Genomic surveillance reveals global spread of macrolide-resistant Bordetella pertussis linked to vaccine changes

ABSTRACT

The resurgence of whooping cough in regions utilizing acellular pertussis vaccines underscores emerging public health challenges. Here, we characterized 178 Bordetella pertussis isolates collected from patients across all age groups in Shanghai (2018–2024) to assess genomic evolution and antibiotic susceptibility. Macrolide resistance to erythromycin, azithromycin, and clarithromycin escalated from ≤50% (pre-2020) to nearly 100% (post-2020), mechanistically linked to the 23S rRNA A2047G mutation. Genome-based analysis identified a genotype MT28-ptxP3-MRBP rapidly dominated post-2020, exhibiting significantly higher prevalence in adults than in other age groups. Phylogenetic analysis of 178 Shanghai and 1,596 global genomes revealed two major lineages corresponding to ptxP1 and ptxP3 alleles. MT28-ptxP3-MRBP cluster was identified in France, Japan, and the United States in 2024, indicating potential cross-border transmission. These findings advocate for integrated surveillance spanning all ages and international borders to contain the global spread of macrolide-resistant B. pertussis.

IMPORTANCE

In recent years, despite high coverage of acellular pertussis vaccines in China, pertussis cases have increased substantially. Drawing on 178 Bordetella pertussis isolates obtained through age-inclusive active surveillance in Shanghai (2018–2024), we found that macrolide resistance rose from ≤50% before 2020 to nearly 100% thereafter, with all resistant isolates harboring the 23S rRNA A2047G mutation. A resistant MT28-ptxP3 lineage became dominant after 2020 (61.7%) and was disproportionately represented among older age groups; the primary affected population shifted from children ≤36 months toward those aged 37 months to 18 years. Incorporating NCBI public genome data, we further observed that this resistant lineage is not confined locally, suggesting a risk of cross-border spread. These findings provide an early warning of the expansion of macrolide-resistant pertussis and underscore the need for age-inclusive, cross-regional genomic surveillance and re-evaluation of diagnostic workflows, antimicrobial stewardship, and immunization strategies.

INTRODUCTION

Whooping cough, an acute respiratory infectious disease caused primarily by Bordetella pertussis (B. pertussis), is characterized by high transmissibility and poses a severe threat to young infants (1). Over the past two decades, countries with high coverage of acellular pertussis (aP) vaccination—including the United States, France, and New Zealand—have witnessed a resurgence of whooping cough from historically low incidence in the early 21st century, with regional outbreaks defining the phenomenon of “pertussis resurgence” (2–4). Since 2012, China has exclusively used aP vaccines (5), maintaining a coverage rate exceeding 97% (6). Paradoxically, reported cases surged from 2016, reaching over 30,000 in 2019—the highest levels since the late 1980s (6, 7).

Notably, B. pertussis can infect or reinfect individuals across all age groups (8), yet the true disease burden in adults remains significantly underestimated (9, 10). Current studies on antibiotic susceptibility and genomic evolution of Chinese B. pertussis isolates predominantly focus on pediatric populations (11–13), with limited systematic investigations across other age demographics. This gap hinders comprehensive assessment of resistance dynamics and transmission risk.

Here, we leveraged Shanghai’s active surveillance system to characterize 178 B. pertussis isolates from patients of diverse age groups (2018–2024). Through antibiotic susceptibility testing and whole-genome analyses, we evaluated trends in macrolide resistance, characterized molecular evolutionary features, and integrated epidemiological data to explore age-related distribution patterns and potential risks of international dissemination.

We observe a dramatic surge in macrolide resistance, with rates escalating from ≤50% (pre-2020) to nearly 100% (post-2020). All resistant isolates harbored the 23S rRNA A2047G mutation, a hallmark of macrolide resistance in B. pertussis. Multilocus variable-number tandem repeat (VNTR) analysis (MLVA) and vaccine antigen genotyping revealed rapid expansion of the MT28-ptxP3 lineage of macrolide-resistant B. pertussis (MRBP) after 2020, with a significantly higher prevalence in adults compared to other age groups; global phylogenetic analysis further demonstrated the detection of this lineage in France, Japan, and the United States in 2024, indicating a potential risk of cross-border transmission. These findings underscore the critical need for continuous, age-stratified surveillance of B. pertussis infections. The rapid emergence and international dissemination of MRBP highlight the urgency of enhancing global collaborative efforts to address this evolving public health challenge.

MATERIALS AND METHODS

Bacterial isolates

Per the Shanghai pertussis active surveillance protocol, six sentinel sites were included from 2018 through 2023, increasing to 10 in 2024 following the addition of four sites. In accordance with the pertussis surveillance case definition, nasopharyngeal swabs were obtained from patients meeting any one of the following four criteria: (i) paroxysmal, spasmodic cough, regardless of duration; (ii) post-tussive vomiting; in severe cases, subconjunctival hemorrhage or ulceration of the lingual frenulum; (iii) neonates or infants with unexplained paroxysmal cyanosis or apnea, often without classic paroxysmal cough; (iv) cough lasting ≥2 weeks with alternative causes excluded. Specimens were immediately transported to the Bacterial Testing Laboratory of the Shanghai Municipal Center for Disease Control and Prevention (Shanghai CDC). Upon receipt, specimens were plated onto charcoal‐selective agar (Qingdao Zhongchuang Huike Biotechnology Co., Ltd., China) and incubated at 36°C in 5% CO₂ for 3–7 days. Following confirmation of B. pertussis IS481-targeted nucleic acid using a commercial PCR assay (Jiangsu Bioperfectus Technologies Co., Ltd., China), single colonies were subcultured onto charcoal agar and preserved in milk-based cryovials at −80°C for subsequent analyses. While highly sensitive for B. pertussis, IS481 is not entirely species-specific. Therefore, all isolates were ultimately confirmed as B. pertussis by whole-genome sequencing. This study was approved by the Ethics Committee of Shanghai CDC (approval no. KY-2025-15).

Antimicrobial sensitivity tests

A 0.5 McFarland standardized suspension of isolated organisms was prepared in API 0.85% NaCl solution (bioMérieux, France) and uniformly plated onto charcoal agar. E-test strips (Liofilchem, Italy) for erythromycin (ERY), azithromycin (AZM), and clarithromycin (CLR) were applied, and plates were incubated at 36°C in 5% CO₂ for 72 h to determine minimum inhibitory concentrations (MICs). Because neither the Clinical and Laboratory Standards Institute (CLSI) nor the European Committee on Antimicrobial Susceptibility Testing currently provides species-specific breakpoints or standardized antimicrobial susceptibility testing (AST) methods for B. pertussis, the resistance interpretation in this study is based on prior studies and is provided for comparative reference only (14). Accordingly, all MICs/AST results are presented solely for surveillance purposes and are not intended to inform clinical therapy. Quality control for macrolide MIC testing was performed with Streptococcus pneumoniae ATCC 49619 in accordance with CLSI recommendations (14, 15).

Whole-genome data sources, assembly, and screening

Genomic DNA from revived B. pertussis isolates was extracted using the QIAamp PowerFecal Pro DNA Kit (QIAGEN, Germany) according to the manufacturer’s instructions. DNA purity and concentration were evaluated using a NanoDrop spectrophotometer (Thermo Fisher Scientific, USA) and a Qubit 4.0 Fluorometer (Thermo Fisher Scientific, USA), respectively. Samples meeting quality criteria were subjected to paired-end 150 bp (PE150) sequencing on both the DNBSEQ-T7 platform (MGI, China) and the NextSeq 2000 platform (Illumina, USA). In addition, publicly available B. pertussis genomes deposited in NCBI from January 2016 through December 2024 were retrieved for comparative analysis.

Raw FASTQ reads were quality-controlled with fastp v0.23.4 and de novo assembled using SPAdes v3.15.5, retaining contigs >1,000 bp. Both newly assembled and downloaded FASTA genome sequences were taxonomically classified with Kraken2 v2.1.3 and assessed for completeness and contamination using CheckM2 v1.0.2. Only assemblies identified as B. pertussis with ≥99.9% completeness and ≤1.5% contamination were included in downstream analyses. In total, 1,774 B. pertussis genomes were analyzed, including 178 isolates sequenced in this study and 1,596 isolates retrieved from NCBI databases (see Table S1 for epidemiological information on the downloaded isolates).

23S rRNA A2047G mutation detection

The A2047G mutation in the B. pertussis 23S rRNA gene was detected by two complementary approaches. First, assembled genomes were aligned to the Tohama I reference (GenBank accession GCA_000195715.1) using nucmer v3.1 to call nucleotide variants. Second, 23S rRNA loci were typed by querying the predefined alleles in the BIGSdb-Pasteur database; isolates classified as allele “13” were designated as harboring the resistance mutation (16).

Multiple locus variable-number tandem repeat analysis

MLVA was performed using the wgsMLVA pipeline as previously described by Weigand et al. (17) and B. pertussis isolates were typed according to the five‐locus VNTR scheme (VNTR1, VNTR3a/VNTR3b, VNTR4, VNTR5, and VNTR6) proposed by Schouls et al. (18).

MLST and vaccine antigen genotyping

Isolates were typed by MLST using the scheme established in the BIGSdb-Pasteur database. Key vaccine antigen loci—including ptxP, ptxA, ptxC, fhaB2400_5550, prn, fim2, fim3, and tcfA—were then extracted from the database definitions to characterize the vaccine antigen genotype of each isolate. Because the full-length fhaB gene (~10,773 bp) is often fragmented during genome assembly, the analysis was restricted to the fhaB2400_5550 fragment as defined in the BIGSdb-Pasteur database (16).

Phylogenetic analysis

Using the Tohama I reference genome (GenBank accession GCA_000195715.1), assembled contigs were aligned with Snippy v4.6.0 using default parameters, and recombinant regions were filtered out with Gubbins v2.4.1. The resulting core single-nucleotide polymorphism(SNP) alignment was used to infer a maximum‐likelihood phylogeny in IQ‐TREE v2.3.6 with automated model selection (-m MFP). Branch support was assessed by 1,000 ultrafast bootstrap replicates (-B 1000) and 1,000 SH-aLRT tests (-alrt 1000). All trees were visualized and edited on the Interactive Tree of Life (iTOL) web server (https://itol.embl.de/, accessed 26 August 2025). Genotypes not listed in the figure legends or not identified by these bioinformatic analyses were collectively designated as “Others.”

Statistical analysis

To ensure consistency and accuracy amid incomplete age data for some pediatric cases, we defined the “parental” age group as 19–40 years and the “grandparental” age group as >40 years. Strains were categorized into three temporal groups—pre-2020 and post-2020—based on prior studies (11). All statistical analyses were performed using SPSS v25.0. Categorical variables were compared by χ² test or Fisher’s exact test, and a two-sided P value <0.01 was considered statistically significant.

RESULTS

Epidemiological characteristics of the 178 culture-confirmed patients

A total of 2,415 nasopharyngeal swabs yielded 178 B. pertussis isolates, with epidemiological characteristics detailed in Table 1. No isolates were recovered in 2020 due to low sampling, whereas the highest number of isolates occurred in 2024 (55/178, 30.90%). In the pre-2020 cohort, 70% of cases occurred in infants ≤36 months of age. In contrast, post-2020 cases predominantly occurred in school‐age children and adolescents (37 months to 18 years, 52.17%). Two age-specific proportions differed significantly between periods (infants: χ² = 44.31, P < 0.01; children/adolescents: χ² = 43.74, P < 0.01), whereas cases in adults (≥19 years) showed no significant change (χ² = 0.091, P > 0.05).

TABLE 1.

Epidemiological characteristics of B. pertussis isolates, Shanghai, 2018–2024^a

Year	Total	2018	2019	2021	2022	2023	2024
Nasopharyngeal swabs	2,301	365	382	343	249	407	555
Culture-confirmed	178	24	26	35	27	11	55
Gender
Male	93	14	11	19	17	7	25
Female	85	10	15	16	10	4	30
Age
<3 months	26	9 (37.50%)	7 (26.92%)	2 (5.71%)	3 (11.11%)	2 (18.18%)	3 (5.45%)
3–6 months	22	5 (20.83%)	7 (26.92%)	7 (20.00%)	2 (7.41%)	0	1 (1.82%)
7–36 months	10	5 (20.83%)	2 (7.69%)	2 (5.71%)	0	0	1 (1.82%)
37 months to 6 years	32	0	1 (3.85%)	11 (31.43%)	7 (25.93%)	2 (18.18%)	11 (20.00%)
7–18 years	41	0	0	1 (2.86%)	10 (37.04%)	3 (27.27%)	27 (49.09%)
19–40 years	19	4 (16.67%)	4 (15.38%)	10 (28.57%)	0	1 (9.09%)	0
>40 years	28	1 (4.17%)	5 (19.23%)	2 (5.71%)	5 (18.52%)	3 (27.27%)	12 (21.82%)

^a In 2020, a total of 114 nasopharyngeal swabs were received; none yielded culture-confirmed B. pertussis.

Antimicrobial susceptibility of B. pertussis and analysis of the A2047G resistance mutation

Of the 178 B. pertussis isolates, 152 (85.39%) exhibited MICs >256 mg/L for all three tested antibiotics, while the remaining isolates showed MICs ≤1 mg/L (Fig. 1). After 2020, macrolide resistance rates surged from ≤50% to nearly 100%. Molecular assays confirmed that the 23S rRNA A2047G mutation was exclusively in all resistant isolates, with no detection in susceptible isolates.

Fig 1.

Macrolide susceptibility of B. pertussis isolates in Shanghai, 2018–2024. Stacked bars indicate the annual percentage of macrolide-sensitive B. pertussis (MSBP) and macrolide-resistant B. pertussis (MRBP) isolates, with the leftmost bar showing the overall (“Total”) distribution. Blue segments denote MSBP (ERY/AZM/CLR MIC ≤1 mg/L) and orange segments denote MRBP (ERY/AZM/CLR MIC ≥256 mg/L); the number within each segment indicates the count of isolates. The y-axis shows percentage of isolates, and the x-axis shows year of collection.

Analysis of MLVA and multilocus sequence typing

All isolates were assigned to ST2 by MLST. In addition, 17 distinct MLVA types (MTs) were identified, excluding nine isolates with unassignable MTs. The predominant types were MT28 (n = 87), MT195 (n = 26), MT60 (n = 22), and MT27 (n = 15) (Table 2). Notably, the prevalence of MT28 increased dramatically from 16% (8/50) pre-2020 to 61.17% (79/128) post-2020, paralleling a rise in macrolide-resistance from 0% (0/8) to 100% (79/79) with this lineage. In contrast, MT195 prevalence declined sharply from 48% (24/50) pre-2020 to 1.56% (2/128) post-2020. MT60, a post-2020 emergent type, accounted for 17.19% (22/128) of isolates, all of which were macrolide-resistant.

TABLE 2.

Macrolide resistance of B. pertussis by MLVA type before and after 2020^a

	MT27		MT28		MT195		MT60		Others
	S	R	S	R	S	R	S	R	S	R
Pre-2020	12 (24%)	0	8 (16%)	0	0	24 (48%)	0	0	3 (6%)	3 (6%)
Post-2020	1 (0.78%)	2 (1.56%)	0	79 (61.72%)	0	2 (1.56%)	0	22 (17.19%)	2 (1.56%)	20 (15.63%)

^a S, susceptible; R, resistance.

Vaccine antigen genotype analysis

Among the eight key vaccine antigen genes analyzed, only allele type 1 was detected for ptxA, fim2, and fim3, leading to their exclusion from further analysis. Six antigen genotype combinations were identified (Table 3). Pre-2020, the most prevalent genotypes were ptxC2/prn2/ptxP3/fhaB2400_5550-1/tcfA2 (36%, 18/50) and ptxC1/prn1/ptxP1/fhaB2400_5550-3/tcfA2 (48%, 24/50). Post-2020, ptxC2/prn150/ptxP3/fhaB2400_5550-1/tcfA2 emerged as the predominant genotype (93.75%, 120/128).

TABLE 3.

Distribution of vaccine‐antigen genotype combinations, 2018–2024

	2018	2019	2021	2022	2023	2024
ptxC2/prn150/ptxP3/fhaB2400_5550-1/tcfA2	0	3 (11.54%)	29 (82.86%)	26 (96.30%)	11 (100%)	54 (98.18%)
ptxC2/prn2/ptxP3/fhaB2400_5550-1/tcfA2	8 (33.33%)	10 (38.46%)	2 (5.71%)	0	0	1 (1.82%)
ptxC1/prn1/ptxP1/fhaB2400_5550-3/tcfA2	13 (54.17%)	11 (42.31%)	3 (8.57%)	1 (3.7%)	0	0
ptxC1/prn166/ptxP1/fhaB2400_5550-3/tcfA2	1 (4.17%)	2 (7.69%)	0	0	0	0
ptxC2/prn149/ptxP3/fhaB2400_5550-1/tcfA2	2 (8.33%)	0	0	0	0	0
ptxC2/prn150/ptxP3/fhaB2400_5550-1/tcfA9	0	0	1 (2.86%)	0	0	0

The temporal shift in antigen genotype composition from pre- to post-2020 can be divided into two main transitions: a change from a roughly equal distribution between ptxC1/ptxP1/fhaB2400_5550-3 (46%, 23/50) and ptxC2/ptxP3/fhaB2400_5550-1 (54%, 27/50) to near-exclusive predominance of ptxC2/ptxP3/fhaB2400_5550-1 (96.88%, 124/128); a switch from predominantly prn1 (48%, 24/50) and prn2 (36%, 18/50) to overwhelmingly prn150 (94.53%, 121/128).

Age‐specific differences in B. pertussis after 2020

Post-2020, the ptxC2/prn150/ptxP3/fhaB2400_5550-1/tcfA2 vaccine antigen genotype dominated across all age groups (93.75%, 120/128), with no significant age-related differences (χ² = 10.48, P > 0.05) (Table 4). Concomitantly, macrolide resistance showed a similar age-independent pattern, with only three isolates retaining susceptibility and no significant variation in resistance rates (χ² = 3.68, P > 0.05) (Table 4).

TABLE 4.

Characteristics of B. pertussis isolates by age group, post-2020

	≤36 months (n = 23)	37 months to 18 years (n = 72)	>18 years (n = 33)	Total (all ages, N = 128)
Vaccine‐antigen genotypes
ptxC1/prn1/ptxP1/fhaB2400_5550-3/tcfA2	2 (8.70%)	2 (2.78%)	0	4 (3.1%)
ptxC2/prn150/ptxP3/fhaB/2400_5550-1/tcfA2	18 (78.26%)	69 (95.83%)	33 (100%)	120 (93.8%)
ptxC2/prn150/ptxP3/fhaB-2400_5550-1/tcfA9	1 (4.35%)	0	0	1 (0.8%)
ptxC2/prn2/ptxP3/fhaB-2400_5550-1/tcfA2	2 (8.70%)	1 (1.39%)	0	3 (2.3%)
MLVA types
MT28	14 (60.87%)	38 (52.78%)	27 (81.82%)	79 (61.7%)
Others	9 (39.13%)	34 (47.22%)	6 (18.18%)	49 (38.3%)
Macrolide resistance
S	2 (8.70%)	1 (1.39%)	0 (100%)	3 (2.3%)
R	21 (91.30%)	71 (98.61%)	33 (0%)	125 (97.7%)

MT28 emerged as the dominant lineage across all age strata post-2020, with all isolates exhibiting macrolide-resistant and carrying the ptxP3 allele (hereafter referred to as MT28-ptxP3-MRBP). Overall, this lineage accounted for 61.72% (79/128) of post-2020 isolates, with a significantly higher prevalence in adults (≥19 years) at 81.82% (27/33) than in children and adolescents (37 months to 18 years) at 52.78% (38/72) (χ² = 8.092, P < 0.01) (Table 4). No significant difference was observed between the ≥19 years group and infants aged ≤36 months (60.87%; P = 0.082). When infants and school‐age children/adolescents were combined into a single <19-year group and compared with the ≥19-year group, the age‐related difference in MT28-ptxP3-MRBP prevalence remained significant (P < 0.01).

Phylogenetic analysis

Phylogenetic analysis of 178 B. pertussis isolates revealed two distinct clades defined by ptxP1 and ptxP3 (Fig. 2). Within the ptxP3 clade, green-highlighted branches denote isolates carrying the A2047G resistance mutation, which closely coincides with the prn150 genotype. No significant age‐related clustering was observed. The phylogeny of 615 Chinese B. pertussis isolates showed no provincial-level clustering, with Shanghai isolates genetically similar to the national population (Fig. 3). The MT28-ptxP3-MRBP lineage was first detected in Beijing in 2019 (SRR27796581, SRR27796588) and 2020 (SRR27796580), with the earliest Shanghai isolate of this lineage also in 2020 (SRR27796580). Global phylogenetic tree (Fig. 4) reveals strong geographic clustering of Chinese isolates. Outside China, only four MT28-ptxP3-MRBP isolates have been identified—ERR13476619 (2024, France), DRR631445 (2024, Japan), SRR32181461 (2024, USA), and SRR32181462 (2024, USA). Among all 1,774 B. pertussis genomes analyzed, MT60 represents 1.41% (25/1,774), all post-2020; 22 of these (88%, 22/25) derive from this study, and the remaining three from Zhejiang Province, China.

Fig 2.

Maximum-likelihood phylogeny of 178 B. pertussis isolates from Shanghai (2018–2024). Branches are colored by ptxP allele clade: ptxP1 (magenta) and macrolide-resistant ptxP3 (green).

Fig 3.

Branches belonging to the ptxP1 clade are shown in magenta. No clear province-level clustering was observed in any of the three phylogenetic trees. (A) Phylogeny of 615 Chinese B. pertussis genomes. MT28-ptxP3-MRBP isolates are highlighted in blue for those collected pre-2020 and in green for those collected in 2020. (B) Phylogeny of 281 Chinese B. pertussis genomes collected during 2016–2020. (C) Phylogeny of 334 Chinese B. pertussis genomes collected during 2021–2024.

Fig 4.

Branches colored in purple represent the non-ptxP3 clade. (A) Phylogeny of 1,774 global B. pertussis isolates. MT28-ptxP3-MRBP isolates from outside China are marked in blue. (B) Phylogeny of 1,215 global B. pertussis genomes collected during 2016–2020. (C) Phylogeny of 559 global B. pertussis genomes collected during 2021–2024.

DISCUSSION

In this study, we systematically analyzed the epidemiological characteristics, antimicrobial susceptibility, and genomic profiles of 178 B. pertussis isolates collected in Shanghai from patients of all age groups between 2018 and 2024. Our findings indicate that, in the pre-2020 period, pertussis cases occurred predominantly in infants aged ≤36 months, whereas in the post-2020 period, they were mainly observed in school-age children and adolescents aged 37 months to 18 years. The MRBP positivity rate rose from 58.33% in 2018 to 94.29% in 2021, and the MT28 and ptxA1/ptxC2/prn150/ptxP3/fhaB2400_5550-1/fim2-1/tcfA2/fim3-1 genotypes rapidly became dominant post-2020. Importantly, this study is the first to demonstrate age-specific differences in the prevalence of the MT28-ptxP3-MRBP lineage in Shanghai after 2020.

In 2011, the first case of MRBP in China was reported in Shandong Province (19). Li et al. conducted a multicenter study across northern and southern China from 2014 to 2016 and found that the MRBP positivity rate reached 91.1% (194/213) in northern isolates versus 64.3% (36/56) in southern isolates (20). Fu et al. reported a 57.5% (81/141) MRBP rate in Shanghai during 2016–2017 (14). However, in the post-2020 period, MRBP positivity exceeded 97% in both northern and southern regions (11, 12, 21). In our study, MRBP positivity remained at 46% pre-2020 and rose to 97.66% post-2020, consistent with the overall southern data and the findings of Fu et al. (11) at Fudan Pediatric Hospital, Shanghai. Despite the high MRBP rates in China, much lower proportions have been reported elsewhere. For example, studies in France (June 2023 to May 2024) and Finland (April to October 2024) found MRBP rates of only 1.5% (1/67) and 0.22% (1/462), respectively (22, 23). Among the 1,159 non-Chinese B. pertussis genomes included in our analysis, only nine (0.78%, 9/1,159) harbored the 23S rRNA A2047G mutation. Nevertheless, this does not imply that MRBP is confined to China. For instance, although India accounts for 26.5% of global pertussis cases (24), relatively few isolates have been characterized and data on antimicrobial susceptibility or A2047G mutation prevalence remain scarce (25), leaving the true global burden of MRBP uncertain.

Currently, approved aP vaccines contain up to five bacterial antigens: Pertussis Toxin (PTX) and four adhesion proteins, including Filamentous Hemagglutinin (FHA), Pertactin (PRN), and Fimbriae Types 2 and 3 (FIM2/3) (26). The genotype of the Chinese B. pertussis vaccine strain is ptxA2/ptxC1/prn1/ptxP1/fhaB2400_5550-1/fim2-1/tcfA2/fim3-1 (27). The approved aP vaccines in China are mainly divided into two types: one containing PTX and FHA (two-component vaccine), and the other containing PTX, FHA, and PRN (three-component vaccine) (28). This also partially explains why, in this study, all strains, except for one identified as a new tcfA9 variant, were consistent with the vaccine strain, featuring fim2-1/tcfA2/fim3-1.

Two studies conducted in Beijing, China, reported that PRN in the region is predominantly prn2, with no detection of prn150 (0/288 and 0/60) (12, 29). However, in a study by Zhou et al. (27) also conducted in Beijing, 100% (44/44) of B. pertussis isolates exhibited prn150, consistent with the findings in this study and the overall trend observed in Shanghai (11). Both prn150 and prn2 are genetically distinct from the vaccine strain, and prn-deficient strains are widely present globally (21, 27, 30, 31). FIM and FHA play crucial roles in allowing B. pertussis to evade immune surveillance during infection and to establish colonization in the respiratory tract (32). Moreover, a mouse model demonstrated that mutants lacking FHA and FIM showed significantly reduced infectivity in the nasal cavity (33). In this study, the proportion of fhaB2400_5550-1 (vaccine strain genotype) increased from 46% (23/50) pre-2020 to 96.88% post-2020. Although this shift was unexpected, similar reports have emerged in several other studies in China (27, 28). Among the 615 Chinese B. pertussis strains included in the analysis, the proportions of fhaB2400_5550-1 were 34.42% (95/276) pre-2020 and 83.23% (278/334) post-2020.

Pre-2020, B. pertussis strains carrying the ptxP3 allele had been reported in China, but ptxP3-MRBP was very rare, with ptxP1-MRBP being predominant (13, 20, 34, 35). In this study, all 23 ptxP3-positive B. pertussis isolates from the pre-2020 period were macrolide-susceptible, while all 27 ptxP1-positive isolates were resistant. Post-2020, pertussis cases in China surged, accompanied by rapid expansion of ptxP3-MRBP, which became the dominant lineage (12, 29, 36), consistent with the overall trends observed in this study. Additionally, this study found that MT28-ptxP3-MRBP was not first detected in Shanghai; it was identified in Beijing as early as 2019. Phylogenetic analysis of 1,159 international B. pertussis isolates included in this study revealed that ptxP3 was predominant globally even before 2020. Notably, MT28-ptxP3-MRBP was first detected in 2024 in Japan, France, and the USA, spanning three continents, indicating a potential for cross-border spread. Miettinen et al. (22) also detected one ptxP3-MRBP isolate in a study of 462 B. pertussis isolates collected from different regions of Finland between April and October 2024, though MLVA typing data were not provided. The global phylogenetic tree reveals that Chinese B. pertussis strains exhibit strong geographic clustering, with MT28-ptxP3-MRBP emerging as a dominant lineage in China from 2019 to 2021 in a remarkably short period, which may indicate a high risk of international spread following its emergence.

In this study, we first observed significant differences in the prevalence of MT28-ptxP3-MRBP across different age groups after 2020. The proportion of MT28-ptxP3-MRBP was significantly higher in the ≥19 years group compared to the 37 months to 18 years group. No significant difference was observed between the ≥19 years group and the ≤36 months group, possibly due to the smaller sample size in the latter group. However, a significant difference remained between the ≥19 years group (60.87%, 27/33) and the <19 years group (54.74%, 52/95), suggesting that MT28-ptxP3-MRBP may have a higher transmission advantage in older age groups.

In this study, MT60, a new genotype emerging post-2020, accounted for 17.19% (22/128) and was the second most common genotype after MT28. However, aside from the samples in this study, only three (0.19%, 3/1596) B. pertussis isolates from Zhejiang Province, China, displayed the MT60 genotype. All MT60 isolates were ptxP3-MRBP, and they showed high homology with MT28-ptxP3-MRBP in the phylogenetic tree, indicating that this lineage should be closely monitored.

There are some limitations to this study: first, our external comparator relied solely on sequences archived in NCBI, rather than on a complete, population-based data set like the one assembled in this study for the selected geographic region; as a result, the public data are vulnerable to selection/reporting bias and may lack geographic or temporal representativeness. Second, the isolation and culture of B. pertussis are challenging, and there may be some bias in the selection of isolates.

Conclusion

In summary, this study is the first to identify significant differences in the prevalence of MT28-ptxP3-MRBP across age groups, suggesting that this B. pertussis genotype may have a transmission advantage in older populations. Moreover, MT28-ptxP3-MRBP has begun to appear in countries outside of China, indicating a risk of international spread. Therefore, we recommend further strengthening active surveillance across all age groups and closely monitoring the global spread of MT28-ptxP3-MRBP to inform and optimize pertussis control strategies.

HIGHLIGHTS

1. After 2020, MT28-ptxP3-MRBP lineage rapidly dominated, comprising 61.7% of isolates.

2. MT28-ptxP3-MRBP exhibits a significant transmission advantage among older individuals.

3. The primary affected group shifted from ≤36 months (pre-2020) to 37 months–18 years (post-2020).

4. Macrolide resistance rose from ≤50% pre-2020 to nearly 100% post-2020, with all resistant isolates carrying the A2047G mutation.

DATA AVAILABILITY

All whole-genome sequencing data from the study are available through the National Center for Biotechnology Information Sequence Read Archive via the project accession number PRJNA1295129 and will become publicly available once accession formalization resumes.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/jcm.01064-25.

Table S1. jcm.01064-25-s0001.xlsx.

Baseline characteristics of the study strains.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.