Chromosome-encoded narrow-spectrum Ambler class A beta-lactamase GIL-1 from Citrobacter gillenii

Abstract

A novel beta-lactamase gene was cloned from the whole-cell DNA of an enterobacterial Citrobacter gillenii reference strain that displayed a weak narrow-spectrum beta-lactam-resistant phenotype and was expressed in Escherichia coli. It encoded a clavulanic acid-inhibited Ambler class A beta-lactamase, GIL-1, with a pI value of 7.5 and a molecular mass of ca. 29 kDa. GIL-1 had the highest percent amino acid sequence identity with TEM-1 and SHV-1, 77%, and 67%, respectively, and only 46%, 31%, and 32% amino acid sequence identity with CKO-1 (C. koseri), CdiA1 (C. diversus), and SED-1 (C. sedlaki), respectively. The substrate profile of the purified GIL-1 was similar to that of beta-lactamases TEM-1 and SHV-1. The blaGIL-1 gene was chromosomally located, as revealed by I-CeuI experiments, and was constitutively expressed at a low level in C. gillenii. No gene homologous to the regulatory ampR genes of chromosomal class C beta-lactamases was found upstream of the blaGIL-1 gene, which fits the noninducibility of beta-lactamase expression in C. gillenii. Rapid amplification of DNA 5' ends analysis of the promoter region revealed putative promoter sequences that diverge from what has been identified as the consensus sequence in E. coli. The blaGIL-1 gene was part of a 5.5-kb DNA fragment bracketed by a 9-bp duplication and inserted between the d-lactate dehydrogenase gene and the ydbH genes; this DNA fragment was absent in other Citrobacter species. This work further illustrates the heterogeneity of beta-lactamases in Citrobacter spp., which may indicate that the variability of Citrobacter species is greater than expected.

FIG. 1.

(A) Schematic representation of recombinant clone pGIL-1 containing the bla_GIL-1-coding region from C. gillenii CIP 106783^T. The orientations of the genes are indicated with arrows, and a solid line indicates the cloning vector. Plasmid-located promoter pLacUV5 is indicated by an arrow. (B) Structure of the promoter of the bla_GIL-1 gene. The +1 sign indicates the mRNA transcription start site, as determined by the 5′-RACE experiment. The deduced −10 and −35 regions of the promoter, the putative ribosomal binding site (RBS), and the translational initiation codon (ATG) are represented in boldface and capital letters. (C) Deduced amino acid sequence of GIL-1. The vertical arrow indicates the cleavage site of the leader peptide. The numbering is according to Ambler et al. (1). Structural elements characteristic of class A β-lactamases are underlined.

FIG. 2.

Phylogenetic tree construction according to parsimony analysis (27) and on the basis of class A β-lactamases. Branch lengths are to scale and are proportional to the number of nucleotides or amino acid changes. The amino acid sequences of the different β-lactamases were recovered from GenBank databases. The percent amino acid identity of each β-lactamase with GIL-1 is indicated into parentheses.

FIG. 3.

Localization of bla_GIL-1 in I-CeuI-generated fragments of C. gillenii CIP 106783^T separated by PFGE. Lane A, I-CeuI restriction pattern of E. coli DH10B, used as a control; lane B, I-CeuI restriction pattern of C. gillenii CIP 106783^T; lane C, hybridization of the I-CeuI restriction pattern of C. gillenii CIP 106783^T with a probe specific for the 16S rRNA gene; lane D, hybridization of the I-CeuI restriction pattern of C. gillenii CIP 106783^T with a probe specific for the bla_GIL-1 gene.

FIG. 4.

(A) Schematic representation of the DNA sequence present between the LDH and ydbH genes in C. gillenii CIP 106783^T (CG), C. braakii CIP 104554^T (CB), and C. freundii P478 (CF). Boxes represent genes, and their orientations are indicated with arrows. Gray triangles represent the DNA sequences that are absent in C. freundii or C. braakii. A 9-bp duplication present on each side of the inserted 5.5-kb DNA sequence is also indicated with shading in the lower part. The small arrows above the genes represent the primers used for PCR amplification. The DNA sequences on both sides of the putative insertion site are provided: CGL, the sequence derived from the left boundary (ldh primer); CGR, the sequence derived from the right boundary (ydbH primer). Identical positions are indicated by asterisks. (B) Schematic representation of the DNA sequence present between the fumarate reductase frdD and the lipocalin blc genes in C. gillenii CIP 106783^T (CG), C. koseri CIP 105177 (CB), and C. murliniae CIP 104556^T. Boxes represent genes, and their orientations are indicated with arrows. A 12-bp imperfect duplication present on each side of the ampC/ampR operon is shaded in the lower part. The small arrows above the genes represent the primers used for PCR amplification. The DNA sequences on both sides of the putative insertion site are provided: CML, the sequence derived from the left boundary (frd primer); CMR the sequence derived from the right boundary (blc primer). Identical positions are indicated by asterisks.

FIG. 5.

Phylogenetic tree construction according to parsimony (27), based on 16S rRNA sequence comparison. Branch lengths are to scale and are proportional to the number of nucleotide changes. The nucleotide sequences were from Warren et al. (30).