Background

Pertussis is a respiratory disease mainly caused by Bordetella pertussis. The incidence of pertussis marked decreased after the whole cell pertussis vaccine (wPV) has introduced all over the world. Since the 1990s, a resurgence of pertussis has appeared in many countries, especially when the acellular pertussis vaccine (aPV) has replaced from the wPV. Furthermore, the circulating B. Pertussis has evolved mainly changed in the vaccine antigen genes proposed by the vaccine-driven, such as the ptxP1 lineage to ptxP3 lineage and also pertactin deficient [1]. Nowadays, the ptxP3 lineage with/or without pertactin deficient strains, which has been proved to be more virulent and reflect selective advantage under the high coverage of aPV vaccination, has emerged globally and raised an important public issue toward an alternative vaccine in pertussis prevention [2].

However, except what happened in some countries like Iran, the ptxP1 lineage was still prevalence in most countries used wPVs [3]. We have reported the ptxP1 strains further shown erythromycin resistance (ER) that emerged in China since 2012. Furthermore, we found that all the ptxP1- ER strains originated from a fhaB3 lineage, which appears to have been selected from the wPV or antibiotic pressure [4]. Interestingly, although the ptxP1-ER strains expanded all the countries of China, the proportions of ptxP3-ES strains varied from less than 10% to about 50% in different areas of China, especially occurred much higher in developed areas [5, 6].

The aPV came in two varieties according to the producing procedures: one is obtained through co-purified procedures so called co-purified aPV, which was used primarily in China and Japan. The other one with purification of each one to five components individually antigen and then blending them in an appropriate ration called purified aPV, which was used in lots of areas all over the world [7]. In China, the co-purified aPV was free and used predominantly since 2006. The purified aPV (Sanofi) was imported and rechargeable since 2011 and supplied much more in developed areas in China. Despite the ptxP3-ES strains were proved to be adapted to the purified aPV all over the world and also been prevalence under wPV vaccination in Iran [3, 8] . Moreover, whether the different proportions of ptxP3-ES strains in China were associated with the types of aPV remains enigmatic.

In this study, we conducted a 5-year retrospective study to survey the dynamic changes in genetic makeup & resistance status of the circulating B. pertussis and further the difference in demographic characteristics between different genotypes in Xi’an China, where co-purified aPV was still prevalence used. We are kind of hoping studies such as this can give more information in consideration of the modified vaccine for global pertussis prevention.

Methods

Study populations, strains and samples

All the patients admitted to Xi’an Children's Hospital for suspected of pertussis from 2012 to 2016 were sampled of nasopharyngeal swabs (NPs) and diagnostic by culture and special PCR for B. pertussis as previous reported [9, 10]. The DNAs from B. pertussis ATCC 9797 and ddH2O were used as positive and negative control in each PCR run [9]. The demographic characteristics were collected if culture and/or special PCR for B. pertusssis was positive. Totally, 702 cases were positive by the special PCR which including 204 cases with both culture and PCR positive of B. pertussis. All the strains and NPs were stored at − 80 °C until to use.

Antibiotic susceptibility test

In-vitro sensitivity of clinical strains against erythromycin was performed and reported as previously [9]. We used B. pertussis ATCC 9797 and Staphylococcus aureus ATCC 25923 as controls.

23S rRNA sequencing and antigen gene typing

The nucleotide position 2047 of the 23S rRNA was performed by DNAs of strains and/or NPs by our previously reported sequencing methods [9]. Cause the A2047G of 23S rRNA was associated with erythromycin resistance, if the nucleotide position 2047 of the 23S rRNA was the wild type as adenine (A), we defined as an erythromycin sensitive B. pertussis infection. A mutation type as guanine (G) of site 2047 was taken for erythromycin resistance B. pertussis infection strain [11, 12]. The allele of ptxP and fhaB was performed by DNAs of strains and/or NPs as previously reported when successful sequencing of 23S rRNA [13].

Statistical analysis

Data were statistically analyzed with SPSS 17.0. Comparisons were performed using χ2 test or one-way analysis of variance (ANOVA). A P value < 0.05 was considered statistical significant.

Results

In total of 204 isolates, 3 strains were ptxP3 allele and the others were all ptxP1 alleles. Four strains, including of 3 ptxP3/fhaB2 and 1 ptxP1/fhaB2 allele, have the MICs against erythromycin lower than 0.023 μg/ml, which refers to sensitive to erythromycin in vitro. The rest of the 200 strains were all resistant to erythromycin with the MICs≥256 μg/ml, which were all ptxP1/fhaB3 allele. All the resistant strains posed an A2047G mutation in 23S rRNA and no mutation occurred this site of the sensitive strains.

Among the 702 NPs for sequencing, 480 obtained both the available sequencing results of 23 rRNA, ptxP and fhaB. The NPs generated the sequencing results as same as obtained by strains from the same patient. There were 449 in 480 specimens (93.5%) shown the allele G in 2047 site of 23 rRNA that defined as erythromycin resistant B. pertussis infection, which also shown the allele of ptxP1/fhaB3. The dynamic changes of proportions of circulating B. pertussis from 2012 to 2016 as shown in Fig. 1.

Fig. 1
figure 1

The proportions of different B. pertussis lineage in B. pertussis population from 2012 to 2016 in Xi’an, China

Full size image

Furthermore, 47 patients were excluded when analysis the difference among demographic characteristics cause of unclear vaccination status. Among the remaining 433 patients, 136 patients have been administrated with least 1 does of co-purifid aPV. Vaccinated with co-purified aPV was found in 133(133/403,33.0%), 1(1/9,11.1%) and 2(2/21,9.5%) in ptxP1/fhaB3-ER, ptxP1/fhaB2-ES and ptxP3/fhaB2-ES infected children, which showed a significant difference (χ2 = 6.87, P = 0.032). (Table 1).

Table 1 The demographic characteristics of children suffering from pertussis with different genetic makeup & erythromycin resistance status of B. pertussis
Full size table

Discussion

Within our study, we discovered that ptxP1-ER strains have been steadily increased to the circulating B. pertussis population from 2012 to 2016 in Xi’an, China. Moreover, unlike that B. pertussis could not only infect the infants that were too young to be vaccinated, but also the infants vaccinated with the purified aPV [14, 15], the ptxP3 strains rarely infected the infants administrated with co-purified aPV from our study.

The ptxP3 strains have been circled predominant all over the world, especially after the replacement of wPV by aPV. However, the ptxP3 strains have spread across the globe seems not only driven by aPV selection, but also by the fitness of ptxP3 strains when compete with erythromycin sensitive non- ptxP3 strains [3, 8]. Furthermore, we assumed that the quality of the wPV between different batch and/or the frequency of international movement of people possibly accelerated the prevalence of ptxP3 strains in Iran too. The increasing incidence of pertussis was also emerged in China from 2013 according to the national infectious diseases case reported system. Besides the A2047G mutation in 23S rRNA occurred in ptxP1-ER B. pertussis strains, a novel fhaB C5330T was also founded in all these strains but didn’t appear in any ptxP3 lineage. This fhaB3 lineage has been proved to be prevalence among China via expansions most likely due to vaccine and/or antibiotic pressure [4]. This study illustrated that the ptxP1/fhaB3-ER strains might be adapted to the co-purified aPV more easily than global ptxP3 strains.

It was reported that the ptxP3 strains have been occurred in 2000 and remains sporadic in this country [16]. According to this study, ptxP3 strains with the decreased proportions have observed from 2012 to 2016 in Xi’an, the western of China. In China, the co-purified aPV was free and predominated used since 2006 while the purified aPV (Sanofi) was available by paid since 2011with rarely market supplied, especially in undeveloped regions of western China, such as Xi’an. The rare of ptxP3 strains in Xi’an after 10 years of co-purified aPV used might give a clue that the co-purified aPV did not give the adaption as purified aPV did in developed countries where ptxP3 was quickly predominant worldwide [8]. The co-purified aPVs have more protein antigen than purified aPVs [17]. Therefore, this study further supported the hypothesis that the small antigen targets of purified aPV could induce the vaccine pressure and vaccine adaption more easily than the more antigen targets vaccine, such as wPV, even the co-purified aPV [18]. Furthermore, among the additional protein antigens of co-purified aPV, most was the out membrane proteins such as BipA and SphB1. Such membrane proteins containing in the outer membrane vesicles (OMVs) of B. pertussis have been suggested as an attracting candidate component of the possible new modified vaccine against pertussis [19, 20]. The latest study further proved that the OMVs can protect against B. pertussis with long term duration, even the global popular ptxP3 and pertactin deficient strains [20]. Japan was the first country to develop aPV (co-purified) in 1981 and to adopt for use in the general population. It has been reported that both of the two types of aPVs was used recently [21]. However, the ptxP3 lineage still holds lower than 50% from 2006 to 2010 until the period of 2011-2014which reached close to 80% [22].

Most of the cases in this study were from the west of China. Otherwise, it was reported that the ptxP1-ER strains contributed to 75.4, 50.7 and 48.6% in the circulating strains in Zhejiang province (Southern of China, 2016), Shanghai (Southern of China, 2016–2017) and Shenzhen (Southern of China,2015–2017), while the rest ES strains were almost ptxP3 strains [5, 23, 24]. No details of the vaccine type were described in these relative high proportion of ptxP3 areas of China. Liking what happened to Japan, we assumed that the purified aPV used was much more in these developed areas of China than in Xi’an, which generate a relative low level of vaccine protection from co-purified aPV in general population. As a result, the proportions of ptxP3 strains were much more.

Consistent with reports in these areas of China, the erythromycin resistant strains were all ptxP1 allele while the ptxP3 strains were all sensitive to erythromycin. As shown in this study, though the average age of ptxP3-ES strains infection group is lower than in ptxP1-ER groups, no statistic significant difference was observed. Furthermore, more than 85% of subjects have taken antibiotics before sampling and detection, no difference was observed between the ptxP3- ES and ptxP1-ER groups (data not shown). Therefore, despite the antibiotic pressure which seems to provide the selective advantage for expansion of erythromycin resistant strains, we pose a hypothesis that the co-purified aPV protect against ptxP3 strains more efficient, which generated a rare chance for ptxP3 strains to be under the antibiotic pressure and further developed to be erythromycin resistance.

However, It is a limitation that the cases of ptxP3 strains were relative too small to give strong evidence about the protection against ptxP3 lineage by co-purified aPV. Furthermore, the ptxP3 with the pertactin (PRN) deficient isolates were widely appeared in some industries countries [25], whether the ptxP3 isolations in this study expressed of PRN were unknown in this study. Lastly, the age of the patients in our study was mainly the infant but not children after at least 5 years of vaccination of co-purified aPV. Thus we can not give powerful support about the protection duration against ptxP3 linage of the co-purified aPV.

Conclusions

In conclusion, this study revealed that the erythromycin resistant B. pertussis have been steadily increased from 2012 to 2016 in Xi’an, western of China. We also pose a hypothesis that the co-purified aPV containing more antigens has the possibility to protect the infant from being ill of pertussis infected by global popular ptxP3 lineage B. pertussis. To be better understanding the effect of co-purified aPV, an international multicenter cohort study should be performed.