| To whom correspondence should be addressed: Tamotsu Yoshimori, Department of Cellular Regulation, Research Institute for Microbial Diseases, Osaka University, 3-1 Yamadaoka, Suita, Osaka, 565-0871, Japan. Tel: +81–6–6879–8293, Fax: +81–6–6879–8294 E-mail: tamyoshi@biken.osaka-u.ac.jp Abbreviations: MV beads, fluorescent beads coated with bacterial membrane vesicles; MβCD, methyl-β-cyclodextrin; GFP, green fluorescent protein; GPI-APs, glycosylphosphatidylinositol-anchored proteins. |
Successful establishment of bacterial infection requires adhesion and entry to host cells by a pathogen, which allows it to occupy various niches within the human bodies. Various bacterial pathogens have been shown to enter non-phagocytic cells by their utilization of a large and diverse array of adhesion and invasion molecules, enabling them to exploit host cell surface components as receptors (Cossart and Sansonetti, 2004). An intracellular location may be necessary for the pathogen to escape from immune surveillance by the host and antibiotic pressure, leading to intracellular persistence, multiplication, or dissemination to adjacent tissues (Cossart and Sansonetti, 2004). Many pathogens that were formerly considered to be extracellular living organisms have recently been shown to invade host cells (Wong and Isberg, 2005). Such pathogens seem to exploit host cell receptors, particularly α5β1 integrin, not only for adherence but also to trigger massive F-actin polymerization that induces cellular engulfment of the pathogens. For example, interaction of α5β1 integrin with bacterial ligands was shown to trigger the recruitment of Rho GTPases Cdc42 and/or Rac, inducing actin rearrangements that eventually permitted invasion by group A streptococcus (Ozeri et al., 2001), Staphylococcus aureus (Agerer et al., 2005), and Escherichia coli (Plancon et al., 2003).
Cellular lipid rafts are membrane sub-domains rich in cholesterol, sphingolipids, and specific membrane proteins, such as glycosylphosphatidylinositol-anchored proteins (GPI-APs), glycosphingolipids (GM1, Gb3, and GD1), and caveolin (Simons and Ikonen, 1997). These rafts are reportedly involved in cholesterol homeostasis, endocytosis, and cell signalling (Lafont et al., 2004; Riethmuller et al., 2006). Recently, lipid rafts were suggested to be exploited as bacterial entry sites by a wide range of pathogens such as Neisseria gonorrhoeae (Grassme et al., 1997), Pseudomonas aeruginosa (Grassme et al., 2003), Escherichia coli (Shin et al., 2000), S. aureus (Kolesnick et al., 2000), Chlamydia trachomatis (Stuart et al., 2003), and Shigella flexneri (Lafont et al., 2002). Multiple steps leading to infection, including bacterial adhesion, perforation of the host cell membrane, and signaling to trigger engulfment of bacteria, likely depend on raft components such as cholesterol, GPI-APs, and caveolin, though the mechanisms that underlie these interactions are only starting to be unraveled.
Periodontitis, one of the most common infectious diseases that afflict humans, is characterized by gingival inflammation, as well as loss of connective tissue and bone from around the roots of teeth, which leads eventually to tooth exfoliation (Genco, 1996). Porphyromonas gingivalis is considered to be a bona fide pathogen that causes several forms of severe periodontitis (Genco, 1996). P. gingivalis can adhere to and enter host cells including epithelial and endothelial cells, which are considered to be mediated by the interaction between bacterial fimbriae and α5β1 integrin (Amano, 2003). Extracellularly, P. gingivalis releases membrane vesicles (MVs), which retain the full components of the outer membrane constituents, including proteins, lipopolysaccharide, muramic acid, the capsule, and fimbriae (Grenier and Mayrand, 1987). We previously used polystyrene fluorescent beads conjugated with MVs (MV beads) as homogenous artificial intruders with epithelial cells to characterize the entry mechanism of P. gingivalis with respect to the host cellular endocytic machinery, cytoskeleton, and lipid rafts (Tsuda et al., 2005). The beads adhered to HeLa epithelial cells and were surrounded by α5β1 integrin, followed by engulfment of the beads through the actin filament and microtubule network, which was suggested to involve phosphatidylinositol 3-kinase (PI3K). We also found that methyl-β-cyclodextrin (MβCD), a cholesterol depletion agent, markedly inhibited entry of the beads into epithelial cells (Tsuda et al., 2005). Those results suggested that P. gingivalis uses both α5β1-integrin and lipid rafts as receptors to invade epithelial cells, thus exploiting the microtubule network.
Nevertheless, a number of questions remain unanswered regarding the entry process, and it is especially unclear whether α5β1-integrin operates in a manner similar to lipid rafts to internalize P. gingivalis or if they function independently of each other. The present study was performed to investigate these issues. Our results extend previous findings in regard to how P. gingivalis enters epithelial cells, and suggest that the entry process is initiated independently of lipid rafts by adhesion of P. gingivalis to α5β1 integrin, after which entry requires lipid raft components, such as cholesterol and sphingolipids, with the recruitment of Rho GTPase Rac1.
HeLa cells were grown in DMEM (Sigma-Aldrich, St Louis, MO) supplemented with 10% FBS (Invitrogen, Carlsbad, CA). GFP-GPI, a fusion construct containing the GFP-anchoring sequence to GFP (Kondoh et al., 1999), was provided by Dr. G. Kondoh (Kyoto University, Kyoto, Japan). GFP-GPI stably expressing HeLa cells were established using LipofectAMINE2000 reagent (Invitrogen) in culture medium containing Geneticin (G418, 1 mg/ml; Invitrogen) after transfection. A sphingolipid deficient CHO cell line (LY-B) and the complemented transformant (LY-B/cLCB) were maintained as described previously (Hanada et al., 1998). The CHO-K1 cell line M2S2, which stably expresses both CD55 and CD59, and its mutant 3B2A, deficient in GPI-anchor biosynthesis, were also cultured as described previously (Nakamura et al., 1997). The dominant negative Rac1 and Cdc42 vectors were provided by Drs. Y. Takai (Osaka University, Osaka, Japan) and H. Miki (University of Tokyo, Tokyo, Japan), respectively. P. gingivalis strain TDC60 was anaerobically grown and its MVs were prepared as described previously (Tsuda et al., 2005).
Alexa Fluor 365 and 580-conjugated sulfate-modified polystyrene beads (diameter; 1.0 μm) and Alexa Fluor 580-conjugated carboxylate-modified beads (diameter; 0.5 μm) were purchased from Molecular Probes Inc. (Eugene, OR). MV-coating for sulfate modified beads was performed as described previously. Coating for carboxy modified beads were performed according to the manufacturer’s protocol. Briefly, 1.35×109 of the beads were washed with PBS, resuspended in 50 μl of 50 mM MES containing 1 μg/ml of MVs, and incubated at room temperature (RT) for 15 mins. EDAC was then added to the mixture at a final concentration of 400 μg/ml and the pH was adjusted to 6.5 and further incubated for 2 hours. To quench the reaction, glycine was added at a final concentration of 0.1 M and the mixture was further incubated at RT for 30 mins. After washing with PBS, MV beads were suspended in 1 ml of 1% BSA/PBS.
HeLa cells at 8×104 or cells of the CHO lines at 1×105 were seeded onto coverslips the day before the assay. To deplete sphingolipids, LY-B, LY-B/cLCB, and wild-type CHO-K1 cells (3×104) were incubated as described previously (Hanada et al., 1998). Before the assay, the cells were washed twice with DMEM and incubated for 30 mins in the presence or absence of inhibitors. MV beads were then added to each well at 1.35×107 for 1-μm beads and 1.45×108 for 0.5-μm beads. After incubation at 37°C for 1 hour, the cells were transferred to wells at 4°C, followed by surface biotinylation, and analyzed using an Olympus FV1000 laser scanning confocal microscope (Olympus, Tokyo, Japan), as described previously (Tsuda et al., 2005). For surface biotinylation, the cells were incubated with membrane impermeable Sulfo-NHS-LC-Biotin (Pierce, Rockford, IL, USA), and then exposed to Alexa Fluor 488-conjugated streptavidin to fluorescently label the cell surfaces and extracellular MV beads. Extracellular magenta colored MV beads, that were attached to cell surfaces and accessible for green biotin-streptavidin labeling, were shown as white in the magenta/green merged images, while completely internalized beads free from biotinylation were detected in their original magenta color. To quantify the numbers of intra- and extracellular-beads, maximal projection images were made and counted manually in each field. At least 10 fields and ~200 cells were analyzed for each experiment, with 3 independent experiments performed. The results were normalized by the results of control experiments.
HeLa cells were washed twice with serum-free DMEM and further incubated with 10 mM MβCD (Sigma-Aldrich), 5 μg/ml of filipin III (Sigma-Aldrich), or 50 μg/ml of nystatin (Sigma-Aldrich) in DMEM at 37°C for 30 mins prior to the assay. All reagents were included in the medium throughout the entry experiments.
Both immunostaining and F-actin staining were performed as described previously. For immunostaining, monoclonal anti-integrin α5 antibody (BD Transduction Laboratories, Lexington, KY) and monoclonal anti-CD55 and CD59 antibodies kindly provided by Dr. T. Kinoshita (Osaka University, Osaka, Japan) were used. Alexa Fluor 568-phalloidin (Molecular Probes, Eugene, OR, USA) was used to stain F-actin. The cells were observed with an Olympus FV1000 laser scanning confocal microscope.
HeLa cells at 7×106 were seeded in 10-cm dishes the day before the assay. After washing with serum-free DMEM, cells were incubated with or without MβCD in DMEM at 37°C for 30 mins, then treated with 1.35×109 MV beads for 40 mins. The cells were washed with ice-cold PBS, scraped, collected, and lysed with 140 μl of cold TNE buffer [25 mM Tris-HCl (pH 7.5), 150 mM NaCl, 5 mM EDTA] containing 1% TritonX-100 and 4% protease inhibitor cocktail (Roche Diagnostics, Basel, Switzerland) for 30 mins on ice. Five hundred and sixty μl of 50% OptiPrep (Sigma-Aldrich) in TNE with 1% TritonX-100 was added to the lysate and it was layered on the bottom of an ultracentrifuge tube to obtain a final concentration of 40%. One ml of 30% OptiPrep in TNE with 1% TritonX-100 was then overlaid, and a final layer of TNE (0.3 ml) was added. The samples were centrifuged at 200,000×g for 2 hours at 4°C and the fractions were collected from top, analyzed by immunoblotting.
Immunoblotting was performed as described previously (Kato et al., 2007). Specific antibodies against Rac1 (Upstate Biotechnology, Lake Placid, NY), Cdc42 (BD Bioscience, San Diego, CA), flotillin (BD Bioscience), α5 integrin (BD Bioscience), and the transferrin receptor (TfR; Zymed Laboratories, San Francisco, CA) were used.
We examined the involvement of raft cholesterol in the interaction of MV beads with epithelial cells using cholesterol sequestration agents (filipin and nystatin) and a cholesterol depletion agent (MβCD). All of the tested raft disrupting reagents significantly prevented entry of the beads into epithelial cells, of which MβCD showed a dramatic inhibition of 98% (Fig. 1A and B). Extraction of membrane cholesterol by MβCD may cause more severe destruction of the lipid raft than the other two reagents. In contrast, for bead adhesion, no inhibitory effects of filipin or nystatin were observed, while that of MβCD was statistically significant, though it was not as notable as the effect on bead entry. These results suggest that raft cholesterol is required for the entry of MV beads, but not their adhesion. To further elucidate the entry mechanism, we analyzed the requirement of sphingolipids, which are enriched in lipid rafts. We used the LY-B cell line, which is defective in sphingolipids biosynthesis, as well as its complemented cell line LY-B/cLCB. The number of magenta colored internalized beads was decreased in the LY-B cells, as compared with the wild-type CHO-K1 cells (Fig. 1B), while the complementation rescued the efficiency of bead entry. These results were quantified and are shown in Fig. 1C. Although the deficiency of sphingolipids significantly inhibited bead entry, bead adhesion was negligibly suppressed. Since both cholesterol and sphingolipids are major components of lipid rafts, these results indicate that lipid rafts play an indispensable role for entry, but not adhesion.
 View Details | Fig. 1. Cholesterol- and sphingolipid-dependent entry of MV beads. (A) Following pretreatment with cholesterol sequestration or depletion agents for 30 mins, HeLa cells were incubated with MV beads (magenta) for 1 hour. The cells were then subjected to biotinylation with Sulfo-NHS-LC-Biotin and treated with Alexa Fluor 488-conjugated streptavidin (green), followed by fixation with paraformaldehyde. By this treatment, the extracellular MV beads on the cell surfaces were also labeled with green color. As a result, the green-magenta merged images of the beads were seen as white ones. Internalized beads were detected with their original magenta color. Adhesion and entry were analyzed using confocal microscopy. (B) The numbers of beads attached to the cell surface and magenta colored internalized beads were determined per cell, with the results presented as a percentage of the control cells not treated with a reagent. At least 10 fields were analyzed and all assays were performed independently on 3 separate occasions. Values are shown as the mean±SE. *P<0.05. (C) Wild-type CHO-K1 cells, and sphingolipid deficient LY-B cells and complementary LY-B/cLCB cells were incubated with MV beads using a method similar to that reported in panel (A). Confocal optical sections with a thickness of 0.18 μm were taken at 0.5 μm intervals. A projection image (x-y plane) is shown. Magenta colored images indicate internalized beads, whereas white colored images show non-internalized beads adhered to the cell surface. (D) The numbers of internalized and adhered beads were determined, and are presented as percentages of wild-type cells. Values are shown as the mean±SE. *P<0.05. |
In our previous study (Tsuda et al., 2005), we found that cellular interaction with MV beads activated cellular actin polymerization and induced formation of actin rich phagocytic cups to engulf the beads. In addition, recruitment of raft markers, ganglioside GM1, and cholera toxin subunit B (CTxB) was also observed around the beads during entry. In this study, we determined which of the lipid rafts and α5β1-integrin functions initiated actin remodeling to form phagocytic cups. To investigate this issue, we used HeLa cells expressing another raft marker GFP-GPI. When GFP-GPI expressing cells were incubated with the MV beads, both GFP-GPI and α5β1-integrin were clearly recruited to surround the beads during entry (Fig. 2A). MβCD treatment abolished the recruitment of GFP-GPI, whereas that of α5β1-integrin around the beads was not inhibited (Fig. 2B). The effect of MβCD treatment on F-actin recruitment was also evaluated. As shown in Fig. 3A, actin rich phagocytic cups were formed to surround internalized beads in the control cells, while MβCD treatment strikingly abolished cellular actin remodeling to form phagocytic cups (Fig. 3B). These results demonstrated that lipid rafts were not involved in the association of α5β1-integrin with the beads, but rather played an essential role in actin mediated uptake of the beads.
 View Details | Fig. 2. Recruitment of α5β1-integrin is not abolished in MβCD-treated HeLa cells. HeLa cells stably expressing GFP-GPI were treated with 10 mM MβCD, then incubated with MV beads for 60 mins. Following incubation, the cells were washed, fixed and stained with an anti-α5β1 integrin antibody. (A) Control HeLa cells (non-treatment). (B) MβCD-treated HeLa cells. Recruitment of GFP-GPI (green) and α5β1-integrin (magenta) around the beads (cyan) is indicated in the boxed areas. Insets show enlarged images of the boxed areas. |
 View Details | Fig. 3. Recruitment of F-actin is abolished in MβCD-treated HeLa cells. HeLa cells expressing GFP-GPI were incubated with MV beads as described in Fig. 2, then stained with Alexa Fluor 568-conjugated phalloidin. Recruitment of GFP-GPI (green) and F-actin (magenta) around the beads is indicated in the boxed areas. Insets show enlarged images of the boxed areas. |
We observed co-localization of GFP-GPI and MV beads (Fig. 2 and 3). Likewise, endogenous GPI-APs, CD55 and CD59, were also detected around the beads (data not shown). GPI-APs reportedly function as cell surface receptors to mediate the entry of several bacteria and toxins into host cells (Grassme et al., 2003). Therefore, we examined if GPI-APs are obligatory in the case of P. gingivalis. For this experiment, we used a CHO cell line deficient in GPI-anchor biosynthesis, due to a mutation in PIG-L (Nakamura et al., 1997). As shown in Fig. 4A, CD59 and CD55 clearly disappeared from the plasma membrane in the mutant cells (M2S2), as compared to the wild-type (3B2A). Both cell lines were subjected to a bead entry assay and the number of internalized magenta colored beads was determined. Unexpectedly, bead entry was not abolished in M2S2 cells (Fig. 4B and C). These results indicate that GPI-APs were not necessary for entry of the beads.
 View Details | Fig. 4. GPI-APs are not essential for MV beads entry. (A) Wild-type (3B2A) and GPI-AP deficient (M2S2) CHO cell lines were stained with antibodies against GPI-APs, CD59 (upper 4 panels), and CD55 (lower 4 panels). (B) 3B2A and M2S2 cells were subjected to bead entry assays using a method similar to that reported in Fig. 1B. (C) The number of internalized beads was determined. |
We previously showed that phosphatidylinositol 3' kinases (PI3Ks) are required for actin mediated entry of MV beads (Tsuda et al., 2005). In addition, PI3K activities were reported to be required to activate small Rho GTPases, Rac, and Cdc42 (Benard et al., 1999), which control actin polymerization during phagocytosis. Therefore, we examined the effects of dominant negative forms of Rac and Cdc42 on bead entry. As shown in Fig. 5, the beads adhered to both types of cells with a similar efficiency as seen with the wild type (data not shown). However, entry was prevented in cells expressing a dominant negative form of Rac, but not in the Cdc42-dominant negative cells (Fig. 5). These results indicated that Rac was involved in bead entry, whereas Cdc42 was not.
 View Details | Fig. 5. Entry of the MV beads is dependent on Rac but not Cdc42. HeLa cells were transiently transfected with myc-Rac (S17N) (A) or myc-Cdc42 (S17N) (B), then subjected to bead entry assays. The cells were permeabilized and labeled with anti-myc mAb (green) and DAPI (cyan). White lines indicate the perimeters of non-transfected cells. Projection images at the mid-height of the cells and vertical (z) optical sections (x-z and y-z planes) are shown. Biotin-streptavidin labeling was not performed in this assay, thus entry efficiency could be evaluated using a z-axis image. Arrows in (B) indicate internalized beads. Bars=10 μm. (C) The number of internalized beads was determined and is presented as a percentage of wild-type cells. Values are shown as the mean±SE. *P<0.05. |
Membrane cholesterol is known to be related to cell actin organization (Kwik et al., 2003) and targeting of active Rac in lipid micro-domains is regulated by integrin signaling (del Pozo et al., 2004). Since the present results suggested that Rac and lipid rafts were involved in actin mediated entry of the beads, we further investigated the effects of cholesterol depletion on the relationship between Rac and lipid rafts. In this experiment, we examined if Rac was recruited to lipid rafts during the bead entry event. After the bead entry assays with or without MβCD, lipid rafts were isolated from HeLa cells and subjected to Western blotting analysis, with TfR used as a marker of non-raft fractions and flotillin used as a raft fraction marker. As shown in Fig. 6, Rac was detected in the non-raft fractions of the non-treated cells. Upon incubation with the MV beads, translocation of Rac to the raft fractions was observed, which was abolished by MβCD treatment. On the other hand, translocation of Cdc42 was not induced under any of the tested conditions. These results suggest involvement of Rac in raft mediated engulfment of the beads. In addition, α5β1-integrin was not translocated to the raft fractions by addition of the beads, suggesting that α5β1-integrin is located in a non-raft domain and plays a role distinct from lipid rafts in the interaction of P. gingivalis with epithelial cells.
 View Details | Fig. 6. Translocation of Rac to lipid rafts is induced by MV beads. HeLa cells were treated with or without 10 mM MβCD, then processed for bead entry assays. Following incubation with MV beads for 40 mins, the cells were lysed and separated by density gradient ultracentrifugation. Fractions were collected from top to bottom and processed for Western blotting using the indicated antibodies. (A) Non-treated control cells. (B) Cells incubated with MV beads. (C) MβCD treated cells incubated with MV beads. |
The plasma membrane of mammalian cells primarily consists of sphingolipids, cholesterol, and other phospholipids, while the former two are believed to be organized into clusters in very small domains, which are termed lipid rafts (Simons and Ikonen, 1997). It is considered that lipid rafts are sites for both adhesion and entry of a number of bacterial species, as shown in several studies that used raft disrupting reagents such as MβCD (Grassme et al., 1997; Grassme et al., 2003; Kolesnick et al., 2000; Lafont et al., 2002; Shin et al., 2000; Stuart et al., 2003). The present findings indicated that both cholesterol and sphingolipids are specifically required for entry of MV beads, whereas their deficiency did not affect bead adhesion (Fig. 1). Another notable finding in the present study is that α5β1-integrin was recruited to the MV beads even when the lipid rafts were disrupted with MβCD and then stably resided in the non-raft membrane domain (Fig. 2). This suggests that P. gingivalis exploits α5β1-integrin for adhesion to the host cell surface, then uses lipid rafts for subsequent entry into the cells. To our knowledge, such a combination of cellular factors shared to mediate bacterial uptake has previously been reported only for Listeria monocytogenes (Seveau et al., 2004), a pathogen that expresses the invasion proteins internalin and InlB, which bind to the host receptors E-cadherin and hepatocyte growth factor receptor (HGF-R), respectively. That report found that internalin interacted with E-cadherin dependent on lipid rafts. In contrast, the interaction of InlB with HGF-R did not require membrane cholesterol, whereas downstream signaling leading to F-actin polymerization was lipid raft dependent, similar to the present findings. Therefore, P. gingivalis is the second known pathogen reported to use a unique combination to share the roles of host receptors.
We found that actin polymerization and phagocytic cup formation induced by the MV beads was suppressed by MβCD. Thus, cholesterol was strongly suggested to be indispensable for actin remodeling to engulf the beads as well as P. gingivalis. This result is supported by a recent finding that cholesterol depletion has an effect on actin organization (Kwik et al., 2003). For a greater understanding of the mechanisms underlying raft mediated entry of P. gingivalis, we attempted to identify the lipid raft components that are involved in bacterial uptake. Glycosphingolipids GM1 (Tsuda et al., 2005), caveolin-1 (Tsuda et al., 2005), and GPI-APs (this study) have been shown to be recruited to the bead entry site. Among these components, sphingolipids were reported to be necessary for cellular engulfment of the beads, which is likely due to the fact that they are basic components of functional rafts (Simons and Ikonen, 1997). As for caveolin-1, it was shown that siRNA mediated knockdown of caveolin-1 expression caused inhibition of P. gingivalis entry into endothelial cells (Tamai et al., 2005). The development of caveolae, which are formed by the polymerization of caveolins, leads to the clustering and invagination of a subset of lipid rafts containing sphingolipid and cholesterol (Parton and Simons, 2007), which is likely required to accommodate the invading pathogen. In contrast, in the present study GPI-APs were not necessary for bead entry. GPI-APs reportedly act as sorting signals for selective endocytosis via a dynamin/clathrin independent pinocytic pathway through a cholesterol sensitive Cdc42 based recruitment of the actin polymerization machinery, which has been termed as the GPI-AP enriched early endosomal compartments pathway (Chadda et al., 2007). In addition, we previously showed that dynamin was essential for bead entry (Tsuda et al., 2005), especially for formation of F-actin rich phagocytic cups (data not shown). Further, bead entry was shown to be dependent on Rac for engulfment in the present experiments. Thus, GPI-APs do not seem to be involved in P. gingivalis entry, which is an interesting characteristic that differs from a number of other bacterial species.
Rac, a member of the Rho family of master actin regulators, promotes recruitment to the site of bacterial entry (Wong and Isberg, 2005). In addition, α5β1-integrin, which interacts with bacteria, has been shown to stimulate Rac1 and/or Cdc42, which promote the invasion of host cells by several bacterial species (Agerer et al., 2005; Del Pozo and Schwartz, 2007; Ozeri et al., 2001; Plancon et al., 2003). In this study, the dominant negative form of Rac inhibited lipid raft mediated uptake of the MV beads, whereas Cdc 42 did not, which suggests that Rac is essential for bead entry. Further, the MV beads induced a translocation of Rac to the raft domain, which was inhibited by MβCD treatment. These results indicate that bead entry directly regulates the recruitment of Rac to lipid rafts and subsequently activates downstream signaling, which initiates cellular engulfment of bacteria. To our knowledge, such a regulatory mechanism has not been reported elsewhere. However, the interplay between integrins and lipid rafts is quite complex, and integrin is one of regulatory molecules of Rac activation (Wong and Isberg, 2005), thus it is necessary to elucidate the exact relationship and sharing roles of these cellular components in a future study, especially in order to examine whether integrins are also involved in the uptake of the beads.
In summary, our findings suggest that lipid raft integrity serves as a platform for Rac dependent actin remodeling to engulf P. gingivalis. They also indicate that the bacterium exploits α5β1-integrin for its adhesion to the host cell surface, then uses lipid rafts for subsequent entry into the cells.
We greatly appreciate the gift of CHO-K1 cell line M2S2 and its mutant 3B2A from Drs. Taroh Kinoshita and Yusuke Maeda (Research Institute for Microbial Diseases, Osaka University). Our work was also a part of the 21st Century COE program entitled “Origination of Frontier BioDentistry” held at Osaka University Graduate School of Dentistry, supported by the Ministry of Education, Culture, Sports, Science, and Technology of Japan.
|
Index