VERC - Res. Pub. Notice #980902

The yeast Zygosaccharomyces has a long history of spoilage in the food industry (Thomas and Davenport, 1985). In recent years, it has received attention among winemakers. In large part, this stems from increased production, transport and utilization of contaminated grape and fruit concentrates.

The currently-recognized species, Zygosaccharomyces bailii, was first described in 1895 by Linder as Saccharomyces bailii. Its affinity with other Saccharomyces (S. acidifaciens, S. elegans) was maintained until the current genus Zygosaccharomyces sp, incorporating eight species, was adopted by Barnett, et. al., in 1983. Of the recognized species, Z. bailii, Z. bisporous, Z. rouxii and Z. florentinus have been isolated from grape must, concentrate and wines sweetened with them.

Zygosaccharomyces sp. is described as osmophilic, suggesting a habitat restricted to high solute (in this case, sugar) environments. The Aw of typical concentrated syrups (>65% soluble solids) ranges from 0.82-0.94 (Kaplow, 1970). Depending upon the solute (sugar, sugar/glycerol, salt), and its relative concentration, Z. rouxii has been reported to grow at Aw ranging from 0.62 (in fructose) and 0.65 (sucrose/glycerol) to 0.86 (in NaCl) (Corry, 1978). Since the yeast grows over a wide range of sugar levels, including those typical of fruit concentrates as well as sweetened wines, it technically should be regarded as osmotolerant or osmoduric.

In nature, Zygosaccharomyces has been isolated from dehydrated or mummified fruit as well as fruit tree exudates. Although present on fruit (Cone, 1996), population increases are noted during and after processing when competitive microorganisms have been either eliminated or their numbers greatly reduced. Thus, raw materials, such as contaminated fruit concentrates, serve to initially introduce the yeast. Once present in the winery, difficult-to-sanitize places in processing lines serve as reservoirs from which continued product contamination may occur. When growing in high sugar environments, the yeast likely grows in the thin film of water at the surface. As growth continues, it generates more water, thereby extending the contamination. Although growth in stored concentrate may go unnoticed at low storage temperatures, warming the product during shipment and/or changes in chemistry upon blend formulation may stimulate previously repressed populations.

The yeast has also been recovered from unexpected sites in the winery. For example, Rankine and Pilone (1973) reported its isolation from a pressure gauge on the post-filtration side of a sterile filter. In this case, the gauge's construction not only allowed the organism to survive steam sterilization, but, to continually reenter sterile-filtered wine enroute to the bottling line. Other unusual sites of isolation include lubricating oils (Beech and Davenport, 1983). Aerosols created by interaction of these contaminated oils and high speed machinery served to disperse the organism throughout the previously sanitized area.

As seen in the following table, Zygosaccharomyces is extraordinarily resistant to common preservatives used in juice, concentrate and wine:


Sulfur Dioxide	>3 mg/L Molecular (at pH 3.4)`^a`
Sorbic Acid	600-800 mg/L`^b,c`
Benzoic Acid	600-1,000 mg/L`^b,c`
Ethanol	>18% (v/v)`^c`
¹Acetic Acid	20-25 g/L`^b,c`

Thomas and Davenport (1985) speculate that the yeast's extradorinary resistance to sulfur dioxide results from formation of extracellular compounds (such as acetaldehyde) that bind SO2 thereby reducing the con-centrations of the molecular form.

Thomas (1983) reported that phenolics and antho-cyanins in red wines may be inhibitory. This supports previous observations of Peynaud and Domercq (1959) suggesting that white wines were at greatest risk of spoilage. By comparison, our survey found concentrate-sweetened rose and red wines to experience greater incidences of refermentation (Fugelsang and Muller, 1994 unpublished results). Further, we found that cork or capsule-finished bottled wines seldom presented a problem whereas the same wine packaged bag-in-box may support growth. The two differ in the area of post-bottling oxygen incursion. Bag-in-box liners are somewhat oxygen permeable whereas incursion of oxygen in cork-finished bottled wine is limited. Further, the greater incidences of spoilage in bag-in-box packaged sweetened red and rose wines suggests the role of flavonoid phenols (and their polymers) as oxygen reservoirs.

Incidences of Zygosaccharomyces spoilage have also been linked to sub-lethal doses of chemical sterilants or steam/hot water protocol that doesn't meet the time x temperature requirements for cell death. In bottled soft drinks, Van Esch (1992) noted that aside from contaminated raw material (i.e., fruit concentrate), 95% of the contamination could be traced to production demands on the bottling facility which lead to shortcutting sanitation. Still others report spontaneous contamination in wines bottled from heat-sanitized lines were standard procedures were closely followed (Annon. 1995-96).

Because of these problems, chemical sterilization techniques have been employed. These include the use of peroxyacetic acid as well as ozonated water. Both sterilants react to produce lethal concentrations of oxygen. Peroxyacids owe much of their lethal properties to the fact that they are soluble in lipid component of the cell membrane. Once inside, they react to form a lethal concentrations of molecular oxygen. Use of chemical sterilants requires thorough rinsing after treatment. Ozone (as ozonated water) has recently been tested on winery bottling lines. Unlike peroxyacids, residual ozone breaks down to molecular oxygen and thus the line requires no special treatment after sterilization (Mahaffey, et. al. 1998). Owing to its unusual resistance, a zero-tolerance for recovery of Zygosaccharomyces in post-bottling samples is recommended (Vilas, 1993).

Over a two year period, 87 samples, representing all stages in processing from crushed fruit to stored juice, concentrate and sweetened packaged wine were sampled for presence of Zygosaccharomyces. As expected, the majority of positive isolations originated in the proximity of concentrate storage tanks and their associated transfer lines, pumps and filters. One isolate was recovered from white grape must (Cone, 1996).

In addition to recovery work, growth studies were also carried out in packaged Zinfandel and Grenache blush-style wines. Alcohol levels in 1L volumes of commercial Zinfandel and Grenache blush-style wines were adjusted from baseline 7.5% (v/v) to 8.0%, 8.5%, 9.0%, 9.5% and 10% alcohol respectively using NSFG. Each was sparged with nitrogen gas to reduce oxygen levels to 0.5 ppm (determined using a YSI Model 57 dissolved oxygen meter) and filtered (0.45 um) into previously steriled mylar bags similar to those used in bag-in-box packaging. Prior to sealing, each was inoculated with stationary phase Zygosaccharomyces at a level of 70 cells/L.

Duplicate lots representing each alcohol level were then incubated for 90 days at 60, 70, 80, and 90^oF. Bags were examined daily for signs of fermentative activity (detected as expansion). The only lots that did not show biological activity were those held at 90šF. At lower temperatures, refermentation followed expected time and temperature relationships (Fugelsang, 1996). In this regard, it has been reported that osmophilic/osmoduric yeasts exhibit little heat resistance (Beuchat, 1982). This combined with our results suggests the potential for HTST thermal processing.

The potential for utilization of carbon monoxide as an alternative to sulfur dioxide (and other preservatives) in control of Zygosaccharomyces and other juice/wine microorganisms was also studied. Carbon monoxide has several advantages, compared with sulfur dioxide, that make it attractive in this regard. First, CO has an electron deficiency on the carbon similar to the electron deficiency of sulfur in sulfur dioxide. In addition, this same carbon has a pair of electrons available which can attack any Lewis acid and/or electrophile. Thus, carbon monoxide can act very much like sulfur dioxide and (additionally) has the ability to act as a strong electrophile. It is this dual chemical activity that makes CO an attractive alternative to sulfur dioxide. Further, once a covalent bond has formed between CO and one of its target molecules, it is usually a carbon-carbon bond and, thus, difficult to hydrolyze. By comparison, carbon-sulfur bonds are polar and hydrolyzable with ease. Thus, compared to sulfur dioxide, no CO can be freed by hydrolysis (in either strong acid or base) from most of its adducts (Hine, 1964). Further, we have determined that virtually no CO remains after treatment; being completely converted to hydrogen and carbon dioxide.

In laboratory studies using 1.5L model juice and wines, initial studies (1994; 1996) have shown that carbon monoxide, at levels of 480 mg/L, was effective in control (defined as no growth for 9+ days at 24^oC) of Zygosaccharomyces bailli among other spoilage yeasts present at levels of 10-20 x 10⁴ CFU/mL (Table 2). Saccharomyces, by comparison was not affected by levels of CO >1,000 mg/L. While we were successful in preventing growth of Zygosaccharomyces for 9 days, full control was not achieved. Growth was reported after 14 days, suggesting the need to retreat.

Based upon initial success of laboratory-scale work, a follow up production scale treatment was carried out in Spring and Summer of 1997.

Treatment System: Since carbon monoxide is relatively insoluble in aqueous solution, tall aspect-ratio (6.4/1) tanks were selected to maximize exposure of yeast to gas. To further maximize contact of gas and cells, a delivery system was designed and built to inject a continuous flow of micron-sized gas. Two identically-configured and adjacent 588 gal. jacketed stainless steel tanks (16'H x 2.5'W) were used for this pilot plant demonstration. Each tank was equipped with a 3" bottom drain and two 2" racking valves positioned 12" and 24" from the tank's bottom. Each tank was also equipped with a sampling port located 24" above the bottom and approximately 30^o with respect to the top racking valve.

Successful introduction of the 480 mg/L CO (see Table 2) required mixing for 2 hrs. To accomplish mixing, a circulation loop was setup as follows:

Gas Dispersion Cylinder: Used to assure delivery of micron-size gas bubblesto the wine, the device was con-structed entirely of stainless steel, consisting of a 2" I.D. pipe containing an array of randomly placed longitudinal baffles whose purpose is to increase aspersion of gas bubbles as provided by a 0.5" gas delivery tube placed on the upstream side of the baffles. The gas inlet was provided with an adapter-housing to hold a 0.45 um sterile filter that assured delivery of sterile, micron-size bubbles to a wine stream flowing through the 2" pipe.

Sterilization of Tanks and Recirculation System: Prior to use, the tanks and circulation system described above were set up and mild solution of caustic soda circulated for 30 minutes. This was followed by a cold-water rinse and neutralization with citric acid solution. Immediately prior to transfer of reconstituted juice, hot (180šF) water was circulated through the system for 30 minutes and drained.

Preparation/Sterilization of Juice: One-hundred gallons of commercial white grape concentrate was reconstituted with water to yield juice of 21.9^oB and transferred to previously described sterilized stainless steel tanks. One-hundred milliliter aliquots of juice were collected for laboratory population density estimates as described under Microbiological Sampling (below).

Each lot was then treated with DMDC at 600 mg/L and the tanks sealed. Twelve hours later, juice samples were collected, membrane filtered and transferred to the appropriate growth medium.

Carbon Monoxide Delivery System: Carbon monoxide (Air Liquide prepared as 4.02% CO in nitrogen) was injected into the system by means of a gas dispersion device (see description, above).

Inoculum Preparation: Eight-hundred milliliter volumes of a stationary phase culture of Zygosaccharomyces bailii was used as inoculum for each tank. After inoculation and thorough mixing (and before introduction of CO), samples from both control and treated tanks were collected to determine viable cell density.

Microbiological Sampling: One-hundred milliliter aliquots from each tank were aseptically collected before and after DMDC treatment and daily for 9 days following treatment with CO. Each sample was examined microscopically and, appropriate dilutions were membrane-filtered (0.45 um) and transferred to the following selective media. Each of the final dilutions was plated in triplicate:

As seen in Figure 1 (Page 4), the lot treated with carbon monoxide yielded viable cell density well below that of the untreated control. The fact that growth was observed was not unexpected given results from earlier laboratory-scale experiments.

The author wishes to thank the California Agricultural Technology Institute (CATI) for its support of this project.

Barnett, J.A., Payne, R.W., and Yarrow, D. Yeasts: Characterization and Identification. Cambridge University Press. 1983.

Beuchat L.R. 1982. Thermal inactivation of yeasts in fruit juices supplemented with food preservatives and sucroses. J. Food Sci. 47:1679-82.

Beech, F.W., and Davenport, R.R. 1983. New prospects and problems in the beverage industry. In: Food Microbiology: Advances and Prospects (T.A. Roberts and F.A. Skinner, Eds.) The Soc. for Appl. Bacteriology Symp. Series No. 11. London Academic Press. pp. 241-56.

Corry, J.E.L., 1978 Relationships of Water Activity to Fungal Growth. In: Food and Beverage Mycology. L.R. Beuchat, ed. Avi Publishing Co., Westport, Conn. pp. 45-83.

Davenport R.R. 1982. Sample size product composition and microbial spoilage. In: Long Ashton Research Station: Seventh Wine Subject Day 'Shelf Life.' R.W. Breech, ed. pp. 1-4.

Hine, J. 1964. Modern Concepts in Chemistry Services. New York: The Ronald Press Co.

Kaplow, M. 1970. Commercial development of intermediate moisture foods. Food Technol. 24:889-93.

Mahaffey, D., McClain, J., and Havens, M. 1998. Ozone, the versatile sanitizer. Pract. Winery and Vineyard XVIII(5) (Jan./Feb.): 89-90.

Muller, C.J. and Fugelsang, K.C. 1994. Microbiological Stabilization of Juice and Wine: Effect of Carbon Monoxide on Spoilage Yeasts. Proceedings of the Office International de la Vigne et du Vin. Paris, France. June, 4-8.

Peynaud, E., and Domerq, S. 1959. A review of microbiological problems in winemaking in France. Am. J. Enol. Vitic. 10:69-77.

Pitt, J.I. 1974. Resistance of some food spoilage yeasts to preservatives. Food Tech. Australia 26:238-41.

Rankine, B.C., and Pilone, D.A. 1973. Saccharomyces bailii, a resistant yeast causing serious spoilage of bottled table wine. Am. J. Enol. Vitic. 24(2):55-58.

Thomas, D.S. 1983. Susceptibility of wines to yeast spoilage. Long Ashton Res. Station Report.

Thomas, S., and Davenport, R.R. 1985. Zygosaccharomyces bailii, a profile of characteristics and spoilage activities. Food Microbiology 2:157-169.

van Esch, F. 1992. Yeast in soft drinks and concentrated fruit juices. Brygmesteren - NR 4:9-20.

Vilas, M. 1993. Bottling line sampling and diagnostic techniques. Vineyard and Winery Management Sep/Oct:33-35.

Zoecklein, B.W., Fugelsang, K.C., Gump, B.H. and Nury, F.S. 1995. Wine Analysis and Production. Chapman & Hall Publ. Co. New York, NY.

Zygosaccharomyces, A Spoilage Yeast Isolated from Grape Juice

Zygosaccharomyces, A Spoilage Yeast
Isolated from Grape Juice