Optimization of conditions for the glycosyltransferase activity of penicillin-binding protein 1a from Thermotoga maritima Julien Offant 1,2,3 , Mohammed Terrak 4 , Adeline Derouaux 4 , Eefjan Breukink 5 , Martine Nguyen-Diste ` che 4 , Andre ´ Zapun 1,2,3 and Thierry Vernet 1,2,3 1 CEA, Institut de Biologie Structurale, Grenoble, France 2 CNRS, Institut de Biologie Structurale, Grenoble, France 3 Universite ´ Joseph Fourier, Institut de Biologie Structurale, Grenoble, France 4 Centre d’Inge ´ nierie des Prote ´ ines, Universite ´ de Lie ` ge, Institut de Chimie, Sart-Tilman Lie ` ge, Belgium 5 Biochemistry of Membranes, Bijvoet Center for Biomolecular Research and Institute of Biomembranes, Utrecht University, Utrecht, The Netherlands Keywords murein; penicillin-binding protein; peptidoglycan; screening Correspondence T. Vernet, Institut de Biologie Structurale, 41 rue Jules Horowitz, 38027 Grenoble, France Fax: +33 4 38 78 54 94 Tel: + 33 4 38 78 96 81 E-mail: thierry.vernet@ibs.fr (Received 31 May 2010, revised 22 July 2010, accepted 18 August 2010) doi:10.1111/j.1742-4658.2010.07817.x Cell wall biosynthesis is a key target for antibacterial drugs. The major constituent of the bacterial wall, peptidoglycan, is a netlike polymer responsible for the size and shape of the cell and for resisting osmotic pres- sure. It consists of glycan chains of repeating disaccharide units cross- linked through short peptide chains. Peptidoglycan assembly is catalyzed by the periplasmic domain of bifunctional class A penicillin-binding pro- teins. Cross-linking of the peptide chains is catalyzed by their transpepti- dase module, which can be inhibited by the most widely used antibiotics, the b-lactams. In contrast, no drug in clinical use inhibits the polymeriza- tion of the glycan chains, catalyzed by their glycosyltransferase module, although it is an obvious target. We report here the purification of the ectodomain of the class A penicillin-binding protein 1a from Thermoto- ga maritima (Tm-1a*), expressed recombinantly in Escherichia coli. A deter- gent screen showed that detergents with shorter aliphatic chains were better solubilizers. Cyclohexyl-hexyl-b-d-maltoside-purified Tm-1a* was found to be monomeric and to have improved thermal stability. A miniaturized, multiwell continuous fluorescence assay of the glycosyltransferase activity was used to screen for optimal reaction conditions. Tm-1a* was active as a glycosyltransferase, catalyzing the formation of glycan chains up to 16 disaccharide units long. Our results emphasize the importance of the detergent in preparing a stable monomeric ectodomain of a class A pen- icillin-binding protein. Our assay could be used to screen collections of compounds for inhibitors of peptidoglycan glycosyltransferases that could serve as the basis for the development of novel antibiotics. Abbreviations AEC, anion exchange chromatography; C 7 G, n-heptyl-b-D-glucopyranoside; CMC, critical micellar concentration; CYMAL-4, cyclohexyl-butyl- b- D-maltoside; CYMAL-5, cyclohexyl-pentyl-b-D-maltoside; CYMAL-6, cyclohexyl-hexyl-b-D-maltoside; GTase, glycosyltransferase; IMAC, immobilized metal-ion affinity chromatography; meso-A 2 pm, meso-diaminopimelic acid; Mtg, membrane-bound monofunctional glycosyltransferase; PBP, penicillin-binding protein; SEC, size exclusion chromatography; TEV, tobacco etch virus; T m, melting temperature; Tm-1a* and Tm-GT1a*, ectodomain of Thermotoga maritima penicillin-binding protein 1a and its GTase domain, respectively; TPase, transpeptidase; TSA, thermal shift assay. 4290 FEBS Journal 277 (2010) 4290–4298 ª 2010 The Authors Journal compilation ª 2010 FEBS Introduction Peptidoglycan biosynthesis is the major target for anti- bacterial drugs, such as the most widely used class of antibiotics, the b-lactams, or the last-resort glycopep- tides (vancomycin). Peptidoglycan is a strong, netlike polymer responsible for maintaining the shape of the bacterial cell and resisting intracellular osmotic pres- sure. It consists of glycan chains of repeating disaccha- ride units cross-linked through short peptide chains [1,2]. Peptidoglycan biosynthesis remains a major tar- get for new antibiotics, as it is both unique and essen- tial to bacteria. The final steps of peptidoglycan assembly are cata- lyzed by penicillin-binding proteins (PBPs), which are membrane-bound enzymes exposed to the extracyto- plasmic medium. For this reason, PBPs are the most interesting targets among the enzymes involved in the biosynthesis of the peptidoglycan, as they are easily accessible to drugs. In addition to a short N-terminal cytoplasmic region and a transmembrane segment, high molecular mass class A PBPs possess two enzymatic modules, a glyco- syltransferase (GTase) and a transpeptidase (TPase) [3]. The former is responsible for the elongation of the glycan strands from the precursor lipid II, a disaccha- ride (MurNAc-b-1,4-GlcNAc) pentapeptide anchored to the membrane by a C 55 undecaprenyl chain via a pyrophosphate. The latter module cross-links the pep- tide chains by a transpeptidation reaction [1,3,4]. The extracytoplasmic region of high molecular mass class B PBPs harbors an N-terminal module of unknown func- tion and a TPase module, but lacks the GTase module. Low molecular mass PBPs contain only a single TPase domain, but with endopeptidase or carboxypeptidase activity. Finally, membrane-bound monofunctional GTases (Mtgs) have also been identified in a few bac- teria [5–8]. Mtgs and GTase modules of class A PBPs belong to family 51 of glycosyltransferases from the CAZy database [9,10]. Penicillin and other b-lactam antibiotics are specific inhibitors of TPase activity. After six decades of b-lac- tam use, bacterial pathogens are now widely resistant to this most broadly used class of drug [11]. GTase activity is an obvious promising target for the development of new antibiotics, but despite many years of effort, no drug candidate is currently available for clinical use. Moenomycin was isolated from Strep- tomyces ghanaensis in 1968 [12], and is the best-charac- terized natural antibiotic that directly inhibits GTase activity. This phosphoglycolipid inhibits GTases at nanomolar concentrations, and has strong antibiotic activity in vitro. It is not used in human medicine, because of poor pharmacokinetics and oral bioavail- ability [13]. In addition, Gram-negative bacteria are not susceptible to moenomycin, as it cannot penetrate the outer membrane of these organisms [14]. Several crystal structures of GTase are now avail- able, with and without bound moenomycin or deriva- tives: the extracellular domain of PBP2 from Staphylococcus aureus [15,16] and of the Mtg from the same organism [17], the GTase module of PBP1a from Aquifex aeolicus [18,19], and the full-length PBP1b from Escherichia coli [20]. For a long time, the functional characterization of the GTase activity of class A PBPs or Mtgs has been limited to basic studies by the lack of availability of the lipid II substrate and analogs [21]. The development of chemically synthesized [22,23] and enzymatically syn- thesized [24,25] lipid II has allowed the development of various in vitro GTase activity assays. Significant mech- anistic insights have been obtained [8,26–33]. In partic- ular, the direction of the glycan chain elongation has been established, where a disaccharide unit is added to the reducing end of the growing chain, using appropri- ately blocked substrate analogs [31,34]. Also, different GTases produce chains with different length distribu- tions [8]. Screening methods for the identification of GTase inhibitors have been proposed [35–37]. As targets of potential new antibacterial drugs, it is desirable to study a greater variety of peptidoglycan GTases from various organisms. We report here the development of a multiwell assay, based on the method of Schwartz et al. [26], to screen for optimal reaction conditions that may differ from enzyme to enzyme. The method is exemplified by a study of the full ectodomain of PBP1a from Thermotoga maritima (Tm-1a*), extending our previous work on the GTase domain from this protein (Tm-GT1a*) [38]. T. maritima is a Gram-negative hyperth ermophilic bacterium is olated from hot sea floors [39]. The fact that proteins from hyperthermophilic organisms are often good candidates for crystallographic structural studies [40] prompted us to purify Tm-1a* in an active form. The properties of the purified enzyme have been determined and com- pared with those of other GTases. Results and Discussion The solubilization efficiency of Tm-1a* is related to the detergent alkyl chain length Tm-1a* has been defined on the basis of sequence alignments, and spans residues Glu34–Gly643. The J. Offant et al. Class A PBP1a from Thermotoga maritima FEBS Journal 277 (2010) 4290–4298 ª 2010 The Authors Journal compilation ª 2010 FEBS 4291 N-terminal region thus defined is likely to include two short b-strands (b1 and b2) that are now known to be part of a five-stranded b-sheet, with other strands being contributed by the TPase domain at the interdo- main junction [15,41]. An alignment with sequences of known crystal structure (S. aureus PBP2 and A. aeoli- cus PBP1a) is shown in Fig. S1. The sequence of T. maritima PBP1a with the predicted domains is shown in Fig. S2. The sequence identities of T. mariti- ma PBP1a with S. aureus PBP2 and A. aeolicus PBP1a are 21% and 25% respectively; for the GTase domain, the identities are 32% and 35%, respectively. Protein expression was performed in E. coli BL21(DE3) CodonPlus-RIL to favor translation of the numerous rare codons (15%) of the Tm-1a* sequence. Initial purification attempts were performed in the presence of 1% of the zwitterionic Chaps detergent, as performed previously for the isolated GTase domain [38]. This allowed solubilization of about 80% of the expressed Tm-1a*. However, immobilized metal-ion affinity chromatography (IMAC)-purified Tm-1a* in Chaps was partially aggregated, as shown by size exclusion chromatography (SEC) analysis (data not shown), as was observed with the isolated GTase domain [38]. To solve this problem, we compared the solubiliza- tion efficiency of 11 detergents. Tm-1a* was solubilized from intact bacteria by sonication in the various deter- gent solutions prior to purification with an Ni 2+ –nitrilo- triacetic acid Superflow column and SDS ⁄ PAGE analysis (Fig. 1). The maximal yield of purification was obtained with Chaps followed by n-heptyl-b-d-gluco- pyranoside (C 7 G) and cyclohexyl-hexyl-b-d-maltoside (CYMAL-4). Interestingly, the solubilization efficiency was related to the length of the carbon chain: the shorter the chain, the better the solubilization. For instance, the solubilization efficiency of Tm-1a* with CYMAL-4 was 40%, which was 70% higher than with cyclohexyl- pentyl-b-d-maltoside (CYMAL-5) and cyclohexyl-hexyl- b-d-maltoside (CYMAL-6), respectively. This pattern was also found for the three alkyl-glucoside and the two alkyl-maltoside detergents. CYMAL-4 was chosen for further comparison with Chaps. Purification of monomeric CYMAL-4-solubilized Tm-1a* Bacteria expressing Tm-1a* were lysed by sonication in the presence of 0.74% CYMAL-4. The first IMAC procedure delivered a protein over 80% pure (Fig. 2A). Most contaminants were eliminated by pre- parative SEC. The N-terminal His 6 -tag was cleaved by the His 6 -tagged tobacco etch virus (TEV) protease, and uncleaved Tm-1a*, the free tag and the protease were retained on the Ni 2+ –nitrilotriacetic acid column. A final anion exchange chromatography (AEC) step removed trace contaminants. CYMAL-4-purified Tm-1a* was eluted as a single symmetrical peak on an analytical SEC column (Fig. 2B), with an apparent molecular mass of 72 kDa (theoretical molecular mass of 69 700 Da). The higher apparent molecular mass is probably attributable to the presence of bound CYMAL-4 molecules. SDS ⁄ PAGE of the SEC peak showed a highly homogeneous protein (Fig. 2B, insert). From 1 L of E. coli culture, 0.5 mg of pure and homogeneous Tm-1a* monomer was obtained. The protein concentrated to 4.4 mgÆmL )1 (60 lm) remained monomeric over time at 4 °C and )80 °C. The TPase domain of Tm-1a* was functional for reac- tion with b-lactams, as its transpeptidase site could be labeled with fluorescent ampicillin (Fig. 2C). CYMAL-4-solubilized Tm-1a* displays elevated thermal stability The thermal shift assay (TSA) is an efficient and easy way to measure the thermal stability of proteins as compared with other biophysical methods, such as CD and microcalorimetry. The method is also amenable to Fig. 1. Detergent screening for the recovery of Tm-1a*. Lysates of E. coli cells overexpressing Tm-1a* were prepared in the presence of various detergents at concentrations twice their CMCs and loaded onto an Ni 2+ IMAC column. (A) SDS ⁄ PAGE of IMAC-eluted fractions, Coomassie-stained 72 kDa Tm-1a* band. (B) Histogram of the intensity of the Tm-1a* bands relative to that obtained with Chaps. Black or shades of gray denote detergents of the same chemical family. C 8 G, n-octyl-b-D-glucopyranoside; C 9 G, n-nonyl-b-D- glucopyranoside; C 10 M, n-decyl-b-D-maltopyranoside; C 12 M, n-dode- cyl-b- D-maltopyranoside; LDAO, lauryldimethylamine-oxide. Class A PBP1a from Thermotoga maritima J. Offant et al. 4292 FEBS Journal 277 (2010) 4290–4298 ª 2010 The Authors Journal compilation ª 2010 FEBS high throughput, and this allows the screening of com- pounds or conditions (pH, ionic strength, etc.) [42,43] that influence protein stability. The TSA is more diffi- cult to implement on detergent-solubilized proteins than without detergent. The fluorescence of the Sypro- Orange probe increases when the molecule is in contact with a nonpolar environment such as exposed hydro- phobic residues. The detection of a fluorescence increase upon denaturation of a protein will depend on the number and hydrophobicity index of exposed resi- dues during unfolding. The micellar phase of a deter- gent can have a similar effect, complicating the TSA results. With Chaps-purified Tm-1a*, two denaturation tran- sitions (T m ) were measured at 61 ± 1 and 81 ± 1 °C. These two transitions might correspond to independent unfolding of the GTase and TPase domains. The T m of Chaps-purified Tm-GT1a* was determined to be 62±1°C. It is therefore tempting to attribute the lower and higher T m values to the unfolding of the GTase and TPase domains, respectively. If this inter- pretation is correct, the TPase domain is more stable than the GTase domain by about 20 °C, and the TPase domain within Tm-1a* does not influence the stability of the GTase domain. The crystal structure of PBP2 from S. aureus [15] shows a narrow neck con- necting, with some flexibility, the GTase and TPase domains, with little contact between the domains. This structure is consistent with the absence of mutual sta- bilizing effect of the domains. The thermal stability of CYMAL-4-purified Tm-1a* also displays two thermal transitions at the higher values of 79 ± 1 and 89 ± 1 °C. These values are compatible with a thermophilic origin of the protein. CYMAL-4-purified Tm-1a* is more stable than Chaps-purified Tm-1a*. This observation, together with the fact that CYMAL-4-purified Tm-1a* does not aggregate, led us to select CYMAL-4 for purification and storage of Tm-1a*. Screening of reaction conditions of Tm-1a* in a multiwell plate format In vitro GTase activity is strongly influenced by the nature and concentration of additives, including deter- gent, dimethylsulfoxide, or metal [26]. We have minia- turized in a multiwell format the continuous fluorescence assay described by Schwartz (2002). This allows the parallel screening of numerous conditions while reducing the use of the limiting reagent lipid II. The assay takes advantage of the higher fluorescence of dansylated lipid II solubilized by detergent micelles than of the free dansylated pentapeptide disaccharide. GTases polymerize glycan chains by transferring the growing chain from its undecaprenyl pyrophosphate anchor onto an additional lipid II unit [31]. Following the hydrolytic action of the muramidase included in the reaction mix, assembled glycan chains are degraded into pentapeptide disaccharide units. As the fluorescence Fig. 2. Purification of CYMAL-4-solubilized Tm-1a*. (A) Coomassie- stained SDS ⁄ PAGE of the initial Ni 2+ IMAC purification: m, molecular mass markers; L, clarified lysate loaded; FT, flowthrough; 1–4, elution fraction with 250 m M imidazole. *The 72 kDa Tm-1a* band. (B) Analytical SEC (Superdex-200 HR 16 ⁄ 30; GE Healthcare) of Tm-1a* purified on Ni 2+ IMAC followed by SEC and AEC. (C) SDS ⁄ PAGE of Tm-1a* labeled with fluorescein–ampicillin (25 l M for 10 min at 37 °C): 1, after Coomassie blue staining; 2, fluorescence image taken with MOLECULAR IMAGER (Bio-Rad). J. Offant et al. Class A PBP1a from Thermotoga maritima FEBS Journal 277 (2010) 4290–4298 ª 2010 The Authors Journal compilation ª 2010 FEBS 4293 of the soluble dansylated pentapeptide disaccharide is lower than that of the dansylated lipid II, the GTase activity can be followed as a decrease of dansyl fluo- rescence. Our multiwell version of the assay is initiated by the addition of the GTase prior to measurement of the flu- orescence over the time course, and allows the parallel monitoring of up to 96 reactions. Visualization of the time courses allows easy comparison of the various reactions (Fig. 3) and an initial selection of the opti- mal conditions. GTase activities were compared by measuring the initial rate of fluorescence decrease (Fig. 3). The optimal conditions that produced the highest initial slope representing the fastest incorporation of lipid II into glycan chains contained 20% dimethylsulf- oxide and decyl-poly(ethylene glycol) at 2.5-fold or 5-fold the critical micellar concentration (CMC). Almost equally good were conditions with 5–10% dimethylsulfoxide and decyl-poly(ethylene glycol) at its CMC. The reaction was completely inhibited in the presence of 10 lm moenomycin (Fig. 3). The specific activity of Tm-1a* was determined in the presence of 2 lm [ 14 C]lipid II to be 2.9 ± 0.5 nmol of lipid II used min )1 ÆmgÆenzyme )1 , which is about 10- fold less than that of E. coli PBP1b (25 ± 5 nmol of lipid II used min )1 ÆmgÆenzyme )1 ). However, this might reflect the fact that the optimal temperature is proba- bly higher for a Tm-1a* that originates from a thermo- philic organism. Determination of the glycan chain size distribution of Tm-1a* products The SDS ⁄ PAGE assay developed by Barrett et al. [30] allows visualization of the length of the synthesized glycan chains, using radiolabeled lipid II as precursor. Tm-1a* was found, in our experimental conditions, to produce rather short chains, with the main product being about 10 disaccharide units long, without detect- able chains longer than 16 units (Fig. 4). PBP1b from E. coli was shown to produce longer chains under the same conditions (about 30 units long), indicating that Tm-1a* is less processive. Of the GTases tested so far, Tm-1a* appears to be the one that synthesizes the shortest chains [8]. Therefore, this enzyme could be useful for applications such as generation of glycan chains with defined size and composition that can be used as substrates for other enzymes (peptidoglycan hydrolases, sortases, or modifying enzymes, such as amidases and muramidases) or as a molecular stan- dards in SDS ⁄ PAGE analysis. It must be noted that the substrates used in this study were either l-lysine-containing lipid II, dansylat- ed on the lysine, or radiolabeled meso-diaminopimelic acid (meso-A 2 pm)–lipid II. It has been found recently that the T. maritima peptidoglycan contains unusual stem peptides, which include d-lysine in nonconven- tional arrangements, as well as unusual cross-links [44]. Future studies will investigate the substrate speci- ficity of the TPase reaction of Tm-1a*. Fig. 3. Screening of conditions for GTase activity of CYMAL-4-purified Tm-1a*. Tm-1a* (250 n M) was incubated in the presence of dansylated lipid II (10 l M)in 50 m M Hepes (pH 7.5), 200 mM NaCl, 10 m M CaCl 2 and muramidase (1 unit) with various combinations of decyl-poly(ethylene glycol) and dimethylsulfoxide concentrations. Fluorescence (excitation at 340 nm and emission at 520 nm) was monitored for 80 min. The best reaction conditions, highlighted in gray, show the greatest initial slopes. Complete inhibition with 10 l M moenomycin is presented below as a negative control. Class A PBP1a from Thermotoga maritima J. Offant et al. 4294 FEBS Journal 277 (2010) 4290–4298 ª 2010 The Authors Journal compilation ª 2010 FEBS GTases that participate in peptidoglycan assembly are obvious targets for the development of novel anti- biotics. To this end, it will be necessary to study multi- ple GTases from diverse microorganisms. We have reported here a preliminary characterization of Tm-1a*. It will be of interest to similarly prepare and characterize the full-length protein, to probe the influ- ence of the transmembrane segment. Also, T. maritima is hyperthermophilic, and future studies of T. maritima PBP1a should investigate how the kinetics and proces- sivity vary with temperature. Our results emphasize the importance of the detergent in preparing a stable monomeric ectodomain of a class A PBP. Optimal reaction conditions for the polymerization of glycan chains by Tm-1a* were found with the use of a multi- well assay. This assay could be used to screen collec- tions of compounds for inhibitors of peptidoglycan glycosyltransferases that could serve as a basis for the development of novel antibiotics. Experimental procedures Gene cloning and protein expression The fragment encoding the extracellular region of PBP1a (accession number AAD35967.1) was PCR-amplified from T. maritima MSB8 genomic DNA with the forward primer 5¢- GAAAATCTGTATTTTCAGGGCGAGGA GAAACT TGTGCCGACC-3¢ and the reverse primer 5¢- TCACCAT CCAATTGATTAACCTCCTTCCATCAAAAACTTTTT CCAGATTTC-3¢. The fragment coding for the isolated GTase was PCR-amplified with the same forward primer and the reverse primer 5¢- TCACCATCCAATTGATTA TTCCGCAGAGTAATTCTCGTATTCCTG-3¢ (sequences required for ligation-independent cloning are underlined). Purified PCR products were introduced into pLIM01 [45] by the ligation-independent cloning method [46], to produce the pTm-1a* and pTm-GT1a* expression vectors encoding Tm-1a* (Glu34–Gly643) and the corresponding GTase domain (Glu34–Thr244) with an N-terminal His 6 -tag fol- lowed by the TEV protease cleavage site. E. coli BL21-CodonPlus(DE3)-RIL cells (Stratagene, Cedar Creek, TX, USA) were transformed with pTm-1a* and pTm-GT1a*, and overnight precultures were diluted 50-fold into fresh LB medium supplemented with antibiot- ics. After growth at 37 °C to an attenuance of 0.8 at 600 nm, protein expression was induced with 0.5 mm iso- propyl-thio-b-d-galactoside, and incubation was continued overnight at 15 °C. Detergent screening for Tm-1a* solubility and purification by Ni 2+ IMAC Detergent screening was performed from 10 mL aliquots of induced bacterial cultures at an attenuance of 4 (600 nm), spun, and resuspended in 1.2 mL of lysis buffer (25 mm Hepes, pH 7.5, 500 mm NaCl) containing the appropriate detergents at a concentration twice their CMC. The 11 fol- lowing detergents were used: Chaps, 1%; lauryldimethyl- amine-oxide, 0.0064%; Triton X-100, 0.03%; C 7 G, 3.8%; n-octyl-b-d-glucopyranoside, 1.37%; n-nonyl-b-d-glucopyr- anoside, 0.4%; n-decyl-b-d-maltopyranoside, 0.174%; n- dodecyl-b-d-maltopyranoside, 0.0174%; CYMAL-4, 0.74%; CYMAL-5, 0.24%; and CYMAL-6, 0.056% (all purchased from Anatrace, Maumee, OH, USA). Samples of 400 lLof the resuspension were placed in 1.1 mL MicroTubes (National Scientific Supply, San Rafael, CA, USA) and simultaneously lysed by sonication, using a microplate horn coupled with an S3000 generator (Misonix, Newtown, CT, USA). The horn was filled with ice-cold water, and cells were lysed by six pulses of 1 min with 2 min intervals (power level set to 7). Lysis supernatants (20 000 g, 30 min, 4 °C) were loaded onto 50 lL of Ni Sepharose High Performance resin pre- packed in a 96-well format His MultiTrap HP (GE Health- care, Little Chalfont, UK). After a wash with 1 mL of 60 mm imidazole in the same buffers, Tm-1a* was eluted with 100 lL of 250 mm, imidazole and eluates were analyzed by SDS ⁄ PAGE. Detergents were evaluated by quantifying the Coomassie-stained bands corresponding to Tm-1a* with a BioRad Geldoc 2000 (BioRad, Hercules, CA, USA). Fig. 4. Glycan chain size distribution of Tm-1a* products. An auto- radiogram of SDS ⁄ PAGE-separated glycan chains is shown. GTase reaction products show that the maximal length of the glycan chains made by Tm-1a* is about 16 disaccharide units (n), with the main product being 10 units long. As a control, glycan chains poly- merized in similar conditions by E. coli PBP1b are shown. J. Offant et al. Class A PBP1a from Thermotoga maritima FEBS Journal 277 (2010) 4290–4298 ª 2010 The Authors Journal compilation ª 2010 FEBS 4295 Purification of Tm-GT1a* and Tm-1a* For the purification of CYMAL-4-solubilized Tm-1a*, a 1 L culture at a final attenuance at 600 nm of 4 was spun and resuspended in 90 mL of 25 mm Hepes (pH 7.5), 500 mm NaCl, 0.74% CYMAL-4 and a pill of Complete EDTA-free protease inhibitors (Roche, Basel, Switzerland). Cells were lysed using a 1 ⁄ 2 inch horn coupled with an S3000 generator (Misonix), with a total sonication time of 4 min (2 s pulses at 10 s intervals) and a power level of 5. After centrifugation (20 000 g, 30 min, 4 °C), the superna- tant was loaded onto 5 mL of Ni 2+ –nitrilotriacetic acid Su- perflow (Qiagen) pre-equilibrated with buffer A (25 mm Hepes, pH 7.5, 500 mm NaCl, 0.37% CYMAL-4). Tm-1a* was eluted with two steps of 60 mm and 250 mm imidazole in buffer A. Further purification steps consisted of SEC on a Superdex-200 preparation grade column (GE Healthcare) equilibrated with buffer A, IMAC following cleavage of the His 6 -tag by the TEV protease (to remove noncleaved Tm-1a* and the protease), and AEC on a 1 mL Resour- ce Q column (GE Healthcare) eluted with a 0–1 m linear gradient in buffer A to remove trace contaminants. CYMAL-4-purified Tm-1a* (0.2 mgÆmL )1 ) was frozen in liquid nitrogen, and stored at )80 °C. A similar protocol was adopted for the purification of Chaps-solubilized Tm-GT1a* and Tm-1a*, but omitting the AEC. Cell lysis was performed with 1% Chaps, and all sub- sequent buffers contained 0.7% Chaps. Final concentra- tions were 0.5 mgÆmL )1 for Tm-GT1a* and 0.4 mgÆmL )1 for Tm-1a*. Purified enzymes were stored at )80 °C. TSA Experiments were carried out using the IQ5 96-well format real-time PCR instrument (BioRad). Briefly, CYMAL-4- purified Tm-1a* (60 lm) and Chaps-purified Tm-GT1a* or Tm-1a* at concentrations ranging from 15 to 40 lm were mixed with 2 lL of 100-fold water-diluted 5000X SYPRO Orange (Molecular Probes, Eugene, Oregon, USA). Sam- ples were heat-denatured from 20 to 100 °C at a rate of 1 °CÆmin )1 , and unfolding was monitored by measuring changes in the fluorescence of SYPRO Orange. The T m val- ues were identified as the maxima of the first derivatives of the fluorescence versus temperature curves. Detergent-con- taining buffers were used as blanks, and their SYPRO Orange fluorescence curves were subtracted from the sam- ple curves. Screening of conditions for the GTase activity The assay developed by Schwartz [26] was adapted to a 96-well format with a medium binding black 96-well microplate (Greiner Bio One, ref. 655076; Frieckenhausen, Germany) in a FLUOstar OPTIMA Microplate reader (BMG Labtech, Offenburg, Germany). The reaction mix (50 lL) included lysine-dansylated lipid II (10 lm) [24] in 50 mm Hepes (pH 7.5), 200 mm NaCl, 10 mm CaCl 2 , 1 unit of N-acetylmuramidase from Streptococcus globisporus (Cal- biochem, Darmstadt, Germany), and various combinations of decyl-poly(ethylene glycol) (0, 1, 2.5, 5, 10, 20 and 40 · CMC) and dimethylsulfoxide (0%, 5%, 10% and 20%) concentrations. Time courses at 30 °C were initiated with the addition of Tm-1a* (250 nm) and followed for 80 min with excitation at 340 nm and emission recorded at 520 nm. Determination of specific GTase activity of Tm-1a* The GTase assay was carried out in triplicate, using the [ 14 C]meso-A 2 pm–lipid II (2 lm; 0.126 lCiÆnmol )1 ) as sub- strate in the following reaction mix: 50 mm Hepes (pH 7.5), 10 mm CaCl 2 , 200 mm NaCl, 0.2% decyl-poly(ethylene gly- col) and 20% dimethylsulfoxide. The reaction was started with addition of 150 nm Tm-1a* and stopped with 12.5 lm moenomycin (Flavomycin; Hoechst, Frankfurt, Germany). The reaction products were separated by TLC in propanol- 2 ⁄ ammonia ⁄ water (6 : 3 : 1), and analyzed with a Molecu- lar Imager (BioRad). Determination of the glycan chain size distribution of Tm-1a* products The reaction conditions were: 50 mm Hepes (pH 7.5), 10 mm CaCl 2 , 200 mm NaCl, penicillin G 1000 UÆmL )1 (500 lgÆmL )1 ), 0.2% decyl-poly(ethylene glycol), 20% dimethylsulfoxide, 4 lm [ 14 C]meso-A 2 pm-lipid II (0.126 lCiÆnmol )1 ), and 2 lm Tm-1a* or 0.5 lm E. coli PBP1b. Samples were collected at various times and analyzed by SDS ⁄ PAGE 9%T 2.6%C (size 20 cm · 20 cm · 1 mm), with an anode buffer of 0.1 m Tris (pH 8.3) and a cathode buffer of 0.1 m Tricine (pH 8.45) and 0.1% SDS [30]. Acknowledgements This work was partly funded by the FP6 EUR-INTA- FAR LSHM-CT-2004-512138 project and the ANR grant PneumoPG ANR-08-BLAN-0201. We thank B. Gallet and M. Noirclerc-Savoye, from the IBS plat- form of the Partnership for Structural Biology and the Institut de Biologie Structurale in Grenoble (PSB ⁄ IBS), for their expertise with the fluorimeter. References 1 van Heijenoort J (2001) Formation of the glycan chains in the synthesis of bacterial peptidoglycan. Glycobiology 11, 25R–36R. 2 Vollmer W, Blanot D & de Pedro MA (2008) Peptido- glycan structure and architecture. FEMS Microbiol Rev 32, 149–167. Class A PBP1a from Thermotoga maritima J. Offant et al. 4296 FEBS Journal 277 (2010) 4290–4298 ª 2010 The Authors Journal compilation ª 2010 FEBS 3 Sauvage E, Kerff F, Terrak M, Ayala JA & Charlier P (2008) The penicillin-binding proteins: structure and role in peptidoglycan biosynthesis. FEMS Microbiol Rev 32, 234–258. 4 Lovering AL, Gretes M & Strynadka NC (2008) Struc- tural details of the glycosyltransferase step of peptido- glycan assembly. Curr Opin Struct Biol 18, 534–543. 5 Derouaux A, Wolf B, Fraipont C, Breukink E, Nguyen-Disteche M & Terrak M (2008) The monofunc- tional glycosyltransferase of Escherichia coli localizes to the cell division site and interacts with penicillin-binding protein 3, FtsW, and FtsN. J Bacteriol 190, 1831–1834. 6 Hara H & Suzuki H (1984) A novel glycan polymerase that synthesizes uncross-linked peptidoglycan in Escher- ichia coli. FEBS Lett 168, 155–160. 7 Terrak M & Nguyen-Disteche M (2006) Kinetic charac- terization of the monofunctional glycosyltransferase from Staphylococcus aureus. J Bacteriol 188, 2528–2532. 8 Wang TS, Manning SA, Walker S & Kahne D (2008) Isolated peptidoglycan glycosyltransferases from differ- ent organisms produce different glycan chain lengths. J Am Chem Soc 130, 14068–14069. 9 Coutinho PM, Deleury E, Davies GJ & Henrissat B (2003) An evolving hierarchical family classification for glycosyltransferases. J Mol Biol 328, 307–317. 10 Henrissat B, Sulzenbacher G & Bourne Y (2008) Glyco- syltransferases, glycoside hydrolases: surprise, surprise! Curr Opin Struct Biol 18, 527–533. 11 Zapun A, Contreras-Martel C & Vernet T (2008) Peni- cillin-binding proteins and beta-lactam resistance. FEMS Microbiol Rev 32, 361–385. 12 Wallhausser KH, Nesemann G, Prave P & Steigler A (1965) Moenomycin, a new antibiotic. I. Fermentation and isolation. Antimicrob Agents Chemother (Bethesda) 5, 734–736. 13 Goldman RC & Gange D (2000) Inhibition of transgly- cosylation involved in bacterial peptidoglycan synthesis. Curr Med Chem 7, 801–820. 14 El-Abadla N, Lampilas M, Hennig L, Findeisen M, Welzel P, Muller D, Markus A & van Heijenoort J (1999) Moenomycin A: the role of the methyl group in the moenuronamide unit and a general discussion of structure–activity relationships. Tetrahedron 55, 699– 722. 15 Lovering AL, de Castro LH, Lim D & Strynadka NC (2007) Structural insight into the transglycosylation step of bacterial cell-wall biosynthesis. Science 315, 1402– 1405. 16 Lovering AL, De Castro L & Strynadka NC (2008) Identification of dynamic structural motifs involved in peptidoglycan glycosyltransfer. J Mol Biol 383, 167– 177. 17 Heaslet H, Shaw B, Mistry A & Miller AA (2009) Characterization of the active site of S. aureus monofunctional glycosyltransferase (Mtg) by site-directed mutation and structural analysis of the protein complexed with moenomycin. J Struct Biol 167, 129–135. 18 Yuan Y, Barrett D, Zhang Y, Kahne D, Sliz P & Walker S (2007) Crystal structure of a peptidoglycan glycosyltransferase suggests a model for processive gly- can chain synthesis. Proc Natl Acad Sci USA 104, 5348–5353. 19 Yuan Y, Fuse S, Ostash B, Sliz P, Kahne D & Walker S (2008) Structural analysis of the contacts anchoring moenomycin to peptidoglycan glycosyltransferases and implications for antibiotic design. ACS Chem Biol 3, 429–436. 20 Sung MT, Lai YT, Huang CY, Chou LY, Shih HW, Cheng WC, Wong CH & Ma C (2009) Crystal structure of the membrane-bound bifunctional transglycosylase PBP1b from Escherichia coli. Proc Natl Acad Sci USA 106, 8824–8829. 21 Suzuki H, van Heijenoort Y, Tamura T, Mizoguchi J, Hirota Y & van Heijenoort J (1980) In vitro peptidogly- can polymerization catalysed by penicillin binding protein 1b of Escherichia coli K-12. FEBS Lett 110, 245–249. 22 Schwartz B, Markwalder JA & Wang Y (2001) Lipid II: total synthesis of the bacterial cell wall precursor and utilization as a substrate for glycosyltransfer and transpeptidation by penicillin binding protein (PBP) 1b of Escherichia coli. J Am Chem Soc 123, 11638–11643. 23 VanNieuwenhze MS, Mauldin SC, Zia-Ebrahimi M, Winger BE, Hornback WJ, Saha SL, Aikins JA & Blaszczak LC (2002) The first total synthesis of lipid II: the final monomeric intermediate in bacterial cell wall biosynthesis. J Am Chem Soc 124, 3656–3660. 24 Breukink E, van Heusden HE, Vollmerhaus PJ, Swiezewska E, Brunner L, Walker S, Heck AJ & de Kruijff B (2003) Lipid II is an intrinsic component of the pore induced by nisin in bacterial membranes. J Biol Chem 278, 19898–19903. 25 Ye XY, Lo MC, Brunner L, Walker D, Kahne D & Walker S (2001) Better substrates for bacterial transglycosylases. J Am Chem Soc 123, 3155–3156. 26 Schwartz B, Markwalder JA, Seitz SP, Wang Y & Stein RL (2002) A kinetic characterization of the glycosyl- transferase activity of Eschericia coli PBP1b and development of a continuous fluorescence assay. Biochemistry 41, 12552–12561. 27 Chen L, Walker D, Sun B, Hu Y, Walker S & Kahne D (2003) Vancomycin analogues active against vanA- resistant strains inhibit bacterial transglycosylase with- out binding substrate. Proc Natl Acad Sci USA 100, 5658–5663. 28 Di Guilmi AM, Dessen A, Dideberg O & Vernet T (2003) Functional characterization of penicillin-binding protein 1b from Streptococcus pneumoniae. J Bacteriol 185, 1650–1658. J. Offant et al. Class A PBP1a from Thermotoga maritima FEBS Journal 277 (2010) 4290–4298 ª 2010 The Authors Journal compilation ª 2010 FEBS 4297 29 Barrett DS, Chen L, Litterman NK & Walker S (2004) Expression and characterization of the isolated glycosyl- transferase module of Escherichia coli PBP1b. Biochem- istry 43, 12375–12381. 30 Barrett D, Wang TS, Yuan Y, Zhang Y, Kahne D & Walker S (2007) Analysis of glycan polymers produced by peptidoglycan glycosyltransferases. J Biol Chem 282, 31964–31971. 31 Perlstein DL, Zhang Y, Wang TS, Kahne DE & Walker S (2007) The direction of glycan chain elongation by peptidoglycan glycosyltransferases. J Am Chem Soc 129, 12674–12675. 32 Terrak M, Sauvage E, Derouaux A, Dehareng D, Bouhss A, Breukink E, Jeanjean S & Nguyen-Disteche M (2008) Importance of the conserved residues in the peptidoglycan glycosyltransferase module of the class A penicillin-binding protein 1b of Escherichia coli. J Biol Chem 283, 28464–28470. 33 Perlstein DL, Wang TS, Doud EH, Kahne D & Walker S (2010) The role of the substrate lipid in processive glycan polymerization by the peptidoglycan glyco- syltransferases. J Am Chem Soc 132, 48–49. 34 Fraipont C, Sapunaric F, Zervosen A, Auger G, Devre- ese B, Lioux T, Blanot D, Mengin-Lecreulx D, Herd- ewijn P, Van Beeumen J et al. (2006) Glycosyl transferase activity of the Escherichia coli penicillin- binding protein 1b: specificity profile for the substrate. Biochemistry 45, 4007–4013. 35 Vollmer W & Holtje JV (2000) A simple screen for murein transglycosylase inhibitors. Antimicrob Agents Chemother 44, 1181–1185. 36 Stachyra T, Michoux F, Michel JM, Biton J, Dini C, Dumas J, Ferrari P & Le Beller D (2002) Development of enzymatic assays based on fluorescence for detection of inhibitors of MraY, MurG and PBP1b in drug screening. Abstr Intersci Conf Antimicrob Agents Chemother 42, F-721. 37 Cheng TJ, Sung MT, Liao HY, Chang YF, Chen CW, Huang CY, Chou LY, Wu YD, Chen YH, Cheng YS et al. (2008) Domain requirement of moenomycin bind- ing to bifunctional transglycosylases and development of high-throughput discovery of antibiotics. Proc Natl Acad Sci USA 105, 431–436. 38 Offant J, Michoux F, Dermiaux A, Biton J & Bourne Y (2006) Functional characterization of the glycosyl- transferase domain of penicillin-binding protein 1a from Thermotoga maritima. Biochim Biophys Acta 1764, 1036–1042. 39 Huber R, Langworthy TA, Ko ¨ nig H, Thomm M, Woese CR, Sleytr UB & Stetter KO (1986) Thermotoga maritima sp. nov. represents a new genus of unique extremely thermophilic eubacteria growing up to 90°C. Arch Microbiol 144, 324–333. 40 Rees DC (2001) Crystallographic analysis of hyper- thermophilic proteins. Methods Enzymol 334, 423–437. 41 Macheboeuf P, Di Guilmi AM, Job V, Vernet T, Dideberg O & Dessen A (2005) Active site restructuring regulates ligand recognition in class A penicillin-binding proteins. Proc Natl Acad Sci USA 102, 577–582. 42 Pantoliano MW, Petrella EC, Kwasnoski JD, Lobanov VS, Myslik J, Graf E, Carver T, Asel E, Springer BA, Lane P et al. (2001) High-density miniaturized thermal shift assays as a general strategy for drug discovery. J Biomol Screen 6, 429–440. 43 Yeh AP, McMillan A & Stowell MH (2006) Rapid and simple protein-stability screens: application to membrane proteins. Acta Crystallogr D Biol Crystallogr 62, 451–457. 44 Boniface A, Parquet C, Arthur M, Mengin-Lecreulx D & Blanot D (2009) The elucidation of the structure of Thermotoga maritima peptidoglycan reveals two novel types of cross-link. J Biol Chem 284, 21856–21862. 45 Attali C, Frolet C, Durmort C, Offant J, Vernet T & Di Guilmi AM (2008) Streptococcus pneumoniae cho- line-binding protein E interaction with plasmino- gen ⁄ plasmin stimulates migration across the extracellular matrix. Infect Immun 76, 466–476. 46 Geertsma ER & Poolman B (2007) High-throughput cloning and expression in recalcitrant bacteria. Nat Methods 4, 705–707. Supporting information The following supplementary material is available: Fig. S1. Sequence alignment of class A PBPs. Fig. S2. Sequence of T. maritima PBP1a with pre- dicted domains. This supplementary material can be found in the online version of this article. Please note: As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. Class A PBP1a from Thermotoga maritima J. Offant et al. 4298 FEBS Journal 277 (2010) 4290–4298 ª 2010 The Authors Journal compilation ª 2010 FEBS . Optimization of conditions for the glycosyltransferase activity of penicillin-binding protein 1a from Thermotoga maritima Julien Offant 1,2,3 , Mohammed Terrak 4 ,. exemplified by a study of the full ectodomain of PBP1a from Thermotoga maritima (Tm -1a* ), extending our previous work on the GTase domain from this protein (Tm-GT1a*) [38]. T. maritima is a Gram-negative. investigate the substrate speci- ficity of the TPase reaction of Tm -1a* . Fig. 3. Screening of conditions for GTase activity of CYMAL-4-purified Tm -1a* . Tm -1a* (250 n M) was incubated in the presence of