57 Journal of Food Protection, Vol. 6, No.,, Pages 57 5 Copyright, International Association for Food Protection Detection of Guaiacol Produced by Alicyclobacillus acidoterrestris in Apple Juice by Sensory and Chromatographic Analyses, and Comparison with Spore and Vegetative Cell Populations RACHEL V. ORR, ROBERT L. SHEWFELT, C. J. HUANG, SEBHAT TEFERA, AND LARRY R. BEUCHAT * Center for Food Safety and Quality Enhancement, Department of Food Science and Technology, University of Georgia, Griffin, Georgia - 797; Center for Food Safety and Quality Enhancement, Department of Food Science and Technology, University of Georgia, Athens, Georgia 6-76; and National Food Processors Association, 5 I Street, N.W., Washington, D.C. 5, USA MS -: Received April /Accepted June ABSTRACT Spoilage of fruit juice by Alicyclobacillus acidoterrestris is characterized by a distinct medicinal or antiseptic off odor attributed to guaiacol, a metabolic by product of the bacterium. Detection of low populations of A. acidoterrestris that would precede sensory detection of guaiacol would enable juice processors to select appropriate processing and storage conditions that would minimize or eliminate spoilage. The objective of this study was to determine the recognition threshold of guaiacol in apple juice by sensory analysis and the population of A. acidoterrestris and incubation time at and 7C necessary for chemical detection of guaiacol. Commercially sterilized apple juice (ph.54.4,..brix) was inoculated with a five-strain mixture of A. acidoterrestris spores (.98 log CFU/ml) and stored at or 7C for up to 6 days. Using an experienced sensory panel and the forced-choice ascending concentration method of limits, the best estimate threshold (BET) for recognition of guaiacol added to uninoculated apple juice was. ppb. Numbers of A. acidoterrestris spores and cells in inoculated juice remained constant during the 6-day storage period; however, the panel detected (P.) guaiacol in juice stored at 7C within 8 days. At three of four sampling times ranging from to 6 days at which the sensory panel detected (P.) guaiacol, concentrations of 8. to.4 ppb were detected by chromatographic analysis. The panel detected (P. to P.) guaiacol in five samples stored at to 7C for 8 to 6 days in which the compound was not detected by chromatographic analyses. It appears that guaiacol content in apple juice inoculated with A. acidoterrestris is not always correlated with numbers of cells, and the limit of sensitivity of chromatographic quantitation of the compound is higher than the BET. In 98, aseptically packaged apple juice produced in Germany spoiled during distribution and storage in what was an unusually long warm season (). Spoilage was characterized as off flavor and slight cloudiness in the juice. The bacterium causing this spoilage was Alicyclobacillus spp. Temperatures at which Alicyclobacillus acidoterrestris will grow range from 5 to 55C (), 5 to 44C (9), and to 55C (), depending upon the strain and composition of the medium. The temperature range for germination of A. acidoterrestris spores has been reported to be 6 to 8C (6), 6 to 5C (), and 4 to 5C (4). Alicyclobacillus spores have been shown to resist exposure to acidic environments and temperatures used in hotfill processes for fruit and vegetable juices, rendering the organism as a cause of potential spoilage in these products. The bacterium has been observed to grow at ph.5 to 5.5 (), ph.5 to 5.8 (5), ph. to 5.8 (), and ph. to 5.5, with an optimum range of ph.5 to 5. (). Walls and Chuyate () observed that A. acidoterrestris grew well in orange serum broth (ph 5.) and on K medium (ph.7) and potato dextrose agar (ph 4.5 and 5.). These research- * Author for correspondence. Tel: 77-4-474; Fax: 77-9-6; E-mail: lbeucha@cfsqe.griffin.peachnet.edu. ers reported a relationship between incubation temperature and ph of the culture medium. When incubated at C, five of five isolates of A. acidoterrestris grew in laboratory broth at ph 5. and two grew at.5, whereas the opposite occurred when broth was incubated at 55C. Spoilage of fruit-based products by Alicyclobacillus has been characterized as the development of an off odor or off flavor described as medicinal () or antiseptic (). A chemical responsible for the off odor has been identified as guaiacol (), which, according to Pettipher et al. (9), at ppb can be detected by sensory means in fruit juices. Guaiacol was detected in orange juice and apple juice when 5 cells of A. acidoterrestris per ml were present (9). Gas is not produced and there is not a substantial change in the ph of fruit juices in which the organism has grown (, ). A survey conducted by Walls and Chuyate () revealed the spoilage went unnoticed until consumer complaints were received. Spoilage was usually described as an off flavor or odor, with or without sediment, and tended to occur most often in the spring or summer. Apple juice was the most frequently spoiled product. Because spoilage of apple juice by Alicyclobacillus is characterized by an off odor caused, in part, by guaiacol, a
58 ORR ET AL. J. Food Prot., Vol. 6, No. study was undertaken to confirm the threshold for human detection of guaiacol. The number of A. acidoterrestris cells per ml of apple juice and the incubation time at and 7C required to produce detectable amounts of guaiacol by sensory evaluation and chromatographic analysis were determined. The ultimate goal was to provide information that can be used to predict the shelf life of pasteurized or hot-filled apple juice containing viable spores of A. acidoterrestris. MATERIALS AND METHODS Procedure for determining odor threshold of guaiacol in apple juice. The odor threshold for recognition of guaiacol was determined by using the forced-choice ascending concentration method of limits described by the American Society for Testing and Materials (). Apple juice (Minute Maid, Coca Cola Company, Atlanta, Ga.) in glass bottles was supplemented with guaiacol (Sigma, St. Louis, Mo.) at concentrations of,.6,.8, 5., 6, and 48 ppb. Prior to analysis of test samples, panelists were presented with apple juice containing 48 ppb guaiacol to enable them to recognize the odor. Tests to determine the recognition threshold of guaiacol odor were conducted using an experienced panel. The panel consisted of females and 9 males with ages ranging from to 55 years. Evaluations were conducted between : and 4: p.m. in a temperature-controlled ( C) sensory panel room containing eight side-by-side booths. Red lighting was used to mask potential differences in color of samples. Coded samples ( ml, C) prepared to 4 h before evaluation were poured into 5-ml polyethylene cups and immediately capped for presentation to 9 panelists. Two of three cups of apple juice presented to each panelist contained no guaiacol, and one cup contained apple juice with a known concentration of guaiacol. Panelists were asked to smell the three samples and decide which sample was different from the other two. A choice was required, even if the panelist was not able to discern a difference, to obtain a full complement of data. Panelists were first presented with a set of three samples, one of which contained 48 ppb guaiacol, followed in random order by four sets of three samples, one sample in each set containing a different concentration of guaiacol than the other two. Each evaluation was repeated three times. Calculations of recognition threshold. The best estimate threshold (BET) was calculated for each panelist using the geometric mean of the last missed concentration and the next highest concentration (). The average panel threshold was determined by calculating the geometric mean of the BET of three replicate tests administered to each panelist, then calculating the BET of individual panelists. Strains and production of spores. Three strains (N-96, N-, N-8) of A. acidoterrestris were obtained from Dr. Isabel Walls (National Food Processors Association, Washington, D.C.) and two strains (OS-CAJ and SAC) were obtained from Ms. Vickie Lewandowski (Cargill, Minneapolis, Minn.). Strain N- 96 (ATCC 495) was isolated from garden soil in Germany. Strains N-, OS-CAJ, and SAC were isolated from apple juice concentrate, and N-8 was isolated from apple-cranberry juice. Spores were produced, harvested, and suspended in sterile distilled water as described by Orr and Beuchat (8). Suspensions containing approximately equal numbers of spores of each strain were stored at C until used as inocula for apple juice. Procedure for inoculating apple juice. One milliliter of a suspension of a five-strain mixture of A. acidoterrestris spores was inoculated into 5 ml of sterile deionized water at 8C and heated for min. This suspension ( ml) was inoculated into 465 ml of apple juice (Minute Maid) in 475-ml glass jars adjusted to or 7C; caps were applied to jars containing inoculated juice at atmospheric pressure. Uninoculated juice served as a control. To determine the number of spores (CFU/ml) in the inoculum, the heated (8C, min) inoculum was serially diluted in.% peptone and surface plated (. ml in duplicate) on K medium (ph.7) (). Plates were inoculated at 4C for 7 days before colonies were counted. Incubation and microbiological analysis of inoculated apple juice. Capped bottles of inoculated and uninoculated apple juice were stored at C or7c for up to 6 days. Samples were analyzed at 4- to 4-day intervals for numbers of vegetative cells and spores of A. acidoterrestris. To determine combined populations of vegetative cells and spores not requiring heat activation, juice was serially diluted in.% peptone, surface plated (. ml in duplicate) on K medium, and incubated for 7 days at 4C. To determine the number of spores present in the juice, 5 ml were heated (8C, min), serially diluted in.% peptone, surface plated (. ml in duplicate) on K medium, and incubated 7 days at 4C. The soluble solids (Brix) content and ph of apple juice were monitored. Soluble solids was determined using an ABBE-L refractometer (Spectronic Instruments, Rochester, N.Y.); ph was determined using an Accumet Basic ph meter (Fisher Scientific, Pittsburgh, Pa.). Evaluation of inoculated apple juice for off odor. A triangle test was conducted to determine if panelists could detect off odor in inoculate apple juice using the same 9 panel members and test protocol described above for uninoculated apple juice supplemented with guaiacol. On each sampling day, each of 6 to 9 panelists was presented with one -ml ( C) sample of apple juice inoculated with A. acidoterrestris and stored at C and two -ml samples of uninoculated (control) juice, also stored at C for the same length of time. Panelists were asked to smell the three samples and choose which was unlike the other two. The same procedure was followed for inoculated and uninoculated juice stored at 7C. If panelists could not detect a difference in aroma between the three samples in each set, they were required to make a choice to provide a complete data set. Three replicates of each set of samples were analyzed on each sampling day. Quantitative analysis of guaiacol content in apple juice. To apple juice (.4 ml) in a 4-ml vial, g of sodium chloride and l ml of phenol (.5 ppm, internal standard) (Sigma) were added. The vial was placed on the receptacle of a solid-phase microextraction fiber holder on a fiber cartridge that was exposed to the headspace of the test sample for 6 min at 6C. The solidphase microextraction needle was then inserted into the injector port of a gas chromatograph (Hewlett-Packard, 589 series II coupled with a 597 series mass spectrophotometer, Palo Alto, Calif.). The following conditions were used: split/splitless injector in a splitless mode at C; desorption time, min; DB5 capillary column,.5 mm 5 m,.5 m (J & W Co., Folsom, Calif.); column temperature of 5C for 5 min, then increased at 4C/min to 5C, then at C/min to 8C; and helium carrier at cm/s. Data were collected in the selective ion-monitoring mode (m/z 8, 9, and 4 for guaiacol and m/z 66 and 94 for phenol, the internal standard) and are reported as ppb of guaiacol in the apple juice.
J. Food Prot., Vol. 6, No. DETECTION OF GUAIACOL IN APPLE JUICE 59 RESULTS AND DISCUSSION Recognition threshold of guaiacol in apple juice. The number of correct responses distinguishing one sample of apple juice containing guaiacol from two samples not containing guaiacol is shown in Figure. As the concentration of guaiacol increased from.6 ppb to 48 ppb, the number of correct responses increased in all three replications. Table shows the BET for recognition of guaiacol odor by each of 9 panelists in three replicate panels. The recognition threshold in the three panels ranged from.84 ppb to.9 ppb, with a mean BET of. ppb. These results are similar to those reported by Pettipher et al. (9), who observed a threshold of guaiacol in orange, apple, and noncarbonated fruit juices to be about ppb. FIGURE. Correct responses (percentage of 6 to 9 panelists) in distinguishing one sample of apple juice containing guaiacol from two samples containing no guaiacol. Statistical analysis. The experiment was replicated three times. Two samples of each replicate were analyzed for numbers of vegetative cells and spores at each sampling time. Significant differences between control and inoculated samples of apple juice were determined using table T7 titled Triangle test for difference: critical number (minimum) of correct answers presented by Meilgaard et al. (7). Data from soluble solids and ph analysis were subjected to the Statistical Analysis System (SAS Institute, Cary, N.C.) for analyses of variance and Duncan s multiple range test to determine significant differences (P.5) between values. Sensory differences of apple juice inoculated with A. acidoterrestris. Significant sensory differences between inoculated and uninoculated apple juice stored at and 7C were detected (Table ). No significant differences were detected until day 8, when two of three bottles of inoculated apple juice stored at C and (P.) and one of three bottles stored at 7C (P.) were recognized by panelists as being different from uninoculated juice. These differences were also significant when data from all three replicates were combined. While some inconsistencies in data were observed, significant differences (P.) between inoculated and uninoculated samples occurred more frequently at 7C than at C, at the higher level of inoculum, and with increased storage time. Results indicate that prevention of recognizable levels of guaiacol in apple juice is more likely in juice held at C for a short time. ph and soluble solids analysis of inoculated and uninoculated apple juice. The ph of inoculated juice ranged from.5 to.56 over the 6-day storage period at and 7C; soluble solids ranged from. to.7brix. Walls and Chuyate () also observed that growth of A. acidoterrestris did not affect the ph of the medium. Microbiological analysis of inoculated juice. Populations of vegetative cells and spores of A. acidoterrestris apple juice from three replicate trials and corresponding significant differences in odor between inoculated and uninoculated juice are shown in Table. Only a few samples of inoculated apple juice held at C had an odor significantly different from the control juice. Panelists detected differences between inoculated and uninoculated apple juice held at C, although populations of vegetative cells and spores did not change substantially during the 6-day incubation period. On days, 5, 9, and 6, significant differences (P.) between inoculated and uninoculated juice held TABLE. BET of guaiacol (ppb) BET of guaiacol (ppb) a Panelist number Panel Panel Panel.5.4. 9. 7.7 8.,, 6 4, 5, 7, 7, 8,,, 4 8, 9,,, 5 6 9,,, 8, 7, 8, 9 4, 5, 6,, 6 7, 5 9,, 4,,, 5, 7 5, 8, 6, 9, 7, 9,,, 8 4, 6, 4 a Mean BET values for 9 panelists were.9,.84, and. ppb for panels,, and, respectively. The mean BET value for all panels was. ppb.
5 ORR ET AL. J. Food Prot., Vol. 6, No. TABLE. Significant sensory (aroma) differences between inoculated and uninoculated apple juice stored at and 7C as determined in a triangle sensory test Incubation time (days) Panelists Number Values ( repetitions) Incubation temperature (C) a 7 5. 5. 8 4 9 8 5 6 8 9 5 7 9 6 6 7 54 57 45 48 57 5 48 5 b c c a Significance levels of, 5,, and.%. b Number of replicates out of three in which there was a significant difference between inoculated and uninoculated (control) juice. c At each level of significance, indicates no significant difference between inoculated and uninoculated (control) juice; indicates significant difference between inoculated and uninoculated (control) juice. at 7C occurred in one of three replicates, but populations of vegetative cells and spores were essentially the same as those in replicates in which no significant difference was detected. Panelists described the off odor of significantly different replicates as being medicinal, a descriptor commonly used to describe the off odor attributed to growth of A. acidoterrestris in juices. It is possible that rates of growth and death are similar, resulting in little net change in populations during the 6-day storage period. New cells may be producing guaiacol that the panelists detected. Apple juice held at 7C, and certainly at C, may not provide an exceptionally favorable environment for vegetative cell growth, and the spores inoculated into the juice may not have germinated. It appears that heat treatment may not be necessary to induce germination of most of the A. acidoterrestris spores in the apple juice. Populations reported as vegetative cells most likely also included spores, because populations of both were similar throughout the 6-day test. Our results are in contrast to those of Pettipher et al. (9) who observed an initial population of.7 to. log CFU/ml increased to 4.8 and 7. log CFU/ml of apple juice stored at 5C for 6 or days, respectively. When stored at 5C, 4.8 and 6.5 log CFU/ml were observed after and 6 days, respectively. Guaiacol was detected by sensory and gas chromatography-mass spectrometry analysis only when 7. log CFU of A. acidoterrestris were present after storage of juice at 5C for days and 6.48 log CFU/ml after storage at 5C for 6 days. In our study, we used a mixture of five strains of A. acidoterrestris, whereas Pettipher et al. (9) use a single strain that was not used in our study. It is possible that the strains used in our study do not grow as well in apple juice as did the strain Pettipher et al. (9) used. It is also possible that rapidly growing A. acidoterrestris may produce only limited amounts of guaiacol, and that production is triggered in stationary-phase cells or when cells are exposed to stress conditions such as suboptimal temperatures for growth, as was the case in our study. We are unable to predict the shelf life of apple juice based on relationships between production of sufficient guaiacol to be detected by panelists and population of A. acidoterrestris, because populations of spores and vegetative cells remained relatively unchanged throughout the 6- day storage period, even though an off odor assumed to be guaiacol was detected. A. acidoterrestris is an aerobic bacterium. Because commercially sterilized apple juice that contains a low amount of dissolved oxygen was stored in glass bottles, the oxygen available was limited throughout the 6-day test. During storage, the bottles were not agitated, which minimized oxygen uptake; therefore, only the cells near the top of the apple juice may have had sufficient oxygen to grow. It is possible that spores settled to the bottom of the bottle where germination and outgrowth may have been prevented or minimized. Apple juice for retail sale is moved and shaken several times between processing and consumption, allowing headspace oxygen to be incorporated into the juice. Temperature fluctuation may also affect the outgrowth of A. acidoterrestris spores. Chromatographic analysis of guaiacol in apple juice. At three of four sampling times ranging from to
J. Food Prot., Vol. 6, No. DETECTION OF GUAIACOL IN APPLE JUICE 5 TABLE. Population (log CFU/ml) of vegetative cells and spores of A. acidoterrestris in three replicate bottles of inoculated apple juice stored at and 7C, and the corresponding significant sensory differences between inoculated apple juice and uninoculated samples as determined in a triangle test Incubation temperature (C) Incubation temperature (7C) Incubation time (days) Repetition No. Population (log CFU/ml) Vegetative cells Spores Level of significance a 5. Population (log CFU/ml) Vegetative cells Spores Level of significance a 5. 4 8 8 5 9 6.99.9.97.97.97.9.74.69.68.9.8.79.7.7.8.8.8.78.8.79.79.98.8.87.8.9.77.85.64.66.6.8.78.74.74.84.86.76.7.75.66.84.8.55.4..4.. 4.54..5.4..9.9..95.94.9.88.86.6.56.98.4.4.7....4..4.7..6.57.8..5.8.7..85. a indicates no significant difference between inoculated and uninoculated (control) juice; indicates significant difference between inoculated and uninoculated (control) juice.
5 ORR ET AL. J. Food Prot., Vol. 6, No. 6 days of storage at which the sensory panel detected (P.) guaiacol (Table ), concentrations of 8. to.4 ppb guaiacol were quantified by chromatographic analysis. The panel detected (P. to P.) guaiacol in five samples stored at to 7C for 8 to 6 days in which the compound was not detected by chromatographic analysis. Thus, the sensitivity of the sensory technique is greater that than of the chromatographic technique. Fruit juices or concentrates that may be judged to have an off odor caused by guaiacol cannot be identified by chromatographic analysis, at least using the technique employed in this study. Further work is needed to determine the effects of temperature and the role of oxygen in the headspace of juice containers on growth and guaiacol production by A. acidoterrestris. Because temperature and the amount of oxygen in the headspace would be expected to affect rates of growth and guaiacol production, it would be of interest to study their combined effects. A study that more closely mimics conditions to which bottled juice is exposed from the point of processing to the consumer would reveal more about the behavior of spores and vegetative cells of A. acidoterrestris. Finally, a study of the effects of nutrient availability on outgrowth of A. acidoterrestris spores and production of guaiacol in fruit juice would provide valuable insight to better understand metabolic activities associated with spoilage. REFERENCES. American Society for Testing and Materials Standard Practice E679. 99. Determination of odor and taste thresholds by a forced choice ascending concentration series method of limits. American Society for Testing and Materials, Philadelphia, Pa.. Brown, K. L. 995. New microbiological spoilage challenges in aseptics: Alicyclobacillus acidoterrestris spoilage in aseptically packed fruit juices, p. 4. In T. Ohlsson (ed.), Advances in aseptic processing and packaging technologies. Proceedings of an International Symposium, Swedish Institute for Food Research, Gutenberg, Sweden.. Cerny, G., W. Hennlich, and K. Poralla. 984. Spoilage of fruit juice by bacilli: isolation and characterization of the spoiling microorganisms. Z. Lebensm. Unters. Forsch. 79:4 7. 4. Deinhard, G., P. Blanz, K. Poralla, and E. Altah. 987. Bacillus acidoterrestris sp. nov., a new thermotolerant acidophile isolated from different soils. Syst. Appl. Microbiol. :47 5. 5. Deinhard, G., J. Saar, W. Krischke, and K. Poralla. 987. Bacillus cycloheptanicus spp. nov., a new thermoacidophile containing cycloheptane fatty acids. Syst. Appl. Microbiol. :68 7. 6. Hippchen, B., A. Roli, and K. Poralla. 98. Occurrence in soil of thermoacidophilic bacilli possessing -cyclo-heptane fatty acids and hopanoids. Arch. Microbiol. 9:5 55. 7. Meilgaard, M., G. V. Civille, and B. T. Carr. 99. Overall difference test: does a sensory difference exist between samples, p. 6 67, 8. In Sensory evaluation techniques. CRC Press, Inc., Boca Raton, Fla. 8. Orr, R. V., and L. R. Beuchat.. Efficacy of disinfectants in killing spores of Alicyclobacillus acidoterrestris and performance of media for supporting colony development by survivors. J. Food Prot. 6:7. 9. Pettipher, G. L., M. E. Osmundson, and J. M. Murphy. 997. Methods for the detection and enumeration of Alicyclobacillus acidoterrestris and investigation of growth and production of taint in fruit juice and fruit juice-containing drinks. Lett. Appl. Microbiol. 4: 85 89.. Pinhatti, M. E. M. C., S. Variane, S. Y. Eguchi, and G. P. Manfio. 997. Detection of acidothermophilic bacilli in industrialized fruit juices. Food Process. 9:5 5.. Splittstoesser, D. F., C. Y. Lee, and J. J. Churey. 998. Control of Alicyclobacillus in the juice industry. Dairy Food Environ. Sanit. 8: 585 587.. Walls, I., and R. Chuyate. 998. Alicyclobacillus historical perspective and preliminary characterization study. Dairy Food Environ. Sanit. 8: 5.. Yamazaki, K., H. Teduka, and H. Shinano. 996. Isolation and identification of Alicyclobacillus acidoterrestris from acidic beverages. Biosci. Biotechnol. Biochem. 6:54 545.