Table 1.
T9SS components predicted using the differential genomics approach.
Fig 1.
Phyre2 modelling of PG1058 predicted structural domains.
Structures are coloured from blue at the N-terminus to red at the C-terminus. A. A TPR domain with three anti-parallel α-helix repeats (α-helix repeats numbered 1–3) and one capping helix (denoted by C) [47]. PG1058 Lys25-Arg153 modelled at 99.9% confidence against c412wA, the myosin chaperone UNC-45 from Caenorhabditis elegans in complex with a Hsp70 peptide [48]. B. A five bladed β–propeller domain (blades numbered 1–5). PG1058 Asp167-Ala438 modelled at 99.9% confidence against c2w8bB, E. coli TolB (a six bladed propeller protein) in complex with Pal [49]. C. The CRD. PG1058 Ile443-Arg528 modelled at 99.5% confidence against cmn8A, Drosophila melanogaster carboxypeptidase d isoform 1b2 short [50]. D. An OmpA_C-like domain. PG1058 Asn536-Val666 modelled at 100.0% confidence against c1r1m1, the OmpA-like domain from RmpM of N. meningitidis [45] with three α-helices (α-helices numbered α1-α3) and an additional α-helix (αn) which has seven additional residues relative to RmpM which are not represented in this model.
Fig 2.
A representative sequence from each of the 70 genera with PG1058 homologues were aligned using CLUSTAL W [51] and a sequence logo was generated using WebLogo 3 [52, 53]. The approximate boundaries of the CRD and OmpA_C-like domains are indicated by arrows. Low homology regions are primarily due to insertions in some sequences, although overall identity in the CRD is lower than seen in the other domains. The three TPR repeats are boxed and the five β–propeller blades are in alternating gray and yellow boxes. The putative peptidoglycan binding motif is boxed in green.
Fig 3.
Phenotypic characterisation of the pg1058 mutant compared to P. gingivalis W50.
A. Unlike P. gingivalis W50, the pg1058 mutant (pg1058) does not produce black pigment when grown on blood agar. The phenotype is restored following pg1058 complementation (pg1058+). B. The pg1058 mutant showed substantially reduced haemagglutination activity relative to P. gingivalis W50. Bacteria were 2-fold serially diluted from 20 to 211 in PBS and photographed after incubation with sheep erythrocytes at RT for 3 h. C. Arg-specific (indicative of the presence of RgpA and RgpB proteinases) and Lys-specific (indicative of the presence of Kgp proteinase) proteinase activity of W50 (wild-type), pg1058 mutant and pg1058+ complement strains in whole-cells and culture supernatants. Units per 2.5 x 1011 cells equates to amount of substrate hydrolysed in μmol/min/2.5 x 1011 cells whilst Units per 100 mL equates to amount of substrate hydrolysed in μmol/min/supernatant derived from 100 mL of culture containing 2.5 x 1011 cells. Inactivation of pg1058 abolished the Arg- and Lys-specific proteolytic activity of the cells, which was restored after pg1058 complementation. D. Absence of EDSL on the P. gingivalis pg1058 mutant. Cryo-EM micrographs representative of the P. gingivalis W50 and pg1058 mutant. Outer membrane (OM), inner membrane (IM), peptidoglycan (PGN) and electron dense surface layer (EDSL) are indicated. Scale bar is 200 nm.
Fig 4.
Aberrant localisation of T9SS substrates in the pg1058 mutant.
P. gingivalis W50 and pg1058 mutant periplasm (A) or total membrane (B) fractions were separated via SDS-PAGE and stained with SimplyBlue™ SafeStain. Each lane was divided into segments and analysed by LC-MS/MS. When the same protein was identified from multiple gel segments the Mascot scores were summed. The total Mascot score for the substrates identified from W50 (black bars) and the pg1058 mutant (white bars) were plotted on a logarithmic axis (left Y-axis). The total Mascot score for the combined substrates and the total Mascot scores for the combined periplasmic or membrane proteins were plotted on a linear axis (right Y-axis). Proteins are indicated by the Locus Tag in the P. gingivalis W83 strain unless previously designated with a protein ID.
Fig 5.
Aberrant localisation of proteinases in subcellular fractions of the pg1058 mutant.
P. gingivalis W50, pg1058 mutant and pg1058+ complement strain cultures were fractionated, separated via SDS-PAGE and immunoblotted on nitrocellulose membranes. The immunoblots have material derived from whole-cells (WC; 2 x 108 cells), total membrane (3 x 108 cells) and soluble (1 x 108 cells) fractions; WC, periplasm and osmotically-shocked cell (OSC) fractions (5 x 108 cells each); WC (2 x 108 cells), culture fluid (CF; from culture containing 2 x 109 cells) and vesicle-free cleared culture fluid (CCF; from culture containing 2 x 109 cells). Immunoblots were probed with either anti-rKgpcat (A, 1/200,000 dilution), pre-adsorbed anti-rRgpAcat (B, 1/10,000 dilution), anti-rKgpA1 (C, 1/10,000 dilution) or anti-rRgpB-CTD422 (D, 1/200,000 dilution) followed by goat anti-rabbit (A) or horse anti-mouse (B, C and D) IgG-conjugated HRP secondary antibodies (1/3,000 dilution).
Fig 6.
Absence of surface-associated Kgp in the pg1058 mutant determined by whole-cell ELISA analysis.
FKWC (109 per well) prepared from P. gingivalis W50 (positive control), pg1058 mutant and kgp mutant (negative control) were probed with two-fold dilutions of anti-rKgpcat antisera (1/5,000–1/80,000) followed by goat anti-rabbit HRP-conjugated IgG (1/3,000) and developed with ABTS chromogenic substrate. The chromogenic reaction was stopped and detected at 405 nm. Controls included no primary antibody control (No 1°Ab), no secondary antibody control (No 2°Ab) and no cell control (No Bacteria). Mean ± SEM (standard error of the mean), N = 3 for each strain.
Fig 7.
Presence of A-LPS in the pg1058 mutant.
A. P. gingivalis W50, pg1058 mutant and pg1058+ complement strain cultures were fractionated, separated via SDS-PAGE and immunoblotted on nitrocellulose membranes (as per Fig 5). Immunoblots were probed with MAb 1B5 (1/10,000 dilution) followed by horse anti-mouse IgG-conjugated HRP secondary antibody (1/3,000 dilution). B. Whole-cell ELISA of P. gingivalis W50 and pg1058 mutant FKWC (109 per well) probed with two-fold dilutions of MAb 1B5 from 1/100–1/1,600 followed by horse anti-mouse HRP-conjugated IgG secondary antibody (1/3,000) and developed with ABTS chromogenic substrate. The chromogenic reaction was stopped and detected at 405 nm. Controls included no primary antibody control (No 1°Ab), no secondary antibody control (No 2°Ab) and no cell control (No Bacteria). Mean ± SEM, N = 3 for each strain. C. Immunogold TEM micrographs representative of the P. gingivalis W50 (N = 19) and pg1058 mutant (N = 29). Probing with MAb 1B5 (1/10,000 dilution) was followed by detection with anti-mouse IgG conjugated to 18 nm colloidal gold particles (1/40 dilution). D. The immunogold particle distribution. Distance between IM and centre of the gold particle measured in nm, distances less than 0 nm correspond to a cytoplasmic and IM localisation, 0–20 nm corresponds to the periplasmic space, 21–120 corresponds to the OM and extracellular environment. Mean ± SEM.
Fig 8.
PG1058 is an OM-associated periplasmic protein.
A. P. gingivalis W50, pg1058 mutant and pg1058+ complement strain cultures were fractionated, separated via SDS-PAGE and immunoblotted on nitrocellulose membranes. The immunoblots have material derived from vesicles (5 x 109 cells); whole-cells (WC; 2 x 108 cells), culture fluid (CF; from culture containing 2 x 109 cells) and vesicle-free cleared culture fluid (CCF; from culture containing 2 x 109 cells); WC (2 x 108 cells), soluble (5 x 109 cells), total membrane (Membrane; 1.6 x 1010 cells), sarkosyl-insoluble membrane (SI; 1.6 x 1010 cells) and sarkosyl-soluble membrane (SS; 1.6 x 1010 cells) fractions; WC (5 x 108 cells), periplasm (5 x 108 cells) and osmotically-shocked cell (OSC; 5 x 108 cells) fractions. Immunoblots were probed with anti-rPG1058 (1/10,000 dilution) and horse anti-mouse IgG-conjugated HRP secondary antibody (1/3,000 dilution). B. Whole-cell ELISA. P. gingivalis W50, pg1058 mutant, pg1058+ complement FKWC or lysed P. gingivalis W50 FKWC as a positive control (2 x 109 cells per well) were probed with anti-rPG1058 antiserum (1/100, 1/200, 1/400, 1/800 and 1/1600 dilutions) followed by horse anti-mouse HRP-conjugated IgG secondary antibody (1/3,000) and developed with ABTS chromogenic substrate. The chromogenic reaction was suspended and detected at 405 nm. Other controls included no primary antibody control (No 1°Ab), no secondary antibody control (No 2°Ab) and no cell control (No Bacteria). Mean ± SEM, N = 3 for each strain except where only one replicate (N = 1) of the positive control (Positive) was assessed. C. P. gingivalis W50 total membrane was separated via sucrose density gradient centrifugation then fractions were separated by SDS-PAGE and immunoblotted on nitrocellulose membrane. Fraction numbers are indicated above the lanes and P refers to the pellet (unequal loading). Immunoblots were probed with anti-rPG1058 (1/10,000 dilution) or anti-Omp41 (1/10,000 dilution) followed by horse anti-mouse IgG-conjugated HRP secondary antibody (1/3,000 dilution).