• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of iaiPermissionsJournals.ASM.orgJournalIAI ArticleJournal InfoAuthorsReviewers
Infect Immun. Jul 2008; 76(7): 3176–3186.
Published online Apr 21, 2008. doi:  10.1128/IAI.01721-07
PMCID: PMC2446735

tRNA Modification by GidA/MnmE Is Necessary for Streptococcus pyogenes Virulence: a New Strategy To Make Live Attenuated Strains [down-pointing small open triangle]


Studies directed at vaccine development and mucosal immunity against Streptococcus pyogenes would benefit from the availability of live attenuated strains. Our approach for production of candidate live attenuated strains was to identify mutations that did not alter growth in vitro and did not alter the overall complement of virulence factors produced but did result in reduced levels of expression of multiple secreted virulence factors. A global reduction but not elimination of expression would likely lead to attenuation while maximizing the number of antigenic targets available for stimulation of immunity. Adaptation of Tn5-based transposome mutagenesis to S. pyogenes with initial screening for reduced expression of the SpeB protease resulted in identification of mutations in gidA, which encodes an enzyme involved in tRNA modification. Reduced SpeB expression was due to delayed onset of speB transcription resulting from reduced translation efficiency of the message for RopB, a transcriptional activator. Overall, GidA mutants had a nearly normal global transcription profile but expressed significantly reduced levels of multiple virulence factors due to impaired translation efficiencies. A translation defect was supported by the observation that mutants lacking MnmE, which functions in the same tRNA modification pathway as GidA, phenocopied GidA deficiency. The mutants stimulated a cytokine response in cultured macrophages identical to that in the wild type, with the exception of reduced levels of tumor necrosis factor alpha and interleukin-23. Significantly, GidA mutants were highly attenuated in the murine ulcer model of soft tissue infection. These characteristics suggest that GidA pathway tRNA modification mutants are attractive candidates for further evaluation as live attenuated strains.

Despite the exquisite sensitivity of Streptococcus pyogenes (group A streptococcus) to penicillin, diseases caused by this organism remain a major public health concern in both developed and developing countries (6). This pathogenic gram-positive bacterium is the causative agent of myriad diseases, ranging from severe invasive diseases (cellulitis, necrotizing faciitis), toxigenic diseases (scarlet fever, toxic shock syndrome), and sequlae (rheumatic heart disease, glomerulonephritis) to more self-limiting diseases, including pyoderma and pharyngitis. A recent analysis of the prevalence of group A streptococcal diseases indicated that the minimum global burden is upwards of 18 million cases of severe disease, with 1.78 million new cases arising each year, resulting in at least 517,000 deaths (6). This degree of mortality places group A streptococci among the major human pathogens, exceeded only by human immunodeficiency virus, Mycobacterium tuberculosis, Plasmodium falciparum, and Streptococcus pneumoniae (6). While children in developing countries bear the greatest burden of severe disease (40), the impact on the health care system of the United States is also considerable, with cases of acute pharyngitis accounting for as many as 11 million office visits per year (3). The burden imposed by group A streptococci on the global population has stimulated considerable interest in the development of new therapies for prevention of these diseases.

A vaccination approach offers many advantages. However, development of a safe and effective group A streptococcal vaccine has met with several significant challenges. Among these challenges are the fact that the pathogenesis of most group A streptococcal diseases is not well understood and may involve an autoimmune component in certain diseases, such as rheumatic fever (11). Similarly, both the contribution of the large number of secreted enzymes, toxins, and surface-associated factors produced by S. pyogenes to the pathogenesis of specific diseases and the importance of the immune response against these components to protective immunity are also not well understood (11). Early studies established that serum immunoglobulin can be opsonic and can promote the neutrophil-mediated killing of S. pyogenes in human blood (25). The antibodies function by neutralizing the antiphagocytic property of a streptococcal surface protein known as the M protein (for a review, see reference 15), and the generation of these antibodies has provided the rationale for several M protein-based vaccine formulations (for a review, see reference 3). However, several issues have complicated the development of vaccines based on the M protein, including (i) the fact that there are over 150 distinct serotypes of M protein and (ii) the fact that M proteins can elicit the production of antibodies that are cross-reactive with human tissues (10).

The use of peptides representing epitopes common to multiple serotypes, the use of multivalent peptides, and the use of peptides that have been engineered to lack prominent cross-reactive epitopes have all been used as strategies to increase the safety and efficacy of M protein vaccines (3). However, continuing concerns have stimulated investigation into other potential formulations, including the use of conserved surface proteins (3), and have spurred several large-scale interrogations of the surface proteome in the search for additional vaccine candidates (34, 47, 51). A combination of these studies with a thorough understanding of the immune response against S. pyogenes would offer a more rationale approach to vaccine design. Of particular utility would be a vaccine capable of preventing S. pyogenes from colonizing the mucosal surfaces of the nasopharynx, which is the primary site of most streptococcus-host interactions (11). However, the factors that contribute to immunity against S. pyogenes at a mucosal surface are not well understood.

Effective immunity at a mucosal surface generally requires engagement of the mucosal immune system, which is responsible for the production of defense molecules, like secretory immunoglobulin A, that are well adapted for defense in the mucosal environment (31). Most mucosal vaccines in current use accomplish this through the use of live attenuated strains that can be administered by a route that mimics the natural route of infection. The attenuated strains typically can undergo the initial stages of pathogenesis but have mutations that restrict them to producing a self-limiting infection with minimal symptomatology (12, 29). Unfortunately, there has been relatively little work directed at the development of live attenuated strains for analyses of immunity to S. pyogenes.

Since S. pyogenes pathogenesis is dependent upon secretion of a large array of surface proteins, extracellular enzymes, and toxins (11), our approach to construction of a candidate for a live attenuated strain was to conduct a screen for a mutant with the following characteristics: (i) it should have no obvious defect in growth characteristics in vitro; (ii) it should have no obvious defect in the number of different virulence factors that it expresses; and (iii) it should have a measurable defect that causes it to produce a large subset of its secreted virulence factors at significantly reduced levels. We reasoned that a global reduction in, but not elimination of, virulence factor expression would render the mutant attenuated, while maximizing the repertory of antigens that could be presented to the immune system. In this report, we describe mutants defective in a tRNA modification pathway as candidates for such a live attenuated strain. The pathway involves the proteins GidA and MnmE, which is responsible for the addition of a carboxymethylaminomethyl group to uridine 34 of a subset of tRNAs that is crucial for the appropriate decoding of many mRNAs (48). While these mutants have no obvious in vitro growth defect and a nearly normal global transcription profile, the level of expression of multiple secreted virulence factors by them is reduced but expression is not eliminated. In a subset of the mutants, reduced expression results from a defect at the level of translation of at least one key transcription regulator. The mutants induce a nearly normal pattern of cytokine expression in macrophages in vitro and are highly attenuated in a murine ulcer model of soft tissue infection. The fact that this pathway is highly conserved suggests that mutation of the genes encoding the factors is a strategy that can be applied to the construction of live attenuated strains of other species of pathogenic streptococci.


Bacterial strains.

The bacterial strains and plasmids used in this study are listed in Table Table1.1. Molecular cloning experiments were performed with Escherichia coli DH5α which was cultured in Luria-Bertani broth (38) at 37°C. For routine culture of S. pyogenes, Todd-Hewitt medium (BBL) supplemented with 0.2% yeast extract (Difco) (THY medium) was employed. Unless otherwise indicated, for growth of S. pyogenes for SpeB activity assays, Western blot analyses, and preparation of RNA for real-time reverse transcription (RT)-PCR C medium was used (37). For growth in liquid media, S. pyogenes was cultured at 37°C in sealed tubes without agitation. To produce solid media, Bacto agar (Difco) was added to a final concentration of 1.4% (wt/vol). When appropriate, antibiotics were added to the media at the following concentrations: spectinomycin, 100 μg/ml for E. coli and S. pyogenes; chloramphenicol, 15 μg/ml for E. coli and 3 μg/ml for S. pyogenes; kanamycin, 50 μg/ml for E. coli and 500 μg/ml for S. pyogenes; and erythromycin, 750 μg/ml for E. coli and 1 μg/ml for S. pyogenes.

Bacterial strains and plasmids

Manipulation and computational analyses of DNA.

Plasmid DNA was isolated via standard techniques and used to transform S. pyogenes as described previously (5). Restriction endonucleases, ligases, and polymerases were used according to the manufacturer's recommendations. The fidelity of all constructs derived by PCR was confirmed by DNA sequencing analyses. All references to genomic loci and predictions of protein function are based on the genome of S. pyogenes strain SF370 (14) and were supported by interrogation of the nonrepetitive sequence database at the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/BLAST/) using a gapped BLAST algorithm (2).

Construction of EZ-Tn5<ΩErm>.

For application of Tn5-based “transposome” mutagenesis (22) to S. pyogenes, the BamHI site of pMOD-2<MCS> (Epicenter Biotechnologies) was used to insert a BamHI fragment that contains the ΩErm element of pJDM-STM (30). The resulting plasmid was designated pMOD-2::Tn5<ΩErm>, and the resulting transposable element was designated EZ-Tn5<ΩErm> (Fig. (Fig.1A).1A). The erythromycin resistance gene of the ΩErm element can be used for selection in both E. coli and S. pyogenes (30).

FIG. 1.
(A) Transposable element EZ-Tn5<ΩErm>. The structure of the EZ-Tn5<ΩErm> transposon developed for transposome mutagenesis (22) is shown. The ΩErm element contains an erythromycin resistance determinant ...

Transposome mutagenesis.

A gel-purified PvuI fragment of pMOD-2::Tn5<ΩErm> that included Tn5<ΩErm> (Fig. (Fig.1A)1A) was mixed in vitro with the EZ-Tn5 transposase according to the directions of the manufacturer (Epicenter Biotechnologies). Electroporation was then used to introduce the transposome complex into competent cells of S. pyogenes that had been prepared by a standard method (5) and then washed once at room temperature in a 15% (vol/vol) glycerol solution containing 1 mM EDTA to remove Mg2+. This additional step was required for stabilization of the transposon-transposase complex prior to its introduction into the streptococcal cell (22). Transformants with insertions of Tn5<ΩErm> were then recovered following overnight incubation at 37°C on THY medium plates supplemented with erythromycin. For mutants of interest, the transposon insertion site was identified by direct sequencing of chromosomal DNA using a transposon-specific primer (SqTM-r) (see Table S1 in the supplemental material) as described previously (18).

Directed mutagenesis and complementation.

Insertional inactivation of specific genes and complementation analyses were conducted using a standard recombinational strategy (8). For mutagenesis, an approximately 0.7-kbp fragment of the gene of interest that lacked both its 5′ and 3′ ends was amplified by PCR, inserted into pSPC18 (37), and then introduced into E. coli. The plasmid was purified and then used to transform S. pyogenes. Since pSPC18 cannot replicate in S. pyogenes, most transformants resistant to spectinomycin have undergone a single homologous recombination event that results in disruption of the targeted gene (8). For complementation of a GidA mutant, an intact copy of gidA was introduced into the chromosome of the mutant as follows. A chromosomal segment containing the entire predicted gidA open reading frame and putative promoter region was amplified by PCR and inserted into the integrational plasmid pCIV2 (34). Transformation followed by recombination of the resulting plasmid into the mutant's chromosome resulted in introduction of an intact copy of gidA. Details for each integrating plasmid are shown in Table Table1,1, and the sequences of primers used for amplification of gene segments are shown in Table S1 in the supplemental material. The chromosomal structure of each mutant was confirmed by PCR using primers for the appropriate sequences. To ensure that mutant phenotypes were specifically due to gene disruption and not due to a nonspecific effect of integration of pSCP18 into the chromosome, a strain with pSPC18 inserted into the intergenic region downstream of recF (SPy_2204) (9), which is referred to here as HSC5spc (Table (Table1),1), was included in selected analyses.

Analysis of protein expression.

Analyses of cell growth and use of protease indicator medium, Western blot analysis of SpeB secretion, and quantitative measurement of SpeB activation and SpeB cysteine protease activity were all performed as described previously (37). Sample preparation and experiments with antisera for Western blot analyses to quantitate expression of streptolysin O (SLO), NAD glycohydrolase (SPN), mitogenic factor (MF), and RopA-hemagglutinin (RopA-HA) were conducted as described elsewhere (22, 35). Cells were fractionated as described previously (36), and M protein in cell wall fractions was analyzed by Western blotting using an antiserum developed against a peptide (QELAKKEEQNKISDASRKG) that was shown to be involved in streptococcal cell aggregation (16). The level of expression of a specific protein in various mutants was expressed relative to its expression in the wild type by comparison of band intensities in digital images of Western blots analyzed using IMAGEJ software (1).

Analysis of transcription.

RNA from the various streptococcal strains was isolated and analyzed using real-time RT-PCR as described previously (27), employing the primers listed in Table S1 in the supplemental material. Transcript abundance was normalized to the results for recA, and the data presented below are the means and standard deviations derived from at least two independent experiments that were performed on different days, with each individual sample analyzed in triplicate. Comparison of global transcriptional profiles using DNA microarrays was conducted as described in detail elsewhere (6, 13, 36).

Infection of macrophages.

Murine bone marrow-derived macrophages were obtained by a standard procedure (7) and were cultured in RPMI 1640 containing l-glutamine (Cellgro Mediatech, Inc.) with 20% (vol/vol) heat-inactivated fetal bovine serum (BioWhittaker, Inc.), 30% (vol/vol) L-cell conditioned medium (a source of macrophage colony-stimulating factor), penicillin G (50 IU ml−1), and streptomycin (50 μg ml−1). Macrophages were cultured for 8 days, harvested with ice-cold phosphate-buffered saline, resuspended in macrophage medium (Dulbecco's minimal essential medium [BioWhittaker, Inc.] containing 50 mM HEPES, 10% fetal bovine serum, and 8 mM l-glutamine), and then plated at a density of 2.5 × 106 cells per well in a six-well plate (catalog no. 3506; Corning Inc.). For infection, overnight cultures of S. pyogenes in THY medium were harvested, washed twice with phosphate-buffered saline, and then resuspended in macrophage medium to an optical density at 600 nm of 0.76. Portions of this suspension ranging from 50 to 100 μl were then added directly to wells containing macrophages. Final multiplicities of infection were determined by plating the streptococcal suspensions on THY agar following brief sonication to disrupt the streptococcal chains. In selected experiments, streptococci were inactivated by exposure to UV light (254 nm, 500 mJ/cm2) prior to incubation with macrophages. This dose exceeds the lethal dose for S. pyogenes (44), which was routinely confirmed by plating on THY agar.

Cytokine assays.

Following 4 h of infection of macrophage cultures, the supernatant fluids were harvested and then sterilized by filtration (Millex-GP; 0.22 μm; Millipore Corp). Cytokine levels in these fluids were determined using a commercial antibody-based multiplex bead assay (Bio-Plex, Mouse Cytokine 23-Plex; catalog no. 171-F11241; Bio-Rad Laboratories) to measure the 23 different murine cytokines that are listed in Fig. Fig.88 and Fig. S2 in the supplemental material. For this analysis the directions of the manufacturer (Bio-Rad Laboratories) were followed, and the analysis was conducted by the High Speed Cell Sorter Core Lab at the Siteman Cancer Center of Washington University. The data presented are the means and standard errors derived from three different infections, which were conducted on different days, and each experiment was analyzed in duplicate. Any differences in the mean values for specific cytokines induced in response to mutants compared with the wild type were tested for significance by the unpaired t test (19), and the null hypothesis was rejected when the P value was <0.05.

FIG. 8.
Cultured macrophages produce reduced levels of TNF-α and IL-12(p40) in response to GidA mutants. The levels of production of the indicated cytokines following 4 h of infection of cultured murine bone marrow-derived macrophages are shown. ...

Murine subcutaneous infection model.

The abilities of selected mutants to cause disease in soft tissue were evaluated by subcutaneous infection of 6- to 8-week-old SKH1 hairless mice (Charles River Labs) as described previously (4). Various strains were analyzed using groups containing five mice, and each experiment was repeated at least twice on different days. The area of the draining ulcers that formed was documented every 24 h by digital photography, and the precise area for each ulcer was calculated from the digital record using MetaMorph image analysis software (version 4.6; Universal Imaging Corp.). Any differences in the areas of ulcers between experimental groups was tested for significance by the Mann-Whitney U test (19), and the null hypothesis was rejected when the P value was <0.05.

Microarray data accession number.

The microarray data have been deposited in the National Center for Biotechnology Information Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo) under GEO Series accession number GSE9678.


Transposome mutagenesis identified the gidA gene.

We adapted a Tn5-based transposome mutagenesis strategy to screen for a mutant with reduced virulence factor expression (see Materials and Methods). This method has the advantages that it is simple, target site selection is random, and, because the inserted element itself does not encode transposase, insertions are stable in the absence of continued antibiotic selection (22). Introduction of a transposome complex generated with EZ-Tn5<ΩErm> (Fig. (Fig.1A)1A) into S. pyogenes HSC5 produced transformants at an efficiency between 103 and104 CFU/μg DNA. An insertion library was then examined for mutants that displayed reduced activity of the secreted SpeB cysteine protease following overnight culture on a protease indicator medium (see Materials and Methods). Approximately 4,000 colonies were examined, and nine mutants with reduced activity were identified that were subsequently found to express cysteine protease activity at levels between 5 and 20% of the wild-type level when they were examined using a quantitative proteolysis assay with a casein substrate (see Materials and Methods) following growth in two different media (THY and C media). Identification of the transposon insertion loci (see Materials and Methods) revealed that six mutants had insertions in the gene encoding RopB, which is a positive transcriptional regulator of SpeB, and three mutants had insertions at different sites within the gidA open reading frame (SPy_2185) (Fig. (Fig.1B),1B), whose orthologous gene products have been implicated in the flavin adenine dinucleotide (FAD)-dependent modification of tRNAs decoding two-family box triplet codons (47). Mutation of gidA in E. coli causes cell elongation and slow growth in rich media supplemented with glucose (45). However, the S. pyogenes mutants did not exhibit any measurable defect in either the growth rate or cell morphology upon culture in several rich media (THY and C media) in the absence or presence of additional glucose (0.5%, wt/vol) (data not shown). Also, like the genomes of many bacterial species (46), the S. pyogenes genome encodes a protein that is highly homologous to, although somewhat smaller than, GidA (SPy_1173) (20.1% identical and 55.1% similar). Insertional inactivation of SPy_1173 did not lead to any observable defect in SpeB biogenesis (data not shown).

Characterization of the gidA chromosomal region.

Examination of the genomic data available for multiple S. pyogenes strains revealed that the structure of the gidA chromosomal region is strongly conserved and that this region includes genes whose products are implicated in the function of ribosomes or DNA polymerase or have an unknown function (Fig. (Fig.1B),1B), all of which may contribute to the biogenesis of SpeB and other virulence factors. Since the EZ::Tn5<ΩErm> element includes strong transcription and translation terminators to ensure generation of strongly polar insertions (Fig. (Fig.1A),1A), it was necessary to examine the contribution of the adjacent genes to the GidA phenotype. Insertion duplication mutations constructed to inactivate gidA and the open reading frame immediately downstream (SPy_2184) (Fig. (Fig.2)2) produced two mutants that both had reduced expression of SpeB (HSC5::pKHC100 and HSC5::pKHC102) (Fig. (Fig.2).2). However, when the amounts of SpeB protein that were secreted into the supernatant were compared using a Western blot analysis, the two mutants had distinct phenotypes. When tested after 10 h of culture, the GidA mutant secreted a reduced amount of SpeB protein, like the original transposon insertion mutants. In contrast, the SPy_2184 mutant secreted wild-type levels of the protease zymogen at this time point but failed to process the inactive precursor to the active protease (data not shown). A comparison of the transcript levels of the genes downstream of gidA (SPy_2184 and rplI) (Fig. (Fig.1B)1B) between the wild type and a transposon insertion mutant (TM-GidA1) (Fig. (Fig.1B)1B) revealed no significant differences when a real-time RT-PCR was used (GidA mutant/wild-type strain expression ratio for SPy_2184, 0.92 ± 0.17; GidA mutant/wild-type strain expression ratio for rplI, 0.91 ± 0.30), showing that insertions into gidA are not polar for expression of downstream genes. Taken together, these data suggested that both GidA and SPy_2184 are required for SpeB biogenesis but contribute to different steps. A requirement for both genes was confirmed by construction of an additional insertion duplication mutation that placed the integrated plasmid between the intact gidA and SPy_2184 open reading frames. The resulting strain (HSC5::pKHC101) (Fig. (Fig.2)2) expressed nearly wild-type levels of SpeB proteolytic activity. Finally, reintroduction of an intact copy of gidA into the chromosome of a GidA mutant (see Materials and Methods) resulted in a complemented strain whose protease activity more closely resembled that of the wild type than that of the GidA mutant on protease indicator plates (Fig. (Fig.3A)3A) and which produced SpeB activity in culture supernatants at nearly wild-type levels (Fig. (Fig.3B).3B). Based on these data, we concluded that the reduced-expression phenotype observed in the transposon insertion mutants was due to the loss of gidA.

FIG. 2.
Confirmation of the contribution of gidA to SpeB expression. The integrational plasmids shown at the top were used to disrupt various regions of the gidA locus in S. pyogenes HSC5. The broken ends of the plasmid-encoded boxes indicate that a gene segment ...
FIG. 3.
Reconstitution of disrupted gidA restores SpeB expression. An intact copy of gidA was inserted into the chromosome of a GidA mutant (see Materials and Methods for details). (A) Protease production on indicator medium. The indicated strains were ...

Transcription of speB is delayed in GidA mutants.

Expression of speB is known to be highly regulated at the level of transcription, and this gene is expressed in a growth phase-dependent pattern (26). Since GidA mutants produced a reduced amount of SpeB protein at the time points described above, it was of interest to determine if the loss of gidA had any effect on the levels or timing of speB transcription. The growth patterns of the strains (the wild type, the GidA mutant, and the SPy_2184 mutant) were almost identical (Fig. (Fig.4;4; see Fig. S1 in the supplemental material), so the levels of speB transcript abundance of the strains at different time points were examined by real-time RT-PCR. The levels of the speB message in the wild type exhibited the expected growth phase pattern, with initiation at the onset of stationary phase and an increase over the next several hours, followed by a gradual decline (Fig. (Fig.4).4). The SPy_2184 mutant, which had reduced SpeB activity but not a reduced level of secreted SpeB protein, exhibited a pattern identical to that of the wild type (Fig. (Fig.4).4). In contrast, the GidA mutant displayed an aberrant pattern of expression. At the onset of stationary phase, the abundance of the speB transcript was over 60-fold less than that of the wild type (Fig. (Fig.4).4). The transcript became more abundant over time, until it reached a level that was not significantly different from the wild-type level when it was measured in 10-h cultures (Fig. (Fig.44).

FIG. 4.
Disruption of gidA results in altered pattern of speB transcription. The relative abundance of the speB message was determined over the course of growth in several strains using real-time RT-PCR. For each strain, transcript abundance is shown relative ...

Disruption of mnmE showed the same phenotype as mutation of gidA.

In addition to GidA, modification of U34 in tRNAs decoding two-family box triplet codons requires additional enzymes, including MnmE (4, 8, 14, 47). This relationship is supported by the observation that GidA and MnmE form a heterotetrameric complex in E. coli (47). To examine whether the SpeB expression defect of GidA mutants was specific to gidA or was a consequence of the loss of the ability to modify tRNA, a strain lacking mnmE was constructed (see Materials and Methods). A Western blot analysis revealed that the MnmE mutant had a SpeB expression defect identical to that of the GidA mutant (Fig. (Fig.5A),5A), indicating that the entire pathway for U34 modification may be important for SpeB biogenesis.

FIG. 5.
Overexpression of RopB restores SpeB expression in GidA pathway mutants. (A) Western blot analysis to detect SpeB in strains containing a plasmid that overexpresses RopB (pRopB-HA) or in the absence of a plasmid (no plasmid). Note that multiple proteolytic ...

GidA influences SpeB expression via the transcription regulator RopB.

In Shigella flexneri, mutation of several genes involved in tRNA modification results in a reduction in expression of selected virulence genes as a consequence of poor translation efficiency of the transcription regulator VirF (13). Expression of such virulence factors can be rescued in Shigella mutants by overexpression of VirF (13). In S. pyogenes, transcription of speB is under control of the transcription activator RopB (26). In the GidA mutant, the abundance of the ropB transcript was unchanged from the abundance in the wild type (Fig. (Fig.6),6), suggesting that the tRNA processing mutants may not efficiently translate this transcription regulator. In this case, overexpression of RopB should rescue speB expression, similar to overexpression of VirF in Shigella tRNA modification mutants. To test this, the GidA and MnmE mutants were transformed with a plasmid that expresses high levels of a RopB protein that has been modified to include an influenza HA tag at its carboxy terminus. This construct produces a functional RopB, as demonstrated by its ability to efficiently complement a ropB deletion (26). When these transformed strains were analyzed, it was found that overexpression of RopB restored the amount of SpeB protein secreted by the mutants to wild-type levels (Fig. (Fig.5A,5A, compare “no plasmid” to “pRopB-HA”). Even though the amount of RopB that the mutant strains produced was sufficient to rescue SpeB expression, these strains produced less than 50% of the amount of RopB-HA made by the wild-type strain transformed by the overexpression plasmid (Fig. (Fig.5B).5B). These data are consistent with the idea that the loss of GidA- and MnmE-dependent tRNA modification alters the pattern of speB transcription via an effect on the translation efficiency of the ropB message.

FIG. 6.
Nearly normal transcription of multiple virulence factors in a GidA mutant. The relative levels of abundance of the messages for the indicated virulence factors in a GidA mutant were determined by real-time RT-PCR following 6 h of culture ...

Reduced production of multiple secreted virulence proteins.

A sodium dodecyl sulfate-polyacrylamide gel electrophoresis-based comparison of cytosolic, cell membrane, and cell wall fractions of GidA and MnmE mutants and the wild type did not reveal any obvious differences in protein profiles (data not shown), consistent with the lack of any overall growth defect in the mutants. Transcription of speB occurs during the early stationary phase of growth; however, many virulence factors are expressed in growing cells. Examination of transcription profiles in growing cells using a microarray analysis indicated that the loss of GidA had a minimal impact on the global transcriptome, with only about 2% of the transcripts examined altered at a level of twofold or greater. Of these, only 0.5% (seven transcripts) were altered at levels greater than threefold and none were altered at levels greater than fivefold (see Tables S2, S3, and S4 in the supplemental material). A lack of an effect on transcription of virulence-associated genes was confirmed by a real-time RT-PCR analysis of a subset of messages, including several messages that encode transcription regulators implicated in virulence (mga, ropB, and covR) and a panel of additional factors representing both intracellular and secreted proteins. This analysis revealed that transcription of this panel of genes varied only about twofold or less from the wild-type transcription at 6 h postinoculation (Fig. (Fig.6).6). In contrast, examination of the levels of several secreted virulence proteins, including SPN, SLO, the surface M protein, and MF, demonstrated that both the GidA and MnmE mutants expressed each of these proteins at levels that were less than 50% of the wild-type levels in the exponential phase (Fig. (Fig.7).7). Taken together, these data suggest that a reduction in expression of multiple virulence proteins, such as SPN, SLO, M protein, and MF, in the exponential phase is influenced at the level of translation by GidA and MnmE and that there may be a greater impact on virulence factor production than on the production of housekeeping proteins that are involved in cell growth.

FIG. 7.
GidA pathway mutants have reduced expression of several virulence proteins. The results of Western blot analyses to detect the levels of expression of the indicated virulence proteins in several GidA pathway mutants are shown. Strains were cultured in ...

Comparison of cytokine profiles from infected macrophages.

It has been shown that S. pyogenes induces a unique cytokine activation program in macrophages (21) and that macrophages have an important role in controlling S. pyogenes infection in a murine model of disease (20). Since macrophages are also a key cell element of the adaptive immune response, the influence of GidA on streptococcus-macrophage interaction was assessed. Murine bone marrow-derived macrophages were infected with either the wild type or the GidA mutant or with these strains following UV treatment to block growth. After 4 h of infection, supernatant fluids were monitored for the presence of a panel of 23 different cytokines and chemokines (see Materials and Methods). Overall, the wild type and the GidA mutant, whether they were UV treated or not treated, induced identical patterns of expression of 21 cytokines at levels of twofold or greater (Fig. (Fig.8;8; see Fig. S2 in the supplemental material). The cytokines not stimulated by both the mutant and the wild type included interleukin 1α (IL-1α) and eotaxin (see Fig. S2 in the supplemental material). For the induced cytokines, the magnitudes of induction were also similar for the wild type and the GidA mutant (see Fig. S2 in the supplemental material), with two exceptions. For both untreated and UV-inactivated cultures, the GidA mutant induced tumor necrosis factor alpha (TNF-α) and IL-12(p40) at levels that were less than 50% of the levels induced by the wild type (Fig. (Fig.8).8). Since IL-12(p40) is a subunit of both IL-12 and IL-23 and since the IL-12(p70) subunit of IL-12 was not differentially expressed (Fig. (Fig.8),8), it was likely that IL-23, rather IL-12, was differentially induced. This was confirmed by real-time RT-PCR analysis of the IL-23(p19) subunit, which showed that this transcript was present in macrophages infected by the GidA mutant at a level that was only 22% ± 15% of the level induced in macrophages infected with the wild type.

GidA mutants are highly attenuated in a murine model of soft tissue infection.

The ability of the GidA mutants to cause disease in soft tissue was evaluated using the murine subcutaneous model. For the strain used in this study (HSC5), injection into the subcutaneous tissue of an SKH1 hairless mouse results in a local draining ulcer, whose area reaches the maximum value 3 days postinfection and which then goes on to heal over the next 10 to 14 days (4). In a comparison of the abilities of a GidA mutant (TM-GidA1) and the wild type to cause disease in this model, it was found that the latter produced lesions of the expected size by day 3 postinfection (Fig. (Fig.9).9). In contrast, the GidA mutant was significantly attenuated in the ability to cause lesions. While a few infected mice exhibited small lesions, the majority of mice infected with the mutant did not exhibit any visible lesions at day 3 postinfection or at any subsequent time point. Identical results were obtained with a second GidA mutant (HSC5::pKHC100) (data not shown). These data indicate that the reduction in virulence factor expression observed in vitro in the GidA mutants translates into a severely attenuated ability to cause disease in soft tissue.

FIG. 9.
GidA mutants are highly attenuated. The ability of a GidA mutant to cause disease in the murine subcutaneous ulcer model is shown. Virulence was evaluated on the basis of the area of the ulcer produced at the time when ulcer formation ...


This analysis showed that S. pyogenes GidA/MnmE pathway mutants have a number of characteristics that make them attractive candidates for further evaluation as live vaccine strains. These characteristics include the fact that they have no obvious defects in growth, the fact that they have minimal alterations to the complement of virulence factors that they produce, the fact that they produce a large subset of these factors at significantly reduced levels, and the fact that they are among the most attenuated mutants that have been analyzed to date using the murine subcutaneous model of soft tissue infection. Their avirulence appears to be the result of a global defect in translation efficiency manifested directly on messages encoding several secreted virulence factors or on messages for essential regulators of virulence gene transcription. Taken together, these observations suggest that tRNA modification mutant strains defective in the GidA pathway may be valuable for probing the mucosal immune response to S. pyogenes.

While the GidA gene was originally isolated as a gene responsible for a glucose-dependent cell division phenotype in E. coli (45), more recent studies have established that GidA is a flavin adenine dinucleotide-dependent enzyme that is required for addition of the carboxymethylaminomethyl group at position 5 of U34 of tRNAs that read mixed codon boxes, including those for Glu, Gln, Lys, Leu, and Arg (4, 29, 32). Also required for this modification is the GTPase MnmE, as both GidA and MnmE mutants of E. coli accumulate the same level of an undermodified form of U34 (48). This modification restricts base pairing at the wobble position with C and U to allow pairing at only G and A residues (63, 64). This prevents both misincorporation and frameshifting, as GidA and MnmE mutants have an elevated frequency of a +2 frameshift (4, 7, 29). Prevention of the latter error is likely to have a greater impact on protein synthesis because while many missense errors can be tolerated in polypeptides, a frameshift generally produces a truncated nonfunctional protein. Hypomodification of U34 increases the rate of frameshifting between two- and fivefold compared with the basal rate in E. coli, depending on the specific codon (43). If these rates are similar in S. pyogenes, then this level of frameshifting is well tolerated, as GidA and MnmE mutants grow at normal rates. However, the observation that virulence factor expression decreased to levels that adversely affected pathogenesis suggests that there is more selective sensitivity of virulence factors than of housekeeping proteins involved in cell growth.

It is possible that secreted proteins are particularly vulnerable to U34 hypomodification. The signal sequence required for protein export includes a positively charged N region and a hydrophobic H region, which in bacteria are dominated by Lys and Leu residues, respectively (32), whose codons both require U34 modifications for efficient decoding. Since many critical virulence factors of S. pyogenes are secreted proteins, a more selective defect in expression of these proteins could result in attenuation of virulence. Alternatively, since many transcriptional regulators have clusters of positively charged Lys and Arg residues in the domains that they use to bind DNA, they may be highly sensitive to the loss of GidA and MnmE, since the codons for these residues require decoding by tRNAs with U34 modifications. Even a modest decrease in the levels of a transcription regulator could have a significant pleiotropic effect on virulence depending on the number target genes that are dependent on the regulator for their transcription. The observation that a reduction in the translation efficiency of the ropB message is at least one factor contributing to the disregulation of expression of the SpeB protease illustrates the vulnerability of transcription regulators to alterations in the level of modified tRNAs. The ska gene was slightly overexpressed in the transcript analysis (Fig. (Fig.6),6), and this might indicate that less of a repressor of ska is produced translationally in the GidA mutant. Similarly, it has been shown that mutations that alter levels of U34 and A37 tRNA modification attenuate the virulence of S. flexneri exclusively via a decrease in the translation efficiency of the gene encoding the VirF transcription activator (13). It has been proposed that tRNA modification can function as a regulatory mechanism to adjust gene expression in response to nutrient deprivation via alterations in pools of cofactors required for the tRNA modification reactions (35). Whether the levels of VirF in S. flexneri or RopB is regulated by this mechanism remains to be determined.

In addition to mutants of S. pyogenes and S. flexneri, tRNA modification mutants of other pathogens have been shown to have robust defects in virulence in the absence of any remarkable defects in growth. For example, defects in virulence and/or virulence factor production associated with defects in tRNA modification have been reported for Agrobacterium tumefaciens (23), Pseudomonas syringae (24), and Aeromonas hydrophila (39). Prominent phenotypes for tRNA modification mutants have also been reported both for nonpathogenic bacteria and for eukaryotic pathogens, including Myxococcus xanthus (46) and the phytopathogenic fungus Colletotrichum lagenarium (42). The observation that tRNA mutants are frequently attenuated, combined with the fact that tRNA modification is a universal feature of all living cells, suggests that mutation of tRNA modification genes may prove to be a general strategy for construction of live attenuated vaccine strains. Since analysis of mucosal immunity against several other pathogenic streptococcal species, including Streptococcus agalactiae and Streptococcus pneumoniae, could benefit from the availability of live attenuated strains, it will be of interest to evaluate virulence factor expression and pathogenesis of GidA and MnmE mutants of these species.

Another characteristic that should make the S. pyogenes tRNA modification mutants useful as live attenuated strains is the fact that they do not have a major defect in the number of different virulence factors that they can express relative to the wild type. This suggests that they can engage the adaptive immune system in a manner similar to the manner of the wild type. This idea is supported by the observation that macrophages respond to the mutant and the wild type in similar ways, as revealed by the nearly identical cytokine profiles observed. The mutants did stimulate several cytokines to a lesser extent than the wild type, most notably TNF-α. Macrophage-derived TNF-α is an important early mediator of the inflammatory cascade, and it is known that S. pyogenes infection stimulates production of TNF-α in murine (26, 59) and baboon (41) models of infection and in human patients with severe invasive disease (33). However, in both humans and baboons, high levels of TNF-α were correlated with more serious disease (44, 54), suggesting that overproduction of TNF-α contributes to pathogenesis. Consistent with this, treatment of septic baboons with an anti-TNF-α monoclonal antibody markedly improved both symptoms and survival (41). Thus, the reduced ability of tRNA modification mutants to stimulate TNF-α may prove to be a useful property for a live attenuated strain.

It is less clear whether the attenuated strains will be useful for vaccination of human populations. Possible complications include the fact that the attenuated strains may also express epitopes that contribute to autoimmune sequlae and the fact that they express only a single serotype of the highly variant M protein molecule. It may be possible to engineer the attenuated strains to produce altered M protein variants lacking known cross-reactive epitopes and to express M protein domains common to multiple serotypes. Regardless, the availability of live attenuated strains, like the tRNA mutants described in this work, should be a valuable resource for further vaccine development.

Supplementary Material

[Supplemental material]


This work was supported by Public Health Service grant 46433 from the National Institutes of Health.

We thank Kwang Cheol Jeong and Joe Vogel for their assistance with macrophage isolation and culture, Molly Ingersoll for her assistance with the analyses of cytokine expression, and Swaine Chen for help with computational analyses.


Editor: J. N. Weiser


[down-pointing small open triangle]Published ahead of print on 21 April 2008.

Supplemental material for this article may be found at http://iai.asm.org/.


1. Abramoff, M. D., P. J. Magelhaes, and S. J. Ram. 2004. Image processing with IMAGEJ. Biophoton Int. 1136-42.
2. Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 253389-3402. [PMC free article] [PubMed]
3. Bisno, A. L., F. Rubin, P. P. Cleary, and J. B. Dale. 2005. Prospects for a group A streptococcal vaccine: rationale, feasibility, and obstacles—report of a National Institute of Allergy and Infectious Disease workshop. Clin. Infect. Dis. 411150-1156. [PubMed]
4. Brenot, A., K. Y. King, B. Janowiak, O. Griffith, and M. G. Caparon. 2004. Contribution of glutathione peroxidase to the virulence of Streptococcus pyogenes. Infect. Immun. 72409-413. [PMC free article] [PubMed]
5. Caparon, M. G., and J. Scott. 1991. Genetic manipulation of pathogenic streptococci. Methods Enzymol. 204556-586. [PubMed]
6. Carapetis, J. R., A. C. Steer, E. K. Mulholland, and M. Weber. 2005. The global burden of group A streptococcal infection. Lancet Infect. Dis. 5685-694. [PubMed]
7. Celada, A., P. W. Gray, E. Rinderknecht, and R. D. Schreiber. 1984. Evidence for a gamma-interferon receptor that regulates macrophage tumoricidal activity. J. Exp. Med. 16055-74. [PMC free article] [PubMed]
8. Cho, K. H., and M. G. Caparon. 2006. Genetics of group A streptococci, p. 59-88. In V. A. Fischetti, R. P. Novick, J. J. Ferretti, D. A. Portnoy, and J. I. Rood (ed.), Gram-positive pathogens, 2nd ed. ASM Press, Washington, DC.
9. Cho, K. H., and M. G. Caparon. 2005. Patterns of virulence gene expression differ between biofilm and tissue communities of Streptococcus pyogenes. Mol. Microbiol. 571545-1556. [PubMed]
10. Cunningham, M. W. 2003. Autoimmunity and molecular mimicry in the pathogenesis of post-streptococcal heart disease. Front. Biosci. 8s533-s543. [PubMed]
11. Cunningham, M. W. 2000. Pathogenesis of group A streptococcal infections. Clin. Microbiol. Rev. 13470-511. [PMC free article] [PubMed]
12. Dietrich, G., M. Griot-Wenk, I. C. Metcalfe, A. B. Lang, and J.-F. Viret. 2003. Experience with registered mucosal vaccines. Vaccine 21678-683. [PubMed]
13. Durand, J. M., B. Dagberg, B. E. Uhlin, and G. R. Bjork. 2000. Transfer RNA modification, temperature and DNA superhelicity have a common target in the regulatory network of the virulence of Shigella flexneri: the expression of the virF gene. Mol. Microbiol. 35924-935. [PubMed]
14. Ferretti, J. J., W. M. McShan, D. Ajdic, D. J. Savic, G. Savic, K. Lyon, C. Primeaux, S. Sezate, A. N. Suvorov, S. Kenton, H. S. Lai, S. P. Lin, Y. Qian, H. G. Jia, F. Z. Najar, Q. Ren, H. Zhu, L. Song, J. White, X. Yuan, S. W. Clifton, B. A. Roe, and R. McLaughlin. 2001. Complete genome sequence of an M1 strain of Streptococcus pyogenes. Proc. Natl. Acad. Sci. USA 984658-4663. [PMC free article] [PubMed]
15. Fischetti, V. A. 1989. Streptococcal M protein: modular design and biological behavior. Clin. Microbiol. Rev. 2285-314. [PMC free article] [PubMed]
16. Frick, I. M., M. Morgelin, and L. Björck. 2000. Virulgent aggregates of Streptococcus pyogenes are generated by homophilic protein-protein interactions. Mol. Microbiol. 371232-1247. [PubMed]
17. Ghosh, J., and M. G. Caparon. 2006. Specificity of Streptococcus pyogenes NAD+ glycohydrolase in cytolysin-mediated translocation. Mol. Microbiol. 621203-1214. [PubMed]
18. Gibson, C. M., and M. G. Caparon. 2002. Alkaline phosphatase reporter transposon for identification of gene encoding secreted proteins in gram-positive microorganisms. Appl. Environ. Microbiol. 68928-932. [PMC free article] [PubMed]
19. Glantz, S. 2002. Primer of biostatistics. McGraw-Hill Co., New York, NY.
20. Goldmann, O., M. Rohde, G. S. Chhatwal, and E. Medina. 2004. Role of macrophages in host resistance to group A streptococci. Infect. Immun. 722956-2963. [PMC free article] [PubMed]
21. Goldmann, O., M. von Kockritz-Blickwede, C. Holtje, G. S. Chhatwal, R. Geffers, and E. Medina. 2007. Transcriptome analysis of murine macrophages in response to infection with Streptococcus pyogenes reveals an unusual activation program. Infect. Immun. 754148-4157. [PMC free article] [PubMed]
22. Goryshin, I. Y., J. Jenrisak, L. M. Hofman, and W. S. Reznikoff. 2000. Insertional transposon mutagenesis by electroporation of released Tn5 transposition complexes. Nat. Biotechnol. 1897-100. [PubMed]
23. Gray, J., J. Wang, and S. B. Gelvin. 1992. Mutation of the miaA gene of Agrobacterium tumefaciens results in reduced vir gene expression. J. Bacteriol. 1741086-1098. [PMC free article] [PubMed]
24. Kinscherf, T. G., and D. K. Willis. 2002. Global regulation by gidA in Pseudomonas syringae. J. Bacteriol. 1842281-2286. [PMC free article] [PubMed]
25. Lancefield, R. C. 1957. Differentiation of group A streptococci with a common R antigen into three serological types, with special reference to the bactericidal test. J. Exp. Med. 106525-544. [PMC free article] [PubMed]
26. Loughman, J. A., and M. G. Caparon. 2007. Contribution of invariant residues to the function of Rgg family transcription regulators. J. Bacteriol. 19650-655. [PMC free article] [PubMed]
27. Loughman, J. A., and M. G. Caparon. 2006. Regulation of SpeB in Streptococcus pyogenes by pH and NaCl: a model for in vivo gene expression. J. Bacteriol. 188399-408. [PMC free article] [PubMed]
28. Lyon, W. R., J. C. Madden, J. C. Levin, J. L. Stein, and M. G. Caparon. 2001. Mutation of luxS affects growth and virulence factor expression in Streptococcus pyogenes. Mol. Microbiol. 42145-157. [PubMed]
29. Medina, E., and C. A. Guzmán. 2001. Use of live bacterial vaccine vectors for antigen delivery: potential and limitations. Vaccine 191573-1580. [PubMed]
30. Miller, J. D., and M. N. Neely. 2005. Large-scale screen highlights the importance of capsule for virulence in the zoonotic pathogen Streptococcus iniae. Infect. Immun. 73921-934. [PMC free article] [PubMed]
31. Nagler-Anderson, C. 2001. Man the barrier! Strategic defenses in the intestinal mucosa. Nat. Rev. Immunol. 159-67. [PubMed]
32. Nielsen, H., J. Engelbrecht, S. Brunak, and G. von Heijne. 1997. Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Eng. 101-6. [PubMed]
33. Norrby-Teglund, A., K. Pauksens, M. Norgren, and S. E. Holm. 1995. Correlation between serum TNF alpha and IL6 levels and severity of group A streptococcal infections. Scand. J. Infect. Dis. 27125-130. [PubMed]
34. Okada, N., R. T. Geist, and M. G. Caparon. 1993. Positive transcriptional control of mry regulates virulence in the group A streptococcus. Mol. Microbiol. 7893-903. [PubMed]
35. Persson, B. C. 1993. Modification of tRNA as a regulatory device. Mol. Microbiol. 81011-1016. [PubMed]
36. Rosch, J. W., and M. G. Caparon. 2005. The ExPortal: an organelle dedicated to protein secretion and folding in Streptococcus pyogenes. Mol. Microbiol. 58959-968. [PubMed]
37. Ruiz, N., B. Wang, A. Pentland, and M. G. Caparon. 1998. Streptolysin O and adherence synergistically modulate proinflammatory responses of keratinocytes to group A streptococci. Mol. Microbiol. 27337-346. [PubMed]
38. Scott, J. 1974. A turbid plaque-forming mutant of P1 that cannot lysogenize Escherichia coli. Virology 62344-349. [PubMed]
39. Sha, J., E. V. Kozlova, A. A. Fadl, J. P. Olano, C. W. Houston, J. W. Peterson, and A. K. Chopra. 2004. Molecular characterization of a glucose-inhibited division gene, gidA, that regulates cytotoxic enterotoxin of Aeromonas hydrophila. Infect. Immun. 721084-1095. [PMC free article] [PubMed]
40. Steer, A. C., M. H. Danchin, and J. R. Carapetis. 2007. Group A streptococcal infections in children. J. Paediatr. Child Care 43203-213. [PubMed]
41. Stevens, D. L., A. E. Bryant, S. P. Hackett, A. Chang, G. Peer, S. Kosanke, T. Emerson, and L. Hinshaw. 1996. Group A streptococcal bacteremia: the role of tumor necrosis factor in shock and organ failure. J. Infect. Dis. 173619-626. [PubMed]
42. Takano, Y., N. Takayanagi, H. Hori, Y. Ikeuchi, T. Suzuki, A. Kimura, and T. Okuno. 2006. A gene involved in modifying transfer RNA is required for fungal pathogenicity and stress tolerance of Colletotrichum lagenarium. Mol. Microbiol. 6081-92. [PubMed]
43. Urbonavicius, J., Q. Qian, J. M. Durand, T. G. Hagervall, and G. R. Bjork. 2001. Improvement of reading frame maintenance is a common function for several tRNA modifications. EMBO J. 204863-4873. [PMC free article] [PubMed]
44. Villarino, A., M. N. Rager, P. A. Grimont, and O. M. Bouvet. 2003. Are UV-induced nonculturable Escherichia coli K-12 cells alive or dead? Eur. J. Biochem. 2702689-2695. [PubMed]
45. von Meyenburg, K., B. B. Jorgensen, J. Nielsen, and F. G. Hansen. 1982. Promoters of the atp operon coding for the membrane-bound ATP synthase of Escherichia coli mapped by Tn10 insertion mutations. Mol. Gen. Genet. 188240-248. [PubMed]
46. White, D. J., R. Merod, B. Thomasson, and P. L. Hartzell. 2001. GidA is an FAD-binding protein involved in development of Myxococcus xanthus. Mol. Microbiol. 42503-517. [PubMed]
47. Yim, L., M. Martinez-Vicente, M. Villarroya, C. Aguado, E. Knecht, and M. E. Armengod. 2003. The GTPase activity and C-terminal cysteine of the Escherichia coli MnmE protein are essential for its tRNA modifying function. J. Biol. Chem. 27828378-28387. [PubMed]
48. Yim, L., I. Moukadiri, G. R. Bjork, and M. E. Armengod. 2006. Further insights into the tRNA modification process controlled by proteins MnmE and GidA of Escherichia coli. Nucleic Acids Res. 345892-5905. [PMC free article] [PubMed]

Articles from Infection and Immunity are provided here courtesy of American Society for Microbiology (ASM)
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...