The Autosow raised miniature swine as a model for assessing the effects of dietary soy trypsin inhibitor.

Toxicological effects of dietary soy trypsin inhibitor (TI) were assessed in male miniature swine, a model chosen for its similarities to human digestive physiology and anatomy. The TI preparation was extracted from defatted raw soy flour. From 1 through 5 weeks of age, piglets were automatically fed either a TI liquid diet [Autosow TI group (ASTI)] or a control liquid diet [Autosow control group (ASC)]. From 6 to 39 weeks of age, these animals received either swine chow and TI or swine chow and control article. The TI diets were formulated to contain a TI activity of approximately 500 mg TI/100 g dry matter. A sow control (SC) group suckled from birth to 6 weeks of age and then fed as the ASC group with swine chow plus control article from 6 to 39 weeks of age. The SC piglets grew faster than ASC piglets during postnatal weeks 1 and 2; however, the ASC piglets were significantly heavier than the SC piglets (P=0.001) at 6 weeks of age. Compared with the ASC group, TI caused a moderate decrease in feed consumption and a moderate but reversible decrease in growth from 2 to 5 weeks of age, but not thereafter. Some control and TI-fed Autosow-reared piglets had loose stools until 6 weeks of age; the effect was significantly greater in the TI-fed group. Otherwise, all swine were active and had normal appearance and behavior.

Rapid fluorogenic enumeration of Escherichia coli in selected, naturally contaminated high moisture foods.

An assay for the enzyme glucuronidase was used to determine the presence of Escherichia coli in selected, naturally contaminated high moisture foods. Raw pork sausage, ground turkey, and ground beef were inoculated into tubes containing the substrate 4-methylumbelliferyl beta-D-glucuronide (MUG) in lauryl tryptose (LT) medium. After incubation at 35 degrees C for 24 h, the inoculated LT-MUG tubes were examined under longwave ultraviolet light for the presence of a fluorogenic glucuronidase end product. A fluorescing tube indicated the presumptive presence of E. coli. The 10 day most probable number method of the AOAC and the LT-MUG procedure gave comparable recoveries of E. coli.

Glucuronidase assay in a rapid MPN determination for recovery of Escherichia coli from selected foods.

Glucuronidase is present in most strains of Escherichia coli but absent in most other enteric microorganisms; therefore, an assay for this enzyme is useful for determining the presence of the organism. The substrate 4-methylumbelliferyl beta-D-glucuronide (MUG) is incorporated into either lauryl tryptose (LT) broth or EC medium; the inoculated tubes are then incubated under specified conditions and examined under longwave UV light for the presence of a fluorogenic glucuronidase end product. When compared with the 10-day most probable number (MPN) procedure of AOAC, the LT-MUG and the EC-MUG tests required 24 and 96 h, respectively, and gave comparable mean log MPN values for samples of crabmeat, sunflower kernels, and walnut pieces. However, false-positive and false-negative reactions were observed with foods tested by both of these rapid methods. Overall, method sensitivity was not compromised by using the LT-MUG rather than the EC-MUG method. Incorporation of 25 micrograms MUG/mL into LT broth resulted in diminished fluorescence of positive reactions, whereas MUG concentrations of 50 and 100 micrograms/mL provided decisive fluorogenic reactions.

Recovery of Salmonella species from nonfat dry milk rehydrated under rapid and reduced pre-enrichment conditions: collaborative study.

A collaborative study was conducted to compare the relative efficiency of the AOAC rapid rehydration method with the reduced rehydration soak method for the recovery of Salmonella species from nonfat dry milk (NFDM). In the AOAC method, a 25 g sample of NFDM is rapidly rehydrated at a 1:9 sample/water ratio and mixed by swirling. After 60 min, the flask contents are adjusted to a pH of 6.8, and 0.45 mL of 1% aqueous brilliant green dye solution is added. The flasks are then incubated at 35 degrees C. In the soak method, a 25 g sample of NFDM is gently added to the sterile brilliant green (BG) water at a 1:9 sample/BG water ratio and allowed to soak undisturbed for 60 min at room temperature before incubation. Twelve collaborators analyzed 3 shipments of samples with the following results for the AOAC and soak methods: shipment 1-31 and 46 positive samples, respectively, with a 48% increase in detection by the soak method; shipment 3-45 and 66 positive samples, respectively, with a 47% increase in detection by the soak method; shipment 2--no significant difference in recovery of Salmonella species by the 2 methods. It is recommended that the official final action method for the detection of Salmonella species, 46.054-46.067, be revised to use the soak method for the analysis of nonfat dry milk.

Microbial hazards associated with bean sprouting.

The behaviour of microorganisms was studied in mung beans and alfalfa seeds before and after germination in modified, commercially available bean-sprouting kits. The microorganism were enumerated by the aerobic plate count (APC) and by total yeast and mold count procedures. Salmonella species were artificially inoculated into selected samples and were enumerated by the most probable number (MPN) method. After germination of the beans or seeds into mature sprouts, significant increases were noted in APCs and in MPN values of Salmonella species. Although counts of yeasts and molds did not increase significantly after germination, these samples show an increase in toxic Aspergillus flavus and potentially toxic Alternaria species. The presence of toxic Penicillium cyclopium molds also increase substantially in 5 samples of a single brand of mung beans. Analysis of selected sprout samples, however, showed no presence of aflatoxin.

Interlaboratory evaluation of the AOAC method and the A-1 procedure for recovery of fecal coliforms from foods.

An interlaboratory evaluation was made of the 96 h AOAC method and the 24 h A-1 procedure for the enumeration of fecal coliforms in samples of yellow corn meal, rye flour, mung beans, raw ground beef, and raw oyster homogenate. Results indicated that the efficiency of the A-1 procedure, measured in terms of recovery of fecal coliforms, and the reproducibility of that recovery were dependent on the particular food being analyzed. Accordingly, until its efficiency can be more fully demonstrated, the A-1 procedure is recommended only as a screening procedure for fecal coliforms in foods.

Pre-enrichment broths for recovery of Salmonella from milk chocolate and edible casein: collaborative study.

A collaborative study was conducted to compare the relative efficiency of nonfat dry milk with brilliant green dye (NFDM-BG and buffered peptone water (BPW) as pre-enrichment broths for recovery of Salmonella from milk chocolate. Lactose broth and modified lactose broth with added 1% NaHCO3 and brilliant green dye were compared as pre-enrichment broths for recovery of Salmonella from edible casein. Two sets of 8 samples each of milk chocolate, containing initial levels of Salmonella ranging from less than 0.03 to 43 organisms/g, were examined by 13 collaborators. Of 104 determinations, 102 (98.1%) and 100 (96.2%) using NFDM-BG and BPW, respectively, were in agreement with sample results of the control laboratory. Two sets of 7 samples each of edible casein, containing initial levels of Salmonella ranging from less than 0.03 to 93 organisms/g, were also examined by the 13 collaborators. Of 91 determinations, 87 (95.6%) and 88 (96.7%) using lactose broth and modified lactose broth, respectively, were in agreement with sample results of the control laboratory. For recovery of Salmonella, therefore, NFDM-BG pre-enrichment is recommended for milk chocolate, and lactose broth is recommended for casein. The proposed revision of official final action method 46.054-46.067 has been adopted official first action.

Comparative efficiency of brilliant green, bismuth sulfite, Salmonella-Shigella, hektoen enteric, and xylose lysine desoxycholate agars for the recovery of Salmonella from foods: collaborative study.

The relative efficiency of brilliant green (BG), bismuth sulfite (BS), Salmonella-Shigella (SS), xylose lysine desoxycholate (XLD), and Hektoen enteric (HE) agars for the recovery of Salmonella from 5 foods was collaboratively studied in 11 laboratories. The analytical efficiency of various paired combinations of the 5 agars was statistically compared according to 3 parameters: (1) productivity or recovery of Salmonella, (2) rate of enumeration of cultures that were false positive for Salmonella, and (3) rate of enumeration of false-negative reactions. In descending order of productivity, the sequential ranking was BS, XLD, HE, BG, and SS agars. In ascending order, the rates of false-positive reactions based on a statistical analysis of paired agar combinations was HE, BS, BG and XLD (tie), and SS agars. Analogously, in ascending order, the sequence of false-negative reaction rates was BS, XLD, HE, BG, and SS agars. The combination of BS, XLD, and HE agars is more efficient for recovery of Salmonella from foods than is the present official combination of BG, BS, and SS agars. The revision of official final action method 46.054 to replace the combination of BG, BS, and SS agars with a combination of BS, XLD, and HE agars has been adopted official first action.

Bacteriological survey of sixty health foods.

A bacteriological survey was performed on 1,960 food samples encompassing 60 types of health foods available in the Baltimore-Washington, D.C., metropolitan area. No consistent bacteriological distinction (aerobic plate counts, total coliform and fecal coliform most probable numbers) was observed between foods labeled as organic (raised on soil with compost or nonchemical fertilizer and without application of pesticides, fungicides, and herbicides) and their counterpart food types bearing no such label. Types and numbers of samples containing Salmonella were: sunflower seeds, 4; soy flour, 3; soy protein powder, 2; soy milk powder, 1; dried active yeast, 1; brewers' years, 1; rye flour, 1; brown rice, 1; and alfalfa seeds,1. The occurrence of this pathogen in three types of soybean products should warrant further investigation of soybean derivatives as potentially significant sources of Salmonella.

Relative productivity of five selective plating agars for the recovery of Salmonella from selected food types.

During a 3-year period, the relative productivity of brilliant green (BG), bismuth sulfite (BS), Salmonella-Shigella (SS), Hektoen enteric (HE), and xylose lysine, desoxycholate (XLD) agars for recovering Salmonella from 9 food types was determined. Following pre-enrichment, selective enrichment of food samples in tetrathionate broth followed by streaking to BS agar was the single most productive selective enrichment broth-agar combination for recovery of Salmonella in 5 of these food types. A study of the performance of these 5 agars used individually and in various combinations, showed that none of the 5 agars used individually nor any of the possible paired combinations of these agars could be used to satisfactorily detect Salmonella in the 9 food types. The use of all 5 agars was not necessary because one combination of 4 agars (BG, BS, HE, and XLD) recoverd 100% of the Salmonella isolates, as compared with the number of Salmonella isolates recovered by the 5-agar combination, in each food category. This particular 4-agar combination, along with two 3-agar combinations (BG, BS, and XLD agars, and BS, HE, and XLD agars), were each able to recover more Salmonella isolates, than the combination of BG, BS, and SS agars, the combination currently recommended by the AOAC. Finally, the relative costs of using these agars, singly and in various combinations, were determined.

Comparative accuracy of five biochemical systems for identifying Salmonella and related foodborne bacteria: collaborative study.

The comparative accuracy of 4 biochemical diagnostic kits (API, Enterotube, Minitek, and Pathotec) and the conventional (AOAC) tube system for identifying primarily Salmonella and other enteric isolates was collaboratively studied. Each of 11 participating analysts received 40 foodborne isolates (25 Salmonella and 15 non-Salmonella cultures), representing a total of 440 cultures examined by each identification system. In decreasing order of accuracy, the overall number of correctly identified cultures with each of the systems was as follows: AOAC, 423 (96.1%), Minitek, 403 (91.6%), Enterotube, 395 (89.8%), API, 394 (89.5%), and Pathotec, 373 (84.8%). A cost analysis showed that all 4 diagnostic kit systems were less expensive than the conventional AOAC tube system for a single culture identification. Three of the diagnostic kits have been adopted as official first action as alternatives to the AOAC biochemical tube system for presumptive generic identification of foodborne Salmonella and for screening and eliminating non-Salmonella isolates. Routine incorporation of any one of the 3 diagnostic kits, however, should be preceded by the demonstration in the analyst's own laboratory of adequate correlation between the kit and the AOAC system.

Usefulness of the stomacher in a microbiological regulatory laboratory.

The relative efficiency of the Waring blender, the Stomacher 400, and the Stomacher 3500 for preparing food samples for microbiological analysis was studied. Comparative aerobic plate count (APC) values were determined on 671 samples, representing 30 categories of foods. Of the 26 categories of nonfatty foods, the blender gave significantly higher geometric mean APC values than those given by the Stomacher 400 and the Stomacher 3500 in 65 and 69 percent of the categories, respectively. In a comparison of the two Stomacher models, the Stomacher 400 gave significantly higher geometric mean APC values than these given by the Stomacher 3500 in 73 percent of the food categories. Addition of Tween 80 to four categories of fatty foods at concentrations of 0.5, 1.0, and 2.0 percent did not raise the APC values given by either model of stomacher to the levels given by the Waring blender. Overall, the efficiency of both models of Stomacher, relative to the blender and to each other, was specific and depended upon the particular food being analyzed.

Comparison of methods for the isolation of Salmonella from imported frog legs.

Four methods of sample preparation were compared for their relative efficiency in recovering Salmonella from imported frog legs. No significant difference (P greater than 0.10) was observed in the efficiency of submersion, blending, and stomaching methods, but rinsing recovered significantly fewer (P less than 0.01) Salmonella-positive frog legs than the other three methods. No significant difference (P less than 0.25) was observed in the number of positive frog legs recovered by selenite cystine or tetrathionate broth. Salmonella-Shigella agar, when streaked from either of these broths, gave significantly fewer (P less than 0.01) Salmonella-positive frog legs than brilliant green, bismuth sulfite, xylose lysine deoxycholate, and Hektoen-enteric agars. Use selective agars, resulted in detection of an additional 11 and 6 Salmonella-positive frog legs, respectively. A variety of serotypes, mostly uncommon, was recovered. One Salmonella serotype (6,14,24:r,i:e,n,z15), possessing a heretofore unreported antigenic formula, was isolated.

Culture method for detection of Salmonella in dried active yeast: collaborative study.

A method for detecting Salmonella in dried active yeast was subjected to collaborative study. This method employs trypticase soy broth as the pre-enrichment medium, a sample-to-broth ratio of 1:10, and subsequent transfers to lauryl sulfate tryptose broth and tetrathionate before streaking onto selective agars. Each collaborating analyst received ten 25 g samples of dried active yeast. Duplicate 25 g samples were each inoculated with Salmonella oranienburg at a low level (28 cells) and a high level (107 cells). Similarly, duplicate 25 g samples were each inoculated with S. senftenberg at a low level (30 cells) and a high level (114 cells). The remaining 2 of 10 samples were not inoculated. Results from 12 of 13 collaborators were evaluated. Only 2 (8.2%) of the 24 low level S. oranienburg samples were reported incorrectly as negative. Twelve of the analysts detected S. senftenberg at both levels and S. oranienburg at the high level in the inoculated samples. Results from 12 collaborators used in the final evaluation show that 117 of 119 (98.3%) collaborative determinations are in agreement. The official final action method for the detection and identification of Salmonella, 46.013-46.026, has been revised official first action to include applicability to dried active yeast.

The moroccan food snail, Helix aspersa, as a source of Salmonella.

A total of 270 samples, nine lots of 30 samples each, of imported Moroccan food snails was examined for the presence of Salmonella. Eighty-four samples (an overall incidence of 31.11%) and all nine lots contained Salmonella. No significant difference (P greater than 0.25) in the number of positive samples was observed by using either selenite cystine both or tetrathionate broth when the samples had been pre-enriched in lactose broth. When used as direct selective enrichments with samples not pre-enriched in lactose broth, tetrathionate broth was significantly (P less than 0.05) more productive than selenite cystine broth. The overall detection of Salmonella-positive samples by direct enrichment was significantly greater (P less than 0.001) than by pre-enrichment. A variety of uncommon serotypes occurrence and incidence, and the concomitant human health potential, of Salmonella in one species of live, imported food snails.