Pertactin-Negative Variants of Bordetella pertussis in the United States

To the Editor:

Global vaccination of children has dramatically reduced the incidence of illness and deaths from Bordetella pertussis, the causative agent of whooping cough. However, increased numbers of cases of whooping cough have recently been reported in several countries, including the United States. Although much attention has been given to the waning immunity associated with the introduction of acellular vaccines,¹ another factor contributing to the outbreaks may be the adaptation of B. pertussis to vaccine selection pressure. Pertactin is a component of acellular vaccines. Pertactin-negative variants of B. pertussis have recently been reported in clinical isolates from Japan, France, and Finland. The variants from Japan and Finland had deletions or insertion sequences in prn1, an allelic variant of the gene encoding pertactin (prn); the French isolates had deletions or truncations in the prn2 allele.^2,3 Pertactin mutants retain lethality in mouse models of infection and are transmissible in humans.²

Table 1. Table 1. Characterization of B. pertussis Isolates from Philadelphia.

We analyzed the pertactin genes from 12 isolates of B. pertussis cultured from specimens from children hospitalized in Philadelphia during 2011 and 2012 (Table 1). The entire coding region for pertactin was amplified and sequenced.^3,4 Pulsed-field gel electrophoresis (PFGE) was performed, with XbaI used as the restriction enzyme.⁴ The PFGE profiles were determined with the use of the database for U.S. isolates at the Centers for Disease Control and Prevention. Pertactin was detected on Western blotting with the use of anti-69K antiserum (National Institute for Biological Standards and Control no. 97/558), with strain 18323 from the World Health Organization serving as the pertactin-positive reference strain.

On Western blotting, 11 of the 12 B. pertussis isolates were negative for pertactin. Sequencing revealed that in 4 of these isolates the insertion sequence (IS) 481 disrupted the pertactin coding region and that in 7 isolates there was a stop codon truncating the protein. PFGE typing showed that 3 isolates in which IS481 was inserted at nucleotide 1613 were identical. The 7 isolates with the stop codon at amino acid position 425 were identical or closely related (>92% of genomic fragments in common). The pertactin allele in all 12 isolates was prn2; mutations were different from the pertactin-negative prn2 isolates from France.⁵ In the United States, prn2 has been the predominant allele since the 1990s.⁵ These variants were not detected on multilocus sequence typing because the mutations were outside the sequenced region.

To our knowledge, this finding represents the first reported occurrence of pertactin-negative variants of B. pertussis in the United States. Isolates of B. pertussis from geographically distinct U.S. regions should be evaluated to determine whether our finding is a local event or represents a more widespread shift in B. pertussis strains. An understanding of the epidemiology and virulence of pertactin-negative variants is crucial to developing the next generation of pertussis vaccines.

Anne Marie Queenan, Ph.D.
Janssen Research and Development, Raritan, NJ
[email protected]jnj.com

Pamela K. Cassiday, M.S.
Centers for Disease Control and Prevention, Atlanta, GA

Alan Evangelista, Ph.D.
St. Christopher's Hospital for Children, Philadelphia, PA