The number of scarlet fever occurrences reported between 2000 and 2006 fluctuated considerably in central Taiwan and throughout the nation. Isolates of Streptococcus pyogenes were collected from scarlet fever patients in central Taiwan and were characterized by emm sequencing and a standardized pulsed-field gel electrophoresis (PFGE) method. National weekly report data were collected for investigating epidemiological trends.
A total of 23 emm types were identified in 1,218 S. pyogenes isolates. The five most prevalent emm types were emm12 (50.4%), emm4 (23.2%), emm1 (16.4%), emm6 (3.8%) and emm22 (3.0%). PFGE analysis with SmaI suggested that, with a few exceptions, strains with a common emm type belonged to the same clone. There were two large emm12 clones, one with DNA resistant to cleavage by SmaI. Each prevalent emm clone had major PFGE strain(s) and many minor strains. Most of the minor strains emerged in the population and disappeared soon after. Even some major strains remained prevalent for only 2–3 years before declining. The large fluctuation of scarlet fever cases between 2000 and 2006 was associated with the shuffling of six prevalent emm clones. In 2003, the dramatic drop in scarlet fever cases in central Taiwan and throughout the whole country was associated with the occurrence of a severe acute respiratory syndrome (SARS) outbreak that occurred between late-February and mid-June in Taiwan.
The occurrences of scarlet fever in central Taiwan in 2000–2006 were primarily caused by five emm types, which accounted for 96.8% of the isolates collected. Most of the S. pyogenes strains (as defined by PFGE genotypes) emerged and lasted for only a few years. The fluctuation in the number of scarlet fever cases during the seven years can be primarily attributed to the shuffling of six prevalent emm clones and to the SARS outbreak in 2003.
Streptococcus pyogenes (Group A streptococcus) is a common pathogen responsible for a number of human suppurative infections, including pharyngitis, impetigo, pyoderma, erysipelas, cellulitis, necrotizing fasciitis, toxic streptococcal syndrome, scarlet fever, septicemia, pneumonia and meningitis. It also causes non-suppurative sequelae, including acute rheumatic fever, acute glomerulonephritis and acute arthritis . Scarlet fever, characterized by a sore throat, skin rash and strawberry tongue, is most prevalent in school children aged four to seven years old. This disease was listed as a notifiable disease in Taiwan until 2007; as such, all cases of scarlet fever had to be reported to the public heath department. According to our records, however, only 9% of the medical centers, regional hospitals and district hospitals in central Taiwan reported cases of scarlet fever to the health authorities between 1996 and 1999. The number of scarlet fever cases is therefore likely to be significantly underreported. Scarlet fever outbreaks frequently occur in young children at day-care centers, kindergartens and elementary schools [2,3] and also occur in adults upon exposure to contaminated food .
Genotyping bacterial isolates with various methods is frequently used to compare the genetic relatedness of bacterial strains and provides useful information for epidemiological studies. In a previous study, we used emm (gene of M protein) sequencing , vir typing  and pulsed-field gel electrophoresis (PFGE) typing to analyze a collection of streptococcal isolates from scarlet fever patients and used these data to build a DNA fingerprint and emm sequence database for long-term disease surveillance . Vir typing has since been abandoned in our lab because it has lower discriminatory power than PFGE and the protocol is difficult to standardize with conventional agarose gel electrophoresis. In contrast, the PFGE protocol for S. pyogenes has been standardized in our laboratory, and a second enzyme, SgrAI, has been found to replace SmaI for analysis of strains with DNA resistant to SmaI digestion . Since PFGE is highly discriminative and emm sequencing provides unambiguous sequence information regarding emm type, we adopted these two genotyping methods to characterize streptococcal isolates and build a Streptococcus pyogenes DNA fingerprint and sequence database for the long-term study of scarlet fever and other streptococcal diseases.
The number of scarlet fever cases in central Taiwan fluctuated greatly between 2000 and 2006. Relative to the number of scarlet fever occurrences in 2000, occurrences increased in 2001 and doubled in 2002, but dramatically dropped in 2003. The number of occurrences increased again since 2004. In this study, we characterized 1,218 isolates collected between 2000–2006 by emm sequencing and PFGE. The bacterial genotyping data and the epidemiological data collected via the Notifiable Disease Reporting System (established by Taiwan Centers for Disease Control (Taiwan CDC)) were used to examine the significant fluctuation in the number of scarlet fever cases between 2000 and 2006.
Epidemiological trend of scarlet fever
Taiwan is an island country populated by 22.9 million people, most of whom reside in the western region (Figure 1A). The population in northern, central, southern, and eastern areas is 10.2, 5.7, 6.4 and 0.6 million, respectively. Nationwide information for all notifiable diseases has been systematically collected since 2000. For accurate analysis, the number of confirmed scarlet fever cases was adjusted by multiplying the number of reported cases and the specimen positive rate. The total, adjusted number of confirmed cases throughout the whole country increased from 716 cases in 2000 to 1,258 in 2002, but dramatically dropped to 771 in 2003 (Table 1). This number increased again in 2004 and, in 2005, reached the high levels seen in 2002. However, the number of cases slightly declined again in 2006. In central Taiwan, the epidemiological trend was similar to the national profile, but fluctuated more dramatically between 2000 and 2004. While the number of scarlet fever cases was 142 in 2000, this number doubled in 2002 but then dropped in 2003 to the levels seen in 2000 (Table 1). The number of cases increased again in 2004 and, in 2006, reached the levels seen in 2002. The number of cases in 2006 was greater than that in 2005 and differed from the national trend. The number of cases in central Taiwan accounted for 18% to 24% of cases throughout the whole country.
Figure 1. (A) Map of Taiwan and population density (B) National weekly reported cases of scarlet fever between 2000 and 2006. The total average throughout 2000–2006 is indicated by a red dashed line. The two stages of the SARS epidemic in 2003 are marked by blue (stage I) and red (stage II) bars.
Table 1. Reported and adjusted confirmed scarlet fever cases in the whole country and in central Taiwan from 2000 to 2006.
The profiles of weekly reported cases revealed that scarlet fever was more prevalent in the winter and spring seasons (2nd – 25th weeks) in 2000–2006. However, there was a remarkable decrease in the number of cases in the 6th and 7th weeks (Figure 1B). This decrease may be due to the long holiday of the traditional lunar New Year and winter break from school, as it is usually from late-January to mid-February (4th – 7th weeks). The weekly reported number of scarlet fever cases in 2002 was mostly higher than the weekly average from 2000 to 2006 (Figure 1B). In 2003, except in the 11th week, the number of weekly reported cases in the first 16 weeks was greater than the average. Furthermore, the number of cases between the 4th and 9th weeks was even higher than that in 2002. After the 16th week, the number of cases in 2003 was below the overall average and was significantly decreased from the 17th to 24th week (mid-April to mid-June). A lower level of reported cases lasted until the first half of year 2004. In early 2003, a severe acute respiratory syndrome (SARS) outbreak occurred in Taiwan. There were two stages for the SARS epidemic: stage I occurred from late-February to mid-April (9th – 16th week), with scattered sporadic cases, and stage II occurred between mid-April and mid-June (17th – 24th week), with severe nosocomial infections in several hospitals. The dramatic decline of scarlet fever notifications in 2003 occurred during the stage II period of the SARS epidemic.
Distribution of emm types among isolates collected in central Taiwan
For each year between 2000 and 2006, 115 to 273 isolates were collected for genotyping in central Taiwan (Table 1). A total of 1,218 isolates were characterized to investigate the distribution of emm types. In total, 23 emm types were identified in the isolates. The five most prevalent emm types, accounting for 96.8% of the collection, were emm12 (50.4%), emm4 (23.2%), emm1 (16.4%), emm6 (3.8%) and emm22 (3.0%) (Table 2). emm12 was the predominant type found between 2000–2001, accounting for 87.1% and 57.1% of the total isolates in 2000 and 2001, respectively. It became the predominant type again in 2005 and 2006, accounting for 69.3% of the isolates in 2006. emm1 was predominant in 2002, emm4 was most prevalent in 2003 and 2004, and emm6 emerged in 2001 but was not detected again after 2003.
Table 2. Distribution of emm types in Streptococcus pyogenes isolates collected in central Taiwan from 2000 to 2006
PFGE and emm genotypes
The 1,218 S. pyogenes isolates were analyzed by PFGE with SmaI to investigate the clonal relationship among the isolates. There were 127 isolates with DNA resistant to SmaI digestion, and their pattern (with only one DNA band) was referred to as a SPYS16.0026 PFGE-SmaI type. The 127 isolates with the SPYS16.0026 genotype were further analyzed by digestion with SgrAI. The genetic relatedness of the bacterial strains was evaluated by the levels of similarity among the PFGE-SmaI patterns. A dendrogram was constructed using the Unweighted Pair Group Method with Arithmatic mean (UPGMA) algorithm. The dendrogram revealed that all of the emm4 and emm6 isolates, as well as the majority of emm1 and emm22 isolates, were each distributed in a unique cluster. However, the emm12 isolates were located in two distinct clusters and two singletons (Figure 2). One of these clusters included 125 emm12 isolates that were resistant to SmaI digestion. Clustering analysis indicated that isolates with a common emm type were, in general, more closely related than those with different emm types. However, there were a few exceptions. Two strains with different emm types (emm101 and st5282) had indistinguishable PFGE-SmaI patterns, and a strain with a stIL103 type was located within the emm1 cluster (Figure 2). stIL103 is an allele of emm1 that lacks the codons encoding the mature M1 7–24 residues (http://www.cdc.gov/ncidod/biotech/strep/strepindex.htm webcite; accessed on April 20th, 2009). Sequence analysis suggests that the st5282 strain could be derived from an emm101 strain via emm gene recombination, as the sequence for the first 26 codons of the st5282 gene were identical to that for the emm101 gene. Two emm12 and one emm22 isolates were distant from the major emm12 and emm22 clusters (Figure 2). The 127 SmaI-resistant isolates were identified to be of emm12, emm1 or emm58 type.
Figure 2. Dendrogram constructed with PFGE-SmaI patterns, with their corresponding emm types and number of isolates obtained between 2000 and 2006. The clustering analysis was performed with BioNumerics using the UPGMA algorithm and the value of Dice predicted similarity of two patterns at settings of 1% optimization and 0.7% position tolerance.
In total, 94 emm:PFGE-SmaI genotypes were identified in the 1,218 isolates. Eight major emm:PFGE genotypes, emm1:SPYS16.0022 (14.9%), emm4:SPYS16.0006 (11.7%), emm4:SPYS16.0008 (8.1%), emm4:SPYS16.0083 (2.6%), emm6:SPYS16.0020 (2.7%), emm12:SPYS16.0013 (29.6%), emm12:SPYS16.0026 (10.3%) and emm12:SPYS16.0087 (2.3%), made up 82.2% of the 1,218 isolates. Five of the major emm:PFGE genotypes were detected throughout the seven years studied. In contrast, most emm:PFGE genotypes lasted for only 1–2 years; they emerged in the population and quickly disappeared.
The 127 SmaI-resistant isolates were discriminated by PFGE with SgrAI into 14 emm12:PFGE-SgrAI, 1 emm1:PFGE and 1 emm58:PFGE types. The 125 emm12 isolates were distributed in two distinct clusters, A and B (Figure 3). Strains within cluster A were quite divergent, with the most divergent types sharing only 65% pattern similarity.
Figure 3. Dendrogram constructed with PFGE-SgrAI patterns, with their corresponding emm types and number of isolates. DNA from these isolates was resistant to SmaI digestion. The clustering analysis was performed with BioNumerics using the UPGMA algorithm and the value of Dice predicted similarity of two patterns at settings of 1% optimization and 0.7% position tolerance.
Distribution of prevalent emm clones over time
In this study, a cluster of strains (as defined by PFGE types) having a common emm type and sharing higher PFGE pattern similarity than others with different emm types were considered to belong to a common emm clone. The stIL103 strain is an exception to this, as it shared high PFGE pattern similarity with the cluster of emm1 strains and was therefore considered to be part of the emm1 clone. Based on the groupings made by the PFGE patterns, six major emm (emm1, emm4, emm6, emm12, emm12* and emm22) clones were identified and are shown in Figure 2. The emm12* clone represents the emm12 strains with DNA resistant to SmaI digestion. The six major emm clones made up 96.5% of the 1,218 isolates. The adjusted number of the annual confirmed cases of scarlet fever in central Taiwan ranged from 142 to 282 between 2000 and 2006 (Table 1), and 115 to 273 isolates were collected each year for genotyping. The number of isolates genotyped was adjusted to the number of annual confirmed cases to investigate the association of the fluctuation of scarlet fever cases and the relative prevalence of the emm clones. As shown in Figure 4, the emm12* and emm12 clones were the most prevalent in 2000. The two clones declined over time and were at their lowest levels in 2003. The emm1 clone was the most prevalent in 2002 and the emm4 clone was predominant in 2003 and 2004. In 2001, although the number of emm12* and emm12 clones declined, the number of emm1 clones increased significantly. The total number of scarlet fever cases in 2002 was doubled that in 2000 and were primarily attributed to an increase in the number of the emm1, emm4 and emm6 clones. The number of cases in 2003 was considerably lower than that in 2002, likely due to a decline in all major clones except for emm4. The number of cases increased significantly again in 2005, and this increase is associated with a dramatic rise in the prevalence of the emm12 clone.
Figure 4. Distribution of emm clones between 2000 and 2006. The number of Streptococcus pyogenes isolates analyzed is adjusted according to the number of adjusted annual confirmed of cases.
The cases of scarlet fever in central Taiwan from 2000 to 2006 were caused by S. pyogenes strains with a limited number of emm types (Table 2). In fact, five prevalent emm types represented 96.8% of the isolates causing scarlet fever during this time period. Of the 23 emm types isolated, 17 made up 99.4% of the isolates. These 17 types were among the 30 most common emm types that caused invasive streptococcal infections in the United States between 2000 and 2004. Twelve of these types accounted for 75.5% of the isolates characterized and were included in the proposed 26-valent vaccine (emm types 1, 1.2, 2, 3, 5, 6, 11, 12, 14, 18, 19, 22, 24, 28, 29, 33, 43, 59, 75, 76, 77, 89, 92, 94, 101, and 114) .
In our previous work on 179 S. pyogenes isolates collected in central Taiwan between 1996 and 1999, the five most common emm types in central Taiwan remained the same, but the frequency changed in the two time periods, 1996–1999 and 2000–2006 . However, the prevalence and distribution of emm types could have geographic variation. Yan et al.  analyzed 77 S. pyogenes isolates collected from scarlet fever patients between 1993 and 2002 in southern Taiwan and found only three emm types among the isolates, with emm1 being the most prevalent type. Chen and colleagues characterized 830 isolates collected between 2001 and 2002 in northern Taiwan and found that the most frequent emm types were emm1 (29.2%), emm4 (24.1%), emm12 (19.0%), emm6 (15.8%), stIL103 (5.7%) and emm22 (1.9%) . In our study, the most common emm types in 427 isolates collected in the same time period in central Taiwan were emm12 (35.6%), emm1 (34.2%), emm4 (18.5%), emm6 (7.5%) and emm11 (0.9%). stIL103 was present in northern Taiwan, but it was not found in the central region during the same time period. Thus, the distribution and frequency of emm types appear to be geographically varied even in such a small country. The geographic variation in the prevalence of emm clones may explain why the epidemiological trend of scarlet fever in 2006 in central Taiwan was different from that in the whole country.
The major emm types were further discriminated into a number of PFGE types, and clustering analysis of the PFGE patterns suggests that the emm1, emm6 and emm4 strains belong to a single clone. The emm12 strains belong to two major clones and two singletons, and emm22 strains belong to one major clone and one singleton (Figure 2). Thus, six emm clones caused most (96.5%) of the scarlet fever cases in central Taiwan during the seven year time period. The fluctuation of scarlet fever cases was associated with the shuffling of the prevalent emm clones (Figure 4). The finding that only a few prevalent M (emm) types caused most occurrences of scarlet fever in a specific location in a given year period, as well as the shuffling of predominant M types, has been observed in many epidemiological studies in the early 20th century . During major epidemics of streptococcal infections in previous years, only a few serotypes predominated, and the strains were rich in M protein, encapsulated and were highly virulent . Type-specific immunity was important for preventing re-infection with the same M type. It is thought that the shuffling of predominant M types is due to the type-specific immunity, leading to the decline of infections with certain M types and the emergence of other virulent M types. In the present study, the prevalence of the emm12*, emm1 and emm6 clones both increased and decreased within one year. In contrast, the emm12 and emm4 clones persisted throughout the seven year period. This phenomenon may be due to the fact that the emm12 and emm4 clones produced less M protein and were less virulent than the emm12*, emm1 and emm6 clones.
The PFGE study also indicates that each of the six emm clones has one predominant PFGE type, except for the emm4 clone, which has two major PFGE types (Figure 2). The less prevalent PFGE genotypes of each emm clone emerged and quickly disappeared. Even some major PFGE genotypes, such as SPYS16.0026 of the emm12* clone, SPYS16.0020 of the emm6 clone and SPYS16.0022 of the emm1 clone, remained prevalent for only 2–3 years before declining. However, the SPYS16.0013 genotype of the emm12 clone did not follow this trend, as it was prevalent throughout 2000–2006 and was most prevalent in 2006. If a newly emerging strain can only prosper in a specific location for a few years, then the emm12:SPYS16.0013 strains isolated during two different time periods should be different. These differences may not be detectable by PFGE analysis. Whether bacterial isolates that prevail for two periods become genetically diversified is an interesting subject and may be studied by other genotyping methods, such as single nucleotide polymorphism, by virulence gene detection and by antimicrobial susceptibility testing.
The SPYS16.0026 isolates, with DNA resistant to SmaI digestion, were found in three emm types, suggesting that they have multiple evolutionary origins. Of the 127 SPYS16.0026 isolates, 125 belonged to the emm12 type. The first isolate resistant to SmaI digestion was identified in central Taiwan in 1998 and was an emm33 type. The emm12:SPYS16.0026 strain was detected for the first time in 1999 . Our previous studies indicated that the emm12:SPYS16.0026 strain is most likely derived from an emm12:SPYS16.0013 strain by an insertion of a large DNA fragment into the genome . The large DNA segment could have carried the gene(s) responsible for DNA methylation and resistance to cleavage by SmaI. These strains were analyzed with SgrAI. Clustering analysis of the PFGE-SgrAI patterns revealed diverse genetic relationships among the emm12:SPYS16.0026 strains (Figure 3). The high genetic divergence suggests that the emm12:SPYS16.0026 strains have derived from multiple origins. Recently, Euler et al.  have shown that resistance to SmaI cleavage is due to the presence of a DNA methyltransferase gene, which is carried on a mobile chimeric element that has transposon- and bacteriophage-like characteristics. This mobile element may explain the high genetic diversity among the SmaI-resistant strains that emerged in such short period of time.
The fluctuation of scarlet fever cases between 2000 and 2006 may be partially explained by the shuffling of several prevalent emm clones. However, the dramatic drop in reported cases in 2003 is difficult to explain. In early 2003, Taiwan was badly hit by a severe SARS outbreak. The SARS epidemic in Taiwan had two distinct stages, with the beginning in the late-February (the 9th week) and the second ending in mid-June . The stage I epidemic occurred from late-February to mid-April (the 9th to 16th week) and consisted of only scattered, sporadic cases, with most of the patients having recently traveled to China. In this stage, the disease did not cause much panic and the level of scarlet fever remained high. In stage II (from mid-April to mid-June or the 17th to 24th week), several clusters of infection occurred via intra-hospital or inter-hospital transmission. Enormous panic spread over the whole country after an outbreak of nosocomial infection was confirmed on the 22nd of April. The disease was subsequently transmitted to several hospitals and spread from the North to the South. The number of scarlet fever cases dropped remarkably during this period. Because a large portion of the SARS infections was associated with hospitals, fear of SARS drove people out of hospitals and public places. This fear and the change of people's behavior may have significantly reduced the number of outpatients and the transmission of many infectious diseases, including scarlet fever. In fact, the SARS outbreak had a long-term effect on the occurrences of scarlet fever. After the SARS epidemic, the number of weekly scarlet fever reports was often lower than the overall average until the first half of 2004.
The occurrences of scarlet fever in central Taiwan between 2000 and 2006 were primarily caused by six emm clones: emm12 (40.0%), emm4 (23.2%), emm1 (16.3%), SmaI-resistant emm12* (10.3%), emm6 (3.8%) and emm22 (2.9%). Each emm clone had predominant PFGE genotype(s), and most minor genotypes within an emm clone emerged and quickly disappeared. The large fluctuation in the number of scarlet fever cases during this time period can be attributed to the shuffling of several prevalent emm clones and to a SARS outbreak in 2003.
Epidemiological data and bacterial strains
Scarlet fever was a notifiable disease in Taiwan until 2007; hospitals and clinics were obligated to report confirmed or suspected cases to the county public health department via a web-based Notifiable Diseases Reporting System established by the Taiwan CDC in 2000. The hospitals and clinics that reported scarlet fever cases were asked to provide throat swab specimens or S. pyogenes isolates to the regional laboratories of the Taiwan CDC for bacterial examination and genotyping. Confirmed cases were those in which S. pyogenes was isolated from the specimens. The number of annual confirmed cases detected through the Notifiable Diseases Reporting System was adjusted by multiplying the number of reported cases and the rate of positive specimens. S. pyogenes isolates used for characterization in this study were obtained directly from hospitals located in central Taiwan through the Notifiable Diseases Reporting System or were recovered from throat swab specimens collected from hospitals and clinics through the Notifiable Diseases Reporting System and the Sentinel Physician Active Reporting System.
The procedure developed by Beall and colleagues  was used to prepare the emm DNA fragments from S. pyogenes isolates for sequencing. The amplified DNA amplicons and primer 1, 5'-TATT(C/G)GCTTAGAAAATTAA-3', were sent to a local biotech company (Mission Biotech Corp. Taipei, Taiwan) for DNA sequencing. The 5' emm sequences (at least the first 240 bases) were subjected to a BLAST comparison with those in the emm database (http://www.cdc.gov/ncidod/biotech/strep/strepindex.htm webcite; accessed on April 20th, 2009) to determine emm type.
S. pyogenes isolates were subjected to PFGE analysis using a previously described protocol . All of the isolates were analyzed by SmaI digestion. Isolates with DNA resistant to SmaI digestion were analyzed with SgrAI. PFGE patterns were recorded using a Kodak digital camera system (Kodak Electrophoresis Documentation and Analysis System 290; Kodak; Rochester, NY, USA) with 1792 × 1200 pixels. The digital PFGE images were then analyzed using BioNumerics software version 4.5 (Applied Maths, Kortrijik, Belgium) and the DNA pattern for each isolate was compared using the computer software. A unique PFGE pattern (genotype) was defined if it contained one or more DNA bands different from the others. The genetic relatedness among isolates is presented in a dendrogram built by clustering the PFGE patterns. The clustering analysis was performed using the UPGMA algorithm provided in the BioNumerics software and the value of Dice predicted similarity of two patterns at settings of 1% optimization and 0.7% position tolerance.
CS Chiou initiated and managed the project, analyzed data and wrote the manuscript. YW Wang worked on emm sequencing, PFGE analysis and data analysis. PL Chen collected and analyzed epidemiological data from the Notifiable Diseases Reporting System. WL Wang worked on PFGE analysis. PF Wu coordinated the laboratory and disease surveillance sectors in Taiwan CDC. HL Wei helped with identification of emm types. All authors have read and approved the final manuscript.
This work was supported by grants DOH94-DC-2025 and DOH94-DC-2026 from the Centers for Disease Control, DOH, Taiwan.
Hsueh PR, Teng LJ, Lee PI, Yang PC, Huang LM, Chang SC, Lee CY, Luh KT: Outbreak of scarlet fever at a hospital day care centre: analysis of strain relatedness with phenotypic and genotypic characteristics.
O'Loughlin RE, Roberson A, Cieslak PR, Lynfield R, Gershman K, Craig A, Albanese BA, Farley MM, Barrett NL, Spina NL, et al.: The epidemiology of invasive group A streptococcal infection and potential vaccine implications: United States, 2000–2004.
Indian J Med Res 2002, 115:215-241. PubMed Abstract