Lack of a surface layer in Tannerella forsythia mutants deficient in the type IX secretion system.

Tannerella forsythia, a Gram-negative anaerobic bacterium, is an important pathogen in periodontal disease. This bacterium possesses genes encoding all known components of the type IX secretion system (T9SS). T. forsythia mutants deficient in genes orthologous to the T9SS-encoding genes porK, porT and sov were constructed. All porK, porT and sov single mutants lacked the surface layer (S-layer) and expressed less-glycosylated versions of the S-layer glycoproteins TfsA and TfsB. In addition, these mutants exhibited decreased haemagglutination and increased biofilm formation. Comparison of the proteins secreted by the porK and WT strains revealed that the secretion of several proteins containing C-terminal domain (CTD)-like sequences is dependent on the porK gene. These results indicate that the T9SS is functional in T. forsythia and contributes to the translocation of CTD proteins to the cell surface or into the extracellular milieu. The Authors.

Introduction

Oral biofilms comprise more than 700 bacterial species and matrix substances and contribute to the development of periodontal disease (Aas ). Although chronic periodontitis is caused by a mixed infection, specific micro-organisms including Porphyromonas gingivalis, Tannerella forsythia and Treponema denticola are considered important for the initiation and progression of chronic periodontitis (Holt & Ebersole, 2005).

P. gingivalis encodes a variety of virulence factors, such as the extracellular and cell-surface cysteine proteinases Arg-gingipain and Lys-gingipain (Potempa ; O'Brien-Simpson et al., 2003). Recently, these proteinases were shown to be secreted by the Por secretion system (PorSS) (Sato ; Sato ). The proteins constituting the PorSS differ from those constituting other secretion systems. The P. gingivalis PorSS includes the PorK, PorL, PorM, PorN, PorP, PorQ, PorT, PorU, PorV (PG27, LptO), PorW and Sov proteins (Sato ). Coding sequences (CDSs) encoding proteins homologous to the P. gingivalis PorSS proteins are present in the genomes of several bacteria in phylum Bacteroidetes (McBride & Zhu, 2013). Therefore, the PorSS has been called the type IX secretion system (T9SS) (Chagnot ).

T. forsythia, an anaerobic Gram-negative bacterium, belongs to phylum Bacteroidetes. It appears to possess the T9SS because it has genes encoding all known components of the T9SS (Sato ). T. forsythia is phylogenetically related to P. gingivalis; however, unlike P. gingivalis, T. forsythia does not form black-pigmented colonies on blood-agar plates. Mixed infection by T. forsythia and P. gingivalis enhanced abscess formation in a murine model (Takemoto ; Yoneda ). T. forsythia encodes multiple potential virulence factors, including the PrtH proteinase and surface components such as surface layer (S-layer) glycoproteins (TfsA and TfsB) and the leucine-rich-repeat protein BspA (Sharma, 2010). Some virulence-related proteins, including TfsA, TfsB and BspA, appear to have C-terminal domains (CTDs) that may function as a recognition signal for the T9SS (Veith ; Shoji ).

In this study, T. forsythia mutants deficient in porK, porT and sov orthologous genes that may be involved in the translocation of CTD proteins such as TfsA, TfsB and BspA to the cell surface were generated. The porK, porT and sov mutant cells exhibited morphological changes and expressed less-glycosylated versions of the S-layer proteins TfsA and TfsB. In the porK mutant, several CTD proteins were not secreted into the extracellular milieu. These results indicate that the T9SS is functional in T. forsythia and is important for the virulence of this bacterium.

Methods

Bacterial strains and culture conditions.

All bacterial strains and plasmids used in this study are listed in Table 1. T. forsythia cells were grown anaerobically (10 % CO2, 10 % H2, and 80 % N2) in enriched brain heart infusion broth (BHI) medium (Sato ) supplemented with 10 µg ml−1 N-acetylmuramic acid (MurNAc) (Sigma-Aldrich) and 5 % (v/v) heat-inactivated calf serum (CS) and on enriched tryptic soy agar (Sato ) supplemented with 10 µg ml−1 MurNAc and 5 % (v/v) defibrinated laked sheep blood. For the selection and maintenance of erythromycin (Em)-resistant T. forsythia strains, Em was added to the medium at a concentration of 5 µg ml−1.

Table 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Description	Reference or source
Bacterial strains
Tannerella forsythia
43037	WT	ATCC
NTF1	porK : : ermF, Emr	This study
NTF2	ΔporT : : ermF, Emr	This study
NTF3	sov : : ermF, Emr	This study
Δ230	ΔtfsA : : cat, Cpr	Sakakibara et al. (2007)
Δ270	ΔtfsB : : cat, Cpr	Sakakibara et al. (2007)
Δ230-270	ΔtfsA tfsB : : cat, Cpr	Sakakibara et al. (2007)
Escherichia coli
XL-1Blue	Host strain for cloning	Stratagene
Plasmids
pCR4 Blunt TOPO	Apr, Kmr, PCR cloning vector	Invitrogen
pBluescript II SK(-)	Apr, cloning vector	Stratagene
pKD1030	Apr, contains porK-5′ region in pBluescript II SK(–)	This study
pKD1031	Apr, contains porK-5′ and -3′ regions in pBluescript II SK(−)	This study
pKD1032	Apr, contains the ermF DNA cassette at the BamHI site of pKD1031	This study
pKD1033	Apr, contains porT-upstream region in pCR4 Blunt TOPO	This study
pKD1034	Apr, contains porT-upstream and -downstream regions in pCR4 Blunt TOPO	This study
pKD1035	Apr, contains the ermF DNA cassette at the BamHI site of pKD1034	This study
pKD1036	Apr, contains 2.0 kb sov fragment in pCR4 Blunt TOPO	This study
pKD1037	Apr, contains the ermF DNA cassette at the BamHI site of pKD1036	This study

Emr, erythromycin resistance; Apr, ampicillin resistance; Cpr, chloramphenicol resistance; Kmr, kanamycin resistance.

Construction of bacterial strains.

Genomic nucleotide sequence data of T. forsythia ATCC 43037 was obtained from the GenBank database (accession number: CP003191). The T. forsythia porK insertion mutant was constructed as follows. A 0.6 kb 5′-terminal region of porK was amplified from the chromosomal DNA of T. forsythia ATCC 43037 using the Pyrobest DNA polymerase (TaKaRa) and PCR using the primers TFporKUF and TFporKUR. The DNA primers used in this study are listed in Table S1 (available in the online Supplementary Material). The amplified DNA was cloned into the pCR4 Blunt TOPO vector (Invitrogen) according to the manufacturer’s instructions and digested with EcoRI and BamHI. The resulting DNA fragment was then inserted into the EcoRI and BamHI sites of pBluesript II SK(−) to generate pKD1030. A 0.8 kb 3′-terminal region of porK was amplified from the chromosomal DNA of ATCC 43037 with the primers TFporKDF and TFporKDR. The amplified DNA was cloned into pCR4 Blunt TOPO and digested with BamHI and NotI. The resulting fragment was then inserted into the BamHI and NotI sites of pKD1030 to generate pKD1031. The 1.1 kb BamHI ermF DNA cassette was inserted into the BamHI site of pKD1031, resulting in pKD1032 (porK : : ermF). The pKD1032 was linearized with NotI and introduced into ATCC 43037 by electroporation to generate the NTF1 strain.

The T. forsythia porT deletion mutant (NTF2) was constructed as described above except that the DNA regions upstream and downstream of porT were amplified by PCR from the chromosomal DNA of the strain ATCC 43037 with the primers TFporTUF and TFporTUR and the primers TFporTDF and TFporTDR, respectively.

The T. forsythia sov insertion mutant was constructed as follows. A 2.0 kb internal region of the sov gene was amplified from the chromosomal DNA of the strain ATCC 43037 by PCR with the primers TFsovF and TFsovR. The amplified DNA was cloned into pCR4 Blunt TOPO to generate pKD1036. The 1.1 kb BamHI ermF DNA cassette was inserted into the BamHI site in the sov region of pKD1036, resulting in pKD1037 (sov : : ermF). The pKD1037 was linearized with NotI and introduced into T. forsythia ATCC 43037 by electroporation to generate the NTF3 strain.

Electron microscopy.

To examine bacterial cell shape, the cells were washed and negatively stained on carbon-coated grids with 1 % ammonium molybdate. To prepare ultrathin sections, the cells were fixed with 4 % paraformaldehyde and 5 % glutaraldehyde in 30 mM HEPES buffer (pH 7.4) overnight at 4 °C. The samples were post-fixed with 1 % osmium tetroxide for 2 h and then with 0.5 % uranyl acetate for 30 min. The fixed cells were dehydrated in a series of 25–100 % ethanol and embedded in Quetol-651 resin (Nisshin EM). The ultrathin sections were stained with 1 % uranyl acetate and 1 % lead citrate. The stained samples (bacterial cells and ultrathin sections) were observed using a JEM-1210 transmission electron microscope (JEOL).

Gel electrophoresis and immunoblot analysis.

SDS-PAGE and immunoblot analyses were performed as previously described (Shoji ). The blotted membranes were treated with anti-TfsA and anti-TfsB antisera (Sakakibara ).

The glycoproteins in SDS-PAGE gels were stained using the Pro-Q Emerald 300 fluorescent stain (Invitrogen). After staining with Pro-Q Emerald 300, total protein staining was performed with SYPRO Ruby (Invitrogen).

Two-dimensional gel electrophoresis (2D-PAGE).

2D-PAGE was performed as described previously (Sato ). T. forsythia strains were grown in serum-free medium. Particle-free culture supernatants were obtained as previously described (Sato ). The proteins in the particle-free culture supernatant fraction were precipitated with 10 % (w/v) trichloroacetic acid at 4 °C. The precipitated proteins were harvested by centrifugation at 4 °C for 20 min, washed three times with cold diethyl ether, dried at room temperature for 30 min and then resuspended in a cell lysis solution (7 M urea, 2 M thiourea, 4 % CHAPS, 1 mM EDTA and 5 mM tributylphosphine). The samples were applied to 13 cm immobilized pH gradient strips (GE Healthcare Bio-Sciences) with a pH range from 4 to 7 (first dimension) swollen with a rehydration solution [7 M urea, 2 M thiourea, 4 %, v/v, CHAPS, 0.5 %, v/v, IPG buffer pH 4 to 7 (GE Healthcare), 1 mM EDTA, 12 µl ml−1 destreak reagent (GE Healthcare)] and bromophenol blue. The 2D electrophoresis (SDS-PAGE) was performed in polyacrylamide gels, and the proteins were stained with Coomassie brilliant blue (CBB) R250 .

MS analysis and database search for protein identification.

Proteins were identified by peptide-mass fingerprinting after in-gel tryptic digestion as previously described (Sato ). Gel plugs containing proteins were subjected to washing with 50 % (v/v) acetonitrile, washing with 100 % acetonitrile, reduction with 10 mM DTT (Wako), alkylation with 55 mM iodoacetamide, washing/dehydration with 50 % (v/v) acetonitrile and digestion for 10 h with 10 µg ml−1 trypsin. The resulting peptides were extracted from the gel plugs with 0.1 % (v/v) trifluoroacetic acid and 50 % (v/v) acetonitrile. The digests were spotted on a MALDI target using α-cyano-4-hydroxycinnamic acid as a matrix. The spectra were acquired on a 4800 MALDI TOF/TOF Analyser (Applied Biosystems). Data analysis and MS database searching were performed using GPS Explorer and Mascot software (Matrix Science) with the significance criteria of the program (P<0.05).

Haemagglutination analysis.

Haemagglutination was assessed as previously described (Shoji ).

Assay for trypsin-like activity.

Trypsin-like activity was assayed as described previously (Grenier, 1995). Trypsin-like activity was measured by monitoring the hydrolysis of the chromogenic synthetic peptide benzoyl-dl-arginine-p-nitroanilide (BAPNA; Peptide Institute) in the presence or absence of various compounds: EDTA (Wako Pure Chemical Industries), N-α-p-tosylamide-2-phenylethyl chloromethyl ketone (TLCK; Wako), leupeptin (Peptide Institute), DTT, iodoacetamide (Wako), SDS (Wako), CaCl2 (Sigma-Aldrich), MgCl2 (Wako) and ZnCl2 (Wako). The cells were harvested by centrifugation (10 000 , 30 min) and suspended in distilled water. The bacterial samples (12.5 µl) were mixed with 125 µl of 150 mM Tris/HCl buffer (pH 7.8), 50 µl of 4 mM BAPNA and 12.5 µl distilled water, and the assay mixtures were incubated at 37 °C for 2 h. The release of p-nitroaniline was determined by measuring the OD405 nm using a microplate reader (Bio-Rad).

Biofilm formation.

Biofilm formation was measured by a microtitre plate biofilm assay using a previously reported protocol with slight modification (O'Toole & Kolter, 1998; Honma ). Briefly, an overnight culture adjusted to an OD595 nm of 1.0 was diluted 1 : 10 with fresh medium. The cells were aliquoted into the wells of a 96-well microtitre plate (250 µl per well) and incubated anaerobically for 1–4 days. After removal of the planktonic cells by washing twice with PBS, the biofilm was stained by incubation with 100 µl of 0.1 % (w/v) crystal violet solution for 5 min. The plate was washed twice with distilled water and destained with 200 µl of 95 % (v/v) ethanol for 5 min. Biofilm mass was evaluated at OD595 nm at 1, 2, 3 and 4 days using a microplate reader. Total biofilm formation was evaluated as the absorbance of crystal violet stained biofilms at OD595 nm divided by the absorbance of total growth (including biofilm and planktonic cells) at OD595 nm.

For visualization by microscopy, T. forsythia biofilms were formed in 4-well Lab Tek II chamber slides (Nunc) as described previously (Honma ). The diluted cells described above were dispensed in each chamber and incubated anaerobically for 3 days. Following incubation, chambers were washed with PBS and biofilms were stained with the SYTO9 LIVE/DEAD Baclight Bacterial Viability kit L7012 (Invitrogen, Molecular Probe). The stain was prepared according the manufacturer’s instruction and coverslips mounted using Vector Shield (Vector Laboratories). The edges of the coverslips were coated with nail polish and the slide was kept in the dark until analysed. Microscope images of fluorescence were examined using ZEISS Axioskop plus microscope mounted with an AxioCam HRc camera with a Plan-Neofluar 20× 0.5 NA objective (ZEISS). The Z-stack images of the biofilm were acquired with confocal laser scanning microscope (LSM510 META, AxioVert200M) with a LD-Achroplan 40× 0.6 NA objective (ZEISS). For three-dimensional reconstruction, the image analysis was done using Z-series image stacks of each biofilm with the LSM image browser and ZEN2009 software.

Statistical analysis.

The data are presented as means±sd. The biofilm formation data were analysed using one-way ANOVA with the Tukey–Kramer multiple-comparison test. P-values less than 0.05 were considered significant.

Results

Construction of T. forsythia mutants deficient in T9SS proteins

The erythromycin-resistance DNA cassette was inserted into the T. forsythia CDSs bfor_c_1_3635, bfor_c_1_6468 and bfor_c_1_12435 (locus tags by the Human Oral Microbiome Database), which were orthologous to P. gingivalis porK, porT and sov, to generate T. forsythia porK (NTF1), porT (NTF2) and sov (NTF3) mutants, respectively (Fig. S1).

Cell morphology of the porK, porT and sov mutants

The bacterial cells were negatively stained with ammonium molybdate and analysed by electron microscopy. WT T. forsythia cells were prolate, ellipsoid-like with sharp ends, whereas the porK, porT and sov mutants were rod-like with round ends (Fig. 1a). The WT cells showed a lattice structure on the cell surface, whereas the porK, porT and sov mutants exhibited amorphous and fragile surfaces. The lattice structure of the WT T. forsythia cell surface is the S-layer, which contains the TfsA and TfsB proteins (Sakakibara ; Sekot ); therefore, the porK, porT and sov mutants appeared to lack the S-layer. Analysis of the ultrathin cross-sections revealed that the WT cells had an S-layer with a thickness of approximately 20 nm on the outer membrane, whereas the porK mutant cells did not possess this structure on the outer membrane (Fig. 1b).

Fig. 1.

Comparison of the cell shape of WT T. forsythia and T9SS-deficient mutants. (a) Electron micrographs of negatively stained cells. The regions indicated by dashed boxes in the upper panels are magnified in the lower panels. Bars, 500 nm (upper), 100 nm (lower). (b) Electron micrographs of ultrathin cell sections. Bars, 100 nm.

SDS-PAGE and immunoblot analyses of whole-cell lysates

The SDS-PAGE profiles of whole-cell lysates revealed that major proteins with molecular masses of 230 and 270 kDa in the WT cells were not present in the porK, porT or sov mutant cells; furthermore, the porK, porT and sov mutant cells contained major proteins with molecular masses of 165 and 205 kDa that were absent in the WT cells (Fig. 2). Immunoblotting analysis using antisera against the TfsA and TfsB proteins revealed that the 230 and 270 kDa proteins in the WT cells were the TfsA and TfsB proteins, respectively, and that the 165 and 205 kDa proteins in the porK, porT and sov mutants were derived from the TfsA and TfsB proteins, respectively. The molecular masses of the TfsA and TfsB proteins without signal peptides are 133.3 kDa and 150.8 kDa, respectively (Lee ); therefore, it was examined whether these proteins were glycosylated (Fig. 3). ProQ-Emerald carbohydrate staining analysis suggested that the 230 kDa TfsA protein and the 270 kDa TfsB protein were strongly glycosylated in WT cells and that the 165 kDa TfsA protein and 205 kDa TfsB protein were also glycosylated in the porK, porT and sov mutants, albeit to a lesser extent.

Fig. 2.

SDS-PAGE and immunoblotting analysis. T. forsythia cell lysates were analysed by SDS-PAGE and immunoblotting using anti-TfsA and anti-TfsB antisera. The proteins were stained with CBB R250.

Fig. 3.

Glycoprotein staining of SDS-PAGE gels of T. forsythia cells with (a) Pro-Q Emerald 300 fluorescent glycoprotein stain and (b) SYPRO Ruby fluorescent total protein stain. The first lane in each gel contains the CandyCane glycoprotein molecular mass standards (Invitrogen), a mixture of glycosylated and non-glycosylated proteins used as a positive control for staining.

2D-PAGE analysis of particle-free culture supernatants

2D-PAGE was performed to analyse the particle-free (membrane-free) culture supernatants from the WT and porK strains (Fig. 4). Time of flight mass spectrometry (TOFMS) analysis revealed the presence of the T. forsythia proteins bfor_c_1_1931 (tentatively named type Nine Secretion System-dependent protein A, NdpA), bfor_c_1_8519 (NdpB), bfor_c_1_10593 (NdpC), bfor_c_1_10600 (NdpD) and bfor_c_1_14540 (NdpE) in the particle-free culture supernatant of the WT cells but not the porK mutant cells (Table 2).

Fig. 4.

2D gel analysis. 2D-PAGE analysis was performed using IPG strips with a pH 4–7 isoelectric point range to analyse the particle-free culture supernatants of the WT and NTF1 (porK) strains. The proteins were stained with CBB R250. Green circles indicate the spots with no, or lower, protein levels in the porK mutant cells compared with the WT cells. Identification of protein spots in Fig. 4 is provided in Table 2.

Table 2.

Identification of protein spots in 2D-gels

Protein spot		HOMD name	Description	Mascot score
WT	NTF1
a	*	bfor_c_1_3670	Hypothetical protein	261	616
b1	*	bfor_c_1_4507	Hypothetical protein	71	1222
b2	*	bfor_c_1_4507	Hypothetical protein	456	162
c1	*	bfor_c_1_4502	Hypothetical protein	559	419
c2	*	bfor_c_1_4502	Hypothetical protein	257	181
c3	*	bfor_c_1_4502	Hypothetical protein	26	–
c1	*	bfor_c_1_4505	Hypothetical protein	277	658
d	*	bfor_c_1_4347	Conserved hypothetical protein	389	266
e	*	bfor_c_1_629	Hypothetical protein	–	165
f	*	bfor_c_1_14680	Conserved hypothetical protein; possible haemagglutinin/haemolysin	–	900
g1	*	bfor_c_1_8519	PorU, conserved hypothetical protein	272	-
g2	*	bfor_c_1_8519	PorU, conserved hypothetical protein	200	-
h	†	bfor_c_1_14540	Thermolysin precursor	216	-
i1	†	bfor_c_1_10600	Lysyl endopeptidase	425	-
i2	†	bfor_c_1_10600	Lysyl endopeptidase	96	-
i3	†	bfor_c_1_10600	Lysyl endopeptidase	300	-
j	†	bfor_c_1_10593	Eukaryotic-like metalloproteinase, karilysin	84	-
k	†	bfor_c_1_1931	Thermolysin; zinc metalloprotease	44	-
l	‡	bfor_c_1_843	Possible lipoprotein	107	264
m	‡	bfor_c_1_2071	Conserved hypothetical protein	182	-
n1	‡	bfor_c_1_2868	Conserved hypothetical protein	357	410
n2	‡	bfor_c_1_2868	Conserved hypothetical protein	449	341
o		bfor_c_1_4765	Zinc protease	129	594

HOMD, Human Oral Microbiome Database.

CTD family protein (Veith ).

Probable CTD family protein.

Antigenic protein (non-CTD) (Veith ).

Haemagglutination

Purified S-layers cause erythrocyte agglutination, and S-layer-deficient mutant strains exhibit decreased haemagglutination activity (Sabet ; Sakakibara ). The haemagglutination activities of the porK, porT and sov mutants on sheep erythrocytes were measured (Fig. 5). The haemagglutination activities of the porK, porT and sov mutants were 12.5 % lower compared with the WT strain. The haemagglutination activities of the tfsA and tfsB single mutants were 50 % of the WT level, whereas the haemagglutination activity of the tfsA tfsB double mutant cells was 25 % of the WT level.

Fig. 5.

Haemaggulutination. Suspensions of T. forsythia cells and serial twofold dilutions in PBS were applied to the wells of a microtitre plate from left to right and mixed with a sheep erythrocyte suspension.

Biofilm formation

The T. forsythia wecC mutant strain, which lacks UDP-N-acetylmannosaminuronic acid dehydrogenase and has truncated S-layer glycans, shows increased biofilm formation (Honma ; Posch ). An examination of the biofilm-forming ability of the porK, porT and sov mutants (Figs 6 and 7) revealed that these mutants showed increased biofilm formation.

Fig. 6.

Biofilm-forming ability. (a) The adherent cells formed after 1, 2, 3 and 4 days were quantified by measuring the OD595 nm of crystal violet (CV). (b) OD595 nm of CV staining/OD595 nm of total bacterial growth. Three independent experiments were performed in quadruplicate or quintuplicate to provide 13 sets of data. The differences in each strain compared with the WT were analysed using one-way ANOVA with the Tukey–Kramer multiple-comparison test. *P<0.01. WT, T. forsythia ATCC 43037; porK, NTF1; porT, NTF2; sov, NTF3.

Fig. 7.

Biofilm analysis by fluorescence microscopy (FM) and confocal laser scanning microscopy (CLSM). Bacterial cells were stained with SYTO 9. (a) FM images of 3 day biofilms. Bars, 50 µm. (b) CLSM images of 3 day biofilms. Biofilm thicknesses of each strain are: WT (3.86 µm), porK (22.4 µm), sov (8.5 µm), porT (14.7 µm). x-z and y-z reconstructions of each biofilm are shown upper and right sides of x-y image. Bars, 50 µm.

Trypsin-like activity

T. forsythia produces an enzymic endopeptidase that degrades BAPNA; this endopeptidase activity was attributed to a trypsin-like proteinase (Grenier, 1995). To examine whether the porK, porT and sov mutations affect the trypsin-like endopeptidase activity of T. forsythia, BAPNA hydrolysis using porK, porT and sov mutant cells in the presence and absence of various chemicals including protease inhibitors was monitored (Table 3). The hydrolysis activities of the porK, porT and sov mutant cells were similar to that of the WT. The BAPNA-hydrolysis activities of the mutant and WT cells were completely or partially suppressed by TLCK, ZnCl2 and leupeptin, indicating that the WT and mutant cells contained similar BAPNA-hydrolysis properties.

Table 3.

Trypsin-like activity

Inhibitor	Concn. (final)	Residual activity (%)*
WT	porK	porT	sov
None	–	100	127.7	126.9	110.4
EDTA	10 mM	107.6	123.1	126.9	105.7
Iodoacetamide	10 mM	113.7	122.3	155.7	101.3
TLCK	1 mM	1.3	4.0	7.3	0.7
Leupeptin	0.01 mM	57.5	72.3	78.1	52.2
CaCl2	10 mM	87.3	103.3	122.5	91.7
ZnCl2	10 mM	28.6	14.2	28.9	19.6
DTT	10 mM	112.3	128.8	99.8	113.7

Trypsin-like activities of various T. forsythia strains with various protease inhibitors and metal ions were determined using BAPNA as a substrate. Enzymic activity of the WT without any additive in the reaction mixture was taken as 100 %.

Discussion

The T9SS (PorSS) was discovered in the periodontal pathogen P. gingivalis (Sato ). Subsequently, homologous genes encoding putative T9SS components were observed in several bacterial species of phylum Bacteroidetes, suggesting that the T9SS is conserved in at least a subset of this phylum (McBride & Zhu, 2013; Chagnot ). The T9SS is related to gliding motility of bacteria of phylum Bacteroidetes (Sato ; Nakane ).

T. forsythia, which belongs to phylum Bacteroidetes, is a member of the ‘red complex’ together with P. gingivalis and T. denticola and is considered a major pathogen underlying periodontal disease. T. forsythia possesses several putative virulence factors such as trypsin-like protease, PrtH protease, sialidases, BspA leucine-rich repeat protein and the S-layer. In this study, we generated three T9SS-deficient T. forsythia mutants in which the porK, porT and sov genes were mutated. All mutations caused identical phenotypes such as absence of the S-layer, decreased haemagglutination activity and increased biofilm formation, suggesting that these properties are related to the T9SS.

The S-layer is a paracrystalline surface-protein array expressed in several bacteria and is thought to function as a protective coat against external sieves and ion traps (Sleytr & Beveridge, 1999; Sabet ; Messner ). The T. forsythia S-layer mediates adhesion to human gingival epithelial cells and subsequent invasion (Sakakibara ) and delays recognition of the bacterium by the host innate immune system (Sekot ). The S-layer contributes to T. forsythia serum resistance and oral bacterial coaggregation (Shimotahira ) and consists of the TfsA and TfsB proteins. S-layer proteins generally contain N-terminal signal peptides, with a few exceptions (Boot & Pouwels, 1996). Because the primary products of the tfsA and tfsB CDSs contain N-terminal signal peptides, these proteins are probably translocated across the inner membrane by a Sec-dependent mechanism. The TfsA and TfsB proteins contain CTD-like sequences at the C terminus, and CTD sequences are a signal for T9SS-mediated translocation across the outer membrane (Shoji ), suggesting that the S-layer proteins are translocated across the outer membrane by the T9SS. Posch observed that the molecular masses of His-tagged TfsA and TfsB proteins expressed in Bacteroides fragilis, which are probably located in the periplasm because of lack of T9SS in B. fragilis (Sato ; McBride & Zhu, 2013), are ~170 kDa and ~200 kDa, respectively, which are consistent with the molecular masses of the TfsA and TfsB proteins in the T. forsythia T9SS-deficient mutants. T. forsythia as well as B. fragilis has an O-glycosylation system (Fletcher ; Coyne ; Posch , 2013). These findings suggest that in T. forsythia cells, the S-layer proteins are primarily O-glycosylated at the inner membrane and/or in the periplasm, and after translocation across the outer membrane by the T9SS, the proteins are further glycosylated on the cell surface.

Using 2D-gel analysis, we observed that the NdpA, NdpB, NdpC, NdpD and NdpE proteins were released into the WT but not the porK culture supernatant. The NdpA and NdpE proteins are thermolysin metallopeptidase homologues; the NdpB protein is a PorU homologue (Sato ; Glew ); the NdpC protein is a karilysin (Karim ) and the NdpD protein is a putative lysyl endopeptidase homologous to P. gingivalis PepK (Nonaka ). These proteins are putative peptidases and possess CTD-like sequences at their C termini. These results indicate that the T9SS is functional in T. forsythia and is involved in the secretion of CTD proteins.

Sabet purified the S-layer from T. forsythia and observed that the S-layer was sufficient to mediate the haemagglutination of sheep erythrocytes. Sakakibara generated tfsA and tfsB single mutants and a tfsA tfsB double mutant and observed that these S-layer-deficient mutants caused decreased haemagglutination of chicken erythrocytes. In this study, it was observed that the porK, porT and sov mutants caused decreased haemagglutination of sheep erythrocytes. The haemagglutination activity of the T9SS-deficient mutants was weaker than that of the S-layer-deficient mutants, suggesting that cell-surface proteins other than TfsA and TfsB, which are secreted by the T9SS, are also involved in haemagglutination.

Honma isolated a wecC mutant that showed increased biofilm formation and observed that the molecular masses of both S-layer proteins were decreased in the wecC mutant. Subsequently, the decreased molecular mass of the S-layer proteins was correlated with truncated S-layer glycans (Posch ). In this study, it was observed that the T9SS-deficient mutants lacked S-layers and contained TfsA and TfsB proteins with decreased molecular masses and reduced glycosylation compared with WT cells; these mutants also showed increased biofilm formation. These results indicate that S-layers or S-layer glycans suppress T. forsythia biofilm formation.

Trypsin-like endopeptidase activity was observed in the T. forsythia cell envelope (Grenier, 1995). The T9SS-deficient mutants showed the same trypsin-like activity as the WT, suggesting that the T9SS is not required for translocation of the trypsin-like enzyme(s) to the cell envelope.

In this study, T. forsythia porK, porT and sov mutant strains were generated, and these mutants were found to lack the S-layer. Several CTD proteins such as thermolysin were not observed in the culture supernatant of the porK mutant cells. These results indicate that the T9SS is functional in T. forsythia and contributes to translocation of the CTD proteins to the cell surface or into the extracellular milieu.