Background

Haemophilus influenzae (H. influenzae) is a Gram-negative coccobacillus that can cause invasive and severe infections, such as bacteria pneumonia, epiglottitis, sinusitis, and similar conditions, in both adults and children, especially in children under 5-years-old [1, 2]. There are two main types of H. influenzae based on the capsule: encapsulated H. influenzae and nontypeable H. influenzae (NTHi). The NTHi cannot be typed by the serum agglutination test. Serotype b H. influenzae (Hib) used to be the main type of H. influenzae in respiratory tract infections, but the epidemiology of invasive H. influenzae disease has changed substantially, with most invasive diseases now caused by non-b serotypes and NTHi strains [3,4,5,6,7,8]. Different types of H. influenzae have different pathogeneses and require different treatments [9, 10]. Detection of H. influenzae by traditional methods like bacterial culture is cheap and facilitates the detection of antibiotic resistance. However, it is time-consuming and may give false-negative results during ongoing antibiotic treatment. Serum agglutination testing was widely used to determine the serotype of encapsulated H. influenzae, and it can also be used to discriminate the encapsulated H. influenzae from NTHi. However, sometimes there are bacterial self-coagulation phenomena and cross agglutination, and serological reagents are more expensive.

Because the specificity of PCR depends on the selection of target genes, it is challenging to design a specific PCR test for the detection of H. influenzae and discriminate encapsulated H. influenzae from NTHi. The gene ompP6, which encodes the outer membrane protein P6, is present in all the H. influenzae strains and has a highly conserved nucleotide sequence. The nucleotide sequence homology of ompP6 was 96.7–100%. OmpP6 is a good target gene to distinguish H. influenzae from other respiratory tract infection bacteria [11,12,13]. The gene bexA encoding the capsule transport protein exists only in encapsulated H. influenzae, and the nucleotide sequence at site 3552–3865 is highly conserved [14,15,16,17]. According to the distribution and sequence of ompP6 and bexA, we designed two pairs of primers to establish a one-step multiplex PCR assay for differential detection of clinical H. influenzae species from nasopharyngeal secretion (NPS) in patients with respiratory tract infections in order to improve the detection rate and identify the predominant H. influenzae strain in Zhejiang Province in China.

Methods

Subjects

This work has obtained the approval from the Research Ethics Committee of Hangzhou Medical College. The approval numbers are 2012–001.The nasopharyngeal secretion (NPS) used in this study have been allowed to collect from patients with respiratory tract infections. Written informed consents have been obtained from the participants.

One thousand five hundred fourteen NPSs were gathered from patients diagnosed with respiratory tract infection disease and treated in the Yiwu Women and Children’s Hospital and the Shaoxing City Keqiao District Hospital of Traditional Chinese Medicine from 2012 to 2016. They all diagnosed with respiratory tract infection, including 289 sinusitis, 108 bronchitis, 83 bronchopneumonia, 262 tonsilitis, 53 HFMD, 133 herpangina, 80 stomatitis, 80 submaxillary lymphadenitis and 426 other respiratory tract diseases. The patients ranged from 0 to 14 years old, and included 825 boys and 689 girls. One thousand fifty-five of them were 5 or under 5 years old, 459 of them were 6 or elder than 6. No antibiotics were used in the 72 h preceding admission. Under nasal endoscopy, nasal meatus secretions were obtained by cotton swabs from these patients, placed in sterile tubes, and immediately sent to the laboratory for pathogen isolation. The same amount of pH 7.6 1% trypsin was added to the samples and incubated for 90 min at 37 °C to homogenize the NPSs, and 200 μL of the homogenized sample was used for bacterial culture. The remains were kept at − 80 °C.

Bacterial culture and identification

Here, 100 μL of each homogenized sample was inoculated on blood plates and chocolate plates under sterile conditions by zoning streak and incubated in 5% CO2 incubators at 37 °C for 24–48 h. The H. influenzae was screened preliminarily by colony morphology of and the characteristics of Gram staining, and then identified by V factor, X factor and V + X factors requirement test. NTHi ATCC49247, encapsulated H. influenzae standard strain ATCC 9327 (Type a), ATCC 9334 (Type b), ATCC 9007 (Type c), ATCC 9332 (Type d), ATCC 8142 (Type e) and ATCC 9833 (Type f), Streptococcus pneumoniae ATCC 49619, Staphylococcus aureus ATCC 29213, Klebsiella pneumoniae ATCC 700603, Pseudomonas aeruginosa ATCC 27853, E. coli ATCC 25922 were donated by department of pathogenic biology, Zhejiang University. The standard H. influenzae strains were cultured with selective medium and other standard strains were cultured on blood plates, in 5% CO2 incubators at 37 °C for 24–48 h.

Serum agglutination test

The antisera of serotypes a, b, c, d, e, and f of H. influenzae were used to determine the serotype of H. influenzae, which we cultured from clinical samples using a slide agglutination test. Serotyping reagents were purchased from TBC in the United States (batch number: IM-H1015).

DNA extraction

The fresh cultures of each standard strain were centrifuged after saline washing, and genomic DNA was extracted by using a Bacterial Genomic DNA Extraction Kit (GENEray). DNA from homogenized samples was also extracted using the same kit. The concentration and purity of DNA were determined by UV spectrophotometry.

One-step multiple PCR assay design

In this PCR assay, we designed two sets of primers of ompP6 and bexA according to the nucleotide sequence of H. influenzae type b reported on GenBank (Table 1). The expected ompP6 amplified product was the whole 459 bp sequence of ompP6 and the expected bexA amplified product was about 343 bp, only the conserved region at 3552–3895 site of bexA gene. All primers were purchased from Thermo Fisher Scientific, Inc. (China). The DNAs of NTHi standard strain ATCC49247, encapsulated H. influenzae standard strain ATCC 9327 (Type a), ATCC 9334 (Type b), ATCC 9007 (Type c), ATCC 9332 (Type d), ATCC 8142 (Type e), and ATCC 9833 (Type f), Streptococcus pneumoniae ATCC 49619, Staphylococcus aureus ATCC 29213, Klebsiella pneumoniae ATCC 700603, Pseudomonas aeruginosa ATCC 27853, and E. coli ATCC 25922 were used to develop the PCR assay. The PCR amplification was performed in 50 μL reactions containing: 2.5 mol/L of dNTP, 250 mol/L of each primer, 2.5 U of Ex-Taq DNA polymerase (Takara), 2 μL target DNA and 1 × PCR buffer. The PCR cycler was set to the following program: 4 min of enzyme activation at 94 °C, followed 35 cycles at 94 °C for 30s, and 54 °C for 30s and 72 °C for 45 s, followed by 1 cycle at 72 °C for 5 in. The PCR product was examined by 2% agarose electrophoresis.

Table 1 One-step multiple PCR primers
Full size table

Repeatability and specificity of the PCR assay

The PCR were repeated three times under the same conditions. The PCR amplification of ompP6 and bexA of standard strains were purified using a PCR product purification kit and then sequenced by Thermo Fisher Scientific, Inc. All the sequenced nucleotide sequences were compared using the sequence reported on GenBank.

Detection limit of PCR assay

H. influenzae standard strain ATCC 9334 (type b) was used to determine the detection limit of the ompP6 PCR assay and multiplex PCR assay. The DNA concentration of ATCC 9334 was diluted 10 times and added to the PCR reaction. The DNA concentration was 1.89 × 106 copies/μL, 1.89 × 105 copies /μL, 1.89 × 104 copies /μL, 1.89 × 103 copies /μL, and 1.89 × 102 copies /μL, respectively. The lowest DNA concentration at which positive products could be amplified was considered the detection limit.

Determination of H. influenzae in clinical samples by the multiplex PCR

We used the multiplex PCR assays we established to detect the H. influenzae subtype of the 1514 clinical NPSs.

Statistical analysis

Statistical analysis was performed using IBM SPSS Statistics ver. 11.0 (IBM Co., Armonk, NY, USA). Statistical significance was defined as P < 0.05. Chi-square test was used in this study. The sensitivity and specificity of the multiplex were measured by comparison to the results of bacterial culture and serum agglutination.

Results

Bacterial culture results of 1514 NPSs from children with respiratory tract infections

Of 1514 NPSs from children with respiratory tract infection, 280 were positive for H. influenzae as indicated by bacteria culture, the positive ratio was 18.49%. The H. influenzae infection rates of boys and girls with respiratory tract infectious diseases were 20% (165/825) and 16.69% (115/689), and there was no significant difference in infection rate between boys and girls (χ2 = 1.2, P > .05) (Table 2).

Table 2 Results of bacterial culture and serum agglutination test
Full size table

Results of serum agglutination test of H. influenzae

According to the serum agglutination test, there were 262 NTHi strains among all 280 H. influenzae strains cultured from NPSs, accounting for 93.57%. There were 18 strains of encapsulated H. influenzae, including 4 type a, 1 type b, 4 type c and 9 type f (Table 2).

Specificity, stability, and detection limit of the PCR assay

All the H. influenzae standard strains had amplification of ompP6 and the standard strains of Staphylococcus aureus, Streptococcus pneumoniae, Klebsiella pneumoniae, E.coli, and Pseudomonas aeruginosa showed negative results in the multiplex PCR assays. The encapsulated H. influenzae determined by serological test had amplification products of bexA. The PCR was performed in triplicate at different times and the same results were found. The results obtained here matched the ompP6 and bexA nucleotide sequences reported on the GenBank. The detection limit of the multiplex PCR assay we used was 1.89 × 103 copies /μL.

H. influenzae detection in nasopharyngeal secretions (NPSs) with one-step multiple PCR assay

Two hundred seventy-three of 1514 NPSs were found to be positive only ompP6 positive and 20 of 1514 were found to be positive for both ompP6 and bexA when assessed using the multiplex PCR scenario established above. The sensitivity and specificity of the multiplex PCR were 100 and 99.86%, respectively (Table 3).

Table 3 Comparison of one-step multiple PCR and bacterial culture plus serum agglutination test of 1514 NPS
Full size table

Discussion

H. influenzae is a common pathogen, causing acute respiratory tract infection and meningitis in children in most developing countries [18]. The epidemiology of invasive H. influenzae disease has changed substantially, with most invasive diseases now caused by non-b serotypes and NTHi strains [3,4,5,6,7,8].The incidence of infectious disease caused by Hib declined and that caused by other serotypes of H. influenzae and NTHi has increased over the past 10 years [4,5,6,7], especially in Europe and North America. NTHi caused 78% of all cases and showed increasing trends among persons < 1 month and > 20 years of age in Europe. Establishing the contribution of NTHi to respiratory tract infections may be helpful in the development of a new vaccine against NTHi that outweighs the cost of long-term healthcare in terms of repeated episodes and sequelae.

Drug resistance of bacteria is a serious problem.β-lactamase production in H. influenzae and NTHi is still the dominant resistance mechanism against β-lactams. The prevalence of β-lactamase in NTHi differs widely worldwide, strains with β-lactamase independent resistance to β-lactams have emerged and are spreading. The emergence and spread of β-lactamase-negative ampicillin-resistant strains in many regions of the world is of substantial concern, potentially necessitating changes to antibiotic treatment guidelines for community-acquired infections of the upper and lower respiratory tract and potentially increasing morbidity associated with invasive NTHi infections [19].

In recent years, the predominant strains of H. influenzae have varied across different regions of China, and the distribution of dominant strains in one place has also been affected by disease type, patient age, and similar factors. It is very important to quickly and accurately distinguish H. influenzae from other bacteria, or encapsulated H. influenzae from NTHi for timely treatment.

Zhejiang is a major economic province in China, located in southeastern China with a large population. In this study, we gathered 1514 NPSs and used traditional detection methods to determine the H. influenzae infection ratio and serotype of H. influenzae in Zhejiang Province in China. According to results of the bacterial culture and serum agglutination test, the H. influenzae infection rate in respiratory tract infections in children ranged from 0 to 14 years was 18.48% (280/1514), of all the H. influenzae we detected from clinical samples, 93.57% (262/280) belonged to NTHi, only 6.43% (18/2280) were encapsulated H. influenzae, and only 0.36% were Hib. It means NTHi is the predominant strain of H. influenza in children respiratory tract infection in Zhejiang province, China.

Because H. influenzae is a fastidious bacterium, the rate of detection of H. influenzae with routine bacterial culture is low. Serum agglutination tests have become widely used to determine the serotype of encapsulated H. influenzae. But sometimes there are bacterial self-coagulation phenomena and cross agglutination, and serological reagents are more expensive. In recent years, PCR has become widely used to detect pathogenic microorganisms in different samples, especially bacteria which are difficult to culture under routine conditions.

Considering of these, we also established a one-step multiple PCR assay by detecting the ompP6 and bexA genes, and we used several kinds of standard strains of respiratory tract infection bacteria, including H. influenza, to examine the stability, specificity, and detection limit of the PCR assay. The results showed that the PCR assay had high stability and specificity, and the detection limit of the PCR assay was 1.89 × 103 copies /μL. The 1514NPSs were used to verify the specificity and sensitivity of the PCR assay. In this study, 293 of 1514 NPSs were ompP6- positive, the detection rate of H. influenzae was 19.35% (293/1514). The H. influenzae positive rate of one-step multiple PCR is higher than bacterial culture, it’s the reason why we want detect H. influenzae by PCR instead of bacterial culture because H. influenzae is a fastidious bacterium and maybe there are some false negative.

Twenty of 1514 NPSs were bexA gene-positive, and the positive rate was 1.32%, compared with the serum agglutination test, 2 NTHi were missed diagnosed. It may be a new finding, and it may be the key note for further researches in PCR and molecular characterization.

Compared with the results of bacterial culture and the serum agglutination test, the sensitivities of ompP6 and bexA detection were both 100%, and the specificity of the multiplex PCR was 99.86%. There was no significant difference in the rate of detection of H. influenzae between the multiplex PCR and the routine bacterial culture (18.49%).

According to our study, and considering the shortages of bacterial culture and serum agglutination test, the one-step multiplex PCR assay we established can be used to identify encapsulated H. influenzae and NTHi in place of routine bacterial culture and serum agglutination testing, but it cannot be used to determine the serotype of encapsulated H. influenzae. It is necessary to design new, effective, specific typing primers, which will require further research.

Conclusions

According to our study, the infection rate of H. influenzae in respiratory tract infections was 18.49%; the most prevalent subtype of H. influenzae in Zhejiang Province was NTHi, account for 93.57%; the sensitivity and specificity of the one-step multiplex PCR assay we established were 100 and 99.86%, respectively; the multiplex PCR can be used as the differential detection of clinical H. influenzae strains, replacing the routine bacterial culture and serum agglutination testing.