakii were low. Nutrient depletion could be ruled out because similar studies demonstrated that S. cerevisiae with higher inoculum levels could still survive (Li 2013). Consequently, it is likely that the typically robust S. cerevisiae may have failed to thrive due to the presence of mycosins or killer toxins produced by W. mrakii (Liu and Tsao 2010), to which S. cerevisiae was sensitive (Yap et al., 2000). Previous studies have also shown that W. mrakii could produce mycotoxins that affects the synthesis of yeast cell walls by inhibiting the β-1,3-glucan synthesis occurring at a budding site or a conjugating tube, resulting in cell death (Yamamoto et al. 1986; Inoue et al. 1995; Kimura et al. 1999; Guyard et al. 2002). Therefore, despite the completion of fermentation, it is likely that W. mrakii might have inhibited the growth of S. cerevisiae due to the accumulation of mycosin. Therefore, it might be necesssary for a S. cerevisiae to be inoculated at a higher dosage. 83 (a) Log CFU/mL 9 8 7 6 5 4 3 2 1 0 0 5 10 Time (Days) 15 20 0 5 10 Time (Days) 15 20 0 5 10 Time (Days) 15 20 Log CFU/mL (b) 10 9 8 7 6 5 4 3 2 1 0 (c) 8 Log CFU/mL 7 6 5 4 3 2 1 0 Figure 5.2 Changes in (a) S. cerevisiae MERIT.ferm for PSF1 ( ), NSF2 ( ) and MCF3 ( ), (b) W. mrakii NCYC 500 cell population during fermentation for PSF1 ( ), NSF2 ( ) and MCF3 ( ), (c) Changes in S. cerevisiae MERIT.ferm ( ) and W. mrakii NCYC 500 ( ) in MCF 1 PSF W.mrakii was inoculated on Day 0, inactivated at Day 14 by ultrasonication. S. cerevisiae was then inoculated at D14. Fermentation was ceased at Day 21. 2 NSF S. cerevisiae on on Day 0, inactivated at Day 7 by ultrasonication. W.mrakii was then inoculated at D14. Fermentation was ceased at Day 21 3 MCF S. cerevisiae and W. mrakii were both simultaneously inoculated on Day 0 in the ratio of 1:1000. Fermentation was ceased at Day 21 84 On the other hand, it is likely that S. cerevisiae may inhibit W. mrakii due to high ethanol production. In addition, studies have indicated that W. mrakii produced and accumulated toxins in the stationary phase (Li 2013). Therefore, it is likely that W. mrakii was unable to kill S. cerevisiae in MCF because S. cerevisiae already reached a high cell density and produced sufficient ethanol to inhibit W. mrakii before the amount of toxin compounds produced by W. mrakii rose above its lethal concentration. This is supported by a study that showed that more than 6% (v/v) ethanol significantly inhibited the growth of W. mrakii (Li 2013). This result was consistent with another report on the behaviour of Williopsis in Japanese sake, demonstrating the inhibitory effect of 6 to 7% ethanol on Williopsis (Inoue et al. 1994). For NSF, S. cerevisiae reached a similar maximal population to those in the single culture fermentations, but NCYC 500 yeast inoculated after the forced demise of the S. cerevisiae yeast was unable to attain the maximal cell population obtained in the single culture fermentations in Chapter 4. Maximal W. mrakii population only reached 106 CFU/mL, as opposed to 108 CFU/mL in the single culture fermentation and PSF. This could be due to the toxic effects of ethanol as discussed previously. 85 5.2.3 Volatile composition A wide variety of volatile compounds could be produced from different inoculation strategies due to the metabolic interactions between the different yeast species (Ciani et al. 2010), and potentially the chemical reactions of the different metabolites and aroma compounds that are produced during the different stages of the fermentation. These volatile compounds include fatty acids, alcohols, aldehydes, esters, ketones, volatile phenols and terpenoids. Although similar compounds were detected in all three fermentations, the amounts of these volatile compounds differed rather significantly. Generally, PSF appeared to produce higher amounts of desirable volatile compounds while NSF produced lower amounts of most volatiles. This may be due to the effects of ethanol toxicity as the amount of ethanol produced in NSF appears to be significantly higher than PSF. The complete list of volatiles and their RPA can be found in Table 5.2 and the impact odourants quantified can be found in Table 5.3. 86 Table 5.2 Complete volatiles for mango wine produced from different inoculation strategies CASA LRIB Compound PSF1 MCF2 NSF3 Peak area (×106) RPA (%) Peak area (×106) RPA (%) 000064-17-5 Subtotal Alcohols 000071-23-8 000078-83-1 000123-51-3 000928-96-1 000505-10-2 000078-70-6 000562-74-3 000098-55-5 000470-08-6 000106-22-9 000106-25-2 000106-24-1 000060-12-8 00499-75-2 1028 Ethanol 1783 ± 137 a 78.219 2187 ± 63 b 73.05 Peak area (×106) 2775 ± 68 c 1039 Propanol 1783 1.24 ± 0.15 a 78.22 0.054 2187 1.37 ± 0.01 a 73.05 0.046 2775 2.79 ± 0.03 c 86.2 0.087 Alcoholic, fusel 0.644 15.5 ± 0.44 b 0.518 21.01 ± 1.08 c 0.653 Fruity, wine-like 43.29 ± 0.65 b 65.97 ± 0.17 c 2.049 Alcoholic, fruity, banana 0.38 ± 0.02 b 0.3 ± 0.01 c 0.009 Green, grassy 1.37 ± 0.01 b 0.82 ± 0.01 c 0.026 Cooked cabbage 1.43 ± 0.07 b 4.8 ± 0.08 c 0.149 Mild floral 0.11 ± 0.03 b 0.24 ± 0.05 c 0.008 Woody 1.23 ± 0.02 b 0.84 ± 0.01 c 0.026 Floral, lilac 0.15 ± 0.01 b 0.13 ± 0.01 a 0.004 Camphor, pine, woody 0.27 ± 0.01 b 0.008 Floral, rosy, citrus 0.89 ± 0.01 c 0.028 Ethereal, cognac, fruity 1.75 ± 0.06 c 0.054 Sweet, floral, rose-like 95.01 ± 4.43 c 2.951 Rose, honey, floral 0.023 0.21 ± 0.01 c 0.006 Spicy, woody, herbal 5.15 195 1172 Isobutyl alcohol 1237 Isoamyl alcohol 1475 cis-3-hexenol 1706 Methionol 1999 Linalool 1691 4-Terpineol 1716 α-Terpineol 1794 β-Fenchol 1867 β-Citronellol 1833 Nerol 1834 Geraniol 2035 2-Phenylethyl alcohol 2236 Carvacrol Subtotal Ethyl esters 14.68 ± 1.49 a 48.48 ± 0.58 a 0.25 ± 0.01 a 0.69 ± 0.02 a 3.16 ± 0.06 a 0.3 ± 0.01 a 0.66 ± 0.04 a 2.126 0.011 0.03 0.139 0.013 0.029 0.12 ± 0.009 a 0.11 ± 0.008 a 0.25 ± 0.01 1.99 ± 0.1 a 000105-54-4 000123-66-0 054653-25-7 064187-83-3 000106-30-9 1009 Ethyl acetate 1034 Ethyl butyrate 1297 Ethyl hexanoate 1392 Ethyl 5-hexenoate 1358 Ethyl cis-3-hexenoate 1369 Ethyl heptanoate 0.005 0.004 0.087 42.53 ± 0.71 a 0.29 ± 0.04 a 1.865 0.013 5.03 59.13 ± 0.15 a 0.41 ± 0.02 a 22.33 ± 1.07 a 0.16 ± 0.007 0.2 ± 0.005 a 0.046 0.048 0.003 0.041 0.005 - 0.59 ± 0.02 b 0.73 ± 0.01 b 87.52 ± 1.56 b 0.68 ± 0.05 b 0.02 0.024 2.922 RPA (%) 86.2 14.24 ± 0.28 b 0.476 1.92 ± 0.11 c 0.06 0.018 0.32 ± 0.01 b 0.011 N.D - 6.28 ± 0.22 b 2.51 ± 0.03 a 0.007 0.009 87 0.21 0.084 0.12 ± 0.002 b 0.16 ± 0.008 b 13.25 ± 0.36 0.05 ± 0 Alcoholic 6.06 2.593 a 0.013 154.34 0.979 N.D 1.445 N.D 0.011 a 114.74 000141-78-6 Odour descriptorC c b 0.003 0.06 ± 0.005 0.005 a 0.19 ± 0.01 c Ethereal, fruity, sweet Sweet, fruity 0.412 Sweet, pineapple, fruity 0.002 Fruity, pineapple 0.002 Fruity, sweet, apple 0.006 Sweet, Table 5.2 (Continued) 000106-32-1 1453 Ethyl octanoate 035194-38-8 1486 Ethyl 7-octenoate a 138.77 ± 2.14 b 0.006 0.36 ± 0.03 b 0.37 ± 0.04 b 246.23 ± 5.47 b 0.07 ± 0.003 000110-38-3 1746 Ethyl decanoate 72.97 ± 0.66 a 5.49 ± 4.13 a 19.66 ± 0.68 a 0.66 ± 0.05 a 0.64 ± 0.05 a 000628-97-7 2373 Ethyl hexadecanoate 0.42 ± 0.01 a 0.019 054546-22-4 2402 Ethyl 9-hexadecenoate 0.87 ± 0.03 a 0.038 000106-33-2 1887 Ethyl dodecanoate 000692-86-4 1986 Ethyl undecenoate 000124-06-1 2201 Ethyl tertradecanoate Subtotal 000110-19-0 1020 Isobutyl acetate 0.241 000142-92-7 1206 n-Hexyl acetate 003681-71-8 1324 cis-3-Hexenyl acetate 000112-06-1 1546 Heptyl acetate 000150-84-5 1659 Citronellyl acetate 000112-17-4 1726 Decyl acetate 000105-87-3 1790 Geranyl acetate 000103-45-7 1862 2-Phenylethyl acetate Subtotal 0.029 0.028 11.483 a 1.36 ± 0.06 5.49 ± 0.27 a 0.12 ± 0.011 1.05 ± 0.11 a 0.21 ± 0.01 a 0.05 ± 0.006 25.15 ± 2.73 a a a 002035-99-6 1762 Isoamyl octanoate 000112-32-3 1803 Octyl formate 030673-38-2 1859 Isobutyl decanoate N.D 2.34 ± 0.11 a 1.53 ± 0.21 a 1.11 ± 0.013 Cognac 36.77 ± 0.82 1.142 Sweet, fruity, pineapple 16.47 ± 0.47 c 0.512 - 0.84 ± 0.01 c 0.026 Sweet, waxy, fruity 0.71 ± 0.01 c 0.022 Waxy, fruity, creamy, cognac 0.5 ± 0.01 c 0.015 Sweet, waxy, creamy c 0.008 0.008 Waxy, balsamic - 0.43 ± 0.03 0.014 0.25 ± 0.01 0.8 ± 0.01 a 0.027 0.26 ± 0.01 c 1.233 0.017 0.023 17.09 130.17 N.D 4.15 ± 0.15 0.11 ± 0.009 0.002 2.64 ± 0.04 b 41.23 ± 1.15 b b Sweet, fruity, banana 0.128 1.52 ± 0.02 0.047 Sweet, fruity, banana 0.13 ± 0.01 c 0.004 Fruity, green 0.098 1.25 ± 0.08 c 0.038 Fresh, green, fruity - N.D 0.139 N.D 0.16 ± 0.02 0.088 N.D 31.1 ± 0.73 c b 0.319 N.D 1.03 ± 0.069 0.005 - b 0.033 - Fresh, green, rum - Floral, green rose, fruity 0.005 - c 0.969 1.862 34.24 0.16 ± 0.009 a 9.5 ± 0.13 0.002 c 0.003 creamy, 4.051 0.06 ± 0.01 c 1.377 55.76 fruity, 0.015 0.142 b 0.009 88 0.003 c a 0.005 0.011 0.09 ± 0.01 0.68 ± 0.04 2.93 ± 0.13 0.067 a Tropical, fruity a 0.241 0.103 0.009 a 0.52 ± 0.01 b - 0.29 ± 0.04 b 0.43 ± 0.01 2.535 0.012 36.94 ± 0.26 b 1.102 Sweet, fruity 2.106 4.25 ± 0.16 0.046 1.817 c b b 0.06 58.52 ± 0.66 c 8.222 b 0.51 ± 0.04 b 1.04 a 4.634 0.012 511.78 0.029 57.81 000624-13-5 1523 Propyl octanoate 63.07 ± 0.6 0.862 0.66 ± 0.04 a 23.71 ± 1.6 0.003 3.2 261.86 000123-92-2 1112 Isoamyl acetate Higher esters 0.14 ± 0.02 a 3.451 000123-29-5 1624 Ethyl nonanoate 067233-91-4 1795 Ethyl 9-decenoate Acetate esters 78.71 ± 1.53 a Waxy, soapy, fatty Floral, rose, lavender Sweet, honey, floral, rosy 1.063 0.13 ± 0.0081 b 2.77 ± 0.07 c 1.59 ± 0.04 a 0.29 ± 0.015 c 0.004 Coconut, cocoa, gin 0.087 Sweet, fruity, 0.049 Fruity, rose, orange 0.009 Oily, sweet, brandy, apricot, cognac Table 5.2 (Continued) 002306-91-4 006309-51-9 006290-37-5 005457-70-5 1973 Isoamyl decanoate 2180 Isoamyl dodecanoate 2261 2-Phenylethyl hexanoate 2419 2-Phenylethyl octanoate Subtotal Organic acids 0.09 ± 0.01 a 0.32 ± 0.01 a 0.36 ± 0.02 a 000064-19-7 000067-43-6 000142-62-1 000124-07-2 000373-49-9 000334-48-5 014436-32-9 000065-85-0 000143-07-7 1549 Acetic acid 1575 Isobutyric acid 1728 Butanoic acid 1890 Hexanoic acid 2170 Octanoic acid 2853 9-Hexadecenoic acid 2390 Decanoic acid 2474 9-Decenoic acid 2455 Benzoic acid 2544 Dodecanoic acid Subtotal 0.004 000123-35-3 000099-86-5 095327-98-3 027400-71-1 000100-42-5 000099-87-6 001195-32-0 000535-77-3 1206 δ-3-Carene 0.016 4.39 ± 0.2 a 0.6 ± 0.04 a 0.46 ± 0.03 a 0.49 ± 0.36 a 13.26 ± 1.25 a 0.27 ± 0.01 a 11.12 ± 1.07 a 0.46 ± 0.03 a 0.5 ± 0.01 a 1.04 ± 0.06 a 1235 α-Terpinene 0.32 ± 0.02 a 1290 β-Ocimene 1403 p-α-Dimethyl styrene 1339 p-Cymene 1529 p-Cymenene 1343 m-Cymene 0.026 b 0.55 ± 0.024 0.23 ± 0.01 0.72 ± 0.027 0.581 b 19.8 ± 0.36 0.012 0.29 ± 0.016 0.487 16.33 ± 0.06 b 2.36 ± 0.05 b 0.87 ± 0.02 b 1.92 ± 0.07 b 0.022 0.046 a N.D 0.23 ± 0.01 1.47 ± 0.16 a 9.19 ± 0.24 a 0.06 ± 0.001 0.019 0.17 ± 0.02 0.014 0.12 ± 0.01 b 0.005 0.22 ± 0.03 b 0.101 ± 0.02 a 0.18 ± 0.03 a 0.064 2.01 ± 0.02 b 0.403 4.25 ± 0.2 b 0.003 N.D 89 1.57 ± 0.03 c 0.049 Waxy, fruity, sweet, cognac 0.005 0.14 ± 0.01 b 0.004 Cognac, green, waxy 0.33 ± 0.01 a 0.01 Sweet, honey, floral - 0.46 ± 0.01 b 0.014 Sweet, waxy, green, cocoa 0.427 7.28 0.024 5.24 ± 0.04 c 0.163 Sharp, pungent, sour, vinegar 0.018 0.55 ± 0.02 b 0.017 Acidic 0.43 ± 0.01 a 0.013 Sharp, dairy-like, cheesy, 0.72 ± 0.45 b 0.022 Sour, fatty, sweat, cheesy 20.43 ± 0.57 c 0.635 Fatty, waxy, rancid 0.09 ± 0.003 0.003 Waxy c 0.236 Rancid, sour, fatty 0.008 b 0.024 0.661 a 0.225 0.01 b 0.545 7.59 ± 0.01 0.079 c 0.029 0.064 2.21 ± 0.1 0.069 Waxy, green, fatty, soapy 0.47 ± 0.06 a 0.015 Faint balsam, urine 1.63 ± 0.15 c 0.051 Fatty, coconut, bay oil 1.462 39.36 b 0.01 a 43.79 0.19 ± 0.01 b a b 0.021 - 0.43 ± 0.02 0.12 ± 0.01 0.72 ± 0.043 b 1.428 1246 β-Myrcene 12.81 0.193 0.02 0.065 - N.D 0.02 N.D b N.D 0.305 a 1254 Limonene 0.15 ± 0.009 0.014 32.59 013466-78-9 1.96 ± 0.12 b 0.052 6.93 000079-31-2 Terpenes 1.18 ± 0.16 a 0.006 1.224 N.D - Sweet citrus 0.006 0.29 ± 0.06 c 0.009 0.004 0.14 ± 0.03 b 0.004 Terpenic, herbaceous, woody, citrus Sharp, terpenic, lemon 0.007 0.19 ± 0.01 b 0.006 Citrius, terpenic, orange note 0.003 0.13 ± 0.02 b 0.004 Citrus, green, lime 0.006 N.D - Sweet, balsam, floral 4.92 ± 0.10 c 0.041 Citrus, terpenic, woody 0.142 7.64 ± 0.05 c 0.237 Spicy, balsamic, musty - N.D - Citrus, terpenic, woody 0.066 Table 5.2 (Continued) 000087-44-5 006753-98-6 017066-67-0 000079-92-5 0.13 ± 0.01 a 0.006 1725 α-caryophyllene 1829 β-Selinene N.D 0.12 ± 0.01 a 0.005 1117 Camphene 0.32 ± 0.03 a 0.75 ± 0.05 a 1695 β-Caryophyllene 000586-62-9 1352 α-Terpinolene Subtotal Aldehydes 14.34 000075-07-0 939 Acetaldehyde 000104-87-0 1773 p-Tolualdehyde 015764-16-6 1836 2, 4-Dimethyl benzaldehyde 000100-52-7 1637 Benzaldehyde Subtotal Ketone 000821-55-6 2-Nonanone 000513-86-0 1401 Acetoin Subtotal Others 1434 cis- Rose oxide trans-Rose oxide 001786-08-9 Subtotal Grand total 2.76 ± 0.06 a 1563 Nerol oxide 0.07 ± 0.009 b 0.002 Woody, clove note 0.92 ± 0.01 0.21 ± 0.03 b 0.031 0.007 N.D 0.15 ± 0.01 c 0.005 Woody Herbal 0.014 0.34 ± 0.01 0.012 9.3 ± 0.38 0.289 0.033 1.86 ± 0.02 0.062 0.64 ± 0.02 0.02 Camphoraceous, minty Citrus, lime, pine 10.93 0.365 23.47 0.121 10.16 ± 0.53 b 0.025 1.39 ± 0.03 b b 0.57 ± 0.04 a 0.49 ± 0.03 a 0.022 0.55 ± 0.01 1.03 ± 0.09 a 0.045 2.11 ± 0.05 b 0.213 0.94 ± 0.05 a 1.52 ± 0.13 a 0.17 ± 0.02 a 1.69 016409-43-1 0.007 0.628 4.85 023696-85-7 1938 β-Damascenone 0.2 ± 0.005 b 0.067 0.007 0.37 ± 0.01 a 0.75 ± 0.05 a 1.32 ± 0.02 0.25 ± 0.33 b 1.57 3.46 ± 0.07 c 0.108 Green apple, fresh 0.046 5.23 ± 0.03 c 0.162 Pungent, fresh, green 0.018 1.3 ± 0.05 c 0.04 Sweet, almond, cherry 0.07 1.07 ± 0.08 a 0.033 Almond, fruity, nutty 0.016 0.19 ± 0.01 b 1.23 ± 0.02 b 0.044 0.008 2.21 ± 0.1 c 0.16 ± 0.02 a 2.37 Sweet, floral, fruity, coconut 0.069 Fruity, sweet, cheese 0.005 Sweet, buttery, creamy 0.074 0.013 0.39 ± 0.02 c 0.012 Rose, geranium-like 0.006 0.11 ± 0.01 c 0.004 Floral, rose-like 0.51 ± 0.01 c 0.016 Floral, orange blossom sweet 0.041 0.06 1 100 3219.6 N.D. Not detected A CAS numbers were obtained from Wiley database library B LRI of all the relative tables was determined on the DB-FFAP column, relative to C5-C40 hydrocarbons. 90 0.343 N.D 0.052 0.38 ± 0.01 b 0.076 1.8 100 2994.94 0.339 b 0.027 0.033 0.729 0.473 11.06 N.D 0.074 0.61 ± 0.02 a 1.73 2280.46 0.041 14.21 cooling, 0.032 100 C Descriptors were retrieved from http://www.thegoodscentscompany.com ANOVA (n=4)at 95% confidence level with the same letters in the same row indicating no significant difference abc 1 PSF W.mrakii was inoculated on Day 0, inactivated at Day 14 by ultrasonication. S. cerevisiae was then inoculated at D14. Fermentation was ceased at Day 21. NSF S. cerevisiae on on Day 0, inactivated at Day 7 by ultrasonication. W.mrakii was then inoculated at D14. Fermentation was ceased at Day 21 3 MCF S. cerevisiae and W. mrakii were both simultaneously inoculated on Day 0 in the ratio of 1:1000. Fermentation was ceased at Day 21 2 91 Table 5.3 Major volatiles quantified for mango wines with different inoculation strategies CASA PSF1 LRIB MCF2 NSF3 mg/L OAV mg/L OAV mg/L Odour threshold OAV (mg/L) OrganolepticsC Alcohol 000078-83-1 1172 Isobutyl Alcohol 8.21 ± 0.61 b 0.21 10.1 ± 0.75 a 0.25 15.32 ± 1.22 a 0.38 40 Fruity, wine-like 000123-51-3 1237 Isoamyl Alcohol 150.28 ± 8.98 b 5.01 83.87 ± 5.01 a 2.8 142.09 ± 17.60 b 4.74 30 Banana-like, sweet, fruity 000078-70-6 1699 Linalool Trace - 0.012 ± 0.001 a 0.46 0.028 ± 0.001 b Esters 1.48 42.46 ± 4.07 9.76 34.51 ± 5.35 a 0.01 0.02 ± 0.01 a 0.86 ± 0.18 a 0.36 ± 0.12 a 1.09 ± 0.44 a a 000060-12-8 2035 2-Phenylethyl Alcohol 14.82 ± 1.42 000141-78-6 1009 Ethyl acetate 73.22 ± 7.04 b 000110-19-0 1020 000123-92-2 1112 000123-66-0 1297 000106-32-1 1453 Acid b Isobutyl acetate Isoamyl acetate Ethyl hexanoate Ethyl octanoate 0.02 ± 0.01 a 0.68 ± 0.14 b 0.61 ± 0.20 b 0.73 ± 0.30 b a 5.05 0.96 ± 0.27 2.28 43.43 366.27 a 1.12 0.025 4.25 44.51 ± 0.69 a 4.45 10 Floral, rose-like 4.6 26.51 ± 2.74 c 3.53 7.5 Ethereal, fruity, sweet, 0.01 ± 0.00 a 0.01 1.6 Sweet, fruity, tropical 0.09 ± 0.03 c 0.3 0.3 Sweet, fruity, banana 0.25 ± 0.17 a 17.58 0.014 Sweet, pineapple c 274 0.002 Sweet, fruity 4.81 0.84 ± 0.07 c 4.18 0.2 Sweet, fruity, apple 0.01 2.85 25.48 546.75 0.55 ± 0.24 Fresh floral, herbal, rosewood 000110-38-3 1746 Ethyl decanoate 1.01 ± 0.28 000103-45-7 1862 2-Phenylethyl acetate 0.77 ± 0.06 b 3.06 0.88 ± 0.07 a 3.52 0.61 ± 0.05 c 2.44 0.25 Sweet, honey, floral, rosy 000106-33-2 1887 Ethyl dodecanoate 0.91 ± 0.09 a 0.16 1.05 ± 0.10 a 0.18 0.95 ± 0.04 a 0.16 5.9 Sweet, waxy, soapy 000064-19-7 1549 Acetic Acid 765.91 ± 33.28 b 3.83 597.6 ± 33.77 a 2.99 777.95 ± 31.29 b Vinegar-like, pungent 000142-62-1 1890 000124-07-2 2170 000334-48-5 2390 000143-07-7 2544 Hexanoic acid Octanoic Acid Decanoic Acid Dodecanoic Acid 4.45 ± 0.04 b 5.4 ± 1.00 b b 4.43 ± 0.22 5.56 ± 0.11 b 1.48 0.61 0.44 0.56 3.13 ± 0.03 a 4.16 ± 0.77 a a 4.07 ± 0.20 5.44 ± 0.11 a 1.04 0.47 0.41 0.54 3.89 200 4.02 ± 0.29 c 1.34 3 6.69 ± 0.51 c 0.76 8.8 Fatty, rancid b 0.43 0.55 10 10 Rancid, sour, fatty Mild fatty, coconut 4.3 ± 0.46 5.53 ± 0.48 b Trace: Compound detected but below linear range. A CAS numbers were obtained from Wiley database library B LRI of all the relative tables was determined on the DB-FFAP column, relative to C5-C40 hydrocarbons. 92 Sour, fatty, cheese C Descriptors were retrieved from http://www.thegoodscentscompany.com ANOVA (n=4)at 95% confidence level with the same letters indicating no significant difference abc 1 PSF W.mrakii was inoculated on Day 0, inactivated at Day 14 by ultrasonication. S. cerevisiae was then inoculated at D14. Fermentation was ceased at Day 21. NSF S. cerevisiae on on Day 0, inactivated at Day 7 by ultrasonication. W.mrakii was then inoculated at D14. Fermentation was ceased at Day 21 3 MCF S. cerevisiae and W. mrakii were both simultaneously inoculated on Day 0 in the ratio of 1:1000. Fermentation was ceased at Day 21 2 93 5.2.3.1 Alcohols The main product of alcoholic fermentation is ethanol. By RPA, PSF produced the least amount of ethanol while NSF produced the most ethanol. Generally, NSF produced the highest amounts of other alcohols while PSF produced the lowest amounts of other alcohols. This may have an effect on wine quality since excessive amounts of alcohols could mask the other volatile aromatic compounds (Swiegers et al. 2005). Ethanol production rate was significantly higher in NSF than in MCF despite both fermentations were dominated by the high ethanol producing S. cerevisiae (Figure 5.3). As discussed previously, this is likely due to the toxic effect of the mycosins secreted by W. mrakii yeast affecting the metabolism of the S. cerevisiae. It is likely that while mycosin accumulation was not rapid enough to kill the S. cerevisiae yeast, mycosins affected S. cerevisiae metabolism, resulting in the slower growth rate and production of ethanol. In MCF, the ethanol content remained unchanged after Day 11, indicating the end of alcoholic fermentation. This is in general agreement with the trend observed for °Brix values (Figure 5.1), which indicated no decrease in TSS after Day 7. There was a slight decline in the amount of ethanol after the inoculation of W. mrakii at Day 7 for NSF (Figure 5.3). This decline could be due to the effects of ultrasonication as the vibrations could have caused the loss of volatile compounds via evaporation. It is unlikely that the ethanol was converted into ethyl acetate by W. mrakii as the amount of ethyl acetate did not increase after Day 7 (Figure 5.6). Although there were residual sugars at the end of the fermentation, this is likely due to the addition of glucose just before the inoculation W. mrakii, which consumed sugars very slowly, hence resulting in the residual sugars. For PSF, rate of ethanol production was very slow for the W. mrakii yeast for the first 14 days of fermentation (Figure 5.3). This is expected as W. mrakii is not a vigorous 94 fermenter. Ethanol production increased rapidly after Day 14, upon the inoculation of the high ethanol producing S. cerevisiae. At the point of cessation, there were only trace amounts of sucrose detected (Table 5.1), indicating a complete fermentation. Ethanol 4.E+09 4.E+09 Peak Area 3.E+09 3.E+09 2.E+09 2.E+09 1.E+09 5.E+08 0.E+00 0 5 10 Time (Days) 15 20 Figure 5.3 Changes in ethanol concentration during PSF1 ( ), NSF2( ), MCF3 ( ) 1 PSF W.mrakii was inoculated on Day 0, inactivated at Day 14 by ultrasonication. S. cerevisiae was then inoculated at D14. Fermentation was ceased at Day 21. 2 NSF S. cerevisiae on on Day 0, inactivated at Day 7 by ultrasonication. W.mrakii was then inoculated at D14. Fermentation was ceased at Day 21 3 MCF S. cerevisiae and W. mrakii were both simultaneously inoculated on Day 0 in the ratio of 1:1000. Fermentation was ceased at Day 21 After ethanol, the major alcohols found in all three mango wines were isoamyl alcohol, 2-phenylethyl alcohol and isobutyl alcohol (Figure 5.4). NSF produced the most fusel alcohol at the fastest rate for the first seven days of the fermentation. MCF also produced significant amounts of fusel alcohols, but at slower rate and lesser amounts than NSF. PSF produced the lowest amounts of fusel alcohols consistently. The trend observed for isoamyl alcohol, 2-phenylethyl alcohol and isobutyl alcohol for NSF was slightly different. Isoamyl alcohol increased gradually throughout the fermentation. 95 Isobutyl alcohol increased rapidly from Day 0 to Day 3 before the rate of increase became more gradual. On the other hand, 2-phenylethyl alcohol showed a modest increase from Day 0 to Day 3 before a drastic increase from Day 3 to Day 7. The initial increase is expected as S. cerevisiae are known to be efficient fusel alcohol producers. After the demise of S. cerevisiae at Day 7, the rate of increase in the fusel alcohols generally decreased. W.mrakii generally do not produce much fusel alcohol. this is also observed in the initial stages of PSF where only W. mrakii was present. Isobutyl alcohol was somewhat different from both isoamyl alcohol and 2-phenylethyl alcohol. Isoamyl alcohol and 2-phenylethyl alcohol only showed substantial increases after the demise of W. mrakii and inoculation of S. cerevisiae. However, isobutyl alcohol showed a steady increase from the beginning of the fermentation. At the end of fermentation, the amount of isobutyl alcohol was similar in NSF and MCF. This was not a trend observed for either isoamyl alcohol or 2-phenylethyl alcohol. These higher alcohols are known for the floral and fruity notes that they impart and their positive contribution to wine flavour if present at desirable levels (below 350 mg/L) (Swiegers et al. 2005). Isoamyl alcohol and 2-phenylethyl alcohol were both detected at levels above their threshold values (Table 5.3) and would likely contribute fruity and floral notes to the overall mango wine flavour profile. On the other hand, isobutyl alcohol was produced at levels below its odour threshold value; hence its influence may be negligible. 96 Isoamyl alcohol Peak Area 1.E+08 8.E+07 6.E+07 4.E+07 2.E+07 0.E+00 0 5 10 15 Time (Days) 20 Peak Area 2-Phenylethyl alcohol 1.E+08 1.E+08 1.E+08 8.E+07 6.E+07 4.E+07 2.E+07 0.E+00 0 5 10 15 Time (Days) 20 Isobutyl alcohol 3.E+07 Peak Area 3.E+07 2.E+07 2.E+07 1.E+07 5.E+06 0.E+00 0 5 10 15 Time (Days) 20 Figure 5.4 Changes in isoamyl alcohol, 2-phenylethyl alcohol, isobutyl alcohol during PSF1 ( ), NSF2( ), MCF3 ( ) 1 PSF W.mrakii was inoculated on Day 0, inactivated at Day 14 by ultrasonication. S. cerevisiae was then inoculated at D14. Fermentation was ceased at Day 21. 2 NSF S. cerevisiae on on Day 0, inactivated at Day 7 by ultrasonication. W.mrakii was then inoculated at D14. Fermentation was ceased at Day 21 3 MCF S. cerevisiae and W. mrakii were both simultaneously inoculated on Day 0 in the ratio of 1:1000. Fermentation was ceased at Day 21 97 Some of the terpene alcohols, or terpenols, detected were gerianol, α-terpineol, nerol, 4terpineol and citronellol (Table 5.3). These terpenols have been identified to be important compounds in the aroma of wine from aromatic grape varieties including Muscat, Riesling, and other aromatic grape varieties (Marais 1983; Swiegers et al. 2005). Amongst these terpenols, linalool was the only one produced at levels above its threshold odour in NSF (Table 5.2); the delayed inoculation of W. mrakii appeared to have a positive effect on the production of terpene alcohols. For all three fermentations, linalool increased from Day 0 to Day 3 before displaying different trends (Figure 5.5). In MCF, linalool decreased until Day 11 before showing a slight increase from Day 11 to Day 21 (Figure 5.5). On the other hand, for PSF, linalool content did not change significantly after Day 3 (Figure 5.5). In NSF, linalool decreased from Day 3 to Day 7 before increasing from Day 7 to Day 21 (Figure 5.5). This appears to imply that W. mrakii has the ability to either release bound linalool glycosides or transform other terpenols such as geraniol and/or nerol into linalool (King and Dickinson 2000). It is unlikely that the increase in linalool was due to de novo synthesis by W. mrakii since the strain has not been reported to possess such a capability. 98 Peak Area Linalool 7.E+06 6.E+06 5.E+06 4.E+06 3.E+06 2.E+06 1.E+06 0.E+00 0 5 10 Time (Days) 15 20 Figure 5.5 Changes in linalool during PSF1 ( ), NSF2( ), MCF3 ( ) 1 PSF W.mrakii was inoculated on Day 0, inactivated at Day 14 by ultrasonication. S. cerevisiae was then inoculated at D14. Fermentation was ceased at Day 21. 2 NSF S. cerevisiae on on Day 0, inactivated at Day 7 by ultrasonication. W.mrakii was then inoculated at D14. Fermentation was ceased at Day 21 3 MCF S. cerevisiae and W. mrakii were both simultaneously inoculated on Day 0 in the ratio of 1:1000. Fermentation was ceased at Day 21 Although the rest of the terpenols were at low levels, it is probable that they may interact in a synergistic manner to contribute positively to the overall flavour profile due to the fruity and floral notes associated with these aromatic compounds. The low levels of these terpenols could also be due to their degradation during the fermentation period (King and Dickinson 2000), such as the trend observed for MCF in this case. Furthermore, the differences between the concentrations of the various alcohols could be due to the different metabolic activities of the yeasts. The amounts of geraniol and nerol were also significantly different amongst the three treatments and this is likely due to geraniol being converted into citronellol and other terpenols while nerol can be converted into α-terpineol, terpinolene and limonene, de novo by S. cerevisiae through an alternative pathway involving leucine (Carrau et al. 2005). 99 S. cerevisiae do not efficiently excrete monoterpenes but synthesize the phosphorylated form of geraniol, geranyl diphosphate (GDP), as intermediate of farnesyl diphosphate synthesis, a key molecule in the isoprenoid pathway that leads to the synthesis of dolicols, ubiquionones, and sterols. It has also been shown that yeasts use terpenes as biosynthetic intermediates for sterol synthesis. Therefore, differences in the terpene profile of wines probably depend on beta-glucosidase activity, terpene bioconversion rate and the percentage of terpenes accumulated by yeasts (Sadoudi, et al. 2012). 5.2.3.2 Esters Esters are important contributors to the fruity flavours of alcoholic beverages (Russell, 2003). The major esters produced in all three fermentations were mainly ethyl octanoate, ethyl decanoate, ethyl dodecanoate, 2-phenylethyl acetate and isoamyl acetate (Table 5.2). The ethyl esters were generally found in the highest concentration in MCF and could be attributed to efficient fatty acid metabolism of S. cerevisiae and the longer exposure time for S. cerevisiae during fermentation as opposed to the shorter time for both NSF and PSF. Amongst these esters, ethyl octanoate could have the most influence on the wine profile due to its very low odour threshold value and the significant amounts produced. Ethyl acetate and ethyl decanoate could also have significant influences on the overall flavour profile as their OAVs all exceeded one ( 100 Table 5.3). The time-course formation of three ethyl esters is presented in Figure 5.6. In PSF, ethyl octanoate and ethyl decanoate both increased from Day 15 to the end of the fermentation with minimal changes during the first 14 days (Figure 5.6). This is likely due to the fatty acid metabolism of the S. cerevisiae yeast inoculated at Day 14. The trend indicated that W. mrakii is a low ethyl ester producer. In MCF, ethyl octanoate and ethyl decanoate increased slowly from Day 0 to Day 3 before showing significant increases from Day 3 onwards (Figure 5.6). This can be explained by the dominant yeast strain present. W. mrakii was the dominant strain in the beginning of the fermentation, hence, resulting in the lacklustre production of ethyl esters. However, as the S. cerevisiae population increased and dominated the fermentation, S. cerevisiae yeast became prominent. Low amounts of all three ethyl esters were produced in NSF throughout the fermentation (Figure 5.6). Although it was hypothesised that the higher S. cerevisiae population in NSF would give rise to higher concentrations of ethyl esters as compared to MCF (Rojas et al. 2003), this was not observed in this study; most esters were found in lower concentrations in NSF (Table 5.2). The likely cause could be the demise of S. cerevisiae at Day 7. The lack of time for S. cerevisiae metabolism could be the likely reason for the low amounts of ethyl ester formation. Furthermore, the subsequently inoculated W. mrakii could possibly have caused the hydrolysis of ethyl esters, as illustrated by the decrease in all three ethyl esters after Day 7. 101 Ethyl octanoate Peak Area 2.E+08 1.E+08 1.E+08 1.E+08 8.E+07 6.E+07 4.E+07 2.E+07 0.E+00 0 5 10 15 Time (Days) 20 Peak Area Ethyl decanoate 4.E+08 4.E+08 3.E+08 3.E+08 2.E+08 2.E+08 1.E+08 5.E+07 0.E+00 0 5 10 15 Time (Days) 20 Peak Area Ethyl acetate 2.E+08 1.E+08 1.E+08 1.E+08 8.E+07 6.E+07 4.E+07 2.E+07 0.E+00 0 5 10 15 Time (Days) 20 25 Figure 5.6 Changes in (a) ethyl octanoate, (b) ethyl decanoate, (c) ethyl acetate during PSF1 ( ), NSF2( ), MCF3 ( ) 1 PSF W.mrakii was inoculated on Day 0, inactivated at Day 14 by ultrasonication. S. cerevisiae was then inoculated at D14. Fermentation was ceased at Day 21. 2 NSF S. cerevisiae on on Day 0, inactivated at Day 7 by ultrasonication. W.mrakii was then inoculated at D14. Fermentation was ceased at Day 21 3 MCF S. cerevisiae and W. mrakii were both simultaneously inoculated on Day 0 in the ratio of 1:1000. Fermentation was ceased at Day 21 102 The major acetate esters produced during fermentation were 2-phenylethyl acetate and isoamyl acetate; these compounds were found in the lowest amounts in NSF, with PSF and MCF displaying mixed trends as illustrated by 2-phenylethyl acetate and isomayl acetate in Figure 5.7. 2-Phenylethyl acetate 7.E+07 Peak Area 6.E+07 5.E+07 4.E+07 3.E+07 2.E+07 1.E+07 0.E+00 0 5 10 15 Time (Days) 20 Peak Area Isoamyl acetate 2.E+08 1.E+08 1.E+08 1.E+08 8.E+07 6.E+07 4.E+07 2.E+07 0.E+00 0 5 10 15 Time (Days) 20 Figure 5.7 Changes in 2-phenylethyl acetate and isoamyl acetate during during PSF1 ( ), NSF2( ), MCF3 ( ) 1 PSF W.mrakii was inoculated on Day 0, inactivated at Day 14 by ultrasonication. S. cerevisiae was then inoculated at D14. Fermentation was ceased at Day 21. 2 NSF S. cerevisiae on on Day 0, inactivated at Day 7 by ultrasonication. W.mrakii was then inoculated at D14. Fermentation was ceased at Day 21 3 MCF S. cerevisiae and W. mrakii were both simultaneously inoculated on Day 0 in the ratio of 1:1000. Fermentation was ceased at Day 21 W. mrakii was dominant in the initial stages of MCF before the more rigorous S. cerevisiae became dominant. The rapid increase of 2-phenylethyl acetate and isoamyl acetate 103 in MCF from Day 0 to Day 7 was due to the acetate ester synthesising activity of W. mrakii. This increase in 2-phenylethyl acetate is more significant in MCF than PSF due to the production of 2-phenylethyl alcohol by S. cerevisiae concurrently. On the other hand, in the absence of W. mrakii in NSF, despite the high amounts of 2-phenylethyl alcohol and isoamyl alcohol present (Figure 5.4), significantly lower amounts of the corresponding acetate esters were produced, likely due to the low acetate ester synthesising activities of S. cerevisiae. The decline in both acetate esters after Day 7 in MCF could be due to esterase activities in S. cerevisiae which dominated the fermentation. In PSF, there were no significant changes in both esters from Day 7 to Day 14. A sharp decline was observed in isoamyl acetate after the inoculation of S. cerevisiae while the decline in 2-phenylethyl acetate was not significant. The discrepancy in the acetate hydrolysing enzymatic activity of S. cerevisiae could be ascribed to the balance between the degradation and synthesis of esters governed by esterase and alcohol acetyltransferase (Fukuda et al. 1998). 5.2.3.3 Acids The main organic acids identified were acetic, octanoic and decanoic acids (Table 5.2). The production of acetic acid was lower in MCF while no significant differences were found between PSF and NSF. It is likely that the fatty acids are derived from yeast anabolic pathways or β-oxidation of higher fatty acids (Tehlivets et al. 2007). Only acetic and hexanoic acids were detected in the mango wine at levels above their odour threshold limits (Table 5.2). Acetic acid increased for the first three days for all three fermentations (Figure 5.8). After which, NSF and MCF both showed significant decreases from Day 3 to Day 7. On the other hand, acetic acid continued increasing steadily until Day 11 when a slight decrease occurred from Day 11 to Day 14. From Day 14 to Day 21, a drastic increase in acetic acid occurred. Although W. mrakii has been reported to produce higher amounts of acetic acid, 104 this was not observed in this study. PSF, in which W. mrakii was present only in the first 14 days, did not show significantly more acetic acid production. The amount of acetic acid also decreased upon the inoculation of S. cerevisiae (Figure 5.8). On the other hand, MCF produced the most acetic acid during the first 14 days of fermentation. This is likely due to the stress that the mycosin produced by the W. mrakii on the S. cerevisiae since acetic acid is a sign of stress and redox imbalance as presented in Chapter 3. However, the amount of acetic acid decreased drastically from Day 15 to Day 21. The reason for this sharp decline is not clear, and further work should be conducted to verify this trend. For NSF, there was a rapid increase in acetic acid from Day 0 to 3 before a significant decrease occurred (Figure 5.8). This could likely be due to the intial adaptation period required for S. cerevisiae; the relatively high sugar concentration resulting in an increase in acetic acid production (Devantier et al. 2005). However, it could be seen that W. mrakii once inoculated, produced large amounts of acetic acid (Day 7 to 11). This is consistent with data presented previously in Chapter 4 stating the W. mrakii produced higher amounts of acetic acid relative to S. cerevisiae. Hexanoic acid showed a different trend from acetic acid (Figure 5.8). Large increases were observed from Day 0 to Day 5 for NSF and MCF. This is likely due to the action of S. cerevisiae; the strain has efficient fatty acid metabolism and produced large amounts of hexanoic acid rapidly once inoculated. A sharp decrease in hexanoic acid content was observed after Day 7 after the inoculation of W. mrakii. In MCF, the amount of hexanoic acid increased in a more gradual manner from Day 7 to Day 11 before decreasing gradually from Day 11 to Day 15 and then decreasing drastically from Day 15 to Day 21. For PSF, there were very little changes in the hexanoic acid content, even after the inoculation of S. 105 cerevisiae at Day 14. Again, the lack of expected metabolic activity from the S. cerevisiae is likely due to the effects of the mycosin produced by the W. mrakii yeasts. Peak Area Acetic acid 9.E+06 8.E+06 7.E+06 6.E+06 5.E+06 4.E+06 3.E+06 2.E+06 1.E+06 0.E+00 0 5 10 15 Time (Days) 20 Peak Area Hexanoic acid 5.E+07 4.E+07 4.E+07 3.E+07 3.E+07 2.E+07 2.E+07 1.E+07 5.E+06 0.E+00 0 5 10 15 Time (Days) 20 Figure 5.8 Changes in acetic acid and hexanoic acid during PSF1 ( ), NSF2( ), MCF3 ( ) 1 PSF W.mrakii was inoculated on Day 0, inactivated at Day 14 by ultrasonication. S. cerevisiae was then inoculated at D14. Fermentation was ceased at Day 21. 2 NSF S. cerevisiae on on Day 0, inactivated at Day 7 by ultrasonication. W.mrakii was then inoculated at D14. Fermentation was ceased at Day 21 3 MCF S. cerevisiae and W. mrakii were both simultaneously inoculated on Day 0 in the ratio of 1:1000. Fermentation was ceased at Day 21 106 5.2.3.4 Terpenes Similar to data presented in previous chapters, most of the terpenes found in fresh mango juice were metabolized to trace levels (Table 5.2). Of interest is the significant increase in the concentration of para-α-dimethyl styrene (Table 5.3). This volatile compound is found naturally in bay leave oil and imparts nutty nuances and hence may provide an interesting flavour note to the mango wine profile. Camphene was also found in significant amounts for NSF despite being found in negligible amounts in PSF and MCF (Table 5.2). The difference in the amounts of terpenes detected after the three different fermentations may prove to be useful in distinguishing between the wines, since terpene profile has been shown to play an important role in differentiating wine varietal character. However, due to their low levels, it is more likely their impact, if any, on the overall aroma flavor profile would be a result of their synergistic effects. 5.2.3.5 Carbonyl compounds Acetaldehyde was produced in significantly higher amounts in MCF relative to NSF and PSF, which had the lowest concentration (Table 5.2). As mentioned previously, acetaldehyde can impart a ‘fresh’ note at low levels but becomes objectionable at higher concentrations. Hence, PSF may prove to be a fermentation strategy to modulate the amount of acetaldehyde produced to suit the desired wine style. Ketones make up a very small proportion of the volatiles identified (Table 5.2). 2Nonanone was the major ketone detected, with NSF producing the most 2-nonanone, followed by PSF and MCF. Varying amounts of acetoin were also detected in the three mango wines, with MCF produced the highest amount, followed by NSF and PSF. Acetoin is responsible for the creamy, buttery note and is considered to be beneficial to wine flavour profile at desired levels. The low level of acetoin is likely due to the lack of malolactic fermentation in this study. However, with the low amounts of these compounds detected, it is 107 unlikely that carbonyl compounds could have much potential impact on the overall wine flavor. 5.2.4 Sensory evaluation The sensory profile of the mango wines is presented in Figure 5.9 and scores are compared by one-way ANOVA. In terms of ranking, NSF wine was found to be more green, winey, sweet and yeasty; PSF wine was considered to be more fruity, waxy and creamy; and MCF wine was deemed to be more terpenic. green 4.0 creamy 3.5 fruity 3.0 2.5 2.0 1.5 acidic waxy 1.0 0.5 0.0 yeasty terpenic sweet rosy winey Figure 5.9 Sensory profile of PSF ( ), MCF ( ) and NSF ( ), mango wine (refer to legend of Fig 1 for definition of PSF, MCF and NSF) For instance, NSF wine contained higher amounts of acetic acid and was deemed to be more slightly more acidic. This is in agreement with most literature stating that acetic acid is 108 largely responsible for the volatile acidity in wine. In addition, NSF did produce significantly more hexanoic and octanoic acids as well (Table 5.3). However, it is interesting to note that NSF was also deemed to be sweet and winey; this could be due to the synergistic effects of the terpenols and terpene hydrocarbons that were retained in higher amounts in NSF. In PSF wine, the significantly larger amounts of ethyl and acetate esters could account for its higher rating for the ‘fruity’ attribute. It was also deemed to be creamy and this could be due to the higher amount of acetoin. However, it was considered waxy, which is generally deemed to be a negative attribute and could be caused by the longer-chain ethyl esters. In addition, the higher rating scored for ‘winey’ in PSF could be due to the higher ethanol content as well. The MCF wine was deemed to be more terpenic (least green). Although the amounts of terpene hydrocarbons were not excessively high in the MCF mango wine, it is likely that the overall synergistic effects of the various terpene hydrocarbons resulted in the terpenic notes. 5.3 Conclusion Different inoculation strategies resulted in different flavour profiles of mango wines. Although most volatile compounds produced were similar, the amounts produced by each inoculation strategy differed. Generally, PSF produced more esters than NSF and MCF. This could be due to the effects of ethanol toxicity that the high ethanol content had on the Williopsis yeast, resulting in the significantly lower concentrations of volatiles produced by NSF. However, there were also certain volatile compounds that were not detected in some of the mango wine. This is likely due to the effects of sequential inoculation on the metabolic process of the yeast, resulting in different biochemical reactions and metabolism. 109 Chapter 6 General conclusions and recommendations This research project provided an insight into how the various fermentation strategies could be employed to optimise the production of mango wine from fresh mango juice. This thesis provides an exposition of techniques and analysis of chemical composition (especially the volatile composition) of mango wine. Different species of yeasts, S. cerevisiae and W. mrakii, were tested for optimising the flavour profile. Several inoculation strategies were also studied for their effects on volatile production and sensory profile. S. cerevisiae was the major yeast in mango wine fermentation and it was not only responsible for converting the sugars into ethanol, but also for producing a wide range of volatiles. S. cerevisiae was efficient to produce alcohols, ethyl esters, and medium-chain fatty acids. It also led to significant decomposition of terpenes which were the signature mango varietal aroma compounds. W. mrakii, on the other hand, was not effective in converting sugars into alcohols; however, it was a prolific producer of acetate esters and branchedchained esters. Some of these compounds include ethyl acetate, isobutyl acetate, isoamyl acetate and 2-phenylethyl acetate. In addition, W. mrakii retained more terpenes in mango wine. However, W. mrakii should be well controlled in case of overproduction of ethyl acetate and acetic acid. This study not only provided details on vinification techniques with using S. cerevisiae and W. mrakii but also on chemical composition of mango wine fermented by both yeast species. In the wine industry, screening and selection of suitable strains is critical to make attractive mango wine. The correct strain should be confirmed before scaleup in the industry. Initial sugar concentration also had a significant effect on volatile and glycerol production. High sugar concentration generally inhibits volatile production due to the toxic 110 effects of ethanol on metabolic activities. However, excessive levels of acetaldehyde and acetic acid can be produced by S. cerevisiae under hyperosmotic conditions which may have a detrimental effect on wine flavour. On the other hand, glycerol production is directly correlated to initial sugar concentrations. Glycerol has a positive effect on the mouthfeel of the wine, and may impact consumers’ perception of wine quality. However, with the data gathered from this study, the use of high sugar concentration to enhance glycerol production may not be an appropriate technique due to the effects on volatile production. Hence, more studies may be required to enhance glycerol production in S. cerevisiae MERIT.ferm. Mixed culture fermentation between S. cerevisiae and W. mrakii produced a wider range of volatile compounds and more balanced aroma. It also provided an opportunity for winemakers to modulate volatile profile by varying inoculation strategy. In addition, an antagonistic effect between S. cerevisiae and W. mrakii was observed by this study. The antagonism could be attributed to ethanol and mycosin, with ethanol produced by S. cerevisiae inhibiting W. mrakii, and mycosin secreted by W. mrakii suppressing S. cerevisiae. These findings will be helpful to optimise the vinification techniques and inoculation strategy. The volatile profiles of the mango wines from these different inoculation strategies were evaluated by a trained panel of flavourists. Different inoculation strategies produced wines with different flavour profile. The sequence of inoculation of S .cerevisiae and W. mrakii also had a significant impact on volatile production. The persistence of W. mrakii without the highly competitive S. cerevisiae allowed the W. mrakii to produce and accumulate sufficient amounts of mycosin which had an inhibitory effect on the growth and metabolic activities of S. cerevisiae in PSF. Consequently, volatile production was affected. However, PSF proved to be a generally better inoculation strategy than NSF. When S. cerevisiae was inoculated first, ethanol 111 production was extremely rapid and the high ethanol content led to severe detrimental effects on the subsequently inoculated W. mrakii. This could be seen by the lower amounts of volatile production for amongst all classes of compounds. Especially prominent was the drastic reduction in the amount of acetate esters. It was originally hypothesized that the delayed inoculation of W. mrakii might allow the final wine to contain higher amounts of acetate esters due to the reduced exposure of acetate ester hydrolysing S. cerevisiae. However, this study proved that this was not the case. A better way to obtain a fruity and floral mango wine might be by blending a W. mrakii fermented wine with a S. cerevisiae fermented wine. All in all, S. cerevisiae and W. mrakii are deemed to have potential in commercial winemaking applications due to their ability to produce desirable volatile and non-volatiles that are deemed to be beneficial to wine quality. However, further studies on how to better regulate the metabolic pathways of these 2 yeast species, either in monocultures, mixed cultures, or sequential fermentation is necessary to reap the most benefits that either or both can offer. Recommendations The mango juice used in this study was not supplemented with additional nitrogenous sources; therefore, the relatively low nitrogenous content may contribute to another additional stress factor, resulting in a stuck fermenation for the high sugar fermentations. In addition, an excessively high initial sugar content could result in excessive amounts of acetaldehyde and acetic acid being produced, which could be detrimental to the wine quality. Studies have shown that adding a nitrogenous source either at the beginning of the fermentation (Bely et al. 2003) or during the exponential phase (Arrizon and Gschaedler 2002) of the fermenation increased fermentation efficiency and reduced volatile acidity. 112 These supplementations includes soluble nutrients such as tryptone, yeast extracts, and a mixture of purine and pyrimidine bases (Thomas, Hynes and Ingledew 1994). Such supplementation could improve the fermentation efficiency of a high sugar medium for a high ethanol content fruit wine. Previous studies have shown that Saccharomyces has the ability to convert geraniol into β-citronellol and other terpenols (Carrau et al. 2005), while nerol and linalool can also be converted into α-terpineol and terpinolene (Carrau et al. 2005). In addition, production of terpenols can occur and has previously been hypothesized to be a result of de novo synthesis by S. cerevisiae through an alternative pathway involving metabolism of L-leucine (Carrau et al. 2005). Fermentation conditions have been known to affect fermentation kinetics and volatile production (Reddy and Reddy 2011). The effects of nitrogen supplementation on volatile production especially its effects on terpene alcohols have not been studied in detail. Further studies in this aspect and its application in the production of mango wine could be investigated to fully develop the ‘varietal’ character of mango wine. While the use of mixed starter cultures in different inoculation sequences has been studied in this research, the duration of the fermentation period for each stage can be investigated to produce the most desirable flavour profile, since it is possible that degradation of some volatile compounds could occur during fermentation. Furthermore, the possibility of the aging process in improving the mango wine profile could be investigated. This would not only optimise the volatile composition and wine flavour profile, it could also lead to process optimisation and economic benefits for wine makers. In addition, another aspect of mango wine that can be studied in greater detail would be the blending of a high ethanol S. cerevisiae mango wine and a low ethanol W. mrakii wine 113 with more floral and fruity attributes. A good blend between the two different styles could potentially lead to synergistic effects, producing a wine that harvests the best of both styles. 114 Bibliography Alexandre, H., I. Rousseauz, and C. Charpentier. “Ethanol adaptation mechanisms in Saccharomyces cerevisiae.” Biotechnology and Applied Biochemistry 20, no. 2 (1994): 173-183. Alves, J. A., L. C. D. O. Lima, D. R. Dias, C. A. Nunes, and R. F. Schwan. "Effects of spontaneous and inoculated fermentation on the volatile profile of lychee (Litchi chinensis Sonn.) fermented beverages." International Journal of Food Science & Technology 45, no. 11 (2010): 1365-2621. Ambesi, A., M. Miranda, V.V. Petrov, and C.W. Slayman. “Biogenesis and function of the yeast plasma membrance H(+)-ATPase.” Journal of Experimental Biology 203, no. 1 (2000): 155-160. Ardö, Y. "Flavour formation by amino acid catabolism." Biotechnology Advances 24, no. 2 (2006): 238-242. Arrizon, L., and A. Gschaedler. “Increasing fermentation efficiency at high sugar concentrations by supplementing an additional source of nitrogen during the exponential phase of the tequila fermenation process.” Canadian Journal of Microbiology 48, no. 11 (2002): 965-970. Baptista, J. A., J. F. D. P. Tavares, and R. C. Carvalho. " Comparison of polyphenols and aroma in red wines from Portuguese mainland versus Azores Islands." Food Research International, 34, no. 4 (2001): 345-355. Bauer, F.F., and I.S. Pretorius. “Yeast stress response and fermenation efficiency: how to survive the making of wine - a review.” South African Journal of Enology and Viticulture 21 , Special issue (2000): 27-51. Bely, M., A. Rinaldi, and D Dubourdieu. “Influence of assimiable nitrogen on volatile acidity production by Saccharomyces cerevisiae during high sugar fermentation.” Journal of Bioscience and Bioengineering 96, no. 6 (2003): 507-512. Bely, M., P. Stoeckle, I. Masneuf-Pomarède, and D. Dubourdieu. "Impact of mixed Torulaspora delbrueckii–Saccharomyces cerevisiae culture on high-sugar fermentation." International Journal of Food Microbiology 122, no. 3 (2008): 312-320. Berthels, N.J., R.R. Cordero Otero, F.F, Bauer, J.M. Thevelein, and I.S. Pretorius. “Discrepancy in glucose and fructose utilisation during fermentatation by Saccharomyces cerevisiae wine yeast strains.” FEMS Yeast Research 4 (2004): 683689. Bisson, L. F., and J. E. Karpel. "Genetics of yeast impacting wine quality." Annual Review of Food Science and Technology, no. 1 (2010): 139-162. Boulton, R.B., V.L. Singleton, L.F. Bisson, and R. E. Kunkee. “Viticulture for Winemakers.” In Principles and Practices of Winemaking, by R.B. Boulton, V.L. Singleton, L.F. Bisson and R. E. Kunkee, 13-64. New York: Springer US, 1996. 115 Buchbauer, G., and A. Ilic. “Biological activities of selected mono- and sesquiterpenes: Possible uses in medicine.” In Natural Products: Phytochemistry, Botany and Metabolism of Alkaloids, Phenolics and Terpenes , 4109-4159. Berlin: Springer Berlin Heidelberg, 2013. Carlson, M., B.C., Osmond, and D. Botstein. “Mutants of yeast defective in sucrose utilization” Genetics 98, no. 1 (1981): 25-40. Carrau, F, M., K. Medina, E. Boido, L. Farina, C. Gaggero, E. Dellacassa, G. Versini, and P.A. Henschke. "De novo synthesis of monoterpenes by Saccharomyces cerevisiae wine yeasts. "FEMS Microbiology Letters 243, no. 1 (2005): 107-115. Cartwright, C., J.-R. Juroszek, M. Beavan, and F. Ruby. “Ethanol dissipates the protonmotive force across the plasma membrane of Saccharomyces cerevisiae.” Journal of General Microbiology 132 , no. 132 (1986): 369-377. Casey, G.P., and W.M. Ingledew. “Ethanol tolerance in yeasts.” Critial Review in Microbiology 13, no. 3 (1986): 219-280. Chauhan, O.P., P.S. Raju, and A.S. Bawa. "Mango Flavor." In Handbook of Fruit and Vegetable Flavors, by Y.H. Hui, edited by Y.H. Hui, F. Chen and L.M.L. Nollet, 319344. New Jersey: John Wiley & Sons, Inc., 2010. Ciani, M., and F. Maccarelli. "Oenological properties of non-Saccharomyces yeasts associated with wine-making." World Journal of Microbiology & Biotechnology 14, no. 2 (1999): 199-203. Clemente-Jimenez, J. M., L. Mingorance-Cazorla, S. Martínez-Rodríguez, F. J. Las HerasVázquez, and F. Rodríguez-Vico. "Influence of sequential yeast mixtures on wine fermentation." International Journal of Food Microbiology 98, no. 3 (2005): 301-308. Coloretti, F., C. Zambonelli, L. Castellari, V. Tini, and S. Rainieri. "The effect of DL-malic acid on the metabolism of L-malic acid during wine alcoholic fermentation." Food Technology and Biotechnology 40, no. 4 (2002): 317-320. Devantier, R., S. Pedersen, and L. Olsson. ". Characterization of very high gravity ethanol fermentation of corn mash. Effect of glucoamylase dosage, pre-saccharification and yeast strain." Applied Microbiology and Biotechnology 68, no. 5 (2005): 622-629. Dubourdieu, D., T. Tominaga, I. Masneuf, C. P. des Gachons, and M. L Murat. "The role of yeasts in grape flavor development during fermentation: the example of Sauvignon blanc." American Journal of Enology and Viticulture 57, no. 1 (2006): 81-88. El Hadi, M. A. M., Zhang, Wu F. J., F. F., H., C. Zhou, and J. Tao. "Advances in fruit aroma volatile research." Molecules 18, no. 7 (2013): 8200-8229. Erten, H., and H. Tanguler. "Influence of Williopsis saturnus yeasts in combination with Saccharomyces cerevisiae on wine fermentation." Letters in Applied Microbiology 50, no. 5 (2010): 474-479. 116 Farkas, G., J. M. Rezessy-Szabo, F. Zakany, and A. Hoschke. "Interaction of Saccharomyces and non-Saccharomyces yeast strains in an alcoholic fermentation process." Acta Alimentaria 34, no. 1 (2005): 81-90. Fernández, M., J. F. Ubeda, and A. I. Briones. "Typing of non-Saccharomyces yeasts with enzymatic activities of interest in wine-making." International Journal of Food Microbiology 59, no. 1 (2000): 29-36. Fleet, G.H., and G.M. Heard. "Yeasts: growth during fermentation." In Wine Microbiology and Biotechnology, by G.H Fleet, 27-54. London: Harwood Academic Press GmbH, 1993. Förster-Fromme, K., and D. Jendrossek. "Catabolism of citronellol and related acyclic terpenoids in Pseudomonads." Applied Microbiology and Biotechnology 87, no. 3 (2010): 859-869. Francis, I. L., and J. L. Newton. "Determining wine aroma from compositional data." Australian Journal of Grape and Wine Research 11, no. 2 (2005): 114-126. Fukuda, K., et al. “Balance of activities of alcohol acetyltransferase and esterase in Saccharomyces cerevisiae is important for production of isoamyl acetate.” Applied and Environmental Microbiology 80, no. 12 (1998): 4076-4078. Gholap, A., and C. Bandyopadhyay. "Comparative assessment of aromatic principles of ripe Alphonso and Langra mango." Journal of Food Science and Technology 49, no. 12 (1975): 262-263. Godoy, H. T., and D. B. Rodriguez-Amaya. "Changes in individual carotenoids on processing and storage of mango (Mangifera indica L.) slices and puree." International Journal of Food Science and Technology 22, no. 5 (1987): 451-460. Gonzalez, R., M. Quirós, and P. Morales. "Yeast respiration of sugars by non-Saccharomyces yeast species: A promising and barely explored approach to lowering alcohol content of wines." Trends in Food Science & Technology 29, no. 1 (2013): 55-61. Gunata, Y. Z., C. L. Bayonove, R. L. Baumes, and R. E. Cordonnier. “Stability of free and bound fractions of some aroma components of grapes cv. Muscat during the wine processing: preliminary results.” American Journal of Enology and Viticulture 37, no. 2 (1986): 112-114. Guyard, C., et al. "Involvement of β-glucans in the wide-spectrum antimicrobial activity of Williopsis saturnus var. mrakii MUCL 41968 killer toxin." Molecular Medicine 8, no. 11 (2002): 686-694. Hallsworth, J.E. “Ethanol induced water stress in yeast.” Journal of Fermentation Bioengineering 85, no.2 (1998): 125-137. Heard, GM, and GH Fleet. "Occurrence and growth of yeast species during the fermentation of some Australian wines." Food Australia 38, no. 1 (1986): 22-25. 117 Herrero, Ó., D. Ramón, and M. Orejas. "Engineering the Saccharomyces cerevisiae isoprenoid pathway for de novo production of aromatic monoterpenes in wine." Metabolic Engineering 10, no. 2 (2008): 78-86. Hohmann, S. “Shaping up: the response of yeast to osmotic stress.” In Yeast Stress Responses, by S. Hohmann and W.H. Mager, 101-145. Austin : R.G. Landes Company, 1997. Hohmann, S. “Shaping up: The response of yeast to osmotic stress.” In Yeast Stress Responses, by S. Hohmann and W.H. Mager, 101-145. Berlin: Springer-Verlag Berlin Heidelberg New York, 2003. Holm Hansen, E., P. Nissen, P. Sommer, J. C. Nielsen, and N. Arneborg. "The effect of oxygen on the survival of non‐Saccharomyces yeasts during mixed culture fermentations of grape juice with Saccharomyces cerevisiae." Journal of Applied Microbiology 9, no. 13 (2001): 541-547. Holzapfel, W. H. "Appropriate starter culture technologies for small-scale fermentation in developing countries." . International Journal of Food Microbiology 75, no. 3 (2002): 197-212. Inoue, S.B., N. Takewakt, T. Takasuka, T. Mio, M. Adachi, Y. Fujii, C. Miyamoto, M. Arisawa, Y. Furuichi, and T. Watanabe. "Characterization and gene cloning of 1, 3‐β‐D‐glucan synthase from Saccharomyces cerevisiae." European Journal of Biochemistry 231, no. 3 (1995): 845-854. Iwase, T., T. Morikawa, H. Fukuda, K. Sasaki, and M. Yoshitake. "Production of fruity odor by genus Williopsis." Journal of the Brewing Society of Japan 90, no. 5 (1995): 394396. Jackson, D. I., and P. B. Lombard. "Environmental and management practices affecting grape composition and wine quality-a review." American Journal of Enology and Viticulture 44, no. 4 (1993): 409-430. Jackson, R. S. Wine Science: Principles and Applications. Oxford: Academic Press, 2008. Jain, V. K., B. Divol, B. A. Prior, and F. F. Bauer. "Elimination of glycerol and replacement with alternative products in ethanol fermentation by Saccharomyces cerevisiae." Journal of Industrial Microbiology & Biotechnology 38, no. 9 (2011): 1427-1435. Jolly, N.P., O.P.H. Augustyn, and I.S. Pretorius. "The role and use of non-Saccharomyces yeasts in wine production." South African Journal of Viticulure 27, no. 1 (2006): 15-39. Jones, P. R., R. Gawel, I. L. Francis, and E. J. Waters. "The influence of interactions between major white wine components on the aroma, flavour and texture of model white wine." Food Quality and Preference 19, no. 6 (2008): 596-607. Jungalwala, F., and H. Cama. "Carotenoids in mango (Mangifera indica L.) fruit." Indian Journal of Chemistry 62, no. 1 (1963): 36-40. Kapsopoulou, K., A. Mourtzini, M. Anthoulas, and E. Nerantzis. "Biological acidification during grape must fermentation using mixed cultures of Kluyveromyces thermotolerans 118 and Saccharomyces cerevisiae." World Journal of Microbiology and Biotechnology 23, no. 5 (2007): 735-739. Kimura, T., T. Komiyama, Y. Furuichi, Y. Iimura, S. Karita, K. Sakka, and K. Ohmiya.. "Nglycosylation is involved in the sensitivity of Saccharomyces cerevisiae to HM-1 killer toxin secreted from Hansenula mrakii IFO 0895." Applied microbiology and biotechnology 51, no. 2 (1999): 176-184. King, A., and J.R Dickinson. ""Biotransformation of monoterpene alcohols by Saccharomyces cerevisiae, Torulaspora delbrueckii and Kluyveromyces lactis." Yeast 16, no. 6 (2000): 499-506. Kulkarni, J. H., H. Singh, and K.L. Chadha. "Preliminary screening of mango varieties for wine making." Journal of Food Science and Technology 17, no. 5 (1980): 218-220. Kurita, O. "Increase of acetate ester‐hydrolysing esterase activity in mixed cultures of Saccharomyces cerevisiae and Pichia anomala." Journal of Applied Microbiology 104, no. 4 (2008): 1051-1058. Kyung, M.Y., C.-L. Rosenfield, and D.C. Knipple. “Ethanol tolerance in the yeast Saccharomyces cerevisiae is dependent on cellular oleic acid content.” Applied and Environmental Microbiology 69, no. 3 (2003): 1499-1503. Lagace, LS, and LF Bisson. "Survey of yeasts acid proteases for effectiveness of wine haze reduction." American Journal of Viticulture and Enology 41, no. 2 (1990): 147–155. Lalel, H. J., Z. Singh, and S. C. Tan. "Glycosidically-bound aroma volatile compounds in the skin and pulp of 'Kensington Pride' mango fruit at different stages of maturity." Postharvest Biology and Technology 29, no. 2 (2003a): 205-218. Lalel, H., Z. Singh, and S. Tan. "Maturity stage at harvest affects fruit ripening, quality and biosynthesis of aroma volatile compounds in 'Kensington Pride' mango." Journal of Horticultural Science and Biotechnology 78, no. 2 (2003b): 225-233. Lambrechts, M. G., and I. S. Pretorius. "Yeast and its importance to wine aroma-a review." South African Journal of Enology and Viticulture, 21., no. 1 (2000): 97-129. Laokhakunjit, N., A. Uthairatakji, O. Kerdchoechuen, A. Chatpaisarn, and S. Photchanachai. Identification of changes in volatile compound in gamma-irradiated mango during storage. Technical report, Bangkok: International Symposium "New Frontier of Irradiated Food and Non-food Products", 2005. Léchaudel, M., and J. Joas. "An overview of preharvest factors influencing mango fruit growth, quality and postharvest behaviour." Brazilian Journal of Plant Physiology 19, no. 4 (2007): 287-298. Lee, P. R., Y. L. Ong, B. Yu, P. Curran, and S. Q. Liu. "Evolution of volatile compounds in papaya wine fermented with three Williopsis saturnus yeasts." International Journal of Food science & Technology 45, no. 10 (2010a): 2032-2041. 119 Lee, P.-R., I. S.-M. Chong, B. Yu, P. Curran, and S.-Q. Liu. "Effects of sequentially inoculated Williopsis saturnus and Saccharomyces cerevisiae on volatile profiles of papaya wine." Food Research International 45, no. 1 (2012b): 177-183. Lee, P.-R., S. H. C. Kho, B. Yu, Curran, P., and S.-Q Liu. "Yeast ratio is a critical factor for sequential fermentation of papaya wine by Williopsis saturnus and Saccharomyces cerevisiae." Microbial Biotechnology 6, no. 4 (2012c): 385-393. Li, X. Impact of Selected Yeasts and Glycosidases on Chemical Composition of Mango Wine. PhD Thesis, Singapore: National University of Singapore, 2013. Li, X., B. Yu, P. Curran, and S. Q Liu. "Chemical and volatile composition of mango wines fermented with different Saccharomyces cerevisiae yeast strains." South African Journal of Enology & Viticulture 32, no. 1 (2011): 117-128. Li, X., B. Yu, P. Curran, and S.-Q. Liu. "Impact of two Williopsis yeast strains on the volatile composition of mango wine." International Journal of Food Science and Technology 47, no. 4 (2012): 808-815. Li, X., L.J. Chan, B. Yu, P. Curran, and S.-Q. Liu. "Fermentation of three varieties of mango juices with a mixture of Saccharomyces cerevisiae and Williopsis saturnus var. mrakii." International Journal of Food Microbiology 158, no. 1 (2012): 28-35. Li, X., Lim, S.L., B. Yu, P. Curran, and S.-Q. Liu. "Mango wine aroma enhancement by pulp contact and β-glucosidase." International Journal of Food Science and Technology 48, no. 11 (2013a): 2258-2266. Li, X., S.L. Lim, B. Yu, P. Curran, and S.-Q. Liu. "Impact of pulp on the chemical profile of mango wine." South African Journal of Enology and Viticulture 34, no. 1 (2013b): 119128. Lilly, M., F. F. Bauer, M. G. Lambrechts, J. H. Swiegers, D. Cozzolino, and I. S. Pretorius. "The effect of increased yeast alcohol acetyltransferase and esterase activity on the flavour profiles of wine and distillates." Yeast 23, no. 9 (2006): 641-659. Liu, S. Q., and M. Tsao. "Biocontrol of spoilage yeasts and moulds by Williopsis saturnus var. saturnus in yoghurt." Nutrition & Food Science 40, no. 2 (2010): 166-175. Magliani, W., S. Conti, L. R. Travassos, and L. Polonelli. "From yeast killer toxins to antibiobodies and beyond." FEMS Microbiology Letters 288, no. 1 (2008): 1-8. Marais, J. "Terpenes in the aroma of grapes and wines: a review." South African Journal of Enology and Viticulture 4, no. 2 (1983): 49-60. Maturano, Y. P., et al. "Multi-enzyme production by pure and mixed cultures of Saccharomyces and non-Saccharomyces yeasts during wine fermentation." International Journal of Food Microbiology 155, no. 1 (2012): 43-50. Mendoza, L. M., M. C. M. de Nadra, E. Bru, and M. E. Farías. "Influence of wine-related physicochemical factors on the growth and metabolism of non-Saccharomyces and Saccharomyces yeasts in mixed culture." Journal of Industrial Microbiology and Biotechnology 36, no. 2 (2009): 229-237. 120 Minetoki, T. "Alcohol acetyl transferase of sake yeast." Journal of the Brewing Society of Japan 87, no. 5 (1992): 334-340. Miyake, T., and T. Shibamoto. "Quantitative analysis of acetaldehyde in foods and beverages." Journal of Agricultural and Food Chemistry 41, no. 11 (1993): 1968-1970. Moreno, J. J., C. Millán, J. M. Ortega, and M. Medina. "Analytical differentiation of wine fermentations using pure and mixed yeast cultures." Journal of Industrial Microbiology and Biotechnology 7, no. 3 (1991): 181-189. Myers, D.K., D.T.M. Lawlor, and P.V. Attfield. "Influence of invertase acitivity and glycerol synthesis and retention on fermentation of media with a high sugar concnetration by Saccharomyces cerevisiae." Applied and Environmental Microbiology 63, no. 1 (1997): 145-150. Nieuwoudt, H. H., B. A. Prior, I. S. Pretorius, and F. F Bauer. "Glycerol in South African table wines: an assessment of its relationship to wine quality." South African Journal for Enology and Viticulture 23, no. 1 (2002): 22-30. Nissen, P., and N. Arneborg. "Characterization of early deaths of non-Saccharomyces yeasts in mixed cultures with Saccharomyces cerevisiae." Archives of Microbiology 180, no. 4 (2003): 257-263. Nissen, P., D. Nielsen, and N Arneborg. "Viable Saccharomyces cerevisiae cells at high concentrations cause early growth arrest of non‐Saccharomyces yeasts in mixed cultures by a cell–cell contact‐mediated mechanism." Yeast 20, no. 4 (2003): 331-341. Obisanya, M. O., J.O. Aina, and G.B Oguntimein. "Production of wine from mango (Magnifera indica L.) using Saccharomyces and Schizosaccharomyces species isolated from palm wine." Journal of Applied Microbiology 63, no. 3 (1987): 191-196. Ong, P. K., and T. E. Acree. "Similarities in the aroma chemistry of Gewürztraminer variety wines and lychee (Litchi chinesis Sonn.) fruit." Journal of Agricultural and Food Chemistry 47, no. 2 (1999): 665-670. Onkarayya, H, and H, Singh. "Screening of mango varieties for Dessert and Madeira-style wines." American Journal of Enology and Viticulture 35, no. 2 (1984): 63-65. Onkarayya, H. "A rapid madeirization process to improve mango dessert wines." Journal of Food Science and Technology 23, no. 3 (1986): 175-176. Perestrelo, R., A. Fernandes, F. F. Albuquerque, J. C. Marques, and J. S. Camara. "Analytical characterization of the aroma of Tinta Negra Mole red wine: Identification of the main odorants compounds." Analytica Chimica Acta 563, no. 1 (2006): 154-164. Pietruszka, M., K Pielech-Przybylska, and J. S Szopa. "Synthesis of higher alcohols during alcoholic fermentation of rye mashes." Zeszyty Naukowe. Chemia Spożywcza i Biotechnologia/Politechnika Łódzka 74, no. 108 (2010): 51-64. Pino, J. A., and J. Mesa. "Contribution of volatile compounds to mango (Mangifera indica L.) aroma." Flavour and Fragrance Journal 21, no. 2 (2006): 207-213. 121 Pino, J. A., and O Queris. "Analysis of volatile compounds of mango wine." Food Chemistry 125, no. 4 (2011): 1141-1146. Pino, J. A., J. Mesa, Y. Muñoz, M. P. Martí, and R. Marbot. "Volatile components from mango (Mangifera indica L.) cultivars." Journal of Agricultural and Food Chemistry 53, no. 6 (2005): 2213-2223. Plata, C., C. Millan, J. C. Mauricio, and J. M. Ortega. "Formation of ethyl acetate and isoamyl acetate by various species of wine yeasts." Food Microbiology 20, no. 2 (2003): 217-224. Pretorius, I. S. "Tailoring wine yeast for the new millennium: novel approaches to the ancient art of winemaking." Yeast 16, no. 8 (2007): 675-729. Pretorius, I., T. Van der Westhuizen, and O. Augustyn. "Yeast biodiversity in vineyards and wineries and its importance to the South African wine industry: A review." South African Journal of Enology and Viticulture 20, no. 2 (1999): 61-70. Quijano, C. E., G. Salamanca, and J. A. Pino. "Aroma volatile constituents of Colombian varieties of mango (Mangifera indica L.)." Flavour and Fragrance Journal 22, no. 5 (2007): 401-406. Radle, F., and H. Schülz. "Glycerol production of various strains of Saccharomyces." American Journal of Enology and Viticulture no. 1 (1982): 36–40. Rapp, A. "Volatile flavour of wine: correlation between instrumental analysis and sensory perception." Food/Nahrung 42, no. 6 (1998): 351-363. Rapp, A., and H. Mandery. "Wine aroma." Experientia 42, no. 8 (1986): 873-884. Reddy, L. V. A., and O. V. S. Reddy. ""Production and characterization of wine from mango fruit (Mangifera indica L)."." World Journal of Microbiology and Biotechnology 21, no. 8-9 (2005): 1345-1350. Reddy, L. V. A., and O. V. S. Reddy. "Effect of fermentation conditions on yeast growth and volatile composition of wine produced from mango Mangifera indica L.) fruit juice." Food and Bioproducts Processing 89, no. 4 (2011): 487-49. Reddy, L.V., and O Reddy. "Production of ethanol from mango (Mangifera indica L.) fruit juice fermentation." Research Journal of Microbiology 2, no. 10 (2007): 763-769. Reddy, L.V., O.V.S. Reddy, and V.K. Joshi. "Production, optimization and characterization of wine from mango (Mangifera indica Linn.)." Natural Product Radiance 8, no. 4 (2009): 426-435. Reddy, L.V.A., and O. V. S. Reddy. "Effect of enzymatic maceration on synthesis of higher alcohols during mango wine fermentation." Journal of Food Quality 32, no. 1 (2009): 34-47. Remize, F., J.L. Roustan, J.M. Sablayrolles, P. Barre, and S. Dequin. "Glycerol overproduction by engineered Saccharomyces cerevisiae wine yeast strains leads to 122 substantial changes in by-product formation and to a stimulation of fermentation rate in stationary phase." Applied and Environmental Microbiology 65, no. 1 (1999): 143-149. Rojas, V., J. V. Gil, F. Piñaga, and P. Manzanares. "Studies on acetate ester production by non-Saccharomyce wine yeasts." International Journal of Food Microbiology 70, no. 3 (2001): 283-289. Romano, P., G. Suzzi, L. Turbanti, and M. Polsinelli. "Acetaldehyde production in Saccharomyces cerevisiae wine yeasts." FEMS Microbiology Letters, 118, no. 3 (1994): 213-218. Rosi, I, M Vinella, and P Domizio. "Characterization of β-glucosidase activity in yeast of oenological origin." Journal of Applied Bacteriology 77, no. 5 (1994): 519-527. Sadoudi, M., R. Tourdot-Maréchal, S. Rousseaux, D. Steyer, J-J Gallardo-Chacón, J. Ballester, S. Vichi, R. Guérin-Schneider, J. Caixach, and H. Alexandre. "Yeast–yeast interactions revealed by aromatic profile analysis of Sauvignon Blanc wine fermented by single or co-culture of non-Saccharomyces and Saccharomyces yeasts."Food microbiology 32, no. 2 (2012): 243-253. Sáenz-Navajas, M. P., E. Campo, P. Fernández-Zurbano, D. Valentin, and V. Ferreira. "An assessment of the effects of wine volatiles on the perception of taste and astringency in wine." Food Chemistry 121, no. 4 (2010): 1139-1149. Saerens, S.M.G., F. Delvaus, K.J, Verstrepen, P. Van Dijck, M. Thevelein, and F.R. Delvaux. “Parameters affecting ethyl ester production by Saccharomyces cerevisiae during fermentation.” Applied and Environmental Microbiology 74, no. 2 (2008): 454-461. Sagar, V., D. Khurdyia, and K. Balakrishnan. "Quality of dehydrated ripe mango slices as affected by packaging material and mode of packaging." Journal of Food Science and Technology 36, no. 1 (1999): 67-70. Sales, K., W. Brandt, E. Rumbak, and G. Lindsey. “The LEA-like protein HSP12 in Saccharomyces cerevisiae has a plasma membrane location and protects membrane against desiccation and ethanol-induced stress.” Biochimica et Biophysica Acta (BBA) Biomembranes 1463, no. 2 (2000): 267-278. Salmon, J.-M., O. Vincent, J. Mauricio, C. Bely, and P. Barre. “Sugar transport inhibition and apparent loss of activity in Saccharomyces cerevisiae as a major limiting factor of enological fermentations.” American Journal of Enology and Viticulture 44, no. 1 (1993): 56-64. Sánchez-Palomo, E., M. C. Diaz-Maroto, and M. S. Pérez-Coello. "Rapid determination of volatile compounds in grapes by HS-SPME coupled with GC–MS." Talanta 66, no. 5 (2005): 1152-1157. Scanes, K.T., S. Hohmann, and B.A. Prior. “Glycerol production by the yeast Saccharomyces cerevisiae and its relevance to wine: a review.” South African Journal of Enology and Viticulture 19 no. 1 (1998): 17-24. 123 Seow, Y. X., P. K. Ong, and S. Q. Liu. " Production of flavour-active methionol from methionine metabolism by yeasts in coconut cream." International Journal of Food Microbiology 143, no. 3 (2010): 235-240. Slaninova, I., S. Sestak, A. Svoboda, and V. Farkas. “Cell wall and cytoskeleton reorganization as a response to hyperosmotic shock in Saccharomyces cerevisiae.” Archives of Microbiology 173, no. 4 (2000): 245-252. Soden, A., I. L. Francis, H. Oakey, and P. A Henschke. "Effects of co‐fermentation with Candida stellata and Saccharomyces cerevisiae on the aroma and composition of Chardonnay wine." Australian Journal of Grape and Wine Research 6, no.1 (2000): 2130. Spencer, J., M. Morris, and W Kennard. "Vitamin C concentration in developing and mature fruits of mango (Mangifera indica L.)." Plant Physiology 31, no. 1 (1956): 79-80. Styger, G., D. Jacobson, B. A. Prior, and F. F. Bauer. "Genetic analysis of the metabolic pathways responsible for aroma metabolite production by Saccharomyces cerevisiae." Applied Microbiology and Biotechnology 97, no. 10 (2013): 4429-4442. Sudheer, K., Yannam, S.P. Reddy, and O.V.S Reddy. "Optimisation of fermentation conditions for mango (Mangifera indica L.) wine production by employing response surface methodology." International Journal of Food Science & Technology 44, no. 11 (2009): 2320-2327. Sumby, K. M., P. R. Grbin, and V. Jiranek. " Microbial modulation of aromatic esters in wine: current knowledge and future prospects." Food Chemistry 121, no. 1 (2010): 116. Swiegers, J. H., E. J. Bartowsky, P. A. Henschke, and I. S. Pretorius. " Yeast and bacterial modulation of wine aroma and flavour." Australian Journal of Grape and Wine Research 11, no. 2 (2005): 139-173. Tan, A. W., P. R. Lee, Y. X. Seow, P. K. Ong, and S. Q. Liu. "Volatile sulphur compounds and pathways of L-methionine catabolism in Williopsis yeasts." Applied Microbiology and Biotechnology 95, no. 4 (2012): 1011-1020. Tandon, D. K., and S. K. Kalra. "Pectin changes during the develoment of mango fruit cv. Dashehari." Journal of Horticultural Science 59 (1984): 283-286. Tehlivets, O., K. Scheuringer, and S. D. Kohlwein. "Fatty acid synthesis and elongation in yeast." Biochimica et Biophysica Acta (BBA)-Molecular and Cell Biology of Lipids 1771, no. 3 (2007): 255-270. Tharanathan, R., H. Yashode, and T. Prabha. "Mango (Mangifera inica L), "The king of fruits"- An overview." Food Reviews International 22, no. 2 (2006): 95-123. This, P., T. Lacombe, and M. R. Thomas. "Historical origins and genetic diversity of wine grapes." TRENDS in Genetics 22, no. 9 (2006): 511-519. Thomas, K. C., S. H. Hynes, and W. M. Ingledew. "Effects of particulate materials and osmoprotectants on very-high-gravity ethanolic fermentation by Saccharomyces 124 cerevisiae." Applied and Environmental Microbiology 60(5), 60, no. 5 (1994): 15191524. Thomas, KC, S H Hynes, and W M Ingledew. “Effects of particulate materials and osmoprotectants on very high gravity ethanolic fermentation by Saccharomyces cerevisiae.” Applied and Environmental Microbiology 60, no. 5 (1994): 1519-1524. Trinh, T. T. T., B. Yu, P. Curran, and S. Q. Liu. "Dynamics of volatile compounds during longan juice fermentation by three yeasts from the genus Williopsis." Acta Alimentaria 40, no. 3 (2011): 385-395. Ugliano, M., E. J. Bartowsky, J. McCarthy, L. Moio, and P. A. Henschke. " Hydrolysis and transformation of grape glycosidically bound volatile compounds during fermentation with three Saccharomyces yeast strains." Journal of Agricultural and Food Chemistry 54, no. 17 (2006): 6322-6331. van Zyl, P. J., S. G. Kilian, and B. A. Prior. "The role of an active transport mechanism in glycerol accumulation during osmoregulation by Zygosaccharomyces rouxii." Applied Microbiology and Biotechnology 34, no. 2 (1990): 231-235. Varakumar, S., K. Naresh, P. S. Variyar, A. Sharma, and O. V. S. Reddy. "Role of malolactic fermentation on the quality of mango (Mangifera indica L.) Wine." Food Biotechnology 27, no. 2 (2013): 119-136. Varakumar, S., N. Kondapalli, and O.V.S. Reddy. "Effect of co-fermentation with Saccharomyces cerevisiae and Torulaspora delbrueckii or Metschnikowia pulcherrima on the aroma and sensory properties of mango wine." Annals of Microbiology 62, no. 4 (2012): 1353-1360. Varakumar, S., Y.S. Kumar, and O.V.S. Reddy. "Carotenoid composition of mango (Mangifera indica L.) wine and its antioxidant activity." Journal of Food Biochemistry 35, no. 5 (2011): 1538-1547. Vásquez-Caicedo, A. L., S. Neidhart, P. Pathomrungsiyounggul, P. Wiriyacharee, A. Chattrakul, P. Sruamsiri, P. Manochai, F. Bangerth, and R. Carle. "Physical, Chemical, and Sensory Properties of Nine Thai Mango Cultivars and Evaluation of their Technological and Nutritional Potential." In International Symposium: Sustaining Food Security and Managing Natural Resources in Southeast Asia Challenges for the 21st Century, pp. 8-11. 2002 Viana, F., J. V. Gil, S. Vallés, and P. Manzanares. " Increasing the levels of 2-phenylethyl acetate in wine through the use of a mixed culture of Hanseniaspora osmophila and Saccharomyces cerevisiae." International Journal of Food Microbiology 135, no. 1 (2009): 68-74. Walker, G. M. Yeast Physiology and Biotechnology. West Sussex: John Wiley & Sons., 1998. Wang, D., Y. Su, J. Hu, and G. Zhao. “Fermentation kinetics of different sugars by apple wine yeast Saccharomyces cerevisiae.” Journal of the Institute of Brewing 110, no. 4 (2004): 340-346. 125 Whiting, G. C. "Organic acid metabolism of yeasts during fermentation of alcoholic beverages—a review." Journal of the Institute of Brewing 82, no. 2 (1976): 84-92. Yamamoto, T., M. Imai, K. Tachibana, and M. Mayumi. "Application of monoclonal antibodies to the isolation and characterization of a killer toxin secreted byHansenula mrakii." FEBS Letters 195, no. 1 (1986): 253-257. Yilmaztekin, M., H. Erten, and T. Cabaroglu. "Enhanced production of isoamyl acetate from beet molasses with addition of fusel oil by Williopsis saturnus var. saturnus." Food Chemistry 112, no. 2 (2009): 290-294. Zoecklein, B. W., J. E. Marcy, J. M. Williams, and Y. Jasinski. "Effect of native yeasts and selected strains of Saccharomyces cerevisiae on glycosyl glucose, potential volatile terpenes, and selected aglycones of White Riesling Vitis vinifera L.) wines." Journal of Food Composition and Analysis 10, no. 1 (1997): 55-65. 126 [...]... 1.5 Research aims and objectives The overall objective of this research project was to study the effects of sugar concentration and inoculation strategies on mango wine fermentation Aim 1: Effect of sugar concentration on mango wine fermentation with S cerevisiae MERIT.ferm — Chapter 3 High sugar concentration or hyperosmotic pressure creates redox imbalance and the efforts of the yeasts to combat this... hydrolysis (Rapp and Mandery 1986) 1.4 Influence of fermentation conditions and yeast strains 1.4.1 Amelioration of must Conventionally, wine grapes are carefully cultivated to contain the optimum amounts of nutrients for the production of a balanced wine However, many factors beyond the 8 control of the viticulturist and winemaker (for instance, weather and climate change) may affect the composition of the... implications on wine quality The effects of initial sugar concentration on yeast biomass accumulation, glycerol and volatile compounds production with special emphasis on acetaldehyde and acetic acid were investigated While it has been hypothesised that high initial sugar concentration will impact wine quality negatively, the specific effects are unknown Hence, the specific effects of high initial sugar concentration. .. focus on yeast growth, sugar consumption, glycerol production and key volatile production in low sugar fermentation (unfortified mango juice with initial TSS of 16.6°B), medium sugar concentration (fortified mango juice with initial TSS of 23°B) and high sugar fermentaiton (fortified mango juice with initial TSS of 30°B) The sugar concentrations were selected for this study based on the typical total... fermentation conditions for the production of mango wine In addition, Reddy et al (2010) also concluded that the aromatic compounds of mango wine were comparable in concentration to those of grape wine Pectinase enzyme treatment was also found to be effective in increasing the yield of juice, the production of ethanol, increasing the yield of higher alcohols (Reddy and Reddy 2009) and 4 improving mango. .. co -fermentation depends on the physiological properties of the individual yeasts – its compatibility with other yeasts and the effects on growth rate and biomass development Suppression of one yeast by the other can result in its reduced metabolic activity and hence decreased impact on the wine characteristics One strategy may be the coinoculation of a weakly fermentative yeast at high ratio to a strongly... of 23°Brix and 30°Brix, respectively A set of control fermentation with no sugar supplementation (low sugar fermentation) was also conducted Each conical flask was inoculated with 105 CFU/mL of S cerevisiae MERIT.ferm then plugged with cotton wool and wrapped with aluminium foil Static fermentation was carried out for 35 days at 20°C The fermentations were conducted in triplicate 2.5 Fermentation of. .. stressful conditions for the propagation and growth of most yeast strains during the onset of fermentation while ethanol toxicity occurs later in the fermentation process Although commercial wine yeasts are selected for their inherent ability to cope with such environmental stresses, yeast metabolism is still significantly affected by fermentation conditions High gravity fermentation or fermentation with... (Reddy and Reddy, 2009), the analysis of volatile composition of wine fermented from Indian mango cultivars (Reddy et al 2010), the effects of fermentation conditions on yeast growth and volatile production (Reddy and Reddy, 2011), and the antioxidant potential of mango wine (Sudheer et al 2012) were reported subsequently These studies concluded that 25°C, pH 5, with 100 ppm SO2 and initial aeration were... lack of studies investigating the effects of sequential inoculation strategies on the production of mango wine involving the use of S cerevisiae MERIT.ferm and W 13 saturnus NCYC 500, Chapter 5 presents information on the differences in yeast growth kinetics, volatile composition and sensory perception in three wines with different inoculation strategies 14 Chapter 2 Materials and Methods 2.1 Mango