Characterization of poxB, a chromosomal-encoded Pseudomonas aeruginosa oxacillinase

Gene 358 (2005) 82 – 92 www.elsevier.com/locate/gene

Characterization of poxB, a chromosomal-encoded Pseudomonas aeruginosa oxacillinase Kok-Fai Kong a, Suriya R. Jayawardena a, Aimee del Puerto a, Lutz Wiehlmann b, Ulrike Laabs b, Burkhard Tu¨mmler b, Kalai Mathee a,* a

Department of Biological Sciences, Florida International University, University Park, Miami, FL 33199, USA b Klinische Forschergruppe, OE 6710, Medizinische Hochschule Hannover, D-30625 Hannover, Germany Received 2 November 2004; received in revised form 11 April 2005; accepted 17 May 2005 Available online 24 August 2005 Received by D.L. Court

Abstract Pseudomonas aeruginosa is the major pathogen associated with morbidity and mortality of patients with cystic fibrosis. One of the reasons for the failure of h-lactam antibiotic regimens appears to be mediated by de-regulation of the ampC gene, encoding the chromosomal Ambler’s Class C h-lactamase. Currently, the AmpC is the only known chromosomal h-lactamase whose expression is regulated by a transcriptional regulator, AmpR. We generated an ampC mutation in the prototypic P. aeruginosa strain PAO1. The mutation in ampC did not abolish the h-lactamase activity entirely suggesting the expression of yet another unreported h-lactamase. Our genomic analysis revealed the presence of an open reading frame encoding a protein with high homology to the Class D h-lactamases, commonly known as oxacillinases. The gene was named poxB for Pseudomonas oxacillinase. Cloning and expression of poxB in Escherichia coli conferred hlactam resistance to the host. We detected the presence of poxB both in clinical and environmental isolates. Our studies show that P. aeruginosa possesses two h-lactamases, AmpC and PoxB, which contribute to its resistance against a wide spectrum of h-lactam antibiotics. D 2005 Elsevier B.V. All rights reserved. Keywords: Antibiotic resistance; ampc; Chromosomal Class D h-lactamase

1. Introduction Pseudomonas aeruginosa is a dominant pathogen in patients with cystic fibrosis (CF), chronic obstructive pulmonary disease, severe burns and patients in intensive care units. The presence of this organism is associated with a high risk of morbidity and mortality in CF. The current treatment against P. aeruginosa includes a combination of h-lactam antibiotics and aminoglycosides (Frederiksen et al., 1996). However, resistance to h-lactam antibiotics is common. P. aeruginosa employs several mechanisms to neutralize h-lactams including multidrug efflux pumps, Abbreviations: bp, base pair(s); CF, Cystic Fibrosis; E, Escherichia; kb, kilobase(s); Gm, Gentamycin; ORF, open reading frame; P, Pseudomonas; PCR, Polymerase Chain Reaction. * Corresponding author. Tel.: +1 305 348 1261; fax: +1 305 348 1986. E-mail address: [email protected] (K. Mathee). 0378-1119/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.gene.2005.05.027

restricted membrane permeability, alteration in the penicillin-binding proteins and expression of h-lactam hydrolyzing enzymes h-lactamases (Lambert, 2002). However, resistance to h-lactam antibiotics due to the expression of hlactamases appears to be by far the most prominent mechanism in CF strains. According to the Nomenclature Committee of the International Union of Biochemistry, h-lactamases are a group of diverse enzymes that hydrolyze amides, amidines and other C– N bonds. Owing to the complexity of these enzymes, four major classification schemes were proposed by Richmond and Sykes, Ambler, Mitsuhashi and Inoue and Bush et al. (Bush et al., 1995). Among these, the Ambler system has been widely adopted. Ambler’s scheme differentiates h-lactamases by primary amino acid sequences into four classes, Classes A, B, C and D (Ambler, 1980). All four classes have been documented in P. aeruginosa.

K.-F. Kong et al. / Gene 358 (2005) 82 – 92

Ambler molecular class A was initially described in Gram-positive plasmids. However, plasmid-, transposonand chromosomally mediated h-lactamases (VHS, PER, TEM and SHV) in Gram-negative bacteria are not uncommon (Couture et al., 1992). This class of enzymes, dubbed penicillinases, exhibits the highest degree of sequence variability and kinetic properties. Plasmid-encoded Ambler class A h-lactamases identified in P. aeruginosa are active against carbenicillin and are hence referred to as CARB or Pseudomonas-specific h-lactamases (Lachapelle et al., 1991). In addition, VHS-, PER- and TEM-derived hlactamases have also been reported in P. aeruginosa (Weldhagen et al., 2003). The Ambler Class B contains a small number of Zn2+ metallo-h-lactamases, whereby their activities could be inhibited by EDTA. IMP-1 is the first metallo-h-lactamase described in P. aeruginosa (Watanabe et al., 1991). Its gene, bla IMP, appears to be dispersed among P. aeruginosa and other gram-negative rods in Japan (Senda et al., 1996). An integron-borne metallo-h-lactamase gene, bla VIM, which was originally described in P. aeruginosa isolated in Italy, gives rise to the resistance of meropenem and imipenem (Tsakris et al., 2000). The Ambler Class C enzymes are active against cephalosporins, hence they are known as cephalosporinases. They are chromosome-encoded and synthesized by most Gram-negative bacteria. The known sequences of these enzymes are highly conserved. The P. aeruginosa Class C cephalosporin-hydrolyzing chromosomal h-lactamase is encoded by ampC (Lodge et al., 1990). The inducible expression of AmpC is regulated by an upstream divergently transcribed gene, ampR that encodes a transcriptional regulator (Lodge et al., 1990). Due to the structural similarity between Class A and Class D enzymes, Couture et al. suggested the use of 26 conserved amino acid residues as the class D standard numbering scheme (DBL numeration) (Couture et al., 1992). This group of h-lactamases is called oxacillinases due to their ability to degrade isoxazolyl h-lactams such as oxacillin and methicillin (Dale and Smith, 1972). Clavulanic acid, on the other hand, serves as a good inhibitor for these enzymes (Medeiros et al., 1985). Plasmid- and transposon-mediated oxacillin-hydrolyzing h-lactamases in P. aeruginosa are common yet complex. A nosocomial outbreak with a P. aeruginosa extendedspectrum of Class D h-lactamases (ESBLs) has been reported (Poirel et al., 2002) and more episodes are expected to arise in the near future. The Ambler Class C enzyme, ampC, has been cloned and sequenced from the prototypic P. aeruginosa PAO1 (Lodge et al., 1990). Here we report the characterization of the ampC mutant that led to the discovery of a second chromosomally encoded h-lactamase belonging to Class D oxacillinases. The identified open reading frame (ORF) PA5514 is named poxB for P. aeruginosa oxacillinase.

83

2. Materials and methods 2.1. Bacterial strains, plasmids and media Escherichia coli used in this study were DH5a (F f80dlacZDM15 D(lacZYA-argF) U169 deoR recA1 endA1 hsdR17(r k , m k+ ) phoA supE44 k-thi-1 gyrA96 relA1; New England Biolabs) and TOP10F’ (F’ {lacI q , Tn10(TetR)} mcrA D(mrr-hsdRMS-mcrBC) f80dlacZDM15 DlacX74 deoR recA1 araD139 D(ara-leu)7697 galU galK rpsL (StrR) endA1 nupG; Invitrogen). P. aeruginosa strains and plasmids constructed in this study were shown in Table 1. Escherichia coli and P. aeruginosa were routinely cultured in Luria-Bertani medium (10 g tryptone, 5 g yeast extract, 5 g NaCl, per liter). Pseudomonas Isolation Agar (Difco) was used in triparental mating experiments. Antibiotics were used at the following concentrations (per milliliter unless indicated otherwise): ampicillin at 50 Ag, tetracycline at 20 Ag, gentamycin at 15 Ag for E. coli and carbenicillin at 300 Ag, gentamycin at 300 Ag and tetracycline at 60 Ag for P. aeruginosa. For induction, 25 Ag and 500 Ag of benzylpenicillin were used for E. coli and P. aeruginosa, respectively. 2.2. DNA manipulations All molecular techniques were performed according to standard protocols (Ausubel et al., 1999). 2.3. Insertional inactivation of ampC gene A 2150-bp ampC fragment was amplified by polymerase chain reaction (PCR) using SBJ05ampCFor (5V-GGAATTCTGAGGCCGCGCGGCAGACGCTTGAACA-3V) and SBJ06ampCRev (5V-CGGGATCCAACCCCGGCGCGGTGGCCAGTCCCGCCAA-3V) primers with flanking EcoRI and BamHI sites (the italicized portion in the sequence of the primers), respectively (Fig. 1). The PCR product was ligated to pCRII-TOPO vector (Invitrogen, CA), generating pSJ02. A 899-bp SalI fragment containing the gentamycin cassette, aacCI, was retrieved from pUCGm (Accession No: U04610) and inserted into the SalI site of ampC. This disrupted the reading frame of ampC in pSJ02. Then, a blunt-ended 3049-bp EcoRI fragment containing the ampC::aacCI was ligated to the SmaI-cut suicide vector, pEX100T (Accession No: U17500), to yield pSJ08. This plasmid was conjugated into P. aeruginosa PAO1 with a helper strain harboring pRK2013 (KmR; ColE1ori-Tra (RK2)+: Figurski and Helinski, 1979). The merodiploids resulting from homologous recombination were selected with PIA containing gentamycin. GmR colonies were then screened for gentamycin resistance and carbenicillin sensitivity by replica plating for the loss of plasmid. The insertion was confirmed by PCR and restriction analysis on the PCR product (data not shown). The PAO1 isogenic

84

K.-F. Kong et al. / Gene 358 (2005) 82 – 92

Table 1 P. aeruginosa strains used and plasmids constructed in this study Strain/Plasmid

Genotype/Source

Reference

Pseudomonas aeruginosa PAO1 Wild-type PKM200 PAOampC::aacCI; GmR 892 poxB1; CF sputum, Hannover, Germany, 1983 K9 poxB2; CF sputum, Husum, Germany, 1985 G7 poxB3; CF sputum, Stade, Germany, 1986 SG31 poxB4; river, Mu¨hlheim, Germany, 1993 DM poxB5, Alg+; CF sputum, Hamburg, Germany, 1984 63741 poxB6; burn wound, intensive care unit, Hannover, Germany, 1990 DSM 1128 poxB7; ear infection, United States, 1980 ATCC 15691 poxB8; PAT, wound, Melbourne, Australia, 1952 ATCC 10145 poxB9; neotype, type strain, Prague, Czech Republic, <1960 ATCC 33818 poxB10; mushroom Agaricus bisporus H2 poxB11; catheter, ward for infectious disease ZW 41 poxB12; CF throat swab, Verona, Italy, 1997 ZW 43 poxB13; CF throat swab, Genoa, Italy, 1997 ZW 49 poxB14; CF throat swab, Verona, Italy, 1997 ZW 77 poxB15, Alg+; CF throat swab, London, UK, 1997 ZW 83 poxB16; CF throat swab, London, UK, 1997 ZW 88 poxB17; CF throat swab, London, UK, 1997 ZW 92 poxB18; CF throat swab, Marseilles, France, 1997 BST 1 poxB19; CF throat swab, Hannover, Germany, 1985 KB 1 poxB20; CF throat swab, Sarstedt, Germany, 1985 MF 6 poxB21; CF throat swab, Bremen, Germany, 1987 641 HD 11/m1 poxB22; water, Mu¨hlheim, Germany, 1992 PT 20 poxB23; water, Mu¨hlheim, Germany, 1992 ATCC 33364 poxB24; Pseudomonas aeruginosa (Schroeter) Migula, serotype 17 ATCC 15522 poxB25; Pseudomonas aeruginosa (Schroeter) Migula, soil ATCC 21472 poxB26; Pseudomonas aeruginosa (Schroeter) Migula, soil from an oil field DSM 1253 poxB27; unknown origin Plasmids pSJ02 pSJ04 pSJ08 pKKF0675 pKKF0677 pKKF0679

ApR; pCRII-TOPO derivative with a 2150-bp fragment containing ampC TcR; pME6030 derivative with EcoRI-digested ampC from pSJ02 ApR, GmR; pEX100T derivative with ampC::aacCI KmR; pCRII-TOPO derivative with 1770-bp fragment containing poxAB operon KmR; pCRII-TOPO derivative with 642-bp fragment containing poxB KmR; pCRII-TOPO derivative with DraI-deleted bla

strain with defective ampC is henceforth referred as PAOampC. 2.4. PCR amplification and cloning of poxB gene Based on the P. aeruginosa PAO1 genome database (www.pseudomonas.com), poxB (PA5514) appears in close proximity (50 bp) to an upstream putative ORF annotated PA5513, named as poxA (Fig. 1). Because these two ORFs could potentially form an operon, we designed primers that PCR-amplified the two ORFs together (poxAB) and the poxB gene alone. KKF28poxAFor (5V-AGCTGCTTGCGCACCTGT-3V) and KKF30poxARev (5V-CGGTTTCG-

(Holloway and Morgan, 1986) This study (Kiewitz and Tummler, 2000; Spangenberg (Kiewitz and Tummler, 2000; Spangenberg (Kiewitz and Tummler, 2000; Spangenberg (Kiewitz and Tummler, 2000; Spangenberg (Kiewitz and Tummler, 2000; Spangenberg (Kiewitz and Tummler, 2000; Spangenberg

et et et et et et

al., al., al., al., al., al.,

1998) 1998) 1998) 1998) 1998) 1998)

(Kiewitz and Tummler, 2000; Spangenberg et al., 1998) (Kiewitz and Tummler, 2000; Spangenberg et al., 1998) (Kiewitz and Tummler, 2000; Spangenberg et al., 1998) (Kiewitz and Tummler, 2000; Spangenberg et al., 1998) (Kiewitz and Tummler, 2000; Spangenberg et al., 1998)

(Liu et al., 1983) US Patent 3,301,766 US Patent 3,729,378

This study This study This study This study This study This study

AATTCACCGTCA-3V) generated a 2170-bp fragment containing both ORFs, whereas KKF29poxAFor (5VCACTGCTTCATGCAGGAAGA-3V) and KKF30poxARev yielded a 1142-bp PCR product containing poxB alone. These PCR amplicons were electrophoresed and purified according to manufacturer’s instructions (Qiagen, CA). The purified products were ligated into pCRII-TOPO and transformed into TOP10F’ (Invitrogen, CA). The successful transformants were screened for orientation with restriction analysis. Only clones with the desired orientation, i.e. the ORFs under the control of the Plac promoter, were selected for further analysis. In addition, because we were concerned that the presence of bla gene in the pCRII-TOPO vector

K.-F. Kong et al. / Gene 358 (2005) 82 – 92

85

Fig. 1. Physical maps of ampC (A) and poxAB (B). All lines are drawn to scale. (A) The 2150-bp fragment corresponds to coordinate 4593700 to 4595849 of the PAO1 genome (www.pseudomonas.com). The horizontal lines below the map show the amount of PAO1 genome DNA present in each plasmid. The restrictions sites, EcoRI and BamH1, flanking the ends, were introduced with the primers used for PCR. Plasmids pSJ02 and pSJ04 are derivatives of pCRIITOPO (ApR, KmR; ColE1ori lacZa; Invitrogen, CA) and pME6030 (TcR; Pseudomonas pVS1 replicon, E. coli p15A replicon; Accession No: AF133830), respectively. The plasmid pSJ04 is used for complementation analysis and is referred to as pAmpC. The gentamycin cassette (4) from pUCGm (ApR, GmR; pUC19 derivative containing gentamycin cassette; Accession No: U04610) was inserted into the SalI site of pSJ02 disrupting ampC. The ampC::aacCI was cloned into a suicide vector, pEX100T (ApR; sacB oriT; Accession No: U17500), creating pSJ08. (B) The 2170-bp fragment contains PAO1 sequence from coordinates 6206069 to 6208240. The horizontal lines below the map show the PAO1 sequence present in pKKF0677 and pKKF0675. The restriction map of the region is based on the PAO1 genome sequence with relevant restriction sites.

might interfere with data interpretation, an internal 712-bp DraI fragment, representing the bla gene, was deleted and transformed into E. coli DH5a. Selection was carried out using kanamycin as this resistance marker was also present in pCRII-TOPO. The plasmid pKKF675 and pKKF677 harbored the poxAB operon and poxB gene, respectively. The negative control, plasmid pKKF0679, contained a self-ligated pCRII-TOPO with deleted bla (Table 1).

PCR products then were sequenced in both directions using the same set of primers. 2.7. Minimal Inhibitory Concentration (MIC) The MIC was determined by the E-test system (AB Biodisk, Sweden) according to the manufacturer’s instructions. 2.8. b-lactamase assay

2.5. DNA sequencing Both strands of the cloned DNA fragment from pKKF675 and pKKF677 were sequenced using an ABI 3.1 Prism BigDye Terminator Cycle Sequencing Ready Reaction Kit (ABI, CA) together with M13 universal primers. The sequences were analyzed using EditView 1.0.1 (ABI, CA) and DNAStar (Wisconsin, WI) computer software. 2.6. PCR amplification of different P. aeruginosa isolates Colony PCR, using the same set of primers (KKF29poxAFor and KKF30poxARev), was performed to determine the presence of poxB in various clinical and environmental P. aeruginosa strains. Each strain was done in triplicate. The

The h-lactamase assay was modified based on previously published protocols (O’Callaghan et al., 1972). Briefly, stationary-phase cultures were diluted 1 : 100 into 10 ml of LB broth and incubated with shaking at 37 -C until the culture density reached an OD600 of 0.6 – 0.8. For induction, the cells were induced with 500 Ag/ml benzylpenicillin for PAOampC and 25 Ag/ml benzyl-penicillin for E. coli. They were incubated for an additional 3 h before harvesting. The cells were washed once with 50 mM sodium phosphate buffer and were then resuspended in a final volume of 2 ml of the same buffer. Following disruption of cells on ice with sonication (three 30-s pulses with 50% maximum power), the cell lysate was centrifuged at 10,000 g for 30 min at 4 -C. A 2 Al aliquot of the h-lactamasecontaining supernatant was added to nitrocefin (final

86

K.-F. Kong et al. / Gene 358 (2005) 82 – 92

concentration, 100 AM) at room temperature in a final volume of 1 ml of assay buffer. The reaction was allowed to take place for 20 min before nitrocefin hydrolysis was measured spectrophotometrically at 482 nm. Concurrently, the total protein was determined with the Bradford assay using the same supernatant. 2.9. Nucleotide sequence accession number The nucleotide sequences reported here have been assigned the following GenBank accession numbers: poxAB, AY596805; poxB, AY596806; poxB1, AY597419; poxB2, AY597420; poxB3, AY597421; poxB4, AY597422; poxB5, AY597423; poxB6, AY597424; poxB7, AY597425; poxB8 , AY597426; poxB9 , AY597427; poxB10 , AY597428; poxB11, AY597429; poxB12, AY597430; poxB13 , AY597431; poxB14 , AY597432; poxB15 , AY597433; poxB16, AY597434; poxB17, AY597435; poxB18 , AY597436; poxB19 , AY597437; poxB20 AY597438; poxB21, AY597439; poxB22, AY597440; poxB23 , AY597441; poxB24 , AY597442; poxB25 , AY597443; poxB26, AY597444; poxB27, AY597445.

3. Results 3.1. Mutation in ampC had no effect on antibiotic sensitivity P. aeruginosa ampC has been cloned and sequenced in the heterologous host, E. coli (Lodge et al., 1990). It has been assumed that AmpC is the sole chromosomal hlactamase produced by P. aeruginosa. We generated an ampC mutation in the wild-type PAO1 to investigate its role in h-lactam resistance. We determined the MIC of PAO1 and the isogenic mutant, PAOampC using three representatives of h-lactam sub-classes: benzyl-penicillin (penicillin group), meropenem (penem group) and cefotaxime (cephalosphorin) (Table 2). As compared to PAO1, PAOampC did not exhibit increased sensitivity towards any of the three h-lactam antibiotics tested. This result was very intriguing as the latter two antibiotics are used therapeutically. 3.2. ampC mutation confers B-lactamase activity The MIC results argued that the ampC mutation had minimal effect on h-lactamase activity, though we predicted otherwise. Thus, we compared the h-lactamase activity of PAOampC to PAO1. In the absence of an inducer, PAO1 expressed a basal-level of h-lactamase activity (Fig. 2). A Table 2 MIC of PAO1 and PAOampC

PAO1 PAOampC

Pen G

MP

CT

>256 >256

0.19 0.19

2 3

Abbreviation: Pen G, pencillin G; MP, meropenem; CT, cefotaxime.

Fig. 2. h-lactamase expression in PAOampC mutant. Assays were performed on the parent PAO1, the mutant PAOampC, and PAOampC (pAmpC) in the absence ( ) and presence (+) of benzyl-penicillin. Cultures at OD600 of 0.6 – 0.8 were induced with 500 Ag/ml benzyl-penicillin for 3 h before harvesting. Assays were performed on sonicated lysate using nitrocefin as a chomogenic substrate. One milliunit of h-lactamase is defined as 1 nmol of nitrocefin hydrolyzed per minute per microgram of protein. Assays were performed in triplicates.

seven-fold induction was observed in the presence of an inducer. Interestingly, the PAOampC mutant continued to express a basal-level of h-lactamase activity in the absence of an inducer that was induced 2.5-fold upon antibiotic challenge. It appears that the mutation in ampC resulted in approximately a 50% loss of activity (Fig. 2). The activity was restored upon complementation with a low-copy plasmid carrying wild-type ampC (pSJ04). These findings led us to postulate that there is a second h-lactamase in the PAO1 genome. A sequence search on the genome database (www.pseudomonas.com) led to the discovery of a second putative h-lactamase at PA5514. 3.3. Analysis of PA5514 According to the genome annotation, PA5514 is a 789bp long ORF, encoding a 29-kDa protein. This putative protein shows 56% similarity to Tn1406 h-lactamase (Couture et al., 1992) and is predicted to have a class D or an oxacillinase active site. However, there are two inframe start-codons (ATG) separated by 21-bp (Fig. 3). All h-lactamases are localized to the periplasm, thus requiring a signal peptide for translocation. A typical signal peptide will have a hydrophilic/charged amino acid residues followed by a hydrophobic core that ends with a cleavage site for signal peptidase (Izard and Kendall, 1994). If the ORF starts at the second ATG, the Kyte-and-Doolitle hydrophobicity plot shows a very hydrophobic domain at the N-terminus. This is not consistent with the presence of a signal peptide. However, if it is translated off the first ATG, the additional seven amino acid residues at the N-terminus fulfills the peptide signals required for translocation across cytoplasm. The predicted signal sequence is likely to be NH2-MTPQDRAMRPLLFSALLLLSGHTQA,SEW with

K.-F. Kong et al. / Gene 358 (2005) 82 – 92

the cleavage site indicated with an arrow (the italicized segment represents the additional seven amino acids). In addition, the longer ORF has a very strong ribosomebinding site immediately upstream as compared to the shorter ORF (Fig. 3). Thus, we proposed that the longer ORF is likely to encode the h-lactamase. We designated the longer ORF as poxB for Pseudomonas oxacillinase.

87

3.6. b-lactamase activity of PoxB

Analysis of the sequences around poxB revealed that the upstream ORF PA5513 is only 49 bp away (Fig. 3). In the absence of promoter-like sequences in the short intergenic region, we proposed that PA5513 and poxB form a two-gene operon (Fig. 1B). A putative j70 promoter sequence is found upstream of PA5513 that may contribute to the basallevel expression of h-lactamase in the absence of h-lactam antibiotics (Fig. 3). The sequence analysis also reveals the presence of a potential stem and loop structure followed by a trail of Ts downstream of poxB, indicating the presence of a U-independent terminator (Fig. 3). PA5513 ORF, or poxA, is 879-bp long encoding a 32.4kDa protein with a calculated pI value of 8.57. The hydrophobicity plot suggests the presence of two transmembrane spanning domains. The absence of any signal sequence (Fig. 3) suggests that it is likely to localize in the inner membrane. BLAST analysis demonstrates that poxA has high homology to a hydrolyase or acyltransferase. Currently, this hydrolyase superfamily constitutes a large number of proteins with similar alpha-beta hydrolyase folding, but their actual functions are yet to be characterized.

To demonstrate if poxB is truly encoding a h-lactamase, we subcloned the poxAB operon (pKKF0675; Fig. 1B) and the poxB alone (pKKF0677; Fig. 1B) into a modified pCRIITOPO (pKKF0679) vector. A 712-bp fragment containing the bla gene from the vector pCRII-TOPO was deleted to prevent potential contribution by this gene in the hlactamase assay. An E. coli strain containing the modified vector alone, pKKF0679, did not express any h-lactamase activity. In the absence of an inducer, the plasmids containing poxAB operon and poxB, pKKF0675 and pKKF0677 respectively, exhibited a significant amount of h-lactamase activity (Fig. 5). This clearly suggests that poxB encodes a h-lactamase. High basal expression may be due to transcription from the putative j70 promoter found upstream of the poxA gene (Fig. 3) or the Plac promoter found upstream in the vector. In the case of the promoterless poxB gene, the expression of the gene has to be driven by Plac. If this argument is true, the activity should increase in the presence of IPTG. As predicted, the amount of h-lactamase activity increased in the presence of IPTG (Fig. 5). In an E. coli strain harboring the poxAB operon, the basal activity remained the same in the absence and presence of IPTG. This suggests that the primary expression is likely to be from the cognate j70 promoter. This result also supports our earlier prediction that these two ORF form an operon. The poxAB operon also appeared not to respond to the presence of inducers such as penicillin G and oxacillin (Fig. 5). This suggests that the expression of this operon is not inducible in the heterologous E. coli host.

3.5. poxB analysis

3.7. poxB is ubiquitous

The poxB gene is 810 bp long encoding a 25-kDa protein with a calculated isoelectric point, pI, of 8.6. A sequence homology search using BLAST revealed that PoxB has the highest homology with Acinetobacter baumannii’s ARI-1, OXA-27, OXA-49 (44% identity, 63% homology; PMID: AJ132105) and Class D h-lactamase of Fusobacterium nucleatum’ FUS-1 (41% identity, 64% homology; PMID: AAP69916). Couture et al. proposed the use of DBL numeration for Class D h-lactamase (Couture et al., 1992). Twenty-six conserved residues are used in this case to distinguish Class D from Class A enzymes. Alignment of PoxB against multiple oxacillinases revealed that 21 out of the 26 residues are identical. The five different residues are 72F&Y (DBL numeration), 93F&L, 95Y&W, 138Y&N and 260L&I. Among these substitutions, the former four had not been reported in any oxacillinases. The 260L&I mutation was previously found in OXA-9 (PMID: AAO24600). Differences in these residues placed PoxB in a new branch in the dendogram constructed using Clustal W (Madison, WI) (Fig. 4). This suggests that poxB belongs to a new kind of oxacillinases, distinctly different from the known OXA enzymes.

Genome diversity of P. aeruginosa has been well documented (Kiewitz and Tummler, 2000). We explored if the poxB gene is PAO1-specific or if it shows a wide distribution. To demonstrate the distribution of poxB, we screened a collection of P. aeruginosa strains. This collection represents 70 strains from various clinical and environmental sources (Table 1). Colony PCR with KKF29poxAFor and KKF30poxARev (Table 1) was performed on each strain to determine the presence of poxB. Forty strains (57%) yielded positive results with expected PCR products of 1142 bp (data not shown). Plasmid extraction using a Qiagen Miniprep kit (Qiagen, CA) was performed to ensure the absence of extrachromosomal elements. This indicates that poxB is commonly found in chromosomes of both clinical and environmental P. aeruginosa isolates. The PCR products were purified and sequenced from both ends.

3.4. Analysis of poxAB operon

3.8. Sequence alignment of poxB alleles We referred to PAO1 poxB as the wild-type gene. It appears that poxB is conserved among the P. aeruginosa

88

K.-F. Kong et al. / Gene 358 (2005) 82 – 92

K.-F. Kong et al. / Gene 358 (2005) 82 – 92

89 PoxB10 PoxB19 PoxB13 PoxB20 PoxB24 PoxB7 PoxB25 PoxB22 PoxB26 PoxB9 PoxB4 PoxB1 PoxB6 PoxB2 PoxB18 PoxB12 PoxB PoxB23 PoxB11 PoxB5 PoxB17 PoxB21 PoxB3 PoxB27 PoxB16 PoxB14 PoxB15 PoxB8 OXA-20 CAC85643 Acinetobacter baumannii OXA-37 AAG33665 Acinetobacter baumannii OXA-3 AAC41449 Pseudomonas aeruginosa OXA-15 AAB05874 Pseudomonas aeruginosa OXA-13-1 AAD02274 P. aeruginosa 391-1 OXA-7 P35695 E. coli OXA-19 AAD02245 P. aeruginosa 191 OXA-5 Q00982 P. aeruginosa OXA-11 Q06778 P. aeruginosa OXA-10 AAB60534 P. aeruginosa OXA-14 AAA93528 P. aeruginosa OXA-16 AAB97924 P. aeruginosa OXA-17 AAC15074 P. aeruginosa OXA-9 AAD22145 Enterobacter aerogens OXA-9 AA024600 E. cloacae OXA-18 AAB58555 P. aeruginosa OXA-45 CAD58780 P. aeruginosa Aox CAA56560 Aeromonas sabria OXA-31 AAK52604 P. aeruginosa OXA-4 AAR12140 Citrobacter freundii OXA-30 AAP51284 E. coli OXA-1 AA084013 Salmonella enterica

795.8 500

200

100

50

20

10

5

2

1

0

Fig. 4. Phylogenetic tree of poxB alleles and other oxacillinases previously reported. All poxB alleles belong to a novel branch that is distantly related to other oxacillinases. The dendogram was generated using Clustal W available in the DNAStar software.

strains. Although none of these strains showed absolute identity with the wild-type poxB, all of the 26 residues designated in the DBL numeration of h-lactamases gave complete conservation in all of the poxB alleles. The change in amino acid sequences ranges from 2 to 35 residues (Table 3). The poxB7 allele from the DSM1128 strain, an ear infection isolate, showed the highest variation, having 35 amino acid substitutions. There were a number of substitutions that appear to have high prevalence. Glutamic acid (E) and aspartic acid (D), for example, appeared at almost the same frequencies at residue 116 of PoxB (53% versus 47%), while at residue 23, threonine (T) is substituted with alanine (A) in nine of the 28 strains analyzed (Table 3). In addition, we analyzed the amino acid exchange ability using Argyle’s ring (Argyle, 1980). Drastic amino acid substitutions were only seen at residue 56 (R Y G/C) and residue 119 (K Y E).

4. Discussion Antibiotic resistance to h-lactam antibiotics in Gramnegative bacteria often involves the interplay of three highly regulated systems, i.e. inducible chromosomal AmpC hlactamase, regulatable membrane permeability and expression of multiple efflux pumps (Okamoto et al., 2001). Besides Class C enzymes, chromosomally encoded inducible classes A, B and D h-lactamases could also be encountered in bacteria. Aeromonas spp. encodes inducible Classes B, C and D h-lactamases simultaneously (Niumsup et al., 2003). Hence, the presence of more than one copy of chromosomal h-lactamases belonging to different classes is not unusual, particularly for environmentally ubiquitous organisms with high intrinsic antibiotic resistance. A previous study has identified the presence of the ampC gene

Fig. 3. Sequence map of poxAB. This is the detailed DNA sequence of pKKF0679 with the corresponding amino acid sequences of PoxA and PoxB. Solid arrows indicate the priming regions used to generate the PCR products. All common restriction sites are shown. The putative promoter is indicated with the E. coli j70 promoter consensus sequence. Boxed amino acids represent the signal peptide recognition site. Dotted arrows represent the hairpin loop of the Uindependent terminator and the following dotted line represents the T string.

90

K.-F. Kong et al. / Gene 358 (2005) 82 – 92

Table 3 Amino acid changes observed in PoxB alleles Residue PAO1 PoxB7c Other Other allelesd a b d number PoxB changes

Fig. 5. PoxB h-lactamase activity in heterologous host E. coli. Assays were performed on E. coli DH5a harboring pKKF0675 (poxB alone), pKKF0677 (poxAB operon), and vector control (pKKF0679) in the absence ( ) and presence (+) of benzyl-penicillin, oxacillin, and IPTG, as described in Materials and methods. Antibiotic challenge was carried out using 25 Ag/ ml of benzyl-penicillin or 100 Ag/ml of oxacillin. Induction with IPTG was done at the final concentration of 100 Ag/ml. Assays were performed on sonicated lysate using nitrocefin as a chomogenic substrate (O’Callaghan et al., 1972). One milliunit of h-lactamase is defined as 1 nmol of nitrocefin hydrolyzed per minute per microgram of protein. Assays were performed in triplicates. The activity of E. coli DH5a(pKKF0679) was negligible in all cases tested.

encoding chromosomal h-lactamase in the P. aeruginosa chromosome (Lodge et al., 1990). In this study, we explored the role of P. aeruginosa ampC in antibiotic resistance. 4.1. P. aeruginosa expresses two chromosomal b-lactamases A previous report showed that inactivation of ampC in PAO1 does not significantly change the susceptibility profile towards the penems (Okamoto et al., 2001). In vitro susceptibility assays on our ampC mutant is in corroboration with this data (Table 2). Furthermore, our study showed that this effect is not limited to the penems, but extends to an array of h-lactam antibiotics, including third generation cephalosporins like cefotaxime. However, Okamoto et al. failed to detect any AmpC h-lactamase activity in their ampC mutant (Okamoto et al., 2001). In contrast, our PAOampC secretes 2.5-fold more h-lactamase upon induction (Fig. 2). Ruling out the contribution of other h-lactamases, Okamoto et al. elegantly showed that the alteration of MIC in ampC mutant is due to the presence of other mechanisms, such as the efflux pumps and outer membrane permeability (Okamoto et al., 2001). However, the reasons for their failure to detect any hlactamases could be due to the choice of the chromogenic substrate, cephaloridin, in their experimental procedure. Our assay, using nitrocefin, enabled the detection of a second inducible h-lactamase. The expression of almost 50% of h-lactamase activity in PAOampC, as compared to PAO1, led us back into the P. aeruginosa genome to search for an ORF that can potentially code for a second h-lactamase. The ORF was designated as poxB. It appeared to be part of a two-gene operon (Fig. 1B). The fragments containing poxB alone and

10 12 13

P L F

L F

20 22 23

S H T

39 43 56

L

PoxB1, PoxB2, PoxB6, PoxB18

F Q A

A

PoxB10, PoxB13, PoxB19, PoxB20, PoxB22, PoxB24, PoxB25, PoxB26

G V R

E M G

C

PoxB3, PoxB8, PoxB14, PoxB15, PoxB16, PoxB27

63 66 79 90

E E N R

A Q H K

K

116

D

E

E

119 136 139 141 142 143 145 146 150 152 155 166 170 174

K E R N V S L G A I V K M R

R Q

E

PoxB10, PoxB13, PoxB19, PoxB20, PoxB24, PoxB25 PoxB1, PoxB2, PoxB3, PoxB5, PoxB6, PoxB8, PoxB12, PoxB14, PoxB15, PoxB16, PoxB17, PoxB21, PoxB25, PoxB27 PoxB10, PoxB13, PoxB19

H D

PoxB10 PoxB25

A

PoxB19, PoxB26

H

PoxB1, PoxB2, PoxB3, PoxB6, PoxB12, PoxB16, PoxB27

177 188 189 192 196 200 202 205 232 237 261 264 267

L A P S A L S G N G S A I

A E Q A D M H D D T A

V

PoxB5, PoxB10

T

PoxB5

a

L A M D G V E L

V

Represents the amino acid location in PoxB. Represents the amino acid residues found in the prototypic P. aeruginosa PAO1 in the positions indicated by the residue number. c The poxB7 allele from the DSM1128 strain, an ear infection isolate, showed the highest variation, having 35 amino acid substitutions. d These two columns indicate all the other substitutions seen and the alleles with those specified changes. b

K.-F. Kong et al. / Gene 358 (2005) 82 – 92

along with the upstream gene poxA were cloned into E. coli DH5a to test for the h-lactamase activity. E. coli serves as a suitable heterologous host because it has a low constitutive level of h-lactamase activity (Honore et al., 1986). The presence of this allele in E. coli DH5a conferred higher resistance and produced detectable h-lactamase activity as compared to the parental strain (Fig. 5), suggesting poxB encodes a h-lactamase. In retrospect, the earliest documentation of the biochemical detection of the chromosomally encoded h-lactamases in P. aeruginosa, formerly known as P. pyocyanea, can be traced as early as the 1960s (Jago et al., 1963). Analysis of the crude extract towards h-lactam agents led to the proposal that P. aeruginosa (NCTC 8203) expresses more than one hlactamase (Jago et al., 1963). Subsequently, crude extracts from the same strain were demonstrated to possess both penicillinase and cephalosporinase activities (Sabath et al., 1965). Although interpreted by the authors as having a single h-lactamase with both the cephalosporinase and penicillinase activities, these observations are consistent with the hypothesis of the presence of two h-lactamases in the chromosome of P. aeruginosa. One of the purified h-lactamases from NCTC 8203 is shown to be 42 kDa with pI of 7.5 (Furth, 1975). These values closely correspond to the physical properties of AmpC. Later, using gel-filtration chromatography, another 32-kDa h-lactamase with a higher pI of 8.5 was purified from two strains, NCTC 8203 and a derepressed carbenicillin-resistant mutant (Berks et al., 1982). Surprisingly, the size and pI of the second h-lactamase correlate well to the predicted values of PoxB. Therefore, in earlier literatures, ampC was the first to be reported, but was not the sole h-lactamase present in P. aeruginosa. The 2.5-fold induction seen in PAOampC suggests that the poxB expression is under an inducible system. But the transfer of the gene in heterologous E. coli resulted in constitutive expression of PoxB activity in the absence and presence of h-lactam antibiotics. This strongly indicates that the expression of poxB is not regulated in E. coli. The lack of inducibility could likely be a result of the absence of ampR in E. coli (Honore et al., 1986). We have demonstrated that a mutation of ampR in PAO1 resulted in high levels of constitutive h-lactamase activity (data not shown). Alternatively, the poxAB operon may be regulated by a twocomponent response-regulatory system similar to ampH and OXA-12 of Aeromonas spp. (Niumsup et al., 2003). Sequence analysis around the poxAB operon shows the presence of two ORFs, PA5511 and PA5512, encoding a twocomponent regulatory system (www.pseudomonas.com). Currently, we are exploring the role of these genes for poxAB expression. 4.2. PoxB may confer penem resistance PoxB has the highest homology (44% identity, 63% homology) to Acinetobacter baumannii’s Class D hlactamases. The acquisition of ARI-1, OXA-27 or OXA-

91

49 renders A. baumannii to be resistant to imipenem and carbapenems (PMID: AJ132105). This data suggests that PoxB, but not AmpC, confers resistance to meropenems (Table 2) or other penems (Okamoto et al., 2001). However, whether the presence of PoxB protects the cells from thirdgeneration cephalosporins (Table 2) or other h-lactam agents is unknown. 4.3. poxB is ubiquitous The identification of the second PoxB h-lactamase in the P. aeruginosa genome led us to study the distribution of this gene in clinical and environmental isolates. We postulated that poxB is likely to be found in clinical isolates due to its important role in antibiotic resistance. Much to our surprise, poxB is found in both clinical and environmental isolates with equal frequency. The ubiquity of the poxB gene suggests that this gene may not be acquired recently upon the introduction of h-lactam antibiotics into the clinical settings, but as a defense mechanism of P. aeruginosa against naturally occurring antibiotics or, as with other hlactamases, it may be that some unknown functions of PoxB are beneficial to the bacterial survival. Furthermore, the sequencing of the poxB alleles from the isolates indicated that all of the alleles deviate little from the prototypic PAO1 PoxB. Though as much as 13% of amino acid substitutions were found in PoxB7 from the strain DSM1128, an ear infection isolate (Kiewitz and Tummler, 2000), this allele still conserved all the 26 residues designated as Class D signatures. Due to the minute changes in all poxB alleles and the conservation of all signature motifs, we suggest that poxB in all the P. aeruginosa isolates examined are likely to be biochemically functional. The prevalence of poxB in both clinical and environmental strains of P. aeruginosa raises an alarm, as the high frequency of horizontal gene transfer among bacteria may likely introduce Class D h-lactamases to other co-inhabiting bacteria species. The wide-distribution perhaps explains the failure of using oxacillin or its derivatives as a therapeutic regimen. In addition, deregulation of this gene may well be the reason for penem resistance. It will be of great interest to determine which genes play a role in regulating the expression of poxAB and the clinical significance of this gene. This report takes us one step closer towards the understanding of the P. aeruginosa’s intrinsic armamentarium against a variety of antibiotics.

Acknowledgements This work has been supported by NIH-MBRS SCORE (S06 GM08205; KM) and Florida International University Teaching Assistantships (KFK and SRJ). We are grateful to Brooke Crandall for her editorial assistance and Mathee Lab Crew for their critical reading of the revised manuscript.

92

K.-F. Kong et al. / Gene 358 (2005) 82 – 92

References Ambler, R.P., 1980. The structure of beta-lactamases. Philos. Trans. R. Soc. Lond., B Biol. Sci. 289, 321 – 331. Argyle, E., 1980. A similarity ring for amino acids based on their evolutionary substitution rates. Orig. Life 10, 357 – 360. Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., Struhl, K., 1999. Current Protocols in Molecular Biology. Greene Publishing Associates Inc. and John Wiley and Sons, Inc. Berks, M., Redhead, K., Abraham, E.P., 1982. Isolation and properties of an inducible and a constitutive beta-lactamase from Pseudomonas aeruginosa. J. Gen. Microbiol. 128, 155 – 159. Bush, K., Jacoby, G., Medeiros, A., 1995. A functional classification scheme for beta-lactamases and its correlation with molecular structure. Antimicrob. Agents Chemother. 39, 1211 – 1233. Couture, F., Lachapelle, J., Levesque, R.C., 1992. Phylogeny of LCR-1 and OXA-5 with class A and class D beta-lactamases. Mol. Microbiol. 6, 1693 – 1705. Dale, J.W., Smith, J.T., 1972. A direct comparison of two unusual R-factormediated-lactamases. Biochem. J. 128, 173 – 174. Figurski, D., Helinski, D.R., 1979. Replication of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in trans. Proc. Natl. Acad. Sci. U. S. A. 76, 1648 – 1652. Frederiksen, B., Lanng, S., Koch, C., Høiby, N., 1996. Improved survival in the Danish center-treated cystic fibrosis patients: results of aggressive treatment. Pediatr. Pulmonol. 21, 153 – 158. Furth, A.J., 1975. Purification and properties of a constitutive betalactamase from Pseudomonas aeruginosa strain Dalgleish. Biochim. Biophys. Acta 377, 431 – 443. Holloway, B.W., Morgan, A.F., 1986. Genome organization in Pseudomonas. Annu. Rev. Microbiol. 40, 79 – 105. Honore, N., Nicolas, M.H., Cole, S.T., 1986. Inducible cephalosporinase production in clinical isolates of Enterobacter cloacae is controlled by a regulatory gene that has been deleted from Escherichia coli. EMBO J. 5, 3709 – 3714. Izard, J.W., Kendall, D.A., 1994. Signal peptides: exquisitely designed transport promoters. Mol. Microbiol. 13, 765 – 773. Jago, M., Migliacci, A., Abraham, E.P., 1963. Production of a cephalosporinase by Pseudomonas pyocyanea. Nature 199, 375. Kiewitz, C., Tummler, B., 2000. Sequence diversity of Pseudomonas aeruginosa: impact on population structure and genome evolution. J. Bacteriol. 182, 3125 – 3135. Lachapelle, J., Dufresne, J., Levesque, R.C., 1991. Characterization of the blaCARB-3 gene encoding the carbenicillinase-3 beta-lactamase of Pseudomonas aeruginosa. Gene 102, 7 – 12. Lambert, P.A., 2002. Mechanism of antibiotic resistance in Pseudomonas aeruginosa. J. R. Soc. Med. 95, 22 – 26.

Liu, P.V., Matsumoto, H., Kusama, H., Bergan, T., 1983. Survey of heatstable, major somatic antigens of Pseudomonas aeruginosa. Int. J. Syst. Bacteriol. 33, 256 – 264. Lodge, J.M., Minchin, S.D., Piddock, L.J., Busby, J.W., 1990. Cloning, sequencing and analysis of the structural gene and regulatory region of the Pseudomonas aeruginosa chromosomal ampC beta-lactamase. Biochem. J. 272, 627 – 631. Medeiros, A.A., Cohenford, M., Jacoby, G.A., 1985. Five novel plasmiddetermined beta-lactamases. Antimicrob. Agents Chemother. 27, 715 – 719. Niumsup, P., Simm, A.M., Nurmahomed, K., Walsh, T.R., Bennett, P.M., Avison, M.B., 2003. Genetic linkage of the penicillinase gene, amp, and blrAB, encoding the regulator of beta-lactamase expression in Aeromonas spp. J. Antimicrob. Chemother. 51, 1351 – 1358. O’Callaghan, C.H., Morris, A., Kirby, S.M., Shingler, A.H., 1972. Novel method for detection of beta-lactamases by using a chromogenic cephalosporin substrate. Antimicrob. Agents Chemother. 1, 283 – 288. Okamoto, K., Gotoh, N., Nishino, T., 2001. Pseudomonas aeruginosa reveals high intrinsic resistance to penem antibiotics: penem resistance mechanisms and their interplay. Antimicrob. Agents Chemother. 45, 1964 – 1971. Poirel, L., Weldhagen, G.F., De Champs, C., Nordmann, P., 2002. A nosocomial outbreak of Pseudomonas aeruginosa isolates expressing the extended-spectrum beta-lactamase GES-2 in South Africa. J. Antimicrob. Chemother. 49, 561 – 565. Sabath, L.D., Jago, M., Abraham, E.P., 1965. Cephalosporinase and penicillinase activities of a beta-lactamase from Pseudomonas pyocyanea. Biochem. J. 96, 739 – 752. Senda, K., et al., 1996. PCR detection of metallo-beta-lactamase gene (blaIMP) in gram-negative rods resistant to broad-spectrum betalactams. J. Clin. Microbiol. 34, 2909 – 2913. Spangenberg, C., Montie, T.C., Tummler, B., 1998. Structural and functional implications of sequence diversity of Pseudomonas aeruginosa genes oriC, ampC and fliC. Electrophoresis 19, 545 – 550. Tsakris, A., et al., 2000. Outbreak of infections caused by Pseudomonas aeruginosa producing VIM-1 carbapenemase in Greece. J. Clin. Microbiol. 38, 1290 – 1292. Watanabe, M., Iyobe, S., Inoue, M., Mitsuhashi, S., 1991. Transferable imipenem resistance in Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 35, 147 – 151. Weldhagen, G.F., Poirel, L., Nordmann, P., 2003. Ambler class A extended-spectrum beta-lactamases in Pseudomonas aeruginosa: novel developments and clinical impact. Antimicrob. Agents Chemother. 47, 2385 – 2392.