The chemotactic behavior of Campylobacterjejuni was determined in the presence of different amino acids, carbohydrates, organic acids, and preparations and constituents of mucin and bile. L-Fucose was the only carbohydrate and L-aspartate, L-cysteine, L-glutamate, and L-serine were the only amino acids producing a chemotactic (positive) response. Several salts of organic acids, including pyruvate, succinate, fumarate, citrate, malate, and a-ketoglutarate, were also chemoattractants, as were bile (beef, chicken, and oxgall) and mucin (bovine gallbladder and hog gastric). Most constituents of bile tested individually were chemorepellents, but the mucin component was chemoattractant. The chemotactic behavior of C. jejuni toward L-fucose, a constituent of both bile and mucin, may be an important factor in the affinity of the organism for the gallbladder and intestinal tract.

Bacterial enumeration and histologic examination of organs and tissues of 8-day-old chicks 7 days after peroral inoculation with Campylobacter jejuni revealed that the organism colonized primarily the lower gastrointestinal tract. The principal sites of localization were the ceca, large intestine, and cloaca, where densely packed cells of C. jejuni were observed in mucus within crypts. Examination of C. jejuni-colonized crypts by transmission electron microscopy revealed that the campylobacters freely pervaded the lumina of crypts without attachment to crypt microvilli. Understanding the mechanism of colonization may lead to approaches that will reduce the incidence of C. jejuni carriage by poultry.

Thirty-one tongues from apparently normal, freshly slaughtered pigs were assayed for the presence of Yersinia enterocolitica by different enrichment and postenrichment techniques. Sixteen different isolates were recovered, including six of serotype 0:8, four of serotype 0:6,30, two of serotype 0:3 phage type IXb, and one each of serotypes 0:13,7, 0:18, and 0:46. One isolate was not typable. Cold enrichment in phosphate-buffered saline followed by treatment with dilute KOH or subsequent enrichment in modified Rappaport broth recovered 12 and 7 isolates, respectively. Four of the same isolates were recovered from the same tongues by both procedures. Cold enrichment without a selective postenrichment treatment recovered two isolates. Direct enrichment in modified Rappaport broth or modified selenite broth was ineffective in recovering yersiniae, as no isolates were obtained by either method. All of the serotype 0:8 isolates were virulent to mice, causing the death of adults after oral challenge. This is the first report that associates Y. enterocolitica serotype 0:8 with a natural reservoir.

An antiserum (WA-SAA) was produced which agglutinated specifically with mouse-virulent but not with avirulent strains of Yersinia enterocolitica. Expression of the antigenic determinant(s) reacting with WA-SAA was temperature dependent; for growth temperatures of 20 to 40°C, agglutination titers were lowest for cultures grown at 20°C and highest for cultures grown at 35 to 40°C. Addition of Ca2+ (2.5 to 10 mM) to the growth medium had little effect on the agglutination titer, and gel diffusion studies with monospecific anti-V serum indicated that V antigen was not likely to be the determinant reacting with WA-SAA. Immunohistological studies of Peyer's patches of mice infected with Y. entero(olitica WA revealed that the antigenic determinant(s) reacting with WA-SAA was expressed in vivo. The strong correlation of agglutination titer with mouse virulence and the expression in vivo of the antigenic determinant(s) reacting with WA-SAA suggest that the antigen(s) may be associated with the pathogenicity of Y. enterocolitica. Yersinia enterocolitica is a widely distributed organism which may be isolated from a variety of foods. Examples include milk (17, 30), fish (20), porcine tongues (12), chicken (23), lamb (15), beef (15), oysters (1, 25), shrimp (25), crab (25), and vegetables used in salads (1). Although the organism is often present in foods and the environment, a large number of the strains isolated from these sources are apparently avirulent (21, 28, 29). Hence, it would be useful to identify characteristics common among virulent Y. enterocolitica that may be used as determinants for developing methods to differentiate virulent from avirulent strains. Several established or presumed virulence determinants have been associated with Yersinia pestis. Included are the ability to produce V and W antigens, capsular or fraction 1 antigen, pesticin, coagulase, and fibrinolysin, and the capacity to absorb hemin (5). Carter et al. (11) have shown that mouse-virulent Y. enterocolitica WA produces V and W antigens that are immunologically identical to the V and W antigens of Y. pestis. These are the only virulence determinants known to be common to these two species; however, it is not known whether V and W antigens are common among all strains of mouse-virulent Y. enterocolitica. Interestingly, although not a determinant of virulence, susceptibility to pesticin, a bacteriocin produced by wild-type strains of Yersinia pestis (4), is also a characteristic common among strains of Y. enterocolitica that are lethal to mice (16). We report here a serological test that identi

A 1-to 3-day enrichment-KOH postenrichment procedure was evaluated and found to be as effective in recovering Yersinia enterocolitica from meats as a 14to 21-day cold enrichment procedure, with or without KOH postenrichment. The shortened procedure consists of enriching 1.0and 25-g samples of meat in phosphate-buffered saline (pH 7.2) at 25°C. After incubation (48 and 72 h for 1.0-g samples and 24 and 48 h for 25-g samples), 0.5-ml portions of enrichment culture were treated with 4.5 ml of 0.25% KOH-0.5% NaCI for 2 min and 0.5% KOH-0.5% NaCl for 15 s, and 0.1-ml portions of treated culture were plated onto MacConkey or CIN agars or both. The procedure effectively recovered 2 to 12 cells of a number of both mouse-virulent and avirulent strains per g of ground beef with aerobic plate counts of approximately 106 to 107 CFU/g. Similarly, the procedure isolated both likely virulent and avirulent strains from porcine tongues (aerobic plate counts of 105 to 107 CFU/g) naturally contaminated with Y.