Characterization of the 6′-N-Aminoglycoside Acetyltransferase Gene aac(6′)-Iq from the Integron of a Natural Multiresistance Plasmid

Abstract

The nucleotide sequence of a newly identified amikacin resistance gene, aac(6′)-Iq (551 bp), is reported. It has 68.4 and 94.4% homology with the aac(6′)-Ia gene and the recently described aac(6′)-Ip gene, respectively. Analysis of its flanking sequences indicated that it is in the first cassette of a class I integron and has an attC site (59-base element) 108 bp in length.

Resistance to amikacin in Klebsiella pneumoniae is widespread and to date has been found to depend exclusively upon production of enzymes which modify the antibiotic (15). Different genes designated aac(6′)-Ib, aac(6′)-Id, aac(6′)-Il, aac(6′)-Ip, ant(4′)-IIa, and aph(3′)-VIb, encoding the production of enzymes that modify amikacin, have been characterized for this species (6, 16). The aac(6′)-Ib gene, which was the most frequently identified gene in several multicenter studies, is located in integrons (8) or in Tn1331 (2, 18, 19). The other genes are extremely rare (16). All aac(6′)-I genes thus far described for gram-negative bacteria are within cassettes, except for aac(6′)-Ic, which is located in chromosomal DNA (15, 16). Cassettes are mobile elements, each of which usually includes a single structural gene, which covers most of the cassette length, and a recombination site known as a 59-base element or attC site. Despite differences in sequence and length, members of the 59-base-element family are active as sites for site-specific recombination events catalyzed by the integron DNA integrase. When the cassettes are integrated, they become part of the integron and, since they usually do not have their own promoters, are expressed from a common promoter region upstream of the attachment site (5, 7). Integrons are usually located on plasmids or on transposons (pVS1, Tn21, and Tn402, etc.) (1, 10, 14); this location allows the rapid spread of cassettes among a wide variety of bacterial species. In the present communication we describe the aac(6′)-Iq gene cassette of a multiresistant strain, the K. pneumoniae K1 isolate from the Clínica Santa Rosa in Buenos Aires, Argentina. This gene is carried in a 50-kb conjugative plasmid that also codes for resistance to cefotaxime, gentamicin, and sulfonamides.

The aminoglycoside resistance profile (16) of K. pneumoniae K1, determined by using aminoglycoside disks from Schering-Plough, Kenilworth, N.J. (16), on Mueller-Hinton agar, suggested the presence of AAC(3)-II and AAC(6′)-I enzymes. By colony hybridization (8) with a labeled oligonucleotide probe (5′-CCT CCG TTA TTG CCT TC), it was shown that the AAC(3)-II activity was due to the aac(3)-IIa (aacC2) gene, which is usually flanked by IS140s and is not part of integrons. By using PCR amplification with primers specific for the integron 5′ conserved sequence (5′-CS) (5′-GGC ATC CAA GCA GCA AG) and 3′ conserved sequence (3′-CS) (5′-AAG CAG ACT TGA CCT GA) (8), we determined that the variable region of the integron was 1,600 bp in length, suggesting the presence of two cassettes. PCR amplification revealed that the ant(3")-Ia (aadA1) gene cassette was the second and last cassette in the integron based on the sizes of the bands obtained: 850 bp with the 5′-CS and ant(3")-Ia (5′-TCG ATG ACG CCA ACT AC) primers and 170 bp with the ant(3")-Ia-3′ (5′-CGC AGA TCA CTT GGA AG) and 3′-CS primers. Its location was confirmed by sequencing the beginning of the gene encoding resistance to both streptomycin and spectinomycin (see below).

The PCR amplification product of the 5′-CS2 (5′-GCC TGA CGA TGC GTG GA) and ant(3")-Ia primers (1,150 bp), made by using cloned Pfu polymerase (Stratagene, La Jolla, Calif.) was cloned into the pCR^TM2.1 cloning vector (original TA cloning kit; Invitrogen Corporation, San Diego, Calif.) in the Escherichia coli NM522 host strain (laboratory collection). This plasmid was named pLQ1001. Since the pCR^TM2.1 vector contains a kanamycin resistance gene, the 1,200-bp EcoRI fragment from pLQ1001 was subcloned into the EcoRI site of the ampicillin resistance-carrying plasmid pTZ19 (Pharmacia-LKB Biotechnology, Uppsala, Sweden), resulting in the plasmid pLQ1002. The disk susceptibility testing of E. coli NM522(pLQ1002) produced a typical AAC(6′)-I profile (Table 1), with low-level expression of enzyme activity. All other molecular procedures were carried out by using previously published protocols (13).

TABLE 1.

Disk susceptibilities of the host strain E. coli NM522(pTZ19), E. coli NM522(pLQ1002), and the parent K. pneumoniae K1 strain

Drug	Zone diam (mm)a
	E. coli NM522(pTZ19)	K. pneumoniae K1	E. coli NM522(pLQ1002)
Neomycin	24	18	25
Kanamycin	26	8	12
Tobramycin	26	9	15
5-Episisomicin	24	16	20
Gentamicin	26	8	28
Amikacin	30	16	18
Isepamicin	30	18	22
Netilmicin	28	12	18
2′-N-Ethylnetilmicin	32	6	18
Apramycin	30	30	28
6′-N-Ethylnetilmicin	30	14	32
Fortimicin A	30	30	32

^aE. coli NM522 (New England Biolabs) transformed with the pTZ19 vector was used as a negative control. The disk susceptibility data show a typical AAC(6′)-I profile of resistance to amikacin and tobramycin for the E. coli NM522(pLQ1002) and the combination AAC(3′)-II and AAC(6′)-I aminoglycoside resistance profile for the K. pneumoniae K1 strain. 

We sequenced the 835-bp PCR amplification product, made directly from K. pneumoniae K1 by using cloned Pfu polymerase, of the 5′-CS and ant(3")-Ia primers after purifying the DNA with the QIAquick kit according to the manufacturer’s instructions (Qiagen Inc., Studio City, Calif.). The sequencing was done on both DNA strands, using an ABI 373 sequencer. The nucleotide sequences were analyzed with the Genetics Computer Group programs. We found an open reading frame (ORF) spanning 551 nucleotides from the start codon, TTG, beginning at position 127 to the stop codon, TAG, ending at position 678. This ORF has 94.4% similarity in 521 bp with the recently described aac(6′)-Ip gene (6), 68.4% similarity in 551 bp with aac(6′)-Ia (17), and 66.9% similarity in 139 bp with aac(6′)-Ii (3). We named the ORF aac(6′)-Iq (Fig. 1). The location of the putative start codon, TTG, of aac(6′)-Iq corresponds to that of the start codon, ATG, of the aac(6′)-Ia gene (positions 476 to 478). The presence of TTG instead of ATG as an initiation codon could explain the lower aminoglycoside resistance activity (9). The putative ribosome-binding site is located 6 bp upstream (positions 117 to 120) from the translation initiation codon of the aac(6′)-Iq gene.

FIG. 1.

Alignment of the nucleotide sequence of the 835-bp PCR product containing the aac(6′)-Iq gene and the sequences of the aac(6′)-Ip (accession no. Z54241) and aac(6′)-Ia (accession no. M18967) genes. Dots indicate identical nucleotides, and dashes indicate gaps introduced to optimize the alignment. The sequences of the 5′ conserved segments from the integrons containing aac(6′)-Ia and aac(6′)-Iq are identical. The start and stop codons are indicated in boldface capital letters. The attC site (59-base element) is underlined. The proposed ribosome-binding site (rbs) of aac(6′)-Iq is also indicated. The seven-base core sites are boxed. The arrows indicate the start and the end of the aac(6′)-Iq cassette.

The first 77 nucleotides of the 5′ flanking sequence of the aac(6′)-Iq gene cassette possess 100% identity with the 5′ conserved segment of the pVS1 integron (1). The sequence between nucleotides 14 and 83 is known as the attI attachment site, where the gene cassettes are inserted (5). It corresponds to the junction between the 5′ conserved segment and the gene cassette, which extends from motif GTTGGGC (positions 77 to 83) to motif GTTAAAC (positions 789 to 795). The length of the aac(6′)-Iq gene cassette, as defined previously (12), extends from position 78 to position 789. At the 3′ end of the aac(6′)-Iq gene, there is a typical palindromic integron attC site (59-base element) 108 bp in length (positions 687 to 795) with an approximately 20-base similarity at each end corresponding to the known consensus sequences (11, 12). This 108-bp sequence possesses 93.6 and 70% similarity with the 109-bp attC site of the aac(6′)-Ip cassette and with the 119-bp attC site of the aac(6′)-Ia cassette (4), respectively, but it does not possess significant similarity, except at the terminal consensus sequences, with any other attC sites. This relevant homology suggests a common ancestor not only for the aac(6′)-Iq, -p, -a, and -i genes but also for the three attC sites. Unlike the other two cassettes, the aac(6′)-Ia cassette has an ORF with unknown function, named orfG, between the aac(6′)-Ia gene and its attC site (4). Finally, downstream of the aac(6′)-Iq cassette, the beginning of the ant(3")-Ia cassette was found at position 790, with the ATG start codon of the ant(3")-Ia gene at positions 799 to 801 (an alternative start codon, GTG, with a better ribosome-binding site, is found at positions 811 to 813).

The AAC(6′) enzyme family can be classified into three subfamilies (6, 16). The AAC(6′)-Iq subfamily contains the AAC(6′)-Iq protein (183 amino acids); the AAC(6′)-Ip protein (173 amino acids) found in Citrobacter freundii, K. pneumoniae, and other species (6) (84.39% identity); the AAC(6′)-Ia protein (185 amino acids) from Citrobacter diversus (17) (62.84% identity); and the AAC(6′)-Ii protein (182 amino acids) from Enterococcus faecium (3) (40.8% identity in a 182-amino-acid overlap) (Fig. 2). We have identified common regions in these proteins, including three motifs. One of these was a large motif, at the carboxyl end of the protein, where the homology was very significant [83.3% similarity and 76.6% identity between the AAC(6′)-Iq and AAC(6′)-Ii motifs].

FIG. 2.

Alignment of the AAC(6′)-Iq (183 amino acids), AAC(6′)-Ip (173 amino acids), AAC(6′)-Ia (185 amino acids), and AAC(6′)-Ii (185 amino acids) proteins. The three motifs identified in these proteins are indicated in boldface capital letters. Asterisks represent the stop codons.

In conclusion, we report here a newly identified gene, aac(6′)-Iq, located within a cassette. It seems that the mechanism involved in the amikacin resistance mediated by AAC(6′)-I activity in gram-negative species is due to a large variety of genes located within cassettes. These cassettes can be located in integrons (8, 12) or, rarely, in the transposon Tn1331 (18, 19).

Nucleotide sequence accession number.

The nucleotide sequence data will appear in the EMBL, GenBank, and DDBJ nucleotide sequence databases under accession no. AF047556.