SssP1, a Streptococcus suis Fimbria-Like Protein Transported by the SecY2/A2 System, Contributes to Bacterial Virulence

Streptococcus suis is an important zoonotic pathogen. Here, we managed to identify key factors to clarify the virulence of S. suis strain CZ130302 from a novel serotype, Chz. Notably, it was shown that a fimbria-like structure was significantly connected to the pathogenicity of the CZ130302 strain by comparative genomics analysis and animal infection assays. The mechanisms of how the CZ130302 strain constructs these fimbria-like structures in the cell surface by genes encoding and production transport were subsequently elucidated. Biosynthesis of the fimbria-like structure was achieved by the production of SssP1 glycoproteins, and its construction was dependent on the SecA2/Y2 secretion system. This study identified a visible fimbria-like protein, SssP1, participating in adhesion to host cells and contributing to the virulence in S. suis. These findings will promote a better understanding of the pathogenesis of S. suis.

ABSTRACT

Streptococcus suis is an important Gram-positive pathogen in the swine industry and is an emerging zoonotic pathogen for humans. In our previous work, we found a virulent S. suis strain, CZ130302, belonging to a novel serotype, Chz, to be associated with acute meningitis in piglets. However, its underlying mechanisms of pathogenesis remain poorly understood. In this study, we sequenced and analyzed the complete genomes of three Chz serotype strains, including strain CZ130302 and two avirulent strains, HN136 and AH681. By genome comparison, we found two putative genomic islands (GIs) uniquely encoded in strain CZ130302 and designated them 50K GI and 58K GI. In mouse infection model, the deletion of 50K and 58K GIs caused 270-fold and 3-fold attenuation of virulence, respectively. Notably, we identified a complete SecY2/A2 system, coupled with its secretory protein SssP1 encoded in the 50K GI, which contributed to the pathogenicity of strain CZ130302. Immunogold electron microscopy and immunofluorescence analyses indicated that SssP1 could form fimbria-like structures that extend outward from the bacterial cell surface. The sssP1 mutation also attenuated bacterial adherence in human laryngeal epithelial (HEp-2) cells and human brain microvessel endothelial cells (HBMECs) compared with the wild type. Furthermore, we showed that two analogous Ig-like subdomains of SssP1 have sialic acid binding capacities. In conclusion, our results revealed that the 50K GI and the inside SecY2/A2 system gene cluster are related to the virulence of strain CZ130302, and we clarified a new S. suis pathogenesis mechanism mediated by the secretion protein SssP1.

IMPORTANCE Streptococcus suis is an important zoonotic pathogen. Here, we managed to identify key factors to clarify the virulence of S. suis strain CZ130302 from a novel serotype, Chz. Notably, it was shown that a fimbria-like structure was significantly connected to the pathogenicity of the CZ130302 strain by comparative genomics analysis and animal infection assays. The mechanisms of how the CZ130302 strain constructs these fimbria-like structures in the cell surface by genes encoding and production transport were subsequently elucidated. Biosynthesis of the fimbria-like structure was achieved by the production of SssP1 glycoproteins, and its construction was dependent on the SecA2/Y2 secretion system. This study identified a visible fimbria-like protein, SssP1, participating in adhesion to host cells and contributing to the virulence in S. suis. These findings will promote a better understanding of the pathogenesis of S. suis.

INTRODUCTION

The Gram-positive bacterium Streptococcus suis is considered to be one of the important bacterial pathogens in the swine industry (1, 2). Many clinical manifestations caused by S. suis infection have been reported, such as arthritis, pneumonia, endocarditis, septicemia, and meningitis (3). Among those manifestations, meningitis is the most lethal and is associated with severe economic loss (1). S. suis is also recognized as an emerging zoonotic pathogen of humans (4). After the first case report of S. suis infection in humans in 1968 at Denmark, human streptococcosis has been continuously reported all over the world. In China, two S. suis epidemic outbreaks occurred in 1998 and 2005 (5–8). Given the alarming fact that S. suis is threatening the swine industry and public health worldwide, studying the pathogenic mechanisms of S. suis is increasingly necessary.

S. suis isolates are serologically classified into 33 serotypes according to their antigenicity to capsular polysaccharides (CPSs) (9–11). Recently, 21 novel cps loci (NCL1 to NCL-20 and Chz) S. suis strains, isolated from diseased or healthy animals, have been identified (12–15). Serotype Chz strains were first reported in China and then in Canada, indicating the existence of dissemination in this serotype during the evolution of S. suis. The Chinese strain CZ130302 is a representative serotype Chz strain and was isolated from an outbreak of clinical acute piglet meningitis in 2013. CZ130302 caused a high fatality rate during that outbreak in pig farm and showed a higher virulence level than the S. suis serotype 2 (SS2) virulent strains P1/7 and HA9801 in a mouse model (12, 16). However, CZ130302 lacks extracellular protein factor (EPF), muramidase-related protein (MRP), and suilysin (SLY), which were three virulence markers widely used in S. suis to evaluate virulence (17, 18). The pathogenic mechanisms of CZ130302 remain to be explored.

Bacterial pathogens require surface proteins to interact with host components (19). The serine-rich repeat glycoproteins (SRRPs) are a particular family of bacterial surface proteins containing hundreds of ordered alternating serine residues, which are transported by the specialized transporter SecA2/Y2 (also known as accessory Sec) system (20). Since 1998, when the first SRR protein Fap1 was found in Streptococcus parasanguinis FW213, more Fap1-like proteins have been identified, predominately in the streptococci of oral cavities (e.g., S. sanguinis, S. pneumoniae, and S. gordonii), yet also in a fair number of pathogenic staphylococcal bacteria (e.g., Staphylococcus aureus and S. epidermidis) (21–26). Several reports demonstrated that the SRR family proteins played critical roles in bacterial pathogenicity by mediating the bacterial nonrepeat (NR) domains of SRRPs in their attachment to host cell surfaces (24, 26). In S. gordonii, the NR domains of Hsa and GspB attach to the sialylated ligands on the platelet membrane, thus enabling this bacterium to effectively colonize the damaged endocardium and cause infective endocarditis via platelet intermediaries (27). Additionally, it was shown that the binding of the NR domain of PsrP from S. pneumoniae to the lung epithelial cell surface factor keratin 10 is important for pneumococcal infection (28).

To investigate the underlying mechanisms helping to understand the strong pathogenicity of virulent S. suis strains, we carried out comparative genomic analyses between strain CZ130302 and two avirulent strains, HN136 and AH681. We identified a new fimbria-like protein, SssP1, transported by the SecY2/A2 system, which was encoded in a predicted 50K GI and required for S. suis virulence. Understanding the role of the SecY2/A2 secretion system and its effectors may help to understand the pathogenesis of S. suis.

RESULTS

CZ130302 may assemble a unique surface complex contributing to its virulence.

In the mouse model, the fatality rate for strain CZ130302 (10/10) was significantly higher than for strains HN136 (0/10) and AH681 (0/10) (P < 0.001), indicating that CZ130302 has a higher virulence level than two other strains (Fig. 1A). It is known that surface components on bacterial cells, such as capsules, flagella, and fimbriae, are important for bacterial contact with environmental niche and involved in bacterial adhesion to and invasion in the host or escape from the host defense systems (19, 29). To visualize the cell surface of the three S. suis Chz isolates, we performed a transmission electron microscopy assay (Fig. 1B). We found that the fimbria-like structure, extending outward from the bacterial surface, was formed in virulent strain CZ130302 but not in two other avirulent S. suis Chz strains HN136 and AH681. In Gram-negative bacteria, such pilus or fimbrial structures are known as essential protective antigens or virulence factors, yet they are seldom observed in micrographs of Gram-positive bacteria (29, 30). Whether this surface structure in CZ130302 is related to virulence or not needs to be further determined.

FIG 1.

Survival curves for BALB/c mice infected with the S. suis isolates and their morphological observation. (A) Mortality curve of lethal challenge with S. suis strains. Each group consisted of 10 mice, and a total of 2.5 × 10⁷ CFU/mouse (∼10 × LD₅₀ of CZ130302) of each strain was injected intraperitoneally. All the mice died within 7 days following the CZ130302 challenge. The log-rank (Mantel-Cox) test was used to compare survival curves between the groups (***, P < 0.001). (B) TEM images of the Streptococcus suis strains. Red arrows point to the fimbria-like structures. Black bars represent 200 nm.

Analysis of the well-known virulence factors identified in S. suis could not distinguish virulent and avirulent strains in S. suis serotype Chz.

To gain insight into possible genomic clues to the pathogenic mechanism of the virulent Chz strain, the complete genomes of CZ130302 and HN136/AH681 were sequenced. Each genome consisted of a single circular chromosome approximately 2.5 Mb in size (see Fig. S1 in the supplemental material), with their general features summarized in Table S1. The number of insertion sequence (IS) element regions found in the genome of CZ130302 was slightly greater than those in the avirulent strains, suggesting that this virulent strain may undergo more abundant structural variation during its evolution.

S. suis encodes a large number of virulence factors important for infection. More than 80 identified or putative virulence associated factors have been reported (31–33). To test the distribution of those virulence factors in serotype Chz strains, we detected 81 virulence-related proteins in the three serotype Chz strains. We found that 79% of the virulence-associated proteins (64/81) presented in CZ130302 (Table S2). Unexpectedly, the virulence-related proteins in two avirulent strains, HN136 (65/81) and AH681 (65/81), are similar to those of CZ130302 in terms of the proportion and distribution. This result suggested that the known virulence factors in S. suis may not efficiently explain the virulence difference in serotype Chz, and the strain CZ130302 may encode unknown virulence factors that uniquely contribute to its virulence in vivo.

Two unique genetic islands, 50K and 58K, were identified in strain CZ130302.

Due to the significant difference in virulence phenotype, a comparative genomic analysis would definitively improve our understanding of the virulence. X-alignment analysis of CZ130302 versus HN136 and AH681 suggested that two avirulent strains are closely related (Fig. 2A), which is consistent with the phylogenetic analysis (Fig. S2). In addition, the Mauve comparison alignment of the CZ130302, HN136, and AH681 genomes revealed a large-scale genomic rearrangement in CZ130302, which occurred in the 487-kb to 1,902-kb region (Fig. 2B). A total of 10 prophage-related gene clusters were detected in the three genomes. In particular, six prophages were located in CZ130302, of which four prophages were situated in the large rearranged segments; this suggested they may contribute to the structural diversification of the CZ130302 genome.

FIG 2.

Genome comparisons of the S. suis strains. (A) Synteny among the CZ130302, HN136, and AH681 genomes. The x axis and y axis show, respectively, positions on the AH681 and CZ130302 genomes (a), positions on the CZ130302 and HN136 genomes (b), and positions on the AH681 and HN136 genomes (c). (B) Mauve comparison diagram of the CZ130302, HN136, and AH681 genomes. Each colored region is a locally colinear block (LCB); same-colored LCBs indicate similar regions of alignment. Those LCBs below the centerline indicate their reverse complement orientation compared with the CZ130302 genome. Compared with the HN136 and AH681 genomes, a wide-scale genomic rearrangement occurred in the CZ130302 genome that is located in the big blue dashed box region (i.e., 0.5 to 1.9 Mb). The prophage-related gene clusters are marked with vertical arrows in the corresponding positions (P_CZ1 to P_AH2). (C) Circle comparison diagram of the CZ130302, HN136, and AH681 genomes. The colored ribbons connect their similar genomic elements. Two unique candidates, 50K GI and 58K GI, are present in CZ130302 but absent in the HN136 and AH681 strains.

It is important to note that the two unique 50K and 58K GIs identified existed only in CZ130302 but were absent in the two avirulent strains (Fig. 2C). Furthermore, both of two unique GIs were located in the area of the rearrangement. We further identified several components that may play potential roles in the pathogenic process by analyzing the two GIs (Fig. S3). In the 50K GI, CVO91_RS04720 and CVO91_RS04725 encoded two cell wall surface anchor family proteins; CVO91_RS04740–CVO91_RS04770 encoded seven glycosyltransferase proteins; and CVO91_RS04775–CVO91_RS04810 encoded a special SecY2/A2 Sec system. In the 58K GI, a defective prophage (designated Ph_CZ5) was identified throughout this island, similar to the Streptococcus oralis phage PH10. We also observed that the prophage harbored some specific sections capable of opportunistic antibiotic resistance and pathogenicity, including the proteolytic subunit ClpP, S-adenosylmethionine synthetase, chloramphenicol O-acetyltransferase, Rgg protein, and the RelBE toxin-antitoxin (TA) system.

The 50K GI plays a key role in the pathogenic process of Chz strain CZ130302.

To test whether those two GIs affect the virulence of CZ130302, knockout mutants for the 50K and 58K GIs were constructed (Δ50K-CZ130302 and Δ58K-CZ130302 mutants, respectively) (Fig. 3A). As shown in Fig. 3B, the 50% lethal dose (LD₅₀) values for CZ130302, the Δ50K-CZ130302 mutant, and the Δ58K-CZ130302 mutant were 2.89 × 10⁶, 7.84 × 10⁸, and 7.86 × 10⁶ CFU/mouse, respectively. Those data clearly suggested that both islands, especially 50K, were involved in the virulence of CZ130302. In addition, we also performed an organ-wide bacterial burden assay using the mouse infection model to test the involvement of the 50K GI in the proliferation of CZ130302 cells in vivo. As Fig. 3C shows, the bacterial loads in organs of the mice challenged with CZ130302 were significantly higher than those of the mice infected with the Δ50K-CZ130302 mutant strain (P < 0.001). Hence, the 50K GI facilities the in vivo bacterial proliferation and contributes to the virulence of strain CZ130302.

FIG 3.

Effects of specific islands on the pathogenicity of strain CZ130302 in mice. (A) PCR confirmation of the genes located in the 50K and 58K GI-deleted regions. Markers are 5,000 bp, 3,000 bp, 2,000 bp, 1,500 bp, 1,000 bp, 750 bp, 500 bp, 250 bp, and 100 bp. (B) LD₅₀ for BALB/c mice infected with the wild-type and mutant strains. Compared to the strain CZ130302, the deletion mutants, especially the Δ50K-CZ130302 mutants, had significantly higher survival over time. The results are indicated as the means ± standard error of the mean (SEM) of the results from 3 independent experiments (*, P < 0.05; ***, P < 0.001). (C) Survival of the wild-type and mutant strains in different mice organs. Treated mice were sacrificed and examined for bacteria in brain (a), lung (b), blood (c), and kidney (d) tissues at 24 h postinfection. The bacterial dissemination of these three strains was significantly different (*, P < 0.05; ***, P < 0.001).

SecY2/A2 system and its putative secretory protein SssP1 contribute to the virulence of Chz strain CZ130302.

The 50K GI was arbitrarily divided into six fragments to construct corresponding deletion mutants ΔP1 to ΔP6, which helped to conduct a preliminary screen of the potential virulence genes (Fig. S4). Our results showed that the deletion of the P5 fragment almost completely abolished the virulence in the mouse infection model (Fig. S4), suggesting a close relationship with the virulence of wild-type (WT) CZ130302. Further analyses of potential function by bioinformatics found a complete set of the SecY2/A2 secretion system encoded in the P5 fragment. The encoding genes of the SecY2/A2 system were retrieved in all S. suis whole/draft genomes available in the NCBI databases (https://www.ncbi.nlm.nih.gov) but were only present in ∼30 S. suis strains, and their biological functions need to be further confirmed.

The core component of the SecY2/A2 system gene in the strain CZ130302, being similar to other known SecY2/A2 system genes, contains two glycosyltransferases genes (gtfA-gtfB) involved in the serine-rich repeat protein (SRRP) glycosylation, six genes (secY2, secA2, and asp1 to asp4) encoding components of the SecY2/A2 complex that are essential for SRRP transport and glycosylation, and a effector protein (Fig. 4A). Furthermore, the BPROM program and reverse transcription-PCR (RT-PCR) analysis indicated that the eight core genes (from secY2 to asp4) form a single operon and share one promoter (Fig. 4B). The effector protein contained a particular KXYKXGKXW signal peptide and a serine-rich repeat adhesion glycoprotein AST domain at its N terminus. Since it belongs to the serine-rich repeat family of proteins, we named it Streptococcus suis serine-rich repeat protein 1 (SssP1). To investigate the functions of SssP1, the ΔsssP1 mutant strain was constructed (>12,000 bp in gene length caused a failure complementary strain construction). To avoid the possibility of polar effect, we detected the expression of flanking genes by quantitative reverse transcription-PCR (qRT-PCR) and found that their transcription levels were not affected (Fig. S6). In addition, we also artificially introduced a termination codon by a single nucleic acid site mutation, sssP1-T182A, to interrupt mRNA translation of sssP1 (Fig. 4C). No significant changes in bacterial growth were detected between the WT and mutant strains when cultured in Todd-Hewitt broth (THB) (Fig. 4D). The LD₅₀ values for the ΔsssP1, ΔsssP1-T182A, and SecY2/A2 system gene mutants (5.78 × 10⁸, 5.11 × 10⁸, and 5.30 × 10⁸ CFU/mouse, respectively), were similar to that of the Δ50K-CZ130302 mutant, which were significantly different from that of WT (Fig. 4E) (P < 0.001). The above-mentioned data suggest that the SecY2/A2 secretory system and its putative secretory protein SssP1 contribute to the virulence of CZ130302.

FIG 4.

Alignments of accessory Sec system gene clusters and the characteristics of SssP1. (A) Phylogenetic relationships of nine streptococcal and staphylococcal species based on the core set of secY2–secA2 genes (blue arrows). Phylogenetic relationships of the SecY2/A2 system gene clusters were obtained by constructing a neighbor-joining tree (bootstrap n = 500) based on a ClustalW alignment of the five core gene sequences. Genes encoding conserved domain proteins are in the same color. Schematic representation of the SssP1 protein is displayed above the gene cluster (S, signal sequence; NR, nonrepeat region; SRR, small repeat region). (B) Eight secY2-asp4 core genes of CZ130302 formed a single operon, as confirmed by RT-PCR. The RNA samples without reverse transcription were performed as the negative control. (C) Site-specific mutagenesis of the sssP1 gene. By artificially changing T to A at the bp 182 site, the ΔsssP1-T182A mutation was created. (D) Effect of the deletion mutants on the growth of Streptococcus suis CZ130302. The results are indicated as the means ± SEM of the results from 3 independent experiments (P > 0.05). (E) Effect of the SssP1 protein and SecY2/A2 system on pathogenicity. The deletion mutants resulted in a significantly high level of survival (***, P < 0.001).

SssP1 is a secretory protein transported by the SecY2/A2 system to construct fimbria-like structures on the cell surface.

The fimbria-like structures localized on the cell surface of CZ130302 by transmission electron microscopy (TEM) but disappeared when the ΔsssP1 and ΔsssP1-T182A deletions were made (Fig. 5A). To verify the relationship between SssP1 and this fimbria-like structure, recombinant SssP1_NR216–781 proteins and anti-SssP1_NR216–781 antiserum were prepared. Indirect immunofluorescence detected SssP1 on the surface of WT CZ130302 but not on the surface of the ΔsssP1 and ΔsssP1-T182A mutants (Fig. 5B). Immunogold electron microscopy with the specific anti-SssP1 antibody revealed the labeled SssP1 located primarily in the fibrillar layer of CZ130302 (Fig. 5C), thus confirming the assembly of fimbria-like structures by SssP1 proteins.

FIG 5.

Bacterial morphological analyses of the wild-type and mutant strains. (A) Analyzing bacterial surface structures with TEM. The bacteria were grown to the logarithmic phase by culturing in THB containing 5% fetal bovine serum and then collected, fixed, and sectioned. Red arrows point to the fimbria-like structures. Bars on the left (magnification, ×10,000) and right (magnification, ×25,000) are equivalent to 500 nm and 200 nm, respectively. (B) SssP1 protein localization on the surface of S. suis. The wild-type CZ130302 strain, the ΔsssP1 mutant strain, ΔsssP1-T182A mutant strain, and the SecY2/A2 system gene mutant strain were analyzed by immunofluorescence with an anti-SssP1 serum. The nucleoid was stained with DAPI (in blue), and the SssP1 protein was detected with an Alexa 488-conjugated goat anti-rabbit IgG (H+L) (in green). White bars represent 10 μm. (C) Immunogold electron micrographs of thin sections. Bacteria were incubated with anti-SssP1 serum and negative serum, and then they were treated with 10-nm goat anti-rabbit gold-conjugated particles, fixed, and sectioned. Red arrows point to the gold particles. Black bars represent 200 nm.

In the present study, after deleting the major genes in the SecY2/A2 system gene cluster (from secY2 to asp4), the qRT-RNA data revealed that the sssP1 gene was downregulated by more than 2-fold in SecY2/A2 system gene mutant than in the WT (Fig. S6). As shown in Fig. 5A and B, indirect immunofluorescence detected SssP1 to be absent from the cell surface in the SecY2/A2 system gene mutant samples. Furthermore, SssP1 was found in the whole-cell proteins but not in the cell wall proteins in the SecY2/A2 system gene mutant, when blotting was performed with the anti-SssP1_NR216–781 antiserum (Fig. S7). These suggested that the SecY2/A2 system plays an important role in the production and secretion of the SssP1 protein, which is consistent with findings of previous studies (20, 34).

SssP1 fimbria-like structure binds to human sialic acids and promotes adherence to host cells.

HEp-2 and human brain microvessel endothelial cells (HBMECs) were used to test how SssP1 affected bacterial adherence on host cells. As depicted in Fig. 6A and B, the adherence of the ΔsssP1 mutant was significantly reduced in both HEp-2 and HBMECs compared with its WT strain (P < 0.01). Further analyses revealed that SssP1 contains two analogous subdomains in its NR domains, with topology and strand inserts similar to the Ig-like fold (Fig. 6C to E). It is notable that the Ig-like subdomain is thought to be capable of binding to mammalian sialic acids (27), which is a clue that these two subdomains in Sssp1 might be involved in microbe-host interactions. Using a dot blot assay, we identified two recombinant proteins that contained Ig-like-1 and Ig-like-2 domains, respectively, that were involved in the binding to sialic acids, a well-known component of the host cell surface (Fig. 6F). Moreover, the obvious adherence of those two recombinant subdomains on the HBMEC surface was observed using fluorescence detection (Fig. 6G), which supports the idea that the SssP1 protein could utilize the Ig-like domains to facilitate the process of adhesion.

FIG 6.

The importance of SssP1 in bacterial adhesion. (A and B) Contribution of the SssP1 protein for adhering to the HEp-2 cells (A) and HBMECs (B). Shown are means ± standard deviation (SD) of the results from three independent experiments. The adhesion level of the SssP1 knockout mutant was significantly lower than that of strain CZ130302 (**, P < 0.01). (C) Alignment of SssP1_Ig-like-1 and SssP1_Ig-like-2 and their secondary structure assignment. Yellow boxes represent the same amino acids, green boxes represent similar kinds of amino acids, and secondary structural elements (β-sheets) in the sequences are shown by the blue (SssP1_Ig-like-1) and red arrows (SssP1_Ig-like-2). (D and E) The putative structures of the SssP1_Ig-like-1 and SssP1_Ig-like-2 proteins. The predicted maps were constructed with the SWISS-MODEL software, and the secondary structures are marked in the corresponding position. (F) Ig-like subdomains binding to saliva. Binding of SssP1_Ig-like-1 and SssP1_Ig-like-2 to immobilized human sialic acid was tested by the dot blot assay. (G) Contribution of the Ig-like subdomains to bacterial adhesion. HBMECs were incubated with SssP1_Ig-like-1 and SssP1_Ig-like-2, while the blank control group was incubated with BSA. The images show the DAPI-stained nuclei in blue and the adhered proteins detected by goat anti-mouse IgG-FITC in green. Red bars represent 100 μm.

DISCUSSION

S. suis is considered an important zoonotic pathogen, and it causes public health problems and economic losses worldwide, especially in countries with high densities of pigs and in countries where raw pork is consumed (35, 36). Increasing awareness of the importance of understanding S. suis pathogenic mechanism will help prevent and control S. suis-related diseases. Serotype Chz strain CZ130302 was shown to cause acute meningitis in piglets (12). In this study, we show that CZ130302 utilizes a SecA2/Y2-dependent fimbria-like structure, formed by the SssP1 protein, to adhere to host cells and thus establish an advantage in the colonization of host niches.

In stark contrast to the multiple substrate types in the canonical Sec system and SecA2-only system, the SecY2/A2 system appears to be specialized for the transport of SRR glycoproteins (37). Normally, most SRRPs contain two nonrepeat regions (i.e., NR1 and NR2) and two serine-rich repeat regions (i.e., SRR1 and SRR2), with the SRR2 domains usually accounting for a relatively large proportion of SRRP (34). However, the SssP1 protein, containing 4,647 amino acid residues, was composed of three NR domains and three SRR domains, which is the longest SRRP, to our knowledge. The NR2 domain of SssP1 exceeded 2,000 amino acids in size, and its isoelectric point (pI) was strongly acidic (Table S3). Several studies have shown that repetitive SRR2 domains can form a straight superhelical skeleton structure that is exposed to O-glycosylation for protein stability, which can also straddle extracellular components, such as capsular polysaccharides, to stretch out the NR domain (38, 39). In our study, immunogold electron microscopy of SssP1_NR216–781 revealed gold particles that were located distally from the bacterial cell wall surface, which is consistent with the literature (Fig. 5C). Next to the SRR2 domains of SssP1, there is an NR3 domain containing 525 amino acids and a SRR3 domain containing 78 truncated SAS(/L)T repeats at the C terminus, instead of a classic LPXTG cell wall anchor domain. Hence, it is worth investigating whether these special NR3-SRR3 domains are involved in stabilizing or interacting with the capsule or other bacterial cell surface components.

Bacterial adherence is an essential step in microbe colonization and invasion (40). The individual NR2 regions of the SRRPs are known to mediate adhesion, and they are highly diverse; correspondingly, there is a great variety of binding substrates in the NR domains on host cells, such as keratin 4, keratin 10, conjugated sialic acid, and other NR domains of similar SRRPs (28, 41–43). Our data suggest that the structure of SssP1NR2 contains two subdomains, NR_Ig-like-1 and NR_Ig-like-2, with analogous β-strands predominating as the secondary structures, and they are similar to the V-set Ig-like fold protein. The Ig-like fold is reportedly adopted by Siglecs, a family of sialic-acid-binding immunoglobulin-like lectins, and is considered able to recognize sialylated glycoconjugates and thus mediate bacterial attachment to host or bacterial cells (44). In our study, we discovered that the recombinant SssP1_Ig-like-1 and SssP1_Ig-like-2 domains showed binding capacities for human sialic acids and possessed obvious adherence to HBMECs. These observations, coupled with the fact that SssP1 inactivation significantly diminished bacterial adherence and virulence, suggest that SssP1 is critical for the colonization and full virulence of the strain CZ130302. Whether the SssP1 protein mediates S. suis infections in pigs through sialic acid binding needs further determination. Deng et al. proposed that some oral streptococci can utilize the Siglec-like domains of SRRPs, such as SrpA, GspB, and Hsa, to preferentially bind platelet sialoglycans, and that the circulating targeted platelets may serve as inadvertent carriers of bacteria which increase the risk of endocarditis (27). SssP1 also encoded two Siglec-like domains, indicating a potential to mediate the direct binding of S. suis to platelets. The platelet-binding role of SssP1 will be addressed in the future.

In summary, a comparative genomic analysis, using three complete genome sequences of a virulent serotype Chz strain CZ130302 and the two avirulent strains HN136 and AH681, was performed to identify two genomic islands, 50K and 58K, in CZ130302. The 50K GI harbors a complete SecY2/A2 secretion system to transport a fimbria-like protein, SssP1, which significantly contributed to adherence to host cells. Animal infection with mouse model confirmed that SssP1 protein is required for the full virulence of strain CZ130302. In conclusion, we uncovered a new pathogenesis in S. suis mediated by the SecY2/A2 secretion system and fimbria-like structure.

MATERIALS AND METHODS

Ethics statement.

Six-week-old female germfree BALB/c mice were purchased from the Comparative Medicine Center of Yangzhou University. All the animal experiments were approved by the Department of Science and Technology of Jiangsu Province, China [permit number SYXK (SU) 2017-0007].

Bacterial strains and culture conditions.

Three serotype Chz strains, CZ130302, HN136, and AH681, were used in this study to perform the comparative genomic analysis (Table 1). All S. suis strains were grown in Todd-Hewitt broth (THB; BD) containing 5% fetal bovine serum or on an agar medium containing 6% (vol/vol) sheep blood, under conditions of 37°C. For selection of the mutants, 100 μg/ml spectinomycin (Spc; Sigma-Aldrich) or 10% (wt/vol) sucrose was added to the medium for S. suis. The Escherichia coli strains were grown on Luria-Bertani (LB; BD) medium at 37°C. For construction of the recombinant plasmid, 50 μg/ml kanamycin (Kan; Sigma-Aldrich) was added to the medium when necessary.

TABLE 1.

Summary of bacterial strains and plasmids

Strain or plasmid	Descriptiona	Source or reference
Bacterial strains
CZ130302	Novel variant serotype Chz of S. suis which caused acute meningitis in piglets	Collected in our lab
HN136	Avirulent strain of S. suis serotype Chz	Collected in our lab
AH681	Avirulent strain of S. suis serotype Chz	Collected in our lab
Δ50K-CZ130302 mutant	Deletion mutant of 50K GI with CZ130302 background, Spcr	This study
Δ58K-CZ130302 mutant	Deletion mutant of 58K GI with CZ130302 background, Spcr	This study
ΔP1 mutant	Deletion mutant of P1 fragment with CZ130302 background, Spcr	This study
ΔP2 mutant	Deletion mutant of P2 fragment with CZ130302 background, Spcr	This study
ΔP3 mutant	Deletion mutant of P3 fragment with CZ130302 background, Spcr	This study
ΔP4 mutant	Deletion mutant of P4 fragment with CZ130302 background, Spcr	This study
ΔP5 mutant	Deletion mutant of P5 fragment with CZ130302 background, Spcr	This study
ΔP6 mutant	Deletion mutant of P6 fragment with CZ130302 background, Spcr	This study
SecY2/A2 system gene mutant	Deletion mutant of SecY2/A2 system gene cluster with CZ130302 background, Spcr	This study
ΔsssP1 mutant	Deletion mutant of sssP1 with CZ130302 background	This study
ΔsssP1-T182A mutant	Termination codon mutant of sssP1 with CZ130302 background	This study
DH5α	Cloning host for maintaining the recombinant plasmids	Invitrogen
BL21(DE3)	Host for expressing recombinant proteins	Invitrogen
Plasmids
pET28a(+)	Expression vector, Kanr	Invitrogen
pET28a-SssP1NR216–781	pET-28a containing sssP1NR216–781 gene, Kanr	This study
pET28a-SssP1Ig-like-1	pET-28a containing sssP1NR1295–1446 gene, Kanr	This study
pET28a-SssP1Ig-like-2	pET-28a containing sssP1NR2071–2212 gene, Kanr	This study

^aSpc^r, spectinomycin resistance cassette; Kan^r, kanamycin resistance cassette.

Genome sequencing, annotation, and comparative genomic analyses.

Total genomic DNA was extracted by using a bacteria DNA kit (Omega), according to the manufacturer's instructions. Genomic DNA was quantified by using the TBS-380 fluorometer (Invitrogen). The library construction and sequencing were performed at Novogene Biotechnology Co., Ltd. (Beijing, China). Whole-genome sequencing was performed on the Illumina HiSeq platforms (45). A TruSeq Nano DNA sample prep kit (Illumina) was employed to construct the library. The library insert size of 400 bp was verified with the Agilent 2100 Bioanalyzer. Illumina sequencing generated 3,180,308, 3,929,292, and 3,341,825 read pairs for strains CZ130302, HN136, and AH681, respectively. The quality of the raw sequencing data was determined by FastQC (46). Contamination reads, such as ones containing adaptors or primers, were filtered by Trimmomatic (47). SOAPdenovo version 2.04 (48) was used to assemble the de novo genomes. The scaffolds were aligned to the published genome of S. suis strain 05ZYH33 (NCBI accession number CP000407.1) to gain linkage information. Gaps between the remaining internal scaffolds were closed based on the GapCloser (https://sourceforge.net/projects/soapdenovo2/files/GapCloser/) and sequenced PCR products. Glimmer 3.0 (49) was used to predict the open reading frames (ORFs). All putative coding sequences were annotated by BLASTALL (50), with a set expected score of 1e−6. The tRNA and rRNA predictions were identified by the tRNAscan-SE (51) and RNAmmer (52), respectively. Prophage-related genes and insertion sequence (IS) elements were found with PHAST (53) and IS Finder (54). The MUMmer 3 package (55) was used for determining the genomic colinearity of the three S. suis genome sequences. A comparative genomic analysis was also carried out by using the Mauve version 2.3.1 (56) and Circos (57) programs. The summarized 81 known virulence factors were queried using BLASTp (https://blast.ncbi.nlm.nih.gov/Blast.cgi) against the three serotype Chz strains. Domains and three-dimensional structures were predicted using the SWISS-MODEL online server (https://www.swissmodel.expasy.org).

Construction of the gene deletion mutants.

To investigate the contribution of islands or genes, a series of mutant strains were created via natural DNA transformation, albeit with some new modifications (58, 59). In our study, the 9-amino-acid peptide ComS13-21 (GNWGKWTDG) stimulated a high level of natural transformation competence of CZ130302. The Δ50K-CZ130302 mutant was established as follows. The up- and downstream sequences were amplified by PCR with primer pairs 50K-L1/50K-L2 and 50K-R1/50K-R2, respectively, from the genomic DNA of strain CZ130302. The spectinomycin gene was amplified from pSET-4S using the 50K-M1 and 50K-M2 primers. The three amplicons were ligated by fusion PCR with the 50K-L1 and 50K-R2 primers. Then, the logarithmic CZ130302 strains were diluted 1:50 into THB medium and grown at 37°C without shaking. The linear fusion DNA fragment used for the Δ50K mutant (1.2 μg) and synthetic peptide (250 μM) were added to the 100-μl bacteria (optical density at 600 nm [OD₆₀₀], 0.035 to 0.058). The composite samples were incubated at 37°C for 2 h under static conditions and then plated in THB-agar medium containing spectinomycin. The other marked deletion mutants (Δ58K, ΔP1-ΔP6, and SecY2/A2 system gene mutants) were generated in CZ130302 in a similar fashion.

However, the unmarked gene mutants (ΔsssP1 and ΔsssP1-T182A) were constructed by two steps of natural DNA transformation. First, the primary marked mutant containing the sacB-spc cassette replacing target genomic loci was selected on the THB plate in the presence of spectinomycin. The sacB gene conferring sucrose sensitivity was used as negative selection (60). Then, the fusion homology fragment without any cassette was transferred to the primary positive mutant for the second transformation, followed by maintaining the transformed bacteria on THB plates containing 10% (wt/vol) sucrose. All primers and linear DNA fragments used in this study are listed in Table 2.

TABLE 2.

Primers used in this study

Primer by function	Sequence (5′–3′)a	Comment
Construction of deletion strains
50K-L1	TAGAGGCGAGTTTCGGACC	Upstream of fusion fragment for Δ50K mutation
50K-L2	GAACACTATTATACCTTCAATCCGCTAAG
50K-M1	GAAGGTATAATAGTGTTCGTGAATACATGTT	Spectinomycin resistance gene
50K-M2	CGTCACTATTTTCTAAAATCTGATTACCA
50K-R1	TTTAGAAAATAGTGACGAGTCAAAAGCAG	Downstream of fusion fragment for Δ50K mutation
50K-R2	GGGCTAGGACATCAAGTAAG
58K-L1	CCTGAGTTGTCATTCGCTGTG	Upstream of fusion fragment for Δ58K mutation
58K-L2	GAACACTAATTTATCGTCACTTTCGGATT
58K-M1	GACGATAAATTAGTGTTCGTGAATACATGTT	Spectinomycin resistance gene
58K-M2	CTGATAATTTTTTCTAAAATCTGATTACCA
58K-R1	TTTAGAAAAAATTATCAGACTGCCGTTAG	Downstream of fusion fragment for Δ58K mutation
58K-R2	ATTGCCCTTGGAAGAGTTAG
P1-L1	CATGCTATTGGCCTAGATCAT	Upstream of fusion fragment for ΔP1 mutant
P1-L2	CGAACACTAGTCACTACTCTGACTGTTGGA
P1-M1	GAGTAGTGACTAGTGTTCGTGAATACATGTT	Spectinomycin resistance gene
P1-M2	TGCGAAGTAATTTTCTAAAATCTGATTACCA
P1-R1	TTAGAAAATTACTTCGCAAATCCTGTCTC	Downstream of fusion fragment for ΔP1 mutation
P1-R2	TCGTTTACGTCTCATAGAACT
P2-L1	TTTCTTGCTTCGAGGTTCA	Upstream of fusion fragment for ΔP2 mutation
P2-L2	GAACACTATAAATACAGACGCATATACGAT
P2-M1	TGTATTTATAGTGTTCGTGAATACATGTT	Spectinomycin resistance gene
P2-M2	ATTCGTCCATTTTCTAAAATCTGATTACCA
P2-R1	TTAGAAAATGGACGAATTATCCAAGAAGAT	Downstream of fusion fragment for ΔP2 mutation
P2-R2	TATGGCACCAATTTCTTATCAT
P3-L1	TGAAAACAACGCTCACCAACAC	Upstream of fusion fragment for ΔP3 mutation
P3-L2	CGAACACTACACGCCTCCCGATTCCGGAGTA
P3-M1	GGAGGCGTGTAGTGTTCGTGAATACATGTT	Spectinomycin resistance gene
P3-M2	CTATACAGTTTTCTAAAATCTGATTACCA
P3-R1	TTAGAAAACTGTATAGTCGAAGCAGTGAAC	Downstream of fusion fragment for ΔP3 mutation
P3-R2	GAACTACGTTCGCTCTATCACC
P4-L1	TTCAAACCTAATGCGTTGGAT	Upstream of fusion fragment for ΔP4 mutation
P4-L2	CGAACACTATTCCCTGCGAAACAAATCTGAT
P4-M1	CGCAGGGAATAGTGTTCGTGAATACATGTT	Spectinomycin resistance gene
P4-M2	ATTTCCTATTTTTCTAAAATCTGATTACCA
P4-R1	TTAGAAAAATAGGAAATTGCTTGTTTCTTT	Downstream of fusion fragment for ΔP4 mutation
P4-R2	GCCAACACAAACATTAAGCTAT
P5-L1	TTAGAAAAGCAAGTACG	Upstream of fusion fragment for ΔP5 mutation
P5-L2	GAACACTACTACTCACTGCTATTTGATTA
P5-M1	CAGTGAGTAGTAGTGTTCGTGAATACATGTT	Spectinomycin resistance gene
P5-M2	CGTACTTGCTTTTCTAAAATCTGATTACCA
P5-R1	TTAGAAAAGCAAGTACGTCCGCATCGACT	Downstream of fusion fragment for ΔP5 mutation
P5-R2	TTGGTACCAACATACGGCATA
P6-L1	TGGAGCTCGTTGACGGAGGAT	Upstream of fusion fragment for ΔP6 mutation
P6-L2	GAACACTAATCTCTAACTACAGAACTGCT
P6-M1	GTTAGAGATTAGTGTTCGTGAATACATGTT	Spectinomycin resistance gene
P6-M2	ATATCTTATTTTCTAAAATCTGATTACCA
P6-R1	TAGAAAATAAGATATTGAAAGCGAGTTA	Downstream of fusion fragment for ΔP6 mutation
P6-R2	GTGACCGCTAAGAGCCGTTCC
sec-L1	CATAGAGGCAATGATTGTCGTT	Upstream of fusion fragment for SecY2/A2 system gene deletion
sec-L2	GAACACTAATCTCTTCATAAGGTGTATCCT
sec-M1	ATGAAGAGATTAGTGTTCGTGAATACATGTT	Spectinomycin resistance gene
sec-M2	ATATCTTTCTTTTCTAAAATCTGATTACCA
sec-R1	TTTAGAAAAGAAAGATATCCATGACGGCTCC	Downstream of fusion fragment for SecY2/A2 system gene deletion
sec-R2	TGCAAGATAATTAAGCCGAGA
sssP1-L1	TTCCACTATTCAAGGCAATC	Upstream of fusion fragment for ΔsssP1 mutation (step 1)
sssP1-L2	CAGCATTATCCGCATCTTGACACGGCTCTTAC
sssP1-M1	TGTCAAGATGCGGATAATGCTGAAAACTCCTTG	Spectinomycin resistance and negative selection marker (step 1)
sssP1-M2	GGTTGTGGACAATCTGATTACCAATTAGAATG
sssP1-R1	GTAATCAGATTGTCCACAACCGGGTGAAC	Downstream of fusion fragment for ΔsssP1 mutation (step 1)
sssP1-R2	GAACGCCCACCATATAGTA
sssP1-LU	ATATATGTACTGCCGAATGTT	Upstream of fusion fragment for ΔsssP1 mutation (step 2)
sssP1-LD	ACGCTATTACTCCTATGCAAAGAATGAGTCACA
sssP1-RU	TTCTTTGCATAGGAGTAATAGCGTTGGTACAGG	Downstream of fusion fragment for ΔsssP1 mutation (step 2)
sssP1-RD	GACTACTTCGCCAGGAAGGGT
sssP1-T182A-L1	ATCTTGATAAGCTCCATAGAG	Upstream of fusion fragment for ΔsssP1-T182A mutation (step 1)
sssP1-T182A-L2	GTTTTCAGCATTATCCACTGTTTCCTGCCCTTTCCCT
sssP1-T182A-M1	GGATAATGCTGAAAACTCCTTG	Spectinomycin resistance and negative selection marker (step 1)
sssP1-T182A-M2	AATCTGATTACCAATTAGAATG
sssP1-T182A-R1	ATTGGTAATCAGATTGTGATTCTTCCAGCCAGTCTG	Downstream of fusion fragment for ΔsssP1-T182A mutation (step 1)
sssP1-T182A-R1	GCGTGATTCTCCACGTAATCG
sssP1-T182A-LU	CGTCTTTAGTCGTCGCGTGAC	Upstream of fusion fragment for ΔsssP1-T182A mutation (step 2)
sssP1-T182A-LD	CTAACATAGCTAACGTTGACTTTCTTATCGC
sssP1-T182A-RU	GTCAACGTTAGCTATGTTAGCTTCAGGAGCG	Downstream of fusion fragment for ΔsssP1-T182A mutation (step 2)
sssP1-T182A-RD	ACCTGTTAGTCTGCCATTGTT
Construction of expression vectors
SssP1NR216–781-L	CgagctcCTAGGGTCTGTTGATGCTTCT	Construction of SssP1NR216–781
SssP1NR216–781-R	CCGctcgagGTAATCACTAGCATTCCCAAA
SssP1NR1295–1446-L	CgagctcGGTCAAGCTGCTTCTATTGAC	Construction of SssP1Ig-like-1
SssP1NR1295–1446-R	CCGctcgagACTGATACTTGTCGTGGTCGG
SssP1NR2071–2212-L	CgagctcGGTGATTCTTTACCTGTTGAT	Construction of SssP1Ig-like-2
SssP1NR2071–2212-R	CCGctcgagATTCCTGTCAGTAACCTTTAA
Cotranscription test
secY2/asp1-L	GGCATTCCGATGTTGGCTAGT	Cotranscription test spanning ORFs of secY2-asp1
secY2/asp1-R	CGATTTGGGCATACCGTTCTC
asp1/asp2-L	TTCCTCAGATTAACCGAATAG	Cotranscription test spanning ORFs of asp1-asp2
asp1/asp2-R	TTACTCGTTCTCAAGCGTTCC
asp2/asp3-L	GCCAAACGTCTATGAATCGAC	Cotranscription test spanning ORFs of asp2-asp3
asp2/asp3-R	CCCGCATTCAATAAGGCAATA
asp3/secA2-L	ACGACGCCTCTACTTATCATA	Cotranscription test spanning ORFs of asp3-secA2
asp3/secA2-R	CTTGGACCTTGTACGGAAACA
secA2/gtfA-L	ATCTACAACAACTCCGCGTCA	Cotranscription test spanning ORFs of secA2-gtfA
secA2/gtfA-R	TGAAACCGATATTCCGTGTCA
gtfA/gtfB-L	ATCTCTTGCCATTCCCTACAG	Cotranscription test spanning ORFs of gtfA-gtfB
gtfA/gtfB-R	CGCTTAGTTGCATAATCATAG
gtfB/asp4-L	CTGCATTTTCATATCGGAGCG	Cotranscription test spanning ORFs of gtfB-asp4
gtfB/asp4-R	TCGGAATAACATAGCGTTCTT

^aUnderlined nucleotides denote reverse complement; lowercase nucleotides denote restriction enzyme sites.

Cotranscription test.

Total RNA was extracted according to the protocol of the bacteria RNA kit (Omega). The primers were designed to span the ORFs of secY2-asp1 asp1-asp2 asp2-asp3 asp3-secA2 secA2-gtfA gtfA-gtfB and gtfB-asp4, listed in Table 2. Then, the total RNAs were reverse transcribed by using the HiScript II cDNA synthesis kit (Vazyme), according to the manufacturer's instructions. The cDNA samples were then used as PCR templates for a cotranscription test. Meanwhile, the RNA samples without reverse transcription were performed as the negative control to confirm that the samples were free of contaminating DNA. The promoter regions were predicted using the BPROM online server (Softberry).

In vivo challenges of BALB/c mice.

To evaluate the contribution of the two putative GIs to the virulence of CZ130302 strain, LD₅₀ assays were performed. The S. suis CZ130302, Δ50K-CZ130302 mutant, and Δ58K-CZ130302 mutant strains were grown to the logarithmic phase and collected, washed three times with phosphate-buffered saline (PBS), and then adjusted to the appropriate doses. In this assay, cohorts of 10 mice (6-week-old female germfree BALB/c) were infected with each S. suis strain dose; their mortality was monitored regularly, approximately every 8 h for 7 days. For the LD₅₀ assay, the mice were injected intraperitoneally with 10-fold serial dilutions (i.e., 10⁵ to 10⁹ CFU in 200 μl of PBS) of the wild-type (WT) strain CZ130302 and of the two mutant strains. For greatest precision, the 2-fold serially diluted suspensions that contained 2.5 × 10⁸ to 2 × 10⁹ CFU/mouse of the Δ50K-CZ130302 mutant strain dose treatments were shown to provide discriminatory power in virulence. The blank-control group was challenged with sterile PBS. The Reed-Muench method was used to calculate the LD₅₀ results. The virulence of the ΔsssP1, ΔsssP1-T182A, and SecY2/A2 system gene mutants was likewise determined by the above-described methods.

Additionally, a bacterial dissemination assay was performed to evaluate the proliferation capacity in vivo. Each group consisted of six mice, and the intravenous injection dose used was 1 × 10⁷ CFU/mouse (∼5× LD₅₀) of the WT and mutant strains. At 24 h postinfection, the diseased mice were anesthetized with isoflurane and euthanized by CO₂. Lungs, kidneys, and brains were harvested, weighed, and homogenized in 1 ml of PBS. Bacteria were isolated from these homogenates and blood by plating serial 10-fold dilutions on a THB-agar medium. From this, we obtained the number of bacteria that had colonized the organs of the mice during systemic infection.

Transmission electron microscopy.

To see the morphological changes in the S. suis strains, transmission electron microscopy (TEM) was used following a known method (61). In brief, the bacteria were grown to logarithmic phase and centrifuged at 5,000 × g for 10 min, and then fixed by 2.5% glutaraldehyde for at least 2 h. Next, the specimens were dehydrated with a graded series of ethanol and transferred to absolute acetone for 20 min. Finally, the samples were placed in capsules that contained an embedding medium for later sectioning. The ensuing sections were observed under TEM at an accelerating voltage of 200 kV. Moreover, immunogold electron microscopy was used to explore the relationship between the Sssp1 protein and the fimbria-like structures. Cells of CZ130302 were first washed with Hanks' balanced salt solution (HBSS) and then incubated for 2 h at 37°C with an anti-SssP1_NR216–781 polyclonal antibody that was diluted at 1:100 in HBSS containing 0.5% bovine serum albumin (BSA). Meanwhile, the blank-control group was disposed with preimmune negative serum. After thoroughly washing them, the cells were labeled for 1 h at 37°C with a 10-nm goat anti-rabbit gold-conjugated secondary antibody (Boster, China). The samples were then fixed with 2% paraformaldehyde and 2% glutaraldehyde in a 0.1 M sodium cacodylate buffer overnight, or for at least 8 h, at 4°C and then processed (as described in the above-mentioned steps). Finally, CZ130302 and its association with the gold particles was examined under a JEOL JEM-101 electron microscope.

Adhesion assays with HEp-2 and HBMECs.

According to the dissemination capacity of S. suis strains in vivo, an adhesion assay was also done following an established method (12). Specifically, the human laryngeal epithelial (HEp-2) cells and human brain microvessel endothelial cells (HBMECs), served as models of bacterial colonization and meningitis, respectively, in vitro. Briefly, these cells were cultured into a monolayer in 24-well tissue culture plates (final density, ∼5 × 10⁵ cells/well); then, they were infected with a treated bacterial suspension at a multiplicity of infection (MOI) of 1:20 (cell to bacterium) at a concentration of 1 × 10⁷ CFU/well, after which the plates were immediately centrifuged at 800 × g for 15 min and then incubated standing at 37°C for 2 h. Next, the monolayers were gently washed five times with sterile PBS to remove any unbound bacteria and treated with 100 μl of 0.25% trypsin-EDTA and 900 μl of sterile deionized water to release all bacteria; this was followed by plating 5-fold serial dilutions onto a THB-agar medium and their bacterial enumerations. The adhesion experiments were repeated three times. The results were expressed as the relative adhesion frequency compared with the adhesion frequency of WT CZ130302 (which was set to 100%).

Dot blot assay.

The immunoblotting for Ig-like subdomains were prepared by immobilizing human sialic acid as dots containing 5 μg on nitrocellulose (0.2-μm pore size; Merck Millipore). When dry, the membranes were blocked with Tris-buffered saline (TBS) containing 5% BSA and then overlaid for 2 h at room temperature with recombinant proteins (SssP1_Ig-like-1 and SssP1_Ig-like-2) at a concentration of 5 μg/ml. Unincubated protein membranes were performed as a negative control. Next, the blots were incubated with the His tag monoclonal antibody (Thermo Fisher) and goat anti-mouse IgG (H+L) (Thermo Fisher) at dilutions of 1:5,000 and 1:10,000, respectively. The signals were then detected using the substrate solution 3,3′-diaminobenzidine (Tiangen, China).

Immunofluorescence.

To validate the localization of SssP1 on the bacterial surface, cells of CZ130302, ΔsssP1 mutant, ΔsssP1-T182A mutant, and SecY2/A2 system gene mutant from the logarithmic-phase culture were centrifuged, washed thrice in PBS, and fixed with 4%-paraformaldehyde dissolved in PBS on a coverslip, as described elsewhere (62). After fixation, the bacterium was blocked with PBS containing 5% BSA and then labeled for 2 h at room temperature with an anti-SssP1_NR216–781 polyclonal antibody (diluted at 1:500 in 5% BSA). Once thoroughly washed, the samples were incubated with a goat anti-rabbit IgG (H+L) secondary antibody, Alexa Fluor 488-conjugated (Thermo Fisher), at a dilution of 1:500 in 5% BSA, followed by 4′,6-diamidino-2-phenylindole (DAPI; KeyGEN BioTECH) for bacterial DNA. Processed samples were mounted onto glass slides with Mowiol and examined by laser scanning confocal microscopy (Leica Sp5 AOBS confocal system).

An immunofluorescence assay was also done to validate whether the recombinant proteins can bind specifically to the HBMECs (63). Briefly, HBMECs were cultured in 24-well tissue culture plates and grown until confluence; they were then incubated with purified recombinant proteins (SssP1_Ig-like-1 and SssP1_Ig-like-2) (100 μg/ml) for 1 h at 37°C. The blank control group was incubated with BSA only. After incubation, the HBMECs were fixed with cold methanol for 20 min at −20°C and labeled with the His tag monoclonal antibody (Thermo Fisher), diluted at 1:500 in PBS, for 1 h at 37°C. Finally, the processed samples were stained with the secondary antibody goat anti-mouse IgG-fluorescein isothiocyanate (IgG-FITC; Thermo Fisher), at a dilution of 1:1,000, followed by DAPI for cellular DNA, before examination under a Carl Zeiss LSM710.

Statistical analyses.

All experiments were repeated at least three times. GraphPad Prism version 5 was used to analyze and plot the data. Log-rank (Mantel-Cox) tests were used to analyze mice survival. Differences between two groups were analyzed with a Student t test. The statistical level of significance was set a priori at a P value of <0.05.

Accession number(s).

The complete genome sequences of strains CZ130302, HN136, and AH681 were deposited in the GenBank database (accession numbers CP024974.1, CP025095.1, and CP025043.1, respectively).