Research Article
1 October 2005

Kinetic Properties of Four Plasmid-Mediated AmpC β-Lactamases

ABSTRACT

The heterologous production in Escherichia coli, the purification, and the kinetic characterization of four plasmid-encoded class C β-lactamases (ACT-1, MIR-1, CMY-2, and CMY-1) were performed. Except for their instability, these enzymes are very similar to the known chromosomally encoded AmpC β-lactamases. Their kinetic parameters did not show major differences from those obtained for the corresponding chromosomal enzymes. However, the Km values of CMY-2 for cefuroxime, cefotaxime, and oxacillin were significantly decreased compared to those of the chromosomal AmpC enzymes. Finally, the susceptibility patterns of different E. coli hosts producing a plasmid- or a chromosome-encoded class C enzyme toward β-lactam antibiotics are mainly due to the overproduction of the β-lactamase in the periplasmic space of the bacteria rather than to a specific catalytic profile of the plasmid-encoded β-lactamases.
β-Lactams illustrate the growing phenomenon of bacterial resistance to antibiotics. Since the beginning of their utilization, over 50 years ago, the introduction of new compounds of this family has resulted in increasingly rapid evolution in the microbial world (4, 12, 19). The principal bacterial defense mechanism is the production of β-lactamases, enzymes which produce a biologically inactive product by hydrolyzing the β-lactam ring (20). On the basis of their primary structures, β-lactamases are grouped into four classes (A, B, C, and D) (8), and the various enzymes exhibit specific activity spectra. Enzymes of classes A, C, and D are active-site serine enzymes, whereas the activities of class B enzymes are zinc dependent (14). The genes coding for the β-lactamases of classes A, B, and D are located either on transferable plasmids or on the bacterial chromosome (14). Initially, class C β-lactamases (AmpC) were described as chromosomal enzymes (6). However, over the last 20 years, more than 20 plasmid-encoded class C enzymes have been identified (18). These proteins may be classified into six subgroups based on amino acid sequence similarities. Each subgroup contains plasmid-borne class C enzymes and their closest chromosomal relatives (18). Clinical isolates harboring plasmidic AmpC exhibit high resistance toward β-lactam antibiotics, such as cephamycins and monobactams.
The present work was focused on an extensive characterization of four plasmid-borne class C β-lactamases; ACT-1 and MIR-1 (subgroup 2), CMY-2 (subgroup 1), and CMY-1 (subgroup 6), in order to determine if their kinetic properties were significantly different from those of the chromosome-encoded enzymes, as suggested before. As will be discussed below, this paper provides the first biochemical data showing that these enzymes have not yet significantly improved their catalytic mechanisms and thus do not deserve the appellation of extended-spectrum β-lactamases, as previously suggested by some authors (1, 7).

MATERIALS AND METHODS

Antibiotics and other chemicals.

Nitrocefin was from Oxoid (Basingstoke, Hants, United Kingdom). Ampicillin, benzylpenicillin, cephalothin, cephaloridine, cefoxitin, oxacillin, cefuroxime, and cefotaxime were from Sigma (St. Louis, MO). Cephalexin and cefazolin were from Eli Lilly and Co. (Indianapolis, IN). Imipenem and kanamycin were from Merck Sharp Dohme (Brussels, Belgium). Aztreonam was from Bristol-Myers Squibb (Brussels, Belgium). The antibiotic extinction molar coefficients were described previously (10, 11). 5-Bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) and isopropyl-β-d-thiogalactopyranoside (IPTG) were from ImmunoSource (Halle, Belgium) and ICN Biomedical Inc. (Aurora, OH), respectively.

Bacterial strains and vectors.

The strains Escherichia coli SNO3 pMG232 (MIR-1 producer), E. coli XL1 pMG261 (ACT-1 producer), and E. coli XL1 pMG262 (CMY-2 producer) were gifts of G. A. Jacoby (Section of Infectious Disease, Lahey Clinic, Burlington, MA). The E. coli strain producing CMY-1 was a gift of Yunsop Chong (College of Medicine, Yonsei University, Seoul, Korea). E. coli DH5α (Gibco-BRL-Life Technologies, Eragny, France) and E. coli BL21(DE3) (Novagen Inc., Madison, WI) were used as recipients for cloning and expression experiments, respectively. The plasmid pGEM-T-easy from Promega (Madison, WI) was used to clone the PCR products and pET26b(+) (Novagen, Inc.) was used to produce the different plasmid-encoded class C β-lactamases. For all the cloning experiments, the E. coli strains carrying the different ampC genes were grown at 37°C in Luria-Bertani (LB) medium in the presence of 50 μg/ml ampicillin or 50 μg/ml kanamycin.

Construction of expression vectors.

The plasmids pMG232, pMG261, and pMG262 were extracted by using the GFX Micro Plasmid Prep kit (Pharmacia, Uppsala, Sweden). The plasmid containing the blaCMY1 gene was isolated with the help of the Sepagene kit (Sanko Junyaku Co., Tokyo, Japan). The different β-lactamase genes were amplified by PCR. The different PCR primers were designed in order to introduce the NdeI and the HindIII restriction sites up- and downstream of the genes, respectively (Table 1). The PCR conditions were as follows: incubation for 1 min at 95°C; 30 cycles of 1 min at 95°C and 1 min at 60°C for blaCMY2, blaACT1, and blaMIR1 or 1 min at 58°C for blaCMY1; and 1 min at 72°C. The PCRs were completed by incubating the reaction mixtures at 72°C for 5 min. Taq DNA polymerase (Eurogentec, Seraing, Belgium) was used for PCR. The amplified fragments were introduced into the polyT pGEM-T-easy vector. The pGEM vectors containing blaACT1, blaCMY1, blaCMY2, and blaMIR1 were called pBC-1, pBC-2, pBC-3, and pBC-4, respectively. They were used to transform E. coli DH5α competent cells. The genes were completely sequenced to verify the presence of the new restriction sites and the absence of unwanted mutations. Read sequencing was performed using an automated laser fluorescent DNA sequencer (Amersham Pharmacia Biosciences, Uppsala, Sweden). The different pBC plasmids were digested by NdeI and HindIII restriction enzymes. The 1.5-kb DNA fragment containing the ampC genes was gel purified and ligated into the pET-26b(+) vector digested by the same restriction enzyme mentioned above. The corresponding plasmids pBC1e, pBC2e, pBC3e, and pBC4e, allowing the overexpression of the ACT-1, CMY-1, CMY-2, and MIR-1 β-lactamases, respectively, were introduced into E. coli BL21(DE3).

Production and purification of ACT-1, MIR-1, CMY-1, and CMY-2.

For all enzymes, a 50-ml preculture was grown overnight in LB medium at 37°C in the presence of kanamycin (50 μg/ml). Two liters of fresh medium (LB for CMY-1 and CMY-2; 2XYT for ACT-1 and MIR-1) was inoculated with the preculture. The cultures were incubated at 28°C for the production of ACT-1 and CMY-2 or at 37°C for the production of CMY-1 and MIR-1. At an A600 of 1, IPTG (0.1 mM final concentration) was added to the cultures. They were incubated for 4, 7, 4, and 5 additional hours for the production of ACT-1, CMY-2, CMY-1, and MIR-1, respectively.
The purification of ACT-1 was performed as follows. The bacteria were harvested by centrifugation at 5,000 × g for 10 min at 4°C. The cells were resuspended in 50 ml of 15 mM sodium phosphate buffer, pH 7 (buffer A), and disrupted with the help of a cellular disruptor (Constant System Ltd., United Kingdom). Cell debris was eliminated by centrifugation at 15,000 × g for 30 min at 4°C. The supernatant was recovered and dialyzed overnight against buffer A at 4°C. The sample was clarified by filtration through a 0.45-μm Durapore membrane filter (Millipore, County Cork, Ireland). The crude extract was then loaded at 5 ml/min on an SP Sepharose fast-flow column (30 × 2.6 cm; Amersham Biosciences, Uppsala, Sweden) previously equilibrated with buffer A. Then, the β-lactamase was eluted by a linear NaCl gradient (0 to 0.3 M) in buffer A over five column volumes at a flow rate of 5 ml/min. The fractions containing the β-lactamase were pooled and concentrated by ultrafiltration on a membrane with a cutoff size of 10 kDa (Millipore, County Cork, Ireland). The sample was then loaded at 1 ml/min onto a 20-ml aminophenylboronic acid-A8530 column (20 by 1.6 cm; Sigma, St. Louis, MO) previously equilibrated in 20 mM triethanolamine-HCl, 0.5 M NaCl, pH 7.0 (buffer B). The column was extensively washed with buffer B until the absorbance of the eluent at 280 nm was lower than 0.1. The β-lactamase was eluted by a linear borate gradient (0 to 0.5 M) in buffer B over five column volumes at a flow rate of 1 ml/min. The fractions exhibiting β-lactamase activity were recovered. The β-lactamase activities of the different fractions were determined as the initial rate of hydrolysis of a 100 μM nitrocefin solution prepared in phosphate buffer. Their purity was estimated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and staining with Coomassie Brilliant Blue R250 from Fluka (Buchs, Switzerland). The fractions containing pure ACT-1 were pooled and dialyzed three times against 1 liter of buffer A. The purification of CMY-2 was done as described for the ACT-1 β-lactamase. In the case of the MIR-1 and CMY-1 β-lactamases, the ion-exchange step was sufficient to obtain a pure enzyme preparation. The purifications of these two enzymes were performed in 15 mM sodium phosphate at pH 7 and 6.5, respectively. The protein concentration was determined by bicinchoninic acid protein assay (Pierce, Rockford, IL) and by measuring their absorbance at 280 nm. The molar extinction coefficients at 280 nm for CMY-2, MIR-1, ACT-1, and CMY-1 were 91,200, 80,000, 81,000, and 80,000 M−1 cm−1, respectively.

Mass spectrometry and N-terminal sequences.

The N-terminal sequences of the purified enzymes were determined using an automated Procise Protein Sequencing System connected to a Macintosh 650 computer (Perkin-Elmer Corp., Wellesley, MA). The exact masses of the ACT-1, MIR-1, CMY-1, and CMY-2 enzymes were determined in positive-ion mode by electron spray ionization-mass spectrometry on a Q-Tof Ultima mass spectrometer (Micromass, United Kingdom) fitted with a nanospray source and using homemade gold-coated borosilicate glass emitters. Before injection into the mass spectrometer, the samples were desalted by performing three cycles of concentration-dilution (fivefold) in 0.1% formic acid-acetonitrile (50:50 [vol/vol]) using an Ultrafree-MC centrifugal filter device (Millipore) with a 10,000 nominal molecular mass limit. Final protein concentrations varied from 2 to 5 μM. Calibration was performed with horse heart myoglobin.

Enzyme titration.

The different β-lactamases (4 μM) were incubated for 30 min at 30°C in the presence of different concentrations of aztreonam (0 to 18 μM) in 50 mM MOPS (morpholinepropanesulfonic acid), pH 7, 50 mM NaCl (buffer M). The solution was diluted with buffer M to a final enzyme concentration of 50 nM; 10 μl of the solution was then added to 490 μl of 100 μM cephalothin in buffer M. The activity was determined as described above.

Kinetic parameters.

All the kinetic measurements were performed at 30°C in 50 mM MOPS, pH 7, 50 mM NaCl (buffer M). The variations in absorbance were measured on a double-beam Uvikon Xl spectrophotometer (BRS, Brussels, Belgium) connected to an Intel Pentium III personal computer. During the kinetic measurements, the enzyme stock solutions were kept on ice in 50 mM MOPS, pH 7, 1 μM bovine serum albumin, 50 mM NaCl. For the good substrates, the kinetic parameters were determined from the initial rates of reactions, using both the Hanes-Woolf linearization of the Henri-Michaelis-Menten equation and a direct nonlinear regression with the hyperbolic equation using the KaleidaGraph program (version 3.5; Synergy Software). Low Km values were determined as competitive inhibition constants, Ki, in the presence of a good reporter substrate (nitrocefin or cephalothin).
When the β-lactam behaved as a poor substrate or inactivator, progressive inactivation of the different enzymes was monitored continuously by measuring the residual activity toward a reporter substrate (100 μM cephalothin). The pseudo-first-order inactivation constants, ki, were computed, and the different constants were calculated as described previously (5).

Susceptibility patterns.

The MICs of β-lactams were determined for E. coli DH5α, E. coli DH5α pMG232 (MIR-1 producer), E. coli DH5α pMG261 (ACT-1 producer), and E. coli pAD7 (chromosomal E. coli K-12 AmpC producer). The in vitro susceptibilities of the different E. coli strains were determined by a broth dilution method (16) in nutrient broth (NB) and LB medium, respectively, with a standard inoculum of 105 CFU/ml. The results were recorded after 20 h of incubation at 37°C.

Theoretical MICs.

As previously described (9), the external antibiotic concentration (Ie) necessary to obtain a periplasmic antibiotic concentration (Ip) is given by equation 1.
\[ \[I_{e}{=}I_{p}{+}\frac{k_{\mathrm{cat}}\ {\cdot}\ E_{0}\ {\cdot}\ I_{p}}{P\ {\cdot}\ A\ {\cdot}\ (K_{m}{+}I_{p})}\] \]
(1)
where P is the permeability of the outer membrane, A is the membrane area by unit weight, kcat and Km are the kinetic parameters of the β-lactamase, and E0 is its concentration by unit of cell weight. When Ip is sufficient to inactivate the essential penicillin binding proteins, the antibiotic concentration in the periplasmic space is lethal (Ip = Ipl), and the external concentration is equal to the MIC (equation 2).
\[ \[\mathrm{MIC}{=}I_{pl}{+}\frac{k_{\mathrm{cat}}\ {\cdot}\ E_{0}\ {\cdot}\ I_{pl}}{P\ {\cdot}\ A\ {\cdot}\ (K_{m}{+}I_{pl})}\] \]
(2)
On the basis of this equation, we computed the theoretical MICs of the different E. coli strains using Ipl values previously established by Frère (9), the kinetic parameters of Galleni et al. and Galleni and Frère (10, 11) for the E. coli AmpC, and our own kinetic parameters for MIR-1. Enzyme concentrations were measured, during the exponential phase, on cells growing in NB medium at 37°C. One milliliter of culture was disrupted by using a nonionic detergent, B-PER II (Pierce, Rockford, IL). After a 5-min centrifugation at 20,000 × g, the supernatant was recovered and the β-lactamase activity was determined as the initial rate of hydrolysis of the saturating concentration of cephalothin prepared in buffer M. It is worth noting that these equations are valid only at the steady state, i.e., when the external concentration of antibiotic is constant, the production of enzyme is constitutive, and the half-life of the acyl-enzyme is short compared to the generation time of the bacteria.

RESULTS AND DISCUSSION

Purification of the plasmid-encoded AmpCs.

The goal of this work was to study the differences between the chromosomal and the ACT-1, MIR-1, CMY-1, and CMY-2 plasmid-encoded class C β-lactamases. Purification of these four enzymes was successfully performed using a method previously established for their chromosomal equivalents, confirming that they conserved similar physicochemical properties. The plasmid-encoded AmpCs exhibit relatively high isoelectric points (>8), allowing their purification at pH 6.5 or 7 with the help of cation-exchange chromatography. All the β-lactamases were purified in one or two chromatographic steps. In the cases of MIR-1 and CMY-1, the enzymes were eluted from an SP Sepharose column at NaCl concentrations of 40 and 75 mM, respectively. The purity of the enzyme preparations was higher than 95%. The purification yields of MIR-1 and CMY-1 were 20 and 60%, respectively. Purification of ACT-1 and CMY-2 was completed by affinity chromatography using 3-aminophenyl boronic acid as a ligand. This compound behaved as a competitive inhibitor with Ki values of 86 and 314 μM for ACT-1 and CMY-2, respectively. The β-lactamases were eluted by washing the column with a solution of 0.5 M sodium borate, 0.5 M NaCl. This step allowed the recovery of pure ACT-1 and CMY-2. Unfortunately, all the enzymes bound to the column could not be recovered. Therefore, the purification yields of ACT-1 and CMY-2 were 22% and 60%, respectively. The low purification yields were also due to the poor stability of the proteins during the different purification steps. The CMY-1 and CMY-2 enzymes precipitated when the protein concentration was higher than 3 mg/ml. The best storage conditions for the plasmid-encoded AmpCs were to keep the enzymes frozen at −20 or −70°C in 50 mM MOPS, pH 7, or 50 mM 2-(cyclohexylamino)ethanesulfonic acid, pH 10, at a concentration higher than 1 μM.
The N-terminal sequences of the different β-lactamases (NH2-PMSEKD, NH2-AAKTEQ, NH2-APMSEK, and NH2-GEAPSV for ACT-1, CMY-2, MIR-1, and CMY-1, respectively) were in agreement with the predicted N-terminal sequences. All the enzymes were found to exhibit the expected masses (39,379 versus 39,780 for ACT-1, 39,855 versus 39,850 for CMY-1, 39,160 versus 39,160 for MIR-1, and 38,755 versus 38,756 for CMY-2). In the case of CMY-1, the various gene manipulations resulted in the addition of five residues (GSGNH) at the C terminus.

Enzymatic titration by aztreonam.

The concentrations of active enzymes were checked by titration with a good inactivator, aztreonam. This experiment allowed us to verify that differences observed between the kinetic parameters of the plasmid-encoded and chromosomal class C enzymes were not due to under- or overestimations of the enzyme concentrations. Figure 1 shows the titration of ACT-1 (5 μM) by aztreonam. Residuals activities were determined using cephalothin as a reporter substrate. As expected, in the presence of excess aztreonam (>5 μM), the residual activity was close to zero (∼3%). In this case, the enzyme was inactivated when the [aztreonam]/[ACT-1] ratio was about 1.25. This result indicated an underestimation of the enzyme concentration, which could result in an overestimation of the kcat values. The presence of a small turnover of the ACT-1-aztreonam complex made it necessary to incubate the enzyme with a large excess of aztreonam to completely inactivate the β-lactamase. Similar results were obtained with the other plasmid-determined class C β-lactamases. [Enzyme]/[aztreonam] ratios of 0.9, 0.8, and 1.1 were found for CMY-2, CMY-1, and MIR-1, respectively. The kinetic data in Tables 2 and 3 were calculated accordingly using the active enzyme concentration, but these corrections did not significantly influence the results.

Kinetic parameters of plasmid-encoded class C β-lactamases.

First, we decided to compare the kinetic behavior of the CMY-1 enzyme we produced to that of the wild-type form of CMY-1. The CMY-1 enzyme was purified from the original strain. Its kinetic parameters versus cephalothin and cefoxitin were comparable to those obtained for the modified CMY-1 enzyme. We therefore decided to perform all the other experiments with this β-lactamase.
Substrate inhibition of the plasmid-encoded class C β-lactamases was observed at high concentrations of nitrocefin. While the data for the chromosomal enzyme fit the classical Henri-Michaelis equation, those for CMY-1 can be fitted to a curve characterized by the following equation:
\[ \[v{=}\frac{V_{\mathrm{max}}\ {\cdot}\ [S]}{[S]{+}([S]^{2}/K_{i}){+}K_{m}}\] \]
(3)
where S, Vmax, Km, and Ki are the substrate concentration, the maximum rate, the Henri-Michaelis constant, and the inhibition constant, respectively. The kcat, Km, and Ki values were estimated to be 2,220 ± 150 s−1, 230 ± 30 μM, and 660 ± 80 μM, respectively. Similar behavior was observed for the ACT-1 and CMY-2 β-lactamases, which indicated that the plasmid-encoded and the chromosomal AmpC β-lactamases have different behaviors toward nitrocefin.
A large number of studies have been performed on the in vitro susceptibility of E. coli strains producing plasmid-mediated class C β-lactamases (18). These investigations indicated that plasmid-encoded enzymes consistently conferred resistance to different β-lactams, such as penicillins, cephalosporins, cephamycins, and even monobactams. Nevertheless, it was not clear how the production of plasmid-derived enzymes modified the resistance patterns of the hosts. According to some authors, the acquisition of bacterial resistance was due to new catalytic properties of these class C β-lactamases (1). Some plasmid-encoded AmpCs received their names for their supposedly improved activities against one substrate or group of substrates. For example, CMY, FOX, and MOX are the acronyms for cephamycinase, cefoxitinase, and moxalactamase, respectively. By contrast, other authors indicated that plasmid-derived AmpCs are distinct from the chromosomal ones only in the localization of their genetic material (18). Table 2 shows the steady-state kinetic parameters of the studied plasmid-encoded and chromosomal class C β-lactamases for good substrates. In the case of the CMY-1 enzyme, we produced a β-lactamase containing five additional amino acids at the C-terminal end. In order to assess if the presence of extra amino acids modified the kinetic properties of the enzyme, we decided to produce and purify the CMY-1 β-lactamase (see above) from the E. coli strain obtained from Yunsop Chong. The steady-state kinetic constants were determined for cephalothin (kcat = 500 ± 50 s−1, Kmmes [measured Km] = 20 ± 4 μM, and kcat/Km = 25 μM−1 s−1), cefoxitin (kcat = 0.07 ± 0.01 s−1, kcat/Km = 1.1 μM−1 s−1, and Kmmes = 0.06 ± 0.015 μM). The data indicated that the presence of five amino acids at the C terminus of CMY-1 did not affect the catalytic properties of the β-lactamase.
The turnover of plasmid-encoded enzymes was typically smaller for penicillin than for cephalosporins. No major differences were seen between the two types of enzymes. In contrast to what was suggested on the basis of the MICs, CMY-1, ACT-1, and MIR-1 have lower catalytic efficiencies toward ampicillin (two to six times smaller) than the other class C enzymes. Nonetheless, some kinetic pecularities were observed, and in particular, the kcat/Km of cephaloridine for the ACT-1 and CMY-2 enzymes was somewhat smaller than for their chromosomal counterparts. Surprisingly, a biphasic kinetic progress curve was observed for the hydrolysis of cefazolin by CMY-2, preventing analysis with a simple kinetic model. This behavior was previously observed for this substrate with the chromosomal AmpC of Serratia marcescens (10). Nevertheless, these results showed that the activity profiles toward good substrates are similar for the tested plasmid-mediated and chromosomal class C β-lactamases.
Kinetic parameters for poor substrates are included in Table 3. kcat and Km values for cefoxitin, cefuroxime, and cefotaxime could be directly obtained for ACT-1, CMY-1, CMY-2, and MIR-1. When determined, the k+3 (rate constant of deacylation) values were similar to kcat, and consequently, k+3 was small compared to k+2 (rate constant of acylation). Also, the rate-limiting step of the hydrolytic pathway corresponds to the deacylation step, i.e., the hydrolysis of the acyl-enzyme complex. The kinetic parameters of ACT-1 and CMY-1 did not present significant differences compared to the chromosomal AmpCs. Nevertheless, CMY-1 exhibited a low Km value and a relatively low rate of acylation by cefuroxime. The CMY-2 enzyme was characterized by relatively low Km values against all the tested poor substrates. With oxacillin, the kcat value (0.015 s−1), which gives a minimum k+3 value, obtained by initial rate measurements was threefold higher than the k+3 value (0.005 s−1) obtained by monitoring the hydrolysis of nitrocefin in the presence of increasing concentrations of oxacillin. Interestingly, the measured and the computed values of Km are in good agreement. At the present time, no clear explanation can be proposed to account for the differences in the k+3 values.
The analysis of our kinetic data confirmed that the four plasmid-encoded β-lactamases exhibited the typical properties attributed to class C β-lactamases. The comparison of the chromosome-encoded enzyme parameters shows that most of the kinetic parameters of the plasmid-encoded enzymes lie between the extreme values obtained for the known chromosomal AmpCs. In particular, the CMY enzymes do not present increased activities toward cefoxitin. None of the four enzymes is less sensitive to aztreonam inactivation. On the contrary, CMY-2 presents the highest acylation rate observed so far for this substrate, and the deacylation constant is very low (k+3 < 10−5 s−1). Interestingly, a few kcat values showed a slight increase with some poor substrates. The catalytic efficiencies of the four plasmid-mediated enzymes toward cefuroxime were higher than those of chromosomal AmpCs. Moreover MIR-1 had a relatively high kcat against cefuroxime, cefotaxime, and oxacillin. However, as the Km values were also increased, its catalytic efficiency was not different from those of the chromosomal enzymes.

MICs of E. coli DH5α strains and overproduction of the plasmid-encoded class C β-lactamases.

The strains producing plasmid-encoded class C β-lactamases are characterized by very high MIC values for the different β-lactam families (18). In agreement with our kinetic study, these high resistances cannot be explained simply by increased activities against these compounds. These data raised a question. Indeed, if the four plasmid-mediated β-lactamases are not more active than their chromosomal equivalents, how can one explain the high MIC values associated with their production? To determine if β-lactamase overproduction was not responsible for these resistance profiles, we compared the MIC values for E. coli DH5α producing either the plasmid-encoded ACT-1 and MIR-1 enzymes or the chromosomal E. coli AmpC encoded by a plasmid (pAD7). This experiment was performed in two media of low or high osmolarity, the nutrient and the Luria-Bertani broths, respectively. First, MIC values obtained for the E. coli AmpC were similar to those obtained with the MIR-1 and ACT-1 enzymes (Table 4) (suppression). Determination of the enzyme concentration showed that all the transformants produced nearly 10 times more β-lactamase than was produced by the overproducer TE18 previously described by Nikaido and Normark (16). Such production may, as previously shown (13, 15, 21), cause a strong increase in resistance, even toward poor substrates.
Second, a comparison was made between the observed and computed MIC values. As shown in Table 4, in the absence of β-lactamase production, the experimental values coincide with the calculated ones, indicating that our results agree with the Ipl values established by Frère (9). For the β-lactamase producers, the calculated and observed MIC values are in good agreement in the cases of ampicillin, cefotaxime, and cefoxitin (a factor of 2 or less) but less good for cephalothin and aztreonam (Table 4). For these two compounds, the predicted resistances of β-lactamase-producing strains were overestimated. For the MIR-1 producer, our theoretical model showed how the slightly increased activities toward cefotaxime and cefuroxime engendered higher MICs. Two pecularities could be observed with the ACT-1 producer. First, MIC values for aztreonam increased less significantly than for other compounds, and second, a very high level of resistance to ampicillin was observed. Unfortunately, the variable production rate of ACT-1 during the culture prevented a correct estimation of the enzyme concentration and thus a prediction of its MIC values.
Finally, considering the effect of the medium, for all substrates tested, higher MIC values were obtained in LB medium. This was especially striking in the case of aztreonam, for which the difference reached a factor of 16. This effect could be correlated with the levels of the OmpF and OmpC porins. Indeed, the production of these proteins is related to the osmolarity of the medium (2). In media of high osmolarity (like LB), the synthesis of the larger OmpF porin is repressed in E. coli. Since this protein enhances the entry of several β-lactam antibiotics (15), it is clear that MIC values must be influenced by the medium osmolarity. In agreement with this hypothesis, we observed that this phenomenon was especially marked for aztreonam. Indeed, its two negative charges retard penetration through the OmpC porin, which is more selective than OmpF (15).

Conclusions

Bacterial resistance to expanded-spectrum β-lactam antibiotics associated with the production of class C β-lactamases has taken two different paths. The first is represented by the production of the GC-1 enzyme. In this case, a tandem tripeptide insertion in the β-lactamase modifies the structure of the omega loop and induces a broadening of the β-lactamase activity (3, 17). The second path is well illustrated by the four enzymes studied in the present paper, showing that resistance could originate from overproduction of plasmid-mediated enzyme. Since their discovery, emergence of new plasmid-encoded β-lactamases has not ceased to occur. In addition, an increasing number of variants have been isolated. These facts enhance the probability of obtaining a true extended-spectrum enzyme. Consequently, these β-lactamases also deserve complete biochemical characterization.
FIG. 1.
FIG. 1. Titration curve of the ACT-1 β-lactamase by aztreonam. The residual activity of ACT-1 versus the [aztreonam]/[ACT-1] ratio is shown. The linear regression allows the determination of the actual concentrations of active enzyme. Eth, theoretical values of enzyme concentration.
TABLE 1.
TABLE 1. Nucleotidic sequences of primers used for PCR
PrimeraSequencebTm (°C)c
ACT-1forNdel5′ -GGGCATATGATGATGACTAAATCCCTTTGCTGCGC-3′65.6
ACT-1revHindIII5′ -CCCAAGCTTCTACAGCGCGCTCAAAATACGTGTATGC-3′64.3
CMY-1forNdel5′ -GGGCATATGCAACAACGACAATCCATCCTGTG-3′62.6
CMY-1revHindIII5′ -CCCAAGCTTCAACCGGCCAACTGCGCCAGGAT-3′66.7
CMY-2forNdel5′ -GGGCATATGATGAAAAAATCGTTATGCTGCGCTCTG-3′65.5
CMY-2revHindIII5′ -CCCAAGCTTATTGCAGCTTTTCAAGAATGCGCCA-3′65.1
MIR-1forNdel5′ -GGGCATATGATGACAAAATCCCTAAGCTGTGCCCT-3′64.5
MIR-1revHindIII5′ -CCCAAGCTTACTGCAGCGCGTCGAGGATACG-3′65.1
a
for, forward primer; rev, reverse primer.
b
Restriction site sequences are underlined.
c
Tm, melting temperature.
TABLE 2.
TABLE 2. Kinetic parameters of plasmid-encoded AmpC β-lactamases for good substratesa
β-LactamaseAntibioticsk cat (s−1)Km (μM)bk cat/Km (μM−1 · s−1)
ACT-1/(P99)Nitrocefin400 ± 80 (780)22 ± 2 (25)18 ± 2 (31)
 Cephaloridine>250 (700)>200 (70)1.3 ± 0.1 (10)
 Cefazolin560 ± 70 (3,000)430 ± 40 (1,500)1.3 ± 0.1 (2)
 Cephalothin460 ± 30 (200)38 ± 4 (9)12 ± 4 (20)
 Cephalexin85 ± 10 (70)42 ± 4 (10)2 ± 0.2 (7)
 Benzylpenicillina55 ± 3 (14)2.1 ± 0.2 (0.6)26 ± 3 (23)
 Ampicillina1 ± 0.2 (0.7)1.7 ± 0.2 (0.4)0.6 ± 0.01 (1.8)
MIR-1/(P99)Nitrocefin765 ± 15 (780)8 ± 0.8 (25)95 ± 5 (31)
 Cephaloridine215 ± 20 (700)93 ± 9 (70)2.3 ± 0.2 (10)
 CefazolinBiphasic kineticBiphasic kineticBiphasic kinetic
 Cephalothin160 ± 20 (200)2.1 ± 0.2 (9)76 ± 14 (20)
 Cephalexin55 ± 5 (70)16 ± 0.2 (10)3.5 ± 0.3 (7)
 Benzylpenicillina14 ± 0.5 (14)0.4 ± 0.04 (0.6)35 ± 5 (23)
 Ampicillina0.55 ± 0.05 (0.7)0.16 ± 0.02 (0.4)3.4 ± 0.1 (1.8)
CMY-1/(CAV-1)Nitrocefin2220 ± 150c230 ± 27c10 ± 1c
 Cephaloridine155 ± 10 (300)110 ± 11 (770)1.4 ± 0.2 (0.39)
 Cefazolin325 ± 854 ± 56 ± 0.6
 Cephalothin480 ± 25 (540)30 ± 3 (500)16 ± 4 (1.08)
 Cephalexin115 ± 614 ± 18 ± 1.2
 Benzylpenicillina13 ± 0.5 (5)1 ± 0.1 (8.7)13 ± 2 (0.57)
 Ampicillina0.45 ± 0.052.2 ± 0.20.2 ± 0.02
CMY-2/(Citrobacter freundii)Nitrocefin765 ± 15 (330)8 ± 0.8 (12)95 ± 5 (28)
 Cephaloridine215 ± 20 (700)93 ± 9 (35)2.3 ± 0.2 (20)
 CefazolinBiphasic kineticBiphasic kineticBiphasic kinetic
 Cephalothin160 ± 20 (210)2.1 ± 0.2 (13)76 ± 14 (16)
 Cephalexin55 ± 5 (150)16 ± 0.2 (12)3.5 ± 0.3 (12,5)
 Benzylpenicillina14 ± 0.5 (31)0.4 ± 0.04 (0.4)35 ± 5 (75)
 Ampicillina0.55 ± 0.05 (6.5)0.16 ± 0.02 (0.2)3.4 ± 0.1 (3)
a
Experiments were performed in 50 mM MOPS, pH 7, 50 mM NaCl. Names and numbers between parentheses correspond to the chromosomal class C β-lactamases and their steady-state kinetic parameters, respectively. Data taken from references 10 and 11.
b
For these substrates, the Km value was determined as a Ki value using cephalothin as the reporter substrate.
c
Computed for the substrate inhibition model.
TABLE 3.
TABLE 3. Kinetic parameters of plasmid-encoded AmpC β-lactamases for poor substratesa
β-LactamaseAntibioticsk cat (s−1)k cat/Km (μM−1 · s−1)k +3 b (s−1)k +2/Kb (μM−1 · s−1)Km calc c (μM)Km mes (μM)
ACT-1/(P99)Cefoxitin0.37 (0.06)0.74 (2.5)NDNDND0.5 (0.024)
 Cefuroxime0.12 (0.05)6.3 (3.1)NDNDND0.019 (0.016)
 Cefotaxime0.05 (0.015)0.7 (1.5)NDNDND0.07 (0.01)
 Imipenem0.011 (0.003)0.030.009 (0.002)0.024 (0.06)0.38 (0.04)0.37
 AztreonamNDND0.0021 (0.0002)0.18 (0.26)0.012 (0.0012)ND
 Oxacillin0.0440.5ND0.65ND0.09
MIR-1/(P99)Cefoxitin0.64 (0.06)0.7 (2.5)NDNDND0.75 (0.024)
 Cefuroxime3.4 (0.05)5.7 (3.1)NDNDND0.6 (0.016)
 Cefotaxime2.7 (0.015)0.67 (1.5)NDNDND4 (0.01)
 Imipenem0.012 (0.003)0.080.007 (0.002)0.05 (0.06)0.14 (0.04)0.15
 AztreonamNDND0.0016 (0.0002)0.218 (0.26)0.008 (0.0012)0.004
 Oxacillin0.181.8NDNDND0.1
CMY-1/(CAV-1)Cefoxitin0.05 (0.5)0.9 (1.25)NDNDND0.055 (0.4)
 Cefuroxime0.024 ± 0.4NDNDND0.005
 Cefotaxime0.01 (0.2)0.67 (2)NDNDND0.015 (0.1)
 Imipenem0.0020.040.0070.0430.160.05
 AztreonamNDND<0.0080.36<0.020.01
 Oxacillin0.0090.450.0160.330.050.02
CMY-2/(Citrobacter freundii)Cefoxitin0.23 (0.32)3.3 (1.3)NDNDND0.07 (0.026)
 Cefuroxime0.017 (50.04)15 (2)0.0146.40.00220.0011 (0.02)
 Cefotaxime0.004 (0.016)3.3 (3.4)0.0032.90.0010.0012 (0.006)
 Imipenem0.033 (0.017)NDBiexponential kineticBiexponential kineticBiexponential kineticND
 AztreonamNDND<0.0060 (0.0003)2 (0.18)<0.003 (0.0014)ND
 Oxacillin0.015250.0056.50.00080.0006
a
Experiments were performed in 50 mM MOPS, pH 7, 50 mM NaCl. ND, not determined. Data taken from references 10 and 11. Names and numbers between parentheses correspond to the chromosomal class C β-lactamases and their steady-state kinetic parameters, respectively. Standard deviations are within 10%.
b
Values obtained by using cephalothin as the reporter substrate.
c
calc, calculated.
TABLE 4.
TABLE 4. Observed and calculated MICs of ampicillin, cefuroxime, cefotaxime, cephalothin, cefoxitin and aztreonam for E. coli DH5α alone and producing MIR-1, ACT-1, and E. coli AmpC
AntibioticMIC (μM)a for strain:           
 DH5α  T1 (MIR-1)  T2 (ACT-1)  T3 (AmpC)  
 O LBO NBCO LBO NBCO LBO NBCO LBO NBC
Ampicillin2.71.40.7-1.43451723485,5142,757ND1,379345840-1000
Cefuroxime4.52.2ND2,2942,294ND28736-72ND573287ND
Cefotaxime0.070.070.08-0.12341341308.44.2ND8.48.44
Aztreonam0.290.040.08-0.14734.6239-180.6ND734.614-16
Cephalothin10-194.81.9-3.72,4472,44735,0002,4472,447ND2,4472,4476,000-6,700
Cefoxitin4.5-94.42.5-5570570850-900285-570142ND570142388
a
Observed (O) MIC values in LB (OLB) or NB (ONB) media are compared with the calculated values (C). The Ipl values were those previously established by Frère (9). The enzyme concentrations were 1.3 and 0.845 nmol/mg (dry weight) for the AmpC and MIR-1 producers, respectively. In the case of the E. coli DH5α strain, no β-lactamase activity was detected. ND, not determined.

Acknowledgments

We are grateful to I. Thamm (Centre d'Ingénierie des Protéines, University of Liège) for her excellent technical assistance.
C.B. holds FRIA fellowships (Brussels). We thank the Belgian Program on Interuniversity Poles of Attraction initiated by the Belgian State, Prime Minister's Office, Science Policy Programming (PAI P5/33, from the Fund for Joint Basic Research of Belgium, contract no. 2.4576.97) and the targeted program COBRA, financed by the European Commission (no. LSHM-CT-2003-503335).

REFERENCES

1.
Bauernfeind, A., Y. Chong, and S. Schweighart. 1989. Extended broad spectrum beta-lactamase in Klebsiella pneumoniae including resistance to cephamycins. Infection17:316-321.
2.
Cai, S. J., and M. Inouye. 2002. EnvZ-OmpR interaction and osmoregulation in Escherichia coli. J. Biol. Chem.277:24155-24161.
3.
Crichlow, G. V., A. P. Kuzin, M. Nukaga, K. Mayama, T. Sawai, and J. R. Knox. 1999. Structure of the extended-spectrum class C beta-lactamase of Enterobacter cloacae GC1, a natural mutant with a tandem tripeptide insertion. Biochemistry38:10256-10261.
4.
Davies, J. 1994. Inactivation of antibiotics and the dissemination of resistance genes. Science264:375-382.
5.
De Meester, F., B. Joris, G. Reckinger, C. Bellefroid-Bourguignon, J. M. Frere, and S. G. Waley. 1987. Automated analysis of enzyme inactivation phenomena. Application to beta-lactamases and DD-peptidases. Biochem. Pharmacol.36:2393-2403.
6.
Edlund, T., T. Grundstrom, and S. Normark. 1979. Isolation and characterization of DNA repetitions carrying the chromosomal beta-lactamase gene of Escherichia coli K-12. Mol. Gen. Genet.173:115-125.
7.
Fosberry, A. P., D. J. Payne, E. J. Lawlor, and J. E. Hodgson. 1994. Cloning and sequence analysis of blaBIL-1, a plasmid-mediated class C beta-lactamase gene in Escherichia coli BS. Antimicrob. Agents Chemother.38:1182-1185.
8.
Frère, J. M. 1995. Beta-lactamases and bacterial resistance to antibiotics. Mol. Microbiol.16:385-395.
9.
Frère, J. M. 1989. Quantitative relationship between sensitivity to beta-lactam antibiotics and beta-lactamase production in gram-negative bacteria-I. Steady-state treatment. Biochem. Pharmacol.38:1415-1426.
10.
Galleni, M., G. Amicosante, and J. M. Frère. 1988. A survey of the kinetic parameters of class C beta-lactamases. Cephalosporins and other beta-lactam compounds. Biochem. J.255:123-129.
11.
Galleni, M., and J. M. Frère. 1988. A survey of the kinetic parameters of class C beta-lactamases. Penicillins. Biochem. J.255:119-122.
12.
Heinemann, J. A. 1999. How antibiotics cause antibiotic resistance. Drug Discov. Today4:72-79.
13.
Lakaye, B., A. Dubus, S. Lepage, S. Groslambert, and J. M. Frère. 1999. When drug inactivation renders the target irrelevant to antibiotic resistance: a case story with beta-lactams. Mol. Microbiol.31:89-101.
14.
Matagne, A., A. Dubus, M. Galleni, and J. M. Frère. 1999. The beta-lactamase cycle: a tale of selective pressure and bacterial ingenuity. Nat. Prod. Rep.16:1-19.
15.
Nikaido, H. 1985. Role of permeability barriers in resistance to beta-lactam antibiotics. Pharmacol. Ther.27:197-231.
16.
Nikaido, H., and S. Normark. 1987. Sensitivity of Escherichia coli to various beta-lactams is determined by the interplay of outer membrane permeability and degradation by periplasmic beta-lactamases: a quantitative predictive treatment. Mol. Microbiol.1:29-36.
17.
Nukaga, M., S. Kumar, K. Nukaga, R. F. Pratt, and J. R. Knox. 2004. Hydrolysis of third-generation cephalosporins by class C beta-lactamases. Structures of a transition state analog of cefotoxamine in wild-type and extended spectrum enzymes. J. Biol. Chem.279:9344-9352.
18.
Philippon, A., G. Arlet, and G. A. Jacoby. 2002. Plasmid-determined AmpC-type beta-lactamases. Antimicrob. Agents Chemother.46:1-11.
19.
Rupp, M. E., and P. D. Fey. 2003. Extended spectrum beta-lactamase (ESBL)-producing Enterobacteriaceae: considerations for diagnosis, prevention and drug treatment. Drugs63:353-365.
20.
Sykes, R. B., and K. Bush. 1982. Physiology, biochemistry and inactivation of β-lactamases, p. 155-208. In R. B. Morin and M. Gorman (ed.), The chemistry and biology of beta-lactam antibiotics. Academic Press, New York, N.Y.
21.
Tracz, D. M., D. A. Boyd, L. Bryden, R. Hizon, S. Giercke, P. Van Caeseele, and M. R. Mulvey. 2005. Increase in ampC promoter strength due to mutations and deletion of the attenuator in a clinical isolate of cefoxitin-resistant Escherichia coli as determined by RT-PCR. J. Antimicrob. Chemother.55:768-772.

Information & Contributors

Information

Published In

cover image Antimicrobial Agents and Chemotherapy
Antimicrobial Agents and Chemotherapy
Volume 49Number 10October 2005
Pages: 4240 - 4246
PubMed: 16189104

History

Received: 1 March 2005
Revision received: 17 May 2005
Accepted: 13 July 2005
Published online: 1 October 2005

Permissions

Request permissions for this article.

Contributors

Authors

Cédric Bauvois
Centre d'Ingénierie des Protéines, Université de Liège, Sart Tilman, Belgium
Akiko Shimizu Ibuka
Department of Food and Nutritional Sciences, University of Shizuoka, 52-1 Yada, Shizuoka 422-8526, Japan
Almeida Celso
Centre d'Ingénierie des Protéines, Université de Liège, Sart Tilman, Belgium
Jimena Alba
Department of Microbiology, Toho University School of Medicine, 5-21-16 Omorinishi, Ota-ku, Tokyo 1438540, Japan
Yoshikazu Ishii
Department of Microbiology, Toho University School of Medicine, 5-21-16 Omorinishi, Ota-ku, Tokyo 1438540, Japan
Jean-Marie Frère
Centre d'Ingénierie des Protéines, Université de Liège, Sart Tilman, Belgium
Moreno Galleni [email protected]
Centre d'Ingénierie des Protéines, Université de Liège, Sart Tilman, Belgium

Metrics & Citations

Metrics

Note:

  • For recently published articles, the TOTAL download count will appear as zero until a new month starts.
  • There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.
  • Citation counts come from the Crossref Cited by service.

Citations

If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.

View Options

Figures and Media

Figures

Media

Tables

Share

Share

Share the article link

Share with email

Email a colleague

Share on social media

American Society for Microbiology ("ASM") is committed to maintaining your confidence and trust with respect to the information we collect from you on websites owned and operated by ASM ("ASM Web Sites") and other sources. This Privacy Policy sets forth the information we collect about you, how we use this information and the choices you have about how we use such information.
FIND OUT MORE about the privacy policy