Antimicrobial activity of CHIR-090, an inhibitor of lipopolysaccharide biosynthesis, against the Burkholderia cepacia complex.

A striking characteristic of cystic fibrosis (CF) is susceptibility to life-limiting bacterial infections of the respiratory tract (13, 26). Members of the Burkholderia cepacia complex are a particular cause of anxiety to CF individuals (8, 14, 23) since they display high resistance to antibiotics and biocides (3, 32), possibly linked to their relatively large ( 8to 9-Mbp) genomes (16). At present, the B. cepacia complex consists of 17 species, the majority of which have been recovered from CF patients (4, 10, 11, 36, 37, 40). As an aid to B. cepacia complex studies, two panels of B. cepacia complex reference isolates (Table 1) have been assembled that include B. multivorans and B. cenocepacia strains, the two most prevalent species responsible for CF infections (11, 22). A promising target for the development of new antibiotics against multiresistant Gramnegative pathogens are the lipopolysaccharide (LPS) biosynthetic and modification pathways (27, 29, 30, 39). LPS (also known as endotoxin) is the major component of the bacterial outer membrane. Nine enzymes are required to form the basic core Kdo2-lipid A, and the first six of these enzymes are essential in Escherichia coli. Also, we discovered that a putative locus involved in Ara4N synthesis and LPS modification was essential to B. cenocepacia (28). The metaldependent UDP-[3-O-(R-3-hydroxymyristoyl)]-N-acetylglucosamine deacetylase (LpxC) (Fig. 1) that catalyzes the second step in lipid A biosynthesis (18, 41) has been targeted and, like many metalloenzymes, can be inhibited by hydroxamate-containing compounds (7, 9, 17). The synthetic antibiotic CHIR-090 (Fig. 1) (N-aroyl-L-threonine hydroxamic acid [international patent WO 2004/062601 A2]) (1) has been shown to be a slow, tight-binding inhibitor of the LpxCs from different species (5, 6, 24) and displayed good antimicrobial activity against several Gram-negative bacteria (6). We determined the activity of CHIR-090 against the B. cepacia complex (Table 1) initially by disc diffusion growth inhibition assay according to published guidelines (2). Individual isolates displayed remarkable differences in susceptibility to CHIR-090, even within a single species. Interestingly, CHIR-090 was active against all representative strains of B. multivorans, B. vietnamiensis, B. dolosa, and B. ambifaria. We prepared a panel of clinically relevant B. multivorans strains for MIC determination and included E. coli and Pseudomonas aeruginosa (Table 2). The CHIR-090 MICs were strain dependent, and the values obtained ranged from 0.1 to 100 g/ml. The LPSs from a number of Burkholderia species display unique structural and inflammatory properties (12, 33); however, there appears to be no correlation between CHIR-090 activity and the LPS profiles of individual strains. For example, CHIR-090 is not active against smooth LPS strain B. cenocepacia K56-2 or its deep-rough LPS derivative SAL1 (20). A BLAST sequence analysis of the Burkholderia genomes (Burkholderia Genome Database) revealed that the LpxC genes are highly conserved and display high sequence homology to LpxCs from P. aeruginosa and E. coli; thus, the reason(s) why CHIR-090 is not active against certain members of the B. cepacia complex remains to be clarified. Our study reports the potential of therapeutic agents against Burkholderia targeted at LPS biosynthesis. Such agents may, possibly in combination TABLE 1. Inhibition of strains of Burkholderia genomovars I to IX by CHIR-090