Challenge testing of the lactoperoxidase system ... - Wiley Online Library

Journal of Applied Microbiology 2001, 91, 735±741

Challenge testing of the lactoperoxidase system in pasteurized milk N.E. Marks, A.S. Grandison and M.J. Lewis School of Food Biosciences, The University of Reading, UK 810/4/01: received 4 April 2001 and accepted 27 April 2001

N . E . M A R K S , A . S . G R A N D I S O N A N D M . J . L E W I S . 2001.

Aims: To determine the role of lactoperoxidase (LP) in inhibiting the growth of microorganisms in pasteurised milk. Methods and Results: Four micro-organisms of importance in the spoilage of pasteurized milk were challenged in lactoperoxidase (LP)-enriched ultra-heat treated (UHT) milk after subsequent pasteurization. Milk samples were stored at the optimum temperatures for growth of the individual bacteria. Pasteurization was carried out at 72°C/15 s and 80°C/15 s to determine the effect of the LP system on the micro-organisms. An active LP system was found to greatly increase the keeping quality (KQ) of milks inoculated with Pseudomonas aeruginosa, Staphylococcus aureus and Streptococcus thermophilus and pasteurized at 72°C, but had little or no effect in milks heated at 80°C, presumably due to virtual inactivation of LP at 80°C. However, pasteurization temperature had no effect on the KQ of milks challenged with Bacillus cereus spores. Conclusions: This study suggests that the LP system, rather than heat-shocking of spores, is responsible for the greater KQ of milk pasteurized at 72°C/15 s compared with 80°C/15 s. Signi®cance and Impact of Study: The study emphasizes the care required in selecting pasteurization temperatures in commercial practice and to avoid the temptation to compensate for inferior quality of raw milk by increasing pasteurization temperature. INTRODUCTION The lactoperoxidase (LP) system is a naturally-occurring antimicrobial system present in raw milk which is active against both Gram-positive and Gram-negative microbes to varying extents (Siragusa and Johnson 1989). Lactoperoxidase is heat sensitive but retains the majority of its activity in milk which has been pasteurized at 72°C/15 s (Barrett et al. 1999). However, as the temperature is increased, the LP activity decreases rapidly until it cannot be detected following treatment at approximately 80°C (Grif®ths 1986). It can therefore be postulated that if the LP system remains active in a pasteurized milk sample, it is able to exert some effect on the ¯ora therein. The spoilage micro¯ora in pasteurized milk is completely different to that found in raw milk. It consists mainly of post-pasteurization contaminants (PPC) (bacteria that con-

Correspondence to: A. Grandison, School of Food Biosciences, The University of Reading, PO Box 226, Reading, Berkshire, RG6 6AP, UK. ã 2001 The Society for Applied Microbiology

taminate the milk after heating), thermoduric micro-organisms (microbes that are able to withstand the pasteurization process) and spore-producing bacteria (vegetative microorganisms which produce spores able to survive heating) (Fredsted et al. 1996). This study examined the effect of the LP system on the KQ of milk pasteurized at different temperatures on two PPC organisms (Pseudomonas aeruginosa and Staphylococcus aureus), one thermophile (Streptococcus thermophilus) and one spore-former (Bacillus cereus). The aim was to determine the possible role of LP in inhibiting the growth of these organisms in pasteurized milk and thus, to examine the possibility that the enzyme has a positive effect on the KQ of pasteurized milk. MATERIALS AND METHODS The basic protocol was to prepare sterilized milk (which has no LP activity), and produce samples, with or without added LP enzyme, pasteurized at either 72°C/15 s or 80°C/15 s. Pure cultures of micro-organisms were added either before

736 N . E . M A R K S ET AL.

or after pasteurization, as appropriate, and the KQ and total viable count (TVC) were assessed during storage. Cultures Pure cultures of Staph. aureus (NCDO 0899), Ps. aeruginosa (NCTC 10299), Strep. thermophilus (NCDO 1469) and B. cereus (NCDO 577), obtained from the departmental culture collection (stored at ±80°C in cryoprotective media), were grown in nutrient broth. Streak plates of the cultures on plate count agar (incubated at 30°C/72 h 0á5 h) followed by microscopic examination con®rmed that the broth contained only the test organism. All cultures were added to sterilized milk to achieve colony counts of approximately 102 colony-forming units (cfu) ml±1. B. cereus spore production method A pure culture of B. cereus was grown on potato starch agar plates for 12 days at an incubation temperature of 30°C to achieve maximum sporulation. The spores were harvested in distilled water, sonicated to disrupt vegetative cells and repeatedly washed in distilled water. Spore suspensions were stored at ±18°C until use, at which point they were thawed, sonicated and washed a further three times. The spores were heat-shocked in thin-walled glass tubes and then added to the milk after pasteurization. In addition, a sample of non-heat-shocked spores was added to a separate milk sample at the same time to determine levels of PPC. The equivalent time for heating was established as 21 s at 72°C and 20 s at 80°C. Milk plus spores was also pasteurized for 63°C/30 min in this batch system to determine the effect of low-temperature, long-time (LTLT) pasteurization on KQ. Heat treatment of milk, addition of LP system constituents and inoculation with micro-organism Separate experiments were carried out for each challenge micro-organism to assess individual bacterial effects on KQ. For each run, 90 l raw whole milk from a local dairy (Stokes, Camberley, UK) was sterilized at 142°C/2 s on an UltraHeat Treatment (UHT) plant (APV, Peterborough, UK) and collected into two sterile churns. This procedure destroyed the natural micro¯ora, but also completely inactivated the LP. Therefore, 15 p.p.m. ®lter-sterilized LP (freeze-dried sample provided by DMV International, Veghel, Belgium) was added to one of the churns (activated sample) to give approximately the same activity as the raw sample, the activity being checked by the LP assay. The remaining churn was the non-activated control. These samples were then pasteurized at 72°C/15 s or 80°C/15 s in a small continuous pasteurizer (model JE301,

APV). Pseudomonas aeruginosa and Staph. aureus were added as post-pasteurization contaminants (PPC) after heating, while Strep. thermophilus was inoculated into the milk before pasteurization. Ideally, B. cereus spores would have been added prior to pasteurization but, due to their potential pathogenic nature, were heat-treated in contained facilities and subsequently added to the pasteurized milk to prevent contamination of the pilot plant. Sodium thiocyanate and hydrogen peroxide were ®ltersterilized and added aseptically to all milk samples after pasteurization to ensure that the LP system was activated where enzyme was present. The solutions of SCN± and H2O2 were prepared in sterile water such that 10 p.p.m. of each was present in the ®nal milk samples. Levels of 10 p.p.m. are recommended by the International Dairy Federation (IDF 1988). Milk samples were then divided into sterile 30 ml containers and stored at the optimum temperature for growth of each organism (25°C for Ps. aeruginosa, 30°C for B. cereus spores and 37°C for Staph. aureus and Strep. thermophilus). The whole procedure was carried out to reduce the risk of PPC. As a check for PPC, samples with no added bacteria or LP enzyme were retained for each pasteurization temperature during each challenge experiment. Lactoperoxidase assay This was measured, using an adaptation of the IDF method (IDF 1994), where the rate of oxidation of 2,2¢azinobis(3ethyl-benzothiazoline-6-sulphonic acid) (ABTS) was measured spectrophotometrically at 412 nm; 1 mmol l±1 ABTS (diammonium salt; Sigma) and 0á3 mmol l±1 H2O2 solutions were prepared daily in 0á1 mmol l±1 phosphate buffer at pH 6á7. All reagents were stored and used at 22°C. Milk (0á1 ml at 4°C) was mixed with 2 ml of the ABTS solution and left for 30 min at 22°C to allow stabilization of casein micelles in the phosphate buffer. Hydrogen peroxide (1 ml) was added and mixed quickly to start the reaction. Absorbance (A412) was recorded at 15 s and again at 75 s. The rate of change of absorbance per min (DA/At) was then calculated (the extinction coef®cient of the ABTS oxidation product at 412 nm is 32á4 ´ 10±3). Determination of keeping quality Titratable acidity and alcohol stability were used as measures of KQ as these parameters have been shown to give reliable and consistent estimates (Barrett et al. 1999). Tests were carried out hourly, KQ being calculated as the mean of the two methods of assessment. KQ was determined in triplicate, and the results are presented as means with standard deviations shown on bar charts.

ã 2001 The Society for Applied Microbiology, Journal of Applied Microbiology, 91, 735±741

CHALLENGING LACTOPEROXIDASE

737

Titratable acidity. To 10 ml milk was added 1 ml phenolphthalein solution (0á5% phenolphthalein in 50% ethanol). This mixture was titrated against 0á111 mol l±1 sodium hydroxide solution until a pink colour was persistent for more than 5 s. The percentage lactic acid was then calculated as titre volume divided by 10. When the lactic acid content of the sample reached 0á18% or higher, the sample was considered to be spoiled. Alcohol stability. This test consisted of mixing equal portions of 68% alcohol (neutralized) with the milk and noting whether coagulation occurred. When the sample coagulated, end of shelf life had been reached. Total viable count (TVC) This was determined hourly using 0á1 ml of ®vefold serial dilutions in Maximum Recovery Diluent (Oxoid) spread on pre-dried Plate Count Agar (Oxoid) plates with sterile spreaders. The resulting plates were inverted and incubated at 30°C for 72 h 0á5 h before counting colonies. TVC values were determined in triplicate and standard deviations were calculated. However, standard deviations were smaller than the symbols used in Figs 2, 4, 6 and 8 and hence were not included. This procedure does not test for the presence of residual damaged cells, which may not be culturable, following treatments. However, measurable growth was observed following all treatments, and it is probable that the presence of such cells would not have been signi®cant to the KQ. RESULTS PPC test micro-organisms

Fig. 2 Total viable count of milks stored at 25°C, challenged with Pseudomonas aeruginosa. (s), 72°C, no LP; (h), 80°C, no LP; (d), 72°C, + LP; (j), 80°C, + LP

Total viable count (log cfu ml–1)

Ps. aeruginosa. The initial inoculation level for this organism was 2á9 ´ 102 cfu ml±1. The non-inoculated milk samples retained as a test for PPC had KQs of more than 110 h

Fig. 1 Keeping quality (h) of milks stored at 25°C, challenged with Pseudomonas aeruginosa

at 25°C each, indicating that PPC had been limited but not totally eliminated. The KQ results for milks inoculated with Ps. aeruginosa can be seen in Fig. 1. All KQ results are quite high considering storage was at 25°C. This was probably because the inoculation level was lower than the raw milk count (4á3 ´ 104 cfu ml±1). The ®rst two columns of Fig. 1 demonstrate that where no LP was present, the KQs were identical (28 h) at the two pasteurization temperatures. In the presence of LP, the sample pasteurized at 80°C had a KQ just 2 h longer than the control samples. However, for the sample that had been pasteurized at 72°C (with 72á4% residual lactoperoxidase activity), the KQ was 46 h, which was 16 h more than the 80°C pasteurized sample (which had no detectable LP activity). Figure 2 shows the pattern of bacterial growth for the above four samples. The curves for the samples with no added LP were virtually identical, as expected. The LP-activated sample pasteurized at 80°C gave a similar curve, but the values were all slightly lower. The LP systemactivated sample pasteurized at 72°C gave a similar growth curve, but the initial levels of Ps. aeruginosa were reduced to

4


Storage time (h)

738 N . E . M A R K S ET AL.

observed in Ps. aeruginosa, growth proceeded immediately and at the same rate as the other samples. Secondly, towards the end of life, the growth rate slowed down in the LP system-activated milk treated at 72°C, giving rise to a longer shelf-life than may have been expected from the ®rst 15 h of growth. Thermophilic test micro-organism (Strep. thermophilus)

Fig. 3 Keeping quality (h) of milks stored at 37°C, challenged with Staphylococcus aureus

less than 1 cfu ml±1; this level was maintained for 5 h, after which growth continued at the same rate as the other samples.


Staph. aureus. The initial inoculation level for this organism was 3á1 ´ 102 cfu ml±1. The samples retained as a test of PPC had a KQ of more than 75 h each. The KQ results for Staph. aureus (Fig. 3) show a similar trend to Ps. aeruginosa, but with lower values for KQ. Again, the KQ values for the two control samples with no added LP were identical (22 h). An increase of 4 h was seen in the LP system-activated sample pasteurized at 80°C, while the sample pasteurized at 72°C had a KQ of 38 h, 12 h longer than the latter. The 72°C sample had a residual enzyme activity of 71á9% compared with an undetectable quantity at 80°C, which presumably led to the increase in KQ at the lower pasteurization temperature. Figure 4 shows the colony counts obtained from the Staph. aureus-challenged milks. The same conclusions can be drawn from this graph as from Fig. 2 for Ps. aeruginosa but with two differences. First, in the 72°C LP systemactivated sample, the initial count was reduced as before, but instead of the LP system delaying growth for 5 h, as

Storage time (h)

The initial inoculum level was 1á7 ´ 102 cfu ml±1. The samples retained as a test for PPC had KQs of 81 and 78 h for the 72°C and 80°C pasteurized samples, respectively. The KQ results for Strep. thermophilus can be seen in Fig. 5. The observed trends are the same as seen for the PPC challenge organisms, Ps. aeruginosa and Staph. aureus. Where the LP was not present, the KQ for both pasteurization temperatures was identical but when LP was present, there was an increase in KQ for both samples. In the case of the sample pasteurized at 80°C, this increase was 4 h, which could have been due to an undetectable amount of residual LP activity. At 72°C, the increase in KQ was 52 h. At this lower pasteurization temperature, there was 72% residual LP activity and the KQ was more than double that of the sample pasteurized at 80°C. Figure 6 shows the colony counts obtained from pasteurized milks challenged with Strep. thermophilus. The observed antimicrobial effect of the LP system for the 72°C pasteurized sample was greater than that for either Ps. aeruginosa or Staph. aureus. The effect appears to be a combination of the three trends observed for the two PPC test micro-organisms seen in Figs 2 and 4, i.e. (i) a reduced initial colony count (in the case of Strep. thermophilus from > 102 cfu ml±1 to < 1 cfu ml±1); (ii) delayed growth of Strep. thermophilus for 12 h before growth commenced at the same rate as in the other samples; (iii) a slowing of growth rate towards the end of storage life.

Fig. 4 Total viable count of milks stored at 37°C, challenged with Staphylococcus aureus. (s), 72°C, no LP; (h), 80°C, no LP; (d), 72°C, + LP; (j), 80°C, + LP



739

treatments. At 63°C/30 min, the KQ was slightly greater than HTST-pasteurized samples for both control samples and samples with added LP. The KQs of control milks pasteurized at 72 and 80°C for 15 s were essentially identical. TVC analysis (Fig. 8) demonstrated that in the samples with added LP, the initial count was reduced from 102 cfu ml±1 to < 1 cfu ml±1; there was no detectable growth for approximately 15 h, after which growth proceeded at the same rate as in the controls. Fig. 5 Keeping quality (h) of milks stored at 37°C, challenged with Streptococcus thermophilus

DISCUSSION The results of samples retained to check for PPC for all four organisms demonstrate that the procedures have generally kept PPC to a minimum and that differences between trials were not due to PPC. Challenge testing with Ps. aeruginosa, Staph. aureus and Strep. thermophilus produced similar trends. With no added enzyme, the keeping quality and growth curves were very similar for each organism following heat treatment at either 72 or 80°C. In the presence of LP, a small increase in KQ occurred following pasteurization at 80°C which was associated with a reduction in initial TVC. Following pasteurization at 72°C, when > 70% of the added LP activity remained, there was a much greater increase in KQ which was associated with reduced initial TVC, a subsequent lag phase and reduced rates of growth, the importance of the different effects varying between the three species. This is consistent with the view that the enzyme works through both bacteriostatic and bactericidal mechanisms (Reiter and HaÈrnulv 1984). It is possible that the small increase in KQ following the 80°C treatment resulted from the effects of residual LP which was below the threshold of detection with the ABTS assay. Overall, these results suggest that the LP system plays an important role in the KQ of milk pasteurized at 72°C for 15 s if spoilage is caused by PPC or heat-resistant thermophiles. The challenge tests with B. cereus produced a different pattern. With the milks with no added LP, the LTLT

Other trends observed for the different sample treatments were the same as for the PPC organisms in that the curves for the control samples were virtually identical, the initial colony count for the 80°C activated sample was reduced, but the growth rate was the same as in the non-activated milks. Heat-shocked spores (B. cereus)

Fig. 6 Total viable count of milks stored at 37°C, challenged with Streptococcus thermophilus. (s), 72°C, no LP; (h), 80°C, no LP; (d), 72°C, + LP; (j), 80°C, + LP

Total viable count (log cfu m–1)

The initial inoculum level of spores was 1 ´ 102 spores ml±1 milk. This was determined by heat-shocking a small quantity of sample at 80°C/10 min and carrying out a TVC measurement. For the non-inoculated samples with no added LP (retained as a test for PPC) and pasteurized at 63, 72 and 80°C, the KQs were 124, 112 and 114 h, respectively. The slightly longer KQ of the LTLT sample indicates that the batch-pasteurized sample contained fewer bacteria than the continuously-pasteurized samples. This probably re¯ects the complexity and dif®culty of cleaning HTST systems, and the fact that it is easier to maintain aseptic conditions in a laboratory than in the departmental pilot plant. The KQ results obtained with this organism are illustrated in Fig. 7. As for all the other organisms investigated, it is clear that the LP system had an effect on the growth of B. cereus by the fact that an addition of LP extended the KQs of the milks to approximately double those observed in the control samples for all three heat 8 7 6 5 4 3 2 1 0 0

5

10

15

20


25

30

Storage time (h)

35

40

45

50

55

740 N . E . M A R K S ET AL.


Fig. 7 Keeping quality (h) of milks stored at 37°C, challenged with Bacillus cereus spores 8 7 6 5 4 3 2 1 0 0

5

10

15

20

25

30

35

40

45

50

55

Storage time (h)

sample had a slightly longer KQ, but the 72 and 80°C treated samples had almost identical KQs. The latter ®nding indicates that the spores were heat-shocked to the same extent at both temperatures. Addition of LP gave rise to large increases in KQ with all three pasteurization temperatures used, but there was no difference between the 72 and 80°C treatments. This result is rather surprising as a lower KQ may have been expected for 80°C, where most of the LP would have been eliminated (note that the detection limit for the assay is 4 lmol l±1 product min±1, but some activity remained after this heat treatment as a colour change could be detected visually in cuvettes on standing). One explanation could be that vegetative B. cereus is susceptible to attack by the LP system even when only a small proportion of normal LP activity is present. These results raise questions about the signi®cance of the LP system in pasteurized milk. It is well established that increasing the temperature of pasteurization leads to faster spoilage of milk (Kessler and Horak 1984; SchroÈder and Bland 1984; Schmidt et al. 1989). Also, Eberhardt and Gallmann (1988) showed that bacterial growth was greater a few days after pasteurization at higher temperatures (80± 90°C) than at 75°C. These apparent anomalies were attributed to stimulation of spore growth and, possibly, destruction of natural inhibitory components in the milk

Fig. 8 Total viable count of milks stored at 30°C, challenged with Bacillus cereus spores. (n), 63°C, no LP; (s),72°C, no LP; (h), 80°C, no LP; (m), 63°C, + LP system; (d),72°C, + LP; (j) 80°C, + LP

(Fredsted et al. 1996). The current study suggests that it is the action of the LP system on PPC bacteria and thermophiles that gives rise to the observed increase in KQ at 72°C compared with 80°C in pasteurized milks, rather than heatshocking of the spores. In fact, the results of studies with B. cereus indicate that heat-shocking was similar at the two temperatures. This emphasizes the need to be vigilant when selecting the pasteurization temperature in commercial practice. For example, there may be a temptation to compensate for inferior quality of raw milk by increasing the pasteurization temperature, which would probably give rise to faster spoilage. ACKNOWLEDGEMENTS This research was supported by a studentship from The University of Reading. The authors would like to thank DMV International (Veghel, Belgium) for enzyme used in this study. REFERENCES Barrett, N., Grandison, A.S. and Lewis, M.J. (1999) Contribution of the lactoperoxidase system to the keeping quality of pasteurised milk. Journal of Dairy Research 66, 73±80.



Eberhardt, P. and Gallmann, P.U. (1988) Haltbarheit und qualitaÈt von pasteurizierter milch. Landwirtschaft Schweiz 1, 13±16. Fredsted, L.B., Rysstad, G. and Eie, T. (1996) Pure-LacTM: the new milk with protected freshness and extended shelf-life. In Heat Treatments and Alternative Methods (Symposium, 1995). pp. 104± 125. Brussels: International Dairy Federation. Grif®ths, M.W. (1986) Use of milk enzymes as indices of heat treatment. Journal of Food Protection 49, 696±705. International Dairy Federation (1988) Code of Practice for the Preservation of Raw Milk by the Lactoperoxidase System. Bulletin Number 234. Brussels: International Dairy Federation. International Dairy Federation (1994) Indigenous Antimicrobial Agents of Milk ± Recent Developments (Seminar 1993). Brussels: International Dairy Federation. Kessler, H.G. and Horak, F.P. (1984) Effect of heat treatment and storage conditions on keeping quality of pasteurized milk. Milchwissenschaft 39, 451±454.

741

Reiter, B. and HaÈrnulv, G. (1984) Lactoperoxidase antibacterial system: Natural occurrence, biological functions and practical applications. Journal of Food Protection 47, 724±732. Schmidt, D., Cromie, S.J. and Dommett, T.W. (1989) Effects of pasteurization and storage conditions on the shelf life and sensory quality of aseptically packaged milk. Australian Journal of Dairy Technology 44, 19±24. SchroÈder, M.A. and Bland, M.A. (1984) Effect of pasteurization temperature on keeping quality of whole milk. Journal of Dairy Research 51, 569±578. Siragusa, G.R. and Johnson, M.G. (1989) Inhibition of Listeria monocytogenes growth by the lactoperoxidase-thiocyanate-H2O2 antimicrobial system. Applied and Environmental Microbiology 55, 2802±2805.
