Fig 1.
Fap1 is not essential for adhesion to sialic acid.
(A) Southern blot of EcoRV-digested gDNA from five different S. oralis subsp. oralis IE-isolates using a DIG-labeled fap1 probe. gDNA from S. oralis subsp. oralis strain ATCC10557 was used as a positive control (+). fap1 was detected in only two IE-isolates. (B) Adherence of five S. oralis subsp. oralis IE-isolates to platelets in the presence of 30 μM CBM40 (+). Adherence was expressed as a percentage relative to the same strain in the absence of CBM40 (-). Values are the means for at least three independent experiments, each performed in triplicate, ± SD. *Statistical significance was tested by two-tailed Student’s t-test. *, P ≤ 0.0015.
Fig 2.
β-1,4-linked galactose serves as a receptor for some S. oralis subsp.
oralis isolates. Binding to β-1,4-linked galactose was tested by the addition of 50 μM CBM71-1.2 following pretreatment with neuraminidase (NPRE) or PBS (-). The graphs show adherence of the SRRP- (A) and SRRP+ (B) IE-isolates to untreated or neuraminidase-pretreated platelets in presence or absence of CBM71-1.2. Adherence is expressed as a percentage relative to binding of the strains to untreated platelets in the absence of CBM71-1.2. Values are the means for at least three independent experiments, each performed in triplicate, ± SD. Statistical significance was tested by two-tailed Student’s t-test. *, P ≤ 0.015; NS, not significant.
Fig 3.
A Sortase A-dependent surface protein(s) is required for binding of S. oralis IE-isolates lacking SRRPs.
(A) a srtA mutant is reduced in binding to platelets. Graphs indicate adherence of IE12 and IE18 srtA mutants and complements, with srtA or the empty vector (pABG5), relative to parental strains. The contribution of SrtA-dependent surface proteins to β-1,4-linked galactose binding was tested by comparing adhesion to neuraminidase-pretreated platelets (NPRE). Adherence data are the means for at least three independent experiments, each performed in triplicate, ± SD. Statistical significance was tested by two-tailed Student’s t-test. *, P ≤ 0.006.(B) Venn diagram showing shared and unique LPxTG-containing proteins in three different S. oralis subsp. oralis IE-isolates. Comparisons were made by BLASTP, proteins with >80% amino acid identity were considered as shared between strains. (C) As demonstrated by PCR, the presence of IE12_1764 and IE18_0557 correlates with the ability of fap1- isolates to bind sialic acid or β-1,4-linked galactose, respectively. rpoB was used as a positive control.
Fig 4.
Csh-like protein is not involved in IE18 binding to β-1,4-linked galactose.
(A) Schematic illustrating the predicted protein architecture of Csh-like from S. oralis subsp. oralis IE18 (IE18_0557). Csh-like consists of an N-terminal secretion signal (SS), followed by the NR2 domain (PDB: 5L2D), 23 CshA-type fibril repeats and a cell wall anchoring motif (LPxTG). (B) The non-repeat region of Csh-like (Csh_NRR) is not involved in binding to β-1,4-linked galactose. Unlike CBM71-1.2, the addition of 30 μM of the recombinantly expressed Csh_NRR did not significantly reduce binding of IE18 to neuraminidase-pretreated (NPRE) platelets. Adherence is expressed as a percentage relative to binding of IE18 to neuraminidase-pretreated (NPRE) platelets in the absence of CBM71-1.2 or Csh_NRR. Values are the means for at least three independent experiments, each performed in triplicate, ± SD. Statistical significance was tested by two-tailed Student’s t-test. *, P ≤ 0.001; NS, not significant.
Fig 5.
AsaA is a novel sialic acid-binding adhesin.
(A) Schematic illustrating the predicted protein architecture of AsaA (IE12_1764) from S. oralis subsp. oralis IE12. AsaA consists of an N-terminal secretion signal (SS), followed by the non-repeat region predicted to contain a FIVAR/CBM domain and two Siglec-like and Unique domains, followed by 31 DUF1542 domains and a C-terminal LPxTG motif. (B) Comparison of the AsaA Siglec-like and Unique domains tertiary structure, predicted by SWISS-MODEL, to the solved structure of SrpA Siglec-like and Unique domain (5KIQ). (C) An IE12 asaA mutant was reduced in binding to platelets. Neuraminidase pretreatment (NPRE) did not further reduced binding of the mutant. (D) Left: schematic illustrating the domains removed in the isogenic IE12 variants. To further evaluate protein production and localization, a 3xFLAG tag (black box) was added to all constructs, including the parental strain (WT-3F). Right: Adhesion of WT-3F and the isogenic AsaA-deletion mutants to platelets. (E) Production of the AsaA deletion variants evaluated by the detection of the 3xFLAG tag in whole-cell lysates by immunodot-blot. (F) Detection of the AsaA-deletion variants on the cell surface by immunofluorescent microscopy. Representative images of Nile red-stained bacteria (red) incubated with anti-FLAG and Alexa Fluor 488-conjugated secondary antibody (green). Bars 1 μm. Adherence is expressed as a percentage relative to binding of the parental strain to untreated platelets. Values are the means for at least three independent experiments, each performed in triplicate, ± SD. Statistical significance was tested by two-tailed Student’s t-test. *, P ≤ 0.0004. NS, not significant.
Fig 6.
The Siglec-like and Unique domains of AsaA mediate direct binding to sialic acid on platelets.
(A) Adhesion of IE12 was reduced in a dose-dependent manner by increasing concentrations of AsaA_NRR. (B) Adhesion of IE12 and the asaA mutant to untreated and neuraminidase-pretreated (NPRE) platelets in the presence of 5 μM AsaA_NRR or AsaA_SU_1–2. Adherence is expressed as a percentage relative to binding of IE12 to untreated platelets in the absence of recombinant proteins. (C) Direct binding of platelets pretreated with PBS or neuraminidase (NPRE) to immobilized AsaA_NRR, AsaA_SU_1–2 or GST. A representative image of platelets binding to immobilized protein is shown. The graph shows the average number of platelets recovered in three independent experiments performed in triplicate. Statistical significance was tested by two-tailed Student’s t-test. *, P ≤ 0.002. NS, not significant.
Fig 7.
AsaA contributes to S. oralis IE12 virulence in a rabbit model of IE.
The two strains indicated were co-inoculated into six rabbits from two independent experiments (three rabbits each) performed on different days. Individual values and geometric means of bacteria recovered from vegetations are shown. Identical symbols indicate bacteria recovered from the same animal. Log-transformed values were analyzed by paired Student’s t-test. *, P ≤ 0.002.
Fig 8.
AsaA is not required for binding to fibrinogen or fibronectin.
The IE12 asaA mutant displayed increased binding to immobilized fibronectin, while binding to fibrinogen was not significantly different from that of IE12. Adherence is expressed as a percentage relative to binding of IE12. Values are the means for at least three independent experiments, each performed in triplicate, ± SD. Statistical significance was tested by two-tailed Student’s t-test. *, P ≤ 0.02; NS, not significant.
Fig 9.
AsaA can serve as an adhesin in other IE-causing bacterial species.
(A) Schematic illustrating the predicted protein architecture of the NRR of AsaA orthologues from G. haemolysans M341 (EGF87895.1), G. elegans ATCC700633 (EEW93886.2), S. pasteuri 915_SPAS (WP_048803571.1) and S. mitis B6 (CBJ22549.1). The percentage of amino acid identity to the NRR of AsaA from S. oralis subsp. oralis IE12 is shown to the right of the schematic. (B) Adhesion to platelets of IE12 and the asaA mutant in the presence of 5 μM Gh_AsaA_NRR. Adherence is expressed as a percentage relative to binding of the same strain in the absence of Gh_AsaA_NRR. Values are the means for at least three independent experiments, each performed in triplicate, ± SD. Statistical significance was tested by two-tailed Student’s t-test. *, P ≤ 0.0015; NS, not significant.
Table 1.
Strains and plasmids.
Table 2.
Primers used in this study.