Growth Characteristics of Saccharomyces rouxii Isolated from

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, May 1983, P Vol. 45, No /83/ $2./ Copyright 1983, American Society for Microbiology Growth Characteristics of Saccharomyces rouxii Isolated from Chocolate Syrup L. RESTAINO,* STEPHANIE BILLS, KARIN TSCHERNEFF, AND LAWRENCE M. LENOVICH Hershey Foods Corporation, Hershey, Pennsylvania 1733 Received 2 November 1982/Accepted 19 January 1983 We investigated the growth parameters of Saccharomyces rouxii isolated from spoiled chocolate syrup. The optimum ph range for S. rouxii was 3.5 to 5.5, whereas the minimum and maximum ph values that permitted growth were 1.5 and 1.5, respectively. For cells grown in and 6% sucrose the optimum water activity (a,) values were.97 and.96, respectively. The optimum temperature for S. rouxii increased with a decreasing a, regardless of whether glucose or sucrose was used as the humectant. The optimum temperatures for S. rouxii were 28 C at an a. of >.995 and 35 C at an a, of.96 to.9 in 2x potato dextrose broth with sucrose. Increasing the sorbate concentration (from.3 to.1%) caused the growth of S. rouxii to become more inhibited between a,s of >.995 and.82. S. rouxii did not grow when the sorbate level was.12% (wt/vol). At lower sorbate levels, the effect of sorbate on the growth of S. rouxii depended on the a, level. Lowering the a, enhanced the resistance of S. rouxii to increasing concentrations of potassium sorbate. Permeability and polyol production are discussed with respect to sorbate tolerance of S. rouxii at different a, levels. The microorganisms that spoil foods containing a low water activity (aw), high acidity, low redox potential, and high carbon/nitrogen ratio are osmotolerant yeasts (31, 37). Since these yeasts compete poorly with other microorganisms in nonselective environments, these parameters become important in favoring the growth of osmotolerant yeasts. Osmotolerant yeasts are the main spoilage organisms in foods such as honey, maple syrup, raw sugar cane, fruit syrups, candy, jams, jellies (37), soy mashes (24), fondants (28), fruit juice concentrates (3), dried fruits (21, 22), and chocolate syrup (unpublished data). Contamination of these food products could originate from highly contaminated raw ingredients, inadequate sanitation due to the difficulty of cleaning a highsugar commodity from equipment, improper packaging and storage facilities, and insects during all phases of production. The spoiled product usually becomes turbid with a strong alcoholic aroma. If the product is enclosed in a hermetically sealed container, the package will swell due to carbon dioxide production. Growth of osmotolerant yeasts could reduce the solid content and alter the solubility of solutes that cause the aw of the food to increase. Osmotolerant yeasts comprise, primarily, members of the genus Saccharomyces, the major species of which include S. rouxii, S. bailii, and S. bisporus. Strains tolerant to very high concentrations of solutes belong to the species S. rouxii (27, 36), the most common spoilage organism of the osmotolerant yeasts. The minimum a, range of this yeast in a fructose syrup at ph 4.8 (35) and in glucose broths (32) is.62 to.65, whereas the optimum aw in sucrose broths is.98 (R. H. Tilbury, M.S. thesis, University of Bristol, England). The solute has an influence on the growth of S. rouxii, for which the limiting concentrations are 2 to 22% for NaCl, 8% for glucose, and 8% for sucrose (24). S. rouxii and other osmotolerant yeasts are classified as mesophiles which can grow at refrigeration temperatures (32). The ph range for most osmotolerant and non-osmotolerant yeasts is 2. to 7. with an optimum ph of 4. to 4.5 (32). Ingram (16) has shown a direct relationship between the tolerance of osmotolerant yeasts to high acidity and reduced aw levels, where the lower the aw, the less tolerant the microorganism is to acidity. Further, Baird-Parker and Kooiman (3) have studied the coupled effects of sorbic acid, ph, and a, on the growth of yeasts. Their data showed that an increase in sucrose (decrease in a,) and a decrease in ph cause sorbate to become more effective against the growth of yeasts. However, the antimycotic effects of potassium sorbate to the growth of S. rouxii depend on the preconditioning of this yeast to sorbate or other weak organic acids (5; J. I. Pitt, personal communication). The objective of this investigation was to elucidate various growth parameters of an osmo- 1614

2 VOL. 45, 1983 tolerant, sorbate-resistant S. rouxii strain isolated from chocolate syrup. MATERIALS AND METHODS Yeast strain and identification scheme. S. rouxii NRRL Y-12691, isolated from chocolate syrup containing.1% potassium sorbate, was used in this investigation. Stock cultures were adapted and maintained on slants of 2x potato dextrose agar (PDA [Difco Laboratories]) (12% glucose added) without or with 6%o (wt/vol) sucrose through at least four transfers and stored at 4 C. Stock cultures were transferred once every 2 months. The identification of the yeast isolate was made by using the carbohydrate fermentation scheme described by Lodder (18), a DNA hybridization assay (17), and sensitivity to.5% acetic acid in malt extract agar plus 2% glucose (26; J. I. Pitt, personal communication). The yeast isolate rapidly (within 5 days) fermented glucose, maltose, and fructose and slowly (28 days) utilized sucrose. After 7 days of incubation at 25 C on malt extract agar containing.5% acetic acid, no visible growth of this yeast strain was observed, indicating sensitivity to this weak organic acid. The DNA hybridization assay as described by Kurtzman et al. (17) showed a 98% DNA base sequence complementarity with the type strain, S. rouxii NRRL Y-229. Preparation of inoculum. S. rouxii cells were transferred from slants to 2 x potato dextrose broth (PDB). Cells maintained on 2x PDA-12% glucose were transferred to 2x PDB-12% glucose, whereas cells from 2x PDA-12% glucose slants with 6o (wt/vol) sucrose were added to 2x PDB-12% glucose containing the corresponding sucrose level. For the experiments involving optimum aw, optimum ph, and ph range, inocula grown in the presence and absence of 6%o sucrose were used, whereas for the other experiments, only cells grown in 2x PDB-12% glucose without sucrose were used. The flasks containing the medium plus the inoculum were incubated at 25 C in a shaking water bath (16 rpm) for 2 days. Stationary-phase cells were transferred to sterile 5-ml polypropylene centrifuge bottles. Cells grown in 6%o sucrose were diluted 1:1 with sterile distilled water to reduce the viscosity. The inocula were centrifuged at 4 to 8 C for 3 min at 6, x g. The supernatant liquid was discarded, and the pellet was resuspended in 1 to 15 ml of sterile distilled water. The absorbancy of the cells was adjusted to.5 at 42 nm, using a Bausch & Lomb Spectronic 2 colorimeter, which corresponded to a cell concentration ranging from 4. x 15 to 1. x 16 cells per ml. Volumes of.3 to.35 ml and.7 ml of cells grown in medium containing and 6%o sucrose, respectively, were used to inoculate 8-ml volumes of the various growth media. Preparation of growth media. The basal medium for all growth experiments was 2x PDB-12% glucose, which was prepared according to manufacturer's specifications. For the experiments involving optimum aw, 2-ml volumes of the basal growth medium were formulated with various sucrose levels (expressed throughout as wt/vol). Glucose and sucrose were added separately and dissolved with the aid of heat. After boiling, duplicate 8-ml volumes for each sucrose level were dispensed into 25-ml Fernbach flasks and autoclaved at 121 C for 1 min. The ph of GROWTH PARAMETERS OF S. ROUXII 1615 the media was adjusted to 5. by using sterile HCI or NaOH. After adjustments, a, levels were determined for each sucrose level by using a Beckman Hygroline recorder (model VFB). The corresponding sucrose and a, levels were % sucrose, >.995 a,; 6% sucrose,.995 a,; 12% sucrose,.99 a,; 18% sucrose,.98 a,; 23% sucrose,.97 a,; 27% sucrose,.96 a,; 3% sucrose,.95 a,; 4% sucrose,.92 a,; and 5% sucrose,.9 a,. The flasks were stored at room temperature with the caps wrapped with Parafilm to prevent a change in aw. After inoculation, the flasks were incubated in a shaking water bath (16 rpm) at 25C. For the experiments involving optimum, minimum, and maximum ph values, 3 liters of the basal medium supplemented with 23 or 27% sucrose was formulated and autoclaved at 121 C for 1 min. With a sterile 25- ml graduated cylinder, 16-ml volumes were dispensed into sterile 25-ml Fembach flasks. ph values ranging from 1. to 12. (in increments of.5 ph units) were obtained with sterile HCl or NaOH. Different concentrations of HCl and NaOH were used to add the smallest possible amount of the base or acid to obtain the desired ph. After ph adjustment, the a, was determined; 23% sucrose gave an aw of.97, and 27% sucrose gave an aw of.96. S. rouxii cells previously grown in or 6%o glucose were inoculated into the ph-adjusted medium, and the flasks were incubated at 25 C in a shaking water bath (16 rpm). The presence of visible growth was monitored daily for 3 days. Double-strength PDB with % sucrose-o glucose, 23% sucrose-o glucose, 9%o sucrose-12% glucose, or the above-mentioned sucrose and glucose levels necessary to obtain a, levels of.98,.97,.96,.95,.92, and.9 was used to determine the optimum temperature versus the a, level of humectant. The growth medium was formulated as in the experiments involving optimum a,. The flasks were placed in 6- ml beakers containing 3 ml of distilled water to reduce temperature fluctuations during incubation. Flasks (one from each growth medium) were placed in incubators at 22, 25, 28, 32, 35, 37, and 4 C. After equilibrating for 24 h, the media were inoculated with S. rouxii previously grown in % sucrose. For studying the effects of the potassium sorbate concentration on the growth of S. rouxii at various aw levels, the basal growth medium plus various sucrose concentrations as described for the experiments on optimum a, was used. In addition, sucrose concentrations of 6 and 7o (wtlvol), corresponding to aws of.86 and.82, respectively, were used. Portions (79 ml) of the basal growth medium with appropriate sucrose concentrations were dissolved in 25-ml Fernbach flasks and autoclaved at 121 C for 1 min. Potassium sorbate granules were dissolved in distilled water at concentrations of 2.4, 4.8, 6.4, 8., 9.6, and 12.o (wt/vol) on the day of use. These solutions were filter sterilized through a.2-,um membrane. After the growth media equilibrated to room temperature, 1. ml was added to each aw level to yield final sorbate concentrations of.3,.6,.8,.1,.12, and.15%. An equal volume of sterile distilled water was added to media in duplicate flasks to yield a control without sorbate. The media were adjusted to ph 5., and the aw levels were calculated. The flasks containing the various media were inoculated with S. rouxii and incubated statically in an air incubator at 32C.

3 1616 RESTAINO ET AL. 2.3( 2.4( 2.5( 2.6( w Z W 3.1 z w O WATER ACTIVITY FIG. 1. Optimum a, levels for S. rouxii cells preconditioned in or 6% sucrose. Cells were grown to the stationary phase in 2x PDB-12% glucose plus % (@) or 6% (U) sucrose. The presence of visible growth was noted daily for 47 days. Plating media, diluent, and enumeration procedures. Plating media and diluents with sucrose were utilized to reduce osmotic shock. Basal phosphate buffer (.3 mm, ph 7.2) diluent and PDA (with 2% glucose added) plating medium were used in all experiments. Regular diluent and plating medium were used when the growth medium contained or 6% sucrose. A 4% (wt/vol) sucrose-phosphate buffer (ph 5.1; a,,.95) was used as the diluent with 12, 18, 23, 27, or 3% sucrose in the growth medium, whereas 6% (wt/vol) sucrose-phosphate buffer (ph 5.1; aw,.93) was used to dilute growth medium containing 4, 5, 6, or 7% sucrose. PDA plating medium containing 3% sucrose (ph 5.2; aw,.99), 45% sucrose (ph 5.2; aw,.95), 5% sucrose (ph 5.2; a,,.94), 6% sucrose (ph 5.2; aw,.92), and 65% sucrose (ph 5.2; aw,.9) was used to enumerate S. rouxii cells from growth media having aws of.97 to.98,.96,.95,.9 and.92, and.82 and.86, respectively. The pour plate technique was utilized to measure the growth of the osmotolerant S. rouxii cells. At various time intervals, 1-ml volumes were pipetted from the different growth media, serially diluted in the appropriate diluent, and plated in duplicate on the corresponding PDA with sucrose. Plates were incubated at 25 C for a minimum of 5 days. Yeast counts were transformed to loglo yeast cells per milliliter. The data were analyzed statistically by comparing the geometric means of duplicate samples. When appropriate, linear regression curves were determined from data for exponentially growing S. rouxii cells. A 95% confidence level was used on all linear regression curves. Growth and death rates (generations per hour) and generation times (in hours [reciprocals of growth or death rates]) were computed from the linear regression curves. RESULTS Optimum a, levels. The optimum a, levels of S. rouxii preconditioned in and 6% sucrose are presented in Fig. 1. S. rouxii cells were preconditioned in the supplemented basal medium with or 6% sucrose for at least four transfers. The preconditioning of S. rouxii cells to or 6% sucrose influenced the growth rate pattern over a ph range. At ph 5., the optimum aws for S. rouxii cells were.97 and.96 for cells previously grown in and 6% sucrose, respectively. Optimum a, levels were determined with glucose and sucrose as the humectants. Growth response to S. rouxii at various ph levels. The effects of different ph levels on the growth of S. rouxii cells previously grown in and 6% sucrose are presented in Table 1. The preconditioning of S. rouxii in or 6% sucrose did not influence the growth response of S. rouxii to different ph levels. The minimum and maximum ph values for growth of S. rouxii in 2x PDB-12% glucose were 1.5 and 1.5, respectively. Since the generation times were similar through several consecutive ph values for S. rouxii cells, an optimum ph could not be calculated; instead, comparable growth rates were obtained over a wide ph range. The optimum ph range was 3.5 to 5.5 for S. rouxii. TABLE 1. APPL. ENVIRON. MICROBIOL. Growth response of S. rouxii at different ph levels in 2x PDB-12% glucose Responsea of inoculum grown: ph Without In 6% sucrose sucrose 1. - NT NT + (8.78) (5.13) + + (2.85) (2.27) ++ (2.68) (2.2) + + (2.47) (2.27) + + (2.6) (2.26) ++ (2.5) (2.16) + + (2.57) (2.15) + + (2.6) (2.23) + + (2.67) (2.25) NT (2.67) NT NT NT a Symbols: + +, normal growth; +, poor growth;-, no growth; NT, not tested. Numbers within parentheses are generation times (in hours).

4 VOL. 45, 1983 CD.3 c cm TEMPERATURE ( C) FIG. 2. Effect of sucrose and glucose on the optimum temperature of S. rouxii. Cells were grown to stationary phase in 2x PDB-12% glucose. Symbols:, no sucrose or glucose; E, 23% sucrose, no glucose;, no sucrose, 12% glucose; U, 23% sucrose, 12% glucose. Optimum temperatures versus a, or carbohydrate. The effects of sucrose and glucose on the optimum temperature for S. rouxii are illustrated in Fig. 2. All experiments involved S. rouxii previously grown without sucrose. The optimum growth temperature for S. rouxii cells in 2 x PDB (a, >.995) with no sucrose and no additional glucose (4% glucose in the basal medium) was 28 C. With the addition of 23% sucrose or 12% glucose, the a, remained at >.995, but the optimum growth temperature increased to 32 C. When both sugars were added together to 2 x PDB, the a, decreased from >.995 to.97, but the optimum growth temperature remained at 32C. The effects of various a, levels on the optimum growth temperature for S. rouxii cells are presented in Fig. 3. The optimum temperature increased from 28 to 35 C with an increase in sucrose concentration. S. rouxii cells grown in 2x PDB with an a, of >.995 had an optimum growth temperature of 28 C, whereas an a, of.97 caused the optimum temperature to increase to 32C. Below the optimum a, level (Fig. 1) for S. rouxii cells (i.e., at aws of.95 to.9), the optimum growth temperature increased to 35 C. For aws above.96, the growth rates (generations per hour) for S. rouxii cells at GROWTH PARAMETERS OF S. ROUXII 1617 temperatures greater than the optimum rapidly decreased; this was not observed in cells growing at an a. of.95,.92, or.9. At 4 C, the growth rates (generations per hour) of S. rouxii in aws of.95,.92, and.9 were.23,.31, and.26, respectively, whereas cells grown in an a, of.97 attained a growth rate of only.4 generations per h. Growth of S. rouxii at different a, and sorbate levels. The effects of different a, levels and sorbate combinations on the growth of S. rouxii cells are presented in Table 2. Increasing the sorbate concentration (from.3 to.1%) accelerated the growth inhibition of S. rouxii in 2x PDB (ph 5.) within an a, range of >.995 to.82. Without sorbate, visible growth of S. rouxii was observed at a,s between >.995 and.9 in 1 day, whereas 2 and 4 days, respectively, were required for visible growth to occur at a,s of.86 and.82. At sorbate concentrations of.12 and.15%, the antimicrobial compound was lethal to S. rouxii during a 47-day incubation period. At a sorbate concentration of.15%, the death rate (generations per hour) ranged from.69 at an a, of.86 to 2.8 at an a, of.99. At.12% sorbate, the death rate increased with an a, ascending from.82 to >.995. At this sorbate concentration, the death rates ranged from 1.89 to 2.22 generations per h at the higher a, levels (.97 and higher), whereas at a,s of.86 ~ s,, TEMPERATURE ( C ) FIG. 3. Effects of a, level on the optimum temperature for S. rouxii cells. Cells were grown to stationary phase in 2x PDB-12% glucose; a,s were >.995 (),.97 (),.95 (-),.92 ([), and.9 (A).

5 1618 RESTAINO ET AL. TABLE 2. Effects of a, levels versus sorbate concentrations on growth of S. rouxiia Effectb with the following potassium sorbate aw concn:.%.3%.6%.8%.1%9.12%.15% > d 3 d d 2 d 15 d d 2 d 6 d 14 d d 2 d 3 d 8 d d 2 d 3 d 5 d 2 d d 2 d 3 d 5 d d 2 d 4 d 5 d 13 d d 3 d 5 d 6 d 13 d d 5 d 6 d 8 d d 8 d a The experiment was performed for 47 days at 32C. bnumbers followed by the letter d (days) indicate the number of days before visible growth could be observed. Numbers without the letter d indicate the death rate (number of generations per hour). and.82, the death rates for S. rouxii cells were.55 and.61, respectively. At sorbate concentrations of.3 to.1%, the day that visible growth of S. rouxii cells was observed depended on the aw level. Growth was initially observed in 2 days in 2x PDB containing.3% sorbate at aws ranging from.92 to.99, whereas 3, 5, and 8 days, respectively, were required for visible growth at aws of >.995 and.9,.86, and.82. At.6% sorbate, aws of.95 to.97 required 3 days before there was visible growth of S. rouxii, whereas death rates of.4 and.41 generations per h, respectively, occurred for S. rouxii at aws of.82 and >.995. At.8 and.1% sorbate, the width of the aw range that supported S. rouxii growth decreased compared with that at the lower sorbate concentrations. At.8% sorbate, aws of.92,.95, and.96 supported the fastest growth rate of S. rouxii, but death occurred at aws of >.99. At.1% sorbate, this yeast propagated fastest at aws of.9 and.92, exemplified by visible growth in 13 days, whereas death occurred for S. rouxii at an aw above.97. APPL. ENVIRON. MICROBIOL. DISCUSSION The most common of the osmotolerant yeasts is S. rouxii (32, 37), which can grow at an aw of.62 when other extrinsic conditions are optimal (27). Using sucrose broths, Tilbury (M.S. thesis) showed that the optimum aw for growth of S. rouxii is.98. In the present investigation, the optimum aw of the S. rouxii strain isolated from chocolate syrup was.97 and.96, respectively, for cells previously grown in and 6% sucrose. Anand and Brown (1) have shown slight variations in growth rate patterns at different a,s for two S. rouxii strains. Slight discrepancies in optimum a,s could be due to the basal growth media, slight genetic variations within species, preconditioning with respect to aw, or a combination of these. The data from this investigation show the influence of acclimation in determining the optimum aw for an osmotolerant yeast strain. When investigating the optimum aw for an osmotolerant yeast, the preconditioning becomes extremely important. The ph growth range of S. rouxii is very wide (1.5 to 1.5) (Table 1), enabling this microorganism to survive or grow in a wide range of foods. This wide ph tolerance can allow this yeast to cross-contaminate a variety of foods under conditions where sofne commodities could contain parameters favorable for osmotolerant yeast spoilage. The optimum ph range of this S. rouxii strain was 3.5 to 5.5, which relates to the ph range (4. to 4.5) of S. rouxii calculated by Tilbury (M.S. thesis). However, the ph optimum range in this investigation was substantially larger than the ph range determined by Tilbury (M.S. thesis), indicating influences of growth media or strain variations or both. The optimum ph range and aw levels of S. rouxii do not coincide with the ph and aw parameters for the types offoods spoiled by this yeast. Other intrinsic or extrinsic characteristics must be involved. Acidic foods such as tomato paste, canned fruits in heavy syrups, margarine, canned cured meats, Gouda cheese, and low-salt fish, pork, and beef products (1, 33) are not normally spoiled by S. rouxii even though their aw levels (.92 to.98) are near the optimum for S. rouxii. However, products that are usually spoiled by S. rouxii have lower aw levels (.8 to.85); these products include honey, maple syrup, fruit syrups, jams, liquid sugar, soy mashes, and chocolate syrups (24, 37; unpublished data). The phs of these low-aw foods are in the range optimal for S. rouxii growth. Since a variety of bacterial strains can proliferate faster than osmotolerant yeasts at aws ranging from.97 to >.995, this yeast cannot compete and spoil a product suitable for bacterial growth. In a harsher environment, i.e., in products containing a low aw and ph, most bacteria either cannot grow or grow very slowly. Spoilage will most likely be caused by yeasts or mold. Additional factors involving the numbers and types of osmotolerant yeasts initially infecting the food, the types of nutrients, the redox potential (liquid or solid food), and the temperature during storage (37) can enable the osmotolerant yeasts to outgrow the xerotolerant molds. As the yeasts proliferate, the production of ethanol and CO2 further creates an environment detrimental to xerotolerant molds (32).

6 VOL. 45, 1983 As noted by other researchers (16, 23), a decrease in a. causes the optimum temperature of S. rouxii to increase (Fig. 2 and 3). Although S. rouxii can rapidly metabolize glucose, whereas sucrose either is not fermented or is fermented extremely slowly (18), the optimum temperature for S. rouxii increases regardless of whether sucrose or glucose is used as the humectant. Recently, increased recovery rates of osmotolerant yeasts from intermediate-a, foods, e.g., fruit juice concentrates, have been reported on media with elevated sugar concentrations (4). Since the optimum growth temperature range for S. rouxii in a broth containing an a, between.9 and.97 is 32 to 35 C, an incubation temperature of 32 C, instead of the traditional 22 to 25 C (34) for 5 days, would be more efficient for enumerating osmotolerant yeasts in the presence of elevated sugar concentrations. In addition, S. bailii, a yeast less osmotolerant than S. rouxii, has an optimum temperature of 32 C at an aw of.96 (unpublished data), increasing the flexibility of this incubation temperature for enumerating a wide range of osmotolerant yeasts. Why was an increased optimum temperature of S. rouxii observed when it was inoculated in a growth medium containing reduced a, levels? One explanation could involve the protective effect of the solute on the metabolic mechanisms and cellular integrity of S. rouxii. The heat resistance of microorganisms (vegetative cells and spores) increases with elevated concentrations of solutes (4, 11, 12, 14). Doyle and Marth (12) and Gibson (14) have explained this increased heat resistance as a dehydrating phenomenon resulting in a greater stability of cellular components, specifically, proteins. In the present investigation, a similar explanation could be proposed for the elevated optimum temperature of S. rouxii with a decreasing aw. However, since a greater concentration of glucose (3%) or sucrose (45%) in the heating menstruum was required to protect S. rouxii from heat (12, 14) than was needed to increase the optimum temperature, a simple dehydration phenomenon of cells may not be the total explanation. A second explanation, involving the production of intracellular polyols (9) by S. rouxii, is also feasible. Intracellular polyols have been shown to function as compatible solutes (6, 7, 9), which relates to the protection of enzymes in yeasts against inhibition or inactivation. The concentrations of intracellular polyols in osmotolerant yeasts can accumulate rapidly in cells growing in decreasing a, levels (6, 13). Only a slight reduction in a, can cause a rapid increase of intracellular polyols in osmotolerant yeasts. When S. rouxii grows in a high sugar concentration, the polyol that accumulates is arabitol (2), GROWTH PARAMETERS OF S. ROUXII 1619 whose concentration is directly proportional to the a, level. Although the mechanism of polyol protection toward enzymes had not been elucidated, this system could be related to heatsensitive enzymes with respect to elevated temperatures. Thus, intracellular polyols could enable S. rouxii to proliferate at a higher temperature, resulting in an increased optimum temperature. With the relationship between potassium sorbate tolerance of S. rouxii and different a, levels, some hypotheses or thoughts can be formulated on the mechanism of sorbate resistance in osmotolerant yeasts, specifically, S. rouxii. In 1977, Warth (38) proposed a mechanism of resistance of S. bailii to benzoic and sorbic acids, involving an inducible energy-requiring system which transports the preservative out of the cell. Since elevated glucose levels in the growth media cause S. bailii to become more resistant to the preservative (25), aerobic respiration of glucose could provide the energy needed to pump the preservative out of the cell (38). In the present investigation, each a, level contained 16% glucose, and the cultures were vigorously shaken to permit sufficient aeration. If Warth's theory (38) can be applied to S. rouxii, the sorbate resistance of this yeast should not be markedly influenced by moderate aw changes. For a particular sublethal sorbate level, the growth rate of S. rouxii should be fastest near the optimum aw level. However, the higher aws (.98 to >.995) enhanced the antimycotic activity of sorbate against the growth of S. rouxii, whereas lower aws (.9 to.92, farther from the optimum) increased the tolerance of this yeast to higher concentrations of sorbate. From the data obtained in this investigation, there are two other possibilities involving the mechanism of resistance of S. rouxii to sorbate. At lower aw levels (higher solute concentrations) the cells shrink (11), causing the pore size on the cell membrane to become smaller, which may retard the flow of sorbate into the cell. If this theory is valid, the induced energy system (Warth's theory [38]) to transport sorbate out of the cell could still be plausible. At lower aw levels the flow of sorbate into the cell will be retarded, allowing the energy formed by glucose respiration to pump the sorbate out at a rate at least equivalent to that of the uptake of the preservative. At the higher aw values and sorbate levels (>.6%), the uptake of sorbate would be too rapid, causing an intracellular accumulation of the preservative. The resistance of S. rouxii to sorbate at different aw values could also involve the production of polyols (6) functioning as a compatible solute (7). Potassium sorbate inhibits various enzyme

7 162 RESTAINO ET AL. systems (2, 15, 19, 39). With decreasing aw levels, increasing amounts of polyols will be produced or retained intracellularly to maintain osmoregulation (8) and will function as a compatible solute protecting various enzymes from the inhibition or inactivation of certain antimicrobial agents. At the higher aws (.98 to >.995), the intracellular polyol levels would be too low to effectively counteract the antimycotic effectiveness of the preservative. When the aw is greatly reduced (<.85), the growth rate of S. rouxii is extremely retarded (low metabolic rate), which could cause a depressed rate of polyol production, allowing the yeast cells to become more sensitive to the preservative. This theory of sorbate resistance is further supported by the data of Bills et al. (5), who have determined the various growth responses of preconditioned S. rouxii cells in media containing different levels of sucrose and sorbate. Further studies are being conducted to analyze intracellular polyols at different aw levels and relate this to sorbate resistance of S. rouxii. As noted above, the aw level of the medium influenced the resistance of S. rouxii to potassium sorbate. S. rouxii can tolerate increasing concentrations of potassium sorbate at decreasing aw levels. The addition of potassium sorbate (.1%) to a liquid, intermediate-moisture food (ph 4.5 to 5.5) such as syrup toppings will not ensure that the growth of osmotolerant yeasts can be retarded. Consequently, a variety of factors must be considered, including the chemical constituents of the food (sugar concentration, salt level, fat content, and ph), processing, packaging, storage temperature and length, and use of additional antimicrobial agents that might enhance or hinder the effectiveness of sorbate (29). ACKNOWLEDGMENT We greatly appreciate the DNA hybridization experiments performed by C. P. Kurtzman. LITERATURE CITED 1. Anand, J. C., and A. D. Brown Growth rate patterns of the so-called osmophilic and non-osmophilicl yeasts in solutions of polyethylene glycol. J. Gen. Microbiol. 52: Azukas, J. J., R. N. Costilow, and H. L. Sadoff Inhibition of alcoholic fermentation by sorbic acid. J. Bacteriol. 81: Baird-Parker, A. C., and W. J. Kooiman Soft drinks, fruit juices, concentrates, and fruit preserves, p In International Commission on Microbiological Specifications for Foods (ed.), Microbial ecology of foods, vol. 2. Academic Press, Inc., New York. 4. Beuchat, L. R Combined effects of solutes and food preservatives on rates of inactivation of and colony formation by heated spores and vegetative cells of molds. Appl. Environ. Microbiol. 41: Bills, S., L. Restaino, and L. M. Lenovich Growth response of an osmotolerant sorbate-resistant yeast, Saccharomyces rouxii, at different sucrose and sorbate levels. APPL. ENVIRON. MICROBIOL. J. Food Prot. 45: Brown, A. D Microbial water relations: features of the intracellular composition of sugar-tolerant yeasts. J. Bacteriol. 118: Brown, A. D Microbial water stress. Bacteriol Rev. 4: Brown, A. D., and M. Edgley Osmoregulation in yeast, p In D. W. Rains, R. C. Valentine, and A. Hollaender (ed.), Genetic engineering of osmoregulation. Plenum Publishing Corp., New York. 9. Brown, A. D., and J. R. Simpson Water relations of sugar-tolerant yeasts: the role of intracellular polyols. J. Gen. Microbiol. 72: Christian, J. H. B Reduced water activity, p In International Commission on Microbiological Specifications for Foods (ed.), Microbial ecology of foods, vol. 1. Academic Press, Inc., New York. 11. Corry, J. E. L The effect of sugars and polyols on the heat resistance and morphology of osmophilic yeasts. J. Appl. Bacteriol. 4: Doyle, M. P., and E. H. Marth Thermal inactivation of condia from Aspergillus flavus and Aspergillus parasiticus. J. Milk Food Technol. 38: Edgley, M., and A. D. Brown Response of xerotolerant and non-tolerant yeasts to water stress. J. Gen. Microbiol. 14: Gibson, B The effect of high sugar concentrations on the heat resistance of vegetative microorganisms. J. Appl. Bacteriol. 36: Harada, K., R. Higuchi, and I. Utsunii Studies on sorbic acid. IV. Inhibition of the respiration of yeast. Agric. Biol. Chem. 32: Ingram, M Physiological properties of osmophilic yeasts. Rev. Ferment. Ind. Aliment. 14: Kurtzman, C. P., M. J. Smiley, C. J. Johnson, L. J. Wickerham, and G. B. Fuson Two new and closely related heterothallic species, Pichia amylophila and Pichia mississippiensis: characterization by hybridization and deoxyribonucleic acid reassociation. Int. J. Syst. Bacteriol. 3: Lodder, J The yeasts. North-Holland Publishing Co., Amsterdam. 19. Martoadipraivito, W., and J. R. Whitaker Potassium sorbate inhibition of yeast alcohol dehydrogenase. Biochim. Biophys. Acta 77: Moran, J. W., and L. D. Witter Effects of sugars on D-arabitol production and glucose metabolism in Saccharomyces rouxii. J. Bacteriol. 138: Mrak, E. M., H. J. Phaff, R. H. Vaughn, and H. N. Hansen Yeasts occurring in souring figs. J. Bacteriol. 44: Nury, F. S., M. W. Miller, and J. E. Brekke Preservative effect of some antimicrobial agents on high-moisture dried fruits. Food Technol. (Chicago) 14: Onishi, H Studies on osmophilic yeasts. V. Factors affecting growth of soy yeasts and others in the environment of a high concentration of sodium chloride (2). Bull. Agric. Chem. Soc. Jpn. 23: Onishi, H Osmophilic yeasts. Adv. Food Res. 12: Pitt, J. I Resistance of some food spoilage yeasts to preservatives. Food Technol. Aust. 26: Pitt, J. I., and K. C. Richardson Spoilage by preservative-resistant yeasts. CSIRO Food Res. Q. 33: Scott, W. J Water relations of food spoilage microorganisms. Adv. Food Res. 7: Seiler, D. A. L Yeast spoilage of bakery products, p In F. A. Skinner, S. M. Passmore, and R. R. Davenport (ed.), Biology and activities of yeasts. Academic Press, Inc., New York. 29. Sofos, J. N., and F. F. Busta Antimicrobial activity of sorbate. J. Food Prot. 44: Splittstoesser, D. F Fruits and fruit products, p In L. R. Beuchat (ed.), Food and beverage mycology. AVI Publishing Co., Westport, Conn.

8 VOL. 45, Tilbury, R. H The microbial stability of intermediate moisture foods with respect to yeasts, p In R. Davies, G. G. Birch, and K. J. Parker (ed.), Intermediate moisture foods. Applied Sciences, London. 32. Tilbury, R. H Xerotolerant (osmophilic) yeasts, p In F. A. Skinner, S. M. Passmore, and R. R. Davenport (ed.), Biology and activities of yeasts. Academic Press, Inc., New York. 33. Troller, J. A., and J. H. B. Christian Water activity and food. Academic Press, Inc., New York. 34. U.S. Food and Drug Administration Bacteriological analytical manual. Bureau of Foods, U.S. Food and Drug Administration, Washington, D.C. 35. Von Schelhorn, M Untersuchungen uber den Verderb wasserarmer Lebensmittel durch osmophile Mikroorganismen. I. Verberb von Lebensmittel durch osmo- GROWTH PARAMETERS OF S. ROUXII 1621 phile Hefen. Z. Lebensm. Unters. Forsch. 91: Walker, H. W Spoilage of food by yeasts. Food Technol. (Chicago) 31: Walker, H. W., and J. C. Ayres Yeasts as spoilage organisms, p In A. H. Rose and J. S. Harrison (ed.), The yeasts, vol. 3. Academic Press, Inc., New York. 38. Warth, A. D Mechanism of resistance of Saccharomyces bailii to benzoic, sorbic, and other weak acids used as food preservatives. J. Appl. Bacteriol. 43: Whitaker, J. R Inhibition of sulfhydryl enzymes with sorbic acid. Food Res. 24: Zottola, E. A Osmophilic microorganisms, p In M. L. Speck (ed.), Compendium of methods for the microbiological examinations of foods. American Public Health Association, Washington, D.C. Downloaded from on June 27, 218 by guest