Group B streptococci (GBS), Streptococcus agalactiae, are a major bacterial pathogen in newborns, pregnant women, the elderly, and immunocompromised patients. GBS infection is the leading cause of bacterial meningitis and sepsis, especially in newborns. The United States Centers for Disease Control and Prevention (CDC) has recommended universal prenatal screening for vaginal and rectal GBS colonization for all pregnant women at 35–37 weeks of gestation, so that infected women can undergo prophylactic antibiotic treatment [1]. GBS is divided into several different subtypes based on capsular polysaccharide (CPS) composition. Recently, some studies have reported that antimicrobial susceptibility results vary according to the GBS subtype [2, 3]. Information on the distribution of GBS subtypes will allow more accurate surveillance and will be essential for the development of protective GBS vaccines [4]. Thus, the detection and subtyping of GBS pathogens is important for the treatment and prevention of GBS infection.

The conventional serotyping test for the detection and serological subtyping of GBS is labor-intensive and time-consuming [5]. Furthermore, only six type-specific agglutinins, against GBS subtypes Ia, Ib, II, III, IV, and V, are routinely available in commercial kits for conventional subtyping; therefore, there are a number of non-typable isolates that go undetected [6]. In order to overcome the disadvantages of the conventional serotyping test, many studies have explored alternative molecular methods for GBS detection and/or subtyping [7–10]. These include the DNA dot blot hybridization method [10], the sequencing method [7], and the multiplex polymerase chain reaction (PCR)-based reverse line blot hybridization method [8, 9]. With the exception of a few studies that used the multiplex PCR-based reverse line blot hybridization method (for the simultaneous detection and subtyping of GBS) [8], most of these studies have been directed toward the subtype classification of clinical isolates that have already been determined as GBS using typical laboratory assays (e.g., characteristic β-hemolysis and colony morphology on blood agar plates [BAP], Gram stain morphology, a negative catalase test, and Lancefield grouping with group-specific antiserum).

In the present study, we developed a rapid genotyping test for simultaneous GBS detection and subtyping, in which PCR and sequencing are performed with the same primers. To verify our rapid genotyping test, we applied both the conventional serotyping test and our rapid genotyping test to the detection and subtyping of 112 clinical GBS isolates. We also investigated the frequency of GBS subtypes, which was determined by the rapid genotyping testing of 318 clinical isolates obtained between March 2008 and April 2009.

Specimens from the lower vagina and rectum were obtained from pregnant women receiving prenatal care at the Seoul National University Bundang Hospital in Korea. Specimens were collected with a cotton swab and were delivered to the laboratory using Amies transport medium without charcoal (KOMED, Seongnam, Korea).

The conventional serotyping test was performed as follows: the swabs were removed from the transport medium and inoculated into LIM broth containing Todd-Hewitt broth powder (0.03 g/mL) with colistin (10 μg/mL) and nalidixic acid (15 μg/mL). The inoculated broth was incubated for 18–24 h at 37°C in ambient air (CO2-free air) and subcultured on sheep BAP for 18–24 h at 37°C in 5% CO2. Typically, a selective broth medium (LIM broth) culture with subsequent subculturing on sheep BAP is the recommended method for GBS detection [1]. GBS was identified by several features, including characteristic β-hemolysis, colony morphology on BAP, Gram stain cell morphology, and a negative catalase test. Testing for the Lancefield group B cell wall antigen using group-specific antisera (Streptex; Remel Inc., Lenexa, KS, USA) was also performed for all isolates to confirm that the identified isolates were GBS. For serological subtyping, pure GBS colonies from the sheep BAP cultures were used to inoculate LIM broth, which was then incubated for 16–20 h at 30°C in ambient air (CO2-free air). Conventional serological subtyping was performed using hemolytic streptococcus group B typing sera (Denka Seiken, Tokyo, Japan). This sera contains six type-specific agglutinins against GBS serological subtypes Ia, Ib, II, III, IV, and V. Subsequent subtyping analysis was performed according to the manufacturer’s instructions.

Our rapid genotyping test was performed as follows: the swabs were removed from the transport medium and inoculated into LIM broth. The inoculated broth was incubated for 18–24 h at 37°C in ambient air (CO2-free air). The LIM broth was centrifuged at 13,000 rpm for 5 min with 500 μL of phosphate-buffered saline (PBS) to obtain a colony pellet. The supernatant was removed, and 100 μL of 5% Chelex (Bio-Rad Laboratories, Hercules, CA, USA) was added to the pellet. After incubation of the pellet with Chelex at 100°C for 20 min, the mixture was centrifuged at 13,000 rpm for 5 min. After centrifugation, the supernatant, which contained the DNA, was removed for DNA amplification and sequencing. We designed ‘GBS primers’ for both the detection and determination of GBS subtypes using the capsular polysaccharide (cps) gene by referring to the GenBank Nucleotide Sequence Database under the accession number AF332908. These primers targeted and amplified type-specific GBS capsular gene sequences approximately 685 bp in length, located between nucleotides 1,495 and 1,892. The forward primer was 5′-GAA TCA CTG GTT TGT GGC-3′ and the reverse primer was 5′-TAA GAC GGT TGA ACT GCT G-3′; both primers were synthesized by Bioneer (Daejeon, Korea). Because GBS primers are unable to distinguish subtype Ia from III-3 and subtype II from III-4, ‘GBS-III primers’ were used to differentiate subtypes III-3 and III-4 from the other subtypes. For the GBS-III primers, we used published oligonucleotide primer pairs targeting partial cps sequences [7]. The DNA extracts were amplified using the GBS primers. The 25-μL PCR reactions contained 2.5 μL of the DNA extract, 0.5 μL of each primer at 0.2 μM, 1 μL of each dNTP at 2.5 mM, 0.625 U of Taq polymerase (Takara Bio Inc., Otsu, Japan), and 2.5 μL of PCR 10× buffer. The mixtures were amplified in a PTC-200 (MJ Research Inc., Waltham, MA, USA) thermal cycler. PCR amplification was performed under the following conditions: denaturation at 95°C for 5 min, and 39 cycles of 95°C for 30 s, 58°C for 30 s, and 72°C for 30 s. Finally, the PCR products were incubated at 72°C for 5 min to allow for DNA extension and complete strand synthesis. The PCR products were visualized by 2.0% agarose gel electrophoresis in Tris-borate-EDTA buffer and with ethidium bromide staining. The visualized PCR products were purified using 2 μL of an ExoSAP-IT purification kit (USB Corp., Cleveland, OH, USA) per 5 μL of the product. The sequencing reaction was performed on 1 μL of the product at a 100-ng/μL concentration using the same ‘GBS primers.’ The cycle sequencing was carried out using a BigDye Terminator v3.1 kit on an ABI Prism 3130xl Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). The generated sequences were compared with the reference sequences of the GBS subtypes. If the colony was identified as subtype Ia/III-3 or subtype II/III-4, the DNA extracts were also amplified using the GBS-III primers.

The results of the subtyping of the 112 isolates identified by conventional serotyping were as follows: subtype Ia, 16 isolates; subtype Ib, 9; subtype II, 7; subtype III, 23; subtype IV, 5; and subtype V, 24. Twenty-eight isolates (25.0%) of these 112 samples were non-typable. In contrast, the rapid genotyping test identified the specific subtypes for all of the 112 isolates. The 23 subtype III isolates identified by the conventional serotyping test were subdivided into subtype III-1 (22 isolates) and subtype III-4 (1 isolate). With the exception of subtype IV isolates in the conventional serotyping test, there was complete agreement between the conventional serotyping test and our rapid genotyping test for the results obtained. Only one out of the five isolates classified as subtype IV by the conventional serotyping test was assigned to subtype IV by the rapid genotyping test. The other four isolates were identified as subtype V by the rapid genotyping test. We amplified the sequencing reference strains of GBS subtypes IV and V using GBS primers and confirmed the proper identification of subtypes IV and V by sequencing the reference strains of these isolates; this sequencing confirmation using reference strains validated our newly designed GBS primers. Therefore, the disagreement between the serological subtyping of the conventional serotyping test and the molecular subtyping of the rapid genotyping test of some isolates of subtype IV may be due to the sensitivity and specificity of the antisera. A recent report for subtype detection by enzyme-linked immunosorbent assays (ELISA) has also demonstrated the possibility of cross-reaction between subtypes IV and V, in agreement with our findings. It is possible that the discrepancies between the results may be due to the sensitivity of the two methods for GBS subtyping [5].

Among the 28 isolates that could not be subtyped using a conventional serotyping test, 9 isolates (32.1%) were assigned to subtype V, which was the most common subtype of non-typable isolates. Several previous studies using molecular methods for GBS subtyping showed similar results (Table 1). These studies have reported that approximately 20–49% of isolates assigned to non-typable isolates by serological subtyping were identified as subtype V using molecular subtyping methods [7–10]. It is believed that these non-typable isolates, especially subtype V, have unique polymorphisms in the cps gene and, thus, show genetic and antigenic diversity [6]. Therefore, we believe that subtype V antigenic diversity may be involved in the cross-reactivity of subtypes IV and V in conventional serotyping tests.

Table 1 Comparison of the results of molecular subtyping for non-typable group B streptococci in conventional serological subtyping
Full size table

Following validation of the GBS primers and the GBS-III primers with the analysis of 112 isolates, we performed the rapid genotyping test for the simultaneous detection and subtyping of 318 clinical GBS isolates obtained between March 2008 and April 2009. The frequencies of GBS subtypes identified by the rapid genotyping test are shown in Table 2. Subtype III was the most commonly observed subtype, followed by subtypes V, Ib, and Ia. The 111 subtype III isolates were subdivided into four subtypes as follows: subtype III-1, 101 isolates (31.8% of total GBS); subtype III-2, 3 isolates (0.9% of total GBS); subtype III-3, 2 isolates (0.6% of total GBS); and subtype III-4, 5 isolates (1.6% of total GBS). Subtype III-1 (91.0% of subtype III) was the most frequently observed subtype. Subtypes IV, VIII, and IX were not observed among these isolates.

Table 2 Comparison of the frequency of molecular subtypes for clinical isolates of group B streptococci
Full size table

Previous studies using conventional serotyping testing to examine the GBS subtypes in Korea have reported that subtypes III, Ib, and Ia are commonly observed in clinical GBS isolates [11, 12]. These studies reported that the frequencies of subtype V were in the range 1.8–8.5%, which is much lower than the frequency of subtype V (27.4%) that was obtained in the present study. As stated above, the cross-reactivity of subtype V in serological subtyping due to antigenic diversity may account for the observed differences in subtype V frequencies between studies. Previous studies using molecular subtyping methods have reported that subtypes III, Ia, and V were commonly observed in clinical GBS isolates (Table 2) [7–10]. In the present study, subtypes III and V accounted for approximately 60% of the clinical GBS isolates. The frequencies of subtypes Ia and Ib were lower than those observed in other studies.

Because some strains do not express a capsule due to inactive gene expression, the subtyping results of genotyping and conventional serotyping tests may be different. This could potentially lead to overtreatment in patients carrying these strains. We believe that our rapid genotyping test will be a useful tool for subtyping of the colonized GBS in pregnant women and for subtyping for the purpose of surveillance for vaccine development.

In conclusion, we have developed a rapid genotyping test for the simultaneous detection and subtyping of GBS by performing PCR amplification and DNA sequencing using the same set of primers. The advantages of using this rapid genotyping test are as follows: a) this method can be used to identify all clinical isolates; and b) it can decrease the amount of time and effort required, as it requires one overnight culture and no serological testing. Furthermore, we studied the frequency of molecular GBS subtypes in Korea using the rapid genotyping test and found that subtypes III and V were commonly observed among the clinical GBS isolates. Our results have provided valuable information that will be helpful in the development of subtype-based GBS vaccines and in the treatment of GBS infection based on the antimicrobial susceptibility of specific subtypes.