Membrane Composition Changes and Physiological Adaptation by Streptococcus mutans Signal Recognition Particle Pathway Mutants

Bacterial strains and media.The S. mutans strains used in this study were UA159 (wild type), AH333 (Δffh), AH358 (ΔscRNA), and AH353 (ΔftsY) (17). The cultures were grown in Todd-Hewitt broth with 0.3% yeast extract (THYE) at 37°C.

Competence assay.Overnight cultures were diluted 20-fold in prewarmed THYE supplemented with 5% heat-inactivated horse serum and incubated at 37°C until the optical density at 600 nm (OD₆₀₀) was approximately 0.2. One microgram of synthetic competence-stimulating peptide (30) was added, and the cultures were incubated for an additional 20 min. Following this incubation, 1 μg of a pDL289-based plasmid, pAH373 (Erm^r Kan^r), was added, and transformants were incubated for another 30 min; this was followed by addition of 0.5 ml fresh THYE supplemented with 5% heat-inactivated horse serum and incubation for another 2 h. The cultures were plated on THYE plates containing appropriate antibiotics and incubated at 37°C for 48 h. The transformation efficiency was expressed as the percentage of transformants of the mutant strain relative to the parent strain.

Isolation of RNA for microarray experiments.Total RNA was isolated from S. mutans strains UA159 and AH333 as follows. Cultures were grown in 15 ml THYE to an OD₆₀₀ of 0.4 to 0.5 and harvested by centrifugation at 4°C. Cell pellets were resuspended in 400 μl of ice-cold, diethyl pyrocarbonate-treated, sterile water, to which 800 μl of RNA Protect reagent (QIAGEN, Inc., Chatsworth, CA) was added, and the preparations were incubated for 5 min at room temperature. Samples were centrifuged at 4°C for 10 min at 10,000 × g and resuspended with 250 μl of Tris-EDTA buffer (pH 7.0), and then each suspension was vortexed and transferred to an RNase-free 1.5-ml microcentrifuge tube containing 250 μl of sterile glass beads (diameter, 0.1 mm; Biospec Products, Inc., Bartlesville, OK), 300 μl of acidic phenol (pH 4.3), and 100 μl of 1% sodium dodecyl sulfate (SDS). The mixture was homogenized with a Mini Bead Beater (Biospec Products, Inc., Bartlesville, OK) three times for 45 s and chilled on ice for 2 min after each cycle. After this the RNA extraction procedure was performed exactly as described previously (1). The resulting RNA pellet was resuspended with 50 μl of sterile, distilled water, subjected to DNase I treatment according to the manufacturer's recommendations (Ambion, Austin, TX), and then purified through an RNeasy column as recommended by the supplier (QIAGEN, Inc., Chatsworth, CA). The RNA was eluted twice from the column with 52 μl of diethyl pyrocarbonate-treated RNase-free water, and the concentration was calculated from the absorbance at 260 nm. One microgram of total RNA was used in a real-time PCR to ascertain that the sample was DNA free.

Microarray experiments and analysis. S. mutans microarray slides were generously provided by The Institute for Genomic Research (TIGR). The microarray consisted of 1,948 70-mer oligonucleotides representing 1,960 open reading frames. The full 70-mer DNA complement per spot was printed four times on the surface of each microarray slide. All microarray experiments and analyses were carried out according to protocols and software provided by TIGR (http://pfgrc.tigr.org/protocol.shtml ) as described previously (1), with slight modifications. A reference RNA that was isolated from 300 ml of UA159 cells grown in THYE to an OD₆₀₀ of 0.4 was used in every experiment. The reference RNA was purified as described above, aliquoted, and stored at −80°C. Our experimental conditions consisted of S. mutans UA159 and AH333 grown in THYE and collected at the mid-exponential phase of growth. All RNAs were purified as described above and used to generate cDNA according to the protocol provided by TIGR (http://pfgrc.tigr.org/protocols.shtml ), with the following minor modifications. The total RNA in each reaction mixture was increased to 10 μg, and the molar ratio of dTTP to 5-(3-aminoallyl)-dUTP was increased to 1:1.5. In addition, Superscript III reverse transcriptase (Invitrogen, Gaithersburg, MD) was used to increase cDNA yields. Purified UA159 and AH333 cDNAs were coupled with indocarbocyanine (Cy3)-dUTP, while reference cDNA was coupled with indodicarbocyanine (Cy5)-dUTP (Amersham Biosciences, Piscataway, NJ). Five Cy3-labeled cDNA samples originating from five different cultures of UA159 or AH333 were hybridized to the arrays along with Cy5-labeled reference cDNA, generating a total of 10 slides. Hybridizations were carried out in the dark for 18 h at 42°C in a water bath. The slides were then washed according to TIGR protocols and scanned using a GenePix scanner (Axon Instruments Inc., Union City, CA) at 532 nm (Cy3 channel) and 635 nm (Cy5 channel). The sensitivity of the photomultiplier tube was adjusted during prescanning at 33% of full power in order to obtain a Cy3/Cy5 ratio of 1:1.

S. mutans microarray data analysis.After the slides were scanned, single-channel images were saved independently. The TIGR Spotfinder software (http://www.tigr.org/software/ ) was used to determine the intensity of each spot, and after the images were overlaid, a spot grid was generated following TIGR recommendations. Intensity values were determined for each spot, and “.mev” and “.tav” files were generated. Data were normalized using LOWESS and iterative log mean centering with default settings, followed by in-slide replicate analysis using TIGR microarray data analysis software (MIDAS; http://www.tigr.org/software/ ). Spots with low intensities or with values above the saturation value were automatically discarded by the program. A statistical analysis was carried out using BRB array tools (http://linus.nci.nih.gov/BRB-ArrayTools.html ) with a cutoff P value of 0.01 for class prediction and class comparison.

Quantitative real-time RT-PCR.Quantitative real-time RT-PCR was used to validate the results of the microarray experiments. RNA was isolated as described previously (1). One microgram of RNA was used for each 20-μl RT reaction mixture. An iScript cDNA synthesis kit containing random primers (Bio-Rad, Hercules, CA) was used to generate cDNA. Real-time PCR experiments were done with triplicate 25-μl mixtures using iQ SYBR green Supermix (Bio-Rad, Hercules, CA), and experiments were carried out with three independent RNA samples. The standard curves for each specific gene were generated as described previously (1). The primers used in all real-time PCR experiments (the primers used in this study will be provided upon request) were designed using the Beacon Designer2.0 software (Premier Biosoft International, Palo Alto, CA). Eight differentially expressed genes were selected based on the microarray analysis, and their expression levels were confirmed by real-time RT-PCR. The correlation coefficients for all genes tested were ≥0.98.

Biofilm formation and analysis. S. mutans parent (UA159) and SRP mutant strains grown to the mid-exponential phase in THYE were labeled with 30 μM hexidium iodide. Following labeling, 100 μl of each cell suspension (8 × 10⁶ CFU) adjusted to an OD₆₀₀ of 0.2 in a chemically defined medium (5) supplemented with 0.5% sucrose at pH 7.0 or pH 5.0 was used to inoculate duplicate wells of a Culture Well chambered cover glass system (Grace BioLabs), and the chamber was incubated at 37°C with light shielding on a rotating platform in an anaerobic chamber for 24 to 48 h. Three sections through two areas per well of each biofilm that developed on the cover glass bottom were observed with a Bio-Rad MRC600 confocal scanning laser microscope (Kr/Ar) system with an MS plan 60 × 1.4 NA objective and reflected laser light at combined wavelengths of 488, 546, and 647 nm, and the thickness of each biofilm was measured using the NIH Image J software (http://rsb.info.nih.gov/ij/docs/intro.html ). A companion technique, grain analysis, was used to measure the total fluorescence within an image slice and was performed using the Area Calculator plugin (http://grove.ufl.edu/∼ksamn2/plugins.html #AREA) for the NIH Image J (v.1.33u) software. This analysis sums each pixel² or “grain” above a specified value of color intensity or “threshold” (in this case set at 15 of a possible 255) to generate a total value for each image slice. The total numbers of grains from the slices of a stack were compiled to obtain the total units of red fluorescence within the field examined. Since there were three fields within each of two wells for each strain, a total of six grain values were averaged to obtain a final grain value for each strain. Each experiment was performed twice.

Preparation of membrane pellets.For membrane preparation, cultures were grown in 4 liters of THYE (pH 7.0) at 37°C until the OD₆₀₀ was 0.5, and then each culture was split in half and centrifuged at 10,000 × g for 10 min at room temperature. One pellet was suspended in 2 liters of THYE (pH 7.0), and the other pellet was suspended in 2 liters of THYE (pH 5.0). The cultures were grown for another 2 h at 37°C and then harvested by centrifugation at 12,000 × g for 10 min at 4°C. Membrane-enriched protein fractions were prepared from protoplasts as described previously (6, 20, 23), with the following modifications. The bacterial cells were harvested by centrifugation at 5,000 × g for 10 min and washed with phosphate-buffered saline. The cells were resuspended in 50 ml of buffer A (20% sucrose, 20 mM Tris [pH 7.0], 10 mm MgCl₂) supplemented with a protease inhibitor cocktail (catalog no. P8465; Sigma, St. Louis, MO), and then in order to reduce cell wall protein contamination, mutanolysin (1 mg; Sigma, St. Louis, MO) and lysozyme (18 mg; Sigma, St. Louis, MO) were added and the mixture was incubated at 37°C with gentle agitation. Protoplast formation was monitored using Gram staining. Once cell wall digestion was complete (ca. 3 h), mixtures were centrifuged as described above. The protoplasts were washed once with 30 ml of buffer A, and each cell pellet was suspended in 15 ml of buffer B (10 mM Tris [pH 8.1], 50 mM MgCl₂, 10 mM glucose). The protease inhibitor cocktail was added, and the protoplasts were lysed by two passages through a French press at 13,000 lb/in². Debris and unbroken protoplasts were removed by centrifugation at 6,000 × g for 10 min at 4°C, and the supernatant was ultracentrifuged at 41,000 × g for 45 min at 4°C. The pellet was suspended in 3 ml of buffer C (10 mM Tris [pH 8.1], 50 mM NaCl, 20 mM MgCl₂) at 4°C and ultracentrifuged at 105,000 × g for 45 min at 4°C. The supernatant was decanted, and the membrane-enriched pellet was washed twice with 2 ml of buffer D (20 mM Tris [pH 7.2], 10 mM MgCl₂) and stored at −70°C until it was used.

Extraction of membrane proteins, 2-DE, and protein identification were carried out exactly as described previously (57).

Diagonal electrophoresis.SDS-polyacrylamide gel electrophoresis (PAGE) in the first and second dimensions was carried out using the method of Spelbrink et al. (44). Briefly, membrane pellets were solubilized in a loading buffer consisting of 12 mM Tris (pH 6.8), 5% glycerol, 0.04% SDS, 14 mM dithiothreitol, and 0.02% bromophenol blue at room temperature for 30 min, centrifuged for 5 min at 12,000 × g, loaded onto a 0.75-mm-thick 12% SDS-polyacrylamide gel, and electrophoresed at 140 V. Lanes were excised from the first-dimension gels and incubated for 1 h in electrophoresis buffer containing 25% trifluorethanol (TFE) to dissociate membrane protein complexes. The excised strips were washed twice with water and twice with electrophoresis buffer and then placed on top of a 1-mm-thick 12% SDS-polyacrylamide gel for electrophoretic separation in the second dimension at 140 V. Gels were stained using either Coomassie blue or SYBR ruby. Off-diagonal spots were excised and sent for identification by quantitative time of flight (QTOF) mass spectroscopy (MS)-MS as described previously (57).

2-DE comparative analysis of wild-type and SRP mutant membrane proteins.To begin to define changes in membrane composition, both to understand the adaptation process and to identify potential SRP pathway substrates, we used a proteomics approach, employing two-dimensional gel electrophoresis analysis of membrane-enriched fractions isolated from wild-type and SRP pathway mutant strains grown under nonstress and acid stress conditions (pH 5.0, 2 h). Equal amounts of protein from membrane preparation fractions were immobilized in narrow-range IPG strips (pH 4 to 7) and resolved by SDS-PAGE. Two-dimensional gel profiles for duplicate gels containing membrane protein extracts from three independent experiments were silver stained and compared. We observed discrete differences between the protein profiles of the wild-type and SRP pathway mutant strains, and we selected for identification the spots whose intensities were increased or decreased in parallel for all three mutants (Δffh, ΔscRNA, and ΔftsY) in at least two of the three replicate experiments for each strain. To provide a reference for spot migration, a representative gel for the wild-type strain grown under nonstress conditions is shown in Fig. 1. The spots whose intensities were increased or decreased similarly in the three mutant strains compared to the wild type were identified; the results are summarized in Table 1 and described below.

• Open in new tab
• Download powerpoint

FIG. 1.

Two-dimensional gel electrophoresis profile of silver-stained membrane-associated proteins from S. mutans strain UA159 grown to the mid-exponential phase without external pH stress. Spots whose intensities were consistently increased or decreased in all three SRP pathway mutants (Δffh, ΔscRNA, and ΔftsY) are indicated by arrows and were identified by QTOF MS-MS (summarized in Table 1). MW, molecular mass.

We also employed a complementary technique (44) to detect stable oligomeric protein complexes associated with the membrane that could be extracted with SDS (Fig. 2). The membrane extracts were separated first by SDS-PAGE, and this was followed by excision and incubation of each lane containing separated proteins in electrophoresis buffer supplemented with TFE to dissociate the oligomeric complexes. Each excised strip was placed on top of an SDS-PAGE gel for second-dimension separation. An examination of off-diagonal spots that were excised from UA159, AH333 (ffh), AH358 (scRNA), and AH353 (ftsY) preparations and analyzed using nanoelectrospray QTOF MS-MS revealed that all samples included enolase and glutamate dehydrogenase and that the wild-type and ffh mutant samples contained an additional protein, elongation factor Tu.

• Open in new tab
• Download powerpoint

FIG. 2.

Diagonal SDS-PAGE analysis of membrane protein extracts from UA159 stained with Coomassie blue. First-dimension strips were treated with 25% TFE to induce dissociation of protein complexes prior to separation in the second dimension. Off-diagonal proteins derived from dissociated complexes were identified by QTOF MS-MS. MW, molecular mass.

Identification of membrane-associated proteins by QTOF MS-MS.We identified spots whose intensities were consistently increased or decreased for all three SRP mutants compared to those obtained for the wild-type strain for a number of membrane-associated proteins by 2-DE following growth under nonstress and acid stress conditions (Table 1). Twenty-eight protein spots were excised from the gels, and 21 different proteins were identified by QTOF MS-MS analysis. In most cases there was agreement between predicted and measured pI and molecular mass data; one exception was spot 20, for which the observed pI (pI 6.1) and the calculated pI (pI 5.5) were different. This protein spot corresponded to a conserved hypothetical protein predicted to be a transcriptional regulator. The observed molecular mass for spot 2, corresponding to biotin carboxyl carrier protein (BCCP), was 25 kDa, while the calculated mass was 15.9 kDa. The observed and calculated pIs of this protein were similar. Five other proteins (spots 4, 5, 7, 20, and 22) were identified as isoforms of one another, most likely due to posttranslational modifications. Generally, each protein spot matched one protein in the S. mutans database; the exception was spot 9 (Fig. 1), which contained more than one protein (enolase and the β subunit of F₁F₀ ATPase). The proteins that were identified were classified into different groups on the basis of their functional activities and included proteins involved in transcription or translation, chaperones, proteases, metabolic enzymes, and stress tolerance proteins, as well as conserved hypothetical proteins (Table 1). There was notable segregation of increased and decreased spot intensities within these categories; the mutants had increased levels of membrane-associated chaperones and a protease but decreased levels of metabolic enzymes, general stress response proteins, and transcription and translation factors.

Transcriptome analysis of the ffh mutant strain.The SRP pathway plays a fundamental role in cotranslational protein translocation in all living cells; therefore, surviving in the absence of this pathway would be expected to involve a complex process of physiological adaptation. The DNA microarray technique was used to identify global changes in gene expression in a representative SRP pathway mutant, AH333 (Δffh), compared to the gene expression in wild-type strain S. mutans UA159. The transcriptome analysis of bacterial cells grown under nonstress conditions revealed that expression of 81 genes was increased and expression of 35 genes was decreased in the ffh mutant strain AH333 compared with the expression in parent strain UA159 (>1.5-fold difference; P < 0.01) (Table 2). The major class of differentially expressed genes comprised hypothetical and conserved hypothetical genes (27 up-regulated genes and 10 down-regulated genes). Other genes were classified into major transcriptional categories (Fig. 3). Up- and down-regulated genes appeared to partition within several of these categories; for example, 10 genes encoding proteases, chaperones, and heat shock factors were induced, 4 genes encoding competence-related proteins were repressed, 10 genes encoding ribosomal proteins and enzymes involved in amino acid and protein biosynthesis were repressed, 28 genes encoding cell envelope and transporter proteins were induced, 9 genes involved in DNA replication or repair and cellular detoxification were induced, and 4 major stress-response regulatory genes were induced.

• Open in new tab
• Download powerpoint

FIG. 3.

Genes differentially expressed in ffh mutant AH333 compared to the UA159 wild-type strain were identified by microarray analysis (P < 0.01; >1.5-fold difference). Functional categories are indicated on the x axis. TC, two-component.

Real-time RT-PCR analysis of selected genes.Microarray data were confirmed using the quantitative real-time RT-PCR technique due to its specificity and sensitivity, which was especially important for low-abundance transcripts. From a total of 116 genes identified in the microarray analysis in which highly or moderately up-regulated or down-regulated genes in the ffh mutant were observed, a set of 8 genes mediating critical functions, such as transcriptional regulation (hrcA and ciaR), the stress response (dnaK, grpE, htrA, and groEL), late competence (comYA), and transport of osmotic stabilizers (opuCA), were selected for real-time RT-PCR validation. The levels of induction or repression of expression compared to the expression in the parent strain UA159 standard were calculated, and the results corroborated the increase or decrease in transcription of each of these genes initially indicated by the microarray data (Table 2). The levels of induction of differentially expressed genes that encode chaperones and proteases and the levels of repression of individual genes involved in competence, as assessed by microarray analysis, are also shown in Table 2 since notable changes in transcription of genes within these two broad categories were observed upon deletion of ffh (Fig. 3). Table 2 shows more comprehensive information regarding other differentially expressed genes within the transcriptional categories listed in Fig. 3 and their relative levels of expression.

Comparison of expression of F₁F₀ ATPase subunits.Acid tolerance in S. mutans is mediated in large part by the F₁F₀ ATPase proton pump (4). S. mutans SRP pathway mutants (Δffh, ΔscRNA, and ΔftsY) are all acid sensitive (14, 17), and all of them exhibit decreased membrane H⁺/ATPase activity in both the presence and the absence of acid stress (10, 17). Hence, the reduced spot intensity of the ATPase β subunit detected by 2-DE in TFE-extracted protoplast-derived membrane fractions prepared from the SRP pathway mutant strains was of particular interest. We showed previously that purified membrane preparations from an ffh mutant exhibited only 26% and 39% of the parental membrane H⁺/ATPase activity when pH 7.0 and pH 5.0 chemostat-grown cells, respectively, were assayed (10); however, nearly identical activities were observed for decryptified whole cells of the parent and mutant strains at each pH, and the activity in both strains increased ∼65% at pH 5.0, which is typical of an acid tolerance response (4, 24). Membrane-integral F₀ subunit c is a known substrate of the SRP pathway in E. coli (50). The decreased spot intensity of the membrane-proximal F₁ β subunit therefore more likely represents diminished membrane localization than decreased expression. To confirm that disruption of a functional SRP pathway does not result in altered levels of expression of proton ATPase components, real-time RT-PCR was used to measure transcription of genes encoding the β and c subunits in ffh mutant strain AH333, and the results were compared to the data for wild-type strain UA159. No substantial differences in the levels of message for either component were detected for these strains. The numbers of message copies/ng total RNA for the mutant and the wild type were 7.6 × 10⁴ and 9.5 × 10⁴ for the ATPase c subunit and 1.1 × 10⁵ and 1.2 × 10⁵ for the ATPase β subunit, respectively.

Assessment of genetic competence in SRP pathway mutants.Analysis of the S. mutans UA159 transcriptome revealed a cluster of four genes involved in late competence whose expression was significantly reduced in the ffh mutant compared to the expression in the wild-type strain (Table 2). To determine the extent of impairment of natural transformation in SRP pathway mutant strains, the transformation efficiency was determined at the same growth stage for all strains. Compared with the transformation frequency of the parent strain UA159, the transformation frequency of ftsY deletion strain AH353 was reduced 25-fold and the transformation frequency of ffh mutant strain AH333 was reduced >140-fold (data not shown). Interestingly, the scRNA deletion in AH358 had substantially less effect (1.25-fold decrease) on the ability of this mutant strain to take up foreign plasmid DNA, suggesting that the genetic competence apparatus of S. mutans is ffh and ftsY dependent but relatively scRNA independent. In other organisms Ffh and scRNA are known to represent components of the signal recognition particle itself; therefore, it is not clear why the scRNA mutant has a different phenotype. Transcriptional analyses of ftsY and scRNA mutants have yet to be performed, so it is not known whether com gene expression in these strains is altered in a manner similar to that in the ffh mutant. Exogenous addition of synthetic competence-stimulating peptide did not improve the transformation frequency of any of the SRP pathway mutant strains.

Biofilm formation by SRP mutant strains.As determined by 2-DE, none of the SRP pathway mutants contained membrane-associated autoinducer-2 production protein LuxS, a key molecule contributing to quorum sensing (3) with a demonstrated link to biofilm formation (34, 54). In addition, the Δffh mutant tested in the microarray analysis exhibited decreased expression of multiple genes involved in genetic competence, another factor known to impact biofilm formation in S. mutans (31). Therefore, we predicted that SRP pathway mutants would show an altered capacity to form biofilms. Biofilm formation was tested using growth in a semisynthetic medium on a glass surface. Deletion of each of the components of the SRP pathway resulted in significant effects on biofilm formation. Thinner biofilms (71 to 78% of the parent strain UA159 thickness) were obtained for all mutant strains at 48 h in the absence of an exogenously applied environmental stressor (Fig. 4 and Table 3). Grain analysis, a companion technique, confirmed these results. Grain analysis measures the total fluorescence in a biofilm image expressed as pixels² (grains) and can help distinguish thinner but more compact biofilms from biofilms that are thinner due to the accumulation of fewer cells. Under acid shock conditions (pH 5.0), UA159 produced a good biofilm by 36 h, while each of the SRP pathway mutants produced only one-third to one-half as much. Again, the grain analysis results supported the finding that the decreases in biofilm thickness represented decreases in cell numbers.

• Open in new tab
• Download powerpoint

FIG. 4.

Representative images of biofilm formation by hexidium-iodide labeled S. mutans strain UA159 and SRP pathway mutant strains. Cells were grown in chambers with glass bottoms for 48 h without external pH stress (A) or for 36 h under acid shock conditions (pH 5.0) (B), and images were obtained using a confocal laser microscope. Overhead (top panels) and side (bottom panels) views are shown.

Molecular chaperones and proteases are up-regulated in SRP pathway mutants.This study in the first comparative analysis of streptococcal membrane preparations derived from wild-type and SRP pathway mutant strains under stress and nonstress conditions. Other groups of workers have described 2-DE analyses of the cytoplasmic protein pool of wild-type S. mutans (15, 28, 29, 47, 48). The data presented above also provide the first evaluation of the transcriptome of a prokaryotic mutant lacking a functional SRP pathway. The combined proteomic analysis of membrane-associated proteins and global gene expression profiles comparing the wild-type and an ffh mutant strain begin the process of identifying potential SRP substrates and revealing mechanisms of physiological adaptation that enable S. mutans to survive without a major protein translocation system. The levels of expression of several chaperones were increased in the SRP pathway mutants. The levels of induction of genes encoding chaperones and proteases in the ffh mutant are summarized in Table 2, which includes data for the genes encoding major chaperones GroEL, GroES, and DnaK, as well as their cofactors, DnaJ and GrpE. Expression of these genes was increased up to 5.5-fold under nonstress conditions. Furthermore, proteomic analysis of cell-free membranes from ffh, scRNA, and ftsY mutants revealed increased levels of DnaK and GroES under acid stress and nonstress growth conditions. Chaperones prevent aggregation of unfolded proteins and assist in the correct refolding of denatured polypeptides. Len et al. (28, 29) found that the levels of the DnaK protein were reduced in the extracellular fraction following acid shock of wild-type S. mutans, while the DnaK and GroEL levels were increased in the cytoplasmic fractions under stress conditions, suggesting that differences in translocation, as well as overall expression levels, may vary under different environmental conditions. The levels of several proteases were also elevated in SRP pathway mutants. Higher levels of the ATP-dependent protease ClpP were detected in mutant membrane preparations, and microarray data indicated that transcription of genes encoding HtrA, ClpP, and ClpC was significantly increased in the absence of Ffh. HtrA, a member of the DegP-DegS-DegQ family of serine proteases localized on the cell surface (9, 22, 39), functions as both a molecular chaperone and a protease, and the switch from one function to the other is temperature dependent. Elevated levels of chaperones and proteases in the absence of the SRP pathway suggest that there is a mechanism that protects against untranslocated or misfolded proteins, as well as possible compensatory roles in protein translocation (17).

The levels of all chaperones were not elevated. With less stringent requirements (P = 0.0117) the microarray data revealed that there was actually a reduction in the level of the ropA-encoded trigger factor in the Δffh mutant (Table 2). This ribosome-associated protein has been reported to function as a chaperone during translocation to eliminate misfolding or aggregation of newly synthesized proteins destined for secretion (49). In Streptococcus pyogenes, trigger factor exhibits peptidyl-prolyl cis-trans-isomerase activity essential for secretion and maturation of the cysteine protease (33), and Wen et al. (55) demonstrated that in S. mutans RopA has an important role during the global stress response, such as acid tolerance, competence development, and biofilm formation.

Protein synthesis machinery is down-regulated in SRP pathway mutants.In membrane preparations of SRP mutants grown under nonstress or acid stress conditions the levels of several proteins involved in transcription and/or translation were decreased or the proteins were absent (Table 1). Since there was no evidence of general protein degradation in the membrane preparations, the decreased levels of specific proteins involved in protein synthesis were of interest. The results of transcriptome analysis of the ffh mutant supported the proteomics data and indicated that there was down-regulation of genes encoding ribosomal proteins and enzymes involved in amino acid and protein biosynthesis. These findings suggest that SRP pathway mutants adapt by modulating the translation machinery. Apparently, in the absence of a functional translocation pathway, protein synthesis is reduced, circumventing accumulation and aggregation of undelivered membrane-targeted or secreted proteins. The repression of key ribosomal proteins and amino acid biosynthetic enzymes suggests that metabolic and biosynthetic functions are slowed down in the impaired cells. These data are supported by the fact that the growth rates of SRP mutants are lower (17), and they are consistent with the results of Mutka and Walter (37), who studied the transcriptome of conditional ffh mutants of Saccharomyces cerevisiae. This is the only known example of a eukaryotic organism able to survive with mutations in the SRP pathway (16). The mutants were slow growing and adapted physiologically by repression of protein synthesis, upregulation of chaperone biosynthesis, and down-regulation of energy-producing pathways.

Potential metabolic changes associated with loss of the SRP pathway.The levels in membrane fractions of the glycolytic enzymes fructose-1,6-bisphosphate aldolase, phosphoglycerate kinase, and glucose-6-phosphate isomerase, as well as adenylate kinase, HPr, and BCCP, were reduced in all SRP pathway mutants grown under nonstress conditions, or the enzymes were absent; however, when the cultures were subjected to a sudden acid challenge (pH 5.0), the levels of fructose-1,6-bisphosphate aldolase, phosphoglycerate kinase, glucose-6-phosphate isomerase, and adenylate kinase did not change, and HPr and BCCP were not detected in membrane fractions of the wild-type strain or any of the mutant strains (Table 1). Since a number of glycolytic enzymes were identified in membrane preparations, it is possible that these enzymes form a functional complex associated with the cytoplasmic membrane of S. mutans, and this suggests that there is functional compartmentalization that increases the efficiency of the glycolytic pathway. It is conceivable that these enzymes associate to form a glycolytic metabolon (43) that may provide thermodynamic efficiency, as well as optimization of metabolic processes for bacteria, as suggested by Shearer et al. (43). Early work by Mowbray and Moses (36) on E. coli demonstrated that there are polyenzymic metabolons in the cytoplasm that are able to efficiently carry out glycolysis. The compartmentalization of glycolytic metabolons was compared with similar structures in muscle tissue (8, 12) and the cell membranes of red blood cells and yeast (12), and it was theorized that the glycolytic metabolon yielded ATP for immediate use in pumping protons out of cells. An alteration in the S. mutans SRP pathway mutant membrane structure could therefore contribute to acid sensitivity due to the lack of immediately available ATP.

Acid tolerance in S. mutans is mediated in large part by the F₁F₀ ATPase proton pump (4, 42). The 2-DE data described above demonstrated that the levels of the ATPase β subunit were decreased in SRP pathway mutant membrane fractions. Enolase and the ATP synthase β subunit comigrated in 2-DE gels, but similar amounts of enolase were identified by diagonal gel electrophoresis. Therefore, the reduced levels of 2-DE protein spot 9 were likely due to reduced levels of the ATPase β subunit. Real-time RT-PCR did not reveal a change in expression of either the membrane-proximal β or membrane-integral c subunit of the ATPase in the absence of Ffh. The current data are consistent with our previously reported observations comparing total cellular and membrane-associated ATPase specific activities in wild-type and SRP pathway mutant strains under nonstress and acid shock conditions (10, 17). Therefore, inefficient membrane localization of proton ATPase components remains the likely explanation for an acid-sensitive phenotype in these mutants. Microarray analysis also did not reveal significant transcriptional differences between the glycolytic pathway enzymes of the wild type and the glycolytic pathway enzymes of the ffh mutant, again suggesting that the 2-DE data for these proteins represent changes in cellular localization rather than altered expression levels.

Components of the SRP pathway influence genetic competence.Transcriptome analysis revealed a cluster of four genes involved in late competence, whose expression was significantly reduced in the ffh mutant compared to the expression in the wild-type strain (Table 2). Three of these genes, comYA, comYD, and comGF, are part of a nine-open reading frame operon, and deletion of each of the first seven genes of this operon eliminates natural transformation in S. mutans (35). The expression of genes involved in early competence, such as comCDE or comX (11), in the ffh mutant was similar to the expression in the wild-type strain, as determined by microarray analysis. A comparison of transformation efficiencies demonstrated the dramatically decreased competence of the ffh and ftsY mutants. The reason that deletion of these two genes had a more profound effect on competence than deletion of the gene encoding the scRNA had is currently unclear since other phenotypic differences, including acid and salt sensitivity and the increase in doubling time under nonstress conditions, were similar in all three mutants (17). More extensive biochemical analyses of the S. mutans SRP and its receptor may help answer this question. Addition of synthetic competence-stimulating peptide did not restore the transformation efficiency, suggesting that ComD, a member of the ComDE two-component regulatory system, may depend on the SRP pathway for proper membrane insertion. Competence-stimulating peptide binding to ComD activates transcription of the transcriptional activator ComX, and DNA uptake is mediated through a late competence system, the ComYA-G complex, whose expression is dependent on ComX (2, 25).

Biofilm formation is impaired in SRP pathway mutants.Bacterial biofilm formation is a complex multifactor process, and in S. mutans it is a key virulence property associated with colonization of the pellicle-coated tooth surface. Numerous gene products, including those involved in adherence, competence, quorum sensing, cell wall synthesis, metabolism, and the stress response, contribute to microbial biofilm formation, and numerous other genes have altered expression within sessile biofilms compared to their expression in planktonic cells (21). The oral cavity is a dynamic and harsh environment, with constant fluctuations in temperature, pH, and nutrient sources; therefore, the ability to respond quickly to stress conditions is critical. SRP pathway mutant strains adapt poorly to sudden shifts in acidic or osmotic environments, raising the question of whether they have a diminished ability to form biofilms. The decreased expression of late competence genes and the inability to detect LuxS in mutant membrane preparations further suggested that biofilm formation is impeded. The quorum-sensing autoinducer-2 protein LuxS is highly conserved in bacteria (21, 34, 46). Quorum sensing regulates important physiological functions in S. mutans, such as the ability to withstand environmental stress conditions (27, 51, 54), competence (51), biofilm formation (56), and expression of virulence factors (38). Both luxS expression (40, 53, 56) and competence gene expression (45) are now well-established factors that play interrelated and pivotal roles in biofilm development by oral streptococci. Although mutations in SRP-related genes did not completely prevent biofilm formation, deletion of SRP pathway components was associated with significant decreases in cellular accumulation under nonstress growth conditions and more profoundly under acid stress growth conditions (Fig. 4 and Table 3), suggesting that the SRP pathway has a role in optimal colonization of the tooth surface.

Taken together, the results presented in this report suggest that optimal insertion of membrane proteins mediated by the S. mutans SRP cotranslational translocation pathway contributes to processes that include pH homeostasis and cell-cell communication and facilitates biofilm formation in this prevalent oral pathogen.