from Cheese - Applied and Environmental Microbiology - American ...

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Nov. 1988, p. 2643-2649

Vol. 54, No. 11

0099-2240/88/112643-07$02.00/0 Copyright C 1988, American Society for Microbiology

Methods for Improved Recovery of Listeria from Cheese

monocytogenes

AHMED E. YOUSEF, ELLIOT T. RYSER, AND ELMER H. MARTH* Department of Food Science and The Food Research Institute, University of Wisconsin-Madison, Madison, Wisconsin 53706 Received 6 May 1988/Accepted 1 August 1988

Method of homogenization (Waring blender versus stomacher), type of diluent (tryptose broth [TB] versus 2% trisodium citrate), and temperature of diluent (20 versus 40°C) were compared for recovery of Listeria monocytogenes from freshly made and ripened Colby cheese. By using direct plating on McBride listeria agar, significantly higher numbers of L. monocytogenes were recovered when cheese samples were (i) homogenized for 2 min with the blender rather than the stomacher (P < 0.01), (ii) diluted in trisodium citrate rather than TB (P < 0.01), and (iii) diluted in diluents at 40 rather than 20°C (P < 0.05). Based on these results, a new diluent/enrichment medium was developed by adding 2% trisodium citrate to TB (TBC). Despite superior results with the blender, biosafety concerns led to use of the stomacher for homogenization of cheese samples; hence, the stomaching time was increased to 3 min. Results obtained by direct plating indicated that recovery of L. monocytogenes from Colby cheese and from curd samples taken during manufacture of brick cheese increased when samples were diluted 1:10 in TBC at 45°C and stomached for 3 min, as compared with similarly treated samples diluted in TB at 25°C. A similar comparison of both diluents for recovery of L. monocytogenes from cold-pack cheese food yielded bacterial counts which were not significantly different. Recovery of L. monocytogenes from cold-enriched (at 4°C for up to 8 weeks) samples of Colby cheese and cold-pack cheese food was generally similar for samples homogenized in TBC or TB. aqueous

Interest in detecting Listeria monocytogenes in dairy products has escalated over the past few years because at least 160 cases of listeriosis, including 56 deaths, were linked to consumption of pasteurized milk (2), Mexican style cheese (4), Vacherin Mont d'Or soft-ripened cheese (Anonymous, Food Chem. News, 7 Dec. 1987), and feta cheese (M. P. Doyle, personal communication) contaminated with the pathogen. In response to these outbreaks, increased surveillance of the dairy industry has led to isolation of L. monocytogenes from cheeses such as Mexican style (Anonymous, Food Chem. News, 10 Mar. 1986), Liederkranz (Anonymous, FDA Enforcement Rep., 4 Sept. 1985), ricotta, Cheddar, and Parmesan (J. G. Nichols, Abstr. J. Food Prot., vol. 50, p. 889, 1987), Brie (Anonymous, FDA Enforcement Rep., 9 Apr. 1986; Anonymous, Food Chem. News, 3 Mar. 1986; Anonymous, Food Chem. News, 17 Feb. 1986; Anonymous, FDA Enforcement Rep., 23 Apr. 1986), and semisoft and soft-ripened types (Anonymous, Food Chem. News, 18 Aug. 1986) imported from France. Recent efforts have focused on developing selective media and shortened enrichment procedures to enumerate L. monocytogenes in food samples containing a mixed microflora; however, little attention has been given to the effects of sample preparation on recovery of L. monocytogenes from cheese by using direct plating. According to standard methods (13), preparation of cheese for microbiological analysis involves diluting a sample of ground cheese in aqueous 2% trisodium citrate at 40 to 45°C, followed by homogenization with a blender or stomacher, and then plating the cheese homogenate on an appropriate medium. Although this method is reliable for quantitating a variety of bacteria, the method lacks the sensitivity to detect the small numbers of L. monocytogenes occasionally found in cheese. Therefore, enrichment of the homogenized cheese sample is *

usually required before the organism can be isolated. Since L. monocytogenes fails to grow in aqueous trisodium citrate, the present standard method is of no value for detecting low levels of the pathogen in cheese. In 1948, Gray et al. (3) observed that L. monocytogenes could be isolated from infected bovine brains that were diluted in tryptose broth (TB) and held at 4°C for several months. This method has since been used in medical laboratories to isolate the pathogen from clinical specimens. Although the extended incubation period makes this procedure of little value to the food industry, cold enrichment is still of academic importance and remains among the most reliable methods for detection of L. monocytogenes. In recent studies concerning the fate of L. monocytogenes during manufacture of cottage (16), Cheddar (14), and Camembert (15) cheese, samples were diluted in TB at ca. 25°C and homogenized in a stomacher for 30 s. If L. monocytogenes was not found initially by direct plating on McBride listeria agar, this procedure allowed possible detection of the pathogen in cheese homogenates during cold enrichment. However, using this method, Ryser and Marth (14, 15, 16) failed to observe expected increases in numbers of L. monocytogenes which occur during cheese manufacture as a result of concentration of cells during curd formation. The apparent insensitivity of this method led to the current study, which examines multiple effects of various factors of sample preparation on recovery of L. monocytogenes from cheese. Results from the first part of this study led to development of a new medium which increased recovery of L. monocytogenes by direct plating and, at the same time, was suitable for cold enrichment. MATERIALS AND METHODS

Media and diluents. The following bacteriological media (Difco Laboratories, Detroit, Mich.) and diluents were prepared, using deionized water: TB, TB containing 2% triso-

Corresponding author. 2643

2644

YOUSEF ET AL.

dium citrate (TBC), aqueous 2% trisodium citrate solution (CS), aqueous 0.5% peptone solution, and tryptose agar. McBride listeria agar was prepared from phenylethanol agar plus 1% glycine anhydride plus 0.05% (for Colby cheese) (7) or 0.5% (for cold-pack cheese food and brick cheese) (6) lithium chloride plus 5.0% defibrinated sheep blood (GIBCO Diagnostics, Madison, Wis.). Cultures. Four strains of L. monocytogenes were used in this study: Scott A (serotype 4b, clinical isolate), V7 (serotype 1, milk isolate), California (CA) (serotype 4b, isolated from Mexican style cheese implicated in the 1985 outbreak of listeriosis in California), and Ohio (OH) (serotype 4b, isolated from recalled Liederkranz cheese manufactured in Ohio). All cultures were maintained at 4°C on tryptose agar slants and were transferred bimonthly. Cultures of L. monocytogenes used to inoculate cold-pack cheese food and milk for manufacture of brick cheese were prepared as described previously (14). Manufacture of cheeses: culture and rennet. A commercial Streptococcus cremoris starter culture (CC6; Marschall Division, Miles Laboratories, Inc., Madison, Wis.) was used to manufacture Colby and brick cheese. The starter culture was incubated for 16 to 18 h at 21°C in previously steamed (45 min, 100°C) reconstituted nonfat dry milk (10% total solids). A sufficient volume of the resulting starter culture was prepared to provide an inoculum of 1% (wt/wt) (for Colby cheese) or 0.25% (wt/wt) (for brick cheese). Microbial rennet from the supplier that provided cultures was used in manufacturing Colby and brick cheese. Colby cheese. Colby cheese was made according to the procedure described by Olson (11) or Kosikowski (5), as indicated in a previous study (22). Pasteurized (75°C for 16 s) whole milk (113 kg) was placed in a pilot-plant-size cheese vat, warmed to 31°C, and inoculated to contain ca. 103 L. monocytogenes (strain CA or V7) CFU/ml, after which lactic starter culture (1%) was added. Each batch of Colby cheese resulted in two 10-lb (ca. 4.6-kg) blocks which were vacuum packaged and ripened at 4°C. Brick cheese. Four duplicate batches of brick cheese were manufactured according to the washed-curd method (10) from 190.5 kg of pasteurized (75°C for 16 s) whole milk. Milk was inoculated to contain ca. S x 102 L. monocytogenes (strain Scott A, V7, CA, or OH) CFU/ml. Starter culture and microbial rennet were added to pasteurized whole milk tempered to 31°C in the cheese vat. After the coagulum was cut, the temperature of the curd-whey mixture was gradually increased from 31 to 39°C. After 30 min of cooking at 39°C, half of the whey was drained and replaced with tap water at 39°C. The whey-water mixture was drained after 30 min, and an approximately equal amount of curd was placed in each of nine hoops. All hoops of cheese were periodically turned for 6 h, after which time the finished cheese was placed in a 22% salt brine at 10°C. After 24 h, blocks of brick cheese were removed from brine, smeared with a culture of Brevibacterium linens (obtained from the surface smear of a commercially produced Limburger cheese), and ripened at 15°C and 95% relative humidity for 2, 3, and 4 weeks to obtain mild, medium, and Limburger-like brick cheese, respectively. Cheese was ripened at 10°C and analyzed for numbers of L. monocytogenes and pH during a 6-month period. Cold pack cheese food. Cold-pack cheese food of three different compositions was manufactured in duplicate 10.4kg (22.9-lb) batches according to the following formula: 8,000 g of Cheddar cheese (aged for 6 to 9 months), 471 g of nonfat dry milk, 471 g of dried whey, 360 g of butter, and 1,100 ml of water (1Sa). Duplicate batches of cheese food

APPL. ENVIRON. MICROBIOL.

were manufactured to contain either 0.30% sorbic acid or 0.30% sodium propionate. Two additional batches were prepared without preservative. All six batches of cheese food had a pH of ca. 5.4. After Cheddar cheese was ground, all ingredients were combined and mixed for 3 min at high speed, using a pilot-plant-sized sausage grinder. Each batch of cheese food was divided into four 2-kg portions which were inoculated to contain ca. S x 102 L. monocytogenes (strain Scott A, V7, CA, or OH) CFU/g. Inoculated portions of cheese food were thoroughly mixed for 10 min to ensure uniform distribution of the organism. Cheese food was then packed into sterile 8-ounce (226.8-g) glass jars and stored at 4°C until time of analysis for L. monocytogenes, pH, moisture, fat, and salt content. Results concerning the fate of L. monocytogenes in cold-pack cheese food during extended storage were reported elsewhere (1Sa). Sampling cheese to test for L. monocytogenes: Colby cheese. Duplicate sets of samples were taken for enumeration of L. monocytogenes at the following points of cheese manufacture: (i) inoculated milk, (ii) curd after cooking, (iii) curd after salting, (iv) curd after pressing (day 0), and (v) cheese after 42 and 84 days of ripening at 4°C. During cheese manufacture, one set of duplicate curd (10-g) samples was placed in sterile (ca. 500-ml capacity) stomacher bags, diluted in 90 ml of TB at 20°C, and homogenized in a stomacher 400 (Tekmar Co., Inc., Cincinnati, Ohio) for 2 min. The remaining set of 10-g samples was placed in an autoclaved stainless-steel blender jar (250-ml capacity), diluted with 90 ml of CS at 40°C, and homogenized with a Waring blender (model 1042; Waring Products Co., Winsted, Conn.) at high speed for 2 min. The influence of the following variables in sample preparation on recovery of L. monocytogenes from 0-, 42-, and 84-day-old samples of Colby cheese was examined: (i) method of homogenization, (ii) type of diluent in which the cheese was homogenized, and (iii) temperature of diluent added to cheese before homogenization. In this study, a section of the cheese block was cut, exterior surfaces (ca. 1-cm thick) were sliced and removed, and then the cheese section was ground and mixed thoroughly. Duplicate 10-g samples of ground cheese were diluted in 90 ml of TB or CS held at either 20 or 40°C and were homogenized as previously described, using a blender or stomacher. Two additional sets of duplicate 10-g samples were taken from 5- to 8-month-old blocks of Colby cheese. One set of samples was diluted in 90 ml of TB at ca. 25°C, whereas the remaining set was diluted in TBC at 45°C. Both sets of samples were homogenized with the stomacher for 3 min. Brick cheese. Two sets of duplicate 10-g samples were taken for enumeration of L. monocytogenes during manufacture of brick cheese according to the following scheme: (i) inoculated milk, (ii) curd after cutting, (iii) curd before washing, (iv) curd at dipping, (v) curd at fourth turn, and (vi) curd at brining. Both sets of samples were diluted in TB or TBC and homogenized with the stomacher as described previously for Colby cheese. Cold-pack cheese food. Six batches of cold-pack cheese food were made as described previously. Each batch was divided into four equal portions and inoculated with one of four strains of L. monocytogenes. Two sets of duplicate 10-g samples were taken from each portion of cheese food after 0, 7, and 14 days of storage at 4°C. Both sets of samples were diluted in TB or TBC and homogenized with the stomacher as described previously for Colby cheese. Detection of L. monocytogenes in cheese. Duplicate 10-g samples taken during manufacture, ripening, or storage of

VOL. 54, 1988

IMPROVED RECOVERY OF L. MONOCYTOGENES FROM CHEESE

cheese were homogenized according to methods indicated earlier. Initial or subsequent dilutions of homogenized cheese samples were surface plated in duplicate on McBride listeria agar. All plates were incubated for 48 h at 35°C in an atmosphere of 5% 02-10% C02-85% N2. Colonies typical of those formed by L. monocytogenes (smooth, bluish gray, slightly raised, translucent, watery consistency, 0.5 to 1.5 mm in diameter, and weakly beta-hemolytic) were counted. Selected colonies were confirmed by the procedure of Ryser and Marth (14), which is based on a positive catalase reaction, tumbling motility in TB, appearance of bluish colonies on tryptose agar, and results of biochemical tests found on API 20S test strips (Analytab Products, Plainview, N.Y.). In addition to direct plating, selected cheese samples were also analyzed for L. monocytogenes by the cold-enrichment method as described previously (14). In this procedure, the original sample diluted 1:10 in TB or TBC was stored at 4°C and surface plated on McBride listeria agar at 2-week intervals for up to 8 weeks. Following 48 h of incubation at 35°C under microaerobic conditions, all plates were examined for colonies resembling L. monocytogenes. If present, the organism was confirmed as described previously (14). Statistical analysis. In our work with Colby cheese, we studied the effect of three factors of cheese sample preparation (two levels each, a 23 factorial design) on recovery of L. monocytogenes (log10) by direct plating. These factors were (i) method of homogenization (stomaching versus blending), (ii) type of diluent in which cheese was homogenized (TB versus CS), and (iii) temperature of diluent (20 verus 40°C). Data were analyzed statistically, using the analysis of variance procedure of SAS statistical programs. The following split-block linear model (19) was used to describe the data: Y,kl = + pi + F, + Tk + PF + ,Tik + FTik + FTijk + bijkl where Y,kl is the dependent variable (log1o CFU of L. monocytogenes per gram of cheese), P3i is batch of cheese (considered blocking factor, since two batches of cheese were made at different times, using different initial populations of L. monocytogenes in milk), F. is the treatment factors (eight combinations of the three factors studied), and Tk is age of cheese (0, 42, and 84 days). The SAS general linear models procedure was used to subdivide further the sum of squares for the F. term so that the factors studied (method of homogenization and diluent type and temperature) and their interactions could be individually tested. The error term used to test each factor of interest was the mean sum of squares of the interaction of this factor with the blocking factor (19). Data on brick cheese and cold-pack cheese food were analyzed statistically by one- or two-way analysis of variance, using MINITAB statistical programs. RESULTS AND DISCUSSION Colby cheese. During manufacture of Colby cheese, samples of curd were prepared for enumeration of L. monocytogenes by two methods. The first method is that described in Standard Methods for the Examination of Dairy Products (13) and involves diluting curd (or cheese) in warm CS and then homogenizing the mixture with a blender. The second method, which is more suitable for handling pathogenic microorganisms, involves homogenizing a diluted curd or cheese sample with a stomacher. Cool TB was previously used in our laboratory as a diluent in the second method (14, 15, 16). The major advantage of this method is that samples

lo$

2645

*

I

C-.b \ 10'-_

0

r

1

2

3

4

Manufacture(h)

0

'X'

20

40

60

80

1 00

Ripening (d)

FIG. 1. Recovery of L. monocytogenes from samples taken during the manufacture and storage of Colby cheese made from pasteurized milk inoculated with strain CA. Symbols: (0) samples homogenized with the blender, using CS at 40°C; (U) samples homogenized with the stomacher, using TB at 20°C. d, Days.

homogenized in TB also can be used for cold enrichment, which is usually needed to detect very small numbers of L. monocytogenes in cheese. Data in Fig. 1 indicate that variable numbers of the pathogen were detected in curd samples analyzed by both methods. Differences in recovery of L. monocytogenes from curd by the two methods increased as cheese manufacture progressed. Apparently, as the curd became firmer, differences between the number of L. monocytogenes recovered by the two methods became larger. Colby cheese was sampled during ripening to study the effect of different parameters in preparing samples on the number of L. monocytogenes recovered from the cheese. The log1o values of the L. monocytogenes count per gram in samples obtained from 0-, 42-, or 84-day-old cheese and prepared by each of eight methods (three factors were varied; each had two levels) are shown in Fig. 2. These results suggest that fewer listeriae were recovered from cheese that was homogenized with the stomacher than from cheese homogenized with the blender. Statistical analysis of the data confirmed that assessment (Table 1). Samples homogenized in CS yielded significantly (P < 0.01) higher numbers of L. monocytogenes than did those prepared in TB. The effect of diluent temperature on recovery of L. monocytogenes from cheese cannot be ascertained directly from data in Fig. 2. Statistical analysis of data, however, shows that significantly (P < 0.05) higher numbers of L. monocytogenes were recovered from samples homogenized in warm (40°C) rather than cool (20°C) diluent. Some of the experimental factors had significant interactions with age of cheese, implying that the effect of these factors was different at different points in the cheese-ripening process. The method of homogenization and the type of diluent each had a significant (P < 0.01) interaction with age of cheese across the three sampling intervals during cheese ripening (0, 42, and 84 days). For example, consider cheese samples diluted in TB at 20°C and homogenized with the stomacher (Al and Bi, Fig. 2) and samples diluted in CS at 40°C and homogenized in the blender (A8 and B8). Differences in recovery of L. monocytogenes from cheese samples prepared by these two methods were quite large at the early stages of ripening (0 and 42 days), but were minimal after 84 days of ripening. This interaction can also be seen in Fig. 1, which may be attributed to increased proteolysis of casein in older rather than younger cheese.


YOUSEF ET AL.

2646

TABLE 2. Ranking and comparison of the eight methods of homogenizing Colby cheese samples by Fisher's least-significant-difference method at P = 0.05 Variables studied

Method' c,. C_D

4

CD

3.75

CD

9D

111111

1

3.50

3.25 3.00 2.75

R1R2R3R4 R5R6R7 R8 B IB2B3B4B5 B6B7B8 METHO

FIG. 2. Average numbers of L. monocytogenes (CFU per gram of cheese) in samples taken during the storage of Colby cheese and homogenized by different methods. The letters A and B designate the first and second batches, respectively. The digits designate the methods of preparing the samples as follows: 1 and 2, stomaching in TB at 20 or 40°C; 3 and 4, stomaching in CS at 20 or 40°C; 5 and 6, blending in TB at 20 or 40°C; 7 and 8, blending in CS at 20 or 40°C, respectively.

Although the split-block design allowed a formal factorial analysis of the treatments, it is useful to order and compare the eight procedures by using Fisher's least-significantdifference method at P = 0.05. Results of this analysis indicate that the numbers of L. monocytogenes recovered from samples prepared with the blender are not significantly different (Table 2). Numbers of L. monocytogenes, recovered from samples diluted in warm CS and homogenized with the stomacher are not significantly different from those recovered with the blender and warm TB or CS. Therefore, if the stomacher is to be used to prepare cheese samples, the samples should be homogenized in warm CS. Thrasher and Richardson (20) reported no significant difference in numbers of organisms recovered from stomached rather than blended Cheddar cheese samples that were diluted 1:10 in CS. von Asperger and Brandl (21) TABLE 1. Results of analysis of variance for recovery of L. monocytogenes from Colby cheese samples prepared by eight different methods, using the split-block model Source

df

Sum of

squares

p

Batch (l)

1

4.810

0.0001

Treatment factors (F) Method of homogenization (Me) Type of diluent (Dil) Temp of diluent (Temp) Me x Dil Me x Temp Dil x Temp Me x Dil x Temp Error 1 (3 x F)

7 1 1 1 1 1 1 1 7

2.770 1.840 0.353 0.061 0.330 0.158 0.011 0.017 0.057

0.0001 0.0001 0.0003 0.0300 0.0004 0.0032 0.2956 0.1899 0.1240

Age of cheese (7) Error 2 (j x 7)

2 2

17.50 0.179

0.0001 0.0001

14

0.963

0.0001

14

0.135

0.0352

Treatment factors-cheese age interaction (F x 7) Error 3 (3 x F x 7)

Bl Bl BJ Bl St St St St

Mean

T(eOC)p

log1 CFUgb

CS

20

TB TB

20 40 40 40 20 40 20

4.05 4.04 4.01 4.01 3.96 3.78 3.67 3.59

Diluent

CS CS CS TB TB

Grouping A A A B A B B C D E

aMethod of homogenization was blending (BI) or stomaching (St). b Average log1o of L. monocytogenes count in 12 samples. Means with the same letters are not significantly different.

obtained lower bacterial counts from stomached rather than blended samples or quarg, Gervais, and cottage cheese that were previously diluted in 2% trisodium citrate at 37°C. In a study by Nanni (8), cheese samples of various ages, ripeness, structure, and consistency were immersed in physiological saline solution 15 min before homogenization with a Stomacher for 10 to 25 min or with an Ultra-Turrax blender for 2 to 4 min. Populations of microorganisms recovered by these two methods were not significantly different. Our data generally indicate that the blender was more effective than the stomacher in preparing curd and cheese samples for enumeratiQn of L. monocytogenes. The type and temperature of the diluent, as well as the age of cheese, had a greater impact on recovery of listeriae from cheese that was homogenized with the stomacher than when the blender was used (Fig. 2). Development of a procedure for homogenizing and cold enriching cheese samples. For years, TB has been the medium of choice for cold enrichment (at 3°C) of clinical and occasionally food specimens so that small numbers of L. monocytogenes could be detected (17). Although cold enrichment in TB is not useful in the food industry (because several weeks are required to complete the test), this method remains among the most reliable for detection of L. monocytogenes and is particularly valuable for detecting injured cells (16). Therefore, TB was chosen as the medium to

modify. Based on our results from comparing methods of homogenization and type and temperature of diluent for recovery of L. monocytogenes by direct plating, the procedure of Ryser and Marth (14) for homogenizing and cold enriching cheese samples was modified as follows: (i) trisodium citrate (2%) was added to TB to improve the emulsification of cheese samples, (ii) temperature of the diluent was increased from 40 to 45°C as described in Standard Methods (13), and (iii) the time of homogenization in the stomacher was increased from 0.5 to 3 min. Although larger numbers of listeriae were routinely detected in samples processed with the blender rather than the stomacher, aerosols containing L. monocytogenes can be produced during blending. Since cases of human listeriosis attributable to the airborne pathogen have been reported (9), preparing contaminated samples with the blender in a poorly equipped laboratory may pose a health hazard to the analyst and other laboratory workers. Other advantages of the stomacher include use of disposable, presterilized, Whirl-pak bags for diluting and cold enriching

2647


VOL. 54, 1988

TABLE 3. Comparison of TB and TBC for detection of L. monocytogenes in cold-pack cheese food, using direct plating (DP) and cold enrichment (CE)

10-

% False-

No. of samples

negative

TBC

TB

Product DP +10

-

TB +

_

+

TBC

_

6 24 18 12 33.3 33.3 3 27 20 10 Colby cheese Cheese food 137 0 127 10 141 0 139 2 7.3 1.4

'-

0

+

CE

DP

CE -_____

2

4

6

TIME (h) FIG. 3. Average numbers of L. monocytogenes (four strains) recovered during manufacture of brick cheese. Symbols: (0) samples diluted in TBC at 45°C and stomached for 3 min; (U) samples diluted in TB at 25°C. and stomached for 3 min.

samples and the relative ease of processing large numbers of samples. Homogenization with the stomacher rather than the blender also was shown to be less injurious to a variety of gram-positive bacteria (18). Deibel and Banwart (1) found that stomaching was markedly less efficient than blending in breaking up clumps and chains of some foodborne pathogens. Yersinia enterocolitica, which is similar to L. monocytogenes in that neither organism forms clumps or chains, was used as a control. After stomaching and blending, 97 and 100% of yersiniae were present as single cells, respectively. Based on our observations and this discussion, we recommend using the stomacher for homogenizing cheese samples thought to contain listeriae. This procedure was evaluated with a variety of cheeses as indicated in the following discussion.

Brick cheese. Results for recovery of L. monocytogenes during manufacture of brick cheese are shown in Fig. 3. Two sets of samples taken at each point during cheese making were compared statistically, using the analysis of variance. Following inoculation of milk, significantly (P < 0.01) higher (2.2- to 3.0-fold) populations of L. monocytogenes were consistently observed in curd samples diluted in TBC at 45°C rather than in TB at 25°C. Approximately 8.5 h after the start of brick cheese manufacture, populations of L. monocytogenes were 22.4- and 7.6-fold greater in curd samples diluted in TBC and TB, respectively, than in the orginal population in the inoculated milk. Data from a previous study (22) indicate that the population of L. monocytogenes in Colby cheese (with a yield of ca. 10%) was ca. 10-fold higher than that in milk from which the cheese was made. This 10-fold increase can be attributed to entrapment of L. monocytogenes in curd during cheese manufacture. Similarly, numbers of L. monocytogenes in brick cheese are expected to be ca. 10-fold higher than in milk from which this cheese is manufactured. Our results indicate that growth of L. monocytogenes during manufacture of brick cheese (22.4-fold increase rather than the expected 10-fold increase resulting from curd formation) was only observed when curd samples were diluted in TBC. Using TB at 25°C, Ryser and Marth failed to detect a 10-fold increase in numbers of L. monocytogenes during manufacture of Cheddar (14) and Camembert (15) cheeses. In all probability, an increase of at least 10-fold would have been observed if the samples had been diluted in TBC at 45°C and stomached for 3 min. Cold-pack cheese food. Average populations of L. monocytogenes observed in 0-, 7-, and 14-day-old cheese food

diluted in TBC and TB were 2.43 and 2.41 log1o CFU/g, respectively, and were not significantly different (results not shown). These results for direct plating were not totally unexpected since cold-pack cheese food was readily spreadable and easily emulsified in both TB and TBC. Since substantially larger populations of L. monocytogenes were recovered during manufacture of brick cheese with TBC rather than TB, the next consideration was to determine the feasibility of using TBC as a cold-enrichment medium. Using 0-, 7-, and 14-day-old samples of cold-pack cheese food in which L. monocytogenes was previously found by direct plating, we failed to detect the organism in 2 of 141 (1.4%) and 10 of 137 (7.3%) samples diluted in TBC and TB, respectively, after 8 weeks of cold enrichment (Table 3). Although cold enrichment was less successful than direct plating for recovery of L. monocytogenes from cheese food, TBC was an acceptable cold-enrichment medium for L. monocytogenes as evidenced by increased recovery of the organism with TBC rather than TB. According to Ralovich (12), trisodium citrate enhances growth of L. monocytogenes in minimal media. Thus, presence of trisodium citrate may be partly responsible for increased recovery of L. monocytogenes from cheese samples diluted in TBC rather than TB. Colby cheese. The experiment with Colby cheese was done to assess the ability of TBC when used as a cold-enrichment medium to recover low levels of L. monocytogenes from a hard cheese. In this study, TBC at 45°C and TB at 25°C were compared for recovery of L. monocytogenes from 5- to 8-month-old samples of Colby cheese, using both direct plating and cold enrichment. Eight batches of Colby cheese, each consisting of two blocks (A and B), were studied, and two samples were taken from each cheese block. Data for direct plating (Fig. 4) indicate that L. monocytogenes was detected in 6 of 30 (20%) and 3 of 30 (10%) samples diluted in TBC and TB, respectively. Five of nine samples positive for L. monocytogenes by direct plating came from the youngest cheese (5 months old, batch 1). In contrast, the organism was not detected by direct plating in the oldest cheese (8 months old, batch 8). With TBC as the diluent, a comparison between cold enrichment and direct plating indicates that L. monocytogenes was not detected by cold enrichment in two of six (33%) samples which were previously positive by direct plating. Similarly, with TB as the diluent, the organism was not detected in one of three (33%) samples which were previously positive by direct plating. Thus, cold enrichment in either medium resulted in 33% of the samples being falsely reported as negative. Cold-enrichment results (Fig. 4) indicate that 18 (60%) and 20 (67%) of 30 samples were positive for L. monocytogenes with TBC and TB, respectively. Two samples in which L. monocytogenes was detected after cold enrichment in TBC were negative after cold enrichment in TB. Similarly, four

2648


YOUSEF ET AL.

inated cheese samples with the stomacher. In addition to allowing greater recovery of listeriae from cheese by direct plating, use of our enrichment medium (TBC) compared favorably with use of TB for recovery of the organism during cold enrichment. ACKNOWLEDGMENTS Our research was supported by the College of Agricultural and Life Sciences; the National Dairy Promotion and Research Board through a grant administered by the Dairy Research Foundation, Rosemont, Ill.; the Wisconsin Milk Marketing Board, Madison; and Kraft, Inc., Glenview, Ill. We thank E. Nordheim for reviewing the statistical analysis of the data. We also thank R. M. Twedt, Food and Drug Administration, Cincinnati, Ohio, for providing strains Scott A and V7; J. H. Silliker, Silliker Laboratories, Carson, Calif., for providing strain CA; and M. P. Doyle, The Food Research Institute, University of Wisconsin-Madison, for providing strain OH.

I U

4

3

2

140

10

60

DP CE TBC

DP CE

TB

DP CE TBC

DP CE TB

FIG. 4. Recovery of L. monocytogenes from eight batches of 5to 8-month-old Colby cheese, each consisting of two blocks (A and B), by direct plating (DP) of samples prepared in TBC at 45°C or TB at 25°C and homogenized with the stomacher for 3 min. All samples

enrichment (CE). Each square represents adjacent vertical squares within the same batch number represent samples from the same batch and block of cheese. Open squares indicate that the organism was not detected by direct plating or cold enrichment. Numbers within open squares indicate the Listeria count (CFU per gram) in cheese as determined by direct plating. Solid squares indicate that the organism was found during 8 weeks of cold enrichment, using either TBC or TB. (Note: Only one block of cheese from batch 5 [block A] was available for Listeria analysis.) were held at 4°C for cold one cheese sample. Two

samples which were positive for the pathogen after cold enrichment in TB were negative following cold enrichment in TBC. In contrast to results with cold-pack cheese food (Table 3), it appears that recovery of L. monocytogenes from Colby cheese was slightly enhanced after cold enrichment in TB rather than TBC. However, since populations of L. monocytogenes were decreasing rapidly in our 5- to 8-month-old Colby cheese, it is possible that the organism could be present in one sample and absent from the other sample taken at the same time from the same block of cheese. Thus, considering that both media resulted in 33% false-negatives for Colby cheese and that TBC was superior to TB for recovery of L. monocytogenes by cold enrichment from cold-pack cheese food, it appears that TB and TBC are equally acceptable as cold-enrichment media. Furthermore, since equal or greater populations of L. monocytogenes were detected by direct plating of cheese samples diluted in TBC at 45°C rather than in TB at 25°C, we recommend diluting ground cheese 1:10 in TBC at 45°C and stomaching the sample for 3 min for detection of L. monocytogenes by direct plating followed by use of the same sample for cold-enrichment procedures. In conclusion, our results show that recovery of L. monocytogenes from cheese curd or ripened cheese can be enhanced by diluting the sample in CS rather than TB. Recovery was further increased by warming the diluent to 40 to 45°C. Although higher populations of L. monocytogenes were detected in blended rather than stomached cheese samples, it appears safer to homogenize Listeria-contam-

LITERATURE CITED 1. Deibel, K. F., and G. W. Banwart. 1982. Comparison of the Stomacher with other systems for breaking clumps and chains in the enumeration of bacteria. J. Food Prot. 45:898-902. 2. Fleming, D. W., S. L. Cochi, K. L. MacDonald, J. Brondum, P. S. Hayes, B. D. Plikaytis, M. B. Holmes, A. Audurier, C. V. Broome, and A. L. Reingold. 1985. Pasteurized milk as a vehicle of infection in an outbreak of listeriosis. N. Engl. J. Med. 312: 404-407. 3. Gray, M. L., H. J. Stafseth, F. Throp, Jr., L. B. Sholl, and W. F. Riley, Jr. 1948. A new technique for isolating Listerellae from the bovine brain. J. Bacteriol. 55:471-476. 4. James, S. M., S. L. Fannin, B. A. Agree, B. Hall, E. Parker, J. Vogt, G. Run, J. Williams, L. Lieb, C. Salminen, T. Prendergast, S. B. Werner, and J. Chin. 1985. Listeriosis outbreak associated with Mexican-style cheese-California. Morbid. Mortal. Weekly Rep. 34:357-359. 5. Kosikowski, F. V. 1977. Cheese and fermented milk foods, 2nd ed. Edwards Brothers, Ann Arbor, Mich. 6. Lee, W. H., and D. McClain. 1986. Improved Listeria monocytogenes selective agar. Appl. Environ. Microbiol. 52:1215-1217. 7. McBride, M. E., and K. F. Girard. 1960. A selective method for the isolation of Listeria monocytogenes from mixed bacterial population. J. Lab. Clin. Med. 55:153-157. 8. Nanni, M. 1975. Use of the stomacher in the dairy industry. Sci. Tec. Lattiero-Casearia 26:368-370. 9. Odegaard, B., R. Grelland, and S. D. Henricksen. 1952. A case of Listeria infection in man, transmitted from sheep. Acta Med. Scand. 67:231-238. 10. Olson, N. F. 1969. Ripened semi-soft cheeses. Chas. Pfizer and Co., Inc., New York. 11. Olson, N. F. 1979. Cheese, p. 39-77. In H. J. Peppler and D. Perlman (ed.), Microbial technology, vol. 2, 2nd ed. Academic Press, Inc., New York. 12. Ralovich, B. 1984. Listeriosis research-present situation and perspective. Akademiai Kiado, Budapest. 13. Richardson, G. H. (ed.). 1985. Standard methods for the examination of dairy products, 15th ed. American Public Health Association, Washington, D.C. 14. Ryser, E. T., and E. H. Marth. 1987. Behavior of Listeria monocytogenes during the manufacture and ripening of Cheddar cheese. J. Food Prot. 50:7-13. 15. Ryser, E. T., and E. H. Marth. 1987. Fate of Listeria monocytogenes during the manufacture and ripening of Camembert cheese. J. Food Prot. 50:372-378. 15a.Ryser, E. T., and E. H. Marth. 1988. Survival of Listeria monocytogenes in cold-pack cheese food during refrigerated storage. J. Food Protect. 51:615-621. 16. Ryser, E. T., E. H. Marth, and M. P. Doyle. 1985. Survival of Listeria monocytogenes during manufacture and storage of cottage cheese. J. Food Prot. 48:746-750, 753. 17. Seeliger, H. P. R. 1961. Listeriosis. Hafner Publishing Co., Inc.,

VOL. 54, 1988


New York. 18. Sharpe, A. N., and G. C. Harshman. 1976. Recovery of Clostridium perfringens, Staphylococcus aureus and molds from foods by the Stomacher: effect of fat content, surfactant concentration, and blending time. Can. Inst. Food Sci. Technol. J. 9:30-34. 19. Steel, R. G. D., and J. H. Torrie. 1983. Principles and procedures of statistics, a biometrical approach. McGraw-Hill Book Co., New York.

2649

20. Thrasher, S., and G. H. Richardson. 1980. Comparative study of the Stomacher and the Waring blender for homogenization of high-fat dairy foods. J. Food Prot. 43:763-764. 21. von Asperger, H., and E. Brandl. 1978. Zur Lebendkeimzahl in Frischkase. 1. Mitteilung. Methodische Untersuchungen. Arch. Lebensmittelhyg. 29:171-176. 22. Yousef, A. E., and E. H. Marth. 1988. Behavior of Listeria monocytogenes during the manufacture and storage of Colby cheese. J. Food Prot. 51:12-15.