Natural Variation in the Promoter of the Gene Encoding the MgaRegulator Alters Host-Pathogen Interactions in Group A StreptococcusCarrier Strains

Anthony R. Flores,a,b Randall J. Olsen,b Andrea Wunsche,b Muthiah Kumaraswami,b Samuel A. Shelburne III,c Ronan K. Carroll,b*James M. Musserb

Section of Infectious Diseases, Department of Pediatrics, Texas Children’s Hospital and Baylor College of Medicine, Houston, Texas, USAa; Center for Molecular andTranslational Human Infectious Diseases Research and Department of Pathology and Genomic Medicine, Houston Methodist Research Institute, Houston, Texas, USAb;Department of Infectious Diseases, Division of Internal Medicine, MD Anderson Cancer Center, Houston, Texas, USAc

Humans commonly carry pathogenic bacteria asymptomatically, but the molecular factors underlying microbial asymp-tomatic carriage are poorly understood. We previously reported that two epidemiologically unassociated serotype M3group A Streptococcus (GAS) carrier strains had an identical 12-bp deletion in the promoter of the gene encoding Mga, aglobal positive gene regulator. Herein, we report on studies designed to test the hypothesis that the identified 12-bp dele-tion in the mga promoter alters GAS virulence, thereby potentially contributing to the asymptomatic carrier phenotype.Using allelic exchange, we introduced the variant promoter into a serotype M3 invasive strain and the wild-type promoterinto an asymptomatic carrier strain. Compared to strains with the wild-type mga promoter, we discovered that strainscontaining the promoter with the 12-bp deletion produced significantly fewer mga and Mga-regulated gene transcripts.Consistent with decreased mga transcripts, strains containing the variant mga promoter were also significantly less viru-lent in in vivo and ex vivo models of GAS disease. Further, we provide evidence that the pleiotropic regulator protein CodYbinds to the mga promoter and that the 12-bp deletion in the mga promoter reduces CodY-mediated mga transcription.We conclude that the naturally occurring 12-bp deletion in the mga promoter significantly alters the pathogen-host inter-action of these asymptomatic carrier strains. Our findings provide new insight into the molecular basis of the carrier stateof an important human pathogen.

Asymptomatic carriage is a common but poorly understoodphenomenon that occurs for many human bacterial patho-

gens, including Neisseria meningitidis (1), Streptococcus pneu-moniae (2), Staphylococcus aureus (3), and Streptococcus pyogenes(group A Streptococcus [GAS]). Carriage of bacterial pathogens isclinically important because in some individuals carriage may be aprelude to invasive infection (4–11). In contrast, some early stud-ies suggest that GAS carrier strains have reduced ability to causedisease (12–14). A combination of pathogen and host factors isbelieved to contribute to the development of asymptomatic car-riage, but we understand little about the molecular basis by whichmajor bacterial pathogens are able to colonize humans for pro-longed periods in the absence of symptoms. The recent develop-ment of whole-genome sequencing techniques has provided anopportunity to generate novel insights into the molecular deter-minants of asymptomatic carriage. Importantly, in stark contrastto the analysis of the molecular genetics of virulence, little efforthas been devoted to examining the bacterial determinants ofasymptomatic carriage on a genome-scale level.

In terms of studying asymptomatic carriage, GAS is the proto-typical model organism, as GAS is responsible for a wide range ofdiseases but may also be carried asymptomatically for prolongedperiods (15, 16). GAS infections vary in severity from superficialskin infections and pharyngitis to necrotizing fasciitis and toxicshock syndrome. GAS causes an estimated 600 million pharyngitiscases each year (17) which, left untreated, may lead to acute rheu-matic fever. Individuals can carry GAS in the oropharynx or nasalepithelium for many months following resolution of clinical dis-ease or may carry GAS with no history of clinical symptoms (15,

16). Depending on the population studied, GAS carriage ratesrange from 5 to 15% in children (18, 19), far exceeding rates ofinvasive disease, making asymptomatic carriage the prevailingstate of GAS on human mucosal and skin surfaces. Paradoxically,few studies have investigated the contributory molecular determi-nants of this common phenomenon.

Although asymptomatic carriage of GAS has been studied for60 years, the bacterial determinants remain poorly understood.Pathogenic GAS strains produce a hyaluronic acid capsule that isstructurally identical to that found in human tissues and thus isnot an immunological target. The immune response to GAS fo-cuses mainly on freely secreted and cell wall-anchored proteins,such as the critical M protein that forms the basis of GAS serotyp-ing. Serotype-specific anti-M protein antibody appears to provideprotection from infection (20, 21), but reinfection with different

Received 29 March 2013 Returned for modification 25 April 2013Accepted 13 August 2013

Published ahead of print 26 August 2013

Editor: A. Camilli

Address correspondence to James M. Musser, JMMusser@HoustonMethodist.org.

* Present address: Ronan K. Carroll, Department of Cell Biology, Microbiology andMolecular Biology, University of South Florida, Tampa, Florida, USA.

Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.00405-13.

Copyright © 2013, American Society for Microbiology. All Rights Reserved.

doi:10.1128/IAI.00405-13

4128 iai.asm.org Infection and Immunity p. 4128–4138 November 2013 Volume 81 Number 11

on October 24, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

GAS serotypes is common given the presence of �140 M proteinserotypes. Interestingly, very early studies demonstrated that GASproduced less type-specific M protein when isolated in the asymp-tomatic carrier stage than did strains recovered in the acute infec-tion phase (13, 14). Despite these early observations, little workhas been done to examine the bacterial molecular mechanism(s)contributing to asymptomatic carriage of GAS in the upper respi-ratory tract.

In an effort to provide new information about asymptomaticGAS carriage, Beres et al. (22) determined the whole-genome se-quence of multiple GAS asymptomatic carrier strains and studiedtheir virulence in an animal model of invasive GAS infection.Compared to strains cultured from patients with invasive infec-tions, GAS carrier strains were significantly less virulent for miceafter intraperitoneal infection. Comparative genome sequencingrevealed that two of the four epidemiologically unassociatedasymptomatic carrier isolates had a 12-bp deletion in the pro-moter region of mga (multiple gene regulator of GAS) (22). Mga isa critical global regulator of many GAS virulence factors and is thekey transcription factor regulating expression of the emm genethat encodes M protein (23, 24). The 12-bp deletion in the carrierstrains is located in a variable-number tandem repeat (VNTR)located between the major promoter start site (25, 26) and thebeginning of the mga coding sequence (Fig. 1A). The effect of this12-bp deletion on mga expression and its contribution to alteredvirulence have not been studied and thus are not known.

Here we report the results of studies designed to test the hy-pothesis that the naturally occurring 12-bp deletion in the mgapromoter region of GAS carrier strains reduces mga and Mga-regulated gene transcript levels and alters GAS virulence in mouseand nonhuman primate models of infection. Our data conclu-sively show that a naturally occurring alteration in the promoter ofa key regulatory protein contributes to the phenotypic differencesobserved between these asymptomatic carrier strains and invasivestrains. These findings provide new insights into a molecular

mechanism contributing to altered virulence and asymptomaticcolonization, a critical, but poorly understood, aspect of bacterialhost-pathogen interaction.

MATERIALS AND METHODSBacterial strains and culture methods. The serotype M3 GAS strainsused are listed in Table S1 in the supplemental material. Asymptomaticcarrier strains MGAS12501 and MGAS12504 were isolated in a popula-tion-based carrier study (27) from unrelated, healthy individuals with norecent history of pharyngitis. The genomes of strain MGAS12501 andMGAS12504 have been sequenced and have a 12-bp deletion in the pro-moter region of mga relative to MGAS315 (22). Serotype M3 strainMGAS315 was isolated in the late 1980s from a patient in the United Stateswith streptococcal toxic shock-like syndrome (28). Strain MGAS315 hasbeen characterized extensively, and its genome has been sequenced (29).The other serotype M3 GAS strains used (MGAS9937, MGAS10863,MGAS10870, and MGAS15049) were cultured from patients with inva-sive infection in Ontario, Canada (22, 30). The genome of each of thesefour strains has been sequenced (30, 31). GAS strains were grown onTrypticase soy agar containing 5% sheep blood agar (SBA) (Becton, Dick-inson, Cockeysville, MD), in Todd-Hewitt broth containing 0.2% (wt/vol) yeast extract (THY) (Difco Laboratories, Detroit, MI), or on THYagar. When needed, GAS medium was supplemented with chloramphen-icol (Sigma-Aldrich, St. Louis, MO) at 10 �g/ml. For cloning experiments,we used Escherichia coli DH5� or TOP10 (Invitrogen) grown in Luria-Bertani (LB) broth or on LB agar (Difco Laboratories) and supplementedwith ampicillin (Sigma-Aldrich) at 100 �g/ml or chloramphenicol (Sig-ma-Aldrich) at 20 �g/ml when appropriate. The E. coli strain used forprotein overexpression was grown in LB broth, and ampicillin was addedto a final concentration of 80 �g/ml.

PCR and allelic exchange in strains MGAS315 and MGAS12504.Primers for PCR and plasmids used in this study are listed in Table S2 inthe supplemental material. Detailed methods regarding the generation ofisoallelic mutants of the mga promoter in MGAS315 and MGAS12504 aregiven in the supplemental material. Briefly, the temperature-sensitive E.coli–Gram-positive shuttle vector pJL1055 (32, 33) was used for allelicreplacement as previously described (34). The mga promoter region and

TAAATGCGTGATCGTGACTGTTTTTTGGATTTTAATGAAAGAATTTGAAAGGCATTTTATGGTATACTTTTTCTTGATAT

AGGTCGTACTGACTTAACGAAATACCTCACGAAATACCTCACGAAATAACAAGAAAATCAAGAAAATTAAAAAGAAGGGT

ATACAAGGTAATGTATGTAAGTAAGTTGTTTACAAGTCAACAGTGGAGAGAACTAAAATTAATCTCATACGTAACAGAAA-1-10

-20ACGAAATACCTC

-30-40-50-60-70-80-90

-100-110-120-130-140-150-160-170

CodY Binding Site12-bp Deletion (Carrier)

Mga translation startVNTR Region

-10-35

605040302010

P2 Start

AACGAAATACCTCACGAAATACCTCACGAAATAACAAGAAAATCAAGAAAATTAAAAAGAAGGGTAACGAAATACCTCACGAAATACCTCACGAAATAACAAGAAAATCAAGAAAATTAAAAAGAAGGGTAACGAAATACCTCACGAAATACCTCACGAAATAACAAGAAAATCAAGAAAATTAAAAAGAAGGGTAACGAAATACCTCACGAAATACCTCACGAAATAACAAGAAAATCAAGAAAATTAAAAAGAAGGGTAACGAAATACCTCACGAAATACCTCACGAAATAACAAGAAAATCAAGAAAATTAAAAAGAAGGGTA------------ACGAAATACCTCACGAAATAACAAGAAAATCAAGAAAATTAAAAAGAAGGGTA------------ACGAAATACCTCACGAAATAACAAGAAAATCAAGAAAATTAAAAAGAAGGGT

MGAS315MGAS9937MGAS10863MGAS10870MGAS15049MGAS12501MGAS12504

Invasive

Carrier

A

B

VNTR Region

FIG 1 The major promoter of mga differs between carrier and invasive GAS strains. (A) The P2 promoter and �10 and �35 positions (green boxes) are known(25, 26) and are indicated in relation to the mga gene start codon (black box). The deletion present in the two carrier strains is shaded and resides in a region (redboxes) between the P2 promoter start site and the beginning of the coding sequence of mga. A putative CodY-binding site (blue box) is indicated adjacent to theVNTR region. The short DNA sequence used in subsequent CodY-binding assays is indicated (black line) beneath the putative CodY-binding site. (B) The VNTRregion residing in the P2 promoter of mga is conserved in invasive GAS strains. The 12-bp deletion in the human carrier strains MGAS12501 and MGAS12504is highlighted in red in the invasive strain sequence. The putative CodY-binding site is indicated by a blue box.

GAS Carrier Strain mga Promoter Variation

November 2013 Volume 81 Number 11 iai.asm.org 4129

on October 24, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

flanking DNA sequence were confirmed in each background using SangerDNA sequencing as previously described (34).

Generation of isogenic codY deletion mutant in MGAS315 andMGAS12504. Primers used for the insertional inactivation of codY arelisted in Table S2 in the supplemental material. Insertional inactivation ofcodY was performed as described previously (35). Briefly, a three-stepprocess was used to generate a PCR fragment containing a spectinomycinresistance cassette (spc) with flanking DNA sequence of codY fromMGAS315 or MGAS12504. The PCR fragment was subsequently clonedinto a plasmid and then used for electroporation of GAS. Isogenic mutantstrains were selected on medium containing spectinomycin, and the genedisruption was confirmed by DNA sequencing.

Animal virulence experiments. Immunocompetent female CD1 mice(Harlan Laboratories) were used for virulence studies as described previ-ously (36). Mice were randomly assigned to treatment groups and inocu-lated in the right hind limb with 1 � 107 CFU of GAS in 100 �l of phos-phate-buffered saline (PBS). Stocks of each strain were prepared at knownCFU and stored at �80°C. Inocula were prepared by diluting frozenstocks in PBS to the desired number of CFU. Near mortality was deter-mined by observation using predefined criteria (36). Mice were eutha-nized with isoflurane. All mouse experiments were approved by the Insti-tutional Animal Care and Use Committee of the Houston MethodistResearch Institute.

Adult cynomolgus macaques (Macaca fasicularis) (Charles River BRF)were used for the nonhuman primate experiments. All monkey experi-ments were performed as previously described (36). We inoculated eachanimal (n � 3 per strain treatment group) with 1 � 108 CFU/kg of strainMGAS315 or MGAS12501. Detailed information regarding experimentswith nonhuman primates is given in the supplemental material. The studyprotocol was approved by the Institutional Animal Care and Use Com-mittee, University of Houston.

Growth in human blood. Experiments assessing the ability to grow inhuman blood were conducted under a Houston Methodist Research In-stitute Institutional Review Board human subjects protocol. Experimentswere carried out as described by Lancefield (37) and Flores et al. (38). Aminimum of three healthy, nonimmune adult donors were used for eachexperiment. Briefly, cultures of strains grown overnight in THY at 37°Csupplemented with 5% CO2 and were used to inoculate fresh, prewarmedTHY. Cultures were grown to mid-exponential phase, and bacteria werepelleted and resuspended in an equal volume of PBS. Each strain wassubsequently diluted to approximately 1 � 103 CFU from which 10 to 100CFU of GAS was used to inoculate 300 ml of fresh human blood. Sampleswere incubated at 37°C with 5% CO2 with gentle rotation for 3 h and thenserially diluted in PBS and immediately plated on SBA for CFU. Themultiplication factor was calculated by dividing the resulting CFU/mlafter 3 h of incubation by the starting inoculum.

RNA isolation and quantitative real-time PCR analysis. To analyzemga transcript over time, GAS strains were grown overnight in THY,diluted 1:100 in fresh THY, and incubated. Samples were taken hourly fordetermination of optical density at 600 nm (OD600) and RNA isolationand purification beginning 2 h postdilution (early exponential phase) forup to 10 h (stationary phase). RNA was isolated and purified using anRNeasy 96 kit according to the manufacturer’s instructions (Qiagen). Thequantity and quality of RNA were determined with an Agilent 2100 Bio-analyzer. Reverse transcription of RNA to produce cDNA was done with ahigh-capacity cDNA reverse transcription kit (Applied Biosystems).

TaqMan quantitative real-time PCR (qRT-PCR) was performed withan ABI 7500 Fast Real-Time system, and proS (39) or tufA (40) was used asthe endogenous control gene. TaqMan primers and probes used in anal-yses are listed in Table S2 in the supplemental material. To compare mgaexpression in asymptomatic carrier strains with invasive strains, thethreshold cycle (��CT) method was used (user bulletin no. 2, ABI Prism7700 Sequence Detection System; Applied Biosystems). All reactions wereperformed in quadruplicate using RNA purified from at least two biologicreplicates.

Purification of CodY from serotype M3 GAS and EMSA. Details ofthe recombinant CodY purification and electrophoretic mobility shift as-says (EMSA) are given in the supplemental material. Briefly, the codingregion of codY was amplified from strain MGAS315 using the primerslisted in Table S2 and was cloned into plasmid pET21b. Oligonucleotidesused to generate oligoduplexes for EMSA were pcodY (5=-TGCAAGGAATTATCAGAAAATTTTCTATT-3=), pmga (5=-TCTTTTTAATTTTCTTGATTTTCTTGTTA-3=), and DNA sequence lacking a CodY operator(5=-ACACACAGGTTGGTTGGTTGGTACACACA-3=). Binding reac-tions were performed at 37°C for 15 min in a 20-�l reaction volume with0.2 �M oligoduplex and increasing concentrations of rCodY.

Western immunoblot analysis for M protein in MGAS12504 andMGAS12504mgawt. GAS strains were grown in THY and cells harvestedin late exponential phase. Cell wall extracts were obtained as previouslydescribed (41). Strain MGAS12502 was used as a negative control, as itharbors a 195-bp deletion in the emm gene deleting the hypervariable Nterminus (22). Equal amounts of total protein were resolved on a 12%SDS-PAGE gel, transferred to a nitrocellulose membrane, and probedwith a previously described polyclonal anti-M protein (31). M protein wasdetected by secondary antibody conjugated with horseradish peroxidaseand visualized by chemiluminescence using Pierce ECL Western blottingsubstrate (Thermo Scientific).

Statistics. Kaplan-Meier analysis was performed to determine statis-tical significance of mortality data. A two-tailed t test was used to comparetranscript levels at individual time points and the multiplication factorsbetween strains grown in human blood. A two-tailed Mann-Whitney testwas used to compare pathology scores for nonhuman primate virulencestudies. A P value less than 0.05 was considered significant for all statisticaltests.

RESULTSAsymptomatic carrier strains MGAS12501 and MGAS12504 areless virulent in mouse and nonhuman primate models of necro-tizing fasciitis. We previously reported that the serotype M3asymptomatic carrier strains MGAS12501 and MGAS12504 had asignificantly higher 90% near-lethal dose (LD90) (i.e., were lessvirulent) for mice than did invasive serotype M3 strains after in-traperitoneal (i.p.) inoculation (22). To study virulence attributesof the asymptomatic carrier strains in more detail, we tested thehypothesis that strains MGAS12501 and MGAS12504 are less vir-ulent in a mouse model of necrotizing fasciitis (NF) (36) than theinvasive strain MGAS315. Consistent with the results obtainedfrom the i.p. mouse model, strain MGAS12501 and MGAS12504were significantly less virulent in the mouse necrotizing fasciitismodel (Fig. 2A). The carrier strains produced significantly lesstissue destruction and abscess formation compared to the exten-sive necrosis observed in mice inoculated with invasive strainMGAS315 (Fig. 2B to D). Similarly, carrier strains MGAS12501and MGAS12504 caused significantly less tissue pathology thaninvasive strain MGAS315 (Fig. 2E to G). Among the animals in-fected with the asymptomatic carrier strains, only two mice had anear-death phenotype, whereas all of the mice inoculated withinvasive strain MGAS315 (P � 0.05) had this phenotype.

GAS is a human-specific pathogen, and several key virulencefactors lack or have poor activity against mouse molecules. Forexample, the critical GAS virulence factor streptokinase is active inhumans and nonhuman primates but not in mice (42, 43). Thus,we next sought to test the hypothesis that the asymptomatic car-rier strain MGAS12501 is less virulent in a nonhuman primatemodel of necrotizing fasciitis (36). The use of strain MGAS12501(versus MGAS12504) in this analysis was an arbitrary choice, asthese two organisms are virtually identical (22). Consistent with

Flores et al.

4130 iai.asm.org Infection and Immunity

on October 24, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

the mouse data, strain MGAS12501 was significantly less virulent inthe nonhuman primate than the invasive strain MGAS315, as as-sessed by lesion volume (Fig. 3A), pathology score (Fig. 3B), andmagnitude and character of tissue destruction (Fig. 3C to H). To-gether, the mouse and nonhuman primate virulence studies clearlydemonstrate a decreased-virulence phenotype in the asymptomaticcarrier strains relative to the invasive strain MGAS315.

Asymptomatic carrier strains MGAS12501 and MGAS12504have significantly decreased mga transcript levels compared tothose of invasive strains. The 12-bp deletion identified in thecarrier strains MGAS12501 and MGAS12504 was not present inany of the genomes of 95 invasive serotype M3 GAS strains re-cently studied (30), suggesting that it is unique to asymptomaticcarrier strains. Further, the finding that the asymptomatic carrierstrains had a 12-bp deletion in the mga promoter suggested thatthe carrier strains may have altered mga expression compared tothat of invasive strains. Given that mga expression is growth phasedependent (44), we tested the hypothesis of decreased mga expres-sion using quantitative real-time reverse transcription-PCR (RT-PCR) with RNA isolated over the entire GAS growth cycle. Themga transcript level increased substantially in invasive strainMGAS315 beginning at the transition from exponential to sta-tionary growth phase and extending into stationary phase (Fig.4A). In striking contrast, the level of mga transcript did not changesignificantly in the two asymptomatic carrier strains. Rather, be-ginning at 7 h (transition to stationary phase), we found between2- and 22-fold less mga transcript in the asymptomatic carrierstrains than in MGAS315 (Fig. 4A and B). To test the hypothesisthat the significant decrease in mga transcript level observed inMGAS12501 and MGAS12504 was unique to the carrier strains,we analyzed mga transcript level in four additional serotype M3invasive strains, all of which lack the 12-bp promoter deletion inmga (see Table S1 and Fig. S1 in the supplemental material). Theasymptomatic carrier strains also had significantly lower levels ofmga transcript than did these other invasive strains (Fig. S1). Fi-nally, to ensure that the differences in mga transcript between theasymptomatic carrier and invasive strains were not an artifactcaused by decreased expression of the endogenous control gene(proS) in stationary phase, we repeated the experiment with a

second endogenous control gene, tufA (40). The results confirmedthe finding of decreased mga transcript levels in the carrier strains(see Fig. S2 in the supplemental material).

Having established that mga expression is significantly de-creased in the growth cycle of the two asymptomatic carrier strainswith the 12-bp deletion in the mga promoter, we next tested thehypothesis that these strains had decreased expression of Mga-regulated genes. Thus, we determined the transcript levels of genesknown to be directly regulated by Mga, including emm, scpA (en-coding C5a peptidase), and sclA (encoding streptococcal collagen-like protein A). The C5a peptidase of GAS cleaves C5a, therebyinhibiting recruitment of phagocytic cells to the site of infection(45, 46), whereas SclA is a cell surface protein involved in adher-ence and internalization (47). Consistent with our hypothesis, thetranscripts of all Mga-regulated genes tested (emm, scpA, and sclA)were significantly decreased in strains MGAS12501 andMGAS12504 compared to MGAS315 (Fig. 4B). Further, consis-tent with peak mga expression in MGAS315 (Fig. 4A), the greatestdifference in emm transcript level was observed in the transitionand stationary phases of growth (Fig. 4B). The same decreasedgene transcript pattern was observed for scpA and sclA (Fig. 4B).However, differences between the scpA and sclA transcripts levelswere far greater than observed for emm. It is possible that theobserved differences in transcript reflect differences in Mga pro-moter affinity between the three genes. Our findings confirm thatthe transcript levels of Mga-regulated genes were decreased in thecarrier isolates compared to invasive strains.

Allelic exchange demonstrates that the 12-bp deletion in mgacontributes to decreased production of mga transcript and de-creased virulence. Given that the carrier strains and invasivestrains have modest genetic differences in addition to the 12-bpmga promoter deletion, the data presented to date do not defini-tively demonstrate that the 12-bp mga promoter deletion is re-sponsible for the observed differences in virulence phenotype ortranscript level. Therefore, we next directly tested the hypothesisthat the 12-bp deletion in the promoter of mga in the asymptom-atic carrier strains results in the observed differences between thecarrier and invasive strains. We used allelic replacement (34) tointroduce the 12-bp deletion into strain MGAS315 to create a

MGAS315 MGAS12501 MGAS12504

BA

0 50 100 150 2000

102030405060708090

100

MGAS315MGAS12501MGAS12504

P < 0.0001

P < 0.0001

Hour

C D

E F G

FIG 2 Asymptomatic carrier strains are significantly less virulent in a necrotizing fasciitis model of GAS infection. (A) Survival curve of mice infected with carrierstrains MGAS12501 and MGAS12504 and invasive strain MGAS315. Ten mice in each group were infected intramuscularly with 1 � 107 CFU, as described inMaterials and Methods. P values are relative to MGAS315 (P � 0.05) as determined by log rank test. (B to D) Gross pathology (original magnification, �4) at 48h and (E to G) histopathology (original magnification, �20) of mouse hind limb lesions at 24 h postinfection. Mice infected with strain MGAS315 had extensivespreading myonecrosis and tissue destruction (B and E). In contrast, the lesions in mice challenged with the asymptomatic carrier strain MGAS12501 (C and F)or MGAS12504 (D and G) had small abscesses confined to the inoculation site. The boxed areas and arrows demarcate a circumscribed, walled-off lesion.

GAS Carrier Strain mga Promoter Variation

November 2013 Volume 81 Number 11 iai.asm.org 4131

on October 24, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

strain (MGAS315�12bp) that differs from its parental strain byonly the 12-bp deletion. The isoallelic mutant MGAS315�12bpwas similar in growth and hyaluronic acid capsule production tothe parental strain MGAS315 (data not shown). Consistent with

the hypothesis, strain MGAS315�12bp had a significantly de-creased mga transcript level compared to that of the parentalstrain MGAS315 (Fig. 5A).

Given the significant differences in the transcript levels of mga,we hypothesized that the introduction of the 12-bp mga promoterdeletion would result in significant virulence differences. As pre-dicted, strain MGAS315�12bp was significantly attenuated com-pared to parental strain MGAS315 following intramuscular inoc-ulation into mice (Fig. 5B). Compared to the invasive strainMGAS315, the isoallelic mutant strain MGAS315�12bp causedsignificantly less tissue destruction (Fig. 5C). Furthermore, incontrast to strain MGAS315, histopathologic analysis demon-strated that the surrounding muscle and soft tissue in the miceinfected with the isoallelic mutant strain retained viability (Fig.5D).

As mentioned previously, GAS is a human-specific pathogen,and the possibility exists that the decreased virulence observed inthe isoallelic mutant may not exist in a model more closely resem-bling infection in humans. Growth in human blood is often usedas a proxy for resistance to phagocytosis (48). Further, given thatthe isoallelic mutant has a significantly reduced mga transcriptand Mga is known to positively regulate multiple virulence factorsnecessary for immune evasion (23, 24), we tested the hypothesisthat the isoallelic mutant MGAS315�12bp would have signifi-cantly reduced ability to survive and grow in human blood. Weobserved a �2-fold reduction (P � 0.01) in the ability ofMGAS315�12bp to grow in human blood compared to that of theparental strain (Fig. 5E). Combined with the mouse virulencedata, these data further demonstrate that the 12-bp deletion in thepromoter of mga contributes to decreased virulence.

Repair of the 12-bp deletion in an asymptomatic carrierstrain results in increased mga and Mga-regulated gene tran-script levels and increased ability to grow in human blood. As ameans to confirm the finding that the 12-bp deletion in the pro-moter of mga contributes to decreased virulence in an invasivestrain of GAS, we next tested the hypothesis that repair of the12-bp deletion in an asymptomatic carrier strain would result inan increase in mga and Mga-regulated gene transcripts. Using al-lelic exchange (34), we constructed a strain that differed from theasymptomatic carrier strain MGAS12504 by only the absence ofthe 12-bp deletion in the promoter of mga (i.e., a wild-type mgapromoter). The isoallelic mutant MGAS12504mgawt was similarin growth and hyaluronic acid capsule production to the parentalstrain MGAS12504. Consistent with our hypothesis, we foundthat the repaired carrier strain, MGAS12504mgawt, produced sig-nificantly greater mga and Mga-regulated gene transcript levelsthan did the parental asymptomatic carrier strain (Fig. 6A). Con-sistent with increased emm transcript, we detected an increase inM protein in MGAS12504mgawt using antibody specific for Mprotein in serotype M3 GAS (Fig. 6B).

We hypothesized that similar to the isoallelic mutantMGAS315�12bp, the carrier strain harboring the 12-bp deletionin the promoter of mga would have a decreased ability to evadephagocytosis and thus grow in human blood compared to that ofthe “repaired” carrier strain containing the wild-type mga pro-moter. Consistent with our hypothesis, we observed a �6-foldincrease (P � 0.01) in the ability of MGAS12504mgawt to grow inhuman blood compared to that of the parental strain (Fig. 6C).Together, our data unambiguously demonstrate that the 12-bp

FIG 3 Asymptomatic carrier strain MGAS12501 is significantly less virulent thanthe invasive strain MGAS315 in a nonhuman primate model of necrotizing fascii-tis. (A) Cynomolgus macaques were inoculated intramuscularly in the anteriorthigh (n � 3 monkeys per strain treatment group), and nonviable lesion volumewas measured. (B) Microscopic pathology was scored using objective criteria asdescribed in Materials and Methods. P values refer to t tests assuming unequalvariance. (C) Specimens taken from the inoculation site of monkeys infected withthe invasive strain MGAS315 show extensive necrosis (boxed region) with com-plete obliteration of the muscle fascicles and intervening fascial planes (originalmagnification, �4). Many nonviable cells (black arrowheads) are seen in the back-ground of necrotic debris. (D) In contrast, monkeys infected with the asymptom-atic carrier strain MGAS12501 have necrosis that is limited to the fascial planes(area between dashed lines), sparing the adjacent muscle tissue, which retainsviability (white arrowheads). (E) At high magnification (original magnification,�10), the tissue from animals infected with strain MGAS315 is completely non-viable (black arrowheads) and extensively infiltrated by polymorphonu-clear leukocytes (circled region). (F) By comparison, muscle tissue fromanimals infected with strain MGAS12501 was relatively preserved, with avery limited pattern of infiltration that was restricted to the major facialplane (black arrows). (G and H) Gram stain confirms that the invasivestrain MGAS315 (left) diffusely infiltrated between individual myocyteswhereas the asymptomatic carrier strain MGAS12501 (right) had a veryrestricted border of progression (original magnification, �20).

Flores et al.

4132 iai.asm.org Infection and Immunity

on October 24, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

deletion contributes to decreased virulence in both invasive andasymptomatic carrier strains.

Decreased mga transcript level with variant mga promotermediated by CodY. It is possible that the 12-bp deletion in theasymptomatic carrier strains alters, either directly or indirectly,binding of a transcription factor, accounting for the observed mgatranscript differences. Mga is known to self-regulate throughbinding to its own promoter (25, 26); however, the previouslyidentified Mga binding sites are distant from the VNTR in whichthe 12-bp deletion occurs in the asymptomatic carrier strains (25,26). Thus, we interrogated the VNTR and immediately adjacentnucleic acid sequence for putative transcription factor bindingsites and discovered a putative binding site for the transcriptionfactor CodY (Fig. 1A). CodY is well described for a diverse array ofGram-positive bacteria in which it is known to repress synthesisand transport of branched-chain amino acids (49). Studies ofCodY in GAS have shown a similar metabolic repression (50–52)but suggest simultaneous activation of multiple virulence genesand virulence gene regulators, including mga (50, 51). However,no studies have shown direct activation of mga through CodY. Wehypothesized that the 12-bp deletion in the mga promoter re-duced or eliminated the ability of CodY to increase mga transcriptlevels. To begin to show CodY regulation of mga transcript levels,we generated isogenic mutants lacking codY in the invasive strainMGAS315 and the asymptomatic carrier strain MGAS12504. Ifour hypothesis is correct, mga transcript levels will be similar be-tween MGAS315�codY and MGAS315�12bp. Likewise, elimina-tion of CodY in the carrier strain MGAS12504, which harbors the12-bp deletion in the mga promoter, should have a minimal effecton mga transcription. Consistent with our hypothesis, we ob-served significantly decreased mga transcript in MGAS315�codYcompared to the wild-type parental strain, and the observed de-

crease was similar to that in the isoallelic mutant MGAS315�12bp(Fig. 7A). Also, we observed no significant decrease in mga tran-script in MGAS12504�codY compared to the parental strain (Fig.7A).

The observed decrease in mga transcripts in strains lackingCodY may be an indirect effect mediated through effects on otherregulators (50, 51). Thus, we next hypothesized that CodY bindsto the mga promoter in a specific manner. CodY protein was pu-rified and used in electrophoretic mobility shift assays (EMSA)with a short, double-stranded DNA oligomer representing theputative CodY binding site in the mga promoter (Fig. 1). CodY haspreviously been shown to bind specifically to its own promoter inserotype M49 GAS (50). As shown in Fig. 7B, CodY binds to itsown promoter in serotype M3 GAS, and consistent with our hy-pothesis, CodY binds to the mga promoter. Importantly, while areduction in the amount of free DNA is observed in the negativecontrol, there is an absence of a distinct shift as observed in pcodYand pmga (Fig. 7B). Together, these data demonstrate that CodYbinds to the mga promoter and increases mga transcription, andthey suggest that mutations in or around the consensus CodYbinding site affect regulation of mga through CodY.

DISCUSSION

Asymptomatic carriage is a common state of many important bac-terial human pathogens, including GAS. For decades, observa-tions have been made regarding the carrier state of these and otherpathogens, but we are in our infancy in understanding the molec-ular factors contributing to asymptomatic bacterial carriage.Studies to date examining asymptomatic carriage of bacterialpathogens have focused on broad genetic characteristics such asthe presence or absence of particular genes and epidemiologiccomparisons between infecting and colonizing strains (53–56).

A B

2 3 4 5 6 7 8 9 10 110

3

6

9

12

15

MGAS315MGAS12501MGAS12504

Exponential Transition Stationary

** **

**

0.01

0.1

1

10

**

** *

Hours

6

9

12

** **

**

0.1

1

**

0.0

0.5

1.0

1.5

2.0

Fold

-cha

nge

in tr

ansc

ript

rela

tive

to p

roS

MGAS315MGAS12501MGAS12504

7 8 9 7 8 9 7 8 9 7 8 9

mga scpAemm sclA

Hour

*P < 0.05

**

** * *

* **

*

* ** *

*

*

* * * * * *

FIG 4 The transcript levels of mga and Mga-regulated genes are decreased in the asymptomatic carrier strains MGAS12501 and MGAS12504 compared to thoseof a serotype-matched invasive strain. RNA was isolated at the time points indicated and TaqMan qRT-PCR was performed as described in Materials andMethods. (A) Transcript levels of mga over the growth phase in the invasive strain MGAS315 and the asymptomatic carrier strains MGAS12501 and MGAS12504with the first time point (3 h) used as a reference. Fold change in mga expression is indicated on the left y axis, while bacterial growth as measured by OD600 isindicated on the right y axis. Asterisks indicate a statistically significant reduction (P � 0.05) in mga transcripts in MGAS12504 (*) or both carrier strains (**)compared to MGAS315 as determined by t test (unequal variance) at the given time point. The stages of growth are shown below the x axis for reference. (B) Foldchange in transcript levels in the asymptomatic carrier strains relative to representative serotype M3 invasive strain MGAS315 for mga, emm, scpA, and sclAassayed at peak mga transcript levels (hours 7 to 9) identified in panel A. Error bars indicate standard deviation for quadruplicate samples performed on twobiologic replicates. Asterisks (*) indicate a statistically significant difference (P � 0.05, t test, unequal variance) in transcript levels compared to that of the invasivestrain.

GAS Carrier Strain mga Promoter Variation

November 2013 Volume 81 Number 11 iai.asm.org 4133

on October 24, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

The data provided herein shed new light on the carrier state ofpathogenic bacteria by demonstrating that a naturally occurringmutation in the promoter region of a virulence gene regulator inhuman carrier strains significantly alters host-pathogen interac-tion. Importantly, this seemingly minor genetic variation wouldnot have been discovered with any of the typical tools currentlyused to analyze carrier strains such as PCR to determine genecontent or multilocus sequence typing to estimate phylogeneticrelationships.

The fact that the identical 12-bp deletion in the promoter ofmga has arisen independently in two GAS carrier strains suggeststhat this genetic change provides an evolutionary advantage forthe asymptomatic carrier state. Furthermore, while it does notreplace the relative lack of asymptomatic carrier DNA sequencesavailable for comparison, the fact that the 12-bp deletion does notoccur in the published invasive serotype M3 genomes (22, 30, 31)suggests that the polymorphism is unique to the carrier strainsexamined and may confer an advantage on the asymptomatic car-riers. Given the transcriptional regulation role of Mga, it is likelythat alterations in the protein levels of genes regulated by Mga,such as M protein, are contributors to this process. As M protein is

known to induce a strong inflammatory response (20), down-regulation of M protein levels might be expected to provide aselective advantage by minimizing the immunogenicity of GASstrains in the oropharynx. As previously noted, other GAS carrierstrains have been found to have decreased M protein levels (13,14). Moreover, a study of experimental GAS pharyngitis in non-human primates found that the emm gene was downregulatedduring the asymptomatic carrier phase of the study (40). Further,Beres et al. (22) reported that an asymptomatic carrier strain con-tained a large deletion in the hypervariable region of the M pro-tein. Given that Mga directly interacts with approximately 10% ofgenes in the GAS genome (57), it is possible that additional genesregulated by Mga contribute to the virulence differences we ob-served in strains with differing promoter structures.

The alteration in the mga promoter of the asymptomatic car-rier strains in this study is due to differences in the number ofvariable-number tandem repeats (VNTRs) (Fig. 1). Because oftheir structure, VNTRs are prone to DNA replication errors re-sulting in high rates of either VNTR extension or deletion. Inprokaryotes, VNTRs in genes encoding many cell surface mole-cules have been recognized as contributing to antigenic variation

0 25 50 75 100 125 150 1750

102030405060708090

100

P < 0.0001

Hour

Perc

ent s

urvi

val

A B

MGAS315

MGAS315∆12bp

C D

MGAS315 MGAS315∆12bp

MGAS315

MGAS315∆12bp

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Expr

essi

on o

f m

ga r

elat

ive

to t

ufA

P = 0.03

0

200

400

600

800

1000

1200

Mul

tiplic

atio

n fa

ctor

P = 0.01

E

FIG 5 Introduction of the 12-bp mga promoter deletion into invasive GAS significantly reduces mga transcript levels and GAS virulence. (A) Transcript level ofmga in early stationary phase determined using TaqMan qRT-PCR as described in Materials and Methods. Error bars indicate standard deviation; P value wasdetermined by t test (unequal variance). (B) Survival curves of mice infected with indicated strains. Results were pooled from experiments in which mice wereinfected intramuscularly with 1 � 107 CFU of MGAS315 or MGAS315�12bp as described in Materials and Methods. P value was determined by log rank test. (C)Gross pathology at 72 h (original magnification, �4) of mouse hind limbs infected with MGAS315 (top) and the strain harboring the 12-bp mutation,MGAS315�12bp (bottom). Boxed regions represent areas of necrosis. (D) Histopathology of hind limb at 48 h postinfection (original magnification, �40) formice given MGAS315 and MGAS315�12bp demonstrated necrotizing fasciitis and acute myonecrosis (within black dashed region). Lesions in mice givenMGAS315 had severe necrotizing myositis and ischemic necrosis (red boxed regions), while MGAS315�12bp had retained viability of surrounding muscle andsoft tissue. (E) Multiplication after growth in human blood. Whole blood was inoculated with the indicated strains as described in Materials and Methods. Shownis multiplication after growth in blood from a single donor. Similar results were obtained with additional donors. Error bars indicate standard deviation of growthperformed in triplicate. P value determined by t test (unequal variance).

Flores et al.

4134 iai.asm.org Infection and Immunity

on October 24, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

through their effects on shifting translation reading frames or al-tering promoter activity (58). The molecular mechanisms bywhich alterations in VNTRs affect promoter activity are not en-tirely clear, but the alterations may occur by creation or elimina-tion of binding sites for transcription factors (59). The VNTR inthe mga promoter is about 20 bp proximal to the major mga tran-scriptional site (25, 26), and based on the current data (Fig. 7), itappears that a larger mga promoter VNTR region allows forgreater activity through the binding of CodY. The exact mecha-nism, e.g., activation by repressor modulation or assisting in RNApolymerase promoter escape, leading to the observed effect ofCodY is unknown but under further investigation. Regardless ofthe mechanism by which the VNTR region affects transcription,inherent instability of the mga promoter VNTR region may allowfor modulation of GAS virulence in response to the diverse envi-ronmental pressures encountered by GAS during various phasesof infection. Our finding of a VNTR that affects transcript of aglobal regulator, rather than a cell surface protein, adds to under-standing of the molecular mechanisms by which pathogenic bac-teria produce variation in virulence phenotype. Further, theVNTR in the promoter of mga, a well-known virulence regulatorof GAS, and its effect on activation through CodY had not beenpreviously described. Thus, our finding that the VNTR affectstranscription and virulence provides a new mechanism of viru-lence control through Mga.

Our data add to existing studies that suggest that increasedvirulence is not always evolutionarily beneficial to microbes. Forexample, our finding that GAS carrier strains downregulate viru-lence factor production is similar to the finding that Pseudomonasaeruginosa strains chronically isolated from a patient with cysticfibrosis had multiple mutations in virulence factor and virulencefactor regulator-encoding genes (60). Further, while our studiesdemonstrate that the 12-bp deletion in the promoter of mga con-tributes to decreased virulence, it is likely that additional factors

are required to complete the asymptomatic carrier phenotype.That is, multiple mutations, as observed in the adaptation of Pseu-domonas aeruginosa (60, 61) and Burkholderia dolosa (62) to theairways of cystic fibrosis patients, may be required for the fullasymptomatic carrier phenotype representing an adaptation togrowth in a human host. In vivo studies are needed with GAS andother important human pathogens to determine the extent thatgenetic variation contributes to the development of asymptomaticcarriage. Such carriage studies challenge the paradigm that onlythe investigation of virulence leads to new insights in bacterialpathogenesis.

Our findings have several implications. First, the decreased ex-pression of emm in the asymptomatic carrier strains suggests thatan M protein-based vaccine strategy may be suboptimal in eradi-cating the GAS carrier state given that such strains may express Mprotein at relatively low levels. Second, our genome-wide ap-proach for evaluating the molecular basis of bacterial carriagehighlights the importance of whole-genome sequencing of clinicalisolates other than those causing severe invasive infections, as hasbeen the typical approach for more than a decade. Similar ap-proaches could be taken with other critical bacterial pathogensthat are often carried asymptomatically such as S. aureus, S. pneu-moniae, and N. meningitidis. Finally, our data demonstrate theimportance of closely examining variation in noncoding regionswhen evaluating whole-genome sequencing data. The 12-bp dele-tion studied in this work is not located in a virulence factor-en-coding gene or even in an open reading frame, yet it had a signif-icant effect on GAS virulence.

The past several decades have seen dramatic advances in un-derstanding how bacterial pathogens cause a diverse array of dis-ease in humans. In contrast, little attention has been given to thestudy of molecular factors contributing to asymptomatic carriage,despite the fact that it is common for many bacterial pathogens.Our data provide new information about the GAS asymptomatic

C

0

20

40

60

80

100

Mul

tiplic

atio

n Fa

ctor

0

1

2

3

4

530456075

Fold

-cha

nge

in tr

ansc

ript

rela

tive

to tu

fA

MGAS12504 MGAS12504mgawt

* *

**

*P < 0.05

P = 0.01

A

mga emm scpA sclA

B

75

50

37

100

MG

AS

1250

4

wt

MG

AS

1250

4mga

MG

AS

1250

2 (n

eg)

Mar

ker (

kDa)

FIG 6 Repair of the 12-bp deletion in asymptomatic carrier strain MGAS12504 results in increased transcripts of mga and Mga-regulated genes and growth inhuman blood. (A) Transcript level of mga and Mga-regulated genes in early stationary phase was determined using TaqMan qRT-PCR as described in Materialsand Methods. P values were determined by t test (unequal variance). (B) Western immunoblot for M protein derived from MGAS12504 and MGAS12504mgawt.Cell wall-associated protein was extracted as described in Materials and Methods. Identical amounts of total protein were used for each sample. MGAS12502 lacksthe N-terminal region of the M protein and was used as a negative control. Proteins were transferred to a membrane and probed with primary and secondaryantibodies as described in Materials and Methods. The expected size of the M protein is 63 kDa. (C) Multiplication after growth in human blood. Whole bloodwas inoculated with the indicated strains as described in Materials and Methods. Shown is the multiplication after growth in blood from a single donor. Similarresults were obtained with additional donors. P value was determined by t test (unequal variance).

GAS Carrier Strain mga Promoter Variation

November 2013 Volume 81 Number 11 iai.asm.org 4135

on October 24, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

carrier phenotype and strongly suggest that subtle genetic varia-tion is a critical contributing factor. Further genome-wide inves-tigation using human carrier isolates may provide additional in-sights into the molecular events underlying the evolution of majorbacterial pathogens.

ACKNOWLEDGMENTS

We gratefully acknowledge pJL1055 from D. L. Kasper and helpful discus-sions with P. Sumby on allelic exchange and TaqMan expression analysis.A. Ayeras, C. Cantu, M. Watkins, T. Humbird, J. Greaver, L. Jenkins, andT. Blasdel assisted with nonhuman primate experiments.

M.K. was supported in part by an American Heart Association Scien-tist Development Grant. A.R.F. was supported by the Pediatric InfectiousDisease Society-St. Jude Children’s Research Hospital Fellowship Award

in Basic Research and the Robert Wood Johnson-Harold Amos MedicalFaculty Development Program Award.

The authors have no conflicts of interest to report.

REFERENCES1. Yazdankhah SP, Caugant DA. 2004. Neisseria meningitidis: an overview

of the carriage state. J. Med. Microbiol. 53:821– 832.2. Bogaert D, van Belkum A, Sluijter M, Luijendijk A, de Groot R, Rumke

HC, Verbrugh HA, Hermans PW. 2004. Colonisation by Streptococcuspneumoniae and Staphylococcus aureus in healthy children. Lancet 363:1871–1872.

3. Graham PL, III, Lin SX, Larson EL. 2006. A U.S. population-basedsurvey of Staphylococcus aureus colonization. Ann. Intern. Med. 144:318 –325.

4. Cockerill FR, III, MacDonald KL, Thompson RL, Roberson F, KohnerPC, Besser-Wiek J, Manahan JM, Musser JM, Schlievert PM, Talbot J,Frankfort B, Steckelberg JM, Wilson WR, Osterholm MT. 1997. Anoutbreak of invasive group A streptococcal disease associated with highcarriage rates of the invasive clone among school-aged children. JAMA277:38 – 43.

5. Crum NF, Russell KL, Kaplan EL, Wallace MR, Wu J, Ashtari P, MorrisDJ, Hale BR. 2005. Pneumonia outbreak associated with group A Strep-tococcus species at a military training facility. Clin. Infect. Dis. 40:511–518.

6. Engelgau MM, Woernle CH, Schwartz B, Vance NJ, Horan JM. 1994.Invasive group A streptococcus carriage in a child care centre after a fatalcase. Arch. Dis. Child. 71:318 –322.

7. Haukness HA, Tanz RR, Thomson RB, Jr, Pierry DK, Kaplan EL, BeallB, Johnson D, Hoe NP, Musser JM, Shulman ST. 2002. The heteroge-neity of endemic community pediatric group A streptococcal pharyngealisolates and their relationship to invasive isolates. J. Infect. Dis. 185:915–920.

8. Jackson LA, Neuzil KM, Yu O, Benson P, Barlow WE, Adams AL,Hanson CA, Mahoney LD, Shay DK, Thompson WW. 2003. Effective-ness of pneumococcal polysaccharide vaccine in older adults. N. Engl. J.Med. 348:1747–1755.

9. Millar EV, Watt JP, Bronsdon MA, Dallas J, Reid R, Santosham M,O’Brien KL. 2008. Indirect effect of 7-valent pneumococcal conjugatevaccine on pneumococcal colonization among unvaccinated householdmembers. Clin. Infect. Dis. 47:989 –996.

10. Poehling KA, Talbot TR, Griffin MR, Craig AS, Whitney CG, Zell E,Lexau CA, Thomas AR, Harrison LH, Reingold AL, Hadler JL, FarleyMM, Anderson BJ, Schaffner W. 2006. Invasive pneumococcal diseaseamong infants before and after introduction of pneumococcal conjugatevaccine. JAMA 295:1668 –1674.

11. Roy S, Kaplan EL, Rodriguez B, Schreiber JR, Salata RA, Palavecino E,John CC. 2003. A family cluster of five cases of group A streptococcalpneumonia. Pediatrics 112:e61– e65. doi:10.1542/peds.112.1.e61.

12. Krause RM, Rammelkamp CH, Jr. 1962. Studies of the carrier statefollowing infection with group A streptococci. II. Infectivity of strepto-cocci isolated during acute pharyngitis and during the carrier state. J. Clin.Invest. 41:575–578.

13. Krause RM, Rammelkamp CH, Jr, Denny FW, Jr, Wannamaker LW.1962. Studies of the carrier state following infection with group A strep-tococci. 1. Effect of climate. J. Clin. Invest. 41:568 –574.

14. Rothbard S, Watson RF. 1948. Variation occurring in group A strep-tococci during human infection. Progressive loss of M substance cor-related with increasing susceptibility to bacteriostasis. J. Exp. Med.87:521–533.

15. Johnson DR, Kurlan R, Leckman J, Kaplan EL. 2010. The humanimmune response to streptococcal extracellular antigens: clinical, diag-nostic, and potential pathogenetic implications. Clin. Infect. Dis. 50:481–490.

16. Martin JM, Green M, Barbadora KA, Wald ER. 2004. Group A strepto-cocci among school-aged children: clinical characteristics and the carrierstate. Pediatrics 114:1212–1219.

17. Carapetis JR. 2005. Current evidence for the burden of group A strepto-coccal diseases. World Health Organization, Geneva, Switzerland.

18. Committee on Infectious Diseases. 2009. Group A streptococcal infec-tions, p 616 – 628. In Pickering LK (ed), Red book, 28th ed. AmericanAcademy of Pediatrics, Elk Grove Village, IL.

19. Shaikh N, Leonard E, Martin JM. 2010. Prevalence of streptococcal

A

B DNArCodY

+ + ++

+++

+ + +++ +++

+ ++- - -

pcodY pmga non-specific DNA

Free

Bound

MGAS1250

4

MGAS1250

4mgaw

t

MGAS1250

4co

dY0

1

2

3

Fold

-cha

nge

in m

ga tr

ansc

ript

rela

tive

to tu

fA

P = 0.13

Carrier

P = 0.015

MGAS315

MGAS315

12bp

MGAS315

codY

0.0

0.5

1.0

1.5

Fold

-cha

nge

in m

ga tr

ansc

ript

rela

tive

to tu

fA

P = 0.022

Invasive

FIG 7 CodY influences transcript levels of mga and binds to the mga pro-moter. (A) Transcript levels of mga in invasive (left, MGAS315) and carrier(right, MGAS12504) isoallelic and isogenic mutants. Transcript levels of mgain early stationary phase were determined using TaqMan qRT-PCR as de-scribed in Materials and Methods. Error bars indicate standard deviation; Pvalues were determined by t test (unequal variance). (B) Purified rCodY pro-tein binds specifically to the operator sequences from the mga promoter. Therecombinant CodY-DNA interactions were probed by electrophoretic mobil-ity shift assay. The DNA probes pcodY and pmga are oligoduplexes with theputative operator sequences of CodY from the promoters of codY and mga,respectively; “nonspecific DNA” is a DNA probe with a random sequence. Areaction mixture containing a 0.2 �M concentration of the probe and purifiedrCodY to a final concentration of either 8.3 �M () or 12.5 �M () wasresolved by 8% native PAGE. The positions of free probe and protein-boundprobe are labeled and indicated by arrows.

Flores et al.

4136 iai.asm.org Infection and Immunity

on October 24, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

pharyngitis and streptococcal carriage in children: a meta-analysis. Pedi-atrics 126:e557– e564. doi:10.1542/peds.2009-2648.

20. Lancefield RC. 1962. Current knowledge of type-specific M antigens ofgroup A streptococci. J. Immunol. 89:307–313.

21. Lancefield RC. 1959. Persistence of type-specific antibodies in manfollowing infection with group A streptococci. J. Exp. Med. 110:271–292.

22. Beres SB, Richter EW, Nagiec MJ, Sumby P, Porcella SF, DeLeo FR,Musser JM. 2006. Molecular genetic anatomy of inter- and intraserotypevariation in the human bacterial pathogen group A Streptococcus. Proc.Natl. Acad. Sci. U. S. A. 103:7059 –7064.

23. Caparon MG, Scott JR. 1987. Identification of a gene that regulatesexpression of M protein, the major virulence determinant of group Astreptococci. Proc. Natl. Acad. Sci. U. S. A. 84:8677– 8681.

24. McIver KS, Heath AS, Green BD, Scott JR. 1995. Specific binding of theactivator Mga to promoter sequences of the emm and scpA genes in thegroup A streptococcus. J. Bacteriol. 177:6619 – 6624.

25. McIver KS, Thurman AS, Scott JR. 1999. Regulation of mga transcrip-tion in the group A streptococcus: specific binding of mga within its ownpromoter and evidence for a negative regulator. J. Bacteriol. 181:5373–5383.

26. Okada N, Geist RT, Caparon MG. 1993. Positive transcriptional controlof mry regulates virulence in the group A streptococcus. Mol. Microbiol.7:893–903.

27. Hoe NP, Fullerton KE, Liu M, Peters JE, Gackstetter GD, Adams GJ,Musser JM. 2003. Molecular genetic analysis of 675 group A streptococ-cus isolates collected in a carrier study at Lackland Air Force Base, SanAntonio, Texas. J. Infect. Dis. 188:818 – 827.

28. Musser JM, Hauser AR, Kim MH, Schlievert PM, Nelson K, SelanderRK. 1991. Streptococcus pyogenes causing toxic-shock-like syndrome andother invasive diseases: clonal diversity and pyrogenic exotoxin expres-sion. Proc. Natl. Acad. Sci. U. S. A. 88:2668 –2672.

29. Beres SB, Sylva GL, Barbian KD, Lei B, Hoff JS, Mammarella ND, LiuMY, Smoot JC, Porcella SF, Parkins LD, Campbell DS, Smith TM,McCormick JK, Leung DY, Schlievert PM, Musser JM. 2002. Genomesequence of a serotype M3 strain of group A Streptococcus: phage-encoded toxins, the high-virulence phenotype, and clone emergence.Proc. Natl. Acad. Sci. U. S. A. 99:10078 –10083.

30. Beres SB, Carroll RK, Shea PR, Sitkiewicz I, Martinez-Gutierrez JC,Low DE, McGeer A, Willey BM, Green K, Tyrrell GJ, Goldman TD,Feldgarden M, Birren BW, Fofanov Y, Boos J, Wheaton WD, HonischC, Musser JM. 2010. Molecular complexity of successive bacterial epi-demics deconvoluted by comparative pathogenomics. Proc. Natl. Acad.Sci. U. S. A. 107:4371– 4376.

31. Beres SB, Sylva GL, Sturdevant DE, Granville CN, Liu M, Ricklefs SM,Whitney AR, Parkins LD, Hoe NP, Adams GJ, Low DE, DeLeo FR,McGeer A, Musser JM. 2004. Genome-wide molecular dissection ofserotype M3 group A Streptococcus strains causing two epidemics of in-vasive infections. Proc. Natl. Acad. Sci. U. S. A. 101:11833–11838.

32. Gryllos I, Grifantini R, Colaprico A, Cary ME, Hakansson A, CareyDW, Suarez-Chavez M, Kalish LA, Mitchell PD, White GL, WesselsMR. 2008. PerR confers phagocytic killing resistance and allows pharyn-geal colonization by group A Streptococcus. PLoS Pathog. 4:e1000145.doi:10.1371/journal.ppat.1000145.

33. Li J, Kasper DL, Ausubel FM, Rosner B, Michel JL. 1997. Inactivation ofthe alpha C protein antigen gene, bca, by a novel shuttle/suicide vectorresults in attenuation of virulence and immunity in group B Streptococ-cus. Proc. Natl. Acad. Sci. U. S. A. 94:13251–13256.

34. Carroll RK, Shelburne SA, III, Olsen RJ, Suber B, Sahasrabhojane P,Kumaraswami M, Beres SB, Shea PR, Flores AR, Musser JM. 2011.Naturally occurring single amino acid replacements in a regulatory pro-tein alter streptococcal gene expression and virulence in mice. J. Clin.Invest. 121:1956 –1968.

35. Lukomski S, Hoe NP, Abdi I, Rurangirwa J, Kordari P, Liu M, Dou SJ,Adams GG, Musser JM. 2000. Nonpolar inactivation of the hypervariablestreptococcal inhibitor of complement gene (sic) in serotype M1 Strepto-coccus pyogenes significantly decreases mouse mucosal colonization. In-fect. Immun. 68:535–542.

36. Olsen RJ, Sitkiewicz I, Ayeras AA, Gonulal VE, Cantu C, Beres SB,Green NM, Lei B, Humbird T, Greaver J, Chang E, Ragasa WP,Montgomery CA, Cartwright J, Jr, McGeer A, Low DE, Whitney AR,Cagle PT, Blasdel TL, DeLeo FR, Musser JM. 2010. Decreased necrotiz-

ing fasciitis capacity caused by a single nucleotide mutation that alters amultiple gene virulence axis. Proc. Natl. Acad. Sci. U. S. A. 107:888 – 893.

37. Lancefield RC. 1957. Differentiation of group A streptococci with a com-mon R antigen into three serological types, with special reference to thebactericidal test. J. Exp. Med. 106:525–544.

38. Flores AR, Jewell BE, Fittipaldi N, Beres SB, Musser JM. 2012. Humandisease isolates of serotype m4 and m22 group a streptococcus lack genesrequired for hyaluronic acid capsule biosynthesis. mBio 3:e00413–12. doi:10.1128/mBio.00413-12.

39. Chaussee MS, Watson RO, Smoot JC, Musser JM. 2001. Identificationof Rgg-regulated exoproteins of Streptococcus pyogenes. Infect. Immun.69:822– 831.

40. Virtaneva K, Porcella SF, Graham MR, Ireland RM, Johnson CA,Ricklefs SM, Babar I, Parkins LD, Romero RA, Corn GJ, Gardner DJ,Bailey JR, Parnell MJ, Musser JM. 2005. Longitudinal analysis of thegroup A Streptococcus transcriptome in experimental pharyngitis in cyn-omolgus macaques. Proc. Natl. Acad. Sci. U. S. A. 102:9014 –9019.

41. Lukomski S, Nakashima K, Abdi I, Cipriano VJ, Ireland RM, Reid SD,Adams GG, Musser JM. 2000. Identification and characterization of thescl gene encoding a group A Streptococcus extracellular protein virulencefactor with similarity to human collagen. Infect. Immun. 68:6542– 6553.

42. Schroeder B, Boyle MD, Sheerin BR, Asbury AC, Lottenberg R. 1999.Species specificity of plasminogen activation and acquisition of surface-associated proteolytic activity by group C streptococci grown in plasma.Infect. Immun. 67:6487– 6495.

43. Wang H, Lottenberg R, Boyle MD. 1995. Analysis of the interaction ofgroup A streptococci with fibrinogen, streptokinase and plasminogen. Mi-crob. Pathog. 18:153–166.

44. McIver KS, Scott JR. 1997. Role of mga in growth phase regulation ofvirulence genes of the group A streptococcus. J. Bacteriol. 179:5178 –5187.

45. Ji Y, McLandsborough L, Kondagunta A, Cleary PP. 1996. C5a pepti-dase alters clearance and trafficking of group A streptococci by infectedmice. Infect. Immun. 64:503–510.

46. Wexler DE, Chenoweth DE, Cleary PP. 1985. Mechanism of action of thegroup A streptococcal C5a inactivator. Proc. Natl. Acad. Sci. U. S. A.82:8144 – 8148.

47. Caswell CC, Lukomska E, Seo NS, Hook M, Lukomski S. 2007. Scl1-dependent internalization of group A Streptococcus via direct interactionswith the alpha2beta(1) integrin enhances pathogen survival and re-emergence. Mol. Microbiol. 64:1319 –1331.

48. Ashbaugh CD, Moser TJ, Shearer MH, White GL, Kennedy RC,Wessels MR. 2000. Bacterial determinants of persistent throat coloni-zation and the associated immune response in a primate model ofhuman group A streptococcal pharyngeal infection. Cell. Microbiol.2:283–292.

49. Stenz L, Francois P, Whiteson K, Wolz C, Linder P, Schrenzel J. 2011.The CodY pleiotropic repressor controls virulence in gram-positivepathogens. FEMS Immunol. Med. Microbiol. 62:123–139.

50. Kreth J, Chen Z, Ferretti J, Malke H. 2011. Counteractive balancing oftranscriptome expression involving CodY and CovRS in Streptococcuspyogenes. J. Bacteriol. 193:4153– 4165.

51. Malke H, Ferretti JJ. 2007. CodY-affected transcriptional gene expressionof Streptococcus pyogenes during growth in human blood. J. Med. Micro-biol. 56:707–714.

52. Malke H, Steiner K, McShan WM, Ferretti JJ. 2006. Linking the nutri-tional status of Streptococcus pyogenes to alteration of transcriptional geneexpression: the action of CodY and RelA. Int. J. Med. Microbiol. 296:259 –275.

53. Caugant DA, Tzanakaki G, Kriz P. 2007. Lessons from meningococcalcarriage studies. FEMS Microbiol. Rev. 31:52– 63.

54. Monecke S, Luedicke C, Slickers P, Ehricht R. 2009. Molecular epide-miology of Staphylococcus aureus in asymptomatic carriers. Eur. J. Clin.Microbiol. Infect. Dis. 28:1159 –1165.

55. Schoen C, Blom J, Claus H, Schramm-Gluck A, Brandt P, Muller T,Goesmann A, Joseph B, Konietzny S, Kurzai O, Schmitt C, Friedrich T,Linke B, Vogel U, Frosch M. 2008. Whole-genome comparison of diseaseand carriage strains provides insights into virulence evolution in Neisseriameningitidis. Proc. Natl. Acad. Sci. U. S. A. 105:3473–3478.

56. Sivaraman K, Venkataraman N, Tsai J, Dewell S, Cole AM. 2008.Genome sequencing and analysis reveals possible determinants of Staph-ylococcus aureus nasal carriage. BMC Genomics 9:433. doi:10.1186/1471-2164-9-433.

57. Ribardo DA, McIver KS. 2006. Defining the Mga regulon: Comparative

GAS Carrier Strain mga Promoter Variation

November 2013 Volume 81 Number 11 iai.asm.org 4137

on October 24, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

transcriptome analysis reveals both direct and indirect regulation by Mgain the group A streptococcus. Mol. Microbiol. 62:491–508.

58. Moxon R, Bayliss C, Hood D. 2006. Bacterial contingency loci: the role ofsimple sequence DNA repeats in bacterial adaptation. Annu. Rev. Genet.40:307–333.

59. Martin P, Makepeace K, Hill SA, Hood DW, Moxon ER. 2005. Micro-satellite instability regulates transcription factor binding and gene expres-sion. Proc. Natl. Acad. Sci. U. S. A. 102:3800 –3804.

60. Smith EE, Buckley DG, Wu Z, Saenphimmachak C, Hoffman LR,D’Argenio DA, Miller SI, Ramsey BW, Speert DP, Moskowitz SM,Burns JL, Kaul R, Olson MV. 2006. Genetic adaptation by Pseudomonas

aeruginosa to the airways of cystic fibrosis patients. Proc. Natl. Acad. Sci.U. S. A. 103:8487– 8492.

61. Yang L, Jelsbak L, Marvig RL, Damkiaer S, Workman CT, Rau MH,Hansen SK, Folkesson A, Johansen HK, Ciofu O, Hoiby N, SommerMO, Molin S. 2011. Evolutionary dynamics of bacteria in a human hostenvironment. Proc. Natl. Acad. Sci. U. S. A. 108:7481–7486.

62. Lieberman TD, Michel JB, Aingaran M, Potter-Bynoe G, Roux D, DavisMR, Jr, Skurnik D, Leiby N, LiPuma JJ, Goldberg JB, McAdam AJ,Priebe GP, Kishony R. 2011. Parallel bacterial evolution within multiplepatients identifies candidate pathogenicity genes. Nat. Genet. 43:1275–1280.

Flores et al.

4138 iai.asm.org Infection and Immunity

on October 24, 2016 by guest

http://iai.asm.org/

Dow

nloaded from