LETTER

Quinupristin-dalfopristin is a mixture of semisynthetic streptogramins A and B that was recently licensed for clinical use in the United States and Europe (3). A related antibiotic, virginiamycin, has been used as a growth promoter for production animals in Europe and the United States, although its use was banned in the European Union in July 1999. Virginiamycin-resistant Enterococcus faecium strains have been isolated from exposed farm animals, raw meat, and hospital patients and are cross-resistant to quinupristin-dalfopristin (1, 2, 5; L. B. Jensen, A. M. Hammerum, F. M. Aarestrup, A. E. van den Bogaard, and E. E. Stobberingh, Letter, Antimicrob. Agents Chemother. 42:3330-3331, 1998; G. Werner, I. Klare, and W. Witte, Letter, Eur. J. Clin. Microbiol. Infect. Dis. 17:401-402, 1998). Resistance to streptogramin A is a prerequisite for resistance to quinupristin-dalfopristin and virginiamycin and is mediated in E. faecium by vat(D) (previously satA) or vat(E) (previously satG), two plasmid-mediated genes that encode acetyltransferases that inactivate streptogramin A (4; G. Werner and W. Witte, Letter, Antimicrob. Agents Chemother. 43:1813-1814, 1999). Available vat(E) sequences are not identical, and we propose designating the alleles in the order of their deposition in the GenBank database: vat(E-1) (accession numbers AF139735, AF229200, and AF242872) and vat(E-2) (AF153312). vat(E-2) differs from vat(E-1) by three nucleotides (99.5% identity), which are predicted to result in two amino acid substitutions (Fig. 1). Two other alleles, each differing from vat(E-1) by two nucleotides, have been reported but have not yet been deposited in GenBank (6).

View larger version (22K): [in this window] [in a new window]   FIG. 1.   Comparison of the amino acid sequences of alleles of the Vat(E) streptogramin A acetyltransferase (see the text for GenBank accession numbers).

We have previously detected vat(E) by PCR in isolates of quinupristin-dalfopristin-resistant E. faecium (MIC  32 µg/ml) from animals and raw meat (n = 10) and also from hospital patients in the United Kingdom (n = 4) (5). A 512-bp internal fragment of vat(E) was amplified from these isolates (5) and subjected to direct cycle sequencing using an ALFexpress DNA sequencer (Amersham Pharmacia Biotech, St. Albans, United Kingdom) and a Thermo Sequenase fluorescence-labeled primer cycle sequencing kit (Amersham Pharmacia Biotech). The sequences, which represented 80% of the vat(E) gene, were compared with those of vat(E-1) and vat(E-2). The sequences of the PCR products from 13 isolates were identical to vat(E-1). However, one isolate, designated E. faecium A41, from a hospital patient, yielded a distinct sequence. Two overlapping fragments of the vat(E) allele from this strain were amplified and cloned into pCR2.1-TOPO (Invitrogen, Groningen, The Netherlands) to yield recombinant plasmid pARL00.31, containing the 512-bp fragment, and pARL00.38, containing a 300-bp fragment spanning the 3' end of vat(E) and extending 137 bp downstream of the stop codon. The latter fragment was amplified with primers 5'-CCA ATT CAA CTC ATC GGA CC-3' and 5'-TAC GAG TAG AGT ACC GCC AG-3' and corresponded to nucleotides 4063 to 4362 of GenBank sequence AF242872. For each fragment, the inserts of three separate clones were sequenced in both directions using a Dye-Labeled ddNTP Terminator Cycle Sequencing Kit (Beckman Coulter UK Ltd., High Wycombe, United Kingdom) and samples were analyzed on a CEQ 2000 automated sequencer (Beckman). Fragments were assembled with ContigExpress (InforMax Inc., Oxford, United Kingdom).

The vat(E) allele from E. faecium A41 has been designated vat(E-3) and deposited in GenBank under the accession number AY008284. It had 20 nucleotide changes (4% divergence) compared with vat(E-1). Fifteen of these changes were silent, but the others resulted in five previously undescribed amino acid substitutions (Fig. 1). The predicted Vat(E-3) peptide had 97% amino acid identity with Vat(E-1) and 96% identity with Vat(E-2). In comparison with sequences downstream of vat(E-1) and vat(E-2), the sequence immediately downstream of the vat(E-3) stop codon had a single base insertion (an additional C after nucleotide 4235 of GenBank sequence AF242872) and two substitutions (both TC changes at nucleotides 4227 and 4253 of GenBank sequence AF242872).

We have confirmed the allelic nature of vat(E) apparent in GenBank submissions and previous reports (2, 6). The vat(E-1) allele was present in all vat(E) PCR-positive E. faecium strains from nonhuman sources studied here and in three of the four clinical isolates. The fourth clinical isolate harbored vat(E-3), which showed greater sequence divergence from vat(E-1) than other previously reported alleles (20 versus 2 or 3 nucleotide changes). In conclusion, isolates of quinupristin-dalfopristin- and virginiamycin-resistant E. faecium that give a vat(E)-specific PCR product should not be assumed to carry identical alleles. Furthermore, we suggest that the epidemiological significance of a vat(E)-positive PCR result cannot be judged accurately in the absence of sequence data.