Few B. pertussis
isolates from China contained nonvaccine type alleles prn2
; those that did were found later. In many countries, the prn1
allele is found in most vaccine strains and predominated before introduction of vaccination. However, the vaccine type strains were gradually replaced by nonvaccine type strains, mainly with allele prn2
, after the introduction of vaccination. The most prevalent type in modern isolates is prn2
). In Taiwan, Pw vaccines have been offered since 1954 (25
). When 168 clinical isolates collected in Taiwan during 1993–2004 were analyzed, prn2
strains accounted for 39% in 1993–1996 and increased to 90% in 1998–2004. In contrast to findings for many countries with long histories of high vaccination coverage, prn2
was first found in China in 2000. The exact reasons for the low frequency of strains with prn2
and their relatively late emergence in China are not known. One explanation might be the low vaccination coverage in China before the 1980s and differences in vaccine coverage between urban and rural areas. Another reason might be the closed borders.
In Japan, divergence of Prn and Ptx between vaccine strains and circulating isolates (26–28
) has been reported. Pw vaccines were introduced in Japan in 1958; the vaccine strain used was prn1
. In 1971, reported vaccination coverage was 50% (8
). In 1976, vaccination coverage dropped to only 9%, and pertussis returned. In 1981, a Pa vaccine containing Ptx and filamentous hemagglutinin was introduced in Japan (27
). The strain used for production of the Pa vaccines was Tohama I (prn1
), isolated in the 1950s. When 107 isolates collected from 1988 through 2001 were studied, the nonvaccine type prn2
appeared in 1994 and was found in 27%–42% of isolates from 1994 through 2001 (26
). A recent study reported similar frequency (33%) of the nonvaccine type prn2/ptxA1
in Japan when 60 isolates collected during 1991–2007 were analyzed (28
TcfA has been shown to be crucial for B. pertussis
). A total of 9 tcfA
alleles have been reported (30,31
), and the most common allele is tcfA2
). Our finding that 94% of isolates studied contained tcfA2
agreed with findings from earlier studies (20,24
). Allele tcfA1
has been described only for reference strain 18323. In our study, 3 clinical isolates recovered during 1953–1958 were found to contain tcfA1
. All 3 isolates were recovered from the same geographic area. Allele tcfA1
contains a 75-bp segment not found in other tcfA
variants. The strain with tcfA1
was postulated to be the progenitor of the strains with tcfA2
Several studies have demonstrated that serotype Fim3 is predominant in vaccinated populations, whereas serotypes Fim2 or Fim2,3 are predominant in unvaccinated populations (14,32,33
). In Sweden, before 1979 when Pw vaccines were first used, 70% of circulating strains were serotype Fim3 (32
). During 1979–1995, when pertussis vaccination was stopped, Fim2 started to increase and reached 64% in the early 1990s. In 1996, when general vaccination with Pa vaccines was reintroduced, prevalence of Fim2 declined and Fim3 strains emerged rapidly. In 2002 and 2003, Fim3 was found in 96% of fully vaccinated persons. In China, before introduction of vaccination, the prevalent serotypes were Fim2,3 and Fim2. After vaccination, the frequency of serotype Fim2,3 decreased and Fim2 became predominant. The possible explanation for the predominance of Fim2 strains after vaccination is that the 2 vaccine strains used in China express Fim2,3 and Fim3. The shift from serotype Fim2 to Fim3 was observed in 1998 and coincided with the introduction of Pa vaccines in this country. Pa vaccines without fimbrial antigens may have some effect on fimbrial serotypes of circulating isolates, as was observed in Sweden (32
); however, the exact reason remains to be shown.
In our study, most strains from China had different PFGE profiles than did those from Europe. However, many PFGE profiles detected among the strains from China fell into the same clusters as those reported in Europe (5,22
). For example, the most common profile, BpCHR15, fell into the same cluster (cluster III) as BpSR23 and BpSR127 (14
). Cluster III includes most strains circulating in Europe before the 1990s (5,32
). In China, strains with BpCHR15 had been prevalent during 1965–1983. Although the strains with BpCHR15 were recovered from several different regions and over 20 years, the possibility that some strains were isolated during outbreaks cannot be excluded.
The major PFGE cluster circulating in Europe since the 1990s is cluster IV (5,32
). Cluster IV can be divided into 3 subgroups (α, β, and γ), the frequency of which differs among countries. However, since the late 1990s in many countries in Europe, subgroup IV-β became more prevalent than the other 2 subgroups (13
). In our study, only 2 isolates with BpSR11 belonged to group IV-β, whereas no isolates fell into subgroups IV-α or IV-γ. This PFGE result correlates with genotyping results.
When we examined the association of 6 common PFGE profiles with different allele combinations, we found that of the 51 isolates with BpCHR15, BpSR23, or BpSR127, 94% contained ptxA1/prn1/tcfA2/fim2–1/fim3–1. Of the 10 isolates with BpCHR6 or BpCHR20, 100% contained ptxA2/prn1/tcfA2/fim2–1/fim3–1. Of the 6 isolates with BpCHR2, all contained ptxA3/prn7/tcfA2/fim2–2/fim3–1.
The emergence of B. pertussis
strains carrying a novel allele (ptxP3
) for the Ptx promoter has been recently observed in countries with long histories of high vaccination coverage, such as the Netherlands (34
). Furthermore, all strains from the Netherlands with BpSR11 were found to carry the ptxP3
allele. In our study, only 2 isolates from China were found to have BpSR11, suggesting that ptxP3
strains are not common in China.
The limitations of this study include the small number of B. pertussis isolates collected during the study period and recent isolates collected mainly from Beijing and its surrounding area. Therefore, our data should be interpreted with caution, and more epidemiologic studies with a larger number of isolates should be conducted in this country.
In conclusion, B. pertussis strains in China may differ from those in countries with long histories of high vaccination coverage. Continuous monitoring of B. pertussis strains is needed.