Overexpression, Purification, and Refolding of a Porphyromonas gingivalis Cysteine Protease from Escherichia coli

Protein Expression and Purification 18, 262–268 (2000) doi:10.1006/prep.2000.1193, available online at http://www.idealibrary.com on

Overexpression, Purification, and Refolding of a Porphyromonas gingivalis Cysteine Protease from Escherichia coli Mai Brigid Margetts, 1 Ian George Barr, and Elizabeth Ann Webb Research and Development Division, CSL Ltd., 45 Poplar Road, Parkville, Victoria 3052, Australia

Received August 5, 1999, and in revised form December 2, 1999

This paper describes the overexpression of the Rgp-1 (arginine) protease domain from Porphyromonas gingivalis. This protease and the related Kgp (lysine) protease, both of which display trypsin-like specificity, have been implicated as major virulence factors and may play a significant role in the etiology of periodontal disease. Both Rgp-1 and Kgp are initially translated as polyproteins, each containing a protease domain and multiple adhesin domains. The Rgp-1 protease domain was expressed in E. coli, purified, refolded, and assayed for activity. These expression studies demonstrated that prior to the formation of inclusion bodies in the E. coli cytoplasm, the protease was proteolytically active and could hydrolyze a specific synthetic substrate. When the Rgp-1 protease domain was purified from inclusion bodies and refolded, it was found to be autolytically active and displayed specific catalytic activity. This is the first report on the expression and purification of active Rgp-1 from E. coli. Polyclonal antisera raised against recombinant protein recognized the native form of the protease in the P. gingivalis strain W50, indicating that the recombinant protein contained some of the antigenic determinants of the native protease. © 2000 Academic Press

Porphyromonas gingivalis is now recognized as an important etiological agent in adult periodontitis, a chronic inflammatory disease of the supporting structures of the teeth. The bacterium is commonly found in the anaerobic environment of the periodontal pocket (1), where it is capable of causing extensive tissue breakdown. P. gingivalis possesses a number of potential virulence factors such as proteases, adhesins, and 1 To whom correspondence should be addressed. Fax: ⫹61 3 9388 2063. E-mail: [email protected].

262

hemaglutinins that may play a significant role in adherence of the cells to oral tissue surfaces and other bacteria (2). As well as being a major cause of tooth loss in adults, periodontal disease is now recognized as a significant risk factor for both cardiovascular disease and preterm labor in humans (3,4). Biochemical investigations of the extracellular and cell-associated proteins of P. gingivalis have led to the identification of a number of distinct proteolytic enzymes. Among these are the cysteine proteases with specificity for arginine (Arg-X) and lysine (Lys-X) containing peptide bonds. Over recent years many reports have described the genetic and biochemical characterization of these proteases (5–14). It has been suggested that up to 95% of the total proteolytic activity of P. gingivalis results from three distinct gene loci and their gene products, Kgp (GenBank Accession No. U54691), the Lys-specific protease, and the Arg-specific proteases Rgp-1 (GenBank Accession No. U15282) and Rgp-2 (GenBank Accession No. U85038). The nomenclature of the proteases is somewhat confusing, largely due to the different P. gingivalis strains used for genetic and biochemical characterization. Consensus nomenclature is now being established, which should alleviate this confusion. At present, Kgp is also referred to as PrtP (5) or PrtK (13), Rgp-1 is also designated PrpR1 (12) or prtR (10), and Rgp-2 is also called PrtRII (14) or prR2 (8). The gene structure of Kgp and Rgp-1 from P. gingivalis H66, which demonstrates the association of proteases and adhesin sequences, is depicted in Fig. 1. Physical separation of Rgp-1, Kgp, and their proteolytically cleaved products from wild-type P. gingivalis cultures has proven to be a difficult task. Autodegradation and the possibility that these proteases degrade each other during purification procedures (11) compound the difficulty in isolating biochemically pure protease. 1046-5928/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

PRODUCTION OF RECOMBINANT Porphyromonas gingivalis Arg PROTEASE

The production of recombinant forms of Rgp-1 and Kgp may be a means of overcoming some of these problems. Past attempts to obtain functionally active recombinant protease have involved the expression of either the catalytic domain alone or the entire polyprotein sequence in several expression systems, including E. coli, Bacteroides, and insect cells. The cloning of the full-length Rgp-1 (Arg protease) polyprotein sequence in E. coli (12) failed to demonstrate expression of active recombinant protease. This was postulated to be due to inappropriate processing of the polyprotein within the host cells. More recently, the full polyprotein sequence of the Kgp gene (Lys protease) was expressed in a Bacteroides species using an E. coli shuttle vector and catalytically active enzyme was demonstrated by fibrinogen zymography (15). One report on the expression of the catalytic domain of Kgp in insect cells described the release of active Kgp into the supernatant of baculovirus-infected cultures (7). Another report described the refolding of recombinant Kgp, derived from E. coli, with protein disulfide isomerase, which led to increased hemoglobin binding activity but reduced protease activity (16). Given that purification of the native Rgp-1 and Kgp proteases from wild-type P. gingivalis cultures is difficult, due largely to the formation of heterogeneous multimeric complexes and autolysis, we independently cloned and expressed the catalytic domain of Rgp-1 in E. coli. The purpose of this work was to purify and refold the recombinant protease, with a view to providing sufficient material for testing as a vaccine candidate in P. gingivalis animal models.

263

FIG. 2. Expression vector pET28a/Rgp-1 encoding N-terminal His 6 and T7 tags fused to the Arg cysteine protease domain.

MATERIALS AND METHODS

Construction of Recombinant E. coli The Rgp-1 protease domain was amplified by PCR using genomic DNA isolated from P. gingivalis W50. Amplification was performed using the forward oligonucleotide 5⬘-GCGC-AGA-TCT-TAC-ACA-CCG-GTAGAG-G-3⬘, incorporating a BglII site on the 5⬘ end of the gene, and the reverse oligonucleotide 5⬘-GCGCGTC-GAC-TTA-GCG-AAG-AAG-TTC-GGG-G-3⬘, incorporating a SalI site on the 3⬘ end of the gene. PCR product was restricted with BglII/SalI and forced into the same sites of the baculovirus transfer vector pBluebacHis2b (Invitrogen Corp.). Inserts were then subcloned into the BamHI/SalI sites of the E. coli expression vector pET28a (Novagen). This created a fusion with an N-terminal hexahistidine (H 6) 2 and T7 tag (Fig. 2). All cloning steps were carried out in the E. coli host ER1793 (a gift from Elizabeth Raleigh, New England Biolabs). Expression of Recombinant E. coli

FIG. 1. Schematic representations of the Rgp-1 and Kgp polyproteins characterized from P. gingivalis strain HG66 (7). Each comprises a prosequence followed by a cysteine protease sequence and four adhesin domains, the latter of which are referred to by their size in kilodaltons (as shown by the numbers within the boxes). The Arg and Lys proteases are 491 and 509 amino acids, respectively, in length. The 15-, 17-, and 27-kDa adhesins are identical in each polyprotein and the 44-kDa adhesin in Rgp-1 has a high homology to the 27/44 hybrid adhesin in the Kgp sequence. This hybrid (44.7 kDa) comprises a fusion of the first 147 amino acids of the 27-kDa adhesin with a sequence which is 99% homologous to the 44-kDa adhesin in Rgp-1.

The pET28a/Rgp-1 construct was transformed into the E. coli expression host HMS174(DE3) (Novagen). Single-colony transformants were used to inoculate 10-ml Luria–Bertani broths containing 50 ␮g/ml kanamycin and cultures were grown at 37°C until the optical density (OD 600) was ⬇1.0. Two milliliters of this 2

Abbreviations used: H 6, hexahistidine; Ni–NTA, nickel–nitrilotriacetic acid; DL-BapNA, N-benzoyl-DL-arginine p-nitroanilide; aa, amino acid; TNG buffer, 50 mM Tris–HCl, 500 mM NaCl, and 8% glycerol, pH 7.8.

264

MARGETTS, BARR, AND WEBB

inoculum was used to seed 20 ml of Terrific Broth containing potassium phosphates plus 50 ␮g/ml kanamycin and cultures were allowed to reach an OD 600 of ⬇0.6 before induction with IPTG (0.1 mM). Uninduced cultures were set up as controls and 1-ml samples were taken at regular intervals postinduction. Cells were harvested by centrifugation at 10,000g for 3 min and the bacterial pellets resuspended to an OD 600 of 8.0 in PBS and stored at ⫺20°C. For purification purposes inclusion bodies were extracted from 500-ml cultures harvested at 3 h postinduction. Solubility Studies One-milliliter cultures (OD 600 ⫽ 8.0) were centrifuged and the pellets resuspended in 0.1 ml of 50 mM Tris–HCl (pH 8.0,) 2 mM EDTA. Lysosyme was added to a concentration of 100 ␮g/ml from a 10 mg/ml freshly prepared stock solution. Ten microliters of 1% Triton X-100 was added and the lysates incubated at 30°C for 15 min. To shear the DNA, the lysates were then placed on an ice bath and sonicated (Virtis Virsonic 475-W cell disrupter) for two 20-s pulses and centrifuged at 12,000g for 15 min at 4°C. The supernatants containing the soluble fraction were mixed with an equal volume of 2⫻ SDS sample-reducing buffer for loading onto SDS–PAGE gels. In addition, the pellets containing the insoluble fraction (inclusion bodies) were resuspended in 1⫻ SDS sample-reducing buffer for SDS–PAGE analysis. Isolation and Solubilization of Inclusion Bodies Bacterial pellets were thawed on ice and resuspended in binding buffer (5 mM imidazole, 500 mM NaCl, 20 mM Tris–HCl, pH 7.9) and then sonicated as described above. Lysates were centrifuged at 20,000g to collect the inclusion bodies. Pellets were resuspended and the process of sonication and centrifugation was repeated twice more to release trapped proteins. Pellets were finally resuspended in binding buffer containing 6 M urea and incubated on ice for 2–3 h with continuous stirring to solubilize proteins. Any remaining insoluble material was removed by centrifuging at 39,000g for 20 min. The supernatant was filtered through a 0.45-␮m membrane before column purification. Nickel Nitrilotriacetic Acid (Ni–NTA) Purification of Solubilized Inclusions Ni–NTA metal affinity chromatography was used to purify the recombinant protease. His-tagged protein was batch bound to an equilibrated Ni–NTA resin (Qiagen), which was then poured into a small column, and unbound protein was eluted under gravity. The column was washed with 10 volumes of binding buffer followed

by 5 column volumes of wash buffer (60 mM imidazole, 500 mM NaCl, 20 mM Tris–HCl, 6 M urea, pH 7.9). Target protein was eluted in buffer containing 1 M imidazole, 500 mM NaCl, 20 mM Tris–HCl, 6 M urea, pH 7.9. Refolding of Recombinant Protease A number of methods for protein refolding were investigated: Rapid or gradual removal of denaturant. Protein was refolded by rapid dialysis into buffer containing 5 mM DTT, 5 mM cysteine, 100 mM Tris–HCl, 150 mM NaCl, 5 mM CaCl 2, pH 7.8 or by gradual removal of urea upon dialysis into buffer containing 50 mM Tris– HCl, 500 mM NaCl, 8% glycerol, pH 7.8. Oxido shuffling using oxidized and reduced glutathione. Protein eluted from the Ni–NTA column was dialyzed into a buffer containing 0.2 M L-arginine, 0.1 M Tris–HCl, 2 mM EDTA, 5 mM GSH, 0.5 mM GSSG, pH 7.8. Cosolvent-assisted protein folding. PEG-8000 was added to the refolding buffers described at a level of 30 g/liter in order to assist refolding. In situ column refolding (17). Bacterial pellets were thawed on ice and resuspended in a buffer containing 6 M guanidine hydrochloride, 20 mM Tris–HCl (pH 7.9), 500 mM NaCl, and 4 mM n-octyl glucoside (Buffer 1). The suspension was sonicated as described previously and centrifuged at 20,000g for 30 min at 4°C. The supernatant was applied to the Ni–NTA resin pre-equilibrated in Buffer 1 and the column was washed with Buffers 2 and 3, respectively. Buffer 2 contained 6 M urea, 20 mM Tris–HCl (pH 7.9), 500 mM NaCl, and 20 mM imidazole. Buffer 3 consisted of 20 mM Tris–HCl (pH 7.9) and 150 mM NaCl. Protein was eluted off the column using a buffer containing 20 mM Tris–HCl (pH 7.9), 150 mM NaCl, and 200 mM imidazole. Enzyme Assays Purified protein was incubated at 25°C with the substrate N-benzoyl-DL-arginine p-nitroanilide (DLBapNA) (Sigma) at a final concentration of 1.0 mM. Absorbance was measured at 410 nm using a Hewlett Packard 8452A diode array spectrophotometer and the amidolytic activity expressed in U, where U ⫽ ␮mol of substrate converted min ⫺1 at 25°C. The protein concentration of purified fractions was determined using the Bradford method (Pierce Coomassie Plus) with BSA as a standard. For crude cell lysates the reaction was assessed by visualization of color intensity rather than by spectroscopy.

PRODUCTION OF RECOMBINANT Porphyromonas gingivalis Arg PROTEASE

FIG. 3. Expression of the pET28a/Rgp-1 construct in E. coli HMS174(DE3). Samples were analyzed by SDS–PAGE on 12% polyacrylamide gels followed by Coomassie blue staining. (A) Whole-cell lysates: Lanes 1, 3, and 5, uninduced samples at 2, 4, and 6 h; Lanes 2, 4, and 6, induced samples at 2, 4, and 6 h postinduction. (B) Separation of E. coli expressed Arg protease into insoluble (Lane 1) and soluble (Lane 2) fractions at 1 h postinduction.

Polyacrylamide Gel Electrophoresis Samples were mixed with an equal volume of 2⫻ sample-reducing buffer, boiled for 10 min at 95°C, and separated electrophoretically by SDS–PAGE as described by Laemmli (18). Western blots were prepared by electroblotting proteins onto nitrocellulose for 1 h at 100 V. Membranes were blocked with 1% casein solution before incubating with primary antibody. Antisera Polyclonal antiserum was raised to the recombinant protein by dosing BALB/c mice subcutaneously twice with 20 ␮g of recombinant protein in Freund’s incomplete adjuvant, 3 weeks apart. Mice were bled 1 week after the second dose and the antiserum generated was used to screen Western blots of whole-cell P. gingivalis W50. Rabbit serum raised to a peptide (FNGGISLANYTGHGSETAWGT conjugated to diphtheria toxoid), found specifically in the catalytic domain of the arginine-specific protease in strain W50, was kindly provided by Dr. Neil O’Brien-Simpson. RESULTS AND DISCUSSION

Expression of pET28a/Rgp-1 in HMS174(DE3) Rgp-1 was expressed with a 34 amino acid (aa) fusion on the N-terminus which contained a H 6 tag and T7 tag for purification and detection, respectively. Protein was produced within large inclusions in the E. coli cytoplasm and Fig. 3A shows SDS–PAGE analysis of the expression pattern for the recombinant protease in

265

E. coli strain HMS174(DE3) over a 6-h period. A major protein band of around 58 kDa (the expected size for the Rgp-1 protease (526 aa)) is shown to be present, indicating the sequence to be extremely well expressed in this E. coli system. Small inclusion bodies were observed in the cytoplasm by transmission electron microscopy as early as 15 min postinduction (data not shown). Isolation of samples into soluble and insoluble fractions clearly demonstrated the recombinant protease to be almost entirely contained in the insoluble fraction from 1 h postinduction onward (Fig. 3B). An interesting observation was the decrease in amount of host cell proteins in these samples. This could be due to proteolytic activity of the protease during early expression and processing or to toxicity of the expressed protein negatively affecting cell growth and hence reducing the total amount of protein. It would appear that the latter is less likely, given that no significant reduction in growth was observed and that the recombinant protease itself continued to be highly produced and transported to inclusion bodies. Qualitative activity assessments were performed on crude cell lysates (freeze–thawed cell pellets) of both uninduced and induced samples using the chromogenic substrate DL-BapNA. The level of specific activity was low and a number of hours were required to demonstrate substrate conversion. Uninduced samples were shown to have around threefold higher activity than induced samples. Furthermore, samples taken at the earlier time points of 15 and 30 min had a higher activity than those taken at the later time points (⬎1 h). In the pET–T7 promoter expression system, target protein production is controlled by IPTG induction, which provides a source of T7 RNA polymerase in the host cell. Even in the absence of IPTG there is some expression of T7 RNA polymerase from the lacUV5 promoter in ␭DE3 lysogens. The observation that protease activity was higher in uninduced samples was probably a result of this apparent “leakiness,” where the target protein is expressed at very low levels and at a slower rate, conditions under which native refolding is likely to occur. Crude cell lysates denatured by the addition of SDS sample-reducing buffer displayed little or no activity, as did lysates from E. coli HMS174(DE3) devoid of plasmid. This expression study indicated the presence of active, recombinant protease molecules in the E. coli cells prior to their sequestration into inclusion bodies and demonstrated the ability of the recombinant Rgp-1 protease to hydrolyze a specific substrate in the presence of host cell proteins.

266

MARGETTS, BARR, AND WEBB

FIG. 4. Inactive and active forms of the recombinant Arg protease. Samples were analyzed by SDS–PAGE on 12% polyacrylamide gels followed by Coomassie blue staining. Lane 1, soluble Arg protease refolded in TNG buffer; Lane 2, soluble Arg protease refolded in buffer containing 5 mM DTT, 5 mM, cysteine, and 5 mM CaCl 2. Note that the latter has undergone substantial proteolytic degradation.

Refolding and Activity of Recombinant Rgp-1 Protease At present it is unknown whether the specific activity of this protease is of importance in its performance as a vaccine candidate. We sought to answer this question by preparing catalytically active recombinant protein for testing in animal protection experiments. Recombinant protease was purified in a one-step procedure by Ni–NTA metal affinity chromatography. The first renaturation strategy applied, primarily to investigate solubility, was the stepwise dialysis from denaturant into buffer containing salts and glycerol (50 mM Tris–HCl, 500 mM NaCl, and 8% glycerol, pH 7.8 (TNG buffer)). Protein remained soluble under these conditions but had neither proteolytic nor substrate specific activity. To promote formation of disulfide bonds representative of the native structure, a buffer was formulated with the low molecular weight thiols GSH and GSSG. The protease was soluble in this buffer but displayed a weak specific activity (⬍0.05 U/mg). In situ column refolding resulted in the same outcome. The addition of high levels of PEG-8000 as a cosolvent resulted in precipitation of the protease. The cysteine proteases of P. gingivalis are known to be activated by reducing agents and stabilized by calcium ions (19,20); hence a buffer was formulated to contain 5 mM DTT, 5 mM cysteine, and 5 mM CaCl 2. Following dialysis into this refolding buffer, the Rgp-1 protease remained soluble and was found to be proteolytically active, undergoing autolytic degradation. Figure 4 shows the degree of proteolysis when compared to soluble Rgp-1 in TNG buffer without the additions. The specific activity of autolytically active recombinant enzyme using the substrate L-BapNA was assayed at 0.14

U/mg, which was marginally higher than that obtained from the other refolding methods attempted. This compares with 78 U/mg for a preparation of highly purified native Arg (Rgp-2) protease (Dr. Yu-Yen Chen, Melbourne University). Autolysis of the recombinant Rgp-1 preparation was prevented by the addition of 100 ␮M N-␣-tosyl-L-lysine chloromethyl ketone, which is an irreversible inhibitor of serine and cysteine proteases, to the purified protein before dialysis (data not shown). Rgp-1 contains 11 cysteine residues, with 1 cysteine residue contained within the active site forming a catalytic triad with histidine and asparagine moieties (7). Hence, formation of correct disulfide linkages is likely to be important in maintaining the activity of the cysteine proteases. Although the reducing environment of the E. coli cytoplasm is not conducive to disulfide bond formation, the recombinant protease was shown to possess a degree of activity, particularly in the earlier stages of expression. This showed the potential for refolding in such a manner as to regain higher amidolytic activity. Another factor to consider is the role of glycosylation in P. gingivalis and whether this posttranslational modification has any effect on the specific activity of the cysteine proteases. Recent work has shown that both the arginine and lysine proteases of P. gingivalis display great diversity in the extent of glycosylation and in the types of sugars present on the individual protease isoforms (21). Whether glycosylation issues impact on structure, stability, or specific activity of the enzymes has yet to be determined. The yield of purified recombinant protease from the one-step affinity chromatography method described was typically 60 –70 mg per 9 –10 g (wet weight) of E. coli (Table 1). Purity, which was estimated by scanning laser densitometry of Coomassie-stained gels, was found to increase from 57% in solubilized inclusion bodies to 80% following chromatography (Fig. 4, Lane 1). Upon refolding to regain activity, autolysis of the protease reduced purity of the full-length product to 40%, while the major degradation products increased to 30 and 25% (Fig. 4, Lane 2). Characterization of Recombinant Rgp-1 Protease Immunoblot analysis using antiserum raised to the protease peptide FNGGISLANYTGHGSETAWGT revealed a major immunoreactive protein at the size expected for the Rgp-1 protease, which was not present in a negative control of E. coli HMS174(DE3) cells (Fig. 5A), thus confirming the identity of the recombinant protein. In addition to this product, other immunoreactive species were present, indicating that, although the protein appeared homogeneous on a Coomassiestained gel, there are, in fact, small amounts of both aggregates and degradation products present. This ob-

PRODUCTION OF RECOMBINANT Porphyromonas gingivalis Arg PROTEASE

267

TABLE 1 Purification of Recombinant Rgp-1

Fraction

Protein (mg)

Percentage recombinant

Amount of recombinant (mg)

Percentage yield

Fold purification

Induced lysate Insoluble fraction Column eluate

10,000 200 70

30 57 80

300 114 56

100 38 18.7

1 1.9 2.7

servation was confirmed by silver-stained gel analysis (Fig. 5B). N-Terminal sequence analysis of the two major species present in the active, refolded protease (Fig. 4, Lane 2: 58 and ⬃54 kDa) showed both to have the expected N-terminal amino acids of the Arg protease, indicating that the smaller product is degraded at the C-terminus. This work demonstrated that it was possible to derive purified, active arginine-specific protease from E. coli and that the adhesin domains of the Rgp-1 locus (22) were not necessary for activity. This finding is consistent with previously reported studies on Kgp in insect cells (7). The fusion of an artificial tag (H 6/T7), for rapid purification, on the N-terminus of the proteins did not appear to adversely affect activity. Antibody Responses to Recombinant Arginine Protease Polyclonal antiserum raised to the recombinant protease was used to evaluate P. gingivalis whole-cell

lysates on Western blots and Fig. 5C shows the reactivity of antiserum raised against recombinant Rgp-1 with P. gingivalis W50. A major band was detected at 50 kDa, consistent with monomeric arginine protease, as it has previously been demonstrated that native Rgp-1 migrates at 50 kDa on SDS–PAGE gels (12). Also detected was a diffuse band running between 70 and 100 kDa. This may represent reactivity with the protein components of highly glycosylated or multimeric forms of the native arginine protease (6). Smaller immunoreactive bands between 36 and 30 kDa may be breakdown products of the 50-kDa monomer. In conclusion, it appears that polyclonal antiserum raised against the recombinant protease recognizes native P. gingivalis proteases under denaturing, reducing conditions. This is encouraging since this antigen, while not highly enzymatically active, may still be a useful periodontal disease vaccine candidate. In the future, it will be of interest to test this recombinant antigen in P. gingivalis animal protection experiments and to compare directly, where possible, to the native purified counterparts. A recent report on the crystallization of RGP-A (23) may provide insights that will help in the refolding of recombinant forms of the protease from inclusion bodies and improve the proteolytic activity. ACKNOWLEDGMENTS We thank our collaborators at the School of Dental Science, University of Melbourne, particularly Professor Eric Reynolds, for valuable discussions and Dr. Yu-Yen Chen and Rita Paolini for enzyme assays and antisera generation. We are also grateful to Ross Hamilton and his team at CSL Ltd. for electron microscopy studies and to Peter Schoofs for providing N-terminal sequence and densitometry analyses.

REFERENCES FIG. 5. Characterization of the recombinant Arg protease. (A) Immunoblot with membrane probed using antiserum raised to a peptide sequence from the arginine-specific protease. Lane 1, recombinant Arg protease; Lane 2, E. coli HMS174(DE3). (B) Lane 3, from the same gel showing silver staining of the recombinant Arg protease. (C) Immunoblot of P. gingivalis W50 whole cells: membrane probed with an antiserum raised to recombinant Arg protease. Note the detection of monomeric Arg protease at 50 kDa and high molecular weight forms at ⬃70 –100 kDa, as well as truncated products (⬍36 kDa).

1. Griffen, A. L., Becker, M. R., Lyons, S. R., Moeschberger, M. L., and Leys, E. J. (1998) Prevalence of Porphyromonas gingivalis and periodontal health status. J. Clin. Microbiol. 36, 3239 –3242. 2. Mayrand, D., and Holt, S. C. (1988) Biology of asaccharolytic black pigmented Bacteroides species. Microbiol. Rev. 52, 134 – 152. 3. Offenbacher, S., Jared, H. L., O’Reilly, P. G., Wells, S. R., Salvi, G. E., Lawrence, H. P., Socransky, S. S., and Beck, J. D. (1988) Potential pathogenic mechanisms of periodontitis associated pregnancy complications. Ann. Periodontol. 3, 233–250.

268

MARGETTS, BARR, AND WEBB

4. Beck, J. D., Offenbacher, S., Williams, R., Gibbs, P., and Garcia, R. (1998) Periodontitis: A risk factor for coronary heart disease? Ann. Periodontol. 3, 127–141. 5. Barkocy-Gallagher, G. A., Han, N., Patti, J. M., Whitlock, A., Progulske-Fox, A., and Lantz, M. S. (1996) Analysis of the prtP gene encoding porphypain, a cysteine proteinase of Porphyromonas gingivalis. J. Bacteriol. 178, 2734 –2741. 6. Aduse-Opoku, J., Rangarajan, M., Young, K. A., and Curtis, M. A. (1998) Maturation of the arginine-specific proteases of Porphyromonas gingivalis W50 is dependent on a functional prR2 protease gene. Infect. Immun. 66, 1594 –1600. 7. Pavloff, N., Pemberton, P. A., Potempa, J., Chen, W. C. A., Pike, R. N., Prochazka, V., Kiefer, M. C., Travis, J., and Barr, P. (1997) Molecular cloning and characterization of Porphyromonas gingivalis lysine-specific gingipain. J. Biol. Chem. 272, 1595–1600. 8. Rangarajan, M., Aduse-Opoku, J., Slaney, J. M., Young, K. A., and Curtis, M. A. (1997) The prpR1 and prR2 arginine-specific protease genes of Porphyromonas gingivalis W50 produce five biochemically distinct enzymes. Mol. Microbiol. 23, 955–965. 9. Pavloff, N., Potempa, J., Pike, R. N., Prochazka, V., Kiefer, M. C., Travis, J., and Barr, P. J. (1995) Molecular cloning and structural characterisation of the Arg-gingipain proteinase of Porphyromonas gingivalis. J. Biol. Chem. 270, 1007–1010. 10. Slakeski, N., Cleal, S. M., and Reynolds, E. C. (1996) Characterization of a Porphyromonas gingivalis gene prtR that encodes an arginine-specific thiol proteinase and multiple adhesins. Biochem. Biophys. Res. Commun. 224, 605– 610. 11. Ciborowski, P., Nishikata, M., Allen, R. D., and Lantz, M. S. (1994) Purification and characterization of two forms of a highmolecular-weight cysteine proteinase (porphypain) from Porphyromonas gingivalis. J. Bacteriol. 176, 4549 – 4557. 12. Aduse-Opoku, J., Muir, J., Slaney, J. M., Rangarajan, M., and Curtis, M. A. (1995) Characterization, genetic analysis, and expression of a protease antigen (PrpR1) of Porphyromonas gingivalis W50. Infect. Immun. 63, 4744 – 4754. 13. Slakeski, N., Cleal, S. M., Bhogal, P. S., and Reynolds, E. C. (1999) Characterization of a Porphyromonas gingivalis gene prtK that encodes a lysine-specific cysteine proteinase and three sequence-related adhesins. Oral Microbiol. Immunol. 14, 92–97.

14. Slakeski, N., Bhogal, P. S., O’Brien-Simpson, N. M., and Reynolds, E. C. (1998) Characterization of a second cell-associated Arg-specific cysteine proteinase of Porphyromonas gingivalis and identification of an adhesin-binding motif involved in association of the prtR and prtK proteinases and adhesins into large complexes. Microbiology 6, 1583–1592. 15. Barkocy-Gallagher, G. A., Foley, J. W., and Lantz, M. S. (1999) Activities of the Porphyromonas gingivalis PrtP proteinase determined by construction of prtP-deficient mutants and expression of the gene in Bacteroides species. J. Bacteriol. 181, 246 – 255. 16. Kuboniwa, M., Amano, A., and Shizukuishi, S. (1998) Hemoglobin-binding protein purified from Porphyromonas gingivalis is identical to lysine-specific cysteine proteinase (lys-gingipain). Biochem. Biophys. Res. Commun. 249, 38 – 43. 17. Holzinger, A., Phillips, K. S., and Weaver, T. E. (1996) Singlestep purification/solubilization of recombinant proteins: Application to surfactant protein B. Biotechniques 20, 804 – 808. 18. Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680 – 685. 19. Otogoto, J., and Kuramitsu, H. K. (1993) Isolation and characterization of the Porphyromonas gingivalis PrtT gene, coding for protease activity. Infect. Immun. 61, 117–123. 20. Chen, Z., Potempa, J., Polanowski, A., Wikstrom, M., and Travis, J. (1992) Purification and characterization of a 50-kDa cysteine proteinase (gingipain) from Porphyromonas gingivalis. J. Biol. Chem. 267, 18896 –18901. 21. Rangarajan, M., Paramonov, N., Hounsell, E. F., and Curtis, M. A. (1999) Diversity in the glycosylation of Arg-X proteases of P. gingivalis. J. Dent. Res. 78, IADR Abstracts, No. 1514, 295. 22. Bhogal, P. S., Slakeski, N., and Reynolds, E. C. (1997) A cellassociated protein complex of Porphyromonas gingivalis W50 composed of Arg- and Lys- specific cysteine proteases and adhesins. Microbiology 143, 2485–2495. 23. Eichinger, A., Beisel, H. G., Jacob, U., Huber, R., Medrano, F. J., Banbula, A., Potempa, J., Travis, J., and Bode, W. (1999) Crystal structure of gingipain R: An Arg-specific bacterial cysteine proteinase with a caspase-like fold. EMBO J. 18(20), 5453–5462.