Full-Genome Dissection of an Epidemic of Severe Invasive Disease Caused by a Hypervirulent, Recently Emerged Clone of Group A Streptococcus

Published:February 13, 2012DOI:https://doi.org/10.1016/j.ajpath.2011.12.037
      Group A Streptococcus (GAS) causes an exceptionally broad range of infections in humans, from relatively mild pharyngitis and skin infections to life-threatening necrotizing fasciitis and toxic shock syndrome. An epidemic of severe invasive human infections caused by type emm59 GAS, heretofore an exceedingly rare cause of disease, spread west to east across Canada over a 3-year period (2006 to 2008). By sequencing the genomes of 601 epidemic, historic, and other emm59 organisms, we discovered that a recently emerged, genetically distinct emm59 clone is responsible for the Canadian epidemic. Using near-real-time genome sequencing, we were able to show spread of the Canadian epidemic clone into the United States. The extensive genome data permitted us to identify patterns of geographic dissemination as well as links between emm59 subclonal lineages that cause infections. Mouse and nonhuman primate models of infection demonstrated that the emerged clone is unusually virulent. Transmission of epidemic emm59 strains may have occurred primarily by skin contact, as suggested by an experimental model of skin transmission. In addition, the emm59 strains had a significantly impaired ability to persist in human saliva and to colonize the oropharynx of mice, and seldom caused human pharyngitis. Our study contributes new information to the rapidly emerging field of molecular pathogenomics of bacterial epidemics and illustrates how full-genome data can be used to precisely illuminate the landscape of strain dissemination during a bacterial epidemic.
      See related Commentary on page 1358.
      Group A Streptococcus (GAS; also known as Streptococcus pyogenes) causes a variety of diseases, ranging from relatively mild pharyngitis and skin infections to life-threatening necrotizing fasciitis and streptococcal toxic shock syndrome. The M protein, encoded by the emm gene, is an important virulence factor and an epidemiological marker that has become the standard for typing GAS isolates.
      • Lancefield R.C.
      The antigenic complex of Streptococcus haemolyticus: I Demonstration of a type-specific substance in extracts of Streptococcus haemolyticus.
      • Scott J.R.
      • Pulliam W.M.
      • Hollingshead S.K.
      • Fischetti V.A.
      Relationship of M protein genes in group A streptococci.
      • Manjula B.N.
      • Acharya A.S.
      • Fairwell T.
      • Fischetti V.A.
      Antigenic domains of the streptococcal Pep M5 protein Localization of epitopes crossreactive with type 6 M protein and identification of a hypervariable region of the M molecule.
      Recently, public health authorities in Canada recognized a striking increase in the frequency and severity of invasive infections caused by serotype M59 (type emm59) GAS. Studies of large numbers of GAS isolates causing invasive infections and pharyngitis worldwide have shown that emm59 strains are a very uncommon cause of human disease.
      • Luca-Harari B.
      • Darenberg J.
      • Neal S.
      • Siljander T.
      • Strakova L.
      • Tanna A.
      • Creti R.
      • Ekelund K.
      • Koliou M.
      • Tassios P.T.
      • van der Linden M.
      • Straut M.
      • Vuopio-Varkila J.
      • Bouvet A.
      • Efstratiou A.
      • Schalén C.
      • Henriques-Normark B.
      • Jasir A.
      Strep-EURO Study Group
      Clinical and microbiological characteristics of severe Streptococcus pyogenes disease in Europe.
      • Steer A.C.
      • Law I.
      • Matatolu L.
      • Beall B.W.
      • Carapetis J.R.
      Global emm type distribution of group A streptococci: systematic review and implications for vaccine development.
      In striking contrast, from June 2006 to December 2008, emm59 strains went from being extremely rare to the most common emm type causing severe infections in Canada; they were eventually recovered from >500 invasive cases.
      • Tyrrell G.J.
      • Lovgren M.
      • St Jean T.
      • Hoang L.
      • Patrick D.M.
      • Horsman G.
      • Van Caeseele P.
      • Sieswerda L.E.
      • McGeer A.
      • Laurence R.A.
      • Bourgault A.M.
      • Low D.E.
      Epidemic of group A Streptococcus M/emm59 causing invasive disease in Canada.
      The rapid emergence and spread of emm59 strains has created a unique opportunity to address several poorly understood aspects of bacterial epidemics. First, what microbiological factors, if any, have contributed? Second, does this epidemic represent the emergence of a single distinct genotype (clone) of emm59 GAS, or have multiple clones emerged simultaneously? Third, if the emm59 strains causing this epidemic have had a clonal origin, what are the rate, extent, and location of genome diversification over the short epidemic period? Finally, do these epidemic emm59 strains have special virulence attributes compared with other emm59 strains? Answers to these questions are important to obtain as we evolve toward a population genomics understanding of bacterial epidemics, and they also are likely to have implications for the public health response to an accidental or deliberate release of a human pathogen.
      Here, we report that this epidemic was caused by the emergence of a distinct, unusually virulent genetic clone of emm59 GAS. We also report dissemination of the Canadian epidemic clone into four states in the United States. Full-genome sequencing of all available epidemic strains allowed us to resolve distinct geographic patterns of subclone dissemination. Our findings add new information to the emerging field of molecular pathogenomics of bacterial epidemics.

      Materials and Methods

      Strains

      We analyzed a total of 601 emm59 strains. For descriptions of the strains and summary of their isolation sites, see Supplemental Tables S1 and S2, respectively (available at http://ajp.amjpathol.org). As part of a passive surveillance project for invasive GAS infections, provincial public health laboratories across Canada submit invasive GAS strains for emm typing to the National Centre for Streptococcus in Edmonton. Thus, the sample of 514 invasive epidemic emm59 GAS strains used in the present study (recovered from 2006 to the beginning of 2010) is not comprehensive; rather, it represents the majority of sterile-site emm59 organisms causing invasive infections during the period. Seven additional emm59 organisms isolated in Canada before the epidemic (1996 and 2004) were also received form the National Centre for Streptococcus. We also analyzed 11 emm59 organisms identified in a convenience sample of 4906 pharyngitis strains cultured since 2003 from six Ontario provincial public health and private diagnostic laboratories throughout the province (Toronto, London, Ottawa, Sudbury, Elliot Lake, and Thunder Bay), as well as 13 emm59 strains identified in a sample of 532 pharyngitis strains recovered from Alberta, Nunavut, and Northwest Territories. Nine sporadic emm59 sterile-site isolates cultured from patients with invasive infections between 1999 and 2008 in Spain
      • Ardanuy C.
      • Rolo D.
      • Domenech A.
      • Tubau F.
      • Ayats J.
      • Grau I.
      • Pallares R.
      • Martin R.
      • Liñares J.
      Trends in invasive group A streptococci in adult patients in Barcelona (abstract) 19th European Congress of Clinical Microbiology and Infectious Diseases, Helsinki, Finland, 16–19 May 2009.
      were also studied. Two emm59 strains from the collection of J.M.M. were analyzed, strain MGAS1882, a U.S. isolate described in 1974,
      • Dillon H.C.
      • Dillon M.S.
      New streptococcal serotypes causing pyoderma and acute glomerulonephritis types 59, 60, and 61.
      and strain MGAS360, an isolate from Japan from the 1960s. Thirty-nine emm59 strains from the Active Bacterial Core surveillance program conducted by the U.S. Centers for Disease Control and Prevention were also studied. These represent virtually all emm59 strains collected by the Active Bacterial Core surveillance program between 2000 and 2009. Six emm59 strains were recovered in 2010 from patients in southeastern Montana with temporally and geographically clustered invasive infections. Strain MGAS5005 is a representative type emm1 GAS strain from the collection of J.M.M.

      Genome Sequencing, Polymorphism Discovery, and Phylogenetic Analysis

      The genomes of historic MGAS1882 and epidemic MGAS15252 emm59 GAS strains were sequenced to closure (GenBank accession no. CP003121 and CP003116, respectively) by using a combination of two massively parallel next-generation sequencing technologies [Illumina (San Diego, CA) sequencing by synthesis
      • Bentley D.R.
      • Balasubramanian S.
      • Swerdlow H.P.
      • Smith G.P.
      • Milton J.
      • Brown C.G.
      • et al.
      Accurate whole human genome sequencing using reversible terminator chemistry.
      and Roche (Indianapolis, IN) 454 Life Sciences pyrosequencing
      • Margulies M.
      • Egholm M.
      • Altman W.E.
      • Attiya S.
      • Bader J.S.
      • Bemben L.A.
      • et al.
      Genome sequencing in microfabricated high-density picolitre reactors.
      ) and conventional Sanger sequencing. De novo genome assembly from the hybrid sequencing technologies was accomplished using MIRA3 software (http://www.chevreux.org/projects_mira.html). Gap closure was conducted using sequence reads generated with an ABI 3730XL genetic analyzer (Applied Biosystems, Foster City, CA) using PCR products that spanned putative gaps in the assembled genomes. The final genome assemblies had an average ∼200-fold coverage of high-quality sequence. Genome sequencing of the remaining 599 emm59 strains was performed using an Illumina GAII instrument according to the manufacturer's instructions, either individually or using a strategy in which DNA from two strains was pooled in a 1:1 ratio. When a pooled-DNA strategy was adopted for genome sequencing, DNA samples of two emm59 strains were mixed in a 1:1 ratio, and then processed according to the Illumina system manufacturer's instructions. TagDust,
      • Lassmann T.
      • Hayashizaki Y.
      • Daub C.O.
      TagDust–a program to eliminate artifacts from next generation sequencing data.
      FastQC (http://www.bioinformatics.bbsrc.ac.uk/projects/fastqc), and FastX (http://hannonlab.cshl.edu/fastx_toolkit) toolkits were used to parse multiplexed sequencing reads, remove barcode information, and perform run quality control analyses. Polymorphism discovery was performed using the variant ascertainment algorithm VAAL
      • Nusbaum C.
      • Ohsumi T.K.
      • Gomez J.
      • Aquadro J.
      • Victor T.C.
      • Warren R.M.
      • Hung D.T.
      • Birren B.W.
      • Lander E.S.
      • Jaffe D.B.
      Sensitive, specific polymorphism discovery in bacteria using massively parallel sequencing.
      or, for genome sequences obtained from pooled DNA samples, the Mapping and Assembly with Quality (MAQ) tool (http://maq.sourceforge.net). Reads were aligned to the reference genome using the Mosaik assembler (http://bioinformatics.bc.edu/marthlab/Mosaik). Unaligned reads were placed into contigs using the Velvet de novo assembler.
      • Zerbino D.R.
      • Birney E.
      Velvet: algorithms for de novo short read assembly using de Bruijn graphs.
      Contigs >100 nucleotides in length were then used to search the National Center for Biotechnology Information (NCBI) nonredundant database using BLAST.
      • Altschul S.F.
      • Madden T.L.
      • Schäffer A.A.
      • Zhang J.
      • Zhang Z.
      • Miller W.
      • Lipman D.J.
      Gapped BLAST and PSI-BLAST: a new generation of protein database search programs.

      Phylogenetic Analysis

      A matrix file containing the genotype of all strains at each polymorphic locus was created from the VAAL polymorphism output data using a custom script. Insertions and deletions (indels) were not considered for phylogenetic analysis. For each individual strain, single-nucleotide polymorphisms (SNPs) were concatenated in order of occurrence relative to the MGAS15252 reference genome and converted to a multiFASTA sequence. ClustalW
      • Thompson J.D.
      • Higgins D.G.
      • Gibson T.J.
      CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice.
      was used to align and generate a guide tree for the FASTA sequences. A neighbor-joining phylogenetic tree was created using SplitsTree.
      • Huson D.H.
      • Bryant D.
      Application of phylogenetic networks in evolutionary studies.

      Growth of emm59 GAS in Human Saliva

      The ability of emm59 GAS strain MGAS15249 to grow and persist in human saliva was evaluated as described previously.
      • Shelburne 3rd, S.A.
      • Granville C.
      • Tokuyama M.
      • Sitkiewicz I.
      • Patel P.
      • Musser J.M.
      Growth characteristics of and virulence factor production by group A Streptococcus during cultivation in human saliva.
      Briefly, GAS was grown overnight in Todd-Hewitt broth supplemented with 0.2% yeast extract (THY; BD Biosciences, Sparks, MD), diluted 1:50 with fresh THY, and grown to mid-logarithmic phase (optical density OD600nm of ∼0.5). A 1:50 dilution of the mid-logarithmic-phase culture was made into 10 mL of human saliva collected from five healthy volunteers as described by Shelburne et al
      • Shelburne 3rd, S.A.
      • Granville C.
      • Tokuyama M.
      • Sitkiewicz I.
      • Patel P.
      • Musser J.M.
      Growth characteristics of and virulence factor production by group A Streptococcus during cultivation in human saliva.
      and pooled to minimize the effects of donor variation. After 4.5 hours of growth, a second 1:50 dilution was made into 10 mL of fresh saliva. Aliquots were removed every 1.5 hours for the first 9 hours, then daily for 13 days and at different time periods thereafter. Samples were serially diluted 10-fold in PBS and plated in duplicate on trypticase soy agar plates supplemented with 5% sheep blood (BD Biosciences). The plates were incubated overnight, and then colonies were enumerated. All incubations were at 37°C with 5% CO2. For comparison purposes, we included type emm1 GAS strain MGAS5005.

      Virulence in Mouse Infection Models

      The virulence of strains MGAS1882, MGAS15249, and MGAS15252 was compared using mouse models of invasive infection, as described previously.
      • Olsen R.J.
      • Musser J.M.
      Molecular pathogenesis of necrotizing fasciitis.
      • Sumby P.
      • Zhang S.
      • Whitney A.R.
      • Falugi F.
      • Grandi G.
      • Graviss E.A.
      • Deleo F.R.
      • Musser J.M.
      A chemokine-degrading extracellular protease made by group A Streptococcus alters pathogenesis by enhancing evasion of the innate immune response.
      To study GAS virulence in a skin and soft tissue model, 4- to 5-week-old immunocompetent SKH1-hrBR hairless female mice, 18 to 20 g body weight (Charles River BRF, Houston, TX), were randomly assigned to treatment groups (n = 15 mice per strain) and inoculated in the subcutaneous tissue overlying the neck with 1 × 107 colony-forming units (CFU) of the indicated GAS strain in 100 μL PBS. Lesion size was measured daily with a caliper for 12 days and then every second day until day 20 after inoculation. Abscess area caused by each strain was compared using generalized estimating equations (Stata 10 software; StataCorp, College Station, TX), with P < 0.05 considered to be statistically significant. To study GAS virulence in a necrotizing fasciitis model, 4- to 5-week-old immunocompetent CD1 mice, 18 to 20 g body weight (Harlan Laboratories, Houston, TX), were randomly assigned to treatment groups (n = 15 mice per strain) and inoculated in the right hindlimb to a uniform depth with 5 × 106 CFU of GAS in 100 μL PBS; mice were monitored for 10 days and sacrificed when near-mortality was reached. Control animals were sham-inoculated with PBS. Mortality caused by each strain was compared using a Kaplan-Meier log rank test (Prism 4; GraphPad Software, La Jolla, CA), with P < 0.05 considered to be statistically significant.
      For the quantitative culture experiment in muscle, 20 additional mice per group per time point were used. Limbs were homogenized using an Omni International (Kennesaw, GA) homogenizer and CFUs were determined by culturing serial dilutions. To study GAS colonization in the mouse oropharynx, female CD1 mice were randomly assigned to treatment groups (n = 9 mice per strain) and inoculated intranasally with 1 × 107 CFU of GAS in 50 μL PBS; CFUs were recovered daily with throat swabs, as described previously.
      • Shelburne 3rd, S.A.
      • Sumby P.
      • Sitkiewicz I.
      • Okorafor N.
      • Granville C.
      • Patel P.
      • Voyich J.
      • Hull R.
      • DeLeo F.R.
      • Musser J.M.
      Maltodextrin utilization plays a key role in the ability of group A Streptococcus to colonize the oropharynx.
      The presence of GAS was confirmed using a latex agglutination assay for the GAS carbohydrate antigen (BD Biosciences).
      To study GAS transmission, a new skin transmission model was developed. Briefly, female SKH1-hrBR mice were treated with a standard 1 cm full-thickness incision, running cranial to caudal, to the epidermis overlying the dorsal-lateral aspect of each hindlimb at the femoral head using a sterile no. 22 scalpel blade (Becton Dickinson, Franklin Lakes, NJ). Each site was visually inspected to confirm a uniform incision. In each cage of five mice, one mouse had 1 × 107 CFU of GAS applied directly to its incisions (seeder mouse). The other four mice were not inoculated (receiver mice), so that infection could occur only by transmission of GAS from the inoculated casemate (n = 16 receiver sites per strain). CFUs were recovered daily with swabs of each incision. The identity of recovered GAS was confirmed by sequencing the emm gene. For all experiments, bacterial doses were prepared at the time of inoculation from previously quantified cryopreserved stock cultures, and the actual dose given was confirmed by CFU enumeration. For histopathological evaluation, skin biopsies or whole limbs were processed using standard methods, slides were examined by a blinded pathologist (R.J.O.), and representative photomicrographs were acquired using a BX5 microscope fitted with a DP70 digital camera (Olympus, Tokyo, Japan). Mouse experiments were approved by the Institutional Animal Care and Use Committee of the Methodist Hospital Research Institute (Houston, TX).

      Virulence in the Nonhuman Primate Infection Model of Necrotizing Fasciitis

      The virulence of strains MGAS1882 and MGAS15252 was assessed in a necrotizing fasciitis model using cynomolgus macaques, as described previously.
      • Olsen R.J.
      • Musser J.M.
      Molecular pathogenesis of necrotizing fasciitis.
      Briefly, adult male cynomolgus macaques (Macaca fascicularis) (Charles River BRF, Houston, TX) were anesthetized, outfitted with a transdermal fentanyl patch (25 µg/h), and inoculated intramuscularly in the anterior thigh to a uniform depth with 1 × 108 CFU/kg of either the MGAS1882 or the MGAS15252 strain (n = 3 per group). The animals were observed continuously, sacrificed in pairs when judged to be near death, and necropsied. The study protocol was approved by the University of Houston Animal Care and Use Committee.

      Results

      Rapid Increase and Spread of Invasive Infections Caused by emm59 GAS throughout Canada

      Although GAS is an important cause of human morbidity and mortality worldwide, emm59 strains rarely cause invasive cases and seldom cause pharyngitis.
      • Luca-Harari B.
      • Darenberg J.
      • Neal S.
      • Siljander T.
      • Strakova L.
      • Tanna A.
      • Creti R.
      • Ekelund K.
      • Koliou M.
      • Tassios P.T.
      • van der Linden M.
      • Straut M.
      • Vuopio-Varkila J.
      • Bouvet A.
      • Efstratiou A.
      • Schalén C.
      • Henriques-Normark B.
      • Jasir A.
      Strep-EURO Study Group
      Clinical and microbiological characteristics of severe Streptococcus pyogenes disease in Europe.
      • Steer A.C.
      • Law I.
      • Matatolu L.
      • Beall B.W.
      • Carapetis J.R.
      Global emm type distribution of group A streptococci: systematic review and implications for vaccine development.
      It was therefore surprising that Canada experienced a rapid, substantial increase in invasive infections caused by emm59 strains. Since 2006, emm59 strains have caused >500 invasive infections in the country, whereas they were almost nonexistent before then.
      • Tyrrell G.J.
      • Lovgren M.
      • St Jean T.
      • Hoang L.
      • Patrick D.M.
      • Horsman G.
      • Van Caeseele P.
      • Sieswerda L.E.
      • McGeer A.
      • Laurence R.A.
      • Bourgault A.M.
      • Low D.E.
      Epidemic of group A Streptococcus M/emm59 causing invasive disease in Canada.
      By 2007, emm59 was the second most common cause of invasive GAS disease in Canada, and in 2008 it became the leading cause (Figure 1A). Early in the epidemic, the western provinces of British Columbia and Alberta accounted for the majority of the cases, but by 2009 the epidemic had spread to eight provinces and two territories (Figure 1B).
      Figure thumbnail gr1
      Figure 1Rapid increase and spread of emm59 invasive GAS infections in Canada. A: The number of reported cases of invasive infection caused by emm59 GAS strains from 2004 through 2009, compared with five other prevalent emm types. Before 2006, emm59 strains rarely caused invasive infections. B: The number of reported invasive emm59 GAS strains by province or territory of isolation. Two western provinces (British Columbia and Alberta) and two central provinces (Saskatchewan and Manitoba) accounted for the majority of the reported cases in 2006 and 2007. In 2008 and 2009, high numbers of infections caused by emm59 strains were identified also in two eastern provinces (Ontario and Quebec). By the end of 2009, the epidemic had extended to Newfoundland and Labrador, New Brunswick, and the Nunavut and Yukon territories.

      DNA Sequence Analysis of Reference Genomes

      It may be hypothesized that the epidemic arose due to a widespread change in the human host, or a change in environmental conditions that created a population now at increased risk for GAS invasive infections. However, the correlation of the increase in invasive infections with the isolation of type emm59 strains, but not with any of the other emm GAS types,
      • Tyrrell G.J.
      • Lovgren M.
      • St Jean T.
      • Hoang L.
      • Patrick D.M.
      • Horsman G.
      • Van Caeseele P.
      • Sieswerda L.E.
      • McGeer A.
      • Laurence R.A.
      • Bourgault A.M.
      • Low D.E.
      Epidemic of group A Streptococcus M/emm59 causing invasive disease in Canada.
      argues against changes in the host as the primary cause of the upsurge. A second hypothesis postulates that epidemic emm59 strains arose from a pre-existing emm59 strain by acquisition of enhanced fitness mutations, or new genetic material encoding one or more factors that enhance virulence and/or facilitate immune evasion by antigenic shift or drift. To begin to differentiate among these and other hypotheses, we sequenced the genomes of a historic (mid-1970s) emm59 strain (MGAS1882) and a contemporary strain isolated during the epidemic (MGAS15252). These strains were chosen arbitrarily based on previous sequence analysis of targeted genes (see below) in several historic and epidemic strains that revealed a common emm allele (emm59.0) and on an identical sequence type (ST172), as defined by multilocus sequence typing (MLST). ST172 is the only sequence type associated to date with emm59 strains in the MLST database for GAS (http://spyogenes.mlst.net).
      The genomes of historic strain MGAS1882 and epidemic strain MGAS15252 are circular chromosomes of 1,781,046 bp and 1,750,847 bp, respectively (Figure 2, A and B). For these two strains, the core genomes (ie, the ∼1670-kbp portion lacking mobile genetic elements that is conserved in gene content relative to other sequenced GAS serotypes) differed by only 118 biallelic SNPs and 12 indels (see Supplemental Table S3 at http://ajp.amjpathol.org). One bacteriophage (1882.1; ∼44 kb), encoding streptococcal pyrogenic exotoxin K (SpeK) and streptococcal phospholipase A2 (SlaA), present in MGAS1882 but not in MGAS15252, and a prophage remnant found in MGAS15252 but not in MGAS1882 (15252.2; ∼22 kb), accounted for the majority of the differences in gene content between strains. The paucity of bacteriophages and other mobile genetic elements in emm59 strains was unexpected, given that they commonly occur in other GAS genomes.
      • Beres S.B.
      • Musser J.M.
      Contribution of exogenous genetic elements to the group A Streptococcus metagenome.
      Indeed, this relative lack of mobile genetic elements is the main explanation for the smaller size of the emm59 genomes, compared with the 1.89-Mbp average size of the 13 other GAS strains for which complete genome sequences are available.
      • Beres S.B.
      • Musser J.M.
      Contribution of exogenous genetic elements to the group A Streptococcus metagenome.
      The genomes of these emm59 strains were not closely related to other sequenced GAS strains (Figure 2C).
      Figure thumbnail gr2
      Figure 2Genome atlases of two reference emm59 GAS strains. Data for historic strain MGAS1882 (A) and contemporary epidemic strain MGAS15252 (B) are expressed as follows, reading from the outermost to the innermost circle. Genome scale (Mbp) is given first (circle 1), with annotated coding sequences (CDSs) encoded on the forward/direct (circle 2, dark blue), and reverse/complementary (circle 3, light blue) chromosomal strands. Reference genome landmarks (circle 4) are labeled as ribosomal RNAs (rRNA 1 to 5), fimbrial locus [fibronectin-collagen-T antigen (FCT)], region of difference (RD), streptolysin S locus (SLS), phage immunity locus [clustered regularly interspaced short palindromic repeat (CRISPR)], tmRNA, multiple gene activator locus (MGA), and prophages (ϕ). Circles 5 and 6 show TBLASTN comparison of the CDSs of the genomes of strain MGAS1882 and MGAS15252 and vice versa (circle 5) and TBLASTN comparison of emm59 CDSs to emm1 strain MGAS5005 genome (circle 6), the available GAS genome showing the lowest number of SNPs in core sequences relative to emm59 genomes; high similarity is shown in red and low similarity in blue. Circle 7 shows percent G+C content, with values greater and less than average shown in dark and light green, respectively. Circle 8 shows GC skew, calculated as (G − C)/(G+C) and averaged over a moving window of 10,000 bp, with excess G and excess C shown in dark and light brown, respectively. C: A phylogenetic network of sequenced GAS genomes, constructed as inferred by neighbor-network analysis using 56,086 concatenated core chromosomal biallelic SNPs; both strain names and emm types are given.

      Restricted Allelic Variation among Targeted Genes

      The rapid transcontinental spread of the epidemic led us to hypothesize that the contemporary emm59 GAS strains recovered in Canada are derived from a single common ancestor. To initially test this hypothesis, we sequenced several genes (ropB, covR, covS, and hasB, all known to be highly polymorphic in strains of other GAS emm types
      • Beres S.B.
      • Carroll R.K.
      • Shea P.R.
      • Sitkiewicz I.
      • Martinez-Gutierrez J.C.
      • Low D.E.
      • McGeer A.
      • Willey B.M.
      • Green K.
      • Tyrrell G.J.
      • Goldman T.D.
      • Feldgarden M.
      • Birren B.W.
      • Fofanov Y.
      • Boos J.
      • Wheaton W.D.
      • Honisch C.
      • Musser J.M.
      Molecular complexity of successive bacterial epidemics deconvoluted by comparative pathogenomics.
      • Shea P.R.
      • Beres S.B.
      • Flores A.R.
      • Ewbank A.L.
      • Gonzalez-Lugo J.H.
      • Martagon-Rosado A.J.
      • Martinez-Gutierrez J.C.
      • Rehman H.A.
      • Serrano-Gonzalez M.
      • Fittipaldi N.
      • Ayers S.D.
      • Webb P.
      • Willey B.M.
      • Low D.E.
      • Musser J.M.
      Distinct signatures of diversifying selection revealed by genome analysis of respiratory tract and invasive bacterial populations.
      ) in 182 randomly chosen Canadian emm59 strains. Consistent with our single-clone hypothesis, we found low levels of allelic variation in these usually highly variable genes (see Supplemental Table S4 at http://ajp.amjpathol.org). The finding was extended by sequence analysis of the emm59 and recP genes in all epidemic strains in our sample, both of which showed a low level of nucleotide polymorphism (see Supplemental Table S4 at http://ajp.amjpathol.org).

      CRISPR Element Analysis

      Clustered regularly interspaced short palindromic repeat (CRISPR) elements mediate bacterial resistance to infection by foreign DNA. CRISPRs have been found to be highly variable among closely related strains in several bacterial species, including GAS.
      • Marraffini L.A.
      • Sontheimer E.J.
      CRISPR interference: RNA-directed adaptive immunity in bacteria and archaea.
      To test the hypothesis that a CRISPR element we identified in the genome of the two sequenced emm59 strains is polymorphic among the strains in our sample, we sequenced the CRISPR element present in 562 emm59 strains; the primers used are given in Table 1. Consistent with this hypothesis, we identified four distinct genetic variants (see Supplemental Figure S1 at http://ajp.amjpathol.org). There was no simple relationship between the variants and their geographic distribution; however, most strains isolated from patients early in the epidemic had the variant 2 CRISPR element. This finding suggests that strains with the variant 2 CRISPR element were the ancestral condition and represent the strain that gave rise to the epidemic.
      Table 1Oligonucleotide Primers Used in the Study of emm59 GAS
      Primer nameSequenceUse
      ropBseqF5′-ATCGCATCTGGCTACTGACC-3′ropB amplification
      ropBregionR5′-GCCCTGGAGCTGTTGAGATA-3′ropB amplification
      covRF5′-TTGACCATAGAGGGCAGAGAA-3′covR amplification
      covRR5′-GGCAATCAGTGTAAAGGCAGA-3′covR amplification
      covSF5′-AATTGACATTCCAGGCAAGG-3′covS amplification
      covSR5′-CGCAAACACCCCAATTCTAT-3′covS amplification
      hasBF5′-GGAATGGGAACACGTAAAA-3′hasB amplification
      hasBR5′-AGTGCTCAATCATACCACCAA-3′hasB amplification
      Emm15′-TATTSGCTTAGAAAATTAA-3′emm amplification
      Emm25′-GCAAGTTCTTCAGCTTGTTT-3′emm amplification
      Core_2384_fwd5′-TTGATGAGTCCATTGGTAAG-3′recP amplification
      Core_2385_rev5′-GTCACATTAGTTTTAATCCC-3′recP amplification
      CRISPR-F15′-TCCGCAGTATATTTTAGA-3′CRISPR amplification
      CRISPR-R15′-CCCATCTGGCTTAATCAT-3′CRISPR amplification
      speK-L5′-GGAACATATATGCTCCCTAGAT-3′speK amplification
      speK-R5′-GTGTGTCTAATGCCACCGTCT-3′speK amplification
      sla-L5′-AATGGAAATTGGCATGAAAG-3′sla amplification
      sla-R5′-CTCTAATAGCATCGGCTACGC-3′sla amplification

      Comparative Genome Sequencing of Canadian emm59 Strains

      We recently sequenced the genomes of 95 invasive emm3 GAS strains and found that the genetic polymorphism data yielded enhanced understanding of successive emm3 GAS epidemics occurring in Ontario.
      • Beres S.B.
      • Carroll R.K.
      • Shea P.R.
      • Sitkiewicz I.
      • Martinez-Gutierrez J.C.
      • Low D.E.
      • McGeer A.
      • Willey B.M.
      • Green K.
      • Tyrrell G.J.
      • Goldman T.D.
      • Feldgarden M.
      • Birren B.W.
      • Fofanov Y.
      • Boos J.
      • Wheaton W.D.
      • Honisch C.
      • Musser J.M.
      Molecular complexity of successive bacterial epidemics deconvoluted by comparative pathogenomics.
      This finding, coupled with the absence of a large number of informative sequence polymorphisms at targeted loci in the emm59 strains, led us to sequence the genomes of 145 additional contemporary emm59 GAS strains. We selected strains to represent the temporal (2006 to early 2010) and geographic (eight Canadian provinces and two territories) extremes of the epidemic. We also sequenced the genomes of the only seven emm59 strains isolated in Canada before the epidemic began (1996 to 2004). Polymorphisms such as biallelic SNPs and indels were identified relative to the core genome of strain MGAS15252 using VAAL.
      • Nusbaum C.
      • Ohsumi T.K.
      • Gomez J.
      • Aquadro J.
      • Victor T.C.
      • Warren R.M.
      • Hung D.T.
      • Birren B.W.
      • Lander E.S.
      • Jaffe D.B.
      Sensitive, specific polymorphism discovery in bacteria using massively parallel sequencing.
      The genomes of all 145 epidemic and 7 pre-epidemic Canadian emm59 strains were virtually identical, differing on average by only 7.3 SNPs and 4.3 indels relative to the reference strain MGAS15252. Genome-wide, we identified a total of 434 polymorphic sites (347 SNPs and 87 indels) relative to strain MGAS15252. Of these 87 indels, 29 were present in one strain only, and they all mapped to a single tRNA locus. The remaining 58 indels were 41 insertions and 17 deletions. Slightly more than half of the indel loci (31/58) were strain-specific; the others were present in more than one strain. Most of the SNPs were strain-specific; only 90 of the SNP loci (25.9%) were present in two or more contemporary strains. Predicted coding sequence accounts for 88.2% of the MGAS15252 genome, and a relatively similar portion of the SNPs (300/347, or 86.4%) occurred in coding sequences. Nonsynonymous (amino acid-altering) SNPs accounted for 75% (225) and synonymous SNPs accounted for 25% (75) of the 300 coding SNPs. Notably, the standalone virulence regulator mga, although representing only 0.11% of the coding sequences (CDSs), accounted for 7% of the SNPs in coding sequences.
      Phylogenetic analysis using the full-genome SNP data showed that the epidemic Canadian emm59 strains are very closely related (Figure 3) and strongly supported the hypothesis that these strains are clonally related by descent. This very close genetic relationship led us to evaluate sequence variation in the genomes of all remaining 392 Canadian contemporary emm59 strains; we applied a recently described pooled-DNA sequencing strategy,
      • Holt K.E.
      • Teo Y.Y.
      • Li H.
      • Nair S.
      • Dougan G.
      • Wain J.
      • Parkhill J.
      Detecting SNPs and estimating allele frequencies in clonal bacterial populations by sequencing pooled DNA.
      in which chromosomal DNA from two strains is mixed in a 1:1 ratio and used to generate a single bar-coded sample for Illumina-based sequencing. Each two-genome sample presented, on average, 24.6 polymorphisms relative to the reference strain MGAS15252, a number consistent with the number of polymorphisms identified when strains were sequenced individually. Thus, we conclude that, regardless of geographic location or year of isolation, all contemporary Canadian emm59 strains studied have descended from a recent, single common ancestral cell that underwent rapid geographic dissemination.
      Figure thumbnail gr3
      Figure 3A distinct emm59 GAS genetic clone is responsible for the Canadian epidemic. A: The inferred genetic relationships among strains based on 474 concatenated SNP loci identified by full-genome sequencing. The resulting phylogenetic tree shows that the 145 analyzed epidemic Canadian strains are monoclonal and genetically distinct from historic emm59 strains (MGAS1882 and MGAS360) and contemporary invasive strains from Spain (MGAS20027 and MGAS20031). In contrast, a Canadian emm59 strain (MGAS20387) isolated in 1996 is closely related to the epidemic strains. B: SNPs identified by full-genome sequencing reveal diversification among epidemic strains that correlates with geographical origin and spread of the Canadian clone into the United States. Circles indicate Canadian isolates; triangles indicate U.S. isolates. Shape size is proportionate to the number of strains.

      The Canadian Clone Is Genetically Distinct from Other emm59 GAS Strains

      Outbreaks of invasive disease caused by emm59 GAS are rare, and normally are geographically confined.
      • Phillips G.
      • Efstratiou A.
      • Tanna A.
      • Beall B.
      • Ferguson J.
      • Roworth M.
      An outbreak of skin sepsis in abattoir workers caused by an ‘unusual’ strain of Streptococcus pyogenes.
      We therefore hypothesized that the epidemic Canadian clone is genetically distinct from other historic emm59 strains and from emm59 organisms causing sporadic infections in other countries. To test this idea, we sequenced the genome of an emm59 strain (MGAS360) isolated in Japan in the 1960s. Consistent with the hypothesis, phylogenetic analysis showed that the contemporary Canadian emm59 strains are genetically distinct from this historic emm59 organism (Figure 3A).
      We also sequenced the genomes of two emm59 strains selected at random from a group of nine emm59 strains causing invasive episodes in Spain from 1999 to 2007.
      • Ardanuy C.
      • Rolo D.
      • Domenech A.
      • Tubau F.
      • Ayats J.
      • Grau I.
      • Pallares R.
      • Martin R.
      • Liñares J.
      Trends in invasive group A streptococci in adult patients in Barcelona (abstract) 19th European Congress of Clinical Microbiology and Infectious Diseases, Helsinki, Finland, 16–19 May 2009.
      Their genomes also differed from those of the Canadian epidemic strains (Figure 3A). However, when we sequenced the genomes of 39 emm59 strains isolated in recent years in the United States, we identified five strains that were genetically very closely related to the Canadian epidemic clone (Figure 3B), including one strain from Oregon and two strains each from California and Minnesota. The phylogenetic data suggest that these five members of the Canadian clone disseminated into the United States on one or two separate occasions (Figure 3B).

      Near-Real-Time Genome Sequencing Reveals Further Spread of the Epidemic Clone into the United States

      As our genomic studies were nearing completion, we became aware of several geographically and temporally clustered cases of invasive GAS infections in Montana, a northern U.S. state that borders British Columbia, Alberta, and Saskatchewan. Within 10 days of isolating the first two strains from infected patients, full-genome sequencing allowed us to unambiguously show that these strains belong to the Canadian emm59 epidemic clone and to identify its closest genetic relatives, which were recovered from patients in British Columbia and Alberta (Figure 3B). The remaining four emm59 strains isolated from Montana were shown to belong to the same clone by the pooled-DNA genome sequencing method.

      The emm59 Epidemic Clone Is Significantly More Virulent in Mouse Models of Invasive Skin Infection and Necrotizing Fasciitis

      Inasmuch as the genome data demonstrated that the epidemic strains are genetically distinct from other emm59 strains analyzed and the epidemiological data suggest a difference in pathogenic behavior, we hypothesized that strains causing the current epidemic have enhanced virulence properties. To test this hypothesis, we compared the virulence of two strains from the current outbreak (MGAS15252 and MGAS15249) to that of historic strain MGAS1882 in a mouse model of skin infection. Mice infected with the epidemic strains had significantly larger lesions, with more extensive tissue necrosis, compared with animals receiving the historic strain (P < 0.001; generalized estimating equation) (Figure 4, A and B). Contemporary epidemic emm59 strains were also more virulent in a mouse model of necrotizing fasciitis (P = 0.047 for MGAS15249 and P = 0.008 for MGAS15252; log rank test) (Figure 4, C and D). These results support the hypothesis that the contemporary emm59 strains are significantly more virulent than historic emm59 strains. To test the hypothesis that the observed differences in virulence correlate with differences in the ability of historic and epidemic strains to replicate in situ in muscle, we assessed GAS burden in mouse muscle after intramuscular inoculation. There was no significant difference in the CFUs recovered per gram of total hindlimb tissue between the experimental groups at any time point (see Supplemental Figure S2 at http://ajp.amjpathol.org). This finding suggests that differences in survival and pathology are not due to preferential replication of the epidemic strain, nor to diminished survival of the historic strain.
      Figure thumbnail gr4
      Figure 4Contemporary epidemic strains are more virulent than the historic strain in mouse models of infection. A: Epidemic strains MGAS15249 and MGAS15252 caused significantly larger lesions than historic strain MGAS1882 in a subcutaneous murine skin infection model. *P < 0.01. n = 15 mice per strain. B: Microscopic examination of the skin lesions obtained at 48 hours after inoculation. The abscess caused by historic strain MGAS1882 is confined to the subcutaneous tissue. In striking contrast, the abscess caused by epidemic strains MGAS15249 or MGAS15252 erodes through the epidermis (arrowheads) and destroys more tissue, extending beyond the lateral and deep margins of the microscopic field (brackets). H&E stain. Original magnification, ×2. C: In an intramuscular infection model, epidemic strains MGAS15249 and MGAS15252 caused significantly more near-mortality, compared with historic strain MGAS1882. **P = 0.047; ***P = 0.008. n = 15 mice per strain. D: Visual (gross) examination of infected mouse hindlimbs (top row). The abscess caused by historic strain MGAS1882 is much smaller than those caused by epidemic strains MGAS15249 or MGAS15252 (white arrowheads). Scale bars: 10 mm. Microscopic examination (bottom row) shows that the small abscess caused by strain MGAS1882 is restricted to the fascial plane (arrowheads), whereas strains MGAS15249 and MGAS15252 breach the fascial plane, disseminate among individual muscle cells (boxed), and destroy the adjacent muscle tissue (encircled). H&E stain. Original magnification, ×10.

      Epidemic emm59 Strains Cause Significantly More Extensive Necrotizing Fasciitis in Cynomolgus Macaques

      Although mouse infection models are informative, GAS is a host-specialist pathogen that causes natural disease only in humans. Some GAS virulence factors, such as streptokinase, have been shown to have modest or no activity in murine infection models.
      • Olsen R.J.
      • Musser J.M.
      Molecular pathogenesis of necrotizing fasciitis.
      • Sun H.
      • Ringdahl U.
      • Homeister J.W.
      • Fay W.P.
      • Engleberg N.C.
      • Yang A.Y.
      • Rozek L.S.
      • Wang X.
      • Sjöbring U.
      • Ginsburg D.
      Plasminogen is a critical host pathogenicity factor for group A streptococcal infection.
      The cynomolgus macaque has become a useful model of human GAS infections and has provided new information about the molecular pathogenesis of pharyngitis and invasive diseases, including necrotizing fasciitis.
      • Sumby P.
      • Tart A.H.
      • Musser J.M.
      A non-human primate model of acute group A Streptococcus pharyngitis.
      • Olsen R.J.
      • Ashraf M.
      • Gonulal V.E.
      • Ayeras A.A.
      • Cantu C.
      • Shea P.R.
      • Carroll R.K.
      • Humbird T.
      • Greaver J.L.
      • Swain J.L.
      • Chang E.
      • Ragasa W.
      • Jenkins L.
      • Lally K.P.
      • Blasdel T.
      • Cagle P.
      • Musser J.M.
      Lower respiratory tract infection in cynomolgus macaques (Macaca fascicularis) infected with group A Streptococcus.
      • Olsen R.J.
      • Sitkiewicz I.
      • Ayeras A.A.
      • Gonulal V.E.
      • Cantu C.
      • Beres S.B.
      • Green N.M.
      • Lei B.
      • Humbird T.
      • Greaver J.
      • Chang E.
      • Ragasa W.P.
      • Montgomery C.A.
      • Cartwright Jr, J.
      • McGeer A.
      • Low D.E.
      • Whitney A.R.
      • Cagle P.T.
      • Blasdel T.L.
      • DeLeo F.R.
      • Musser J.M.
      Decreased necrotizing fasciitis capacity caused by a single nucleotide mutation that alters a multiple gene virulence axis.
      We therefore sought to confirm our hypothesis that the contemporary epidemic strains are more virulent than historic emm59 GAS strains by using this nonhuman primate model of necrotizing fasciitis. Consistent with the hypothesis, animals that were inoculated intramuscularly with epidemic strain MGAS15252 had significantly larger muscle lesions and more severe myonecrosis than those inoculated with MGAS1882 (P < 0.05; Mann-Whitney test) (Figure 5, A and B). In accord with our findings in the mouse model, we did not find significant differences in the ability of historic and epidemic and strains to replicate in situ in the macaque muscle (Figure 5C).
      Figure thumbnail gr5
      Figure 5Contemporary epidemic emm59 strains are more virulent than a historic emm59 strain in a nonhuman primate model of necrotizing fasciitis. A: Epidemic strain MGAS15252 (green) caused significantly larger lesions than historic strain MGAS1882 (red). P < 0.05. n = 3 animals per strain. B: Microscopic examination of the infected tissue at the inoculation site confirmed the virulence data obtained with the mouse model. The small abscess caused by historic strain MGAS1882 was restricted to the fascial plane (arrows). In contrast, epidemic strain MGAS15252 breached the fascial plane and extended the severe necrotizing infection into the muscle tissue (encircled). Hemorrhage into the necrotic muscle tissue (boxed) was a characteristic feature. H&E stain. Original magnification, ×10. C: Historic strain MGAS1882 and epidemic strain MGAS15252 were recovered in similar numbers from the infected muscle tissue of cynomolgus macaques.

      The Increase in Invasive emm59 Disease in Canada Was Not Accompanied by a Similar Increase in emm59 Pharyngitis

      Study of epidemic emm1 GAS disease in Finland has shown that the most abundant subclones causing pyogenic infections and septicemia were common in the pool of subclones causing pharyngitis.
      • Hoe N.P.
      • Vuopio-Varkila J.
      • Vaara M.
      • Grigsby D.
      • De Lorenzo D.
      • Fu Y.X.
      • Dou S.J.
      • Pan X.
      • Nakashima K.
      • Musser J.M.
      Distribution of streptococcal inhibitor of complement variants in pharyngitis and invasive isolates in an epidemic of serotype M1 group A Streptococcus infection.
      Pharyngitis is by far the most abundant human infection caused by GAS, and one of the most important mechanisms of GAS transmission involves direct contact with nose and throat discharges of an infected individual.
      • Hamburger Jr, M.
      • Green M.J.
      • Hamburger V.G.
      The problem of the dangerous carrier of hemolytic streptococci; spread of infection by individuals with strongly positive nose cultures who expelled large numbers of hemolytic streptococci.
      • Hamburger Jr, M.
      • Robertson O.H.
      Expulsion of group A hemolytic streptococci in droplets and droplet nuclei by sneezing, coughing and talking.
      To begin to investigate the mechanisms by which invasive emm59 may have disseminated during the ongoing epidemic, we evaluated the hypothesis that emm59 strains were also responsible for cases of pharyngitis in Canada during the same time period. To that end, we sequenced the emm gene in a convenience sample of 5438 isolates recovered from patients with GAS sore throat in Canada during 2006 to 2010, including Alberta, Ontario, Nunavut, and the Northwest Territories. Unexpectedly, we identified only 24 (<0.5%) emm59 pharyngeal strains. Of note, these strains were first isolated in 2008, which is 2 years after the invasive epidemic began, suggesting that these strains are likely derived from invasive emm59 strains. Genome sequencing of these 24 pharyngeal isolates demonstrated that they are members of the invasive epidemic clone (data not shown).

      Compared with a Typical GAS Strain Causing Pharyngitis, Type emm59 GAS Organisms Are Impaired in Survival in Human Saliva

      The interaction of GAS with saliva is crucial for the establishment of pharyngitis and the subsequent transmission of infectious organisms.
      • Hamburger Jr, M.
      • Green M.J.
      • Hamburger V.G.
      The problem of the dangerous carrier of hemolytic streptococci; spread of infection by individuals with strongly positive nose cultures who expelled large numbers of hemolytic streptococci.
      • Hamburger M.
      • Robertson O.H.
      The expulsion of beta hemolytic streptococci by sneezing.
      For example, type emm1 GAS, the leading cause of GAS pharyngitis in many studies,
      • Shulman S.T.
      • Tanz R.R.
      • Dale J.B.
      • Beall B.
      • Kabat W.
      • Kabat K.
      • Cederlund E.
      • Patel D.
      • Rippe J.
      • Li Z.
      • Sakota V.
      North American Streptococcal Pharyngitis Surveillance Group
      Seven-year surveillance of North American pediatric group A streptococcal pharyngitis isolates.
      • Shulman S.T.
      • Tanz R.R.
      • Kabat W.
      • Kabat K.
      • Cederlund E.
      • Patel D.
      • Li Z.
      • Sakota V.
      • Dale J.B.
      • Beall B.
      US Streptococcal Pharyngitis Surveillance Group
      Group A streptococcal pharyngitis serotype surveillance in North America, 2000–2002.
      has been shown to survive in saliva for up to 4 weeks.
      • Shelburne 3rd, S.A.
      • Granville C.
      • Tokuyama M.
      • Sitkiewicz I.
      • Patel P.
      • Musser J.M.
      Growth characteristics of and virulence factor production by group A Streptococcus during cultivation in human saliva.
      Inasmuch as our previous findings demonstrated a low prevalence of emm59 GAS pharyngitis in Canada, we hypothesized that emm59 GAS organisms are impaired in their ability to persist in human saliva. Consistent with this hypothesis, no viable emm59 organisms were recovered from human saliva after only 11 days after inoculation, whereas the emm1 strain survived in saliva for at least 63 days at relatively high titers (Figure 6A).
      Figure thumbnail gr6
      Figure 6Type emm59 GAS organisms are impaired in survival in human saliva and in colonization of the oropharynx of experimentally inoculated mice. A: The emm59 strain MGAS15249 (green circles) survived for only 10 days in human saliva, whereas a type emm1 strain MGAS5005 (blue triangles) persisted in human saliva for at least 63 days at relatively high titers (>8 × 103 CFU/mL). B: After intranasal inoculation, the emm59 strain colonized the oropharynx of significant fewer mice than the emm1 strain. P < 0.01. n = 9 mice per strain. C: The emm59 strain MGAS15249 was recovered on swabs of the oropharynx of infected mice at significantly lower numbers than the type emm1 strain MGAS5005. P < 0.05. n = 9 mice per strain.

      Compared with a Typical GAS Strain Causing Pharyngitis, Type emm59 GAS Organisms Are Significantly Less Able to Colonize the Oropharynx of Mice

      Because emm59 organisms had decreased ability to survive in saliva, we hypothesized that strains of emm59 GAS also have reduced ability to colonize the oropharynx. To test this idea, we used a mouse model of pharyngeal colonization in which CD1 mice were inoculated intranasally with either emm59 strain MGAS15249 or emm1 strain MGAS5005. The results were consistent with our hypothesis, showing that the emm59 strain MGAS15249 colonized to a significantly lower extent the oropharynx of significantly fewer animals than the emm1 GAS strain MGAS5005 (Figure 6, B and C).

      Skin Spread of emm59 Organisms in a Mouse Model of GAS Transmission

      The above results suggested that pharyngeal transmission might not be a major driver of emm59 GAS dissemination. Another mechanism of GAS transmission can involve direct contact with the skin of an infected individual. To test the hypothesis that emm59 organisms can spread via infected skin lesions, we developed a mouse skin transmission model. We used strain MGAS15249 as a representative emm59 strain. For comparison purposes, we included the emm1 strain MGAS5005. Two groups of five animals each were used per strain. One animal per group was inoculated. The inoculated (seeder) animals were then placed in cages containing four additional naïve (receiver) mice. Consistent with our hypothesis and suggesting that skin-to-skin transmission is a likely route of spread for emm59 GAS, swabbing revealed rapid colonization of all of the receiver mice, as soon as 12 hours after the seeder mice were introduced into the cages (Figure 7A). No significant differences were found in the rate of transmission or the number of receiver animals acquiring the infection by exposure to the skin of seeder mice between emm59 and emm1 GAS strains (Figure 7A). CFU counts were also similar between groups (Figure 7B). Microscopic examination of the resulting skin lesions confirmed the presence of a localized infection (Figure 7, C and D).
      Figure thumbnail gr7
      Figure 7Spread of type emm59 GAS strain from experimentally inoculated (seeder) to naïve (receiver) mice in a mouse model of skin transmission. Seeder (n = 2) and receiver (n = 8) mice were treated with a standard incision to the epidermis overlying each hindlimb (skin sites), but only seeder animals were inoculated with GAS bacteria. A: The emm59 GAS strain MGAS15249 rapidly colonized >90% of the skin sites of receiver mice, within 12 hours after introduction of the seeder mice. No significant differences were found in the number of sites colonized or in the rate of transmission between the emm59 strain MGAS15249 (green circles) and the emm1 strain MGAS5005 (blue triangles). B: No significant differences were observed in the number of CFUs recovered on swabs of skin sites of receiver mice between the emm59 strain MGAS15249 (green circles) and the emm1 strain MGAS5005 (blue triangles). C: Microscopic examination of the infected skin after GAS transmission confirmed that the seeder mice developed necrotizing GAS infections. The localized infection, characterized by a dense inflammatory infiltrate and moderate edema, is confined to the dermis and epidermis (black arrows) and is covered by a crust containing many entrapped polymorphonuclear leukocytes (red arrows). H&E stain. Original magnification, ×10. D: Gram staining of the boxed region in C highlights the presence of GAS organisms in the tissue adjacent to the necrotizing lesion. Original magnification, ×40.

      Discussion

      Full-genome analysis has become the method of choice to investigate many biomedically relevant but previously inaccessible phenomena, including cancer causation, human disorders of unknown genetic basis, and epidemics.
      • Jones S.
      • Zhang X.
      • Parsons D.W.
      • Lin J.C.
      • Leary R.J.
      • Angenendt P.
      • et al.
      Core signaling pathways in human pancreatic cancers revealed by global genomic analyses.
      • Ley T.J.
      • Mardis E.R.
      • Ding L.
      • Fulton B.
      • McLellan M.D.
      • Chen K.
      • et al.
      DNA sequencing of a cytogenetically normal acute myeloid leukaemia genome.
      The Cancer Genome Atlas Research Network
      Comprehensive genomic characterization defines human glioblastoma genes and core pathways.
      • Musser J.M.
      • Shelburne 3rd, S.A.
      A decade of molecular pathogenomic analysis of group A Streptococcus.
      • Harris S.R.
      • Feil E.J.
      • Holden M.T.
      • Quail M.A.
      • Nickerson E.K.
      • Chantratita N.
      • Gardete S.
      • Tavares A.
      • Day N.
      • Lindsay J.A.
      • Edgeworth J.D.
      • de Lencastre H.
      • Parkhill J.
      • Peacock S.J.
      • Bentley S.D.
      Evolution of MRSA during hospital transmission and intercontinental spread.
      • Croucher N.J.
      • Harris S.R.
      • Fraser C.
      • Quail M.A.
      • Burton J.
      • van der Linden M.
      • McGee L.
      • von Gottberg A.
      • Song J.H.
      • Ko K.S.
      • Pichon B.
      • Baker S.
      • Parry C.M.
      • Lambertsen L.M.
      • Shahinas D.
      • Pillai D.R.
      • Mitchell T.J.
      • Dougan G.
      • Tomasz A.
      • Klugman K.P.
      • Parkhill J.
      • Hanage W.P.
      • Bentley S.D.
      Rapid pneumococcal evolution in response to clinical interventions.
      • Gardy J.L.
      • Johnston J.C.
      • Ho Sui S.J.
      • Cook V.J.
      • Shah L.
      • Brodkin E.
      • Rempel S.
      • Moore R.
      • Zhao Y.
      • Holt R.
      • Varhol R.
      • Birol I.
      • Lem M.
      • Sharma M.K.
      • Elwood K.
      • Jones S.J.
      • Brinkman F.S.
      • Brunham R.C.
      • Tang P.
      Whole-genome sequencing and social-network analysis of a tuberculosis outbreak [Erratum appeared in N Engl J Med 2011 Jun 2;364(22):2174].
      • Nelson M.I.
      • Tan Y.
      • Ghedin E.
      • Wentworth D.E.
      • St George K.
      • Edelman L.
      • Beck E.T.
      • Fan J.
      • Lam T.T.
      • Kumar S.
      • Spiro D.J.
      • Simonsen L.
      • Viboud C.
      • Holmes E.C.
      • Henrickson K.J.
      • Musser J.M.
      Phylogeography of the spring and fall waves of the H1N1/09 pandemic influenza virus in the United States.
      • Rasko D.A.
      • Webster D.R.
      • Sahl J.W.
      • Bashir A.
      • Boisen N.
      • Scheutz F.
      • Paxinos E.E.
      • Sebra R.
      • Chin C.S.
      • Iliopoulos D.
      • Klammer A.
      • Peluso P.
      • Lee L.
      • Kislyuk A.O.
      • Bullard J.
      • Kasarskis A.
      • Wang S.
      • Eid J.
      • Rank D.
      • Redman J.C.
      • Steyert S.R.
      • Frimodt-Møller J.
      • Struve C.
      • Petersen A.M.
      • Krogfelt K.A.
      • Nataro J.P.
      • Schadt E.E.
      • Waldor M.K.
      Origins of the E. coli strain causing an outbreak of hemolytic-uremic syndrome in Germany.
      With the present work, we show that a genetically distinct, hypervirulent clone of emm59 GAS recently emerged in western Canada, rapidly spread countrywide, crossed into the United States, and caused severe invasive infections in both countries. Full-genome data permitted us to delineate patterns of geographic dissemination in widely dispersed areas. Because the epidemic strains differed from one another, on average, by only 7.3 SNPs and 4.3 indels, these patterns were identifiable only by sequencing the entire genomes of the strains.
      We and others have used GAS as a model organism to investigate molecular processes underpinning infection specificity and bacterial epidemics.
      • Musser J.M.
      • Shelburne 3rd, S.A.
      A decade of molecular pathogenomic analysis of group A Streptococcus.
      Although new knowledge has accrued, we are far from achieving a comprehensive understanding of the molecular genetic events that produce bacterial strains that differ in virulence and have enhanced potential to cause epidemics. The precise molecular basis underlying the enhanced virulence of the epidemic emm59 strains remains unidentified by the present study. After comparison of the genome sequences of the historic and epidemic strains, including sequences corresponding to the core genome and mobile genetic elements, we were unable to identify genetic regions or single genes encoding known or putative virulence factors that are present in the epidemic strains but absent in the genome of the historic strain. Emergence of epidemic emm59 strains could have been caused, at least in part, by antigenic changes that allowed the bacteria to escape from population immunity, a common mechanism associated with epidemics of a variety of microbial pathogens.
      • Centurion-Lara A.
      • LaFond R.E.
      • Hevner K.
      • Godornes C.
      • Molini B.J.
      • Van Voorhis W.C.
      • Lukehart S.A.
      Gene conversion: a mechanism for generation of heterogeneity in the tprK gene of Treponema pallidum during infection.
      • Futse J.E.
      • Brayton K.A.
      • Dark M.J.
      • Knowles Jr, D.P.
      • Palmer G.H.
      Superinfection as a driver of genomic diversification in antigenically variant pathogens.
      • Johnson W.E.
      • Desrosiers R.C.
      Viral persistence: HIV's strategies of immune system evasion.
      However, although we have not systematically performed an analysis of antigenic differences, the genomic data did not reveal differences between historic and epidemic emm59 strains in known or putative GAS antigenic proteins such as the M protein, streptococcal C5a peptidase ScpA, pilus subunits, and most fibronectin-binding proteins.
      • Steer A.C.
      • Batzloff M.R.
      • Mulholland K.
      • Carapetis J.R.
      Group A streptococcal vaccines: facts versus fantasy.
      Recently, Olsen et al
      • Olsen R.J.
      • Sitkiewicz I.
      • Ayeras A.A.
      • Gonulal V.E.
      • Cantu C.
      • Beres S.B.
      • Green N.M.
      • Lei B.
      • Humbird T.
      • Greaver J.
      • Chang E.
      • Ragasa W.P.
      • Montgomery C.A.
      • Cartwright Jr, J.
      • McGeer A.
      • Low D.E.
      • Whitney A.R.
      • Cagle P.T.
      • Blasdel T.L.
      • DeLeo F.R.
      • Musser J.M.
      Decreased necrotizing fasciitis capacity caused by a single nucleotide mutation that alters a multiple gene virulence axis.
      demonstrated that a unique, naturally occurring SNP in the mtsR gene results in premature termination of the translation of MtsR, which alters a multiple gene virulence axis and results in decreased ability of emm3 GAS strains to cause necrotizing fasciitis. It is important to note that, compared with epidemic strain MGAS15252, the historic strain MGAS1882 has 118 biallelic SNPs and 12 indels. These naturally occurring polymorphisms are thus the focus of ongoing research in our laboratory, aimed at deciphering the molecular basis of the enhanced virulence of epidemic emm59 GAS strains.
      One of our discoveries was that, as a population, the contemporary emm59 epidemic-related isolates have accumulated relatively few genetic polymorphisms, despite wide geographic dissemination and at least 5 years in which to diverge from their common ancestor. What factors might contribute to the restricted variation among the emm59 organisms causing this epidemic? One possibility is that emm59 GAS have a decreased spontaneous mutation frequency. However, we have no evidence that supports this hypothesis. Indeed, by growing emm59 organisms in vitro on antibiotic-containing media, we observed that it is possible to isolate spontaneous antibiotic-resistant derivatives (data not shown). A second possibility is that sufficient time has not elapsed, over the course of the 5 years represented by our emm59 strain sample, to produce considerable genomic differentiation. Although we favor the idea that the relatively short time period involved in this epidemic likely contributed to constrained diversity, it is unlikely to be the sole explanation. Of note, in the case of type emm3 GAS, which are common causes of pharyngitis and episodes of invasive infections in Canada and the United States, a far higher level of genome-wide polymorphisms accumulated over the course of 5 years than in the emm59 strains.
      • Beres S.B.
      • Carroll R.K.
      • Shea P.R.
      • Sitkiewicz I.
      • Martinez-Gutierrez J.C.
      • Low D.E.
      • McGeer A.
      • Willey B.M.
      • Green K.
      • Tyrrell G.J.
      • Goldman T.D.
      • Feldgarden M.
      • Birren B.W.
      • Fofanov Y.
      • Boos J.
      • Wheaton W.D.
      • Honisch C.
      • Musser J.M.
      Molecular complexity of successive bacterial epidemics deconvoluted by comparative pathogenomics.
      Another key contributor to the amount of genetic diversity present in organisms is the population size.
      • Denamur E.
      • Matic I.
      Evolution of mutation rates in bacteria.
      It is possible that a relatively restricted bacterial population size has contributed to the very modest emm59 genetic variation in the epidemic strains. Is there evidence that emm59 organisms have a small population size? GAS is a human-specific bacterium that commonly inhabits either the upper respiratory tract or skin. Importantly, emm59 organisms are rare causes of pharyngitis in all geographical areas studied. To study this issue in more detail in Canada, we sequenced the emm gene in >5000 pharyngitis strains recovered from patients with GAS sore throat in Canada, including Alberta, Ontario, Nunavut, and Northwest Territories and identified only 24 (<0.5%) emm59 strains. Thus, emm59 strains are rarely responsible for pharyngitis, by far the most abundant human infection caused by GAS. These data are consistent with our hypothesis of restricted emm59 population size.
      One of the most remarkable and enigmatic features of this epidemic was the relatively rapid spread across vast geographic areas of emm59 organisms causing invasive disease. We show here that emm59 GAS strains are impaired in survival in human saliva and less able to colonize the oropharynx of experimentally inoculated mice than strains of a GAS type that commonly causes throat infections. Together with the limited number of pharyngitis cases in Canada, these data argue against a major involvement of pharyngitis patients in transmission of epidemic emm59 GAS strains. It must be noted, however, that our data cannot rule out the hypothesis of dissemination of emm59 organisms from healthy carriers harboring the bacteria asymptomatically in the oropharynx or nares. Tissue tropism in GAS has long been described.
      • Wannamaker L.W.
      Differences between streptococcal infections of the throat and of the skin (second of two parts).
      • Wannamaker L.W.
      Differences between streptococcal infections of the throat and of the skin I.
      Moreover, markers of niche specificity defining throat, skin, and generalist GAS strains have been proposed, based on the chromosomal arrangement of emm and emm-like genes (and their lineage), and in the presence of different variants of the mga and rofA/nra loci (encoding transcription regulators of gene expression), the fibronectin-collagen-T antigen region, and the sof gene (encoding serum opacity factor).
      • Bessen D.E.
      • Lizano S.
      Tissue tropisms in group A streptococcal infections.
      Data from our genome investigation indicate that emm59 epidemic strains may be defined as generalists, based solely on the presence of these markers. However, our saliva and animal experiments and previous reports
      • Tyrrell G.J.
      • Lovgren M.
      • St Jean T.
      • Hoang L.
      • Patrick D.M.
      • Horsman G.
      • Van Caeseele P.
      • Sieswerda L.E.
      • McGeer A.
      • Laurence R.A.
      • Bourgault A.M.
      • Low D.E.
      Epidemic of group A Streptococcus M/emm59 causing invasive disease in Canada.
      • Dillon H.C.
      • Dillon M.S.
      New streptococcal serotypes causing pyoderma and acute glomerulonephritis types 59, 60, and 61.
      • Phillips G.
      • Efstratiou A.
      • Tanna A.
      • Beall B.
      • Ferguson J.
      • Roworth M.
      An outbreak of skin sepsis in abattoir workers caused by an ‘unusual’ strain of Streptococcus pyogenes.
      suggest that emm59 strains have a predilection for skin. Indeed, on discovery of type emm59 GAS, it was shown that emm59 strains were more common among patients with pyoderma than in those with acute pharyngitis.
      • Dillon H.C.
      • Dillon M.S.
      New streptococcal serotypes causing pyoderma and acute glomerulonephritis types 59, 60, and 61.
      During the Canadian epidemic, emm59 strains were recovered in higher percentages from patients with abscess and soft-tissue infections than all other emm types.
      • Tyrrell G.J.
      • Lovgren M.
      • St Jean T.
      • Hoang L.
      • Patrick D.M.
      • Horsman G.
      • Van Caeseele P.
      • Sieswerda L.E.
      • McGeer A.
      • Laurence R.A.
      • Bourgault A.M.
      • Low D.E.
      Epidemic of group A Streptococcus M/emm59 causing invasive disease in Canada.
      A predominance of skin infections was also observed in a small emm59 outbreak in Scotland.
      • Phillips G.
      • Efstratiou A.
      • Tanna A.
      • Beall B.
      • Ferguson J.
      • Roworth M.
      An outbreak of skin sepsis in abattoir workers caused by an ‘unusual’ strain of Streptococcus pyogenes.
      In our present study, using a mouse model of skin transmission, we were able to show colonization of the skin of naïve mice by emm59 GAS organisms after contact with a skin-infected mouse. Carriage of emm59 on the skin of healthy individuals is also a possibility. Indeed, we were able to cultivate emm59 strains from the intact skin of mice for as many as 3 days after inoculation (data not shown). Thus, based on our aggregate results, we hypothesize that skin transmission is the major driver of epidemic emm59 GAS transmission.
      In summary, we report that an emm59 GAS epidemic, caused by the emergence of a distinct genetic clone of emm59 GAS, arose in Canada and spread to four states in the United States. The epidemic strains were highly virulent and caused more severe disease than a historic emm59 organism in the mouse skin infection model and in both the murine and nonhuman primate infection models of necrotizing fasciitis. Full-genome sequencing of all available epidemic strains unambiguously demonstrated a clonal origin of the epidemic and permitted us to resolve distinct geographic patterns of subclone dissemination. We show here that full-genome sequencing provides key data economically and rapidly that can be used to provide precise information about the population genomic landscape of bacterial epidemics, whether natural, accidental, or deliberate.

      Acknowledgments

      We thank public health and hospital laboratory personnel in Canada and Montana for their efforts, Gerold Spire for providing strains, Chris van Beneden and Ruth Lynfield for their help with strain epidemiological information, Jovanka Voyich-Kane and Irene Harrison for assistance, Kathryn Stockbauer for suggestions, Robert K. Selander for critical reading of an early version of the manuscript, and Tammy L. Humbird, Jamie L. Greaver, and Terry L. Blasdel for help with animal experiments.

      Supplementary data

      • Supplemental Figure S1

        Clustered regularly interspaced short palindromic repeat (CRISPR) analysis reveals differences among emm59 strains. Four variants were detected among emm59 strains (boxed area at approximately 4K in the genetic scheme). A: CRISPR variant 1, found in the historic emm59 strain MGAS15252, has four direct repeat (DR) consensus sequences and three spacer regions. B: CRISPR variants 2, 3, and 4. The vast majority of epidemic strains have a CRISPR variant that contains one DR and one spacer less than the historic strain (variant 2). Two other variants were identified among the epidemic strains (variants 3 and 4), each of which may be derived from variant 2 by loss of one DR and one spacer. No simple relationship was identified between the CRISPR variants and the geographic and/or temporal distribution of the strains. C: The DR consensus sequence of each CRISPR region, as well as the spacer sequences.

      • Supplemental Figure S2

        GAS burden in mouse muscle after intramuscular inoculation. There were no significant differences in the number of CFUs recovered per gram of total hindlimb muscle tissue between epidemic strain MGAS15252 and historic strain MGAS1882 at any time point. n = 20 animals per strain per time point.

      References

        • Lancefield R.C.
        The antigenic complex of Streptococcus haemolyticus: I.
        J Exp Med. 1928; 47: 91-103
        • Scott J.R.
        • Pulliam W.M.
        • Hollingshead S.K.
        • Fischetti V.A.
        Relationship of M protein genes in group A streptococci.
        Proc Natl Acad Sci USA. 1985; 82: 1822-1826
        • Manjula B.N.
        • Acharya A.S.
        • Fairwell T.
        • Fischetti V.A.
        Antigenic domains of the streptococcal Pep M5 protein.
        J Exp Med. 1986; 163: 129-138
        • Luca-Harari B.
        • Darenberg J.
        • Neal S.
        • Siljander T.
        • Strakova L.
        • Tanna A.
        • Creti R.
        • Ekelund K.
        • Koliou M.
        • Tassios P.T.
        • van der Linden M.
        • Straut M.
        • Vuopio-Varkila J.
        • Bouvet A.
        • Efstratiou A.
        • Schalén C.
        • Henriques-Normark B.
        • Jasir A.
        • Strep-EURO Study Group
        Clinical and microbiological characteristics of severe Streptococcus pyogenes disease in Europe.
        J Clin Microbiol. 2009; 47: 1155-1165
        • Steer A.C.
        • Law I.
        • Matatolu L.
        • Beall B.W.
        • Carapetis J.R.
        Global emm type distribution of group A streptococci: systematic review and implications for vaccine development.
        Lancet Infect Dis. 2009; 9: 611-616
        • Tyrrell G.J.
        • Lovgren M.
        • St Jean T.
        • Hoang L.
        • Patrick D.M.
        • Horsman G.
        • Van Caeseele P.
        • Sieswerda L.E.
        • McGeer A.
        • Laurence R.A.
        • Bourgault A.M.
        • Low D.E.
        Epidemic of group A Streptococcus M/emm59 causing invasive disease in Canada.
        Clin Infect Dis. 2010; 51: 1290-1297
        • Ardanuy C.
        • Rolo D.
        • Domenech A.
        • Tubau F.
        • Ayats J.
        • Grau I.
        • Pallares R.
        • Martin R.
        • Liñares J.
        Trends in invasive group A streptococci in adult patients in Barcelona (abstract).
        Clin Microbiol Infect. 2009; 15: S495
        • Dillon H.C.
        • Dillon M.S.
        New streptococcal serotypes causing pyoderma and acute glomerulonephritis types 59, 60, and 61.
        Infect Immun. 1974; 9: 1070-1078
        • Bentley D.R.
        • Balasubramanian S.
        • Swerdlow H.P.
        • Smith G.P.
        • Milton J.
        • Brown C.G.
        • et al.
        Accurate whole human genome sequencing using reversible terminator chemistry.
        Nature. 2008; 456: 53-59
        • Margulies M.
        • Egholm M.
        • Altman W.E.
        • Attiya S.
        • Bader J.S.
        • Bemben L.A.
        • et al.
        Genome sequencing in microfabricated high-density picolitre reactors.
        Nature. 2005; 437: 376-380
        • Lassmann T.
        • Hayashizaki Y.
        • Daub C.O.
        TagDust–a program to eliminate artifacts from next generation sequencing data.
        Bioinformatics. 2009; 25: 2839-2840
        • Nusbaum C.
        • Ohsumi T.K.
        • Gomez J.
        • Aquadro J.
        • Victor T.C.
        • Warren R.M.
        • Hung D.T.
        • Birren B.W.
        • Lander E.S.
        • Jaffe D.B.
        Sensitive, specific polymorphism discovery in bacteria using massively parallel sequencing.
        Nat Methods. 2009; 6: 67-69
        • Zerbino D.R.
        • Birney E.
        Velvet: algorithms for de novo short read assembly using de Bruijn graphs.
        Genome Res. 2008; 18: 821-829
        • Altschul S.F.
        • Madden T.L.
        • Schäffer A.A.
        • Zhang J.
        • Zhang Z.
        • Miller W.
        • Lipman D.J.
        Gapped BLAST and PSI-BLAST: a new generation of protein database search programs.
        Nucleic Acids Res. 1997; 25: 3389-3402
        • Thompson J.D.
        • Higgins D.G.
        • Gibson T.J.
        CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice.
        Nucleic Acids Res. 1994; 22: 4673-4680
        • Huson D.H.
        • Bryant D.
        Application of phylogenetic networks in evolutionary studies.
        Mol Biol Evol. 2006; 23: 254-267
        • Shelburne 3rd, S.A.
        • Granville C.
        • Tokuyama M.
        • Sitkiewicz I.
        • Patel P.
        • Musser J.M.
        Growth characteristics of and virulence factor production by group A Streptococcus during cultivation in human saliva.
        Infect Immun. 2005; 73: 4723-4731
        • Olsen R.J.
        • Musser J.M.
        Molecular pathogenesis of necrotizing fasciitis.
        Annu Rev Pathol. 2010; 5: 1-31
        • Sumby P.
        • Zhang S.
        • Whitney A.R.
        • Falugi F.
        • Grandi G.
        • Graviss E.A.
        • Deleo F.R.
        • Musser J.M.
        A chemokine-degrading extracellular protease made by group A Streptococcus alters pathogenesis by enhancing evasion of the innate immune response.
        Infect Immun. 2008; 76: 978-985
        • Shelburne 3rd, S.A.
        • Sumby P.
        • Sitkiewicz I.
        • Okorafor N.
        • Granville C.
        • Patel P.
        • Voyich J.
        • Hull R.
        • DeLeo F.R.
        • Musser J.M.
        Maltodextrin utilization plays a key role in the ability of group A Streptococcus to colonize the oropharynx.
        Infect Immun. 2006; 74: 4605-4614
        • Beres S.B.
        • Musser J.M.
        Contribution of exogenous genetic elements to the group A Streptococcus metagenome.
        PLoS ONE. 2007; 2: e800
        • Beres S.B.
        • Carroll R.K.
        • Shea P.R.
        • Sitkiewicz I.
        • Martinez-Gutierrez J.C.
        • Low D.E.
        • McGeer A.
        • Willey B.M.
        • Green K.
        • Tyrrell G.J.
        • Goldman T.D.
        • Feldgarden M.
        • Birren B.W.
        • Fofanov Y.
        • Boos J.
        • Wheaton W.D.
        • Honisch C.
        • Musser J.M.
        Molecular complexity of successive bacterial epidemics deconvoluted by comparative pathogenomics.
        Proc Natl Acad Sci USA. 2010; 107: 4371-4376
        • Shea P.R.
        • Beres S.B.
        • Flores A.R.
        • Ewbank A.L.
        • Gonzalez-Lugo J.H.
        • Martagon-Rosado A.J.
        • Martinez-Gutierrez J.C.
        • Rehman H.A.
        • Serrano-Gonzalez M.
        • Fittipaldi N.
        • Ayers S.D.
        • Webb P.
        • Willey B.M.
        • Low D.E.
        • Musser J.M.
        Distinct signatures of diversifying selection revealed by genome analysis of respiratory tract and invasive bacterial populations.
        Proc Natl Acad Sci USA. 2011; 108: 5039-5044
        • Marraffini L.A.
        • Sontheimer E.J.
        CRISPR interference: RNA-directed adaptive immunity in bacteria and archaea.
        Nat Rev Genet. 2010; 11: 181-190
        • Holt K.E.
        • Teo Y.Y.
        • Li H.
        • Nair S.
        • Dougan G.
        • Wain J.
        • Parkhill J.
        Detecting SNPs and estimating allele frequencies in clonal bacterial populations by sequencing pooled DNA.
        Bioinformatics. 2009; 25: 2074-2075
        • Phillips G.
        • Efstratiou A.
        • Tanna A.
        • Beall B.
        • Ferguson J.
        • Roworth M.
        An outbreak of skin sepsis in abattoir workers caused by an ‘unusual’ strain of Streptococcus pyogenes.
        J Med Microbiol. 2000; 49: 371-374
        • Sun H.
        • Ringdahl U.
        • Homeister J.W.
        • Fay W.P.
        • Engleberg N.C.
        • Yang A.Y.
        • Rozek L.S.
        • Wang X.
        • Sjöbring U.
        • Ginsburg D.
        Plasminogen is a critical host pathogenicity factor for group A streptococcal infection.
        Science. 2004; 305: 1283-1286
        • Sumby P.
        • Tart A.H.
        • Musser J.M.
        A non-human primate model of acute group A Streptococcus pharyngitis.
        Methods Mol Biol. 2008; 431: 255-267
        • Olsen R.J.
        • Ashraf M.
        • Gonulal V.E.
        • Ayeras A.A.
        • Cantu C.
        • Shea P.R.
        • Carroll R.K.
        • Humbird T.
        • Greaver J.L.
        • Swain J.L.
        • Chang E.
        • Ragasa W.
        • Jenkins L.
        • Lally K.P.
        • Blasdel T.
        • Cagle P.
        • Musser J.M.
        Lower respiratory tract infection in cynomolgus macaques (Macaca fascicularis) infected with group A Streptococcus.
        Microb Pathog. 2010; 49: 336-347
        • Olsen R.J.
        • Sitkiewicz I.
        • Ayeras A.A.
        • Gonulal V.E.
        • Cantu C.
        • Beres S.B.
        • Green N.M.
        • Lei B.
        • Humbird T.
        • Greaver J.
        • Chang E.
        • Ragasa W.P.
        • Montgomery C.A.
        • Cartwright Jr, J.
        • McGeer A.
        • Low D.E.
        • Whitney A.R.
        • Cagle P.T.
        • Blasdel T.L.
        • DeLeo F.R.
        • Musser J.M.
        Decreased necrotizing fasciitis capacity caused by a single nucleotide mutation that alters a multiple gene virulence axis.
        Proc Natl Acad Sci USA. 2010; 107: 888-893
        • Hoe N.P.
        • Vuopio-Varkila J.
        • Vaara M.
        • Grigsby D.
        • De Lorenzo D.
        • Fu Y.X.
        • Dou S.J.
        • Pan X.
        • Nakashima K.
        • Musser J.M.
        Distribution of streptococcal inhibitor of complement variants in pharyngitis and invasive isolates in an epidemic of serotype M1 group A Streptococcus infection.
        J Infect Dis. 2001; 183: 633-639
        • Hamburger Jr, M.
        • Green M.J.
        • Hamburger V.G.
        The problem of the dangerous carrier of hemolytic streptococci; spread of infection by individuals with strongly positive nose cultures who expelled large numbers of hemolytic streptococci.
        J Infect Dis. 1945; 77: 96-108
        • Hamburger Jr, M.
        • Robertson O.H.
        Expulsion of group A hemolytic streptococci in droplets and droplet nuclei by sneezing, coughing and talking.
        Am J Med. 1948; 4: 690-701
        • Hamburger M.
        • Robertson O.H.
        The expulsion of beta hemolytic streptococci by sneezing.
        Proc Annu Meet Cent Soc Clin Res US. 1947; 20: 72
        • Shulman S.T.
        • Tanz R.R.
        • Dale J.B.
        • Beall B.
        • Kabat W.
        • Kabat K.
        • Cederlund E.
        • Patel D.
        • Rippe J.
        • Li Z.
        • Sakota V.
        • North American Streptococcal Pharyngitis Surveillance Group
        Seven-year surveillance of North American pediatric group A streptococcal pharyngitis isolates.
        Clin Infect Dis. 2009; 49: 78-84
        • Shulman S.T.
        • Tanz R.R.
        • Kabat W.
        • Kabat K.
        • Cederlund E.
        • Patel D.
        • Li Z.
        • Sakota V.
        • Dale J.B.
        • Beall B.
        • US Streptococcal Pharyngitis Surveillance Group
        Group A streptococcal pharyngitis serotype surveillance in North America, 2000–2002.
        Clin Infect Dis. 2004; 39: 325-332
        • Jones S.
        • Zhang X.
        • Parsons D.W.
        • Lin J.C.
        • Leary R.J.
        • Angenendt P.
        • et al.
        Core signaling pathways in human pancreatic cancers revealed by global genomic analyses.
        Science. 2008; 321: 1801-1806
        • Ley T.J.
        • Mardis E.R.
        • Ding L.
        • Fulton B.
        • McLellan M.D.
        • Chen K.
        • et al.
        DNA sequencing of a cytogenetically normal acute myeloid leukaemia genome.
        Nature. 2008; 456: 66-72
        • The Cancer Genome Atlas Research Network
        Comprehensive genomic characterization defines human glioblastoma genes and core pathways.
        Nature. 2008; 455: 1061-1068
        • Musser J.M.
        • Shelburne 3rd, S.A.
        A decade of molecular pathogenomic analysis of group A Streptococcus.
        J Clin Invest. 2009; 119: 2455-2463
        • Harris S.R.
        • Feil E.J.
        • Holden M.T.
        • Quail M.A.
        • Nickerson E.K.
        • Chantratita N.
        • Gardete S.
        • Tavares A.
        • Day N.
        • Lindsay J.A.
        • Edgeworth J.D.
        • de Lencastre H.
        • Parkhill J.
        • Peacock S.J.
        • Bentley S.D.
        Evolution of MRSA during hospital transmission and intercontinental spread.
        Science. 2010; 327: 469-474
        • Croucher N.J.
        • Harris S.R.
        • Fraser C.
        • Quail M.A.
        • Burton J.
        • van der Linden M.
        • McGee L.
        • von Gottberg A.
        • Song J.H.
        • Ko K.S.
        • Pichon B.
        • Baker S.
        • Parry C.M.
        • Lambertsen L.M.
        • Shahinas D.
        • Pillai D.R.
        • Mitchell T.J.
        • Dougan G.
        • Tomasz A.
        • Klugman K.P.
        • Parkhill J.
        • Hanage W.P.
        • Bentley S.D.
        Rapid pneumococcal evolution in response to clinical interventions.
        Science. 2011; 331: 430-434
        • Gardy J.L.
        • Johnston J.C.
        • Ho Sui S.J.
        • Cook V.J.
        • Shah L.
        • Brodkin E.
        • Rempel S.
        • Moore R.
        • Zhao Y.
        • Holt R.
        • Varhol R.
        • Birol I.
        • Lem M.
        • Sharma M.K.
        • Elwood K.
        • Jones S.J.
        • Brinkman F.S.
        • Brunham R.C.
        • Tang P.
        Whole-genome sequencing and social-network analysis of a tuberculosis outbreak [Erratum appeared in N Engl J Med 2011 Jun 2;364(22):2174].
        N Engl J Med. 2011; 364: 730-739
        • Nelson M.I.
        • Tan Y.
        • Ghedin E.
        • Wentworth D.E.
        • St George K.
        • Edelman L.
        • Beck E.T.
        • Fan J.
        • Lam T.T.
        • Kumar S.
        • Spiro D.J.
        • Simonsen L.
        • Viboud C.
        • Holmes E.C.
        • Henrickson K.J.
        • Musser J.M.
        Phylogeography of the spring and fall waves of the H1N1/09 pandemic influenza virus in the United States.
        J Virol. 2011; 85: 828-834
        • Rasko D.A.
        • Webster D.R.
        • Sahl J.W.
        • Bashir A.
        • Boisen N.
        • Scheutz F.
        • Paxinos E.E.
        • Sebra R.
        • Chin C.S.
        • Iliopoulos D.
        • Klammer A.
        • Peluso P.
        • Lee L.
        • Kislyuk A.O.
        • Bullard J.
        • Kasarskis A.
        • Wang S.
        • Eid J.
        • Rank D.
        • Redman J.C.
        • Steyert S.R.
        • Frimodt-Møller J.
        • Struve C.
        • Petersen A.M.
        • Krogfelt K.A.
        • Nataro J.P.
        • Schadt E.E.
        • Waldor M.K.
        Origins of the E. coli strain causing an outbreak of hemolytic-uremic syndrome in Germany.
        N Engl J Med. 2011; 365: 709-717
        • Centurion-Lara A.
        • LaFond R.E.
        • Hevner K.
        • Godornes C.
        • Molini B.J.
        • Van Voorhis W.C.
        • Lukehart S.A.
        Gene conversion: a mechanism for generation of heterogeneity in the tprK gene of Treponema pallidum during infection.
        Mol Microbiol. 2004; 52: 1579-1596
        • Futse J.E.
        • Brayton K.A.
        • Dark M.J.
        • Knowles Jr, D.P.
        • Palmer G.H.
        Superinfection as a driver of genomic diversification in antigenically variant pathogens.
        Proc Natl Acad Sci USA. 2008; 105: 2123-2127
        • Johnson W.E.
        • Desrosiers R.C.
        Viral persistence: HIV's strategies of immune system evasion.
        Annu Rev Med. 2002; 53: 499-518
        • Steer A.C.
        • Batzloff M.R.
        • Mulholland K.
        • Carapetis J.R.
        Group A streptococcal vaccines: facts versus fantasy.
        Curr Opin Infect Dis. 2009; 22: 544-552
        • Denamur E.
        • Matic I.
        Evolution of mutation rates in bacteria.
        Mol Microbiol. 2006; 60: 820-827
        • Wannamaker L.W.
        Differences between streptococcal infections of the throat and of the skin (second of two parts).
        N Engl J Med. 1970; 282: 78-85
        • Wannamaker L.W.
        Differences between streptococcal infections of the throat and of the skin.
        N Engl J Med. 1970; 282: 23-31
        • Bessen D.E.
        • Lizano S.
        Tissue tropisms in group A streptococcal infections.
        Future Microbiol. 2010; 5: 623-638

      Linked Article