Mutations in Bartonella bacilliformis gyrB confer resistance to coumermycin A1

Abstract

This study describes the first isolation and characterization of spontaneous mutants conferring natural resistance to an antibiotic for any Bartonella species. The Bartonella bacilliformis gyrB gene, which encodes the B subunit of DNA gyrase, was cloned and sequenced. The gyrB open reading frame (ORF) is 2,079 bp and encodes a deduced amino acid sequence of 692 residues, corresponding to a predicted protein of approximately 77.5 kDa. Sequence alignment indicates that B. bacilliformis GyrB is most similar to the GyrB protein from Bacillus subtilis (40.1% amino acid sequence identity) and that it contains the longest N-terminal tail (52 residues) of any GyrB characterized to date. The cloned B. bacilliformis gyrB was expressed in an Escherichia coli S30 cell extract and was able to functionally complement a temperature-sensitive E. coli Cour gyrB mutant (strain N4177). We isolated and characterized spontaneous mutants of B. bacilliformis resistant to coumermycin A1, an antibiotic that targets GyrB. Sequence analysis of gyrB from 12 Cour mutants of B. bacilliformis identified single nucleotide transitions at three separate loci in the ORF. The predicted amino acid substitutions resulting from these transitions are Gly to Ser at position 124 (Gly124-->Ser), Arg184-->Gln, and Thr214-->Ala or Thr214-->Ile, which are analogous to mutated residues found in previously characterized resistant gyrB genes from Borrelia burgdorferi, E. coli, Staphylococcus aureus, and Haloferax sp. The Cour mutants are three to five times more resistant to coumermycin A1 than the wild-type parental strain.

FIG. 1

Nucleotide and predicted amino acid sequence of B. bacilliformis gyrB. The nucleotide sequence of a 2,250-bp fragment containing the wild-type coumermycin A₁-sensitive B. bacilliformis gyrB is shown. Nucleotides within the 2,079-bp ORF are given in uppercase letters, and the deduced 692-residue amino acid sequence is shown below each corresponding codon. Putative consensus regulatory elements are indicated (−35, −10, ribosomal binding site [RBS]). The stop codon is marked with an asterisk. The three codons (and their corresponding amino acids) in which single nucleotide substitutions resulting in coumermycin A₁ resistance were found are boxed. The unusually long 52-residue N terminus is shown in boldface type. The predicted molecular mass of the mature protein is 77.5 kDa. The GenBank accession number for the gyrB gene is U82225.

FIG. 2

Multiple alignment of B. bacilliformis GyrB with B. subtilis, E. coli, and M. tuberculosis GyrB. Multiple alignment of B. bacilliformis GyrB (Barba) with B. subtilis GyrB (Bacsu), E. coli GyrB (Ecoli), and M. tuberculosis (Myctu) generated with CLUSTAL W 1.6 (45) and formatted with BOXSHADE 3.21 (18). Identical amino acid residues are shown as white on black, conserved residues are shown as black on grey, and introduced gaps are shown as dots. Note the unusual 53-residue N-terminal extension that is similar in length to the N terminus of M. tuberculosis GyrB. The first universally conserved residue (E. coli of Tyr5) is indicated by the arrowhead. GenBank accession numbers for these GyrB sequences are U82225 (B. bacilliformis) D26185 (B. subtilis), AE000447 (E. coli), and X78888 (M. tuberculosis).

FIG. 3

Detection of the gyrB gene in the B. bacilliformis chromosome by DNA hybridization. (A) Ethidium bromide-stained agarose gel (1%, wt/vol) containing λ HindIII size standards (lane 1), BamHI-digested chromosomal DNA of E. coli HB101 (lane 2), BamHI-digested chromosomal DNA of B. bacilliformis KC583 (lane 3), BamHI-digested chromosomal DNA of B. bacilliformis KC584 (lane 4), no DNA (lane 5), and 2,410-bp PCR fragment containing the entire B. bacilliformis gyrB ORF. (B) The corresponding autoradiograph following DNA hybridization with the described 2,410-bp PCR fragment labeled with [³²P]dCTP. The lanes are the same as those panel A. Note the hybridization signal in both B. bacilliformis strains (lanes 3 and 4).