OUP user menu

Plasmid-mediated florfenicol and ceftriaxone resistance encoded by the floR and blaCMY-2 genes in Salmonella enterica serovars Typhimurium and Newport isolated in the United States

Benoît Doublet, Alessandra Carattoli, Jean M. Whichard, David G. White, Sylvie Baucheron, Elisabeth Chaslus-Dancla, Axel Cloeckaert
DOI: http://dx.doi.org/10.1111/j.1574-6968.2004.tb09496.x 301-305 First published online: 1 April 2004


Multidrug resistance plasmids carrying the blaCMY-2 gene have been identified in Salmonella enterica serovars Typhimurium and Newport from the United States. This gene confers decreased susceptibility to ceftriaxone, and is most often found in strains with concomitant resistance to ampicillin, chloramphenicol, streptomycin, sulfamethoxazole and tetracycline. The blaCMY-2-carrying plasmids studied here were shown to also carry the florfenicol resistance gene, floR, on a genetic structure previously identified in Escherichia coli plasmids in Europe. These data indicate that the use of different antimicrobial agents, including phenicols, may serve to maintain multidrug resistance plasmids on which extended-spectrum cephalosporin resistance determinants co-exist with other resistance genes in Salmonella.

  • Salmonella
  • Multidrug resistance
  • Phenicols
  • Cephalosporins
  • Plasmid
  • Conjugation

1 Introduction

Salmonella enterica and Escherichia coli exhibiting resistance to extended-spectrum cephalosporins are an emergent problem worldwide. Until 1996, resistance to extended-spectrum cephalosporins was rarely reported among Salmonella. In 2000, the emergence of domestically acquired infections by extended-spectrum cephalosporin-resistant Salmonella carrying a plasmid-mediated CMY-2 AmpC β-lactamase was reported by the National Antimicrobial Resistance Monitoring System (NARMS) in the United States [1]. Molecular and phenotypic analysis of extended-spectrum cephalosporin-resistant strains revealed several distinct serovars and chromosomal DNA patterns, suggesting that this resistance phenotype is present among genetically diverse strains [13]. It has been recently demonstrated that the blaCMY genes in Salmonella reside on different large plasmids, which were divided into three groups (type A, B and C) on the basis of restriction patterns and blaCMY-2 gene hybridization profiles [4].

The occurrence of CMY-2-mediated cephalosporin resistance in Salmonella has now been reported in Canada [5], Spain [6], Romania [7], and Taiwan [8]. In 2002, the US Centers for Disease Control and Prevention investigated an outbreak of multidrug-resistant S. enterica serovar Newport (S. Newport) infections associated with eating raw or undercooked ground beef [9]. This multidrug resistance phenotype included resistance to streptomycin, sulfamethoxazole, tetracycline, ampicillin, chloramphenicol and decreased susceptibility to extended-spectrum cephalosporins. Researchers in Pennsylvania, USA have also identified plasmid-mediated extended-spectrum cephalosporin resistance conferred by the blaCMY-2 gene in S. Newport strains isolated from cattle [2].

The prevalence of extended-spectrum cephalosporin resistant (including ceftiofur-resistance) Salmonella from food animals has increased (URL: http://www.arru.saa.ars.usda.gov). Ceftiofur is an extended-spectrum cephalosporin approved for use in veterinary medicine [10,11], and emergence of these strains may be related to use of this drug. Since the blaCMY genes confer decreased susceptibility to both ceftiofur and ceftriaxone, the use of ceftiofur has the potential to select for Salmonella cross-resistant to ceftriaxone [4]. However, more detailed scientific analyses are needed to determine the extent of non-cephalosporin antimicrobial resistance determinants that may be present on the blaCMY-2-carrying plasmids as well. In fact, many plasmids that contain blaCMY-2 genes have been reported to confer additional resistance to aminoglycosides, phenicols, tetracyclines and sulfonamides. Ceftriaxone-resistant Salmonella strains recently isolated from animals in Canada [5] and retail foods in the US [11] have also been reported to be resistant to florfenicol. Florfenicol, a veterinary fluorinated analog of thiamphenicol and chloramphenicol, was approved by the United States Food and Drug Administration (FDA) in 1996 for treatment of bovine respiratory pathogens such as Mannheimia haemolytica, Pasteurella multocida, and Haemophilus somnus. Florfenicol resistance has recently been identified among veterinary E. coli isolates isolated from calves with diarrhea, and the floR resistance gene, located on conjugative and non-conjugative plasmids, has been identified among florfenicol-resistant E. coli from poultry and cattle isolated in the US and Europe [1214]. The floR gene has also been identified in the Salmonella genomic island 1 (SGI1) of S. Typhimurium Definitive Type 104 (DT104) and of other S. enterica serovars, i.e. Agona, Albany, and Paratyphi B [1517]. A plasmid encoding a highly related floR gene (pp-flo) was previously identified in Japan from Photobacterium damselae subsp. piscicida[18]. Analysis of the DNA sequences upstream and downstream of the floR gene in E. coli plasmids revealed the presence of two open reading frames (orfA′ and orfA) that show considerable homology to transposase proteins, suggesting that the floR gene could be located within a mobile genetic element [12].

Since there have been increasing reports of blaCMY-2-carrying plasmids that possess additional resistance phenotypes, including phenicol resistance, we sought to determine possible genetic linkages of the florfenicol resistance gene, floR, with blaCMY-2 genes conferring resistance to extended-spectrum cephalosporins, and to compare the florfenicol resistance genes and their flanking regions with other previously described florfenicol resistance genotypes.

2 Materials and methods

Seven isolates were chosen among the blaCMY-2-positive unrelated strains isolated from humans through the NARMS project at CDC in the 1996–1999 period, and from retail foods by the Center for Veterinary Medicine, US Food and Drug Administration in 2000, using the following criteria: they were resistant to florfenicol, showing a minimal inhibitory concentration (MIC) ≥ 32 μg/ml; they were previously characterized at molecular level for the presence of plasmid located blaCMY-2 gene [4,5,11]; they demonstrated the transfer of florfenicol resistance by conjugation to an E. coli susceptible strain. The plasmids in the S. Typhimurium strains 2039 and 3977 and the plasmid in the S. Newport strain 4528 were previously classified as blaCMY-2-positive type A and C, respectively [4]. The phenotypic characteristics of the selected strains are reported in Table 1. Molecular methods used to characterize the plasmids, floR and flanking regions were those described previously [12,1517,19].

View this table:
Table 1

Ceftriaxone-resistant Salmonella strains used in this study

StrainSerovarAntibiotic resistance phenotypeFf MICb (μg/ml)Cm MIC (μg/ml)
1290Typhimurium DT208ApCmFfSmSuTcGmKmFxCro128>256
1291Typhimurium DT208ApCmFfSmSuTcGmKmFxCro64>256
  • Abbreviations: Ap, ampicillin; Cm, chloramphenicol; Ff, florfenicol; Sm, streptomycin; Sp, spectinomycin; Su, sulfonamides; Tc, tetracycline; Tm, trimethoprim; Gm, gentamicin; Km, kanamycin; Cro, ceftriaxone; Fx, Cefoxitin.

  • aStrains labeled as TC1 are E. coli transconjugant strains.

  • bMIC breakpoints of Ff and Cm: susceptible MIC≤8 μg/ml; resistant MIC≤32 μg/ml.

3 Results and discussion

All strains and their derivative E. coli transconjugants were confirmed to possess the floR gene by PCR using previously published specific primers [12]. PCR products obtained from strains 2039 and 3977 were sequenced demonstrating that the floR gene was almost identical (1 bp different at position 3746) to that previously described in E. coli BN10660 plasmid isolated from cattle in France (EMBL Accession No. AF231986), but differed from that located within the Salmonella Genomic Island SGI1 of S. Typhimurium DT104 by 20 bp (EMBL Accession No. AF261825). Furthermore, all the florfenicol-resistant Salmonella strains of this study were found to lack the SGI1, as determined by PCR amplification and Southern blot hybridization experiments (data not shown) [1517].

Minimal inhibitory concentrations of both chloramphenicol and florfenicol for Salmonella strains showing plasmid-mediated florfenicol resistance, as well as their respective transconjugants, are reported in Table 1. Particularly high MIC levels for both florfenicol and chloramphenicol (both 256 μg/ml) were observed for the three S. Newport strains (4528, 5313 and 7073), and the same MIC levels were maintained by their relative transconjugants (4528TC1, 5313TC1 and 7073TC1). Lower MIC levels were observed for transconjugants obtained from the four S. Typhimurium strains (2039TC1, 3977TC1, 1290TC1 and 1291TC1). The origin of this discrepancy is not known but it could be due to the differential expression of the floR resistance gene in the E. coli recipient cells or additional contribution of other efflux mechanisms in the wild-type isolates. Florfenicol and chloramphenicol MICs for the host strain of E. coli used for conjugation were 4 μg/ml.

To gain insight into the nature of florfenicol resistance in these strains, all plasmids were purified, digested with BglI and compared with the E. coli plasmid BN10660 previously described to carry the floR-gene associated to the orfA and orfA′ genes [12]. Although the plasmids obtained from Salmonella strains were not identical (Fig. 1, panel (a), lanes 2–8), restriction analysis suggested that they were similar to one another but different from the E. coli BN10660 plasmid (Fig. 1, panel (a), lane 1). Plasmid profiles were hybridized with specific probes as described previously [4,19]. These included the blaCMY-2 gene and the orfA′–floR–orfA gene cluster, the latter obtained from the cloned floR plasmid pEF03 (Fig. 1). The blaCMY-2 probe hybridized with a BglI fragment of ca. 5 kb in the four S. Typhimurium strains (Fig. 1, panel (b), lanes 2–5) and with a distinct BglI band of ca. 8 kb in the three S. Newport strains (Fig. 1, panel (b), lanes 6–8), suggesting different localizations of the blaCMY-2 gene on these plasmids. As expected, the E. coli BN10660 plasmid was negative for the presence of the blaCMY-2 gene (Fig. 1, panel (b), lane 1). Hybridization with the orfA′–floR–orfA specific probe yielded the same pattern for all seven transconjugants and the E. coli BN10660 plasmid (Fig. 1, panel (c), lanes 1–8). This result indicates that the floR-resistance determinant and flanking transposase regions are apparently well-conserved in all the blaCMY-2-carrying plasmids analysed, and are consistent with the pattern previously described on the E. coli BN10660 plasmid isolated from cattle in Europe [12].

Figure 1

(a) BglI restriction profiles of plasmids extracted from E. coli BN10660 (lane 1), from transconjugants 2039TC1 (lane 2), 3977TC1 (lane 3), 1290TC1 (lane 4), 1291TC1 (lane 5), 4528TC1 (lane 6), 5313TC1 (lane 7) and 7073TC1 (lane 8). (b) Southern blot hybridization with the blaCMY-2probe (consisting of the blaCMY-2 internal PCR product) of BglI digested plasmids. (c) Southern blot hybridization with the pEF03 probe of BglI digested plasmids. A schematic view of the gene organization of the floR locus on the 6522 bp EcoRI–BamHI pEF03 inset is shown below [12]. The sizes and locations of the three BglI (b) fragments are indicated. The reading frames for orfA′, floR, and orfA are shown as arrows.

In summary, we have identified the florfenicol floR resistance gene on blaCMY-2-carrying plasmids conferring extended-spectrum cephalosporin (e.g. ceftiofur and ceftriaxone) resistance to Salmonella isolated from humans, animals, and foods in the US. Having these resistance genes co-exist on the same plasmid may allow the bacterial host to be positively selected by either individual antimicrobial agent. Thus, the application of a phenicol selective pressure in a food animal production setting could contribute to the dissemination of the blaCMY-2-plasmids, and vice versa. The floR gene also confers cross-resistance to chloramphenicol, raising the possibility that chloramphenicol could have previously selected the floR gene long before the introduction of florfenicol for veterinary use, although chloramphenicol use in all food-producing animals has been prohibited in the US since 1984. Additional antimicrobial resistance phenotypes were also observed among tested transconjugants (e.g. aminoglycosides, sulfa drugs, tetracyclines), demonstrating the continued need to determine what other antimicrobial resistance genes co-exist on the blaCMY-2-carrying plasmids in addition to those for florfenicol.

Our results also demonstrate that the floR gene on blaCMY-2-plasmids is flanked by the orfA′ and orfA sequences consistent with the genetic organization previously described for different E. coli plasmids circulating in animals in Europe that lack cephalosporin resistance determinants [12]. Protein sequence homologies of the floR-flanking open reading frames suggest that they could be transposases. This means that the floR gene is likely located within a mobile element, which could facilitate its transfer from one plasmid to another or to the bacterial chromosome [19].

The use of antimicrobial agents in food animals has drawn increased scientific and public scrutiny in recent years. The appearance and dissemination of multidrug-resistant foodborne bacterial pathogens is of particular concern. Once antimicrobial use puts pressure on a microbe, resistance may quickly emerge in that species [20]. The presence of more than one antimicrobial resistance gene on a plasmid allows for co-selection of multidrug resistance phenotypes. Circulation of linked resistance genes such as blaCMY-2and floR among bacteria in food animal production environments is significant in light of the possibility of contamination of food products with these organisms. Our findings further highlight the potential ramifications of antimicrobial selective pressures in food animal production settings and the important interface between veterinary and human medicine.


We thank C. Mouline for expert technical assistance. This study was funded by INRA, projet Transversalité and by ISS, Progetto Ricerca Intramurale No. 2012/RI.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
View Abstract