Bacterial spoilage of wine and approaches to minimize it
Abstract
Bacteria are part of the natural microbial ecosystem of wine and play an important role in winemaking by reducing wine acidity and contributing to aroma and flavour. Conversely, they can cause numerous unwelcome wine spoilage problems, which reduce wine quality and value. Lactic acid bacteria, especially Oenococcus oeni, contribute positively to wine sensory characters, but other species, such as Lactobacillus sp. and Pediococcus sp can produce undesirable volatile compounds. Consequences of bacterial wine spoilage include mousy taint, bitterness, geranium notes, volatile acidity, oily and slimy-texture, and overt buttery characters. Management of wine spoilage bacteria can be as simple as manipulating wine acidity or adding sulfur dioxide. However, to control the more recalcitrant bacteria, several other technologies can be explored including pulsed electric fields, ultrahigh pressure, ultrasound or UV irradiation, and natural products, including bacteriocins and lysozyme.
Introduction
Winemaking has a long history dating back over 7000 years. Although the concept of transforming grape must into wine is not difficult to understand, production of a flavoursome and stable wine that does not spoil during storage requires considerable expertise on the part of the winemaker. Vinification practices today are not vastly different from those of ancient Egyptians and Greeks, however, the contemporary winemaker has much greater control at critical stages from grape harvest to bottling when bacteria can proliferate.
The main role of micro-organisms in winemaking is to convert grape sugars to alcohol, reduce wine acidity and introduce interesting and desirable aroma and flavours to the wine. Although grape must has a relatively complete nutrient composition, it can support only a limited number of micro-organisms, and wine, with its limited nutrients, is even less inviting. The strongest selection pressures against yeast and bacteria in grape must are high sugar content and low pH, whereas, in wine, it is high ethanol, acidity, SO2 content and limited nutrients. One of the aims of winemaking is to minimize potential for microbial spoilage and this review focuses on bacterial wine spoilage and explores options for curtailing the growth of unwanted bacteria.
Wine-associated micro-organisms
Yeast and bacteria found in grape must and wine originate from the vineyard, grapes, and winery processing equipment (Fleet 1993). This ‘natural microflora’ includes several dozen species of yeast, with Saccharomyces cerevisiae being predominant. Lactic acid and acetic acid bacteria (AAB) are the only families of bacteria found in grape must and wine. These include four genera of lactic acid bacteria (LAB), Lactobacillus, Leuconostoc, Oenococcus and Pediococcus and two genera of AAB, Acetobacter and Gluconobacter.
Bacterial wine spoilage
Many secondary metabolites produced by bacteria are volatile and potentially affect wine sensory qualities; this review will focus on undesirable flavour compounds. Figure 1 summarizes the pathways for bacterial metabolism of wine spoilage compounds and Table 1 lists these compounds, their sensory descriptors and aroma threshold concentrations.
Summary of bacterial pathways leading to spoilage aroma and flavour compounds of wine. Pathways are complied from several sources (Sponholz 1993; Costello and Henschke 2002; Wisselink et al. 2002; Swiegers et al. 2005; Walling et al. 2005a; Bartowsky and Pretorius 2008).
Growth of Lactobacillus, Pediococcus and even some Oenococcus species in wine, usually following malolactic fermentation, gives rise to numerous spoilage scenarios, as they can form undesirable aroma and flavour compounds. All AAB species are considered spoilage bacteria. Fortunately, the occurrences of most spoilage scenarios are uncommon and can be avoided with correct hygiene management during the vinification and maturation process.
Desirability of a compound in wine is dependent on concentration and wine style (Francis and Newton 2005). For example, the buttery or butterscotch aromas of the carbonyl compound, diacetyl (2,3-butanedione) which is an intermediate metabolite of citric acid metabolism in LAB (Ramos et al. 1995), can add pleasant aromas and complexity to wine at concentrations below 4 mg l−1. However, above this, diacetyl in wine may become objectionable with overt buttery notes (Martineau et al. 1995). A variety of factors, including that the winemaker can control, particularly during malolactic fermentation, affect the concentration of diacetyl. The bacterial strain used, oxygen exposure, fermentation temperature and duration of malolactic fermentation impact on diacetyl production (Bartowsky and Henschke 2004b).
Acetic acid, acetaldehyde and ethyl acetate are the main spoilage compounds produced by wine-associated AAB species. Acetic acid and acetaldehyde are formed from the oxidative metabolism of ethanol (Adachi et al. 1978). In addition, AAB can form the ethyl ester of acetic acid, ethyl acetate, which has a pungent, solvent-like aroma, reminiscent of nail lacquer remover (Francis and Newton 2005). Wine is at high risk of spoilage by AAB during prolonged barrel maturation if wine is not topped up and monitored regularly, but poor management during bottling and storage of red wine can give rise to spoilage because of the proliferation of Acetobacter pasteurianus (Bartowsky and Henschke 2004a, 2008). Small increases in acetic acid can be observed during alcoholic fermentation because of yeast metabolism, or after the completion of malolactic fermentation, usually from citric acid metabolism by LAB (Ramos and Santos 1996).
Mousy wines result from the metabolism of ornithine and lysine, leading to the formation of extremely potent and unpleasant nitrogen-heterocylic compounds [2-acetyltetrahydropyridine (ACTPY), 2-acetyl-1-pyrroline (ACPY) and 2-ethyltetrahydropyridine (ETPY)] (Costello et al. 2001; Costello and Henschke 2002). These compounds are perceived on the back palate as a persistent aftertaste reminiscent of caged mice (Tucknott 1977) because of interactions with the mouth environment; an increase in pH renders the compounds volatile. Production of the nitrogen-heterocylic compounds appears to be limited to the heterofermentative LAB (O. oeni and some species of Lactobacillus) (Costello et al. 2001; Costello and Henschke 2002). Wine associated Dekkera and Brettanomyces yeast have also been shown to produce these compounds (Grbin and Henschke 2000), however, they do not appear to be the major source of this problem.
Sorbic acid (2,4-hexadienoic acid) can be used as a chemical preservative in sweetened wines at bottling to prevent yeast fermentation after packaging. However, several LAB species, including O. oeni strains, are able to metabolize sorbic acid resulting in the formation of 2-ethoxyhexa-3,5-diene, which has an odour reminiscent of crushed geranium leaves (Pelargonium spp.) (Riesen 1992). Thus, care is needed when bottling wine preserved with sorbic acid to ensure that the bacterial population has been eliminated.
Metabolism of several sugars and polyols by bacteria can result in wine spoilage. Bitterness in wine can develop from metabolism of glycerol, mainly by Lactobacillus sp. The bitter taste is thought to result from the reaction of red wine phenolics with acrolein (Fig. 1) (Sponholz 1993). This type of wine spoilage is often referred to as ‘amertume’ (bitter in French).
Fructose metabolism by heterofermentative LAB including O. oeni can result in the formation of mannitol, a six-carbon sugar alcohol (Wisselink et al. 2002). Mannitol tainted wine is complex as it is usually also accompanied by high acetic acid, d-lactic acid, n-propanol and 2-butanol (Sponholz 1993). Such spoiled wine can also be perceived as having a slimy texture with a vinegar-estery aroma and slightly sweet taste. The switch to mannitol formation via heterolactic and mannitol fermentation occurs at the metabolic level, is growth rate related and maintenance of the redox balance (Richter et al. 2003).
A wine with a viscous and thick texture is referred to as ‘ropy’; this is because of the presence of excess exopolysaccharides such as β-d-glucan. In this context, the production of exopolysaccharides is almost exclusively because of Pediococcus growth in wine, which is prompted by high pH. The production of β-d-glucan and its polymerization are well characterized (Walling et al. 2005a) and shown to be as a result of the presence of a plasmid carrying the dps (glucosyltransferase) gene (Walling et al. 2005b). Recently, some O. oeni strains have been isolated carrying the dps gene (Walling et al. 2005b).
Removal or inhibition of unsolicited wine bacteria
How best to avoid wine spoilage is not always clear-cut. Even appropriate hygiene practices and the chemically harsh nature of wine cannot be relied on as a deterrent to unwanted bacteria and winemaking regulations may further limit the options for intervention available to the winemaker. As an initial barrier, the high ethanol concentrations (up to 16% v/v), high wine acidity (pH as low as 2·9) can inhibit development of bacterial populations, however, in wines with lower ethanol concentrations and low acidity (above pH 3·7), it can be challenging to arrest bacterial growth. Storage of wine at temperatures below 15°C might assist with minimizing the ability of bacteria to proliferate in wine, but will also delay wine maturation. Unlike the treatment of wort in beer brewing, grape must is not pasteurized prior to yeast inoculation. Heating wine prior to bottling has been explored, including flash pasteurization, however, concerns on the impact of this on wine sensory characteristics have meant that this technology is not widely used (Ribéreau-Gayon et al. 2006).
Treatment of wine with chemical inhibitors or natural products
Traditionally, sulfur dioxide has been used to control unwanted micro-organisms during winemaking, where it is usually added to bins of machine-harvested grapes and after malolactic fermentation. Sulfur dioxide acts as both an antimicrobial agent and an antioxidant in wine (Romano and Suzzi 1993). However, several bacterial species are resistant to high concentrations of sulfur dioxide. Physical removal of micro-organisms through filtration of juice or wine can also be used. However, filtration typically is mainly conducted prior to bottling and hence is not used to remove micro-organisms during winemaking.
An overall trend to reduce the use of sulfur dioxide, mainly for public health concerns, and a move in recent years to reduce the use of filtration, as some winemakers feel that filtration might impact unfavourably on wine flavour, has seen the search for alternative methodologies including chemical inhibitors and physical means to curb bacterial wine spoilage.
There are several chemical inhibitors and natural products that can be used for the control of bacteria in wine (Table 2). Although these options have great potential to reduce or eliminate bacterial populations, they are additives, and as such, legislative approval is required for their use in winemaking.
| Controlling agent | Mechanism of action |
|---|---|
| Traditional | |
| Sulfur dioxide | Inhibits the development of bacteria |
| Filtration | Physical removal of bacteria from wine |
| Chemical | |
| Dimethyl dicarbonate (DMDC) | Reacts irreversibly with the amino groups on active sites of enzymes |
| Natural products | |
| Lysozyme | Disrupts cell wall synthesis causing cell lysis |
| Bacteriocins | Alters cell wall components causing cell lysis |
| Up and coming physical technologies | |
| Ultrahigh pressure | Causes damage to cytoplasmic membrane and inactivates enzymes |
| High power ultrasound | Sound waves cause thinning of cell membranes, localized heating and production of free radicals |
| UV irradiation | Damages DNA |
| Pulsed electric fields | Dielectrical breakdown of cell membranes |
Dimethyl dicarbonate is a chemical inhibitor of micro-organisms (Daudt and Ough 1980) by inactivating cellular enzymes. It hydrolyses to methanol and carbon dioxide, natural constituents of grape juice and wine that do not affect wine flavour or colour. Although dimethyl dicarbonate is approved for use in most winemaking countries, the effectiveness of dimethyl dicarbonate varies between species and strain. Studies in grape must demonstrated that bacteria were more resistant than yeast to dimethyl dicarbonate (500–1000, 150–400 mg l−1, respectively) (Delfini et al. 2002). More recent studies in red wine suggest that the permitted rate of dimethyl dicarbonate addition (200 mg l−1) does not effectively inhibit LAB or AAB (Costa et al. 2008) implying that dimethyl dicarbonate might not be a good preservative against undesired bacterial contamination of wine. Other disadvantages of the use of dimethyl dicarbonate in wine are its low solubility in water, and, potential toxicity after ingestion or inhalation during treatment of wine.
‘Natural products’ such as lysozyme and bacteriocins to inhibit bacterial growth have been successfully utilized in various pharmaceutical and food industries for almost 50 years, and lysozyme has recently been approved for use in winemaking (maximum addition rate: 500 mg l−1). Lysozyme, a small single peptide with muramidase activity, is ineffective against eukaryotic cells; that is it cannot be used to control spoilage yeast, such as Dekkera/Brettanomyces (McKenzie and White 1991). Structural differences between the cell wall of Gram-positive and Gram-negative bacteria also limit its use for controlling AAB species.
Lysozyme can be added at various stages throughout grape vinification to inhibit LAB (Gerbaux et al. 1997; Bartowsky 2003). Different LAB vary in their susceptibility to lysozyme in wine (Bartowsky 2003), however, uses of lysozyme include the inhibition of Lactobacillus species during alcoholic fermentation thus reducing the risk of increased volatile acidity, delaying or blocking the onset of malolactic fermentation, controlling LAB populations during sluggish or stuck alcoholic fermentation, and to inhibit the onset of malolactic fermentation postbottling (Gerbaux et al. 1999). The aroma of wine is not affected by the addition of lysozyme (Bartowsky et al. 2004). As with all treatments of wine, the addition of lysozyme must be considered carefully; it is able to bind with tannins and polyphenols in red wines and typically results in a slight decrease in wine colour or might result in the formation of a wine haze (Gerbaux et al. 2000; Bartowsky 2003; Bartowsky et al. 2004).
Bacteriocins, such as nisin, pediocin and plantaricin, produced by some LAB, are small polypeptides that are inhibitory to other bacterial species. These polypeptides act on the cell wall of bacteria to induce cell lysis (Bruno et al. 1992). Species of Lactobacillus and Pediococcus are more resistant to nisin than O. oeni strains (Mendes Faia and Radler 1990), and pediocin and plantarincin have been shown to successfully kill O. oeni cells (Nel et al. 2002). A combination of nisin and sulfur dioxide has been proposed as a means to reduce the use of sulfur dioxide in winemaking (Rojo-Bezares et al. 2007). More recently, a bacteriocin-like inhibitory substance has been shown to be affective against wine Lactobacillus species (Yurdugül and Bozoglu 2002, 2008). Although the use of bacteriocins to control LAB in wine has great potential, its use has not yet been approved in winemaking.
Oenological products, such as phenolic compounds, have been demonstrated to have antimicrobial activity against pathogenic bacteria (Papadopoulou et al. 2005; Vaquero et al. 2007) and several compound types (hydroxycinnamic and hydroxybenzoic acids) can hinder wine bacterial growth (Vivas et al. 1997; Reguant et al. 2000). Limited investigations have been undertaken in using individual phenolic compounds to control spoilage bacteria (Garcia-Ruiz et al. 2008).
Alternative technologies to eliminate bacteria from wine
There is an array of emerging technologies that have been used successfully in several food and beverage industries for eliminating micro-organisms (Cheftel 1995; Smelt 1998) and could be considered for removing micro-organisms from wine. These include ultrahigh-pressure processing, ultrasound, ultraviolet irradiation and pulsed electric fields (Table 2).
Ultrahigh-pressure treatment was recognized as a potential preservation technique almost a century ago when it was demonstrated that microbial spoilage of milk could be delayed following ultrahigh-pressure treatment (Hite 1899). The applied pressure causes inactivation of micro-organisms and enzymes while not affecting flavour molecules and vitamins (Tauscher 1995). Its antimicrobial effects are primarily as a result of cytoplasmic membrane damage (Hoover et al. 1989). Ultrahigh-pressure technology has been successfully applied in fruit juices (Smelt 1998), desserts and rice cakes (Cheftel 1995), and has been reviewed as an application in cheese manufacture (Stewart et al. 2006).
High-power ultrasound uses frequencies in the range 20 to 100 kHz and has the ability to cause the formation and collapse (cavitation) of high-energy microbubbles and can be used in food processing to inactivate microbes (Piyasena et al. 2003). The mechanism of microbial killing is mainly because of thinning of cell membranes, localized heating and production of free radicals (Fellows 2000; Butz and Tauscher 2002). This technique has been shown to inactivate numerous food-related micro-organisms and has recently been proposed as an option for consideration in the wine industry (Jiranek et al. 2008).
Ultraviolet irradiation has been shown to significantly reduce LAB populations (including Lactobacillus sp.) in recirculated brines (Gailunas et al. 2008) and killing fungi on harvested grapes (Valero et al. 2007). However, it has not been extensively investigated for sterilizing wine. Sensory effects of UV exposure on wine will need to be examined as beer or milk stored in clear glass bottles exposed to light can develop ‘light struck’ off flavour (Cardoso et al. 2006).
Pulsed electric field technology has been used in beverage industries, as a means of sterilizing the product. It has been explored as an alternative to pasteurization in the production of fruit drinks, which can lead to losses in nutritional and organoleptic qualities. This technology involves application of short pulses (1–10 ms) of high- or low-intensity electric field to foods placed between two electrodes in batch, or in continuous flow systems, at low-processing temperatures (<50°C). Pulsed electric field has been used in combination with natural antimicrobials (bacteriocins, enzymes, lysozyme) to enhance the micro-biocidal protection of fruit juices (Liang et al. 2006; Mosqueda-Melgar et al. 2008). Recent trials have also shown that pulsed electric field in combination with low concentrations of SO2 does not negatively influence the formation of volatile compounds in grape must (Garde-Cerdan et al. 2008). Thus, pulsed electric field technology could be further explored as a means to eliminate spoilage bacteria from wine during wine storage prior to bottling.
Conclusion and future directions
Bacterial wine spoilage continues to be of concern in grape vinification. Consumer reactions to the use of chemical preservatives in wine will be on-going challenges for the winemaker. Managing wine acidity is one mechanism with which bacterial spoilage can be controlled, however, addition of acid to grape must and wine is subject to regulations in numerous countries and may be limited in some wine types because of impacts on wine style. Technologies based on UV irradiation, pressure and electric fields have been successfully employed in numerous beverage industries to sterilize products and recent trials in grape must or wine have been encouraging. Health concerns and changing regulatory requirements provide further motivation for the winemaking community to seek alternative ways to limit the proliferation of wine spoilage bacteria and these emerging technologies might provide support in this quest.
Acknowledgements
This project was supported by Australia’s grapegrowers and winemakers through their investment agency, the Grape and Wine Research and Development Corporation, with matching funds from the Australian Government. The author is appreciative for critical comments by Drs Paul Chambers and Paul Henschke during the preparation of the manuscript.
