DnaJ of Streptococcus suis Type 2 Contributes to Cell Adhesion and Thermotolerance

Abstract

To examine if the molecular chaperone DnaK operon proteins of Streptococcus suis type 2 (SS2) are involved in adhesion to host cells, the abundance values of these proteins from the surface of two SS2 strains of different adhesion capability were compared. Their roles in growth and adhesion to human laryngeal epithelial cell line HEp-2 cells were investigated on SS2 strain HA9801 and its mutants with DnaK operon genes partially knocked-out (PKO mutant) under heat stress. The major difference was that DnaJ was more abundant in strain HA9801 than in strain JX0811. Pretreatment of the bacteria with hyperimmune sera to DnaJ, but not with those to other proteins, could significantly reduce SS2 adhesion to HEp-2 cells. PKO of dnaJ g ene resulted in decreased SS2 growth at 37℃ and 42℃, and reduced its adhesion to HEp-2 cells. The wild-type strain stressed at 42℃ had increased expression of DnaJ on its surface and elevated adhesion to HEp-2 cells, which was also inhibitable by DnaJ specific antiserum. These results indicate that the DnaJ of S. suis type 2 is important not only for thermotolerance but also for adhesion to host cells. Because DnaJ expression is increased upon temperature upshift with increased exposure on the bacterial surface, the febrile conditions of the cases with systemic infections might help facilitate bacterial adhesion to host cells. DnaJ could be one of the potential candidates as a subunit vaccine because of its good immunogenicity.

Introduction

Adhesion to host cells and colonization of tissues are prerequisites for bacterial pathogens to successfully establish infection [18,26]. Streptococcus suis type 2 (SS2), a zoonotic pathogen causing septicemia, meningitis, or toxic shock syndrome in swine and humans, may enter the body through the respiratory tract and reach the central nervous system via blood circulation [10,11]. Adherence to the respiratory epithelia is the first step for its infection. SS2 contains multiple mechanisms to bind to host cells via a number of adhesion-associated proteins such as dipeptidyl peptidase IV [8], enolase [5,6], fibronectin-binding protein [12], glyceraldehyde-3-phosphate dehydrogenase [1], sortase A [25], glutamine synthetase [27], amylopullulanase [7], fused deoxyribonucleotide triphosphate pyrophosphatase/unknown domain-containing protein HAM1 [2], and autolysin [16].

Molecular chaperone DnaK operon proteins include HrcA, GrpE, DnaK, and DnaJ. HrcA is a heat-shock transcription repressor for the other four chaperone proteins by its interaction with the controlling inverted repeat of chaperone expression (CIRCE) element in front of hrcA [23]. GrpE, a nuclotide exchange factor and heat shock protein, acts as a co-chaperone to help DnaK release ADP [9]. DnaK possesses weak ATPase activity that is stimulated by its interaction with DnaJ [22]. Thus, the nucleotide- and substrate-bound state of DnaK is regulated by co-chaperones DnaJ and GrpE for its functions in response to stresses. A recent study reported that DnaK might be involved in the adhesion of S S 2 to HEp-2 cells [2]. This led us to explore the roles of other chaperone proteins of the DnaK operon in SS2 pathogenicity. Here, we demonstrate that DnaJ contributes not only to thermotolerance but also to adhesion of SS2 in cultured HEp-2 cells.

 

Materials and Methods

Bacterial Strains, Growth Conditions, and Cell Culture

Streptococcus suis type 2 strain JX0811 was isolated from the lung of a diseased pig in Zhejiang Province, China [31]. The SS2 HA9801, isolated from a pig with septicemia in Jiangsu Province, is a well-characterized type 2 strain and was kindly provided by Dr. Chengping Lu at Nanjing Agriculture University, Jiangsu, China [3,16,34,36]. Unless otherwise specified, all strains were routinely grown in brain heart infusion (BHI; Difco, UK) broth at 37℃. Spectinomycin (Spc) at 50 µg/ml, chloramphenicol (Cm) at 8 µg/ml, kanamycin at 50 µg/ml, and ampicillin at 100 µg/ml for E. coli as well as Spc at 100 µg/ml and Cm at 4 µg/ml for S. suis type 2 were used, where necessary, in vector construction and mutant screening. The human laryngeal epithelial cell line HEp-2 was cultured in 1640 medium supplemented with 10% fetal bovine serum (FBS; Gibco, USA) at 37℃ and 5% CO2.

For heat stress, the wild-type HA9801 strain or its mutants were first cultured at 37℃ until OD600 of 0.35. The cultures were then divided into two equal portions, one shifted to 42℃ and the other kept at 37℃. They were incubated for three additional hours. All cultures were washed and adjusted to the same OD600 for adhesion assay, ELISA, and preparation of cell lysates for western blotting.

Preparation of Recombinant Proteins and Hyperimmune Sera

The region of the DnaK operon genes hrcA, grpE, dnaK, and dnaJ in the genome of S. suis type 2 is shown in Fig. 1. The ORFs of these genes and gapdh encoding control protein glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were amplified by PCR from the HA9801 genomic DNA using primer pairs (with BamHI or EcoRI and XhoI) listed in Table S1. The PCR products were digested with corresponding restriction enzymes, ligated into the pET-30a expression vector, and transformed into E. coli BL21. The transformants were selected on LB agar with 50 µg/ml kanamycin. Positive clones were obtained by resistant phenotype and verified by PCR and sequencing. E. coli strains containing the recombinant plasmids were inoculated into individual flasks with LB broth. When the cultures reached 0.6 at OD600, isopropyl-β-D-thiogalactopyranoside (Sangon, China) was added at the final concentration of 1 mM and growth was continued at 37℃ for 4 h. The cultures were then pelleted by centrifugation and resuspended with appropriate volumes of 50 mM phosphate buffered saline (PBS) at pH 7.2. The bacterial suspensions were subjected to ultrasonic treatment to release the target proteins, which were eventually purified on a Ni-NTA agarose column as described previously [2]. Hyperimmune sera to each of the recombinant proteins or to the whole bacterial cells were obtained from New Zealand White rabbits after four times of subcutaneous immunization at a 2-week interval with 200 µg of purified recombinant proteins or inactivated whole bacterial cells (1010 of S. suis type 2 JX0811) mixed with Freund’s complete or incomplete adjuvant. All hyperimmune sera had titers of more than 1:10,000 as measured by ELISA. Preimmune serum samples were collected as negative control. The animal experiment was approved by the Laboratory Animal Management Committee of Zhejiang University (Approval No. 20120232).

Fig. 1.Organization of the DnaK operon gene region in the genome of S. suis type 2. Gene positions shown are relative to the European strain P1/7 (SSU0278-SSU0281) from blood culture from a pig dying with meningitis.

Construction and Identification of S. suis Type 2 Mutants

The deletion mutants of hrcA, grpE, dnaK, dnaJ, or the entire dnaK cluster (hrcA to dnaJ) from strain HA9801 were constructed as previously described [30]. The screening process for the deletion mutants is depicted in Fig. S1. Briefly, the 5’ and 3’ flanking regions of each gene were amplified and ligated into the pSET4s vector using corresponding primer pairs (Table S2). Each of the resulting plasmids was transformed into the HA9801 strain by electroporation, and the transformants were grown to mid-exponential phase at 28℃ in the presence of Spc. The cultures were then diluted by 1,000-fold in fresh BHI broth without antibiotics and subjected to further incubation at 28℃ for 5 to 10 generations in order to maximize the chances of genetic replacement. In order to cure the vector, bacterial cells that showed putative double cross-over by PCR were diluted by 1,000-fold in BHI medium without antibiotics and followed by shifting the diluted suspensions to 37℃ for 5-8 rounds of overnight incubation, each subcultured by 1:1,000 dilution. At the end of the last subculture, bacterial suspensions were streaked onto nonselective BHI agar plates and incubated at 37℃. About 200 colonies were then screened for loss of the vector by spotting on BHI agar plates with and without Spc at 37℃.

Growth Assay

The wild-type strain HA9801 and its mutants were first grown for 6-8 h and their OD600 adjusted to 0.4. Each of the cultures was then diluted 1:50 in BHI, dispensed into microplate wells of 6 replicate wells per culture, and incubated at 37℃ and 42℃. Growth was monitored by measuring the OD600 at 2-h interval up to 8 h.

Preparation of Surface-Associated Proteins and Whole Bacterial Lysates

The mutants and wild-type strain cultured at 37℃ for 6-7 h were prepared for SDS-PAGE/western blotting to visualize expression of the respective proteins on the surface and in the whole bacterial cells. The wild-type strain, stressed at 42℃ for 3 h, was used to examine the DnaJ expression level upon heat stress. After incubation or stress, 50 ml of bacterial cultures was washed twice with PBS and resuspended in 1 ml of surface-associated protein extraction buffer containing 30 mM Tris-HCl (pH 7.5), 2 mM MgCl2, 25% sucrose, lysozyme (3 mg/ml), and 5 mM phenylmethanesulfonyl fluoride (Beyotime, China), similar to the surface separation buffer described previously [2]. The mixtures were then incubated at 37℃ for 1.5 h and the supernatant samples (surface-associated proteins, SAP) and the precipitate samples (intracellular proteins, ICP) were then collected separately by centrifugation at 12,000 rpm and 4℃ for 10 min and stored at -80℃ for later use. The whole bacterial lysates was obtained by homogenizing the above mixtures in a Precellys 24 Homoginizer (Bertin, France) in the presence of 0.1 mm ceramic beads at the speed of 6,000 rpm for 5 times, each lasting for 30 sec. The supernatant samples were collected after centrifugation and stored at -80℃ for later use.

ELISA

To assay the abundance of DnaK operon proteins on the bacterial surface, 100 µl of live cells of strains HA9801 and JX0811 at the mid-log phase (109 CFU/ml) in 50 mM carbonate-bicarbonate buffer (pH 9.6) was coated onto the microplate wells and incubated at 4℃ overnight. After blocking for 2 h with 10 mM PBS containing 0.1% Tween-20 (PBST) and 5% non-fat milk, each of the antisera to the recombinant DnaK operon proteins (1:10 dilution in 0.5% non-fat milk PBST) was added to the wells as the primary antibody (6 wells/antibody) and incubated at 37℃ for 2 h. The antisera to the whole SS2 bacterial cells and pre-immune sera at the same dilution were used as positive and negative controls, respectively. The HRP-conjugated anti-rabbit IgG antibody (MultiSciences, China) was added and incubated at 37℃ for 1 h. The plate was then washed and 100 µl of 3,3’,5,5’-tetramethylbenzidine reagent per well was added. Color development proceeded for 10 min and was stopped by adding 100 µl of 2 M H2SO4. Absorbance was measured at 450 nm.

Quantitative PCR

Real-time quantitative PCR (qPCR) was used to further identify if the target genes in the putative mutants were down-regulated, using the primers listed in Table S3. Total RNA was extracted from the logarithmic phase culture (OD600nm 0.4) from each mutant using the RNAsimple total RNA kit (Tiangen Biotech, China). The purity and quantity of RNA were measured using the Nanodrop spectrophotometer (Thermo, USA). Only intact RNA samples with a 260:280 nm ratio between 2.0 and 2.2 were used for cDNA synthesis.

The M-MLV First-Strand cDNA synthesis kit (Promega, USA) was used to synthesize the first-strand cDNA from total RNAs. qPCR was conducted to examine the relative mRNA expression of the target genes with 16S rRNA serving as control. The PCR mixtures were made up of 7.5 µl of SYBR green mix (Bio-Rad, USA), 0.5 µl of 10 µM primers (Table S3), and 0.5 µl of cDNA, and subjected to 95℃ pre-denaturation for 3 min, followed by 40 cycles at 95℃ for 10 sec, optimal annealing temperature for each gene (hrcA 55℃, grpE 55℃, dnaK 59.5℃, dnaJ 55℃, and 16S-RNA 57.6℃) for 10 sec, and 72℃ for 10-15 sec in the iQ5 real-time multicolor PCR detector (Bio-Rad, USA). All reactions were conducted in triplicate and the target gene mRNA expression in each mutant was calculated using the 2 -∆∆Ct method [21] and shown as fold-changes relative to the parent strain HA9801 (mean ± SD).

Western Blotting Assay

The total protein content of the whole bacterial lysate or its mutants as well as their SAP and ICP extractions were quantified using the Bradford method. The protein samples (10 µg each) were subjected to SDS-PAGE on 10% gels and transferred onto PVDF membranes (Millipore, USA). After blocking with Tris-buffered saline containing 0.1% Tween 20 and 5% non-fat milk, the membranes were incubated with each of the antisera to recombinant proteins and HRP-conjugated anti-rabbit IgG (MultiSciences, China) and followed by detection with the chemiluminescence substrate (Themo, USA).

Adhesion Assay

Adhesion of S. suis type 2 strain HA9801 or its mutants to HEp-2 cells was determined as previously described [16] with some modifications. Mid-log phase bacteria (107 CFU/well) were added onto HEp-2 cell monolayers grown in 24-well culture plates at a multiplicity of infection of 100:1. The plates were centrifuged at 800 ×g for 10 min and then incubated for 1 h at 37℃ and 5% CO2. The wells were then washed four times with PBS and the HEp-2 cells were disrupted by repeated pipetting in 1 ml of sterile distilled water. The cell lysates were collected, vortex-mixed for 1 min to release all bacterial cells, and serially diluted in PBS. Appropriate dilutions were plated on BHI agar plates for enumeration of viable bacteria. Percent adhesion was difined as (CFUAdh/CFUTotal) × 100% [31].

To evaluate inhibition of S. suis type 2 adhesion to the cells by antisera to the recombinant DnaK operon proteins, the strain HA9801 cells (108 CFU/ml) were pre-incubated with each antiserum (1:10 dilution) or with antiserum to the whole bacterial cells (positive control) and pre-immune serum (negative control) for 30 min at 37℃. The remaining procedures for adhesion assay were the same as aforementioned.

Confocal Microscopy

Cell monolayers grown on coverslips were infected with strain HA9801 or JX0811 for 1 h at 37℃ and 5% CO2. The cells were then fixed with 5% paraformaldehyde for 30 min at 37℃ and permeabilized with 0.2% Triton X-100 for 15 min at room temperature. Bacterial cells were stained with hyperimmune sera to the whole S. suis type 2 cells and Alexa Flour 488-conjugated anti-rabbit IgG (Invitrogen, USA). The actin filaments were stained with Alexa Flour 568 phalloidin (Invitrogen, USA) and the nucleus with 4,6-diamino-2-phenyl indole (Invitrogen, USA). After final washing, all coverslips were mounted onto the slides, sealed with nail-polish, and observed under an FV1000-IX81 laser confocal microscope (Olympus, Japan).

The presence of DnaK operon proteins on the bacterial surface was examined according to a previous method [24]. Mid-log phase strain HA9801 cells were washed twice with PBS and adjusted to 0.4 at OD600 . The bacterial suspensions in Eppendorf tubes were spun and resuspended in the blocking agent (PBS containing 10% FBS) for 1 h at 37℃. Bacterial cells were spun and pellets resuspended in 100 µl of 1:1,000 diluted SYTO 9/DNA (Invitrogen, USA) for 10 min, followed by three washings with PBS. The DnaK operon proteins and GAPDH (as positive control of surface protein) [1,35] were probed by specific hyperimmune rabbit sera (1:10) and Alexa Flour 568-conjugated goat anti-rabbit IgG (Invitrogen, USA). Stained bacterial cells were spotted on glass slides and observed using the confocal microscope.

 

Results

The Two S. suis Type 2 Strains Differed in Their Adhesion to Cultured Cells

S. suis type 2 strain HA9801 was significantly of higher adhesion to HEp-2 cells than the strain JX0811, as shown by plate counting (7.6% vs. 0.2%, p < 0.01) (Fig. 2A). This was supported by immunofluorescence staining of more adhered HA9801 bacteria to the cells than JX0811 (Fig. 2B).

Fig. 2.Adhesion to HEp-2 cell monolayers of S. suis type 2 strains HA9801 and JX0811 as assessed by plate counting (A) and immunofluorescence probed with hyperimmune sera to whole bacterial cells (B). Each data set on panel A is shown as the mean ± SD of three independent experiments. The two-asterisk marks mean statistical difference at p < 0.01 between the means of the two strains using a two-tailed t-test [28].

Structural Characteristics of DnaK Operon Genes in S. suis Type 2 Strains HA9801 and JX0811

Putative dnaK operon genes were identified from the genomes of the SS2 strains HA9801 and JX0811 based on the known S. suis type 2 genomes (Fig. 1). BLASTP indicated that all the DnaK operon proteins of HA9801 showed identical sequence with those in the known SS2 strains, whereas there were two amino acids differences of DnaJ in JX0811 (Fig. S2). However, when we further analyzed the hydrophilicity and surface probability of the amino acid sequences of HA9801 and JX0811 by the Protean module of DNAStar, no difference was found (Fig. S3). Transmembrance helix analysis (http://www.cbs.dtu.dk/services/TMHMM-2.0/) showed that the DnaK operon proteins were all surface-associated proteins, similar to the known protein GAPDH [1,35] (Fig. S4). However, no signal peptide was found in all these proteins, including GAPDH (http:// www.predisi.de/) (Fig. S5).

Expression of DnaK Operon Proteins on Bacterial Surface and Their Role in Adhesion

The DnaK operon proteins are located on the bacterial cell surface in both strains but differ in their abundance, as shown by ELISA. Notably, DnaJ was not only the most abundant protein of the operon on the surface of HA9801, but also of higher abundance than the counterparts on strain JX0811 (OD450 2.05 vs. 1.42, p < 0.01) (Fig. 3). Immunostaining also revealed that the DnaJ protein was seen to co-localize with more bacterial cells than the other DnaK operon proteins (Fig. 4A), though it was less apparent as compared with the positive control serum to GAPDH [1]. Moreover, western blotting showed that all DnaK operon proteins were detected in the surface-associated fractions, similar to the known surface-associated protein GAPDH [1,35] (Fig. 4B).

Fig. 3.Expression of DnaK operon proteins on the surface of S. suis type 2 strains HA9801 and JX0811, as analyzed by indirect ELISA with their hyperimmune sera to corresponding recombinant DnaK operon proteins. Pooled pre-immune sera and hyperimmune sera to whole S. suis type 2 cells were used as positive and negative controls, respectively. Each data set is shown as the mean ± SD of three independent experiments. The two-asterisk marks mean statistical difference at p < 0.01 between the means of the two strains using a two-tailed t-test [28].

Fig. 4.Characterization of DnaK operon proteins on the surface of S. suis type 2 strain HA9801 by confocal immunofluorescent co-localization (A) and western blotting (B) assays. (A) The surface of HA9801 was probed by the hyperimmune sera to corresponding recombinant DnaK operon proteins and Alexa Flour 568-conjugated secondary antibody (red). Bacterial cells were stained with SYTO9/DNA (green). Pre-immune sera and phosphate-buffered saline were used as negative controls. (B) Western blotting analysis of the DnaK operon proteins in the surface-associated protein extractions of the wild-type strain HA9801 probed with hyperimmune sera to corresponding recombinant DnaK operon proteins. GAPDH probed with its polyclonal antibody was used as the positive control.

To examine the contribution of DnaK operon proteins in bacterial adhesion of HEp-2 cells, hyperimmune sera to corresponding recombinant proteins were used in the adhesion assay. Only the sera to DnaJ, but not to other proteins, were able to markedly inhibit SS2 adherence (p < 0.01, as compared with the pre-immune sera), though the inhibitory effect was less pronounced than the SS2 whole cell antisera (Fig. 5). Therefore, we consider the protein DnaJ as a contributing factor to SS2 adhesion.

Fig. 5.Inhibition of S. suis type 2 strain HA9801 adhesion to HEp-2 cells by hyperimmune sera to corresponding recombinant DnaK operon proteins or to the whole S. suis type 2 bacterial cells (positive control). Pooled pre-immune serum was used as the negative control (negative). PBS means no serum was used. Each data set is shown as the mean ± SD of three independent experiments. The two-asterisk marks mean statistical difference at p < 0.01 between the means of the target antibodies and the negative control sera using a two-tailed t-test [28].

Down-Regulation of the DnaK Operon Proteins

Unfortunately, we were not able to obtain deletion mutants lacking hrcA, grpE, dnaK, or dnaJ after five months of screening. We only got partially knocked-out (PKO) strains, since the putative colonies on Spc- agar were PCR positive for the spc gene or the target genes, although there was no visible growth of the corresponding replica on Spc+ agar on the 2nd day of incubation. Later, we found that extended incubation of the Spc+ agar plates for 3-4 days would still show small colonies, indicating that part of the bacterial population from the original colonies still possessed the recombinant plasmids. This led us to have months of repeated screening of the putative colonies on the non-antibiotic plates based on the small colonies on the replica Spc+ agar plates.

For verification of deletion of the target genes, the primer pairs on the flanking regions upstream and downstream of each target gene (named outer-flanking primers), and the primer pairs targeting the middle region of the spc gene or the target genes were used for PCR identification of DnaK family gene deletion or PKO mutants (Table S4). With the deletion mutants, the PCR products using outer-flanking primers were shorter, and no fragment of spc or of the target gene could be amplified. With PKO strains, PCR with the outer-flanking primers could generate fragments of two sizes. The long ones represented fragments from the original genome, and the short ones from the genome void of the target genes (all bands were cut and sequenced for verification) (Fig. S6A). PCR of the PKO strains with the primer pairs targeting the middle region of gene spc or the target genes could still amplify the fragment of spc or of the target genes (Figs. S6B and S6C). These results suggest that we were not able to obtain the clones of pure mutant strains even with repeated screening, and that the cultures represented mixtures of deletion mutant cells and wild-type bacterial cells that might contain recombinant plasmids. This is why we could still see reduced levels of target genes mRNA (Fig. 6) and target proteins (Fig. 7), with the degree of reduction depending on the ratio of non-mutated cells to mutated ones in the population. PKO of the hrcA gene reduced its own mRNA and protein levels and did not significantly affect the level of other genes (Figs. 6 and 7), although this gene is known to negatively regulate the downstream genes [33].

Fig. 6.Relative mRNA expression of the DnaK operon protein genes in S. suis type 2 wild-type strain HA9801 and its PKO mutants. The results are presented as fold-changes relative to wild-type strain HA9801. Each datum set is the mean ± SD (n = 3) with p < 0.05 (*) or p < 0.01 (**).

Fig. 7.Western blotting analysis of the DnaK operon proteins in whole cell lysates of S. suis type 2 wild-type strain HA9801 and its PKO mutants probed with hyperimmune sera to corresponding recombinant DnaK operon proteins. GAPDH probed with its polyclonal antibody was used as the control.

Down-Regulation of DnaK Operon Proteins Led to Reduced Cell Adhesion

Adhesion to HEp-2 cells was compared between the wild-type and the PKO strains under the same conditions. Adhesion of the PKO strains of dnaJ and hrcA-dnaJ was only half of the wild-type strain HA9801 (p < 0.05 and p < 0.01, respectively) (Fig. 8). The dnaK PKO strain also showed significant reduction of adherence, as compared with the wild-type strain (p < 0.05) (Fig. 8).

Fig. 8.Adhesion to HEp-2 cell monolayers of S. suis type 2 wild-type strain HA9801 (wt) and its PKO mutants by plate counting (A) and confocal immunofluorescence (B) assays. Significant difference (*, p < 0.05; **, p < 0.01) of the mean values of three independent experiments between the wild-type strain and the PKO mutant was determined with a two-tailed t-test [28].

Down-Regulation of DnaJ Led to Reduced Growth at Elevated Temperature

In other bacterial species such as Streptococcus intermedius and Synechocystis sp., complete deletion of dnaK or dnaJ could result in growth defects [4,32]. Therefore, we compared the growth of the PKO strains at 37℃ and 42℃. The wild-type strain and the hrcA, grpE, and dnaK PKO strains showed better growth at 42℃ than at 37℃. Only the dnaJ PKO strain showed apparent growth defect at both temperatures in comparison with the wild-type strain, and its growth at 42℃ was even inferior to that of the wild-type strain at 37℃, indicating that DnaJ might play a major role in SS2 growth at elevated temperature (Fig. 9). The hrcA-dnaJ PKO strain also displayed a certain degree of reduced growth at 42℃.

Fig. 9.Growth of S. suis type 2 wild-type strain HA9801 (wt) and its PKO mutants under normal (37℃) or heat stress condition (42℃) by measuring OD600. The experiment was repeated three times with similar trends, and data representing one typical experiment are shown.

Heat-Stress Up-Regulated DnaJ Expression and Increased Adhesion to HEp-2 Cells

To examine if heat stress at 42℃ increases DnaJ expression and affects SS2 adhesion to cultured cells, we subjected the wild-type strain HA9801 to both temperatures at 37℃ and 42℃ for comparison of DnaJ expression. Three hours of heat stress could significantly induce DnaJ expression in the whole bacterial lysate (western blotting, Fig. 10A) and on the bacterial surface (ELISA, Fig. 10B). The heat-stressed HA9801 strain also showed increased adhesion to HEp-2 cell monolayers (Fig. 11A), which was inhibitable by specific antiserum (Fig. 11B), indicating that DnaJ could play a role in SS2 interaction to host cells.

Fig. 10.Effects of heat stress at 42℃ on DnaJ expression of the whole S. suis type 2 strain HA9801 (A) by western blotting probed with hyperimmune sera to recombinant DnaJ and GAPDH proteins, and by indirect ELISA (B) with anti-DnaJ antibody. The experiment of Panel B was repeated three times with similar trends, and data representing one typical experiment, each having three replicate wells, are shown (**, p < 0.01 as compared with 37℃ control).

Fig. 11.Heat stress at 42℃ for 3 h increased adhesion of S. suis type 2 strain HA9801 to HEp-2 cell monolayers (A), which was inhibitable by specific antiserum (B). Data in panel A are shown from two independent experiments, each having three replicate wells, and those in panel B are from three independent experiments (*, p < 0.05; **, p < 0.01). Negative means the pre-immune control serum.

 

Discussion

DnaK operon proteins are multifunctional chaperones involved in protein folding and transport in prokaryotes like E. coli [13]. However, their distribution over the bacterial surface and their functions other than protein folding are largely unknown. Knockout of the dnaK-dnaJ operon in Salmonella enterica serovar Typhimurium affected growth, and the mutant strain did not replicate in cultured macrophages and failed to colonize mice [29]. In Campylobacter jejuni, inactivation of the dnaJ gene affected its colonization to chicken gut [17]. A recent study showed that DnaK could be a surface molecule participating in adhesion of S. suis type 2 to cultured cells [2]. Here, we provide clear evidence that DnaJ is the most abundant protein of the DnaK operon in S. suis type 2 strain HA9801 on the bacterial surface that is involved in adhesion to HEp-2 cell monolayers and responsive to heat stress.

Our preliminary experiments showed that the major difference of abundance of the DnaK operon proteins on the bacterial surface was DnaJ, much higher on the strain HA9801 than JX0811. As a molecular co-chaperone, DnaJ protein was distributed mainly in the intracellular proteins, but some were present on the surface of SS2 (Fig. S7), which led us to postulate that DnaJ could be an important molecule in interaction to the host cells. Adhesion inhibition assay indicated that hyperimmune sera to DnaJ, but not those to other proteins of the operon, could markedly reduce SS2 adhesion to HEp-2 cells. The dnaJ PKO strain resulted not only in decreased growth in the medium at both temperatures of 37℃ and 42℃ but also in reduced adhesion to cultured cell monolayers. The wild-type strain subjected to heat stress at 42℃ had increased expression of DnaJ on its cell surface and elevated adhesion to HEp-2 cells. These results indicate that DnaJ in SS2 might play a role in adhesion or host colonization, similar to Salmonella [29] or C. jejuni [17]. In contrast to the previous study [2], we did not see apparent involvement of DnaK in adhesion. This may be due to strain-dependent difference of its abundance on the bacterial surface, because this protein is present in less amount on strain HA9801 than on JX0811 (Fig. 3).

The other proteins in the operon (HrcA and GrpE), which had similar or lower expression in HA9801 than in JX0811, did not show significant growth defects upon thermal upshift when their genes were partially knocked out. HrcA is a negative regulation factor for the dnaK operon in several gram-negative and gram-positive bacteria such as S. mutans [19] or C. jejuni [14]. We did not see significantly increased expression of DnaK, DnaJ, and GrpE in the hrcA partially knock-out mutant, although there is a consensus CIRCE element preceding the +113 bp region upstream of the cluster [14]. There was also no change in the growth profiles of the mutant strain at both temperatures of 37℃ and 42℃, as compared with the wild-type strain. In C. jejuni, deletion of hrcA did not affect the expression of DnaK or its survival at 50℃ [14]. These results seem to suggest that there might be post-transcriptional mechanisms that regulate protein translation, probably by affecting mRNA stability, as seen in Bacillus subtilis [15].

We were not able to obtain mutants with expected full deletion of the genes of the dnaK cluster. Amplification of the shorter products with outer-flanking primers suggests that the target genes were actually deleted from the chromosomal DNA (sequence-confirmed; data not shown). However, longer fragments using the same primers were also amplified, suggesting the presence of target genes in the chromosome as well. Decreased expression at the mRNA and protein levels in the PKO strains is also indicative of the knock-down of the target genes. This might indicate that the single clones on the screening plates were still mixtures of bacterial cells having the target genes knocked out and undeleted, possibly as a result of inseparability of the chained bacterial cells. Lemos et al. [20] could not recover S. mutans strains lacking dnaK or groEL by homologous recombination.

In conclusion, the present study reveals that the DnaJ of S. suis type 2 is important not only for protein folding and thermotolerance but also for adhesion to host cells. Because DnaJ expression is increased upon temperature upshift with increased exposure on the bacterial surface, the febrile conditions of the host in systemic infections could facilitate bacterial adhesion to host cells. Further work is needed to examine if DnaJ is also predominant over other DnaK operon proteins on the surface of more S. suis type 2 strains or even other serotypes of S. suis. If the protein is predominant and common in S. suis strains, it could be a good candidate as a subunit vaccine because of its good immunogenicity.