Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 49 (7), 2594-601

Use of bexB to Detect the Capsule Locus in Haemophilus Influenzae

Affiliations

Use of bexB to Detect the Capsule Locus in Haemophilus Influenzae

Gregg S Davis et al. J Clin Microbiol.

Abstract

Haemophilus influenzae strains are classified as typeable or nontypeable H. influenzae (NTHI) based upon the presence or absence of capsule. In addition to serotyping, which is subject to false-positive results, typeable strains can be identified through the detection of the capsular export gene bexA and one of six capsule-specific genes, but this method is resource intensive, especially in characterizing large numbers of strains. To address these challenges, we developed a bexB-based method to differentiate true NTHI strains from typeable strains. We validated a PCR-based method to detect bexB in 10 strains whose capsule status was well defined. Among 40 strains that were previously serotype positive in clinical microbiology laboratories, 5 lacked bexA, bexB, and capsule type-specific genes by PCR analysis and thus likely represent false-positive serotyping results. Among 94 additional otitis media, commensal, and serotype b-negative invasive strains, 85 were bexA and bexB negative and 9 contained either a complete or partial capsule locus, i.e., 8 were bexA and bexB positive and 1 was bexA negative but bexB positive. Finally, we adapted the method for use in a high-throughput DNA hybridization-based microarray method, which showed 98.75 and 97.5% concordance to the PCR methods for bexA and bexB, respectively. In addition, bexB showed 84% or greater nucleotide identity among strains containing the capsule locus. In this study, we demonstrate that bexB is a reliable proxy for the capsule locus and that its detection provides a simple and reliable method for differentiating strains that lack the entire capsule locus from those containing a partial or complete capsule locus.

Figures

Fig. 1
Fig. 1
Schematic representations of the capsule locus. (A) The capsule locus can be categorized based upon its location in the H. influenzae genome relative to either an IS1016 element or sodC. In one arrangement, the cap locus is flanked by IS1016 insertion element sequences. In the second arrangement, sodC is upstream of the cap locus. (B) The cap locus is divided into three regions, namely, regions I, II, and III; both bexA and bexB are located within region I. Previously described bexA-specific primers produce a 343-bp amplicon (8), and the bexB primers used in this study amplify a 567-bp product. (C) Representative cap locus arrangements. (i) Duplicated IS1016-cap locus harboring a bexA partial deletion in one copy. (ii) Single-copy IS1016-cap arrangement. *, partial bexA deletion precludes annealing with the standard bexA 5′ primer. (D) Representative PCR results. Eagan is a type b strain, containing intact bexA, bexB, and the serotype b-specific cap gene; 86-028NP is an NTHI strain lacking bexA and bexB; strain Rd is a type d strain in which the entire cap region has been deleted; AAr64 is a bexA-negative, type b capsule-deficient variant; and F22477 is a bexA-negative, type f capsule-deficient variant. Invitrogen 1-kb Plus (Invitrogen Corporation, Carlsbad, CA) served as the DNA ladder.
Fig. 2
Fig. 2
Substitutions within bexB primer annealing regions. The top row of each sequence set represents the actual PCR primer used (Table 1); dots represent agreement with the primer sequence. Primers were designed using publicly available type b nucleotide sequences. The bexB.1R and bexB.1F primer set amplified bexB from all six capsular types. There is a nucleotide substitution localized to the 3′ end of the forward primer (bexB.1R) relative to type a bexB, but there were no substitutions in the 3′-most region of either primer.
Fig. 3
Fig. 3
The evolutionary history of bexB was inferred using the maximum parsimony method (7). The 100% consensus tree from the 45 most parsimonious trees (tree length = 148) is shown. The consistency index is 0.965217, the retention index is 0.981651, and the composite index is 0.947507 for all sites and the 98 parsimony-informative sites. The percentages of replicate trees in which the associated taxa clustered together in the bootstrap test (5,000 replicates) are shown next to the branches (10). A total of 666 nucleotides were included in the final data set, and 98 nucleotides were parsimony informative. Phylogenetic analyses were conducted in MEGA4 (41), using all codon positions; there were no gaps in the sequence alignment.
Fig. 4
Fig. 4
Suggested workflow for characterizing H. influenzae strain collections with regard to the capsule locus. Gene names listed within an oval represent detection of that gene via a PCR- or probe-based technique. Identification of “true” NTHI strains can be achieved by screening for the presence of bexB; strains lacking bexB represent true NTHI strains or typeable strains in which the entire capsule locus has been deleted. In the second step, the presence of bexA differentiates potentially serotypeable strains (bexA positive) from capsule-deficient but genetically typeable strains (nonserotypeable, bexA negative, cap locus positive). Among bexB-positive strains, capsular type can be assigned by screening for capsule-specific genes residing within region II of the capsule locus (8). Finally, if necessary, production of a functional capsule can be verified with traditional type-specific serotyping (8).

Similar articles

See all similar articles

Cited by 16 articles

See all "Cited by" articles

Publication types

MeSH terms

Substances

Associated data

LinkOut - more resources

Feedback