Generation of the blaCTX-M-116 hybrid gene.

Members of the globally important CTX-M group of extended-spectrum -lactamases fall into at least six subgroups (CTX-M-1, -2, -8, -9, -25, and KLUC [ 90% amino acid identity between subgroups]), with members of each subgroup differing by one or a few amino acid changes (1). Ancestral members of the different blaCTX-M subgroups appear to have been captured from the chromosomes of different Kluyvera species, with subsequent mutations leading to variants with advantageous phenotypes (1). Due to their relatedness, blaCTX-M genes could also evolve by homologous recombination, and the sequences of three hybrids of blaCTX-M-1 group and blaCTX-M-9 group genes have been previously reported (1). The recently described CTX-M-116 was proposed to be composed of parts of the CTX-M-1 group proteins CTX-M-23 and CTX-M-22 (2). Fursova et al. suggested that recombination to create blaCTX-M-116 occurred at a “putative 7-bp site of recombination (GTTAAAT)” referred to as a “core site; attC” (positions 436 to 442) in the accompanying GenBank entry (JF966749); i.e., they seem to suggest that recombination has occurred at a sequence found in the middle of blaCTX-M-1 group genes that happens to fit the integron/gene cassette-type core site consensus sequence (GTTRRRY) but which is not a full attC site that would be the normal substrate for intI-mediated recombination. We agree that blaCTX-M-116 could be a hybrid of blaCTX-M-23, which has a nucleotide sequence distinct from those of other blaCTX-M-1 group genes, and blaCTX-M-22 but believe that this would have occurred by homologous recombination, as for blaCTX-M-64 (3). The site of the crossover cannot be precisely defined, due to identity between the blaCTX-M-23 and blaCTX-M-22 sequences in the relevant region, but would lie between position 240, where the match to blaCTX-M-22 begins, and position 312, where the match to blaCTX-M-23 ends (see Fig. 1 in reference 2). Fursova et al. go on to suggest that other blaCTX-M-1 group genes (and some blaCTX-M-8 and blaCTX-M-25 group genes) are also made up of 5= and 3= “moieties” from different genes delineated by the GTTAAAT sequence (or GTTGAGT for blaCTX-M-8 and blaCTX-M-25 group genes). This boundary is artificial, and we believe that this analysis is misleading. The sequence of the blaCTX-M-3 gene from GenBank accession no. Y10278 is proposed to include the 5= moiety of blaCTX-M-15. As blaCTX-M genes (and other bla genes) are named from the protein sequence and different nucleotide sequences can encode the same protein, this gene was designated blaCTX-M-3a (4). blaCTX-M-3a and flanking regions are 100% identical to a region from the Kluyvera ascobata chromosome (AJ632119), and this gene has been proposed as a progenitor of other blaCTX-M-1 group genes (5). blaCTX-M-15 could then be derived from blaCTX-M-3a by a single mutation leading to the D240G substitution that results in increased resistance to ceftazidime (6). The sequences of other genes within each group listed in Table 1 of reference 2 (unmarked, own 5= and 3=moieties; *, 5=moiety contributed by another blaCTX-M gene; **, both 5= and 3= moieties contributed by other blaCTX-M genes) could similarly be derived from blaCTX-M-3a or blaCTX-M-15 by a single nucleotide change resulting in one amino acid substitution (Fig. 1).