F. nucleatum ATCC 10953, P. gingivalis ATCC 33277, P. gingivalis clinical isolate KUMC-P4 [10], Prevotella intermedia ATCC 25611, and T. denticola ATCC 33521 were grown and counted as previously described [6]. For the fluorescence studies, bacteria were stained with 5µM of 5-(and-6)-carboxy-fluorescein diacetate succinimidyl ester (CFSE) (Molecular Probes, Carlsbad, CA, USA) before being used. Immortalized human gingival keratinocyte HOK-16B cells were maintained in keratinocyte growth culture medium (Clonetics, San Diego, CA, USA) containing supplementary growth factors. For bacterial infection, HOK-16B cells were plated one day before infection. At 80% confluence, the cells were infected with live bacteria at a multiplicity of infection of 1000 in the medium without antibiotics. The infected cells were cultured at 37℃ in a water-saturated atmosphere of 95% air and 5% CO₂.

HSA and human α-1 antitrypsin isolated from human plasma were purchased from Sigma (Saint Louis, MO, USA). Intravenous immunoglobulin (IVIG from Green Cross, Yongin, Korea) was used as pooled human immunoglobulin G (phIgG). Sterile human serum purchased from Sigma was used after heat inactivation.

The flow cytometric adhesion and invasion assay was performed as previously described [6]. Briefly, HOK-16B cells were plated at 6×10⁴ cells/well in triplicate in 24-well plates. For the adhesion assay, HOK-16B cells were incubated with CFSE-labeled bacteria in the absence or presence of various serum components for 1 hour on ice. The cells were then washed, detached with trypsin-ethylenediaminetetraacetic acid and immediately analyzed by flow cytometry. For the invasion assay, HOK-16B cells were infected with CFSE-labeled bacteria in the absence or presence of various serum components or N-acetylcysteine (NAC from Sigma) for 24 hours at 37℃. The cells were detached, washed, and then analyzed by flow cytometry after quenching the fluorescence of the bacteria bound on the surface with 300 µL of 0.4% trypan blue. Cells exposed to the same amount of CFSE-labeled bacteria after fixing with 3.7% formaldehyde served as a negative control. The viability of the HOK-16B cells was determined based on their forward scatter and the FL-3 fluorescence of the trypan blue staining, and bacterial invasion was analyzed only for the live cells. It was confirmed that bacterial infection or serum components did not affect the viability of HOK-16B cells.

HOK-16B cells were plated at 2.5×10⁴ cells/well in 96-well flat bottom black plates (NUNC, Roskilde, Denmark). The next day, the cells were stained with 10µM 5-(and-6)-carboxy-2',7'-dichlorodihydrofluorescein diacetate (carboxy-H₂DCFDA, Life Technologies Korea, Seoul, Korea) for 45 minutes, washed, and then infected with CFSE-labeled F. nucleatum in the absence or presence of various serum components or NAC. In our preliminary study, the basal and F. nucleatum-induced reactive oxygen species (ROS) levels increased continuously during the 24 hours infection period. Therefore, the DCF fluorescence was read at 24 hours using a FLUOstar OPTIMA (BMG LABTECH, Ortenberg, Germany).

HOK-16 B cells were plated at 2×10⁶ cells per 100-mm plate. At 80% confluence, the cells were infected with F. nucleatum in the absence or presence of HSA, phIgG, or NAC for the indicated time. The cells were lysed with ice cold 1× lysis buffer. The lysates were incubated with Rhotekin-bound agarose (Cell Biolabs, San Diego, CA, USA) to isolate Rho-GTP and with PAK-PBD-bound agarose (Cell Biolabs) to isolate both Rac-GTP and Cdc42-GTP. The GTPγs-and GDPγs-treated cell lysates were included as positive and negative controls, respectively. The agarose/protein complex was collected by centrifugation, washed, and then subjected to western blotting using anti-Rac1, anti-RhoA, or anti-Cdc42 antibodies.

The experimental protocol and animal handling procedures were approved by the Seoul National University Animal Care and Use Committee (#SNU-130911-7) and conducted in accordance with the recommendations in the Guide from the Institute of Laboratory Animal Resources, Seoul National University, under the Laboratory Animals Act 9025 of the Republic of Korea. Six-week-old female Balb/c mice were purchased from Orient Bio (Seongnam, Korea) and housed under specific-pathogen-free conditions in the Laboratory Animal Facility at the School of Dentistry, Seoul National University. Experimental periodontitis was induced as described by Baker et al. [11] with slight modification. The mice (n=40) were given sulphamethoxazole/trimethoprim (Samil, Ansan, Korea) in their drinking water for 10 days. Following a four-day resting period, mice were randomly divided into four groups: Sham, periodontitis (PD), PD_HSA+phIgG, or PD_NAC. The sham group received 2% carboxymethylcellulose in phosphate buffered saline; the other groups received oral inoculation of P. gingivalis KUMC-P4 (10⁹ cells) plus T. denticola (10⁹ cells) in 100 µL of 2% carboxymethylcellulose six times on alternative days. During the inoculation period of two weeks, the PD_HSA+phIgG and PD_NAC groups were fed with 6.6-mg/mL HSA plus 1-mg/mL phIgG and 10mM NAC in their drinking water, respectively. All mice were euthanized four weeks after the last inoculation. The right half of the maxilla was defleshed, immersed overnight in 3% hydrogen peroxide and stained with 1% methylene blue. The images of the lingual sites were taken under a dissecting microscope (40×). The distance from the cementoenamel junction to the alveolar bone crest at six predetermined lingual sites below each cusp were blindly measured with CellSens digital imaging software (Olympus Korea, Seoul, Korea) on coded images.

The differences between the control and experimental groups in the in vitro studies were determined using a two-tailed Student t-test. All animal data groups passed a normality test. Thus, the one-way analysis of variance and post hoc tests were applied to determine the between-group differences. P<0.05 was adopted as the significance level.

To identify the responsible component(s) of human serum that inhibit the bacterial invasion of gingival epithelial cells, three major components of human serum, including albumin, IgG, and α1-antitrypsin, were chosen and their effect on the invasion of CFSE-labeled F. nucleatum into immortalized human gingival epithelial HOK-16B cells was examined by flow cytometric invasion assay. F. nucleatum was chosen because of its high invasive ability among the several oral species with known invasive capacity [6]. The selected concentrations of each serum component are equivalent to those present in 5%, 10%, and 15% of human serum, respectively [12]. Both HSA and phIgG inhibited the invasion of F. nucleatum in a dose-dependent manner, while α1-antitrypsin had no effect (Fig. 1A left panel). In repeated experiments, the HSA significantly inhibited the invasion of F. nucleatum but less effectively than 2% human serum at all three concentrations. The phIgG was more effective than the HSA, inhibiting the invasion of F. nucleatum almost completely at 1.5 mg/mL (Fig. 1A right panel).

We next investigated whether the inhibitory effect of HSA and phIgG on bacterial invasion could be extended to other bacterial species. P. intermedia, P. gingivalis, and T. denticola are increased in periodontal lesions and have substantial invasive capacities [6,13]. HSA significantly decreased the invasion of P. intermedia and P. gingivalis but not that of T. denticola, while phIgG was effective for all three species. In addition, both HSA and phIgG were also effective for the P. gingivalis clinical isolate KUMC-P4 (Fig. 1B).

Bacterial adhesion is a critical step prior to invasion into host cells. The effect of HSA and phIgG on bacterial adhesion was thus examined. The HSA significantly reduced the adhesion of F. nucleatum onto HOK-16B cells in a dose-dependent manner. In contrast to our expectations, phIgG did not inhibit bacterial adhesion (Fig. 2A).

The antioxidant function of HSA is well-characterized [14]. ROS are now recognized to play essential roles in signal transduction throughout various physiological processes, including the actin cytoskeleton reorganization that is required for bacterial invasion [15,16]. Thus, the effect of HSA and phIgG on the levels of intracellular ROS in HOK-16B cells was determined by DCF fluorescence. Infection with F. nucleatum up-regulated ROS in HOK-16B cells. All three doses of HSA and 2% human serum equivalently reduced more than 50% of both the basal and F. nucleatum-induced ROS. In contrast, phIgG slightly increased the basal levels of ROS in a dose dependent manner (Fig. 2B). To confirm the role of ROS in bacterial invasion, HOK-16B cells were infected with F. nucleatum in the presence of NAC, a free radical scavenger. NAC inhibited both the levels of ROS and the invasion of F. nucleatum in a dose-dependent manner (Fig. 2C, D).

The bacterial invasion of host cells depends on the dynamic rearrangement of the actin cytoskeleton. The Rho family GTPases, including RhoA, Rac1, and Cdc42, are important regulators of actin cytoskeleton dynamics [17]. To examine the involvement of RhoA, Rac1, and Cdc42 in F. nucleatum invasion, HOK-16B cells were stimulated with F. nucleatum for 0.5, 1, or 2 hours, and the activation of RhoA, Rac1, and Cdc42 was examined using a pull-down assay. F. nucleatum induced activation of Rac1 and concomitant inactivation of RhoA, while Cdc42 was not affected (Fig. 3A). Subsequently, the effect of HSA, phIgG, and NAC on the F. nucleatum-induced Rac1 activation was examined. All three reagents reduced the amount of activated Rac1 upon infection with F. nucleatum (Fig. 3B), corroborating their inhibitory effect on invasion.

We tested whether these reagents with inhibitory effects on bacterial invasion could prevent periodontitis in mice. Periodontitis was induced by oral infection with the periodontal pathogens P. gingivalis and T. denticola and the effect of either HSA+phIgG or NAC on alveolar bone loss was examined. Because HSA was more effective than phIgG at inhibiting P. gingivalis invasion but was not effective on T. denticola, HSA and phIgG were administered together. P. gingivalis and T. denticola resulted in significant alveolar bone destruction. Both the PD_HSA+phIgG and PD_NAC groups showed significantly lower amounts of alveolar bone loss than the PD group (P=0.03 and P=0.006, respectively). In particular, NAC completely prevented the alveolar bone loss induced by the periodontal pathogens (Fig. 4).