International Journal of Food Microbiology, 14 ( 1991 ) 19-26 © 1991 Elsevier Science Publishers B.V. All rights reserved 0168-1605/91/$03.50
IO
FOOD 00426
Production of staphylococcal enterotoxin D in foods by low-enterotoxin-producing staphylococci J o s e L u i z P e r e i r a 1, S o n i a P. S a l z b e r g 1 a n d M e r l i n S. B e r g d o l l 2 1 FEA UNICAMP C.P. 6121, Carapinas, Brasil, and 2 Food Research Institute, University o f Wisconsin, Madison, Wisconsin, U.S.A. (Received 5 March 1991; accepted 22 May 1991)
The goal of this investigation was to determine whether staphylococcal strains producing enterotoxins at nanogram levels per milliliter in laboratory medium, not detectable by gel diffusion methods, could produce sufficient enterotoxin in foods to result in food poisoning. Three Iow-enterotoxin D (SED)-producing strains were selected for this research because this enterotoxin is produced in smaller amounts than the other enterotoxins. The foods used were cream pie and cooked ham. divided into two portions, sterile and non-sterile. Each portion was inoculated with known concentrations of the staphylococcal strains under study and incubated for 48 h at 25, 30, and 37 ° C. Samples were taken after 24 and 48 h. Enterotoxin was detectable in both sterilized and unsterilized cream and ham after 24 h at 37 ° C with an inoculum of 103/g. Some strains produced detectable amounts of enterotoxin in the sterilized foods after 24 h at 30 ° C and some produced detectable amounts of enterotoxin in the sterilized foods after 24 h at 25°C with inocula of 104/g. It can be concluded that staphylococcal strains producing enterotoxin at ng/ml levels in laboratory medium, not detectable by gel diffusion methods, can produce sufficient enterotoxin (rig/g) in foods to cause food poisoning. Key words: Staphylococcus; Enterotoxin producer, ng/ml level
Introduction
Gel diffusion methods (Casman et ah, 1969; Robbins et al., 1974) are the standard procedures for the detection of enterotoxin production by staphylococci, along with production methods that produce relatively large amounts of toxin, such as the membrane-over-agar and sac culture methods (Robbins et al., 1974). The gel diffusion methods, such as the optimum sensitivity plate (OSP) method (Robbins et al., 1974), are relatively easy to perform and do not require special reagents other than the specific antisera and the partially purified toxin. The sensitivity of 0.1-0.5 /zg/ml had been considered sufficient to identify enterotoxin-producing strains. The advent of methods with sensitivities of 1 ng/ml or less revealed that some strains produced only ng/ml quantities of enterotoxin (Igarashi et al., 1986). Using the enzyme-linked immunosorbent assay (ELISA), it was shown that 26 out Correspondence address: M.S. Bergdoll, Food Research Institute, University of Wisconsin, Madison, Wl 53706, U.S.A.
20 of 110 strains that were negative for production of the identified enterotoxins by the gel diffusion methods produced low levels of indentified enterotoxins (Kokan and Bergdoll, 1987). Previously these strains were considered producers of unidentified enterotoxins because they produced an emetic reaction in monkeys. Some of these strains were isolated from foods that had been implicated in food-poisoning outbreaks (Kokan and Bergdoll, 1987). Although strains may produce g g / m l of enterotoxin in laboratory medium, usually only n g / g of enterotoxin is produced in foods. The question arose as to whether this low level of production in laboratory medium could account for the food poisoning or if it was due to the production of an additional unidentified enterotoxin. The amount of enterotoxin present in foods implicated in food-poisoning outbreaks varies from over 10 n g / g to less than 0.5 n g / g (Evenson et al., 1988). Hence, if these strains produced at least 1 n g / g of enterotoxin in food, this enterotoxin would be considered the source of the outbreak. This is important because it would indicate that gel diffusion methods are inadequate to identify some strains that produce enterotoxin. This research was undertaken to determine whether strains that produce n g / m l quantities of enterotoxin in laboratory medium could produce detectable amounts of enterotoxin in foods.
Materials and Methods
Bacterial cultures Four low-enterotoxin D (SED)-producing Staphylococcus aureus strains, FRI-396 and FRI-427 (both isolated from fish), FRI-464 (human isolate), FRI-542 (isolated from cheese) and a standard SED-producing strain FRI-472 were used.
SED production in laboratory media Shake flasks. 200 ml of 3% tryptone + 1% yeast extract (TYE) in 1-1 Erlenmeyer flasks was inoculated with 105 cells, six flasks for each strain (one for each time period); incubation was at 3 7 ° C on a rotary shaker at 250 rpm. 50-ml samples, removed at 8, 12, 16, 20, 24, and 48 h, were centrifuged, dialysed, concentrated to 50% of the original volume with polyethylene glycol 6000, iyophilized, and redissolved in 1 ml of T N T buffer (0.01 Tris, 0.5 M NaCI, 0.1% Tween 20, 0.1% sodium azide, pH 7.5). Sac culture. 70 ml of T Y E medium was inserted in cellulose sausage casing, 3.5 cm flat width (Union Carbide Corp., Chicago, IL); the tube was placed in a 300-ml Erlenmeyer flask; 19 ml of phosphate-saline buffer (PBS) was added. Six flasks were prepared for each strain (one for each time period) and autoclaved. The PBS was inoculated with 1 ml (104 cfu) of each strain. The flasks were incubated on a rotary shaker at 250 rpm at 37 ° C. 10-ml samples were taken from the inoculated buffer culture at 8, 12, 16, 20, 24, and 48 h and centrifuged.
SED production in foods Cream pie filling. The cream filling was prepared as follows: 400 g sucrose, 100 g starch, 8 ml of 1% Tintanil (a yellow food dye), 1 egg yolk, and 2 1 of skim milk
21 were homogenized and heated to the desired consistency; 50-g portions were placed in 24 100-ml beakers for each strain; 12 were inoculated directly, six with 103 c f u / g and six with 104 cfu/g; the other 12 were sterilized at 121 ° C for 15 min and inoculated, six with 103 c f u / g and six with 104 cfu/g. Incubation was at ~ , 30, and 3 7 ° C for 24 and 48 h. One sterilized and one unsterilized cream, both uninoculated, were used as controls; incubation was at 37 °C for 24 and 48 h. Baked ham. 50-g slices of baked ham were placed in 24 petri dishes; 12 were sterilized at 121°C for 15 min and 12 were left unsterilized. Six samples each of the sterilized and unsterilized hamwere inoculated with 103 cfu/g, one sample for each of the six strains and the other six samples of each type were inoculated with 104 cfu/g. One sterilized and one unsterilized sample were used as controls, with incubation at 37 o C for 24 and 48 h.
Staphylococcal and mesophilic counts Mesophilic counts of samples diluted and homogenized in PBS were made by the standard plate count method using plate count agar (Difco); staphylococcal counts were made using Baird-Parker agar (Difco).
SED determination Optimum sensitivity plate (OSP) method.
The OSP method is a modification of the Ouchterlony gel diffusion plate method, with a sensitivity of about 0 . 5 / z g / m l (Robbins et al., 1974). Enzyme-linked immunosorbent assay (ELISA). The Bomelli ball ELISA kit (Dr. Bomelli AG, Stationstrasse 12, CH-3097 Liebefeld, Bern, Switzerland) was used for SED analysis. The sensitivity of this method is 0.1-1.0 ng/ml. For assay of foods, 50 g of food was homogenized with an equal volume of T N T buffer and centrifuged. The supernatant fluid was analysed for SED.
Results
SED production in laboratory medium For shake flasks all cultures were at the end of the exponential phase after 16 h incubation (data not shown). Strain FRI-472 produced about four times as much SED as the low-producing strains. SED was detectable by ELISA in the concentrated supernatant fluids after 8 h incubation for strains FRI-396. 427, and 472 when the cell counts were greater than 106 cfu/ml (Table I). SED production by strain FRI-464 was detectable after 12 h incubation when the cell count was 4.3 x 107 cfu/ml. For strain FRI-542, SED was detectable after 16 h (Data for cell counts not included). For sac cultures only strain FRI-472 produced detectable SED by the OSP method after 8 h of incubation (Table II). SED was detectable by OSP in the concentrated extracts from strain FRI-396 and 427 after 20 h incubation. SED production by strains FRI-464 and 542 was not detectable by OSP. SED produc-
22 TABLE I Production of SED n g / m l by shake-flask and sac culture methods a Strain
Method
FRI-396
1b 2 c 1 2 1 2 1 2 1 2
FRI-427 FRI-464 FRI-542 FRI-472
a b c d
Time of incubation (h) 8
12
16
20
24
48
7 68 10 93 nd d 3 nd nd 29 210
12 118 16 154 6 64 nd nd 60 405
37 202 31 218 15 165 6 63 112 2150
72 506 51 375 30 286 13 87 320 3280
80 762 53 496 33 352 17 106 326 3692
89 789 55 530 35 377 20 114 340 3800
SED assayed by ELISA. Shake-flask method. Sac culture method. Not detectable.
tion was detectable by ELISA after 8 h incubation for all strains except strain FRI-542; production by this strain was detectable only after 16 h incubation.
SED production in cream pie filling and baked ham SED was produced by all strains in the sterilized cream filling at all three temperatures with an inoculum of 104 cfu/g (Table liD. SED was not produced by
TABLE II Detection of SED production ( n g / m l ) by gel diffusion and ELISA a Strain
FRI-396 FRI-427 FRI-464 FR1-542 FRI-472
a b c d
Method
1b 2 c 1 2 1 2 1 2 1 2
Time of incubation (h) 8
12
16
20
24
48
68 nd d 93 nd 3 nd nd nd 210 < 2000
118 nd 154 nd 64 nd nd nd 405 < 2000
202 nd 218 nd 165 nd 63 nd 2150 < 2000
506 < 2000 375 < 2000 286 nd 87 nd 3280 > 2000
762 < 2000 496 < 2000 352 nd 106 nd 3690 > 2000
789
IO
FOOD 00426
Production of staphylococcal enterotoxin D in foods by low-enterotoxin-producing staphylococci J o s e L u i z P e r e i r a 1, S o n i a P. S a l z b e r g 1 a n d M e r l i n S. B e r g d o l l 2 1 FEA UNICAMP C.P. 6121, Carapinas, Brasil, and 2 Food Research Institute, University o f Wisconsin, Madison, Wisconsin, U.S.A. (Received 5 March 1991; accepted 22 May 1991)
The goal of this investigation was to determine whether staphylococcal strains producing enterotoxins at nanogram levels per milliliter in laboratory medium, not detectable by gel diffusion methods, could produce sufficient enterotoxin in foods to result in food poisoning. Three Iow-enterotoxin D (SED)-producing strains were selected for this research because this enterotoxin is produced in smaller amounts than the other enterotoxins. The foods used were cream pie and cooked ham. divided into two portions, sterile and non-sterile. Each portion was inoculated with known concentrations of the staphylococcal strains under study and incubated for 48 h at 25, 30, and 37 ° C. Samples were taken after 24 and 48 h. Enterotoxin was detectable in both sterilized and unsterilized cream and ham after 24 h at 37 ° C with an inoculum of 103/g. Some strains produced detectable amounts of enterotoxin in the sterilized foods after 24 h at 30 ° C and some produced detectable amounts of enterotoxin in the sterilized foods after 24 h at 25°C with inocula of 104/g. It can be concluded that staphylococcal strains producing enterotoxin at ng/ml levels in laboratory medium, not detectable by gel diffusion methods, can produce sufficient enterotoxin (rig/g) in foods to cause food poisoning. Key words: Staphylococcus; Enterotoxin producer, ng/ml level
Introduction
Gel diffusion methods (Casman et ah, 1969; Robbins et al., 1974) are the standard procedures for the detection of enterotoxin production by staphylococci, along with production methods that produce relatively large amounts of toxin, such as the membrane-over-agar and sac culture methods (Robbins et al., 1974). The gel diffusion methods, such as the optimum sensitivity plate (OSP) method (Robbins et al., 1974), are relatively easy to perform and do not require special reagents other than the specific antisera and the partially purified toxin. The sensitivity of 0.1-0.5 /zg/ml had been considered sufficient to identify enterotoxin-producing strains. The advent of methods with sensitivities of 1 ng/ml or less revealed that some strains produced only ng/ml quantities of enterotoxin (Igarashi et al., 1986). Using the enzyme-linked immunosorbent assay (ELISA), it was shown that 26 out Correspondence address: M.S. Bergdoll, Food Research Institute, University of Wisconsin, Madison, Wl 53706, U.S.A.
20 of 110 strains that were negative for production of the identified enterotoxins by the gel diffusion methods produced low levels of indentified enterotoxins (Kokan and Bergdoll, 1987). Previously these strains were considered producers of unidentified enterotoxins because they produced an emetic reaction in monkeys. Some of these strains were isolated from foods that had been implicated in food-poisoning outbreaks (Kokan and Bergdoll, 1987). Although strains may produce g g / m l of enterotoxin in laboratory medium, usually only n g / g of enterotoxin is produced in foods. The question arose as to whether this low level of production in laboratory medium could account for the food poisoning or if it was due to the production of an additional unidentified enterotoxin. The amount of enterotoxin present in foods implicated in food-poisoning outbreaks varies from over 10 n g / g to less than 0.5 n g / g (Evenson et al., 1988). Hence, if these strains produced at least 1 n g / g of enterotoxin in food, this enterotoxin would be considered the source of the outbreak. This is important because it would indicate that gel diffusion methods are inadequate to identify some strains that produce enterotoxin. This research was undertaken to determine whether strains that produce n g / m l quantities of enterotoxin in laboratory medium could produce detectable amounts of enterotoxin in foods.
Materials and Methods
Bacterial cultures Four low-enterotoxin D (SED)-producing Staphylococcus aureus strains, FRI-396 and FRI-427 (both isolated from fish), FRI-464 (human isolate), FRI-542 (isolated from cheese) and a standard SED-producing strain FRI-472 were used.
SED production in laboratory media Shake flasks. 200 ml of 3% tryptone + 1% yeast extract (TYE) in 1-1 Erlenmeyer flasks was inoculated with 105 cells, six flasks for each strain (one for each time period); incubation was at 3 7 ° C on a rotary shaker at 250 rpm. 50-ml samples, removed at 8, 12, 16, 20, 24, and 48 h, were centrifuged, dialysed, concentrated to 50% of the original volume with polyethylene glycol 6000, iyophilized, and redissolved in 1 ml of T N T buffer (0.01 Tris, 0.5 M NaCI, 0.1% Tween 20, 0.1% sodium azide, pH 7.5). Sac culture. 70 ml of T Y E medium was inserted in cellulose sausage casing, 3.5 cm flat width (Union Carbide Corp., Chicago, IL); the tube was placed in a 300-ml Erlenmeyer flask; 19 ml of phosphate-saline buffer (PBS) was added. Six flasks were prepared for each strain (one for each time period) and autoclaved. The PBS was inoculated with 1 ml (104 cfu) of each strain. The flasks were incubated on a rotary shaker at 250 rpm at 37 ° C. 10-ml samples were taken from the inoculated buffer culture at 8, 12, 16, 20, 24, and 48 h and centrifuged.
SED production in foods Cream pie filling. The cream filling was prepared as follows: 400 g sucrose, 100 g starch, 8 ml of 1% Tintanil (a yellow food dye), 1 egg yolk, and 2 1 of skim milk
21 were homogenized and heated to the desired consistency; 50-g portions were placed in 24 100-ml beakers for each strain; 12 were inoculated directly, six with 103 c f u / g and six with 104 cfu/g; the other 12 were sterilized at 121 ° C for 15 min and inoculated, six with 103 c f u / g and six with 104 cfu/g. Incubation was at ~ , 30, and 3 7 ° C for 24 and 48 h. One sterilized and one unsterilized cream, both uninoculated, were used as controls; incubation was at 37 °C for 24 and 48 h. Baked ham. 50-g slices of baked ham were placed in 24 petri dishes; 12 were sterilized at 121°C for 15 min and 12 were left unsterilized. Six samples each of the sterilized and unsterilized hamwere inoculated with 103 cfu/g, one sample for each of the six strains and the other six samples of each type were inoculated with 104 cfu/g. One sterilized and one unsterilized sample were used as controls, with incubation at 37 o C for 24 and 48 h.
Staphylococcal and mesophilic counts Mesophilic counts of samples diluted and homogenized in PBS were made by the standard plate count method using plate count agar (Difco); staphylococcal counts were made using Baird-Parker agar (Difco).
SED determination Optimum sensitivity plate (OSP) method.
The OSP method is a modification of the Ouchterlony gel diffusion plate method, with a sensitivity of about 0 . 5 / z g / m l (Robbins et al., 1974). Enzyme-linked immunosorbent assay (ELISA). The Bomelli ball ELISA kit (Dr. Bomelli AG, Stationstrasse 12, CH-3097 Liebefeld, Bern, Switzerland) was used for SED analysis. The sensitivity of this method is 0.1-1.0 ng/ml. For assay of foods, 50 g of food was homogenized with an equal volume of T N T buffer and centrifuged. The supernatant fluid was analysed for SED.
Results
SED production in laboratory medium For shake flasks all cultures were at the end of the exponential phase after 16 h incubation (data not shown). Strain FRI-472 produced about four times as much SED as the low-producing strains. SED was detectable by ELISA in the concentrated supernatant fluids after 8 h incubation for strains FRI-396. 427, and 472 when the cell counts were greater than 106 cfu/ml (Table I). SED production by strain FRI-464 was detectable after 12 h incubation when the cell count was 4.3 x 107 cfu/ml. For strain FRI-542, SED was detectable after 16 h (Data for cell counts not included). For sac cultures only strain FRI-472 produced detectable SED by the OSP method after 8 h of incubation (Table II). SED was detectable by OSP in the concentrated extracts from strain FRI-396 and 427 after 20 h incubation. SED production by strains FRI-464 and 542 was not detectable by OSP. SED produc-
22 TABLE I Production of SED n g / m l by shake-flask and sac culture methods a Strain
Method
FRI-396
1b 2 c 1 2 1 2 1 2 1 2
FRI-427 FRI-464 FRI-542 FRI-472
a b c d
Time of incubation (h) 8
12
16
20
24
48
7 68 10 93 nd d 3 nd nd 29 210
12 118 16 154 6 64 nd nd 60 405
37 202 31 218 15 165 6 63 112 2150
72 506 51 375 30 286 13 87 320 3280
80 762 53 496 33 352 17 106 326 3692
89 789 55 530 35 377 20 114 340 3800
SED assayed by ELISA. Shake-flask method. Sac culture method. Not detectable.
tion was detectable by ELISA after 8 h incubation for all strains except strain FRI-542; production by this strain was detectable only after 16 h incubation.
SED production in cream pie filling and baked ham SED was produced by all strains in the sterilized cream filling at all three temperatures with an inoculum of 104 cfu/g (Table liD. SED was not produced by
TABLE II Detection of SED production ( n g / m l ) by gel diffusion and ELISA a Strain
FRI-396 FRI-427 FRI-464 FR1-542 FRI-472
a b c d
Method
1b 2 c 1 2 1 2 1 2 1 2
Time of incubation (h) 8
12
16
20
24
48
68 nd d 93 nd 3 nd nd nd 210 < 2000
118 nd 154 nd 64 nd nd nd 405 < 2000
202 nd 218 nd 165 nd 63 nd 2150 < 2000
506 < 2000 375 < 2000 286 nd 87 nd 3280 > 2000
762 < 2000 496 < 2000 352 nd 106 nd 3690 > 2000
789
