Portal de Periódicos da CAPES

Resistance to β-Lactam Antibiotics and Its Mediation by the Sensor Domain of the Transmembrane BlaR Signaling Pathway in Staphylococcus aureus

2003; Elsevier BV; Volume: 278; Issue: 20 Linguagem: Inglês

10.1074/jbc.m300611200

1083-351X

Dasantila Golemi‐Kotra, Joo Young, Samy O. Meroueh, Sergei B. Vakulenko, Shahriar Mobashery,

Antibiotics Pharmacokinetics and Efficacy

Staphylococci, a leading cause of infections worldwide, have devised two mechanisms for resistance to β-lactam antibiotics. One is production of β-lactamases, hydrolytic resistance enzymes, and the other is the expression of penicillin-binding protein 2a (PBP 2a), which is not susceptible to inhibition by β-lactam antibiotics. The β-lactam sensor-transducer (BlaR), an integral membrane protein, binds β-lactam antibiotics on the cell surface and transduces the information to the cytoplasm, where gene expression is derepressed for both β-lactamase and penicillin-binding protein 2a. The gene for the sensor domain of the sensor-transducer protein (BlaRS) of Staphylococcus aureuswas cloned, and the protein was purified to homogeneity. It is shown that β-lactam antibiotics covalently modify the BlaRSprotein. The protein was shown to contain the unusual carboxylated lysine that activates the active site serine residue for acylation by the β-lactam antibiotics. The details of the kinetics of interactions of the BlaRS protein with a series of β-lactam antibiotics were investigated. The protein undergoes acylation by β-lactam antibiotics with microscopic rate constants (k2) of 1–26 s−1, yet the deacylation process was essentially irreversible within one cell cycle. The protein undergoes a significant conformational change on binding with β-lactam antibiotics, a process that commences at the preacylation complex and reaches its full effect after protein acylation has been accomplished. These conformational changes are likely to be central to the signal transduction events when the organism is exposed to the β-lactam antibiotic. Staphylococci, a leading cause of infections worldwide, have devised two mechanisms for resistance to β-lactam antibiotics. One is production of β-lactamases, hydrolytic resistance enzymes, and the other is the expression of penicillin-binding protein 2a (PBP 2a), which is not susceptible to inhibition by β-lactam antibiotics. The β-lactam sensor-transducer (BlaR), an integral membrane protein, binds β-lactam antibiotics on the cell surface and transduces the information to the cytoplasm, where gene expression is derepressed for both β-lactamase and penicillin-binding protein 2a. The gene for the sensor domain of the sensor-transducer protein (BlaRS) of Staphylococcus aureuswas cloned, and the protein was purified to homogeneity. It is shown that β-lactam antibiotics covalently modify the BlaRSprotein. The protein was shown to contain the unusual carboxylated lysine that activates the active site serine residue for acylation by the β-lactam antibiotics. The details of the kinetics of interactions of the BlaRS protein with a series of β-lactam antibiotics were investigated. The protein undergoes acylation by β-lactam antibiotics with microscopic rate constants (k2) of 1–26 s−1, yet the deacylation process was essentially irreversible within one cell cycle. The protein undergoes a significant conformational change on binding with β-lactam antibiotics, a process that commences at the preacylation complex and reaches its full effect after protein acylation has been accomplished. These conformational changes are likely to be central to the signal transduction events when the organism is exposed to the β-lactam antibiotic. penicillin-binding protein 2-{[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]amino}ethanesulfonic acid Staphylococci are the most common cause of bacterial infections in the United States (1Boyce J.M. Crossley K.B. Archer G.L. The Staphylococci in Human Disease. Churchill Livingstone, New York1997: 309-330Google Scholar). Among these organisms, methicillin-resistant Staphylococcus aureus has received notoriety since it is currently the scourge of hospitals. Staphylococci have acquired multiple drug resistance genes over the past few decades such that methicillin-resistant S. aureus can usually be treated only by glycopeptides, such as vancomycin (2Novick R.P. Schlievert P. Ruzin A. Microbes Infect. 2001; 7: 585-594Crossref Scopus (116) Google Scholar), or by oxazolidinones, such as linezolid (3Fung H.B. Kirschenbaum H.L. Ojofeitimi B.O. Clin. Ther. 2001; 23: 356-391Abstract Full Text PDF PubMed Scopus (100) Google Scholar). It is disconcerting that variants of methicillin-resistant S. aureus that have become at least partially resistant to glycopeptide (4Walsh T.R. Blomström A. Qwärnström A.H.P. Wootton M. Howe R.A. MacGowan A.P. Diekema D.J. J. Clin. Microbiol. 2002; 4: 2439-2444Google Scholar) and oxazolidinone antibiotics are being identified (5Tsiodras S. Gold H.S. Sakoulas G. Eliopoulos G.M. Wennersten C. Vankataraman L. Moellering Jr., R.C. Ferraro M.J. Lancet. 2001; 358: 207-208Abstract Full Text Full Text PDF PubMed Scopus (907) Google Scholar).β-Lactam antibiotics target penicillin-binding proteins (PBPs)1 for inhibition. Staphylococci have become resistant to β-lactam antibiotics by two parallel mechanisms. First, they produce β-lactamases, enzymes that hydrolytically destroy these antibiotics (6Bush K. Mobashery S. Adv. Exp. Med. Biol. 1998; 456: 71-98Crossref PubMed Scopus (108) Google Scholar, 7Kotra L.P. Samama J.P. Mobashery S. Lewis A. Salyers A. Haber H. Wax R.G. Bacterial Resistance to Antimicrobials, Mechanisms, Genetics, Medical Practice and Public Health. Marcel Dekker, New York2001: 123-159Google Scholar). Second, they have acquired a novel PBP (referred to as PBP 2a) that carries out the physiological functions of the four existing PBPs in staphylococci yet is not inhibited in vivo by any of the clinically used β-lactam antibiotics. The acquisition of the gene for PBP 2a in staphylococci took place only once from an unknown source (8Kreiswirth B. Kornblum J. Arbeit R.D. Eisner W. Maslow J.N. McGeer A. Low D.E. Novick R.P. Science. 1993; 259: 227-230Crossref PubMed Scopus (328) Google Scholar).The blaZ and mecA genes encode the staphylococcal β-lactamase and PBP 2a, respectively. Transcription of these genes is regulated by signal-transducing integral membrane proteins BlaR and MecR and their respective transcriptional repressor proteins BlaI and MecI. BlaR has a β-lactam-binding domain on the surface of the plasma membrane and a zinc protease domain in the cytoplasm (Fig.1) (9Zhang H.Z. Hackbarth C.J. Chansky K.M. Chambers H.F. Science. 2001; 291: 1962-1965Crossref PubMed Scopus (177) Google Scholar). The β-lactam antibiotic acylates a specific active site serine on the cell surface domain, which transduces a signal to the cytoplasmic domain. The zinc-dependent cytoplasmic protease domain, which hydrolyzes the repressor proteins, precipitates transcription of the genes for the two resistance proteins. This proteolytic signaling pathway is unique in bacteria (10Archer G.L. Bosilevac J.M. Science. 2001; 291: 1915-1916Crossref PubMed Scopus (43) Google Scholar). We have cloned, expressed, and purified to homogeneity the cell surface sensor domain of the BlaR protein (referred to as BlaRS hereafter). As an unusual feature, this protein is carboxylated at the side chain of an active site lysine (Lys-392; a carbamate on the side chain), which is critical for active site serine acylation by the β-lactam antibiotic. The kinetics of these processes are reported herein for the first time. Furthermore, binding of the antibiotic to BlaRS entails a significant conformational change, a process that is likely to play a role in the signal transduction mechanism from the cell surface to the cytoplasm.RESULTS AND DISCUSSIONWe cloned the sensor domain of the signal transducer protein BlaR from S. aureus (spans amino acids 331–581), which we refer to as the BlaRS protein. The cloned protein was expressed in the cytoplasm of E. coli. The protein was purified to homogeneity in two chromatographic steps and was highly soluble (up to 26 mg/ml). We routinely obtain 50 mg of pure protein from one liter of growth medium. The C terminus of BlaR from Bacillus licheniformis has also been cloned (23Joris B. Ledent P. Kobayshi T. Lampen J.O. Ghysen J.M. FEMS Microbiol. Lett. 1990; 70: 107-114Crossref Google Scholar, 24Zhu Y. Curran I.H. Joris B. Ghysen J.M. Lampen O. J. Bacteriol. 1990; 172: 1137-1141Crossref PubMed Google Scholar).Carboxylation of Lysine Side Chain in the BlaR ProteinThe BlaRS protein is related to the OXA family of β-lactamases, enzymes of resistance to β-lactam antibiotics (24Zhu Y. Curran I.H. Joris B. Ghysen J.M. Lampen O. J. Bacteriol. 1990; 172: 1137-1141Crossref PubMed Google Scholar,25Massova I. Mobashery S. Antimicrob. Agents Chemother. 1998; 42: 1-17Crossref PubMed Scopus (15) Google Scholar). The active site peptide sequence of Ser-X-X-Lys, which is a known minimal motif for these proteins that undergo acylation at the serine residue, is present in both (25Massova I. Mobashery S. Antimicrob. Agents Chemother. 1998; 42: 1-17Crossref PubMed Scopus (15) Google Scholar). The x-ray structures for the OXA-10 β-lactamase (26Golemi D. Maveyraud L. Vakulenko S.B. Samama J.P. Mobashery S. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 14280-14285Crossref PubMed Scopus (196) Google Scholar,27Maveyraud L. Golemi D. Ishiwata A. Meroueh O. Mobashery S. Samama J.P. J. Am. Chem. Soc. 2002; 124: 2461-2465Crossref PubMed Scopus (66) Google Scholar) reveal that the active site lysine is carboxylated on its side chain (i.e. the carbamate product of reaction with carbon dioxide). The side chain of lysine in the OXA-10 β-lactamase is sequestered in an unusual environment made up of five hydrophobic amino acid side chains (Phe-69, Val-117, Phe-120, Trp-154, and Leu-155) that is believed to lower the pKa of the lysine side chain such that it exists in the free base form that undergoes reaction with carbon dioxide (26Golemi D. Maveyraud L. Vakulenko S.B. Samama J.P. Mobashery S. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 14280-14285Crossref PubMed Scopus (196) Google Scholar). The side chain of the carboxylated lysine and that of serine are in contact, and the former activates the latter for enzyme acylation by β-lactam antibiotics (26Golemi D. Maveyraud L. Vakulenko S.B. Samama J.P. Mobashery S. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 14280-14285Crossref PubMed Scopus (196) Google Scholar). The requisite amino acids in the Ser-X-X-Lys motif and the five hydrophobic sites, among others, are conserved among the many OXA β-lactamases and the BlaR protein (25Massova I. Mobashery S. Antimicrob. Agents Chemother. 1998; 42: 1-17Crossref PubMed Scopus (15) Google Scholar). A pertinent question now is whether the sensor domain of the BlaR protein is also carboxylated at the corresponding lysine residue.A diagnostic test for carboxylation of the lysine side chain is by13C NMR, which detects a distinctive signal. The13C NMR experiment indicated that lysine carboxylation is seen in the BlaRS protein, as shown by the presence of a diagnostic resonance at 164 ppm (Fig. 2). The same experiment was carried out with the K392A mutant protein, and unexpectedly, we observed that the NMR signal at 164 ppm was not entirely eliminated (Fig. 2B). The integrations of the carbamate signals in Fig. 2 indicated that the wild type enzyme (Fig.2A) had approximately two carboxylated lysines to one in the mutant protein. Therefore, under the NMR experiment conditions, two lysines in the wild-type protein exist in the free base forms, which undergo carboxylation in the presence of the13C-labeled carbon dioxide, one of which is at position 392.Figure 2The 13C NMR spectra of the wild-type BlaRS (A) and the K392A mutant proteins (at 1 mm) (B) in 10 mm sodium phosphate, 0.1 mm EDTA, supplemented with 20 mmNaH13CO3.View Large Image Figure ViewerDownload (PPT)We resorted to binding of radioactive carbon dioxide to the BlaRS and the K392A mutant proteins. Analogously to the case of the OXA-10 β-lactamases, the expectation was that the active site carboxylated lysine would be stabilized by specific interactions. On the other hand, the other carboxylated lysine seen in the NMR experiment might have experienced carboxylation in an adventitious process and could be back-titrated by non-radioactive carbon dioxide. Here, each protein was incubated at pH 4.5 to facilitate the decarboxylation of lysine followed by reconstitution of the protein by the radioactively labeled carbon dioxide. The workup was made in the presence of non-labeled carbon dioxide. We were able to measure an average of 0.9 equivalents of radioactive label incorporated per each of the wild-type protein molecule. By the same procedure, no label was introduced into the K392A mutant protein. This argued that carboxylation of the protein was indeed at the Lys-392 position and that there may be another exposed lysine with reduced pKa in this protein that may be partiallyand unstably carboxylated under the conditions of the NMR experiment (i.e. high carbon dioxide concentration in the NMR tube). The cloned protein has a total of 31 lysine residues.In the case of the OXA-10 β-lactamase, proximity of one of the oxygen atoms of the lysine carbamate to Trp-154 was useful in quantitative fluorescence quenching studies of the interaction of carbon dioxide with the enzyme (26Golemi D. Maveyraud L. Vakulenko S.B. Samama J.P. Mobashery S. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 14280-14285Crossref PubMed Scopus (196) Google Scholar). Tryptophan 475 of the BlaR protein corresponds to Trp-154 of the OXA-10 β-lactamase, so we felt that the dissociation constant for carbon dioxide of the BlaRS protein may be evaluated by fluorescence analyses. The intrinsic fluorescence of BlaRS was quenched upon addition of sodium bicarbonate as the source of carbon dioxide in a saturable fashion. Data fit to Equation 1 revealed a Kd of 0.6 ± 0.2 μm (Fig. 3). Considering that the physiological concentration of carbon dioxide is 1.3 mm (28Tien M. Berlett B.S. Levine R.L. Chock P.B. Stadtman E.R. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 7809-7814Crossref PubMed Scopus (157) Google Scholar), this indicates that BlaRS is fully carboxylated in vivo. It is significant to note that the K392A mutant protein did not give the tryptophan fluorescence quench, which is indicative of the fact that the dissociation constant that was evaluated for the wild-type protein was for carboxylation of residue 392 and is a further validation that this residue is indeed carboxylated in the wild-type protein. As will be described below, carboxylated Lys-392 is the active site base that activates the serine for protein acylation. This is now only the second example of a protein, after the OXA-10 β-lactamases (26Golemi D. Maveyraud L. Vakulenko S.B. Samama J.P. Mobashery S. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 14280-14285Crossref PubMed Scopus (196) Google Scholar), that uses the highly uncommon carboxylated lysine as a basic residue to facilitate reactions in the active site. The few other proteins having carboxylated lysine use the modified amino acid as metal ligand or for hydrogen bonding in the protein structure.Figure 3Relative quenching of the intrinsic tryptophan fluorescence of BlaRS protein (1 μm) versus total concentration of carbon dioxide (μm). Data were fit to a single binding site model as per the equation described under "Experimental Procedures."View Large Image Figure ViewerDownload (PPT)The collective information in the preceding paragraphs made possible the generation of a homology-based computational model for the sensor domain of the BlaR protein (Fig. 4). In comparison with the structure of the OXA-10 β-lactamase, the arrangements of the side chains of serine and the carboxylated lysine and the hydrophobic environment around the lysine, including the proximity of the carboxylated lysine and the tryptophan residue, are preserved.Figure 4A, stereo view of the computational model for the β-lactam binding site of the BlaR protein. A Connolly water-accessible surface (in green) was constructed around the hydrophobic pocket of the binding site residues. The side chains of the residues that make up the hydrophobic pocket are shown in orange-capped sticks. Carboxylated Lys-392 (shown in the middle of the Connolly surface), Ser-389, and Trp-475 are color-coded according to atom types and shown in capped-sticks representation (white,red, and blue correspond to carbon, oxygen, and nitrogen, respectively). Hydrogen bonding interactions between carboxylated Lys-392 and Ser-389 and Trp-475 are represented with awhite dashed line. The protein is shown in a purple tube representation. In B, a similar perspective from the x-ray structure for the OXA-10 β-lactamase is given for comparison.View Large Image Figure ViewerDownload (PPT)Kinetics of Interactions of β-Lactam Antibiotics and the BlaRS ProteinIn light of the information that the sensor domain of the BlaR protein has a carboxylated lysine, it is conceivable that the protein at the end of each individual purification protocol would be carboxylated to varying degrees, since the process is reversible. This point was documented by observing typically a 2-fold enhancement of the rate of interactions of the BlaRSprotein with β-lactam antibiotics by supplementation of the buffer with sodium bicarbonate (as a source of carbon dioxide). As argued above, the BlaR protein is fully carboxylated in vivo, and the fact that some of the carboxylation of the protein is reversed during the purification is an artifact. Therefore, we have supplemented the reaction mixtures for the kinetic studies with bicarbonate to generate the fully carboxylated and active form of the protein for all kinetic determinations.BOCILLIN FL, a fluorescent penicillin, was used to further study and analyze the mode of action of BlaRS protein. This molecule modifies BlaRS covalently, as would any β-lactam antibiotic, whereby the protein would migrate through an SDS-polyacrylamide gel to allow quantitative detection by Fluorimager. Titration of the BlaRS protein with BOCILLIN FL revealed saturation and also indicated a one-to-one modification of the protein by the antibiotic. The wild-type and K393A mutant BlaRSproteins were acylated by BOCILIN FL, as revealed by Fluorimager. The K393A has a residual level of activity (see below) that accounts for this observation. In contrast, incubation of the S389A mutant variant with BOCILIN FL did not give a fluoregenic band. A previous study based on sequence analysis of the BlaR from B. licheniformis with a class D β-lactamase had suggested that residue Ser-389 (BlaR numbering according to S. aureus) might be the modification site by β-lactams (23Joris B. Ledent P. Kobayshi T. Lampen J.O. Ghysen J.M. FEMS Microbiol. Lett. 1990; 70: 107-114Crossref Google Scholar). The experiments reported herein clearly reveal Ser-389 to be the serine-active site residue that is acylated.The kinetics of interactions of several β-lactam molecules (three penicillins, three cephalosporins, and one carbapenem) with the BlaRS protein were investigated (TableI). Stopped-flow kinetics distinguished between a rapid enzyme acylation event and a substantially slower deacylation step. Acylation of the active site proceeded with microscopic rate constants (i.e. k2) of 1–26 s−1, which indicate rapid t1/2 values for acylation of 27–690 ms for the β-lactams that we studied. The deacylation rate constants (i.e. k3) for the same β-lactam molecules are listed in Table I, corresponding to t1/2 values of ∼12–240 min. The k3 for oxacillin would appear to be representative of most of the substrates studied, with a value of (4.8 ± 0.6) × 10−5 s−1, which corresponds to a t1/2 value of 240 min. In light of the fact that typical strains of S. aureus double their population sizes in 20–30 min under favorable growth conditions, this indicates that a single acylation event per each molecule of the β-lactam signal sensor-transducer protein accounts for the biological consequences per each generation of bacterial growth.Table IKinetic parameters of BlaRS with the β-lactam antibioticsSubstratek2k3Ksk2/Kss−1(s−1)×105μm(m−1s−1)× 10−5Nitrocefin26 ± 692 ± 1024 ± 911 ± 4FA-penicillin1.6 ± 0.18 ± 413 ± 21.2 ± 0.2Ampicillin1.0 ± 0.19 ± 123 ± 20.4 ± 0.1Oxacillin18 ± 14.8 ± 0.649 ± 53.7 ± 0.4Cefepime1.4 ± 0.112.6 ± 0.325 ± 30.6 ± 0.1Ceftazidime4.1 ± 0.59.1 ± 0.468 ± 160.6 ± 0.2Imipenem6 ± 110 ± 1183 ± 660.3 ± 0.1The rate constants were determined in the presence of 50 mmNaHCO3 at pH 7.0. The concentration of the protein in the assays was 1 μm. The kinetic parameters for substrates other than nitrocefin were determined in competition experiments with nitrocefin. (See "Experimental Procedures" for more details.) Open table in a new tab It is a feature of the signal sensor-transducer protein that it is activated by all β-lactam antibiotics (29McKinney T.K. Sharma V.K. Craig W.A. Archer G.L. J. Bacteriol. 2001; 183: 6862-6868Crossref PubMed Scopus (105) Google Scholar). Consistent with this information, the dissociation constants (i.e. Ks) for various β-lactam antibiotics are in the micromolar range, which are attainable in the milieu where the bacteria grow. The dissociation constants are practically in the same range for the three penicillins and three cephalosporins that we tested, whereas the carbapenem imipenem shows a higher value (Table I). The second-order rate constants (i.e. k2/Ks) for the encounter of the β-lactam molecules and the BlaRS protein were typically 104 to 106m−1 s−1, indicative of a very favorable process.As per the computational model and the foregoing evidence, we decided to evaluate the effect of Lys-392 on the kinetics of BlaRSacylation. In the case of the OXA-10 β-lactamase, the mutational change of the corresponding carboxylated lysine resulted in an inactive enzyme that did not experience acylation in the active site by the β-lactam antibiotics (26Golemi D. Maveyraud L. Vakulenko S.B. Samama J.P. Mobashery S. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 14280-14285Crossref PubMed Scopus (196) Google Scholar). Similarly to the case of the OXA-10 β-lactamase, a mutational change of Lys-392 to Ala in BlaRS resulted in a protein that was severely impaired in acylation of the active site serine as evaluated for oxacillin (k2 = 0.0026 ± 0.0005 s−1 andKs = 43 ± 14 μm). The rate constant for acylation was attenuated for the mutant protein by 6730-fold with no change in Ks. Since the mutant protein has the same conformation as the wild-type protein by circular dichroic analyses (see below), the attenuation onk2 may be ascribed to poor activation of serine in the mutant protein.Conformational Change in the BlaRS ProteinSignal transduction from one side of the membrane to the other necessitates communication between the surface and the cytoplasmic domains. A means to this communication is by conformational change of the membrane-bound protein after binding to the β-lactam antibiotic. As shown in Fig.5, the BlaRS protein is prone to significant conformational change on binding to the β-lactam antibiotics. The conformational change commences upon binding to the β-lactam antibiotic at the preacylation complex and reaches its full extent on protein acylation. If the acyl-protein species is allowed to undergo its sluggish deacylation, the protein returns to the native conformation (data not shown). As indicated earlier, the K392A mutant variant of the BlaRS protein is severely deficient in the acylation step. This mutant variant would not experience acylation by β-lactam antibiotics during the course of the CD experiment. However, the non-covalent binding by a β-lactam antibiotic, for example by oxacillin (Fig. 5B), resulted in a discernable change in the CD spectrum of the protein (similar results were seen with the S389A mutant protein, which does not have the opportunity to give the acyl-protein species; see Supplemental Material). The minima at 208 and 222 nm, which are due to the helices, were enhanced, and the maximum at 195 nm, due to β-sheets, sharpened (Fig. 5B). These data argue for the enhancement of secondary structures (helicity and β-sheets) in the protein on non-covalent binding by the antibiotic. Upon acylation of the protein by oxacillin (Fig.5A), these effects were enhanced further, but the native state returned upon deacylation. Similar results and trends were noted for all β-lactam antibiotics shown in Table I (see Supplemental Material), so the effects of the conformational change on the BlaRS protein are shared by all of these antibiotics. In light of the fact that this conformational change is significant and is generally seen regardless of the nature of the β-lactam antibiotic, we believe that it is likely that it plays a role in the signal transduction process. However, we acknowledge the fact that in the whole cell context, other factors may play a role as well.Figure 5Circular dichroic spectra of the wild-type BlaR protein (2 μm, solid line) and the wild-type BlaRS protein (2 μm) incubated with oxacillin (30 μm, broken line) (A) of the K392A mutant variant of the BlaRSprotein (2 μm, solid line) and the K392A mutant (2 μm) incubated with oxacillin (30 μm, broken line) (B) and of the S389A mutant variant of the BlaR protein (2 μm,solid line) and the S389A mutant (2 μm) incubated with oxacillin (30 μm, broken line) (C). All the spectra were corrected for the small contribution from the antibiotic in the mixture.View Large Image Figure ViewerDownload (PPT)The essence of signal transduction is the switch between inactive and active forms of a given protein. In the case of the β-lactam signal sensor-transducer, a key question is how binding of the β-lactam antibiotic to the sensor domain facilitates signal transduction. Two classical models for signal transduction have been proposed (30Volkman B.F. Lipson D. Wemmer D.E. Kern D. Science. 2001; 291: 2429-2433Crossref PubMed Scopus (534) Google Scholar). In one, ligand binding induces the formation of a new conformation in the protein. In the other, an equilibrium mixture of conformational states exists, and the ligand binding shifts the equilibrium in favor of the active form. The data presented here for the BlaR protein are consistent with either the induced-fit or the population shift model. It is conceivable that the protein switches its conformation on binding by the β-lactam antibiotic during the formation of the preacylation complex, which reaches its maximal effect after acylation of the enzyme in the active site (the induced-fit model). Alternatively, the CD spectrum of the native BlaR protein may be due to the distribution of several preexisting conformational states, which binding of the β-lactam antibiotics to one would shift the equilibrium in favor of the active structure (the population shift model). The discrimination between these two models for the case of the BlaR protein should await availability of structural information in the future.We have described in this report the dynamic nature of the sensor domain of the BlaR protein from staphylococci. This protein undergoes structural rearrangement on binding to a wide range of β-lactam antibiotics, the implications of which for the signal transduction event remain to be studied by structural biologists. We have shown that β-lactam antibiotics modify the protein covalently and essentially irreversibly within a bacterial population doubling time at Ser-389. The covalent modification of the sensor domain is facilitated by an uncommon carboxylated lysine at position 392 within the antibiotic binding site. The means by which the BlaR system carries out its signal sensing and transducing processes is unique to Gram-positive bacteria. We have provided herein insights into how this specific protein inS. aureus facilitates the manifestation of resistance to β-lactam antibiotics, a process that remains a challenge in clinical treatment of infections caused by this organism. Staphylococci are the most common cause of bacterial infections in the United States (1Boyce J.M. Crossley K.B. Archer G.L. The Staphylococci in Human Disease. Churchill Livingstone, New York1997: 309-330Google Scholar). Among these organisms, methicillin-resistant Staphylococcus aureus has received notoriety since it is currently the scourge of hospitals. Staphylococci have acquired multiple drug resistance genes over the past few decades such that methicillin-resistant S. aureus can usually be treated only by glycopeptides, such as vancomycin (2Novick R.P. Schlievert P. Ruzin A. Microbes Infect. 2001; 7: 585-594Crossref Scopus (116) Google Scholar), or by oxazolidinones, such as linezolid (3Fung H.B. Kirschenbaum H.L. Ojofeitimi B.O. Clin. Ther. 2001; 23: 356-391Abstract Full Text PDF PubMed Scopus (100) Google Scholar). It is disconcerting that variants of methicillin-resistant S. aureus that have become at least partially resistant to glycopeptide (4Walsh T.R. Blomström A. Qwärnström A.H.P. Wootton M. Howe R.A. MacGowan A.P. Diekema D.J. J. Clin. Microbiol. 2002; 4: 2439-2444Google Scholar) and oxazolidinone antibiotics are being identified (5Tsiodras S. Gold H.S. Sakoulas G. Eliopoulos G.M. Wennersten C. Vankataraman L. Moellering Jr., R.C. Ferraro M.J. Lancet. 2001; 358: 207-208Abstract Full Text Full Text PDF PubMed Scopus (907) Google Scholar). β-Lactam antibiotics target penicillin-binding proteins (PBPs)1 for inhibition. Staphylococci have become resistant to β-lactam antibiotics by two parallel mechanisms. First, they produce β-lactamases, enzymes that hydrolytically destroy these antibiotics (6Bush K. Mobashery S. Adv. Exp. Med. Biol. 1998; 456: 71-98Crossref PubMed Scopus (108) Google Scholar, 7Kotra L.P. Samama J.P.