- Top of page
Despite the effectiveness of the Hib vaccine, multiple amplification of the capb locus contributes to vaccine failure. However, there has been no report on the effect of Hib locus amplification in Japan. We examined 24 Hib strains from Japanese children with invasive diseases due to Hib. Although all strains showed the same capb sequence, Southern blot analysis showed that four strains (16.7%) harbored multiple copies (more than two) of the capb locus. Careful analysis of the locus in circulating Hib strains is necessary now that the Hib vaccine has been introduced into Japan.
Hib occasionally causes invasive bacterial diseases such as meningitis, epiglottitis and sepsis, especially among young children. Hib conjugate vaccines, which consist of capsule polysaccharide conjugated with carrier protein, are very effective and safe. Since the Hib conjugate vaccine was introduced in Europe and America in the 1990s, the incidence of invasive Hib disease has decreased dramatically in many countries (1). However, despite the efficacy of the Hib vaccine, an increased number of cases of the rare invasive Hib diseases (i.e. cases of true vaccine failure) have now been reported in Europe in fully vaccinated children (2–5). Although possibly contributory host factors such as lower avidity of the anti-Hib antibody are known to occur (6, 7), amplification of the capsulation locus may also have contributed to vaccine failure (8, 9).
Type b polysaccharide capsules, polymers of PRP, are cell-surface components that serve as major virulence factors against host defense mechanisms. The genes involved in Hib capsule expression are found within the capb locus, an 18-kb DNA segment of the chromosome (10). Most invasive Hib strains contain a partial duplication of the capb locus which consists of one intact copy of the locus, and a second copy with a 1.2-kb deletion region containing the bexA gene and an IS1016 insertion element that flanks the locus (10). Polysaccharide capsule production relates to the number of copies of the locus (11). Recently, Cerquetti et al. reported that amplification of the capb locus to as many as three to five copies is associated with vaccine failure (8, 9). In addition, Schouls et al. found two variants of the capsular gene cluster, designated type I and type II, which were assessed by considerable sequence divergence in the hcsA and hcsB genes of the capb locus. They found that type I strains carry approximately twice as much capsular polysaccharide on the cell surface as type II strains (12).
In Japan, the Hib conjugate vaccine was licensed in January 2007, and introduced in December 2008; however, the vaccination plan has not yet been fully implemented. Although 55% of bacterial meningitis cases in children in Japan were caused by Hib (13), there has been no national survey of strains isolated from patients with invasive Hib diseases including meningitis. Furthermore, there are no reports on the amplification or sequence divergence of the capb locus. The principle aim of this study was to analyze the number of capb copies, and to assess sequence divergence in the hcsA and hcsB genes of Hib strains isolated from children with Hib diseases in our district before the introduction of the Hib conjugate vaccine.
A total of 24 Hib strains isolated between November 2004 and May 2009 from 24 children with invasive Hib diseases who had not received Hib conjugate vaccine in Kagoshima Prefecture, Japan, were collected and examined. Of these strains, 15 were isolated from CSF and 9 from blood. The strains were epidemiologically unrelated and individually stored at −80°C. All isolates were identified as serotype b by PCR capsular genotyping (14). PFGE was performed using a CHEF-DR 3 apparatus (Nippon Bio-Rad Laboratories, Tokyo, Japan) according to previously reported methodology (15). Briefly, DNA was digested by SmaI and separated on 1% agarose gels by PFGE under the following conditions: current range, 100 to 130 mA at 14°C for 16 hr; initial switch time, 5.3 s, linearly increasing to a final switch time of 49.9 s; angle, 120°; field strength, 6 volts/cm. The gels were stained with ethidium bromide and photographed. A lambda with a size range of 48.5 kb to 1 Mb (BME, Rockland, ME, USA) was used as a size marker. For interpretation of banding patterns separated by PFGE, we referred to the criteria of Tenover et al. (16).
Two variants of the capb locus DNA sequence, type I and type II, were determined by PCR using two primer sets targeting the hcsA gene which could discriminate between the two capsular genotypes as described in a previous report (12). The DNA sequences of the PCR products were determined by an ABI Prism 310 sequencer (Applied Biosystems Japan, Tokyo, Japan).
The number of capb locus copies was detected by Southern blotting analysis according to previously reported methods (8). Because KpnI and SmaI restriction sites flank the capb locus, extracted DNA in an agarose plug was digested with these enzymes, separated by PFGE, and transferred to a nylon membrane. A Hib capsule-specific 480-bp probe was constructed by PCR (14) and labeled with DIG using a DIG high prime DNA labeling kit (Roche Diagnostics, Mannheim, Germany). The membrane was hybridized with the probe and visualized by chemiluminescent detection using a DIG detection kit (Roche Diagnostics). The Kpn I/Sma I fragment of a two copy strain was expected to be 45-kb, because it includes two repeats of the locus (18 + 17 kb) plus additional segments (∼10 kb) upstream and downstream of the cap region (17). Three-, four-, and five-copy fragments showed increased size in 18-kb increments for each additional copy (63, 81, and 99-kb, respectively) (8).
A summary of results is shown in Table 1. The type I-associated hcsA gene was found in all of the strains examined. The DNA sequences of all the PCR products were completely identical. PFGE analysis showed nine distinctive restriction patterns (A to I) among the 24 isolates. Fourteen strains with the A pattern were divided into A1 subtype (13 strains) and the closely-related A2 subtype (one strain). Southern blotting analysis demonstrated that 20 strains showed a two-copy arrangement of the capb locus (45-kb), two strains showed three copies (63-kb), and the other two showed four copies (81-kb) (Fig. 1). The incidence of multiple-copy strains (>two copies) among examined strains was 16.7% (4/24). All of the strains with the dominant PFGE pattern (A1) possessed two copies, while one with the closely-related A2 subtype harbored four copies. The other three strains with multiple copies showed minor PFGE patterns (B, G or I). All the patients infected by strains with multiple copies were treated successfully without neurological or physical sequelae.
Table 1. Sequence type and number of copies of the capb locus of the 21 Haemophilus influenzae strains examined in this study
|No. of cases||No. of strains||Detected date (Year/month)||Age (months)||specimen||disease||Ampicillin susceptibility||PFGE pattern||the capb locus|
|Sequence type||Size of band||No. of copies|
| 1||C1650||2004/11||14||blood||bacteremia||R†||H||I||45 kb||2|
| 2||K4646||2005/7|| 9||blood||meningitis||R||G||I||81 kb||4|
| 3||K5003||2005/11||53||blood||meningitis||S‡||A1||I||45 kb||2|
| 4||K5154||2006/1||17||CSF||meningitis||S||D||I||45 kb||2|
| 5||K5221||2006/1|| 5||CSF||meningitis||S||B||I||45 kb||2|
| 6||K5331||2006/2||24||CSF||meningitis||S||E||I||45 kb||2|
| 7||K5545||2006/4||12||blood||cellulitis||-||A1||I||45 kb||2|
| 8||K5625||2006/5||31||CSF||meningitis||R||F||I||45 kb||2|
| 9||K5905||2006/9||19||CSF||meningitis||S||A1||I||45 kb||2|
|10||K6066||2006/11|| 7||CSF||meningitis||S||B||I||63 kb||3|
|15||K6892||2007/12|| 9||CSF||meningitis||R||A1||I||45 kb||2|
|17||K6934||2008/1|| 2||CSF||meningitis||R||A1||I||45 kb||2|
|19||K7448||2008/7|| 8||CSF||meningitis||S||C||I||45 kb||2|
|20||K7450||2008/7|| 7||CSF||meningitis||S||A1||I||45 kb||2|
|22||K7639||2009/4|| 4||blood||meningitis||S||A2||I||81 kb||4|
|24||K7721||2009/5|| 4||blood||bacteremia||S||I||I||63 kb||3|
Figure 1. Examples of Southern blot analysis of DNA from Haemophilus influenzae type b strains digested with KpnI/SmaI, separated by PFGE, and hybridized with the 480-bp DIG-labeled capb probe. Strain K6066 in lane 2 and strain K7721 in lane 17 showed three-copy arrangement of the capb locus (ca. 63-kb). Strain K4646 in lane 14 and K7639 in lane 15 had four-locus copies (ca. 81-kb). Other strains had two copies (ca. 45-kb).
Download figure to PowerPoint
Amplified capb sequences were detected more frequently among strains from children with true vaccine failure than among those from unvaccinated children (24% vs. 10%) in the United Kingdom (8). Furthermore, the proportion of strains with multiple copies of the capb locus increased over time in Italy (9). Amplification of the capb locus is associated with decreased susceptibility to complement-mediated lysis and decreased complement-mediated opsonization (11). Thus, amplification of the capb locus may result in the overcoming of host defenses and contribute to vaccine failure. We have found that Hib strains with multiple (three or four) copies of the capb locus were present in Japan before the introduction of the Hib conjugate vaccine. The incidence of 16.7% (4/24) of multiple-copy strains found in our study is slightly higher than that found in the UK between 1991 and 1992 before routine immunization was introduced (10.1%, 9/89) (8). In our study, most of the multiple-copy strains showed rare PFGE patterns. Thus these strains might be selected and involved in vaccine failure after the introduction of Hib conjugate vaccination in Japan.
Sequence typing of the capb locus is based on the considerable sequence divergence in the hcsA and hcsB genes, which are involved in the transport of capsular polysaccharides across the outer membrane (18). Schouls et al. have reported that type II strains display less expression of capsular polysaccharide than do type I, and were isolated only during the pre-vaccination era in the Netherlands (12). The greater polysaccharide expression may have provided a selective advantage for type I strains, resulting in the rapid elimination of type II. In addition, there have been remarkable differences in the geographic distribution of type I and type II; with a higher incidence in the United States (73%) than the Netherlands (5%) of type II among Hib strains isolated from patients (12). While we did not find type II strains in this study, more Hib strains should be evaluated to clarify the exact incidence.
To our knowledge, this is the first study to investigate capb locus copy number in invasive Hib strains isolated in Japan. We found that multiple-copy strains were in existence in Japan before the introduction of Hib conjugate vaccine. Molecular epidemiological surveillance of invasive Hib strains after the introduction of vaccines will allow prompt detection of any changes in bacterial properties. In addition, because higher antibody concentrations may be required to protect against Hib disease caused by strains with multiple copies of the capb locus, we strongly recommend the complete implementation of Hib vaccination in young children in Japan.