Typing of nonencapsulated haemophilus strains by repetitive-element sequence-based PCR using intergenic dyad sequences

Abstract

Intergenic dyad sequences (IDS) are short repeated elements that have been described for several Haemophilus genomes and for only two other bacterial genera. We developed a repetitive-element sequence-based PCR using an IDS-specific primer as a typing method (IDS-PCR) for nonencapsulated Haemophilus strains and compared this technique with pulsed-field gel electrophoresis (PFGE) of DNA restricted with SmaI. IDS-PCR was rapid, easy to perform, and reproducible, with a high discriminatory capacity for nontypeable Haemophilus influenzae (NTHI) strains. The 69 NTHI strains tested generated 65 different banding patterns. Epidemiologically related strains gave similar or identical fingerprints, and all of the unrelated strains except two showed different patterns. These results were in agreement with those obtained by PFGE. For 20 genital strains usually identified as being biotype IV NTHI and belonging to a cryptic genospecies of Haemophilus with remarkable genetic homogeneity, four bands were significantly present and six bands were significantly absent from the fingerprints. The 20 strains were gathered in 11 closely related profiles, whereas PFGE provided no band when DNA was treated with SmaI. IDS-PCR improved the differentiation previously obtained within this species by ribotyping and multilocus enzyme electrophoresis. Our findings suggest that IDS-PCR is a rapid, reliable, and discriminatory method for typing NTHI strains and is currently the most efficient method for distinguishing strains within the cryptic genospecies of HAEMOPHILUS:

FIG. 1.

Banding patterns obtained by IDS-PCR for 18 Haemophilus strains. Lane 1, cryptic genital strain C2; lanes 2 to 8, genital strains G1 and G35 to G40; lanes 9 to 18, nongenital strains A3 to A7 and CH3 to CH7; lane B, reagent blank; lanes M, size markers.

FIG. 2.

Banding patterns obtained by IDS-PCR. Lanes 1 and 2, identical profiles of strains G6 and G7; lanes 3 and 4, identical profiles of strains G22 and G23; lanes 5 to 7, closely related profiles of strains G24, G25, and G26; lanes 8 and 9, identical profiles of strains G29 and G30; lanes B, reagent blank; lanes M; size markers.

FIG. 3.

Banding patterns obtained by IDS-PCR with the cryptic genital strains C2, C3, C7, C8, and C10 (lane 1), C4, C6, and C14 (lane 2), C19 (lane 3), C5 (lane 4), C9, C11, and C12 (lane 5), C13 and C20 (lane 6), C17 (lane 7), C1 (lane 8), C15 (lane 9), C16 (lane 10), and C18 (lane 11). Lane B, reagent blank; lanes M, size markers. The sizes of the two fragments common to the 20 cryptic genital strains are indicated on the right.

FIG. 4.

Reproducibility of IDS-PCR. Shown are banding patterns obtained on two different days with 10 Haemophilus strains. Lanes 1 to 3, cryptic genital strains C4, C11, and C19; lanes 4 to 6, genital strains G24, G30, and G40; lane 7, strain Rd; lanes 8 to 10, nongenital strains A8, CH8, and CH9; lanes B, reagent blank; lanes M, size markers.

FIG. 5.

Banding patterns obtained by PFGE of DNA restricted with SmaI for 13 Haemophilus strains. Lanes 1 to 3, genital strains G1, G35, and G39; lanes 4 to 7, nongenital strains A4, A7, CH3, and CH6; lane 8, cryptic genital strain C7; lanes 9 and 10, identical profiles of strains G29 and G30; lanes 11 to 13, identical profiles of strains G24, G25, and G26.