To find the exact substrate specificities of three species of tripartite efflux systems of Pseudomonas aeruginosa, MexAB-OprM, MexCD-OprJ, and MexXY-OprM, we constructed a series of isogenic mutants, each of which constitutively overproduced one of the three efflux systems and lacked the other two, and their isogenic mutants, which lacked all these systems. Comparison of the susceptibilities of the constructed mutants to 52 antimicrobial agents belonging to various groups suggested the following substrate specificities. All of the efflux systems extrude a wide variety of antimicrobial agent groups, i.e., quinolones, macrolides, tetracyclines, lincomycin, chloramphenicol, most penicillins (all but carbenicillin and sulbenicillin), most cephems (all but cefsulodin and ceftazidime), meropenem, and S-4661, but none of them extrude polymyxin B or imipenem. Extrusion of aminoglycosides is specific to MexXY-OprM, and extrusion of a group of the ␤-lactams, i.e., carbenicillin, sulbenicillin, ceftazidime, moxalactam, and aztreonam, is specific to MexAB-OprM. Moreover, MexAB-OprM and MexCD-OprJ extrude novobiocin, cefsulodin, and flomoxef, while MexXY-OprM does not. These substrate specificities are distinct from those reported previously.Several tripartite efflux systems coded on the chromosome play important roles in the intrinsic and acquired resistance in Pseudomonas aeruginosa. Each system consists of a cytoplasmic membrane component of the resistance-nodulation-division family presumed to function as a transporter, an outer membrane component presumed to form channels, and a membrane fusion protein presumed to link the two membrane proteins for reviews, see references (13, 14, and 15). The MexAMexB-OprM efflux system (7, 16) contributes to both intrinsic and acquired resistance in P. aeruginosa, while the MexCMexD-OprJ (17) and MexE-MexF-OprN (6) efflux systems contribute only to the acquired resistance in this bacterium. Studies with mutants that overproduce or lack MexAB-OprM demonstrated that this efflux system extrudes quinolones, macrolides, tetracycline, chloramphenicol, novobiocin, and most ␤-lactams but not imipenem (7,8,22). Studies with mutants overproducing MexCD-OprJ demonstrated that this efflux system extrudes quinolones, erythromycin, tetracycline, chloramphenicol, and expanded-spectrum cephems such as cefpirome (9, 10). Furthermore, characterization of mutants lacking the mexA-mexB-oprM region demonstrated that the MexCD-OprJ efflux system extrudes ceftazidime and cefoperazone as well as cefpirome, while it does not extrude carbenicillin and aztreonam (3). Recently, Aires et al. (1) and our group (12) showed that MexX-MexY extrudes aminoglycosides, tetracycline, and erythromycin in cooperation with spontaneously expressed OprM, thereby contributing to the intrinsic resistance to these agents in P. aeruginosa. Our group also showed that no expression of MexXY is observed in wild-type strains but that the expression can be induced by subinhibitory concentrations of its substrates such as tetracycline a...

OprJ, overproduced in nfxB multidrug-resistant strains of Pseudomonas aeruginosa, and OprK, overproduced in the multidrug-resistant strain K385, were demonstrated to be immunologically cross-reactive using an OprJ-specific monoclonal antibody. Treatment of the purified proteins with trypsin or chymotrypsin yielded virtually indistinguishable digestion patterns, and the N-terminal sequence of two trypsin fragments was identical for both proteins, indicating that OprJ and OprK share identity. The N-terminal amino acid sequences were used to facilitate cloning of the oprJ gene on a 5kbp Kpnl fragment and a 10 kbp BamHl fragment. Nucleotide sequencing of portions of these fragments revealed that oprJ was the terminal gene in a putative three-gene operon, mexC-mexD-oprJ. The predicted mexC-mexD-oprJ gene products exhibit homology to the MexA-MexB-OprM components of the multidrug-resistance efflux pump of P. aeruginosa (43-46% identity). Consistent with an implied role for mexC-mexD-oprJ in drug efflux, the mexC-mexD-oprJ-hyperexpressing strain K385 showed reduced accumulation of a variety of antibiotics as compared with its parent strain, and this drug 'exclusion' was abrogated by energy inhibitors. The mexC and oprJ products are putative lipoproteins of a molecular mass of 40,707 and 51,742 Da, respectively, while mexD was predicted to encode a protein of 111 936 Da. Sequencing upstream of mexC revealed the presence of the nfxB gene transcribed divergently from the efflux genes. Overproduction of OprJ and the attendant multiple-antibiotic resistance of strain K385 was shown to result from a point mutation in nfxB, resulting in a H87-->R change in the predicted NfxB polypeptide. OprJ overproduction and multidrug resistance in K385 was reversed by the cloned nfxB gene, suggesting that nfxB encodes a repressor of mexC-mexD-oprJ expression. Consistent with this, the cloned nfxB gene repressed synthesis of a mexC-lacZ fusion in Escherichia coli. nfxB also repressed expression of a nfxB-lacZ fusion, indicating that NfxB negatively regulates its own expression. These data indicate that the multidrug resistance of nfxB strains is due to overexpression of an efflux operon, mexC-mexD-oprJ, encoding components of a second efflux pump in P. aeruginosa.

To test the possibility that MexX-MexY, a new set of efflux system components, is associated with OprM and contributes to intrinsic resistance in Pseudomonas aeruginosa, we constructed a series of isogenic mutants lacking mexXY and/or mexAB and/or oprM from a laboratory strain PAO1, and examined their susceptibilities to ofloxacin, tetracycline, erythromycin, gentamicin, and streptomycin. Loss of either MexXY or OprM from the MexAB-deficient mutant increased susceptibility to all agents tested, whereas loss of MexXY from the MexAB-OprM-deficient mutant caused no change in susceptibility. Introduction of an OprM expression plasmid decreased the susceptibility of the mexAB-oprM-deficient-/mexXY-maintaining mutant, yet caused no change in the susceptibility of a mexAB-oprM-and mexXY-deficient double mutant. Immunoblot analysis using anti-MexX polyclonal rabbit serum generated against synthetic oligopeptides detected expression of MexX in the PAO1 cells grown in medium containing tetracycline, erythromycin, or gentamicin, although expression of MexX was undetectable in the cells incubated in medium without any agent. These results suggest that MexXY induced by these agents is functionally associated with spontaneously expressed OprM and contributes to the intrinsic resistance to these agents.A variety of multicomponent efflux systems are coded on the chromosomes of gram-negative bacteria and contribute to intrinsic and acquired resistance against antimicrobial agents, disinfectants, organic solvents, and heavy metals (17)(18)(19)(20)24). Pseudomonas aeruginosa is a clinically significant pathogen exhibiting highly intrinsic resistance to various antimicrobial agents. One of the mechanisms contributing to its intrinsic resistance is a spontaneous expression of the efflux system MexA-MexB-OprM encoded on a mexA-mexB-oprM operon of the chromosome of P. aeruginosa (10). This system energydependently extrudes many antimicrobial agents from the cell interior in cooperation with the periplasmic, inner membrane, and outer membrane components organized through the two membranes. While the disruption of each component gene of the mexA-mexB-oprM operon increases the susceptibility to many antimicrobial agents, the disruption of oprM increases the susceptibility to a greater extent than the disruption of mexA or mexB (5,21,30,31). Due to the presence of a weak promoter in the mexB gene upstream of oprM gene, the polar effect from the disruptions of mexA and mexB does not entirely suppress the expression of the oprM gene (31). Thus, OprM can contribute to the intrinsic resistance by cooperation with unknown periplasmic and inner membrane components. Recently, mexG-mexH (GenBank accession no. AF073776 To test the possibility that MexXY is associated with OprM and contributes to intrinsic resistance, we constructed a series of isogenic mutants lacking mexXY and/or mexAB and/or oprM from laboratory strain PAO1 and compared their susceptibilities to antimicrobial agents. We also showed that the expression of MexXY is induced by expo...

Active surveillance of Vibrio parahaemolyticus infection among hospitalized patients in Calcutta, India, was initiated in January 1994. The incidence of cases of V. parahaemolyticus infection suddenly increased in February 1996 and has remained high since then. One hundred thirty-four strains of V. parahaemolyticus isolated from January 1994 to August 1996 were examined for serovar, the presence of the thermostable direct hemolysin gene (tdh) and tdh-related hemolysin genes (trh1 and trh2), production of urease, and antibiogram. Strains of the O3:K6 serovar appeared for the first time in February 1996. The O3:K6 serovar strains accounted for 50 to 80% of the strains isolated during the high-incidence period (February to August 1996). All of the serovar O3:K6 strains carried the tdh gene but not the trh genes and did not produce urease. All of the isolates except two were sensitive to all of the antibiotics tested. These and the results of analysis by an arbitrarily primed PCR method indicated that the O3:K6 serovar strains belong to a unique clone. When the O3:K6 serovar strains, isolated from travelers arriving in Japan from Southeast Asian countries, were compared by the arbitrarily primed PCR method, the strains isolated between 1982 and 1993 were distinct from Calcutta O3:K6 while the strains isolated in 1995 and 1996 were indistinguishable from the Calcutta O3:K6 strains. The results suggest that this unique O3:K6 clone may have become prevalent not only in Calcutta but also in Southeast Asian countries very recently. Not only the O3:K6 strains but also the non-O3:K6, tdh-bearing strains isolated in 1996 produced thermostable direct hemolysin at high levels, and thus the level of hemolysin produced does not appear to have influenced the high incidence of serovar O3:K6 strains.