JOURNAL OF CLINICAL MICROBIOLOGY, Dec. 1992, p. 3058-3064 0095-1137/92/123058-07$02.00/0 Copyright X 1992, American Society for Microbiology
Vol. 30, No. 12
Comparison of Epidemiological Markers of Salmonella Strains Isolated from Different Sources in Spain JUAN J. BORREGO,1* DOLORES CASTRO,1 MANUEL JIMENEZ-NOTARIO,2 ANTONIO LUQUE,l EDUARDO MARTINEZ-MANZANARES,1 CARMEN RODRIGUEZ-AVIAL,3 AND JUAN J. PICAZO3 Departamento de Microbiologia, Universidad de Mdlaga, Campus Universitario Teatinos, 29071-Mdlaga,1 Centro de Salud de Fuengirola, Servicio Andaluz de Salud, Fuengirola (Mdlaga), 2 and Servicio de Microbiologia, Hospital Clinico Universitario San Carlos, Plaza Cristo Rey sin, 28040-Madrid,3 Spain Received 4 April 1992/Accepted 1 September 1992
A comparative study of the phage types, antimicrobial susceptibility patterns, and plasmid profiles of 171 strains of Salmonela isolated from food, epidemic outbreaks, and water-contaminated environments as well as sporadic human isolates was carried out to determine the most adequate marker in epidemiological investigations. Typing based on the plasmid profiles appears to be the most effective method for grouping strains with the same serotype obtained from a single outbreak and from environmental sources. However, none of the three markers tested allow us total discrimination and identification of related strains from a common source for epidemiological tracing. Therefore, the combined use of the three methods is necessary for determining whether common source isolates are related.
Salmonella infections appear to be one of the most typical examples of an enteric disease that is transmitted from animals to humans. The transmission occurs both through food, such as meat, dairy products, and egg by-products (16, 18), and by contact between animals and humans directly through the fecal-oral route (26) or indirectly via fecally contaminated waters (22). More than 2,000 serovars of Salmonella have been described, and all are considered to be potentially pathogenic for animals, including humans. Salmonellosis in humans can produce symptoms ranging in severity from intestinal disturbances to death (44). Infections caused by Salmonella strains, mainly S. enteritidis, have increased over the last few years (36). Bacterial epidemic strains have traditionally been defined on the basis of phenotypic properties of the isolates, such as serotype and biotype. However, these determinations have not always been useful in epidemiological tracing of the epidemic strains (14, 17, 25, 47). To evaluate the epidemiological relationship among isolates from a common source, several different subtyping methods have been applied, such as serotyping (39), biotyping and colicine typing (4, 19), phage typing (46), antimicrobial resistance patterns (12, 13, 42), and plasmid analysis (8, 31). Several studies have demonstrated that plasmid profile analysis of the Salmonella isolates could be a reliable epidemiological tool for the differentiation of epidemic and nonepidemic strains from outbreaks (9, 14, 35, 43) or to elucidate the epidemiology of these food-borne pathogens (41). However, only a few studies of the application of this marker to environmental, food, and animal isolates have been carried out (7, 29, 34). To study the reliability of the subtyping methods as useful epidemiological markers, we compared Salmonella strains isolated from contaminated food, outbreaks, sporadic human cases, and fecally contaminated waters by use of their phage types (PTs), plasmid profiles (PPs), and antimicrobial resistance patterns (APs). *
MATERIALS AND METHODS Bacterial strains. Salmonella strains were selected and separated into the following four groups according to their sources and epidemiological significance: (i) isolates from contaminated food (n = 26); (ii) isolates from different outbreaks in various locations of the provinces of Malaga and Madrid (Spain) and previously described by Castro et al. (10) (n = 34); (iii) sporadic human Salmonella isolates (n = 38); and (iv) isolates from contaminated natural fresh and marine waters (Guadalhorce River and Mediterranean Sea, respectively) (n = 72). All isolates were biochemically confirmed by using the API 20E system (API Systems S.A., Montalieu-Vercieu, France) and were serologically characterized by the slide agglutination test with Salmonella polyvalent 0 and group antisera (Difco Laboratories, Detroit, Mich.) as well as by the tube agglutination test with H antisera (Difco). All strains were sent to the National Reference Center of Salmonella (Majadahonda, Madrid, Spain) for confirmation of serotyping results. Antimicrobial susceptibility testing. The disk diffusion method (6) on Mueller-Hinton agar (Difco) was used to test the resistance of the isolates to the following antimicrobial agents (supplied by Biomerieux, Madrid, Spain): colistin (Cl), 10 ,ug; nalidixic acid (Na), 30 ,ug; gentamicin (Gm), 10 ,ug; streptomycin (Sm), 10 ,ug; kanamycin (Km), 30 p,g; tetracycline (Tc), 30 ,ug; chloramphenicol (Cm), 30 p,g; ampicillin (Ap), 10 ,ug; carbenicillin (Cb), 100 ,Lg; cephalothin (Cf), 30 ,ug; trimethoprim-sulfamethoxazole (SXT), 25 ,ug; tobramycin (Tm), 10 ,ug; and neomycin (Nm), 30 ,ug. Phage typing. PTs were determined by the use of 25 typing phages belonging to a new phage typing system developed by Castro et al. (10). PTs were established and identified by the patterns of reaction to the typing phages described previously (10). PP analysis. Plasmid DNA was extracted from lysed Salmonella isolates by a modification of the technique described by Kado and Liu (23) and performed by Toranzo et al. (49). Extracted plasmid DNA was electrophoresed for 2 h at 150 mA on a 0.7% horizontal agarose gel (Sigma Chemical Co., St. Louis, Mo.) in Tris-acetate buffer (40 mM Tris base, 2 mM disodium EDTA; adjusted to pH 7.9 with glacial acetic
Corresponding author. 3058
EPIDEMIOLOGICAL MARKERS OF SALMONELLA STRAINS
VOL. 30, 1992
TABLE 1. Serotype distribution and numbers of PTs, antimicrobial resistance patterns, and plasmid profiles of the Salmonella isolates according to their origins Source and serotype
No. of H isolates islts PTs
No.Pof:
P
APs
PPs
Food S. entenitidis S. virchow S. typhimuium S. typhi S. anatum S. infantis S. newport Total
16 3 2 2 1 1 1 26
15 3 2 2 1 1 1 25
11 3 2 1 1 1 1 20
14 3 2 1 1 1 1 23
Outbreak no. 1, S. enteritidis Outbreak no. 2, S. virchow Outbreak no. 3, S. enteritidis Outbreak no. 9, S. enteritidis Outbreak no. 11, S. enteritidis Outbreak no. 12, S. enteritidis Outbreak total
13 1 9 6 3 2 34
2 1 6 1 3 1 14
4 1 5 6 3 1 20
8 1 2 2 2 1 16
22 15 1 1 39
8 10 1 1 20
12 8 1 1 22
5 8 1 1 15
17 13 9 7 7 3 2 2 1 11 72
12 6 6 5 6 3 2 2 1 9 52
12 9 5 4 4 3 2 2 1 8 50
9 3 3 6 3 2 2 1 1 5 35
Sporadic strains S. entenitidis S. typhimurium S. virchow S. infantis Total
Environmental S. typhimurium S. enteritidis S. london S. blockley S. ohio S. weltevreden S. infantis S. thompson S. bovismorbificans Self-agglutinable Total
acid). Then, the gels were stained with ethidium bromide solution (0.5 ,ug/ml) and photographed under a UV transilluminator (LKB, Pharmacia, Barcelona, Spain) with Plus-XPan film and a 23 A Wratten filter. The approximate molecular masses of the plasmids (in megadaltons) were determined by comparison with plasmids with known molecular masses, e.g., R40a (96 MDa) and plasmids from Escherichia coli V517 ranging from 35.8 to 1.4 MDa. The nonparametric Friedman test, chi-square test, and t test (one tail) were used for statistical analysis by using StatView 512 Plus software and a Macintosh SE computer. RESULTS Table 1 summarizes the serotype distribution and the number of PTs, APs, and PPs of the four groups of Salmonella isolates selected (food, outbreak, sporadic, and environmental strains). The Salmonella serotypes most frequently isolated were S. enteritidis (48.8%) and S. typhimurium (20%), although only S. enteritidis was repeatedly isolated from all the sample groups. Food isolates presented a wide variety of PTs, APs, and PPs, and no epidemiological relationship was observed among Salmonella isolates (chi-square = 3.71; P > 0.95).
3059
Similar results were obtained with the environmental strains, which presented a high number of plasmidless strains. On the other hand, the isolates from outbreaks possessed common serotypes, PTs, and PPs but a great variety of APs. Numerous PPs were found in the 171 isolates. Many of them were shown to harbor large plasmids (from 82.8 to 110 MDa) that have generally been related to virulence properties (25, 28, 33). The large plasmids were not used for pattern comparison of isolates because of their proven instability on subculture (24, 38). Plasmids under 20 MDa were used for comparison of PPs. All the Salmonella isolates were grouped into 29 PPs. Patterns containing from one to five plasmids with low molecular masses were found in 72 isolates. The most frequently detected plasmid groups were D5 (two plasmids of 2.0 and 2.2 MDa) in 12.5% of strains with low-molecular-mass plasmids and 03 (one 2.0-MDa plasmid) in 7.6% of the isolates tested. On the other hand, plasmids of 1.5, 2.0, 2.2, and 3.7 MDa were frequently present in the different pattern groups, corresponding to 30.6, 33.3, 25, and 19.4% of the strains, respectively. The correlation between serotype and PP is shown in Table 2. Only five serotypes, S. enteritidis, S. typhimurium, S. virchow, S. blockley, and S. ohio, could be subdivided into 16, 13, 3, 5, and 2 plasmid groups, respectively. The other serotypes presented only one plasmid group. S. typhi, S. thompson, S. weltevreden, and the nonserological typeable strains could not be included in any established plasmid group. On the other hand, all plasmid groups except groups 02, 03, 05, 06, 07, D2, D3, D5, D7, and D9 were limited to a single serotype. The relationship between the serotypes and the PPs according to the isolation source is also given in Table 2. No plasmid group was common to the isolates from all the sources, although the 03, 05, and D5 groups were detected from strains isolated from three sources. Several PPs were common to two sources, such as 02, 06, and D2 for sporadic and environmental strains; 07 and 08 for sporadic and food isolates; D9 for strains isolated from outbreak and sporadic cases; and D3 for outbreak and environmental strains. All the serotypes could be divided into several PTs, e.g., S. enteritidis into 37 PTs, 29 of which possessed 1 strain each; 5 PTs with 2 strains each; and 3 PTs with more than 5 strains each (PT 84 with 9 strains, PT 56 with 13 strains, and PT 115 with 20 strains). S. typhimurium isolates were divided among 20 PTs, 13 of which possessed one isolate each, 3 PTs with two isolates each, PTs 114 and 115 with three isolates each, PT 122 with four isolates each, and PT 61 with five isolates. S. london could be divided into five PTs, three of which had one isolate each, and PTs 110 and 135 with three isolates each. The rest of the serotypes were divided into different PTs comprising one or two isolates each. Resistance to one or more of the antimicrobial agents tested was detected in 140 of the 171 Salmonella strains screened (81.9%). According to their origins, the antibiotic resistance proportions were 80.8% for food isolates, 79.5% for sporadic isolates, 73.5% for outbreak isolates, and 87.5% for environmental isolates. With regard to the resistance of the isolates to individual antibiotics, the resistance to Sm was the most commonly found (overall resistance, 50.9%); this was followed by resistance to Tc and Na (33.3 and 31.0%, respectively). On the other hand, only one strain was found to be Gm resistant. It is interesting that the distribution of antimicrobial resistance of the Salmonella isolates varied according to their origins. Regarding their APs, the environmental Salmonella
3060
BORREGO ET AL.
J. CLIN. MICROBIOL.
TABLE 2. Correlation between serotype and PPs of the
Salmonella isolates No. of Source and serotype
Food S. enteritidis
typhimurium S. virchow S. infantis S. typhi
S.
S. S.
anatum newport
Outbreak S. enteritidis
S.
virchow
isolates tested
strains
16
1 2 10 1 1 1 1 1 1
2 3 1 2 1 1
33
1
stan
1 2 3 4 5 9 1
03, 04, 05, D7 D5 NP 08, NP 07, D5, NP NP NP 05 Dll D8, Ti, T5 05, D9, D10 D6 03, D5 D3 NP 05
Sporadic strains enteritidis
22
S. typhimurium
15
S.
S. virchow S. infantis
1 1
Environmental S. enteritidis
13
S. typhimunum
17
S. infantis S. blockley S. ohio
2 7 7
london
9
S.
S. weltevreden S. thompson S. bovismorbificans Self-agglutinable
3 2 1 11
1 18 1 2 6 1 1
02, 06, T2, T7 NP 07, D2, D7, T3, T4 D1, D9 NP NP D5
1 12 1 3 10 2 1 1 2 4 1 8 3 2 1 11
05 NP
06, 09, D4, T6 01 NP NP
02, 03, Fl, F2, NP D3 NP 02 D2 NP NP NP 02 NP
a NP, Plasmidless strains or strains with plasmids which were not included into an established PPs. The following PPs were established (molecular mass expressed in megadaltons): 01 (0.7), 02 (1.5), 03 (2.0), 04 (2.2), 05 (2.4), 06 (2.6), 07 (3.7), 08 (5.2), 09 (4.1), Dl (1.2, 1.5), D2 (1.5, 1.8), D3 (1.5, 2.2), D4 (1.5, 3.7), D5 (2.0, 2.2), D6 (2.0, 2.6), D7 (2.6, 3.7), D8 (1.8, 3.7), D9 (1.8, 2.4), D10 (2.0, 4.4), Dll (4.1, 5.2), Ti (1.0, 1.5, 2.2), T2 (1.5, 2.0, 2.2), T3 (3.1, 3.7, 4.1), T4 (2.0, 2.2, 3.7), T5 (1.8, 3.7, 4.8), T6 (1.2, 1.5, 2.6), T7 (1.0, 1.8, 2.6), Fl (1.5, 2.0, 3.1, 3.7), and F2 (1.0, 1.5, 3.5, 3.7, 4.4).
strains differed sharply from the strains isolated from other sources. The strains isolated from water exhibited a higher level of resistance to Sm (62.5%), Tc (45.8%), SXT (5.6%), and Tm (13.9%), but all of the strains were susceptible to Cf. Salmonella isolates from sporadic cases presented a higher level of resistance to Cb (25.6%) and Ap (28.2%), and the outbreak isolates showed increased resistance to Km (52.9%) and Nm (23.5%).
The most frequent APs displayed by the S. enteritidis
strains was susceptibility to all the antimicrobial agents tested (19 strains); this was followed by resistance to Na (8 strains). In the case of S. typhimurium, the APs most frequently obtained was the combination Sm-Na with six strains. Among S. ohio and S. london isolates, the predominant AP was resistance to Sm and Tm, respectively. The distribution of the multiresistant patterns within the different serotypes is given in Table 3. None of these combinations was common to strains from all sources, although resistance triplets, such as Sm-Cm-Tc and Sm-CmNa, were detected in strains isolated from food and outbreaks, and the combination Sm-Tc-Na was distributed among isolates from environmental and sporadic human samples. The pattern Sm-Tc-Na was the most common in environmental strains, whereas in sporadic and outbreak strains, the most common patterns were Sm-Km-Nm, SmNa-Km, and Sm-Cm-Tc-Cb-Ap. Several PPs could be identified within one PT, and various PTs possessed the same PP, with a total of 24 subtypes among 34 isolates from outbreaks (Table 4). The most common PP of the S. enteritidis isolated from humans involved in disease outbreaks, PP D3, was included only in PT 56. On the contrary, the most common PT, PT 56, could be subtyped into six PPs. In Table 5, the relationship between the APs and the PPs of the outbreak isolates is given. The most frequently detected APs among S. enteritidis isolates were susceptibility to all the antimicrobial agents (n = 8) and the resistant pattern Sm-Km-Na (n = 6). The susceptible strains could be subtyped into four PPs, with PP D5 being dominant; the strains with the Sm-Km-Na resistance pattern could be subtyped into four PPs. On the other hand, D3 was the most frequently observed PP (n = 5), and the strains with this PP were included in three APs, with resistance to Na being the most common (n = 3). From the statistical analysis (Table 6), it can be deduced that only significant relationships were established between the markers AP and PT and the markers PP and PT in isolates from outbreaks, between AP and PP in isolates from sporadic, and between AP and PP and PP and PT in isolates from environmental sources. DISCUSSION For an optimal bacteriological diagnosis and successful epidemiological tracing, the classification of species into smaller units is of great importance. Serological and biochemical typing of Salmonella isolates can only be regarded as a first step in this respect (51). However, modern typing methods based on the analysis of genotypic characteristics may provide a useful tool for epidemiological studies. In the study described here, it was shown that the incidence of serotypes, PPs, PTs, and APs is extremely high in Salmonella isolates from all the sources examined (Table 1). Similar data have been published previously (17, 29, 45). We would expect that, in outbreaks in which there is a strong epidemiological association with a common exposure (food or water), the Salmonella isolates would prove to be the same regardless of the test used. However, the comparison of strains from five outbreaks with strains isolated from sporadic human cases, water, and food brings to light the absence of a standard by which a definite classification of related or unrelated can be obtained (Table 1). Thus, isolates with the same serotype, PT, PP, and AP have different sources. On the other hand, two strains which have the same serotype may differ by the loss or gain of a plasmid or by any change in phage susceptibility (7, 48).
EPIDEMIOLOGICAL MARKERS OF SALMONELLA STRAINS
VOL. 30, 1992
3061
TABLE 3. Distribution of the multiresistant patterns (three or more agents) in the different serotypes of Salmonella AP
Serotype
S. virchow S. enteritidis S. enteritidis
Sm-Cf-Tc Sm-Cm-Tc Sm-Cm-Na Sm-Cm-Na Sm-Cf-Na Sm-Tm-Tc Sm-Tc-Na Sm-Tc-Na Sm-Na-Ar Sm-Cm-SX, Sm-Km-Nr' Sm-Km-NiSm-Cb-Ap Sm-Na-Km Sm-Na-Km Sm-Na-Km
Food
Outbreaks and sporadic human isolates
No. of No. (MMa) isolates of plasmids
Sero e
1 1 1
No. (MM) of plasmids
No. of isolates
2 (2.0, 2.2) 1 (17.7) S. enteritidis 2 (17.7, 36) S. typhimurium S. typhimurium
2 1 1 1
1 (3.7) 2 (2.6, 3.7) 0 1 (2.6)
typhimurium
1
0
S. typhimurium
1
1 (3.7)
S. S. S. S.
enteritidis enteritidis enteritidis enteritidis enteritidis enteritidis
2 1 1 1 1 1
1 1 1 2 3 5
S. S. S. S.
enteritidis enteritidis enteritidis enteritidis
2 1 1 1
0 2 (2.0, 36) 2 (2.0, 36) 2 (2.0, 36)
S. enteritidis
Environmental
S
No. of No. (MM) isolates of plasmids
erotye
S. enteritidis
1 1 2
2 (1.5, 2.0) 0 0
S. weltevreden
1
1 (85)
S. ohio
1
1 (1.5)
Self-agglutinable S. london
1 1
S. typhimurinum S. typhimuinum
1 1
1 (39.5) 3 (1.5, 1.8, 32.1) 2 (1.5, 3.7) 3 (1.2, 1.5, 3.7)
typhimurium
1
S. ohio S.
S. S.
Cl-Na-Tc Cl-Na-Ap Nm-Na-Km Nm-Tc-Km Na-Tc-Km Sm-Cm-Na-Tc Sm-Cm-Na-Tc Sm-Cm-Km-Nm Sm-Tc-Km-Na
S. enteritidis S. anatum
1 1
S. typhimurium
(32.1) (36) (56) (2.0, 36) (1.8, 2.4, 16.6) (2.0, 4.4, 14.8, 30, 33.2)
1 (36) 1 (2.6)
Sm-Cm-Tc-SXT Sm-Cm-Tc-SXT Sm-Km-Na-Nm Sm-Km-Na-Tc Sm-Km-Na-Tc
S. enteritidis S. enteritidis S. typhimurium
Cf-Tc-Cb-Ap Tm-Na-Cb-Ap Tc-Na-Cb-Ap Sm-Km-Na-Nm-Tc
S.
Sm-Cm-Tc-Cb-Ap Sm-Cm-Tc-Cb-Ap Sm-Cm-Tc-Cb-Ap Sm-Km-Cb-Ap-Tc-Tm-SXT
S.
S. S. S.
S.
S.
enteritidis enteritidis enteritidis enteritidis
1
1 1
1 1 1 1
enteritidis enteritidis enteritidis
1 1 1
0 0 4 (1.8, 2.4, 16.6, 42.7) 1 (56) 2 (12, 56) 3 (19.4, 21, 42) 5 (2.0, 4.4, 14.8, 30, 33.2) 2 (1.2, 1.5) 3 (1.2, 1.5, 55) 4 (1.8, 2.4, 24, 58) S.
3 (2.6, 40.2,
110) a
MM, molecular
mass
(in megadaltons).
TABLE 4. PPs related to PTs of outbreak isolates
Agarose gel electrophoresis of DNA from 171 Salmonella strains isolated from different sources revealed a heterogeneous plasmid population (Table 2). These results are consistent with those of other studies of plasmids from strains of members of the family Enterobacteriaceae (27) and Salmonella isolates (43). Plasmid analysis of bacteria has gained acceptance as a tool for identifying Salmonella isolates (40). Many investigators have reported the utility of plasmid profile analysis in the study of the epidemiology of infection by Salmonella (9, 34), as well as in the definition and identification of bacteria originating from the same clone (32). However, in the present study we were unable to establish strong evidence of the epidemiology of the strains by means of their plasmid profiles (Table 2). These findings are in contrast to those of previous reports (31, 35, 43), whose authors were able, by means of plasmid profiles, to group S. typhimurium and S. enteritidis strains from com-
No. of isolates with the following
PP
20 21
NP1b 03 05 07 D3
D5 D6 D8 D9 D10
Ti
T5
22 23
35
36 41
Pl':
48 49 50 54 56 68 79
1
1
1
1
1
1
1
0
0
0
1
1
0
0
0 0
0 0
0 0
0 0
3 0
0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 1
0 0 0 0 0 1
0 0 0 0 0 0
0 0 0 0 0 0 1 0
0 2 0 5 3 1 0 0
0 0 2 0 0 0 0 0
0
0
0
0
0
1
1
0
0
0 0
0 0
0 0
0 0
0 1
0 0
0 0
1 0 0 0 0 0 0 0 0 0 0
0 0
0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
0 0
0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
0 0
1 0
0 0 1 0 0 0 0 0 0
0 0 0
a According to the mnemonic pattern described by Castro et al. (10). b NP, plasmidless strains or strains with plasmids which were not included in an established PP.
J. CLIN. MICROBIOL.
BORREGO ET AL.
3062
TABLE 5. APs related to PPs of outbreak isolates AP
No. of isolates with the following PP:
NPa 03 05 07 D3 D5 D6 D8 D9 D10 Ti T5
Susceptible
1 1 0 0 1 2 0 0 3 0 Km-Nm-Tc 0 0 Km-Na-Tc 0 Cm-Sm-Tc 1 Sm-Km-Na-Nm Sm-Km-Na-Nm-Tc 0
Km Tc Na Km-Na Sm-Km Km-Nm Na-Tc Sm-Km-Na Km-Na-Nm
0 0 0 0 0 0 0 0 1 1 1 1 0 0 0
2 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 1 0 3 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 2 0 0 0 0 0
4 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 1 0 0 0 0 0 0 0 0
0 0 0 0 0 1 0 0 1 0 0 0 0 0 0
0 0 0 0 0 0 0 0 1 0 0 0 0 0 1
0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 1 0 0 0 0 0 0 0 0 0 0
a NP, plasmidless strains or strains with plasmids which were not included in an established PP.
mon sources and to characterize the spread of such strains. Olsvik et al. (31) established that the presence of a single plasmid cannot always be used to characterize a clone, and it may be necessary to carry out restriction endonuclease digests when strains contain plasmids with similar molecular weights. In the present study, we compared the small plasmids (less than 20 MDa) present in Salmonella strains because these plasmids are prevalent in members of the family Enterobacteriaceae (20, 27). Small plasmids (less than 5 MDa) are found in both antibiotic-resistant and antibiotic-susceptible Salmonella strains, but the functions of several of these small plasmids are still unknown. Taylor et al. (43) reported that the strains that were devoid of small plasmids contained one high-molecular-mass plasmid (greater than 50 MDa for the drug-resistant strains and greater than 20 MDa for the drug-susceptible strains). Such plasmids should be capable of self-transfer, since the transfer operon of the F plasmid comprises about 15 MDa of DNA (52). We have observed that, while most Salmonella strains susceptible to antibiotics were plasmidless (42.8%), a high number of these strains (25%) contained plasmids with both high and low molecular masses. In the case of drug-resistant strains, 52 did not harbor plasmids (37.1%), 32 contained only small plasmids (22.9%), and 23 possessed only large plasmids (16.4%); the remaining strains presented both small and large plasmids TABLE 6. Statistical analysis of the epidemiologic markers studied Friedman test
(mean rank)
Source
Food Outbreak
Sporadic Environmental a p = 0.ooo01. 0.0005. c P = 0.002.
bp =
dp
=
0.095.
e p = 0.028.
AP/PT
PP/PT
AP/PP
AP
PP
PT
1.57 27.17a 2.40 2.64
1.07
0.64 15.0 9.8c 18.63e
1.79 2.25 2.12 2.15
1.93 1.83 1.62 1.40
2.29 1.92 2.25 2.45
22.0b 1.52
14.87d
(23.6%). These strains may, in fact, harbor plasmid aggregates consisting of a transfer factor and small plasmids encoding drug resistance (2). Some small plasmids may be formed by dissociation of the large plasmids present in the same host cell (43). The spread of drug resistance in Salmonella isolates is a relatively recent phenomenon. Current investigations show a sharp increase in antibiotic resistance among Salmonella strains, yielding proportions greater than 80% (12, 42). In this study, resistance to one or more antibiotics was recorded in 81.9% of the strains tested, demonstrating the epidemic spread and persistence of resistant clones of Salmonella in Spain. Several investigators have used the resistance typing of Salmonella strains for epidemiological purposes, mainly to characterize clones with a specific resistance (1, 37). However, we found that antimicrobial resistance testing was less specific than testing with other epidemiological markers in the identification of related isolates. These findings agree with those obtained by other investigators (17, 35). Resistance to different antimicrobial agents has been reported to be mediated by plasmids. Thus, Sm and sulfonamide resistance in Salmonella isolates is coded for by a plasmid of approximately 5.5 MDa (3, 15). Tc resistance is mediated in S. typhimurium by a plasmid of 24 MDa (31) or by one of 96 MDa (29) and by two plasmids of 50 and 3 MDa in S. dublin (30). Sm and Tc resistances are encoded by two plasmids of 71.1 and 2.5 MDa (21) or by one of 62 MDa (29) or 140 MDa (17). In addition, multiple drug resistance patterns have been related to the presence of R-type plasmids linked to cryptic plasmids (43). In the present study, resistance to three or more antimicrobial agents was observed in 43 strains, the most frequent resistance pattern being the profiles Sm-Km-Nm, Sm-Na-Km, and Sm-Cm-TcCb-Ap with three isolates each (Table 3). The Sm-Km-Nm resistance pattern was observed in two strains of S. ententidis with one plasmid of 32.1 MDa and in one strain that harbored a 36-MDa plasmid. The Sm-Na-Km resistance profile was observed in three strains of S. entenitidis with various plasmid contents, one strain with two plasmids (2 and 36 MDa), one strain with three plasmids (1.8, 2.3 and 16.6 MDa), and a final strain with five plasmids (2, 4.4, 14.8, 30, and 33.2 MDa). Finally, the multiple resistance pattern Sm-Cm-Tc-Cb-Ap was recorded in three strains of S. typhimunum with two, three, and four plasmids each with molecular masses of 1.2 and 1.5 MDa; 1.2, 1.5, and 55 MDa; and 1.8, 2.3, 24, and 58 MDa, respectively. Plasmids with similar molecular masses linked to antimicrobial resistance have been reported by other investigators. Holmberg et al. (17) described several resistance profiles associated with the presence of plasmids. Thus, the most common pattern Sm-Tc-Cb-Ap was observed in strains with one plasmid (140 MDa), two plasmids (5.7 and 73 MDa), and three plasmids (4, 5.5, and 87 MDa). In our study, this resistance pattern seemed to be linked to the presence of low-molecular-mass plasmids (from 1.2 to 1.8 MDa). Nakamura et al. (29) described a plasmid-linked resistance to Sm-Cm-Su-Tc (the plasmid was 110 MDa). A similar plasmid was obtained in our study in a multiply resistant strain of S. typhimurium, with resistance to Sm, Km, Cb, Ap, Tc, Tm, and SXT, although the strain also harbored two additional plasmids of 2.6 and 40.2 MDa. It is worthwhile to note that resistance to SXT was displayed only by environmental strains and was always linked to the presence of large plasmids. Similar results have been obtained in other enterobacterial strains (50).
VOL. 30, 1992
EPIDEMIOLOGICAL MARKERS OF SALMONELLA STRAINS
Phage typing has been the selected method of differentiating among serovars in the reference laboratory. This technique is rapid and provides a reproducible and highly discriminatory method of subdivision (47). Phage typing has been particularly useful in the epidemiological investigation of several salmonellosis outbreaks (5, 11). In the present study, we used a phage typing scheme developed in our laboratory (10) which permitted us to discriminate and group the serotypes of Salmonella with similar drug resistance and plasmid profiles into different phage types. For isolates from outbreaks, mainly from outbreak 1 (n = 14 isolates), phage typing was the most efficient technique for establishing the fact that the isolates were identical. However, in environmental strains, plasmid analysis may be preferable, and antimicrobial resistance could even serve as a good marker of related strains in food-borne isolates. Certain PPs and APs were seen repeatedly in the various outbreak strains studied, which suggests that they are relatively stable markers. ACKNOWLEDGMENTS This work was supported by a grant from United Nations Environment Programme/World Health Organization. We thank M. A. Morifiigo for help with this study and reviewing the manuscript and the National Reference Center of Salmonella (Majadahonda, Madrid, Spain) for the serological confirmation of the isolates.
REFERENCES 1. Anderson, E. S. 1975. The problem and implications of chloramphenicol resistance in the typhoid bacillus. J. Hyg. Camb. 74:289-299. 2. Anderson, E. S., and M. J. Lewis. 1965. Drug resistance and its transfer in Salmonella typhimurium. Nature (London) 206:579583. 3. Araki, Y., M. Inoue, and H. Hashimoto. 1987. Characterization and mobilization of nonconjugative plasmids encoding resistance to streptomycin and sulfanilamide. Microbiol. Immunol. 31:1107-1111. 4. Baker, R. M., and D. C. Old. 1979. Biotyping and colicine typing of Salmonella typhimurium strains of phage type 141 isolated in Scotland. J. Med. Microbiol. 12:265-276. 5. Baker, R. M., D. C. Old, and J. C. M. Sharp. 1980. Phage type/biotype groups of Salmonella typhimurium in Scotland 1974-6: variation during spread of epidemic clones. J. Hyg. Camb. 84:115-125. 6. Barry, A. L., and C. Thornberry. 1991. Susceptibility tests: diffusion test procedures, p. 1117-1125. In A. Balows, W. J. Hausler, Jr., K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (ed.), Manual of clinical microbiology, 5th ed. American Society for Microbiology, Washington, D.C. 7. Bezanson, G., R. Khakhria, and R. Lacroix. 1982. Involvement of plasmids on determining bacteriophage sensitivity in Salmonella typhimuium: genetics and physical analysis of phagovar 204. Can. J. Microbiol. 28:993-1001. 8. Bezanson, G., R. Khakhria, and D. Pagnutti. 1983. Plasmid profiles of value in differentiating Salmonella muenster isolates. J. Clin. Microbiol. 17:1159-1160. 9. Brumner, F., A. Margadant, R. Peduzzi, and J. C. Piffaretti. 1983. The plasmid pattern as an epidemiological tool for Salmonella typhimurium epidemics. Comparison with the lysotype. J. Infect. Dis. 148:7-11. 10. Castro, D., M. A. Morifiigo, E. Martinez-Manzanares, R. Cornax, M. C. Balebona, A. Luque, and J. J. Borrego. 1992. Development and application of a new scheme of phages for typing and differentiating Salmonella strains from different sources. J. Clin. Microbiol. 30:1418-1423. 11. Chart, H., E. J. Threlfall, and B. Rowe. 1989. Virulence of Salmonella enteritidis phage type 4 is related to possession of a 38 MDa plasmid. FEMS Microbiol. Lett. 58:299-304. 12. Chugh, T. D., and A. Suheir. 1983. Drug resistance among
3063
Salmonellae in Kuwait. Trop. Geogr. Med. 35:37-41. 13. Falbo, V., A. Caprioli, F. MondeUo, M. L. Cacace, S. Luzi, and D. Greco. 1982. Antimicrobial resistance among Salmonella isolates from hospitals in Rome. J. Hyg. Camb. 88:275-284. 14. Farrar, W. E., Jr. 1983. Molecular analysis of plasmids in epidemiologic investigations. J. Infect. Dis. 148:1-6. 15. Grinter, N. J., and P. T. Barth. 1976. Characterization of SmSu plasmids by restriction endonuclease cleavage and compatibility testing. J. Bacteriol. 128:394-400. 16. Gustavsen, S., and 0. Breen. 1984. Investigations of an outbreak of Salmonella oranienbwg infections in Norway, caused by contaminated black pepper. Am. J. Epidemiol. 119:806-812. 17. Holmberg, S. D., I. K. Wachsmuth, F. W. Hickman-Brenner, and M. L. Cohen. 1984. Comparison of plasmid profile analysis, phage typing, and antimicrobial susceptibility testing in characterizing Salmonella typhimurinum isolates from outbreaks. J. Clin. Microbiol. 19:100-104. 18. Holmberg, S. D., J. G. Wells, and M. L. Cohen. 1984. Animalto-man transmission of antimicrobial-resistant Salmonella: investigation of US outbreaks, 1971-1983. Science 225:883-885. 19. Ishiguro, N., G. Sato, K. Takeuchi, and A. Nakayama. 1979. A longitudinal epizootiological study of Salmonella infection on a piggery: a study on the mode of contamination by biotyping of Salnonella typhimunum and by the antibiogram. Jpn. J. Vet. Sci. 41:261-272. 20. Jamieson, A. F., D. A. Brenner, P. L. Bergquist, and H. E. D. Lane. 1979. Characterization of plasmids from antibiotic-resistant Shigella isolated by agarose gel electrophoresis. J. Gen. Microbiol. 113:73-81. 21. Jayaratne, A. H. G. P., D. L. Collins-Thompson, and J. T. Trevors. 1987. Plasmid-encoded transferable multiple antibiotic resistance in Escherichia coli isolated from meat. Syst. Appl. Microbiol. 10:78-84. 22. Johnston, W. S., G. F. Hopkins, G. K. MacLachlan, and J. C. M. Sharp. 1986. Salmonella in sewage effluent and the relationship to animal and human disease in the north of Scotland. Vet. Rec. 119:201-203. 23. Kado, C. I., and S. T. Lu. 1981. Rapid procedure for detection and isolation of large and small plasmids. J. Bacteriol. 145:13651373. 24. Litwin, C. M., A. L. Storm, S. Chipowsky, and K. J. Ryan. 1991. Molecular epidemiology of Shigella infections: plasmid profiles, serotype correlation, and restriction endonuclease analysis. J. Clin. Microbiol. 29:104-108. 25. Lujan, R., A. Echeita, M. A. Usera, J. V. Martinez-Suarez, R. Alonso, and J. A. Saez-Nieto. 1990. Plasmid profiles as an epidemiological marker for Salnonella enterica serotype enteritidis foodborne outbreaks. Microbiol. SEM 6:45-50. 26. Lyons, R. W., C. L. Samples, N. H. deSilva, K. H. Ross, E. M. Julian, and P. J. Checko. 1980. An epidemic of resistant Salnonella in a nursery: animal-to-human spread. JAMA 243:
546-547.
27. Moller, J. K., H. F. Jorgenson, C. Christiansen, G. Chnstiansen, A. L. Bak, and A. Senderup. 1978. Characterization of plasmids from wild-type Enterobacteriaceae, p. 257-261. In D. Schlessinger (ed.), Microbiology-1978. American Society for Microbiology, Washington, D.C. 28. Nakamura, M., S. Sato, T. Ohya, S. Suzuki, and S. Ikeda. 1985. Possible relationship of a 36-megadalton Salmonella entenitidis plasmid to virulence in mice. Infect. Immun. 47:831-833. 29. Nakamura, M., S. Sato, T. Ohya, S. Suzuki, and S. Ikeda. 1986. Plasmid profile analysis in epidemiological studies of animal Salnonella typhimurium infection in Japan. J. Clin. Microbiol. 23:360-365. 30. Nastasi, A., M. R. Vldlafrate, C. Mammina, M. F. Massenti, D. Ooliva, and G. Scarlata. 1987. Molecular relationship among Salmonella dublin isolates identified at the Center of Enterobacteriaceae of Palermo during the years 1971-85. Epidemiol. Infect. 99:283-290. 31. Olsvik, O., H. Sorum, K. Birkness, K. Wachsmuth, M. Fjolstad, J. Lassen, K. Fossum, and J. C. Feeley. 1985. Plasmid characterization of SalmoneUla typhimurinum transmitted from animals to humans. J. Clin. Microbiol. 22:336-338.
3064
J. CLIN. MICROBIOL.
BORREGO ET AL.
32. Orskov, F., and I. Orskov. 1983. Summary of a workshop on the clone concept in the epidemiology, taxonomy, and evaluation of the Enterobacteriaceae and other bacteria. J. Infect. Dis. 149: 346-357. 33. Popoff, M., I. Miras, C. Coynault, C. Lasselin, and P. Paradon. 1981. Molecular relationship between virulence plasamids of Salmonella serotypes typhimurium and dublin and large plasmids of other Salmonella serotypes. Ann. Microbiol. Inst. Pasteur 135A:389-398. 34. Riley, L. W., G. T. DiFerdinando, Jr., T. M. DeMelfi, and M. L. Cohen. 1983. Evaluation of isolated cases of salmonellosis by plasmid profile analysis: introduction and transmission of a bacterial clone by precooked roast beef. J. Infect. Dis. 148:1217. 35. Rodrigue, D. C., D. N. Cameron, N. D. Puhr, F. W. Brenner, M. E. St. Louis, I. K. Wachsmuth, and R. V. Tauxe. 1992. Comparison of plasmid profiles, phage types, and antimicrobial resistance patterns of Salmonella enteritidis isolates in the United States. J. Clin. Microbiol. 30:854-857. 36. Rodrigue, D. C., R. V. Tauxe, and B. Rowe. 1990. International increase in Salmonella enteritidis; a new pandemic?. Epidemiol. Infect. 105:21-27. 37. Rowe, B., and E. J. Threlfall. 1984. Drug resistance in gramnegative aerobic bacilli. Br. Med. Bull. 40:68-76. 38. Sansonetti, P. J., D. J. Kopecko, and S. B. Formal. 1981. Shigella sonnei plasmids: evidence that a large plasmid is necessary for virulence. Infect. Immun. 34:75-83. 39. Saxena, S. N., M. L. Mago, L. N. Rao Bhau, S. Ahuja, and H. Singh. 1983. Salmonella serotypes prevalent in India during 1978-81. Indian J. Med. Res. 77:10-18. 40. Schaberg, D. R., L. S. Tompkins, and S. Falkow. 1981. Use of agarose gel electrophoresis of plasmid deoxyribonucleic acid to fingerprint gram-negative bacilli. J. Clin. Microbiol. 13:11051108. 41. Scheinbach, S., and S. I. Hong. 1988. Detection of resident populations of Salmonella and Escherichia coli in food ingredients by plasmid analysis. Food Microbiol. 5:235-241. 42. Sojka, W. J., C. Wray, and I. McLaren. 1986. A survey of drug
43.
44.
45. 46.
47.
48.
49. 50.
51. 52.
resistance in Salmonellae isolated from animals in England and Wales in 1982 and 1983. Br. Vet. J. 142:371-380. Taylor, D. E., J. G. Levine, and K. L. Kouvelos. 1982. Incidence of plasmid DNA in Salmonella strains isolated from clinical sources in Ontario, Canada, during 1979 and 1980. Can. J. Microbiol. 28:1150-1157. Taylor, D. N., C. Bopp, K. Birkness, and M. L. Cohen. 1984. An outbreak of salmonellosis associated with fatality in a healthy child: a large dose and severe illness. Am. J. Epidemiol. 119:907-912. Terakado, N., T. Ohya, H. Ueda, and Y. Isayama. 1980. A survey on drug resistance and R plasmids in Salmonella isolated from domestic animals in Japan. Jpn. J. Vet. Sci. 42:543-550. Thiel, W., and S. Recla. 1984. Epidemiologie und lysotypie von Salmonella typhimunum in Osterreich 1980-1982. Zentralb. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. Reihe A 258:544-563. Threlfall, E. J., and J. A. Frost. 1990. The identification, typing and fingerprinting of Salmonella: laboratory aspects and epidemiological applications. J. Appl. Bacteriol. 68:5-16. Threlfall, E. J., L. R. Ward, and B. Rowe. 1978. Epidemic spread of chloramphenicol-resistant strain of Salmonella typhimurium phage type 204 in bovine animals in Britain. Vet. Rec. 103:438-440. Toranzo, A. E., J. L. Barja, R. R. Coiwell, and F. M. Hetrick. 1983. Characterization of plasmids in bacterial fish pathogens. Infect. Immun. 39:184-192. Tschape, H., E. Tietze, R. Prager, and H. Heier. 1986. Occurrence in water and waste water of E. coli and coliform bacteria carrying R-plasmids. Part 3. Plasmids encoding gentamicin, trimethoprim or streptothricin resistance. Acta Hydrochim. Hydrobiol. 14:167-174. Wachsmuth, K. 1986. Molecular epidemiology of bacterial infections; examples of methodology and investigations of outbreaks. Rev. Infect. Dis. 8:682-692. Willetts, N., and R. A. Skurray. 1980. The conjugation system of F-like plasmids. Annu. Rev. Genet. 14:41-76.
Vol. 30, No. 12
Comparison of Epidemiological Markers of Salmonella Strains Isolated from Different Sources in Spain JUAN J. BORREGO,1* DOLORES CASTRO,1 MANUEL JIMENEZ-NOTARIO,2 ANTONIO LUQUE,l EDUARDO MARTINEZ-MANZANARES,1 CARMEN RODRIGUEZ-AVIAL,3 AND JUAN J. PICAZO3 Departamento de Microbiologia, Universidad de Mdlaga, Campus Universitario Teatinos, 29071-Mdlaga,1 Centro de Salud de Fuengirola, Servicio Andaluz de Salud, Fuengirola (Mdlaga), 2 and Servicio de Microbiologia, Hospital Clinico Universitario San Carlos, Plaza Cristo Rey sin, 28040-Madrid,3 Spain Received 4 April 1992/Accepted 1 September 1992
A comparative study of the phage types, antimicrobial susceptibility patterns, and plasmid profiles of 171 strains of Salmonela isolated from food, epidemic outbreaks, and water-contaminated environments as well as sporadic human isolates was carried out to determine the most adequate marker in epidemiological investigations. Typing based on the plasmid profiles appears to be the most effective method for grouping strains with the same serotype obtained from a single outbreak and from environmental sources. However, none of the three markers tested allow us total discrimination and identification of related strains from a common source for epidemiological tracing. Therefore, the combined use of the three methods is necessary for determining whether common source isolates are related.
Salmonella infections appear to be one of the most typical examples of an enteric disease that is transmitted from animals to humans. The transmission occurs both through food, such as meat, dairy products, and egg by-products (16, 18), and by contact between animals and humans directly through the fecal-oral route (26) or indirectly via fecally contaminated waters (22). More than 2,000 serovars of Salmonella have been described, and all are considered to be potentially pathogenic for animals, including humans. Salmonellosis in humans can produce symptoms ranging in severity from intestinal disturbances to death (44). Infections caused by Salmonella strains, mainly S. enteritidis, have increased over the last few years (36). Bacterial epidemic strains have traditionally been defined on the basis of phenotypic properties of the isolates, such as serotype and biotype. However, these determinations have not always been useful in epidemiological tracing of the epidemic strains (14, 17, 25, 47). To evaluate the epidemiological relationship among isolates from a common source, several different subtyping methods have been applied, such as serotyping (39), biotyping and colicine typing (4, 19), phage typing (46), antimicrobial resistance patterns (12, 13, 42), and plasmid analysis (8, 31). Several studies have demonstrated that plasmid profile analysis of the Salmonella isolates could be a reliable epidemiological tool for the differentiation of epidemic and nonepidemic strains from outbreaks (9, 14, 35, 43) or to elucidate the epidemiology of these food-borne pathogens (41). However, only a few studies of the application of this marker to environmental, food, and animal isolates have been carried out (7, 29, 34). To study the reliability of the subtyping methods as useful epidemiological markers, we compared Salmonella strains isolated from contaminated food, outbreaks, sporadic human cases, and fecally contaminated waters by use of their phage types (PTs), plasmid profiles (PPs), and antimicrobial resistance patterns (APs). *
MATERIALS AND METHODS Bacterial strains. Salmonella strains were selected and separated into the following four groups according to their sources and epidemiological significance: (i) isolates from contaminated food (n = 26); (ii) isolates from different outbreaks in various locations of the provinces of Malaga and Madrid (Spain) and previously described by Castro et al. (10) (n = 34); (iii) sporadic human Salmonella isolates (n = 38); and (iv) isolates from contaminated natural fresh and marine waters (Guadalhorce River and Mediterranean Sea, respectively) (n = 72). All isolates were biochemically confirmed by using the API 20E system (API Systems S.A., Montalieu-Vercieu, France) and were serologically characterized by the slide agglutination test with Salmonella polyvalent 0 and group antisera (Difco Laboratories, Detroit, Mich.) as well as by the tube agglutination test with H antisera (Difco). All strains were sent to the National Reference Center of Salmonella (Majadahonda, Madrid, Spain) for confirmation of serotyping results. Antimicrobial susceptibility testing. The disk diffusion method (6) on Mueller-Hinton agar (Difco) was used to test the resistance of the isolates to the following antimicrobial agents (supplied by Biomerieux, Madrid, Spain): colistin (Cl), 10 ,ug; nalidixic acid (Na), 30 ,ug; gentamicin (Gm), 10 ,ug; streptomycin (Sm), 10 ,ug; kanamycin (Km), 30 p,g; tetracycline (Tc), 30 ,ug; chloramphenicol (Cm), 30 p,g; ampicillin (Ap), 10 ,ug; carbenicillin (Cb), 100 ,Lg; cephalothin (Cf), 30 ,ug; trimethoprim-sulfamethoxazole (SXT), 25 ,ug; tobramycin (Tm), 10 ,ug; and neomycin (Nm), 30 ,ug. Phage typing. PTs were determined by the use of 25 typing phages belonging to a new phage typing system developed by Castro et al. (10). PTs were established and identified by the patterns of reaction to the typing phages described previously (10). PP analysis. Plasmid DNA was extracted from lysed Salmonella isolates by a modification of the technique described by Kado and Liu (23) and performed by Toranzo et al. (49). Extracted plasmid DNA was electrophoresed for 2 h at 150 mA on a 0.7% horizontal agarose gel (Sigma Chemical Co., St. Louis, Mo.) in Tris-acetate buffer (40 mM Tris base, 2 mM disodium EDTA; adjusted to pH 7.9 with glacial acetic
Corresponding author. 3058
EPIDEMIOLOGICAL MARKERS OF SALMONELLA STRAINS
VOL. 30, 1992
TABLE 1. Serotype distribution and numbers of PTs, antimicrobial resistance patterns, and plasmid profiles of the Salmonella isolates according to their origins Source and serotype
No. of H isolates islts PTs
No.Pof:
P
APs
PPs
Food S. entenitidis S. virchow S. typhimuium S. typhi S. anatum S. infantis S. newport Total
16 3 2 2 1 1 1 26
15 3 2 2 1 1 1 25
11 3 2 1 1 1 1 20
14 3 2 1 1 1 1 23
Outbreak no. 1, S. enteritidis Outbreak no. 2, S. virchow Outbreak no. 3, S. enteritidis Outbreak no. 9, S. enteritidis Outbreak no. 11, S. enteritidis Outbreak no. 12, S. enteritidis Outbreak total
13 1 9 6 3 2 34
2 1 6 1 3 1 14
4 1 5 6 3 1 20
8 1 2 2 2 1 16
22 15 1 1 39
8 10 1 1 20
12 8 1 1 22
5 8 1 1 15
17 13 9 7 7 3 2 2 1 11 72
12 6 6 5 6 3 2 2 1 9 52
12 9 5 4 4 3 2 2 1 8 50
9 3 3 6 3 2 2 1 1 5 35
Sporadic strains S. entenitidis S. typhimurium S. virchow S. infantis Total
Environmental S. typhimurium S. enteritidis S. london S. blockley S. ohio S. weltevreden S. infantis S. thompson S. bovismorbificans Self-agglutinable Total
acid). Then, the gels were stained with ethidium bromide solution (0.5 ,ug/ml) and photographed under a UV transilluminator (LKB, Pharmacia, Barcelona, Spain) with Plus-XPan film and a 23 A Wratten filter. The approximate molecular masses of the plasmids (in megadaltons) were determined by comparison with plasmids with known molecular masses, e.g., R40a (96 MDa) and plasmids from Escherichia coli V517 ranging from 35.8 to 1.4 MDa. The nonparametric Friedman test, chi-square test, and t test (one tail) were used for statistical analysis by using StatView 512 Plus software and a Macintosh SE computer. RESULTS Table 1 summarizes the serotype distribution and the number of PTs, APs, and PPs of the four groups of Salmonella isolates selected (food, outbreak, sporadic, and environmental strains). The Salmonella serotypes most frequently isolated were S. enteritidis (48.8%) and S. typhimurium (20%), although only S. enteritidis was repeatedly isolated from all the sample groups. Food isolates presented a wide variety of PTs, APs, and PPs, and no epidemiological relationship was observed among Salmonella isolates (chi-square = 3.71; P > 0.95).
3059
Similar results were obtained with the environmental strains, which presented a high number of plasmidless strains. On the other hand, the isolates from outbreaks possessed common serotypes, PTs, and PPs but a great variety of APs. Numerous PPs were found in the 171 isolates. Many of them were shown to harbor large plasmids (from 82.8 to 110 MDa) that have generally been related to virulence properties (25, 28, 33). The large plasmids were not used for pattern comparison of isolates because of their proven instability on subculture (24, 38). Plasmids under 20 MDa were used for comparison of PPs. All the Salmonella isolates were grouped into 29 PPs. Patterns containing from one to five plasmids with low molecular masses were found in 72 isolates. The most frequently detected plasmid groups were D5 (two plasmids of 2.0 and 2.2 MDa) in 12.5% of strains with low-molecular-mass plasmids and 03 (one 2.0-MDa plasmid) in 7.6% of the isolates tested. On the other hand, plasmids of 1.5, 2.0, 2.2, and 3.7 MDa were frequently present in the different pattern groups, corresponding to 30.6, 33.3, 25, and 19.4% of the strains, respectively. The correlation between serotype and PP is shown in Table 2. Only five serotypes, S. enteritidis, S. typhimurium, S. virchow, S. blockley, and S. ohio, could be subdivided into 16, 13, 3, 5, and 2 plasmid groups, respectively. The other serotypes presented only one plasmid group. S. typhi, S. thompson, S. weltevreden, and the nonserological typeable strains could not be included in any established plasmid group. On the other hand, all plasmid groups except groups 02, 03, 05, 06, 07, D2, D3, D5, D7, and D9 were limited to a single serotype. The relationship between the serotypes and the PPs according to the isolation source is also given in Table 2. No plasmid group was common to the isolates from all the sources, although the 03, 05, and D5 groups were detected from strains isolated from three sources. Several PPs were common to two sources, such as 02, 06, and D2 for sporadic and environmental strains; 07 and 08 for sporadic and food isolates; D9 for strains isolated from outbreak and sporadic cases; and D3 for outbreak and environmental strains. All the serotypes could be divided into several PTs, e.g., S. enteritidis into 37 PTs, 29 of which possessed 1 strain each; 5 PTs with 2 strains each; and 3 PTs with more than 5 strains each (PT 84 with 9 strains, PT 56 with 13 strains, and PT 115 with 20 strains). S. typhimurium isolates were divided among 20 PTs, 13 of which possessed one isolate each, 3 PTs with two isolates each, PTs 114 and 115 with three isolates each, PT 122 with four isolates each, and PT 61 with five isolates. S. london could be divided into five PTs, three of which had one isolate each, and PTs 110 and 135 with three isolates each. The rest of the serotypes were divided into different PTs comprising one or two isolates each. Resistance to one or more of the antimicrobial agents tested was detected in 140 of the 171 Salmonella strains screened (81.9%). According to their origins, the antibiotic resistance proportions were 80.8% for food isolates, 79.5% for sporadic isolates, 73.5% for outbreak isolates, and 87.5% for environmental isolates. With regard to the resistance of the isolates to individual antibiotics, the resistance to Sm was the most commonly found (overall resistance, 50.9%); this was followed by resistance to Tc and Na (33.3 and 31.0%, respectively). On the other hand, only one strain was found to be Gm resistant. It is interesting that the distribution of antimicrobial resistance of the Salmonella isolates varied according to their origins. Regarding their APs, the environmental Salmonella
3060
BORREGO ET AL.
J. CLIN. MICROBIOL.
TABLE 2. Correlation between serotype and PPs of the
Salmonella isolates No. of Source and serotype
Food S. enteritidis
typhimurium S. virchow S. infantis S. typhi
S.
S. S.
anatum newport
Outbreak S. enteritidis
S.
virchow
isolates tested
strains
16
1 2 10 1 1 1 1 1 1
2 3 1 2 1 1
33
1
stan
1 2 3 4 5 9 1
03, 04, 05, D7 D5 NP 08, NP 07, D5, NP NP NP 05 Dll D8, Ti, T5 05, D9, D10 D6 03, D5 D3 NP 05
Sporadic strains enteritidis
22
S. typhimurium
15
S.
S. virchow S. infantis
1 1
Environmental S. enteritidis
13
S. typhimunum
17
S. infantis S. blockley S. ohio
2 7 7
london
9
S.
S. weltevreden S. thompson S. bovismorbificans Self-agglutinable
3 2 1 11
1 18 1 2 6 1 1
02, 06, T2, T7 NP 07, D2, D7, T3, T4 D1, D9 NP NP D5
1 12 1 3 10 2 1 1 2 4 1 8 3 2 1 11
05 NP
06, 09, D4, T6 01 NP NP
02, 03, Fl, F2, NP D3 NP 02 D2 NP NP NP 02 NP
a NP, Plasmidless strains or strains with plasmids which were not included into an established PPs. The following PPs were established (molecular mass expressed in megadaltons): 01 (0.7), 02 (1.5), 03 (2.0), 04 (2.2), 05 (2.4), 06 (2.6), 07 (3.7), 08 (5.2), 09 (4.1), Dl (1.2, 1.5), D2 (1.5, 1.8), D3 (1.5, 2.2), D4 (1.5, 3.7), D5 (2.0, 2.2), D6 (2.0, 2.6), D7 (2.6, 3.7), D8 (1.8, 3.7), D9 (1.8, 2.4), D10 (2.0, 4.4), Dll (4.1, 5.2), Ti (1.0, 1.5, 2.2), T2 (1.5, 2.0, 2.2), T3 (3.1, 3.7, 4.1), T4 (2.0, 2.2, 3.7), T5 (1.8, 3.7, 4.8), T6 (1.2, 1.5, 2.6), T7 (1.0, 1.8, 2.6), Fl (1.5, 2.0, 3.1, 3.7), and F2 (1.0, 1.5, 3.5, 3.7, 4.4).
strains differed sharply from the strains isolated from other sources. The strains isolated from water exhibited a higher level of resistance to Sm (62.5%), Tc (45.8%), SXT (5.6%), and Tm (13.9%), but all of the strains were susceptible to Cf. Salmonella isolates from sporadic cases presented a higher level of resistance to Cb (25.6%) and Ap (28.2%), and the outbreak isolates showed increased resistance to Km (52.9%) and Nm (23.5%).
The most frequent APs displayed by the S. enteritidis
strains was susceptibility to all the antimicrobial agents tested (19 strains); this was followed by resistance to Na (8 strains). In the case of S. typhimurium, the APs most frequently obtained was the combination Sm-Na with six strains. Among S. ohio and S. london isolates, the predominant AP was resistance to Sm and Tm, respectively. The distribution of the multiresistant patterns within the different serotypes is given in Table 3. None of these combinations was common to strains from all sources, although resistance triplets, such as Sm-Cm-Tc and Sm-CmNa, were detected in strains isolated from food and outbreaks, and the combination Sm-Tc-Na was distributed among isolates from environmental and sporadic human samples. The pattern Sm-Tc-Na was the most common in environmental strains, whereas in sporadic and outbreak strains, the most common patterns were Sm-Km-Nm, SmNa-Km, and Sm-Cm-Tc-Cb-Ap. Several PPs could be identified within one PT, and various PTs possessed the same PP, with a total of 24 subtypes among 34 isolates from outbreaks (Table 4). The most common PP of the S. enteritidis isolated from humans involved in disease outbreaks, PP D3, was included only in PT 56. On the contrary, the most common PT, PT 56, could be subtyped into six PPs. In Table 5, the relationship between the APs and the PPs of the outbreak isolates is given. The most frequently detected APs among S. enteritidis isolates were susceptibility to all the antimicrobial agents (n = 8) and the resistant pattern Sm-Km-Na (n = 6). The susceptible strains could be subtyped into four PPs, with PP D5 being dominant; the strains with the Sm-Km-Na resistance pattern could be subtyped into four PPs. On the other hand, D3 was the most frequently observed PP (n = 5), and the strains with this PP were included in three APs, with resistance to Na being the most common (n = 3). From the statistical analysis (Table 6), it can be deduced that only significant relationships were established between the markers AP and PT and the markers PP and PT in isolates from outbreaks, between AP and PP in isolates from sporadic, and between AP and PP and PP and PT in isolates from environmental sources. DISCUSSION For an optimal bacteriological diagnosis and successful epidemiological tracing, the classification of species into smaller units is of great importance. Serological and biochemical typing of Salmonella isolates can only be regarded as a first step in this respect (51). However, modern typing methods based on the analysis of genotypic characteristics may provide a useful tool for epidemiological studies. In the study described here, it was shown that the incidence of serotypes, PPs, PTs, and APs is extremely high in Salmonella isolates from all the sources examined (Table 1). Similar data have been published previously (17, 29, 45). We would expect that, in outbreaks in which there is a strong epidemiological association with a common exposure (food or water), the Salmonella isolates would prove to be the same regardless of the test used. However, the comparison of strains from five outbreaks with strains isolated from sporadic human cases, water, and food brings to light the absence of a standard by which a definite classification of related or unrelated can be obtained (Table 1). Thus, isolates with the same serotype, PT, PP, and AP have different sources. On the other hand, two strains which have the same serotype may differ by the loss or gain of a plasmid or by any change in phage susceptibility (7, 48).
EPIDEMIOLOGICAL MARKERS OF SALMONELLA STRAINS
VOL. 30, 1992
3061
TABLE 3. Distribution of the multiresistant patterns (three or more agents) in the different serotypes of Salmonella AP
Serotype
S. virchow S. enteritidis S. enteritidis
Sm-Cf-Tc Sm-Cm-Tc Sm-Cm-Na Sm-Cm-Na Sm-Cf-Na Sm-Tm-Tc Sm-Tc-Na Sm-Tc-Na Sm-Na-Ar Sm-Cm-SX, Sm-Km-Nr' Sm-Km-NiSm-Cb-Ap Sm-Na-Km Sm-Na-Km Sm-Na-Km
Food
Outbreaks and sporadic human isolates
No. of No. (MMa) isolates of plasmids
Sero e
1 1 1
No. (MM) of plasmids
No. of isolates
2 (2.0, 2.2) 1 (17.7) S. enteritidis 2 (17.7, 36) S. typhimurium S. typhimurium
2 1 1 1
1 (3.7) 2 (2.6, 3.7) 0 1 (2.6)
typhimurium
1
0
S. typhimurium
1
1 (3.7)
S. S. S. S.
enteritidis enteritidis enteritidis enteritidis enteritidis enteritidis
2 1 1 1 1 1
1 1 1 2 3 5
S. S. S. S.
enteritidis enteritidis enteritidis enteritidis
2 1 1 1
0 2 (2.0, 36) 2 (2.0, 36) 2 (2.0, 36)
S. enteritidis
Environmental
S
No. of No. (MM) isolates of plasmids
erotye
S. enteritidis
1 1 2
2 (1.5, 2.0) 0 0
S. weltevreden
1
1 (85)
S. ohio
1
1 (1.5)
Self-agglutinable S. london
1 1
S. typhimurinum S. typhimuinum
1 1
1 (39.5) 3 (1.5, 1.8, 32.1) 2 (1.5, 3.7) 3 (1.2, 1.5, 3.7)
typhimurium
1
S. ohio S.
S. S.
Cl-Na-Tc Cl-Na-Ap Nm-Na-Km Nm-Tc-Km Na-Tc-Km Sm-Cm-Na-Tc Sm-Cm-Na-Tc Sm-Cm-Km-Nm Sm-Tc-Km-Na
S. enteritidis S. anatum
1 1
S. typhimurium
(32.1) (36) (56) (2.0, 36) (1.8, 2.4, 16.6) (2.0, 4.4, 14.8, 30, 33.2)
1 (36) 1 (2.6)
Sm-Cm-Tc-SXT Sm-Cm-Tc-SXT Sm-Km-Na-Nm Sm-Km-Na-Tc Sm-Km-Na-Tc
S. enteritidis S. enteritidis S. typhimurium
Cf-Tc-Cb-Ap Tm-Na-Cb-Ap Tc-Na-Cb-Ap Sm-Km-Na-Nm-Tc
S.
Sm-Cm-Tc-Cb-Ap Sm-Cm-Tc-Cb-Ap Sm-Cm-Tc-Cb-Ap Sm-Km-Cb-Ap-Tc-Tm-SXT
S.
S. S. S.
S.
S.
enteritidis enteritidis enteritidis enteritidis
1
1 1
1 1 1 1
enteritidis enteritidis enteritidis
1 1 1
0 0 4 (1.8, 2.4, 16.6, 42.7) 1 (56) 2 (12, 56) 3 (19.4, 21, 42) 5 (2.0, 4.4, 14.8, 30, 33.2) 2 (1.2, 1.5) 3 (1.2, 1.5, 55) 4 (1.8, 2.4, 24, 58) S.
3 (2.6, 40.2,
110) a
MM, molecular
mass
(in megadaltons).
TABLE 4. PPs related to PTs of outbreak isolates
Agarose gel electrophoresis of DNA from 171 Salmonella strains isolated from different sources revealed a heterogeneous plasmid population (Table 2). These results are consistent with those of other studies of plasmids from strains of members of the family Enterobacteriaceae (27) and Salmonella isolates (43). Plasmid analysis of bacteria has gained acceptance as a tool for identifying Salmonella isolates (40). Many investigators have reported the utility of plasmid profile analysis in the study of the epidemiology of infection by Salmonella (9, 34), as well as in the definition and identification of bacteria originating from the same clone (32). However, in the present study we were unable to establish strong evidence of the epidemiology of the strains by means of their plasmid profiles (Table 2). These findings are in contrast to those of previous reports (31, 35, 43), whose authors were able, by means of plasmid profiles, to group S. typhimurium and S. enteritidis strains from com-
No. of isolates with the following
PP
20 21
NP1b 03 05 07 D3
D5 D6 D8 D9 D10
Ti
T5
22 23
35
36 41
Pl':
48 49 50 54 56 68 79
1
1
1
1
1
1
1
0
0
0
1
1
0
0
0 0
0 0
0 0
0 0
3 0
0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 1
0 0 0 0 0 1
0 0 0 0 0 0
0 0 0 0 0 0 1 0
0 2 0 5 3 1 0 0
0 0 2 0 0 0 0 0
0
0
0
0
0
1
1
0
0
0 0
0 0
0 0
0 0
0 1
0 0
0 0
1 0 0 0 0 0 0 0 0 0 0
0 0
0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
0 0
0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
0 0
1 0
0 0 1 0 0 0 0 0 0
0 0 0
a According to the mnemonic pattern described by Castro et al. (10). b NP, plasmidless strains or strains with plasmids which were not included in an established PP.
J. CLIN. MICROBIOL.
BORREGO ET AL.
3062
TABLE 5. APs related to PPs of outbreak isolates AP
No. of isolates with the following PP:
NPa 03 05 07 D3 D5 D6 D8 D9 D10 Ti T5
Susceptible
1 1 0 0 1 2 0 0 3 0 Km-Nm-Tc 0 0 Km-Na-Tc 0 Cm-Sm-Tc 1 Sm-Km-Na-Nm Sm-Km-Na-Nm-Tc 0
Km Tc Na Km-Na Sm-Km Km-Nm Na-Tc Sm-Km-Na Km-Na-Nm
0 0 0 0 0 0 0 0 1 1 1 1 0 0 0
2 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 1 0 3 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 2 0 0 0 0 0
4 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 1 0 0 0 0 0 0 0 0
0 0 0 0 0 1 0 0 1 0 0 0 0 0 0
0 0 0 0 0 0 0 0 1 0 0 0 0 0 1
0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 1 0 0 0 0 0 0 0 0 0 0
a NP, plasmidless strains or strains with plasmids which were not included in an established PP.
mon sources and to characterize the spread of such strains. Olsvik et al. (31) established that the presence of a single plasmid cannot always be used to characterize a clone, and it may be necessary to carry out restriction endonuclease digests when strains contain plasmids with similar molecular weights. In the present study, we compared the small plasmids (less than 20 MDa) present in Salmonella strains because these plasmids are prevalent in members of the family Enterobacteriaceae (20, 27). Small plasmids (less than 5 MDa) are found in both antibiotic-resistant and antibiotic-susceptible Salmonella strains, but the functions of several of these small plasmids are still unknown. Taylor et al. (43) reported that the strains that were devoid of small plasmids contained one high-molecular-mass plasmid (greater than 50 MDa for the drug-resistant strains and greater than 20 MDa for the drug-susceptible strains). Such plasmids should be capable of self-transfer, since the transfer operon of the F plasmid comprises about 15 MDa of DNA (52). We have observed that, while most Salmonella strains susceptible to antibiotics were plasmidless (42.8%), a high number of these strains (25%) contained plasmids with both high and low molecular masses. In the case of drug-resistant strains, 52 did not harbor plasmids (37.1%), 32 contained only small plasmids (22.9%), and 23 possessed only large plasmids (16.4%); the remaining strains presented both small and large plasmids TABLE 6. Statistical analysis of the epidemiologic markers studied Friedman test
(mean rank)
Source
Food Outbreak
Sporadic Environmental a p = 0.ooo01. 0.0005. c P = 0.002.
bp =
dp
=
0.095.
e p = 0.028.
AP/PT
PP/PT
AP/PP
AP
PP
PT
1.57 27.17a 2.40 2.64
1.07
0.64 15.0 9.8c 18.63e
1.79 2.25 2.12 2.15
1.93 1.83 1.62 1.40
2.29 1.92 2.25 2.45
22.0b 1.52
14.87d
(23.6%). These strains may, in fact, harbor plasmid aggregates consisting of a transfer factor and small plasmids encoding drug resistance (2). Some small plasmids may be formed by dissociation of the large plasmids present in the same host cell (43). The spread of drug resistance in Salmonella isolates is a relatively recent phenomenon. Current investigations show a sharp increase in antibiotic resistance among Salmonella strains, yielding proportions greater than 80% (12, 42). In this study, resistance to one or more antibiotics was recorded in 81.9% of the strains tested, demonstrating the epidemic spread and persistence of resistant clones of Salmonella in Spain. Several investigators have used the resistance typing of Salmonella strains for epidemiological purposes, mainly to characterize clones with a specific resistance (1, 37). However, we found that antimicrobial resistance testing was less specific than testing with other epidemiological markers in the identification of related isolates. These findings agree with those obtained by other investigators (17, 35). Resistance to different antimicrobial agents has been reported to be mediated by plasmids. Thus, Sm and sulfonamide resistance in Salmonella isolates is coded for by a plasmid of approximately 5.5 MDa (3, 15). Tc resistance is mediated in S. typhimurium by a plasmid of 24 MDa (31) or by one of 96 MDa (29) and by two plasmids of 50 and 3 MDa in S. dublin (30). Sm and Tc resistances are encoded by two plasmids of 71.1 and 2.5 MDa (21) or by one of 62 MDa (29) or 140 MDa (17). In addition, multiple drug resistance patterns have been related to the presence of R-type plasmids linked to cryptic plasmids (43). In the present study, resistance to three or more antimicrobial agents was observed in 43 strains, the most frequent resistance pattern being the profiles Sm-Km-Nm, Sm-Na-Km, and Sm-Cm-TcCb-Ap with three isolates each (Table 3). The Sm-Km-Nm resistance pattern was observed in two strains of S. ententidis with one plasmid of 32.1 MDa and in one strain that harbored a 36-MDa plasmid. The Sm-Na-Km resistance profile was observed in three strains of S. entenitidis with various plasmid contents, one strain with two plasmids (2 and 36 MDa), one strain with three plasmids (1.8, 2.3 and 16.6 MDa), and a final strain with five plasmids (2, 4.4, 14.8, 30, and 33.2 MDa). Finally, the multiple resistance pattern Sm-Cm-Tc-Cb-Ap was recorded in three strains of S. typhimunum with two, three, and four plasmids each with molecular masses of 1.2 and 1.5 MDa; 1.2, 1.5, and 55 MDa; and 1.8, 2.3, 24, and 58 MDa, respectively. Plasmids with similar molecular masses linked to antimicrobial resistance have been reported by other investigators. Holmberg et al. (17) described several resistance profiles associated with the presence of plasmids. Thus, the most common pattern Sm-Tc-Cb-Ap was observed in strains with one plasmid (140 MDa), two plasmids (5.7 and 73 MDa), and three plasmids (4, 5.5, and 87 MDa). In our study, this resistance pattern seemed to be linked to the presence of low-molecular-mass plasmids (from 1.2 to 1.8 MDa). Nakamura et al. (29) described a plasmid-linked resistance to Sm-Cm-Su-Tc (the plasmid was 110 MDa). A similar plasmid was obtained in our study in a multiply resistant strain of S. typhimurium, with resistance to Sm, Km, Cb, Ap, Tc, Tm, and SXT, although the strain also harbored two additional plasmids of 2.6 and 40.2 MDa. It is worthwhile to note that resistance to SXT was displayed only by environmental strains and was always linked to the presence of large plasmids. Similar results have been obtained in other enterobacterial strains (50).
VOL. 30, 1992
EPIDEMIOLOGICAL MARKERS OF SALMONELLA STRAINS
Phage typing has been the selected method of differentiating among serovars in the reference laboratory. This technique is rapid and provides a reproducible and highly discriminatory method of subdivision (47). Phage typing has been particularly useful in the epidemiological investigation of several salmonellosis outbreaks (5, 11). In the present study, we used a phage typing scheme developed in our laboratory (10) which permitted us to discriminate and group the serotypes of Salmonella with similar drug resistance and plasmid profiles into different phage types. For isolates from outbreaks, mainly from outbreak 1 (n = 14 isolates), phage typing was the most efficient technique for establishing the fact that the isolates were identical. However, in environmental strains, plasmid analysis may be preferable, and antimicrobial resistance could even serve as a good marker of related strains in food-borne isolates. Certain PPs and APs were seen repeatedly in the various outbreak strains studied, which suggests that they are relatively stable markers. ACKNOWLEDGMENTS This work was supported by a grant from United Nations Environment Programme/World Health Organization. We thank M. A. Morifiigo for help with this study and reviewing the manuscript and the National Reference Center of Salmonella (Majadahonda, Madrid, Spain) for the serological confirmation of the isolates.
REFERENCES 1. Anderson, E. S. 1975. The problem and implications of chloramphenicol resistance in the typhoid bacillus. J. Hyg. Camb. 74:289-299. 2. Anderson, E. S., and M. J. Lewis. 1965. Drug resistance and its transfer in Salmonella typhimurium. Nature (London) 206:579583. 3. Araki, Y., M. Inoue, and H. Hashimoto. 1987. Characterization and mobilization of nonconjugative plasmids encoding resistance to streptomycin and sulfanilamide. Microbiol. Immunol. 31:1107-1111. 4. Baker, R. M., and D. C. Old. 1979. Biotyping and colicine typing of Salmonella typhimurium strains of phage type 141 isolated in Scotland. J. Med. Microbiol. 12:265-276. 5. Baker, R. M., D. C. Old, and J. C. M. Sharp. 1980. Phage type/biotype groups of Salmonella typhimurium in Scotland 1974-6: variation during spread of epidemic clones. J. Hyg. Camb. 84:115-125. 6. Barry, A. L., and C. Thornberry. 1991. Susceptibility tests: diffusion test procedures, p. 1117-1125. In A. Balows, W. J. Hausler, Jr., K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (ed.), Manual of clinical microbiology, 5th ed. American Society for Microbiology, Washington, D.C. 7. Bezanson, G., R. Khakhria, and R. Lacroix. 1982. Involvement of plasmids on determining bacteriophage sensitivity in Salmonella typhimuium: genetics and physical analysis of phagovar 204. Can. J. Microbiol. 28:993-1001. 8. Bezanson, G., R. Khakhria, and D. Pagnutti. 1983. Plasmid profiles of value in differentiating Salmonella muenster isolates. J. Clin. Microbiol. 17:1159-1160. 9. Brumner, F., A. Margadant, R. Peduzzi, and J. C. Piffaretti. 1983. The plasmid pattern as an epidemiological tool for Salmonella typhimurium epidemics. Comparison with the lysotype. J. Infect. Dis. 148:7-11. 10. Castro, D., M. A. Morifiigo, E. Martinez-Manzanares, R. Cornax, M. C. Balebona, A. Luque, and J. J. Borrego. 1992. Development and application of a new scheme of phages for typing and differentiating Salmonella strains from different sources. J. Clin. Microbiol. 30:1418-1423. 11. Chart, H., E. J. Threlfall, and B. Rowe. 1989. Virulence of Salmonella enteritidis phage type 4 is related to possession of a 38 MDa plasmid. FEMS Microbiol. Lett. 58:299-304. 12. Chugh, T. D., and A. Suheir. 1983. Drug resistance among
3063
Salmonellae in Kuwait. Trop. Geogr. Med. 35:37-41. 13. Falbo, V., A. Caprioli, F. MondeUo, M. L. Cacace, S. Luzi, and D. Greco. 1982. Antimicrobial resistance among Salmonella isolates from hospitals in Rome. J. Hyg. Camb. 88:275-284. 14. Farrar, W. E., Jr. 1983. Molecular analysis of plasmids in epidemiologic investigations. J. Infect. Dis. 148:1-6. 15. Grinter, N. J., and P. T. Barth. 1976. Characterization of SmSu plasmids by restriction endonuclease cleavage and compatibility testing. J. Bacteriol. 128:394-400. 16. Gustavsen, S., and 0. Breen. 1984. Investigations of an outbreak of Salmonella oranienbwg infections in Norway, caused by contaminated black pepper. Am. J. Epidemiol. 119:806-812. 17. Holmberg, S. D., I. K. Wachsmuth, F. W. Hickman-Brenner, and M. L. Cohen. 1984. Comparison of plasmid profile analysis, phage typing, and antimicrobial susceptibility testing in characterizing Salmonella typhimurinum isolates from outbreaks. J. Clin. Microbiol. 19:100-104. 18. Holmberg, S. D., J. G. Wells, and M. L. Cohen. 1984. Animalto-man transmission of antimicrobial-resistant Salmonella: investigation of US outbreaks, 1971-1983. Science 225:883-885. 19. Ishiguro, N., G. Sato, K. Takeuchi, and A. Nakayama. 1979. A longitudinal epizootiological study of Salmonella infection on a piggery: a study on the mode of contamination by biotyping of Salnonella typhimunum and by the antibiogram. Jpn. J. Vet. Sci. 41:261-272. 20. Jamieson, A. F., D. A. Brenner, P. L. Bergquist, and H. E. D. Lane. 1979. Characterization of plasmids from antibiotic-resistant Shigella isolated by agarose gel electrophoresis. J. Gen. Microbiol. 113:73-81. 21. Jayaratne, A. H. G. P., D. L. Collins-Thompson, and J. T. Trevors. 1987. Plasmid-encoded transferable multiple antibiotic resistance in Escherichia coli isolated from meat. Syst. Appl. Microbiol. 10:78-84. 22. Johnston, W. S., G. F. Hopkins, G. K. MacLachlan, and J. C. M. Sharp. 1986. Salmonella in sewage effluent and the relationship to animal and human disease in the north of Scotland. Vet. Rec. 119:201-203. 23. Kado, C. I., and S. T. Lu. 1981. Rapid procedure for detection and isolation of large and small plasmids. J. Bacteriol. 145:13651373. 24. Litwin, C. M., A. L. Storm, S. Chipowsky, and K. J. Ryan. 1991. Molecular epidemiology of Shigella infections: plasmid profiles, serotype correlation, and restriction endonuclease analysis. J. Clin. Microbiol. 29:104-108. 25. Lujan, R., A. Echeita, M. A. Usera, J. V. Martinez-Suarez, R. Alonso, and J. A. Saez-Nieto. 1990. Plasmid profiles as an epidemiological marker for Salnonella enterica serotype enteritidis foodborne outbreaks. Microbiol. SEM 6:45-50. 26. Lyons, R. W., C. L. Samples, N. H. deSilva, K. H. Ross, E. M. Julian, and P. J. Checko. 1980. An epidemic of resistant Salnonella in a nursery: animal-to-human spread. JAMA 243:
546-547.
27. Moller, J. K., H. F. Jorgenson, C. Christiansen, G. Chnstiansen, A. L. Bak, and A. Senderup. 1978. Characterization of plasmids from wild-type Enterobacteriaceae, p. 257-261. In D. Schlessinger (ed.), Microbiology-1978. American Society for Microbiology, Washington, D.C. 28. Nakamura, M., S. Sato, T. Ohya, S. Suzuki, and S. Ikeda. 1985. Possible relationship of a 36-megadalton Salmonella entenitidis plasmid to virulence in mice. Infect. Immun. 47:831-833. 29. Nakamura, M., S. Sato, T. Ohya, S. Suzuki, and S. Ikeda. 1986. Plasmid profile analysis in epidemiological studies of animal Salnonella typhimurium infection in Japan. J. Clin. Microbiol. 23:360-365. 30. Nastasi, A., M. R. Vldlafrate, C. Mammina, M. F. Massenti, D. Ooliva, and G. Scarlata. 1987. Molecular relationship among Salmonella dublin isolates identified at the Center of Enterobacteriaceae of Palermo during the years 1971-85. Epidemiol. Infect. 99:283-290. 31. Olsvik, O., H. Sorum, K. Birkness, K. Wachsmuth, M. Fjolstad, J. Lassen, K. Fossum, and J. C. Feeley. 1985. Plasmid characterization of SalmoneUla typhimurinum transmitted from animals to humans. J. Clin. Microbiol. 22:336-338.
3064
J. CLIN. MICROBIOL.
BORREGO ET AL.
32. Orskov, F., and I. Orskov. 1983. Summary of a workshop on the clone concept in the epidemiology, taxonomy, and evaluation of the Enterobacteriaceae and other bacteria. J. Infect. Dis. 149: 346-357. 33. Popoff, M., I. Miras, C. Coynault, C. Lasselin, and P. Paradon. 1981. Molecular relationship between virulence plasamids of Salmonella serotypes typhimurium and dublin and large plasmids of other Salmonella serotypes. Ann. Microbiol. Inst. Pasteur 135A:389-398. 34. Riley, L. W., G. T. DiFerdinando, Jr., T. M. DeMelfi, and M. L. Cohen. 1983. Evaluation of isolated cases of salmonellosis by plasmid profile analysis: introduction and transmission of a bacterial clone by precooked roast beef. J. Infect. Dis. 148:1217. 35. Rodrigue, D. C., D. N. Cameron, N. D. Puhr, F. W. Brenner, M. E. St. Louis, I. K. Wachsmuth, and R. V. Tauxe. 1992. Comparison of plasmid profiles, phage types, and antimicrobial resistance patterns of Salmonella enteritidis isolates in the United States. J. Clin. Microbiol. 30:854-857. 36. Rodrigue, D. C., R. V. Tauxe, and B. Rowe. 1990. International increase in Salmonella enteritidis; a new pandemic?. Epidemiol. Infect. 105:21-27. 37. Rowe, B., and E. J. Threlfall. 1984. Drug resistance in gramnegative aerobic bacilli. Br. Med. Bull. 40:68-76. 38. Sansonetti, P. J., D. J. Kopecko, and S. B. Formal. 1981. Shigella sonnei plasmids: evidence that a large plasmid is necessary for virulence. Infect. Immun. 34:75-83. 39. Saxena, S. N., M. L. Mago, L. N. Rao Bhau, S. Ahuja, and H. Singh. 1983. Salmonella serotypes prevalent in India during 1978-81. Indian J. Med. Res. 77:10-18. 40. Schaberg, D. R., L. S. Tompkins, and S. Falkow. 1981. Use of agarose gel electrophoresis of plasmid deoxyribonucleic acid to fingerprint gram-negative bacilli. J. Clin. Microbiol. 13:11051108. 41. Scheinbach, S., and S. I. Hong. 1988. Detection of resident populations of Salmonella and Escherichia coli in food ingredients by plasmid analysis. Food Microbiol. 5:235-241. 42. Sojka, W. J., C. Wray, and I. McLaren. 1986. A survey of drug
43.
44.
45. 46.
47.
48.
49. 50.
51. 52.
resistance in Salmonellae isolated from animals in England and Wales in 1982 and 1983. Br. Vet. J. 142:371-380. Taylor, D. E., J. G. Levine, and K. L. Kouvelos. 1982. Incidence of plasmid DNA in Salmonella strains isolated from clinical sources in Ontario, Canada, during 1979 and 1980. Can. J. Microbiol. 28:1150-1157. Taylor, D. N., C. Bopp, K. Birkness, and M. L. Cohen. 1984. An outbreak of salmonellosis associated with fatality in a healthy child: a large dose and severe illness. Am. J. Epidemiol. 119:907-912. Terakado, N., T. Ohya, H. Ueda, and Y. Isayama. 1980. A survey on drug resistance and R plasmids in Salmonella isolated from domestic animals in Japan. Jpn. J. Vet. Sci. 42:543-550. Thiel, W., and S. Recla. 1984. Epidemiologie und lysotypie von Salmonella typhimunum in Osterreich 1980-1982. Zentralb. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. Reihe A 258:544-563. Threlfall, E. J., and J. A. Frost. 1990. The identification, typing and fingerprinting of Salmonella: laboratory aspects and epidemiological applications. J. Appl. Bacteriol. 68:5-16. Threlfall, E. J., L. R. Ward, and B. Rowe. 1978. Epidemic spread of chloramphenicol-resistant strain of Salmonella typhimurium phage type 204 in bovine animals in Britain. Vet. Rec. 103:438-440. Toranzo, A. E., J. L. Barja, R. R. Coiwell, and F. M. Hetrick. 1983. Characterization of plasmids in bacterial fish pathogens. Infect. Immun. 39:184-192. Tschape, H., E. Tietze, R. Prager, and H. Heier. 1986. Occurrence in water and waste water of E. coli and coliform bacteria carrying R-plasmids. Part 3. Plasmids encoding gentamicin, trimethoprim or streptothricin resistance. Acta Hydrochim. Hydrobiol. 14:167-174. Wachsmuth, K. 1986. Molecular epidemiology of bacterial infections; examples of methodology and investigations of outbreaks. Rev. Infect. Dis. 8:682-692. Willetts, N., and R. A. Skurray. 1980. The conjugation system of F-like plasmids. Annu. Rev. Genet. 14:41-76.
