After transformation of the linearized plasmid p?EpsC to P. gingivalis W83 the epsC insertional mutation was confirmed by specific PCR amplifications and agarose gel electrophoresis of the products (data not shown). Primer combinations epsC BamHI F ? PG0119 R and EryF F ? epsC EcoRI R (Table 1) ensured that a 1.2 Kb fragment of p?EpsC had been integrated by double crossover at PG0120 (epsC) as expected, replacing the intact copy with the insertionally inactivated copy (Figure 1).

To examine if the mutation had an influence on the growth characteristics of the epsC mutant both W83 and the epsC mutant were grown in brain heart infusion broth supplemented with hemin (5 ?g/ml) and menadione (1 ?g/ml) (BHI+H/M). Phase-contrast microscopy revealed that the mutant grows in aggregates, but no difference in growth rate was observed.

The potential polar effect of the insertional inactivation on the down stream gene of epsC named hup-1 was examined. Total RNA was extracted from W83 and the epsC mutant in the early exponential phase and the hup-1 expression levels were evaluated by Real-Time PCR. No significant difference in expression of hup-1 was found between W83 and the epsC mutant (data not shown).

To show the effect of capsule-loss on the surface structure of P. gingivalis the hydrophobicity of the epsC mutant was tested by the capacity to adhere to hexadecane. While 3% of W83 cells was shown to adhere to hexadecane more than 60% of the epsC mutant cells was adhered to hexadecane. 19% of the complemented mutant cells was adhered to hexadecane (see Additional file 1).

Reactivity with the CPS-specific polyclonal rabbit antisera against P. gingivalis serotypes K1-K6 [⁸,⁹] was examined for W83 and the epsC mutant. The epsC mutant was not recognized by any of the antisera including the K1 antiserum, whereas the wild type strain was only recognized by the K1 antiserum (Figure 2). Differences in CPS characteristics were also studied by Percoll density gradient centrifugation, which can reveal density differences between encapsulated and non-encapsulted bacteroides strains [²⁴]. Percoll density gradient centrifugation analyses of W83 and the epsC mutant showed that the density of the mutant had been changed (Figure 3). Where W83 mostly settled at the 20-30% interface, the epsC mutant settled at the 50-60% interface. Note that the appearance of W83 is diffuse and not restricted to the 20-30% interface. The mutant settles as a compact and granulous layer.

To conclusively examine the absence of CPS in the epsC mutant, light microscopy was performed using India ink in combination with fuchsine staining (Figure 4). The negative India ink staining allows direct visualization of the capsule, appearing as a light halo surrounding the P. gingivalis cell. Fuchsine is used to stain the cell body. The halos around the W83 wild type strain are clearly visible in the phase contrast microscopic picture, whereas halos are absent around the epsC mutant. The intact epsC gene in trans under control of the CP25 promoter rescues the wild-type phenotype enabling the complemented mutant to produce a K1 capsule again (Figures 2 and 4).

To study the effect of the epsC deletion on the host immune response six hour infection studies of human gingival fibroblasts with W83 and the epsC mutant were performed. Figure 5 shows IL-1?, IL-6 and IL-8 expression of infected gingival fibroblasts relative to the non-infected negative control which is set to 1 and normalized against expression of housekeeping gene GAPDH.

At multiplicity of infection (MOI) 1000:1 of both strains a small induction of the tested genes could be detected compared to the non-infected control, but significant induction for all three genes was found when MOI 10.000:1 was used for infection.At MOI 1000:1 IL-6 and IL-8 expression showed a significantly higher induction (150-fold and 37-fold induction respectively) in the cells challenged with the epsC mutant when compared to the wild-type (6-fold and 2-fold induction respectively), IL-1? did not show a difference compared to the wild-type. However, when gingival fibroblasts were challenged with MOI 10.000 bacteria all three tested genes showed a significantly higher induction in the cells challenged with the epsC mutant than with W83 (figure 5). When fibroblasts were challenged with the complemented mutant the response was almost completely restored to wild-type levels (see Additional file 2).

Sedimentation of the epsC mutant in comparison to W83 was analyzed in the same buffer as used in the infection experiments. No significant sedimentation differences were found between W83 and the epsC mutant within the 6 hours needed for infection of the fibroblasts (data not shown).

Since infections were done with viable P. gingivalis, survival of the bacteria during the 6-hour aerobic period of infection in DMEM medium had to be ensured. Therefore a 6-hour survival experiment was performed in the 24-well plates used for the fibroblast challenge. On average 60-75% of W83, epsC mutant and complemented mutant cells survived for 6 hours in Dulbecco's modified Eagle's Medium (DMEM; Sigma Chemical Co.) supplemented with 10% fetal calf serum (FCS) (see Additional file 3).

P. gingivalis strains were grown either on 5% horse blood agar plates (Oxoid no. 2, Basingstoke, UK) supplemented with hemin (5 ?g/ml) and menadione (1 ?g/ml) (BA+H/M plates) or BHI+H/M, both, at 37?C in an anaerobic atmosphere of 80% N₂, 10% H₂, and 10% CO₂. Mutants were selected in the presence of 5 ?g/ml erythromycin. Complemented mutants were selected in the presence of 50 ?g/ml gentamycin and 1 ?g/ml tetracycline. Purity of P. gingivalis liquid and plate-grown cultures was routinely checked by gram staining and microscopic examination.

Escherichia coli DH5? was used for maintenance and construction of plasmids. DH5? was cultured in Luria-Bertani (LB) broth or on solid medium (LB broth with addition of 1.5% agar). Ampicillin (Na⁺ salt; 100 ?g/ml) was added to the growth media to select for pUC-derived plasmids. E. coli S17-1 grown on LB supplemented with 5 ?g/ml tetracycline carrying the complementation construct pT-PG0120 was used for conjugation with P. gingivalis.

The gingival fibroblasts (HGF1 and HGF2) used in this study were collected from extracted third molars of two periodontally healthy subjects with a high pro-inflammatory immunological response when challenged with P. gingivalis [²⁰]. Donors had given written informed consent, and the study was approved by the VUmc Medical Ethical committee.

Genomic DNA from P. gingivalis strains was isolated from plate-grown bacteria using the DNeasy tissue kit (Qiagen Benelux BV). The DNA concentration of all samples after purification was between 20 ng/?l and 60 ng/?l.

To make an insertional knockout of epsC in the W83 wild type strain we constructed plasmid p?EpsC. Primers epsC BamHI-F and epsC EcoRI-R (see table 1 for details) were used to amplify the 1.2 Kb epsC gene from P. gingivalis W83 genomic DNA in a PCR reaction. Pfu polymerase (Fermentas, GmbH, St. Leon-Rot, Germany) was used according to the manufacturer's protocol with 100 ng of genomic DNA. The PCR program started with 95?C for 5 min and then 25 cycles of 95?C, 55?C and 72?C for 30 s, 30 s and 2.5 min respectively and was ended by one step of 72?C for 5 min. The amplified fragment was cleaned using the Qiagen PCR purification kit (Qiagen Benelux B.V.) and restricted with BamHI and EcoRI. This restricted epsC gene fragment was ligated into BamHI-EcoRI restricted pGEX-6p-3 plasmid to yield pGEX-PG0120.

The 1.2 Kb EryF erythromycin resistance cassettes for use in P. gingivalis was amplified from plasmid pEP4351 using primers EryF ClaI F and EryF ClaI R. and after restriction with ClaI this fragment was ligated into the ClaI-restricted pGEX-PG0120 plasmid yielding p?EpsC. The ScaI-linearized p?EpsC plasmid was used for insertional inactivation of epsC in P. gingivalis strain W83.

The 120 bp artificial constitutive CP25 promoter [³⁷] was amplified from plasmid pDM15 [³⁸] using primers CP25 ClaI F and CP25 AscI R. The intact epsC 1.2 Kb gene was amplified from genomic DNA of P. gingivalis strain W83 using primers epsC AscI F and epsC SpeI R. After ligation of these fragments into cloning vector pJET1.2 (Fermentas, GmbH, St. Leon-Rot, Germany) the constructed expression cassette was cut out with XhoI and HindIII and ligated into the SalI and HindIII digested pT-COW shuttle plasmid [³⁹] to yield the complementation construct pT-PG0120.

BHI+H/M was inoculated with P. gingivalis W83 from a 6-day-old blood agar plate. This pre-culture was anaerobically incubated at 37?C for 2 days. 2 ml of the pre-culture was used to inoculate a 100 ml culture. The next day this culture was used to inoculate 2 ? 100 ml of fresh BHI+H/M to an OD₆₉₀ of 0.2. After six hours of anaerobic incubation at 37?C the cells were harvested by centrifugation in mid-exponential phase. The pellet was washed two times in 20 ml EPB (10% glycerol, 1 mM MgCl₂) and after that resuspended in 2 ml of EPB. Aliquots of 200 ?l were stored at -80?C and used for electroporation.

200 ng of PstI digested p?EpsC was added to 200 ?l of W83 P. gingivalis cells. The mixture was transferred to a 2 mm electroporation cuvette and electroporated using an Electro Cell Manipulator 600 (BTX Instrument Division, Holliston, MA, USA; 25 ?F, 2.5 kV, 186 ?). 1 ml of BHI+H/M was added immediately after the pulse. The cells were left for recovery anaerobically at 37?C for 18 hours. The suspension was plated on BA+H/M plates with 5 ?g/ml erythromycin for selection of the transformants. The authenticity of the insertional knockout epsC mutants was verified using primer combinations epsC BamHI F ? PG0119 R and EryF ClaI F ? epsC EcoRI R. Furthermore, using Real-Time PCR, the expression of the downstream gene hup-1 in both W83 and the epsC mutant was monitored using primers hup-1 F and hup-1 R to exclude polar effects. W83 and the epsC mutant were grown till early exponential phase. The cell pellets were collected by centrifugation and resuspended in RLT buffer (Qiagen, Benelux B. V.). The cells were disrupted using a Fast Prep Cell Disrupter (Bio 101, Thermo electron corporation, Milford, USA) and centrifuged, the total RNA was extracted from the supernatant according to the manufacturer's protocol of Qiagen RNeasy^? mini kit (Qiagen Benelux B.V.). The residual contaminating genomic DNA was removed by Turbo DNA-free? kit (Ambion, Austin, USA). mRNA was then reverse transcribed using the Fermentas first-strand cDNA synthesis kit (Fermentas GmbH, St. Leon-Rot, Germany) according to the manufacturer's protocol.

The synthesized cDNA was further analyzed using Real-Time PCR with gene-specific primers on an ABI Prism 7000 Sequence Detecting System (Applied Biosystems, Nieuwerkerk a/d lJssel, The Netherlands). Gene expression was normalized to the expression of glucokinase (glk), amplified with primers glk F and glk R [⁴⁰]. The relative hup-1 expression levels of W83 from three independent experiments were compared in duplicate to those of the epsC mutant.

To complement the epsC mutant, plasmid pT-PG0120 was transferred into the mutant by conjugation following a protocol described earlier [⁴¹], with slight modifications. For selection of P. gingivalis after the over-night conjugation we used 50 ?g/ml of gentamycin in our blood agar plates instead of 150 ?g/ml. Integrity of the trans-conjugants was confirmed by colony PCR and plasmid isolation combined with restriction analysis using a plasmid isolation kit (Qiagen Benelux B.V.).

Percoll density gradients were in principle prepared as described by Patrick and Reid [²⁴]. In short, a 9:1 stock solution of Percoll (Pharmacia, Biotech AB, Uppsala, Sweden) was prepared with 1.5 M NaCl. Solutions containing 80, 70, 60, 50, 40, 30, 20 and 10% Percoll in 0.15 NaCl were prepared from the stock. In an open top 14 ml polycarbonate tube (Kontron instruments, Milan, Italy) 1.5 ml of each of the solutions was carefully layered on top of the previous starting with 80%. 1 ml of an anaerobically grown over night culture of wild type and the epsC mutant concentrated to an OD₆₉₀ of 4 in PBS was added to the top of the 10% layer and centrifuged for one hour at 8000 ? g at 20?C in a Centrikon TST 41.14 rotor (Kontron instruments, Milan, Italy) using a Centrikon T-1170 (Kontron instruments, Milan, Italy) centrifuge.

W83, the epsC mutant and the complemented mutant were grown 18 hours in BHI+H/M. The bacteria were washed twice in PBS after which the OD₆₀₀ was set to 0.5. After addition of 150 ?l n-hexadecane to 3 ml of this suspension the mix was vortexed 30 seconds, rested for 5 seconds and vortexed for 25 seconds. After exactly 10 minutes incubation at room temperature a sample was taken to measure the OD₆₀₀ of the aqueous phase. The percentage of bacteria adhered to hexadecane was calculated by the formula: (OD₆₀₀ before-OD₆₀₀ after)/OD₆₀₀ before ? 100%. The presented data in Additional figure 1 were collected from two experiments using triplicate measurements.

Serotyping of P. gingivalis was based on the detection of the six described K-antigens [⁸,⁹]. In short, serotype-specific, polyclonal antisera were obtained after immunization of rabbits with whole bacterial cells of the six P. gingivalis type strains [⁴²]. Bacterial antigens for double immunodiffusion tests were prepared as described previously [⁸]. Immunodiffusion was carried out in 1% agarose (Sigma Chemical Co., St. Louis, MO, type 1, low EEO) in 50 mM Tris-HCl buffer (pH 8.6). 10 ?l antiserum and 10 ?l of antigen were loaded and allowed to diffuse and precipitate for 48 hours at room temperature.

P. gingivalis cells were taken from 4 day-old plates and resuspended in 1 ml of PBS. On a glass slide 10 ?l of this suspension was mixed with 10 ?l of India ink (Talens, Apeldoorn, The Netherlands) and using another glass slide a thin film was made. The film was air-dried. A drop of 0.2% fuchsine was carefully added onto the film and removed after 2 minutes by decanting. Then the film was air-dried. Pictures were taken with a Leica DC500 camera on a Zeiss Axioskop using phase-contrast.

Pre-cultures of W83 and the epsC mutant were grown anaerobically for 18 hours in BHI+H/M at 37?C. The pre-cultures were diluted to an OD₆₉₀ of 0.05 in duplo in fresh BHI+H/M and incubated anaerobically at 37?C. Every few hours the OD₆₉₀ was measured and a sample was taken for cfu-counts.

W83 and the epsC mutant were grown anaerobically for 18 hours in BHI+H/M at 37?C. After 3 wash steps in phosphate buffered saline (PBS) the OD₆₉₀ was standardized to 5 in DMEM with 10% FCS. 10 ml of this culture was added to 40 ml DMEM with 10% FCS in a 100 ml flask to set the OD₆₉₀ to 1. The cultures were incubated standing still at 37?C for six hours. At regular time intervals, a 200 ?l sample was taken 0.5 cm from the liquid surface and the decrease of the OD₆₉₀ values was determined as a measure for sedimentation.

W83, the epsC mutant and the complemented mutant were grown anaerobically for 18 hours in BHI+H/M at 37?C. After 2 wash steps in phosphate buffered saline (PBS) the pellets were resuspended in DMEM with 10% FCS to an OD₆₉₀ of 0.05 as used in fibroblast infections at MOI 10.000:1. 500 ?l of these suspensions was incubated at 37?C in a humidified atmosphere of 5% CO₂ in air. Samples for cfu-counts were taken at t = 0 hours, t = 3 hours and t = 6 hours and dilutions were plated on BA+H/M plates.

Bacteria were grown overnight for 18 hours in BHI+H/M. The bacterial cells were washed three times in PBS and then used to infect gingival fibroblasts at MOIs of 1000:1 and 10.000:1 (bacteria cells: fibroblasts) in a total volume of 500 ?l DMEM with 10% FCS in 24-well plates. The plates were incubated for 6 hours at 37?C in a humidified atmosphere of 5% CO₂ in air. The cells were washed twice with cold PBS. Then 350 ?l lysis buffer (1% ?-mercapthanol in RLT buffer) was added to the cells according to the protocol of Qiagen RNeasy^? mini kit (Qiagen Benelux B.V.) after which the plate was stored at -80?C for later use.

mRNA was isolated from the gingival fibroblast lysates according to the manufacturer's protocol of Qiagen RNeasy^? mini kit (Qiagen Benelux B.V.). The mRNA concentrations of the samples were determined using the Nanodrop ND_1000 (Isogen Life Science). mRNA was reverse transcribed using the Fermentas first-strand cDNA synthesis kit (Fermentas GmbH, St. Leon-Rot, Germany) according to the manufacturer's protocol.

cDNA synthesized from mRNA isolated from gingival fibroblasts after infection with P. gingivalis was analyzed in quadruple using Real-Time PCR with gene-specific primers on a ABI Prism 7000 Sequence Detecting System (Applied Biosystems, Nieuwerkerk a/d lJssel, The Netherlands). Reactions were performed with 2 ng cDNA in a total volume of 8 ?l containing SYBR Green PCR Master Mix (Applied Biosystems) and 0.99 pM of each primer. After activation of the AmpliTaq Gold DNA polymerase for 10 minutes at 94?C, 40 cycles were run of a two step PCR consisting of a denaturation step at 95?C for 30 seconds and annealing and extension step at 60?C for 1 minute. Predicted product sizes were in the 100-200 bp range. Subsequently the PCR products were subjected to melting curve analysis to test if any unspecific PCR products were generated. The PCR reactions of the different amplicons had equal efficiencies. Samples were normalized for the expression of housekeeping gene GAPDH, which is not affected by the experimental conditions, by calculating the ? Ct (Ct _{housekeeping gene} - Ct _{gene of interest}) and expression of the different genes is expressed as 2^-(?Ct). Fold increase in gene expression (induction) was expressed by 2 ^-(??Ct), wherein ??Ct = ?Ct_challenged- average Ct-value _{non-challenged}.

Differences in gene induction between multiple groups were tested by one-way analysis of variance (ANOVA) and Bonferroni's Multiple Comparison Test. Tests were performed with GraphPad Prism version 4.00 for Windows, GraphPad Software, San Diego California USA. Differences were considered significant at p < 0.01.

In bold: restriction sites used in this study