A bla_VEB-1 Variant, bla_VEB-6, Associated with Repeated Elements in a Complex Genetic Structure

Abstract

bla_VEB-6 was found on the Proteus mirabilis chromosome in a context similar to those of bla_VEB-1a and bla_VEB-1b, in a truncated gene cassette flanked by 135-bp elements and duplications of the 3′-conserved segment of class 1 integrons. A linked aacA4-aadB-dfrA1-orfC cassette array includes components of Tn1331, illustrating the complex mosaicism of multiresistance regions.

Resistance to extended-spectrum β-lactams is an important clinical problem. In addition to the large TEM, SHV, and CTX families, several minor extended-spectrum β-lactamases have been identified (22), including VEB. All VEB enzymes identified to date (Table 1) are minor variants of VEB-1, which confers a high level of resistance to ceftazidime, cefotaxime, and aztreonam (26). bla_VEB genes have been identified in a variety of species of Enterobacteriaceae and in nonfermenting bacilli from Asia, Europe, the Middle East, Africa, and North and South America (22) (Table 2) on both plasmids and the chromosome.

TABLE 1.

Amino acid differences among VEB variants

VEB variant	Amino acid position for indicated type of peptidea					GenBank accession no.b	Reference or source
	Leader		Mature
	18	19	25	56	104
1	Ile	Val	Thr	Leu	Thr	AF010416	26
1a	Val	Val	Thr	Leu	Thr	AF324833	28
1b	Val	Glu	Thr	Leu	Thr	AF324834	28
2	Ile	Val	Ala	Leu	Thr	AY027870	8
3	Val	Val	Thr	Phe	Thr	AY536519	11
4	Ile	Val	Ala	Leu	Met	EF136375	1
5	Val	Glu	Thr	Leu	Met	EF420108
6	Val	Val	Ala	Leu	Met	EU259884	This work

^aPositions were numbered consecutively from the start codon. Amino acids differing from those in VEB-1 are in boldface.

^bAccession numbers in boldface type are listed at http://www.lahey.org/Studies.

TABLE 2.

Distribution and contexts of bla_VEB genes

blaVEB variant	Speciesa	Country	GenBank accession no.	Genetic contextd	Locationf	Reference(s) or sourceg
1	E. coli	Vietnam	AF205943b	qacI-aadB-aacA1/orfG-veb-aadB-arr2-cmlA-oxa10/aadA1	P	20 and 26
	K. pneumoniae	Vietnam	Not available	Not available	P	26
	P. mirabilis	Vietnam	AF220758	qacI-aadB-aacA1/orfG-veb-aadB-?-oxa10-?	P	18
	A. xylosoxidans	France	DQ393569	dfrA14-veb-aadB	P	23
	A. baumannii	France	CT025832c	veb-aadB-arr2-cmlA-oxa10/aadA1	C	7
	A. baumannii	Belgium	Not available	veb-aadB-arr2-cmlA-oxa10/aadA1	C	19
	E. coli	Canada	Not available	oxa10/aadA1-qacI-aadB-aacA1/orfG-veb-aadB-arr2-cmlA-oxa10/aadA1	P	27
	E. coli	Canada	Not available	Not available	NA	25
	P. aeruginosa	China	AY536743	veb-aadB-oxa10/aadA1	NA
	P. aeruginosa	China	DQ333895	Incomplete veb gene sequence	NA
	P. aeruginosa	Thailand	AF133699	veb-aadB	C	21
	P. aeruginosa	Thailand	AF078527	veb-aadB-arr2-cmlA-?	NA	34
	P. aeruginosa	Thailand	Not available	veb-aadB-arr2-cmlA-oxa10/aadA1	C	8
	E. coli	Thailand	Not available	Various cassette arrays were partially characterized by PCR	P	9
K. pneumoniae	Thailand
E. cloacae	Thailand
E. sakazakii	Thailand
	P. aeruginosa	Bulgaria	Not available	Not available	NA	31
1a	A. baumannii	Argentina	Not available	PCR suggests that veb is not located in a cassette array or typical Re/3′-CS structure	C	24
	P. aeruginosa	Bangladesh	DQ315788	Re	C	17
	P. aeruginosa	India	AY444815	Re	C	2
	P. aeruginosa	Kuwait	AF324833	PCR suggests that veb is the last cassette in an array	P	28
	P. aeruginosa	United Kingdom	Not available	Not available	NA	36
1b	P. aeruginosa	Kuwait	AF324834	PCR suggests that veb is the last cassette in an array	C	28
	C. freundii	Kuwait	Not available	Not available	NA	4
	P. mirabilis	Korea	Not available	Cassette array; aadB also present	C	13
	P. stuartii	Algeria	Not available	Re	P	3
2	P. aeruginosa	Thailand	AY027870	veb-aadB-arr2-cmlA-oxa10/aadA1	C	8
	P. aeruginosa	Thailand	Not available	?-veb-aadB-arr2-cmlA-oxa10/aadA1	C	8
3	E. cloacae	China	AY536519	No arraye	C	11
	P. aeruginosa	China	Not available	Not available	NA	12
	A. baumannii	Taiwan	Not available	Not in cassette array?		10
4	P. mirabilis	Spain	EF136375	veb gene sequence only	C	1
5	E. coli	United States	EF420108	Sequence includes start of veb cassette, suggesting a cassette array	NA
6	P. mirabilis	Australia	EU259884	Re	C	This work

^aK. pneumoniae, Klebsiella pneumoniae; A. baumannii, Acinetobacter baumannii; A. xylosoxidans, Achromobacter xylosoxidans; C. freundii, Citrobacter freundii; E. cloacae, Enterobacter cloacae; E. sakazakii, Enterobacter sakazakii.

^bGenBank accession no. AF205943 contains a longer sequence from the same strain as in AF010416 (Table 1).

^cPartial genome sequence of A. baumannii AYE. The same sequence is also available from the whole-genome sequence of AYE under GenBank accession no. CU459141.

^dContexts available from the GenBank entry or a published paper are listed. ?, the remainder of the array was not determined; Re, a truncated veb cassette is found in a structure with repeat elements (Fig. 1A).

^eThe sequence includes the first 7 bp of the bla_VEB-3 cassette, but the cassette is interrupted after 937 of 1,070 bp and the attC site is missing. A short region (34 bp) matching the right end of ISCR1 immediately follows, and then the left end of IS6100, probably explaining why an amplicon carrying bla_VEB-3 was not obtained using primers in the 5′-CS and 3′-CS (11).

^fP, bla_VEB gene found on a plasmid obtained by conjugation or electroporation; C, bla_VEB gene in a whole-genome sequence or found on the chromosome by hybridization (in boldface type) or presumed to be on the chromosome from the absence of transconjugants/transformants and/or plasmid DNA; NA, information not available.

^gReference 8 includes several other cassette arrays, but the bla_VEB variant is not specified.

bla_VEB-1 was first described in a gene cassette in a class 1 integron (26), and most other examples of bla_VEB genes where enough sequence data are available are also found in cassette arrays in class 1 integrons. These arrays are mostly related (Table 2), containing different combinations from a limited set of cassettes in different configurations, suggesting rearrangements mediated by both homologous and IntI-catalyzed recombination. The bla_VEB cassette is followed by the aadB cassette in almost all of these arrays, and the 5′-conserved segment (5′-CS) is interrupted in several cases by IS1999, with or without IS2000 (21).

bla_VEB-1a and bla_VEB-1b cassettes missing the first 7 bp have been found outside arrays in regions containing one or more 135-bp repeat elements (Re1, Re2, Re3) (Fig. 1A). In both Pseudomonas aeruginosa 10.2 (GenBank accession no. AY444815) (2) and Providencia stuartii B1 (3), Re1 and Re2 (71% identical) flank a region containing a truncated bla_VEB cassette, and Re2 and Re3 (98% identical to Re1) flank a region apparently largely derived from Tn1721 (GenBank accession no. X61367) that includes the tetA(A) gene truncated at both ends. This whole structure is flanked by duplications of the 3′-CS of class 1 integrons (3′-CS-a and 3′-CS-b) (Fig. 1A), and strain 10.2 has an extra copy of Re1. In P. aeruginosa TL-1 (GenBank accession no. DQ315788) (17), only Re1 is present, and bla_VEB-1a is linked to ISCR1, the aphA6 gene, and the aadA6-orfD cassette array. A strong promoter in Re1 was found to drive expression of bla_VEB-1a (2), and the Re and/or the 3′-CS duplications in the first two structures may provide an alternative means for the movement of bla_VEB genes by homologous recombination.

FIG. 1.

(A) Genetic structures containing bla_VEB genes and repeat elements. The species and isolate names are shown on the left. Insertion sequences are shown as pointed boxes labeled with the insertion sequence number, and Re1, Re2, and Re3 and their orientations are indicated. Gene cassettes are shown as open boxes, with small, filled boxes representing their attC sites. The extents of other resistance genes are shown, and the 5′-CS and 3′-CS of class 1 integrons and transposon fragments are labeled. Selected PCR products used for mapping are indicated by lines labeled with the primers used to amplify them. The region marked with an asterisk represents the first 1 to 111/114 bp of the 3′-CS. The structures for P. aeruginosa strains TL-1 and 10. 2 were drawn from sequences available under GenBank accession no. DQ315788 and AY444815, respectively, and P. stuartii B1 was drawn from the work of Aubert et al. (3). No additional information about flanking regions is available for 10.2 or B1. (B) Structure of Tn1331 and model for creation of a circular cassette containing aacA4 fused to the start of the bla_TEM-1a gene. Circular molecules are not shown to scale. Tn1331 was drawn from the sequence available under GenBank accession no. AF479774.

We have previously reported a novel bla_VEB variant, designated bla_VEB-6, in a Proteus mirabilis clinical isolate (JIE273) that is resistant to cefotaxime and ceftazidime (37). VEB-6 is essentially identical to VEB-4 (1), the A52G variation, predicting only a conservative amino acid substitution (Ile18Val) in the leader peptide.

Repeated attempts to transfer bla_VEB-6 from JIE273 to DH5αRf (a rifampin-resistant variant of Escherichia coli DH5α) by conjugation (38) and by electroporation with alkaline lysis preparations were unsuccessful. Whole-cell DNA from JIE273 was digested with I-CeuI (New England Biolabs, Ipswich, MA) (15), electrophoresed (for 36 h at a switch time of 5 to 60 s) (CHEF-DR II; Bio-Rad, Hercules, CA), and transferred to a Hybond-N+ membrane (GE Healthcare, Piscataway, NJ). Digoxigenin-labeled (DIG DNA labeling and detection kit; Roche, Penzberg, Germany) bla_VEB and 16S rRNA PCR amplicons (see Table 3) hybridized to the same (ca. 291-kb) I-CeuI fragment (data not shown), suggesting that bla_VEB-6 is located on the chromosome of JIE273.

TABLE 3.

Primers used in this work

Primer	Sequence (5′-3′)	Target	Reference or source
VEB-F	CGACTTCCATTTCCCGATGC	blaVEB	19
VEB-R	GGACTCTGCAACAAATACGC
hep58	TCATGGCTTGTTATGACTGT	5′-CS	35
hep59	GTAGGGCTTATTATGCACGC	3′-CS
Int1F	CAGTGGACATAAGCCTGTTC	intI1	14
3CS-F	CTATTGGTCTCGGTGTCG	3′-CS	6
3CS-R	ATCGTTCAGGTAGCCCAC	3′-CS
orf5-R	ACGAAGGTCTCCGCGAATGTC	3′-CS	This work
A	AGAGTTTGATCHTGGYTYAGA	16S rRNA	16
B	ACGGYTACCTTGTTACGACTT

PCR mapping (Table 3; Fig. 1A), including by long-range PCR (Expand Long Template PCR system; Roche, Mannheim, Germany), and sequencing revealed a genetic context most similar to bla_VEB-1b in P. stuartii B1. The aacA4-aadB-dfrA1-orfC array found adjacent to the 3′-CS-a was preceded by an unusual structure: the end of the 5′-CS is followed by the first 111 to 114 bp of the 3′-CS (Fig. 1) and then by 517 to 520 bp from the transposon Tn3 (three A residues could be derived from either the 3′-CS or Tn3). The Tn3 region includes the end of tnpR and the start of bla_TEM-1a and is followed by a short sequence (AAACAAAG) derived from the attI1 sites of class 1 integrons (Fig. 1A).

These components are all found in a different configuration in Tn1331 (32) and related transposons. Tn1331 (Fig. 1B) is a derivative of Tn3, with a 517- to 520-bp duplication at the end of the Tn3 tnpR gene and at the start of the bla_TEM-1a gene flanking the aacA4-aadA1-bla_OXA-9 cassette array and the first 111 to 114 bp of the 3′-CS (Fig. 1B). The sequence AAACAAAG derived from the attI1 site links the first 19 bp of bla_TEM to the start of the aacA4 cassette, providing both a promoter and a ribosome binding site for aacA4 expression and giving an AacA4 protein with a 17-amino-acid N-terminal extension (33). The aadA1 and bla_OXA-9 cassettes are separated by part of the attI1 site rather than a complete aadA1 attC site.

The configuration seen in pJIE273 may be derived from Tn1331 or a related transposon, but in Tn1331, the 111- to 114-bp 3′-CS region follows the cassette array, while in JIE273, it precedes the cassette array. This region would be brought to the start of the aacA4 cassette in a circular molecule created by homologous recombination between the duplicated regions of Tn1331 (Fig. 1B). Tn1331.2 has a duplication of the tnpR-aacA4-aadA1-bla_OXA-9 region, and experimentally observed conversion to Tn1331 (30) could also occur by loss of this circular molecule. Loss of aadA1-bla_OXA-9 from the circular molecule would effectively give an extended aacA4 cassette and could occur either before or after circularization. It has recently been demonstrated that aadA1-bla_OXA-9 can be readily excised from Tn1331 (29), and circles containing more than one cassette were found to separate into individual cassettes before insertion into class 1 integrons (5). The extended aacA4 cassette could then be inserted in front of an aadB-dfrA1-orfC array by IntI1-mediated recombination to give the configuration seen in pJIE273.

The only example of an aadB-dfrA1-orfC array currently in GenBank has three nucleotide differences from the corresponding part of the aacA4-aadB-dfrA1-orfC array in JIE273 and is found in P. aeruginosa TL-1 (GenBank accession no. DQ315789, where orfC is referred to as orfX), although it is not linked to bla_VEB-1a (17). It is also interesting that in P. aeruginosa TL-1, the aadA6-orfD array is linked to bla_VEB-1a, while a related array, aacA8-bla_OXA-2a-aacA7-aadA6-orfD, was found in P. aeruginosa 10.2 but could not be linked to bla_VEB-1a by PCR. These similarities suggest multiple recombination events. The association of bla_VEB-6 with a class 1 integron and components of both Tn1721 and Tn1331 illustrates recombinations and rearrangements of a limited set of different components which, in addition to the actions of individual mobile elements, all contribute to the mosaicism characteristic of many complex multiresistance regions.

The patient carrying JIE273 was born in and had recently traveled to India, and bla_VEB was not detected in other enterobacterial isolates from Sydney (38). P. aeruginosa strains 10.2 and TL-1 were also isolated from patients in India/Bangladesh (2, 17), and the similarities between the genetic contexts of bla_VEB-6 in JIE273 and bla_VEB-1a in these strains suggest that this structure, if not this bacterial strain, was acquired in that region.

Nucleotide sequence accession number.

The nucleotide sequence surrounding bla_VEB-6 has been added to GenBank accession no. EU259884.