Correctly Identifying the Streptothricin Resistance Gene Cassette

If analysis of the gene cassettes located in class 1 or class 2 integrons found in surveys of antibiotic resistance genes is to be useful in tracking the prevalence and movement of resistance genes, the cassettes must be accurately identified. This can be simply achieved by citing the GenBank accession numbers for the relevant reference cassettes listed in available compilations (e.g., references 6 and 9). However, confusion is arising around the single cassette known to include a streptothricin acetyltransferase gene (GenBank accession no. X15995), which was first identified in transposons Tn1825 and Tn1826 isolated from bacteria resistant to streptothricin (4, 12). These transposons are close relatives of Tn7 (14), and the sat cassette was subsequently identified in Tn7 (3, 11, 13), where it confers only modest levels of resistance to the antibiotic (11, 13). Thus, Tn7, Tn1825, and Tn1826 are all class 2 integrons that appear to differ only in the composition of the cassette array (see reference 9) (Fig. 1A).

FIG. 1.

Maps of integrons containing the sat2 or estX cassette. Thick black lines represent the 5′ CS (intI region), and the attI1 site is represented by an open box with a number indicating the integron class. Each cassette is shown as a narrow open or hatched box with an adjacent filled box representing the 59-be (59-base element). Labeled arrows indicate the positions of the genes. The positions of the TAG stop codon at the end of estX and the mutant version (TAT) in Tn1825 are indicated. The maps of Tn7, Tn1825, and Tn1826 were drawn using sequences from GenBank accession nos. AP002527 (Tn7), X56185 (Tn1825), and X15995 plus additional map information from reference 14 for Tn1826. Maps for the remaining cassette arrays are from the GenBank accession numbers shown. The estX gene/cassette is annotated as sat in AB121039 and AB161461-3 but has been corrected in AY090896; sat2 is named sat1 in reference 1a and AB161461-3.

The confusion has arisen because proteins of different sizes were produced by the cloned streptothricin resistance determinants of Tn1825 and Tn1826, and the genes were named sat1 and sat2 (15). Tn1826, in which the sat cassette is first in the cassette array, produces the shorter Sat-2 protein, which is sufficient to confer streptothricin resistance. In Tn1825 the sat cassette is preceded by another previously unnamed cassette (estX in Fig. 1), and the protein produced by this cassette is fused to Sat-2, forming the longer Sat-1 (15). Further sequences of cassette arrays equivalent to that of Tn1825 are now available (GenBank accession nos. AB161461 to AB161463[1]), and it appears that the protein fusion, which may increase the level of resistance, results from a mutation in the termination codon of the estX gene of Tn1825, leading to translational readthrough (Fig. 1A). Thus, sat2 is the only known cassette-associated streptothricin resistance gene, and sat1 is an estX-sat2 fusion.

Recently, the first cassette of Tn1825 has been found in class 1 integrons in GenBank accession nos. AY090896 (2) and AB121039 (1) (see Fig. 1B) but has been recorded as a sat cassette. In fact, the predicted 280-amino-acid polypeptide is not an acetyltransferase but is over 40% identical to several predicted proteins annotated as putative esterases or hydrolases of the α/β fold superfamily (8). These are encoded in the genomes of Mesorhizobium loti, Bacillus cereus, Bacillus anthracis, and Bacillus thuringiensis and in a Sinorhizobium meliloti plasmid (Fig. 2). We have therefore named the cassette gene estX for esterase X. EstX is 42% identical to a protein encoded in multidrug resistance plasmids from Escherichia coli (7) and over 30% identical to RdmC, DauP, and DnrP from Streptomyces species. The latter are methylesterases that catalyze one of the final “tailoring” reactions in the biosynthesis of rhodomycin, daunomycin, and daunorubicin, respectively (see reference 5). Whether EstX can inactivate any known antibiotic remains to be established.

FIG. 2.

Alignment of proteins related to estX. Residues conserved in all sequences are shown as black on white. The residues that constitute the catalytic triad of RdmC (Ser102, Asp248, and His276) are indicated by asterisks in the top line, and Ser102 lies at the center of the G-X-S-X-G motif typical of a serine hydrolase active site (5). The proteins are translated from the following sources: RdmC (GenBank accession no. U10405); DauP (L35154); SMa, product of SMa1327 from Sinorhizobium meliloti plasmid pSymA (AE007259); mll, product of mll3776 from Mesorhizobium loti (AP003002); BA, product of BA2738 from Bacillus anthracis (AE017032), representing all of the homologous (>88% identical) Bacillus proteins; pTZ, “RdmC homologue” from pTZ3721 (AB020531) and pTZ3723 (AB038654); EstX, cassette-encoded protein (AY090896). The completely conserved residues shown are also conserved in all additional homologues (not shown): DnrP (L40425), 94.6% identical to DauP; products of BCE2771 (AE017273) and BC2784 (AE017007) from B. cereus; EstA (AE017355) from B. thuringiensis; and all other variants of EstX (AB161461 to AB161463 and AB121039).

Reports of the presence of a sat cassette should be treated with caution, unless the standard sat2 cassette (GenBank accession no. X15995) is cited (e.g., in reference 10) or an accompanying database entry is available for confirmation (e.g., AB186119). An approximately 0.65-kb cassette array PCR product obtained with standard 5′-conserved segment (CS) and 3′-CS primers (e.g., references 16 and 17) is, however, consistent with the presence of the sat2 cassette.