OUP user menu

Sublingual vaccination with fusion protein consisting of the functional domain of hemagglutinin A of Porphyromonas gingivalis and Escherichia coli maltose-binding protein elicits protective immunity in the oral cavity

Satoshi Yuzawa, Tomoko Kurita-Ochiai, Tomomi Hashizume, Ryoki Kobayashi, Yoshimitsu Abiko, Masafumi Yamamoto
DOI: http://dx.doi.org/10.1111/j.1574-695X.2011.00895.x 265-272 First published online: 1 March 2012


This study demonstrated that sublingual immunization with a fusion protein, 25k-hagA-MBP, which consists of a 25-kDa antigenic region of hemagglutinin A purified from Porphyromonas gingivalis fused to maltose-binding protein (MBP) originating from Escherichia coli as an adjuvant, elicited protective immune responses. Immunization with 25k-hagA-MBP induced high levels of antigen-specific serum IgG and IgA, as well as salivary IgA. High level titers of serum IgG and IgA were also induced for almost 1 year. In an IgG subclass analysis, sublingual immunization with 25k-hagA-MBP induced both IgG1 and IgG2b antibody responses. Additionally, numerous antigen-specific IgA antibody-forming cells were detected from the salivary gland 7 days after the final immunization. Mononuclear cells isolated from submandibular lymph nodes (SMLs) showed significant levels of proliferation upon restimulation with 25k-hagA-MBP. An analysis of cytokine responses showed that antigen-specific mononuclear cells isolated from SMLs produced significantly high levels of IL-4, IFN-γ, and TGF-β. These results indicate that sublingual immunization with 25k-hagA-MBP induces efficient protective immunity against P. gingivalis infection in the oral cavity via Th1-type and Th2-type cytokine production.

  • hemagglutinin A
  • Porphyromonas gingivalis
  • sublingual immunization


Periodontal disease is a chronic inflammatory malady that causes both alveolar bone absorption followed by tooth loss, as well as systemic diseases such as cardiac disease (Destefano et al., 1993), diabetes mellitus (Roeder & Dennison, 1998), osteoporosis (Krejci, 1996; Reddy, 2002), and premature, low-birth-weight babies (Offenbacher et al., 1996). Therefore, prevention or treatment of periodontal disease is very important for maintaining health. Porphyromonas gingivalis, which is a gram-negative and asaccharolytic anaerobic bacterium with high adherence activity to erythrocytes and epithelial cells, is one of the major virulent bacteria causing periodontal disease. It exerts virulence through fimbriae, lipopolysaccharides, outer membrane proteins, and outer membrane vesicles (Holt et al., 1999). Hemagglutinin protein, which is expressed on the cell surface of P. gingivalis, regulates bacterial adhesion to the host cells, as well as agglutinates and hemolyzes erythrocytes. Multiple hemagglutinin genes have been cloned from P. gingivalis by functional screening (Lee et al., 1996; Lépine et al., 1996; Song et al., 2005). Among these, hemagglutinin A (hagA) is thought to possess a functional domain and thus to be a potential candidate for periodontal vaccination.

Previous studies have demonstrated the efficacy of mucosal immunization for delivering vaccines, which induces mucosal and systemic immune responses via oral (Yamamoto et al., 1997; Liu et al., 2010), nasal (Koizumi et al., 2008; Momoi et al., 2008), and sublingual routes (Cuburu et al., 2007; Song et al., 2008; Zhang et al., 2009). Of the vaccination methods available, sublingual vaccination has recently been reported to induce significant antibody (Ab) production in nasal, bronchial, and oral mucosa (Cuburu et al., 2007; Zhang et al., 2009). Moreover, sublingual immunization does not allow the antigen to be carried to the central nervous system to the same extent as nasal vaccination (Cuburu et al., 2007). Finally, sublingual vaccines require much less of the antigen than is required for intragastric vaccination. Also, sublingual mucosa have been proposed to be more permeable to low-molecular-weight drugs (Zhang et al., 2002) and small immunogenic peptides than the cheek mucosa (Squier, 1991), a general oral mucosa that contains dendritic cells (DCs). DCs take up foreign antigens in the submucosal region, which migrate to the regional lymph nodes, where the antigen is presented to T-lymphocytes by DCs to activate the adaptive immune responses (Song et al., 2009).

The simultaneous application of adjuvants with an antigen can efficiently induce an antigen-specific immune response. Maltose-binding protein (MBP) is a high affinity maltose/maltodextrin-binding protein and a periplasmic receptor for the capture and transport of maltodextrins from the periplasmic space in gram-negative bacteria (Fox et al., 2001; Fernandez et al., 2007). MBP was recently reported to act as an adjuvant that elicits innate immunity through Toll-like receptor 4 (TLR4) (Fernandez et al., 2007). Given that MBP can easily be prepared by taking advantage of its characteristic binding to maltose (Zhu et al., 2007), as well as the enhanced solubility and stability of fusion proteins, MBP is used to facilitate the production and delivery of subunit vaccines against various pathogenic bacteria and viruses (Fox et al., 2001; Routzahn & Waugh, 2002). Although hagA was originally easy to aggregate as an inclusion body (Fox et al., 2001), even the minimal antigenic region of the 25-kDa protein, the fusion form of the 25k-hagA-MBP protein used in this study, is drastically easier to dissolve under hydrophilic conditions. Therefore, we analyzed the immune responses induced by the fusion protein 25k-hagA-MBP, which comprises the 25-kDa antigenic region of hagA purified from P. gingivalis, including the hemagglutinin-associated minimum motif ‘PVQNLT’ amino acid sequence in the Ab recognition sites (Shibata et al., 1999) as well as MBP from Escherichia coli, to assess the potential sublingual vaccine for preventing P. gingivalis infection.

Materials and methods


Female 8–11-week-old BALB/c mice were purchased from Sankyo Laboratory Services (Tokyo, Japan) and maintained under pathogen-free conditions in the experimental facility of Nihon University School of Dentistry at Matsudo. Mice received sterile food and water. All animals were maintained and used in accordance with the Guidelines for the Care and Use of Laboratory Animals (Nihon University School of Dentistry at Matsudo).

Antigen purification

Plasmid pMD157-expressing 25k-hagA-MBP was kindly provided by Dr Yoshimitsu Abiko (Nihon University). The antigen was purified using a p-MAL4 protein purification kit (Bio-Rad, Hercules, CA) (Riggs, 2000; Suyama et al., 2004; Kobayashi et al., 2006). Briefly, a gene fragment encoding the functional domain of hemagglutinin originating from hagA on the surface of P. gingivalis was inserted into the p-MAL plasmid pMD157, followed by transfection to E. coli and incubation. After 1 or 2 days of incubation, the E. coli suspension was centrifuged and the pellet was homogenized. The homogenized suspension was subjected to the dialysis treatment, gel-filtration chromatography, and ion-exchange chromatography. Finally, isolation of the antigen was performed using amylose resin column affinity chromatography, and 25k-hagA was obtained via cleavage treatment of 25k-hagA-MBP using Factor Xa (New England BioLabs, Ipswich, MA).

Immunization and sample collection

For sublingual immunization on days 0, 7, and 14, mice were anesthetized with pentobarbital, and 30 µL of phosphate-buffered saline (PBS) containing 50 µg of 25k-hagA-MBP was delivered with a micropipette applied against the ventral side of the tongue while directed toward the floor of the mouth. Mice were immunized with 7.5 µL of antigen four times (total volume = 30 µL). Ten minutes of interval were set between each administrations. A nonimmunized group was PBS treated. Animals were maintained with their heads placed in ante flexion for 30 s during each delivery. Serum and saliva were collected from each group to examine the 25k-hagA-MBP-specific Ab responses.

25 k-hagA-MBP-specific Ab responses

Ab titers were detected using an enzyme-linked immunosorbent assay (ELISA) as described previously (Maeba et al., 2005). Briefly, plates were coated with 25k-hagA-MBP (5 µg mL−1). After washing with PBS containing 0.05% Tween 20, plates were blocked with PBS containing 1% bovine serum albumin. Next, serial dilutions of serum or saliva samples were added in duplicate. The starting dilution of the serum was 1 : 26, while that of the saliva was 1 : 22. The plates were incubated for 5 h at room temperature, washed, and then incubated with horseradish peroxidase-labeled goat anti-mouse heavy chain γ, γ1, γ2a, γ2b, γ3, or α-specific antibodies (Southern Biotechnology Associates, Birmingham, AL) at 4 °C for 20 h. Finally, 2,2′-azino-bis (3-ethylbenz-thiazoline-6-sulfonic acid) (ABTS) with H2O2 (Moss, Inc., Pasadena, MD) was added for color development. Endpoint titers were expressed as the reciprocal log2 of the last dilution, which gave an optical density at 415 nm of 0.1 greater than that of nonimmunized control samples after 15 min of incubation.

Measurement of Ab-forming cells (AFCs)

Single-cell suspensions were obtained from the salivary gland 7 days after the last immunization. Briefly, salivary glands were carefully extracted, teased apart, and dissociated using 0.3 mg mL−1 collagenase (Nitta Gelatin Co. Ltd, Osaka, Japan) in RPMI-1640 (Wako Pure Chemical Industries Ltd, Osaka Japan). Mononuclear cells were obtained at the interface of the 50% and 75% layers of a discontinuous Percoll gradient (GE Healthcare UK, Ltd, Little Chalfont, Buckinghamshire, UK) (Maeba et al., 2005). To assess the numbers of antigen-specific AFCs, an enzyme-linked immunospot (ELISPOT) assay was performed as described previously (Yamamoto et al., 1997). Briefly, 96-well nitrocellulose plates (BD Biosciences, Franklin Lakes, NJ) were coated with 25k-hagA-MBP (5 µg mL−1), incubated at 4 °C for 20 h, and then washed extensively before being blocked with RPMI-1640 containing 10% fetal calf serum. After 30 min, the blocking solution was discarded, and cell suspensions at various dilutions were added to wells and incubated at 37 °C for 4 h under 5% CO2 in moist air. The cells were washed and then incubated with horseradish peroxidase-conjugated goat anti-mouse heavy chain α-specific antibodies (Southern Biotechnology Associates) at 4 °C for 20 h. Following incubation, the plates were washed with PBS and developed adding 3-amino-9-ethylcarbazole dissolved in 0.1 M sodium acetate buffer containing H2O2 to each well (Moss, Inc.). Plates were incubated at room temperature for 30 min and washed with distilled water, and AFCs were then counted with the aid of a stereomicroscope (Olympus, Tokyo, Japan).

k-hagA-MBP-specific mononuclear cell responses

Mononuclear cells were isolated 7 days after the final immunization from submandibular lymph nodes (SMLs) of the immunized mice, adjusted to a concentration of 5 × 106 cells mL−1, and cultured with 5 µg mL−1 of 25k-hagA-MBP in RPMI-1640 medium containing 10% fetal bovine serum, 50 µM 2-mercaptoethanol, 15 mM HEPES, 2 mM l-glutamine, 100 U mL−1 penicillin, 100 µg mL−1 streptomycin, and 10 U mL−1 of recombinant IL-2 (Genzyme, Cambridge, MA). Cultures were incubated for 4 days at 37 °C under 5% CO2 in air. To measure the 25k-hagA-MBP-specific cell proliferation, 1.0 µCi of [3H]thymidine was added to the culture 18 h before harvesting, and the incorporated radioactivity was measured by scintillation counting. Four-day culture supernatants were also collected and centrifuged to remove cell debris. The IL-4, IFN-γ, and TGF-β cytokine levels of the culture supernatants were then determined by cytokine-specific ELISA kit (Pierce Endogen; Pierce Biotechnology, Rockford, IL) as described previously (Hashizume et al., 2008).

Oral infection

Mice were orally infected with P. gingivalis as described previously (Du et al., 2011), with minor modifications. Briefly, mice were given ad libitum access to ionized water containing sulfamethoxazole/trimethoprim (Sulfatrim; Goldline Laboratories, Fort Lauderdale, FL) at 10 mL per pint for 10 days. This was followed by a 3-day antibiotic-free period. Mice were then administered 109 CFU of P. gingivalis suspended in 100 µL of PBS with 2% carboxymethylcellulose via oral topical application. Mice were inoculated five times a week (from Monday to Friday) for 3 weeks, for a total of 15 inoculations. Control groups included sham-infected mice, which received antibiotic pretreatment and carboxymethylcellulose without P. gingivalis.

Measurement of alveolar bone loss

Horizontal bone loss around the maxillary molars was assessed by the morphometric method as described previously (Klausen et al., 1989). The distance from the cementoenamel junction (CEJ) to the alveolar bone crest (ABC) was measured at a total of 14 buccal sites per mouse. Bone measurements were performed a total of three times by two evaluators using a random and blinded protocol.

Statistical analysis

Results are expressed as means ± standard deviation (SD) and were compared using an unpaired Student's t test.


Induction of 25k-hagA-MBP-specific Ab responses

To determine the effectiveness of the sublingual immunization, mice were immunized with 25k-hagA, 25k-hagA-MBP, or PBS. Sublingual immunization with 25k-hagA-MBP induced significant serum IgG and IgA 7 days after the final immunization (Fig. 1a). In contrast, 25k-hagA-immunized and nonimmunized mice induced low or no detectable titers, respectively, after sublingual immunization. In addition, the serum IgG and IgA Ab responses induced by 25k-hagA-MBP persisted for almost 1 year (Fig. 1b). When the subclasses of antigen-specific IgG antibodies induced by sublingual 25k-hagA or 25k-hagA-MBP challenge were determined, all IgG subclasses were significantly enhanced in 25k-hagA-MBP group. On the other hand, 25k-hagA-immunized group showed a low level of IgG1 (and sparse IgG2b) (Fig. 1c).

Figure 1

Sublingual administration of 25k-hagA-MBP induces antigen-specific antibody responses in serum IgG and IgA on day 21 (a). Mice were immunized with 50 µg of 25k-hagA-MBP on days 0, 7, and 14, sublingually. Serum samples were collected 1 week after the third immunization and assayed for the detection of IgG and IgA. Long-term detection of sublingually administered 25k-hagA-MBP-specific immunoglobulin G and A Ab responses in serum (b). IgG subclass responses in mice given 25k-hagA-MBP (c). Results are expressed as means ± SD for four mice per group in a total of three experiments. **P < 0.001 vs. mice given 25k-hagA alone. N.D., not detectable.

Sublingual immunization of 25k-hagA-MBP induced high levels of 25k-hagA-MBP-specific IgA Ab responses in saliva (Fig. 2a). In contrast, essentially no IgA was detected in the saliva of mice sublingually treated with 25k-hagA or PBS. The most 25k-hagA-MBP-specific IgA AFCs were detected in the salivary glands suspensions (Fig. 2b).

Figure 1

25k-hagA-MBP-specific immunoglobulin A (IgA) antibody responses in saliva (a) and numbers of IgA AFCs in salivary glands (b). Mice were sublingually immunized with 25k-hagA or 25k-hagA-MBP, and samples were collected 7 days after the last immunization followed by the detection of 25k-hagA-MBP-specific IgA titers. Mononuclear cells from salivary glands were assessed for IgA AFCs. Results are expressed as means ± SD for four mice per group in a total of three experiments. *P < 0.01 vs. mice given PBS. N.D., not detectable.

k-hagA-MBP-specific mononuclear cell responses

As sublingual immunization with 25k-hagA-MBP elicited 25k-hagA-MBP-specific Ab responses in both mucosal and systemic compartments, establishing the nature of the T cell help supporting the responses was important. When mononuclear cells from the SMLs of immunized mice were restimulated with 25k-hagA-MBP in vitro, significant levels of proliferative responses were induced (Fig. 3a). In contrast, no significant proliferation or cytokine production was observed in hagA-immunized mice (data not shown). Furthermore, mononuclear cells isolated from SMLs immunized with 25k-hagA-MBP showed higher production of IL-4, IFN-γ, and TGF-β (Fig. 3b). These data indicate that sublingually immunized 25k-hagA-MBP-specific Th1-type and Th2-type responses are induced in SMLs.

25k-hagA-MBP-specific mononuclear cell responses in SMLs (a). Groups of mice were immunized sublingually with 25k-hagA-MBP as described in the legend to Fig. 1. Mononuclear cells were isolated from the SMLs of nonimmunized or 25k-hagA-MBP-immunized mice and were cultured with 25k-hagA-MBP. To measure cell proliferation, 1.0 µCi of [3H]thymidine was added to the culture 18 h before harvesting, and incorporated radioactivity was measured by scintillation counting. To analyze cytokine synthesis, culture supernatants were harvested, and the levels of secreted cytokines were assessed by cytokine-specific ELISA (b). Results are expressed as means ± SD for five mice per group in a total of three experiments. *P < 0.01; **P < 0.001 vs. mice given PBS.

Sublingual immunization with 25k-hagA-MBP reduces alveolar bone loss following the oral infection of P. gingivalis

Given that sublingual immunization with 25k-hagA-MBP elicited long-term antigen-specific Ab responses in sera, we sought to determine whether these antibodies were capable of suppressing the alveolar bone absorption caused by P. gingivalis infection. Thus, mice given 25k-hagA, 25k-hagA-MBP, and PBS were infected orally with P. gingivalis 7 days after the last immunization. Mice immunized with 25k-hagA-MBP showed a significant protection and reduced bone loss caused by P. gingivalis infection (Fig. 4). In contrast, mice immunized with 25k-hagA alone did not show the reduced level of bone loss by P. gingivalis infection. These findings indicate that sublingual immunization with 25k-hagA-MBP is protective against oral infection by P. gingivalis.

Decreased levels of Porphyromonas gingivalis-induced alveolar bone loss by sublingual immunization with 25k-hagA-MBP. Mice were immunized sublingually with either 25k-hagA, 25k-hagA-MBP, or PBS, as described in the legend to Fig. 1. Seven days after immunization, mice were inoculated orally with 109 CFU of P. gingivalis in 2% carboxymethylcellulose, as described in the . Control mice were sham-infected mice inoculated with 2% carboxymethylcellulose only. The distance (µm) from the CEJ to ABC was measured at 14 predetermined sites in defleshed maxilla and was totaled for each mouse. The results are expressed as means ± SD for eight mice per group in a total of three experiments. *P < 0.01 vs. mice given PBS.


Previous studies have demonstrated that intragastric and nasal immunizations are very potent methods to deliver drugs or small molecules to the systemic environment (Koizumi et al., 2008; Momoi et al., 2008; Liu et al., 2010). However, intragastrically administered antigens must be subjected to degradation processes prior to absorption through the lamina propria or Peyer's patches. This requires that a mouse be challenged with a much greater amount of antigen than other routes, which may induce immune tolerance (Mestecky et al., 1996; McSorley & Garside, 1999). In contrast, nasal administration is a well-established route of mucosal immunization because antigens are not subjected to such degradation processes. However, nasally administered antigens, such as the cholera toxin or influenza vaccine, threaten to migrate to the olfactory nerve and on to the central nervous system given their affinity for nerve tissue (van Ginkel et al., 2000; Mutsch et al., 2004).

These drawbacks make sublingual vaccination a superior alternative given that a much lower dose is required than for intragastric vaccination. Sublingual mucosa are permeable to drugs and can deliver low-molecular-weight molecules to the bloodstream while avoiding enterohepatic circulation and the immediate destruction of ingested molecules by gastric acid or partial first-pass effects of hepatic metabolism (Cuburu et al., 2007). Moreover, sublingually administered antigens have no propensity to migrate to the central nervous system (Cuburu et al., 2007). In addition to these advantages, sublingual vaccination induces substantially greater immune responses compared with nasal vaccination. Together, these advantages indicate that sublingual administration is an effective means of delivering drugs or low-molecular-weight molecules to protect against infectious diseases.

MBP has been used as a chaperone component in vaccines to enhance Ag-specific humoral and cellular immune responses (Seong et al., 1997; Rico et al., 1998). Therefore, we assessed the efficacy of sublingual immunization with the fusion protein 25k-hagA-MBP. Our results demonstrate that a sublingual challenge with 25k-hagA-MBP elicited high titers of the 25k-hagA-MBP-specific serum IgG and IgA Ab responses. Furthermore, these antibodies persisted for almost 1 year. As MBP adjuvanticity is mediated via signaling through TLR4 (Fernandez et al., 2007), we also tested whether the antigen-specific immune responses are induced in TLR-4 (the receptor of MBP) KO mice. As expected, neither antigen-specific IgG nor IgA antibodies were detected after sublingual immunization in these mice (S. Yuzawa, T. Kurita-Ochiai, T. Hashizume, R. Kobayashi, Y. Abiko & M. Yamamoto, unpublished data). A significantly high salivary IgA Ab titer was associated with the number of 25k-hagA-MBP-specific Ab-producing cells in the salivary gland.

Our results also showed that predominant mononuclear cell proliferation and cytokine production occurred in SMLs, in which 25k-hagA-MBP-specific helper T cells produced significant IL-4 and IFN-γ, which favor Th1-type and Th2-type responses, together with the increased production of TGF-β. In accordance with the production of cytokines, antigen-specific serum IgG subclasses are predominant in IgG1, IgG2b, and relatively lower in IgG2a responses. These results are also in accordance with previous observations that sublingual immunization might favor the induction of both Th1-type and Th2-type responses (Cuburu et al., 2007; Zhang et al., 2009). In contrast, nasal vaccination with 25k-hagA-MBP exhibited Th2-type responses owing to the predominant production of IL-4 with no IFN-γ (Du et al., 2011). This discrepancy may indicate that the induction of Th1-type and Th2-type responses is determined by the route of the vaccine rather than the properties of the vaccine antigens. Therefore, antigens should be administered in the most effective way to induce the suitable immune response. Additionally, TGF-β has been shown to play key roles in IgG2b production and IgA class switch. After sublingual immunization with 25k-hagA-MBP, it is surely confirmed that IgA and IgG2b production was increased in accordance with the level of TGF-β.

In summary, this study provides evidence that sublingual immunization with the fusion protein 25k-hagA-MBP augmented the activity of IFN-γ-producing Th1- and IL-4-producing Th2-type cells for the induction of serum IgG, IgA, and mucosal IgA Ab responses. Furthermore, 25k-hagA-MBP-specific immune responses provided protective immunity against alveolar bone loss after P. gingivalis infection. These results suggest that sublingual immunization with 25k-hagA-MBP may be a candidate for an efficient and safe vaccine against periodontal infection.


We thank Mitsuo Hayakawa for help with the antigen preparation. This work was supported by an ‘Academic Frontier’ Project for Private Universities matching fund subsidy from the Ministry of Education, Culture, Sports, Science and Technology, Japan, 2007–2011.


  • Editor: Richard Marconi


View Abstract