Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Clustered Regularly Interspaced Short Palindromic Repeats Are emm Type-Specific in Highly Prevalent Group A Streptococci

  • Po-Xing Zheng,

    Affiliation Institute of Basic Medical Sciences, College of Medicine, National Cheng Kung University, Tainan, Taiwan

  • Yuen-Chi Chan,

    Affiliation Department of Medical Laboratory Science and Biotechnology, College of Medicine, National Cheng Kung University, Tainan, Taiwan

  • Chien-Shun Chiou,

    Affiliation Center for Research, Diagnostics and Vaccine Development, Centers for Disease Control, Taichung, Taiwan

  • Chuan Chiang-Ni,

    Affiliation Department of Microbiology and Immunology, College of Medicine, Chang Gung University, Tao-Yuan, Taiwan

  • Shu-Ying Wang,

    Affiliation Department of Microbiology and Immunology, College of Medicine, National Cheng Kung University, Tainan, Taiwan

  • Pei-Jane Tsai,

    Affiliation Department of Medical Laboratory Science and Biotechnology, College of Medicine, National Cheng Kung University, Tainan, Taiwan

  • Woei-Jer Chuang,

    Affiliation Department of Biochemistry and Molecular Biology, College of Medicine, National Cheng Kung University, Tainan, Taiwan

  • Yee-Shin Lin,

    Affiliation Department of Microbiology and Immunology, College of Medicine, National Cheng Kung University, Tainan, Taiwan

  • Ching-Chuan Liu,

    Affiliation Department of Pediatrics, College of Medicine, National Cheng Kung University, Tainan, Taiwan

  • Jiunn-Jong Wu

    jjwu@mail.ncku.edu.tw

    Affiliations Institute of Basic Medical Sciences, College of Medicine, National Cheng Kung University, Tainan, Taiwan, Department of Medical Laboratory Science and Biotechnology, College of Medicine, National Cheng Kung University, Tainan, Taiwan, Center of Infectious Disease and Signaling Research, National Cheng Kung University, Tainan, Taiwan, Department of Biotechnology and Laboratory Science in Medicine, School of Biomedical Science and Engineering, National Yang-Ming University, Taipei, Taiwan

Clustered Regularly Interspaced Short Palindromic Repeats Are emm Type-Specific in Highly Prevalent Group A Streptococci

  • Po-Xing Zheng, 
  • Yuen-Chi Chan, 
  • Chien-Shun Chiou, 
  • Chuan Chiang-Ni, 
  • Shu-Ying Wang, 
  • Pei-Jane Tsai, 
  • Woei-Jer Chuang, 
  • Yee-Shin Lin, 
  • Ching-Chuan Liu, 
  • Jiunn-Jong Wu
PLOS
x

Abstract

Clustered regularly interspaced short palindromic repeats (CRISPR) are the bacterial adaptive immune system against foreign nucleic acids. Given the variable nature of CRISPR, it could be a good marker for molecular epidemiology. Group A streptococcus is one of the major human pathogens. It has two CRISPR loci, including CRISPR01 and CRISPR02. The aim of this study was to analyze the distribution of CRISPR-associated gene cassettes (cas) and CRISPR arrays in highly prevalent emm types. The cas cassette and CRISPR array in two CRISPR loci were analyzed in a total of 332 strains, including emm1, emm3, emm4, emm12, and emm28 strains. The CRISPR type was defined by the spacer content of each CRISPR array. All strains had at least one cas cassette or CRISPR array. More than 90% of the spacers were found in one emm type, specifically. Comparing the consistency between emm and CRISPR types by Simpson’s index of diversity and the adjusted Wallace coefficient, CRISPR01 type was concordant to emm type, and CRISPR02 showed unidirectional congruence to emm type, suggesting that at least for the majority of isolates causing infection in high income countries, the emm type can be inferred from CRISPR analysis, which can further discriminate isolates sharing the same emm type.

Introduction

Clustered regularly interspaced short palindromic repeats (CRISPR) are composed of serial spacer sequences flanked by repeats. CRISPR is now considered to be the prokaryotic adaptive immune system against foreign nucleic acid [1]. The mechanism involves two steps. The first step is an immunization. Prokaryotes can acquire small fragments from invading sequences, including phages or plasmids, to become new spacers in CRISPR. The second step is immunity. CRISPR can be transcribed from a leader promoter into a long RNA, and further processed into small RNAs containing one spacer and a partial repeat. Mature small RNAs, together with serial CRISPR-associated proteins (Cas), can recognize and degrade invading sequences complementary to spacer sequences. Therefore, CRISPR is like molecular “vaccination cards”, recording bacteria-virus interactions in spacer-repeat units [2], and providing adaptive immunity against foreign nucleic acids.

Since CRISPR is highly polymorphic, it has been used for bacterial typing. In Mycobacterium tuberculosis, 43 spacer sequences are used to differentiate epidemic clones, which is named spoligotyping [3]. In Escherichia coli, CRISPR is congruent with the evolutionary divergence of shiga toxin-producing E. coli, and has been used to detect hemorrhagic E. coli [4]. CRISPR typing also has good discriminatory powers in Salmonella, Lactobacillus buchneri, Campylobacter jejuni, Propionibacterium acnes, and Erwinia amylovora [5].

Group A streptococcus (GAS, Streptococcus pyogenes) is a Gram positive coccus. The diseases caused by GAS are varied, including pharyngitis, necrotizing fasciitis, and streptococcal toxic shock syndrome [6]. Traditionally, the serotype of GAS is determined by the surface virulence factor, M protein. However, this is limited by the availability of M typing sera. In recent years, sequencing of the emm gene, which encodes the M protein, has largely replaced the serotyping of the M protein [7]. The emm sequence typing is focused on the highly variable region of M protein. Two hundred and twenty three different emm types have been identified as of now, worldwide [8]. The prevalence rate of emm types is different among different regions, but the emm1 type has been the most common type since the 1980s, followed by emm12, emm28, emm3, and emm4 [9]. These were the most prevalent emm types in high-income countries. In Asia, emm1, emm4, and emm12 types are the most common types [9], and these types account for 79~89% of the isolates from scarlet fever and pediatric infections in central and southern Taiwan [10, 11].

In GAS, there are two CRISPR loci, which are named CRISPR01 and CRISPR02 [12]. Each CRISPR locus has its own cas cassette and CRISPR array. Experimental data demonstrate that the CRISPR01 locus of GAS strain SF370 can digest the invading nucleic acid, whereas the function of the CRISPR02 locus remains unclear [13]. The spacer contents and number of spacers are also associated with erythromycin susceptibility in emm12, emm75, and emm92 strains [14]. The aim of this study was to analyze the association between CRISPR and emm types in GAS, and we found that the emm type and spacer content of two CRISPR loci were associated.

Materials and Methods

Bacterial characterization

A total of 151 GAS isolates, including strain A20 (with a complete genome sequence) [15], were collected from National Cheng Kung University Hospital in southern Taiwan during 1994–2008 (S1 Table). Strain characterization, including identification, PFGE, and emm typing, was described in previous work [14]. A total of 170 non-redundant GAS strains with incomplete genome sequences and 11 strains with complete genome sequences were obtained from GenBank in the National Center for Biotechnology Information. The accession number and strain information of incomplete- and completely-sequenced strains are listed in S2 and S3 Tables, respectively.

Determination of emm type

PCR amplification and sequencing of emm genes were performed as in previous descriptions [8]. To determine the emm type of strains with complete or incomplete genome sequences, the primer sequence of emm-Seq2 (tattcgcttagaaaattaaaaacagg) and emm-2 (gcaagttcttcagcttgttt) were used to search against GAS contigs by “somewhat similar sequences BLASTN” with default parameters. The targeted sequences from the emm-Seq2 to emm-2 region were used to search against the emm type-specific CDC database (http://www2a.cdc.gov/ncidod/biotech/strepblast.asp) to determine the emm type.

Determination of cas cassette and CRISPR array in two CRISPR loci

The spacer contents of CRISPR01 and CRISPR02 were determined as described [14]. Briefly, CRISPR01 was amplified with primer CRISPR1-3 (cggtacaattcttgtgctcgaa) and CRISPR1-4 (tcaatggcgtttaacttgatgg) and sequenced with CRISPR1-1 (tgagaaacccgaagtgaa). CRISPR02 was amplified with primer CRISPR2-1 (tctgtgacacccgcagaattt) and CRISPR2-2 (aaaccagccccgtaacctaaa) and sequenced with CRISPR2-1. The spacer and repeat sequences were determined with the CRISPRtionary tool and CRISPR finder tool [16]. The presence of cas cassettes was determined by polymerase chain reaction in a previous study [14].

In silico analysis was used to determine the presence of CRISPR in strains with complete or incomplete genome sequences. The cas cassettes from CRISPR01 and CRISPR02 in MGAS9429 were used to search against the complete or incomplete genome sequences by megaBLASTN with default parameters. The contigs with cas cassettes were chosen. The presence of a CRISPR array in contigs with cas cassettes was identified by the CRISPR recognition tool and CRISPRFinder [17, 18]. The spacers in CRISPR01 and CRISPR02 loci were further identified by the CRISPRtionary tool [16]. Each unique spacer was designated with a specific numeral code.

Estimation of diversity index and grouping comparison coefficients

Simpson’s index of diversity was used to measure the discriminatory ability of different typing methods [19]. This index indicates the probability that two strains which were randomly selected from a population belong to different types. The confidence interval (CI) and p value of Simpson’s index were calculated according to a previous study [20]. To compare different typing methods, the Adjusted Wallace coefficient was used to describe the directional relationship of congruence between two typing methods [21]. A higher score indicates two different typing methods have higher congruence. The CI of the adjusted Wallace coefficient was estimated as described previously [22]. The analysis was performed at Comparing Partitions (http://darwin.phyloviz.net/ComparingPartitions/index.php?link=Home).

Statistical analysis

The Fisher's exact test was performed in SPSS software, version 17.0 (SPSS Inc., Chicago, IL, USA). A p value lower than 0.05 was considered to indicate statistical significance.

Results

Presence of CRISPR in GAS strains

A total of 182 non-redundant emm1, emm3, emm4, emm12, and emm28 strains with complete or incomplete genome sequences from NCBI were collected. Since these strains were not isolated in Taiwan (except emm1 strain A20, having a complete genome sequence), they were designated as “foreign” strains. To increase the strain collection, a total of 150 local strains, including 34, 49, and 67 strains of emm1, emm4, and emm12 types, respectively, were also included, which were designated as “local” strains.

Among 332 strains, all strains had at least one cas cassette or CRISPR array in the CRISPR01 or CRISPR02 loci (Table 1). The prevalence of a CRISPR array in the CRISPR01 locus was significantly less than in the CRISPR02 locus, while having a cas cassette in the CRISPR01 locus was more common than in the CRISPR02 locus (Table 1, p <0.001). All of the emm1, emm12, and emm28 strains had two intact CRISPR loci, including CRISPR array and cas cassette, except one emm1 strain (DSM 20565, S2 Table). emm3 and emm4 strains only had a cas cassette in the CRISPR01 locus. emm4 strains had a cas cassette and CRISPR array in the CRISPR02 locus, whereas emm3 strains did not have a CRISPR02 locus (Table 1). Thus, strains having these highly prevalent emm types had at least one cas cassette or CRISPR array.

thumbnail
Table 1. The prevalence of cas cassettes and CRISPR arrays in different emm types and regions of isolation.

http://dx.doi.org/10.1371/journal.pone.0145223.t001

Analysis of the spacer contents in CRISPR01 and CRISPR02 loci

A total of 14 and 21 unique spacer sequences were found in CRISPR01 and CRISPR02 loci, respectively, from all strains. In the CRISPR01 locus, 92.9% (13/14) of spacers were only found in one emm type, and 90.5% (19/21) of CRISPR02 spacers were specifically found in one emm type (S4 Table), suggesting that most CRISPR spacers were emm type-specific.

To further analyze the association between CRISPR and emm types, specific spacer contents were defined as a CRISPR type. In addition, the “CRISPRa type” was further defined as the combination of spacer contents from two CRISPR loci. Thus, based on spacer contents, there were 3 CRISPR types in one strain. Different numeral codes were used to represent different CRISPR types (Table 2). All CRISPR types were emm type-specific, except for CRISPR01 type 37, which corresponds to all isolates with a cas cassette but no CRISPR array in the CRISPR01 locus.

thumbnail
Table 2. The distribution of emm type, CRISPR01, CRISPR02, and CRISPRa type among local and foreign GAS isolates.

http://dx.doi.org/10.1371/journal.pone.0145223.t002

Association between emm and CRISPR

CRISPR01 type 12 and CRISPR02 type 8 accounted for 68.5% (61/89) and 98.9% (88/89) of all emm1 strains, respectively (Table 2). CRISPR01 type 23 and CRISPR02 type 17 accounted for 100% (5/5) and 80% (4/5) of all emm28 strains, respectively (Table 2). In emm12, 95.5% (127/133) of the strains had CRISPR01 type 4. Although emm12 strains had diverse CRISPR02 loci, spacer No.212 from CRISPR02 was found in 96.2% of all emm12 isolates (128 of 133, Table 2).

All emm3 strains only had a cas cassette in the CRISPR01 locus. All emm4 strains had a cas cassette in the CRISPR01 locus, and had a CRISPR array and cas cassette in the CRISPR02 locus (Table 1). In emm4 strains, 78.4% (40/51) had CRISPR02 type 13 (Table 2). Since most emm4 strains were isolated from the local region, to rule out the possibility that CRISPR02 type 13 was the same clone, Smal-digested PFGE typing was performed. The local emm4 strains with CRISPR02 type 13 had different PFGE types (data not shown), suggesting that the strains with a conserved CRISPR02 type were not closely related. Together, these results suggest that each emm type had their dominant CRISPR type.

To compare the emm and the three CRISPR types, Simpson’s index of diversity was used. CRISPR01 and emm type presented a similar discriminatory power, according to the Simpson's ID (Table 3). However, CRISPR02 and CRISPRa types had a significantly higher discriminatory power when compared to emm type (Table 3). The association between emm and CRISPR01 types was supported by the adjusted Wallace method, which showed bidirectional congruence (Fig 1, the coefficients and 95% CI are listed in S5 Table), indicating that emm and CRISPR01 types were correlated. In addition, the adjusted Wallace coefficients from CRISPR02 and CRISPRa to emm type showed strong unidirectional congruence (Fig 1, the coefficients and 95% CI are listed in S5 Table), suggesting that the CRISPR02 and CRISPRa types can be used to infer the emm type among the highly prevalent emm types analyzed in this study.

thumbnail
Fig 1. Correlations among emm, CRISPRa, CRISPR01, and CRISPR02 types.

The congruence among different typing methods was estimated by Adjusted Wallace coefficients. The dash, thin, and thick arrows indicate Adjusted Wallace coefficients of 0.6~0.7, 0.7~0.9, and >0.9, respectively.

http://dx.doi.org/10.1371/journal.pone.0145223.g001

To further demonstrate the robust relationship between CRISPR and emm types, the Simpson’s ID and adjusted Wallace coefficient between local and foreign strains were compared. Results showed that most values were not significantly different between local and foreign strains, except for the adjusted Wallace coefficients of emm to CRISPR02 and CRISPRa (Figs. A and B in S1 File). This suggests that the strong associations between CRISPS02/CRISPRa and emm type described for the whole dataset were robust and independent of region.

thumbnail
Table 3. Simpson’s index of diversity (ID) of emm, CRISPRa, CRISPR01, and CRISPR02 types.

http://dx.doi.org/10.1371/journal.pone.0145223.t003

Discussion

In this study, we have demonstrated that highly prevalent emm types of GAS strains had at least one cas cassette or CRISPR array, and more than 90% of spacers were emm type-specific. Based on Simpson’s index of diversity and adjusted Wallace coefficient, CRISPR01 and emm types were associated, and CRISPR02 showed strong unidirectional congruence to emm type, suggesting that CRISPR typing can be used as an alternative way to infer the emm type of GAS, at least among the highly prevalent emm types included in this study. Furthermore, because of the higher Simpson’s ID of CRISPR02 and CRISPRa, and the strong unidirectional congruence from CRISPR02 and CRISPRa to emm type, CRISPR typing can be also used to discriminate isolates of the same emm type. Further studies including more genetic lineages are required to compare the CRISPR typing with different typing methods, such as PFGE, multilocus sequence typing, or exotoxin profiles to demonstrate the value of CRISPR typing.

In the CRISPR01 locus of GAS, Hoe et al analyzed 30 emm1 strains collected from Texas. They found 9 different spacers [23]. Interestingly, 8 spacers of the Texas collection were comparable to our spacers in emm1 strains, further supporting that the spacers were emm type-specific.

Although CRISPR01 and emm type showed high congruence, all emm3 and emm4 strains had CRISPR01 type 37, indicating that the CRISPR01 type cannot distinguish between the emm3 and emm4 strains. Since emm4 strains had an intact CRISPR02 locus, when combined with the information from CRISPR02, emm3 and emm4 strains can be distinguished. In addition, the erythromycin resistance of emm12 strains has been associated with the spacer contents and number of spacers which were often found in the CRISPR02 type [14]. Therefore, CRISPR02 can not only distinguish the emm type, but also predict macrolide susceptibility, namely in emm12 strains.

Since combination of CRISPR01 and CRISPR02 types, or CRISPR02 type only can infer emm types, we propose that CRISPR typing can be performed in two alternative ways. First, the CRISPR01 and CRISPR02 arrays are amplified with primer CRISPR1-3 and CRISPR1-4, CRISPR2-1 and CRISPR2-2, respectively. The gel electrophoresis is performed to confirm the presence of two CRISPR arrays. The CRISPR01 array is further sequenced to determine its spacers. In our study, the sequence of CRISPR01 spacers combined with the PCR detection of the CRISPR02 array was sufficient to differentiate highly prevalent emm types, except one emm1 strain. Second, given the high value of the respective adjusted Wallace coefficient, CRISPR typing can be performed by amplifying and sequencing of CRISPR02 array only. It is not required to detect cas genes for inferring emm type. Furthermore, since the number of spacers was negatively associated with erythromycin susceptibility [14], the size of the CRISPR02 array obtained by PCR and gel electrophoresis can be used to infer the erythromycin susceptibility in emm12 strains.

Since the CRISPR array results from the bacteria-phage interaction, and phage diversity is tremendous, CRISPR spacers should be geographically specific [2426]. However, our study showed the association between emm and CRISPR was robust among different regions, suggesting the CRISPR of GAS is independent of region. Interestingly, when the human gut microbiome was analyzed, unrelated people from different countries shared ~22% of spacers [27]. The reasons leading to form “conserved spacers” are still unclear. Possibly these strains shared recent common ancestry, or the conserved pattern was due to the slow insertion and deletion of spacers [28]. Population genomic studies showed that the contemporary emm1 and emm12 GAS strains originated in the early twentieth century [29, 30], suggesting recent common ancestry might exist in GAS. The number of spacers in GAS was less than in other streptococci [12], indicating that acquiring new spacers in GAS is not as efficient as in other streptococci, which may be due to the slow insertion and deletion of spacers. Another possibility to explain the robust association between CRISPR and emm types may be that the patterns of prophage-encoded virulence factors are associated with emm type [31, 32]. Since CRISPR arrays are expected to be closely associated with prophage content, it is possible that the association between CRISPR and emm type is due to the association between emm type and phage content. Further studies are required to support these hypotheses.

Our results suggest that CRISPR typing is a valuable typing method for GAS, although several limitations should be acknowledged in this study. Since most emm4 strains were collected locally and all emm28 strains were foreign, more geographically diverse strains of these emm types are required to analyze the association between emm and CRISPR types. In addition, strains of other less common emm types should also be included to further analyze this association. The correlation between emm and CRISPR types might be underrated due to the methodology used to identify CRISPR. In Listeria monocytogenes, cas gene-independent CRISPR is found [33]. However in our experiments, only the contigs with a cas cassette were chosen to analyze the presence of a CRISPR array. Therefore, the strains with a CRISPR array but without a cas cassette were missed in our study. Furthermore, studies from Africa and the Pacific region reported a much higher diversity of emm types, without a clear dominance of particular types [9]. Additionally, emm5, emm6, and emm18 strains did not have a CRISPR01 locus [12], which would limit the application of CRISPR typing in several emm types.

In summary, CRISPR spacers were emm type-specific in highly prevalent GAS, suggesting that at least for the majority of isolates causing infection in high income countries, the emm type can be inferred from CRISPR typing, which can further discriminate isolates sharing the same emm type.

Supporting Information

S1 File. Comparison of Simpson’s ID (Figure A in S1 File) and adjusted Wallace coefficients (Figure B in S1 File) between local and foreign strains.

doi:10.1371/journal.pone.0145223.s001

(DOCX)

S1 Table. The emm type, year of isolation, and CRISPR information of local strains used in this study.

doi:10.1371/journal.pone.0145223.s002

(DOCX)

S2 Table. The emm type and CRISPR information of foreign strains with incomplete genomes used in this study.

doi:10.1371/journal.pone.0145223.s003

(DOCX)

S3 Table. The emm type, accession number, and CRISPR information of strains with complete genomes used in this study

doi:10.1371/journal.pone.0145223.s004

(DOCX)

S4 Table. Spacer sequences from CRISPR01 and CRISPR02 and their relationship to emm type

doi:10.1371/journal.pone.0145223.s005

(DOCX)

S5 Table. Adjusted Wallace coefficients and jackknife pseudo-value 95% confident interval (CI) for the emm, CRISPRa, CRISPR01, and CRISPR02 types among all foreign and local strains.

doi:10.1371/journal.pone.0145223.s006

(DOCX)

Acknowledgments

We are very grateful to Robert M. Jonas (Texas Lutheran University) for helpful comments on the manuscript.

Author Contributions

Conceived and designed the experiments: PXZ CCN. Performed the experiments: PXZ YCC CSC. Analyzed the data: PXZ SYW PJT WJC YSL CCL JJW. Contributed reagents/materials/analysis tools: PXZ CSC. Wrote the paper: PXZ JJW.

References

  1. 1. Wiedenheft B, Sternberg SH, Doudna JA. RNA-guided genetic silencing systems in bacteria and archaea. Nature. 2012;482(7385):331–8. doi: 10.1038/nature10886 pmid:22337052.
  2. 2. Vale PF, Little TJ. CRISPR-mediated phage resistance and the ghost of coevolution past. Proc R Soc B-Biol Sci. 2010;277(1691):2097–103. Epub 2010/03/20. rspb.2010.0055 [pii] doi: 10.1098/rspb.2010.0055 pmid:20236977; PubMed Central PMCID: PMC2880148.
  3. 3. Kamerbeek J, Schouls L, Kolk A, van Agterveld M, van Soolingen D, Kuijper S, et al. Simultaneous detection and strain differentiation of Mycobacterium tuberculosis for diagnosis and epidemiology. J Clin Microbiol. 1997;35(4):907–14. pmid:9157152; PubMed Central PMCID: PMC229700.
  4. 4. Delannoy S, Beutin L, Fach P. Use of clustered regularly interspaced short palindromic repeat sequence polymorphisms for specific detection of enterohemorrhagic Escherichia coli strains of serotypes O26:H11, O45:H2, O103:H2, O111:H8, O121:H19, O145:H28, and O157:H7 by real-time PCR. J Clin Microbiol. 2012;50(12):4035–40. doi: 10.1128/JCM.02097-12 pmid:23035199; PubMed Central PMCID: PMC3503007.
  5. 5. Shariat N, Dudley EG. CRISPRs: molecular signatures used for pathogen subtyping. Appl Environ Microbiol. 2014;80(2):430–9. doi: 10.1128/AEM.02790-13 pmid:24162568; PubMed Central PMCID: PMC3911090.
  6. 6. Cole JN, Barnett TC, Nizet V, Walker MJ. Molecular insight into invasive group A streptococcal disease. Nat Rev Microbiol. 2011;9(10):724–36. Epub 2011/09/17. doi: 10.1038/nrmicro2648 nrmicro2648 [pii]. pmid:21921933.
  7. 7. Beall B, Gherardi G, Lovgren M, Facklam RR, Forwick BA, Tyrrell GJ. emm and sof gene sequence variation in relation to serological typing of opacity-factor-positive group A streptococci. Microbiology. 2000;146 (Pt 5):1195–209. pmid:10832648.
  8. 8. McMillan DJ, Dreze PA, Vu T, Bessen DE, Guglielmini J, Steer AC, et al. Updated model of group A streptococcus M proteins based on a comprehensive worldwide study. Clin Microbiol Infect. 2013;19(5):E222–9. doi: 10.1111/1469-0691.12134 pmid:23464795.
  9. 9. Steer AC, Law I, Matatolu L, Beall BW, Carapetis JR. Global emm type distribution of group A streptococci: systematic review and implications for vaccine development. Lancet Infect Dis. 2009;9(10):611–6. doi: 10.1016/S1473-3099(09)70178-1 pmid:19778763.
  10. 10. Chiou CS, Liao TL, Wang TH, Chang HL, Liao JC, Li CC. Epidemiology and molecular characterization of Streptococcus pyogenes recovered from scarlet fever patients in central Taiwan from 1996 to 1999. J Clin Microbiol. 2004;42(9):3998–4006. doi: 10.1128/JCM.42.9.3998–4006.2004 pmid:15364982; PubMed Central PMCID: PMC516299.
  11. 11. Lin HC, Wang SM, Lin YL, Lin YS, Wu JJ, Chuang WJ, et al. Group A streptococcal infection caused by emm1 strains among children in southern Taiwan. Eur J Clin Microbiol Infect Dis. 2008;27(12):1253–6. Epub 2008/07/04. doi: 10.1007/s10096-008-0561-z pmid:18597127.
  12. 12. Nozawa T, Furukawa N, Aikawa C, Watanabe T, Haobam B, Kurokawa K, et al. CRISPR inhibition of prophage acquisition in Streptococcus pyogenes. PLoS One. 2011;6(5):e19543. doi: 10.1371/journal.pone.0019543 pmid:21573110; PubMed Central PMCID: PMC3089615.
  13. 13. Deltcheva E, Chylinski K, Sharma CM, Gonzales K, Chao Y, Pirzada ZA, et al. CRISPR RNA maturation by trans-encoded small RNA and host factor RNase III. Nature. 2011;471(7340):602–7. Epub 2011/04/02. doi: 10.1038/nature09886 pmid:21455174; PubMed Central PMCID: PMC3070239.
  14. 14. Zheng PX, Chiang-Ni C, Wang SY, Tsai PJ, Kuo CF, Chuang WJ, et al. Arrangement and number of clustered regularly interspaced short palindromic repeat spacers are associated with erythromycin susceptibility in emm12, emm75 and emm92 of group A streptococcus. Clin Microbiol Infect. 2014;20(6):516–23. doi: 10.1111/1469-0691.12379 pmid:24118239.
  15. 15. Zheng PX, Chung KT, Chiang-Ni C, Wang SY, Tsai PJ, Chuang WJ, et al. Complete genome sequence of emm1 Streptococcus pyogenes A20, a strain with an intact two-component system, CovRS, isolated from a patient with necrotizing fasciitis. Genome Announc. 2013;1(1). Epub 2013/02/14. doi: 10.1128/genomeA.00149-12 pmid:23405303; PubMed Central PMCID: PMC3569289.
  16. 16. Grissa I, Vergnaud G, Pourcel C. The CRISPRdb database and tools to display CRISPRs and to generate dictionaries of spacers and repeats. BMC Bioinformatics. 2007;8:172–82. Epub 2007/05/25. doi: 10.1186/1471-2105-8-172 pmid:17521438; PubMed Central PMCID: PMC1892036.
  17. 17. Bland C, Ramsey TL, Sabree F, Lowe M, Brown K, Kyrpides NC, et al. CRISPR recognition tool (CRT): a tool for automatic detection of clustered regularly interspaced palindromic repeats. BMC Bioinformatics. 2007;8:209–17. Epub 2007/06/20. doi: 10.1186/1471-2105-8-209 pmid:17577412; PubMed Central PMCID: PMC1924867.
  18. 18. Grissa I, Vergnaud G, Pourcel C. CRISPRFinder: a web tool to identify clustered regularly interspaced short palindromic repeats. Nucleic Acids Res. 2007;35:W52–W7. doi: 10.1093/Nar/Gkm360 pmid:WOS:000255311500012.
  19. 19. Hunter PR, Gaston MA. Numerical index of the discriminatory ability of typing systems: an application of Simpson's index of diversity. J Clin Microbiol. 1988;26(11):2465–6. pmid:3069867; PubMed Central PMCID: PMC266921.
  20. 20. Grundmann H, Hori S, Tanner G. Determining confidence intervals when measuring genetic diversity and the discriminatory abilities of typing methods for microorganisms. J Clin Microbiol. 2001;39(11):4190–2. doi: 10.1128/JCM.39.11.4190–4192.2001 pmid:11682558; PubMed Central PMCID: PMC88515.
  21. 21. Severiano A, Pinto FR, Ramirez M, Carrico JA. Adjusted Wallace coefficient as a measure of congruence between typing methods. J Clin Microbiol. 2011;49(11):3997–4000. doi: 10.1128/JCM.00624-11 pmid:21918028; PubMed Central PMCID: PMC3209087.
  22. 22. Severiano A, Carrico JA, Robinson DA, Ramirez M, Pinto FR. Evaluation of jackknife and bootstrap for defining confidence intervals for pairwise agreement measures. PLoS One. 2011;6(5):e19539. doi: 10.1371/journal.pone.0019539 pmid:21611165; PubMed Central PMCID: PMC3097183.
  23. 23. Hoe N, Nakashima K, Grigsby D, Pan X, Dou SJ, Naidich S, et al. Rapid molecular genetic subtyping of serotype M1 group A Streptococcus strains. Emerg Infect Dis. 1999;5(2):254–63. doi: 10.3201/eid0502.990210 pmid:10221878; PubMed Central PMCID: PMC2640708.
  24. 24. Rho M, Wu YW, Tang H, Doak TG, Ye Y. Diverse CRISPRs evolving in human microbiomes. PLoS Genet. 2012;8(6):e1002441. doi: 10.1371/journal.pgen.1002441 pmid:22719260; PubMed Central PMCID: PMC3374615.
  25. 25. Pride DT, Sun CL, Salzman J, Rao N, Loomer P, Armitage GC, et al. Analysis of streptococcal CRISPRs from human saliva reveals substantial sequence diversity within and between subjects over time. Genome Res. 2011;21(1):126–36. doi: 10.1101/gr.111732.110 pmid:21149389; PubMed Central PMCID: PMC3012920.
  26. 26. Rezzonico F, Smits TH, Duffy B. Diversity, evolution, and functionality of clustered regularly interspaced short palindromic repeat (CRISPR) regions in the fire blight pathogen Erwinia amylovora. Appl Environ Microbiol. 2011;77(11):3819–29. doi: 10.1128/AEM.00177-11 pmid:21460108; PubMed Central PMCID: PMCPMC3127596.
  27. 27. Stern A, Mick E, Tirosh I, Sagy O, Sorek R. CRISPR targeting reveals a reservoir of common phages associated with the human gut microbiome. Genome Res. 2012;22(10):1985–94. doi: 10.1101/gr.138297.112 pmid:22732228; PubMed Central PMCID: PMC3460193.
  28. 28. Mick E, Stern A, Sorek R. Holding a grudge: persisting anti-phage CRISPR immunity in multiple human gut microbiomes. RNA Biol. 2013;10(5):900–6. doi: 10.4161/rna.23929 pmid:23439321; PubMed Central PMCID: PMC3737347.
  29. 29. Davies MR, Holden MT, Coupland P, Chen JH, Venturini C, Barnett TC, et al. Emergence of scarlet fever Streptococcus pyogenes emm12 clones in Hong Kong is associated with toxin acquisition and multidrug resistance. Nat Genet. 2015;47(1):84–7. doi: 10.1038/ng.3147 pmid:25401300.
  30. 30. Nasser W, Beres SB, Olsen RJ, Dean MA, Rice KA, Long SW, et al. Evolutionary pathway to increased virulence and epidemic group A Streptococcus disease derived from 3,615 genome sequences. Proc Natl Acad Sci U S A. 2014;111(17):E1768–76. doi: 10.1073/pnas.1403138111 pmid:24733896; PubMed Central PMCID: PMC4035937.
  31. 31. Plainvert C, Doloy A, Loubinoux J, Lepoutre A, Collobert G, Touak G, et al. Invasive group A streptococcal infections in adults, France (2006–2010). Clin Microbiol Infect. 2012;18(7):702–10. doi: 10.1111/j.1469-0691.2011.03624.x pmid:WOS:000305285200027.
  32. 32. Friaes A, Pinto FR, Silva-Costa C, Ramirez M, Melo-Cristino J. Superantigen gene complement of Streptococcus pyogenes-relationship with other typing methods and short-term stability. Eur J Clin Microbiol Infect Dis. 2013;32(1):115–25. doi: 10.1007/s10096-012-1726-3 pmid:WOS:000313653100016.
  33. 33. Sesto N, Touchon M, Andrade JM, Kondo J, Rocha EP, Arraiano CM, et al. A PNPase dependent CRISPR System in Listeria. PLoS Genet. 2014;10(1):e1004065. doi: 10.1371/journal.pgen.1004065 pmid:24415952; PubMed Central PMCID: PMC3886909.