MALOLACTIC FERMENTATION AND BIOGENIC AMINE FORMATION. SCREENING FOR LOW BIOGENIC AMINE FORMING Oenococcus oeni STRAINS.

Transcription

1 Universität für Bodenkultur, Wien University of Natural Resources and Life Sciences, Vienna Department of Food Sciences and Technology Laboratory of Food Biotechnology MALOLACTIC FERMENTATION AND BIOGENIC AMINE FORMATION. SCREENING FOR LOW BIOGENIC AMINE FORMING Oenococcus oeni STRAINS. Master Thesis Submitted by Nuno Miguel Barreira Braz da Silva Bakk. techn. Supervised by o. Univ. Prof. Dipl.-Chem. Dr. rer. nat. Klaus D. Kulbe Vienna, July 2012

2 Acknowledgements It is a pleasure to thank all the people who helped me during my studies at the University of Natural Resources and Life Sciences, Vienna and made this thesis possible. First and foremost, I want to offer my sincere gratitude to Prof. Dr. Klaus Dieter Kulbe for the assistance and for offering me the opportunity to join the FWF project which provided me important financial support during my thesis. Very special thanks go to my supervisor Dr. Andrés del Hierro for his expertise, his inspiration and his efforts to explain things clearly and simply. He was more than a supervisor and became a really good friend. Without him this thesis would have never been possible. I am also truly in debt to Dr. DI Herbert Michlmayr for all the patience, the support and all the skills he transmitted to me, that added considerably to my graduated experience. Moreover, I wish to thank the other members of the wine group, DI Christina Schümann, DI Kateryna Rogowicz and DI Veronika Rogl for the help and support but more importantly for the friendship within this group that provided a funny and stimulating environment in which to learn and grow. Furthermore, I would also like to thank all the other student colleagues of the laboratory of Food and Biotechnology for the help and for providing a good working climate. Special thanks go to Dr. Reinhard Eder for the collaboration in this project. I wish to thank all my working colleagues from the department of Chemistry, Biology and Wine Cellar from the Federal College and Research Institute for Viticulture and Pomology, Klosterneuburg, Austria, for the support and pleasant working atmosphere. Despite the geographical distance my family was always nearby. I would like to thank my parents, Felizmina Maria and Domingos Lima as well as my sister Maria José and my brother Francisco for all the support they provided me in the course of my academic career. It is to them that I dedicate this thesis. To my girlfriend Angelika, you were my shadow and my light, you supported, encouraged, understood and loved me at every moment during this process that seemed endless at times. I finally thank my close friends of Wolfganggasse 7 and Dijkgraaf 5C who shared good and bad moments with me during this entire journey. 2

3 Table of Contents Index of Tables... 7 Index of Figures... 8 Abbreviations Introduction Hazard Identification Background Formation of Biogenic Amines Decarboxylases Histidine decarboxylase Tyrosine decarboxylase Ornithine decarboxylase Lysine decarboxylase Phenylalanine decarboxylase Metabolic pathway of decarboxylation Main scenarios leading to Biogenic Amine formation in wine Substrate availability Viticulture conditions Fermentation Alcoholic fermentation Malolactic fermentation Aging and storage Wine preservatives Sulphur dioxide Dimethyl Dicarbonate Wine lees Time Filtration Type of barrel Temperature

4 Factors affecting growth of Biogenic Amine producers Grape skin maceration Pectolytic enzymes Phenolics Wine physicochemical composition Wine influence on the decarboxylase gene expression Glucose and L-malic acid Lactic acid Ethanol Sulphur dioxide and ph Toxicology of Biogenic Amines Histamine Synthesis and metabolism in humans Physiological functions Toxic response Toxic dose Tyramine and Phenylethylamine Synthesis and metabolism in humans Physiological functions Toxic response Toxic dose Putrescine and Cadaverine Synthesis and metabolism in humans Physiological functions Toxic response Toxic dose Analytical tools for detection of Biogenic Amines in wine Extraction techniques Rapid and semi-quantitative methods Screening methods using selective media Enzymatic methods Quantitative methods Chromatography methods

5 Thin-layer Chromatography Liquid Chromatography Capillary Electrophoresis Polymerase Chain Reaction Control of Biogenic Amines production Experimental overview Materials and Methods Material Media Buffers Isolation of Lactic Acid Bacteria Inoculum preparation Wine preparation Sampling Methods Microscope Molecular methods Chemical methods Panel tasting Results Bacterial growth S Amplified Ribosomal DNA Restriction Analysis Method Wine parameters Fourier Transform Infrared spectroscopy analysis Analyses from the Institute of Chemistry Analyses from the Institute of Oenology Chemical analyses Biogenic Amine Production Alcohols and ethyl acetate Phenolics

6 4.6. Wine aroma Discussion and Conclusion Identification and characterization of Lactic Acid Bacteria Comparison between the two cellars Amino acids Lactic Acid Bacteria Biogenic amines Wine aroma Summary Zusammenfassung References Annex

7 Index of Tables Table 1.1: Biogenic amines in foods and their pharmacological effects. 12 Table 1.2: Biogenic amine producing microorganisms found in wine. 15 Table 1.3: Wine ph determines the quantity of free sulphur dioxide needed to produce 0.8 mg/l of molecular sulphur dioxide. 32 Table 1.4: Biogenic amine content in different food products. 40 Table 1.5: Dose response of histamine alcoholic drinks in patients after oral administration. 46 Table 3.1: Different Man, Rogosa and Sharpe media used. 59 Table 3.2: Origin of the selected strains. 60 Table 4.1: Summary of the identification of Lactic Acid Bacteria isolates by restriction DNA patterns after digestion with BFaI and MSeI. 72 Table 4.2: Control wine analyses. 73 Table 4.3: Amino acid concentrations (mg/l) of the wines in the different cellars after malolactic fermentation. 82 Table 4.4: Biogenic amine content of the wines produced in two different cellars after malolactic fermentation. 83 Table 4.5: Higher alcohols and esters (ethyl acetate) measured in the final wines. 84 Table 4.6: Phenolic compounds measured in finished wines. 85 Table 4.7: Results of the wine descriptive analyses. 87 7

8 Index of Figures Figure 1.1: Chemical structure of some of the most important biogenic amines from the oenological point of view. 13 Figure 1.2: Pathways of arginine metabolism in bacteria. 19 Figure 1.3: Biogenic amine biosynthesis pathway in bacteria. 21 Figure 1.4: Schematic representation of the main biochemical mechanisms of yeast metabolism during alcoholic fermentation. 24 Figure 1.5: The average concentration for histamine, putrescine, tyramine, cadaverine and phenylethylamine in the seven stages of industrial winemaking. 26 Figure 1.6: Schematic description of the major pathways of hexose (glucose) fermentation by lactic acid bacteria. 28 Figure 1.7: Oenococcus oeni viewed with phase contrast microscopy at a magnification of 1000x. 29 Figure 1.8: The different forms of sulphur dioxide according with the ph values. 31 Figure 1.9: Sulphur dioxide presented in the different forms. 31 Figure 1.10: Effect of biogenic amines in the human intestinal tract. 41 Figure 1.11: Schematic representation of the histamine metabolism. 43 Figure 1.12: Multiplex Polimerase Chain Reaction amplification of wine bacteria producing histamine, tyramine and putrescine. 54 Figure 3.1: Cultured strains before inoculation into the wine. 61 Figure 3.2: The wineries where the experiments took place. 62 Figure 3.3: Statistical wine tasting tool in order to evaluate the wine according with different defined aromas. 69 Figure 3.4: Sensory spider diagram for wine. 69 Figure 4.1: Restriction pattern after BfaI digestion isolated from Austrian wines. 71 Figure 4.2: Restriction pattern after MseI digestion isolated from Austrian wines. 71 Figure 4.3: FTIR analyses from strain K

9 Figure 4.4: FTIR analyses from strain SK3. 75 Figure 4.5: Final sulphur dioxide concentrations of wines from the Institute of Chemistry. 77 Figure 4.6: FTIR analyses from the strain Figure 4.7: FTIR analyses from the strain SK3. 78 Figure 4.8: Final sulphur dioxide concentrations from wines of the Institute of Oenology. 80 Figure 4.9: Aroma spider plots of wines from the Institute of Oenology 88 Figure 4.10: Aroma spider plot of wines from the Institute of Chemistry. 88 Figure 9.1: FTIR analyses from strain 551 (Institute of Chemistry). 110 Figure 9.2: FTIR analyses from strain 433 (Institute of Chemistry). 111 Figure 9.3: FTIR analyses from strain 78 (Institute of Chemistry). 112 Figure 9.4: FTIR analyses from the strain K1 (Institute of Oenology). 113 Figure 9.5: FTIR analyses from the strain 433 (Institute of Oenology). 114 Figure 9.6: FTIR analyses from the strain 78 (Institute of Oenology)

10 Abbreviations aadc AF ADI Amino acid decarboxylase Alcoholic fermentation Agmatine deiminase ARDRA Amplified rdna restriction analysis BA BOKU Biogenic Amines Universität für Bodenkultur, Wien University of Natural Resources and Applied Life Sciences, Vienna Bp CE DAO FTIR GC HPLC HDC KC LAB MAO MLF ODC OTC PCR PLP TDC TLC base pair Capillary electrophoresis Diamino oxidase Fourier Transform Infrared Spectroscopy Gas chromatography High-performance liquid chromatography Histidine decarboxylase Carbamate kinase Lactic Acid Bacteria Monoamino oxidase Malolactic fermentation Ornithine decarboxylase Ornithine transcarbomylase Polymerase Chain Reaction Pyridoxal-5-phosphate Tyrosine decarboxylase Thin-layer chromatography 10

11 1. Introduction BAs are defined as low molecular weight organic bases that widely occur in nature. They are formed and degraded as a result of the normal metabolism of microorganisms, plants and animals. In food they are mostly produced during fermentation, aging and storage as a result of microbial decarboxylation of amino acids. Wine, as well as other fermented foods is produced due to the beneficial effects of fermentation such as preservation, hygienic safety and improved nutritive value and aroma (Moreno-Arribas et al., 2009). MLF of wine can be defined as the enzymatic conversion of L-malic acid into L-lactic acid. This secondary process normally follows alcoholic fermentation and favours the deacidification of high acid wines. Malolactic fermentation is normally conducted by lactic acid bacteria, especially by Oenococcus oeni, a species that is able to withstand the low ph (<3.5), high ethanol (>10% v/v) and high sulphur dioxide levels (50 mg/l) present in wine (Ribéreau-Gayon et al., 2006). However, lactic acid bacteria also seem to be the main cause for the formation of biogenic amines in wines; their formation, however, is strain dependent (Beneduce et al., 2010). Several viticultural and oenological factors may have an impact on the level of biogenic amines formed in wine. Factors such as viticultural conditions, aging practices, grape skin maceration and nutrient additions can increase the precursor amino acids in grapes and wine. Other factors such as ph, temperature, nutrient status, substrate availability, fermentation and the use of preservatives may influence the diversity, growth and decarboxylase activity of microorganisms that can potentially produce biogenic amines (Smit, 2007). Low concentrations of BA are usually tolerated by the human body, once they are efficiently detoxified by mono- and diamine oxidase in the intestinal tract. Biogenic amines are able to cause intoxication when ingested in high amounts especially in susceptible individuals. Histamine is responsible for most of the serious intoxication reactions such as rash, oedema, headache, and vomiting. Tyramine and phenylethylamine can lead to symptoms connected with the release of neuradrenaline from the sympathetic nervous system. Putrescine and cadaverine although not being toxic themselves, can cause a depreciation in wine aroma and enhance the toxicity of other biogenic amines by interacting with catabolic enzymes (EFSA, 2011). Consumer demands in order to increase food safety have led several countries to create guidelines for biogenic amines. The fact that no global policy has been created yet can obstruct future exports to such countries, making biogenic amines a potential economic threat (Martín-Álvarez et al., 2005). Several simple and rapid methods including analytical chromatography and PCR based methods have been developed in order to detect biogenic amine producing bacteria in food (Marcobal et al., 2006c). Applying such methods, the objective of the present work was to screen selected O. oeni strains in regard to their capacities to produce BA. Using the same conditions as applied in commercial wineries, the central aim of the present thesis was to investigate the formation of biogenic amines during MLF. 11

12 1.2. Hazard Identification Background Biogenic Amines (BA) are low molecular weight organic bases derived internally from plants as well as fruits and vegetables. Those organic bases can be divided in several groups according to their chemical structure such as aromatic (tyramine and phenylethylamine), aliphatic (putrescine and cadaverine) or heterocyclic (histamine and tryptamine), or according to the number of amino groups into monoamines (phenylethylamine and tyramine) and diamines (histamine, cadaverine and putrescine). BA can also be classified into volatile (phenylethylamine) and non-volatile (histamine, cadaverine, putrescine, spermine, agmatine, tryptamine and tyramine) biogenic amines (Moreno-Arrivas and Polo, 2009). Figure 1.1 shows the chemical structures of the most important biogenic amines. Biogenic amines are essential metabolic constituents when present at low concentrations in order to ensure physiological function in animals, plants and microorganisms. Nevertheless, these biological produced amines can have adverse effects when present at high concentrations and pose a health risk for sensitive individuals. Table 1.1 gives an overview on which precursor produces which amine and the pharmacological effects of each amine. Table 1.1: Biogenic amines in foods and their pharmacological effects (adapted from Salabi, 1996). Amine Precursor Pharmacological effects Histamine Tyramine Putrescine and Cadaverine Phenylethylamine Histidine Tyrosine Ornithine and Lysine Phenylalanine Liberates adrenaline and noradrenaline Stimulates both sensory and motor neurons Controls gastric acid secretion Increases the cardiac output Causes lacrimation and salivation Increases respiration Hypotension Depreciation of the wine aroma Potentiate the toxicity of other amines Releases neuradrenaline from the sympathetic nervous system Increases the blood pressure Causes migraine 12

13 Figure 1.1: Chemical structure of some of the most important biogenic amines from the oenological point of view. Aliphatic amines Putrescine Cadaverine Aromatic amines Tyramine Phenylethylamine Heterocyclic amine Histamine 13

14 Formation of Biogenic Amines There are three main requirements for the formation of biogenic amines: (1) the availability of free amino acids (precursors), (2) the existence of decarboxylase positive microorganisms and (3) the presence of suitable pre-conditions that allow bacterial growth, bacterial activity decarboxylase and synthesis (Anli and Bayram, 2008). In the set of metabolic pathways that break down amino acids into smaller units, several reactions are pointed, namely decarboxylation, transamination, deamination and desulphuration (Anli and Bayram, 2008). Many microorganisms possessing decarboxylase enzymes are able to produce biogenic amines. This reaction generally occurs in acidic media increasing the microorganism survival and favouring microbial growth by enhancing the ph. Furthermore, amino acid decarboxylation presents a lack of substrate specificity and provides ph homeostasis independently of the amino acid available (Foster et al., 1991; Park et al., 1996). Biogenic amines formation among different organisms The production of BA among LAB is strain specific and characteristic among certain species. Several studies pointed horizontal gene transfer as accountant for their dissemination within strains (Coton and Coton, 2009; Lucas et al., 2005; Marcobal et al., 2006b). In a general manner, putrescine and cadaverine are mainly found in enterobacteria while tyramine is mostly reported in enterococci. In the case of wine histamine, putrescine, cadaverine and tyramine have been reported as the most important BA produced by decarboxylation of the respective amino acids. 14

15 Table 1.2: Biogenic amine producing microorganisms found in wine (adapted from Ladero et al., 2010). Wine Biogenic amines Producer microorganism Histamine Oenococcus oeni, Lactobacillus hilgardii, Pediococcus parvulus Tyramine and Phenylethylamine Putrescine Cadaverine Lactobacillus brevis, Lactobacillus hilgardii, Leuconostoc mesenteroides, Lactobacillus plantarum, Enterococcus faecium Lactobacillus brevis, Lactobacillus hilgardii, Leuconostoc mesenteroides, Lactobacillus plantarum, O. Oeni, Lactobacillus buchneri, Lactobacillus zeae Enterobacteriaceae Decarboxylases Most of the amino acid decarboxylases belonging to animals and gram-negative bacteria are pyridoxal-phosphate-dependent enzyme group, which use pyridoxal-5-phosphate (PLP) as a coenzyme. In wines BA are produced by diverse lactic acid bacteria (Gram positive bacteria) but only by strains carrying specific metabolic pathways that convert the precursor amine acid into BA. The histidine decarboxylase (HDC) pathway producing histamine, the tyrosine decarboxylase (TDC) pathway producing tyramine and two independent pathways producing putrescine: (1) the ornithine decarboxylase and (2) the agmatine deiminase (ADI) pathways (Arena and Manca de Nadra, 2001; Lonvaud-Funel, 2001; Smit et al., 2008). Moreover, the production of cadaverine and phenylethylamine is correspondently originated by lysine and phenylalanine decarboxylases although found to a less extent in wine. 15

16 Histidine decarboxylase Histidine decarboxylase (HDC) is the only enzyme considered in this framework that falls into two different classes: (1) enzymes belonging to gram negative bacteria which are stimulated by PLP and (2) those belonging to gram positive bacteria, which use a covalent bound pyruvoyl group as a cofactor in the reaction and are PLP independent (Smit, 2007). In regard to the last point, several studies have been performed in Clostridium perfringens (Recsei et al., 1984); Lactobacillus 30A (Chang & Snell, 1968); Micrococcus sp. (Prozorouski & Jörnvall, 1975) and Oenococcus oeni (Rollan et al., 1995) showing that HDCs from gram-positive bacteria are pyruvoyl-dependent. Histidine decarboxylase was firstly isolated from wine lactic acid bacteria by Lonvaud- Funel and Joyeux (1994) from a histamine producing strain (O. oeni 9204). Later, Coton et al. (1998) purified HDC to homogeneity and provided important molecular data. The enzyme that decarboxylates histidine to produce histamine is a single polypeptide of 315 amino acids. The gene sequence aided researchers to develop rapid and specific detection systems based on polymerase chain reaction (PCR) to detect potential histamine-producing bacteria from wine (Le Jeune et al., 1995; Coton et al., 1998). The substrate affinity of HDC is directly correlated to the concentration of histidine, with the binding to the active site being favoured as the concentration of histidine increases. On the other hand, histamine acts as a competitive inhibitor of the antiport histidine/histamine at the cell membrane and decreases the HDC activity (Rollan et al., 1995). Lucas et al. (2005) showed that related HDC activity of the Lactobacillus hilgardii 0006 (isolated from wine) is due to the presence of an 80-kb plasmid. In this plasmid the decarboxylase gene was located as part of a four-gene cluster. The same authors suggest that plasmid-encoded histidine decarboxylase could be transferred horizontally and the random distribution of HDC bacteria could be explained by the location of the gene on an unstable plasmid. 16

17 Tyrosine decarboxylase The enzyme that converts tyrosine into tyramine was first investigated in wine lactic acid bacteria by Moreno Arribas and Lonvaud-Funel (1999). The identification was done on the strain Lactobacillus brevis IOEB 9809 isolated from wine. The same strain has been further purified and sequenced (Lucas and Lonvaud-Funel, 2002; Lucas et al., 2003), suggesting Lactobacillus brevis IOEB 9809 as the first well characterized bacterial tyrosine decarboxylase gene. Tyrosine decarboxylase (TDC) enzyme is mainly active in a ph range from 3.0 to 7.0 with an optimum ph of 5.0. Its activity is enhanced by the substrate (L-tyrosine) and the coenzyme (pyridoxal-5-phosphate). This enzyme is also able to use phenylalanine as substrate to produce phenylethylamine which is responsible for its presence in some food products. Recent studies have been able to confirm this dual activity with the expression of TDC enzyme from Enterococcus faecium RM58 in E. coli which showed phenylalanine and tyrosine decarboxylases activities (Marcobal et al., 2004; Marcobal et al., 2006a). All LAB identified as tyramine producers in wine were able to perform malolactic fermentation (Moreno Arribas et al, 2000). This research conducted showed that none of the strains in their study were identified as O. oeni, but all were Lactobacillus brevis and Lactobacillus hilgardii. TDC in Lactobacillus brevis is highly specific for tyrosine (Moreno Arribas and Lonvaud-Funel, 1999). According to the literature no tyramine producing O. oeni has been reported so far in vivo (Moreno-Arribas et al., 2003) suggesting that O. oeni has a low distribution of the metabolic ability to produce tyramine. Only in laboratory medium O. oeni DSM 2025 has shown to produce tyramine (Choudhury et al., 1990). 17

18 Ornithine decarboxylase Ornithine decarboxylase (ODC) is the enzyme that decarboxylates ornithine to produce putrescine. This enzyme has been identified by Marcobal et al. (2004) the first isolated O. oeni putrescine producing strain from ropy Spanish red wine - a discovery that led to the first report of the presence of ODC gene in the genome of O. oeni and detectable by PCR (Marcobal et al., 2004, 2005). Putrescine is considered to be the most abundant BA found in fermented food. In the case of wine many researchers have found putrescine to be present in 100% of the samples analyzed (Gloria et al., 1998; Soufleros et al., 1998; Vasquez-Lasa et al., 1998; Soleas et al., 1999; Moreno & Azpilicueta, 2004; Landete et al., 2005; Bover-Cid et al., 2006; González Marco & Ancín Azpilicueta, 2006; Alcaide-Hidalgo et al., 2007). Despite this, only one strain of wine lactic acid bacteria has been reported in the literature to possess the ornithine decarboxylase gene: O. oeni BIF-I83 (Marcobal et al., 2004). In the same research it is shown that the ornithine decarboxylase gene is rarely present in the genome of O. oeni. Moreover, results showed ornithine to be present at low levels in wine, which indicates that the levels of putrescine will be correspondingly low. A possible explanation for this is the alternative production of putrescine through the arginine deiminase pathway at wine ph (3.2) (Mangani et al., 2005). Most of the strains are able to produce putrescine dominant during MLF such as Oenococcus oeni, although several studies have shown also the genus Lactobacillus to be part of this group (Arena and Manca de Nadra, 2001). The same author reported that Lactobacillus hilgardii isolated from wine was able to produce putrescine from the intermediates ornithine and agmatine. Arginine deiminase A reason for putrescine to be one of the most produced BA in wine is because it can be produced from O. oeni not only through the ornithine pathway but also through the arginine deiminase pathway (Mangani et al., 2005). The same authors described that strains that can produce putrescine (also) possess the complete enzyme system to convert arginine (major amino acid in wine) to putrescine. In this case, the bacterial strain must have all three enzymes necessary to catabolise arginine via the arginine deiminase (ADI) pathway and be active under wine conditions. These enzymes are 18

19 arginine deiminase, ornithine transcarbomoylase (OTC) and carbamate kinase (CK). This process is an association and exchange between strains able to metabolize arginine to ornithine but unable to produce putrescine together with strains that cannot degrade arginine but can produce putrescine from ornithine. However, some O. oeni strains can be deficient on one of these enzymes or the catabolic pathway of arginine is inhibited by the low ph, metabolism still occurs among O. oeni species. This could be explained by an association and genetic transfer exchange among strains that can on the one hand metabolize arginine to ornithine but cannot produce putrescine and strains that on the other hand cannot degrade arginine but can produce putrescine from ornithine. The production of putrescine by metabolic association can occur slowly after the end of malolactic fermentation while on the contrary, conversion of ornithine to putrescine by ODC of a single O. oeni strain was found to occur simultaneously with malic acid degradation at a faster rate (Mangani et al., 2005). The authors concluded that although commercial starter cultures were unable of metabolizing arginine or did not posses ornithine decarboxylase, metabolizes could still occur with indigenous strains leading to an important accumulation of putrescine in wine. Arginine 3 Agmatine Citruline 4 N- Carbamoylputrescine Ornithine 7 Putrescine Figure 1.2: Pathways of arginine metabolism in bacteria. (1) Arginine deiminase; (2) Ornithine transcarbamylase; (3) arginine decarboxylase; (4) Agmatine deiminase; (5) Agmatinase; (6) N- Carbamoylputrescine hydrolase; (7) Ornithine decarboxylase; (8) Anabolic system; (adapted from Arena and Manca de Nadra, 2001). 19

20 Lysine decarboxylase The first oligonucleotides for the PCR detection of lysine decarboxylase encoding genes in foodborne bacteria were identified by De Las Ribas et al. (2006). The primers were designed to amplify the genes coding for lysine decarboxylase by alignment of lysine decarboxylase proteins in two groups: first the Enterobactericeae and second bacteria from the genera Bacillus, Clostridium, Listeria and Staphylococcus. The method has been further extended in order to detect cadaverine producing bacteria in wine Phenylalanine decarboxylase The production of phenylethylamine is often associated with tyramine production in LAB. The explanation for this is correlated with the fact that phenylalanine is also substrate for tyrosine decarboxylase and the production of phenylethylamine results from a secondary reaction (Landete et al., 2007a). According to Moreno-Arribas (2000) and Landet et al. (2007a) the formation of phenylethylamine by wine LAB seem to be very unusual, with Lactobacillus brevis and Lactobacillus hilgardii being the only two species to date associated with high level of phenylethylamine production in wines. No other phenylethylamine decarboxylase enzymes studies have been reported (Landete et al., 2007a). 20

21 Metabolic pathway of decarboxylation The process characterized first for an amino acid decarboxylation is followed by the extraction of the decarboxylated product and addition of the amino acid substrate trough an inner membrane antiporter. Figure 1.3: Biogenic amine biosynthesis pathway in bacteria. aadc (amino acid decarboxylase) ( EFSA 2011). Mollenar et al. (1993) reported the presence of a Histidine/Histamine antiporter (HdcP) in Lactobacillus buchneri ST2A. Descriptions of the cell metabolism of this bacterium have shown coupled reactions of histidine decarboxylase and histidine/histamine exchange to produce a transmenbrane ph gradient (inside alkaline) and an electrical potential (inside negative) such as a proton motive force from a secondary metabolic energy generation. Several authors have identified similar metabolic pathways for other decarboxylases: Wolken et al. (2006) described the tyrosine/tyramine transport mechanism on Lactobacillus brevis, while Soksawatmaekhin et al. (2004) characterized the putrescine and cadaverine transport of protein in Escherichia coli. 21

22 Main scenarios leading to Biogenic Amine formation in wine Substrate availability The availability of free amino acid is one of the necessary conditions for the formation of BA. They are of paramount importance during alcoholic fermentation (AF) as they act as a source of nitrogen for yeasts and can also be metabolized by LAB during MLF (Moreno-Arribas et al., 2009). Amino acid can also influence the aroma composition of the wines. Its presence in wines mainly occurs as salts which are odourless. However, at the ph prevailing in the mouth, amines are partly liberated and their flavour becomes apparent. Amino acid in wine can have a variety of origins. They may arise from the raw material or be produced during the fermentation process (Herbert et al., 2005). Those indigenous to the grape can be totally metabolized by yeasts during the growth phase. Some of them are excreted by live yeast at the end of fermentation or released by proteolysis during yeast autolysis. Others can be produced by enzymatic degradation of grape proteins (Moreno-Arribas et al., 2009) Viticulture conditions Several authors report the presence of amines already in grapes, namely histamine and tyramine (Vidal-Carou et al., 1990) as well as putrescine and other polyamines (Halász et al., 1994; Bover-Cid et al., 2006). Putrescine, cadaverine and spermidine have been found in the pericarp as well as in the seeds of grape barriers (Gloria et al., 1998; Kiss et al., 2006). Biogenic amines seem to be also influenced by many factors during grape production such as the grape variety, degree of maturation of grapes and the soil type (Gloria et al., 1998; Herbert et al., 2005). Regarding the soil, Adams (1991) presented a study in which potassium deficiencies in the soil could be connected to the increase in putrescine concentration in plants. Moreover, nitrogen fertilization treatments can cause an increase in grape amino acid and amino concentrations (Spayd et al., 1994; Soufleros et al., 2007). According to Herbert et al. (2005) grape varieties with higher levels of assimilable amino acid have been found to gain a highest final concentration of BA. 22

23 Fermentation Winemaking normally combines two fermentation processes. First the alcoholic fermentation carried out by yeast and second the malolactic fermentation carried out by LAB Alcoholic fermentation Alcoholic fermentation is defined as the biological process that transforms sugars especially hexoses (glucose and fructose) into ethanol and carbon dioxide among other large number of by-products such as glycerol, butylene, glycol, aldehydes etc. The chemical expression of alcoholic fermentation was first reported by Gay-Lussac (1810) and can be written as follows: C 6 H 12 O 6 2 CO CH 3 CH 2 OH + 2 ATP KJ mol According to this equation, 1 mol sugar gives 173 KJ of which 26% is stored as biological forms useful for living cells and 74% is freed as heat (Staniszewska, 2005). The fermentation proceeds most rapidly at a temperature between 24 and 38ºC, but wines have a much finer bouquet when they are produced at lower temperature and slower fermentation. The wine ancient production was always made by fermentation with indigenous yeast present on the grape when harvested or introduced from the equipment during vinification. This situation has changed when in the late 80s, selected yeast were introduced to perform alcoholic fermentation. Nowadays this is a common practice in most wineries with Saccharomyces cerevisiae strain representing the majority of commercial wine yeast used (Moreno-Arribas & Polo, 2005). Yeast A large variety of indigenous yeast species as well as commercial Saccharomyces cerevisiae strain can grow and perform alcoholic fermentation in wine. Figure 1.4 describes the main biochemical mechanism of yeast metabolism during AF. When the 23

24 concentration of sugars is high such as in the must S. cerevisiae can only metabolize by the fermentation route. The transformation of glucose into ethanol occurs by glycolysis, with the production of pyruvate, which is later transformed in ethanol and carbon dioxide. During the fermentation of grape sugars, the yeast produces from the pyruvate a low concentration of a range of volatile compounds that make up the so-called fermentation bouquet. Among this main group of compounds the best studied are higher alcohols, fatty acids, aldehydes and esters Figure 1.4: Schematic representation of the main biochemical mechanisms of yeast metabolism during alcoholic fermentation (Moreno Arribas and Polo, 2005). Several research studies have been conducted on the formation of biogenic amines by yeast, where most only compared different yeast species and quantified only histamine (Torrea & Ancín, 2002). The level of Biogenic Amine in wine is directly correlated with the amount of amino acid precursors present in the medium (Smit, 2007). Therefore it is important to identify and characterize the evolution of free amino acids during alcoholic fermentation. Free amino acids are of paramount importance since they 24

25 constitute the main components of the nitrogenous fraction of musts and wine. According to the literature they can have two different origins: (1) free amino acids already present in the grape with an increase of their content during grape maturation or (2) being released at the end of fermentation by yeast autolysis (Herbert et al., 2005). The two most abundant free amino acids are arginine and proline and both have glutamic acid as a precursor. About 30% to 85% of the total amino acid content of wine is proline. Proline and its hydroxylated derivative hydroxyproline have been suggested as discriminating indicator among wines made from different grape varieties. This fact arises from the non-metabolization of proline by yeast during fermentation (Moreno- Arribas et al., 1998). At the beginning of fermentation, yeasts use ammonium salt for their development followed by the nitrogen from free amino acids. Namely arginine, glutamic acid, glutamine, aspartic acid, asparagine, threonine and serine are preferentially assimilated. In parallel, by enzymatic process yeasts degrade the proteins to peptides and to amino acids until the fermentation finishes. This leads to the consumption of nearly all free amino acids during the first step of fermentation. In the time between the end of the fermentation and the racking there was an increase in the free amino acid content associated to yeast autolysis (Dizy and polo, 1996). Soufleros et al. (1998), reported that yeast could have an indirect effect on BA production by using and secreting some amino acids during fermentation and autolysis, altering the composition of grape must that would therefore change the concentration of the precursor amino acid that can be used by other microorganisms in further fermentation steps. Many authors reported no remarkable increase in the concentration of BA during alcoholic fermentation. Herbert et al. (2005) and Marcobal et al. (2006b) conclude that the commercial as well as indigenous yeast were not responsible for most of the BA found in wine. On the other hand, the most considerable increase on the BA concentration occurs during both spontaneous and inoculated MLF conducted by O. oeni strain. This fact illustrated in figure 1.5 was been reported by Marcobal et al., (2006b) who together with other authors, considered the MLF as the most important step in the increase of BA concentration, when compared with the contributions by alcoholic fermentation and/or ageing (Cilliers &Van Wyk, 1985; Landet et al., 2005; Alcaide-Hidalgo et al., 2007). 25

26 Figure 1.5: The average concentration for histamine, putrescine, tyramine, cadaverine and phenylethylamine in the seven stages of industrial winemaking: (1) must; (2) after AF; (3) during AF; (4) after MLF; (5) after MLF with the addition of SO 2 ; (6) after 6 months of ageing; (7) after 12 months of ageing; (adapted from Marcobal et al., 2006b). 26

27 Malolactic fermentation Malolatic fermentation (MLF) in wine results in the direct transformation of malic acid (dicarboxylic acid) into lactic acid (monocarboxylic acid) and CO 2 that is catalysed by the malolatic enzyme. This reduction of malic acid to lactic acid cannot be considered from a biochemical perspective a real fermentation, but rather an enzymatic reaction performed by LAB. The LAB associated with grape must and wines belong to four genera: Lactobacillus, Leuconostoc, Pediococcus and Oenococcus (Lonvaud-Funel, 1999). However, in most cases MLF is performed by Oenococcus oeni, since this species can withstand low ph (<3.5), high ethanol (>10 vol.%) and high SO 2 levels (50 mg/l) found in wine (Davis et al., 1986; Wibowo et al., 1985). The most important benefits of MLF are the deacidification of high acid wines that also produces a change in its organoleptic characteristics, since the bitter taste of the malic acid is replaced by the smoother taste of the lactic acid, enhancing the complexity of the wine (Lonvaud- Funel 1999; Moreno-Arribas and Polo 2005) and also releasing other bound volatile aromas (Lerm et al., 2010). LAB cannot grow at the same time as yeast and they multiply only during the final stages of the alcoholic fermentation or just after it (Lonvaud-Funel, 1995). There are many effects and relations in wine which limit their growth. In order to be able to survive and maintain their viability, the bacteria must produce and store energy under adverse wine conditions such as low ph, low temperature, presence of SO 2 and ethanol (Britz and Tracey, 1990; Linskes and Jackson, 1988; Vaillant et al., 1995). The ph of the wine is highly selective; therefore wines with ph values below 3.5 lead to a population mainly composed by Oenococcus oeni (Maicas et al., 2001), while in wines with ph values above 3.5 various species of Pediococcus and Lactobacillus might be present (Henick-Kling, 1988). LAB in wine produce energy from the fermentation of carbohydrates such as the residual sugar which are not fermented by the yeast and in resting cells of O. oeni by the proton force generated by citrate and malate metabolism (Salema, 1996). The homofermentative bacteria (Pediococcus and several species of Lactobacillus) follow the glycolysis pathway, which yields 2 mol of lactate and 2 mol of ATP from 1 mol of glucose. On the other hand, the main gain from 1 mol of glucose through the heterofermentative bacteria (Leuconostoc, Lactobacillus and Oenococcus) is 1 mol of lactate, 1 mol of carbon dioxide (CO 2 ) and 1 mol of acetate and ethanol, generating only 1 mol ATP (Lui, 2002). 27

28 GLUCOSE FRUCTOSE-1,6-P GLUCOSE-6-P 2 x TRIOSE-3-P 6-P-GLUCONATE 2 x PYRUVATE XYLULOSE-5-P 2 x LACTATE TRIOSE-3-P ACETYL-P PYRUVATE ETHANOL LACTATE HOMOFERMENTATION HETEROFERMENTATION Figure 1.6: Schematic description of the major pathways of hexose (glucose) fermentation by lactic acid bacteria. Adapted from Fugelsang and Edwards (2007). Lactic Acid Bacteria Low numbers of LAB present in healthy grapes decline during alcoholic fermentation surviving only the specie (Oenococcus oeni) that are able to overcome the difficult conditions of wine such as high alcohol and SO 2 levels and is responsible for conducting malolactic fermentation. LAB can also be transferred to winery equipment and be present in significant numbers (Wibowo et al., 1985) Whenever the wine ph is higher than 3.5, other species rather than O. oeni can be present in the MLF. Those species such as Pediococcus and Lactobacillus are normally associated with spoilage when growing to levels of 10 6 to 10 8 cell/ml and having an opposing interaction with O. oeni (Wibowo et al., 1985; Lonvaud-Funel, 1999). Several studies have been performed in order to correlate biogenic amines production in wine with the different species of lactic acid bacteria responsible for MLF. A study of Delfini, 1989 made an association of Pediococcus species (Pediococcus cerevisiae) with the production of histamine in wine. Another study conducted by Landete et al. (2007c) showed that other strain of the genera Pediococcus was able to produce a higher amount of histamine. 28

29 The first discover of Leuconostoc to be implicated in biogenic amines production was done by Moreno-Arribas et al. (2003) when Leuconostoc mesenteroides was found to have a high potential to produce tyramine in wine. Commercial O. oeni strains have always been chosen to perform MLF regarding their properties to produce low biogenic amines content. Several authors agree on the fact that O. oeni used as a starter culture was able to produce none of the most important biogenic amines in wine (histamine, putrescine or tyramine) (Moreno-Arribas et al., 2003). The same study compares the inoculation of commercial strains with spontaneous malolatic fermentation and results revealed a decrease of biogenic amines formation when commercial strains were used. Starter cultures were able to eliminate the indigenous bacteria or induce biogenic amines degradation that was produced by undesirable strains (Moreno-Arribas et al., 2003). The use of starter cultures has many advantages in wine such as the control of the beginning of MLF, the quality of wine to be impaired by the development of contaminating bacteria and the organoleptic characteristics of wine to be selected (Moreno-Arribas et al., 2009). Figure 1.7: Oenococcus oeni viewed with phase contrast microscopy at a magnification of 1000x (Fugelsang and Edwards, 2007, photograph provided by WineBugs LLC). 29

30 On the other hand, a study from Guerrini at al. (2002) discovered that O. oeni were able to produce BA in wine and that this ability varies among strains. The study was performed under optimal growth condition in a synthetic medium using 44 strains. Twenty six strains were able to produce up to 33 mg/l of histamine with putrescine and cadaverine being also produced by some strains. This research is in accordance with a previous study conducted by Coton et al. (1998), where the results showed that O. oeni strains isolated from wine have histamine producing capabilities. Conclusions could be archived by many authors suggesting that some LAB strains have the ability to produce different amines (Moreno-Arribas et al., 2000; Guerrini et al., 2002). This characteristic seems to be strain-dependent and not species-specific. This can be explained by the fact that decarboxylase activities are possibly randomly distributed among the different genera of Lactobacillus, Pediococcus and Oenococcus. Gale (1946) mention that LAB strains have more than one amino acid decarboxylase activity under specific culture conditions Aging and storage During wine aging and storage, the concentrations of BA have been reported to increase (Landet at al., 2005; Gerbaux & Monamy 2000; Scheiblhofer 2012). In the particular case of histamine the increase was consistent and noticed only several months after MLF had ended. On the other hand, putrescine and tyramine increased their concentration right after MLF had finished (Herbert et al., 2005) Wine preservatives Sulphur dioxide SO 2 is one of the oldest compounds used in food and beverage industries due to its antioxidant and antimicrobial properties. This chemical preservative became widely used in the wine industry through the addition of sulphite or bisulphite in order to inhibit the growth of unwanted yeast and bacteria. Sulphite can be present in three different forms in the aqueous solution depending of the ph equilibrium (Zoecklein et al., 1995). As showed in figure 1.8, when the ph value is low, sulphite exists mainly as molecular SO 2, at intermediate ph values bisulphite ions are present and with high values of ph as sulphite ions are the main form (Suzzi and Romano, 1992). 30

31 Figure 1.8: The different forms of sulphur dioxide according with the ph values (Einsenman, L., 2001). The relationship between molecular sulphur dioxide, bisulphite and sulphite and free, bound and total sulphur dioxide can be seen below in figure 1.9. A part of the total sulphur dioxide is always bound to acetaldehyde, pigments, sugars and other materials present in the wine. The bound sulphur dioxide does not possess anti-microbial activity and is therefore of little interest for the winemaker. The remaining sulphur dioxide in the wine is present in the free state and it is of particular interest to the winemaker because free sulphur dioxide is the active form of sulphur dioxide. For normal wine ph values (3.0 to 4.0), 94 to 99 percent of the free sulphur dioxide exists in the bisulphite form while ½ to 6 percent is present in the molecular form. The sulphite form of SO 2 can be neglected at this ph once it is inferior to 0.1 percent. (Fleet, 1993). Total SO 2 Molecular Bisulphite Sulphite Stable Unstable SO 2 1- HNO 3 2- SO 3 compounds compounds (acetaldehyde) (pigments, sugars, etc.) Free SO 2 Bound SO 2 Figure 1.9: Sulphur dioxide presented in the different forms (adapted from Einsenman, L., 2001). 31

32 The use of SO 2 should be judicious once the concentration added to the different stages of vinification should be sufficient to inhibit the unwanted spoilage species but not the yeast conducting the alcoholic fermentation or the LAB responsible for the malolactic fermentation (Thomas & Davenport, 1985). Most winemakers add an initial dose of 30 to 50 milligrams of free SO 2 at the crusher (during reductive winemaking) in order to suppress the growth of non-saccharomyces yeast so that S. cerevisiae can proliferate and dominate the fermentation. Table 1.3: Wine ph determines the quantity of free sulphur dioxide needed to produce 0.8 mg/l of molecular sulphur dioxide (adapted from Einsenman, L., 2001). Wine ph Free SO 2 needed for 0.8 mg/l molecular SO (mg/l) Sulphur dioxide has been reported as an effective antimicrobial agent against LAB associated with must and wine and they are more sensitive than yeasts. The sensitivity of LAB to SO 2 can vary among the different strains. Lactobacillus and Pediococcus are more resistant than Oenococcus (Wibowo et al., 1985; Davis et al., 1988). Most commercial wineries maintain at least 0.8 milligrams of molecular sulphur dioxide per litre wine in order to provide good microbial stability. For example, table 1.3 shows that when wine has a ph of 3.3, then 26 mg/l of free sulphur dioxide produces 0.8 mg/l of molecular SO 2. The inhibitory activity of SO 2 is enhanced by the low ph and the high ethanol content (Britz and Tracey, 1990). 32

33 The addition of SO 2 in order to control LAB development depends on the style of wine to be made. Therefore, if wine must undergo MLF, the SO 2 should be added after MLF is finished. Total SO 2 concentrations of mg/l have been reported by Davis et al. (1985) to significantly retard the growth of O. oeni and therefore MLF. Gerbaux & Monamy (2000) reported that the level of SO 2 added to wine after MLF was not enough to stop all the biochemical process and enzymatic activity. Moreover, the high ph of several wines decreases SO 2 efficacy leading to an increase of microbial activity and therefore of BA production in sulphated wines during aging. The judicious use of SO 2 ensures a high quality and microbial stable wine as end product. Nevertheless, there is a worldwide trend in order to reduce SO 2 levels as health risks and organoleptic changes are associated to their use, as well as to find an alternative for SO 2 preservation that would suit producer s and consumer s demands Dimethyl Dicarbonate Diethyl dicarbonate (DEDC) was the well known preservative approved in USA in In 1972, the U. S. Food and Drug Administration (FDA) rescinded the approval of DEDC for food use due to the formation of cancerogenic compound ethyl carbonate. Dimetyl dicarbonate (DMDC) was proposed then as a possible alternative to DEDC as cold sterilizing agent for the use in packaging juices and alcoholic beverages. The common trade name for this product is Velcorin (LanXess GmbH). DMDC is a colourless volatile liquid, with a faint odour, molecular weight of g/mol, specific gravity of DMDC has been described as being able to inhibit the development of yeast and LAB, and can be used as a stabilizing agent in wine in order to ensure a controllable fermentation with the desired inoculated starter cultures. Renouf et al. (2008) and Eder et al. (2011) reported that the dose of SO 2 could be reduced in some wines when applied in combination with DMDC. DMDC mode of action is the denaturation of the fermentative pathway enzymes such as glyceraldehyde-3-phophate dehydrogenase and alcohol dehydrogenase (Porter & Ough, 1982). The predominant reaction of DMDC in aqueous solution is the hydrolysis to form methanol and carbonic dioxide. In alcoholic solutions, DMDC also forms minute amounts of ethyl methyl carbonate that can therefore be used to account for the amount of DMDC applied in sterilizing the wine before bottling. The microbial effect of 33

34 DMDC depends of several factors such as strains of microorganisms present, the initial cell concentration (CFU/mL), the ph, the temperature and the ethanol content (Daudt and Ough, 1980; Ough et al., 1978; Porter and Ough, 1982; Threlfall and Morris, 2002). Terrel et al. (1993) reported that higher alcohol and temperature levels would work synergistically with DMDC in order to reduce the time needed for killing. Tests have shown DMDC at concentrations between 10 and 100 mg/l to be effective as a growth inhibitor for many types of yeasts, while its activity against bacteria is less effective. Both yeast and malolatic bacteria re-growth in bottled wine can be controlled by the judicious use of DMDC and SO 2 (Ough et al., 1988). These two compounds exert additive effects, and their use in combination requires less of both to achieve adequate microbiological control at bottling. At concentrations of 50 mg/l of free SO 2 and 100 mg/l of DMDC, control of bacteria and yeast on bottling wine with ph lower than 3.6 is excellent (Threlfall and Morris, 2002). However, the inactivation depends of the initial cell count, when the concentration of microorganisms exceed 10 6 CFU/mL and the molecular SO 2 is lower than 1 mg/l the maximum dose allowed for DMDC (200 mg/l) is not effective against the most resistant species (Costa et al., 2008). DMDC may be added to wine for one or more of the following purposes: a) microbial stabilization of bottled wine, b) preventing of development of undesirable yeasts and bacteria, and c) blocking the fermentation of sweet and semi-dry wine (Commission Implementing Regulation (EU) No. 315/2012 of April 2012 amending Regulation (EC) No. 606/2009 laying down certain detailed rules for implementing Council Regulation (EC) No. 479/2008 as regards the categories of grapevine products, oenological practices and the applicable restrictions.) Wine lees Another reason pointed for the increase of BA during aging arises from the winemaking practice of aging wine in contact with yeast lees. The process known as yeast autolysis follows alcoholic fermentation and favours the growth of LAB due to the release of vitamins and nitrogen compounds. The released of amino acids are used by LAB as nutrients or energy sources. Most of these amino acids include precursors of BA (Lonvaud-Funel & Joyeux, 1994). The increase of tyramine and putrescine has been reported by Bauza et al. (1995) in wine inoculated with bacteria trough the addition of 34

35 lees. Marcobal et al. (2004) isolated the first ornithine decarboxylase gene responsible for putrescine producing LAB from wine lees. In another study Martin-Alvarez et al. (2006) left the wines in contact with lees for two months after alcoholic fermentation, before aging in barrels and concluded that the concentrations of methylamine and putrescine were high. On the opposite, tyramine concentration was significantly low. This fact could be explained by the consumption of tyramine by residual organisms from lees for the production of carbon skeletons or amino groups. Normally during ageing no culturable cells can be detected. However, Cotton et al. (1998) discover histidine decarboxylase to be active during this phase and therefore leading to the production of BA during ageing. Alcaide-Hidalgo et al. (2007) present a study where the higher mean levels for histamine could be found. Malolatic fermentation has been performed in the wine in tanks followed by ageing in the presence of lees mixed periodically. Results showed an increase of putrescine in wines aged in the presence of lees when compared with stable wines without lees contact and where putrescine remains stable Time Many authors could agree that after an initial increase of BA during storage, a general decrease or stabilization on the BA concentration could be observed (Gerbaux & Monamy, 2000; Landete et al., 2005; Marcobal et al., 2006b). The degradation of BA during ageing is done by oxidase enzymes present in some bacteria (Vidal-Carou et al., 1991; Moreno & Azpilicueta, 2004) Filtration Filtration is a common technique used during wine-making in order to provide stabilization. Physical stabilization prevents the formation of deposits after bottling and microbiological stabilization by eliminating yeast and bacteria that destroy wine taste. The influence of filtration on BA formation has been evaluated by Moreno & Azpilicueta (2004) on a study where filtered and unfiltered wine aged in barrels for 234 days were compared. These authors suggested that the presence of diatomaceous earth normally used as filter material could influence the evolution of biogenic amines during 35

36 ageing by absorbing cationic amino acids and protein on its surface. On the other hand, unfiltered wine presented skin residues that could be rich in amino acids precursors. Yet, the findings results were not significant enough to prove that turbidity could influence the accumulation of biogenic amines during ageing Type of barrel The type of barrel (American, French Allier and French Never oak) did not have any influence in the content of biogenic amines (Moreno & Azpilicueta, 2004) Temperature According to the literature (González Marco & Ancín Azpilicueta, 2006), high and fluctuating temperatures can result on unwanted microbial chemical and enzymatic reactions of wine components and seriously decrease the quality of the final product. However, in regard to wine, it has been found out that storage temperature only has a small effect on amine concentration. According to a study conducted by González Marco and Ancín Azpilicueta (2006), histamine concentration in wine at room temperature (20ºC) was considerably higher than that of wine stored at more extreme temperatures of 4ºC and 35ºC. In a wine stored for a total of 105 days, the formation and degradation of amines in wine mainly took place during the first 45 days of storage. This arises because after alcoholic and malolatic fermentation microorganisms with decarboxylase activity still remain in the wine (Lonvaud-Funel, 2001). Last but not least the use of temperature in order to preserve the final product while stored can be applied (e.g. pasteurization). This process has, however, no preventive impact on the reduction of the amines and decarboxylases enzymes; hence, they are heat stable (Brink et al., 1990) Factors affecting growth of Biogenic Amine producers Grape skin maceration Grape skin maceration facilitates the extraction of nutrients, flavourings and other constituents from the pulp skin and seeds (Moreno-Arribas et al., 2009). The major 36

37 factors influencing this extraction are temperature and duration. The use of cold maceration where grape must is left in contact with grape skin at cold temperature (4ºC) with a short contact time minimizes flavanoide uptake and therefore limits potential bitterness and astringency (Moreno-Arribas et al., 2009). Cold maceration occurs mainly before alcoholic fermentation. However, extended maceration after alcoholic fermentation can also be used at cold temperature to extend the extraction period (Smit, 2007). Results obtained about the influence of cold maceration in biogenic amine formation seem to be contradictory. Some authors like Soleas et al. (1999) found no correlation between the extend of skin contact and the concentration of BA while Bauza et al. (1995) and Martín-Álvarez et al. (2006) suggested that longer maceration could increase the formation of biogenic amines Pectolytic enzymes Pectolytic enzymes may be added to grape must right after crushing in order to enhance the yield of juice extraction. These enzymes can also facilitate wine clarification, extract more grape derived compounds such as phenols and facilitate pressing and filtration (Moreno-Arribas et al., 2009). Addition of pectolytic enzymes to an extent of 2g/100kg grape must was found to have to no influence on the concentration of BA Phenolics The phenolic content is extracted from the grape skin and seeds during maceration and fermentation and extended maceration (Smit, 2007). Phenols are strong antioxidants, affecting the colour and flavour of the wine. They are also responsible for several reactions during fermentation that could inhibit or stimulate the growth of microorganisms. For example, phenols in red grapes, namely anthocyanins, are able to stimulate fermentation while procyanidins in grapes can be slightly inhibitory for fermentation. The influence of phenol contents on biogenic amine formation by wine lactic acid bacteria was first studied by Alberto et al This study put special attention on the effect of phenols in the agmatine metabolism of Lactobacillus hilgardii X1B. According to these authors, agmatine degradation to putrescine does not involve an amino acid decarboxylase. Therefore, a direct correlation could be observed from 37

38 this conversion that was lower in the presence of phenols, with the exception of gallic acid and quercetin. Phenolics are strong antioxidants and the diminution of putrescine formation in the presence of phenols is due to the ability of phenols to protect the cells against oxidative stress (Alberto et al., 2007). Phenolic decarboxylation could also, in accordance with the same study, compete with the enzyme involved in the conversion of agmatine to putrescine (N-Carbamoylputrescine decarboxylase), and therefore block the formation of putrescine. Alberto et al., 2007 could conclude from the study that the presence of phenolic contents was significant enough to cause a reduction on putrescine formation in red wine Wine physicochemical composition Wine influence on the decarboxylase gene expression The presence of HDC in wine does not assure for itself that histamine will be necessarily formed (Cotton et al., 1998). HDC is normally use by the bacterial cell to produce additional energy when the conditions of growth are unfavourable (Konings et al., 1997) Glucose and L-malic acid Several studies present contradictory data concerning the influence of substrate (Glucose and L-malic acid) on the production of BA. Lonvaud-Funel & Joyeux (1994) discovered that histamine production by LAB would increase under a shortage of substrate. On the other hand, Moreno-Arribas et al. (2000) found out that more tyramine was produced by a tyrosine decarboxylase positive Lactobacillus strain in the presence of glucose. A fact explained by the energy provided to support the enzyme activity. However, glucose had no effect on HDC enzyme of O. oeni (Rollan et al., 1995), malic acid was an important figure to activate arginine catabolism by O. oeni and therefore increases the formation of putrescine (Mangani et al., 2005) Lactic acid Many groups of research agree on saying that lactic acid at levels present in wines after malolactic fermentation is able to inhibit histidine and tyramine decarboxylase (Rollan 38

39 et al., 1995; Moreno-Arribas & Lonvaud-Funel, 1999). However, citric acid does not seem to influence ornithine decarboxylase activity (Mangani et al., 2005) Ethanol The influence of ethanol on the decarboxylase activity of BA seems to be connected with the levels of ethanol in wine. Lonvaud-Funel & Joyeux (1994) suggested that HDC activity would be enhanced by levels of ethanol up to 10%, justifying that the conditions for histamine transport inside the cells are more favourable due to the fluidification of the cell membrane by ethanol. On the contrary, Rollan et al. (1995) showed that high ethanol concentration (12% or more) could be an inhibiting factor of the HDC activity by altering the physiochemical properties of the membrane and reduce histidine transport Sulphur dioxide and ph SO 2 added to the must does not affect the concentrations of BA in wine during alcoholic fermentation (Garde-Cerdán et al., 2006). On the other hand, studies about the complete vinification of industrial wines concluded that SO 2 is important to prevent the formation of BA by reducing lactic acid bacteria in wine (Marcobal et al., 2006b). This could also be proved by Vidal-Carou et al., (1990), who found the highest concentrations of biogenic amines in red wines with low levels of SO 2. A direct correlation could be observed on the decrease of tyramine and histamine when the levels of SO 2 were enhanced. The concentrations of BA seem to be also affected by the ph. The presence of a high ph in wine (>3.5) corresponds to a more diverse bacterial micro flora and the growth and survival of decarboxylase positive bacteria becomes more likely (Wibowo et al., 1985; Lonvaud-Funel & Joyeux, 1994; Gardini et al., 2005; Landete et al., 2005; Martín-Álvarez et al., 2006). Studies show that wines with high histamine content (>10 mg/l) had a ph of 3.6 or higher (Cilliers &Van Wyk 1985; Landete et al., 2005). Contrary to this, Gardini et al. (2005) found that at a higher ph and increased SO 2 the concentration of tyramine would decrease, while at lower ph the concentration of tyramine would enhance with an increase of SO 2. 39

40 1.3. Toxicology of Biogenic Amines Many advantages arise from the production of fermented foods and beverages such as enhancing of hygiene, natural preservation, organoleptic quality and nutritive value. Nevertheless, the metabolic activity of the organisms involved may give space for the formation of undesirable compounds such as BA (Spano et al., 2010). Other fermented foods than wine are consumed more often and with higher levels of BA, such as fish, fish products, cheese, beer and fermented foodstuffs (M.H. Silla, 1995). Table 1.4: Biogenic amine content in different food products (adapted from Ladero et al., 2010). BA (mg L -1 o mg Kg -1 ) Food Tyramine Histamine Putrescine Cadaverine Phenylethylamine Vegetables Fish Spinach Ketchup Fresh fish Tuna (canned) Fermented drinks Must Wine White wine Red wine Beer Fermented meat Dry fermented sausages Chorizo Dry cured ham Dairy products Yoghurt - < Feta cheese Ripened raw milk cheese

41 However, the presence of biogenic amines in wine has received more attention because of the undesirable physiological effects in sensitive humans. This, especially when ethanol and acetaldehyde are present enhance the effects on health by directly or indirectly blocking the enzymes responsible for the detoxification of these compounds (Ladero et al., 2010). The human organism easily tolerates low levels of biogenic amines once these are efficiently metabolized by mono- and diaminoxidase in the intestinal tract. Nevertheless, there are differences in the human sensitivity to intoxication by biogenic amines leading to several pharmacological reactions that can take place after excess intake of these compounds (Moreno-Arribas et al., 2009). According to the literature, histamine is the most studied biogenic amine (Hernández-Orte et al., 2008). Figure 1.10: Effect of biogenic amines in the human intestinal tract (Ladero et al., 2010). 41

42 Histamine Synthesis and metabolism in humans The synthesis of histamine is done by pyridoxal phosphate containing L-histidine decarboxylase from the amino acid histidine in the mast cells, basophils, platelets, histaminerfic neurons, and enterochromaffine cells. Histamine can be metabolized by monoamine oxidase (MAO) through ring methylation, by histamine-nmethyltransferase (HNMT) and by diamino oxidase (DAO) through oxidative deamination, depending on its localization (Maintz and Novak, 2007). MAO as well as DAO are important enzymes for the reduction and inactivation of extracellular histamine originated from ingested histamine rich food (Schwelberger et al., 1998; Schwelberger and Bodner, 1998; Klocker et al., 2005). The metabolism of histamine can be inhibited by other biogenic amines that may also be present in wine. Tyramine putrecine and cadaverine favor the passage of histamine from the intestines into the systemic circulation by competing for the binding sites in gastrointestinal tract or by interfering with the catabolism of histamine by saturating the activity of diamine oxidase (Moreno-Arribas et al., 2009). Other compounds such as alcohol, acetaldehyde and anti-depressive drugs are able to cause influence on amino oxidase enzymes given to its lack of specificity (Ten Brink et al., 1990; Straub et al., 1995). 42

43 Figure 1.11: Schematic representation of the histamine metabolism. The biogenic amine histidine is synthesized by decarboxylation of the amino acid histidine catalyzed by L-histidine decarboxylase (HDC) (1). Histamine can be metabolized by extracellular oxidative deamination of the primary amino group by diamine oxidase (DAO) (2), or intracellular methylation of the imidazole ring by histamine-nmethyltransferase (HNMT) (3). Thus an insufficient enzyme activity caused by enzyme deficiency or inhibition may lead to accumulation of histamine. Both enzymes can be inhibited by their respective reaction products in a negative feedbackloop (4). N-Methylhistamine is oxidatively deaminated to N- methyl-imidazole acetaldehyde by monoamine oxidase B (MAO B) (5) or by DAO (6). Because the methylation pathway takes place in the cytosolic compartment of cells, MAO B has been suggested to catalyze this reaction in vivo (Tsujikawa et al., 1999). 43

44 Physiological functions Histamine is normally present in low concentrations in human body and acts both as a local hormone and a neurotransmitter (EFSA, 2011). Mast cells, blood cells and neurons in the brain have been found to contain histamine. The interaction of histamine with specific receptors on target cells modulates a variety of functions such as regulation of body temperature, stomach volume, stomach ph and brain activity. Moreover, histamine is involved in allergic reactions that affect the contraction of the smooth muscle cells, dilatation of blood vessels and thus an outflow of blood plasma into surrounding tissues (Jorgensen et al., 2007; Rangachari, 1992) Toxic response The consumption of histamine by healthy persons leads to a quick detoxification in the human body by amino oxidases or through conjugation. The amino oxidase catalyses deamination of biogenic amines to produce an aldehyde, hydrogen peroxide and ammonia (Gardini et al., 2005). This consumption can, however, lead to symptoms of histamine intoxication as a result of high levels of histamine ingested or if the normal catabolic routes are inhibit or genetically deficient (Ten Brink et al 1990). Symptoms of histamine intoxication are related with effects on the blood vessels and smooth muscles, such as headaches, low blood pressure, heart palpitations, vomiting, urticaria, diarrhoea and oedema (Maintz and Novak, 2007). The factors that can lead to histamine intolerance are correlated with the inactivation of DAO activity, mainly because of genetic predisposition, gastrointestinal diseases or medication which inhibits DAO. Histamine intolerance leads to allergic reactions even when histamine is ingested in small amounts by healthy individuals (Maintz and Novak, 2007). Intoxification is characterized by an incubation period ranging from a few minutes to some hours and the symptoms are normally felt for a few hours (EFSA 2011). Maintz and Novak (2007) suggested dietary histamine to be correlated with the pathogenesis migrane in individuals suffering from DAO deficiencies. Food rich in histamine has been reported in migraine patients to cause headache which could be alleviated by a histamine free diet (Wantke et al., 1993). Several trials use histamine containing alcoholic beverages such as wine to investigate the effect on healthy individuals. No significant results have 44

45 been found, although 12 of the 40 patients with histamine intolerance demonstrate clear symptoms of dizziness, headache, nausea and itching (Lüthy and Schlatter, 1983) Toxic dose The toxic dose in alcoholic beverages is considered to be between 8 and 20 mg/l for histamine (Soufleros et at., 1998), although healthy individuals can tolerate 129 mg/l of histamine taken orally before symptoms occur but only 7 µg administered intravenously (Kanny et al., 2001). Although several studies conclude to have found no correlation between the ingestion of biogenic amines and the wine tolerance (Kanny et al., 2001; Jansen et al., 2003), the same authors agree that wine contain compounds such as ethanol and acetaldehyde that can stimulate the release of histamine in the body. Other toxic effects of histamine in wine have been associated with mouth feel. The identification of high histamine concentration in commercial wines could be done by well trained wine assessors in a study conducted by Rohn et al. (2005). Although no specific taste could be attributed to histamine, irritation of the deep throat and crawling of the tongue has been described by wine tasters for the presence of histamine. Biogenic amines in wine are also important from a financial point of view, since they could cause commercial import and export difficulties. Until today, no legal limits for biogenic amines have been established yet, leading several countries to draw up their own recommendation, especially for histamine. While Switzerland and Austria reject wines that contain more than 10 mg/l, other lower top limits have been recommended in Germany (2 mg/l), The Netherlands (3 mg/l), Belgium (5-6 mg/l) and France (8 mg/l) (Lehtonen, 1996) 45

46 Table 1.5: Dose response of histamine alcoholic drinks in patients after oral administration (adapted from EFSA, 2011). Administration Histamine (amount ingested) Symptoms Number of subjects showing symptoms/ total number of subjects White wine 0,12 4,2 mg No statistically significant effects n.a./20 healthy volunteers Red wine 100 mg No statistical significant effects 0/2 healthy volunteers Sparkling wine 4 mg Dizziness, headache, nausea and itching 12/40 patients with histamine intolerance Tyramine and Phenylethylamine Synthesis and metabolism in humans Tyramine and phenylethylamine belong to the group of trace amines, which are part of a family of endogenous compounds with strong structural similarities to classical monoamine neurotransmitters. These amines are synthesized in humans from their corresponding amino acid (tyrosine and phenylalanine respectively). Their catabolism is mediated by monoamine oxidase (MAO) which encloses two isozymes (MAO-A and MAO-B). These isozymes have different locations and substrate specificity. The preferred substrates of MAO-A are polar aromatic amines and they are mainly present in the intestine, stomach and placenta. Instead, non polar aromatic amines are used as substrate for MAO-B that is mainly present in the brain. These BA are also subjected to N-methylation by N-methyltransferases, generating the sympathetic neurotransmitter, noradrenaline (Broadley, 2010) Physiological functions Although the effects of these low physiological concentrations have been difficult to prove, it has been considered that they serve to maintain the neuronal activity of monoamine neurotransmitters with defined physiological limits. Branchek and 46

47 Blackburn (2003) suggested these amines to have an important influence in the human disorders with the physiological role mainly dependent of their localization. If trace amines are present at the brain level, they will act as indirect sympathomimetics through the release of noradrenalin, which cause vasoconstriction and transient hypertension. On the other hand, at the gastrointestinal level, trace amines cause vasodilatation of mesenteric vascular bed, increasing vascular flow (Broadley, 2010) Toxic response In general the toxic response of these amines in humans is mainly associated with schizophrenia, depression, attention deficit disorder and Parkinson s disease (Branchek and Blackburn, 2003). At the vasoconstriction level, the effect of tyramine and phenylethylamine cause hypertension but also other symptoms such as headache, perspiration, vomiting and pupil dilatation have been reported. The clinical symptoms appear after 30 minutes to a few hours following biogenic amines consumption and normally disappear within a maximum of 24 hours (EFSA, 2011) Toxic dose In healthy men, an average of 500 mg of orally administrated tyramine is required to increase the systolic blood pressure by at least 30 mmhg. The amount required to archive the same level in woman might be lower once they are more sensitive to tyramine (van den Berg et al., 2003). According to some literature (McCabe, 1986), a threshold concentration of 125 mg kg -1 has been indicated to cause an effect in normal individuals while a concentration of 6 mg Kg -1 has been pointed has potential toxic when ingested in combination with MAOIs Putrescine and Cadaverine Synthesis and metabolism in humans Putrescine and cadaverine are polyamines that can be formed either by polyamine biosynthesis or by decarboxylation of the correspondent amino acid (Bardocz, 1995). Putrescine formation requires the free amino acid ornithine and the enzyme ornithine 47

48 decarboxylase while the amino acid lysine is decarboxylated by lysine decarboxylase to form cadaverine. During the metabolism putrescine and cadaverine are oxidized by diamine oxidase to form aldehydes (EFSA, 2011) Physiological functions In humans, putrescine promotes the physiological role of polyamines, and all act together in the regulation of the cell growth, cell division and tumour promotion (Halasz and Barath, 2002). Putrescine is an important component of all mammalian cells and is tightly linked to the proliferation of neoplasms in the gastrointestinal tract (GIT). Löser et al. (1999) reported intestinal and colonic mucosa to have a special demand for putrescine due to high proliferation rates. Although the role of cadaverine is less known, they can also contribute to the physiological role of polyamines. Hölttä and Pohjanpelto (1983) reported the possibility of cadaverine to substitute putrescine in bacterial systems Toxic response Toxic effects resulting from the action of putrescine and cadaverine seem to be less strong that the ones described for histamine and tyramine. Adverse effects such as hypotension, bradycardia, lockjaw and paresis of the extremities have been reported (Shalaby, 1996; Halász and Barath, 2002). Though, the most important feature of putrescine and cadaverine in relation to food is the potential to promote other amines, especially histamine (Hui and Taylor, 1985; Chu and Bjeldanes, 1981) Toxic dose According to the literature (EFSA, 2011), no human studies have been conducted on toxic dose for alimentary putrescine and cadaverine and only few studies have been published concerning animals. In this regard Til et al. (1997) performed a subacute oral toxicity study with Wistar rats where the no-observed-adverse-effect level (NOAEL) was found to be 180 mg/kg body weight day for cadaverine and putrescine. 48

49 1.4. Analytical tools for detection of Biogenic Amines in wine The main reasons for determination of biogenic amines in foods is to monitor potential food poisoning, to verify the food production process (including HACCP) and as a measure for quality control of raw materials, intermediates and end products. Several methods have been developed for the analysis of biogenic amines, with some analyzing only one specific biogenic amine (histamine or tyramine), while others could detect more than one simultaneously (EFSA, 2011). According to the literature, spectrofluorimetric techniques were first used for the detection of biogenic amines (Vidal-Carou et al., 1990). Nowadays, chromatographic methods are the most common method used for detection of several biogenic amines simultaneously. The identification of biogenic amines in wines can be done by molecular methods which detect the presence of decarboxylase-positive microorganisms by targeting the gene coding for these amino acid decarboxylase enzymes Extraction techniques Prior to detection, most analytical methods require some form of extraction of the biogenic amines from the food matrix. The complexity of the food matrix as well as the low concentrations of biogenic amines in food products, and especially in wine, are the critical points in obtaining adequate recoveries of all biogenic amines. According to the literature (EFSA, 2011), reported extraction procedures consist of the use of acids (trichloracetic acid, hydrochloric, perchloric, thiodipropionic, or methansulfonic acids), solvents (petroleum ether, chloroform or methanol) and filtration. In the case of chromatography analysis several cleaning or preconcentration steps must be undertaken prior to detection. Those steps include the use of liquid-liquid extraction (LLE) or solidphase extraction (SPE) when isolating biogenic amines in wines (Costantini et al., 2010) Rapid and semi-quantitative methods Screening methods using selective media The use of selective media was one of the first methods developed for qualitative biogenic amine detection. This method initially created for histamine detection relies on 49

50 the use of a differential agar medium containing precursor amino acid and a ph indicator (purple bromocresol), where an increase in ph as a result of biogenic amine formation can be easily noticed by a change in colour (Smit, 2007). This deferential media has been used to screen lactic acid bacteria producing biogenic amines as well as decarboxylase enzyme activity. Nevertheless, methods based on deferential media have been reported as unreliable due to the generation of false positive and false negative results (EFSA, 2011; Moreno-Arribas et al., 2003). The demand of an isolation step followed by a subsequent growth of the biogenic amine producing organism characterizes this method as time consuming to perform Enzymatic methods Enzymatic methods were first reported in fish to quantify histamine production. The application of these methods to wine and musts resulted in false positives until Landete et al. (2004), developed a direct enzymatic test to be used in wine. This test performed by a sequential activity of diamine oxidase and peroxidase enzymes results in the breakdown of histamine and production of a colour change. Quantification is done by the correlation between optical density and histamine concentration Quantitative methods Chromatography methods Several chromatography methods have been reported as being able to detect biogenic amines in foods such as thin-layer chromatography (TLC), liquid chromatography and capillary electrophoresis Thin-layer Chromatography Thin-layer chromatography was one of the first techniques developed to simply and rapidly detect the presence of biogenic amines in foods (Hálasz et al., 1994). Recently, Garcia-Moruno et al. (2005) have developed a rapid and qualitative TLC method with the particularity of determining bacteria producing biogenic amines in liquid culture media containing the amino acid precursor. Improvements have been done by reducing 50

51 the number of extraction steps from the bacterial supernatants. TLC separates and identifies the fluorescent dansyl derivates of histamine, tyramine, putrescine and phenylethylamine Liquid Chromatography High performance liquid chromatography (HPLC) is the most common method used in laboratories given to its ability to obtain the most information about amines in wine and musts (Lethonen 1996; Vásquez-Lasa et al., 1998; Torrea and Ancín 2002; Marcobal et al., 2005; Hernández-Orte et al., 2006; Gómez-Alonso et al., 2007). HPLC has been subjected to inter-laboratory trails, providing linear sensitives over wide ranges, with detection limit around 0.1 mg/kg. The use of HPLC based assay has been cited as a reference method for histamine in Regulation (EC) No 2073/2005 (microbiological criteria for foodstuffs) and is currently used as a reference technique by the international organization of vine and wine (OIV). Although HPLC requires specialized equipment and skills, the analytical method of choice for official control should be used due to its ability to quantify all biogenic amines. HPLC usually includes pre- or post-column derivatisation and fluorimetric detection of the corresponding derivates in order to reduce preparation and analysis time and to improve resolution of biogenic amine peaks in the chromatogram (Soleas et al., 1999; Marcobal et al., 2005). According to the literature, derivatizing agents such as dansylchloride and O-phythaldialdehyde (OPA) are the most used reagents. In the case of dansyl-chloride, the reaction is carried out before chromatographic separation while with OPA it can be done before or after the column (Vázquez-Lasa et al., 1998; Marcobal et al., 2005). Several other derivatizing reagents such as fluorenylmethylchloroformate (FMOC) (Bauza et al., 1995), aminoquinolyl-nhydroxysuccinimidyl carbamate (AQC) (Hernández-Orte et al., 2006) and diethyl ethoxymethylenemalonate (DEEMM) (Gómez-Alonso et al., 2007) are also used. Of all these derivatizing agents, OPA is the most use due to its rapid one-step derivatization and the possibility of automating the reaction, which increases the reproducibility of the analytical methods (Moreno-Arribas et al., 2009). 51

52 Capillary Electrophoresis The main advantages of using capillary electrophoresis (CE) methods on biogenic amines analysis relies on the short analysis time, high separation efficiencies and reduced consumption of reagents. The main drawback of this method is the lack of sensitivity which can be overcome by the coupling the CE to mass spectrometry (MS) detection instead of UV-detection. The use of CE with different detection methods has been reported by Önal (2007) in the detection of biogenic amine in foods. Kováks et al. (1999) reported CE application to several food matrices including wine. Kvasnicka and Voldrich (2006) reported a CE method that used conductometric detection and which requires no derivatisation or sample cleaning steps. This direct method is sensitive and can detect the present of biogenic amines in foods and wine in shorter time. In comparison with HPLC, this method has the direct detection as advantage, once HPLC requires derivatisation of the amine due to aliphatic biogenic amines containing no chromophores that absorb significantly in the UV visual region. Moreover, a high performance capillary electrophoresis (HPCE) method exists to determine biogenic amines in wine and other foodstuffs (Kóvacs et al., 1999) Polymerase Chain Reaction Molecular methods such as PCR and DNA hybridization have become important methods in the detection of biogenic amines and biogenic amines producing lactic acid bacteria. These methods archived an early and rapid detection of the mentioned bacteria which is an important feature for preventing the accumulation of biogenic amines in wine and other food products. Although PCR technique cannot determine quantitative or qualitative amounts of biogenic amines, it can be used to estimate the potential risk of amine formation (Smit, 2007). PCR techniques have been developed to detect bacterial amino acid decarboxylase gene in a rapid, sensitive, simple and accurate way. In comparison with traditional methods, molecular methods are independent of the culture conditions what leads to a detection of biogenic amine producer even under conditions where the lactic acid bacteria had lost the ability to produce biogenic amines (Lonvaud-Funel and Joyeux, 1994; Izquierdo-Pulido et al., 1997). 52

53 The fact that some DNA sequences of selected genes are highly conservative allows PCR application to detect specific genes in different organisms. Some conserved genes regions that code for histidine-, tyrosine-, ornithine-, and lysine decarboxylase, have been detected in different bacteria leading to the design of specific oligonucleotides (Landete et al., 2007b). In order to detect histidine producing LAB, Le Jeune et al. (1995), created several oligonucleotide primers (CL1, CL2, JV16HC and JV17HC) based on the comparison of the nucleotide and amino acid sequences of Lactobacillus 30a and Clostridium perfringens and the amino acid sequences of Lactobacillus buchneri and Micrococcus sp. The primers allowed for amplification of a HDC gene fragment using DNA of these bacterial strains. Improvements of the reduction time of these tests have been done by Cotton et al. (1998) by applying them directly in wine samples. Later, Landete et al. (2005) modified the original primers described by Le Jeune at al. (1995) to identify the presence of O.oeni histamine producing strains. The new primer set composed by CL1mod/JV17HC was able to amplify all histamine producing O. oeni strains present in the PCR test. A similar approach has been taken by Lucas and Lonvaud-Funel (2002) to detect tyramine producing bacteria mainly belonging to the genera Lactobacillus and Enterococcus. These authors designed a degenerated primer set (P2-for/P1-rev) to detect tdc gene fragments in L. brevis strains present in wine. Another set of primers has been developed by Marcobal et al. (2004) in order to identify ornitine decarboxylase gene (odc) in the putrescine-production O. oeni RM83 strain. Further research from the same authors lead to the development of a method useful for the detection of putrescine production bacteria present in a wine bacterial collection. The identification of lysine decarboxylase gene using a PCR assay has been first described by De las Rivas et al. (2006) in order to determine the presence of cadaverine producing microorganisms in food. Multiplex PCR assays provide a useful technique for routine detection of bacterial strains that are potential producers of histamine, tyramine, putrescine and cadaverine in wine. The main advantage of this technique is the detection of all target genes at one time in the same PCR assay, resulting in the reduction of reagent s quantities and labour costs. Multiplex amplification has been first developed by Marcobal et al. (2005) and 53

54 later improved by De las Rivas et al. (2005) that include primers for the detection of Pyridoxalphosphate dependent histamine decarboxylases for gram negative bacteria. The fact that molecular methods are becoming more reliable, fast and an effective tool in amine detection media constitutes an important alternative for the identification and detection of biogenic amine producing bacteria (Landete et al., 2007b). Figure 1.12: Multiplex Polymerase Chain Reaction amplification of wine bacteria producing histamine, tyramine and putrescine. Sample (1) containing DNA fragments from histamine and putrescine producer Lactobacillus 30a and the tyramine-producer Lactobacillus brevis CECT 5354, has been amplified by multiplex PCR. Negative sample (2) without biogenic amine producer bacteria. On the right side a DNA marker (λ HindIII/EcoRI) is included. Identification of the fragments on the left side (Landete et al., 2010). 54

55 1.5. Control of Biogenic Amines production As mentioned before, biogenic amines in food are mainly produced by amino acid decarboxylase positive organisms. Accumulation of biogenic amines in foods and special in wine permits the microbial growth and activity during manufacturing, handling and storage. The final contents of the different amines in foods are directly correlated with the nature of the product, the microorganisms present and the environmental conditions. On the whole, difficulties have been found in characterize each technological factor effect on aminogenesis during food fermentation, ripening and storage due to the complexity of the aminogenesis process (Smit, 2007) On the overall process of biogenic amine control, two main approaches are considered. First of all the evaluation and monitoring of the hygienic quality of the raw materials and production process in order to limit the contamination of fermented food products with aminogenic microorganisms. Secondly, specific production techniques are implemented in order to prevent the growth and minimize the decarboxylase activity of microorganisms with aminogenic potential (EFSA, 2011). According to the literature, mostly measures used to control biogenic amines in food are focused on the food processing level, including the handling of raw materials and the fermentation process as they constitute the most important factors for biogenic amines accumulation in fermented foods. The storage is also considered but to a less extend once it is dependent of the above mentioned previous stages (EFSA, 2011). Regarding wine raw materials, suggestions have been made on regulating the soil parameters and the vine nutritional status, as measures to control the accumulation of biogenic amines in grapes. This accumulation in musts or wines could also be controlled by reducing practices that lead to an increase of amino acid extraction such as grape skin maceration or lees contact (Smit, 2007). Bodmer et al. (1999) reported the application of Hazard-Analysis and Critical-Control-Points (HACCP) concept to the multistep process of wine production with special focus on the histamine formation. They were motivated by an unacceptable histamine content in wine, which had as raw material fresh and good quality grape juice with almost neglectable histamine content. The HACCP concept has been developed in order to identify risks and hazards during production and storage of foods which could lead to reduce the quality of products or even harm the consumers. The HACCP system constitute the strategy for improvement 55

56 of good quality and its composed by several steps: execute hazard analysis, determine critical control points and control limits, building up monitoring systems, determine corrective actions, establishment of verification criteria and documentation (EFSA, 2011). HACCP is also part of the DIN EN ISO 9000 quality management system. During the fermentation apart from the bacteria present that can contaminate the process the own microorganisms responsible for fermentation can also show aminogenic activity. Furthermore, fermentation is also characterized by several processes such proteolytic activities, yeast lysis and acidification, that increase the availability of precursor free amino acids and favours the decarboxylation reaction (Ten Brink et al., 1990). Presently the use of starter cultures has been reported as the most common and reliable practice in order to control the fermentation process in wine. The microorganisms used as starter cultures must be selected appropriated for each type of product, considering not only their technological competences (influence in organoleptic characteristics and competitiveness) but also safety requirements such as the inability to produce biogenic amines. Starter cultures interact between different background microbial populations and therefore, they can influence the biogenic amine concentration by the competitive suppression of amine producing bacteria, limiting their growth and inhibiting their activity. This results in a dominant microbial group free of the potential to form biogenic amines. O. oeni is a common example of LAB that occurs naturally in wine and is the dominant bacterial species found during the MLF (Smit, 2007). Despite of that, histamine accumulation in wine has also been associated with O. oeni (Simona et al., 2002). Husnik et al. (2006) reported several indigenous lactic acid bacteria, including O. oeni being able to produce BA during MLF. These authors proposed the application of a genetically engineered malolatic wine yeast strain capable of the complete degradation of L-malic acid in wine. This measure would allow the yeast to prevent the formation of biogenic amines by excluding the need of LAB to perform MLF. Nevertheless, the unacceptance of the public for genetically modified organisms in wine producing countries renders this option as unavailable (EFSA, 2011). In another study Leuschner et al. (1998) considered the option to use some microorganisms, among them LAB, which had the capacity to degrade biogenic amines by amine oxidase enzyme activity. Unfortunately, most of these microorganisms are 56

57 restricted to aerobic bacteria which are of limited use in fermented foods such as wine which needs an anaerobic environment. Although the inoculation of O. oeni starter cultures helps to reduce the accumulation of BA in wine, the co-inoculation of O. oeni starter cultures together with alcoholic fermentation leads to an improvement to curb biogenic amines formation even more than normal post alcoholic fermentation inoculation for MLF (Van der Merwe, 2007). Other factors exist that constitute a viable option to control the potential problem of biogenic amines by inhibiting the growth of decarboxylase positive indigenous bacteria and other possible contaminating organisms. Sulphur dioxide is known to inhibit lactic acid bacteria in wine due to its antimicrobial properties. The antibacterial activity of SO 2 is ph dependent and will decrease when the ph increases (Smit, 2007). Lysozyme is an enzyme responsible for the lysis of the cell walls of Gram-positive bacteria including wine lactic acid bacteria and the use of this enzyme can successfully lead to a delay or inhibition of the LAB growth, especially when used together with SO 2 (Delfini et al., 2004; Ribérau-Gayon at al., 2006). The use of antimicrobial peptides (bacteriocins) produced by some strains of LAB has also been considered for the control of biogenic amines. Rojo-Bezares et al. (2007) explored the synergetic effect of nisin and metabisulphite on growth inhibition of wine spoilage LAB for wine preservation but its application in wine has not yet been authorized. The use of natural phenolic compounds as antimicrobial agents to control the growth of LAB in wine has also been reported (García-Ruiz et al., 2007). 57

58 2. Experimental overview The aim of this work was to examine the effect of different selected lactic acid bacteria (Oenococcus oeni) on amine generation. Amines, amino acids and different volatile compounds were analyzed in an attempt to explain this amine generation. Industrial scale wines obtained from a mixture of different grape varieties (Cuvée) were used. MLF took place in a favourable environment, with addition of SO 2 in order to reduce the activity of natural microbiota before inoculating bacteria of O. oeni species. All malolactic fermentations were conducted in 30 L glass balloons. LAB isolation: Storage cell culture (-70ºC) Selective agar media Microscope 16s PCR Wine (HBLA Klosterneuburg) Wine preparation Velcorin + SO 2 Selected starter cultures Wine inoculation Malolactic fermentation Chemical and sensoric analysis 58

59 3. Materials and Methods 3.1 Material Media A commercial Man Rogosa Sharpe (MRS) Broth modified, Vegitone (Fluka # 38944) was used to grow and isolate LAB from Austrian wines. A specific modified MRS media for isolation of Oenococcus oeni was used (table 3.1). Table 3.1: Different Man, Rogosa and Sharpe media used MRS medium with tomato juice MRS Agar Yeast extract 5.00 g/l Yeast extract 5.00 g/l Tryptone g/l Tryptone g/l Diammoniumcitrate 3.50 g/l Diammoniumcitrate 3.50 g/l MgSO 4 x 7 H 2 O 0.20 g/l MgSO 4 x 7 H 2 O 0.20 g/l MnSO 4 x H 2 O 0.05 g/l MnSO 4 x H 2 O 0.05 g/l Tween g/l Tween g/l D/L Malic acid 5.00 g/l Glucose g/l Glucose g/l Fructose 5.00 g/l Fructose 5.00 g/l Tomato juice ml/l Tomato juice ml/l Agar g/l ph 4.8 ph 4.8 Tomato juice (Sweet Valley biological obtained from the local supermarket without any preserving agents) was autoclaved separately and added after autoclaving in order to avoid Maillard reactions. The ph of the media and tomato juice were adjusted to ph 3.8 before autoclaving. After autoclaving and under the laminar 10% V/V ethanol was added to the MRS Broth. 59

60 Buffers Several buffers were used within the practical work and are listed here: Citric acid-phosphate buffer, ph 7.0, 20 mm g citric acid g sodium phosphate dibasic dihydrate Citric acid-phosphate buffer, ph 5.5, 200 mm 1.06 g citric acid 2.58 g sodium phosphate dibasic dihydrate 50x TAE 242 g/l Tris 57.1 ml/l glacial acetic acid 100 ml/l EDTA (0.5 M, ph 8.0) Isolation of Lactic Acid Bacteria Lactic acid bacteria have been isolated from wines of Austrian wineries and identified. They were enriched in MRS Broth with Vegitone (Fluka # 38944) and isolated on MRS agar plates with 200 mg/l cycloheximide. In order to storage cells for a long period, they are first harvested by centrifugation (2 ml) and aseptically transferred to a 1.5 ml Eppendorf tube with 1 ml of MRS media with 20% v/v glycerol added. The pellet is resuspended and then frozen at -70ºC at the Institute of Food Biotechnology, Muthgasse 11, 1190 Wien. A total of 531 strains have been examined and characterized according to the genetic of the microorganisms. The activity of the isolated single colonies was checked with all strains present in the cell collection belonging to the genus Oenococcus, Lactobacillus and Pediococcus. All the five selected strains used in this experiment belong to the species Oenococcus oeni. Table 3.2: Origin of the selected strains Strains 551b 78a SK3 α-enoferm 433a K1 Origin HBLA St. Laurent wild ferment Andreas Schreiner, Weingut Sonnenmulde GOLS Chardonnay in BSA (slow wild MLF ferment) Herrenwald. Biostart Bianco SK3 Comercial strain from Lallemand Cabernet / Gumpoldskirchen WBS Krems wild MLF Control isolated from HBLA Klosterneuburg cellars 60

61 Inoculum preparation The five selected bacteria (551, 78, 433, SK3, 433, K1) with the exception of the commercial strain (α-enoferm) have been cultivated during a approximated period of 30 days at the Institute of Food Biotechnology, Muthgasse 11, Wien. Systematic reinoculations of the cells have been done in MRS broth in order to produce enough biomass. The colonies of the interested bacteria were first inoculated in 10 ml tubes and grown in the 25ºC room. After harvesting, the 10 ml test tubes by centrifuging a further inoculation step was performed on 250 ml MRS broth glass bottles. The volume of media was then increased after growth (broth in 1L bottle) and further cultivated into 1L media at 20ºC 14 days. At this stage, the cells were harvested by centrifugation and washed 3 times with isotonic salt solution (0.9% NaCl). Figure 3.1: Cultured strains before inoculation into the wine Wine preparation The experiments were carried out in two different cellars. The use of different conditions is here relevant for the final wine product. Both institute cellars belong to the Höhere Bundeslehranstalt (HBLA) und das Bundesamt für Wein und Obstbau in Klosterneuburg. Institute of Oenology White wine (700L) with a total concentration of 40 mg/l SO 2 was filled into 18 glass ballons (35L). Before adding the inoculums all the balloons were sterile with 200 mg/l 61

62 velcorin. Balloons were inoculated with five different O. oeni strains (three balloons per strain) with the exception of one commercial strain balloon (enoferm alpha, Lallemand) and two negative controls. The commercial malolactic fermentation cultures added to wine during small scale vinifications were inoculated according to the instructions of the manufacturer at the maximum recommended value all the other triplicates were inoculated with the cell suspensions grown in MRS. Institute of Chemistry White wine (400L) with a maximal amount of total sulphur dioxide (40 mg/l) was filled into 18 glass balloons (20L). Five selected strains were inoculated in triplicate one commercial strain (enoferm alpha, Lallemand) and two controls. The sterilization of the wine was done by means of velcorine (200mg/L) addition. Figure 3.2: The wineries where the experiments took place. Institute of Chemistry (left) and Institute of Oenology (right) Sampling Representative samples for FTIR analysis were drawn in sterile sample vials of the wine during malolactic fermentation. After four months of ageing time, wines were sample for the analyses of biogenic amines and also FTIR spectroscopy. Samples of all wines were taken after malolactic fermentation for amino acid analyses. 62

63 3.2. Methods Several methods were used in order to select the desired colonies for the experiment Microscope Wet Mount: Take from the small tube an bacterial sample and put it on the microscope slide; and examine it under the microscope with a 600 magnification. Dry Mount: With a bacterial oase smear 1 colony from the agar plate on the slide on a water drop. Let it dry and fix it with a flame 7 to 8 times. Add contrast tinction (safranin) and rinse off the excess of tinction and dry. Pour oil and examine under the microscope with 100x oil immersion objective Molecular methods DNA preparation Microbial pellets from cell cultures were prepared and the DNA was extracted using the following technique: TWO STEP METHOD 1.5 ml bacterial culture was harvest and centrifuge (spin down 5 minutes at 1500 rpm in an Eppendorf micro centrifuge). The microbial pellet was suspended in 300 µl STET buffer (8% w/v Sucrose, 5% w/v Triton X-100, 50 mm EDTA sodium salt,50 mm Tris HCl ph 8.0, sterilize with filter, store at 4ºC) and then frozen over night. Lysozyme 0.2mg (10 μl from STOCK solution 10 mg/ml STET) was added to the pellet mixed with the vortex and incubate with shaker at 37 ºC for 30 minutes. After 0.1mg Proteinase K (10 µl from STOCK solution 10 mg/ml STET), mixed and incubated with shaker at 50 ºC for 30 minutes. Heat for 5 minutes at 99 ºC on a heat block and then cooled down to room temperature. The DNA was precipitate by the addition of 240 µl of isopropanol. The mixed sample was centrifuged at 1500g for 15 minutes at 20 o C. The DNA pellet was washed with 70% ethanol (add 240 µl ETOH 70% to pellet, mix with pipette) and centrifuge at 1500 g for10 minutes at 4 ºC. The pellet was air dried and re-suspended in 500 µl of sterile TE buffer for PCR amplification of the 16S ribosomal DNA sequence. For long term storage of the DNA stock solutions a STET buffer with 30% Glycerol is used, and stored in aliquots at 4 ºC or -20 ºC. 63

64 Polymerase Chain Reaction ARDRA Identification of wine LAB following Rodas et al. (2003): Amplification and restriction analysis of 16S-rDNA gene Complete 16S rrna genes were amplified as described by Rodas et al. (2003) for the 16S-ARDRA technique. The primers used for complete sequencing of the 16S rrna gene were: AGA GTT TGA TCC TGG CTC AG (PA) and AAG GAG GTG ATC CAG CCG CA (PH). DNA amplification was carried out in 50 μl PCR mixture containing 25 μl of Red Taq polymerase mix, 1 μl of each primer, 20 μl of H 2 O and 3 μl of the cell suspension (final dilution of 1:2000). PCR was performed in a Biometra T3 Thermal cycler (University of Natural Resources and Applied Life Sciences, Vienna). Each cycle consisted of an initial denaturation time of 3 min at 94 C followed by 35 cycles of amplification comprising a denaturation step for 30s at 94 C, annealing at 56 C for 30s, and extension at 72 C for 1 min. Reactions were completed with 5 min elongation at 72 C. PCR products were resolved by electrophoresis in 0.8% (w/v) Roti Agarose in 0.5 TBE (45 mm Tris-HCl, 45 mm boric acid and 1 mm EDTA ph 8.0) gels, stained with ethidium bromide (0.5 μg ml -1 ). A constant voltage was applied to the system and fragment separation was performed using the two-phase program. The images were digitalized with GelDoc 200, BioRad (University of Natural Resources and Applied Life Sciences, Vienna) and the DNAs were quantified using 1Kb DNA Mass Ladder from New England Biolabs (NEB) as molecular weight standard. Restriction Analysis Restriction of the amplified fragment was carried out overnight at 37 C in 20 μl volume containing 1µL BfaI (New England Biolabs, 5U/µL), 2 µl buffer 4 (New England Biolabs, USA) and 17 µl PCR-product. Samples were further analysed and digested with MseI restriction enzyme. Again, 20 µl reaction volume was made containing 0.3 µl MseI (New England Biolabs, 10 U/µL), 2 µl buffer 2 (New England Biolabs USA), 0.2 µl BSA (New England Biolabs) and 17.5 µl PCR-Product. Restriction fragments patterns were analysed in 2.0% (w/v) Roti Agarose in 0.5 TBE using 100 bp ladder (New England Biolabs, USA) for samples digested with BfaI and 50 bp ladder (New England Biolabs, USA) for samples digested with MseI. After performing electrophoresis at 80 V during 1.5 hours, samples were digitalized with a Gel Doc (Bio Rad, Gel Doc 2000 software Quality One 4.3.0). 64

65 Chemical methods High performance liquid chromatography (HPLC) The analyses of biogenic amines and amino acid contents in wine were performed by HPLC based on the method described by Alberto et al. (2000). Samples were diluted 10 times and filtered through a 0.22 µm syringe filter prior to derivatisation and column injection. 200 mg of o-phtaldialdehyde (OPA) (Sigma, Germany) has been used as derivatizing reagent, dissolved in 9 ml methanol, 1 ml 0.1M. Sodium tetra-borate (ph 10) and 160 µl 2-mercaptoethanol. The 25 µl of the diluted sample reacted with 25 µl of the derivatizing reagent for exactly 45 seconds and 25 µl of this solution was injected immediately thereafter. The derivatisation process was automated by the use of an autosampler. Fourier Transform Infrared spectroscopy (FTIR) analysis The apparatus used is a WineScan FT 120 type from the company Foss (DK Hillerod). The FTIR is a very recent method of wine analyses whose parameters are not determined with the help of a specific chemical reaction but indirectly over the infrared spectrum of a wine sample against standardized spectral data. The spectrum used for the wine analyses is in a wavelength range of 2.5 to 100 µm and thus belongs to the middle part of the infrared wavelength range. Compared to the classical infrared spectroscopy, the peculiarity of the fourier transform infrared spectroscopy is that the measurement is done with the help of an infrared meter in polychromatic infrared light and that the infrared spectrum is only reckoned over the recorded interference pattern by means of the fast fourier transformation. The big advantage of this method over the classical dispersive measurement in monochromatic infrared light which changes the wave length gradually is that less time is required. Furthermore, the signal-to-noise ratio is way better at FTIR (Baumgartner et al., 2001; Eder and Brandes, 2003). After transforming the interference pattern into the corresponding spectrum those wave lengths are picked which are most appropriate for determining the substances to be investigated; the absorbance is evaluated by a neuronal network. The fact that the evaluation of a neuronal network is done leads to the fact that the device needs to be trained with known comparison samples whose ingredients are determined classically. The current FTIR devices for wine analysis are already equipped with a basic calibration but are additionally trained with practical samples that are as similar as 65

66 possible to the future samples like that a higher precision can be reached. On top of that, the results are continuously validated by the measurement of known standard samples so that deviations can be detected as soon as possible. Wine Scan FT 120 is an analysis device for the control of vinification and of the final products; it can determine several ingredients within a time span of 30 seconds to two minutes. Only a small quantity of samples (approximately 50 ml for a duplicate determination) and nearly no consumables are needed. These together with the easy handling, the quick sample preparation (samples only need to be pre-filtered), the high repeatability and comparability are the great advantages of this analytical method. The disadvantages are the high costs of equipment approximately 70,000 Euro and the complex calibration that needs to be done several times. Moreover, this method has a comparably low sensitivity at detection limits in the range of 0,1-0,2 g/l and depends relatively strong on the matrix. Therefore, the ingredients that can be reliably detected have to be found in quite high concentration in the samples (Baumgartner et al., 2001; Eder and Brandes, 2003). Carrying out the FTIR measurement with the WineScan FT (Foss) Approximately 50 ml of the sample were taken and filtered with a pleaded filter (Ø 150 mm). The sample was put on the specimen wheel of the WineScan. Young wine analysis calibration program was chosen at the computer. By means of a peristaltic pump the sample was guided through the temperature unit, warmed up to the analysis temperature and guided to the measuring cell. 30 ml were needed for each measurement. Two measurements were carried out and the mean score was calculated. The following parameters were determined at the young wine: density, alcohol, reducing sugars, fructose, glucose, titratable acidity, ph, volatile acidity, tartaric acid, malic acid and lactic acid. Must and wine samples were filtered using filter paper with a grading 20 to 25 µm and a diameter of 185 mm (Shleicher and Schuell, catalogue number ). All the wines composed by different strains have been scanned with FTIR before and during malolactic fermentation until its completion (taken as the point when the malic acid concentration was equal to or lower than 0.3 g/l). The main goal of these scans was for routine monitoring and evolution of wine chemical compounds (particularly malic acid 66

67 and lactic acid concentrations). Final scanning was been performed after an ageing period of four months. Triplicate scans were obtained for each sample Panel tasting Based on the three basic criteria look, smell and taste, the wine can be divided into a variety of single impressions and so it can be characterized in more detail. Smell: Pure tone Intensity + pure, clean - musty, impure, foul, faulty + sealed, decent, unobtrusive, frothy, distinct, aromatic, intense, accentuated bouquet tender - no smell, neutral, small bouquet, perfumed, intrusive Overall character: winy, flowery, fruity, velvety, strong Taste: Acid + fresh, classy, steely - flat, hard, pointed, sharp, astringent, immature Sweetness + dry, semidry, decent sweetness, clear sugar residue, extra sweet, noble sweet - crude, sticky, sugar water Tannin + velvet, soft tannin, round, slightly bitter, bittersweet - bulky, astringent furry, rough, bitter Harmony + simple, true, clean, palatable, balanced, complex, elegant - small, disharmonious, angular, rough, crude, bland Age + young, strong, juvenile, built out, reap - cider, tired, limp, oxidative, no peak Body + fine, tender, elegant, dense, strong, rich in extracts, corpulent - Empty, slim, weak, crude Alcohol + light, moderate, strong, powerful, heavy, forceful - poor, slim, burnt, spirit Consequences: quality faulty-weak-mediocre-good-very good-excellent Ripeness: immature-juvenile-ripe-on the peak-degrading-degraded (Steidl, 2001). 67

68 Duo-Trio Test From 3 prepared samples: one is the reference (R), the others are coded. One of the two coded samples is identical to the reference sample. The participants are asked to identify the right sample Quantitative descriptive analyses This method tries to objectify and standardize the opinions of the subjects as far as possible. Therefore, following process is applied: (summarized after Quandt, 1999) Use of open, unstructured rating scales. Use of selected and product specific trained testers who determine not only the examination but also the descriptive attributes. Prevention of hedonistically ambiguous attributes such as harmonic, heady or aromatic. Preparation and completion of the tasting by a trained administrator. Repeated examination of the samples whereas the tester should have the closed testing cabins at his disposal. The use of statistic evaluation methods in order to capture the samples and effects influenced by the taste as well as display of the results in form of spider plots. Fruit less more Acidic less more Butter less more Phenolics less more H 2 S less more Harmony less more 68

69 Figure 3.3: Statistical wine tasting tool in order to evaluate the wine according with different defined aromas. Fruit 10 8 Harmony Acid H2S Butter Phenolics Figure 3.4: Sensory spider plot diagram for wine. 69

70 4. Results 4.1. Bacterial growth The first step of this work was to identify lactic acid bacteria (LAB) isolated from Austrian wine with the target to find bacteria of the genius Oenococcus spp. for further investigation. Isolated bacteria from the wine cell collection from Austrian vintage 2007 with ellipsoidal to spherical morphology that usually occur in pairs or chains were firstly analyzed. Selected bacteria from the cell collection were streaked on MRS agar plates and two phenotypic similar colonies of each plate were analyzed. Microscopic observations together with 16S-ARDRA method were used to select the desired strains S Amplified Ribosomal DNA Restriction Analysis Method DNA from a total of 531 LAB wine isolates was extracted and bacteria were isolated by amplifying a 1500 bp part of the 16S rdna using universal primers for that sequence, following the 16S-ARDRA method developed by Rodas et al. (2003). This sequence was further digested with BFaI and MSeI. Some LAB showed a similar restriction pattern and were therefore put into groups. In order to identify the samples, the results were compared with experimental restriction clusters of the publication by Rodas et al. (2003). From this cell collection a total of 90 strains have been identified as O. oeni. From those 90 strains, five have been selected for the experiment. The following figures 4.1 and 4.2 indicate some of the electrophoresis pictures after BFaI and MSeI digestion. Observation from the gel electrophoresis, digested with BfaI show two typical bands for O. oeni between 500 and 600 bp. The third and fourth bands are correspondently at 200 bp and 100 bp. Further identification with the restriction enzyme MseI is performed in order to insure the accuracy of the results when comparing the strain with the database. Results show one typical band at the size of 600 bp and all the other three below 300 bp. 70

71 Marker Marker 600 bp 500 bp 200 bp Figure 4.1: Restriction pattern after BfaI digestion isolated from Austrian wines. The marker is a 100 bp standard provided by New England Biolabs. Lanes 1 to 6 have been all identified as Oenococcus oeni according with Rodas et al. (2003) Marker 700 bp 250 bp 200 bp 150 bp Figure 4.2: Restriction pattern after MseI digestion isolated from Austrian wines. The marker is a 50 bp standard provided by New England Biolabs. Lanes 1 to 6 have been all identified as Oenococcus oeni according with Rodas et al. (2003). 71