Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, May 2008, p. 1806-1811, Vol. 52, No. 5
0066-4804/08/$08.00+0 doi:10.1128/AAC.01381-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Antibiotic Resistance Unit, Department of Infectious Diseases, National Institute of Health Dr. Ricardo Jorge, Lisbon, Portugal,1 University Clermont1, UFR Médecine, Laboratoire de bactériologie, EA3844, Clermont-Ferrand F-63001, France,2 CHU Clermont-Ferrand, Centre de Biologie, Laboratoire de bactériologie clinique, Clermont-Ferrand F-63003, France,3 Laboratory of Microbiology, Hospital de Santa Maria, Lisbon, Portugal4
Received 25 October 2007/ Returned for modification 14 January 2008/ Accepted 25 February 2008
|
|
|---|
atom of Ser130 around 3.5 Å away from the key O
atom of the reactive serine (Ser70). This movement may therefore decrease the susceptibility to clavulanic acid by preventing cross-linking between Ser130 and Ser70. |
|
|---|
TEM and SHV enzymes initially had preferential activity against penicillins, as in the case of enzymes SHV-1 and TEM-1. Oxyimino β-lactams that are resistant to their hydrolytic activity and β-lactam inhibitors, such as clavulanic acid and tazobactam, have been developed to get around the activities of these enzymes (6, 9, 26). Nevertheless, the presence of point mutations in TEM and SHV enzymes has expanded the substrate spectrum to include oxyimino β-lactams and/or has conferred resistance to the inhibitors (3, 6, 9, 26).
More than 28 inhibitor-resistant TEM enzymes have been detected. They harbor amino acid substitutions at positions 69, 130, 165, 182, 244, 275, and/or 276 that confer resistance to inhibitors (9, 15). Only three natural inhibitor-resistant SHVs (IRS) have been reported (http://www.lahey.org/studies). It has been proposed that substitutions at positions 69, 130, and 187 are involved in their resistance to inhibitors (10, 14, 31). IRS enzymes have also been constructed in vitro by site saturation mutagenesis at position 244 (37).
In this study, we performed a phenotypic, molecular, and biochemical characterization of the new IRS-type β-lactamase SHV-72 from a clinical K. pneumoniae strain and investigated by molecular-dynamics simulations (MDSs) the role of the Lys234Arg substitution in its resistance to clavulanic acid.
|
|
|---|
ampC and plasmid pBK-CMV (Stratagene, Amsterdam, The Netherlands) were used for the cloning experiments (Table 1) (13). blaSHV-1-encoding E. coli C600 was used as a control strain for PCR. |
View this table: [in a new window] |
TABLE 1. Strains and plasmids used in this study
|
Isoelectric focusing. Cell extracts were obtained by ultrasonic treatment, and isoelectric focusing was performed with polyacrylamide gels containing ampholines with a pH range of 3.5 to 9.5, as previously described (4), with IRT-2 (pI 5.2), TEM-1 (5.4), TEM-2 (pI 5.6), OXA-1 (pI 7.4), SHV-1 (pI 7.6), CTX-M-15 (pI 8.9), and AmpC (pI 9.2) as standards.
Amplification, sequencing, and cloning of β-lactamase genes.
The β-lactamase-encoding genes were detected and sequenced, as described elsewhere (28). The complete SHV-encoding open reading frame of the blaSHV-1 and blaSHV-72 genes was amplified with the specific primers SHVsf and SHVsr (28). The blaSHV-1 and blaSHV-72 genes were cloned as follows: proofreading Isis DNA polymerase (Qbiogene, Irvine, CA) was used for blaSHV-72, and proofreading iProofTM high-fidelity DNA polymerase (Bio-Rad Laboratories Inc., Hercules, CA) was used for blaSHV-1. PCR products were ligated in the SmaI site of the plasmid pBK-CMV, and the recombinant plasmids were electroporated in E. coli DH5
ampC. The transformants harboring the recombinant SHV-encoding plasmids (pBK-SHV-72 and pBK-SHV-1) were selected on Mueller-Hinton agar supplemented with 30 µg/ml kanamycin and 16 µg/ml ticarcillin. The sequence and the orientation of the inserted open reading frames were determined from PCR experiments, which were performed with different combinations of the primers pBK-CMV1' (5'-CTAGTGGATCCAAAGAATTCAAAAAGC-3'), pBK-CMV2' (5'-AATTGGGTACACTTACCTGGTACCC-3'), SHVsf, and SHVsr.
β-Lactamase preparation. The SHV-producing clones were grown for 18 h at 37°C in 6 liters of Luria-Bertani broth complemented with yeast extract, 30 µg/ml of kanamycin, and 16 µg/ml of ticarcillin. After centrifugation, bacterial pellets were suspended with MES-NaOH (20 mM) (pH 6) and disrupted by ultrasonic treatment as previously described (5). The extract was then clarified by centrifugation and treated with DNase I (Roche Applied Science, Meylan, France). Purification was carried out by ion-exchange chromatography with an SP Sepharose column or a HiPrep 16/10 SP HF column (Amersham Pharmacia Biotech) and gel filtration chromatography with a Superose 12 or HiPrep 16/60 Sephacryl S-100 HR column (Amersham Pharmacia Biotech), using a fast-protein liquid chromatography system as previously described (4). The protein concentration was estimated by using the BCA protein assay kit (Pierce, Rockford, IL). The purity of enzymes was estimated by using sodium dodecyl sulfate-polyacrylamide gel electrophoresis as previously described (4).
Determination of β-lactamase kinetic constants. The Michaelis constant (Km) and catalytic activity (kcat) of the SHV-1 and SHV-72 enzymes were obtained with purified extracts by a computerized microacidimetric method, using a 702 SM Titrino pHstat apparatus (Metrohm, Herisau, Switzerland) (20). These kinetic parameters were determined by analysis of the complete hydrolysis time courses, and the kinetic progress curves were fitted by nonlinear least-squares regression (20). The concentrations of the β-lactamase inhibitors (clavulanate and tazobactam) necessary to inhibit the enzyme activity by 50% (IC50s) were determined as described elsewhere (4), using 200 µM ticarcillin as the reporter substrate. The kinetic constants were determined three times for each substrate tested.
MDS. The model of the mutant enzyme (SHV-72) was constructed on the basis of the SHV-1 crystal structure (19). The SHV-72 and SHV-1 enzymes were solvated with water in a periodic cubic box that was large enough to contain the system and 1 nm of solvent on all sides. Version 1.8.2 of the VMD software package was used to manipulate the two systems (16). The GROMACS software package, version 3.2 (24), and the geometric and charge parameters of the OPLSAA (optimized potentials for liquid simulations in all-atom) force field (18) were used to carry out all energy minimizations and MDSs. TIP3P parameters were used for the water molecules (17). The particle-mesh Ewald method was used to treat long-range electrostatics (12). All covalent bond lengths were constrained by the SHAKE algorithm (32) with a relative tolerance of 10–4. The systems were equilibrated as reported previously (27), and MDSs of 400 ps were then made with a time step of 1.5 fs and coordinates collected every 0.0015 ps. The velocities of all atoms were generated from a Maxwellian distribution. The temperature was kept constant at 300 K, while the pressure was kept constant by the weak coupling constant of 1 bar using Berendsen's algorithms (25).
Nucleotide sequence accession number. The new blaSHV nucleotide sequence was submitted to the EMBL nucleotide sequence database as blaSHV-72 with accession number AM176547.
|
|
|---|
|
View this table: [in a new window] |
TABLE 2. MICs of β-lactam antibiotics for the clinical K. pneumoniae strain INSRA1229, SHV-72- and SHV-1-producing transformants, and the recipient E. coli DH5 ampC
|
-URA1229, exhibited a β-lactam resistance phenotype similar to that of the clinical strain (Table 2).
Biochemical properties of β-lactamase SHV-72.
The clinical strain and the corresponding clone produced only β-lactamases of pI 7.6, which is compatible with the amino acid sequence of SHV-72. SHV-72 and SHV-1 were purified from the E. coli clones by ion exchange and gel filtration. The rate of purity was estimated to be
96% for SHV-72 and
95% for SHV-1 on a sodium dodecyl sulfate-polyacrylamide gel as a band of 28 kDa, which corresponded to the molecular mass deduced from the amino acid sequence (data not shown).
The kinetic parameters of SHV-72 were determined for seven β-lactams and compared with those of SHV-1 (Table 3). kcat values of SHV-72 against penicillins were higher than those of SHV-1, and Km values were comparable or slightly higher (19 to 43 µM versus 11 to 31 µM). Overall, the catalytic efficiency against penicillins was slightly higher for SHV-72 (kcat/Km, 36 to 286 µM–1·s–1) than for SHV-1 (kcat/Km, 20 to 84 µM–1·s–1), except for piperacillin (kcat/Km, 45 µM–1·s–1 versus 62 µM–1·s–1). In contrast with penicillins, SHV-72 had a lower catalytic efficiency against cephalothin (11-fold) than SHV-1. Neither SHV-1 nor SHV-72 exhibited catalytic activity against oxyimino β-lactams.
|
View this table: [in a new window] |
TABLE 3. Kinetic parameters of SHV-72 and SHV-1
|
MDS.
The enzyme SHV-72 was modeled from the crystallographic structure of SHV-1 (19). The behaviors of SHV-72 and SHV-1 were compared during MDSs of 400 ps at a temperature of 300 K. The MDS for each model were checked for stability by monitoring several overall properties, such as the radius of gyration, secondary structure, root mean squared deviation (RMSD) from the initial structure, and the kinetic and potential energies (data not shown). These parameters were found to stabilize after about 100 ps, and hence, data from 300 ps were used for all subsequent analyses. The radius of gyration and the RMSDs of C
atoms were similar to those for the crystallographic structure of SHV-1 (Table 4). The secondary structure was also preserved during the simulation (Table 4). The relatively small deviations were further evidence of the inherent stability of the model and indicated that the dynamic structures of the models remained in the realm of the crystal SHV-1 geometry during the course of the simulation. The largest fluctuations were localized in loops connecting the secondary structure elements, as is usual in MDSs of proteins (data not shown). The introduction of the Ala146Val and Lys234Arg substitutions caused no overall or large-scale deviation of the dynamic properties.
|
View this table: [in a new window] |
TABLE 4. Summary of statistical data for 300-ps MDSs
|
of positions 70 and 140, 16.7 Å), did not modify the positioning of surrounding residues. Residue Arg234 is located in the catalytic site and adopted the conformation observed in the crystallographic structure of the class A β-lactamase PSE-4 (Fig. 1A) (24). Overall, the architectures of the active site were identical for the two enzymes SHV-72 and SHV-1. All residues of the active site except residue Ser130 had similar positioning in SHV-1 and SHV-72.
![]() View larger version (15K): [in a new window] |
FIG. 1. The two conformations of the Ser130 side chain. (A) SHV-72 exhibiting a Ser130 1 angle value of –145°. (B) SHV-1 exhibiting a Ser130 1 angle value of –60°. 1 is the angle between the plane containing the atoms N, CA, and CB and the plane containing the atoms CA, CB, and O .
|
1 angle of Ser130 was around –145° in the starting models (Fig. 1A and 2A and C), as observed in the crystal structures of class A enzymes such as the TEM-, SHV- and CTX-M-type enzymes (11, 19, 29). After 130 ps of MDS, the Ser130
1 angle of SHV-72 increased to –61 ± 11° and thereafter remained stable until the end of the simulation (Fig. 2A). In SHV-1, the
1 angle kept the value of –145 ± 12° during almost all of the simulation (Fig. 2C).
![]() View larger version (17K): [in a new window] |
FIG. 2. 1 angle of residue Ser130 during 300-ps MDSs. (A) SHV-72 with a starting 1 angle of –148°. (B) SHV-72 with a starting 1 angle of –60°. (C) SHV-1 with a starting 1 angle of –148°. (D) SHV-1 with a starting 1 angle of –60°.
|
1 angle of Ser130 was set at –61° in the starting models by manual modeling (Fig. 1B and 2B and D). For SHV-72, the
1 angle value of Ser130 was –67 ± 10° and stable (Fig. 2B). In contrast, the Ser130
1 angle of SHV-1 decreased after 160 ps from –58 ± 10° to –153 ± 20° and remained stable until the end of the MDS (Fig. 2D). The Lys234Arg substitution thus stabilized an atypical conformation of the Ser130 side chain. This conformation was characterized by an
1 angle between –50° and –77°, which moved the O
atom of Ser130 around 3.5 Å away from the key O
atom of the reactive serine (Ser70). |
|
|---|
atoms 146 and 70,
17 Å), and the Lys234Arg substitution in the catalytic pocket. To our knowledge, this last substitution is observed for the first time in a natural SHV/TEM-type β-lactamase (23). The resulting new enzyme induced a resistance phenotype compatible with that of an inhibitor-resistant penicillinase. No effect was observed on the imipenem MIC, despite the presence of the Ala146Val substitution (30). The kinetic constant confirmed a high catalytic efficiency against penicillins, no significant activity against oxyimino β-lactams, and a decreased susceptibility to clavulanate in comparison with results for SHV-1. Only the three SHV-type enzymes SHV-10, SHV-26, and SVH-49 have been previously described as resistant to inhibitors. In contrast to SHV-72, these enzymes have an increase in Km values and/or a decrease in kcat values against penicillins (10, 14, 31). In SHV-10 and SHV-49, the amino acid substitutions Ser130Gly and Met69Ile reduced activity against β-lactam substrates (14, 31). The introduction of Lys234Arg in TEM-1 by site-directed-mutagenesis experiments induce a 10-fold decrease of the affinity against penicillins (23). The mutations observed in SHV-72 did not significantly affect Km values and did not decrease catalytic activities against penicillins.
Among IRSs, SHV-10 is the most resistant (with an IC50 41-fold higher than that for SHV-1) to inhibitors and SHV-26 the least (with an IC50 threefold higher than that for SHV-1) (10, 30). SHV-72, like SHV-49 (14), was 10-fold more resistant to clavulanic acid than SHV-1. The Ser130Gly substitution in SHV-10 also induces resistance to inhibitors in TEM-, OXY- and CTX-M-type enzymes (2, 22, 31, 33). The mechanism of inhibition by clavulanic acid is based on the formation of a covalent cross-link between O
atoms of Ser70 and Ser130 by residual atoms of the inhibitor. In enzymes harboring Gly130, this residue, which is deprived of the side chain, prevents cross-linking with position 70 (34-36). In SHV-49, the Met69Ile substitution is responsible for resistance to inhibitors (14). By analogy with inhibitor-resistant TEMs (38), substitution at position 69 may decrease the susceptibility to inhibitors because of the modification of Ser130 side-chain positioning. In SHV-72, the Lys234Arg substitution is probably responsible for resistance to inhibitors owing to its location in the catalytic pocket and in the vicinity of residue 130.
To understand the role of Arg234 in resistance to clavulanic acid, SHV-72 was modeled from the SHV-1 crystal structure (19) and analyzed during MDSs. SHV-72 and SHV-1 models exhibited different behaviors only in the local region of residue 234. In Lys234-harboring class A β-lactamases, the conformation of the Ser130 side chain is such that its
1 values are in the range of –120.5° to –163.5°, as for SHV-1 (–140°). In SHV-72 MDSs, an alternative conformation of the Ser 130 side chain (
1
–64°) appeared because of a hydrogen bond with Arg234, as previously observed in the crystal structure of the Arg234-harboring enzyme PSE-4 (24). This alternative conformation is probably stabilized because of the restricted ability of Arg234 to move and establish a hydrogen bond, since hydrogen bonds are favorable only in the plane of the rigid arginine guanidium group. This conformation is probably involved in the weak susceptibility to inhibitors of SHV-72.
The change in the
1 angle by around –64° moved the Ser130 O
atom away from the reactive Ser70 O
atom. This movement of the O
atom of Ser130, which is the ultimate covalent attachment point for the inhibitors, may therefore prevent the cross-link. Such a resistance mechanism has been previously proposed for TEM-32, for which the Met69Ile substitution induces, by another mechanism, the same movement of the Ser130 side chain (
1 angle, –64°) (38). The effects of the Met69Ile, Ser130Gly, and Lys234Arg substitutions therefore share a common logic for inhibitor resistance.
The O
atom of Ser130 plays a role in the hydrolysis of β-lactams (21). In SHV-72, this role could presumably be disrupted. However, the enzyme did not lose its catalytic efficiency. This behavior may be explained by the coexistence of the two Ser130 conformations and/or the replacement of the Ser130 O
atom by a water molecule, as observed in the crystal structures of the enzymes deprived of Ser130, such as SHV-10 and TEM-76 (35, 36).
In conclusion, we report a new SHV-type penicillinase resistant to clavulanic acid. The resistance to clavulanic acid is induced by the Lys234Arg substitution, which probably affects the positioning of the Ser130 side chain, a key element of the inhibition reaction mediated by clavulanic acid.
This work was supported financially by the project POCTI/2001/ESP/43037 from Fundação para a Ciência e a Tecnologia, Lisbon, Portugal, awarded to M. Caniça, and by a grant from the Ministère de l'Education Nationale, de l'Enseignement Supérieur et de la Recherche, France. N. Mendonça and V. Manageiro were supported by the grants BIC 03/2003-I from NIH Dr. Ricardo Jorge and SFRH/BD/32578/2006 from Fundação para a Ciência e a Tecnologia, Lisbon, Portugal, respectively, and both were also the recipients of short-term research fellowship grants from the Federation of European Microbiological Societies (FEMS).
Published ahead of print on 3 March 2008. ![]()
|
|
|---|
Gly. Antimicrob. Agents Chemother. 45:2269-2275.This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»