Inactivation of Mycobacterium avium subsp. paratuberculosis in Cow's Milk by Means of High Hydrostatic Pressure at Mild Temperatures

Abstract

Two strains of Mycobacterium avium subsp. paratuberculosis (3644/02 and ATCC 19698) were inoculated (approximately 6 log CFU/ml) into sterilized milk to evaluate inactivation by high hydrostatic pressure. Reductions of M. avium subsp. paratuberculosis increased with pressure level. Significant differences were also found between M. avium subsp. paratuberculosis strains and between the media used. Average reductions of 4 log CFU/ml after treatment with 500 MPa are comparable to those caused by thermal treatments.

Paratuberculosis (Johne's disease) is a chronic granulomatous enteritis caused by Mycobacterium avium subspecies paratuberculosis that affects cattle and other ruminants. Its prevalence in food-producing animals is increasing globally, causing significant financial losses (17). Crohn's disease is a chronic regional enteritis affecting around 5 per 100,000 humans in developed countries (3, 7, 8). Due to the clinical and pathological similarities between the diseases, involvement of M. avium subsp. paratuberculosis in the etiology of Crohn's disease has been debated for many years. The introduction of specific PCR together with optimized processing procedures significantly increased the implication of M. avium subsp. paratuberculosis in disease causation (6, 7, 17, 18). Milk could be a potential vehicle of transmission of M. avium subsp. paratuberculosis from animals with Johne's disease to humans. The presence of M. avium subsp. paratuberculosis DNA in human patients without symptoms of inflammatory bowel disease suggests that this widespread bacterium is present in contaminated water, milk, and other sources (4, 23). Recent reports add evidence to this hypothesis and include the isolation of M. avium subsp. paratuberculosis from breast milk and intestinal and lymph nodes from patients with Crohn's disease. Viable M. avium subsp. paratuberculosis has also been detected in peripheral blood in a substantial proportion of individuals with Crohn's disease (18, 22, 23, 29, 30). The dairy industry is also concerned about M. avium subsp. paratuberculosis due to reports that suggest it may not be effectively inactivated by high-temperature, short-time pasteurization (72°C, 15 s). There is controversy regarding this particular issue, although differences in inactivation may be explained by different pasteurization times and equipment, different detection methods and selective media, separation of clumps, different initial loads, variation in heat sensitivity between M. avium subsp. paratuberculosis strains, and the use of freshly isolated or laboratory-adapted strains of M. avium subsp. paratuberculosis (5, 15, 16, 19).

In the last 15 years, treatment of milk and dairy products with high hydrostatic pressure (HHP) has been the subject of intensive research (24, 36), mostly involving gram-positive or -negative cells. However, acid-resistant microorganisms such as Mycobacterium spp. have never been included in these studies. Our objective was to determine the possibility of reducing counts of M. avium subsp. paratuberculosis in milk by means of moderate HHP treatments carried out at mild temperatures.

Strains and preparation of inocula.

The strains used were Mycobacterium avium subsp. paratuberculosis ATCC 19698 and Mycobacterium avium subsp. paratuberculosis 3644/02, a Neiker isolate from a natural case of clinical paratuberculosis in cattle, determined to be of cattle type (C) by IS1311 PCR-restriction endonuclease analysis (20) and of type [1-1] according to pulsed-field gel electrophoresis analysis (32). Both strains were grown in Middlebrook 7H9 broth (Difco, Detroit, MI) supplemented with Middlebrook oleic acid-albumin-dextrose-catalase enrichment (Becton, Dickinson and Company, MD) and 0.05% Tween 80 (Panreac Quimica SA, Barcelona, Spain) (7H9 medium) for 6 and 12 weeks, and then the MacFarland scale (Densimat; Bio-Mérieux, Marcy l'Etoile, France) was used to estimate cell concentration. The final inoculums were 8.5 log CFU/ml and 8.6 log CFU/ml for strains 3644/02 and 19698, respectively.

Inoculation.

Sixteen 30-ml aliquots of ultra-high-temperature-treated milk were inoculated with 1 ml of each strain suspension. Afterwards, the milk was dispensed in sterile, hermetic, 30-ml low-density polyethylene bottles (Azlon; Bibby Sterilin, Ltd., United Kingdom) and vacuum packed in a triple layer of plastic bags (bb4.l; Cryovac Packaging, Sant Boi de Llobregat, Spain). Inoculated samples were stored at 6 ± 2°C.

HHP treatments.

Milk samples were pressurized 24 h after inoculation by using discontinuous HHP equipment (model S-FL-850-9-W; Stansted Fluid Power, Stansted, United Kingdom). On each run, two milk samples were placed inside the product canister. They were stabilized at the required temperature (5 or 20°C) and then immediately subjected to 10 min of HHP treatment at 300, 400, or 500 MPa (as measured on the digital and analogue pressure indicators). Control samples not subjected to HHP (0.1 MPa) were used as the reference for initial M. avium subsp. paratuberculosis counts. The time of pressure buildup was 24 s per 100 MPa; the decompression time was 30 to 35 s.

Microbiological analysis.

At 24 h after pressurization, microbiological analysis was carried out. After vigorous shaking, a 1-ml aliquot of each treated spiked milk sample was diluted in 9 ml of sterile distilled water. Further 10-fold dilutions up to 1:100,000 were made for each aliquot. Then, 0.2 ml of four dilutions (1:1, 1:10, 1:1,000, and 1:100,000) were inoculated onto two tubes of Herrold′s egg yolk medium (HEY) supplemented with 2 μg/ml of mycobactin J (Allied Monitor, Inc., Fayette, MO) and onto another two tubes of agar-solidified 7H9 (Bacto Agar; Becton, Dickinson and Company, MD) and incubated at 37°C. Colonies of M. avium subsp. paratuberculosis were counted after an incubation period of 16 weeks. Some tubes were discarded due to contamination or desiccation during the incubation period. An aliquot of similarly treated, uninoculated ultra-high-temperature-treated milk and pressurization liquid were used as negative controls.

Statistical analysis.

Colony counts per tube were transformed into number of CFU in the original milk sample. This calculated inoculum value was assigned to all tubes with 100 colonies or more because confluent growth made individual colony count beyond this number impossible. Only dilutions with different counts in at least two levels of pressure were used for statistical analysis after logarithmic transformation. This resulted in using dilutions of 1/1, 1/10, and 1/1,000 on both media for strain 3644/02 and dilutions of 1/1,000 and 1/100,000 on 7H9 and of 1/1 and 1/10 on HEY for strain 19698. The general linear models procedure of the SAS System version 8.0 (SAS Institute Inc., Cary, NC) was used to test main effects of strain, pressure, temperature, and culture medium, as well as their interactions, on the final colony count logarithm. A Dunnett test was used to assess the significance of differences between the control level and each other level of each main factor, and a Tukey-Kramer multirange test was used for testing the differences between paired-interaction marginal means.

The overall analysis of variance including all effects and interactions showed that strain (P < 0.0006), culture medium (P < 0.0001), and pressure (P < 0.0001) had strong effects on the colony counts. However, the assayed temperature levels had no significant effects (P = 0.2783), and therefore this effect was eliminated from further statistical analysis (data not shown). As can be deduced from Table 1, observing data for both strains and culture media, 300-, 400-, and 500-MPa treatments caused significant (P < 0.05) reductions of M. avium subsp. paratuberculosis counts of 0.92, 1.59, and 4.58 log CFU/ml, respectively, compared to control samples as determined by the Dunnett test. Strain 3644/02 had an average recovery of 4.54 log CFU/ml, which was 1.68 log CFU/ml higher than the recovery of strain 19698 (P < 0.05), independently of other factors. For both strains and all pressures, the tubes of 7H9 medium had a recovery rate 1.72 log CFU/ml higher than those of HEY medium. However, there was a strong interaction between strain and culture medium (P < 0.0001); nonsignificant differences in recovery occurred for strain 3644/02 in both media (5.01 log CFU/ml on 7H9 versus 4.24 log CFU/ml on HEY, for all pressures), while very few colonies of strain 19698 grew on HEY compared to 7H9 medium (4.49 versus 1.27 log CFU/ml, respectively). Therefore, another analysis was carried out, considering only results on 7H9 for strain and pressure factors. This analysis confirmed the significant differences between control and 500-MPa-treated samples for both strains (4.85 log CFU/ml; P < 0.05), but this was not the case for the other two pressure levels. The analysis of interactions further supported this result, since the differences between control and 500-MPa-treated samples were 4.07 log CFU/ml (P = 0.01) and 5.20 log CFU/ml (P < 0.001) for strains 3644/02 and 19698, respectively (Table 1).

TABLE 1.

HHP treatment effects on counts of Mycobacterium avium subsp. paratuberculosis strains 3644/02 and 19698 in different culture media^a

Medium	Pressure (MPa)	Strain 3644/02			Strain 19698			Both strains
		No. of replicates	Counts (log CFU/ml)		No. of replicates	Counts (log CFU/ml)		No. of replicates	Counts (log CFU/ml)
			Mean	SD		Mean	SD		Mean	SD
7H9	0.1	30	5.90	1.66	20	6.10	2.63	50	5.98	2.08
	300	7	5.22	2.74	14	5.49	3.02	21	5.40	2.86
	400	19	4.37	2.53	13	5.37	3.07	32	4.78	2.76
	500	5	1.83	1.03	16	0.90	1.98	21	1.13	1.82
	All	61	5.01	2.33	63	4.49	3.37	124	4.75	2.91
HEY	0.1	30	6.28	1.24	20	1.98	1.43	50	4.56	2.50
	300	19	5.66	2.00	15	1.23	1.06	34	3.70	2.76
	400	24	3.94	2.67	16	1.11	0.93	40	2.81	2.55
	500	21	0.39	0.65	14	0.50	0.83	35	0.44	0.72
	All	94	4.24	2.87	65	1.27	1.22	159	3.03	2.75
Both	0.1	60	6.09	1.46	40	4.04	2.95	100	5.27	2.39
	300	26	5.54	2.17	29	3.29	3.08	55	4.35	2.90
	400	43	4.13	2.59	29	3.02	3.02	72	3.68	2.80
	500	26	0.67	0.92	30	0.72	1.54	56	0.69	1.28
	All	155	4.54	2.69	128	2.86	2.98	283	3.78	2.94

^aThere were no significant differences between treatments at 5 and 20°C.

The counts of M. avium subsp. paratuberculosis obtained at 5°C were similar to those obtained at 20°C. This fact turns out to be very useful, since it means that mild HHP treatments could be applied to milk at refrigeration temperatures, thus minimizing organoleptic and nutritional alteration. As can be seen in Table 1, HHP inactivation of both strains of M. avium subsp. paratuberculosis was proportional to the increase in the pressure applied, an effect which was also observed in both culture media. This is generally the tendency observed among vegetative bacteria (12, 31). However, there were strong interactions between strain and medium, which indicate that assay conditions are very important for the assessment of inactivation of M. avium subsp. paratuberculosis. This was quite evident for strain 19698 in HEY, where even the counts of control samples were reduced by 4.12 log CFU/ml relative to their growth on 7H9 (Table 1). In contrast, strain 3644/02 grew similarly in both media. Preference of different strains of M. avium subsp. paratuberculosis for primary isolation media has been observed before (1).

Different piezotolerances between strains of the same species have been observed by many researchers (2, 9, 27). From the way that both M. avium subsp. paratuberculosis strains responded to HHP treatments, it seems that strain 3644/02 was slightly more pressure resistant than strain 19698.

The mean reduction obtained for strain 3644/02 at 500 MPa was 4.07 log CFU/ml. This value suggests the possibility of inactivating this microorganism by means of HHP when its load in raw milk is less than 4 log CFU/ml. Many authors estimate that raw bulk tank milk could reach levels of up to 4 and 5 log CFU/ml of M. avium subsp. paratuberculosis cells (11, 13, 14, 16, 21, 25, 34). This is mostly due to fecal contamination and also, to a lesser extent, to cells secreted within the udders (5, 33, 35); Sweeney et al. (35) found 2 to 8 CFU/ml (0.3 to 0.9 log CFU/ml) of M. avium subsp. paratuberculosis cells in milk samples aseptically obtained from the udders of nine subclinical animals.

Nevertheless, HHP inactivation values are comparable to those obtained in pasteurization studies (11, 17) and to those obtained with pulsed electric fields (28), but they have the advantage of using relatively lower temperatures. It is still possible to apply pressures higher than 500 MPa to raw milk. HHP technology could also be applied to soft-curd cheeses traditionally made from raw milk in order to improve its safety. The work of Donaghy et al. (10) states that M. avium subsp. paratuberculosis cells survive short-term curing when present in high (up to 5 log CFU/ml) or low (2 log CFU/ml) numbers as well. Due to the isostatic condition of HHP treatments (which means that pressure is homogeneously applied to every point of the sample), large differences are not expected when larger samples and HHP machines are used. However, aggregated M. avium subsp. paratuberculosis cells could still interfere and modify the reductions obtained. Moreover, temperature gradients due to the adiabatic heating of the pressure chamber may vary in laboratory and industrial equipment. This is why pilot-scale experiments such as those made with pasteurization treatments (15, 26) may be useful. Further research should be done in order to determine the nature of HHP injuries in this sort of vegetative cell and to evaluate the behavior of pressurized M. avium subsp. paratuberculosis cells during storage life at refrigeration temperatures.