Vital Surveillances: Phylogenetic Analysis of Serogroup O5 Vibrio cholerae that Caused Successive Cholera Outbreaks — Guangdong Province, China, 2020–2021

Stools or anal swabs from patients and asymptomatic persons were collected. Environment samples in the kitchen and environmental water were collected. Oral consent was obtained from each eligible patient.

All 66 strains sequenced in the study were recovered on Luria-Bertani (LB) agar (in a −80 °C freezer). A single colony was transferred to LB broth and was incubated at 37 °C with shaking at 200 revolutions per minute (RPM). Genomic DNA was extracted from overnight cultures with a Qiagen kit (QIAamp Mini Kit, Hilden, Germany) according to the manufacturer’s instructions.

A total amount of 0.2 μg DNA per sample was used as input material for the DNA library preparations. Sequencing library was generated using NEB Next^® Ultra™ DNA Library Prep Kit for Illumina (NEB, Ipswich, USA) following the manufacturer’s recommendations and index codes were added to each sample. Briefly, genomic DNA sample was fragmented by sonication to a size of 350 bp. Then DNA fragments were end-polished, A-tailed, and ligated with the full-length adapter for Illumina sequencing, followed by further polymerase chain reaction (PCR) amplification. After PCR products were purified by AMPure XP system (Beckman Coulter, Beverly, USA), DNA concentration was measured by Qubit^® 3.0 Flurometer (Invitrogen, Carlsbad, USA), and libraries were analyzed for size distribution by NGS3K/Caliper and quantified by real-time PCR (3 nM).

The clustering of the index-coded samples was performed on a cBot Cluster Generation System using Illumina PE Cluster Kit (Illumina, San Diego, USA) according to the manufacturer’s instructions. After cluster generation, the DNA libraries were sequenced on Illumina NovaSeq6000 platform and 150 bp paired-end reads were generated.

Short reads were assembled de novo into contigs and scaffolds and a whole-genome alignment for all of the genomes investigated was built in this study. The whole-genome sequence of N16961 was used as the reference to call SNVs.

Gastroenteritis outbreaks were reported in Qingyuan City, Guangdong Province in 2020 and 2021. The gastroenteritis outbreak in 2020 occurred on September 9 and lasted for 18 days in a vocational-technical college. A total of 137 cases were reported. In these cases, 97.7% of them complained of mild diarrheal and 62.3% complained of abdominal pain. A total of 303 samples from patients, asymptomatic persons, environment samples in the kitchen, and environmental water were collected and sent to the laboratory for screening of norovirus, rotavirus, Salmonella, pathogenic E. coli, Shigella, Campylobacter jejuni, Vibrio parahaemolyticus, and Vibrio cholerae. V. cholerae strains were isolated from 32 samples. In 2021, another outbreak occurred during the period from September 28 to September 30 in a vocational-technical college, which is about 2 kilometers away from the former one in Qingyuan City. A total of 79 patients in this college were reported to have diarrhea. A total of 47 anal swabs of the case-patients were collected and sent to the diagnostic laboratory for detection of norovirus, rotavirus, Salmonella, pathogenic E. coli, Shigella, Campylobacter jejuni, Vibrio parahaemolyticus, and Vibrio cholerae. Of the 47 samples, 34 anal swabs of the case-patients were positive for V. cholerae by PCR assay. All of the strains did not agglutinate with either O1 or O139 specific antiserum. Therefore, we determined that these strains were non-O1/non-O139 V. cholerae. In both of these two outbreaks, water has been suspected as the infection source; however, no V. cholerae was isolated from water samples.

A maximum likelihood (ML) tree was reconstructed based on the SNVs identified in the non-repetitive and non-recombinant core-genome of 40 publicly available representative V. cholerae and 66 V. cholerae isolated in these two outbreaks using IQ-TREE. It showed that 97% (64/66) of the V. cholerae isolated in these two outbreaks formed a tight cluster and the left two were far away from this cluster and differed from each other (Figure 1A). In order to investigate the detailed relationship among V. cholerae in these two outbreaks, we built an ML tree based on the SNVs in the non-repetitive and non-recombinant core-genome of the 64 V. cholerae. The V. cholerae isolated in 2021 formed 2 clusters that included 25 and 8 strains, respectively, and the left one was in a cluster including another 6 V. cholerae isolated in 2020. We also detected the presence of the virulence genes and pathogenic islands that were common in the pandemic clones of V. cholerae. It showed that none of the cholera toxin genes (ctxA and ctxB), the CTXΦ, the Vibrio pathogenic island (VPI)1 and 2, and the Vibrio seventh pandemic island I and II (VSPI and VSPII) were detected in all of the 64 V. cholerae (Figure 1B). However, we detected T3SS in all of the 64 V. cholerae (Figure 1B). Moreover, we detected qnrVC5, dfrA31, and varG in almost all of the 64 strains except that qnrVC, instead of qnrVC5, was detected in 1 strain isolated in 2020. BLAST analysis indicated that qnrVC5 and dfrA31 were located on an integron. In order to determine the serogroup, we tried to agglutinate these strains with O1 and O139 antigenic serum, it turned out that these strains were not O1 or O139 serogroup. The PCR result gave the same results. As we could not get the antigenic serum of other serogroups of V. cholerae, we extracted the O-PS sequence of these strains and did BLAST analysis against GenBank. The results showed that the 64 V. cholerae in the same tight cluster in Figure 1A showed best hits to the O-PS sequence of a serogroup O5 V. cholerae (Figure 1C). Therefore, we concluded that these 64 strains were serogroup O5. As to the left 2 V. cholerae, we could not determine their serogroups. We also determined that the ST of these serogroup O5 V. cholerae was ST88, and we could not get the ST of the left 2 non-O5 genomes as no information was available in the database.

One important characteristic of the 2 cholera outbreaks was that they were reported in 2 successive years in almost the same place at almost the same time. The genomic analysis indicated that those strains isolated in two outbreaks were closely related. V. cholerae is a natural habitat of estuary water, especially in places rich in plankton (1). These pieces of information indicated that an unknown contaminated source existed. Therefore, enhanced surveillance, which focuses on food and environmental water, should be carried out next year.

Till now, the determination of serogroup of V. cholerae mainly relies on the agglutination reaction. However, due to the poor availability and cost of serogroup-specific serum, it is difficult to get information on the serogroup of V. cholerae other than O1 and O139. As the price of whole-genome sequencing decreased rapidly in these years, it is affordable to carry out WGS of the representative strains that caused outbreaks (7). Moreover, the component and structure of the O-PS coding sequences determined the various serogroups (2). Therefore, it is possible to establish a correlation between the serogroups and the O-PS sequences. A similar study of Salmonella has shown consistency between traditional serovar typing and STs (8). This could be achieved by the WGS of a reference collection of representative strains of different serogroups of V. cholerae. Ultimately, a molecular-serogroup scheme could replace the existing serogroup typing scheme.

The reported cholera cases decreased dramatically after the year 2010 in China, which is an indicator that cholera has been under control in this country (6). The cholera cases caused by pathogenic O1 and O139 V. cholerae were rare since the year 2010; however, those caused by non-toxigenic O1 and O139 or non-O1/non-O139 V. cholerae were reported occasionally. For example, cases caused by non-O1/O139 V. cholerae were reported every year since 2012 in Guangdong (data not shown). Therefore, the strategy for cholera prevention and control in China nowadays should be adjusted and attention should also be paid to non-toxigenic O1 and O139 or non-O1/non-O139 V. cholerae, particularly those that carry T3SS and/or T6SS (9). As the non-toxigenic V. cholerae, especially those carrying the TCP gene cluster and those carrying pre-CTXΦ, could be converted to toxigenic strains by transduction, more attention should also be paid to such strains.

No conflicts of interest.

Figure 1. 

Phylogenetic analysis and molecular characteristics of serogroup O5 V. cholerae isolated in two outbreaks (A) Phylogenetic tree of non-O1/non-O139 V. cholerae. (B) Phylogenetic tree of serogroup O5 V. cholerae. (C) Illustration of the O-PS biosynthesis gene locus in V. cholerae isolated in these two outbreaks.

Notes: Maximum likelihood (ML) tree of 66 non-O1/non-O139 V. cholerae isolated in the 2 outbreaks, and the publicly available representative V. cholerae was constructed based on the SNVs in the core genome. Those isolated in this study were in red. This ML tree was constructed based on the SNVs in the non-repetitive and non-recombinant core-genome of 106 non-O1/non-O139 V. cholerae. O1 El Tor strain N16961 was used as reference. The ML tree was constructed on the SNVs identified in the non-repetitive, non-recombinant core-genome of the 66 most closely-related serogroup O5 V. cholerae. The presence and absence of virulence genes and mobile genetic elements usually detected in pandemic V. cholerae and antibiotic resistance genes were annotated on the right of the ML tree. Comparison between O-PS biosynthesis gene locus in serogroup O5 V. cholerae CO545 and that in a representative strain which was isolated in the outbreak in 2021. Abbreviations: MGEs=mobile genetic element; ARGs=antibiotic resistance genes; SNVs=single nucleiotide variation.