The adhesin pertactin (Prn) is one of the major virulence factors of Bordetella pertussis, the etiological agent of whooping cough. However, a significant prevalence of Prn-deficient (Prn−) B. pertussis was observed in Japan. The Prn− isolate was first discovered in 1997, and 33 (27%) Prn− isolates were identified among 121 B. pertussis isolates collected from 1990 to 2009. Sequence analysis revealed that all the Prn− isolates harbor exclusively the vaccine-type prn1 allele and that loss of Prn expression is caused by 2 different mutations: an 84-bp deletion of the prn signal sequence (prn1ΔSS, n = 24) and an IS481 insertion in prn1 (prn1::IS481, n = 9). The frequency of Prn− isolates, notably those harboring prn1ΔSS, significantly increased since the early 2000s, and Prn− isolates were subsequently found nationwide. Multilocus variable-number tandem repeat analysis (MLVA) revealed that 24 (73%) of 33 Prn− isolates belong to MLVA-186, and 6 and 3 Prn− isolates belong to MLVA-194 and MLVA-226, respectively. The 3 MLVA types are phylogenetically closely related, suggesting that the 2 Prn− clinical strains (harboring prn1ΔSS and prn1::IS481) have clonally expanded in Japan. Growth competition assays in vitro also demonstrated that Prn− isolates have a higher growth potential than the Prn+ back-mutants from which they were derived. Our observations suggested that human host factors (genetic factors and immune status) that select for Prn− strains have arisen and that Prn expression is not essential for fitness under these conditions.

A large pertussis epidemic occurred between 2008 and 2010 in Japan. To investigate epidemic strains, we analyzed 33 Bordetella pertussis isolates from the epidemic period by sequencing virulence-associated genes (fim3, ptxP, ptxA, and prn) and performing multilocus variable-number tandem repeat analysis (MLVA), and compared these results with those of 101 isolates from non-epidemic, earlier and later time periods. DNA sequencing of the fim3 allele revealed that the frequency of fim3B was 4.3%, 12.8%, 30.3%, and 5.1% within isolates in 2002–2004, 2005–2007, 2008–2010, and 2011–2012, respectively. The isolation rate of the fim3B strain therefore temporarily increased during the epidemic period 2008–2010. In contrast, the frequencies of the virulence-associated allelic variants, ptxP3, ptxA1, and prn2, increased with time during overall study period, indicating that these variants were not directly involved in the occurrence of the 2008–2010 epidemic. MLVA genotyping in combination with analysis of allele types showed that the prevalence of an MT27d strain temporarily increased in the epidemic period, and that this strain carried virulence-associated allelic variants (fim3B, ptxP3, ptxA1, and prn2) also identified in recent epidemic strains of Australia, Europe, and the US. Phenotypic analyses revealed that the serotype Fim3 strain was predominant (≥87%) during all the periods studied, and that the frequency of adhesion pertactin (Prn) non-expressing B. pertussis decreased by half in the epidemic period. All MT27d strains expressed Prn and Fim3 proteins, suggesting that B. pertussis MT27d strains expressing Prn and Fim3B have the potential to cause large epidemics worldwide.

We developed a loop-mediated isothermal amplification (LAMP) method to detect Bordetella pertussis infection. This LAMP assay detected B. pertussis with high sensitivity, but not other Bordetella species. Among nasopharyngeal swab samples from subjects with suspected pertussis, LAMP results showed a high level of agreement with results of conventional PCR. This method is a rapid, sensitive, and specific method for diagnosis of B. pertussis infection even in clinical laboratories with no specific equipment.

We describe the epidemiology of a pertussis outbreak in Japan in 2010–2011 and Bordetella holmesii transmission. Six patients were infected; 4 patients were students and a teacher at the same junior high school. Epidemiologic links were found between 5 patients. B. holmesii may have been transmitted from person to person.

Bordetella pertussis is the aetiologic agent of whooping cough, a common cause of severe respiratory illness in children and prolonged mild cough in adults. To understand some of the reasons for differences in clinical symptoms between adults and children, we measured B. pertussis DNA loads in nasopharyngeal swabs (NPS) from 19 adults and 40 children (including 14 infants) by quantitative IS481 real-time PCR. All cases had been pre-diagnosed with the B. pertussis-specific loop-mediated isothermal amplification method. The mean PCR threshold cycles for adult and child NPS were 34.9 and 27.1, respectively, indicating a significantly lower B. pertussis DNA load in adults than in children (p <0.001). Moreover, adults had very low DNA loads during both early and later stages of the disease. When corresponding bacterial loads in NPS were calculated for B. pertussis Tohama cells using a standard curve, the mean number of bacterial cells taken with a rayon-tipped swab from an adult, older child and infant was estimated to be 320 (95% CI 120-910), 2.1 × 10⁴(95% CI 5.3 × 10³ to 8.3 × 10⁴) and 1.1 × 10⁶ cells (95% CI 1.2 × 10⁵ to 8.9 × 10⁶), respectively. This indicates that the B. pertussis load in NPS is closely correlated with patient age. Our observations suggest that adult pertussis is characterized by a lower bacterial load in the nasopharynx, resulting in milder symptoms and negative cultures.

No abstract

A loop-mediated isothermal amplification (LAMP) assay for simple detection of Bordetella holmesii was developed. This assay discriminates between B. holmesii and other Bordetella species and successfully detect B. holmesii DNA in nasopharyngeal swab samples from subjects with suspected pertussis. The LAMP assay results were in complete agreement with the results of previously published real-time PCR assay, indicating that the former is a powerful tool for the accurate diagnosis and surveillance of B. holmesii.

An ELISA that measures anti-PT IgG antibody has been used widely for the serodiagnosis of pertussis; however, the IgG-based ELISA is inadequate for patients during the acute phase of the disease because of the slow response of anti-PT IgG antibodies. To solve this problem, we developed a novel IgM-capture ELISA that measures serum anti-Bordetella pertussis Vag8 IgM levels for the accurate and early diagnosis of pertussis. First, we confirmed that Vag8 was highly expressed in all B. pertussis isolates tested (n = 30), but little or none in other Bordetella species, and that DTaP vaccines did not induce anti-Vag8 IgG antibodies in mice (i.e. the antibody level could be unaffected by the vaccination). To determine the immune response to Vag8 in B. pertussis infection, anti-Vag8 IgM levels were compared between 38 patients (acute phase of pertussis) and 29 healthy individuals using the anti-Vag8 IgM-capture ELISA. The results revealed that the anti-Vag8 IgM levels were significantly higher in the patients compared with the healthy individuals (P < 0.001). ROC analysis also showed that the anti-Vag8 IgM-capture ELISA has higher diagnostic accuracy (AUC, 0.92) than a commercial anti-PT IgG ELISA kit. Moreover, it was shown that anti-Vag8 IgM antibodies were induced earlier than anti-PT IgG antibodies on sequential patients' sera. These data indicate that our novel anti-Vag8 IgM-capture ELISA is a potentially useful tool for making the accurate and early diagnosis of B. pertussis infection.