Heat Resistance of Juice Spoilage Microorganisms

Transcription

1 1271 Journal of Food Protection, Vol. 65, No. 8, 2002, Pages Copyright Q, International Association for Food Protection Heat Resistance of Juice Spoilage Microorganisms ADRIENNE E. H. SHEARER, ALEJANDRO S. MAZZOTTA, ROLENDA CHUYATE, AND DAVID E. GOMBAS* Center for the Development of Research Policy and New Technologies, National Food Processors Association, 1350 I Street N.W., Suite 300, Washington, D.C , USA MS : Received 11 October 2001/Accepted 30 March 2002 ABSTRACT The heat resistance of various yeasts (Saccharomyces cerevisiae, Rhodotorula mucilaginosa, Torulaspora delbrueckii, and Zygosaccharomyces rouxii), molds (Penicillium citrinum, Penicillium roquefortii, and Aspergillus niger), and lactic acid bacteria (Lactobacillus fermentum and Lactobacillus plantarum) obtained from spoiled acid or acidi ed food products was determined in 0.1 M citrate buffer at ph values of,, and. S. cerevisiae was the most heat resistant of the microorganisms in citrate buffer, and its heat resistance was further evaluated in apple, grapefruit, calcium-forti ed apple, and tomato juices as well as in a juice base with high fructose corn syrup. Decimal reduction times (D-values) and changes in temperature required to change the D-value (z-values) for S. cerevisiae were higher in the juices than in citrate buffer at all ph values tested. The D 578C(1358F) -values varied from 9.4 min in the juice product with ph 2.8 to 32 min in a calcium-added apple juice with ph 3.9. The S. cerevisiae strain used in this study can be used in thermal-death-time experiments in acidic products to calculate process conditions and in challenge tests to validate the calculated temperatures and hold times during processing. Commercial sterility of hot- ll-hold processed juices is achieved by the heat pasteurization of the juice and the subsequent inactivation of spoilage organisms on the container and closure by the heat transferred from the hot- lled juice. Yeasts, molds, and acid-tolerant bacteria associated with raw fruit, juice concentrates, and the juice processing plant environment are the spoilage microorganisms of concern for the adequate heat processing of juices and their containers. The spoilage of pasteurized, single-strength juice products by Penicillium and Cladosporium has been reported, although other mold genera have been isolated from citrus juices, including Aspergillus, Alternaria, Byssochlamys, and Aureobasidium, among others (13). Adequate oxygen in the container headspace is necessary for the proliferation of molds in single-strength juice products. According to a survey on the yeast ora of frozen fruit juice concentrates, the isolates recovered represented 12 genera and 21 species of yeast (3). The ve most frequently isolated yeast species included Saccharomyces cerevisiae (24.7%), Candida stellata (22.1%), Zygosaccharomyces rouxii (14.3%), Torulaspora delbrueckii, and Rhodotorula mucilaginosa (3). Hot- ll-hold pasteurization conditions vary by processor, but typical processing conditions for citrus juices range from 85 to 958C for 15 to 60 s (6). Commercial sterilization of the container is typically achieved by lling the heated juice product and holding for 30 s to 2 min before cooling. These pasteurization parameters were established to ensure * Author for correspondence. Tel: ; Fax: ; dgombas@nfpa-food.org. Present address: University of Delaware, Department of Animal and Food Sciences, Newark, DE 19717, USA. Present address: General Mills, James Ford Bell Technical Center, 9000 Plymouth Avenue North, Minneapolis, MN , USA. the inactivation of pectinmethylesterase, thus stabilizing the cloud of orange juice and other nonclari ed juices. The processing conditions were also found to be adequate for microbial inactivation (6) and are in excess of those needed to inactivate pathogenic microorganisms, such as Escherichia coli O157:H7, Salmonella, and Listeria monocytogenes (10). While the parameters for juice pasteurization for nonclari ed juices may be driven by the minimum enzyme denaturation conditions, the hot- ll-hold portion of the process could be optimized according to the requirements for the inactivation of spoilage microorganisms. It is desirable to minimize the heat treatment applied to juices for product quality and producer bene ts. Consumers may perceive unpasteurized or lightly heated juices to have a higher avor quality than juices treated by more extreme heat (12). Reduced heat processing could also conserve energy expended during processing and create more packaging options. Glass containers, although sometimes given a heat pretreatment before lling to prevent shattering, can withstand the temperatures of hot- lled juices. Likewise, metal containers can be used for packaging fruit juice products, provided the cans have an acid-resistant lacquer (17). Lightweight, shatterproof plastic bottles, however, cannot withstand high ll temperatures (11), and ll temperatures in excess of the thermal tolerance for plastic can result in relaxation and distortion of the container. While there are several incentives to optimize the heatprocessing conditions for shelf-stable, single-strength fruit juice products, the data needed to do so are lacking. A considerable number of studies have been conducted on the heat resistance of mold and yeast ascospores, which are the most heat-resistant structures of fungi (5, 9, 15, 20, 21). However, existing processing conditions are not adequate for the destruction of all ascospores (6), and contamination

2 1272 SHEARER ET AL. J. Food Prot., Vol. 65, No. 8 and subsequent spoilage by ascospores must be prevented by stringent sanitation. Limited data are available on the heat resistance of vegetative yeasts (2, 8, 15, 16, 23, 24) and mold conidia (1) in model systems. Studies concerned with such organisms have generally focused on a particular variable in a buffer or broth system, such as the sugar composition (24) or synergistic effects with preservatives (1, 2). Put et al. (16) reported extensive heat inactivation data but without the decimal reduction times (D-values) or the changes in temperature required to change the D-value (zvalues) necessary to calculate process conditions. No data were found on the heat resistance of vegetative yeasts and mold conidia in actual fruit juice products. This study investigated the relative heat resistance of yeast, mold, and lactic acid bacteria (LAB) spoilage isolates in buffer over the ph range of to. The most heatresistant organism in the buffer system, S. cerevisiae, was then evaluated for heat resistance in ve types of juice products. MATERIALS AND METHODS Microbial strains and culture conditions. Yeast isolates of spoiled acid or acidi ed food products (R. mucilaginosa, S. cerevisiae, Z. rouxii, and T. delbrueckii) were obtained from Dr. Larry Beuchat, University of Georgia, Grif n, Ga. Yeast identi cations were con rmed with the API 20C system (BioMerieux, Hazelwood, Mo.). Mold cultures (Penicillium roquefortii, Penicillium citrinum, and Aspergillus niger) were obtained from spoiled acid food products of commercial processors. Molds were identi ed microscopically (18) and macroscopically by characteristic growth on malt extract agar, Czapek yeast agar, and 25% glycerol nitrate agar according to Pitt and Hocking (14). LAB (Lactobacillus fermentum and Lactobacillus plantarum) were obtained from spoiled acid products of commercial processors. LAB were identi ed with the API 50C system (BioMerieux). Yeasts were grown on Sabouraud dextrose agar (Difco Laboratories, Sparks, Md.) for no longer than 4 days at 308C. Vegetative cells were harvested from the Sabouraud dextrose agar with phosphate-buffered saline ph 7 and a sterile bent glass rod and were stored at 48C for a maximum of 4 weeks. The suspension population was determined by pour plating on Sabouraud dextrose agar after appropriate decimal dilutions and incubation at 308C for 5 days. Molds were grown on yeast mold (YM) agar (Difco) for approximately 10 days at 258C. Conidia were harvested from the YM agar with Butter eld s buffered phosphate with 0.1% Tween 20 (Fisher Scienti c, Pittsburgh, Pa.) and a sterile bent glass rod, ltered through sterile cheesecloth, and stored under refrigeration for a maximum of 5 weeks. Suspensions were enumerated on YM agar. LAB were stored on deman Rogosa Sharpe (MRS) agar (Difco) slants under refrigeration. For each experiment, LAB were inoculated into MRS broth and incubated at 308C for approximately 17 h to the stationary phase of growth. Cells were pelleted and washed twice with citrate buffer. Suspension populations were determined by pour plating in MRS agar. Heat resistance studies. For studies in buffer, the heat resistances of the yeasts, molds, and LAB were determined in 0.1 M citrate buffer at ph,, and using an endpoint procedure in thermal-death-time(tdt) tubes. Sterile TDT glass tubes (9 mm i.d. by 10 cm) were lled with 1 ml of culture suspension (10 5 CFU/ml) in buffer. The tubes were heat sealed, inserted in holders, suspended from a rod, and fully immersed in an oil bath preheated to the desired temperature according to the method of Scott and Bernard (19) using ve temperatures and 10 tubes for each of six time intervals. Preliminary experiments were done with LAB using three to ve temperatures and ve tubes for each of six time intervals. Tubes were cooled in an ice bath immediately after heating. A reference tube containing a copper-constantan thermocouple in the approximate center of the heating menstruum was heated simultaneously with the sample tubes to monitor the buffer temperature. Temperature data were collected with a validator (Kaye Instruments, Bedford, Mass.). After cooling, the contents of each TDT tube were aseptically transferred to 9 ml YM broth adjusted to the same ph as the heating menstruum with 0.1 M citric acid (Fisher) for fungi. LAB were incubated in MRS broth with the ph adjusted as for the yeasts and molds. The subcultured tubes were incubated at 308C and monitored for the growth of survivors as indicated by turbidity. Turbidity was evaluated in comparison to positive controls of inoculated recovery broth and negative controls of uninoculated recovery broth tubes. Incubation was carried out 1 week beyond the observance of new samples positive for growth (or a minimum of 21 days). The population of the initial inoculum was enumerated by pour plating in YM agar for fungi and in MRS agar for LAB. The heat resistance of S. cerevisiae in juice products was determined in tomato juice at ph 4.2 and 4.5, in grapefruit juice (Brix ca. 10.4) at ph 3.3, in apple juice (Brix ca. 11.7) at ph and 3.9, in commercial apple juice enriched with calcium hydroxide at ph 3.9, and in a juice product (Brix ca. 1) at ph 2.8 containing juice base, malic acid, high fructose corn syrup, and water. The ph values were adjusted with 1 M or 0.1 M NaOH. TDT experiments for S. cerevisiae in juices were conducted using an endpoint procedure in 1,000-ml three-neck asks, following a previously described procedure (10), to eliminate comeup time considerations. After the heating time, samples were recovered in tubes containing the same juice and incubated at 308C. The initial concentration of cells was determined by pour plating the inoculum with tempered Sabouraud dextrose agar after appropriate decimal dilutions and incubating the plates overnight at 308C. TDT experiments with S. cerevisiae in juices were run as in buffer studies with 10 recovery tubes per interval to increase con dence. The heat resistance in some juices was determined in triplicate (Table 1). Calculation of D- and z-values. At each temperature, D- values were determined by the formula D 5 t/(log A 2 log B), where t is the heating time in minutes (corrected for come-up time for submerged TDT tubes); A is the initial number of organisms in the ask or tubes; and B is the number of surviving cells after heat treatment. B was estimated by the most probable number procedure of Halvorson and Ziegler (7) using the formula B 5 ln(total number of tubes for an interval/number of sterile tubes). For experiments in which all samples were positive at one time interval and all were negative at the next time interval, the D- value was estimated by assuming one negative at the last allpositive interval and one positive at the rst all-negative time interval, calculating separate D-values for each assumption, and averaging the values. For the determination of D-values not tested experimentally, the log D-values of those determined experimentally were plotted against the respective temperatures, and the best- t line was drawn. The z-values were calculated as the negative inverse slope of the linear regression line for the log D- values over the range of heating temperatures tested. RESULTS D- and z-values of the fungi in citrate buffer are presented in Table 2. Because of the wide variation in heat

3 J. Food Prot., Vol. 65, No. 8 HEAT RESISTANCE OF JUICE SPOILAGE MICROORGANISMS 1273 TABLE 1. Heat resistance of Saccharomyces cerevisiae in juice products a Product ph D-value (min) at temp (8C/8F) 57/135 60/140 63/145 z-value (8C/8F) Tomato /9.4 Tomato /7.7 Grapefruit juice /11 Apple juice /7.2 Apple juice /11 Calcium-forti ed apple juice /9.2 Juice product /11 a Values with standard deviations are averages of three independent experiments. Values without standard deviations are results from a single trial with 10 recovery tubes per time interval. Z-values were calculated from the plots of the averages of the D-values. The r- squared values for z-value determination were 0.98 or greater. resistances of the yeasts and molds, various temperature ranges were used in the experiments; therefore, the D-values presented in Table 2 are extrapolated or interpolated values from the linear regression line for log D-values for the temperatures tested experimentally. D-values are presented at one temperature for direct comparison of all microorganisms. The r-squared values for the regression lines of the D-values versus temperature for buffer studies were 0.95 or greater. Preliminary studies demonstrated that the heat resistances of LAB were lower than those of P. citrinum (data not shown); therefore, no further studies were conducted with LAB. The change in buffer ph did not consistently correlate with an increase or decrease in heat resistance among the fungi. The heat resistance of S. cerevisiae was at least 2.5 times greater in citrate buffer than the other organisms tested and was selected for subsequent studies in the juices. The D- and z-values of S. cerevisiae in the juices and juice product are presented in Table 1. The heat resistance of S. cerevisiae in apple juice at ph 3.9 most closely resembled that of the same microorganism in the citrate buffer at ph (with z-values near 48C and D 608C(1408F) values near 2.0 to 2.5 min). Otherwise, S. cerevisiae had greater heat resistance in the juice products than in citrate buffer, as indicated by higher D-values as well as higher z-values. In juices, higher D-values did not correlate with higher ph values. DISCUSSION Determination of the appropriate processing conditions for hot- lled shelf-stable juices can be achieved by trial and error for new products and processing methods. However, it is more ef cient for process optimization and new product development to determine the heat resistance of microorganisms of importance to the product over a range of conditions. The aim of this study was to evaluate the heat resistances of a variety of microorganisms isolated from spoiled products. The relative heat resistance was rst determined in citrate buffer at a variety of ph values, and then further studies were conducted with the most resistant microorganism in actual juice products preserved solely by heat pasteurization. An extensive study of the heat resistance of 120 yeast strains in phosphate citrate buffer (ph 4.5) by Put et al. (16) concluded that S. cerevisiae was not only the most frequently isolated and most commonly organism associated with the spoilage of heat-processed soft drinks and fruit juices among other Saccharomyces spp., but also the most heat resistant. It is dif cult to compare the heat resistance determined in the present study to that determined by Put et al. (16) because of the differences in experimental methods and reporting. At ph 4.5, S. cerevisiae (10 5 initial population) had a 50% survival rate after treatment at 658C for 10 min but was completely inactivated after 20 min (16). The determination of D- and z-values in the present study allows for calculation and better prediction of resistance at temperatures not tested experimentally. The heat resistance of S. cerevisiae in the various juices demonstrates the complexity of factors that in uence heat resistance. The behavior of S. cerevisiae in citrate buffer at ph was most similar to that in apple juice at ph 3.9. In all other products, the resistance was greater than in buffer, with both higher D- and z-values. Heat resistance in the various juices appeared to be dependent on composition as well as ph, as no consistent trend was observed solely as a function of ph. The focus of this study was not aimed at elucidating the effects of the individual variables or the mechanisms of inhibition; however, data reported in the literature support the observation that several parameters may affect heat resistance, including acid (8) and sugar composition (24). No data on the effect of calcium addition to fruit juices on the heat resistance of spoilage microorganisms were found in the literature. The present study indicated that S. cerevisiae heat resistance in calcium-added apple juice was more than twice that of apple juice at the same ph (3.9) or a reduced ph () (Table 1). Whether calcium interacts with other inhibitory chemical constituents of apple juice or plays a more biochemically protective role for yeast cells during heating is unclear but is worthy of further investigation. The possible roles of calcium in yeast activity are

4 1274 SHEARER ET AL. J. Food Prot., Vol. 65, No. 8 TABLE 2. Calculated heat resistance of molds and yeasts in 0.1 M citrate buffer at various ph levels Organism Experimental heating range (8C) ph D-value (min) at temp 608C/1408F z-value (8C/8F) Penicillium citrinum Torulaspora delbrueckii Rhodotorula mucilaginosa Zygosaccharomyces rouxii Penicillium roquefortii Aspergillus niger Saccharomyces cerevisiae / / / / / / / / / / /4.3 / / / / / / /6.6 /7.2 /6.4 a ND, not determined; all values from single nal determination following preliminary trials. broad, and some of these proposed functions are associated with yeast stability and resistance. For example, calcium may generate the sporulation signal by increased in ux of calcium ions (22) and may stabilize cell membranes (4). It is not known how calcium affected the S. cerevisiae in this study, but if calcium could have in uenced sporulation initiation or stabilized membranes within the time frame of the experiment, either of these effects may reasonably be hypothesized to increase heat resistance. Processors considering nutrient forti cation of juice products should evaluate the potential effect on the heat resistance of microorganisms. This study focused on the heat resistance of vegetative fungi and LAB as primary spoilage microorganisms of fruit juice products within the ph range of to. S. cerevisiae, as observed in the present study, had a greater heat resistance than bacterial pathogens of concern in the same products (10). A process of 3 s at 71.18C (z-value C) can inactivate 5 logs of Salmonella, E. coli O157:H7, or L. monocytogenes (10). Process conditions established for the destruction of the spoilage microorganisms must always be in excess of those needed for the destruction of pathogens and should be reevaluated with any change in product or processing. There are other spoilage microorganisms of importance in some of the products tested, including ascospores, which may be 100-fold more resistant (20) than vegetative fungi, as well as Alicyclobacillus, a spore former capable of outgrowth in some juices even at ph values less than (25). Additionally, tomato juice at ph 4.5 may potentially support the outgrowth of Bacillus coagulans spores. Therefore, stringent sanitation as well as additional inactivation treatments or inhibitory methods in excess of the heat treatment evaluated for this study may be necessary for these products before, during, or after processing. ACKNOWLEDGMENTS The authors gratefully acknowledge the assistance provided by Donald Kautter, Jr., Virginia N. Scott, Karen Suit, and members of the National Food Processors Association Juice Products Committee. REFERENCES 1. Beuchat, L. R In uence of potassium sorbate and sodium benzoate on heat inactivation of Aspergillus avus, Penicillium puberulum, and Geotrichum candidum. J. Food Prot. 44: Beuchat, L. R Synergistic effects of potassium sorbate and sodium benzoate on thermal inactivation of yeasts. J. Food Sci. 46: Deak, T., and L. R. Beuchat Yeasts associated with fruit juice concentrates. J. Food Prot. 56: Frausto da Silva, J. J. R., and R. J. P. Williams (ed.) Calcium: controls and triggers, p In The biological chemistry of the elements, the inorganic chemistry of life. Clarendon Press, Oxford. 5. Garg, N., S. K. Kalra, and D. K. Tandon Heat resistance studies of spoilage yeasts from preserved mango slices. J. Food Sci. Technol. 34: Graumlich, T. R., J. E. Marcy, and J. P. Adams Aseptically packaged orange juice and concentrate: a review of the in uence of processing and packaging conditions on quality. J. Agric. Food Chem. 34: Halvorson, H. O., and N. R. Ziegler Application of statistics to problems in bacteriology. J. Bacteriol. 25: Juven, B. J., J. Kanner, and H. Weisslowicz In uence of orange juice composition on the thermal resistance of spoilage yeasts. J. Food Sci. 43: , King, A. D., Jr., and W. U. Halbrook Ascospore heat resistance and control measures for Talaromyces avus isolated from fruit juice concentrate. J. Food Sci. 52: , Mazzotta, A. S Thermal inactivation of stationary-phase and

5 J. Food Prot., Vol. 65, No. 8 HEAT RESISTANCE OF JUICE SPOILAGE MICROORGANISMS 1275 acid-adapted Escherichia coli O157:H7, Salmonella and Listeria monocytogenes in fruit juices. J. Food Prot. 64: McLellan, M. R., L. R. Lind, and R. W. Kime A shel ife evaluation of an oriented polyethylene terephthalate package for use with hot lled apple juice. J. Food Sci. 52: Parish, M. E Microbiological concerns in citrus juice processing. Food Technol. 45: Parish, M. E., and D. P. Higgins Yeasts and molds isolated from spoiling citrus products and by-products. J. Food Prot. 52: Pitt, J. I., and A. D. Hocking Fungi and food spoilage. Academic Press, Sydney. 15. Put, H. M. C., and J. De Jong Heat resistance studies of yeasts; vegetative cells versus ascospores: erythromycin inhibition of sporulation in Kluyveromyces and Saccharomyces species. J. Appl. Bacteriol. 53: Put, H. M. C., J. De Jong, F.E.M.J. Sand, and A. M. Van Grinsven Heat resistance studies on yeast spp. causing spoilage in soft drinks. J. Appl. Bacteriol. 40: Robertson, G. L. (ed.) Packaging of beverages, p In Food packaging principles and practice. Marcel Dekker, Inc., New York. 18. Samson, R. A., E. S. Hoekstra, J. C. Frisvad, and O. Filtenborg (ed.) Methods for the detection and isolation of food-borne fungi, p In Introduction to food-borne fungi, 4th ed. Centraalbureau Voor Schemmelcultures, Baarn, The Netherlands. 19. Scott, V. N., and D. T. Bernard Heat resistance of spores of non-proteolytic type B Clostridium botulinum. J. Food Prot. 45: Splittstoesser, D. F., S. B. Leasor, and K. M. J. Swanson Effect of food composition on the heat resistance of yeast ascospores. J. Food Sci. 51: Splittstoesser, D. F., and C. M. Splittstoesser Ascospores of Byssochlamys fulva compared with those of a heat resistant Aspergillus. J. Food Sci. 42: Suizu, T., H. Tsutsumi, A. Kawado, K. Suginami, S. Imayasu, and K. Murata Calcium ion in ux during sporulation in the yeast Saccharomyces cerevisiae. Can. J. Microbiol. 41: Torok, T., and A. D. King, Jr Thermal inactivation kinetics of food-borne yeasts. J. Food Sci. 56:6 9, Torreggiani, D., and R. T. Toledo In uence of sugars on heat inactivation, injury and repair of Saccharomyces cerevisiae. J. Food Sci. 51: Walls, I., and R. Chuyate Alicyclobacillus historical perspective and preliminary characterization study. Dairy Food Environ. Sanit. 18: