The High Resolution Crystal Structure for Class A β-Lactamase PER-1 Reveals the Bases for Its Increase in Breadth of Activity
2000; Elsevier BV; Volume: 275; Issue: 36 Linguagem: Inglês
10.1074/jbc.m003802200
ISSN1083-351X
AutoresS. Tranier, Anne-Typhaine Bouthors, Laurent Maveyraud, Valérie Guillet, Wladimir Sougakoff, Jean‐Pierre Samama,
Tópico(s)Antimicrobial Resistance in Staphylococcus
ResumoThe treatment of infectious diseases by β-lactam antibiotics is continuously challenged by the emergence and dissemination of new β-lactamases. In most cases, the cephalosporinase activity of class A enzymes results from a few mutations in the TEM and SHV penicillinases. The PER-1 β-lactamase was characterized as a class A enzyme displaying a cephalosporinase activity. This activity was, however, insensitive to the mutations of residues known to be critical for providing extended substrate profiles to TEM and SHV. The x-ray structure of the protein, solved at 1.9-Å resolution, reveals that two of the most conserved features in class A β-lactamases are not present in this enzyme: the fold of the Ω-loop and the cis conformation of the peptide bond between residues 166 and 167. The new fold of the Ω-loop and the insertion of four residues at the edge of strand S3 generate a broad cavity that may easily accommodate the bulky substituents of cephalosporin substrates. The trans conformation of the 166–167 bond is related to the presence of an aspartic acid at position 136. Selection of class A enzymes based on the occurrence of both Asp136 and Asn179 identifies a subgroup of enzymes with high sequence homology. The treatment of infectious diseases by β-lactam antibiotics is continuously challenged by the emergence and dissemination of new β-lactamases. In most cases, the cephalosporinase activity of class A enzymes results from a few mutations in the TEM and SHV penicillinases. The PER-1 β-lactamase was characterized as a class A enzyme displaying a cephalosporinase activity. This activity was, however, insensitive to the mutations of residues known to be critical for providing extended substrate profiles to TEM and SHV. The x-ray structure of the protein, solved at 1.9-Å resolution, reveals that two of the most conserved features in class A β-lactamases are not present in this enzyme: the fold of the Ω-loop and the cis conformation of the peptide bond between residues 166 and 167. The new fold of the Ω-loop and the insertion of four residues at the edge of strand S3 generate a broad cavity that may easily accommodate the bulky substituents of cephalosporin substrates. The trans conformation of the 166–167 bond is related to the presence of an aspartic acid at position 136. Selection of class A enzymes based on the occurrence of both Asp136 and Asn179 identifies a subgroup of enzymes with high sequence homology. Pseudomonas aeruginosa, a prevalent microorganism responsible for several infections in humans, is intrinsically resistant to many antibiotics (1Carmeli Y. Troillet N. Eliopoulos G.M. Samore M.H. Antimicrob. Agents Chemother. 1999; 43: 1379-1382Crossref PubMed Google Scholar). This resistance simultaneously arises from the low permeability of its outer membrane (2Yoshimura F. Nikaido H. J. Bacteriol. 1982; 152: 636-642Crossref PubMed Google Scholar), the existence of an active efflux system (3Li X.-Z. Ma D. Livermore D.M. Nikaido H. Antimicrob. Agents Chemother. 1994; 38: 1742-1752Crossref PubMed Scopus (245) Google Scholar, 4Li X.-Z. Livermore D.M. Nikaido H. Antimicrob. Agents Chemother. 1994; 38: 1732-1741Crossref PubMed Scopus (302) Google Scholar), and the production of a chromosomally encoded inducible class C β-lactamase (5Sabath L.D. Jago M. Abraham E.P. Biochem. J. 1965; 96: 739-752Crossref PubMed Scopus (133) Google Scholar). This bacterium reinforces its resistance to β-lactam antibiotics by expressing a chromosomally encoded class A cephalosporinase, PER-1 (for P seudomonas extendedresistance).PER-1 was first detected in 1993 in a strain isolated from a Turkish patient (6Nordmann P. Ronco E. Naas T. Duport C. Michel-Briand Y. Labia R. Antimicrob. Agents Chemother. 1993; 37: 962-969Crossref PubMed Scopus (189) Google Scholar). It was subsequently identified in nosocomial strains ofSalmonella typhimurium and Acinetobacter baumaniiin Turkey (7Vahaboglu H. Dodanli S. Eroglu C. Öztürk R. Soyletir G. Yildirim I. Avkan V. J. Clin. Microbiol. 1996; 34: 2942-2946Crossref PubMed Google Scholar, 8Vahaboglu H. Öztürk R. Aygün G. Coskunkan F. Yaman A. Kaygusuz A. Leblebicioglu H. Balik I. Aydin K. Otkun M. Antimicrob. Agents Chemother. 1997; 41: 2265-2269Crossref PubMed Google Scholar) and more recently in France (9Poirel L. Karim A. Mercat A. Le Thomas I. Vahaboglu H. Richard C. Nordmann P. J. Antimicrob. Chemother. 1999; 43: 157-158Crossref PubMed Scopus (65) Google Scholar). PER-2, a closely related enzyme sharing 86% homology, has been found in South America (10Bauernfeind A. Stemplinger I. Jungwirth R. Mangold P. Amann S. Akalin E. Ang Ö. Bal C. Casellas J.M. Antimicrob. Agents Chemother. 1996; 40: 616-620Crossref PubMed Google Scholar). These enzymes efficiently hydrolyze penicillins and cephalosporins but not cephamycins or carbapenems, and they are susceptible to clavulanic acid inhibition (6Nordmann P. Ronco E. Naas T. Duport C. Michel-Briand Y. Labia R. Antimicrob. Agents Chemother. 1993; 37: 962-969Crossref PubMed Scopus (189) Google Scholar, 10Bauernfeind A. Stemplinger I. Jungwirth R. Mangold P. Amann S. Akalin E. Ang Ö. Bal C. Casellas J.M. Antimicrob. Agents Chemother. 1996; 40: 616-620Crossref PubMed Google Scholar). PER enzymes belong to the class A β-lactamases (10Bauernfeind A. Stemplinger I. Jungwirth R. Mangold P. Amann S. Akalin E. Ang Ö. Bal C. Casellas J.M. Antimicrob. Agents Chemother. 1996; 40: 616-620Crossref PubMed Google Scholar, 11Nordmann P. Naas T. Antimicrob. Agents Chemother. 1994; 38: 104-114Crossref PubMed Scopus (119) Google Scholar, 12Ambler R.P. Coulson A.F.W. Frère J.-M. Ghuysen J.-M. Joris B. Forsman M. Levesque R.C. Tiraby G. Waley S.G. Biochem. J. 1991; 276: 269-272Crossref PubMed Scopus (838) Google Scholar), but the sequence identity with the TEM and SHV enzymes is only 27% (11Nordmann P. Naas T. Antimicrob. Agents Chemother. 1994; 38: 104-114Crossref PubMed Scopus (119) Google Scholar). It was proposed that PER enzymes and various β-lactamases from Bacteroides sp. (13Rogers M.B. Parker A.C. Smith C.J. Antimicrob. Agents Chemother. 1993; 37: 2391-2400Crossref PubMed Scopus (69) Google Scholar, 14Parker A.C. Smith C.J. Antimicrob. Agents Chemother. 1993; 37: 1028-1036Crossref PubMed Scopus (103) Google Scholar, 15Smith C.J. Bennet T.K. Parker A.C. Antimicrob. Agents Chemother. 1994; 38: 1711-1715Crossref PubMed Scopus (35) Google Scholar) may constitute a subgroup of the class A enzymes (11Nordmann P. Naas T. Antimicrob. Agents Chemother. 1994; 38: 104-114Crossref PubMed Scopus (119) Google Scholar).Class A β-lactamases hydrolyze β-lactam antibiotics through a double displacement mechanism with a transient acylation of the catalytic Ser70 residue. Eight x-ray structures of apo-enzymes, TEM-1 and TOHO-1 from Escherichia coli (16Jelsch C. Mourey L. Masson J.-M. Samama J.-P. Proteins. 1993; 16: 364-383Crossref PubMed Scopus (359) Google Scholar,17Ibuka A. Taguchi A. Ishiguro M. Fushinobu S. Ishii Y. Kamitori S. Okuyama K. Yamaguchi K. Konno M. Matsuzawa H. J. Mol. Biol. 1999; 285: 2079-2087Crossref PubMed Scopus (85) Google Scholar), PC1 from Staphylococcus aureus (18Herzberg O. J. Mol. Biol. 1991; 217: 701-719Crossref PubMed Scopus (205) Google Scholar), SHV fromKlebsiella pneumoniae (19Kuzin A.P. Nukaga M. Nukaga Y. Hujer A.M. Bonomo R.A. Knox J.A. Biochemistry. 1999; 38: 5720-5727Crossref PubMed Scopus (108) Google Scholar), NMC-A from Enterobacter cloacae (20Swarén P. Maveyraud L. Raquet X. Cabantous S. Duez C. Pédelacq J.-D. Mariotte-Boyer S. Mourey L. Labia R. Nicolas-Chanoine M.-H. Nordmann P. Frère J.-M. Samama J.-P. J. Biol. Chem. 1998; 273: 26714-26738Abstract Full Text Full Text PDF PubMed Scopus (80) Google Scholar), MFO from Mycobacterium fortuitum (PDB entry 1MFO), BLIC from Bacillus licheniformis (21Moews P.C. Knox J.R. Dideberg O. Charlier P. Frère J.-M. Proteins. 1990; 7: 156-171Crossref PubMed Scopus (158) Google Scholar), and SAG from Streptomyces albus G (22Dideberg O. Charlier P. Wéry J.-P. Dehottay P. Dusart J. Erpicum T. Frère J.-M. Ghuysen J.-M. Biochem. J. 1987; 245: 911-913Crossref PubMed Scopus (114) Google Scholar) have been solved. These enzymes display a very similar fold, and detailed comparisons of the structures were helpful in relating some significant differences in substrate profile to local structural features (23Mourey L. Miyashita K. Swarén P. Bulychev A. Samama J.-P. Mobashery S. J. Am. Chem. Soc. 1998; 120: 9382-9383Crossref Scopus (50) Google Scholar, 24Knox J.R. Moews P.C. Frère J.-M. Chem. Biol. 1996; 3: 937-947Abstract Full Text PDF PubMed Scopus (143) Google Scholar). The structural impact of point mutations leading to some 100 extended spectrum enzymes in the TEM and SHV families is less documented than the possible function of the invariant residues of the catalytic machinery (25Swarén P. Maveyraud L. Guillet V. Masson J.-M. Mourey L. Samama J.-P. Structure. 1995; 3: 603-613Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar, 26Maveyraud L. Pratt R.F. Samama J.-P. Biochemistry. 1998; 37: 2622-2628Crossref PubMed Scopus (83) Google Scholar, 27Vanwetswinkel S. Avalle B. Fastrez J. J. Mol. Biol. 2000; 295: 527-540Crossref PubMed Scopus (26) Google Scholar, 28Knox J.R. Antimicrob. Agents. Chemother. 1995; 39: 2593-2601Crossref PubMed Scopus (298) Google Scholar). Several studies supported the hypothesis that the cephalosporinase activity of these extended substrate enzymes was related to an increased flexibility of the Ω-loop region and to alterations of the S3 strand, two of the regions lining the active site (24Knox J.R. Moews P.C. Frère J.-M. Chem. Biol. 1996; 3: 937-947Abstract Full Text PDF PubMed Scopus (143) Google Scholar, 29Banerjee S. Pieter U. Kapadia G. Pannell L.K. Herzberg O. Biochemistry. 1998; 37: 3286-3296Crossref PubMed Scopus (84) Google Scholar, 30Maveyraud L. Saves I. Burlet-Schiltz O. Swarén P. Masson J.-M. Delaire M. Mourey L. Promé J.-C. Samama J.-P. J. Biol. Chem. 1996; 271: 10482-10489Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar, 31Chen C.C.H. Herzberg O. Protein Eng. 1999; 12: 573-579Crossref PubMed Scopus (13) Google Scholar, 32Saves I. Burlet-Schiltz O. Maveyraud L. Samama J.-P. Promé J.-C. Masson J.-M. Biochemistry. 1995; 34: 11660-11667Crossref PubMed Scopus (51) Google Scholar, 33Raquet X. Lamotte-Brasseur J. Fonzé E. Goussard S. Courvalin P. Frère J.-M. J. Mol. Biol. 1994; 244: 625-639Crossref PubMed Scopus (124) Google Scholar).Surprisingly, site-directed mutagenesis on the PER-1 enzyme showed that none of the residues responsible for the cephalosporinase activity in the TEM and SHV families (104, 164, 179, 238, and 240) were implicated in the substrate profile of this enzyme (34Bouthors A.-T. Dagoneau-Blanchard N. Naas T. Nordmann P. Jarlier V. Sougakoff W. Biochem. J. 1998; 330: 1443-1449Crossref PubMed Scopus (33) Google Scholar, 35Bouthors A.-T. Delettré J. Mugnier P. Jarlier V. Sougakoff W. Protein Eng. 1999; 12: 313-318Crossref PubMed Scopus (25) Google Scholar). The 1.9-Å x-ray structure of PER-1 presented in this report reveals a completely new fold of the Ω-loop region, so far one of the most conserved features in the class A enzymes. Structure and sequence analysis suggests that PER-1 defines a group of class A β-lactamases that can be recognized by the presence of an aspartic acid residue at position 136. The insertion of four amino acids in the 238–242 region, together with the Ω-loop fold, led to a significant increase in size of the substrate binding pocket.DISCUSSIONAlthough PER-1 is a typical class A enzyme with respect to the conservation of the catalytic machinery, its three-dimensional structure revealed that this superfamily is not as homogeneous in structure as it was thought. It was unexpected to find a new fold of the Ω-loop, a region considered to be a canonical motif in these enzymes. This finding suggested to classify β-lactamase sequences according to the occurence of an aspartic acid at position 136. Several enzymes with significant sequence identity (40%) were identified that form a subgroup in the class A superfamily (Figs. 4 and6). These “PER-like” enzymes (11Nordmann P. Naas T. Antimicrob. Agents Chemother. 1994; 38: 104-114Crossref PubMed Scopus (119) Google Scholar) comprise PER-2 from S. typhimurium (10Bauernfeind A. Stemplinger I. Jungwirth R. Mangold P. Amann S. Akalin E. Ang Ö. Bal C. Casellas J.M. Antimicrob. Agents Chemother. 1996; 40: 616-620Crossref PubMed Google Scholar), VEB-1 from E. coli (57Poirel L. Naas T. Guibert M. Chaibi E.B. Labia R. Nordmann P. Antimicrob. Agents Chemother. 1999; 43: 573-581Crossref PubMed Google Scholar), CME-1 from Chryseobacterium(flavobacterium) meningosepticum (58Rossolini G.M. Franceschini N. Lauretti L. Caravelli B. Riccio M.L. Galleni M. Frere J.-M. Amicosante G. Antimicrob. Agents Chemother. 1999; 43: 2193-2199Crossref PubMed Google Scholar), TLA-1 fromE. coli (59Silva J. Aguilar C. Ayala G. Estrada M.A. Garza-Ramos U. Lara-Lemus R. Ledezma L. Antimicrob. Agents Chemother. 2000; 44: 997-1003Crossref PubMed Scopus (70) Google Scholar), CBLA from Bacteroides uniformis(15Smith C.J. Bennet T.K. Parker A.C. Antimicrob. Agents Chemother. 1994; 38: 1711-1715Crossref PubMed Scopus (35) Google Scholar), CEPA from Bacteroides fragilis (13Rogers M.B. Parker A.C. Smith C.J. Antimicrob. Agents Chemother. 1993; 37: 2391-2400Crossref PubMed Scopus (69) Google Scholar), and CFXA fromBacteroides vulgatus (14Parker A.C. Smith C.J. Antimicrob. Agents Chemother. 1993; 37: 1028-1036Crossref PubMed Scopus (103) Google Scholar). The phylogenetic tree shown in Fig. 6 indicates that these proteins constitute a distinct cluster of enzymes, and the x-ray structure of PER-1 suggests that this partition reflects major structural differences compared with the TEM-like β-lactamases. On the contrary, there is no major partition between the penicillinases and the NMC-A, IMI-1 (60Rasmussen B.A. Bush K. Keeney D. Yang Y. Hare R. O'Gara C. Medeiros A.A. Antimicrob. Agents Chemother. 1996; 40: 2080-2086Crossref PubMed Google Scholar), and Sme-1 (61Naas T. Vedel L. Sougakoff W. Livermore D.M. Nordmann P. Antimicrob. Agents Chemother. 1994; 38: 1262-1270Crossref PubMed Scopus (180) Google Scholar) carbapenemases, which may be identified by the presence of cysteine residues at positions 69 and 238 (Fig. 4). It agrees with the recent structure determination of NMC-A (20Swarén P. Maveyraud L. Raquet X. Cabantous S. Duez C. Pédelacq J.-D. Mariotte-Boyer S. Mourey L. Labia R. Nicolas-Chanoine M.-H. Nordmann P. Frère J.-M. Samama J.-P. J. Biol. Chem. 1998; 273: 26714-26738Abstract Full Text Full Text PDF PubMed Scopus (80) Google Scholar), which revealed that the carbapenemase activity was associated with subtle structural modifications. For reasons discussed below, we suggest that the presence of both Asp136 and Asn179 should identify any new β-lactamase as member of the PER subgroup.Previous studies on class A β-lactamases, including x-ray structure determinations, led to the conclusion that the cisconformation of the 166–167 peptide bond was mandatory for the proper location of Glu166 in the active site, and therefore for catalysis. The N136A mutant of the PC1 enzyme was found to accumulate acyl-enzyme adducts but had no hydrolytic activity against penicillin and cefotaxime (29Banerjee S. Pieter U. Kapadia G. Pannell L.K. Herzberg O. Biochemistry. 1998; 37: 3286-3296Crossref PubMed Scopus (84) Google Scholar). Recent kinetic analysis of randomly generated Ω-loop TEM mutants concluded that the isomerization of the 166–167 peptide bond controlled one conformation of this loop that was compatible with fast acylation of the protein (27Vanwetswinkel S. Avalle B. Fastrez J. J. Mol. Biol. 2000; 295: 527-540Crossref PubMed Scopus (26) Google Scholar). The PER-1 structure presented here reveals that a functional enzyme can accommodate thetrans conformation of the 166–167 peptide bond through interactions with Asp136 and maintains the position of Glu166 that preserves the molecular basis of the catalytic mechanism.The trans conformation goes with a different sequence of the 161–181 (Ω-loop) region, which displays a new fold compared with the typical class A enzymes. In those cases, the Ω-loop is stabilized by salt bridge interactions, and the R164S mutation is frequently observed in naturally occurring TEM and SHV mutants with extended spectrum (62Jacoby G.A. Medeiros A.A. Antimicrob. Agents Chemother. 1991; 35: 1697-1704Crossref PubMed Scopus (649) Google Scholar). The disruption of the Arg164–Asp179 salt bridge is assumed to increase the flexibility of the Ω-loop, thereby favoring the binding of third generation cephalosporins (63Vakulenko S.B. Toth M. Taibi P. Mobashery S. Lerner S.A. Antimicrob. Agents Chemother. 1995; 39: 1878-1880Crossref PubMed Scopus (34) Google Scholar, 64Vakulenko S.B. Taibi P. Toth M. Massova I. Lerner S.A. Mobashery S. J. Biol. Chem. 1999; 274: 23052-23060Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar). In PC1, the β-lactamase from S. aureus, disruption of this single salt bridge by the D179N mutation led to a disordered conformation of the Ω-loop (29Banerjee S. Pieter U. Kapadia G. Pannell L.K. Herzberg O. Biochemistry. 1998; 37: 3286-3296Crossref PubMed Scopus (84) Google Scholar,65Herzberg O. Kapadia G. Blanco B. Smith T.S. Coulson A. Biochemistry. 1991; 30: 9503-9509Crossref PubMed Scopus (67) Google Scholar). There are no salt bridges in PER-1 where the stability of the Ω-loop fold stems from the presence of secondary structure elements complemented by the hydrogen bond interactions of the side chain of Asn179 to the main chain nitrogen atoms of Val163 and Ala164 and to the main chain oxygen atom of Ala164 (Fig. 5).In retrospect and in view of the major structural differences between the PER-1 and the TEM enzymes, it is not surprising that site-directed mutagenesis of PER-1 (34Bouthors A.-T. Dagoneau-Blanchard N. Naas T. Nordmann P. Jarlier V. Sougakoff W. Biochem. J. 1998; 330: 1443-1449Crossref PubMed Scopus (33) Google Scholar, 35Bouthors A.-T. Delettré J. Mugnier P. Jarlier V. Sougakoff W. Protein Eng. 1999; 12: 313-318Crossref PubMed Scopus (25) Google Scholar) could not relate the cephalosporinase activity of PER-1 to the residue type at positions 164, 179, 238, and 240. These residues, which provide extended spectrum profiles in the TEM and SHV enzymes, have different spatial location and environments in these three-dimensional structures. Protein engineering in PER-1 nevertheless identified two residues whose mutations had significant kinetic effects. Replacement of Thr104 with Glu completely abolished the catalytic activity on penicillins and reduced thek cat values for cephalosporins by a factor of 50–700 (34Bouthors A.-T. Dagoneau-Blanchard N. Naas T. Nordmann P. Jarlier V. Sougakoff W. Biochem. J. 1998; 330: 1443-1449Crossref PubMed Scopus (33) Google Scholar). According to the x-ray structure, this mutation should in any case disrupt the interaction of the γ-hydroxyl group of Thr104 with Asn132. These observations shed light on the previously suggested importance for catalysis in class A enzymes of the hydrogen bond interaction between residue 104 and Asn132 (50Petit A. Maveyraud L. Lenfant F. Samama J.-P. Labia R. Masson J.-M. Biochem. J. 1995; 305: 33-40Crossref PubMed Scopus (50) Google Scholar, 66Matagne A. Frère J.-M. Biochim. Biophys. Acta. 1995; 1246: 109-127Crossref PubMed Scopus (104) Google Scholar). It involves the main chain oxygen atom of residue 104 in all structures except PER-1, and its contribution could evidently not be demonstrated by protein engineering in these enzymes. Replacement of Thr237 by an alanine residue increased thek cat/K m of the mutant protein on cephalosporin substrates by 10–100-fold (35Bouthors A.-T. Delettré J. Mugnier P. Jarlier V. Sougakoff W. Protein Eng. 1999; 12: 313-318Crossref PubMed Scopus (25) Google Scholar). Interestingly, it is the reverse mutation, A237T, that improved hydrolysis on cephalosporin substrates in the TEM enzyme (67Bush K. Jacoby G.A. J. Antimicrob. Chemother. 1997; 39: 1-3Crossref PubMed Scopus (82) Google Scholar).The relocation of the side chain of residue 104, the fold of the Ω-loop, and the conformation of the 238–242 region in PER-1 seem in direct relationship with the cephalosporinase activity of this enzyme. Indeed, these folding features generate a cavity, filled with several water molecules in the apoenzyme, that is precisely located in the area expected to bind the bulky side chain of third and fourth generation cephalosporins (33Raquet X. Lamotte-Brasseur J. Fonzé E. Goussard S. Courvalin P. Frère J.-M. J. Mol. Biol. 1994; 244: 625-639Crossref PubMed Scopus (124) Google Scholar) (Fig. 7). According to the superposed TEM and PER-1 structures (Fig.8), this observation supports the proposal that the displacement of the Ω-loop in the TEM enzyme should favor the binding of cephalosporin substrates.Figure 7Stereo view of the solvent accessible surface area in a modeled PER-1-cefotaxime acyl-enzyme complex. This view illustrates the large cavity where the substituents of third generation cephalosporins bind to the protein. The surface was computed using MSMS (M. F. Sanner, The Scripps Research Institute) and displayed with DINO (A. Philippsen, University of Basel, Switzerland).View Large Image Figure ViewerDownload (PPT)Figure 8The superposition of the active sites in PER-1 (black) and TEM (red) shows the significant enlargement of the substrate binding cavity in PER-1.View Large Image Figure ViewerDownload (PPT)The structure determinations of the PER-1 β-lactamase breaks the assumption of a unique fold of the class A enzymes. This finding promoted the suggestion that class A β-lactamases may be categorized into three subgroups according to their kinetic properties as penicillinases (TEM/SHV group), cephalosporinases (PER group), and carbapenemases (NMC-A group). These groups have sequence signatures that should help in assigning any new enzymes into either of them. From the microbiological point of view, the simultaneous occurrence in bacteria of extended spectrum TEM or SHV and PER enzymes should complicate identification of the resistance enzymes based on the antibiogram patterns. Pseudomonas aeruginosa, a prevalent microorganism responsible for several infections in humans, is intrinsically resistant to many antibiotics (1Carmeli Y. Troillet N. Eliopoulos G.M. Samore M.H. Antimicrob. Agents Chemother. 1999; 43: 1379-1382Crossref PubMed Google Scholar). This resistance simultaneously arises from the low permeability of its outer membrane (2Yoshimura F. Nikaido H. J. Bacteriol. 1982; 152: 636-642Crossref PubMed Google Scholar), the existence of an active efflux system (3Li X.-Z. Ma D. Livermore D.M. Nikaido H. Antimicrob. Agents Chemother. 1994; 38: 1742-1752Crossref PubMed Scopus (245) Google Scholar, 4Li X.-Z. Livermore D.M. Nikaido H. Antimicrob. Agents Chemother. 1994; 38: 1732-1741Crossref PubMed Scopus (302) Google Scholar), and the production of a chromosomally encoded inducible class C β-lactamase (5Sabath L.D. Jago M. Abraham E.P. Biochem. J. 1965; 96: 739-752Crossref PubMed Scopus (133) Google Scholar). This bacterium reinforces its resistance to β-lactam antibiotics by expressing a chromosomally encoded class A cephalosporinase, PER-1 (for P seudomonas extendedresistance). PER-1 was first detected in 1993 in a strain isolated from a Turkish patient (6Nordmann P. Ronco E. Naas T. Duport C. Michel-Briand Y. Labia R. Antimicrob. Agents Chemother. 1993; 37: 962-969Crossref PubMed Scopus (189) Google Scholar). It was subsequently identified in nosocomial strains ofSalmonella typhimurium and Acinetobacter baumaniiin Turkey (7Vahaboglu H. Dodanli S. Eroglu C. Öztürk R. Soyletir G. Yildirim I. Avkan V. J. Clin. Microbiol. 1996; 34: 2942-2946Crossref PubMed Google Scholar, 8Vahaboglu H. Öztürk R. Aygün G. Coskunkan F. Yaman A. Kaygusuz A. Leblebicioglu H. Balik I. Aydin K. Otkun M. Antimicrob. Agents Chemother. 1997; 41: 2265-2269Crossref PubMed Google Scholar) and more recently in France (9Poirel L. Karim A. Mercat A. Le Thomas I. Vahaboglu H. Richard C. Nordmann P. J. Antimicrob. Chemother. 1999; 43: 157-158Crossref PubMed Scopus (65) Google Scholar). PER-2, a closely related enzyme sharing 86% homology, has been found in South America (10Bauernfeind A. Stemplinger I. Jungwirth R. Mangold P. Amann S. Akalin E. Ang Ö. Bal C. Casellas J.M. Antimicrob. Agents Chemother. 1996; 40: 616-620Crossref PubMed Google Scholar). These enzymes efficiently hydrolyze penicillins and cephalosporins but not cephamycins or carbapenems, and they are susceptible to clavulanic acid inhibition (6Nordmann P. Ronco E. Naas T. Duport C. Michel-Briand Y. Labia R. Antimicrob. Agents Chemother. 1993; 37: 962-969Crossref PubMed Scopus (189) Google Scholar, 10Bauernfeind A. Stemplinger I. Jungwirth R. Mangold P. Amann S. Akalin E. Ang Ö. Bal C. Casellas J.M. Antimicrob. Agents Chemother. 1996; 40: 616-620Crossref PubMed Google Scholar). PER enzymes belong to the class A β-lactamases (10Bauernfeind A. Stemplinger I. Jungwirth R. Mangold P. Amann S. Akalin E. Ang Ö. Bal C. Casellas J.M. Antimicrob. Agents Chemother. 1996; 40: 616-620Crossref PubMed Google Scholar, 11Nordmann P. Naas T. Antimicrob. Agents Chemother. 1994; 38: 104-114Crossref PubMed Scopus (119) Google Scholar, 12Ambler R.P. Coulson A.F.W. Frère J.-M. Ghuysen J.-M. Joris B. Forsman M. Levesque R.C. Tiraby G. Waley S.G. Biochem. J. 1991; 276: 269-272Crossref PubMed Scopus (838) Google Scholar), but the sequence identity with the TEM and SHV enzymes is only 27% (11Nordmann P. Naas T. Antimicrob. Agents Chemother. 1994; 38: 104-114Crossref PubMed Scopus (119) Google Scholar). It was proposed that PER enzymes and various β-lactamases from Bacteroides sp. (13Rogers M.B. Parker A.C. Smith C.J. Antimicrob. Agents Chemother. 1993; 37: 2391-2400Crossref PubMed Scopus (69) Google Scholar, 14Parker A.C. Smith C.J. Antimicrob. Agents Chemother. 1993; 37: 1028-1036Crossref PubMed Scopus (103) Google Scholar, 15Smith C.J. Bennet T.K. Parker A.C. Antimicrob. Agents Chemother. 1994; 38: 1711-1715Crossref PubMed Scopus (35) Google Scholar) may constitute a subgroup of the class A enzymes (11Nordmann P. Naas T. Antimicrob. Agents Chemother. 1994; 38: 104-114Crossref PubMed Scopus (119) Google Scholar). Class A β-lactamases hydrolyze β-lactam antibiotics through a double displacement mechanism with a transient acylation of the catalytic Ser70 residue. Eight x-ray structures of apo-enzymes, TEM-1 and TOHO-1 from Escherichia coli (16Jelsch C. Mourey L. Masson J.-M. Samama J.-P. Proteins. 1993; 16: 364-383Crossref PubMed Scopus (359) Google Scholar,17Ibuka A. Taguchi A. Ishiguro M. Fushinobu S. Ishii Y. Kamitori S. Okuyama K. Yamaguchi K. Konno M. Matsuzawa H. J. Mol. Biol. 1999; 285: 2079-2087Crossref PubMed Scopus (85) Google Scholar), PC1 from Staphylococcus aureus (18Herzberg O. J. Mol. Biol. 1991; 217: 701-719Crossref PubMed Scopus (205) Google Scholar), SHV fromKlebsiella pneumoniae (19Kuzin A.P. Nukaga M. Nukaga Y. Hujer A.M. Bonomo R.A. Knox J.A. Biochemistry. 1999; 38: 5720-5727Crossref PubMed Scopus (108) Google Scholar), NMC-A from Enterobacter cloacae (20Swarén P. Maveyraud L. Raquet X. Cabantous S. Duez C. Pédelacq J.-D. Mariotte-Boyer S. Mourey L. Labia R. Nicolas-Chanoine M.-H. Nordmann P. Frère J.-M. Samama J.-P. J. Biol. Chem. 1998; 273: 26714-26738Abstract Full Text Full Text PDF PubMed Scopus (80) Google Scholar), MFO from Mycobacterium fortuitum (PDB entry 1MFO), BLIC from Bacillus licheniformis (21Moews P.C. Knox J.R. Dideberg O. Charlier P. Frère J.-M. Proteins. 1990; 7: 156-171Crossref PubMed Scopus (158) Google Scholar), and SAG from Streptomyces albus G (22Dideberg O. Charlier P. Wéry J.-P. Dehottay P. Dusart J. Erpicum T. Frère J.-M. Ghuysen J.-M. Biochem. J. 1987; 245: 911-913Crossref PubMed Scopus (114) Google Scholar) have been solved. These enzymes display a very similar fold, and detailed comparisons of the structures were helpful in relating some significant differences in substrate profile to local structural features (23Mourey L. Miyashita K. Swarén P. Bulychev A. Samama J.-P. Mobashery S. J. Am. Chem. Soc. 1998; 120: 9382-9383Crossref Scopus (50) Google Scholar, 24Knox J.R. Moews P.C. Frère J.-M. Chem. Biol. 1996; 3: 937-947Abstract Full Text PDF PubMed Scopus (143) Google Scholar). The structural impact of point mutations leading to some 100 extended spectrum enzymes in the TEM and SHV families is less documented than the possible function of the invariant residues of the catalytic machinery (25Swarén P. Maveyraud L. Guillet V. Masson J.-M. Mourey L. Samama J.-P. Structure. 1995; 3: 603-613Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar, 26Maveyraud L. Pratt R.F. Samama J.-P. Biochemistry. 1998; 37: 2622-2628Crossref PubMed Scopus (83) Google Scholar, 27Vanwetswinkel S. Avalle B. Fastrez J. J. Mol. Biol. 2000; 295: 527-540Crossref PubMed Scopus (26) Google Scholar, 28Knox J.R. Antimicrob. Agents. Chemother. 1995; 39: 2593-2601Crossref PubMed Scopus (298) Google Scholar). Several studies supported the hypothesis that the cephalosporinase activity of these extended substrate enzymes was related to an increased flexibility of the Ω-loop region and to alterations of the S3 strand, two of the regions lining the active site (24Knox J.R. Moews P.C. Frère J.-M. Chem. Biol. 1996; 3: 937-947Abstract Full Text PDF PubMed Scopus (143) Google Scholar, 29Banerjee S. Pieter U. Kapadia G. Pannell L.K. Herzberg O. Biochemistry. 1998; 37: 3286-3296Crossref PubMed Scopus (84) Google Scholar, 30Maveyraud L. Saves I. Burlet-Schiltz O. Swarén P. Masson J.-M. Delaire M. Mourey L. Promé J.-C. Samama J.-P. J. Biol. Chem. 1996; 271: 10482-10489Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar, 31Chen C.C.H. Herzberg O. Protein Eng. 1999; 12: 573-579Crossref PubMed Scopus (13) Google Scholar, 32Saves I. Burlet-Schiltz O. Maveyraud L. Samama J.-P. Promé J.-C. Masson J.-M. Biochemistry. 1995; 34: 11660-11667Crossref PubMed Scopus (51) Google Scholar, 33Raquet X. Lamotte-Brasseur J. Fonzé E. Goussard S. Courvalin P. Frère J.-M. J. Mol. Biol. 1994; 244: 625-639Crossref PubMed Scopus (124) Google Scholar). Surprisingly, site-directed mutagenesis on the PER-1 enzyme showed that none of the residues responsible for the cephalosporinase activity in the TEM and SHV families (104, 164, 179, 238, and 240) were implicated in the substrate profile of this enzyme (34Bouthors A.-T. Dagoneau-Blanchard N. Naas T. Nordmann P. Jarlier V. Sougakoff W. Biochem. J. 1998; 330: 1443-1449Crossref PubMed Scopus (33) Google Scholar, 35Bouthors A.-T. Delettré J. Mugnier P. Jarlier V. Sougakoff W. Protein Eng. 1999; 12: 313-318Crossref PubMed Scopus (25) Google Scholar). The 1.9-Å x-ray structure of PER-1 presented in this report reveals a completely new fold of the Ω-loop region, so far one of the most conserved features in the class A enzymes. Structure and sequence analysis suggests that PER-1 defines a group of class A β-lactamases that can be recognized by the presence of an aspartic acid residue at position 136. The insertion of four amino acids in the 238–242 region, together with the Ω-loop fold, led to a significant increase in size of the substrate binding pocket. DISCUSSIONAlthough PER-1 is a typical class A enzyme with respect to the conservation of the catalytic machinery, its three-dimensional structure revealed that this superfamily is not as homogeneous in structure as it was thought. It was unexpected to find a new fold of the Ω-loop, a region considered to be a canonical motif in these enzymes. This finding suggested to classify β-lactamase sequences according to the occurence of an aspartic acid at position 136. Several enzymes with significant sequence identity (40%) were identified that form a subgroup in the class A superfamily (Figs. 4 and6). These “PER-like” enzymes (11Nordmann P. Naas T. Antimicrob. Agents Chemother. 1994; 38: 104-114Crossref PubMed Scopus (119) Google Scholar) comprise PER-2 from S. typhimurium (10Bauernfeind A. Stemplinger I. Jungwirth R. Mangold P. Amann S. Akalin E. Ang Ö. Bal C. Casellas J.M. Antimicrob. Agents Chemother. 1996; 40: 616-620Crossref PubMed Google Scholar), VEB-1 from E. coli (57Poirel L. Naas T. Guibert M. Chaibi E.B. Labia R. Nordmann P. Antimicrob. Agents Chemother. 1999; 43: 573-581Crossref PubMed Google Scholar), CME-1 from Chryseobacterium(flavobacterium) meningosepticum (58Rossolini G.M. Franceschini N. Lauretti L. Caravelli B. Riccio M.L. Galleni M. Frere J.-M. Amicosante G. Antimicrob. Agents Chemother. 1999; 43: 2193-2199Crossref PubMed Google Scholar), TLA-1 fromE. coli (59Silva J. Aguilar C. Ayala G. Estrada M.A. Garza-Ramos U. Lara-Lemus R. Ledezma L. Antimicrob. Agents Chemother. 2000; 44: 997-1003Crossref PubMed Scopus (70) Google Scholar), CBLA from Bacteroides uniformis(15Smith C.J. Bennet T.K. Parker A.C. Antimicrob. Agents Chemother. 1994; 38: 1711-1715Crossref PubMed Scopus (35) Google Scholar), CEPA from Bacteroides fragilis (13Rogers M.B. Parker A.C. Smith C.J. Antimicrob. Agents Chemother. 1993; 37: 2391-2400Crossref PubMed Scopus (69) Google Scholar), and CFXA fromBacteroides vulgatus (14Parker A.C. Smith C.J. Antimicrob. Agents Chemother. 1993; 37: 1028-1036Crossref PubMed Scopus (103) Google Scholar). The phylogenetic tree shown in Fig. 6 indicates that these proteins constitute a distinct cluster of enzymes, and the x-ray structure of PER-1 suggests that this partition reflects major structural differences compared with the TEM-like β-lactamases. On the contrary, there is no major partition between the penicillinases and the NMC-A, IMI-1 (60Rasmussen B.A. Bush K. Keeney D. Yang Y. Hare R. O'Gara C. Medeiros A.A. Antimicrob. Agents Chemother. 1996; 40: 2080-2086Crossref PubMed Google Scholar), and Sme-1 (61Naas T. Vedel L. Sougakoff W. Livermore D.M. Nordmann P. Antimicrob. Agents Chemother. 1994; 38: 1262-1270Crossref PubMed Scopus (180) Google Scholar) carbapenemases, which may be identified by the presence of cysteine residues at positions 69 and 238 (Fig. 4). It agrees with the recent structure determination of NMC-A (20Swarén P. Maveyraud L. Raquet X. Cabantous S. Duez C. Pédelacq J.-D. Mariotte-Boyer S. Mourey L. Labia R. Nicolas-Chanoine M.-H. Nordmann P. Frère J.-M. Samama J.-P. J. Biol. Chem. 1998; 273: 26714-26738Abstract Full Text Full Text PDF PubMed Scopus (80) Google Scholar), which revealed that the carbapenemase activity was associated with subtle structural modifications. For reasons discussed below, we suggest that the presence of both Asp136 and Asn179 should identify any new β-lactamase as member of the PER subgroup.Previous studies on class A β-lactamases, including x-ray structure determinations, led to the conclusion that the cisconformation of the 166–167 peptide bond was mandatory for the proper location of Glu166 in the active site, and therefore for catalysis. The N136A mutant of the PC1 enzyme was found to accumulate acyl-enzyme adducts but had no hydrolytic activity against penicillin and cefotaxime (29Banerjee S. Pieter U. Kapadia G. Pannell L.K. Herzberg O. Biochemistry. 1998; 37: 3286-3296Crossref PubMed Scopus (84) Google Scholar). Recent kinetic analysis of randomly generated Ω-loop TEM mutants concluded that the isomerization of the 166–167 peptide bond controlled one conformation of this loop that was compatible with fast acylation of the protein (27Vanwetswinkel S. Avalle B. Fastrez J. J. Mol. Biol. 2000; 295: 527-540Crossref PubMed Scopus (26) Google Scholar). The PER-1 structure presented here reveals that a functional enzyme can accommodate thetrans conformation of the 166–167 peptide bond through interactions with Asp136 and maintains the position of Glu166 that preserves the molecular basis of the catalytic mechanism.The trans conformation goes with a different sequence of the 161–181 (Ω-loop) region, which displays a new fold compared with the typical class A enzymes. In those cases, the Ω-loop is stabilized by salt bridge interactions, and the R164S mutation is frequently observed in naturally occurring TEM and SHV mutants with extended spectrum (62Jacoby G.A. Medeiros A.A. Antimicrob. Agents Chemother. 1991; 35: 1697-1704Crossref PubMed Scopus (649) Google Scholar). The disruption of the Arg164–Asp179 salt bridge is assumed to increase the flexibility of the Ω-loop, thereby favoring the binding of third generation cephalosporins (63Vakulenko S.B. Toth M. Taibi P. Mobashery S. Lerner S.A. Antimicrob. Agents Chemother. 1995; 39: 1878-1880Crossref PubMed Scopus (34) Google Scholar, 64Vakulenko S.B. Taibi P. Toth M. Massova I. Lerner S.A. Mobashery S. J. Biol. Chem. 1999; 274: 23052-23060Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar). In PC1, the β-lactamase from S. aureus, disruption of this single salt bridge by the D179N mutation led to a disordered conformation of the Ω-loop (29Banerjee S. Pieter U. Kapadia G. Pannell L.K. Herzberg O. Biochemistry. 1998; 37: 3286-3296Crossref PubMed Scopus (84) Google Scholar,65Herzberg O. Kapadia G. Blanco B. Smith T.S. Coulson A. Biochemistry. 1991; 30: 9503-9509Crossref PubMed Scopus (67) Google Scholar). There are no salt bridges in PER-1 where the stability of the Ω-loop fold stems from the presence of secondary structure elements complemented by the hydrogen bond interactions of the side chain of Asn179 to the main chain nitrogen atoms of Val163 and Ala164 and to the main chain oxygen atom of Ala164 (Fig. 5).In retrospect and in view of the major structural differences between the PER-1 and the TEM enzymes, it is not surprising that site-directed mutagenesis of PER-1 (34Bouthors A.-T. Dagoneau-Blanchard N. Naas T. Nordmann P. Jarlier V. Sougakoff W. Biochem. J. 1998; 330: 1443-1449Crossref PubMed Scopus (33) Google Scholar, 35Bouthors A.-T. Delettré J. Mugnier P. Jarlier V. Sougakoff W. Protein Eng. 1999; 12: 313-318Crossref PubMed Scopus (25) Google Scholar) could not relate the cephalosporinase activity of PER-1 to the residue type at positions 164, 179, 238, and 240. These residues, which provide extended spectrum profiles in the TEM and SHV enzymes, have different spatial location and environments in these three-dimensional structures. Protein engineering in PER-1 nevertheless identified two residues whose mutations had significant kinetic effects. Replacement of Thr104 with Glu completely abolished the catalytic activity on penicillins and reduced thek cat values for cephalosporins by a factor of 50–700 (34Bouthors A.-T. Dagoneau-Blanchard N. Naas T. Nordmann P. Jarlier V. Sougakoff W. Biochem. J. 1998; 330: 1443-1449Crossref PubMed Scopus (33) Google Scholar). According to the x-ray structure, this mutation should in any case disrupt the interaction of the γ-hydroxyl group of Thr104 with Asn132. These observations shed light on the previously suggested importance for catalysis in class A enzymes of the hydrogen bond interaction between residue 104 and Asn132 (50Petit A. Maveyraud L. Lenfant F. Samama J.-P. Labia R. Masson J.-M. Biochem. J. 1995; 305: 33-40Crossref PubMed Scopus (50) Google Scholar, 66Matagne A. Frère J.-M. Biochim. Biophys. Acta. 1995; 1246: 109-127Crossref PubMed Scopus (104) Google Scholar). It involves the main chain oxygen atom of residue 104 in all structures except PER-1, and its contribution could evidently not be demonstrated by protein engineering in these enzymes. Replacement of Thr237 by an alanine residue increased thek cat/K m of the mutant protein on cephalosporin substrates by 10–100-fold (35Bouthors A.-T. Delettré J. Mugnier P. Jarlier V. Sougakoff W. Protein Eng. 1999; 12: 313-318Crossref PubMed Scopus (25) Google Scholar). Interestingly, it is the reverse mutation, A237T, that improved hydrolysis on cephalosporin substrates in the TEM enzyme (67Bush K. Jacoby G.A. J. Antimicrob. Chemother. 1997; 39: 1-3Crossref PubMed Scopus (82) Google Scholar).The relocation of the side chain of residue 104, the fold of the Ω-loop, and the conformation of the 238–242 region in PER-1 seem in direct relationship with the cephalosporinase activity of this enzyme. Indeed, these folding features generate a cavity, filled with several water molecules in the apoenzyme, that is precisely located in the area expected to bind the bulky side chain of third and fourth generation cephalosporins (33Raquet X. Lamotte-Brasseur J. Fonzé E. Goussard S. Courvalin P. Frère J.-M. J. Mol. Biol. 1994; 244: 625-639Crossref PubMed Scopus (124) Google Scholar) (Fig. 7). According to the superposed TEM and PER-1 structures (Fig.8), this observation supports the proposal that the displacement of the Ω-loop in the TEM enzyme should favor the binding of cephalosporin substrates.Figure 8The superposition of the active sites in PER-1 (black) and TEM (red) shows the significant enlargement of the substrate binding cavity in PER-1.View Large Image Figure ViewerDownload (PPT)The structure determinations of the PER-1 β-lactamase breaks the assumption of a unique fold of the class A enzymes. This finding promoted the suggestion that class A β-lactamases may be categorized into three subgroups according to their kinetic properties as penicillinases (TEM/SHV group), cephalosporinases (PER group), and carbapenemases (NMC-A group). These groups have sequence signatures that should help in assigning any new enzymes into either of them. From the microbiological point of view, the simultaneous occurrence in bacteria of extended spectrum TEM or SHV and PER enzymes should complicate identification of the resistance enzymes based on the antibiogram patterns. Although PER-1 is a typical class A enzyme with respect to the conservation of the catalytic machinery, its three-dimensional structure revealed that this superfamily is not as homogeneous in structure as it was thought. It was unexpected to find a new fold of the Ω-loop, a region considered to be a canonical motif in these enzymes. This finding suggested to classify β-lactamase sequences according to the occurence of an aspartic acid at position 136. Several enzymes with significant sequence identity (40%) were identified that form a subgroup in the class A superfamily (Figs. 4 and6). These “PER-like” enzymes (11Nordmann P. Naas T. Antimicrob. Agents Chemother. 1994; 38: 104-114Crossref PubMed Scopus (119) Google Scholar) comprise PER-2 from S. typhimurium (10Bauernfeind A. Stemplinger I. Jungwirth R. Mangold P. Amann S. Akalin E. Ang Ö. Bal C. Casellas J.M. Antimicrob. Agents Chemother. 1996; 40: 616-620Crossref PubMed Google Scholar), VEB-1 from E. coli (57Poirel L. Naas T. Guibert M. Chaibi E.B. Labia R. Nordmann P. Antimicrob. Agents Chemother. 1999; 43: 573-581Crossref PubMed Google Scholar), CME-1 from Chryseobacterium(flavobacterium) meningosepticum (58Rossolini G.M. Franceschini N. Lauretti L. Caravelli B. Riccio M.L. Galleni M. Frere J.-M. Amicosante G. Antimicrob. Agents Chemother. 1999; 43: 2193-2199Crossref PubMed Google Scholar), TLA-1 fromE. coli (59Silva J. Aguilar C. Ayala G. Estrada M.A. Garza-Ramos U. Lara-Lemus R. Ledezma L. Antimicrob. Agents Chemother. 2000; 44: 997-1003Crossref PubMed Scopus (70) Google Scholar), CBLA from Bacteroides uniformis(15Smith C.J. Bennet T.K. Parker A.C. Antimicrob. Agents Chemother. 1994; 38: 1711-1715Crossref PubMed Scopus (35) Google Scholar), CEPA from Bacteroides fragilis (13Rogers M.B. Parker A.C. Smith C.J. Antimicrob. Agents Chemother. 1993; 37: 2391-2400Crossref PubMed Scopus (69) Google Scholar), and CFXA fromBacteroides vulgatus (14Parker A.C. Smith C.J. Antimicrob. Agents Chemother. 1993; 37: 1028-1036Crossref PubMed Scopus (103) Google Scholar). The phylogenetic tree shown in Fig. 6 indicates that these proteins constitute a distinct cluster of enzymes, and the x-ray structure of PER-1 suggests that this partition reflects major structural differences compared with the TEM-like β-lactamases. On the contrary, there is no major partition between the penicillinases and the NMC-A, IMI-1 (60Rasmussen B.A. Bush K. Keeney D. Yang Y. Hare R. O'Gara C. Medeiros A.A. Antimicrob. Agents Chemother. 1996; 40: 2080-2086Crossref PubMed Google Scholar), and Sme-1 (61Naas T. Vedel L. Sougakoff W. Livermore D.M. Nordmann P. Antimicrob. Agents Chemother. 1994; 38: 1262-1270Crossref PubMed Scopus (180) Google Scholar) carbapenemases, which may be identified by the presence of cysteine residues at positions 69 and 238 (Fig. 4). It agrees with the recent structure determination of NMC-A (20Swarén P. Maveyraud L. Raquet X. Cabantous S. Duez C. Pédelacq J.-D. Mariotte-Boyer S. Mourey L. Labia R. Nicolas-Chanoine M.-H. Nordmann P. Frère J.-M. Samama J.-P. J. Biol. Chem. 1998; 273: 26714-26738Abstract Full Text Full Text PDF PubMed Scopus (80) Google Scholar), which revealed that the carbapenemase activity was associated with subtle structural modifications. For reasons discussed below, we suggest that the presence of both Asp136 and Asn179 should identify any new β-lactamase as member of the PER subgroup. Previous studies on class A β-lactamases, including x-ray structure determinations, led to the conclusion that the cisconformation of the 166–167 peptide bond was mandatory for the proper location of Glu166 in the active site, and therefore for catalysis. The N136A mutant of the PC1 enzyme was found to accumulate acyl-enzyme adducts but had no hydrolytic activity against penicillin and cefotaxime (29Banerjee S. Pieter U. Kapadia G. Pannell L.K. Herzberg O. Biochemistry. 1998; 37: 3286-3296Crossref PubMed Scopus (84) Google Scholar). Recent kinetic analysis of randomly generated Ω-loop TEM mutants concluded that the isomerization of the 166–167 peptide bond controlled one conformation of this loop that was compatible with fast acylation of the protein (27Vanwetswinkel S. Avalle B. Fastrez J. J. Mol. Biol. 2000; 295: 527-540Crossref PubMed Scopus (26) Google Scholar). The PER-1 structure presented here reveals that a functional enzyme can accommodate thetrans conformation of the 166–167 peptide bond through interactions with Asp136 and maintains the position of Glu166 that preserves the molecular basis of the catalytic mechanism. The trans conformation goes with a different sequence of the 161–181 (Ω-loop) region, which displays a new fold compared with the typical class A enzymes. In those cases, the Ω-loop is stabilized by salt bridge interactions, and the R164S mutation is frequently observed in naturally occurring TEM and SHV mutants with extended spectrum (62Jacoby G.A. Medeiros A.A. Antimicrob. Agents Chemother. 1991; 35: 1697-1704Crossref PubMed Scopus (649) Google Scholar). The disruption of the Arg164–Asp179 salt bridge is assumed to increase the flexibility of the Ω-loop, thereby favoring the binding of third generation cephalosporins (63Vakulenko S.B. Toth M. Taibi P. Mobashery S. Lerner S.A. Antimicrob. Agents Chemother. 1995; 39: 1878-1880Crossref PubMed Scopus (34) Google Scholar, 64Vakulenko S.B. Taibi P. Toth M. Massova I. Lerner S.A. Mobashery S. J. Biol. Chem. 1999; 274: 23052-23060Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar). In PC1, the β-lactamase from S. aureus, disruption of this single salt bridge by the D179N mutation led to a disordered conformation of the Ω-loop (29Banerjee S. Pieter U. Kapadia G. Pannell L.K. Herzberg O. Biochemistry. 1998; 37: 3286-3296Crossref PubMed Scopus (84) Google Scholar,65Herzberg O. Kapadia G. Blanco B. Smith T.S. Coulson A. Biochemistry. 1991; 30: 9503-9509Crossref PubMed Scopus (67) Google Scholar). There are no salt bridges in PER-1 where the stability of the Ω-loop fold stems from the presence of secondary structure elements complemented by the hydrogen bond interactions of the side chain of Asn179 to the main chain nitrogen atoms of Val163 and Ala164 and to the main chain oxygen atom of Ala164 (Fig. 5). In retrospect and in view of the major structural differences between the PER-1 and the TEM enzymes, it is not surprising that site-directed mutagenesis of PER-1 (34Bouthors A.-T. Dagoneau-Blanchard N. Naas T. Nordmann P. Jarlier V. Sougakoff W. Biochem. J. 1998; 330: 1443-1449Crossref PubMed Scopus (33) Google Scholar, 35Bouthors A.-T. Delettré J. Mugnier P. Jarlier V. Sougakoff W. Protein Eng. 1999; 12: 313-318Crossref PubMed Scopus (25) Google Scholar) could not relate the cephalosporinase activity of PER-1 to the residue type at positions 164, 179, 238, and 240. These residues, which provide extended spectrum profiles in the TEM and SHV enzymes, have different spatial location and environments in these three-dimensional structures. Protein engineering in PER-1 nevertheless identified two residues whose mutations had significant kinetic effects. Replacement of Thr104 with Glu completely abolished the catalytic activity on penicillins and reduced thek cat values for cephalosporins by a factor of 50–700 (34Bouthors A.-T. Dagoneau-Blanchard N. Naas T. Nordmann P. Jarlier V. Sougakoff W. Biochem. J. 1998; 330: 1443-1449Crossref PubMed Scopus (33) Google Scholar). According to the x-ray structure, this mutation should in any case disrupt the interaction of the γ-hydroxyl group of Thr104 with Asn132. These observations shed light on the previously suggested importance for catalysis in class A enzymes of the hydrogen bond interaction between residue 104 and Asn132 (50Petit A. Maveyraud L. Lenfant F. Samama J.-P. Labia R. Masson J.-M. Biochem. J. 1995; 305: 33-40Crossref PubMed Scopus (50) Google Scholar, 66Matagne A. Frère J.-M. Biochim. Biophys. Acta. 1995; 1246: 109-127Crossref PubMed Scopus (104) Google Scholar). It involves the main chain oxygen atom of residue 104 in all structures except PER-1, and its contribution could evidently not be demonstrated by protein engineering in these enzymes. Replacement of Thr237 by an alanine residue increased thek cat/K m of the mutant protein on cephalosporin substrates by 10–100-fold (35Bouthors A.-T. Delettré J. Mugnier P. Jarlier V. Sougakoff W. Protein Eng. 1999; 12: 313-318Crossref PubMed Scopus (25) Google Scholar). Interestingly, it is the reverse mutation, A237T, that improved hydrolysis on cephalosporin substrates in the TEM enzyme (67Bush K. Jacoby G.A. J. Antimicrob. Chemother. 1997; 39: 1-3Crossref PubMed Scopus (82) Google Scholar). The relocation of the side chain of residue 104, the fold of the Ω-loop, and the conformation of the 238–242 region in PER-1 seem in direct relationship with the cephalosporinase activity of this enzyme. Indeed, these folding features generate a cavity, filled with several water molecules in the apoenzyme, that is precisely located in the area expected to bind the bulky side chain of third and fourth generation cephalosporins (33Raquet X. Lamotte-Brasseur J. Fonzé E. Goussard S. Courvalin P. Frère J.-M. J. Mol. Biol. 1994; 244: 625-639Crossref PubMed Scopus (124) Google Scholar) (Fig. 7). According to the superposed TEM and PER-1 structures (Fig.8), this observation supports the proposal that the displacement of the Ω-loop in the TEM enzyme should favor the binding of cephalosporin substrates. The structure determinations of the PER-1 β-lactamase breaks the assumption of a unique fold of the class A enzymes. This finding promoted the suggestion that class A β-lactamases may be categorized into three subgroups according to their kinetic properties as penicillinases (TEM/SHV group), cephalosporinases (PER group), and carbapenemases (NMC-A group). These groups have sequence signatures that should help in assigning any new enzymes into either of them. From the microbiological point of view, the simultaneous occurrence in bacteria of extended spectrum TEM or SHV and PER enzymes should complicate identification of the resistance enzymes based on the antibiogram patterns.
Referência(s)