The Effect of Nisin on the Keeping Quality of Reduced Heat-Treated Milks

Transcription

1 213 Journal of Food Protection, Vol. 64, No. 2, 2001, Pages Copyright q, International Association for Food Protection The Effect of Nisin on the Keeping Quality of Reduced Heat-Treated Milks T. I. WIRJANTORO, 1 M. J. LEWIS, 1 * A. S. GRANDISON, 1 G. C. WILLIAMS, 2 AND J. DELVES-BROUGHTON 2 1 School of Food Biosciences, The University of Reading, Reading RG66AP, England; and 2 Danisco Cultor, Beaminster DT83DZ, Dorset, England MS : Received 2 May 2000/Accepted 6 September 2000 ABSTRACT Milk was subjected to a combination process involving reduced heat treatment (RHT) of and nisin (75 and 150 IU ml 21 ). The microbial and other quality aspects were compared with a RHT control (without nisin) and with a ultrahigh temperature (UHT) milk processed at 1428C for 2 s. Nisin was found to inhibit microbial growth for products stored without refrigeration, and RHT-nisin samples stored at 8C showed very low spoilage rates during 150 days, although not low enough to satisfy requirements for commercial sterility. RHT-nisin samples could be distinguished from and were preferred to the UHT control. Signi cant browning occurred during storage at 8C and above but was less in the RHT-nisin milk samples compared with the UHT milk. In RHT-nisin milk samples stored at 20 and 108C, no microbial could be detected in most samples after storage for 1 year. The effectiveness of this combination of RHT, nisin, and low storage temperatures against gram-positive spore-forming bacteria suggests potential for use of nisin in extended shelf life products. Nisin is a natural polypeptide bacteriocin that is used as a food preservative in more than 50 countries, including the European Union and the United States (12). It shows antimicrobial against gram-positive bacteria and is particularly active against spore formers, preventing spore outgrowth (10, 11, 17). Bacterial spores will survive pasteurization, and a small proportion may survive some commercial sterilization processes, so nisin can be effective as a preservative for heat-treated foods. Ultrahigh temperature (UHT) milk is typically processed at 1428C for 2 s. When this is packaged aseptically, it gives rise to a product that is commercially sterile. UHT milk has a distinct cooked avor that is not liked by people used to drinking pasteurized milk, and it is widely acknowledged that consumer acceptance of milk is in uenced more by its avor than by any other attribute (19, 24). The intensity of this cooked avor could be reduced by decreasing the processing temperature. Although this would also improve the avor, there is the possibility of survival of heatresistant spoilage and pathogenic microorganisms in reduced heat treatment (RHT) milks. Processing conditions of 1158C for 2 s have been found to be useful for extending shelf life (6), and conditions of 1208C for 2 s are used commercially to produce extended shelf life milks, which still retain the fresh nature of pasteurized milk. However, such milks still require refrigerated storage and will spoil quickly at ambient temperature because of the survival of thermoduric and spore-forming spoilage bacteria. Nisin has been shown to be effective at increasing the shelf life of milk processed at conditions slightly more severe than pasteurization. In fact, milk processed at 1158C for 2 s with nisin and stored at 108C was still in good condition after 28 days, with a microbial count * Author for correspondence. alewismj@afnovell.reading.ac.uk. of less than 10 ml 21 (). Since nisin has been shown to be very effective at controlling the of gram-positive bacteria, its mode of action being described more as sporostatic rather than sporicidal (5, 12, 23), it is highly likely that it would be effective in inhibiting those microorganisms that would survive such a RHT, provided that there are no postprocessing contaminants (16, 27). It is also more effective against heat-shocked or slightly damaged cells (28). Therefore, we thought it was appropriate to evaluate the keeping quality of milk subjected to RHT conditions (), with and without nisin addition, stored at a range of temperatures, refrigerated (2 and 108C), ambient (208C), and elevated ( to 558C), to determine its effectiveness. MATERIALS AND METHODS Milk processing. Preliminary experiments were performed on a plate pilot UHT plant (APV Ltd, Crawley, UK) with a number of different raw milk samples. These results showed that milk heated to 1208C for 2 s with 150 IU showed a much better keeping quality at 8C compared with a control milk, which showed 100% spoilage within 7 to 8 days. Similar ndings were observed for milks stored at 208C. The main trial involved the production of control UHT milk (1428C for 2 s) and RHT milk samples (). Raw milk was collected in two pallecons (1,000 liters) from local farms in Cheshire 1 day before processing and stored at 28C. Samples were taken from both pallecons for subsequent testing of microbiological quality. Nisin was added as Nisaplin (supplied by Aplin & Barrett, Danisco Cultor, Beaminster). Two levels (75 and 150 IU ml 21 ) were added to the RHT milk in addition to a control without nisin. In all trials, nisin was added with good mixing to the milk at least 1 h before processing. The main trials were performed using a plate-type heat exchanger operating at a ow rate of 720 liters h 21. The temperature pro le in the nal heating section was

2 214 WIRJANTORO ET AL. J. Food Prot., Vol. 64, No. 2 from 80 to 1178C (or 1428C) in 47 s and in the rst cooling section from 117 (142) to 508C in 20 s. These temperature pro les were used to evaluate the Fo values and browning indexes for the UHT and RHT milks. The browning index is expressed as an equivalent time (s) at 1218C (18), the z-value being 26.28C. RHT milks produced under these conditions had a similar Fo value to those produced at 1208C on the APV plant used in the preliminary experiments. To eliminate cross-contamination by spores that may survive at the RHT, the equipment downstream of the holding tube and the aseptic tank were resterilized between batches. The heat-treated milk then was packed aseptically into 125-ml containers, using a TetraPak lling machine (model TBA 9), to eliminate postprocessing contamination. More than 1,200 samples were collected for each set of processing conditions. The sterilized milk samples were stored at temperatures from 2 to 558C. The effects of storage conditions on a number of different quality parameters were monitored for up to 150 days. Some samples were also analyzed after 14 months. Microbiological analysis. Microbial counts on the raw milk were determined by making serial 1 in 10 dilutions of milk samples using maximum recovery diluent (Oxoid CM733), then 0.1 ml of the diluents were surface spreaded onto milk plate count agar (Oxoid CM681). The plates were incubated aerobically at 8C for 72 h. For the aerobic spores, milk samples were initially heated at 808C for 10 min and rapidly cooled to room temperature before the same procedure was repeated. The main microbiological analysis for heat-treated milks involved a streaking method, where 10 ml of sample (a loop) was streaked onto a dry agar medium. The agar medium was MPCA (Oxoid CM681) or plate count agar (Oxoid CM325). After streaking, plates were incubated inverted at 8C for 48 h for all milk samples. For the heat-treated milk samples that had been previously stored at 42 and 558C, a second set of plates was prepared and incubated inverted at the same storage temperature as the milk samples for 48 h. Separate colonies of microorganisms that survived the RHT were isolated, puri ed by transferring at least three times on PCA plates, examined under a microscope, and identi ed using API strips (API 50 CHB for Bacillus spp. from biomérieux UK Ltd., Basingstoke, England). ph analysis. Changes in ph result from microbial growth and to a lesser extent from Maillard browning. Therefore, this was chosen as a quick and simple method to monitor microbial growth and quality changes in the heat-treated samples. Milk samples that were previously stored at different incubation temperatures were allowed to reach 208C before measurement (Kent EIL 7045/46). Nisin assay. The nisin concentration in the milk samples was determined by an agar plate diffusion method using Micrococcus luteus as a test microorganism (14), whereby the concentration of nisin was determined from the diameter of the inhibition zone produced. Color analysis. Color was measured using a Hunter Color Quest instrument version 2.6 (Hunter Associates Laboratory, Inc., Reston, Virginia), using a D65 standard illumination lamp and an observer angle of 108. It was calibrated using a standard white tile, a gray tile, and a black cavity. Results are expressed in terms of L* (0, black, to 100, white), a* (160, red, to 260, green), and b* (260, blue, to 160, yellow) values (21). Total color difference (DE*) between two samples is given by the following equation: DE* 5 [(L* 1 2 L* 2 ) 2 1 (a* 1 2 a* 2 ) 2 1 (b* 1 2 b* 2 ) 2 ] 0.5 Dissolved oxygen. Dissolved oxygen concentration was measured using a Uniprobe Type G2 dissolved oxygen electrode. Samples were allowed to come to room temperature and readings were taken 10 min after immersion of the probe. Sensory analysis. Sensory analysis was performed to determine whether panelists could distinguish between the UHT control milk and the RHT-nisin milk samples. A triangle test (22) was conducted separately for milks between the control UHT milk and the RHT-nisin milks, using 15 to 20 panelists. Milks incubated at 10, 20, and 8C were evaluated. A further study was conducted for milks stored at 8C, using a scaling method. Milks were evaluated between 1 and 2 months after production to allow any reactive sulphydryl groups to oxidize and to be representative of when consumers might drink the product (3). Statistical analysis. The data from the triangular test were compared with statistical tables for that speci c test (22). A completely randomized design with three replications for the color study or six replications for sensory attributes in the scaling method was used (9). Analysis of variance of experimental data was carried out using SAS Release 6.10 TS Level 0019 computer program system (SAS Institute, Cary, N.C.). RESULTS The microbiological results for the two pallecons of raw milk showed that the raw milk quality was good (i.e.,,20,000 CFU/ml) (16), with a maximum aerobic viable count of 3.0 log CFU ml 21 and a spore count of 2.0 log CFU ml 21 in both samples. After 24-h storage at 8C, no microbial could be detected in heated milks using the streaking method, suggesting that there was no gross postprocessing contamination. Nisin assays on the raw milk before heat treatment showed values of 80 and 162 IU ml 21, which were close to the target values of 75 and 150 IU ml 21. Immediately following heat treatment, nisin was reduced to 64 and 137 IU ml 21, which represents a loss of about 20 and 15.5%, respectively. The ph of all heat-treated samples was 6.60 to The milks were subjected to the following processing conditions, calculated from their temperature-time pro les: UHT milk, Fo and browning index ; and RHT-nisin milks, Fo and browning index The keeping quality of UHT and RHT milks: storage at C. The main keeping quality trial was undertaken at 8C for up to 151 days. Table 1 shows the microbiological results during this period; each day s results are based on analysis of 20 separate samples. Failure of heat treatment would arise either from survival and growth of heat-resistant spores or postprocessing contamination. Heat treatment regulations in the United Kingdom (1) and the European Community require that UHT milks show a count of less than 100 CFU ml 21 after 15 days of incubation at 8C or 7 days at 558C. Before the main trial, preliminary experiments on a UHT pilot plant had established that such failures of heat treatment quickly resulted in spoiled samples when incubated at 8C. Plates streaked with a 10-ml sample were either clear, indicating a count less than 100 CFU ml 21, or heavily contaminated. The control UHT process resulted in 100% of the samples being acceptable

3 J. Food Prot., Vol. 64, No. 2 NISIN AND REDUCED HEAT-TREATED MILKS 215 TABLE 1. Plates tested by streaking method for heat-treated milk samples stored at 8C storage temperature Day 1428C for 2 s plus 75 IU plus 150 IU a Not done. 5/20 9/20 9/20 18/20 a 3/20 3/20 throughout the 151 days of storage at 8C, as indicated by clear plates. In contrast, the RHT control started to show considerable signs of spoilage after only 4 days, with % spoilage after 1 week and 90% spoilage after 2 weeks of storage; sampling ceased after 25 days. However, the addition of 150 IU was very effective at preventing the outgrowth of spores for 151 days of storage, with only a very low number of spoiled samples during this period. The same also applied to 75 IU, although the number of spoiled samples was slightly higher after about 80 days. Microorganisms from the RHT control stored at 8C were isolated for further study. Microscopical observations showed that they were gram-positive bacilli and spore formers. Another test with 3% H 2 O 2 indicated that the microorganisms were catalase positive. A further test with the API kits identi ed Bacillus subtilis, Bacillus licheniformis, and Bacillus amyloliquefaciens. This identi cation provides further evidence that nisin can increase the keeping quality of the RHT milks because nisin can inhibit the outgrowth of spores of most gram-positive bacteria, which may survive RHT. The ph was also monitored at regular intervals, again based on 20 separate samples (Fig. 1), since it is common commercial practice to examine samples for signs of microbial growth following incubation (ph change, organoleptic defects, blown packs) (2). The ph of the RHT control was signi cantly lower than the rest of the samples after 13 days (mean , with 7 samples below ph 6.50), and after 25 days it had decreased to 5.62, with all samples below ph No further samples were analyzed after 25 days. At this time, the RHT-nisin samples were at ph 6.54, and the UHT control was at The ph of all these milks continued to decrease slightly during storage, probably due to the Maillard browning reaction. After about 80 days, it was observed that RHT milk with only 75 IU was showing a slightly higher level of spoilage, with a FIGURE 1. ph of heat-treated milks during storage at 8C. l, 1428C for 2 s; m, ; n, and 75 IU ; 3, and 150 IU. small number (1 to 3 of 20) having decreased to below ph 6.0. After 100 days, there appeared to be no major differences in recorded ph values between the UHT control and the RHT milk with 150 IU. Further microbiological analysis was done on those RHT-nisin samples that had shown microbial growth. The surviving microorganisms were gram-positive bacilli and catalase positive, the most predominant being B. licheniformis; Bacillus lentus and Bacillus stearothermophilus were also identi ed. Since a higher incidence of spoilage was noted at the lower level of nisin addition, spore growth and germination could occur due to nisin depletion with storage. Some loss of nisin was observed during storage (Table 2). This was also temperature dependent, with higher retentions at lower storage temperatures. The two RHT-nisin samples showed 14.1 and 24.8% nisin retention (percentage of the concentration immediately following processing) after days at 8C. Nisin retention was 5.6 to 10% at 428C and less than 1% at 558C. Nisin degradation TABLE 2. Nisin in the RHT milk samples processed at during storage time (day) temperature (8C) Milk with 75 IU Nisin (IU ml 21 ) ,1 Residual nisin (%) ,1.6 Milk with 150 IU Nisin (IU ml 21 ) ,1 Residual nisin (%) ,0.7

4 216 WIRJANTORO ET AL. J. Food Prot., Vol. 64, No. 2 TABLE 3. ph and microbiology results of heat-treated milk samples stored at 42 and 558C a temperature (8C) time (day) 1428C for 2 s plus 75 IU plus 150 IU 55 ph (n 5 2) / ph (n 5 5) / ph (n 5 5) / ph (n 5 2) / ph (n 5 5) / a The plates for samples stored at 42 and 558C were incubated at 42 and 558C, respectively. during storage seemed to increase between 20 and 8C. Retention at 208C and below exceeded 63% after days. at 42 and 55 C. At 42 and 558C, microbiological and ph results indicated that spoilage was much more rapid for the RHT-control milk than at 8C. These changes are illustrated in Table 3. In fact, the mean ph dropped most rapidly at 428C, decreasing to 5.65 after 5 days, as opposed to 5.92 at 558C. At both temperatures, all samples examined had spoiled ( ve of ve samples) after 7 days. It appears that the optimum growth range for microorganisms surviving RHT in this trial is in the thermophilic region (42 to 558C). In contrast, no spoilage was detected in the RHT-nisin milks after 14 days at 558C and 8 days at 428C, although fewer samples were analyzed compared with 8C. The samples also showed similar ph values to the UHT control samples. However, considerable browning was observed in all these samples, and the ph after 14 days at 558C had decreased from 6.62 to below This change was greater than that found after more than 100 days at 8C. No microbial spoilage was observed in the period taken for the milks to become very visibly brown. Also the fall in ph was greater for the UHT control compared with the RHT-nisin milks, suggesting that more browning took place in the UHT control. When the RHTcontrol samples stored at 558C were streaked on the agar media, no microbial growth occurred when the plates were incubated at 8C. Streaked samples from similar milk stored at 428C did show some microbial growth at 8C. These results further suggested that microorganisms surviving in these trials were predominantly thermophilic in nature. Twenty samples of UHT and RHT-nisin milks stored at 558C for more than 137 days were examined for microbial and all found to be clear, although they were unacceptably brown, with a ph below 5.5. It was interesting that RHT-nisin milks stored showed no signs of microbial, even though the residual nisin level was low. It would appear that the combination of nisin and heat caused suf cient cell damage to prevent any further outgrowth of thermophilic spores, even when nisin levels were subsequently reduced. at 10 and 20 C. RHT-nisin milks stored at 108C for 14 months showed no spoilage (based on analysis of 100 samples). All samples were inspected, and no spoiled off-odors were detected, showing its effectiveness in extended shelf life products. For samples stored at 208C, one spoiled sample was found from 100, suggesting that it would be also effective in situations where the product was subject to temperature abuse during refrigerated storage, for example, in transporting food from the retail outlet to home. Sensory evaluation of RHT. A triangular test was set up using 18 panelists to see whether they could distinguish between the control UHT milk and the RHT milk with 150 IU. The rst test was carried out with the milk samples that had been stored for 55 days at 8C, which is some time after the initial cooked avor would have disappeared but before the onset of stale and oxidized avors (3). Twenty-three of 36 answers selected the correct sample (panelists could pick the odd sample from three samples), which shows that the RHT-nisin milk differed signi cantly from the control milk (P, 0.001) (22). Shortly afterward, the procedure was repeated for milk samples stored at 10 and 208C. Panelists could also distinguish between the UHT milk and RHT-nisin milk at both temperatures (27 of 41 at 208C and 22 of 34 at 108C), which are both signi cant (P, 0.001) (22). A second taste panel was conducted to study which of the two milks was most liked after 56 days of storage at 8C. Six panelists were asked to score this, using a scaling

5 J. Food Prot., Vol. 64, No. 2 NISIN AND REDUCED HEAT-TREATED MILKS 217 TABLE 4. Color measurement of heat-treated milk samples after 75 days at different storage temperatures temperature (8C) Treatments L* value a* value b* value C for 2 s, Control AB a A AB HJ HI HIJ 9.02 RSW 8.77 QRSUW 8.69 QRTUV C for 2 s, Control AB A A J I I 9.03 SW 8.68 QTUV 8.79 QRSUW C for 2 s, Control BC A A K HI HI 8.93 QRSW 8.37 TUV 8.40 TUV 1428C for 2 s, Control C A A 0.10 L HI HI 9.03 W 8.38 V 8.55 UV C for 2 s, Control E D D 2.74 N 1.56 M 1.67 M Y X X C for 2 s, Control G F F 7.44 P 6.36 Q 6.44 Q A Z Z a Different letters within a column indicate signi cant differences (P # 0.05). method. The results showed that the panel could signi - cantly identify that the UHT milk had more cooked avor (P, 0.035) and preferred more the RHT-nisin milk, although this preference was not signi cantly different. Color and dissolved oxygen. The browning index for the RHT-nisin (13.8) and UHT (84.4) milks was well below the threshold value of 400 (13), so both heat treatments produced no immediate browning. The color of the milks stored at different temperatures was also measured after 75 days (Table 4). At 28C, there was no signi cant difference between the RHT-nisin milks and UHT milks in terms of their L*, a*, and b* values. For RHT-nisin milks, there was also no signi cant difference in L* values (lightness) at storage temperatures between 2 and 8C, suggesting that color changes were minimal for RHT samples during this temperature range. However, color changes were more pronounced above 8C, the brown color becoming much TABLE 5. Concentration of dissolved oxygen (percentage of saturation) in heat-treated milk samples after 150 days at different storage temperatures a temperature (8C) C for 2 s plus 75 IU ml 21 of nisin plus 150 IU ml 21 of nisin a The dissolved O 2 values were 83.4% for fresh raw milk and 77.1% for fresh pasteurized milk. more pronounced as temperature increased, as seen from decreasing L* and increasing a* and b* values. L*, a*, and b* values were signi cantly different between UHT milks and RHT-nisin milks for all storage temperatures above 208C. The following color difference values between the UHT milk and the nisin-rht milk (75 IU ) were found at different temperatures: 1.35 (208C), 1.99 (8C), 3.45 (428C), and 4. (558C). The difference became greater as the storage temperature increased and differences could be easily perceived at 42 and 558C. Maillard browning reactions lead to a reduction in ph, which in exceptional circumstances may be signi cant to cause coagulation to occur. Pagliarni et al. (25) reported that the visual perception threshold of browning corresponded to a DE* of 3.8, although noticeable differences were found at lower values in this study, i.e., 3.45 at 428C. Nisin concentration did not appear to affect the DE* value, with values ranging between 0.08 and 0.68, so the RHT is responsible for the slower browning during storage. Dissolved oxygen measurements were taken on samples stored for 150 days at different storage temperatures (Table 5). Dissolved oxygen concentration decreases as storage temperature increases because of higher oxygen utilization in oxidation reactions. It is interesting that dissolved oxygen levels are higher in the RHT-nisin milks than in UHT milks at the lower storage temperatures but similar at 8C and higher storage temperatures. There was still considerable dissolved oxygen remaining in milks stored at 208C after this time. DISCUSSION These results showed clearly that nisin was effective at controlling microbial in RHT milks. Target spoil-

6 218 WIRJANTORO ET AL. J. Food Prot., Vol. 64, No. 2 age rates for commercially sterile products are less than 1 in However, low levels of spoilage were found (about 1 in 40 throughout the 150-day storage), suggesting the presence of one or more spore-forming bacteria, e.g., B. licheniformis, which was tolerant to 150 IU. The D-value for B. licheniformis has been reported to be 11.1 s at 1158C and the z-value to equal 8.0 in skim milk (4). According to this, increasing the temperature from 117 to 1208C would reduce the D-value by a factor of 2.37 and more than double the number of decimal reductions achieved. In this trial, this temperature increase should further reduce the spoilage rate for RHT-nisin samples, so there is scope for further optimization of the process to achieve commercial sterility. Such reduced combination processes would only be advocated for milk products, because these conditions do not comply with a full-botulinum cook (Fo 5 3) required for low-acid foods (ph. 4.5). This speci city is stressed for milk because Clostridium botulinum is rarely found in raw milk and cases of botulism in heat-treated milks have not been recorded. It would not be advocated for other lowacid foods, such as meats, sh, and vegetable products. However, nisin s ef cacy in preventing C. botulinum spore outgrowth and toxin formation has been demonstrated by Taylor et al. (29). In milk, in addition to the hurdles of heat and nisin, there are also the inhibitory effects that arise from relatively high dissolved oxygen levels and from high redox potentials found in sound milk. C. botulinum grows optimally at redox potential of 2350 mv, but growth initiation may occur in the range of 1 to 1250 mv. Most research on the inhibitory effects of oxygen on C. botulinum has been done with modi ed atmosphere storage. There is little direct experimental evidence on the effects of oxygen in solution on its growth and toxin production. Note that dissolved oxygen is still present in sound milk after considerable storage, so it is suggested that suitable conditions for C. botulinum growth would only be established if the product had already spoiled. Nisin has been shown in this work and by others (12) to be particularly effective against species of thermophilic bacteria, such as B. stearothermophilus. It has been suggested that sterilization processes should be more severe (e.g., Fo. 10) in countries with hot climates (2) to minimize spoilage due to thermophilic spore-forming bacteria. Thus, there is scope for adding nisin to UHT products, without compromising the minimum botulinum cook, which should reduce the browning potential at higher storage temperatures and generally improve the quality. More recently Bacillus sporothermodurans has been recognized as causing UHT products not to comply with the microbiological speci cations (15, 26). This is a mesophile nonpathogenic microorganism that has been reported to reach counts of 10 5 ml 21, without causing any changes in the sensory characteristics of milk. The raw milk used in the main trial was high quality. Spore counts in raw milk usually range between 100 (found here) and 1,000 CFU ml 21 and rarely exceed 5,000 CFU ml 21 (7, 20). Some preliminary trials were performed with raw milk of poorer quality than that used in the main trial (3.6 and 5.4 log CFU ml 21 ) and showed that addition of 150 IU was effective in controlling microbial compared with the control, although fewer samples were analyzed. It is well established that psychrotrophic bacteria produce extracellular proteases (8), and one trial with such a milk (6.8 log CFU ml 21 ) showed that nisin was not effective, even though the count immediately after heat treatment was less than 1 CFU ml 21. These proteases may interfere with the of nisin, which itself is a polypeptide. Also, since nisin is positively charged in milk, there could be interactions between it and negatively charged species, such as casein and other proteins in milk, which may further reduce its. Some of the earlier trials demonstrated that it was essential to eliminate postprocessing contamination, since nisin was rendered ineffective when this occurred. Similar results have been found when eliminating postprocessing contamination in pasteurized cream (27). There was no evidence of any signi cant postprocessing contamination in the main trial. Since nisin was found to be very effective at controlling microbial at 10 and 208C, it would be very appropriate for extended shelf life products, and the very low level of microbial found would probably exceed current expectations for these products. Therefore, there is potential for using nisin in extended shelf life products to further improve keeping quality and to provide some additional safeguard against poor temperature control in the cold chain and even to tolerate ambient storage (208C) for some time. No off- avors related to lipase or protease were noticed during prolonged refrigerated storage, and nisin losses and cooked avor development would be less at these lower storage temperatures. ACKNOWLEDGMENTS The authors thank Dr. D. B. MacDougall for his generous advice regarding the sensory evaluation and color measurement. The project was supported by Danisco Cultor, Beaminster, Dorset, England, and The University of Reading, Reading, England. REFERENCES 1. Anonymous Food, milk and dairies: the dairy products (hygiene) regulations. Statutory Instruments (SI) no HMSO, London. 2. Anonymous Development and use of microbiological criteria for foods. Institute of Food Science and Technology (UK), London. 3. Ashton, T. R Practical experience: the processing and aseptic packaging of sterile milk in the United Kingdom. J. Soc. Dairy Technol. 18: Behringer, R., and H. G. Kessler Thermal destruction of spores of selected Bacillus strains in skimmilk and skimmilk concentrate. Int. Dairy J. 2: Beuchat, L. R., M. R. S. Clavero, and C. B. Jaquette Effects of nisin and temperature on survival, growth, and enterotoxin production characteristics of psychrotrophic Bacillus cereus in beef gravy. Appl. Environ. Microbiol. 63: Borde-Lekona, B., M. J. Lewis, and W. F. Harrigan Keeping quality of pasteurised and high pasteurised milk, p In A. T. Andrews and J. Varley (ed.), Biochemistry of milk products. Royal Society of Chemistry, Cambridge. 7. Bramley, A. J., and C. H. McKinnon The microbiology of raw milk, p In R. K. Robinson (ed.), Dairy microbiology, vol. 1. The microbiology of milk. Elsevier Applied Science, London.

7 J. Food Prot., Vol. 64, No. 2 NISIN AND REDUCED HEAT-TREATED MILKS Burton, H Ultra-high-temperature processing of milk and milk products. Elsevier Applied Science, London. 9. Cochran, W. G., and G. M. Cox Experimental designs. Wiley, New York. 10. Delves-Broughton, J Nisin and its uses as a food preservative. Food Technol. 44: Delves-Broughton, J Nisin and its application as a food preservative. J. Soc. Dairy Technol. 43: Delves-Broughton, J., P. Blackburn, R. J. Evans, and J. Hugenholtz Applications of the bacteriocin, nisin. Antonie van Leeuwenhoek 69: Fink, R., and H. G. Kessler Comparison methods for distinguishing UHT treatment and sterilisation of milk. Milchwissenschaft 43: Fowler, G. G., B. Jarvis, and S. Tramer The assay of nisin in foods. Soc. Appl. Bacteriol. Tech. Ser. 8: Hammer, P., F. Lembke, G. Suhren, and W. Heeschen Characterisation of heat resistant mesophilic Bacillus species affecting the quality of UHT milk. In Heat treatments and alternative methods. International Dairy Federation/Federation Internationale de Laiterie, Brussels, Belgium, no Harding, F The impact of raw milk quality on product quality, p In F. Harding (ed.), Milk quality. Blackie Academic & Professional, London. 17. Hurst, A., and D. G. Hoover Nisin, p In P. M. Davidson and A. L. Branen (ed.), Antimicrobials in foods. Marcel Dekker, New York. 18. Kessler, H. G, and P. Horak Objective assessment of UHT treatment of milk by standardization of bacteriological and chemical effects. Milchwissenschaft 36: Lewis, M. J Heat treatment of milk, p In R. K. Robinson (ed.), Modern dairy technology. Chapman & Hall, London. 20. Lewis, M. J Microbiological issues associated with heat treated milks. Int. J. Dairy Technol. 52: MacDougall, D. B Colour vision and appearance measurement, p In J. R. Piggott (ed.), Sensory analysis of foods. Elsevier Applied Science, London. 22. Meilgaard, M., G. V. Civille, and B. T. Carr Sensory evaluation techniques. CRC Press, Boca Raton, Fla. 23. Montville, T. J., A. M. Rogers, and A. Okereke Differential sensitivity of Clostridium botulinum strains to nisin is not biotypeassociated. J. Food Prot. 55: Nursten, H. E The avor of milk and diary products, part 1: milk of different kinds, powder, butter and cream. Int. J. Dairy Technol. 50: Pagliarni, E., M. Vernile, and C. Peri Kinetic study on colour changes in milk due to heat: a research note. J. Food Sci. 55: Pettersson, B., F. Lembke, P. Hammer, E. Stackebrandt, and F. G. Priest Bacillus sporothermodurans, a new species producing highly heat-resistant endospores. Int. J. Syst. Bacteriol. 46: Phillips, J. D., M. W. Grif ths, and D. D. Muir Effects of nisin on the shelf-life of pasteurized double cream. J. Soc. Dairy Technol. 36: Scott, V. N., and S. L. Taylor Effect of nisin on the outgrowth of Clostridium botulinum spores. J. Food Sci. 46: , Taylor, L. Y., D. D. Cann, and B. J. Welch Antibotulinal properties of nisin in fresh sh packaged in an atmosphere of carbon dioxide. J. Food Prot. 53: Wirjantoro, T. I., and M. J. Lewis Effect of nisin and high temperature pasteurisation on the shelf-life of whole milk. J. Soc. Dairy Technol. 4: