Specific In Situ Visualization of Plasma Cells Producing Antibodies against Porphyromonas gingivalis in Gingival Radicular Cyst

Abstract

The enzyme-labeled antigen method was applied to visualize plasma cells producing antibodies to Porphyromonas gingivalis, flora of the human oral cavity. Antibodies to P. gingivalis have reportedly been detected in sera of patients with periodontitis. Biotinylated bacterial antigens, Ag53, and four gingipain domains (Arg-pro, Arg-hgp, Lys-pro, and Lys-hgp) were prepared by the cell-free protein synthesis system using the wheat germ extract. In paraformaldehyde-fixed frozen sections of rat lymph nodes experimentally immunized with Ag53-positive and Ag53-negative P. gingivalis, plasma cells were labeled with biotinylated Arg-hgp and Lys-hgp. Antibodies to Ag53 were detected only in the nodes immunized with Ag53-positive bacteria. In two of eight lesions of gingival radicular cyst with inflammatory infiltration, CD138-positive plasma cells in frozen sections were signalized for Arg-hgp and Lys-hgp. An absorption study using unlabeled antigens confirmed the specificity of staining. The AlphaScreen method identified the same-type antibodies in tissue extracts but not in sera. Antibodies to Ag53, Arg-pro, and Lys-pro were undetectable. In two cases, serum antibodies to Arg-hgp and Lys-hgp were AlphaScreen positive, whereas plasma cells were scarcely observed within the lesions. These findings indicate the validity of the enzyme-labeled antigen method. This is the very first application of this novel histochemical technique to human clinical samples.

A variety of infectious, autoimmune, and neoplastic lesions contain numerous antibody-producing plasma cells as part of inflammatory cells. In most cases, target antigens of the antibodies produced by these plasma cells in the lesion remain unknown. The antibodies produced within the lesion should directly be related to pathogenesis.

The enzyme-labeled antigen method is a reversed form of the enzyme-labeled antibody method or immunoperoxi- dase technique. This histochemical technique was first proposed as early as 1968, in an experimental model to detect anti–horseradish peroxidase (HRP) antibody-producing cells in the lymph nodes of rats immunized with HRP (Leduce et al. 1968; Straus 1968). However, no further application has been reported since we revived this technique in 2009 to visualize the site of specific antibody production in the lymph nodes and spleen of rats immunized with HRP, keyhole limpet hemocyanin (KLH), and ovalbumin (Mizutani et al. 2009). Biotinylated KLH and ovalbumin were prepared as probes for this technique. When the labeled antigen is available, the enzyme-labeled antigen method can thus visualize the site of antibody production in sections of human clinical samples.

In the present study, we analyzed the site of antibody production against Porphyromonas gingivalis (Bacteroides gingivalis) in human surgical specimens of gingival radicular cyst. P. gingivalis is a black-pigmented, non-motile, obligatory anaerobic gram-negative bacillus normally residing in the human oral cavity and abnormally colonizing in lesions of pyorrheal gingivitis/periodontitis (Cutler et al. 1995; Lamont and Jenkinson 1998; Holt et al. 1999; Slots and Ting 1999). Biofilm formation was noted when colonizing in the gingival lesion (Lamont and Jenkinson 1998). Antibodies to P. gingivalis have been detected in the serum of patients with periodontitis (Kurihara et al. 1991; Schenkein 2006). Radicular cyst, also called an apical cyst, is associated with dental caries and often shows marked inflammatory reaction with dense plasma cell infiltration (Dahlen 2002). Microbiological analysis of radicular cyst indicated some 50 species of bacterial colonization, mainly anaerobes such as Porphyromonas (or Bacteroides), Fusobacterium, Tannerella, Prevotella, Peptostreptococcus, Actinomyces, and so on (Jung et al. 2000; Dahlen 2002; Wayman et al. 2002; Sunde et al. 2003; Noguchi et al. 2005).

The enzyme-labeled antigen method was herein applied to demonstrating plasma cells producing specific antibodies to P. gingivalis–related antigens in prefixed frozen sections of radicular cyst. We chose the following five proteins as target antigens, including Ag53, a 53-kDa membrane-associated protein (Kurihara et al. 1991; Oyaizu et al. 2001), and P. gingivalis–associated cysteine proteases, Arg-gingipain and Lys-gingipain (Chen et al. 2001; O’Brien-Simpson et al. 2001; Grenier and Tanabe 2010). Rat lymph nodes experimentally immunized with Ag53-positive and Ag53-negative P. gingivalis were also successfully evaluated.

For effectively preparing biotin-labeled antigens, we used the cell-free protein synthesis system using the wheat germ extract, which has been established by the Cell-Free Science and Technology Research Center, Ehime University (Sawasaki, Hasegawa, et al. 2002). The specific antibodies in the serum were easily identified with the AlphaScreen method (Matsuoka et al. 2010).

Application of the enzyme-labeled antigen method may help the pathogenesis be solved. Plasma cells infiltrating in the lesion of autoimmune disease should produce autoantibodies, those in the infectious lesion should produce antibodies against pathogens, and those in the neoplastic lesion may produce antibodies against tumor cells. Once the labeled antigen is available, our methodology may have powerful potential in morphological and histochemical research and also in clinicopathological applications.

Material and Methods

Bacterial Strains and Growth Condition

Two strains of P. gingivalis ATCC33277 and FDC381 and Streptococcus mutans 854S were obtained from American Type Culture Collection (Manassas, VA). P. gingivalis ATCC33277 lacks Ag53 expression (Naito et al. 2008), and FDC381 expresses Ag53 (Oyaizu et al. 2001). The bacteria were maintained under anaerobic conditions in the Department of Pathophysiology–Periodontal Science, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama, Japan. Briefly, P. gingivalis was grown in GAM broth (Nissui Seiyaku, Tokyo, Japan) supplemented with 0.0005% hemin and 0.0001% vitamin K3 at 37C for 24 to 48 hr in an anaerobic box (model ANX-1; Hirasawa Works, Tokyo, Japan) containing 80% N₂, 10% H₂, and 10% CO₂, as previously described (Kato et al. 2008). S. mutans was grown in Tripticate Soy Broth (BD Biosciences, San Jose, CA) supplemented with 0.5% yeast extract (BD Biosciences), as previously described (Yoshida and Kuramitsu 2002). DNA was purified from P. gingivalis FDC381 and S. mutans 854S.

Experimental Animals

Male Sprague Dawley rats aged 5 weeks and weighing 150 g (Chubu Kagaku, Nagoya, Japan) were kept in the animal laboratory of Fujita Health University, Toyoake, Japan. The animal experiments were conducted in accordance with the Guidelines for the Management of Laboratory Animals at Fujita Health University (acknowledgment no. M2104, April 2008).

Immunization and Tissue Sampling

The animals were immunized with two strains of P. gingivalis, Ag53 positive and Ag53 negative. The boiled bacteria emulsified in Freund’s complete (first challenge) or incomplete (second and third challenges) adjuvant were injected three times in 5 weeks into footpads of the rats (n=3 in the respective group), and the axillary and popliteal lymph nodes, as well as the serum, were sampled 14 days after the third immunization. Rats (n=2) immunized only with Freund’s adjuvant served as negative controls.

Under the diethyl ether anesthesia, part of the axillary and popliteal lymph nodes was sampled and fixed in buffered 4% paraformaldehyde, pH 7.2, at 4C for 4 hr to prepare frozen sections on a cryostat. The remaining part of the lymph nodes was homogenized in 10 mM phosphate-buffered saline (PBS), pH 7.2, to prepare tissue extracts. Blood was sampled from renal veins of the rats. The sera were separated by centrifugation for 10 min at 3,000 rpm and stored at –80C until use.

Western Blotting

Western blotting was performed, as previously described (Kato et al. 2008). P. gingivalis or S. mutans was totally lysed by adding 100 µL of 6× sodium dodecyl sulfate (SDS) sample buffer to 500 µL of the reaction mixture, and then boiling the samples for 5 min followed. The final composition of SDS sample buffer after mixing was 2% SDS, 58.3 mM Tris-HCl (pH 6.8), 6% glycerol, 5% 2-mercaptoethanol, 0.002% bromophenol blue, and protease inhibitor mix (Complete TM; Roche Applied Science, Indianapolis, IN). Aliquots of these samples (10 µg/lane) were separated by SDS–polyacrylamide gel electrophoresis on 12% polyacrylamide slab gels. The separated proteins were immediately transferred electrophoretically to polyvinylidene difluoride (PVDF) membranes (Millipore, Bedford, MA) using TB buffer (25 mM Tris, 192 mM glycine, and 20% methanol, pH 8.4). Proteins were transferred at 100 V for 1 hr at 4C. Membranes were blocked for 1 hr at room temperature with 5% skim milk in TBST (20 mM Tris-HCl, pH 7.6, containing 150 mM NaCl and 0.05% Tween-20). The blocking buffer was removed, and the membranes were incubated with the appropriate primary antiserum (the serum of P. gingivalis–immunized rat or control rat; 1:100,000 dilution) at 4C overnight in the blocking buffer. The membranes were subsequently washed three times (10 min/wash) with TBST and then incubated with the secondary antibody (goat anti-rat IgG-HRP conjugate, GE Healthcare, Piscataway, NJ; 1:100,000 dilution) in the blocking buffer for 1 hr at room temperature. After rinsing three times in TBST, the HRP activity was visualized by incubating the membranes for 5 min at room temperature in Super Signal West Pico Chemiluminescent Substrate detection system (Pierce, Rockford, IL) followed by autoradiography. At the end of these experiments, the membranes were stained with Coomassie brilliant blue (Bio-Rad, Hercules, CA) to confirm that equal amounts of protein were loaded in each lane of the gel.

Patients and Surgical Specimens

A total of 10 patients (M:F = 6:4) suffering from dental caries–associated radicular cyst were recruited. The age of the patients ranged from 28 to 60 years (mean, 46). Six occurred in the upper jaw and four in the mandible. Incisor teeth were affected in five cases, followed by molar (n=4) and premolar (n=1). The size varied from 0.8 to 3.4 cm (mean, 2.0) in diameter. The association of periodontitis was evaluated on X-ray film according to the degree of absorption of the alveolar bone. Periodontitis was graded as none (n=4), mild (n=2), moderate (n=4), and severe (n=0).

The surgical specimens of radicular cyst were divided into three pieces. Half was fixed in 10% formalin for histopathological diagnosis. The remaining tissue portions were used for preparing both fixed frozen sections and tissue extracts in the same way as for the rat experiment. Of 10 surgical specimens of radicular cyst, 8 showed dense infiltration of plasma cells within the lesions. Tissue extraction was performed in 6. The serum was sampled from the respective cases.

Written informed consent was obtained from each patient, and the analysis using human specimens was authorized by the Committee of Ethics for Clinical and Epidemiological Study, Fujita Health University School of Medicine (acknowledgment no. 07-102, October 2007).

Target Bacterial Proteins (Antigens)

In the present study, a total of five proteins derived from P. gingivalis were analyzed: These included Ag53 and four gingipain components such as the proteinase domain of Arg-gingipain (Arg-pro), the hemagglutinin/adhesin domain of Arg-gingipain (Arg-hgp), the proteinase domain of Lys-gingipain (Lys-pro), and the hemagglutinin/adhesin domain of Lys-gingipain (Lys-hgp). For the negative control purpose, SpaP, a representative pathogenic protein derived from S. mutans and dihydrofolate reductase (DHFR) of Escherichia coli origin were evaluated. The molecular weight of the respective proteins is as follows: Ag53 = 53 kDa, Arg-pro = 44 kDa, Arg-hgp = 103 kDa, Lys-pro = 51 kDa, Lys-hgp = 103 kDa, SpaP = 185 kDa, and DHFR = 24 kDa.

Plasmid Construction for the Cell-Free Protein Synthesis

Six bacterial genes encoding Ag53, Arg-pro, Arg-hgp, Lys-pro, Lys-hgp, and SpaP were amplified from genomic DNA of P. gingivalis FDC381 or S. mutans 854S by polymerase chain reaction (PCR) with PrimeStar HS DNA polymerase (Takara Bio, Otsu, Japan) or Ex Taq polymerase (Takara Bio), using the primers as indicated in Table 1. The sense primers contained the restriction enzyme recognition sequence (SalI or XhoI) and biotin ligase recognition sequence, whereas the restriction enzyme recognition site (NotI) was included in the antisense primers. The PCR products were digested with SalI and NotI or XhoI and NotI and then cloned into the corresponding sites of the pEU-E01-His-MCS vector (CellFree Sciences, Matsuyama, Japan). Nucleotide sequences of the DNA inserts in all plasmid constructs were subsequently confirmed by using the ABI PRISM 310 Genetic Analyzer using the BigDye terminator v1.1 Cycle sequence kit (Applied Biosystems, Foster City, CA). Plasmid DNA for transcription reactions was purified using Plasmid Midi Kit (QIAGEN, Stanford, CA).

Table 1.

The sequences of the primer pairs used for PCR amplification of P. gingivalis-related proteins, SpaP and DHFR

Target Gene	Sense/Antisense	Restriction Site	Sequence
Ag53	Sense	SalI	GAGAGTCGACATGGGCCTGAACGACATCTTCGAGGCCCAGAA GATCGAGTGGCACGAAATGAAGTTAAACAAAATGTT
	Antisense	NotI	GAGAGCGGCCGCTTAGAATTCGATATCATAGTTATG
Arg-pro	Sense	XhoI	GAGACTCGAGATGGGCCTGAACGACATCTTCGAGGCCCAGAAGATCGAGTGG CACGAATACACACCGGTAGAGGAAAAAC
	Antisense	NotI	GAGAGCGGCCGCTTAGCGAAGAAGTTCGGGGGCATCGC
Arg-hgp	Sense	XhoI	GAGACTCGAGATGGGCCTGAACGACATCTTCGAGGCCCAGAAGATCGAGT GGCACGAAAGCGGTCAGGCCGAGATTGTTC
	Antisense	NotI	GAGAGCGGCCGCTTACTTTACAGCGAGTTTCTC
Lys-pro	Sense	XhoI	GAGACTCGAGATGGGCCTGAACGACATCTTCGAGGCCCAGAAGATCGAGT GGCACGAAGATGTTTATACAGATCATGGCG
	Antisense	NotI	GAGAGCGGCCGCTTAACGTACATCGTTTGCAGGTTC
Lys-hgp	Sense	XhoI	GAGACTCGAGATGGGCCTGAACGACATCTTCGAGGCCCAGAAGATCGAGT GGCACGAAGCCAACGAAGCCAAGGTTGTGC
	Antisense	NotI	GAGAGCGGCCGCTTACTTGATAGCGAGTTTCTC
SpaP	Sense	XhoI	GAGACTCGAGATGGGCCTGAACGACATCTTCGAGGCCCAGAAGATCGAGT GGCACGAAATGAAAGTCAAAAAAACTTACGG
	Antisense	NotI	GAGAGCGGCCGCTCAATCTTTCTTAGCCTTTAAG
DHFR	Sense	XhoI	GAGACTCGAGGGCCTGAACGACATCTTCGAGGCCCAGAAGATCGAGTGG CACGAACTCCACCCACCACCACCAATGATCAGTCTGATTGCGGC
	Antisense	SpeI	GAGAACTAGTTTACCGCCGCTCCAGAATCT

Bold: restriction enzyme recognition sequence. Underline: biotin ligase recognition sequence.

The DHFR gene of E. coli was also PCR amplified (Sawasaki, Ogasawara, et al. 2002) with LA Taq polymerase (Takara Bio) using the primers, as indicated in Table 1. The PCR product was digested with XhoI and SpeI and then cloned into the corresponding sites of the pEU-E01-His-MCS vector. Nucleotide sequence analysis and plasmid DNA purification were carried out as mentioned above.

Cell-Free Protein Synthesis System

The cell-free protein synthesis system has been established in the Cell-Free Science and Technology Research Center, Ehime University (Sawasaki, Hasegawa, et al. 2002; Sawasaki, Ogasawara, et al. 2002). Briefly, the wheat germ extract (ENDEXT kits; CellFree Sciences) effectively synthesized proteins encoded by genomic bacterial DNA in 96-well plastic plates. The plasmid vectors, incorporated with target DNA and the biotin ligase recognition sequence, were harvested from E. coli and used as templates for the protein synthesis. Transcription by SP6 RNA polymerase to yield messenger RNA (mRNA) and translation to yield protein products using the wheat germ extract followed. Finally, biotin labeling was performed through biotin ligase activity to obtain a target protein that a single biotin was labeled on the N-terminal site. The N-terminally biotin-labeled target proteins lacking sugar residues functioned as probes for the enzyme-labeled antigen method. The translation mixtures were directly used without purification or concentration.

For biotin labeling of proteins, the previously described bilayer diffusion system was employed (Sawasaki, Hasegawa, et al. 2002; Sawasaki et al. 2008; Matsuoka et al. 2010). Briefly, 2 µg of biotin protein ligase (BirA, GenBank accession no. NP0312927) produced by the wheat cell-free system was added to the translation layer, and 5 µM D-biotin (Nacalai Tesque, Kyoto, Japan) were supplemented to both the translation and substrate layers.

Purification of Bacterial Proteins

Purification of biotinylated proteins (Sawasaki et al. 2008; Matsuoka et al. 2010) was performed for Ag53, Arg-pro, Arg-hgp, Lys-pro, Lys-hgp, and DHFR to compare the efficiency as the primary reagent and to know the appropriate concentration as the probe for the enzyme-labeled antigen method.

For purification, a polyhistidine tag was added to the N-terminal of the biotinylated proteins on a specialized vector (pEU-E01-His-bls-ORF subcloned from pEU-E01-His-MCS). Ni²⁺-affinity chromatography using Ni-NTA sepharose (GE Healthcare) was employed for one-step purification of the tagged proteins. The elution buffer contained 20 mM Na-phosphate (pH 7.5), 0.3 M NaCl, and 500 mM imidazole (Nacalai Tesque). The purified biotinylated proteins (concentrations around 200 µg/mL) were electrophoresed on polyacrylamide gel under SDS, and Coomassie brilliant blue staining for proteins was done. The separated proteins were also transferred to a Hybond-LFP PVDF membrane (GE Healthcare). After blocking with 5% skim milk in PBS at 4C overnight, the membranes were soaked in PBS containing 2 µg/mL streptavidin Alexa Fluor 488 conjugate (STA–Alexa 488; Invitrogen, Carlsbad, CA) and were rinsed three times with PBS containing 0.05% Tween-20. The biotinylated proteins on membranes were detected by the Typhoon 9400 imaging system (GE Healthcare), according to the manufacturer’s protocol.

The cell-free protein synthesis and polyhistidine-tagged protein purification using imidazole buffer for elution were carried out using an automatic robot, Protemist DTII (CellFree Sciences), basically according to the manufacturer’s instructions. The processes of in vitro transcription (for 6 hr), synthesis of biotinylated proteins (for 20 hr), and purification (for 4–6 hr) were automated with this machine.

The AlphaScreen Method

The AlphaScreen assay (Xu et al. 2009; Matsuoka et al. 2010) was performed at the Cell-Free Science and Technology Research Center, Ehime University, according to the manufacturer’s protocol (PerkinElmer Life and Analytical Sciences, Boston, MA). Reactions were carried out in 25 µL of reaction volume in 384-well Optiwell microtiter plates (PerkinElmer Life and Analytical Sciences). For the antigen-antibody reaction, the translation mixture containing the biotinylated protein was mixed with the 1:600 diluted sera or 1:300 diluted tissue extracts of patients and rats in 15 µL of reaction buffer (100 mM Tris-HCl, pH 8.0, containing 0.01% Tween-20 and 0.1% bovine serum albumin) and incubated at 26C for 30 min. Subsequently, 10 µL of streptavidin-coated donor beads and protein G-conjugated acceptor beads (PerkinElmer Life and Analytical Sciences) was added to a final concentration at 12 µg/mL per well and incubated at 26C for 1 hr in a dark box. Fluorescence emission was measured with the EnVision plate reader (PerkinElmer Life and Analytical Sciences), and the resultant data were analyzed using the AlphaScreen detection program. All repetitive mechanical procedures were performed by a Biomek FX robotic workstation (Beckman Coulter, Fullerton, CA).

Only when the antibodies were reactive with the N-terminally biotinylated antigens, fluorescent signals at 520 to 620 nm emitted. The wells with antigen-antibody reaction were thus easily recognized. The data were expressed as the signal intensity ratio when compared with the DHFR control (basal level) (Sawasaki, Hasegawa, et al. 2002; Sawasaki, Ogasawara, et al. 2002; Matsuoka et al. 2010). In the present study, a ratio no less than 2.0 was regarded as positive.

Enzyme-Labeled Antigen Method

Paraformaldehyde-fixed frozen sections with or without proteinase K pretreatment were incubated for 1 hr at room temperature with unpurified crude solutions (translation mixtures) containing biotinylated proteins, which were synthesized in the cell-free protein synthesis system in the well. HRP-labeled streptavidin was incubated as the secondary reagent for 1 hr at room temperature. After rinsing with PBS, the site of antigen-antibody reaction was visualized in 50 mM Tris-HCl buffer, pH 7.6, containing 20 mg/dL diaminobenzidine hydrochloride and 0.006% hydrogen peroxide. The nuclei were lightly counterstained with Mayer’s hematoxylin.

The effect of proteinase K pretreatment was evaluated by comparing with untreated sections. In the present study, we employed the digestion with 4 µg/mL proteinase K in 50 mM Tris-buffered saline, pH 7.6, for 15 min at room temperature, as reported previously (Mizutani et al. 2009).

To know the suitable concentration of the antigen solution for this technique, purified biotinylated proteins of Ag53, Arg-hgp, and Lys-hgp were incubated at the final concentrations of 1, 10, 50, and 100 µg protein/mL. For the human study, purified antigens were used on most occasions, except for SpaP.

Double Immunofluorescence Study

Double immunofluorescence staining was performed to show that cells labeled with biotinylated antigens express CD138 (syndecan-1) on the plasma membrane. CD138 is a known plasma cell marker, although epithelial cells are also labeled (Sebestyén et al. 1999). Sections were incubated with 1:100 diluted mouse anti-human CD138 monoclonal antibody (clone MI15; Dako, Glostrup, Denmark) and then with Alexa Fluor 568 (red)–labeled goat anti-mouse IgG antibody (Invitrogen, Eugene, OR). Subsequently, the sections were incubated with biotinylated antigens and next with Alexa Fluor 488 (green)–labeled streptavidin (Invitrogen). Proteinase K treatment was omitted to retain CD138 immunoreactivity. In every step, the reagents were incubated for 1 hr at room temperature. Finally, the sections were mounted in hydrophilic mounting medium (Prolong Gold Antifade Reagent with DAPI, Invitrogen). The nuclei were stained blue with 4′,6-diamidino-2′-phenylindole (DAPI). The immunofluorescence was observed under a microscope (AXIO Imager A1; Carl Zeiss, Oberkochen, Germany).

Absorption Experiment

To confirm the specificity of the enzyme-labeled antigen method, we performed the absorption experiment. We prepared five kinds of crude unlabeled antigen solutions (Ag53, Arg-hgp, Lys-hgp, Arg-pro, and Lys-pro) by the cell-free wheat germ system. The crude biotinylated antigen solutions (Ag53, Arg-hgp, and Lys-hgp) were diluted with the unlabeled antigen solutions or PBS at 1:5, and plasma cells producing antibodies to Ag53, Arg-hgp, or Lys-hgp in rat lymph nodes and human radicular cyst lesions (cases 3 and 6) were visualized. Expectedly, an excess amount of the corresponding antigens abolished the specific staining, but indifferent antigens did not.

Results

Synthesis and Purification of Bacterial Proteins

Five P. gingivalis proteins, including Ag53, Arg-pro, Arg-hgp, Lys-pro, and Lys-hgp, as well as SpaP and DHFR as controls, were synthesized and biotinylated with the cell-free, wheat germ extract–mediated protein synthesis system. Crude solutions (translation mixtures) in the wells were used for screening for the enzyme-labeled antigen method and the absorption experiment. To obtain purified biotinylated proteins (except for SpaP), polyhistidine-tagged proteins were harvested by reacting in multiple wells.

Figure 1 demonstrates results of the electrophoretic analysis of the purified biotinylated proteins. Protein bands showing appropriate molecular weights were visualized with both Coomassie brilliant blue staining and Western blot analysis using streptavidin–Alexa 488.

Figure 1.

Electrophoretic analysis of purified biotinylated proteins used in the present study. Protein bands showing appropriate molecular weights are visualized with both Coomassie brilliant blue (CBB) staining (left panel) and Western blot analysis using streptavidin–Alexa 488 (STA-Alexa488) (right panel). The estimated molecular weight of the proteins is Ag53 = 53 kDa, Lys-pro = 51 kDa, Lys-hgp = 103 kDa, Arg-pro = 44 kDa, Arg-hgp = 103 kDa, and DHFR = 24 kDa. The band of Arg-pro is relatively weak (arrowhead in the right panel). M, molecular weight markers.

Rat Experiment

Western Blotting

Western blot analysis was performed to analyze reactivity of sera of the immunized rats against bacterial lysates (Fig. 2). Multiple bands were seen on the blot membrane when P. gingivalis ATCC33277-injected rat sera (n=3) were reacted to P. gingivalis lysates, whereas no bands were observed against S. mutans 854S lysates. The sera of the control rats injected adjuvant alone (n=2) were also unreactive.

Figure 2.

Western blot analysis of bacterial lysates with immunized rat sera. Upper panels: autoradiographic demonstration of antibodies to Porphyromonas gingivalis in rat sera. Lower panels: Coomassie brilliant blue staining. (a) The sera of Ag53-negative P. gingivalis ATCC33277-injected rats (n=3) showed multiple bands reactive to the whole-cell lysates of P. gingivalis, but the sera of control rats injected adjuvant alone (n=2) did not. (b) The sera of P. gingivalis–injected and control rats did not react to the whole-cell lysates of Streptococcus mutans.

AlphaScreen Method

When the signal intensity ratio relative to the DHFR control was no less than 2.0, the signals in the AlphaScreen method were regarded as positive. The extracts of axillary and popliteal lymph nodes of rats immunized with Ag53-positive and Ag53-negative P. gingivalis, as well as the extracts from the control rats, were evaluated. The sera of rats immunized with Ag53-positive bacteria and of control rats were also checked by the AlphaScreen method. The sera of rats immunized with Ag53-negative bacteria, used for the Western blot analysis, were not examined.

As shown in Table 2, in two of three of rats immunized with Ag53-negative P. gingivalis, the antibodies to Arg-hgp and Lys-hgp were detected in the lymph node extract with the AlphaScreen method, whereas antibodies to Ag53, Arg-pro, Lys-pro, and SpaP were undetectable. In all three rats immunized with Ag53-positive bacteria, antibodies to Ag53 were raised in the lymph node extract and serum, whereas antibody titers against gingipain-related proteins were below the cutoff level. No antibody elevation was noted in the lymph node extracts and sera of the control rats (n=2).

Table 2.

Summary of the AlphaScreen Method and the Enzyme-Labeled Antigen Method in the Rat Experiment

	Serum					Tissue Extract
Ag	Control		Ag53(+)Pg			Control				Ag53(–)Pg						Ag53(+)Pg
Animal	1	2	1	2	3	1		2		1		2		3		1		2		3
Site						Ax	Po	Ax	Po	Ax	Po	Ax	Po	Ax	Po	Ax	Po	Ax	Po	Ax	Po
AlphaScreen method
Ag53	1.1	1.1	7.7	9.5	20.5	1.3	1.5	1.2	1.5	1.4	1.1	1.2	1.1	1.4	1.0	1.9	2.2	3.0	2.9	4.4	NE
Arg-pro	1.0	0.9	0.9	1.3	1.1	1.0	1.2	1.0	1.0	1.1	0.9	0.9	0.9	1.2	1.0	0.9	1.0	0.9	1.0	1.1	NE
Arg-hgp	0.9	1.0	3.3	4.2	4.4	1.0	1.1	1.0	1.0	1.9	1.3	1.5	2.9	3.0	2.5	1.3	1.4	1.5	1.4	1.5	NE
Lys-pro	0.9	1.0	1.1	1.4	1.2	1.0	1.3	1.1	1.1	1.2	1.0	1.0	1.0	1.1	1.0	0.9	1.1	1.0	0.9	1.1	NE
Lys-hgp	1.0	1.0	3.5	4.0	5.1	1.1	1.4	1.1	1.3	1.6	1.2	1.3	2.1	2.8	2.2	1.2	1.3	1.3	1.2	1.6	NE
SpaP	0.9	0.9	1.0	1.2	0.9	1.0	1.0	0.9	1.1	1.1	0.9	0.8	1.0	1.2	0.9	1.0	0.9	1.1	0.9	1.1	NE
Enzyme-labeled antigen method
Ag53						−	−	−	−	−	−	−	−	−	−	+	+	+	+	+	+
Arg-pro						−	−	−	−	−	−	−	+	−	−	−	−	−	−	−	−
Arg-hgp						−	−	−	−	+	+	+	+	+	+	+	+	−	+	+	+
Lys-pro						−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−
Lys-hgp						−	−	−	−	+	+	+	+	+	+	+	+	+	+	+	+
SpaP						−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−

Ag53(+)Pg, Ag53-positive Porphyromonas gingivalis immunized rats; Ag53(–)Pg, Ag53-negative P. gingivalis immunized rats; Ax, axillary lymph node; Po, popliteal lymph node; NE, not examined. No serum was evaluated in Ag53(–)Pg. AlphaScreen method: Data no less than 2.0 are regarded as positive, represented with bold figures. + indicates positive reaction with the enzyme-labeled antigen method.

Unpurified Protein Solutions as Probes for the Enzyme-Labeled Antigen Method and the Effect of Proteinase K Pretreatment

At first, we checked the crude biotinylated protein solutions (the supernatant of the cell-free translation mixture in each well) as probes for the enzyme-labeled antigen method. In fixed frozen sections of the lymph nodes of rats immunized with Ag53-positive P. gingivalis, the cytoplasm of plasma cells was positively labeled with biotinylated Ag53, Arg-hgp, and Lys-hgp. Then, the effectiveness of proteinase K pretreatment was evaluated. Proteinase K (4 µg/mL) pretreatment for 15 min at room temperature increased detectability of the specific antibody in plasma cells in fixed frozen sections (Fig. 3). Inhibition of endogenous peroxidase and endogenous biotin activities was confirmed by the negativity in proteinase K–pretreated negative control sections incubated with buffer alone.

Figure 3.

The effect of proteinase K pretreatment (4 µg/mL, for 15 min at room temperature) to retrieve the reactivity in the enzyme-labeled antigen method. Unpurified translation mixtures recruited from the cell-free wheat germ extract system were used as probes. Plasma cells in rat axillary lymph node immunized with Ag53-positive Porphyromonas gingivalis exhibit production of anti-Ag53 antibodies. When compared with the untreated frozen section (a), the reactivity is stronger in the proteinase K–pretreated frozen sections (b). Panel c shows proteinase K–pretreated negative control specimen without applying antigen solution. Inhibition of endogenous peroxidase and endogenous biotin activities is shown. Bar indicates 50 µm.

Evaluation of Appropriate Concentration of Biotinylated Antigen Solution

In rat lymph nodes with positive signals, the appropriate concentration of the biotinylated antigen solution was evaluated. For this purpose, purified biotinylated proteins of Ag53, Arg-hgp, and Lys-hgp (original protein concentration around 200 µg/mL) were used, and the antigen solutions adjusted with concentrations of 1, 10, 50, and 100 µg/mL were applied to fixed frozen sections. The concentration at 100 µg/mL gave the best result for Ag53, whereas 50 µg/mL was regarded suitable for detecting antibodies against Arg-hgp and Lys-hgp in frozen sections (Fig. 4).

Figure 4.

Evaluation of the suitable concentration of biotinylated protein solutions in the enzyme-labeled antigen method. Purified biotinylated Arg-hgp was diluted at the concentration of 10 µg/mL (a) and 50 µg/mL (b) and applied to proteinase K–pretreated frozen sections of the axillary lymph node of rat immunized with Ag53-positive Porphyromonas gingivalis. Signals are very weak when the concentration is low (10 µg/mL). The concentration at 50 µg/mL is suitable to get strong signals and low background. Bar indicates 50 µm.

We decided to use purified Arg-pro and Lys-pro with a concentration at 50 µg/mL for further study because the original concentration of the purified proteins was around 70 µg/mL.

Application of the Enzyme-Labeled Antigen Method to Rat Lymph Nodes

By the enzyme-labeled antigen method, antibodies specific to Arg-hgp and Lys-hgp were consistently demonstrated in plasma cells in the axillary and popliteal lymph nodes of all rats (n=6) immunized with Ag53-positive and Ag53-negative P. gingivalis. A good number of positively labeled plasma cells were distributed mainly in the medulla of the nodes. Plasma cells producing antibodies to Arg-pro were demonstrated in the popliteal lymph node of one rat immunized with Ag53-negative bacteria. Antibodies to Ag53 were detected only in the lymph nodes of Ag53-immunized rats (n=3). The lymph nodes in the control rats (n=2) did not react with any labeled antigen.

The results are summarized in Table 2, and representative features are illustrated in Figure 5.

Figure 5.

Visualization of plasma cells producing antibodies against Porphyromonas gingivalis–related proteins in proteinase K–pretreated frozen sections of rat popliteal lymph nodes with the enzyme-labeled antigen method. Anti-Ag53 antibodies are demonstrated in nodal tissue immunized with Ag53-positive strain (a) but undetectable in tissue immunized with Ag53-negative strain (b). Anti-Arg-hgp antibodies (c, d) and anti-Lys-hgp antibodies (e, f) are seen in a good number of plasma cells in tissues immunized with either the Ag53-positive (left panels: c and e) or Ag53-negative strains (right panels: d and f). Bar indicates 50 µm.

Analysis of 10 Radicular Cyst Lesions

The results of both the AlphaScreen method and the enzyme-labeled antigen method in human cases are summarized in Table 3. Sera and frozen sections were evaluated in all 10 cases, whereas tissue extracts were available in 6.

Table 3.

Summary of the AlphaScreen Method and the Enzyme-Labeled Antigen Method in 10 Patients with Radicular Cyst

Case	1		2		3		4		5		6		7		8		9		10
Age/sex	49M		42M		52M		50M		43M		38F		60F		47M		47F		28F
Site	UM		UI		UI		UI		UI		LP		LM		LM		LM		UI
Size (cm)	3.4		1.6		2.2		1.9		1.7		2.1		1.9		2.8		1.4		0.8
Periodontitis	Moderate		Moderate		Mild		Moderate		Mild		None		None		Moderate		None		None
Sample	T	S	T	S	T	S	T	S	T	S	T	S	T	S	T	S	T	S	T	S
AlphaScreen method
Ag53	1.4	1.7	NE	1.7	1.3	1.7	1.0	2.1	NE	1.4	0.9	1.3	NE	1.5	NE	1.4	0.9	1.4	0.8	1.2
Arg-pro	1.1	1.1	NE	1.3	1.1	1.2	0.7	1.3	NE	1.2	1.2	1.2	NE	1.2	NE	1.4	0.8	1.1	0.8	1.1
Arg-hgp	1.1	1.6	NE	1.3	7.7	1.6	1.2	2.5	NE	1.3	7.4	1.7	NE	1.4	NE	4.2	1.0	1.1	1.0	1.4
Lys-pro	1.2	1.4	NE	1.4	1.2	1.3	0.8	1.5	NE	1.2	1.0	1.2	NE	1.3	NE	8.5	1.5	1.2	1.0	1.2
Lys-hgp	1.2	1.8	NE	1.5	3.4	1.8	1.2	2.5	NE	1.3	8.6	1.7	NE	1.4	NE	7.0	1.0	1.3	1.7	1.6
SpaP	1.1	1.2	NE	1.3	1.1	1.3	0.7	1.4	NE	1.2	0.8	1.0	NE	1.3	NE	4.2	1.0	1.3	0.9	1.3
Enzyme-labeled antigen method
Ag53	−		−		−		NP		−		−		−		NP		−		−
Arg-pro	−		−		−		NP		−		−		−		NP		−		−
Arg-hgp	−		−		+		NP		−		+		−		NP		−		−
Lys-pro	−		−		−		NP		−		−		−		NP		−		−
Lys-hgp	−		−		−		NP		−		+		−		NP		−		−
SpaP	−		−		−		NP		−		−		−		NP		−		−

Site: UM, upper molar; UI, upper incisor; LP, lower premolar; LM, lower molar. Sample: T, tissue extract; S, serum. NE, not examined; NP, no plasma cells seen in frozen sections. AlphaScreen method: Data no less than 2.0 are regarded as positive, represented with bold figures. + indicates positive reaction with the enzyme-labeled antigen method.

With the AlphaScreen method, antibodies specific to Arg-hgp and Lys-hgp were demonstrated in the tissue extract from two of six cases examined (cases 3 and 6), but the specific antibodies were not elevated in the serum in these two cases. In contrast, specific antibodies were elevated in the serum in two cases (cases 4 and 8): In case 4, low serum titers against Ag53, Arg-hgp, and Lys-hgp were detected, and in case 8, high titers against Arg-hgp, Lys-hgp, Lys-pro, and SpaP were shown. In these two cases, however, plasma cells were scarcely observed in frozen sections of the surgically removed radicular cyst. When the degree of periodontitis was reviewed, these two cases complicated moderate-degree periodontitis. Although cases 1 and 2 also showed moderate-degree periodontitis, all the cases with mild-degree (n=2) or no (n=4) periodontitis were AlphaScreen negative.

On proteinase K–pretreated fixed frozen sections, plasma cells were infiltrated in a total of eight lesions of radicular cyst. Purified biotinylated proteins were used as probes, except for SpaP, for which crude (unpurified) solutions were applied. In cases 3 and 6, positive signals for Arg-hgp were detected by the enzyme-labeled antigen method. The cytoplasm of plasma cells, often clustered beneath the squamous epithelial lining, showed homogeneous positivity. Antibodies against Lys-hgp were also visualized in a small number of plasma cells in case 6. The AlphaScreen titration was very high (more than 7) in these positive cases. No signals were demonstrated in the other sections. Representative features in case 6 are illustrated in Figure 6.

Figure 6.

Visualization of plasma cells producing antibodies against Porphyromonas gingivalis–related proteins in proteinase K–pretreated frozen sections of radicular cyst (case 6) with the enzyme-labeled antigen method. Plasma cells are clustered beneath the squamous epithelial lining in the gingival cystic lesion (a: hematoxylin and eosin staining). Antibodies against Ag53 (b) are negative in the lesion, whereas antibodies against Arg-hgp (c) and Lys-hgp (d) are localized in the cytoplasm of plasma cells clustered just beneath the squamous epithelial lining. Bar indicates 50 µm.

The results of the double immunofluorescence study are illustrated in Figure 7. Inflammatory cells showing cytoplasmic signals positive for Arg-hgp in cases 3 and 6, as well as the cells positive for Lys-hgp in case 6, were labeled for CD138 mainly along the plasma membrane. The plasmacytic nature of the signal-positive cells was thus confirmed. Squamous epithelial cells were also CD138 positive.

Figure 7.

Double immunofluorescence analysis with biotinylated antigens and anti-CD138 monoclonal antibody in radicular cyst lesions (cases 3 and 6). The biotinylated antigens are stained green with Alexa Fluor 488 (a: Arg-hgp in case 3, d: Lys-hgp in case 6), whereas CD138 is immunolocalized in red with Alexa Fluor 568 (b, e). Panels c and f demonstrate merged pictures for both immunofluorescence. CD138 is expressed along the plasma membrane of the plasma cells positive for anti-Arg-hgp or anti-Lys-hgp. Arrows show the same plasma cells. Bar indicates 25 µm.

When paraffin sections were reviewed, the histological features of cases 3 and 6 were indistinguishable from other cases. All but one (case 8) showed dense infiltration of plasma cells in the cyst wall. In case 4, plasma cells were plentiful in paraffin sections in contrast to frozen sections. Foamy cells were clustered in cases 5 and 10, and neutrophils were rich in case 7.

Absorption Test for Confirming the Specificity of Staining

The positive anti-Ag53 signals in plasma cells of rat lymph nodes were eliminated only with an excess amount of unlabeled Ag53 antigen. The positive signals of anti-Arg-hgp and anti-Lys-hgp reactivities were not eliminated with unlabeled Ag53, Arg-pro, and Lys-pro in both rat and human tissues. When biotinylated Arg-hgp or Lys-hgp was diluted with unlabeled Arg-hgp or Lys-hgp, a considerable percentage of the positive signals in plasma cells in rat and human tissues disappeared or became significantly weakened, indicating the production of antibodies against the common epitope between Arg-hgp and Lys-hgp (Fig. 8).

Figure 8.

Absorption experiment (I) in rat axillary lymph node, showing the specificity of the enzyme-labeled antigen method (a–c: anti-Ag53 reactivity, d–f: anti-Arg-hgp reactivity). Anti-Ag53 reactivity (a) is mostly abolished with unlabeled Ag53 (b) but kept unchanged after addition of unlabeled Lys-hgp (c). Anti-Arg-hgp reactivity (d) is totally abolished with Arg-hgp (e). The addition of unlabeled Lys-hgp significantly weakens the signals in more than half of plasma cells, but some plasma cells remain strongly reactive (f). Bar indicates 50 µm.

In the radicular cyst of case 6 and some lymph nodes of the rat (Fig. 8f), in addition to antibodies directed toward the common epitope, antibodies specific to either Arg-hgp or Lys-hgp were also demonstrated. In some areas of the case 6 lesion, clustered plasma cells were reactive with both biotinylated Arg-hgp and Lys-hgp, and the reactivity was considerably weakened after incubating with unlabeled Arg-hgp or Lys-hgp. In other areas, dispersed plasma cells were reactive with biotinylated Arg-hgp but not with biotinylated Lys-hgp, and the reactivity was weakened with unlabeled Arg-hgp but not with unlabeled Lys-hgp (Fig. 9). The plasma cells in the latter region were considered to produce antibodies against an Arg-hgp-specific epitope. Similarly, antibodies against Lys-hgp-specific epitopes were also recognized in plasma cells in case 6. In the case 3 lesion where anti-Arg-hgp signals were detected without signals against Lys-hgp, anti-Arg-hgp reactivity was abolished with Arg-hgp but not with Lys-hgp, again proving antibody production specific to the epitope unique in Arg-hgp.

Figure 9.

Absorption experiment (II) in radicular cyst in case 6, showing region specificity of the antibodies (a–d: area 1, e–h: area 2). Plasma cells with anti-Arg-hgp reactivity are scattered in area 1 (a) and densely clustered in area 2 (e). Addition of unlabeled Arg-hgp significantly abolishes the reactivity (b, f). Unlabeled Lys-hgp scarcely affects the reactivity in area 1 (c) but significantly weakens the reactivity in area 2 (g). Anti-Lys-hgp reactivity is not demonstrated in area 1 (d), whereas the same clusters are labeled with biotinylated Lys-hgp in area 2 (h). It is concluded that the plasma cells in area 1 produce antibodies against an epitope specific to Arg-hgp, whereas another epitope common to both Arg-hgp and Lys-hgp is recognized by the clonally clustered plasma cells in area 2. Bar indicates 100 µm.

Discussion

We documented herein the novel application of the AlphaScreen method and the enzyme-labeled antigen method for demonstrating antibody production against P. gingivalis in the rat experimental system and in human radicular cyst lesions.

P. gingivalis is a black-pigmented, obligatory anaerobic gram-negative bacillus, often colonizing in the affected periodontal space in periodontitis or chronic gingivitis (Cutler et al. 1995; Slots and Ting 2000). Ag53 is a major outer membrane protein of P. gingivalis (Kurihara et al. 1991; Oyaizu et al. 2001), and positive immune response to Ag53 has been reported in patients with periodontitis (Kurihara et al. 1991; Schenkein 2006). P. gingivalis also expresses and secretes cysteine proteinases named gingipains (Chen et al. 2001; O’Brien-Simpson et al. 2001; Grenier and Tanabe 2010). Two major isoforms of gingipains are known: arginine-specific cysteine proteinase A (RgpA) and lysine-specific cysteine proteinase (Kgp). Each form consists of the proteinase domain (pro) and the hemagglutinin/adhesin domain (hgp). The hgps have been implicated as adhesins that actuate colonization of the epithelium lining the gingival sulcus. It has been reported that in cases of severe chronic periodontitis, the level of RgpA in periodontal pocket fluid correlated with the load of P. gingivalis (Guentsch et al. 2011). In the present study, we synthesized five P. gingivalis proteins such as Ag53, Arg-pro (pro in RgpA), Arg-hgp (hgp in RgpA), Lys-pro (pro in Kgp), and Lys-hgp (hgp in Kgp).

SpaP, a pathogenic major cell wall protein of S. mutans (Lee et al. 1988; Galaviz et al. 2002), was used as a negative control. The gram-positive cocci, residents in the human oral cavity, cause dental caries, and SpaP plays a major role in cell adherence to tooth surfaces and in sucrose-induced cell aggregation. DHFR, a folic acid–metabolizing enzyme reducing dihydrofolate to tetrahydrofolate, plays a housekeeping role in all proliferating cells (Feder et al. 1989). At the Cell-Free Science and Technology Research Center, Ehime University, E. coli–derived DHFR, synthesized with the cell-free protein synthesis system, has been used as a negative control (background level) in the AlphaScreen method (Sawasaki, Hasegawa, et al. 2002; Matsuoka et al. 2010).

The merits of the AlphaScreen method are as follows (Matsuoka et al. 2010). 1) In the reaction mixture, the antigens remain intact and thus functional. 2) Only small aliquot of sample (0.025 µL serum per protein) is needed. 3) The antibody detection is highly efficient since no rinsing is needed, and multiple assays can be done simultaneously by using 384-well plates. 4) The sensitivity and specificity of detection are very high with low background.

Roughly speaking, the results obtained with both the AlphaScreen method and the enzyme-labeled antigen method corresponded. In the rat model, the enzyme-labeled antigen method was more sensitive than the AlphaScreen method, whereas in human cases, the AlphaScreen method seemed to be superior to the enzyme-labeled antigen method. The authors believe that with the aid of the AlphaScreen method, the validity and specificity of the new histochemical technique were proven in the present study. This might be the very first application of the enzyme-labeled antigen method to human tissues.

The following findings were particularly noteworthy from a technical and pathophysiological viewpoint.

1. Methodologically, the suitable technical sequences of the enzyme-labeled antigen method using paraformaldehyde-fixed frozen sections were comparable with those we recently reported in the rat experimental model (Mizutani et al. 2009). Namely, proteinase K pretreatment enhanced the detectability, and the suitable concentration of the labeled antigen solution ranged from 50 to 100 µg/mL.

2. In the rat experiment, Ag53 antibodies were demonstrated only in the animals immunized with the Ag53-positive bacterial strain. In human cases, antibodies against Ag53 were undetectable in both the sera and tissues. The lack of anti-Ag53 antibody reaction in humans may be explained as follows: The infection of Ag53-positive P. gingivalis is not as frequent as expected in human lesions, or Ag53 is pathogenic in periodontitis but not in dental caries–associated radicular cyst.

3. In both rat and human, the Arg-hgp and Lys-hgp portions of P. gingivalis–derived gingipains were immunogenic when compared with the Arg-pro and Lys-pro portions. Reportedly, in patients with periodontitis, serum antibodies to Arg-hgp were raised (O’Brien-Simpson et al. 2000; Inagaki 2001), whereas antibodies to Arg-pro were not provoked (Inagaki 2001). In contrast, serum antibodies immunoprotective against P. gingivalis infection were detected when mice were experimentally immunized with synthetic peptides from Arg-pro (Genco et al. 1998; O’Brien-Simpson et al. 2001).

4. In patients 3 and 6, anti-Arg-hgp and anti-Lys-hgp antibodies were detected in CD138-positive plasma cells accumulated in the gingival tissue but not in the serum. It is hypothesized that these antibodies were produced locally but not secreted into the blood. The fact indicates the importance of evaluating the diseased tissue itself with this novel histochemical technique, the enzyme-labeled antigen method.

5. The pathogenetic significance of P. gingivalis infection in radicular cyst remains to be solved since only 2 of 10 lesions evaluated showed the production of anti-gingipain antibodies. P. gingivalis has so far been isolated from the root canal and the lesion of radicular cyst (Jung et al. 2000; Dahlen 2002; Wayman et al. 2002; Sunde et al. 2003; Noguchi et al. 2005), but the isolation is inconsistent: Noguchi et al. (2005) reported that biofilm infection of P. gingivalis was associated with refractory periapical periodontitis. Jung et al. (2000) reported that P. gingivalis was isolated from 26.3% of root canal infection and that P. gingivalis infection was significantly related to symptom manifestation. Our data suggest that infection of P. gingivalis and immune reaction to this pathogen may be involved in the pathogenesis of part (one-fifth) of the radicular cyst lesions.

6. Another question is the reason why the elevated serum antibody titers against gingipains were demonstrated in two cases, in which plasma cells were scarcely infiltrated in frozen sections of radicular cyst (although one of these two lesions showed plasma cell infiltration in paraffin sections). It may be related to the complication of periodontitis, because both of these cases showed a moderate degree of periodontal bone absorption.

7. Antibodies to SpaP of S. mutans origin were undetectable in the patients’ sera and radicular cyst lesions, except for one case (case 8) in which serum antibody titer against SpaP was elevated. We should explain the reason for such unexpected negative data on the dental caries–causing bacteria by continuing further analysis.

8. We should discuss the specificity of the method. The absorption experiment using an excess amount of unlabeled antigens supported the specificity of the present technique, we believe. In rat and human tissues, antibodies against the epitope common between Arg-hgp and Lys-hgp were produced in a good number of plasma cells, whereas the other plasma cells contained antibodies specific to epitopes unique in either Arg-hgp or Lys-hgp. In fact, the homology of the amino acid sequence of both domains is estimated 76% (758/1007) (O’Brien-Simpson et al. 2001). According to the protein sequence analysis using the Basic Local Alignment Search Tool (BLAST), the homology between Arg-pro and Lys-pro was 30% (148/501). Lower homology data were observed among the other antigens examined, although regional short segment homology was seen: for example, 73% (40/55) between Arg-hgp and Lys-pro, 55% (19/35) between Arg-hgp/Lys-hgp and Arg-pro, and 24% (16/67) between Ag53 and Arg-hgp/Lys-hgp. It is intriguing to say that the region (or epitope) specificity was able to be shown with our histochemical technique. When the difference in the amino acid sequence is considered, Ag53 and gingipains can be regarded as indifferent antigens, and hgp and pro regions in gingipains also belong to indifferent proteins for each other.

9. It may happen that the specific antibodies demonstrated in plasma cells in rat and human tissues belong to immunoglobulins that were originally raised against closely related proteins coming from different bacteria. To exactly confirm the specificity of the antibodies detected in tissue sections, we should extend our study to perform bacterial isolation from the diseased tooth root and radicular cyst lesions, and antigens cross-reactive to Ag53 or gingipain domains should be screened in P. gingivalis–related or unrelated pathogens by analyzing the gene database. When labeled antigens strictly specific to certain pathogens are not available or their strict specificity is unknown, the positive findings with plural probes may strengthen the specificity in the enzyme-labeled antigen method, as in the present study and also for the immunoperoxidase technique.

To use the enzyme-labeled antigen method as a ubiquitous technique, one must consider the method of preparing labeled antigens.

When the pathogenic proteins are known, the sequence of transcription of DNA to yield mRNA and translation of mRNA to yield protein products can be introduced. The cell-free protein synthesis system employing the wheat germ extract gives an efficient tool for this purpose (Sawasaki, Hasegawa, et al. 2002; Sawasaki, Ogasawara, et al. 2002). Once the plasmid construction is ready, crude or purified biotinylated proteins can be synthesized automatically within 36 hr.

For preparing the labeled antigens for analyzing autoimmune diseases in which the pathogenic antigens are not necessarily known, human/mouse libraries of autoimmune-related antigens are useful: The Cell-Free Science and Technology Research Center at Ehime University has established those libraries containing more than 2000 biotinylated proteins by applying the cell-free protein synthesis system (Matsuoka K, Endo Y, Sawasaki T, unpublished data). The target antigens can be screened in the human library with the patient’s serum and diseased tissue extract: The AlphaScreen method is quite suitable for this purpose (Matsuoka et al. 2010).

In conclusion, the significance of our idea on the enzyme-labeled antigen method includes the following:

1. The target antigens (or even epitopes) of the antibodies produced within the lesion are detectable, so that the pathogenesis of the lesion can be analyzed in view of the site of antibody production.

2. Analysis of the serum and tissue extract with the AlphaScreen method provides the specificity confirmation of this novel histochemical technique.

3. The method for preparing labeled probes for the enzyme-labeled antigen method, the wheat germ extract–mediated cell-free protein synthesis system, is relatively simple and automated, so that the site of antibody production can be visualized specifically by applying this technique.

4. Our approach is applicable to a variety of lesions with plasma cell infiltration and provides a novel histochemical tool for analyzing disease process.

The possible target lesions of the enzyme-labeled antigen method include 1) the autoimmune diseases, such as rheumatoid arthritis, Hashimoto thyroiditis, Sjögren syndrome, ulcerative colitis, and autoimmune gastritis; 2) infectious diseases, such as Helicobacter pylori–induced gastritis and syphilis; and 3) malignant tumors showing heavy infiltration of plasma cells in the stroma, including Epstein-Barr virus–related tumors (nasopharyngeal carcinoma, gastric carcinoma with lymphoid stroma, Hodgkin’s lymphoma, and nasal malignant lymphoma), human papilloma virus–related uterine cervical carcinoma, mucosa-associated lymphoid tissue lymphoma, and multiple myeloma (plasmacytoma).

We sincerely hope that the enzyme-labeled antigen method is widely used as a novel histochemical technique.