• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Mol Microbiol. Author manuscript; available in PMC Dec 4, 2009.
Published in final edited form as:
PMCID: PMC2788772

Large-scale identification of serotype 4 Streptococcus pneumoniae virulence factors


Streptococcus pneumoniae (the pneumococcus) is carried in the nasopharynx of healthy individuals, but can spread to other host sites and lead to pneumonia, bacteraemia, otitis media and meningitis. Although it is logical to think a priori that differential gene expression would contribute to the ability of this pathogen to colonize different sites, in fact very few genes have been demonstrated to play tissue specific roles in virulence or carriage. Using signature-tagged mutagenesis to screen 6149 mariner-transposon insertion strains, we identified 387 mutants attenuated for infection in a murine model of pneumonia. Among these mutants are ones with disruptions in a number of putative tissue-specific transcriptional regulators, surface proteins, metabolic proteins and proteins of unknown function, most of which had not previously been associated with virulence. A subset of these, including most of those with insertions in putative transcriptional regulators, was examined for phenotypes in murine models of bacteraemia and nasopharyngeal carriage. Four classes of mutants defective in infection models of the: (I) lung, (II) lung and blood, (III) lung and nasopharynx, and (IV) all three tissues were identified, thus demonstrating the existence of tissue-specific pneumococcal virulence factors. Included in these strains were two with disruptions in a genetic locus that putatively codes for a transcriptional regulator, three surface proteins and three sortase homologues. Mutation analysis revealed that three of the seven genes in this locus are virulence factors that are specific to mucosal surfaces.


The identification of novel virulence determinants can facilitate the development of new vaccines and drug treatments, which are especially needed for organisms in which antibiotic resistance is prevalent. Antibiotic resistance is emerging rapidly in the respiratory pathogen Streptococcus pneumoniae (Tomasz, 1999), which is the causative agent of a number of diseases ranging in severity from normally benign otitis media to highly lethal meningitis. The major virulence factor required for all pneumococcal disease is the extracellular polysaccharide capsule, which protects colonizing or infecting bacteria from phagocytosis. In addition to the polysaccharide capsule, many surface exposed protein factors have been implicated in pneumococcal disease, however, knowledge of precise roles of many of these proteins in different in vivo niches is limited (Jedrzejas, 2001; McCullers and Tuomanen, 2001).

Signature-tagged mutagenesis (STM) represents one of several recently developed techniques for the identification of genes important for infection (Unsworth and Holden, 2000). Two STM screens in S. pneumoniae have been reported, one in a serotype 19 and one in a serotype 3 strain, and have identified key factors in virulence while screening only a limited number of mutants (Polissi et al., 1998; Lau et al., 2001). Here we report our findings of STM on a serotype 4 encapsulated clinical isolate. This work expands on previous studies by screening four times as many mutant strains in a murine model of pneumonia, resulting in the identification of 387 attenuated mutants. A number of the identified strains were tested for additional in vivo phenotypes in murine models of bacteraemia and nasopharyngeal carriage. Among the 387 attenuated strains are several known virulence factors of S. pneumoniae, and in some cases, we find additional roles in virulence that were previously unknown. Additionally, a large percentage of the genes identified code for hypothetical proteins or proteins of unknown function, thus greatly expanding our knowledge of the role of these proteins. Finally, we have identified a large number of putative tissue-specific regulatory factors, revealing the importance of differential gene regulation for the colonization of different host sites. Additional experiments identify one such regulator and neighbouring genes as tissue specific virulence factors. Included in these genes are three surface proteins and three sortase homologues.


Construction of mutant pools

To generate a large number of S. pneumoniae transposon insertion strains, chromosomal DNA was prepared from strain AC353, a streptomycin-resistant derivative of TIGR4 (Tettelin et al., 2001), and mutagenized by in vitro transposition with magellan2. magellan2, a mini-transposon derivative of mariner, inserts into the pneumococcal chromosome in a highly random manner (data not shown), requiring only a TA dinucleotide at the insertion site (Lampe et al., 1996). Transposon mutagenesis was performed as described (Akerley et al., 1998), except that 63 magellan2 derivatives, each containing a unique 40 basepair (bp) signature tag were used. Following transposition, mutagenized DNA was transformed into naturally competent AC353 as described (Bricker and Camilli, 1999). Approximately 100 insertion strains were sequentially collected from each of the 63 magellan2 derivatives into the wells of microtitre plates, resulting in 100 pools of 63 signature tagged insertion strains for STM screening.

Determination of colonization bottlenecks

In animal infections, the population of bacteria that can initially survive and begin to multiply is restricted to those that are able to overcome certain barriers during transit to the site of infection. This phenomenon is commonly referred to as a ‘colonization bottleneck’ (although ‘colonization’ is an inaccurate term in cases like pneumococcal pneumonia that are acute infections and of limited duration). Because STM depends on all strains in the starting inoculum having an equal opportunity to infect a particular tissue, the population dynamics of AC353 in the murine lung were analysed to determine whether a bottleneck existed. To address this, a group of female Swiss Webster adult mice were infected with a single STM pool of 63 unique strains at a dose of 105 colony-forming units (CFU) administered intranasally. At various times following inoculation, pairs of mice were euthanized and the number of CFU in the lungs from each animal was enumerated. After 12 h, the mice appeared healthy and no bacteria could be cultured from the lungs, suggesting that a severe bottleneck exists. In an attempt to circumvent this bottleneck, we increased the inoculum to 2 × 107 CFU and determined the number of CFU per mouse lung as above. The larger inoculum resulted in the successful infection of all mice, as between 104 and 107 CFU were recovered from all animals at all time points until the mice became moribund after approximately 48 h. Accordingly, all subsequent lung infection experiments were performed with an inoculum of 2 × 107 CFU.

A second variable that we assessed was the potential for a limited number of strains to out-grow all others after initial adherence, thus preventing all 63 strains from being equally represented at a late stage of infection. To test this possibility, four mice were infected with a single STM pool, and the complexity of the bacterial populations remaining in the lungs of each mouse at a late stage of infection was determined and compared. The presence or absence of each strain in the lungs was assessed by recovery of the signature tags and hybridization to a master signature tag dot blot as described in the Experimental procedures. The full input pool strain complexity was maintained in all four mice after 44 h of infection, with the exception of a few strains absent from all mice that represent bona fide attenuated strains (data not shown). Therefore, a pool complexity of 63 strains administered at 2 × 107 CFU/mouse results in all 63 strains having an equal opportunity to adhere and multiply in the mouse lung, and strains that fail to be recovered after 44 h are attenuated. Of note, the pool complexity used here is intermediate to that chosen for two prior STM screens in S. pneumoniae. Polissi et al. (1998) used a pool complexity of 50 strains, mutagenized by plasmid insertion-duplication, to infect BALB/c mice in a murine model of pneumonia. In the other study, Lau et al. (2001) used a pool complexity of 96 strains, also mutagenized by plasmid insertion-duplication, to infect CD-1 mice in murine models of pneumonia and bacteraemia.

Another potentially powerful application of STM is to track a population of bacteria that initially infect a single site, and subsequently spread to other sites in the animal. From such an analysis, it is possible to determine if a systemic infection is clonal or due to a larger founder population. If the latter was true for the case of S. pneumoniae spreading from the lung to the bloodstream, then two simultaneous STM screens could be conducted after intranasal inoculation; one in the lung and one in the blood. Instead, we found that in mice infected with the same STM pool, the population of bacteria recovered from the bloodstream is randomly composed of only a few strains from the input inoculum (data not shown).

Selection of avirulent strains

To identify pneumococcal genes essential for lung infection in mice, an STM screen was done. In total, 100 pools comprising 6149 strains were screened. Each pool was inoculated into two mice and the bacteria were recovered after 44 h by plating homogenized lung tissue from each animal on Tryptic-Soy Agar blood plates. Chromosomal DNA was purified from the combined outputs from each animal and used as template DNA for the PCR amplification of the signature tags as described in Experimental procedures. A similar procedure was followed to obtain signature tags from the input population of bacteria. The amplified signature tags were used to probe nitrocellulose dot blots containing all 63 tags and attenuated strains were identified by visually examining output blots for spots exhibiting a decreased hybridization signal compared to the input blot.

In the primary (1°) round of screening, 2101 candidate attenuated strains were identified in a non-stringent manner, i.e. all strains that gave a noticeably reduced signal on the output blot compared to the input blot were selected. A more stringent secondary (2°) screen was then done on 2080 of these candidate attenuated strains. For this, the 2080 strains were assembled into smaller 2° pools of 40 strains and each pool was inoculated intranasally into two mice at 2 × 107 CFU. After amplifying the signature tags from the input and output bacteria, and hybridization to the master dot blot membranes, 1265 strains were selected that had a highly reduced output signal relative to the input signal. These virulence attenuated strains represent 20% of the total number of strains initially screened.

In order to narrow the focus of study, a re-examination of the 2° screen dot blot films was done to identify the subset of strains that were highly attenuated as determined by lack of any hybridization signal on the output blots. Through this analysis, 387 strains were identified, representing 6.3% of the total strains screened. The sites of transposon insertion in 337 of the 387 strains were determined by arbitrary-primed PCR and DNA sequencing of the magellan2/genome junctions essentially as described (Merrell et al., 2002). Table 1 lists these 337 strains, along with information on the gene disrupted in each and a functional classification based on the TIGR4 genome sequence release (Tettelin et al., 2001).

Table 1
Streptococcus pneumoniae genes essential for lung infection.

Quantification of virulence defects of selected mutants

To validate the results of our STM screen and to quantify the degree of virulence attenuation of individual strains, competition assays were done. The transposon insertion mutations from 17 of the 337 highly attenuated strains, including 12 strains with disruptions in putative transcriptional regulators, were backcrossed into the wild-type strain and tested by competition assay as follows. Each of the mutant strains was mixed with the wild-type strain at a 1:1 ratio, and inoculated intranasally into four or more mice and simultaneously into Todd Hewitt-Yeast extract (THY) broth. Bacteria were enumerated from the lungs at 44 h and after 5 h from THY broth by plating serial dilutions on media selective for both wild-type and test strains, and then replica plating the colonies to media selective for only the test strain. The in vivo competitive index (CI) was calculated by dividing the ratio of mutant to wild-type bacteria recovered from the lungs by the ratio of mutant to wild-type bacteria that were inoculated into each animal. Similarly, the in vitro CI was calculated using THY broth cultures in order to assess general growth defects. The geometric means of the CIs for each strain are listed in Table 2; a mean CI of less than 1 indicates a defect in virulence (or growth in vitro) of the test strain. Of the 17 strains examined, 16 were attenuated for lung infection. Furthermore, 13 were out-competed by greater than 10-fold, confirming that the selected strains are highly attenuated when tested against the wild-type parental strain. None of the 17 strains tested suffered gross defects in multiplication in broth in vitro. This analysis indicates that the majority of the 387 mutant strains identified by two successive rounds of STM screening are reproducibly attenuated in competition assays. Hence, the genes disrupted or whose expression is affected by the transposon insertion in these strains should be considered bona fide virulence factors.

Table 2
Competition analysis of virulence gene tissue specificity of selected mutants identified by STM.

Determination of virulence phenotypes in other infection models

After confirming the attenuation of virulence of the selected strains in the murine lung, we sought to determine if these strains have global virulence defects, or if they could be categorized into classes based on in vivo phenotypes in other animal assays. To this end, most of the set of confirmed lung attenuated strains, and many additional strains, were tested in competition assays in murine models of bacteraemia and nasopharyngeal carriage. For each animal model, mutant and wild-type bacteria were prepared exactly as described for the lung infections, however, different inoculum sizes were utilized to assure proper representation of each strain. For the bacteraemia model of infection, mice were inoculated with 106 CFU by intraperitoneal injection (i.p). Alternatively, 108 CFU were inoculated intranasally into mice using a small inoculum volume for the nasopharyngeal carriage model. After 20 h for bacteraemia and 7 days for nasopharyngeal carriage, the bacteria were recovered from blood or nasopharyngeal washes, respectively, and CIs were determined as described above.

Of the 24 strains that were tested in the bacteraemia model, half were attenuated, albeit to varying degrees (Table 2). Four strains, two with insertions in transcriptional regulators (STM119 and STM210) and two with insertions in biosynthetic genes (STM4 and STM208), had severe virulence defects of greater than 40-fold. Of the remaining eight attenuated strains, three had intermediate defects (14 to 19-fold) and five were only slightly attenuated. The remaining 12 strains that were tested against the parental strain were not attenuated. Together, these data show that we have identified a set of tissue specific virulence factors, including several putative transcriptional regulators.

Fourteen strains that were attenuated for lung infection were tested for their ability to colonize the nasopharynx in competition with the wild-type strain. As in the bacteraemia model, not all of the strains that were attenuated for lung infection were attenuated for colonization of the nasopharynx. Ten of the tested strains were deficient at colonizing the nasopharynx, and all 10 exhibited greater than a 10-fold colonization defect. Interestingly, these 10 strains had differing phenotypes in lung infection and bacteraemia. Six were attenuated in all three of the animal models (Class IV), including two mutants in a putative transcriptional regulators (STM38 and STM210) and a third in the response regulator of a two-component signal transduction system (STM185). The remaining four strains were not attenuated when tested in the bacteraemia model, but were each severely out-competed by the wild-type strain in the two animal models of infection that involve interactions with mucosal surfaces.

Identification of rlrA and a sortase homologue required for infection

Of the transcriptional regulators identified by STM and tested in additional animal models, one (STM64) putatively codes for a protein with 49% similarity to RofA and Nra from S. pyogenes (Fogg et al., 1994; Podbielski et al., 1999). This mutant strain was out-competed by the parental strain in both the pneumonia and nasopharyngeal carriage models, but not the bacteraemia model. A greater virulence defect was observed in the nasopharynx, where the rlrA strain was out-competed 14-fold (Table 2). These findings suggest that RlrA regulates one or more genes that are important for the interaction of S. pneumoniae with mucosal surfaces in the respiratory tract.

In some strains of S. pyogenes, rofA regulates the expression of a divergently transcribed gene coding for Protein F, a factor that mediates attachment to fibronectin (Fogg et al., 1994). The pneumococcal rofA homologue, herein named rlrA, for rofA-like regulator, is divergently transcribed from six genes (SP0462 to SP0468), three of which have very weak homology to microbial surface components recognizing adhesive matrix molecules (MSCRAMMs), and three that have homology to sortases (Fig. 1). Sortases are enzymes that catalyse the covalent linkage of a family of secreted proteins that contain an LPXTG motif to the bacterial cell wall (Mazmanian et al., 2001). Although these three S. pneumoniae sortases were apparent from the genome sequence (Pallen et al., 2001; Tettelin et al., 2001), no experimental data to characterize these has been reported. In work to be reported elsewhere, we find that RlrA regulates the transcription of these six genes, and thus we have named the three putative MSCRAMMs rrgA, rrgB, and rrgC for RlrA-regulated gene. In addition, based on strong homology to sortases, the existence of a fourth sortase elsewhere in the chromosome, and our results below, we herein name the three flanking genes srtB. srtC, srtD. Interestingly, one of the mutants identified in the STM screen mapped to srtD (STM65 in Table 1), the terminal gene in the locus.

Fig. 1
A. Schematic representation of the S. pneumoniae rlrA locus. rlrA is divergently transcribed from at least six different genes indicated by black arrows, and the entire locus is flanked by two IS1167 elements. The left element contains a frameshift mutation ...

Given that the rlrA and srtD strains were attenuated for infection and colonization of mucosal surfaces, we sought to determine if any of the other genes in the locus were also attenuated in the same mouse tissues. We used magellan5, a mini-mariner transposon conferring spectinomycin resistance to create insertions by in vitro transposition into two PCR products spanning the entire rlrA locus. The transposition products were transformed into naturally competent AC353, and 55 transposon insertion strains were selected. Each transposon insertion was coarsely mapped by PCR using a template specific primer and a transposon specific primer, and the junction sequence of the transposon/magellan5 junctions in selected strains was determined by DNA sequencing. We obtained transposon insertions throughout the locus, including numerous insertions in each gene in the locus, thus demonstrating that neither the sortase genes, nor the rrg genes are essential for growth of S. pneumoniae in vitro.

Each of the transposon insertion strains shown in Fig. 1, except for rlrA, was tested by competition assay in the murine model of pneumonia, as described above. The results from these experiments are shown in Fig. 2A. Of the six strains tested, only the rrgA and the srtD strains had a virulence defect in the murine lung, exhibiting a 18-fold and a 24-fold attenuation respectively (Fig. 2A). The same six strains were also tested for colonization defects in the nasopharynx. In these assays, the rrgA strain was again attenuated, exhibiting a fivefold defect, and the srtB strain had a modest twofold defect (Fig. 2B). The other four genes were not required in either animal model. On the contrary, transposon insertions in srtC and rrgC resulted in a phenotype of hypercolonization in the nasopharynx. The basis by which these strains outcompete the wild-type strain was not investigated. In a further set of competition experiments, we tested the rrgA, srtB, and srtD strains for defects in survival during bacteraemia. We found that none of these were attenuated in this model (Fig. 2B). These data support the model whereby these factors are specific to the interaction of S. pneumoniae and mucosal surfaces.

Fig. 2
Analysis of rlrA locus mutants in animal models of lung infection (A), nasopharyngeal carriage and bacteraemia (B). The in vivo competitive index (CI) was calculated as described in the text; each circle represents the CI for a single mouse in each set ...

Signature-tagged mutagenesis fails to isolate virulence attenuated acapsular strains

The extracellular polysaccharide capsule plays an absolute role in the pathogenesis of S. pneumoniae. The majority of the biosynthetic genes coding for the serotype 4 capsule (SP0337 to SP0353) appear to be organized into a single operon of approximately 15 kilobases (kb), representing about 0.7% of the TIGR4 genome. In our STM screen and in two smaller scale STM screens in S. pneumoniae (Polissi et al., 1998; Lau et al., 2001) virulence attenuated acapsular mutants were not found. Our negative results led us to hypothesize that either the capsule was not required by TIGR4 for lung infection of Swiss Webster mice or that magellan2 insertions into the capsular operon were deleterious to growth in vitro, and therefore could not be isolated.

To discern between these two possibilities, a 9.9-kb fragment of the capsule operon was amplified with primers cpsF1 and cpsR1 by PCR, and used as template DNA for in vitro magellan2 transposon mutagenesis. The transposition products were transformed into wild-type bacteria and transposon insertion strains were selected on media containing Cm. To our surprise, no transformants were recovered after the standard 24 h of growth. After an additional 24 h, however, a small number of transformants appeared. In contrast, in a parallel experiment, transposition into a non-essential 10 kb segment of the genome yielded a large number of transformants after 24 h of growth (data not shown). Thus we conclude that disruption of the TIGR4 capsular operon by magellan2 is inhibitory to colony formation in the experimental conditions used, and that the failure to recover attenuated acapsular mutants is probably caused by the low plating efficiency of these strains following transformation. The growth rate of an acapsular mutant, AC846 (see below), was found to be equivalent to the wild-type strain when grown individually or in co-culture experiments in THY broth showing that the low frequency selection of these strains is not simply due a general growth defect. Whether a similar phenomenon occurred in the other two STM screens remains unknown.

Several of the acapsular mutants isolated after 48 h of growth in the above experiment were confirmed by mapping of the magellan2 insertion and by negative Quellung reactions (data not shown). An acapsular strain containing a disruption of cps4E (AC846, Table 3) was tested in competition assays in the murine lung and bacteraemia models. In both instances the acapsular mutant was severely attenuated (CI ≤ 0.04 and ≤0.001 for pneumonia and bacteraemia models respectively), confirming the importance of serotype 4 capsule in our animal models.

Table 3
Relevant strains, plasmids and primers used in this study.


Novel insights into the pathogenesis of S. pneumoniae are likely to aid in the development of new antibiotic treatments and vaccines. Current knowledge of factors implicated in virulence have led to promising developments towards new protein based vaccines (Briles et al., 2000a, b), however, an understanding of how most of these factors contribute to and function during infection is still lacking. In this study, we have greatly expanded the knowledge base of genes that are essential for virulence in a murine model of pneumonia by completing an STM screen; the third of its kind in this organism, but by far the most extensive.

Surprisingly, 20% of the 6147 strains screened by STM had a noticeable virulence defect. The large percentage of attenuated strains isolated in this screen is much higher than the 1–7% observed in similar screens in other Gram-positive pathogens (Mei et al., 1997; Jones et al., 2000; Autret et al., 2001). Of note, however, each of the previous pneumococcal STM screens identified approximately 10% of strains as attenuated (Polissi et al., 1998; Lau et al., 2001). The difference between prior pneumococcal STM screens and ours may result from more stringent cut-offs for selecting attenuated strains in the latter studies. Additionally, as has been suggested by others, it is conceivable that the use of a polar transposon mutagen may contribute to a higher percentage of attenuated strains within a library compared to mutants isolated by plasmid insertion-duplication (Paton and Giammarinaro, 2001). Insertion of a polar transposon into a genetic locus not only disrupts the gene harbouring the insertion, but also downstream genes that are co-transcribed with that gene. With plasmid insertion-duplication, not all insertions will result in a gene or operon null mutation, as for example plasmids containing either the 5′ end of a gene or containing a promoter region will likely regenerate a wild-type copy of the same gene or promoter following recombination.

To learn more about additional in vivo roles for some of the virulence genes identified herein, mutants were grouped based on their phenotypes in murine models of nasopharyngeal carriage and bacteraemia. In total, 25 different strains were tested in multiple animal models using competition assays, and these are grouped by class in Table 2. From this a picture emerges of the tissue specificity that many S. pneumoniae virulence factors play.

One striking feature of these classes is that many transcriptional regulators are found in each of the four classes, reinforcing the idea that tissue specific regulation of virulence factors is important for pneumococcal pathogenesis. Of the 16 putative transcriptional regulators identified in our screen, only two, smrC and SP2142 (STM119 and STM256) have been previously identified in S. pneumoniae, and thus most have unknown targets of regulation. Similarly, most of the two component signal transduction systems (TCSTS) in S. pneumoniae also have unknown targets of regulation. In our screen, we implicate roles for five of the 13 S. pneumoniae TCSTSs in lung infection (Lange et al., 1999; Throup et al., 2000). In addition to the insertions isolated in rr01, rr07, zmpR, and comD (STM185, STM29, STM90, and STM281), the insertion in strain STM237 could potentially have polar effects on hk11/rr11. We tested four of these five strains in competition assays. Three strains were attenuated in competition assays in the lung, but they each had different phenotypes in the bacteraemia model; zmpR was not attenuated, rr01 was attenuated fourfold, and STM237 was attenuated 14-fold. Additionally, rr01 was severely out-competed by the wild-type strain during nasopharyngeal colonization, making it the only one of the three TCSTSs tested that was required in all three models.

Mutants in most of the TCSTSs have been tested previously for avirulent phenotypes. comDE, which is involved in the induction of natural competence (Pestova et al., 1996), was previously shown to attenuate virulence in both lung infection and bacteraemia (Bartilson et al., 2001; Lau et al., 2001). Two groups identified each of the S. pneumoniae TCSTS by sequence homology and tested mutant TCSTS strains in different animal models (Lange et al., 1999; Throup et al., 2000). Throup et al. (2000) used the respiratory tract infection (RTI) model to test TCSTS mutants for virulence defects, which employs single strain infections and relies on the titre of the mutant strain compared to the wild-type strain following 48 h of infection to determine the virulence phenotype. By this assay, mutations in rr01 and zmpR each attenuated virulence, which is consistent with our findings, however, a mutation in hk11/rr11 did not. Lange et al. (1999) examined mutant TCSTS strains for defects during systemic infection by determining the mean survival time of mice infected with each strain. None of the TCSTS mutant strains tested by this assay were attenuated, which conflicts with our observed phenotype of an rr01 strain and STM237. These differences are best explained by the fact that competition assays measure the ratio of the mutant and wild-type strain following co-infection to assess attenuation, which is a more sensitive test for the loss of virulence than survival curves. Additionally, because our insertion in STM237 is upstream of the coding sequence for hk11/rr11 in a putative ABC transporter, it is possible that our strain has a more severe phenotype than an insertion in either hk11 or rr11 alone. In light of our findings and those of others (Throup et al., 2000; Bartilson et al., 2001; Lau et al., 2001), it is interesting to speculate that these three two-component systems play important roles in the adaptation of S. pneumoniae to different host environments by sensing different extracellular signals that in turn result in differential virulence gene regulation.

The sequencing project of the TIGR4 strain identified a small number of loci that are not conserved in two other pneumococcal strains (Tettelin et al., 2001). One such locus encodes rlrA (SP0461), a rofA-like transcriptional regulator, and six divergently transcribed genes including three putative MSCRAMM surface proteins. Our screen identified two genes in this locus, rlrA and srtD, and we subsequently tested the other five genes for roles during infection. Three of the flanking genes, rrgA, rrgB, and rrgC, code for putative surface proteins that are homologous to MSCRAMM family members, and thus we predict that they may be involved in the attachment of S. pneumoniae to mucosal surfaces. Consistent with this hypothesis, the rrgA strain was attenuated in both the pneumonia and the nasopharynx carriage model, but not the bacteraemia model.

In addition to having homology to MSCRAMMs, RrgA, RrgB, and RrgC have sorting signals that are characteristic of proteins that are anchored to the gram-positive cell wall by sortases (Fig. 1B) (Mazmanian et al., 2001; Pallen et al., 2001). The sorting signal is composed of a C-terminal sequence consisting of an LPXTG motif, followed by a stretch of hydrophobic residues, and a series of charged residues (Schneewind et al., 1993). RrgA, RrgB, and RrgC each have these characteristics, except that the leucine is replaced by a tyrosine, isoleucine and valine respectively (Fig. 1B). As at least one cell-wall anchored protein (RrgA) is required for infection and colonization, one would predict that one or more sortases should also be required. Consistent with this hypothesis, we found that a mutation in srtD resulted in a severe defect in the ability of S. pneumoniae to infect the lung. Together with the observed phenotypes of other strains with mutations in the rlrA locus, these data suggest a specific role this locus in the interaction of S. pneumoniae with mucosal surfaces.

In S. aureus, it has been elegantly shown that sortases are transpeptidases that anchor target proteins by cleaving the peptide bond between the threonine and glycine of the LPXTG and covalently anchoring the threonine to the cell wall. Through the genomic sequence analysis of numerous Gram-positive organisms it is evident that multiple sortase paralogues are common within single strains, including TIGR4 (Pallen et al., 2001; Mazmanian et al., 2002). In addition to srtBCD, the TIGR4 genomic sequence also contains a fourth sortase, srtA. SrtA is found in at least two other S. pneumoniae strains, R6 and D39, which do not contain srtBCD (Hoskins et al., 2001).

The role that multiple sortase paralogues play in protein anchoring has been studied in three different species thus far. In S. aureus there are two known sortases, SrtA anchors the majority of the LPXTG containing proteins, while SrtB has only been shown to anchor a single protein that contains an asparagine substituted for the leucine and the glycine in the LPXTG motif resulting in an NPQTN sequence. Furthermore, srtB is transcriptionally regulated in response to changing iron conditions, rather than being expressed constitutively (Mazmanian et al., 2002). In S. suis, five sortase homologues have been identified (Osaki et al., 2002), and as in S. aureus, the majority of the anchored surface proteins are dependent upon a single sortase, SrtA. Lastly, Barnett and Scott have recently shown that SrtA and SrtB from S. pyogenes also anchor different subsets of LPXTG containing surface proteins (Barnett and Scott, 2002). Given these findings in other organisms, we suspect that SrtA will anchor most LPXTG containing proteins in S. pneumoniae, and the remaining sortases may then anchor a specific set of surface proteins in different environmental conditions in response to different environmental cues. It is tempting to speculate that the role for SrtB, SrtC, and SrtD proteins in TIGR4 is to anchor the (L)PXTG-motif proteins RrgA, RrgB, and RrgC, which are coded by the genes flanking srtBCD.

The three independent STM screens in S. pneumoniae have resulted in the combined screening of over 8500 strains for virulence defects in a number of different assays. Remarkably, there is very little overlap in the sets of genes that have been identified as essential virulence factors in each of these screens. Only 10 of the 231 unique genes identified here were also reported in previous S. pneumoniae STM screens (Polissi et al., 1998; Lau et al., 2001). The lack of significant overlap between the three screens is likely the result of two factors; (1) the number of S. pneumoniae genes that are crucial to survival in vivo is probably large, such that the combined STM screens have not approached saturation yet, and (2) the different mutagenesis strategies employed (transposon vs. plasmid insertion-duplication) are responsible for mostly distinct sets of genes being disrupted. Regardless of the underlying causes, the lack of significant overlap between the three STM screens suggests that many additional factors linked to virulence remain to be identified.

Experimental procedures

Bacterial strains, plasmids and DNA manipulations

Strains, plasmids, and primers used in this study are listed in Table 3. All S. pneumoniae strains used and constructed in this study are derivatives of TIGR4, a serotype 4 clinical isolate. Antibiotic concentrations used in this study were as follows: chloramphenicol (Cm) 4 µg ml−1, streptomycin (Sm) 100 µg ml−1, and spectinomycin (Spc) 200 µg ml−1 for S. pneumoniae; Cm 10 µg ml−1 and Spc 100 µg ml−1 for E. coli. All DNA manipulations were carried out according to standard protocols (Sambrook et al., 1998). Signature-tags were PCR amplified from a plasmid preparation of pUTmTn5Km2 (Hensel et al., 1995) using primers P6 and P7. Amplification conditions for PCR were as follows: 30 cycles of 96°C for 30 s, 94°C for 20 s, 52°C for 45 s, and 72°C for 10 s, followed by a final dwell at 72°C for 15 min. PCR products were ethanol precipitated and resuspended in Bgl II buffer (New England Biolabs), and digested with Bgl II overnight at 37°C. Plasmid pEMcat was digested overnight with Bgl II. Both the linearized plasmid and the signature-tags were gel purified, and the former was dephosphorylated using shrimp alkaline phosphatase according to the manufacturer's instructions (Boehringer Mannheim). Purified signature-tags were ligated into the vector overnight with T4 DNA ligase (New England Biolabs) and the ligation mixture was introduced into E. coli DH5αλpir via electroporation. Transformants were selected on Luria–Bertani (LB) agar plates supplemented with Cm.

Transformants that contained uniquely tagged mini-transposons on pEMcat were isolated as follows. Colony purified transformants were grown overnight in four microtitre plates, each plate comprised a pool. Next, 4 µl of each well was spotted onto Duralon nitrocellulose membranes, one pool per membrane (Stratagene). Membranes (one per pool) were transferred onto filter paper (Whatman) saturated in denaturation solution (0.5 N NaOH, 1.5 M NaCl) for 10 min, 0.1% SDS for 3 min, and lastly, neutralization solution [1.0 M Tris HCl (pH 7.5), 1.5 M NaCl] for 3 min, at which time, DNA was cross-linked to membranes in a UV Stratalinker (Stratagene). The membranes were incubated in 3 × SSC, 0.1% SDS for 1 h, and cellular debris was gently removed from the membranes by rubbing with Kimwipes (Kimberly Clarke). Probe was generated from each pool using primers P6 and P7 by dioxygenin(DIG)-dUTP labelling PCR as described by the manufacturer (Roche). Cross-reacting signature-tags were eliminated between each of the pools by successive hybridizations of probe from one pool to blots with tags from another pool. From these hybridizations, 129 strains that did not cross-hybridize were randomly assembled into two new pools, and screened for cross-hybridizing signature-tags as above. Finally, 93 strains were selected that did not cross-hybridize and that contained signature-tags were isolated that amplify well by PCR. To generate master dot blots for hybridization of input and output signature tags, the unique 40 bp signature-tag of each magellan2 transposon was purified and spotted onto membranes as described (Merrell et al., 2002). All membranes were stored at 4°C.

In vitro transposon mutagenesis, DNA transformation, and pool construction

Plasmid DNA was purified from E. coli strains harbouring each of the 93 uniquely tagged magellan2 elements using Qiagen mini plasmid preparation kit according to the manufacturer's instructions (Qiagen). Streptococcus pneumoniae genomic DNA was isolated from AC353 as follows: AC353 was grown in 40 ml of THY (Todd Hewitt broth, 0.5% yeast extract) supplemented with Sm and 5 µl ml−1 Oxyrase (Oxyrase) statically in a candle extinction jar. Cells were washed in sterile dH20, resuspended in 200 µl of lysis buffer (0.1% deoxycholate, 0.01% SDS, 0.15 M NaCl) and incubated at 37°C for 10 min. Next, 0.9 ml of SSC was added and samples were incubated an additional 10 min at 65°C. The cell lysate was phenol-extracted, chloroform-extracted and ethanol-precipitated. Precipitated DNA was washed in 70% ethanol, and resuspended in 200 µl of 50 mM Tris-HCl (pH 7.5), 5 mM CaCl2. Ten microlitres of proteinase K (10 mg ml−1) and 2 µl of RNase (100 mg ml−1) were added and the mixture was incubated at 37°C for 10 min EDTA was added to 10 mM to stop the reaction, and the lysate was again extracted with phenol and chloroform and ethanol-precipitated.

In vitro magellan2 transposition reactions were carried out with purified MarC9 transposase, 500 ng of target AC353 genomic DNA and 1 µg of each of pEMcat derivative separately, essentially as described (Lampe et al., 1999). Reactions were ethanol-precipitated and resuspended in gap repair buffer [50 mM Tris (pH 7.8), 10 mM MgCl2, 1 mM DTT, 100 nM dNTP, and 50 ng of BSA]. Repair of transposition product gaps was performed as described (Akerley et al., 1998), except that E. coli DNA ligase (NEB) was used in place of T4 DNA ligase. Repaired transposition products were transformed into naturally competent AC353 as described (Bricker and Camilli, 1999). Of the 93 pEMcat derivatives used in the above procedure, only 63 reproducibly yielded sufficient numbers of transformants. Therefore, CmR colonies were picked from these 63 transformations only, and statically grown to late logarithmic phase in 200 µl of THY in 96-well microtitre plates in candle extinction jars, and subsequently frozen after the addition of glycerol to 20% (v/v). This entire procedure was repeated three times to assemble 100 pools of 63 mutant strains to be used for STM screening as described below. For the assembly of 2° pools, 1 µl of frozen cells from the appropriate well was inoculated into 200 µl of THY in a 96-well microtitre plate and grown for 5 h to log phase as above. Glycerol was added to 20% (v/v) and the plates were stored at −75°C.

magellan5 transposon insertions into the rlrA locus were generated identically to the magellan2 mutagenesis, except that two different 7 kb PCR products were used as target DNA. Polymerase chain reaction products were amplified from AC353 with primer sets TNPAB-F/REG2-R and REG2-F/PFL-R (Table 3) and purified using the Qiagen PCR purification kit according to manufacturers guidelines. In vitro transposition, gap repair, and natural transformation were carried out exactly as for magellan2.

Animal infections

In all animal infections 6 to 10-week-old female Swiss Webster mice were used (Taconic Laboratories). Mice were provided with continuous food and water, and housed according to the Tufts University Department of Laboratory Animal Medicine guidelines. Pools were prepared for infection by resuspending ~1 µl of frozen cells in 25 µl of THY, and plating 5 µl of each strain as a discrete spot on a blood agar plate [Blood Agar Base no. 2 (Difco) and 5% defibrinated sheep blood] supplemented with Cm and Sm. Following overnight growth, the entire pool was resuspended in THY and adjusted to OD600 ≈ 0.85 (approximately 5 × 108 CFU ml−1), the remainder of this resuspension was saved and used to assess the complexity of the input population of bacteria (see below). In the determination of colonization bottlenecks, and the 1° and 2° STM screens, 40 µl of each resuspended pool was inoculated intranasally into two lightly anaesthetized mice using methoxyflurane inhalation. The infections were carried out for 44 h at which time, mice were sacrificed by CO2 asphyxiation. Both lungs from each animal were aseptically removed, and homogenized in 5 ml of THY-glycerol (20% v/v) and stored at −75°C.

Serial dilutions of bacteria recovered from each mouse were plated on blood plates supplemented with Sm and Cm, such that a semi-confluent lawn of colonies was obtained. Bacteria were recovered with THY, genomic DNA from input and output bacteria were prepared using the DNAEasy Tissue kit according to the manufacturers tissue preparation protocol (Qiagen). Recovered genomic DNA was used as template for PCR amplification of the signature tags. DIG-dUTP was incorporated during the PCR as described above and signature-tag master blots were probed as described (Merrell et al., 2002).

Competition experiments

Prior to competition experiments, magellan2 insertion mutations were back-crossed into AC353 as follows: genomic DNA was prepared from each selected mutant strain as above, and was used to transform natural competent AC353 as described (Bricker and Camilli, 1999). Mutant and wild-type (AC353) strains were grown separately on blood agar plates with appropriate antibiotics, and recovered and prepared for infection, identically to the input pools above. Before infection, mutant bacteria and AC353 were mixed in a 1:1 ratio, and inoculated at the following doses: 1 × 107 CFU for lung infections, 5 × 105 for i.p. infections, and 1 × 108 CFU for nasopharyngeal inoculation. Infections (i.p) were carried out for 20 h and nasopharyngeal carriage infections for 7 days. Bacteria from systemic infections were recovered from the bloodstream by cardiac puncture. Nasopharyngeal colonized bacteria were recovered by washing the nasopharynx with 400 µl of sterile phosphate-buffered saline essentially as described (Wu et al., 1997). In conjunction with each in vivo competition, an in vitro competition was carried out as follows: 40 µl of each mixture was inoculated into 10 ml of THY supplemented with Sm (50 µg ml−1) and Oxyrase (5 µl ml−1) and grown (~9 doublings) to mid-log phase for 5 h. Following each experiment, the ratio of mutant to wild-type bacteria, for both in vitro and in vivo competitions, was determined by first plating recovered bacteria on TSA blood plates with Sm, and subsequently replica-plating colonies to plates with Sm or Sm and Cm. Competitive indices were calculated as the ratio of mutant to wild-type bacteria recovered from each animal (in vivo CI) or from THY broth (in vitro CI) adjusted by the input ratio.

Arbitrary-primed PCR, DNA sequencing, and sequence analysis

For each of the 387 strains determined to be highly attenuated by STM screening, we attempted to amplify one magellan2/genomic junctional sequence by arbitrary-primed PCR and determined its sequence as described (Merrell et al., 2002). The primer pairs used for the arbitrary-primed PCRs (ARB1/MAG2F3 and ARB2/MAGF4) are listed in Table 3. DNA sequencing of arbitrary primed PCR products was performed by the W. M. Keck Facility at Yale University. Obtained nucleotide sequence was used to identify the precise site of the magellan2 insertion in the TIGR4 genome sequence. The predicted protein sequence of each disrupted ORF was used to search the non-redundant NCBI protein database by blastp.

The site of the magellan5 transposon insertions in the rlrA locus were first determined by PCR using either TNPAB-F or PFLA-R with the primer marOUT, which anneals to either end of the mini-mariner transposons, followed by gel electrophoresis. Select PCR products were then purified with the Qiagen PCR purification and the DNA sequence of the transposon junction was determined using marOUT by the Tufts University Core Sequencing Facility.


This research was supported by Pew Scholars Award P0168SC and the Center for Gastroenterology Research on Absorptive and Secretory Processes, NEMC (P30DK34928). We are grateful to K.S. Lew for the construction of the cps4E mutant strain and to M. Prudhomme for the purified MarC9. We thank David Holden for the pUTmTn5Km2 plasmid library, S. H. Lee, D. S. Merrell, and M. Angelichio for technical assistance and thoughtful discussions, the members of the Waldor Laboratory for countless helpful discussions, and J. Mecsas for critically reading the manuscript and help with statistical analysis.


  • Adamou JE, Heinrichs JH, Erwin AL, Walsh W, Gayle T, Dormitzer M, et al. Identification and characterization of a novel family of pneumococcal proteins that are protective against sepsis. Infect Immun. 2001;69:949–958. [PMC free article] [PubMed]
  • Akerley BJ, Rubin EJ, Camilli A, Lampe DJ, Robertson HM, Mekalanos JJ. Systematic identification of essential genes by in vitro mariner mutagenesis. Proc Natl Acad Sci USA. 1998;95:8927–8932. [PMC free article] [PubMed]
  • Autret N, Dubail I, Trieu-Cuot P, Berche P, Charbit A. Identification of new genes involved in the virulence of Listeria monocytogenes by signature-tagged transposon mutagenesis. Infect Immun. 2001;69:2054–2065. [PMC free article] [PubMed]
  • Barnett TC, Scott JR. Differential recognition of surface proteins in Streptococcus pyogenes by two sortase gene homologs. J Bacteriol. 2002;184:2181–2191. [PMC free article] [PubMed]
  • Bartilson M, Marra A, Christine J, Asundi JS, Schneider WP, Hromockyj AE. Differential fluorescence induction reveals Streptococcus pneumoniae loci regulated by competence stimulatory peptide. Mol Microbiol. 2001;39:126–135. [PubMed]
  • Berry AM, Paton JC. Additive attenuation of virulence of Streptococcus pneumoniae by mutation of the genes encoding pneumolysin and other putative pneumococcal virulence proteins. Infect Immun. 2000;68:133–140. [PMC free article] [PubMed]
  • Berry AM, Lock RA, Thomas SM, Rajan DP, Hansman D, Paton JC. Cloning and nucleotide sequence of the Streptococcus pneumoniae hyaluronidase gene and purification of the enzyme from recombinant Escherichia coli. Infect Immun. 1994;62:1101–1108. [PMC free article] [PubMed]
  • Bongaerts RJ, Heinz HP, Hadding U, Zysk G. Antigenicity, expression, and molecular characterization of surface-located pullulanase of Streptococcus pneumoniae. Infect Immun. 2000;68:7141–7143. [PMC free article] [PubMed]
  • Bricker AL, Camilli A. Transformation of a type 4 encapsulated strain of Streptococcus pneumoniae. FEMS Microbiol Lett. 1999;172:131–135. [PubMed]
  • Briles DE, Hollingshead S, Brooks-Walter A, Nabors GS, Ferguson L, Schilling M, et al. The potential to use PspA and other pneumococcal proteins to elicit protection against pneumococcal infection. Vaccine. 2000a;18:1707–1711. [PubMed]
  • Briles DE, Hollingshead SK, Nabors GS, Paton JC, Brooks-Walter A. The potential for using protein vaccines to protect against otitis media caused by Streptococcus pneumoniae. Vaccine. 2000b;19:S87–S95. [PubMed]
  • Brown JS, Gilliland SM, Holden DW. A Streptococcus pneumoniae pathogenicity island encoding an ABC transporter involved in iron uptake and virulence. Mol Microbiol. 2001;40:572–585. [PubMed]
  • Cundell DR, Pearce BJ, Sandros J, Naughton AM, Masure HR. Peptide permeases from Streptococcus pneumoniae affect adherence to eucaryotic cells. Infect Immun. 1995;63:2493–2498. [PMC free article] [PubMed]
  • Dintilhac A, Alloing G, Granadel C, Claverys JP. Competence and virulence of Streptococcus pneumoniae: Adc and PsaA mutants exhibit a requirement for Zn and Mn resulting from inactivation of putative ABC metal permeases. Mol Microbiol. 1997;25:727–739. [PubMed]
  • Fogg GC, Gibson CM, Caparon MG. The identification of rofA, a positive-acting regulatory component of prtF expression: use of an m gamma delta-based shuttle mutagenesis strategy in Streptococcus pyogenes. Mol Microbiol. 1994;11:671–684. [PubMed]
  • Gosink KK, Mann ER, Guglielmo C, Tuomanen EI, Masure HR. Role of novel choline binding proteins in virulence of Streptococcus pneumoniae. Infect Immun. 2000;68:5690–5695. [PMC free article] [PubMed]
  • Hammerschmidt S, Bethe G, Remane PH, Chhatwal GS. Identification of pneumococcal surface protein A as a lactoferrin- binding protein of Streptococcus pneumoniae. Infect Immun. 1999;67:1683–1687. [PMC free article] [PubMed]
  • Hanahan D. Studies on transformation of Escherichia coli with plasmids. J Mol Biol. 1983;166:557–580. [PubMed]
  • Hensel M, Shea JE, Gleeson C, Jones MD, Dalton E, Holden DW. Simultaneous identification of bacterial virulence genes by negative selection. Science. 1995;269:400–403. [PubMed]
  • Hollingshead SK, Becker R, Briles DE. Diversity of PspA: mosaic genes and evidence for past recombination in Streptococcus pneumoniae. Infect Immun. 2000;68:5889–5900. [PMC free article] [PubMed]
  • Hoskins J, Alborn WE, Jr, Arnold J, Blaszczak LC, Burgett S, DeHoff BS, et al. Genome of the bacterium Streptococcus pneumoniae strain R6. J Bacteriol. 2001;183:5709–5717. [PMC free article] [PubMed]
  • Jedrzejas MJ. Pneumococcal virulence factors: structure and function. Microbiol Mol Biol Rev. 2001;65:187–207. [PMC free article] [PubMed]
  • Jones AL, Knoll KM, Rubens CE. Identification of Streptococcus agalactiae virulence genes in the neonatal rat sepsis model using signature-tagged mutagenesis. Mol Microbiol. 2000;37:1444–1455. [PubMed]
  • Kolter R, Inuzuka M, Helinski DR. Trans-complementation-dependent replication of a low molecular weight origin fragment from plasmid R6K. Cell. 1978;15:1199–1208. [PubMed]
  • Lampe DJ, Churchill ME, Robertson HM. A purified mariner transposase is sufficient to mediate transposition in vitro. EMBO J. 1996;15:5470–5479. [PMC free article] [PubMed]
  • Lampe DJ, Akerley BJ, Rubin EJ, Mekalanos JJ, Robertson HM. Hyperactive transposase mutants of the Himar1 mariner transposon. Proc Natl Acad Sci USA. 1999;96:11428–11433. [PMC free article] [PubMed]
  • Lange R, Wagner C, de Saizieu A, Flint N, Molnos J, Stieger M, et al. Domain organization and molecular characterization of 13 two-component systems identified by genome sequencing of Streptococcus pneumoniae. Gene. 1999;237:223–234. [PubMed]
  • Lau GW, Haataja S, Lonetto M, Kensit SE, Marra A, Bryant AP, et al. A functional genomic analysis of type 3 Streptococcus pneumoniae virulence. Mol Microbiol. 2001;40:555–571. [PubMed]
  • Martin B, Prudhomme M, Alloing G, Granadel C, Claverys JP. Cross-regulation of competence pheromone production and export in the early control of transformation in Streptococcus pneumoniae. Mol Microbiol. 2000;38:867–878. [PubMed]
  • Mazmanian SK, Ton-That H, Schneewind O. Sortase-catalysed anchoring of surface proteins to the cell wall of Staphylococcus aureus. Mol Microbiol. 2001;40:1049–1057. [PubMed]
  • Mazmanian SK, Ton-That H, Su K, Schneewind O. An iron-regulated sortase anchors a class of surface protein during Staphylococcus aureus pathogenesis. Proc Natl Acad Sci USA. 2002;99:2293–2298. [PMC free article] [PubMed]
  • McCullers JA, Tuomanen EI. Molecular pathogenesis of pneumococcal pneumonia. Front Biosci. 2001;6:D877–D889. [PubMed]
  • Mei JM, Nourbakhsh F, Ford CW, Holden DW. Identification of Staphylococcus aureus virulence genes in a murine model of bacteraemia using signature-tagged mutagenesis. Mol Microbiol. 1997;26:399–407. [PubMed]
  • Merrell DS, Hava DL, Camilli A. Identification of novel factors involved in colonization and acid tolerance of Vibrio cholerae. Mol Microbiol. 2002;43:1471–1491. [PubMed]
  • Osaki M, Takamatsu D, Shimoji Y, Sekizaki T. Characterization of Streptococcus suis genes encoding proteins homologous to sortase of Gram-Positive bacteria. J Bacteriol. 2002;184:971–982. [PMC free article] [PubMed]
  • Pallen MJ, Lam AC, Antonio M, Dunbar K. An embarrassment of sortases – a richness of substrates? Trends Microbiol. 2001;9:97–101. [PubMed]
  • Paton JC, Giammarinaro P. Genome-based analysis of pneumococcal virulence factors: the quest for novel vaccine antigens and drug targets. Trends Microbiol. 2001;9:515–518. [PubMed]
  • Pestova EV, Havarstein LS, Morrison DA. Regulation of competence for genetic transformation in Streptococcus pneumoniae by an auto-induced peptide pheromone and a two-component regulatory system. Mol Microbiol. 1996;21:853–862. [PubMed]
  • Podbielski A, Woischnik M, Leonard BA, Schmidt KH. Characterization of nra, a global negative regulator gene in group A streptococci. Mol Microbiol. 1999;31:1051–1064. [PubMed]
  • Polissi A, Pontiggia A, Feger G, Altieri M, Mottl H, Ferrari L, Simon D. Large-scale identification of virulence genes from Streptococcus pneumoniae. Infect Immun. 1998;66:5620–5629. [PMC free article] [PubMed]
  • Poulsen K, Reinholdt J, Jespersgaard C, Boye K, Brown TA, Hauge M, Kilian M. A comprehensive genetic study of streptococcal immunoglobulin A1 proteases: evidence for recombination within and between species. Infect Immun. 1998;66:181–190. [PMC free article] [PubMed]
  • Rosenow C, Ryan P, Weiser JN, Johnson S, Fontan P, Ortqvist A, Masure HR. Contribution of novel choline-binding proteins to adherence, colonization and immunogenicity of Streptococcus pneumoniae. Mol Microbiol. 1997;25:819–829. [PubMed]
  • Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: a Laboratory Manual. 2nd edn. Plainview, NY: Cold Spring Harbor Laboratory Press; 1998.
  • Sanchez-Beato AR, Lopez R, Garcia JL. Molecular characterization of PcpA: a novel choline-binding protein of Streptococcus pneumoniae. FEMS Microbiol Lett. 1998;164:207–214. [PubMed]
  • Schneewind O, Mihaylova-Petkov D, Model P. Cell wall sorting signals in surface proteins of Gram-positive bacteria. EMBO J. 1993;12:4803–4811. [PMC free article] [PubMed]
  • Tettelin H, Nelson KE, Paulsen IT, Eisen JA, Read TD, Peterson S, et al. Complete genome sequence of a virulent isolate of Streptococcus pneumoniae. Science. 2001;293:498–506. [PubMed]
  • Throup JP, Koretke KK, Bryant AP, Ingraham KA, Chalker AF, Ge Y, et al. A genomic analysis of two-component signal transduction in Streptococcus pneumoniae. Mol Microbiol. 2000;35:566–576. [PubMed]
  • Tomasz A. New faces of an old pathogen: emergence and spread of multidrug-resistant Streptococcus pneumoniae. Am J Med. 1999;107:55S–62S. [PubMed]
  • Unsworth KE, Holden DW. Identification and analysis of bacterial virulence genes in vivo. Philos Trans R Soc Lond B Biol Sci. 2000;355:613–622. [PMC free article] [PubMed]
  • Walker JA, Allen RL, Falmagne P, Johnson MK, Boulnois GJ. Molecular cloning, characterization, and complete nucleotide sequence of the gene for pneumolysin, the sulfhydryl-activated toxin of Streptococcus pneumoniae. Infect Immun. 1987;55:1184–1189. [PMC free article] [PubMed]
  • Wizemann TM, Heinrichs JH, Adamou JE, Erwin AL, Kunsch C, Choi GH, et al. Use of a whole genome approach to identify vaccine molecules affording protection against Streptococcus pneumoniae infection. Infect Immun. 2001;69:1593–1598. [PMC free article] [PubMed]
  • Wu HY, Virolainen A, Mathews B, King J, Russell MW, Briles DE. Establishment of a Streptococcus pneumoniae nasopharyngeal colonization model in adult mice. Microb Pathog. 1997;23:127–137. [PubMed]
  • Yesilkaya H, Kadioglu A, Gingles N, Alexander JE, Mitchell TJ, Andrew PW. Role of manganese-containing superoxide dismutase in oxidative stress and virulence of Streptococcus pneumoniae. Infect Immun. 2000;68:2819–2826. [PMC free article] [PubMed]
  • Yother J, White JM. Novel surface attachment mechanism of the Streptococcus pneumoniae protein PspA. J Bacteriol. 1994;176:2976–2985. [PMC free article] [PubMed]
  • Zysk G, Bongaerts RJ, ten Thoren E, Bethe G, Hakenbeck R, Heinz HP. Detection of 23 immunogenic pneumococcal proteins using convalescent-phase serum. Infect Immun. 2000;68:3740–3743. [PMC free article] [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...