AdpC is a Prevotella intermedia 17 leucine-rich repeat internalin-like protein

Abstract

The oral bacterium Prevotella intermedia attaches to and invades gingival epithelial cells, fibroblasts, and endothelial cells. Several genes encoding proteins that mediate both the adhesion and invasion processes are carried on the genome of this bacterium. Here, we characterized one such protein, AdpC, belonging to the leucine-rich repeat (LRR) protein family. Bioinformatics analysis revealed that this protein shares similarity with the Treponema pallidum LRR (LRR(TP)) family of proteins and contains six LRRs. Despite the absence of a signal peptide, this protein is localized on the bacterial outer membrane, indicating that it is transported through an atypical secretion mechanism. The recombinant form of this protein (rAdpC) was shown to bind fibrinogen. In addition, the heterologous host strain Escherichia coli BL21 expressing rAdpC (V2846) invaded fibroblast NIH 3T3 cells at a 40-fold-higher frequency than control E. coli BL21 cells expressing a sham P. intermedia 17 protein. Although similar results were obtained by using human umbilical vein endothelial cells (HUVECs), only a 3-fold-increased invasion of V2846 into oral epithelial HN4 cells was observed. Thus, AdpC-mediated invasion is cell specific. This work demonstrated that AdpC is an important invasin protein of P. intermedia 17.

FIG. 1.

Characteristics of P. intermedia 17 AdpC. (A) adpC genomic locus. Gene designations are according to the Oralgen database (www.oralgen.lanl.gov). (B) Alignment of P. intermedia 17 AdpC and Tannerella forsythia BspA. Leucine amino acid residues are shown in boldface type. The positions of the six LRR regions in AdpC are highlighted in gray.

FIG. 2.

Genetic organization of the adpC transcriptional start site. Sequences upstream of adpC are denoted in italic type. The transcriptional (+1) and translational (Met) start sites are indicated. Sequences used to design primers are underlined, and the designations of the primers are included underneath the sequences.

FIG. 3.

Purification of rAdpC. Shown is an SDS-PAGE analysis of the E. coli V2846 cell lysate before (lane 2) and after (lane 3) IPTG induction. Purified AdpC after nickel agarose (Ni-NTA) affinity chromatography is shown in lane 4. A molecular mass marker is shown in lane 1.

FIG. 4.

rAdpC binding to ECM proteins. Shown is data from ELISA of rAdpC binding to various ECM ligands. (A) Binding of various ligands to rAdpC. Different amounts of rAdpC ranging from 3 to 75 μg were deposited into wells of a 96-well plate. Solutions (100 μl) containing 20 μg/ml of various ECM proteins were then added to the wells, and the binding was detected by using antibody specific for the ECM proteins followed by a secondary AP-conjugated antibody. The amounts of bound proteins were then determined based on the fluorescent signal intensity derived from a substrate for AP. (B) Specificity of AdpC binding to fibrinogen. Various amounts of rAdpC and BSA were deposited into the wells of a 96-well plate, and the plate was then incubated with solutions (100 μl) containing 20 μg/ml of fibrinogen. Bound fibrinogen was detected as described above for A. (C) Competitive inhibition of rAdpC binding to fibrinogen using various concentrations of soluble rAdpC. A 96-well plate coated with rAdpC (2 μg/well) was incubated with fibrinogen (5 μg/ml) mixed with various amounts of rAdpC (0.25 mg/ml used to make dilutions ranging from 1 to 1:128). Bound fibrinogen was detected as described above for A.

FIG. 5.

Cellular localization of AdpC. Proteins were separated by SDS-PAGE and analyzed by Western blotting. AdpC was detected with affinity-purified rabbit anti-rAdpC. (A) E. coli V2921 whole-cell lysate proteins (lanes 1, 2, 3, and 4) and outer membrane proteins (lanes 5 and 6), which were derived from cultures grown in the presence (lanes 1, 3, and 5) or absence (lanes 2, 4, and 6) of IPTG. Proteins were derived from E. coli V2921 (lanes 1 and 2) and E. coli V2846 (lanes 3, 4, 5, and 6). (B) P. intermedia 17 outer membrane proteins (lane 2) and E. coli V2846 cell lysate proteins (lane 3). A molecular mass marker is shown in lane 1.

FIG. 6.

Surface location of AdpC determined by dot blot analysis. Serial dilutions of bacterial cells were deposited onto a nitrocellulose membrane, and the presence of AdpC was detected by using an anti-AdpC antibody. (A) E. coli containing plasmid pET30 (V2720). (B) Prevotella intermedia 17. (C) IPTG-induced E. coli containing recombinant plasmid pET30-adpC (V2846). (D) IPTG-induced E. coli containing recombinant plasmid pET30-adpC (V2846). The cells were protease treated prior to deposition onto the membrane.

FIG. 7.

Interaction of E. coli strains with eukaryotic cells. Eukaryotic cells were infected with E. coli V2846 (AdpC), V2848 (AdpD), and V2720 (pET30 only) as described in Materials and Methods. Total association (A, C, and E) and invasion (B, D, and F) of E. coli strains into host cells were determined by the quantification of viable bacteria by counting of CFU. Results from an experiment performed in triplicate are shown. (A and B) NIH 3T3 fibroblasts. (C and D) HUVECs. (E and F) Oral epithelial HN4 cells.

FIG. 7.

Interaction of E. coli strains with eukaryotic cells. Eukaryotic cells were infected with E. coli V2846 (AdpC), V2848 (AdpD), and V2720 (pET30 only) as described in Materials and Methods. Total association (A, C, and E) and invasion (B, D, and F) of E. coli strains into host cells were determined by the quantification of viable bacteria by counting of CFU. Results from an experiment performed in triplicate are shown. (A and B) NIH 3T3 fibroblasts. (C and D) HUVECs. (E and F) Oral epithelial HN4 cells.

FIG. 8.

Interaction of E. coli strains with HUVECs. HUVECs were infected with fluorescently labeled E. coli V2846 (AdpC) and V2720 (pET30 only) as described in Materials and Methods. Bacterial invasion was determined by visualization of the cells using confocal microscopy and counting of fluorescently labeled bacteria at 0.5-μm intervals for each HUVEC. Bacterial invasion of at least 30 HUVECs was assessed.