OUP user menu

Characterization of Bordetella holmesii isolates from patients with pertussis-like illness in the Netherlands

Frits R. Mooi, Sylvia Bruisten, Ineke Linde, Frans Reubsaet, Kees Heuvelman, Saskia van der Lee, Audrey J. King
DOI: http://dx.doi.org/10.1111/j.1574-695X.2011.00911.x 289-291 First published online: 1 March 2012


Bordetella holmesii is a recently described human pathogen mainly isolated from blood. However, in the US and Canada,B. holmesii has also been cultured from the nasopharynx of patients with pertussis-like symptoms. To the best of our knowledge, respiratory isolates from Europe have not been characterized. Here, we report the isolation and characterization of B. holmesii from Dutch patients with pertussis-like illness. Species determination was confirmed by 16S rRNA gene sequencing and detection by PCR of IS481 and bhoE, a gene not found inBordetella pertussis but present in B. holmesii. Comparative genomic hybridization (CGH) with microarrays revealed that the Dutch isolates formed a cluster distinct from isolates from the US and UK suggesting a distinct population or an epidemiological relationship between the Dutch isolates. All isolates contained a locus involved in iron uptake, previously suggested to originate from B. pertussis. The causes for the apparent increase in the isolation of B. holmesii are discussed.

  • pertussis
  • whooping cough
  • Bordetellaholmesii

Bordetella holmesii is a recently described human pathogen first isolated in 1983 from the blood of a septicemic patient (Weyant et al., 1995). Phylogeny based on 16S rRNA genes placed B. holmesii closest to Bordetella pertussis the causative agent of whooping cough. However, multi-locus sequence typing and cellular fatty acid analysis suggested a more distant evolutionary relationship with B. pertussis and placed B. holmesii closest to Bordetella hinzii, which is only sporadically isolated from humans (Gerlach et al., 2001, 2004; Diavatopoulos et al., 2006). The anomalously high sequence identity of the B. holmesii and B. pertussis 16S rRNA genes has been suggested to be due to lateral transfer of 16S rRNA genes from B. pertussis, together with IS481 and a 66 kb putative pathogenicity island (Diavatopoulos et al., 2006). As the pathogenicity island contained a cluster of eight alcaligin genes, involved in iron uptake, it was designated IUI for Iron Uptake Island (Diavatopoulos et al., 2006). The acquisition of the IUI was suggested to have contributed to the emergence of B. holmesii as a human pathogen (Diavatopoulos et al., 2006). B. holmesii has mostly been associated with septicemia in patients with underlying conditions but has also been isolated from the sputum of patients with respiratory failure suggesting it can cause respiratory disease (Tang et al., 1998). Indeed, B. holmesii has been isolated from the nasopharynx of patients with pertussis-like symptoms in the US and Canada (Yih et al., 1999; Fishman & Jamieson, 2007). In contrast, in a survey involving 11 319 nasopharyngeal swabs from Finnish and Dutch patients, no B. holmesii DNA was detected with PCR, suggesting a low prevalence of B. holmesii in these countries (Antila et al., 2006). Recently, we isolated B. holmesii from nasopharyncheal swabs from three patients with pertussis-like symptoms. As European respiratory isolates of B. holmesii have not been described, they were characterized and compared with American strains. Strain characteristics are given in Supporting Information, Table S1.

In the period June to August 2009, B. holmesii was isolated from nasopharyngeal swabs from three patients (ages 15, 16, and 27 years), with pertussis-like symptoms (Table 1). In all cases, the symptoms were protracted (paroxysmal) coughing, while one patient reported vomiting. The patients lived in the proximity of Amsterdam and the strains were isolated within a period of 3 months. This may suggest that the strains are epidemiologically related. The strains were characterized with PCR and primers used are given in Table S1. PCR for the insertion sequence element IS481, known to be present in both B. holmesii and B. pertussis (Reischl et al., 2001), was positive. The partial 16S rRNA gene sequence of the three isolates was identical to that of the B. holmesii type strain (ATCC 51541), but slightly different from that of the B. pertussis, consistent with previous findings (Diavatopoulos et al., 2006). Further bhoE, a gene not found in the B. pertussis but present in B. holmesii (Diavatopoulos et al., 2006), was detected in all three B. holmesii strains with PCR. Comparative genomic hybridization, using a pan genomic Bordetella microarray (King et al., 2010), with B. holmesii strains from the UK, the US, and the Netherlands, confirmed the presence of the alcaligin genes in all strains (Fig. 1, Table S2). As expected, pertussis toxin genes were absent from the B. holmesii strains (Fig. 1). Clustering based on the CGH data showed that the Dutch isolates formed a separate cluster possibly suggesting the presence of a distinct population (Fig. 1). Alternatively, this may suggest that the strains are epidemiologically related.

Figure 1

Comparative genomic hybridization of Bordetella holmesii isolates. Only a selection of genes is shown. The alc, genes have previously been shown to be present in B. holmesii and were probably acquired from B. pertussis (Diavatopoulos et al., 2006). The genes for pertussis toxin (ptxA-E) are found in B. pertussis, but not B. holmesii. Each column represents one B. holmesii strain. Country of origin, year of isolation, and strain designation are indicated. Gene designations are shown on the right. Red, black, and green indicate increased, similar, or decreased hybridization of the test strain compared with the reference strains. Clustering was based on the total microarray (King et al., 2010).

View this table:
Table 1

Patient data and clinical symptoms

SourceSampling dateAge patient (year)SexSymptoms
Amsterdam regionJune 200916FemaleProtracted, paroxysmal coughing
Amsterdam regionJuly 200927FemaleProtracted, paroxysmal coughing, vomiting
Amsterdam regionAugust 200915MaleProtracted coughing

Both PFGE and multi-locus sequence typing suggest that B. holmesii shows very little genetic diversity (Mazengia et al., 2000; Diavatopoulos et al., 2006). It is tempting to speculate that the low degree of diversity is the result of a population bottle neck caused by the invasion of a new nice, i.e. Homo sapiens. A similar hypothesis has been put forward to explain the monomorphic nature of B. pertussis (Mooi, 2010). In contrast to B. pertussis, relatively little is known about virulence factors produced by B. holmesii. However, the bvg locus, involved in the regulation of virulence genes in B. pertussis, has also been detected in B. holmesii (Gerlach et al., 2004). This is also true for the gene for filamentous hemagglutinin the product of which has been implicated in attachment and modulation of the host immune response (McGuirk & Mills, 2000; Link et al., 2007). In contrast to B. pertussis, but like Bordetella bronchiseptica, B. holmesii is able to cause both respiratory and invasive disease.

It is unclear whether the apparent increased isolation of B. holmesii reflects a higher circulation rate, or increased awareness and improved detection. Acellular vaccines probably confer little or no protection against Bordetella species other than B. pertussis (Khelef et al., 1993; David et al., 2004) and their introduction may have changed the competitive balance between Bordetellae infecting humans in favor of non-B. pertussis Bordetellae such as B. holmesii, Bordetella parapertussis, and B. bronchiseptica. Thus, it is important to monitor the prevalence of all Bordetella species that cause respiratory disease in humans. IS481 is widely used to detect B. pertussis by means of PCR but is also present in B. holmesii. The bho genes, found only in B. holmesii, may be used to distinguish between the two species.

Supporting Information

Additional Supporting Information may be found in the online version of this article:

Table S1. Primers used in this study.

Table S2. Comparative genomic hybridization.

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.


  • Editor: Nicholas Carbonetti


View Abstract