Both the unique and repeat regions of the Porphyromonas gingivalis hemagglutinin A are involved in adhesion and invasion of host cells

Abstract

Porphyromonas gingivalis is one of the major etiologic agents of adult periodontitis and has been associated with cardiovascular diseases. It expresses multiple hemagglutinins that are significant virulence factors that play an important role in bacterial attachment and invasion of host cells. The objective of this study was to determine the impact of P. gingivalis hemagglutinin A (HagA) on the attachment to and invasion of human coronary artery endothelial cells (HCAEC) and gingival epithelial cells (GEC).

Bacterial strains expressing the HagA protein (or subunits), including Escherichia coli carrying plasmid pEKS5, E. coli carrying plasmid ST2, and Salmonella enterica serovar Typhimurium with plasmid pNM1.1 were used in this study. The strains were tested for their ability to attach to and invade HCAEC and GEC using antibiotic protection assays. In addition, the unique 5′ N-terminal non-repeated segment of HagA was purified in recombinant form and a monoclonal antibody was created against the polypeptide. The monoclonal antibody against the unique portion of HagA was tested for inhibitory activity in these assays.

The attachment of both E. coli strains expressing HagA fragment to host cells was significantly increased compared to their respective controls. However they did not invade GEC or HCAEC. Interestingly, HagA expression in the Salmonella strain increased both adherence to and invasion of HCAEC, which may be due to the presence of the entire hagA ORF. A monoclonal antibody against the unique 5′ N-terminal portion of HagA reduced invasion. Further experiments are needed to determine the role of the unique and the repeat segments of P. gingivalis HagA.

1. Introduction

Porphyromonas gingivalis is a major etiologic agent of chronic and severe adult periodontitis, an important cause of tooth loss [1–5]. Periodontal infections have been associated with systemic conditions such as atherosclerotic heart disease [6–9] and a higher risk of preterm low birth-weight babies [10]. More specifically, inoculation with P. gingivalis accelerates atherosclerotic development in mice [11–12] and DNA from this microorganism is recovered from aortic tissue of infected mice [11]. Live P. gingivalis and evidence of its DNA are also detected from human atherosclerotic plaques [13–14]. Furthermore, upon infection by P. gingivalis, foam cell formation by macrophages in the vascular walls was enhanced and may possibly initiate or exacerbate the atherosclerotic process [15].

P. gingivalis has been shown to invade various types of cells including gingival epithelial cells [16–18] and aortic and heart endothelial cells [19–20]. P. gingivalis is found within the cytoplasm in gingival epithelial cells [17] or either free or in the cytoplasm in pocket epithelial cells [18]. In contrast, P. gingivalis replicates in endocytic vacuoles of endothelial cells [19–20]. Given these differences, additional studies are warranted to study and compare the initial interactions between P. gingivalis and various types of host cells. The adherence of P. gingivalis to host tissue cells is a crucial step in the pathogenesis of infection. It enables the microorganism to colonize host tissues and secure critical nutrients [21]. Several virulence factors of P. gingivalis have been characterized and shown to play a role in adhesion [21]. The fimbriae of P. gingivalis strains such as 381 mediate adhesion/invasion of host cells whereas nonfimbriated strains have a reduced ability to invade [19, 22–23]. Furthermore, monoclonal antibodies against the fimbriae blocked the adherence to buccal epithelial cells [24] and a mutation in the fimA gene reduced adherence of P. gingivalis to gingival epithelial cells [22, 25]. However, the expression of FimA is not sufficient for invasion [26]. In another study, the fimbriae and the hemagglutinin adhesin HA-Ag2 were also shown to mediate the adhesion to epithelial cells [27].

Microorganisms such as P. gingivalis may use hemagglutinins to adhere to erythrocytes or other cells and to acquire nutrients [28–29]. Multiple hemagglutinins have been identified in P. gingivalis [30–32]. HagB has been shown to be involved in adherence of P. gingivalis to HCAEC [33]. HagA and HagD are 73.8% identical [34] and share homology to cysteine protease (gingipain) genes [35–36]. Another hemagglutinin, HagE, shares a 523-aa region with 93% homology to HagA [34]. The hagA gene encodes a large protein of predicted molecular mass of 283.3 kDa containing multiple contiguous direct repeats (hemagglutinin A repeat; HArep) of 440 to 456 amino acids, with hemagglutinin activity for each of the repeats [37]. The multiple repeat units might be involved in antigenic variation. A mAb, which recognizes the GVSPKVCKDVTVEGSNEFAPVQNLT epitope present in HRgpA and Kgp as well as in the HagA gene product, inhibits hemagglutination and confers passive immunization against colonization by P. gingivalis [38]. The PVQNLT motif has been found to elicit a protective immune response against P. gingivalis colonization [39]. Due to its importance, HagA was expressed in an immunogenic form in a Salmonella enterica serovar Typhimurium avirulent vaccine strain. This recombinant major virulence associated surface protein was recognized by serum from a periodontal patient [40]. This vaccine strain could be used to develop a protective vaccine against P. gingivalis infection.

Even though the repeat units of HagA have been recognized to have adhesin properties necessary for hemagglutination activities, the importance of HagA in the colonization process, more specifically its role in adhesion and invasion of human host cells, has not yet been determined. In this study, we demonstrate that HagA is involved in adhesion to host cells and for the first time address the function of the heretofore unstudied N-terminal unique segment of the protein.

2. Materials and methods

2.1. Bacterial strains and cell culture conditions

P. gingivalis strain 381 was grown anaerobically on blood agar plates (BAP; Difco Laboratories, Detroit, MI) or in brain heart infusion broth (Difco) supplemented with 5 μg hemin ml⁻¹, and 5 μg vitamin K1 ml⁻¹, as described previously [29]. Clindamycin was added to the media at 5 μg ml⁻¹ to maintain the HagA mutant of P. gingivalis 381.

In a previous study, a 3,164 bp fragment from P. gingivalis 381 hagA has been cloned into vector pUC8 [31]. Using rabbit anti-P. gingivalis 381 antisera, insert antigen expression was confirmed, including in absence of IPTG [31]. The resulting strains harboring clones, E. coli-ST2, and a vector-only control E. coli-pUC8 were used in the present study. E. coli-pEKS5, E. coli-pET19b, and E. coli-HagAU are described below and represented in Fig. 1. E. coli strains were grown aerobically overnight at 37 °C in Luria-Bertani (LB) plates or in LB broth (Difco) containing 100 μg·ml⁻¹ ampicillin except for E. coli-HagAU which was grown in presence of 50 μg·ml⁻¹ kanamycin.

Fig. 1.

Diagram depicting the various portions of the hagA gene that have been used in this study. In P. gingivalis 381, the 7,887 bp hagA gene include 4 direct repeats (HArep) represented as shaded areas. The whole hagA gene has been cloned into S. enterica pNM1.1 whereas only portions of the hagA gene, as illustrated above, have been cloned into strains E. coli-ST2 (nt 705-3872; see [37]), E. coli-pEKS5 (nt 705-3872), and E. coli HagAU (nt 1-1233; see [37]). The nt numbers are provided for each clone in relationship to the HagA molecule depicted at the top of the Figure.

A vaccine strain of Salmonella enterica serovar Typhimurium χ4072 (S. enterica pNM1.1), which expresses the functional hemagglutinin A of P. gingivalis, and it respective control strain harboring an empty vector were described in a previous study and grown in LB broth or plates [40]. A diagram of S. enterica pNM1.1 insert can be found in Fig. 1.

Human coronary artery endothelial cells (HCAEC; Cambrex, Walkersville, MD) were cultured in endothelial cell basal medium-2 (EBM-2; Cambrex) supplemented with EGM-2-MV singlequots (Cambrex) as described by the manufacturer. HCAEC were maintained at 37 °C with 5% CO₂ in a humidified atmosphere. Human gingival epithelial cells (GEC), a primary cell line kindly provided by Dr. Richard Lamont (University of Florida, Department of Oral Biology), was prepared from gingival explants and cultured as monolayers in serum-free keratinocyte growth medium (Cambrex Bio Science, Walkersville, Inc., Walkersville, MD) as described previously [17].

2.2. P. gingivalis HagA mutant construction

Primers were designed using the Oligo 4.0 software (MBI, Cascade, Colorado). A 756-bp fragment of hagA starting at nt 292 of the coding sequence was amplified by PCR from purified genomic DNA of P. gingivalis 381 with the forward primer (5′-AAAACTGCAGTGGTTTGCCAACGGAAC-3′) and the reverse primer (5′-TGCTCTAGAATCCGAGGGTTTCTTCCAG-3′) (underlined sequences indicate PstI and XbaI sites for the forward and reverse primers respectively). This fragment was cloned into the PstI/XbaI site of the suicide vector pVA3000 to obtain a 6.06-kb plasmid which was subsequently transferred from E. coli S17-1 into P. gingivalis 381 by conjugation as described previously [41]. Clone A4 was designated as P. gingivalis HagA mutant after the mutation was confirmed by Southern hybridization (data not shown) by using the Genius DNA labeling and detection kit (Roche Diagnostics, Indianapolis, IN) according to the manufacturer’s directions. All restriction and modification enzymes were purchased from Promega Corporation (Madison, WI).

2.3. Construction and expression of HagA translational fusion in E. coli expression vector

i) HagA from E. coli-pEKS5

The cloning of the hagA coding sequence fragment in an E. coli expression vector were performed as follows. A genomic clone of hagA in pUC18 [37] was used as DNA template. PCR amplification of a 3.2 kb hagA subfragment was performed with primers bearing restriction enzymatic sites NdeI in the forward primer (5′-GGCCATATGAGCTTCGATACGCAAAC-3′) and XhoI plus a stop codon in the reverse primer (5′-CGCGCTCGAGTTACGCAGGCAAATCTACC-3′) in a final volume of 50 μl containing 2.35 mM MgCl₂, 0.3 μM of each primer, 0.4 mM of each dNTP, 1.25 unit Taq DNA Polymerase (Promega, Madison, WI) and 0.78 unit Pfu DNA Polymerase (Stratagene, La Jolla, CA). The PCR conditions used were as follows: an initial denaturation cycle of 1 min at 94 °C, followed by 30 cycles of 30 sec at 94 °C, 30 sec at 55 °C, and 6 min at 68 °C, with a final extension of 10 min at 72 °C in a PTC-100 thermal cycler (MJ Research, Watertown, MA). The PCR amplicon was purified using the QIAquick PCR purification kit (QIAGEN Inc., Valencia, CA) and initially cloned into the PCR cloning vector pT7Blue (Novagen, Madison, WI) to be subsequently excised with NdeI and XhoI, and directionally subcloned into the same restriction sites in pET19b (Novagen). Transformants in E. coli Novablue were screened by PCR for an insert in the proper orientation using the universal primer T7 (from vector) and above reverse primer (from insert). The plasmid from chosen clone pEKS5 was extracted, transformed into the expression strain E. coli BL21(DE3) and named E. coli-pEKS5. The authenticity of the expressed protein was confirmed by SDS-PAGE and Western blot (not shown) using mAb 61BG1.3 (generously provided by Dr. Gmür, University of Zürich), a mAb shown to react against the repeat regions of the HagA gene [42].

ii) Unique fragment of HagA from E. coli-HagAU

The first 1,233 bp of hagA were cloned into pET30a, using SacI and XhoI cloning sites and the cloning procedure above, and the recombinant plasmid was transformed into E. coli. This fragment does not include any of the repeat segments normally present in the HagA gene and was named HagAU (U for unique) since this fragment is unique in P. gingivalis whereas the repeat sequences are also present in rgpA and kgp. Using the above protocol, the protein was purified, and migrated at approximately 46 kDa on SDS-polyacrylamide gels (data not shown). This purified protein was used to produce monoclonal antibody HL1899 (see below).

2.4. Monoclonal antibody production

Mouse monoclonal antibody (mAb) HL1899 against the purified rHagAU was produced by standard protocols utilized by the Interdisciplinary Center for Biotechnology Research (ICBR) Hybridoma Core Laboratory at the University Florida [43–44] and described previously [33]. Briefly, Balb/cByj mice were immunized subcutaneously with either 25 μg or 50 μg of the antigen using the MPL + TDM adjuvant system (Sigma-Aldrich Company Ltd, St. Louis, MO) or 10 μg of the antigen mixed with the ImmunEasy adjuvant system (Qiagen). Spleen cells from the immunized mice were fused with myeloma cells. Hybridoma cells were screened by ELISA for the presence of antibodies that bound to the immunogen. Screening was also performed by Western blot, as described below. Hybridomas were selected and cloned using a single cell per well procedure.

2.5. Antibody purification

The mAb HL1899 was purified by FPLC on a Protein A Cartridge (Bio-Rad Laboratories, Inc. Hercules, CA) using ImmunoPure® IgG binding and elution buffers as described by the manufacturer (Pierce, Rockford, IL). The purified mAb was dialyzed in 10 mM sodium phosphate and 0.15 M sodium chloride, pH 7.5, and then concentrated using PEG 8000. Purity was determined by running them on gels, as described below. Purified mAb concentrations were determined using a bicinchoninic acid protein determination assay (Pierce, Rockford, IL).

2.6. Gel electrophoresis and Western immunoblot analysis

SDS-PAGE was carried out following the Laemmli system [45] using 4–20 % Ready Gel Tris HCl gels (Bio-Rad). Gels were stained with Coomassie brilliant blue R-250 or transferred onto PVDF membranes (PerkinElmer Life Sciences Inc., Boston, MA) [43]. The specific proteins, HagA and HagAU band, were subsequently detected by incubating with mAb 61BG1.3 and HL1899, respectively, followed by immunostaining with an anti-mouse IgG conjugated with alkaline phosphatase (ICN Biomedicals, Aurora, OH). The membranes were revealed using 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (Sigma-Aldrich) as the substrate.

2.7. Immunoelectron microscopy

Immunoelectron microscopy was used to localize the HagAU antigen in the wild-type and HagA mutant of P. gingivalis 381. Briefly, bacteria were grown overnight, harvested, and incubated on 400 nm nickel hexagonal grids. Grids were blocked with 1 % percent non-fat dry milk for 10 min, incubated with a 1/200 dilution of mAb HL1899 for 1 h at room temperature, and then washed 3 times with PBS. Anti-mouse IgG (Jackson ImmunoResearch Laboratories, West Grove, PA) labeled with 18-nm diameter gold particles was added and incubated for 1 h at room temperature and then washed three times with PBS. Controls for non-specific binding of the secondary gold-labeled antibody were performed without pre-incubation with the first antibody (data not shown). The grids were examined using a Hitachi 7000 transmission electron microscope at the ICBR Electron Microscopy Core Laboratory of the University of Florida.

2.8. Adhesion and invasion assays

HCAEC and GEC were seeded in 24-well tissue culture plates at 10⁵ cells per well and incubated for 18 h. Overnight broth cultures of P. gingivalis 381 (wild type and HagA mutant), E. coli-ST2, E. coli-pUC8 (control), E. coli-pEKS5, and E. coli-pET19b (control) were centrifuged and the pellet resuspended in antibiotic-free EBM-2 medium to a concentration of 10⁷ c.f.u. ml⁻¹ as determined with a spectrophotometer (Shimadzu, Kyoto, Japan). The HCAEC were washed with PBS and infected with 1 ml per well of the above P. gingivalis or E. coli strains. Three wells per bacterial strain were infected. The plates were incubated aerobically at 37 °C for 90 min and then washed 3 times with PBS. Cells were lysed with sterile H₂O for 20 min at 37 °C.

For the invasion assay, the wells were treated with 300 μg·ml⁻¹ gentamycin and 200 μg·ml⁻¹ metronidazole for 60 min at 37 °C to kill all extracellular bacteria prior to lysing the cells with sterile H₂O. Decimal serial dilutions were plated and c.f.u. were enumerated. Each assay was performed in duplicate. The adhesion or invasion ratios were then calculated as a percentage of the number of bacteria recovered/number of bacteria in the inoculum.

2.9. Competitive inhibition assay

In the antibody inhibition assay, the mAb HL1899 against HagAU, and bovine serum albumin (BSA, Sigma; control) were diluted in 5 fold-increasing concentrations of 10 to 250 μg·ml⁻¹, and added to the HCAEC. After a 30 min pre-incubation at 37 °C, P. gingivalis 381 was added to each well and the adhesion and invasion assays were performed as described above.

2.10. Statistical analysis

Differences between groups were determined by analysis of variance. Normality and equal variance of the data were confirmed in preliminary analysis. When indicated, multiple pairwise comparisons were done using the Student-Newman-Keuls test (SigmaStat® 3.0, SPSS Inc, Chicago, IL). For all comparisons, values of P<0.05 were considered significant.

3. Results

3.1. Adhesion and invasion assays

To determine whether HagA is involved in adhesion to and/or invasion of host cells, P. gingivalis 381 (wild type and HagA mutant) as well as two recombinant bacterial strains expressing HagA (E. coli-ST2 and E. coli-pEKS5) were used to infect HCAEC and GEC. Vector-only strains E. coli-pUC8 and E. coli-pET19b were used as negative controls respectively. The attachment and invasion of P. gingivalis wild type and the HagA mutant to HCAEC were compared. No statistically significant difference was observed between the two bacterial strains for adhesion (P=0.132) or invasion (P=0.093) (data not shown). However, a statistically significant difference (P<0.035) between the E. coli-pUC8 control and the E. coli-ST2 expressing the HagA protein was demonstrated for the adhesion assay to GEC (Fig. 2a, group 1). Furthermore, a statistically significant difference (P<0.0019) between the E. coli-pET19b control and the E. coli-pEKS5 expressing the HagA protein was observed (Fig. 2a, group 2). The presence of rHagA by these two E. coli strains increased adherence to GEC by a factor of 3.

Fig. 2.

Attachment of different E. coli strains to (a) GEC and (b) HCAEC. E. coli-pUC8 and E.coli-ST2 (Group 1) and E. coli-pUC19b and E. coli-pEKS5 (Group 2) were compared. Results are reported as the average (± standard deviation) of the attachment to host cells. Using GEC or HCAEC, a statistically significant difference (P<0.05) between the E. coli control and E. coli expressing HagA strains from each group was observed.

Adherence of the E. coli-ST2 to HCAEC was increased by about 2.5 times when compared to that of the E. coli-pUC8 control (P<0.002) (Fig. 2b, group 1). Also, a statistically significant difference (P<0.009) between the E. coli-pET19b control and the E. coli-pEKS5 expressing the HagA protein was observed, as adherence was increased by a factor of 4 (Fig. 2b, group 2). The increase in adhesion to GEC and HCAEC for the HagA-expressing strains did not lead to increase of GEC or HCAEC invasion (data not shown).

In contrast, HagA facilitated both adherence and invasion of host cells when cloned into a S. enterica strain (Fig. 3). The adherence and invasion of the HagA expressing S. enterica pNM1.1 strain was 3 and 4 times greater, respectively, than that of the negative control strains, showing statistically significant differences.

Fig. 3.

Salmonella enterica strains attachment (a) and invasion (b) to HCAEC. S. enterica control (solid bar) and pNM1.1 were compared. Results are reported as the average (± standard deviation) of the attachment or invasion to host cells. A statistically significant difference (P<0.05) between the both strains was observed during both the attachment and invasion steps.

3.2. Competitive inhibition assays

No HagAU-specific antibodies have been described in the literature, only antibodies recognizing the HArep in HagA or in the gingipains. MAb HL1899 developed against the unique 5′ N-terminal portion of Hemagglutinin A (HagAU) was first used here in transmission electron microscopy to localize HagA. As presented in Fig. 4, immunogold labeling indicates that HagA is a surface expressed antigen.

Fig. 4.

Transmission electron micrographs of wild-type (a) and HagA mutant (b) of P. gingivalis after a 1 h incubation with mAb HL1899 and followed by labeling with gold particles conjugated to a mouse IgG antibody.

Confirming the antibody specificity, the affinity purified mAb HL1899 against bound only to the P. gingivalis wild type strain and not to the HagAU mutant (Fig. 4). Interestingly, mAb HL1899 did not inhibit the attachment of P. gingivalis 381 to HCAEC. However, at a concentration of 250 μg·ml⁻¹, invasion was sharply reduced when compared to the control and this difference was statistically significant. BSA, which was used a control, did not affect the attachment to or invasion of host cells by P. gingivalis (Fig. 5).

Fig. 5.

Inhibition of the (a) attachment or (b) invasion of P. gingivalis 381 to HCAEC by pre-incubating cells with various concentrations of BSA (solid bars) or mAb HL1899 (dotted bars). Results are reported as the average (± standard deviation) of the attachment or invasion to host cells. The adhesion and invasion values for each P. gingivalis group without inhibitors were individually normalized to 100% and used to as control to derive the other adhesion and invasion values. The asterisk indicates a statistically significant difference (P<0.01) between that group and the group without inhibitors.

4. Discussion

In order to determine the role of the hemagglutinin A of P. gingivalis in colonization, a HagA mutant as well as several strains expressing recombinant HagA were tested for adherence and invasion of tissue culture cells.

We initially tested two P. gingivalis strains, a natural “mutant” containing only 2 repeats in its HagA gene, strain AJW4, and the database strain (W83) for adherence and invasion, using SEM. In previous studies, the strain with the lowest number of HagA repeats, AJW4, showed lowest invasion of endothelial cells [26], suggesting a role of the hemagglutinin repeats. The expression of key surface adhesins on invasive strains is suggested to be linked to invasive ability [46]. Indeed, a single S. enterica surface protein, Rck enabled non-invasive E. coli and Rck-coated beads to adhere to and invade host cells [47]. In this study, no difference could be observed between the P. gingivalis parent strain and the HagA mutant in adherence and invasion of GEC and HCAEC. As P. gingivalis possesses rgpA and kgp, both of which encode HA peptides with extensive homology to the HArep of HagA [48], the HagA mutant is functionally not a HA knockout because of the redundant activities. Consequently, it is expected that these HAs compensate for the defect in the hemagglutinin A expression. Other adhesins, such as the hemagglutinin B, previously shown to be involved in attachment to HCAEC could also compensate for the HagA defect [33].

However, other approaches in this study demonstrated that the HagA molecule is involved in attachment to host cells, as the Salmonella pNM1.1 strain expressing a complete HagA molecule (Fig. 1), which includes the unique portion as well as the HA repeats, attached better to cells than the control strain. In addition, both recombinant E. coli strains expressing portions of HagA also demonstrated increased adherence. As both recombinant E. coli strains had a section of the HA repeat in common, this suggest that the region involved in attachment to host cells is included in HArep. Furthermore, mAb HL1899 against the unique region (HagAU) was not able to inhibit adhesion of P. gingivalis to cells (see below for more details).

The native HagA protein in P. gingivalis is surface-associated, as we have shown using EM [31]. E. coli ST2 strain has two putative transmembrane domains and hemagglutinating properties which are associated with surface localization of the protein [31]. E. coli pEKS5 carries the same insert in expression vector therefore we expect to have similar localization. Based on these data, it was not surprising to observe, using EM, that the (entire) HagA, expressed in S. enterica, was also surface-associated [40].

The Salmonella strain expressing HagA invaded better than its control but no difference in invasion could be detected with the recombinant E. coli strains and their control strains. The various mechanisms of invasion as well as host cell specificity expressed by different bacterial species might explain why the S. enterica strain invaded HCAEC in this study while the E. coli did not, even when expressing HagA. It is likely that another bacterial surface protein, present in Salmonella but not in E. coli, participates in the invasion process. The increase of both adherence to and invasion of HCAEC by S. enterica expressing HagA may thus suggest an invasion mechanism supported in S. enterica but not in E. coli. Alternatively, as suggested by the repeat numbers, the entire HagA molecule might be required for invasion to occur. Finally, there is a possibility that different domains carry different functions as in other large multifunctional bacterial virulence proteins, such as the GAS serum opacity factor (SOF) protein [49].

As depicted in Fig. 1, the HagA molecule is large and contains a unique portion located at the N-terminal end followed by four repetitive HA fragments. To determine which portion of HagA mediate adherence, inhibition assays with a monoclonal antibody targeting the N-terminal unique portion of the HagA molecule as well as inhibition experiments with a peptide common to repeat regions of HagA were performed. The mAb HL1899 binds to P. gingivalis 381 but does not bind to the HagA mutant indicating the specificity of mAb HL1899 to hagA and not the rgp/kgp HAs. The same mAb did not affect attachment of P. gingivalis 381 to HCAEC but inhibited invasion in a dose-dependent manner. It has been reported that an (unrelated) antibody blocks binding of purified P. gingivalis RgpA protein to human cells at 10X lower concentration [50], indicating more limited contribution of the HagAU epitopes to the adhesion and higher contribution to invasion. As this mAb targets the unique part of HagA, it is possible that this region of the hemagglutinin facilitates invasion of host cells but not attachment. The reason for the inhibitory effect of the antibody but not of the mutation is most likely the presence in the genome of other ORFs that contain partially homologous to HagAU sequences, such as the genes for HagE and gingipains (both arginine- and lysine-specific).

Hemagglutinin A is also composed of four direct repeats proposed to contain the hemagglutinin functional domain. The functional motif for hemagglutination from strain P. gingivalis FDC381 was previously identified as “PVQNLT” [51]. This motif can also be found in rgp, kgp, and the hag gene family [51]. Also, the PVQNLT motif is part of the HGP44, which was shown to bind asialoglycophorin found on senescent erythrocytes. Additionally, HGP44B can bind host proteins such as fibronectin, fibrinogen, laminin, and collagen type V which might be important for adhesion and invasion of host cells [52]. It has been previously observed that mAb 1A1, produced against P. gingivalis whole cells, strongly inhibited agglutination of human erythrocytes [53]. Previously, specific antibodies produced from hen egg yolk antibodies against the recombinant 122k-HagA and the synthetic peptide containing PVQNLT inhibited the hemagglutinating activity of P. gingivalis [54]. In addition, further analysis demonstrated that the residue GVSPKVCKDVTVEGSNEFAPVQNLT was the epitope recognized by mAb 1A1 [53]. However, when using this amino stretch as a synthetic peptide, the binding of mAb 1A1 to P. gingivalis was not inhibited. Moreover, the hemagglutination property of P. gingivalis was not decreased [53].

Direct and indirect influence of molecules involved in adherence, hemagglutination and hemolysin has been proposed for P. gingivalis [55]. Hemagglutinins have been shown to be required for colonization of host cells in other microorganisms [56] and vaccines have been designed against them [57]. Previous studies suggest that further prophylactic and therapeutic approaches against periodontal diseases associated with P. gingivalis could be developed based on hemagglutinins. For example, human alpha- and beta- defensins were shown to bind P. gingivalis rHagB and enhance the IgG response in mice which could potentially lead to P. gingivalis clearance from the host [58–59]. HagA has been recognized by serum from a periodontal patient while we have previously reported that HagA expression in a S. enterica serovar Enterica vaccine strain was able to induce a humoral immune response against this hemagglutinin [40]. In conclusion, we demonstrate the importance of continuing investigation of both the unique and the repeat segment of HagA and their role in adhesion to and invasion of human cells and suggest that the unique N-terminal region is important for P. gingivalis invasion. The study supports the concept that HagA is an important molecule involved in invasion of host cells and could be a target to prevent colonization and subsequently disease.

Table 1.

Strains and plasmids used in this study

STRAINS	DESCRIPTION	REFERENCE
P. gingivalis FDC 381	Clinical isolate from a periodontitis patient	[60]
P. gingivalis 381	hagA mutant, clone A4	This study
E. coli ST2	Genomic HagA clone from P. gingivalis 381 (3.1 kb)	[31]
E. coli S17-1	F− pro recA1 (r−m+) RP4-2, integrated (Tc::Mu) Km::Tn7[SmrTpr]	[41]
E. coli BL21	E. coli B F− dcm ompT hsdS(rB− mB−) gal λ(DE3)	[61]
S. enterica χ4072	S. enterica serovar Typhimurium	[62]
	χ4072 SR-11 (Δcya Δcrp Δasd)
PLASMIDS
pNM1.1	pYA292 carrying 7.9 kb full-size hagA gene in SalI-SacI cloning sites	[40]
pHagAU	pET30a carrying 1.2 kb of hagA gene (5′ end)	This study
pVA3000	ermF/ermAM, suicide vector for Bacteroides, 5.3 kb	[41]
pEKS5	pET19b carrying 3.2 kb of hagA gene (see Fig. 1)	This study

Highlights.

• P. gingivalis is a pathogen involved in periodontal disease, and likely, cardiovascular disease.

• HagA is one of the prominent adhesins in P. gingivalis.

• This study investigated the role of both the unique and repeat regions of HagA in adherence/invasion.

• The data demonstrate that both the unique and repeat portions of HagA are involved in adherence but that only the unique portion is important for invasion.

Abbreviations

HagA

hemagglutinin

GEC

epithelial cells

HCAEC

human coronary artery endothelial cells

HArep

hemagglutinin A repeat

LB

Luria-Bertani

mAb

monoclonal antibody

IBCR

Interdisciplinary Center for Biotechnology Research