sign in sign up
<<
Home , Aureolysin, Chymotrypsin, Elastase, Serine protease, Trypsin, Zymogen

NAOSITE: Nagasaki University's Academic Output SITE

Amino acid residues modulating the activities of staphylococcal glutamyl Title .

Ono, Toshio; Ohara-Nemoto, Yuko; Shimoyama, Yu; Okawara, Hisami; Author(s) Kobayakawa, Takeshi; Baba, Tomomi T; Kimura, Shigenobu; Nemoto, Takayuki K

Citation Biological chemistry, 391(10), pp.1221-1232; 2010

Issue Date 2010-10

URL http://hdl.handle.net/10069/24535

Right © 2010 by Walter de Gruyter Berlin New York 2010.

This document is downloaded at: 2020-09-18T02:19:01Z

http://naosite.lb.nagasaki-u.ac.jp BIOLOGICAL CHEMISTRY

Founded in 1877 by Felix Hoppe-Seyler as EDITOR-IN-CHIEF Zeitschriftfur PhysioLogische Chemie H. Sies, DUsseldorf

Felix Hoppe-Seyler (1825-1895) was a pioneer of , EXECUTIVE EDITORS remembered not only for hi s di scovery of hemoglobin and F. U. Hart!, Martinsried hi s contributions to the chemical characterization of many other S. LudlVig, MUnster biological compounds and processes but also for having been the K. Sandhoff, Bonn mentor of Friedrich Miescher and Albrecht Kossel. In hi s preface B. Turk, Ljubljana to the first issue of Zeitschrift fur Physiologische Chemie, A. Wittinghofer, Dortmund Feli x Hoppe-Seyler coined the term Biochemistry (,Biochemie' ) for the then newly emerging di scipline. ASSOCIATE EDITORS (GBM STUDY GROUPS) C.-M. Becker, Erlangen F. Bier, Nuthetal G. BoL, Berlin R. Brandt, OsnabrUck R. Frank, Braunschweig T. Friedrich, Freiburg K. Giehl, Giessen F.-G. Hanisch, Cologne D. Heinz, Braunschweig D. KLostermeier, Basel I. Koch, Frankfurt/Main T. Roitsch, WUrzburg C. SeideL, DUsseldorf R. Tampe, Frankfurt/Main D. WoLf, Stuttgart

EDITORIAL BOARD W Baumeister, Martinsried B. Brune, Frankfurt/Main B. Bukau, Heidelberg H. Fritz, Munich C. Hammann, Darmstadt D. Haussingel; DUsseldorf R . Kornberg, Stanford A. Kruger, Munich V. MagdoLen, Munich M. Muschen, Los Angeles G. PejLer, Uppsala N. Pfanner, Freiburg R. Pike, Melbourne J. Potempa, Krakow M. Ryan, Melbourne P. Saftig, Kiel W Schaffnel; ZUrich Biological Chemi stry is associated C. Sommerhoff, Munich with the Gesell schaft fUr Biochemie und D. WilLbold, DUsseldorf Moleku larbiologie e.V (GBM)

DE GRUYTER BioI. Chern., Vol. 391, pp. 1221-1232, October 2010· Copyright © by Walter de Gruyter· Berlin· New York. DOI10.1515/BC .2010.116

Amino acid residues modulating the activities of staphylococcal glutamyl endopeptidases

2 Toshio Ono\ Yuko Ohara-Nemoto\ Yu Shimoyama , important virulence factors. One of these abundantly secreted Hisami Okawara\ Takeshi KobayakawaI, extracellular is a , GluV8, also l 2 Tomomi T. Baba , Shigenobu Kimura and known as V8 protease/SspNGlu-C (Drapeau et aI. , 1972; 1 Takayuki K. Nemoto ,:;, Rice et aI., 2001). GluV8 contributes to the growth and sur­ vival of this microorganism in animal models (Coulter et aI., I Department of Oral , Course of 1998) and plays a key role in degrading -binding Medical and Dental Sciences, Nagasaki University and A, which are adhesion molecules on the Graduate School of Biomedical Sciences, 1-7-1 Sakamoto, staphylococcal surface (Karlsson et aI. , 2001). S. aureus Nagasaki 852-8588, Japan triggers the production of inhibitory 2 Department of Pathogenesis and Control of Oral Diseases, against GluV8 in mice, suggesting that the immunological Division of Oral Microbiology and , Iwate response to GluV8 could be an important host-defense mech­ Medical University School of Dentistry, 1-3-27 Chuodori, anism (Calander et aI., 2008). Morioka 020-8505, Japan GluV8 belongs to the glutamyl I family 8 Corresponding author (EC 3.4.21.19), whereby members specifically cleave the e-mail: [email protected] carboxyl side of the bond of negatively charged resi­ dues, i.e., and, much less potently, Abstract (S tennicke and Breddam, 1998). In staphylococci, GluV8 family members are expressed in S. aureus (Drapeau et aI. , The glutamyl endopeptidase family of from staph­ 1972), Staphylococcus epidermidis (designated GluSE; Sasa­ ylococci has been shown to be important virulence deter­ ki et aI. , 1998; Dubin et aI. , 2001; Ohara-Nemoto et aI., minants of pathogenic family members, such as 2002), and Staphylococcus wa rneri (GluSW; Yokoi et aI. , . Previous studies have identified the 2001). Members of the GluV8 family are also expressed in N-terminus and residues from positions 185- 195 as poten­ Bacillus licheniformis (Kakudo et aI., 1992; Svendsen and tially important regions that determine the activity of three Breddam, 1992), Bacillus intermedius (Rebrikov et aI., 1999; members of the family. and sequencing of the new Demidyuk et aI. , 2004), Enterococcus faecalis (Kawalec et family members from Staphylococcus caprae (GluScpr) and aI., 2005), Streptomyces fradiae (Kitadokoro et aI., 1993), Staphylococcus cohnii (GluScoh) revealed that the N-termi­ and Streptomyces griseus (Svendsen et aI. , 1991 ; Suzuki et nal Val residue is maintained in all family members. Mutants aI. , 1994), suggesting important roles of these proteases in of the GluV8 from S. aureus with altered N-terminal the life of these . Therefore, it is reasonable to pos­ residues, including amino acids with similar properties, were tulate that other staphylococcal species also produce these inactive, indicating that the Val residue is specifically endopeptidases. In fact, the genome project of Staphylococ­ required at the N-terminus of this enzyme family in order cus saprophyticus revealed the existence of the encod­ for them to function correctly. Recombinant GluScpr was ing a GluV8 family endopeptidase (Figure 1), although the found to have peptidase activity intermediate between GluV8 protein expression of it has not been achieved yet (Kuroda and GluSE from Staphylococcus epidermis and to be some­ et aI., 2005). what less specific in its substrate requirements than other Drapeau (1978) found that maturation of GluV8 family members. The 185-195 region was found to contrib­ is mediated by a family enzyme, aureolysin, that ute to the activity of GluScpr, although other regions of the exhibits a preference for hydrophobic amino acids in the PI' enzyme must also playa role in defining the activity. Our position of its substrates. In accordance with this finding, results strongly indicate the importance of the N-terminal and recombinant GluV8 expressed in can be the 185-195 region in the activity of the glutamyl endopep­ efficiently processed into the mature form by thermolysin tidases of staphylococci. treatment in vitro (Nemoto et aI., 2008). Because V is com­ Keywords: processing; Staphylococcus aureus; monly located at the N-termini of all mature enzymes studied Staphylococcus cap rae; Staphylococcus cohnii; thus far (Ono et aI. , 2008; Figure 1), the thermolysin family Staphylococcus epidermidis; thermolysin. enzymes could be responsible for processing of all staphy­ lococcal glutamyl endopeptidases. In addition to a ro le of the pro-mature border, the protonated a -amino group of VI was Introduction reported to be important for the recognition of a substrate peptide (Prasad et aI. , 2004). However, if these two are the Staphylococcus aureus, a frequent pathogen in dis­ entire roles of VI, any hydrophobic amino acids (L, Y, I, ease, produces extracellular proteases, which are regarded as and F) and A can be located at the N-terminus of mature 1222 T. Ono et al.

A -68 -60 pre -50 -40 -30 pro -20 -10 -1 I. I ~ ' ~ I I . ' S . aureus MKGKFLKIfSSLFVATLTTJI.TLVSSPAANALSSKAHD.-HPQQTQSSKQQTPKIQ-KGG LKPLEQREHAN S . epidermidis !'1KKRFLSICTNTlAALATTTMVN--TSYAKTDTESHNHSSLGTENKNVLDINS-SSHNIKPSQ_ KSYPS S . warner i MKVKFFTASSLLIATLTS.lI.TLIN-- PAIiAETTSSTDNHQQTTQSQQQKTPKI D-KGNNVKPVEKKERAN S . saprop.IJyticus !'ILKRKLRGGLIFIIALLCTVISINGNTYAATQTHQSTDQSAQSEDJI.KVG------TNQIKTR S . caprae MRKFFLSIGFITIAILATTVLLN--ISQASEKSTPQADSMHQASHKKVTNDNGKANKKVK?LGSRKFPS S . cohnii MLKNILVS- I I IGTLLFLSSAINA-EVQAQNQNDnlKSSQQANESKQVG------TATPR * 10 20 30 40 50 60 70 I I I I S . aureus VILPNNPRHOITDTT_'GHYAPVTYIQVEAPTG--TFIASGVVVGKDTLLTN~'vVDATHGDPHJI. LK .ZI.FPSAI S . epidermidis VILPNN 'RHOIFKTTQGHYDAVSFIYIPIDGG--YI~SGSGVVVGENEILTN H,TVNGAKGNPRNISVHPSAK S . warneri VILPNNDRHOINDTTLGHYAPVTFVQVQSNEG--TFIASGVVVGKDKLLTN HrvDATHGKPRALKAFPSAV S . saprop.IJyticus WLPNDDRTQIE. TTNGHYQSVGYISIGDNIA------TGVVIDKNTVLTN H,T.l'NLSEGN---HNFSPJI.AQ S. caprae VI LPNNNRHQI SNTQEGHYAPVS FVK IYNDSGDLVASGSG IVVGEKE I LTN H IVDAJI.HN . PDNLGVAPSSK S . cohnii VNLPGENRHQIENTQSSHYQS IGY I EKGETIA------TGVVVGKNTVLTN TVNNGDGK --- LS FAP.ZI.AE * ** * ** * ** * * ***** * 80 90 100 110 120 130 140 I I I I I I I S . aureus QDNY PNGGFTAEQ ITKY SGEIlAI VKFS PNEQNKH I GEVVKPAT~·IS -NAETQVNQ I TVTGYPGlllill S . epidermidis ENDYP. GKFVGQEIIPYPGN AILRVSPNEIiNQIiIGQWKPATISSNTDT!UNEXITVTGYPG.QlS£1. S . warneri NQNNYP GGFTGEQITKYPGN~ IVKFSPNNQNQNIGEWTPATLSDNADTSQNQPITVTGYPGDKPL S . saprop.IJyticus NENTI·\PYGTFSEKEIEVYPGN ALlIiLNKNKDEQSVGD'NQPATLKDASAVTKDI1PITVTGYPGDKSL S. caprae SESEHPNGEFKGQEIKKYPGN SILKVQPNEDNKKIGDVVTPAKISSNSNTQTDQDITVTGYPG.QlS£1. S. cohnii . EKSFPYGTFTEKE\NAYPGD VLVHFNKNNEGQSVGDWTPATIGETSTVSSNDPITVTGYPGDKPL *'1'"******** * * * '* * * *: * 150 160 170 180 190 200 210 I I I I ! U I I I S . aUl.-eus ~SKGKITYLKGEAJolQYDLSTTGGISPVFNEKNEVIGIHNGGVPNEFNGJI'VFINENVR.!ti:I.lillNI S . epidel.-midis ATWilESVGKVVYIGGEELRYDLSTVGG S..,SPVFNGKNQVIGIHYGGVDNKYNSSVYINDFVQQFLRNNI S . ;Iarnel.-i .~TI1WESRGKITQIQGEDJ.mY DLSTTGG S -,S PVFNS RNEVI G I HvlGGAANSYNGAVFINN. 'VQNFLKQN I S . saprop.IJyticus .~TI1WESKGQI LNTNTTEFEYNJI.STFGG. S..,S PVFNEKNEVI GI HQGG I EGESNS.lI.VAl'ITDDVLS FIKKNK S . caprae .~TI-1WESKGKILSINGKQI·1TYDLSTYGG S SPVFNENNEVIGIHYGGVEHKIiNSJI.VFINEDVQKFLKQNI S . cohnii ATWilESSGKVI SNNDS I LTYSJI.GTFGG - S PVFNAKNELI GLHYGG I EGESNNAVAITGEVLQFI KEHX *****w "* *: *Jt*******1r * ** * ** * 220 230 240 250 260 268

S . aureus ~AKDDQPNKPD -PDNPNKPDNPNKPDEPNNPDKPNKPDNPD. GDNNNSDN?DA.~ S . epidermidis PDINIQ S . warnel.-i EDIHFSNSD--N DDN------DKNNGTDDNNN----NN-----KDDDNNYDNPDA .~ S.saprop.IJyticus S S . caprae PSIQINEKN- TNLEKSHD------GN ·NYKKA.lI. S. cohnii V 0.0979 B GluSW (S. warneri) 0.1714

'----- GluV8 (S. aureus) 0.1340 rO._28_2_5______GluScoh (S. cohnii )

0.2760 0.0224 GluSsap (S. saprophyticus) 0.2288

F::..;..:'------GluScpr (S. caprae)

'--__------GluSE (S. epidermidis)

Figure 1 Alignment of the amino acid sequences of staphylococcal glutamyl endopeptidases. (A) The deduced amino acid sequences of glutamyl endopeptidases from S. aureus (Carmona and Gray, 1987), S. epidermidis (Ohara­ Nemoto et al.. 2002), S. warn.eri (Yokoi et aI. , 2001 ), S. saprophylicus (Kuroda et aI. , 2005), S. caprae and S. cohn.ii are aligned. Hyphens are introduced for maximal matching. Identi cal amino acids among the six endopeptidases are indicated by asteri sks. pre, preseqllence; pro, prosequence. Amino acids forming the of serine proteases are boxed. Arrows indicate the amino acids at positions 185, 188 , and 189, which define the proteolyti c activities of GluV8, GluSE, and GluSW (Nemoto et aI. , 2009). Amino acid sequences taken into consideration for the synthesis of the degenerate primers are underlined. (B) Phylogenic tree of the six staphylococcal glutamyl en do pep­ tidases prepared by comparison with V I-A2 16 of Gill V8 and its equi valent regions of other proteases. The tree was generated by a di stant­ cased method using a neighbor-joining algorithm. Numbers of the branches represent length between two taxa. Proteolytic activity of staphylococcal glutamyl endopeptidases 1223

enzymes. To address this issue, we expressed GluV8 with ATG typical start codon did not exist at the region equivalent the V I L substitution by site-directed mutagenesis and stud­ to that of GluV8; however, ITG instead of ATG was regard­ ied the processing of proGluV8 and the proteolytic activity. ed as the start codon, because (i) ITG and GTG are able to In addition to VI residue, we recently defined three amino function as the start codon in , (ii) the Shine-Dal­ acids at positions 185, 188, and 189 (as the numbers of garno sequence (Shine and Dalgarno, 1975) existed at the mature GluV8) involved in the alteration of the proteolytic position equivalent to that of other staphylococcal glutamyl activities of GluV8, GluSE, and GluSW by modification of endopeptidases (data not shown), and (iii) ITG is also used their Kill values (Nemoto et aI. , 2009). In combination of as the start codon of the glutamyl endopeptidase from B. W185, V188, and P189, GluV8 exhibited the smallest Kill licheniformis (Kakudo et aI., 1992). Therefore, we propose value. In the present study, we also aimed to elucidate wheth­ that the open reading frame of the GluScoh gene is com­ er these three amino acids are generally major determinants posed of 771 bp encoding a 256-amino acid protein. This of the proteolytic activities of staphylococcal glutamyl endo­ number is the smallest among the six staphylococcal gluta­ peptidases. Toward this goal, for the first time, we cloned my I endopeptidases reported to date. the homologous from Staphylococcus caprae (desig­ The deduced amino acid sequences of GluScpr and nated GluScpr) and Staphylococcus cohnii (designated GluScoh were aligned together with those of the four other GluScoh), and then compared their protease activities in staphylococcal glutamyl endopeptidases (Figure 1). The pre-, detail with those of GluV8 and GluSE. pro-, and mature regions of GluScpr were predicted to pos­ sess 27, 40, and 238 amino acids, respectively, and those of GluScoh possessed 27, 25, and 204 amino acids, respective­ Results ly. A 15-amino acid deletion was present in the pro sequence of GluScoh starting at the position identical to the 13-amino Cloning of the genes encoding the glutamyl acid deletion of glutamyl endopeptidase from S. saprophy­ endopeptidases from S. caprae and S. cohnii tic us (tentatively designated GluSsap). The amino acid iden­ tity among the proteases was maximal at the mature region; The genes encoding glutamyl endopeptidases from S. caprae and the identity of GluScpr and GluScoh to that of GluV8 and S. cohnii were cloned by performing successive peR in was 60.1 % and 44.9%, respectively. By contrast, amino acid combination with a pair of degenerate primers and gene­ sequences were least conserved at the prosequence region: walking techniques (Table 1). The resu ltant 4042-bp EcoRV the identity of the prosequences of GluScpr and GluScoh to DNA fragment from S. caprae carried the open reading that of GluV8 was 17.5% and 18.0%, respectively, and the frame of the GluScpr gene composed of 918 bp encoding a identity of presequences of GluScpr and GluScoh to that of 305-amino acid protein. A 1952-bp DNA fragment that GluV8 was 27.6% and 24. 1%, respectively, and only three potentially encoded a glutamyl endopeptidase GluScoh was amino acid residues, M-68 (first M), L-53, and A-40, which cloned from S. cohnii. In the putative GluScoh gene, the is a putative recognition site by , were con-

Table 1 Primers used for degenerate PCR, cloning, and expression.

Primer Sequence

Degenerate PCR 5VSdegVl 5'-GT(c/t)ATA(c/t)TACCCalt)AA(c /t) (alg)AT(alc)G(alt)CA(c/t)CAAAT-3' 3VSdegF225 5'-A(alt)AT(g/t)(g/alt)ATATCCalt)(g/t)(g/c)(alt)ATATT(g/t)T(g/t)T(c/t)TTA(alg)GAA-3' 3VSdegSl46 5' -ACTTTCCCA CAT( t/a)GT( t/al g)GCC t/a)A( clg/a) ( alt )GGTTT( alg)TC-3'

Cloning 3ScprGSPI 5' -ATCTGGGTTATT ATGTGCGGCATCAAC-3 ' 5ScprGSP2 5' -GGAGATAAACCTTTAGCAACAATGTGG-3' 5ScprEco 5' -GATATCCCCTCGACTTAACACTCACAA-3' 3ScprEco 5' -GATA TCTGCCGTCGCACGTCCGTCTTG-3' 3ScohGSPI 5'-GCCGATGGATTGGTAATGACTAGATTG-3' 5ScohGSP2 5' -CCTGCCACAATTGGTGAAACTAGTACT-3' 5ScohPvu 5'-CAGCTGATCGAAAATGTTGACTTAG-3' 3ScohDra 5'-TTTAAATCAATCGCTCAATATGGTT-3'

Expression 5GluScprM-6SBgl 5' -GGTAGATCTATGAGAAAATTTTTTTTATCTA-3' 3Glu ScprA237Bgi 5' -ACTAGA TCTTGCTGCTTTTTTATAATTG-3' 5GluScohL-67Bam 5' -ACAGGATCCAAC ATTGTGTTCATTGATAAA- 3' 3GluScohV211Bam 5' -ACAGGATCCA ACATTGTGTTCATTGATAAA-3 ' Small letters in parentheses represent mixtures of bases . Underlined sequences represent the endonuclease restric­ tion site. 1224 T. Ono et al.

served out of 52-68 residues of the preprosequences among at the N-l/S-I-VI bond (Ohara-Nemoto et aI. , 2008; Ono et six endopeptidases (Figure I). aI. , 2008). To investigate the role of VI, K residue was intro­ In the mature region, 69 amino acids were conserved. duced at -2 and -3 positions of GluV8 molecule (designated These include the catalytic triad (H5 I, D93, and S 169) of GluV8 P-2K and H-3K, respectively) (Table 2). After ther­ family serine proteases and HI84 (Figure 1). molysin treatment, 32-kDa proGluV8 converted to 28-kDa The replacement of H184 is reported to drastically reduce species, of which N-terminus was VI. These molecules the kca, value of the endopeptidase from B. intermedius exhibited the full activity as the case of non-mutated GluV8 (Demidyuk et aI. , 2004). Continuous residues at 129- 148 entity (Figure 2A). Although treatment resulted in the and 166-175 are completely identical among the six endo­ conversion to 28-kDa proteases, the N-termini of these peptidases. In the latter segment, conserved G166, G167, or GluV8, GluV8 P-2K, and GluV8 H-3K species were S-4, G 170, most likely G 167, could participate in forming an S-I, and A-2, respectively, and they did not show any pro­ of family (Tamada et aI., teolytic activity. These results clearly demonstrated that 2009). attachment of even one residue, S-1 in this case, to VI com­ The phylogenic analysis on the mature proteases without pletely abolished the Glu-specific proteolytic activity, and the C-terminal repeat demonstrated the kinship between thus suggested that V residue at the N-terminus of mature GluV8 and GluSW, that between GluScoh and GluSsap, and GluV8 was required to achieve the proteolytic activity. that between GluScpr and GluSE (Figure 1). The kinship We then investigated whether V I can be compensated by between GluV8 and GluSW was further ascertained by the other hydrophobic amino acids. Here, L was selected for presence of the related hydrophilic segments at the C-ter­ substitution, because L residue is an excellent target for minal region, although its role remains unknown. The kin­ aureolysin as well as thermolysin (Drapeau, 1978; Stennicke ship between GluScoh and GluSsap was further indicated by and Breddam, 1998). GluV8 VIL mutants carTying a trypsin­ their 13-15-amino acid deletions in the proregion and the cleavage site at positions -I, -2, and -3 (designated GluV8 identical C-terminal termination. Despite the kinship S-IKIVIL, P-2K1V IL, and H-3K1VIL, respectively) were between GluScpr and GluSE, GluScpr contained 20-amino expressed. The treatment of these V lL mutants with ther­ acid hydrophilic tag beyond the C-terrrunus of GluSE. This molysin as well as trypsin again produced the 28-kDa forms. phylogenicity of staphylococcal glutamyl endopeptidases Unexpectedly, the 28-kDa species of VIL mutants processed was basically identical to that calculated by using 16S rRNA by thermolysin were a mixture of molecules with the or the hsp60 gene (Ghebremedhin et aI. , 2008). N-termini of Ll and 12 (Table 2 and Figure 2B), indicating that the Ll-I2 bond became hydrolyzed. This 28-kDa species Role of the N-terminal amino acid for mature GluV8 showed 5%-10% of the activity of GluV8 (Figure 2). After I is completely conserved among the six staphylo­ trypsin treatment, the peptide bonds on the C-terminal side coccal endopeptidases (Figure I) and GluV8, GluSE, and of K or R residues were always processed. Among them, the GluSW acquire the enzymatic activity through the cleavage S-IKiVIL mutant, of which N-terminus was L, exerted the

Table 2 Amino acid sequence of GluV8 deri vatives and their N-terminal residues after processing.

GluV8 species Amino acid sequence and N-terminal residue after putative cleavage state cleavage with

Thermolysin Trypsin

-3 -I I GluV8' --- E Q R E HA N t VI L P N--- mutSb --- Q Q RT S H PS t VI L P N--- VI S-4 P-2K --- Q Q RT S H KT S t V 1 L P N--- VI S-I H-3K --- Q Q R· S K· /:,. S t V 1 L P N--- VI A-2 VIL --- Q Q RT S H P Sa- 1 L P N--- Ll , I2c S-4 S-IK/VIL --- Q Q R· S H P KT t L 1 L P N--- Ll « 12c Ll P-2K/VIL --- Q Q RT S H KT Sa-l L P N--- Ll < 12c S- l H-3K/V IL --- Q Q RT S KT /:,. S t L 1 L P N--- n.d. n.d. "Carmona and Gray (1987). bFive substitutions (D-33H, E-7Q, E-4S, A-2P, and N- IS) were introduced into wild type GluV8 (Ono et aI. , 2008). cA mi xtu re of two polypeptides starting with L or 1. Amino acid substitutions introduced into prosequence of GluV8 are underlined. Arrows and arrowheads indicate potential processing sites mediated by thermolysin and trypsin , respectively. n.d., not determined. Proteolytic activity of staphylococcal glutamyl endopeptidases 1225

A32j·_------_--~~------~I~' 28 1 ------~ ----- C:1 150

~ 100 J?: '> U 50

0 M n n n Th - + - - + - - + - + + - + - + Tryp - - + - - + - - + + --- + + + GluV8 K-2 K-3 L1 K-1L1 K-2L 1 K-3L1

Thermolysin P-2K 10

'c 8 S-1 :::J 6 ~ I jg 4 :0 4: 2 , I I o '\.\,I,I,\lo! \ .. ~...... \I....., """"VI..::i PJ\.\.~',:"",,-"")"\,J\...I,,-~ -,...... j-....A"--:v...w.- ......

V1L 10 I i j~"~ :~f i'!!~~cc~.,~~~_ J 0 6 9 12 15 18 21 6 9 12 15 18 21 Retention time Retention time

Figure 2 Processing and proteolyti c activities of GluV8 and its derivatives. GluV8 and its derivatives were incubated at O°C without protease or at 37°C with 0.3 f.lg of thermolysin (Th) or trypsin (Tryp), as described in the materials and methods section. (A) Proteins (l f.lg) were separated by SDS-PAGE and stained with CBB. The apparent molecular masses of major products are shown on the left. The proteolytic acti vities of the samples were determined with LLE-MCA and expressed as percent±SD (n=3) of the activity (4700±60 Uf.lg-') of thermolysin-processed GluV8. (B) GluV8 P-2K and GluV8 VIL treated with either thermolysin or trypsin were subjected to the N-terminal sequencing by Edman degradation. Phenylthiohydantoin-amino acids at the first cycle were subj ected to reversed phase liquid chromatography. activity (31.8±0.9%, mean±SD, n= 3). In contrast, trypsin­ mature GluV8 certainly played important roles, which were treated 28-kDa GluV8 P-2K!VIL and H-3K!VIL exhibited distinct from the protection of VI from the thermolysin no activity due to amino acids optionally attached to Ll. cleavage. Therefore, these results suggested at least two roles of VI residue in GluV8: V is more preferable to L for the precise Expression and processing of recombinant GluScpr processing event and is directly involved in the proteolytic and GluScoh activity. The findings described above demonstrated that the V 1- An attempt to express the wild type GluScpr in E. coli did 12 bond was protected from thermolysin cleavage in GluV8. not succeed, suggesting the toxicity of this protease due to The three-dimensional structure of mature GluV8 shows that the conserved TI64 and N193 are located adjacent to VI autoproteolysis, as reported for GluV8 (Nemoto et a!., 2008). «2 A) and are in contact with the a-amino group of V I To suppress the autoproteolysis, we introduced substitutions through hydrogen bonding (Prasad et a!., 2004). Accordingly, of D-18 to K (D-18K) and D-32 to H (D-32H) into the pro­ we speculated that the interaction of TI64 and NI93 with sequence of GluScpr (designated GluScpr-mut). Consequent­ VI protects the VI-I2 bond from thermolysin cleavage. ly, several bands ranging at 30- 38 kDa were recovered GluV8 mutants with either TI64A or NI93A and Tl64AI (Figure 3). Thermolysin treatment induced the conversion NI93A substitutions were expressed to test this assumption. into two bands migrating at 29 and 30 kDa accompanied Although thermolysin induced the degradation from the 33- with an acquisition of the protease activity. The N-tenninal kDa form to the 28-kDa form, none of the mutants possessed amino acid of the 38-kDa form of GluScpr-mut was A-32, any proteolytic activities. N-tenninal sequencing revealed suggesting the cleavage at the Q-33-A-32 bond (Table 3). that their N-terminal amino acids were V I in all cases (data The 35-kDa species was a mi xture of two polypeptides start­ not shown). Therefore, TI64 and N 193 adjacent to V 1 in ing at A-32 and K-22. Valine I was the N-terminus of both 1226 T. Ono et al.

A GluScpr-mut B GluScoh °C 0 37 37 °C 0 37 37 Th ~~~ __~~~~~~~~~k Da Th ~---=--====~;;;~:Jd..-...., kDa 97 97 I •- 67 67 iii 45 15 - _30* 33- ~~- - • ==-=-- -29 29- --- -=--- - ~ 30

20.1 20.1 14.3 - 14.3 2 3 4 5 6 7 8 9 -M 1 234 5 6 7 89M C D kDa 45 kDa 33 -+ 30 ------.., ~ .. 30 29 -+ , -----I 2 3 4 5 6 7 8 9 M 2 3 4 5 6 7 8 9 M E F 150 100 c c ".Q "0 75 1:3 t5 ro 100 ro .l= .l= 2- 2- 50 .:;;~ 50 .:;;~ n n 25 -0: ' n -0: 0 n 0 .... n I 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9

Figure 3 Thermolysin processing and protease activities of GluScpr and GluScoh. Purified GluScpr-mut (A, C, E) and GluScoh (B, D, F) at 10 fLg were incubated for 4 h without thermolysin (Th) at O°C (lane I), at 37°C (lane 2) or at 37°C with 0.001 (lane 3), 0.003 (lane 4), 0.01 (lane 5), 0.03 (lane 6), 0.1 (lane 7), 0.3 (lane 8) or 1 fLg (lane 9) of thermolysin. Aliquots (l fLg) of the samples were separated by SDS-PAGE and stained with CBB (A and B) or subjected to zymography (C and D). Asterisks show the position of thermolysin. The apparent molecular masses of the major bands are indicated. (E and F) The proteolytic activities (means±SD, n=3) of the samples were determined as described in the materials and methods section. Samples for columns 1-9 in panels (E) and (F) are identical to those in lanes 1-9 in panels (A) and (B), respectively.

29- and 30-kDa species upon thermolysin treatment. Because terminus of the 33-kDa GluScoh was Q-39, indicating the GluScpr possessed two E residues at positions 217 and 224 cleavage at the A-40-Q-39 bond by signal peptidase. The (Figure 1), which were adjacent to the C-terminal position N-terminus of 29-kDa GluScoh was unable to be determined of GluSE (Q216), we suppose that the cleavage at either the because of its limited production. E217-K218 or E223- K224 bond of the 30-kDa species pro­ Following the thermolysin treatment, the recoveries of the duced the 29-kDa species. mature forms significantly differed among recombinant pro­ On the expression of the full-length form of GluScoh, a teins, and the band intensities of mature GluScpr and 33-kDa proenzyme was obtained. When 33-kDa GluScoh GluScoh were 33 .8% and 2.4% of the inputs, respectively was treated with thermolysin, a faint 29-kDa band showing (Figure 4). These figures were much lower than those of the proteolytic activity was recovered (Figure 3). The N- GluV8 (9l.9%) and GluSE (82.5%). Taken this breakdown

Table 3 N-terminal sequences of recombinant GluScpr and GluScoh.

Species kDa Thermolysin Detected amino acids Determined sequence

GluScpr-mut 38 AHSMHQA Q-33/A-32HSMHQA-26 35" AHXM Q-33/A-32 BSM-29 KYTN K-23/K-22 YTN-19 30 + YILPNNN S- IIYJ ILPNNN7 29 + YILP S-JIYI ILP4 GluScoh 33 QNQNDT A-40IQ-39NQNDT-34 aA mixture of two polypeptides. X, not identified. The amino acid substituted in GluScpr-mut (D-3IH) is underlined. Slashes represent cleavage sites. Proteolytic activity of staphylococcal glutamyl endopeptidases 1227

A kDa B kDa

97 97 67 67 46 46 -- iii- ----=.-, - 30 30 20.1 20.1 14.3 14.3 1 2 3 4 5 6 7 8 9 10 M 1 2 3 4 5 6 7 8 9 10 M Th - + - + + + - + Th - + - + - + - + - + D C 150 4000 c ·iii - ~ 3000 C e ·en 100 0- C <;" 2 OJ .s ::i 2000 OJ 2- > 50 :g .c;; 1000 "iii 0:: ~

Figure 4 Comparison of the proteolytic activities of staphylococcal glutamyl endopeptidases. GluV8 (lanes I and 2), GluSE (lanes 3 and 4), GluSW (lanes 5 and 6), GluScpr-I11ut (lanes 7 and 8), and GluScoh (lanes 9 and 10) at 10 J.Lg were incubated at O°C without thermolysin (Th) or at 37°C with 0.3 J.Lg of thermolysin. Aliquots (I J.Lg) of the samples were separated by SDS-PAGE. Separated protein s were stained with CBB (A) or subjected to zymography (B). (C) The amounts of unprocessed and thermolysin-processed 28-29-kDa bands for GluV8, GluSE, and GluScoh, the 35-kDa band for GluSW, and the sum of 28- and 30-kDa bands for GluScpr-mut were quantified by PDQuest. (D) The proteolytic activities of thermolysin-treated enzymes toward 20 J.LM LLE-MCA were measured, and the specific activities (means±SD, n=3) were calculated in consideration of their recoveries determined by PDQuest. Columns I-to in panel (C) and 2-8 in (D) correspond to lanes 1- 10 of panel (A). in calculation, the specific activity of GluScpr-mut was QAR-MCA to some extent, whereas GluScoh as well as defined to be 1073±2 U//-Lg protein, which corresponded to GluV8 hardly hydrolyzed these substrates. The proteolytic 30.4% of that of GluV8, but was II-fold higher than that of activities of GluScpr and GluScoh were most effectively sup­ GluSE. GluScpr carried Y185, V188, and El89 and GluScoh pressed by diisopropyl fluorophosphate (DFP), followed by carried Y185 , 1188, and E189, and thus these proteases were 4-(2-aminoethyl)benzenesulfonyl fluoride (AEBSF) (Figure expected to possess activities lower than GluV8 (WI85, 5B). Tosyl-L- chloromethyl ketone (TPCK) V188, and P189) and comparable to GluSE (Y185 , V188, (0 .1 mg/m l), tosyl-L-Iysine chloromethyl ketone (TLCK) and D 189). We did not calculate the specific activity of (50 /-Lg/ml) , phenylmethylsulfonyl fluoride (PMSF) (l 1111\1) , GluScoh, because reliable data could not be expected due to (0.2 U/ml), (0.1 mg/rnl), pepstatin its intrinsic instability. (0.01 mg/ml), and dithiothreitol (5 111M) did not inhibit the These observations indicated protease susceptibility of activities of GluScpr and GluScoh. Recombinant GluV8, GluScoh. Then, this property was compared among the GluScpr, and GluScoh possessed an identical pH optimum. GluV8 family members by using constitutive-inactive S 169A Exceptionally, a slight shift to basic pH was observed on mutants. As a result, GluScoh S169A was completely GluSE (Figure 6C). degraded by both thermolysin and GluV8 treatments, where­ as GluV8, GluSE, and GluSW mutants remained undegraded Effect of amino acid substitutions at position 185 (Figure 5). Therefore, these results indicated the intrinsic and 189 on the activity of GluScpr structural instability of GluScoh. We previously reported that three amino acids at positions Characteristics of proteolytic activities of GluScpr 185, 188, and 189 are involved in the determination of the and GluScoh proteolytic activity of glutamyl endopeptidases (Nemoto et aI., 2009). Because V 188 was conserved in GluScpr, Glu V8, Substrate specificity was examined in GluScpr and GluScoh. and GluSE, here we focused on the remaining two positions Both proteases efficiently hydrolyzed LLE-MCA, but scarce­ at 185 and 189. The 33 -35 kDa species of GluScpr-WP, ly cleaved AE-MCA (Figure 6A). Among other substrates which mimicked the combination of GluV8, was scarcely tested, GluScpr hydrolyzed YVAD-MCA, AAN-MCA, and detected in the lysate; and the band was detected by immuno- 1228 T. Ono et al.

kDar-______-. cells increased after 3 h and finally overcame the growth of 97 _ GluScpr-mut-producing cells, this increase did not accom­ pany the production of GluScpr (data not shown).

Effect of amino acid substit utions at residues -- 190-195 of GluSE to those of GluScpr on the activity 3D _ -- --- We next investigated the reason why the specific activity of GluScpr was higher than that of GluSE. We previously 20.1 ~ ______~ reported that four amino acid substitutions of GluSE to those M1234123412341234 of GluV8 (KI91E, Y192F, S194G, and S195A) induced a 2- GluV8 GluSE GluSW GluScoh fold increase in the proteolytic activity (Nemoto et aI. , 2009). S169A S169A S169A S169A By analogy, we suspected that amino acid substitutions at residues 190-195 were involved in the enhancement of the Figure 5 Protease sensitivity of GluScoh. protease activity of GluScpr from that of GluSE. The sub­ The S 169A mutation was introduced to GluV8, GluSE, GluSW, and stitutions of N190H, Yl92H, and S195A were introduced GluScoh. Proteases (10 J.L g) were incubated at O°C (lane I) or 37°C (lane 2) without thermolysin or at 37°C with 0.3 J.Lg of thermolysin into GluSE (designated GluSE-HHA), and the mutated pro­ (lane 3) or thermolysin-acti vated GluV8 (lane 4) . After 4 h, aliqllots tease was expressed (Figure 7 A). The specific acti vity of (I J.L g) of the samples were separated by SDS-PAGE. GluSE-HHA was 4.5-fold increased from that of GluSE, although this activity was only 26.2% of that of GluScpr­ mut. This difference was consistently observed at the sub­ blotting only after Talon-affinity purification (data not strate concentrations up to 0.2 mM (data not shown). shown). The growth rates of E. coli cells were investigated GluSE-HHA most potently hydrolyzed LLE-MCA and after induction of GluScpr, GluScpr-mut, and GluScpr-WP scarcely cleaved the remaining MCA substrates. (Figure 7B). E. coli expressing GluScpr-mut grew fastest fol­ lowed by those with GluScpr, and cells expressing GluScpr­ WP did not grow under the induction conditions. These Discussion findings strongly suggested that the heavy toxicity against the host s was caused by an enhanced protease activity To date, four members of the glutamyl endopeptidases have of GluScpr-WP. Although the growth of GluScpr-producing been identified in staphylococci: one in S. aureus and three

A 20 B 125

100 15 ~ ~ 75 ~ 10 .:;~ :~ 13 U 50

0 0 GluV8 GluScpr-mut GluScoh GluScpr-mut GluScoh C 100 rti 75 ~ /II / 1 .:;~ 50 /.1 l n Iii' I . \\

Figure 6 Proteolytic characteristics of GlliScpr and GlliScoh. (A) Protease activities (mean±SD, n=3) of GlllV8, GlliScpr-mllt, and GluScoh for AE-MCA (open), YVAD-MCA (c losed), AAN-MCA (hatched), and QAR-MCA (hyphened) were compared with respective acti vities for LLE-MCA (100 %). (8) Thermolys in-treated GluScpr­ mut and GluScoh were preincubated without (open) and with 3 mM DFP (c losed) or I mM AE8SF (hatched) at O°e. After 30 min , the reaction was started by an add ition of 20 J.LM LLE-MCA. (C) Proteolytic activities of GluV8 (open circle); GlliSE (closed circle); GlliScpr (open square); and GluScoh (closed sq uare) at pH 4- 10.5 were determined with triplicate samples in 50 mM buffers of Na-acetate (pH 4-5.5); Na-phosphate (pH 5.5-8); Tris-HCI (pH 8-9); and glyci ne-NaOH (pH 9-10.5). Proteolytic activity of staphylococcal g lutamyl endopeptidases 1229

A 180 190 200 GI uVS " FINE GIuS;: YINQ GIuSN "FINN GluSsap VIGIHQGGI vAHT.12 GluScpr VIGIHYG FINE Gluscoh AI TG

GluSE-" lP' VIGIHWGGV PN ~N ~V YI N Q GluSE-EFGA .I. VI GIH.XGGV ..I2N EFNGAVXIN.12 GluScp r-'A1P VIGIH WGGV P H li HN~AVF ! NE " GIuSE - HH .~ VIGIH YGGVQ HK HN~,A VY INQ B c 0.4 2500 c ·iD 2000 0.3 e 0.

'0) 1500 g 0.2 E c:i c:i Time (h)

Figure 7 Effect of amino acid substitutions in staphylococcal glutamyl endopeptidases. (A) Amino acids sequences at residues IS0-200 are compared among the six staphylococcal glutamyl endopeptidases. Amino acids char­ acteristic of GluVS are represented by bold letters and those of GluSE are underlined. Asterisks and arrowheads show amino acids substituted in GluScpr-WP and GluSE-HHA, respectively. Residues 190- 195 of GluSE, which are potentially involved in the alteration in the proteolytic activity of GluScpr, are boxed. aGluSE-WP and GluSE-EFGA are mutated GluSE, which mimicked residues ISO-190 and 190-195 of GluVS, respectively (Nemoto et aI. , 2009). (B) Growth curves of E. coli carrying the expression plasmids for GluScpr (open circles), GluScpr-mut (filled circles), and GluScpl~WP (open squares) in Luria-Bertani broth in the presence of 0.2 111M isopropyl [3-D-thiogalacto­ pyranoside containing 50 /-Lg/ml of ampicillin. The bacteria were cultured at 30c e with vigorous shaking. (C) Proteolytic activities (mean±SD, n=3) of aliquots (0.25 /-Lg) of GluSE, GluScpr, and GluSE-HHA (10 /-Lg) preincubated at oce without thermolysin (Th) or at 37 ce with 0.1 or 0.3 /-Lg of thermolysin for 4 h were determined. in -negative staphylococci, i.e., S. epidermidis, S. between V and L, these results strongly suggested that the warneri, and S. saprophyticus. The present study demonstrat­ residual group of VI is required for the enzyme activity ed the existence of two additional GluV8 family members in itself. coagulase-negative staphylococci S. cap rae and S. cohnii. The N-terminal tripeptide of staphylococcal glutamyl These bacteria occasionally cause urinary tract infection and endopeptidases is V I-(I/V /M)2-L3 (Figure 1), implicating bacterial endocarditis (Kloos and Bannerman, 1994). The that the bond between V 1 and (I/V IM)2, as well as (I/V IM)2 number of amino acids comprising GluScoh was the smallest and L3 could be cleaved by thermolysin. Although this pos­ among the six endopeptidases; 15 amino acids were deleted sibility had not been consciously recognized, the present in the prosequence, and the typical ATG codon did not exist. study demonstrated that V I residue prevented both cleav­ However, the E. coli expression showed that the putative ages. We had speculated that the Tl64 and N 193 being locat­ GluScoh gene was sufficient to produce a proteolytically ed sterically adjacent to V I ascertained the restricted active molecule. processing at the N-I-VI bond. However, GluV8 with Although the N-tenninus of GluSW was initially reported Tl64A and Nl93A substitutions was invariably processed at to be R-3 (Yokoi et aI., 2001), our study on recombinant the N- I-VI bond. Thus, although the mechanism that sup­ GluSW demonstrated that the species starting at V I was ports the precise processing remains unknown, we currently enzymatically active, whereas that staring at R-3 was inac­ suspect that hydrophobic interaction with the side chain of tive (Ono et aI. , 2008). Consequently, V I was maintained in VI could be important for the maturation processing of the all family members. The present study demonstrated that VI GluV8 fmnily. was required for lim.ited between the S-I-V I GluV8 family of enzymes including GluScpr and GluScoh bond, because the Ll-I2 bond was additionally cleaved in most potently hydrolyzed a Glu-containing peptide substrate, the V IL mutant. Moreover, mature GluV8 with the N-ter­ i.e., LLE-MCA. Noticeably, GluScpr hydrolyzed QAR-MCA minus of L exerted 32% of the activity of that harboring V I as well as YVAD-MCA and AAN-MCA to some extent, (Figure 2). Substitutions from V to F, G, S or A completely indicating its broader substrate specificity. The PI acceptance abrogated the proteolytic activity (Nemoto et aI. , 2008). of R was also observed with GluScoh to a lesser extent (Fig­ Therefore, taken into consideration the structural similarity ure SA). In addition, it was reported that commercially sup- 1230 T. Ono et a l.

plied GluV8 exhibited a subtle hydrolyzing activity toward were purchased from Genenet (Fukuoka, Japan) and Greiner Bio­ QAR-MCA (Wildeboer et aI. , 2009). These findings suggest One GmbH (Fri ckenhausen, Germany) . that GluV8 fam il y members could accept positively charged am ino acids at PI positi on, and that this property could have Bacterial strains and culture conditions some significance under certain conditions. Among protease inhibitors tested, irreversible serine pro­ S. cap rae GTC 37S, S. cohnii subsp. cohnii GTC 24S, S. aureus tease inhibitors, i.e., DFP and AEBSF, efficiently suppressed ATCC 2S923, S. epidermidis ATCC 14990, and S. \IIarneri JCM the protease activities of GluScpr an d GluScoh. Other inhib­ 241S were used for this study. Staphylococci were cultured in Todd­ itors that are reported to be effective on serine proteases, Hewitt broth (Becton Dickinson, Sparks, MD, USA). For the expression of recombinant proteases, E. coli XL I Blue was used. such as PMSF, leupeptin, aprotinin, TPLK, and TLCK did not suppress their activities. Hence, the unique inhibitory effect of DFP on GluSE (Ohara-Nemoto et aI. , 2002) was PCR cloning of the genes encoding glutamyl common to the staphylococcal endopeptidases. endopeptidases from S. caprae and S. cohnii In addition to VI and the catalytic triad, the three amino Degenerate S'- and 3'-0Iigonucleotide primers (SVSdegV I and acid residues at positions 185, 188 , and 189 dictate the pro­ 3VSdegF22S) encoding V i-lle ll and F22S-A214 ofGluVS, respec­ teolytic activity of GluV8, GiuSE, and GluSW (Nemoto et tively, were synthesized in consideration of the nucleotide homology aI. , 2009). These seven amino acid residues participate in the of GluVS, GluSE, and GluSW (Table I). Genome DNA of S. cap rae formation of the substrate-binding pocket of GluV8. From ( 100 ng) prepared as described previously (Ikeda et aI., 2004) was comparison of the amino acid sequ ences, we speculated that used as a template. Following an initial denaturation at 9S oC for the activity of GluScpr carrying Y185, V188, and E 189 IS min, SO cycles of 94°C for 20 s; SO°C for IS s, and n oc for should be much lower than that of GluV8 and comparable 3 min were performed with a HotStar PCR system (Qiagen Inc.). to GiuSE. In fact, the specific activity of GluScpr toward As a result, a 6S0-bp DNA fragment was ampli fied . The fragment LLE-MCA was one-third of that of GluV8, and the muta­ was in serted in pB luntTOPO (Invitrogen, Carlsbad, CA, USA) and tions of Y 185W and E l 89P mimicking the amino acids of sequenced. Based on the nucleotide sequence of the 6S0-bp frag­ GluV8 seemed to enh ance the protease activity. This study ment, two gene-specific primers (3ScprGSPI and SScprGSP2; Table further demonstrated that the acti vity of GluSE carrying tri­ I) were synthesized for amplification of the genome coupled in both directions with an adaptor primer (S'-GTA ATA CGA CTC ACT ple substitutions of N190H, Yl92H, and S195A (GluSE­ ATA GGG C-3'). Thereafter, 2-kb DNA fragments in both directi ons HHA) was 4.5-fold increased from the wild type, indicating were obtained fro m the EcoRV-digested genome DNA by using a th at the 190-195 region is also in volved in modulation of Genome Walker Uni versal kit (Clontech Laboratories Inc.). Again , the protease activity. However, because the activity of the primer set (SScprEco and 3ScprEco) was synthesized based on GluSE-HHA could reach approximately 30% of that of the sequence data from the two ends of the 2-kb DNA fragments GluScpr, regions in addition to amino acid residues 185- 195 for an overall amplification (Table I). Finally, a 4-kb genome DNA should be certainly involved in modul ation of the proteolytic of S. caprae was amplified and in serted into the Sma! site of activity. The three-dimensional structure of GluV8 revealed pUC IS. The nucleotide sequence of the 4042-bp in sert was that H184WGGVP189 and E191FNGA195 form an anti-par­ determined. all el l3-sheet structure (Prasad et aI. , 2004). The former Degenerate primers (SVSdegVI and 3VSdegS 146; Table I) cor­ l3 -strand takes part in the substrate-binding pocket together responding to amino acids at positions V I-Ill and D137-S 146 of with the catalytic triad and VI. In contrast, the latter l3-strand GluVS, respectively, were used for the first PCR step with genome DNA of S. cohnii (100 ng), and a 4S0-bp DNA fragment was is largely buried behind the former l3-strand. Accordingly, obtained. The fragment was in serted into pGEM-TEasy (Promega the effect of the latter l3-strand seems indirectly mediated Corp ., Madison, WI, USA) and sequenced. Based on the sequence possibly through the former strand. of the 4S0-bp fragment, two gene-specific primers (3ScohGSP I and SScohGSP2) were synthesized and used for gene-walking in com­ bination with the adaptor pri mer (S'-GTA ATA CGA CTC ACT Materials and methods ATA GGG C-3'). Consequently, a I-kb DNA fragment in the S'­ direction was obtai ned from the Pvull-digested genome DNA; and a O.S-kb DNA in the 3'-direction was obtained from Dral-digested Materials genome DNA by usi ng the Genome Walker Uni versal kit (Clontech Laboratories Tn c.). Again , the primer set (SScohPvu and 3ScohDra) The materials used and their sources were as fo ll ows: expression was synthesized for the overall ampli fication based on the sequence vector pQE60, from Qiagen Inc. (Chatsworth, CA, USA); low­ at the two ends of the DNA fragments. Finall y, the 19S2-bp DNA molecular-weight molecul ar markers, from GE Healthcare region of S. cohnii genome was cloned and sequenced. (B uchinghamshire, UK); restri cti on enzymes and DNA-modify ing enzymes, from Nippon Gene (Tokyo, Japan); KOD Plus DNA pol­ ymerase, fro m Toyobo (Tokyo, Japan); MCA peptide substrates and Construction of the expression plasm ids for leupeptin , from the Peptide Institute Inc. (Osaka, Japan); Genome glutamyl endopeptidases Walker Uni versal kit and Talon metal affinity resin , from Clontech Laboratories Inc. (Palo Alto, CA, USA); azocasein, thennolys in Recombinant Glu Scpr and GluScoh were expressed in E. coli as the from Bacillus thermoproteolyticus rokko, trypsin, DFP, AEBSF fu ll-length for m with N- and C-terminal tags (MGGS and (Pefabloc® sq, PMSF, aprotinin, pepstatin, TPCK, and TLCK, GSRSHHHHHH, respecti vely) by use of the pQE60 expression vec­ from Sigma-Aldrich (S t. Loui s, MO, USA); oli gonucleotide primers tor, as described previously (Nemoto et aI. , 200S). Briefl y, the DNA Proteolytic activity of staphylococcal glutamyl endopeptidases 1231

fragment carrying the full-length form of GluScpr was amplified 0.2 ml of 50 mM Tris-HCI (pH 8.0) and 5 mM EDTA at 3TC for with 5'-primer (5GluScprM-68Bgl) and 3'-primer (3Glu­ I h. EDTA was then added to the reaction mi xture to inactivate ScprA237BgI) (Table I), which carried BglII sites, and then inserted thermolysin (Fontana, 1988). The fluorescence was measured with into a BamHI site of pQE60 (designated pQE60-GluScpr). The excitation at 380 nm and emission at 460 nm with a Fluorescence DNA caITying the full-length form of G luScoh was amplified with Photometer F-4000 (Hitachi, Tokyo, Japan). One unit was defined 5' -primer (5GluScohL-67Bam) and 3' -primer (3GluScoh V21IBam), as the activity that produces I fmol of product per min. For cal­ which carried BamHI sites, and then in serted into a BamHI site of culation of the specific activity, the bands visuali zed with CBB cor­ pQE60 (designated pQE60-GluScoh). responding to mature proteases were quantified with a PDQuest Full-length GluV8 with five substitutions in its prosequence (D- software (Bio-Rad, Hercules, CA, USA) by using BSA separated 3 1H , E-7Q, E-4S, A-2P, and N-IS) and deletion of the C-terminal on SDS-PAGE as a standard. 52 residues was expressed in E. coli as described previously (Ono et aI., 2008). N-Terminal amino acid sequencing

In vitro mutagenesis by peR N-terminal amino acid sequences were determined after separation of recombinant proteins by SDS-PAGE and transference to a poly­ In vitro mutagenesis was performed as described previously (Nemo­ vinylidene difluoride membrane (Sequi-Blot PVDF Membrane, Bio­ to et aI. , 2008) by PCR with mutated primers to substitute the nucle­ Rad). After having been stained with CBB, the bands were excised; otides of pQE60-GluScpr encoding D-31 and D-18 to H-31 and and the respecti ve materials were then directl y sequenced with a K-18 (designated pQE60-GluScpr-mut), the nucleotides of pQE60- model Procise 49XcLC protein sequencer (ABI, Foster City, CA, GluScoh encoding D-35 and E-25 to S (designated pQE60-Glu­ USA). Scoh-mut), the nucleotides of pQE60-GluScpr-mut encoding Y 185 and EI89 to WI85 and PI 89 (designated pQE60-GluScpr-WP), and Protein concentration determination the nucleotides of pQE60-GluSE encoding N 190, Y192, and S 195 to H190, HI92 and AI95 (designated pQE60-GluSE-HHA), the Protein concentrations were determined by the bicinchoninic acid nucleotides of pQE60-GluV8 encoding the amino acids at positions method using bovi ne serum albumin as the standard (Pierce, Rock­ -3, -2, -I , and I to K at -3, -2, and - I positions or L at position I ford, IL, USA). For quantification of affinity-purified samples, (Nemoto et aI. , 2008). SI69, Tl64, and NI93 of GluV8 were sub­ bovine albumin was dissolved in either 0.1 M imidazole (pH 8.0) stituted to A. Similarly, S 169 of GluScoh, GluSE, and Gl uSW were containing 10% (v/v) to adjust buffer compositi ons. mutated to A. All mutagenesis and truncation were confirmed by DNA sequencing.

Expression and purification of recombinant Acknowledgments proteases

The expression and purification of recombinant proteins in E. coli This research was supported by grants-in-aid for scientific research was performed by Talon affin ity chromatography as reported pre­ from the Ministry of Education, Science, Sports, and Culture of viously (Nemoto et aI., 2008, 2009). The purified proteins were Japan (to T.K.N., Y.O-N., and S.K.). stored at -80°C until used.

Sodium dodecyl sulfate-polyacrylamide gel References electrophoresis (SOS-PAGE) and zymography

Samples ( I fLg) were separated by electrophoresis in the presence Calander, A-M., Dubin, G. , Potempa, J. , and Tarkowski, A. (2008). of 0.1 % (w/v) SDS at a polyacrylamide concentration of 12.5% (wi Staphylococcus aureus infection triggers production of neutral­ v), and then stained with Coomassie Brilliant Blue (CBB). For izi ng, V8 protease-specific antibodies. FEMS Immunol. Med. zymography, SDS-PAGE was performed by using a polyacrylamide Microbiol. 52, 267-272. gel of 12.5% (w/v) containing 1% (w/v) of azocasein (Rice et aI., Carmona, C. and Gray, G.L. (1987). Nucleotide sequence of the 2001). After SDS-PAGE, the gel was incubated twice at 25°C with serine protease gene of Staphylococcus aureus, strain V8. 100 ml of 2.5% (w/v) Triton X-IOO for 20 min each time, twice for Nucleic Acids Res. IS, 6757. 10 min each time with 100 ml of 50 111M Tri s-HCI (pH 7.8) con­ Coulter, S.N., Schwan, WR., Ng, E.Y. , Langhorne, M.H., Ritchie, taining 30 mM NaCI, and then incubated in 100 ml of the latter H.D., Westbrock-Wadman, S., Hufnagle, WO., Folger, K.R., buffer at 37°C overnight. Finall y, non-hydrolyzed azocasein in the Bayer, A.S., and Stover, C.K. (1998). Staphylococcus aureus gel was visualized with CBB. genetic loci impacting growth and survival in mUltiple infection environments. Mol. Microbiol. 30, 393 -404. In vitro processing and measurement of Demidyuk, LV. , Romanova, D.V. , Nosovskaya, E.A., Chestukhina, protease activity G.G., Kuranova, LP. , and Kostrov, S.v. (2004). Modification of substrate- of glutamyl endopeptidase from Bacillus Unless otherwise stated, in vitro processing of recombinant proen­ intermedius. Protein Eng. Des. Sel. 17, 411-416. zymes and subsequent protease assay were performed as follows: Drapeau, G.R. (1978). Role of a metalloprotease in activation of the recombinant proteins (10 fLg) were incubated for 4 h in 0.1 ml of precursor of Staphylococcal protease. J. Bacteriol. 136, 607- buffer A [1 0 mM sodium borate (pH 8.0) containing 0.005% (v/v) 613.

Triton X-IOO and 2 mM CaS04 ] with 0.3 fLg of thermolysin (weight Drapeau, G.R., Boily, Y., and Houmard , J. ( 1972). Purification and ratio 33: I) at 3TC. Thereafter, aliquots of glutamyl endopeptidases properties of an extracellular protease of Staphylococcus aureus. (0.25 fLg as a proform) were incubated with 20 fLM LLE-MCA in 1. BioI. Chem. 247, 6720-6726. 1232 T. Ono et al.

Dubin, G., Chmiel, D., Mak, P. , Rakwalska, M., Rzychon, M. , and Ohara-Nemoto, Y , Ono, T. , Shimoyama, Y , Kimura, S., and Nemo­ Dubin, A. (2001). Molecular cloning and biochemical charac­ to, T.K. (2008). Homologous and heterologous expression and terization of proteases from Staphylococcus epidermidis. BioI. maturation processing of extracellular glutamyl endopeptidase of Chem.382, 1575-1582. Staphylococcus epidennidis. BioI. Chem. 389, 1217- 12 19. Fontana, A. (1988). Structure and stability of thermophilic enzymes: Ono, T., Nemoto, T.K., Shimoyama, Y, Kimura, S., and Ohara­ studies on thermolys in . Biophys. Chem. 29, 18 1-193. Nemoto, Y (2008). An Escherichia coli expression system for Ghebremedhin, B., Layer, F. , Konig, w. , and Konig, B. (2008). glutamyl endopeptidases optimized by complete suppression of Genetic classification and di stinguishing of Staphylococcus spe­ autodegradation. Anal. Biochem. 381, 74- 80. cies based on different partial gap, 16S rRNA, hsp60, rpoB , Prasad, L., Leduc, Y , Hayakawa, K. , and Delbaere, L.T. (2004). sodA, and tuf gene sequences. 1. Clin. Microbiol. 46, 1019- The structure of a uni versally employed enzyme: V8 protease 1025. from Staphylococcus aureus. Acta Crystallogr. Sect. D 60, Ikeda, Y , Ohara-Nemoto, Y, Kimura, S., Ishibashi , K., and Kikuchi, 256-259. K. (2004). PCR-based identification of Staphylococcus epider­ Rebrikov, D.Y., Akimkina, T.Y. , Shevelev, A.B., Demidyuk, LY. , midis targeting gseA encoding the glutamic-acid-specific prote­ Bushueva, A.M., Kostrov, S.Y., Chestukhina, G.G., and Stepa­ ase. Can. J. Microbiol. 50, 493- 498. no v, Y.M. (1999). Bacillus intermedius glutamyl endopeptidase. Kakudo, S., Kikuchi, N., Kitadokoro, K., Fujiwara, T. , Nakamura, Molecular cloning and nucleotide sequence of the structural E., Okamoto, H., Shin, M., Tamaki, M., Teraoka, H., Tsuzuki, gene. J. Protein Chem. 18, 21-27. H., et al. (1992). Purification, characterization, cloning, and Rice, K. , Peralta, R. , Bast, D. , Azavedo, J. , and McGav in, M. expression of a gl utamic ac id-specific protease from Bacillus (200 I). Description of Staphylococcus serine protease (ssp) licheniformis ATCC 14580. J. BioI. Chem. 267, 23782- operon in Staphylococcus aureus and nonpolar inacti vation of 23788. sspA -encoded serine protease. Infect. Immun. 69, 159-169. Karlsson, A., Saravia-Otten, P , Tegmark, K. , Morfeldt, E., and Sasaki, M., Ohara-Nemoto, Y , Tajika, S., Kobayashi , M. , Ikeda, Y , Arvidson, S. (2001). Decreased amounts of cell wall-associated Kaneko, M. , and Takagi T. (1998). Purification and characteri­ protein A and fibronectin-binding proteins in Staphylococcus zation of a glutamic acid-specific protease from Staphylococcus aureus sarA mutants due to up-regulation of extracellular pro­ epidermidis. Jpn. J. Oral BioI. 40, 542-548. teases. Infect. Immun. 69, 4742- 4748. Shine, J. and Dalgarno, L. (1975). Terminal-sequence analysis of Kawalec, M., Potempa, J. , Moon, 1.L., Travis , J., and Murray, B.E. bacterial ribosomal RNA. Correlati on between the 3'-terminal­ (2005). Molecular diversity of a putative virulence factor: puri­ polypyrimidine sequence of 16-S RNA and translational speci­ fication and characterization of isoforms of an extracellular ser­ ficity of the ribosome. Eur. J. Biochem. 57, 221 - 230. ine glutamyl endopeptidase of Enterococcus faecalis with Stennicke, H.R. and Breddam, K. (1998). Glutamyl endopeptidase different enzymatic activities. J. Bacteriol. 187, 266- 275. 1. In: Handbook of Proteolytic Enzymes, AJ. Barrett, N.D. Raw­ Kitadokoro, K. , Nakayama, E., Tamaki, M. , Horii, T. , Okamoto, H., lings and J.F. Woess ner, eds. (S an Diego, CA, USA: Academic Shin, M., Sato , T. , Fujiwara, T. , Tsuzuki, H., Yoshida, N. et al. Press), pp. 243- 246. (1993). Purification, characterization, and molecular cloning of Suzuki, Y , Yabuta, M., and Ohsuye, K. (1994). Cloning and expres­ an acidic amino acid-specific proteinase from Streptomyces fra­ sion of the gene encoding the glutamic acid- specific protease diae ATCC14544. Biochim. Biophys. Acta 1163, 149-157. from Streptomyces griseus ATCCI01 37. Gene 150, 149- 151. Kloos, W.E. and Bannerman, T.L. (1994). Update on clinical sig­ Svendsen, 1. and Breddam, K. (1992). Isolation and amino acid nificance of coagulase-negative staphylococci. Clin. Microbiol. sequence of a glutamic acid specific endopeptidase from Bacil­ Rev. 7, 117- 140. lus licheniformis. Eur. J. Biochem. 204, 165-1 7 1. Kuroda, M., Yamashita, A. , Hirakawa, H. , Kumano, M., Morikawa, Svendsen, 1. , Jensen, M.R. , and Breddam, K. (1991). The primary K., Higashide, M. , Maruyama, A. , Inose, Y, Matobe, K., Toh, structure of the glutamic acid-specific protease of Streptomyces H., et al. (2005). Whole genome sequence of Staphylococcus griseus. FEBS Lett. 292, 165- 167. saprophyticus revealed the pathogenesis of uncomplicated uri ­ Tamada, T. , Kinoshita, T., Kurihara, K. , Adachi, M., Ohhara, T. , nary tract infection. Proc. Natl. Acad. Sci. USA 102, 13272- Imai, K. , Kuroki, R. , and Tada, T. (2009). Combined high-reso­ 13277. lution neutron and X-ray analysis of inhibited confirms Nemoto, T.K. , Ohara-Nemoto, Y , Ono, T. , Kobayakawa, T. , Shi­ the active-si te oxyanion hole but rul es against a low-barrier moyama, Y , Kimura, S. , and Takagi, T. (2008). Characteri zation . J. Am. Chem. Soc. 131 , 11033-11040. of the glutamyl endopeptidase from Staphylococcus aureus Wildeboer, D., Jeganathan, F. , Price, R.G., and Abuknesha, R.A. expressed in Escherichia coli. FEBS 1. 275, 573- 587. (2009). Characterization of bacterial proteases wi th a panel of Nemoto, T.K. , Ono, T. , Shimoyama, Y , Kimura, S., and Ohara­ fluorescent peptide substrate. Anal. Biochem. 384, 321 - 328. Nemoto, Y (2009). Determination of three amino acids that Yokoi , K. , Kakikawa, M. , Kimoto, H., Watanabe, K., Yasukawa, H., caused the alteration of proteolytic activities of staphylococcal Yamakawa, A. , Taketo, A., and Kodaira, K. (2001 ). Genetic and glutamyl endopeptidases. BioI. Chem. 390, 277- 285 . biochemical characterization of glutamyl endopeptidase of Ohara-Nemoto, Y , Ikeda, Y , Kobayashi, M. , Sasaki, M. , Tajika, S., Staphylococcus warneri M. Gene 281, 115- 122. and Kimura, S. (2002). Characterization and molecular cloning of a glutamyl endopeptidase from Staphylococcus epidermidis. Microb. Pathog. 33, 33-41. Received February 7, 20 10; accepted June 14, 20 10