Ethics statement

Animal experiments were approved by the Seoul National University Hospital Institutional Animal Care and Use Committee (SNUH IACUC No.12–0331-C1A03) and performed in strict accordance with the recommendations in the National Guide Line for the care and use of laboratory animals. Ethical approval for this work was granted by the Institutional Review Boards of Seoul National University Hospital (IRB no. 0–1001–039–307).

Cell culture

HeLa cells (ATCC CCL-2, American Type Culture Collection), L929 cells (ATCC NCTC929), Vero cells (ATCC CCL-81), and ECV304, an endothelial cell-like cell line [37], were maintained in DMEM (Welgene, Daegu, Korea) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Welgene), 100 U/mL penicillin and 100 μg/mL streptomycin (Gibco BRL) at 37°C in 5% CO₂.

Preparation of O. tsutsugamushi

The Boryong, Karp, and Kato strains of O. tsutsugamushi were purified using a modified Percoll gradient purification method [32,38]. O. tsutsugamushi was propagated in L929 cells. At 3 to 4 days post-infection, infectivity was determined using an indirect immunofluorescence assay (see below). When an infection rate of > 90% was achieved, the cells were harvested by centrifugation at 6000 × g for 20 min. The cell pellet was resuspended with Tris-sucrose (TS) buffer (33 mM Tris-Cl (pH 7.4) and 0.25 M sucrose) and homogenized using 100 strokes of a Polytron homogenizer (Wheaton Inc., Millville, NJ) followed by centrifugation at 200 × g for 5 min. The supernatant was then mixed with 40% Percoll (Pharmacia Fine Chemicals, Uppsala, Sweden) in TS buffer and centrifuged at 25,000 × g for 60 min. The bacterial band was collected and centrifuged at 77,000 × g for 30 min. The bacterial pellet was washed 3 times in TS buffer, resuspended in DMEM and stored in liquid nitrogen until use [39]. The infectivity titer of the inoculum was determined as previously described [40], with minor modifications. Infected-cell-counting units (icu) were calculated as follows: [(total number of cells used for infection) × (percentage of infected cells) × (dilution of the O. tsutsugamushi suspension)]/100. For infection assays, 1.0 × 10⁷ icu of O. tsutsugamushi were used to infect cells cultured in 6-well plates containing 1.0 × 10⁶ of host cells.

Sequence analysis

Nucleotide sequences of scaA genes amplified from Gilliam, Karp, and Kato strains were deposited to GenBank under accession no. KM591910, KM591911, and KM591912, respectively. The scaA gene sequence from the Boryong strain was obtained from its genomic sequences (GenBank accession no. AM494475.1). Sequences of tsa56 genes from each strain (AM494475.1 for Boryong, AY956315.1 for Gilliam, AY836148.1 for Karp, and GU120147.1 for Kato strain) were also used for comparative analysis. Nucleotide sequence alignments for constructing phylogenetic trees were processed by Clustal W with maximum likelihood method implemented in MEGA6 software [41]. The similarity and identity of those nucleotides and amino acids was calculated through Matrix Global Alignment Tool (MatGAT) version 2.03 [42]. The aligned nucleotide sequences were evaluated in SimPlot version 3.5.1 with Kimura (2-parameter) and Empiric Ts/Tv ratio settings [43]. The aligned amino acid sequences were analyzed through the BLOSUM62-referenced 100 amino acid sliding window analysis. The output values were calculated from R-Project (http://www.r-project.org/). Line graphs were visualized by GraphPad Prism software version (Graph-Pad Software Inc., La Jolla, CA). Repeat sequences within a scaA gene were also analyzed using Tandem Repeat Finder software [44].

Cloning and expression of recombinant antigens

scaA, scaB, scaC, scaE, and tsa56 were amplified from the genomic DNA of O. tsutsugamushi Boryong strain by PCR using the primer pairs listed in S1 Table. The PCR products were cloned into pET-28a or pGEX4T-1 vector (Novagen, Gibbstown, NJ). Full length scaA genes were also amplified from the genomes of Boryong, Gilliam, Karp, and Kato strains for sequence comparison. All constructs were sequenced to confirm in-frame cloning. Recombinant Sca and TSA56 proteins were purified from E.coli BL21 (DE3) harboring a recombinant plasmid encoding each bacterial protein. Following induction with isopropyl β-D-thiogalactoside (IPTG) (0.1 mM, Duchefa, Zwijndrecht, Netherlands) at 16°C for 16 h, the proteins were purified using Ni-nitrilotriacetic acid His-resin (Qiagen, Calrsbad, CA) or glutathione-Sepharose 4B columns (GE Healthcare, Piscataway, NJ) according to manufacturer’s instructions. The purified proteins were dialyzed against phosphate-buffered saline (PBS) in an Aside-A-Lyzer Dialysis Cassette (Therrmo scientific, Rockford, IL) at 4°C for overnight. After dialysis, purified proteins were treated with endotoxin removal columns (Thermo scientific) and endotoxin contamination was determined using the QCL-1000 kit (Lonza, Bloemfontein, South Africa) according to manufacturer’s instructions. All protein contained less than < 0.05 EU/mg of endotoxin.

Antibodies and reagents

Both preimmune mouse serum and anti-Sca polyclonal mouse serum (produced from Balb/c mice immunized with purified Sca proteins; Cosmogenetech, Seoul, South Korea) were used for the experiments. Human sera were prepared from scrub typhus patients following institutional review board approval. Horseradish peroxidase (HRP)-conjugated anti-mouse or anti-human IgG secondary antibodies (Santa Cruz Biotech Inc., Santa Cruz, CA) were used for immunoblotting [32]. The Alexa Fluor 488- or Alexa Fluor 594-conjugated anti-mouse, and-human antibodies used in the immunofluorescence assays were purchased from Molecular Probes (Invitrogen). For the beadbinding assay, Fluoresbrite microparticles (1 μm; Polyscience Inc., Warrington, PA) containing rhodamine were conjugated to GST or GST-ScaA proteins by using a PolyLink protein coupling kit (Polyscience Inc.) in accordance with the manufacturer’s instructions.

Bead-binding assay

HeLa cells (2.4 × 10⁵ cells in a 24-well plate) were incubated with Fluoresbrite microparticles (416 μg/well) conjugated to GST or GST-Sca proteins for 1 h, washed extensively with PBS, and fixed with 4% paraformaldehyde for 15 min [32]. Cells were subsequently incubated with ToPro-3 (Molecular Probes) for nuclear staining and observed under a confocal microscope or analyzed using a FACScan (Becton Dickinson).

Cellular adhesion and invasion assays

Bacterial adhesion and invasion assays were performed as previously described [32]. Briefly, E. coli strains harboring a vector or pET28a encoding scaA gene were induced with IPTG and added to confluent monolayers of ECV304, HeLa, and Vero cells in serum-free media. Portions of the bacterium-containing media were plated to determine the number of CFU added to each host cell monolayer. Contact between bacteria and the mammalian cells was synchronized by centrifugation at 200 × g, and the preparations were incubated at 37°C for either 20 min or 60 min for the adherence and invasion assays, respectively. For the invasion assays, infected cells were washed extensively with PBS and incubated for 2 h with complete medium supplemented with 100 μg/ml of gentamicin to kill any extracellular bacteria [32]. For all E. coli assays, infected cells were washed extensively with PBS and the bacteria liberated by incubation with 0.1% Triton X-100 in sterile water. The lysate was then plated on LB agar to enumerate the cell-associated bacteria. The results were expressed as the percentages of bacteria recovered relative to the number of bacteria in the initial inoculum [32].

For antibody neutralization assays, HeLa cells or ECV304 cells were grown in a 24-well plate (2.4 × 10⁵ cells/well) and infected with O. tsutsugamushi or E. coli expressing ScaA in the presence of 1:100-diluted preimmune or anti-Sca polyclonal mouse serum. Association of E. coli with host cells were measured by CFU assays as mentioned above. To detect intracellular O. tsutsugamushi, infected cells were stained by differential immunofluorescence assay [38]. First, cells were washed three times with PBS, fixed with 4% paraformaldehyde, and incubated with an anti-TSA56 antibody, followed by Alexa Fluor 488-conjugated goat anti-mouse IgG to stain the cell-surface associated-bacteria. Next, cells were permeabilized in a 0.2% Triton X-100 solution for 15 min and incubated with scrub typhus patients’ sera for 1 h, followed by Alexa Fluor 633-conjugated goat anti-human IgG to stain intracellular bacteria. Cells were observed using an Olympus FV1000 laser confocal microscope (Olympus, Tokyo, Japan) and analyzed using the Fluoview software (Olympus).

Immunofluorescence microscopy

Immunofluorescence microscopy was used to visualize O. tsutsugamushi. HeLa cells infected with O. tsutsugamushi were washed with PBS and fixed with 4% paraformaldehyde incubated with pooled scrub typhus human serum or anti-ScaA immune serum for 1 h, followed by incubation with Alexa Fluor 488-conjugated goat anti-mouse IgG and Alexa Fluor 594-conjugated mouse anti-rabbit IgG (Invitrogen) [32]. In some experiments, recombinant E. coli was stained with preimmune mouse serum, anti-ScaA serum, or anti-E. coli serum, followed by incubation with Alexa Fluor 488-conjugated mouse anti-rabbit IgG (Invitrogen) [32]. Cells were examined under an Olympus FV1000 laser scanning confocal microscope (Olympus). Images of cell sections were analyzed and processed using the Olympus Fluoview software (Olympus).

ELISA

To determine the titer of antibodies specific to ScaA or TSA56 in the sera of immunized mice, immunoassay plates (96-well plates; Nunc, Rochester, NY) were coated with 100 μl of purified antigen at a concentration of 5μg/ml at 4°C overnight. The plates were then blocked for 2 h at room temperature with PBS containing 5% skim milk. 100 μl of serum samples serially diluted in 2-fold were incubated for 2 h at room temperature. After washing with PBS containing 0.05% Tween20 (PBST), horseradish peroxidase (HRP)-conjugated goat anti-mouse IgM, IgG₁, or IgG_2c (Santa Cruz Biotechnology, Santa Cruz, CA) was added and incubated for 2h at room temperature. Wells were washed with PBST and incubated with 3,3′,5,5′-tetramethylbenzidine (TMB) peroxidase substrate solution (KPL, Gaithersburg, MD) for 10 min. The reactions were stopped by addition of 1M phosphoric acid solution. Absorbances were measured at 450 nm using a microplate reader (Beckman Coulter Inc., Fullerton, CA).

Immunization of mice and challenges

For immunization experiments, 6- to 8-week-old female C57BL/6 mice (Orient Bio Inc., Seongnam, South Korea) were used. Groups (n = 5) of mice were immunized subcutaneously in the hind leg three times at two weeks interval. 20 μg of purified ScaA, ScaC, or TSA56 proteins in PBS emulsified 1:1 with 2% alhydrogel adjuvant (Invitrogen) was used for each immunization. Blood was collected from immunized mice after one week after each immunization to determine serum antibody titer. One week after the final immunization, mice were challenged intraperitoneally with 10 × or 100 × LD₅₀ of O. tsutsugamushi strains. Body weight and mice survival was monitored for one month after bacterial challenge.

Statistical analysis

The data was analyzed using the Graph Pad Prism 5.01 software. Statistical analysis of all the experimental data except survival rate was performed using the two-tailed Student’s t-test with 95% confidence interval. Data are expressed as the mean ± standard deviation. Statistical analysis on survival rates were performed using the Mantel-Cox Log Rank test. A p-value of < 0.05 was considered statistically significant.

ScaA is expressed in O. tsutsugamushi

In order to examine whether the ScaA is expressed in O. tsutsugamushi, we generated a polyclonal anti-ScaA antiserum by immunizing mice with purified ScaA passenger domain (amino acids 30 to 1000). The specificity of this antiserum was confirmed by ELISA and immunoblot analysis using the recombinant Sca proteins (S1 Fig). To identify endogenous ScaA protein in O. tsutsugamushi, the anti-ScaA serum was reacted with the cell lysates of O. tsutsugamushi-infected cells. Immunoblot analysis showed that a ~ 150 kDa protein was recognized by the anti-ScaA serum in infected cells but not in uninfected control (Fig. 1A). The full-length ScaA protein was predicted to have a mass of 156 kDa. Anti-ScaA serum was also weakly reacted with a few bands lower than ~ 150 kDa, suggesting a cross-reactive antigens or fragmented ScaA protein in infected cells. The TSA56 protein, a major outer membrane protein of O. tsutsugamushi, was used as a positive control [32]. To further confirm the specificity of the anti-ScaA antiserum for O. tsutsugamushi, intracellular bacteria were stained using the pooled sera of scrub typhus patients together with anti-ScaA serum or preimmune mouse serum. As shown in Fig. 1B, anti-ScaA serum readily detected the bacteria within the host cells, whereas the preimmune serum did not. In addition, we found that the ScaA proteins were located on the periphery of bacterial cells (Fig. 1B. lower panels, inset boxes). Taken together, these results confirm that the scaA gene is actively translated in O. tsutsugamushi within eukaryotic host cells and that the protein might be expressed on the outer membrane of the bacteria.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Expression of ScaA by O. tsutsugamushi.

(A) Immunoblot analysis of whole proteins from L929 cells infected with O. tsutsugamushi proteins by using anti-ScaA serum (right panel). Anti-ScaA serum detected a protein with a molecular mass of approximately 150 kDa. Immunoblotting using anti-TAS56 was performed as a control (left panel). (B) Immunofluorescence confocal microscopy using preimmune serum or anti-ScaA serum (α-ScaA) showed ScaA in the O. tsutsugamushi-infected L929 cells. The left panels show bacteria stained with the pooled sera of scrub typhus patients (α-OT). Magnified images are shown in the lower panels (inset boxes). Scale bars, 5 μm.

https://doi.org/10.1371/journal.pntd.0003585.g001

ScaA mediates bacterial adhesion to host cells

Recently, several studies reported that rickettsial Sca proteins mediate bacterial adherence to and/or invasion into mammalian host cells [32,45]. Therefore, we examined whether the O. tsutsugamushi ScaA protein could function as a virulence factor for bacterial adhesion and/or invasion. First, we performed a bead-binding assay using fluorescent microbeads (1 μm in diameter) covalently conjugated to either purified GST or GST-ScaA. Incubation of HeLa cells with GST-ScaA-conjugated beads resulted in marked binding to the host cells, even after extensive washing. The control beads linked to GST alone interacted only weakly with the HeLa cells (Fig. 2A) [32]. The interaction of the fluorescent beads with the host cells was quantified using flow cytometry (Fig. 2B). After fixation, the mean fluorescence intensity (MFI) of the HeLa cells incubated with beads conjugated to GST-ScaA dramatically increased (MFI = 50.1) compared to that of cells incubated with beads conjugated to GST (MFI = 13.6) or that of untreated cells [32].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Adhesion function of ScaA.

(A) HeLa cells were incubated with fluorescent microbeads coated with GST or GST-ScaA (ScaA) for 1 h, washed extensively, and fixed. Cell-bound microbeads (green) were visualized by fluorescence microscopy after staining of cell nuclei (blue). Scale bars, 10 μm. (B) Relative binding of the microbeads coated with GST (dotted line) or GST-ScaA (thick line) to HeLa cells was quantified directly using fluorescence-activated cell sorter (FACS) analysis. The gray histogram represents unbound cells (cells not incubated with microbeads). (C) Immunofluorescence microscopy using an anti-ScaA antibody revealed the presence of ScaA on the surface of the recombinant E. coli (lower panels). Preimmune serum did not detect the recombinant protein (upper panels). Scale bars, 5 μm. (D) E. coli transformed with the pET28a vector or with pScaA was induced with IPTG and incubated with HeLa cells. After being washed to remove adherent bacteria, the cells were fixed, permeabilized, and stained with an anti-E. coli antibody (green) and ToPro-3 for nuclear staining (blue). Scale bars, 10 μm. (E) CFU-based quantification of adherent E. coli transformed with the vector or pScaA was performed. The results are presented as percentages of adherent bacteria relative to the total bacterial input. Data are representative of three independent assays for each of the host cells. **, p < 0.01. (F) Inclusion of anti-ScaA serum in the medium (α-ScaA) significantly inhibited adhesion of E. coli expressing ScaA into host cells. After addition of anti-ScaA or preimmune serum into infection media, CFU-based quantification of adherent E. coli transformed with the vector or pScaA was performed. **, p < 0.01.

https://doi.org/10.1371/journal.pntd.0003585.g002

To further verify the role of the scaA gene in bacterial adherence to host cells, we utilized a heterologous E. coli expression system [32]. The entire O. tsutsugamushi scaA open reading frame was cloned into the IPTG-inducible expression vector, pET-28a, to yield the plasmid pScaA as previously described [32]. ScaA was expressed in the E. coli strain BL21(DE3) and analyzed using anti-ScaA serum and confocal microscopy after fixation with 4% paraformaldehyde. As shown in Fig. 2C, ScaA was readily detectable on the surface of all the recombinant E. coli cells by anti-ScaA serum (lower panels) but not on bacteria harboring empty vector (upper panels). We next examined the ability of ScaA-expressing E. coli to adhere to monolayers of nonphagocytic host cells [32]. Epithelial (HeLa and Vero) and endothelial (ECV304) cells were incubated with recombinant E. coli harboring an empty vector or pScaA. The cells were then washed extensively to remove non-adherent bacteria, fixed, and analyzed under a confocal microscope after staining with an anti-E. coli antibody and the nuclear stain ToPro-3. Immunofluorescence analysis revealed that ScaA expression resulted in an increase in the number of adherent E. coli bacteria (Fig. 2D). This ScaA-mediated enhanced adhesion was verified by removing the adherent bacteria from the live host cells and counting them using a CFU-based quantification assay [32]. The assay confirmed that ScaA expression significantly increased bacterial adherence to HeLa cells (Fig. 2E) and other cell lines (S2A Fig). Therefore, the expression of O. tsutsugamushi ScaA on the outer surface of E. coli enhances bacterial adherence to nonphagocytic host cells. We also examined whether our anti-ScaA antibody could neutralize the adhesion of bacteria expressing ScaA to host cells. Treatment of the recombinant bacteria with the anti-ScaA antibody for 1 h before adding it to host cells reduced bacterial adhesion by approximately four fold (Fig. 2F) when compared to treatment with nonimmune serum, indicating that the anti-SacA antibody can block ScaA-mediated bacterial adhesion to host cells.

ScaA vaccination provides protective immunity against O. tsutsugmaushi infection

In order to confirm the neutralizing effect of anti-ScaA antibody on O. tsutsugamushi infection, HeLa cells were infected with the pathogen in the presence of various anti-Sca antibodies or nonimmune serum. At 4 h after infection, bacterial infection was examined by confocal microscopy after differential immunoflourescent staining and the O. tsutsugamushi/host cell ratio was determined (Fig. 3A and 3B). Presence of anti-ScaA antibody in the infection media significantly inhibited O. tsutsugamushi infection of host cells. The number of bacteria per cell was reduced by approximately 50% compared with the control group treated with nonimmune serum, whereas other anti-Sca antibodies failed to significantly inhibit bacterial infection, indicating that the anti-ScaA antibody could provide specific protective effect against bacterial infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Protective role of anti-ScaA immunity.

(A) Anti-ScaA antibody inhibited O. tsutsugamushi infection into host cells. HeLa cells were infected with the pathogen in the presence of the indicated anti-Sca antibodies or nonimmune serum. At 4 h after infection, bacterial infection was examined using confocal microscopy after differential immunoflourescent staining (see materials and methods). (B) The O. tsutsugamushi per host cell ratio was determined from three independent experiments in (A). **, p < 0.01. (C) Survival curves of immunized mice following lethal challenge with O. tsutsugamushi. Mice (n = 5/group) were immunized with the indicated antigen from the Boryong strain and challenged intraperitoneally with 100 x LD₅₀ of O. tsutsugamushi Boryong strain. Their survival was monitored until all the surviving mice recovered from the disease. This graph is a representative survival curve of two experiments. **, p < 0.01 when compared with non-immunized group (PBS).

https://doi.org/10.1371/journal.pntd.0003585.g003

This protective effect of the anti-ScaA antibody against O. tsutsugamushi infection was further investigated in vivo by challenging mice with 100 × LD₅₀ of O. tsutsugamushi at 7 d after immunization. All the vaccine antigens were derived from the O. tsutsugamushi Boryong strain and mice were challenged with the same bacterial strain. As shown in Fig. 3C, a significant level of protection against the homologous strain was observed in the ScaA-immunized group as well as in the TSA56-immunized mice. In contrast, ScaC immunization did not provide any significant protection. Therefore, vaccination with a ScaA antigen could provide protective immunity against homologous strain infection as efficiently as TSA56, a dominant membrane antigen of O. tsutsugamushi.

Next, we tested whether the candidate bacterial antigens could provide protective immunity against heterologous strain infection. Each group of mice were immunized with the indicated antigens derived from the O. tsutsugamushi Boryong strain and then challenged with a low (10 × LD₅₀) or high (100 × LD₅₀) dose of Boryong, Karp, or Kato strains (Fig. 4 and S3 Fig). We confirmed significant increases of both type 1 (IgG_2C) [46] and type 2 (IgG₁) antibodies against ScaA and/or TSA56 after third immunization (S4 Fig). Following infection with O. tsutsugamushi, mock-immunized mice began to lose body weight between 8–12 d after inoculation, depending on the bacterial doses and strains challenged, and lost 10–25% of body weight before they expired (S5 Fig). All the unimmunized mice had expired by 10–17 d after infection. The immunized mice survived after infection with O. tsutsugamushi maintained normal body weight during the experiment, but the expired ones rapidly lost their body weight from 4–8 d before death. All the mice immunized with ScaA or TSA56 were protected from the homologous Boryong strain regardless of infection dose, and were also protected against low dose (10 × LD₅₀) Karp strain infection. When mice were challenged with high dose (100 × LD₅₀) Karp strain infection, all the mice immunized with both ScaA and TSA56 survived and 80% of ScaA-immunized mice were protected. Although TSA56 immunization also provide significant protection (40% survival, p value = 0.017) compared to mock-immunized control group, the protection level was significantly (p value = 0.049) lower than that afforded by vaccination with both ScaA and TSA56 antigens. In the groups challenged with low dose Kato strain, groups immunized with both ScaA and TSA56 showed the best protective effect (60% survival) and TSA56 immunization provided only 20% survival. Immunization with ScaA also provided significant protection (40% survival). Although the level of protection afforded by ScaA (median survival = 22 d) was higher than that of TSA56 (median survival = 19 d), the difference was not statistically significant (p > 0.05). In contrast, ScaA vaccination (median survival = 18 d) significantly prolonged the survival of mice compared to TSA56 and mock-immunization (p < 0.01, median survival = 15 d in both groups) when mice were challenged with high dose of Kato strain. Immunization of TSA56 together with ScaA provided a similar level of protection as that observed in the ScaA immunization group even though all the challenged mice ultimately succumbed to pathogen infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Protective role of ScaA or combined immunization against heterologous strain infection.

Mice (n = 5/group) were immunized with the indicated antigens and challenged intraperitoneally with 10 x LD₅₀ (A) or 100 x LD₅₀ (B) of O. tsutsugamushi. Mice were immunized with antigens from the Boryong strain and challenged with the indicated strains (BR: Boryong, KP: Karp, KT: Kato). p value and median survival are summarized in S3 Fig. *, p < 0.05; **, p < 0.01.

https://doi.org/10.1371/journal.pntd.0003585.g004