Skip to main content
NCBI home page
Acta Crystallographica Section D: Structural Biology logoLink to Acta Crystallographica Section D: Structural Biology
. 2020 Jan 1;76(Pt 1):28–40. doi: 10.1107/S2059798319015055

Structural insights into the role of the N-terminus in the activation and function of extracellular serine protease from Staphylococcus epidermidis

Kartik Manne a, Sthanam V L Narayana a,*
PMCID: PMC6939437  PMID: 31909741

Extracellular serine protease (Esp) is a glutamyl endopeptidase from Staphylococcus epidermidis that plays a key role in inhibiting the growth and formation of Staphylococcus aureus biofilms. Here, crystal structures of the Esp zymogen and its N-terminal locked variants are presented, and details are given of the role of the unusually long N-terminus in the activation and function of Esp.

Keywords: extracellular serine protease, glutamyl endopeptidase, zymogen, crystal structure, Staphylococcus epidermidis

Abstract

Extracellular serine protease (Esp) from Staphylococcus epidermidis is a glutamyl endopeptidase that inhibits the growth and formation of S. aureus biofilms. Previously, crystal structures of the matured and active Esp have been determined. Interestingly, many of the staphylococcal glutamyl endopeptidase zymogens, including V8 from Staphylococcus aureus and Esp from S. epidermidis, contain unusually long pro-peptide segments; however, their function is not known. With the aim of elucidating the function of these pro-peptide segments, crystal structures of the Esp zymogen (Pro-Esp) and its variants were determined. It was observed that the N-terminus of the Pro-Esp crystal structure is flexible and is not associated with the main body of the enzyme, unlike in the known active Esp structure. In addition, the loops that border the putative substrate-binding pocket of Pro-Esp are flexible and disordered; the structural components that are responsible for enzyme specificity and efficiency in serine proteases are disordered in Pro-Esp. However, the N-terminal locked Pro-Esp variants exhibit a rigid substrate-binding pocket similar to the active Esp structure and regain activity. These structural studies highlight the role of the N-terminus in stabilizing the structural components responsible for the activity and specificity of staphylococcal glutamyl endopeptidases.

1. Introduction  

Glutamyl endopeptidases (GSEs) hydrolyze the peptide bonds formed by the α-carboxyl groups of Glu and Asp residues; however, Glu–X bonds are cleaved more efficiently than Asp–X bonds, where X can be any amino acid. Bacterial GSEs belong to the serine protease family and are related to chymotrypsin-like serine proteases. They have been isolated from a number of bacterial species such as Staphylococcus aureus (Drapeau et al., 1972 ▸), Staphylococcus epidermidis (Dubin et al., 2001 ▸; Ohara-Nemoto et al., 2002 ▸), Bacillus licheniformis (Kakudo et al., 1992 ▸; Svendsen & Breddam, 1992 ▸), Entero­coccus faecalis (Kawalec et al., 2005 ▸) and Streptomyces griseus (Suzuki et al., 1994 ▸; Svendsen et al., 1991 ▸). The V8 protease from S. aureus and SprE from E. faecalis have been identified as important virulence factors (Calander et al., 2008 ▸; Kawalec et al., 2005 ▸; Ohara-Nemoto et al., 2002 ▸; Qin et al., 2000 ▸) along with many other members of this enzyme family. The V8 protease is essential for the activation of the cysteine protease SspB, a virulence factor of S. aureus (Massimi et al., 2002 ▸). V8 also exhibits 25% sequence identity to toxins A and B (ETA and ETB). It has been implicated in staphylococcal scalded skin syndrome (Bukowski et al., 2010 ▸; Dancer et al., 1990 ▸; Lee et al., 1987 ▸; O’Toole & Foster, 1987 ▸). Interestingly, S. epidermidis GSE, also known as extracellular serine protease (Esp), has 69% sequence identity to V8 and has been shown to inhibit biofilm formation and nasal coloniz­ation by S. aureus (Iwase et al., 2010 ▸).

Many crystal structures of mature GSEs are available. The crystal structures of ETA, ETB (Cavarelli et al., 1997 ▸) and V8 from S. aureus (Prasad et al., 2004 ▸), BIEP from B. intermedius (Meijers et al., 2004 ▸) and Esp from S. epidermidis (Chen et al., 2013 ▸) reveal that they exhibit a unique substrate-binding pocket; their N-terminus is implicated in the specificity of the enzyme towards negatively charged (Glu and Asp) residues of the substrate.

The activation pathway of chymotrypsin-like serine proteases has been extensively studied, in which the inactive enzyme is generally converted to its mature form by losing the N-terminal pro-peptide segment (Neurath & Walsh, 1976 ▸). Interestingly, the pro-peptide segments of bacterial GSEs are unusually long, at more than 60 residues, and could function as inhibitors of the proteolytic domain (Gasanov et al., 2008 ▸; Nickerson et al., 2007 ▸; Ohara-Nemoto et al., 2008 ▸), similar to as seen in human cysteine proteases, where the pro-peptide acts as an intramolecular chaperone (Wiederanders et al., 2003 ▸) and as an inhibitor keeping the protease inactive (Khan & James, 1998 ▸). Deletion and mutational analyses suggested a similar chaperone-like role for the extra-long pro-peptide of S. aureus V8 (Ohara-Nemoto et al., 2008 ▸). Meanwhile, the GSE from S. griseus has been suggested to be activated auto-catalytically by a cleavage between the N-terminal pro-peptide and the mature domain (Rouf et al., 2012 ▸; Stennicke et al., 1996 ▸; Svendsen et al., 1991 ▸); however, the activation of the GSEs from S. aureus (Drapeau, 1978 ▸; Nickerson et al., 2007 ▸; Ono et al., 2008 ▸; Rouf et al., 2012 ▸), B. intermedius (Gasanov et al., 2008 ▸; Park et al., 2004 ▸; Trachuk et al., 2005 ▸) and Entero­coccus (Kawalec et al., 2005 ▸) involves both autocatalytic truncation of the pro-peptide further towards the N-terminus and a final heterocatalytic cleavage of the remaining part of the pro-peptide by other proteases. Modification of the heterocatalytic cleavage site of B. intermedius GSE has been shown to affect the production of active GSE by B. subtilis cells (Velishaeva et al., 2008 ▸), suggesting that the cellular production of active GSE largely relies on the pro-peptide structure and its cleavage sites.

Although Esp from S. epidermidis is known to eradicate even pre-existing S. aureus biofilms (Iwase et al., 2010 ▸), the highly homologous S. aureus V8 exhibits no such inhibitory effect on S. epidermidis biofilms or colonization. In the light of the high morbidity and mortality associated with multi-drug-resistant strains of S. aureus (MRSA; Klevens et al., 2007 ▸), there is an urgent need for the identification of a new class of inhibitors that do not elicit bacterial resistance. Understanding the structural correlation of Esp maturation and its functional role in preventing S. aureus biofilm formation and growth may provide such a lead.

Our laboratory has previously determined the three-dimensional crystal structure of active Esp by molecular-replacement methods using S. aureus V8 (PDB entry 1qy6; Prasad et al., 2004 ▸) as a starting model (Chen et al., 2013 ▸). Similar to the eukaryotic serine proteases of the chymotrypsin family, Esp exhibits a β-barrel fold with two domains (each consisting of six antiparallel β-strands) and an α-helix at the C-terminus (Perona et al., 1995 ▸; Perona & Craik, 1995 ▸; Rühlmann et al., 1973 ▸). The essential functional features of the putative catalytic triad residues (Ser235, Asp159 and His117), the oxyanion hole and the S1 pocket (substrate-binding region) are present. However, distinct from the eukaryotic serine proteases, the conserved intra-domain disulfide bonds responsible for structural rigidity are absent in Esp and V8 (Chen et al., 2013 ▸; Perona & Craik, 1995 ▸; Prasad et al., 2004 ▸). Nemoto and coworkers identified the catalytic Ser237 (Ser235 in Esp) and N-terminal Val69 (Val67 in Esp) of V8 as essential for substrate cleavage (Ohara-Nemoto et al., 2008 ▸). N-terminal truncation to Ile70 (Ile68 in Esp) renders V8 inactive and mutants with an altered N-terminus, replacing Val69 with conserved substitutions, were also inactive, indicating a strict requirement for the N-terminal Val69 residue (Ono et al., 2010 ▸). Moreover, we investigated the mechanism of Esp-mediated S. aureus biofilm inhibition and identified autolysin (Atl) as the major target of Esp during biofilm disassembly (Chen et al., 2013 ▸). Although the biofilm formed by an S. aureus atl mutant was dramatically reduced compared with that of the wild-type strain, the residual biofilm formed by the atl mutant was sensitive to Esp, suggesting that Esp also degrades other biofilm formation-associated proteins such as coagulase (Coa), protein A (SpA) and clumping factor A (ClfA) (Chen et al., 2013 ▸).

The disposition of the N-terminus, which associates with the respective S1 pockets more intimately than in other active serine proteases (Prasad et al., 2004 ▸; Chen et al., 2013 ▸), is identical in V8 and Esp. The amino group of the N-terminal Val67 of Esp (Val69 in V8) is suitably positioned to act as an acceptor of the negative charge of the side chain of the P1 residue Glu. Similarly, the His250 (His252 in V8), Tyr226 (Tyr228 in V8) and Thr230 (Thr232 in V8) residues are highly conserved across GSEs. Modification of His250, which makes hydrogen bonds to the side chains of the conserved Tyr226 (Tyr228 in V8) and Thr230 (Thr232 in V8), significantly affects the catalytic activity (k cat decreases more than 600-fold) but has a low impact on substrate binding (K m increases approximately fivefold; Demidyuk et al., 2004 ▸), suggesting that it may not play a crucial role in holding the negatively charged substrate residue, but it seems to be significant for accurate positioning of the scissile bond with respect to the nucleophile Ser235. Modification of the N-terminal Val69 in V8 has been shown to substantially reduce the hydrolytic efficiency of Glu-carrying substrates, indicating its importance. Furthermore, its α-amino group probably makes a very significant contribution to the recognition of the charged substrate (Nemoto et al., 2008 ▸; Ono et al., 2010 ▸).

As mentioned earlier, the pro-enzymes of Esp and V8, and all other known bacterial glutamyl endopeptidases (Kawalec et al., 2005 ▸; Liu et al., 2016 ▸), are equipped with unusually long 68–69-residue N-termini, similar to some cysteine proteases, that need to be clipped for enzyme activation. However, their structural role in the transition from zymogen to active enzyme and the structural changes that the enzyme undergoes owing to the loss of pro-peptide are unclear; the pro-peptide could either be flexible, free-floating as observed in the zymogens of chymotrypsin-like serine proteases, or rigid and intimately associated with the protease body and act as an inhibitor, as seen in cysteine protease zymogens. In this paper, we determine crystal structures of Pro-Esp and its variants to localize the pro-peptide in relation to the main body of the enzyme and to understand the structure–function relationships of such an extended N-terminus and specifically the residues responsible for Pro-Esp inactivity and activation.

2. Materials and methods  

2.1. Cloning and mutations  

Full-length Pro-Esp was cloned in the pET-28b vector, as described previously (Chen et al., 2013 ▸). Other constructs of Pro-Esp were cloned in pET-28b, pMCSG7 and pET-21b plasmids using XhoI and NdeI (New England BioLabs) restriction sites with a His6 tag at the N-terminus and/or the C-terminus. Mutations were made using the Change-IT Multiple Mutation Site-Directed Mutagenesis Kit (USB Affymetrix; Chen & Ruffner, 1998 ▸). The clones obtained and the mutations performed were confirmed by DNA sequencing (Table 1 ▸).

Table 1. Recombinant production information.

Construct Plasmid N- and C-terminal tags Mutations
Full-length Esp, Ala27–Gln282 pET-28b; kanamycin-resistant His6 tag at the N-terminus NA
Full-length EspS235A, Val67–Gln282 pET-28b; kanamycin-resistant His6 tag at the N-terminus Ser235 to Ala
EspΔ55, Ile56–Gln282 pMCSG7 ligation-independent cloning; ampicillin-resistant His6 tag at the N-terminus followed by TEV cleavage site NA
EspΔ55S235A, Ile56–Gln282 pMCSG7 ligation-independent cloning; ampicillin-resistant His6 tag at the N-terminus followed by TEV cleavage site Ser235 to Ala
EspΔ55S66V, Ile56–Gln282 pMCSG7 ligation-independent cloning; ampicillin-resistant His6 tag at the N-terminus followed by TEV cleavage site Ser66 to Val
EspΔ55CysM1, Ile56–Gln282 pET-21b; ampicillin-resistant His6 tags at the N- and C-termini with TEV and thrombin cleavage sites, respectively Val67 to Cys; Thr230 to Cys
EspΔ65CysM1, Ser66–Gln282 pET-21b; ampicillin-resistant His6 tags at the N- and C-termini with TEV and thrombin cleavage sites, respectively Val67 to Cys; Thr230 to Cys
EspΔ65CysM2, Ser66–Gln282 pET-21b; ampicillin-resistant His6 tags at the N- and C-termini with TEV and thrombin cleavage sites, respectively Val67 to Cys; Thr230 to Cys
EspΔ58CysM2, Ser59–Gln282 pET-21b; ampicillin-resistant His6 tags at the N- and C-termini with TEV and thrombin cleavage sites, respectively Val67 to Cys; Asp255 to Cys
EspΔ55CysM2, Ile56–Gln282 pET-21b; ampicillin-resistant Thrombin cleavage site followed by His6 tag at the C-terminus Val67 to Cys; Asp255 to Cys
Open in a new tab

2.2. Expression and purification  

The respective plasmids were transformed into Escherichia coli BL21 (DE3) cells for expression. 1 l Luria–Bertani (LB) broth supplemented with 100 mg ml−1 ampicillin or 50 mg ml−1 kanamycin (depending on the plasmid) was inoculated with an overnight culture (10 ml) of stationary-phase E. coli. The cells were grown at 37°C until the culture reached an OD600 nm of ∼0.5, at which point the temperature was reduced to 30°C. When the OD600 nm reached ∼0.7, protein expression was induced by the addition of 1 mM isopropyl β-d-1-thio­galactopyranoside (IPTG) and the culture was incubated overnight at 30°C. Harvesting of the cells was carried out by centrifugation at 3210g (Beckman Allegra 6R Centrifuge) for 30 min at 4°C; the cells were resuspended in lysis buffer consisting of 50 mM Tris pH 8.0, 250 mM NaCl, 10%(v/v) glycerol and an EDTA-free protease-inhibitor cocktail tablet (Roche). The cells were lysed by sonication on ice and the soluble supernatant was collected by centrifugation at 48 384g (Beckman Avanti J-25 centrifuge) for 1 h at 4°C. The clear lysate was allowed to pass through a pre-packed 5 ml HisTrap FF column (GE Healthcare) pre-equilibrated with lysis buffer at a flow rate of 1 ml min−1 to enable binding of the recombinant His-tagged proteins to the beads. The protein-bound column was washed with ten bed volumes of buffer A (50 mM Tris pH 8.0, 50 mM NaCl), followed by a wash with ten bed volumes of buffer consisting of 50 mM pH 8.0, 50 mM NaCl, 37.5 mM imidazole. The bound protein was eluted with a linear gradient of 15–100% buffer B and the fractions containing the recombinant protein were checked for purity and molecular weight on a 12% SDS–PAGE gel (Laemmli, 1970 ▸). These samples were further purified by gel-filtration chromatography (HiPrep 16/60 Sephacryl S-100 HR column; GE Healthcare) with 50 mM Tris pH 8.0, 150 mM NaCl, 10% glycerol.

2.3. Enzyme activity  

The activity of the purified Pro-Esp variant recombinants was tested using azocasein, a nonspecific substrate (Rice et al., 2001 ▸). Proteins (40 µg per reaction) were incubated at 37°C for 1 h with 1%(w/v) azocasein (Sigma) in a total volume of 400 µl. Following incubation, 400 µl 5% trichloroacetic acid (TCA) was added to quench the reaction. Upon centrifugation at 14 000g for 10 min, the soluble material was used to check the activity by measuring the absorbance at 440 nm using a spectrophotometer.

The enzyme activity of purified Pro-Esp and its variant recombinants (0.2 µg per reaction) was measured using 1 mM Z-Leu-Leu-Glu-AMC (where Z is a carboxyl/Boc protecting group and AMC is 7-amino-4-methylcoumarin, a detectable fluorescent group; Enzo Life Sciences) in 50 mM Tris–HCl pH 8.0 containing 5 mM EDTA in a total volume of 200 µl. EDTA was added to suppress the activity of thermolysin and other nonspecific metalloproteases (Fontana, 1988 ▸). The reaction mixture was incubated at 37°C for 1 h before measuring the fluorescence with excitation at 370 nm and emission at 460 nm using a FLUOstar OPTIMA (BMG Labtech) in a 384-well hard-shell black plate.

2.4. Protein crystallization and X-ray diffraction data collection  

Purified Pro-Esp and its variants were concentrated using an Amicon ultracentrifugation system and initial crystallization conditions were screened by the sitting-drop vapor-diffusion method at 20°C using a Crystal Phoenix robot (Art Robbins Instruments). Commercially available crystallization screens such as The PEGs and PEGs II Suites (Qiagen), JCSG Core I and II Suites (Qiagen), Crystal Screen HT (Hampton Research) and Wizard Classic 1 and 2 (Rigaku Reagents) were used for screening. The crystallization conditions were optimized using the hanging-drop vapor-diffusion method at 22°C and good-quality single crystals were produced after optimization. Native diffraction data were collected at 100 K using an appropriate cryoprotectant either in-house or on the SER-CAT 22-ID beamline at the Advanced Photon Source. Optimized crystallization conditions, the cryoprotectant used for data collection at 100 K and the diffraction data statistics for the Esp variants are summarized in Table 2 ▸.

Table 2. Crystallization conditions and data-collection and refinement statistics.

Values in parentheses are for the highest resolution shell.

  EspS235A† EspΔ55S235A-pro EspΔ55S66V-pro EspΔ65CysM2† EspΔ58CysM2†
Crystallization conditions
 Composition of reservoir solution 20% PEG 8000, 0.1 M HEPES pH 7.5 0.1 M magnesium chloride, 0.1 M sodium acetate pH 4.6, 30% PEG 4000 20% PEG 3350, 0.225 M ammonium nitrate 25% PEG MME 5000, 0.2 M lithium sulfate, 0.1 M Tris pH 8.5 1 M lithium chloride, 0.1 M Tris pH 8.5, 20% PEG 6000
 Cryoprotectant 15% glycerol 20% glycerol 35% ethylene glycol 15% glycerol 20% PEG 400
Data collection
 Wavelength (Å) 1.0 1.0 1.54 1.0 1.0
 Resolution range (Å) 34.6–1.200 (1.243–1.200) 30.4–1.850 (1.916–1.850) 38.4–2.199 (2.278–2.199) 36.8–2.066 (2.140–2.066) 36.2–2.080 (2.154–2.080)
 Space group P1211 P41212 P1211 P6122 P6122
  a, b, c (Å) 39.6, 61.0, 42.5 69.1, 69.1, 127.3 42.6, 61.1, 75.5 42.5, 42.5, 373.4 42.0, 42.0, 378.4
 α, β, γ (°) 90.0, 98.7, 90.0 90.0, 90.0, 90.0 90.0, 94.7, 90.0 90.0, 90.0, 120.0 90.0, 90.0, 120.0
 No. of molecules in the asymmetric unit 1 1 2 1 1
 Total reflections 119842 (10950) 47575 (4664) 28089 (2054) 193282 (4750) 25524 (1948)
 Unique reflections 60633 (5576) 25463 (2522) 18892 (1779) 13121 (1075) 12812 (1016)
 Multiplicity 2.0 (2.0) 1.9 (1.8) 1.5 (1.2) 14.7 (4.4) 2.0 (1.9)
 Completeness (%) 97 (89) 94 (95) 95 (91) 97 (82) 97 (81)
 Mean I/σ(I) 30.8 (13.8) 8.8 (1.2) 9.7 (4.5) 49.0 (17.2) 39.7 (13.4)
 Wilson B factor (Å2) 9.8 23.9 11.8 16.9 19.3
  R merge (%) 1.5 (3.8) 6.2 (61.8) 4.1 (8.8) 4.2 (5.6) 1.0 (3.8)
  R meas (%) 2.2 (5.4) 8.8 (87.4) 5.8 (12.5) 4.4 (6.2) 1.4 (5.4)
 CC1/2 99.8 (99.5) 99.7 (43.6) 99.4 (96.6) 99.8 (99.7) 99.9 (99.2)
 CC* 99.9 (99.9) 99.9 (77.9) 99.8 (99.1) 100.0 (99.9) 100.0 (99.8)
Refinement
 Reflections used in refinement 60628 (5576) 25460 (2522) 18887 (1779) 13114 (1075) 12810 (1014)
 Reflections used for R free 1799 (166) 1999 (199) 980 (118) 1312 (108) 1281 (101)
  R work (%) 15.9 (17.0) 19.3 (28.7) 17.0 (17.5) 14.7 (16.3) 16.0 (19.4)
  R free (%) 16.4 (18.2) 22.2 (31.5) 23.2 (25.7) 18.6 (21.7) 21.5 (24.8)
 CC (work) 0.95 (0.93) 0.96 (0.72) 0.95 (0.92) 0.96 (0.93) 0.95 (0.91)
 CC (free) 0.94 (0.93) 0.95 (0.73) 0.91 (0.85) 0.94 (0.90) 0.92 (0.87)
 No. of non-H atoms
  Protein 1646 1620 3145 1670 1704
  Water 299 190 275 207 186
  Total 1945 1810 3420 1877 1890
 R.m.s. deviations
  Bonds (Å) 0.004 0.007 0.007 0.009 0.007
  Angles (°) 0.83 0.82 0.79 0.86 0.82
 Ramachandran plot
  Most favored (%) 96.2 96.6 96.4 95.8 96
  Allowed (%) 3.8 3.4 3.6 4.2 3.7
  Outliers (%) 0 0 0 0 0
PDB code 6pym 6q24 6q12 6tya 6u1b
Open in a new tab
†

Friedel mates were averaged when calculating reflection statistics.

2.5. Structure determination  

Data indexing, integration and scaling were performed using XDS (Kabsch, 2010 ▸) and iMosflm (Battye et al., 2011 ▸) and the crystal structures were solved by the molecular-replacement technique (Rossmann, 1990 ▸) using the mature Esp structure (PDB entry 4jcn; Chen et al., 2013 ▸) as the starting model. Model building, refinement and validation were performed with the CCP4 suite (Collaborative Computational Project, Number 4, 1994 ▸; Winn et al., 2011 ▸; Vagin et al., 2004 ▸; Matthews, 1968 ▸), Phenix (Liebschner et al., 2019 ▸) and Coot (Emsley & Cowtan, 2004 ▸; Emsley et al., 2010 ▸).

2.6. N-terminal sequencing  

Protein sequencing was performed using an ABI 494 protein sequencer. Samples were transferred onto a PVDF membrane and shipped to Tufts University core facility for sequencing.

2.7. Sequence and structure comparison  

The primary sequences of the Pro-Esp constructs were compared using MUSCLE (Edgar, 2004 ▸), an online multiple sequence-alignment tool from The European Bioinformatics Institute (EMB-EBI; McWilliam et al., 2013 ▸). Crystal structure comparison was performed using Coot (Emsley et al., 2010 ▸), PyMOL (v.1.8; Schrödinger; http://www.pymol.org) and UCSF Chimera (Pettersen et al., 2004 ▸). Figures were generated using PyMOL.

2.8. Docking studies  

Molecular-docking studies were performed using an open-source program called AutoDock Vina (Trott & Olson, 2010 ▸). The high-resolution crystal structure of EspS235A was used as the minimized receptor structure and the peptide Z-Leu-Leu-Glu-AMC was docked into it. The structures were optimized using AutoDock parameters such as atom types, torsion modes and partial charges, and the models were saved in PDBQT format. A grid of dimensions 62 × 34 × 46 Å was generated to identify xyz coordinates (x = 19.331, y = 19.55, z = 1.444) around the substrate-binding region. To find the most favorable interactions, the Lamarckian genetic algorithm (LGA) was selected with default parameters in AutoDock Vina. The best pose was selected depending on the highest binding energy criterion.

3. Results  

3.1. Primary sequences, data collection and refinement statistics of Esp variants  

The primary sequences of the Pro-Esp constructs resulting from cloning and expression were compared using T-Coffee (Notredame et al., 2000 ▸) and BoxShade from ExPASy (Fig. 1 ▸), and the primary structures of the purified proteins are depicted in Fig. 2 ▸ with the thermolysin cleavage site, N-terminal pro-peptide and mutations labeled. Purified full-length Pro-Esp and full-length Pro-EspS235A were treated with thermolysin to cleave between Ser66 and Val67 in order to generate mature Esp and mature EspS235A, respectively. Before setting up the crystallization screens, the N- and C-terminal His6 tags were removed where applicable using TEV protease and thrombin, respectively. Protein purity was checked on an SDS–PAGE gel and the data-collection and refinement statistics for the protein crystals are given in Table 2 ▸.

Figure 1.

Figure 1

Open in a new tab

Primary-sequence comparison of recombinant Esp variants. Sequence similarity is highlighted in red and the mutations S66V, V67C, T230C, S235A and D255C in their respective constructs are indicated in boxes. The missing N-terminal residues are indicated by dotted lines.

Figure 2.

Figure 2

Open in a new tab

Schematic representation of the structures of recombinant Esp variants. The thermolysin cleavage site is indicated with an arrow and the N-­terminal pro-peptide residues are labeled. Mutated residues are indicated with red lines and labeled.

3.2. Structural comparison of Esp with Pro-Esp  

Mature EspS235A was crystallized and its three-dimensional structure was determined by molecular replacement. Except for the Ala235 residue, the mature EspS235A structure was identical to the three-dimensional structure of mature Esp previously determined in our laboratory (PDB entry 4jcn; Fig. 3 ▸ c; Chen et al., 2013 ▸).

Figure 3.

Figure 3

Open in a new tab

(a) EspS235A crystals. (b) Crystal structure overlay of EspS235A (blue) and mature Esp (PDB entry 4jcn; green). His117, Asp159 and Ala235 are represented as sticks in magenta. Loops 1 and 2 are labeled.

It has previously been shown that an undecapeptide of the endogenous pro-peptide (Ile56–Ser66) mimicked the native pro-peptide (66 residues) in suppressing the proteolytic activity completely and supporting efficient cleavage at the Ser66-Val67 bond by thermolysin (Ohara-Nemoto et al., 2008 ▸). Firstly, we tried to crystallize recombinant EspΔ55 (Ile56–Gln282), but it consistently underwent autoproteolysis to produce mature Esp crystals. To prevent the autoproteolysis, we cloned two variants. In one variant we mutated the active site Ser235 to Ala (EspΔ55S235A) to prevent autocleavage, and in the other we mutated Ser66 to Val (EspΔ55S66V) to block the cleavage of the Ser66-Val67 bond (Figs. 1 ▸ and 2 ▸). We were able to determine crystal structures of both EspΔ55S66V and EspΔ55S235A successfully. The EspΔ55S235A recombinant crystallized in space group P41212 with one molecule in the asymmetric unit, while the crystals of the EspΔ55S66V recombinant belonged to space group P21, with two molecules in the asymmetric unit related by a twofold noncrystallographic axis. The root-mean-square deviation (r.m.s.d.) between EspΔ55S235A and EspΔ55S66V was 0.34–0.36 Å, with some interesting structural differences between the two.

The crystal structure of EspΔ55S66V, which is referred to as Pro-Esp from here on, has two molecules in the asymmetric unit (A and B). Molecules A and B are almost identical (r.m.s.d. of 0.23 Å) and exhibit significant differences from the mature Esp structure (PDB entry 4jcn). The substrate-binding sites of both Pro-Esp molecules in the asymmetric unit (A and B) were solvent-exposed. More importantly, no electron density was observed for residues Ser63–Asn72 in both molecules A and B, and thus the Val67 residue critical for substrate recognition (Nemoto et al., 2008 ▸; Ono et al., 2010 ▸) was not associated with the S1 pocket as seen in the mature Esp structure. The electron density for the loop 1 residues Ser229–Gly233 in molecule A and Leu228–Thr230 in molecule B and the loop 2 residues Val254–Lys257 in both molecules A and B was weak, indicating flexibility (Fig. 4 ▸ b). Significantly, the disorder in loop 1 of molecules A and B also extended to the respective oxyanion-hole residues (Val231–Ser235); a disordered oxyanion-hole conformation is deleterious to serine protease activity and efficiency (Perona & Craik, 1995 ▸; Fig. 4 ▸ b).

Figure 4.

Figure 4

Open in a new tab

(a) Pro-Esp crystals. (b) Crystal structure overlay of Pro-Esp (red) and mature Esp (green). The catalytic triad and the N-terminal residues Asn73 in Pro-Esp and Val67 in mature Esp are represented as sticks in magenta. The N-­terminus of Pro-Esp is repositioned, destabilizing loops 1 and 2. The oxyanion hole is in a nonfunctional conformation. (b) is enlarged to highlight the residues on the N-terminus and loops 1 and 2.

The EspΔ55S235A recombinant crystallized with one molecule in the asymmetric unit and exhibited no enzyme activity in an azocasein assay, similar to Pro-Esp, and the electron density was also missing for its N-terminal residues Ser63–Asn72. Unlike the Pro-Esp structure, loop 1 and loop 2 of EspΔ55S235A had clear, well defined electron density; loop 2 was almost identical to the mature Esp structure. However, loop 1, which contained the conserved ‘oxyanion-hole’ region, moved inwards compared with the mature Esp structure. This significant movement was owing to crystal-packing contacts from a neighboring symmetry-related molecule.

Interestingly, the loop 2 residue region of the neighboring symmetry-related molecule is seen to associate with the S1 pocket of the EspΔ55S235A reference molecule. The side chain of Asp255 in the intruding loop is suitably positioned to interact with the conserved His250 of the S1 pocket, which has been implicated in positioning the P1 residue Glu in the mature Esp–substrate complex (Demidyuk et al., 2004 ▸). Also, the Asp255-Asn256 peptide bond of the intruding loop is positioned such that the carbonyl atom of Asp255 is at a suitable distance (∼3 Å) for nucleophilic attack by the catalytic Ser235. This serendipitous observation may provide a partial glimpse of the substrate-bound active-site structure of Esp (Fig. 5 ▸).

Figure 5.

Figure 5

Open in a new tab

The crystal structure of EspΔ55S235A (cyan) is superimposed (r.m.s.d. of 0.41 Å) on that of mature Esp (PDB entry 4jcn; orange). Val67, His117, Asp159 and Ser235 of mature Esp are labeled and represented as sticks in red. In EspΔ55S235A, Asn73 and His250 (represented as sticks in red) are labeled. Loop 2 of the symmetry-related molecule (magenta) of EspΔ55S235A is indicated with an arrow, and Asp255 and Asn256 are labeled and represented as sticks in yellow. The inward movement of loop 1 and the closed conformation of the oxyanion hole are shown as an arrow and a dotted ellipse, respectively. The contacts between residues Asp255–His250 (2.82 Å) and Asn256–Ser235 (2.52 Å) are represented as dotted lines.

Thus, both the EspΔ55S235A and EspΔ55S66V (Pro-Esp) structures revealed a flexible N-terminus and a distorted loop 1 and loop 2 in the S1 pocket, similar to other chymotrypsin-like serine protease zymogen structures.

3.3. Crystal structures of locked N-terminal Pro-Esp variants  

To check the significance of such a uniquely positioned N-terminus, associating with and covering the bottom of the S1 pocket, as seen in the structures of mature GSE enzymes, we created cysteine mutants of Pro-Esp. We expected the enzyme activity to be retained when the N-terminus is locked in a position similar to mature Esp but with extra N-terminal residues. Firstly, we mutated Val67 and Thr230 (loop 2) to cysteine (EspΔ65CysM1; Figs. 1 ▸ and 2 ▸) and crystallized the recombinant. However, the crystal structure of the EspΔ65CysM1 recombinant revealed no disulfide bond between Cys67 and Cys230, and there was no electron density for residues Ser66–Asn72 of the N-terminus, similar to the Pro-Esp crystal structure.

Although there was no interaction between Asp255 on loop 2 and the N-terminal Val67, their Cα atoms were in close proximity to form a disulfide bond when mutated to cysteines (EspΔ65CysM2; Figs. 1 ▸ and 2 ▸). The crystal structure of EspΔ65CysM2 containing just one residue from the pro-segment (Ser66) revealed a disulfide bond between the side-chain –SH groups of residues Cys67 and Cys255. Furthermore, loops 1 and 2 were stabilized and the oxyanion hole adopted a functionally active and open conformation.

The crystal structure of EspΔ58CysM2, similar to that of EspΔ65CysM2, had a disulfide bond between the side-chain –SH groups of Cys67 and Cys255. It also showed clearly defined electron densities for the residues involved in the formation of loops 1 and 2. The oxyanion hole in this variant also adopted an open and functionally active conformation similar to that in mature Esp and EspΔ66CysM2. Additionally, we observed that the six residues in the pro-peptide segment, Asn61–Ser66, were positioned on top of the S1 pocket, potentially blocking its access to the substrate.

4. Discussion  

4.1. A structural perspective on Pro-Esp activation  

It has previously been reported that residues Met1–Ala27 of the native Esp form a pre-segment containing a signal sequence and that Lys28–Ser66 is a pro-segment, suggesting that Esp is secreted as a 32 kDa (Lys28–Gln282) pro-enzyme (Ohara-Nemoto et al., 2008 ▸). Therefore, we cloned the recombinant pro-enzyme Ala27–Gln282 (Fig. 3 ▸) but, as detected by N-terminal sequencing (Table 3 ▸), the purified pro-enzyme converted to a 29 kDa intermediate (Ile49–Gln282). This intermediate pro-enzyme is referred to as full-length Esp in the subsequent discussion. Additionally, Table 3 ▸ details the regions with undefined electron density at the N-terminus of the respective crystal structures.

Table 3. N-terminal sequencing and definition from electron density.

  N-terminal sequencing  
Esp variant Sample Crystal N-terminus in the crystal structure (electron density)
Full-length Esp I49NSSS53, missing residues Ala27–Asp48 NA NA
Esp mature NA NA V67ILPN71
EspS235A NA NA V67ILPN71
EspΔ55S235A NA NA N73RHQI77, undefined density for residues Ile56–Asn72
EspΔ55S66V S63YPVV67, missing residues Ile56–Lys62 NA N73RHQI77, undefined density for residues Ser63–Asn72
EspΔ55CysM1 SI56KPS59 NA NA
EspΔ65CysM1 S66CILP70 S66CILP70 N73RHQI77, undefined density for residues Ser66–Asn72
EspΔ65CysM2 S66CILP70 NA S66CILP70
EspΔ58CysM2 N61KSYP65, missing residues Ser59 and Glu60 N61KSYP65 N61KSYP65
EspΔ55CysM2 MI56KPS59 MI56KPS59 P65SCIL69, undefined density for residues Ile56–Tyr64
Open in a new tab

We were not able to crystallize full-length Esp, perhaps owing to the unusually long and flexible pro-peptide. According to Nemoto et al. (2008 ▸), the undecapeptide of the endogenous pro-peptide was adequate to completely suppress the proteolytic activity and to support efficient cleavage at the Ser66-Val67 bond by thermolysin (Ohara-Nemoto et al., 2008 ▸). Interestingly, full-length Esp displayed residual activity in the azocasein assay which was about 35-fold lower than that of mature Esp (Fig. 6 ▸ a). This was similar to the zymogens of the staphylococcal SplB protease (Pustelny et al., 2014 ▸) and bovine trypsin (Fehlhammer et al., 1977 ▸), in which the spatial arrangement of the catalytic residues (Ser214, Asp102, His57 and Ser195; chymotrypsin numbering) in the zymogens was almost similar, but with slightly distorted or inadequately formed S1 pockets compared with the mature enzymes. Such an arrangement in the zymogen could also explain the complete absence of activity against the specific substrate (Fig. 6 ▸ b) for the full-length Esp recombinant. The EspS235A recombinant showed no activity against azocasein and specific substrate as expected, but this raised a pertinent question. Does the long pro-peptide in full-length Esp block the active site by association as in cathepsins (Fox et al., 1992 ▸), but is not tight enough and ‘leaks’ residual activity, or is it as in trypsin­ogen with a flexible N-terminus and a disordered S1 pocket, hence the weak enzyme activity?

Figure 6.

Figure 6

Open in a new tab

Protease activities (mean ± SD, n = 3) of Esp variants before and after incubation with thermolysin overnight at room temperature using (a) the nonspecific substrate azocasein and (b) the specific substrate Z-LLE-AMC. Samples are labeled as follows: 1, full-length Esp; 2, full-length Esp + thermolysin; 3, EspS235A; 4, EspΔ55; 5, EspΔ55 + thermolysin; 6, EspΔ55S235A; 7, EspΔ55S66V (Pro-Esp); 8, EspΔ55CysM1, 9, EspΔ65CysM2; 10, EspΔ58CysM2; 11, EspΔ55CysM2. Full-length Esp and EspΔ55 are activated after cleavage at the thermolysin site.

As expected, the purified EspΔ55S235A and EspΔ55S66V recombinants were inactive (Figs. 7 ▸ a and 7 ▸ b) despite the loss of seven residues (Ile56–Lys62) in EspΔ55S66V, as identified from N-terminal sequencing (Table 3 ▸). The crystal structures of both variants showed density starting from Asn73 of the N-terminus, suggesting a flexible nature for the pro-segment. The stabilization of the N-terminus beyond Asn73 was mainly owing to the presence of Arg74 and its salt bridge with Asp230. Interestingly, in the mature Esp structure and also in the EspΔ55S235A structure loops 1 and 2 had clear density, and Arg74, in addition to Asp230, also interacted with the Asn234 side chain through a hydrogen-bonding network mediated by a conserved water molecule. This Arg74–Asn234 interaction was reminiscent of and was spatially similar to the interaction between the N-terminal Ile NH2 positive charge and Asp194 (chymotrypsin numbering), as seen in all activated serine proteases, an interaction that is responsible for stabilization of the S1 pocket and for the critical ‘oxyanion-hole’ formation during conversion from the zymogen to the active enzyme (Volanakis & Narayana, 1996 ▸). However, the N-terminal positive charge in GSEs, a product of pro-peptide loss during the transformation from the zymogen to the active enzyme, was directly involved in the stable association with the S1 pocket but not through interaction with Asn234 (Asp194; chymotrypsin numbering). It is interesting to note that owing to Arg74–Asn234 interactions, loop 1 in EspΔ55S235A had clear density. However, in the absence of anchoring by the N-terminal Val67, this particular loop, which was highly conserved spatially in all serine proteases, had shifted inwards to accommodate the crystallographic contacts and substrate-like peptide (Fig. 5 ▸) intrusion into the S1 pocket.

Figure 7.

Figure 7

Open in a new tab

(a) EspΔ58CysM2 (cyan) is superimposed onto the Pro-Esp structure (red). (b) EspΔ65CysM2 (yellow) is superimposed onto EspΔ58CysM2 (cyan). The disulfide bonds between Cys67 and Cys255 are represented as sticks in yellow in (a) and magenta in (b). In both of the CysM2 structures the N-­terminus is locked stabilizing loops 1 and 2, and the oxyanion hole is in the functional conformation. The pro-peptide extension blocks the S1 pocket in EspΔ58CysM2. (a) and (b) are enlarged to highlight the residues at the N-terminus and loops 1 and 2. (c) Crystals of the N-terminal locked Esp variant.

4.2. Elucidating the significance of the N-terminal Val67 residue in the activity  

The position of the N-terminus, as observed in the Pro-Esp structure, begged the questions: if we secure the N-terminus in its preferred position, will it stabilize loops 1 and 2 and can the enzyme activity be regained? One way to address these questions was by securing the N-terminus with disulfide bonds. Our initial attempts to lock the N-terminus by mutating Val67 to Cys and Thr230 of loop 1 to Cys (EspΔ65CysM1) failed. We suspect that the side chains of Cys67 and Cys230 were not optimally oriented to form a disulfide bond, which could reinforce the significance of the spatial orientation and positioning of Val67, as suggested by Ono et al. (2010 ▸); any replacement of the Val69 residue of V8 (Val67 in Esp) resulted in a loss of the enzyme activity. The crystal structure of EspΔ65CysM1 was similar to that of Pro-Esp, in which electron density for Ser66–Asn72 was missing despite the presence of these residues before and after crystallization (Table 3 ▸).

Next, we locked the N-terminus by replacing Val67 and Asp255 (loop 2) with Cys residues. This purified recombinant Pro-Esp variant with eight pro-peptide residues Ser59–Ser66 (EspΔ58CysM2) had no enzymatic activity (Figs. 6 ▸ a and 6 ▸ b). Even though N-terminal sequencing of the EspΔ58CysM2 recombinant revealed a loss of the Ser59 and Gln60 residues (Table 3 ▸), the EspΔ58CysM2 crystal structure revealed an S1 pocket and catalytic apparatus similar to those of mature Esp with the six-residue pro-segment Asn61–Ser66 blocking the S1 pocket (Fig. 7 ▸ b). Interestingly, the EspΔ65CysM2 recombinant regained partial activity in the azocasein assay, which was tenfold higher than that of full-length Esp and only 2.5-fold lower than that of mature Esp (Fig. 6 ▸ a). The EspΔ65CysM2 crystal structure, in contrast to that of EspΔ65CysM1, displayed a disulfide bond between Cys67 and Cys255 and stabilized loops 1 and 2 and the overall structure (Fig. 7 ▸ a). It also resembled that of mature Esp. However, EspΔ65CysM2 retained one residue from the pro-segment (Ser66) owing to cloning constraints. Upon inserting the TEV cleavage site (ENLYFQS) and treating the EspΔ65CysM2 protein with TEV protease, the bond between the Gln and Ser residues was cleaved, leaving the serine from the TEV cleavage site at the N-terminus. Serendipitously, this matched with the Ser66 residue on the pro-peptide of Esp. The reduced activity of recombinant EspΔ65CysM2 could be explained with the help of its crystal structure, in which the terminal Ser66 of the locked N-terminus was seen to partially block the S1 pocket. The side-chain hydroxyl group of Ser66 interacted with the catalytic Ser235 through a network of water-mediated hydrogen bonds. Owing to the high structural similarity between EspΔ65CysM2 and mature Esp, we can suggest that a Pro-Esp cysteine mutant without the extra Ser66 residue would regain full activity. The crystal structures of all of the Esp variants discussed so far were superimposed onto that of EspS235A for comparison (Fig. 8 ▸).

Figure 8.

Figure 8

Open in a new tab

Superimposing the crystal structures of all of the Esp variants on EspS235A (blue) enlarged to highlight the residues at the N-terminus and loops 1 and 2. Pro-Esp (EspΔ55S66V) is in red (r.m.s.d. of 0.2 Å), EspΔ65CysM2 is in yellow (r.m.s.d. of 0.15 Å) and EspΔ58CysM2 is in cyan (r.m.s.d. of 0.2 Å). Disulfide bonds and catalytic triad residues are represented as sticks in magenta.

4.3. AutoDock: model of the Z-LLE-AMC–EspS235A complex structure and future direction  

Esp cleaves autolysin to inhibit S. aureus biofilm formation (Chen et al., 2013 ▸), and the substrate Z-LLE-AMC that mimics the autolysin cleavage site was used to test for protease activity. We tried to crystallize the EspS235A–Z-LLE-AMC complex by co-crystallization and by soaking the substrate into EspS235A crystals, but we were not successful.

We used the AutoDockVina docking program and proposed a model of the EspS235A–Z-LLE-AMC complex. The docked substrate Z-LLE-AMC was positioned appropriately in the S1 pocket of EspS235A. The N-terminal Val67 amino group was suitably positioned to act as an acceptor of the negative charge of the P1 residue (Glu) side chain of the substrate. The carboxylate group of Glu in Z-LLE-AMC was also at a hydrogen-bond distance from the Ser235/Ala235 residue (Fig. 9 ▸). These interactions validate the interaction of Ser235 with the glutamate of Z-LLE-AMC and the subsequent cleavage of the bond between glutamate and AMC, releasing the fluorogenic group AMC. We suspect that the glutamate-harboring loop/region of autolysin would be cleaved similarly, rendering it inactive.

Figure 9.

Figure 9

Open in a new tab

Docking of Z-LLE-AMC in EspS235A using AutoDock Vina. EspS235A is in gold, mature Esp is in cyan and Z-LLE-AMC is in orange. Ser235 of mature Esp and Ala235, Val67 and His250 of EspS235A are represented as sticks and colored magenta.

5. Conclusions  

S. epidermidis extracellular serine protease (Esp) closely resembles V8 and belongs to a subfamily of chymotrypsin-like serine proteases. It is also a glutamyl endopeptidase that hydrolyzes the peptide bond next to a glutamate residue. Expressed as an inactive zymogen and activated upon losing the pro-peptide segment, Esp has been shown to cleave the autolysin from S. aureus and inhibit its biofilm formation. The previously determined mature Esp structure identified an intimate association between the N-terminus (Val67–Gln76) and the substrate-binding S1 pocket (Chen et al., 2013 ▸). A structural perspective on the the activation of Pro-Esp to Esp, as presented in this study, reveals a mechanism similar to that of the typical serine protease trypsin. The residues on loop 1 and loop 2 in the Pro-Esp structures EspΔ55S235A and EspΔ55S66V were either flexible or disordered, and the respective oxyanion hole was in a nonfunctional conformation. These crucial segments of the enzyme need to be stabilized before it can function efficiently. The crystal structure of EspΔ55S235A also revealed the interaction of the residues in the substrate-binding S1 pocket, Ser235 and His250, with Asn256 and Asp255, respectively, from a symmetry-related molecule, giving us a possible glimpse into the substrate-bound Esp structure.

Furthermore, an N-terminal locked Pro-Esp variant with eight pro-peptide residues did not display any activity, but the crystal structure revealed stabilization of the regions around the S1 pocket (loops 1 and 2 and the oxyanion hole). However, the crystal structure of an N-terminal locked variant with one pro-peptide residue revealed interaction between Ser66 and the active-site serine Ser235, explaining its partial activity despite having a stabilized S1 pocket. Considering the similarity between the mature structures, we can assume that the activation of the V8 protease zymogen would be similar to that of the Esp zymogen. Overall, we observe that the unusually long pro-peptide segment in bacterial GSEs is flexible and does not associate with the main body of the enzyme, as seen in many cysteine proteases, and the zymogens of GSEs undergo structural changes in the substrate-binding pocket similar to the pro-enzymes of chymotrypsin-like serine proteases upon the loss of their pro-peptide segments. However, the resultant N-termini associate with the substrate-binding pockets of GSEs in a unique way to dictate the specificity and efficiency of the enzyme.

Supplementary Material

PDB reference: extracellular serine protease, active-site serine mutant, 6pym

PDB reference: zymogen mutant, S66V, 6q12

PDB reference: zymogen mutant, S235A, 6q24

PDB reference: N-terminal locked with one pro-peptide residue, V67C, D255C, 6tya

PDB reference: N-terminal locked with eight pro-peptide residues, V67C, D255C, 6u1b

Funding Statement

This work was funded by National Institute of Allergy and Infectious Diseases grant R01-AI106808.

References

  1. Battye, T. G. G., Kontogiannis, L., Johnson, O., Powell, H. R. & Leslie, A. G. W. (2011). Acta Cryst. D67, 271–281. [DOI] [PMC free article] [PubMed]
  2. Bukowski, M., Wladyka, B. & Dubin, G. (2010). Toxins, 2, 1148–1165. [DOI] [PMC free article] [PubMed]
  3. Calander, A. M., Dubin, G., Potempa, J. & Tarkowski, A. (2008). FEMS Immunol. Med. Microbiol. 52, 267–272. [DOI] [PubMed]
  4. Cavarelli, J., Prévost, G., Bourguet, W., Moulinier, L., Chevrier, B., Delagoutte, B., Bilwes, A., Mourey, L., Rifai, S., Piémont, Y. & Moras, D. (1997). Structure, 5, 813–824. [DOI] [PubMed]
  5. Chen, C., Krishnan, V., Macon, K., Manne, K., Narayana, S. V. L. & Schneewind, O. (2013). J. Biol. Chem. 288, 29440–29452. [DOI] [PMC free article] [PubMed]
  6. Chen, Z. & Ruffner, D. E. (1998). Nucleic Acids Res. 26, 1126–1127. [PMC free article] [PubMed]
  7. Collaborative Computational Project, Number 4 (1994). Acta Cryst. D50, 760–763.
  8. Dancer, S. J., Garratt, R., Saldanha, J., Jhoti, H. & Evans, R. (1990). FEBS Lett. 268, 129–132. [DOI] [PubMed]
  9. Demidyuk, I. V., Romanova, D. V., Nosovskaya, E. A., Chestukhina, G. G., Kuranova, I. P. & Kostrov, S. V. (2004). Protein Eng. Des. Sel. 17, 411–416. [DOI] [PubMed]
  10. Drapeau, G. R. (1978). J. Bacteriol. 136, 607–613. [DOI] [PMC free article] [PubMed]
  11. Drapeau, G. R., Boily, Y. & Houmard, J. (1972). J. Biol. Chem. 247, 6720–6726. [PubMed]
  12. Dubin, G., Chmiel, D., Mak, P., Rakwalska, M., Rzychon, M. & Dubin, A. (2001). Biol. Chem. 382, 1575–1582. [DOI] [PubMed]
  13. Edgar, R. C. (2004). BMC Bioinformatics, 5, 113. [DOI] [PMC free article] [PubMed]
  14. Emsley, P. & Cowtan, K. (2004). Acta Cryst. D60, 2126–2132. [DOI] [PubMed]
  15. Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. (2010). Acta Cryst. D66, 486–501. [DOI] [PMC free article] [PubMed]
  16. Fehlhammer, H., Bode, W. & Huber, R. (1977). J. Mol. Biol. 111, 415–438. [DOI] [PubMed]
  17. Fontana, A. (1988). Biophys. Chem. 29, 181–193. [DOI] [PubMed]
  18. Fox, T., de Miguel, E., Mort, J. S. & Storer, A. C. (1992). Biochemistry, 31, 12571–12576. [DOI] [PubMed]
  19. Gasanov, E. V., Demidyuk, I. V., Shubin, A. V., Kozlovskiy, V. I., Leonova, O. G. & Kostrov, S. V. (2008). Protein Eng. Des. Sel. 21, 653–658. [DOI] [PubMed]
  20. Iwase, T., Uehara, Y., Shinji, H., Tajima, A., Seo, H., Takada, K., Agata, T. & Mizunoe, Y. (2010). Nature (London), 465, 346–349. [DOI] [PubMed]
  21. Kabsch, W. (2010). Acta Cryst. D66, 125–132. [DOI] [PMC free article] [PubMed]
  22. Kakudo, S., Kikuchi, N., Kitadokoro, K., Fujiwara, T., Nakamura, E., Okamoto, H., Shin, M., Tamaki, M., Teraoka, H., Tsuzuki, H. & Yoshida, N. (1992). J. Biol. Chem. 267, 23782–23788. [PubMed]
  23. Kawalec, M., Potempa, J., Moon, J. L., Travis, J. & Murray, B. E. (2005). J. Bacteriol. 187, 266–275. [DOI] [PMC free article] [PubMed]
  24. Khan, A. R. & James, M. N. (1998). Protein Sci. 7, 815–836. [DOI] [PMC free article] [PubMed]
  25. Klevens, R. M., Morrison, M. A., Nadle, J., Petit, S., Gershman, K., Ray, S., Harrison, L. H., Lynfield, R., Dumyati, G., Townes, J. M., Craig, A. S., Zell, E. R., Fosheim, G. E., McDougal, L. K., Carey, R. B., Fridkin, S. K. & Active Bacterial Core Surveillance (ABCs) MRSA Investigators (2007). JAMA, 298, 1763–1771. [DOI] [PubMed]
  26. Laemmli, U. K. (1970). Nature (London), 227, 680–685. [DOI] [PubMed]
  27. Lee, C. Y., Schmidt, J. J., Johnson-Winegar, A. D., Spero, L. & Iandolo, J. J. (1987). J. Bacteriol. 169, 3904–3909. [DOI] [PMC free article] [PubMed]
  28. Liebschner, D., Afonine, P. V., Baker, M. L., Bunkóczi, G., Chen, V. B., Croll, T. I., Hintze, B., Hung, L.-W., Jain, S., McCoy, A. J., Moriarty, N. W., Oeffner, R. D., Poon, B. K., Prisant, M. G., Read, R. J., Richardson, J. S., Richardson, D. C., Sammito, M. D., Sobolev, O. V., Stockwell, D. H., Terwilliger, T. C., Urzhumtsev, A. G., Videau, L. L., Williams, C. J. & Adams, P. D. (2019). Acta Cryst. D75, 861–877.
  29. Liu, F., Zhao, Z.-S., Ren, Y., Cheng, G., Tang, X.-F. & Tang, B. (2016). Appl. Microbiol. Biotechnol. 100, 10429–10441. [DOI] [PubMed]
  30. Massimi, I., Park, E., Rice, K., Müller-Esterl, W., Sauder, D. & McGavin, M. J. (2002). J. Biol. Chem. 277, 41770–41777. [DOI] [PubMed]
  31. Matthews, B. W. (1968). J. Mol. Biol. 33, 491–497. [DOI] [PubMed]
  32. McWilliam, H., Li, W., Uludag, M., Squizzato, S., Park, Y. M., Buso, N., Cowley, A. P. & Lopez, R. (2013). Nucleic Acids Res. 41, W597–W600. [DOI] [PMC free article] [PubMed]
  33. Meijers, R., Blagova, E. V., Levdikov, V. M., Rudenskaya, G. N., Chestukhina, G. G., Akimkina, T. V., Kostrov, S. V., Lamzin, V. S. & Kuranova, I. P. (2004). Biochemistry, 43, 2784–2791. [DOI] [PubMed]
  34. Nemoto, T. K., Ohara-Nemoto, Y., Ono, T., Kobayakawa, T., Shimoyama, Y., Kimura, S. & Takagi, T. (2008). FEBS J. 275, 573–587. [DOI] [PubMed]
  35. Neurath, H. & Walsh, K. A. (1976). Proc. Natl Acad. Sci. USA, 73, 3825–3832. [DOI] [PMC free article] [PubMed]
  36. Nickerson, N. N., Prasad, L., Jacob, L., Delbaere, L. T. & McGavin, M. J. (2007). J. Biol. Chem. 282, 34129–34138. [DOI] [PubMed]
  37. Notredame, C., Higgins, D. G. & Heringa, J. (2000). J. Mol. Biol. 302, 205–217. [DOI] [PubMed]
  38. Ohara-Nemoto, Y., Ikeda, Y., Kobayashi, M., Sasaki, M., Tajika, S. & Kimura, S. (2002). Microb. Pathog. 33, 33–41. [DOI] [PubMed]
  39. Ohara-Nemoto, Y., Ono, T., Shimoyama, Y., Kimura, S. & Nemoto, T. K. (2008). Biol. Chem. 389, 1209–1217. [DOI] [PubMed]
  40. Ono, H. K., Omoe, K., Imanishi, K., Iwakabe, Y., Hu, D.-L., Kato, H., Saito, N., Nakane, A., Uchiyama, T. & Shinagawa, K. (2008). Infect. Immun. 76, 4999–5005. [DOI] [PMC free article] [PubMed]
  41. Ono, T., Ohara-Nemoto, Y., Shimoyama, Y., Okawara, H., Kobayakawa, T., Baba, T. T., Kimura, S. & Nemoto, T. K. (2010). Biol. Chem. 391, 1221–1232. [DOI] [PubMed]
  42. O’Toole, P. W. & Foster, T. J. (1987). J. Bacteriol. 169, 3910–3915. [DOI] [PMC free article] [PubMed]
  43. Park, C. H., Lee, S. J., Lee, S. G., Lee, W. S. & Byun, S. M. (2004). J. Bacteriol. 186, 6457–6464. [DOI] [PMC free article] [PubMed]
  44. Perona, J. J. & Craik, C. S. (1995). Protein Sci. 4, 337–360. [DOI] [PMC free article] [PubMed]
  45. Perona, J. J., Hedstrom, L., Rutter, W. J. & Fletterick, R. J. (1995). Biochemistry, 34, 1489–1499. [DOI] [PubMed]
  46. Pettersen, E. F., Goddard, T. D., Huang, C. C., Couch, G. S., Greenblatt, D. M., Meng, E. C. & Ferrin, T. E. (2004). J. Comput. Chem. 25, 1605–1612. [DOI] [PubMed]
  47. Prasad, L., Leduc, Y., Hayakawa, K. & Delbaere, L. T. J. (2004). Acta Cryst. D60, 256–259. [DOI] [PubMed]
  48. Pustelny, K., Zdzalik, M., Stach, N., Stec-Niemczyk, J., Cichon, P., Czarna, A., Popowicz, G., Mak, P., Drag, M., Salvesen, G. S., Wladyka, B., Potempa, J., Dubin, A. & Dubin, G. (2014). J. Biol. Chem. 289, 15544–15553. [DOI] [PMC free article] [PubMed]
  49. Qin, X., Singh, K. V., Weinstock, G. M. & Murray, B. E. (2000). Infect. Immun. 68, 2579–2586. [DOI] [PMC free article] [PubMed]
  50. Rice, K., Peralta, R., Bast, D., de Azavedo, J. & McGavin, M. J. (2001). Infect. Immun. 69, 159–169. [DOI] [PMC free article] [PubMed]
  51. Rossmann, M. G. (1990). Acta Cryst. A46, 73–82. [DOI] [PubMed]
  52. Rouf, S. M., Ohara-Nemoto, Y., Shimoyama, Y., Kimura, S., Ono, T. & Nemoto, T. K. (2012). Indian J. Biochem. Biophys. 49, 421–427. [PubMed]
  53. Rühlmann, A., Kukla, D., Schwager, P., Bartels, K. & Huber, R. (1973). J. Mol. Biol. 77, 417–436. [DOI] [PubMed]
  54. Stennicke, H. R., Birktoft, J. J. & Breddam, K. (1996). Protein Sci. 5, 2266–2275. [DOI] [PMC free article] [PubMed]
  55. Suzuki, Y., Yabuta, M. & Ohsuye, K. (1994). Gene, 150, 149–151. [DOI] [PubMed]
  56. Svendsen, I. & Breddam, K. (1992). Eur. J. Biochem. 204, 165–171. [DOI] [PubMed]
  57. Svendsen, I., Jensen, M. R. & Breddam, K. (1991). FEBS Lett. 292, 165–167. [DOI] [PubMed]
  58. Trachuk, L. A., Shcheglov, A. S., Milgotina, E. I. & Chestukhina, G. G. (2005). Biochimie, 87, 529–537. [DOI] [PubMed]
  59. Trott, O. & Olson, A. J. (2010). J. Comput. Chem. 31, 455–461. [DOI] [PMC free article] [PubMed]
  60. Vagin, A. A., Steiner, R. A., Lebedev, A. A., Potterton, L., McNicholas, S., Long, F. & Murshudov, G. N. (2004). Acta Cryst. D60, 2184–2195. [DOI] [PubMed]
  61. Velishaeva, N. S., Gasanov, E. V., Gromova, T. I. & Demidiuk, I. V. (2008). Bioorg. Khim. 34, 786–791. [DOI] [PubMed]
  62. Volanakis, J. E. & Narayana, S. V. L. (1996). Protein Sci. 5, 553–564. [DOI] [PMC free article] [PubMed]
  63. Wiederanders, B., Kaulmann, G. & Schilling, K. (2003). Curr. Protein Pept. Sci. 4, 309–326. [DOI] [PubMed]
  64. Winn, M. D., Ballard, C. C., Cowtan, K. D., Dodson, E. J., Emsley, P., Evans, P. R., Keegan, R. M., Krissinel, E. B., Leslie, A. G. W., McCoy, A., McNicholas, S. J., Murshudov, G. N., Pannu, N. S., Potterton, E. A., Powell, H. R., Read, R. J., Vagin, A. & Wilson, K. S. (2011). Acta Cryst. D67, 235–242. [DOI] [PMC free article] [PubMed]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

PDB reference: extracellular serine protease, active-site serine mutant, 6pym

PDB reference: zymogen mutant, S66V, 6q12

PDB reference: zymogen mutant, S235A, 6q24

PDB reference: N-terminal locked with one pro-peptide residue, V67C, D255C, 6tya

PDB reference: N-terminal locked with eight pro-peptide residues, V67C, D255C, 6u1b

Articles from Acta Crystallographica. Section D, Structural Biology are provided here courtesy of International Union of Crystallography

RESOURCES