gingivalis prtt Gene, Coding for Protease Activity

Transcription

1 INFECrION AND IMMUNITY, Jan. 1993, p /93/ $02.00/0 Copyright 1993, American Society for Microbiology Vol. 61, No. 1 Isolation and Characterization of the Porphyromonas gingivalis prtt Gene, Coding for Protease Activity JUN-ICHI OTOGOTO AND HOWARD K. KURAMITSU* Department of Pediatric Dentistry, University of Texas Health Science Center at San Antonio, San Antonio, Texas Received 8 June 1992/Accepted 3 November 1992 The prtt gene, coding for trypsinlike proteolytic activity, has been isolated from Porphyromonas gingivalis ATCC This gene is present immediately downstream from the sod gene on a 5.9-kb DNA fragment from the organism isolated in Escherichia coli. The complete nucleotide sequence of the gene was determined, and the deduced amino acid sequence of the enzyme corresponds to a 53.9-kDa protein with an estimated pl of Gelatin-sodium dodecyl sulfate-polyacrylamide gel electrophoresis zymography also indicated a similar molecular size for the protease. The enzyme was purified to near homogeneity following anion-exchange and gel-filtration chromatography. The purified enzyme also exhibited a single protein species with a size of approximately 53 kda. Enzyme activity was strongly dependent upon the presence of reducing agents (dithiothreitol, cysteine, and 2-mercaptoethanol) and was also stimulated in the presence of calcium ions. A comparison of the properties of the prtt gene product with comparable parameters of proteases previously purified from different strains of P. gingivalis suggested that the cloned protease represents a previously uncharacterized enzyme. Although human periodontal diseases appear to result from complex interactions between the host and a variety of anaerobic microorganisms, a variety of approaches have strongly suggested an important role for Porphyromonas gingivalis in periodontal tissue destruction (28). Several potential virulence factors, including the elaboration of protease activity, have been identified for these organisms (17). These enzymes may play a role in periodontal disease by degrading protective host immunogloblins (10, 15), hydrolyzing host proteins to provide required amino acids for growth (4), and aiding in the destruction of host connective tissue (30). One prominent enzymatic activity expressed by these organisms (6, 7, 31, 32, 36) is that related to the hydrolysis of the synthetic trypsin substrate BAPNA (N-abenzoyl-DL-arginine-p-nitroanilide). However, several laboratories have demonstrated that enzymes with distinct properties present in both the culture fluids (7, 8, 21, 31) and associated with the cellular membranes of P. gingivalis (7, 8, 19, 22, 26, 29) express BAPNA-hydrolyzing activity. In view of the possibility of autodegradation of the proteases, it is not clear how many distinct proteases are produced by each strain of these organisms. One approach to answering this question is to utilize molecular genetic techniques to isolate the genes for the individual proteases. Several recent communications (23, 32) have reported on the cloning of protease genes from P. gingivalis. In the latter investigation, we identified a protease activity expressed from the Escherichia coli plasmid psl containing a 5.9-kb DNA fragment from P. gingivalis ATCC The present study demonstrates that this activity is not mediated by the collagenase expressed by theprtc gene present on the insert fragment but is due to a distinct gene, prtt, located downstream from the sod gene recently characterized on this same DNA fragment (5). The properties of this protease are compared with those of other proteases previously characterized from different strains of P. gingivalis. * Corresponding author. 117 MATERIALS AND METHODS Bacteria and plasmids. E. coli HB101 and MV1184 (34) were maintained and grown in Luria-Bertani (LB) broth as previously described (1). Plasmid vectors puc18 and puc19 (35) were used for subcloning and expression studies while pbluescript SK+ and KS+ (Stratagene, La Jolla, Calif.) were utilized for nucleotide sequencing. DNA manipulations. Restriction endonuclease digestion and ligation of DNA fragments were carried out according to the directions of the suppliers. The isolation of DNA and transformation of E. coli cells were carried out as previously described (1). Enzyme assays. For routine enzyme assays, E. coli strains harboring recombinant plasmids were grown in 10 ml of LB medium containing ampicillin (50,ug/ml) and 0.14 mm isopropyl-p-d-thiogalactopyranoside (IPTG) at 37 C for 18 h. The cultures were harvested by centrifugation at 5,000 x g for 10 min, and the cells were resuspended in 1 ml of 50 mm Tris-HCl buffer (TB) (ph 8.0) and sonicated for a total time of 10 min (with intermittent cooling) in a Microson Ultrasonic Disruptor (Heat Systems, Farmingdale, N.Y.). The supernatant fluids obtained following centrifugation at 5,000 x g for 5 min served as the crude extracts. Subsequently, this crude extract was centrifuged at 160,000 x g for 1 h, and the supernatant fluid was used as the membrane-free fraction. The resulting pellet was mixed with TB containing 1% Triton X-100 and centrifuged at 160,000 x g for 1 h, and the supernatant fluid was used as the membrane-bound fraction. Trypsinlike protease activity was determined following hydrolysis of the synthetic chromogenic substrate BAPNA (Sigma Chemical Co., St. Louis, Mo.). The enzyme samples (0.1 ml) were mixed with 0.7 ml of reaction mixture (5 mm CaCl2, 2 mm mercaptoethanol, 0.2 mm BAPNA in TB) and incubated at 37 C for 30 min. The reaction was terminated following the addition of 0.2 ml of 50% acetic acid, and the absorbance of the solutions was measured at 410 nm. One unit of enzyme was defined as the amount of enzyme required to release 1.0,umol ofp-nitroanilide per min at 37 C in the standard assay.

2 118 OTOGOTO AND KURAMITSU Two assays were utilized to determine gelatinase activity. Gelatinase activity was routinely determined by visual detection of the dissolution of the gelatin matrix of X-ray film as previously described (20). Samples (20,ul) were spotted onto the film, incubated for 18 h at 37 C, and flushed with water; the film was observed for zones of clearing, indicating gelatinase activity. In addition, gelatinase activity was also determined following gelatin-sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) zymography (9). Following electrophoresis of enzyme samples in 0.1% gelatin-conjugated gels, the gels were washed gently with 2.5% Triton X-100 (twice for 30 min) to remove SDS and finally with detergent containing TB. The washed gel was then incubated for 18 h at 37 C in 0.1 M glycine-naoh (ph 8.3) and stained with Coomassie brilliant blue R to visualize gelatinase bands. The degradation of potential protein substrates by the cloned enzyme was examined following SDS-PAGE analysis of the reaction mixtures (12). Utilizing a 10% acrylamide separating gel and a 4.5% stacking gel, reaction mixtures (containing 15,ul of enzyme and 5 pg of protein substrate incubated for 24 h at 37 C) were heated at 100 C for 3 mn and subjected to electrophoresis (16). The gels were then stained with Coomassie brilliant blue R to visualize protein degradation. Enzyme purification procedure. E. coli subclone KS1 (see Fig. 2) was grown in 4 liters of LB broth at 37 C for 18 h, and the washed cells were resuspended in 20 ml of TB buffer. The bacterial cells were sonicated, and the suspensions were centrifuged at 5,000 x g for 15 min to remove the cellular debris and produce a crude extract. The supernatant fluids were then filtered through a Millipore filter (0.22-p,m-poresize), and the filtrate was injected into a Mono-Q HR5/5 anion-exchange chromatography column (Pharmacia-LKB Biotechnology, Inc., Piscataway, N.J.). The column was eluted with a gradient of 0 to 1 M NaCl in TB, and the fractions exhibiting BAPNA hydrolysis and gelatinase activities were pooled and concentrated through a Centricon-10 ultrafilter (Amicon Corp., Danvers, Mass.). The partially purified protease present in the flow-through fractions was next applied to a TSK-G3000SW gel-filtration column (Pharmacia-LKB) and eluted with TB. The active fractions were again pooled, concentrated through a Centricon ultrafilter, and stored at -20 C. Protein concentrations were routinely determined by the modified Bradford method (3), with bovine serum albumin as a standard. Cystatin was kindly provided by M. J. Levine (State University of New York, Buffalo, N.Y.) and histatin-5 was provided by Y. Kuboki (Hokkaido University, Sapporo, Japan). DNA sequence analysis. The complete nucleotide sequence of the prtt gene was determined from both DNA strands by utilizing the dideoxy nucleotide sequencing strategy (24). Overlapping DNA fragments were subcloned into pbluescript SK+ or KS+ and deleted with exonuclease III and mung bean nuclease (11). Sequencing was carried out with cx-35s-datp, Sequenase T7 DNA polymerase (United States Biochemical Corp., Cleveland, Ohio), and universal sequencing primers. Sequence analysis was carried out with the Pustell Sequence Analysis Programs (International Biotech, Inc., New Haven, Conn.). Nucleotide sequence accession number. The P. gingivalis prtt gene sequence has been deposited in the GenBank data base under accession number M KDa FIG. 1. Gelatin-SDS zymography of crude extracts containing proteins expressed from recombinant plasmids. Zymography was carried out as described in the text. Lanes: 1, HB101 (ppl-lambda); 2, JM109 (puc19); 3, KS1, JM109 (puc19); 4, HB101 (psl). See Fig. 2 for orientation of DNA fragments. RESULTS Isolation of the prtt gene. The 5.9-kb DNA fragment containing the prtc (32) and sod (5) genes of P. gingivalis was originally isolated following screening of plasmid clone banks for protease activity on skim-milk-agar plates. However, the collagenase activity expressed from the prtc gene did not degrade the major milk protein casein or gelatin (32). By contrast, an enzyme expressed from plasmid psl from the original clone displayed gelatinase activity (Fig. 1). Furthermore, deletion of the region downstream from the sod gene resulted in the loss of gelatinase activity (32). These results suggested that a gene coding for protease activity was present in the deleted region. In order to confirm this possibility, various subcloned fragments from the psl insert were isolated and inserted into puc18 and puc19 (Fig. 2). These results indicated that insertion of the 1.1-kb KpnI-PstI or 3.7-kb Kjpnl-SalI fragment into puc19 resulted in subclones expressing BAPNA hydrolysis activity, but only the latter fragment expressed gelatinase activity (Table 1). Subsequent nucleotide sequencing (see below) revealed that the former fragment expressed a truncated, but enzymatically active, proteinase. However, insertion of both DNA fragments into puc18 did not result in similar activities. This suggested that transcription of the gene responsible for the protease activity was initiated from the lac promoter of the plasmid and not from a P. gingivalis promoter. In addition, sod KP= PS prtt I_I INFECT. IMMUN. KSII ---I KS2 = 0 1Kb FIG. 2. Subcloning analysis of the prtt gene. Fragments from the 5.9-kb P. gingivalis DNA insert of plasmid ps1 were subcloned into either puc18 or puc19. prtc, collagenase gene; sod, superoxide dismutase gene.

3 VOL. 61, 1993 TABLE 1. Activities of subcloned fragmentsa Subclone Vet Trypsinlike pro- Gelatinase Hemagglutination fragment ector tease activityb activity' activityd KP puc puc PS puc puc KS1 puc puc KS2' puc puc puc puc19 pslf ppl lambda ppl lambda a The DNA fragments were subcloned into the vectors and the resulting plasmid-containing strains were examined for indicated activities. E. coli cells containing the recombinant plasmids were grown in 100 ml of LB medium containing ampicillin (50 p.g/ml) and 0.14 mm IPTG at 37'C for 18 h. The cultures were harvested, suspended in 1 ml of 50 mm TB (ph 8.0), and sonicated. b Trypsinlike protease activity was detected following hydrolysis of the chromogenic substrate BAPNA. c Each sample was assayed for enzymatic activity by the X-ray film method. d Hemagglutinating activity was assayed using sheep erythrocytes (19). e Plasmid containing deletion of the KpnI-SalI fragment. f Reference 32. most of the activity in crude extracts was located in the membrane-bound fraction (data not shown). Plasmid psl also expressed hemagglutinin activity, which was expressed from subclones containing the PstI-SalI fragment in puc18 and the KpnI-SalI fragment in puc18. Therefore, these results suggested that the putative protease was not responsible for hemagglutinin activity. Likewise, deletion of the KpnI-SalI fragment indicated that both activities were encoded by genes located downstream from the sod gene. Nucleotide sequencing of the prtt gene. Both DNA strands downstream from the sod gene were sequenced, and one long open reading frame corresponding to a 53.9-kDa protein was detected (Fig. 3). This molecular size was approximately equal to that estimated for the gelatinase for which activity was expressed from psl following gelatin-sds- PAGE analysis (Fig. 1). The protein would be initiated at base position 239 (Fig. 3), which was 8 bp downstream from an AGA sequence found in E. coli ribosome binding sites (27). However, sequences corresponding to -10 and -35 promoter consensus sequences could not be detected upstream from the gene. In addition, no sequence typical of bacterial signal sequences could be detected in the amino terminus of the deduced protein. The unprocessed protein appears to be highly basic, with an estimated pi of This gene has been named prtt (protease with trypsinlike activity). The calculated G+C ratio for the prtt gene was 45.4%, which corresponds well to the ratio of 46 to 48% previously determined for chromosomal DNA from strains of P. gingivalis (25) and the value of 48% determined from the sod gene of strain (5). Purification of the trypsinlike protease. In order to charac- prtt PROTEASE 119 terize the protein product of the prtt gene, the protease was purified from the E. coli subclone containing plasmid pks1. Since the deduced amino acid sequence indicated that the enzyme was highly basic, crude extracts of the subclone were initially chromatographed on a Mono-Q anion-exchange column. As predicted from the pi of the cloned enzyme, the protease was not retained by the column at ph 8.5. The flow-through fractions were pooled, concentrated, and then subjected to chromatography on a TSK G3000 gel-filtration column. Several peaks of BAPNA hydrolytic activity were detected, but only one of these displayed gelatinase activity. This latter fraction was concentrated and represented an approximately 64-fold purification of protease activity, with an overall yield of 11% (Table 2). Analysis of the concentrated enzyme following SDS-PAGE yielded a single protein band with an estimated molecular size of 53 kda (Fig. 4). This value is similar to that estimated from the deduced amino acid sequence of theprtt gene (Fig. 3) and that determined following gelatin zymography of crude extracts of the original psl clone (Fig.1). Characterization of the purified protease. Maximum protease activity of the purified enzyme was observed near ph 8.5 (data not shown) in the presence of Ca2+ (Table 3). Mg2+ ions were somewhat inhibitory, while Mn2' had only a slight stimulatory effect on enzyme activity. The enzyme did not appear to be a metalloenzyme since EDTA did not inhibit activity. The synthetic protease inhibitors TLCK (N-a-tosyl- L-lySyl chloromethyl ketone) and PMSF (phenylmethylsulfonyl fluoride) produced little or only marginal inhibitory activity. Essential sulfhydryl groups did not appear to be important for maximum activity since PCMB (p-chloromercuribenzoic acid) produced only a small inhibitory effect. However, like several previously characterized P. gingivalis proteases (31), the prtt gene product was strongly stimulated in the presence of the reducing agents dithiothreitol, cysteine, and 2-mercaptoethanol. Two salivary peptides, histatin and cystatin (18), which have been demonstrated to influence protease activity, had little effect on the trypsinlike protease. Despite the strong hydrolytic activity on the trypsin substrate BAPNA by the enzyme (Table 1), little activity against the homologous lysine-containing substrate BLPNA (Nbenzoyl-DL-lysine-p-nitroanilide) could be detected (data not shown). Therefore, unlike classical trypsin, the prtt gene product appears to be specific for arginine-containing peptide bonds. Since the protease could hydrolyze gelatin, it was of interest to determine whether the enzyme was active on native collagens. Analysis of proteolysis following SDS- PAGE revealed that the purified enzyme did not degrade native type I or type IV collagens (data not shown) following incubation at either 30 or 37 C for 18 h. In addition, the enzyme did not hydrolyze laminin, fibronectin, lysozyme, bovine serum albumin, or human immunoglobulin G or M. However, the enzyme could be shown to degrade casein in addition to gelatin (Fig. 5 and 6). Two fractions from the gel-filtration column (fractions 1 and 12) exhibited BAPNA hydrolytic activity but no gelatinase activity (Fig. 5). These fractions may represent altered forms of the P. gingivalis proteinase or E. coli endogenous proteinases. Furthermore, only a low degree of apparent self-digestion of the enzyme was detected following incubation for 15 h at 4 or 37 C. DISCUSSION The present communication describes the isolation of the P. gingivalis prtt gene, coding for a protease with activity

4 120 OTOGOTO AND KURAMITSU INFEcr. IMMUN. A BqlII A GAT CTT TGG AGT ATT GTT GAC TGG GAT ATTl GTA GAA TCT CGG TAT TAA GTA ACC CCA TTG TGC ACT TTG CAC AAT ACA TAA GGT ATA TGC CTG TGC CAA GAA CCG ATC GGG TGT CTC GGC AGG GCT TCT TCT TTT TCT CTT TTC GTT GTT CAC TAA CAG CCG AAT CAA AGC AAA AGA AAA AAG AAA CGG T'! TTC CCT CAA TCC TAT CAA GCC T'm TCA GAA AAG ATC AGG AAC ATG TAC CAT GTA TAT CCG AAT TGG TTT AGG CTT CAA AAT TT! CCC GTT MNt Tyr His Val Tyr Pro Asn Trp Phe Arg Leu Gln Asn Phe Pro Val TGC TAC TCC TCA AAA CGC GGT TCG GAA ACT TTT TTA TTT TGG CGT GGG AAG ACC AAA Cys Tyr Ser Ser Lys Ar; Gly Ser Glu Thr Ph. Lou Ph. Trp Arg Gly Lys Thr Lys AAA TTC TCA CGC CAC AAC GAA AAA AAT CTC GCG CCA CTT TTT CAG GGA ATA CGC GCC Lys Ph. S.r Arg His Asn Glu Lys Asn Lou Ala Pro L.u Phe Gln Gly Ile Arg Ala ACA ATC GGA GCT TTT CCG GTT CGT ATT T'! TTG AAT AGC TCC TTT TGC AAA CAA GCC Thr Ile Gly Ala Phe Pro Val Arg Ile Ph. Lou Asn Ser Ser Ph. Cys Lys Gln Ala * * 0 * * 0 GAG GCA ATC GAC AAA AAT CTG CCG AGG ATT CAC CTT CCG GGA AAA GAC TI! CCT GAG Glu Ala Ile Asp Lys Asn Lou Pro Arg Ile His Lou Pro Gly Lys Asp Pho Pro Glu *0 * * 0 * GAA AAG ATC GAA AAG ATT CGT TCA CAC ATA CAT TTA CGC AAA AAT TGT CCG ACA AAA Glu Lys Ile Glu Lys Ile Arg Ser His I1e His Lou Arg Lys Asn Cys Pro Thr Lys A"A TTA CAT TTG CAC CCG AAT AAA GAT GGA CAG GTA CCT CAG CTG GAT ACA GCA ATA Arg Lou His Lou His Pro Asn Lys Asp Gly Gln Val Ala Gln Lou Asp Thr Ala Ile GCC TTC GA GCT ATC GGC CTG GGG TCG AAT CCC AGC CTG ATC ACC TCA AAA AGG GCA Ala Phe Glu Ala Ile Gly Lou Gly Ser Asn Pro Ser Lou I1- Thr S-r Lys Arg Ala TGC GCA AGT ATC CTT TTN TAT CTC TGT AAC TI! AAT AAT CGT CCT TI! TTA TCA TTA Cys Ala Ser Ile Lou Lou Tyr Lou Cys Asn Ph. Asn Asn Arg Pro Pho Lou Sor L.u OTC GAC GTT AAA GAG aat GAC GAT TAT CCT TCT TCT CCC ATC TCC TGC TTC GAA CAA Val Asp Val Lys Glu Asp Clu Asp Tyr Pro Ser S-r Ala I1 Sor Cys Pho Glu Gln CAA AAC CAA aat GCA ATA ATC AAA COT ATC TTC TAC ACc TTA GcG CTA TTA TTA CTG Gln Asn Gln Asn Ala Il Hot Lys Ar; I1 Pho Tyr Thr Lou Gly Lou Lou Lou Lou a TCT CTC CCT ATN cc CAa WCA GGA cco GTG ACA CGA TCA AAG CCG AAC A"A CN cta Cys Lou Pro Not Lou Gln Ala Gly Pro Val Thr Ar Sor Lys Pro Asn Ar; Lu Lou FIG. 3. Nucleotide sequence and deduced amino acid sequence of the P. gingivalis prtt gene. A potential ribosome binding site (SD) as well as the locations of the restriction endonuclease sites is indicated. against the synthetic trypsin substrate BAPNA. Several ally inhibited by leupeptin (Table 3). Furthermore, all of proteases with similar specificity isolated from different P. these enzymes are serine proteases. By contrast, the prtt gingivalis strains have also been described previously (2, 6, protease is insensitive to inhibitors of this class of enzymes, 7, 31, 33, 36). These enzymes differ in terms of cellular such as PMSF (Table 3). location and sensitivity to inhibitors, as well as in estimated The molecular size of the cloned prtt protease estimated molecular size. A comparison of the properties of these following SDS-PAGE analysis of the purified enzyme (Fig. enzymes with those of the prtt protease suggests that the 4) and gelatin-sds-page zymography (Fig. 1) and from the latter enzyme may represent a unique enzyme which has not deduced amino acid sequence is consistent with a value of 53 been previously identified. For example, all the P. gingivalis kda. This size is most similar to that estimated for the trypsinlike proteases examined have been reported to be trypsinlike protease purified from strain 381 of P. gingivalis strongly inhibited by the synthetic protease inhibitor leupep- (33). However, because of the multiple proteases produced tin (13, 19, 33). However, the prtt protease is only margin- by these organisms (9), caution should be exercised in

5 VOL. 61, 1993 prtt PROTEASE 121 B AGA ACT m TTG CCA AAC GAC AAC CCA CGT TGT CTT CAT CGA CTG CGA GTC TCC GGA Arg Thr Phe Lou Pro Asn Asp Asn Pro Arg Cys Lou His Arg Leu Arg Val Ser Gly * * Pstl * * * TGG ATT TCG m ACA AAG CTG CAG AAA GAG AGG AGG CAC TAT TCT TCG TTT TCA ATC Trp Ile Ser Phe Thr Lys Leu Gln Lys Glu Arg Arg His Tyr Ser Ser Phe Sr Ile GAG GAG AGA AAG ACG GAT TTC TCC TCG TCG CAG CGG ATT ATC GGT TCC CGG AGT GAT Glu Glu Arg Lys Thr Asp Phe Ser Ser Ser Gln Arg Ile Ile Gly Ser Arg Ser Asp CGG ATA TGC TTT CkA GGA AAA CTT CGT ATG GGC GTA TGC CGG ACA ATC TCA CGG GGT Arg Ile Cys Ph- Gln Gly Lys Lou Arg Not Gly Val Cys Arg Thr Ile Sr Arg Gly * * * * * TGT CAA AGG CTA GAa CGT GAA ATG CTT GCT GTA ATG GAC GGC AAG GCA GAG CCG ATA Cys Gln Arg Lou Glu Arg Glu Not Lou Ala Val Hot Asp Gly Lys Ala Glu Pro Ile GAT CCT ATC CGT GAA GAC MAG CCT ACA CGG ACC TGC CAT CAT CCA TTG CCC CTA TTT Asp Pro Ile Arg Glu Asp Lys Pro Thr Arg Thr Cys His His Pro Lou Pro Lou Phc * * * BamI * * * TGG AAA CGG GCG AAC ATG CAT CGG ATC COT ACG ATA CCG GGT GCT TTA TTG CTG GCC Trp Lys Arg Aio Asn Met His Arg Iie Arg Thr Ile Pro Gly Ala Lou Lou Lou Ala * * * * * GGC TCT ATG CCT ATG ACA ACC TCT TOG TCA GTA GAG TAT CCT CAT CGC CTG ACT CCT Gly Sr Met Pro Mot Thr Thr Scr Trp Sr Val Glu Tyr Pro His Arg Lou Thr Pro GTG CCA AGC GAa CCC GGA TTG AGT AGA TTG AGA TAC MAA TTA CCG TTA CTC ATG CCA Val Pro Ser Glu Pro Gly Lou Sr Arg Lou Arg Tyr Lys Lou Pro Lou Liu Not Pro * * * * * CCC CAT CCC CAG TTG AAG TGG aac GTT CCO TCT GGT TCO TAT CCA TCO CAA ACG AAA Pro His Pro Gln Lou Lys Trp Asn Val Pro Sr Gly Sr Tyr Pro Sr Gln Thr Lys S GCA TGT CCA TCG ACC GTC TGC ACC GGC ATA ATA CAC AGO TCT GTT TTC TGC CAG TTC Ala Cys Pro Sr Thr Val Cys Thr Gly I1- I1 His Arg Sr Val Ph- Cys Gln Phe * * 0 CTT ACG AAT CAT GTC TTC ATT CTA GCC GGT TCO CTG GTG ACO GTC TTA GCG GTG CAA Lou Thr Asn His Val Phe Ile Lou Ala Gly Ser Lou Val Thr Val Lou Ala Val Gln * * * 0 0 * CGO ATT CGC TAC ATG ACA GCG CAT GAC AGT GGT CTA CTA ACA GAT CGG AAA GCA CGA Arg Ile Arg Tyr Net Thr Ala His Asp Sr Gly Lou Lou Thr Asp Arq Lys Ala Arg * * * 0 * * CAT GTA TTC CTC MAT GTA TGA TCA aac GTC AAT ATT GAA TCA GGA GCT TGT AAA aat His Val Phe Lou Asn Val determining the precise molecular size for proteins purified from these organisms. Despite the differences detected between the clonedprtt protease and the trypsinlike proteases previously character- TABLE GAC AM GCO CAA GC CCC GTC CTG GAA TO T FIG. 3-Continued. Purification of the trypsinlike protease Total Total Sp act Yield Purification Purification step protein activity (U/mg) (%) (fold) (mg) (U) Crude extract Mono-Q HR5/ TSK G3000SW ized from P. gingivalis, all of these enzymes are strongly stimulated by reducing agents. However, by contrast to most of the trypsinlike proteases from these organisms, the cloned enzyme and the enzyme purified from strain 381 (33) are not strongly inhibited by sulfhydryl reagents. Therefore, it appears that the prtt protease is not a classical serine protease or cysteine protease. In addition, this enzyme is not a classical trypsin since it does not degrade the synthetic substrate BLPNA (containing a lysine residue instead of arginine). Therefore, the protease appears to hydrolyze peptide bonds containing arginine at the carboxyl side of the hydrolyzed peptide bond. Recently, two proteases purified from P. gingivalis have been reported to display similar substrate specificity (4, 13). However, it is not clear whether

6 122 OTOGOTO AND KURAMITSU INFEcr. IMMUN KDa FIG. 4. SDS-PAGE analysis of the purified enzyme. Purified enzyme (2.Lg) from the TSK-G3000SW column was concentrated by a Centricon-10 ultrafilter. Molecular size standards (in kilodaltons) are indicated on the right. the prtt gene product is identical to either of these two enzymes since distinct differences (substrate specificity and preliminary amino acid sequence comparisons) between the enzymes have been detected. Several of the trypsinlike proteases isolated from P. gingivalis were isolated from culture supernatant fluids (7, 13, 21, 31) while others were isolated from cellular fractions (7, 8, 19, 22, 26, 29). However, the demonstrations that these organisms release considerable amounts of membrane vesicles containing membrane-associated proteins into the culture fluids (30) make it difficult to assess the cellular location of some of these enzymes. Therefore, the cellular location of the prtt protease in strain has not been determined. The majority of the protease activity expressed from the cloned prtt gene was associated with the E. coli membrane fraction, although a significant amount was also present in the membrane-free fraction (data not shown). Therefore, the protease could be primarily membrane associated in E. coli but released following the sonication treatment. TABLE Effects of enzyme inhibitorsa Effector Concn (mm) Relative activity None 100 TLCK Leupeptin Cystatin Histatin EDTA PCMB PMSF Dithiothreitol L-Cysteine Mercaptoethanol CaCl MgCl MnCl a Purified protease was assayed for BAPNA hydrolyzing activity in the presence of the indicated effectors in the absence of Ca and 2-mercaptoethanol. In addition, the inhibitors were preincubated with the enzyme before the addition of the substrate. 200K1 X) 97~ ' *-* A4 0 ' - Gelatin FIG. 5. The effects of the trypsinlike protease on gelatin. Fractions from the TSK-G3000SW gel-filtration column were assayed for gelatinase activity following SDS-PAGE. Lanes: 1, fraction 1 plus gelatin; 2, fraction 12 plus gelatin; 3, fraction 20 (prtt gene product) plus gelatin; 4, gelatin alone. Preliminary characterization of a protease gene isolated from P. gingivalis W83 has been recently reported (23). However, a comparison of the properties of this enzyme with those of theprtt protease revealed that their molecular sizes following SDS-PAGE analysis are distinct and that both enzymes exhibit different sensitivities to EDTA and 2-mercaptoethanol. Recent Southern blot analysis utilizing plasmid psl as a probe has suggested that the prtc, sod, and prtt genes are present in the same relative orientation on the chromosome of representative strains from the three major serotypes of P. gingivalis (14). In addition, the results of that investigation using Northern (RNA) blot analysis indicated that the mrna coding for the prtc collagenase may be large enough to include the prtt protease. Therefore, possible coexpression of two enzymes involved in collagen degradation (gelatin is denatured type I collagen) would suggest that the major role of the prtt protease may be the digestion of collagen fragments generated from the action of the prtc collagenase. However, it is of interest that no apparent signal sequence could be detected in the deduced amino acid sequence of the proteinase. Therefore, it is not clear how this protein would be compartmentalized (outer membrane vesicles) to function in exogenous collagen breakdown. Further characterization of the mrna corresponding to these genes as well as the specificities of these two enzymes will be required to confirm this hypothesis. In addition, the recent development of a gene transfer system for generating specific mutants in P. gingivalis (5a) will allow direct testing of this hypothesis with the prtc and prtt genes. Recently, it has been reported that the hemagglutinin of P. gingivalis 381 is identical to a trypsinlike protease produced t4' 21KflaKl) _ (53KDa)n3a - - *w 46 r* 21.5 FIG. 6. Effects of trypsinlike protease on casein and self-digestion. Protein degradation was analyzed by SDS-PAGE. Lanes: 1, 2 p.g of purified protease stored at 4 C; 2, 2,ug of enzyme incubated at 37 C for 18 h; 3 and 4, 2,ug of enzyme incubated with 1,ug of casein at 37'C for 18 h; 5, 1,ug of casein.

7 VOL. 61, 1993 by these organisms (19). Therefore, it was of great interest that the P. gingivalis 5.9-kb DNA fragment on plasmid psl expressed both BAPNA hydrolysis as well as hemagglutinating activities (Table 1). However, subcloning of this fragment revealed that the prtt protease was not responsible for hemagglutinin activity and that this latter activity was expressed from a gene downstream from the prtt gene. Preliminary sequencing of this region has revealed the presence of an open reading frame corresponding to a truncated protein, and further characterization of this putative gene is now in progress. ACKNOWLEDGMENTS We express our appreciation to T. Kato for advice in several aspects of the current study. This investigation was supported in part by National Institutes of Health grant DE REFERENCES 1. Aoki, H., T. Shiroza, M. Hayakawa, S. Sato, and H. K. Kuramitsu Cloning of a Streptococcus mutans glucosyltransferase gene coding for insoluble glucan synthesis. Infect. Immun. 53: Birkedal-Hansen, H., R E. Taylor, J. J. Zambon, P. K. Barua, and M. E. Neiders Characterization of collagenolytic activity from strains of Bacteroides gingivalis. J. Periodontal. Res. 23: Bradford, M. H A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: Chen, Z., J. Potempa, A. Polanowski, S. Renvert, M. Wikstrom, and J. Travis Stimulation of proteinase and amidase activities in Porphyromonas (Bacteroides) gingivalis by amino acids and peptides. Infect. Immun. 59: Choi, J., N. Takahashi, T. Kato, and H. K. Kuramitsu Isolation, expression, and nucleotide sequence of the sod gene from Porphyromonas gingivalis. Infect. Immun. 59: a.Dyer, D. Personal communication. 6. Endo, J., M. Otsuka, M. Sato, and R. Nakamura Cleavage action of a trypsin-like protease from Bacteroidesgingivalis 381 on reduced egg-white lysozyme. Arch. Oral Biol. 34: Fujimura, S., and T. Nakamura Isolation and characterization of a protease from Bacteroides gingivalis. Infect. Immun. 55: Fujimura, S., and T. Nakamura Purification and characterization of a 43 kda protease of Porphyromonas gingivalis. Oral Microbiol. Immunol. 5: Grenier, D., G. Chao, and B. M. McBride Characterization of sodium dodecyl sulfate-stable Bacteroides gingivalis proteases by polyacrylamide gel electrophoresis. Infect. Immun. 57: Grenier, D., D. Mayrand, and B. C. McBride Further studies on the degradation of immunoglobulins by black-pigmented Bacteroides. Oral Microbiol. Immunol. 4: Heinkoff, S Unidirectional digestion with exonuclease II creates targeted breakpoints for DNA sequencing. Gene 28: Heussen, C., and E. B. Dowdle Electrophoretic analysis of plasminogen activators in polyacrylamide gels containing sodium dodecyl sulfate and copolymerized substrates. Anal. Biochem. 102: Hinode, D., H. Hayashi, and R. Nakamura Purification and characterization of three types of proteases from culture supernatants of Porphyromonas gingivalis. Infect. Immun. 59: Kato, T., N. Takahashi, and H. K. Kuramitsu Sequence analysis and characterization of the Porphyromonas gingivalis prtc gene expressing a novel collagenase activity. J. Bacteriol. 174: Kilian, M Degradation of immunoglobulins Al, A2, and G by suspected principal periodontal pathogens. Infect. Immun. 34: prtt PROTEASE Laemmli, U. K Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227: Mayrand, D., and S. C. Holt Biology of asaccharolytic black-pigmented Bacteroides species. Microbiol. Rev. 52: Nishikata, M., T. Kanehira, H. Oh, H. Tani, M. Tazaki, and Y. Kuboki Salivary histatin as an inhibitor of a protease produced by the oral bacterium Bacteroides gingivalis. Biochem. Biophys. Res. Commun. 174: Nishikata, M., and F. Yoshimura Characterization of Porphyromonas (Bacteroides) gingivalis hemagglutinin as a protease. Biochem. Biophys. Res. Commun. 178: Norqvist, A., B. Norrman, and H. Wolf-Walz Identification and characterization of a zinc metalloprotease associated with invasion by the fish pathogen Vibrio anguillarum. Infect. Immun. 58: Ono, M., K. Okuda, and I. Takazoe Purification and characterization of a thiol-protease from Bacteroides gingivalis strain 381. Oral Microbiol. Immunol. 2: Otsuka, M., J. Endo, D. Hinode, A. Nagata, R. Maehara, M. Sato, and R. Nakamura Isolation and characterization of protease from culture supematant of Bacteroides gingivalis. J. Periodontal. Res. 22: Park, Y., and B. C. McBride Cloning of a Porphyromonas (Bacteroides) gingivalis protease gene and characterization of its product. FEMS Microbiol. Lett. 92: Sanger, F., S. Nicklen, and A. R Coulson DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74: Shah, H. N., and M. D. Collins Proposal for reclassification of Bacteroides asaccharolyticus, Bacteroides gingivalis, and Bacteroides endodontalis in a new genus, Porphyromonas. Int. J. Syst. Bacteriol. 38: Shah, H. N., and S. E. Gharbia Lysis of erythrocytes by the secreted cysteine proteinase of Porphyromonas gingivalis W83. FEMS Microbiol. Lett. 61: Shine, J., and L. Dalgarno The 3' terminal sequence of Escherichia coli 16S ribosomal RNA: complementarity to nonsense triplets and ribosome binding sites. Proc. Natl. Acad. Sci. USA 71: Slots, J., and R. J. Genco Black-pigmented Bacteroides species, Capnocytophaga species, and Actinobacillus actinomycetemcomitans in human periodontal disease: virulence factors in colonization, survival, and tissue destruction. J. Dent. Res. 63: Smalley, J. W., and A. J. Birss Trypsin-like enzyme activity of the extracellular membrane vesicles of Bacteroides gingivalis W50. J. Gen. Microbiol. 133: Smalley, J. W., A. J. Birss, and C. A. Shuttleworth The degradation of type I collagen and human plasma fibronectin by the trypsinlike enzyme and extracellular membrane vesicles of Bacteroides gingivalis W50. Arch. Oral Biol. 33: Sorsa, T., V. J. Uitto, K. Suomalainen, H. Turto, and S. Lindy A trypsinlike protease from Bacteroides gingivalis: partial purification and characterization. J. Periodontal Res. 22: Takahashi, N., T. Kato, and H. K. Kuramitsu Isolation and preliminary characterization of the Porphyromonas gingivalis prtc gene expressing collagenase activity. FEMS Microbiol. Lett. 84: Tsutsui, H., T. Kinouchi, Y. Wakano, and Y. Ohnishi Purification and characterization of a protease from Bacteroides gingivalis 381. Infect. Immun. 55: Vieira, J., and J. Messing The puc plasmids, an M13mp7-derived system for insertion mutagenesis and sequencing with synthetic universal primers. Gene 19: Vieira, J., and J. Messing Production of single-stranded plasmid DNA. Methods Enzymol. 153: Yoshimura, F., M. Nishikata, T. Suzuki, C. I. Hoover, and E. Newbrun Characterization of a trypsin-like protease from the bacterium Bacteroides gingivalis isolated from human dental plaque. Arch. Oral Biol. 29: