This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Introduction
Haemophilus influenzae is a common pathogen of community-acquired infections in children (Fu et al. 2021), which is the leading cause of pneumonia, otitis media, conjunctivitis, meningitis, septicemia, and other diseases (Vallejo et al. 2019). H. influenzae is classified in various schemes, of which capsule type and biotype are the most common. Bacterial typing is an essential basis for disease epidemiology and vaccine development. Using various biochemical reactions, such as with the indigo substrate, indole, and ornithine decarboxylase, H. influenzae can be divided into eight biotypes, namely I–VIII. According to the properties of capsular polysaccharides, H. influenzae can also be divided into six serotypes (from a to f) and non-typeable H. influenzae strains (NTHi) (Wen et al. 2020). The drug resistance, disease characteristics, invasiveness, and infection-vulnerable tissues differ between these bio- and serotypes (Crandall et al. 2019; Fuji et al. 2021). The literature data indicates a relationship between the selected biotypes and the sites of their colonization, as well as their association with specific infectious problems (Brabender et al. 1984; Mojgani et al. 2011; Crandall et al. 2019; Heliodoro et al. 2020; Li et al. 2020; Fuji et al. 2021). NTHi has replaced H. influenzae type b (Hib) as the leading cause of invasive infection in children; however, it has less concern for pediatricians (Butler et al. 2018; Heliodoro et al. 2020). Different regions and populations showed mixed H. influenzae typing results. The classification of H. influenzae in children in the Kunming region of China is unclear.
In recent years, the incidence of H. influenzae infection in children has increased, and profiles of antimicrobial resistance among children in China demonstrated that the proportion of H. influenzae in all pathogenic bacteria has risen from one-fourth in 2017 to one-third in 2018. It seriously endangered children’s health in China (Fu et al. 2021). With the widespread use of antibiotics, the resistance pattern of H. influenzae to commonly used antibiotics has also undergone significant changes. Common resistance patterns include β-lacta- mase-negative ampicillin-resistant (BLNAR), β-lactamase-positive ampicillin-resistant (BLPAR), β-lactamase- negative ampicillin-sensitive (BLNAS), β-lactamase-producing amoxicillin-clavulanate-resistant (BLPACR), and BL-nonproducing ampicillin-intermediate (BLNAI) (Kim et al. 2007). Resistance mechanisms are also becoming more complex, which also complicates the choice of antibiotics for treatment. The mechanism of drug resistance for H. influenzae included the production of β-lactamase and mutations of the fstI gene, which encodes a penicillin-binding protein (PBP) of low affinity (Li et al. 2020). The former is mainly mediated by the TEM-1 and ROB-1 genes, and the latter mainly expresses the PBP3-S and PBP3-BLN mutations in the ftsI gene (Søndergaard and Nørskov-Lauritsen 2016).
Nevertheless, rare drug resistance has not yet been detected in H. influenzae strains in southwestern China. The replacement phenomenon and capsular switching are widespread in Streptococcus pneumoniae, another microorganism associated with meningitis and pneumonia. Therefore, it is required to monitor the spreading of H. influenzae in the post-vaccine period to know the circulating serotypes in the community. It is also desired to know which serotypes are circulating in the region to study this data compared to other regional data. It would help to analyze the factors causing the serotype variation.
This paper studied the serotypes, common drug resistance patterns, and drug resistance genes of H. influenzae strains in Kunming. Besides, the scientific basis for the diagnosis, the prevalence and drug resistance of H. influenzae in children, as well as the prevention and treatment and prevention of these bacteria, was provided.
Experimental
Materials and Methods
Patients and clinical isolates. This study was approved by the institutional Review Board, and written informed consent was obtained from each participant. Ethical approval was obtained from the Zhu Futang Clinics of Pediatrics. The patients were recruited at the Children’s Hospital affiliated with Kunming Medical University (Kunming, China), where the publication of data generated from this study was approved. Informed consent was obtained from the participants with the option to withdraw from the study at any time. A total of 800 H. influenzae strains isolated from children of 0–15 years, diagnosed with H. influenzae infection according to the Zhu Futang Clinics of Pediatrics, were recruited at the Children’s Hospital affiliated with Kunming Medical University between 2018–2021. H. influenzae isolates were identified by satellite phenomenon and the VITEK®2 NH card (bioMérieux, France) (Fuji et al. 2021). The ATB identification cards were used to test the drug sensitivity. H. influenzae ATCC® 49247™ was used as the reference strain for the quality control.
Capsular serotyping and genotyping. Capsule typing was carried out on 24-h subcultures with the Phadebact™ Haemophilus Test 50 kit (TaKaRa Company, Japan) (Collins and Kelly 1983). Strains were reported as having a type b capsule, a non-type-b capsule, or as non-typeable isolate. The classic slide agglutination method was used for serotyping (Crandall et al. 2019). H. influenzae bacterial suspension was mixed with specific antiserum of H. influenzae type a, b, c, d, e, f. Slide agglutination occurred within 1 minute, which indicated that the corresponding capsular serotype was positive. If neither aggregated, it was judged as an NTHi strain. All the isolated strains were serotyped by amplifying the bexA capsular gene with PCR, as described previously. Strains that PCR could not identify were classified as non-typeable (NTHi) (Falla et al. 1994). PCR was also done to identify the corresponding capsular genes of H. influenzae of type “a” to type “f” (Falla et al. 1994). The primer sequences of capsular genes in H. influenzae are provided in Table I (Fella et al. 1994).
The primers for capsular genes of Haemophilus influenzae (Fella et al. 1994).
Biotyping. The biotypes of the isolates were determined by the method described by Kilian (Kilian et al. 1979; Kilian 1985). Liquid media prepared in the laboratory were inoculated with 24-h old or fewer subcultures. Test cultures were incubated in the air for 4 h at 35°C. This biotyping system uses the presence of preformed enzymes, and because growth is not required, the media were not supplemented with X and V factors. According to the difference of reaction results in biochemical reaction tubes of indole, ornithine decarboxylase, and indole, H. influenzae strains could be divided into eight biotypes: I (+, +, –), II (+, +, –), III (–, +, –), IV (–, +, +), V (+, –, +), VI (–, –, +), VII (+, –, –), VIII (–, –, –,).
Antibiotics susceptibility testing and β-lactamase assay. The minimum inhibitory concentrations (MICs) were determined by the microdilution method according to the CLSI guideline (CLSI 2011). Eighteen antibiotics were tested, and the results were interpreted according to CLSI standards (Søndergaard and Nørskov-Lauritsen 2016). H. influenzae ATCC® 49247™ and ATCC® 49766™ were used as the quality controls. Production of β-lactamase was detected with the cephalosporin thiophene paper (Oxoid Company, UK). When the paper turned red, it was a positive indicator that the bacteria produced β-lactamase (Yamada et al. 2020). The antibiotic resistance-encoding genes were identified using the primers described in Table II in a conventional PCR.
Statistical analysis. We utilized the chi-squared test to compare the differences among groups. The measurement data were expressed as a mean ± standard deviation. A p-value of < 0.05 was considered statistically significant. All the analyses were carried out within the SPSS v22.0 (IBM, USA).
Results
Clinical characteristics of pediatric patients. This study characterized 800 H. influenzae strains, of which most came from respiratory specimens (738, 92.3%) (sputum and bronchoalveolar lavage fluid). The second primary source of strains was wound secretions (31, 3.8%). The remaining 12 strains (1.5%) were from ocular secretions, 11 isolates were from blood (1.4%), and 8 (1.0%) were from the cerebrospinal fluid. H. influenzae strains were identified by the presence of the P6 gene. The bexA gene found in all the subtypes for H. influenzae was also sought, and the isolates, which showed the absence of the bexA gene, were classified as NTHi isolates. Almost 82% (n = 652) of H. influenzae isolates were recovered from 2–15 years old children’s samples. The remaining 18.5% of ilates (n = 148) were isolated from children of 0–2 years and were vulnerable to H. influenzae meningitides and pneumonia. Only these isolates were used for typing purposes.
Typing. One hundred forty-eight H. influenzae strains isolated from children (0–2 years) were selected for serotyping. It appeared that 132 strains (89.2%) were NTHi and nine strains (6.1%) were H. influenzae type b, whereas four strains (2.7%) were H. influenzae type e, two strains (1.3%) were H. influenzae type f, and one strain (0.7%) was H. influenzae type c1. The corresponding capsule-encoding genes were detected accurately (Fig. 1).
Fig. 1.
Electrophoregram of Haemophilus influenzae capsular genes’ amplicons, including NT0Hi, Hie, Hib, Hif, and Hic strains.
The biotyping showed the presence of 64 (43.2%) type II strains, 55 (37.2%) type III strains, 12 (8.1%) type I strains, and eight (5.4%) type IV strains. Type V and VI were detected in four (2.7%) strains, while only one strain was of type VII. Two blood isolates were identified as NTHi and were of biotypes II and IV. The others blood isolates were H. influenzae type b and biotype I. Six wound isolates were nontypeable and of type III. All ocular isolates were of biotype II and were non-typeable.
Drug resistance patterns. All the 800 strains were tested for antibiotic resistance using the minimum inhibitory concentrations (MICs) method according to the CLSI guideline (CLSI 2011) using eighteen antibiotics. Among them, 474 (59.5%) strains were β-lactamase positive. In the group of β-lactamase positive strains, 295 were multi-drug resistant (MDR) to numerous antibiotics, such as sulfamethoxazole/trimethoprim, ampicillin, cefaclor, cefuroxime, amoxicillin-clavulanic acid, tetracycline, and chloramphenicol. The rate of MDR in the β-lactamase-positive strains was significantly higher than that of β-lactamase-negative strains (326 (40.5 %)). Among β-lactamase-negative strains, 106 were multi-drug resistant.
Based on the resistance patterns towards ampicillin, amoxicillin, and clavulanic acid, the strains were classified into BLPAR (43.5%), BLNAS (24.6%), BLPACR (16.5%), BLNAR (14.8%), and BLNAI (0.6%) (Table III). The TEM-1-type bla gene, PBP3-BLN (ftsI) gene, PBP3s gene, and ROB-1-type bla gene were detected in 54.1%, 18.9%, 11.8%, and 6.9% of 800 H. influenzae strains, respectively. The proportion of BLNAS was 12% of TEM-1-positive strains; 16% of ROB-1-positive strains. The BLNAR detection rate was 60% for PBP3BLN-positive strains, and 40% for PBP3S-positive strains. The BLPACR phenotype was detected in 56% of TEM-1 and PBP3BLN-positive strains. Simultaneously, 28% of β-lactamase-positive strains had both TEM-1 and PBP3s, while 12% of β-lactamase-positive strains had two TEM-1 and PBP3s genes. The PBP3S and PBP3BLN genes were seen together in 4% of the β-lactamase-positive strain (Fig. 2).
Drug resistance rates of Haemophilus influenzae strains [n (%)].
– indicates a statistically significant difference;
– indicates a particularly significant statistical difference
Fig. 2.
Electrophoregram of Haemophilus influenzae drug-resistance genes (in the BLNAR and BLPACR phenotype strains).
Discussion
H. influenzae causes a common community-acquired infection in children and has regional strong characteristics in drug resistance and serotypes (Falla et al. 1994). In Mainland China, limited surveys focused on the epidemiologic data of H. influenzae from children, especially those under five years. The surveillance of serotype distribution and prevalence of drug-resistant strains in the general population is critical for the public health governmental departments to develop appropriate prevention protocols for H. influenzae infection.
Ampicillin and second-generation cephalosporins are commonly used drugs for treating H. influenzae infections. However, they may not be suitable for treatment due to high resistance found in the clinical strains (Heinz 2018; Ubukata et al. 2019). Antibiotics such as third generation cephalosporins and amoxicillin-clavulanate are recommended for children in this area. The results of capsule typing by serological methods and PCR were consistent. The prevalent strains in Yunnan were NTHi, while the capsular H. influenzae strains were rare, consistent with the literature (Fally et al. 2021). The current vaccine against H. influenzae can only prevent Hib infection. This vaccine is no longer suitable for the current epidemiological situation, where NTHi has become the prevalent type (Dong et al. 2020). Research on new and effective vaccines against NTHi has become the current focus for preventing and treating these bacteria (Bakaletz et al. 2018). According to the difference in the reaction results of H. influenzae with the indigo substrate, indole, and ornithine decarboxylase, the bacteria can be divided into eight biotypes, namely I–VIII. The results of this work showed that types II and III were the most prevalent biotypes in children in this area and are different from those reported in the literature. The H. influenzae strains from adults in Kunming are mainly of types III and IV, in the neighboring city of Chengdu are mainly of types I and IV, and in Iran are mainly of types I–III (Langereis and de Jonge 2020). Different biotypes cause different manifesting diseases. This study shows that the Hib and biotype I strains are mainly detected in blood and cerebrospinal fluid samples, indicating a specific relationship in their invasion. Type III was mainly detected in wound secretion specimens, which may be associated with cellulitis. Type II was detected in all ocular secretion specimens, which may be related to ocular conjunctivitis. Unlike those reported in the literature, type I often causes acute lower respiratory tract infection and sepsis in children, type II causes explosive respiratory disease, type III causes tissue cellulitis, and type IV causes neonatal sepsis. However, there are also limits to this study. The above findings’ mechanisms were not clarified and still need further exploration.
There were regional differences in the detec- tion rate of β-lactamases, drug-resistance genes, and drug-resistance patterns. The rate of antibiotic resistance and multidrug resistance of β-lactamase-positive strains was higher than that of β-lactamase-negative strains, indicating that the production of β-lactamase is one of the important reasons for the increased drug resistance and multidrug resistance of H. influenzae (Mojgani et al. 2011). In this study, the detection rate of β-lactamase was slightly lower than the national average level of China, as reported in 2022 (Zhou et al. 2022). The detection rate of drug resistance gene TEM-1 was the highest, indicating that the drug resistance phenotype BLPAR mediated by TEM-1 β-lactamase-producing strains was the primary drug resistance mechanism in H. influenzae (Fu et al. 2021). Mutations in the ftsI gene (PBP3S and PBP3BLN) result in the BLNAR phenotype (Nørskov-Lauritsen et al. 2021). The β-lactamase production and ftsI gene mutations work together to result in BLPACR, which is consistent with the resistance mechanism reported in the literature (Schotte et al. 2019). The higher detection rate of BLPAR shows that the most common drug resistance pattern in this region is BLPAR, while BLPACR and BLNAR are also popular, which may be one of the reasons for multidrug resistance (Cherkaoui et al. 2015). This situation requires rigorous attention.
In conclusion, H. nfluenzae is one of the most common pathogens of childhood diseases, invading respiratory tract infections. The common biotypes in this region are II and III, and the common serotype is NTHi. The antibiotic resistance and multi-resistance levels are severe. The resistance rate to ampicillin and second-generation antibiotics is high, so third-generation cephalosporins and amoxicillin-clavulanic acid are recommended for the treatment. The detection rate of drug-resistant phenotypes BLPAR, BLNAR, and BLPACR was high, and BLPAR was the most prevalent. The TEM-1, ROB-1-type β-lactamase, and PBP3BLN, PBP3S resistance genes contributed to the observed drug resistance. In short, this region’s drug resistance phenotypes and mechanisms are diverse, and the drug resistance situation is complex. It is necessary to monitor the H. nfluenzae drug-resistance phenotypes continuously.
Conclusions
Non-typeable H. influenzae strains are an important cause of bacteremic pneumonia and meningitis in children 0–2 years from the Kunming region. The dominant biotypes were II and III. The BLPAR was the most often detected resistant phenotype among NTHi isolates.
