Figure 1

Figure 1. Phylogenomics of clinical and environmental Vibrio cholerae O1 El Tor isolates from the Lake Tanganyika basin, Africa. A) Maximum-likelihood phylogeny of 1,366 seventh pandemic V. choleraeO1...

We analyzed 96 V. cholerae O1 isolates collected during 2015–2020 in DRC (86 clinical isolates, including 39 collected in 2018–2020) and Tanzania (10 environmental isolates from fish and lake water) (Appendix 1; Appendix 2 Table 1). We subjected the isolates to antimicrobial susceptibility testing, whole-genome sequencing, genomic characterization, and phylogenetic analyses, as previously described (11,12) (Appendix 1). We performed a phylogenetic analysis of these genomes within a global collection of 1,366 7PET V. cholerae O1 genomes (Appendix 2 Table 2), including another 130 genomes from DRC collected during 1984–2017. We based the final maximum-likelihood phylogenetic tree on 10,352 single-nucleotide variants distributed over the nonrepetitive, nonrecombinant core genome (Figure 1, panel A).

Figure 2

Figure 2. A) Spatiotemporal dynamics of the AFR10 clades of Vibrio choleraeO1 in the African Great Lakes Region, Africa, 1998–2020. Circle size indicates the number of isolates at the location...

Phylogenetic analysis of the 96 genomes of V. cholerae O1 isolates showed that all belonged to 7PET sublineage AFR10 (Figure 1, panel A). Within the limits of our sampling, sublineage AFR5, which circulated actively in the AGLR during the 1980s–1990s (11), appears to be extinct in the region, whereas sublineage AFR13, reported in 2015 in Uganda and Tanzania (12), has not yet spread to DRC. Since 1998, the endemicity of the AFR10 sublineage in the Tanganyika basin and surrounding countries has led to microevolution; an analysis of 357 AFR10 genomes from our global dataset revealed the presence of 5 clades, AFR10a–AFR10e (Figure 1, panel B; Appendix 1 Figure 1). Clades AFR10a, AFR10b, and AFR10c were mostly associated with the eastern AGLR countries. The isolates of these clades were of sequence type (ST) 69 (Figure 1, panel B). Clades AFR10d and AFR10e predominated in DRC and the Lake Tanganyika basin. Clade AFR10d is of ST515, essentially Inaba serotype, and was widespread in DRC and neighboring countries (Figure 2, panel A), as previously reported (13). It was the only clade found in the Lake Kivu and Lake Edward basins. AFR10e strains are of ST69, Ogawa serotype, and were essentially restricted to the Lake Tanganyika basin, confirming previous findings (13). A further pangenome analysis of the AFR10 isolates revealed no clade-specific gain or loss of genes (Appendix 1 Figure 2). AFR10e strains have gradually replaced AFR10d strains in the region since 2014; all V. cholerae O1 strains obtained from the Tanganyika basin by 2017, as well as those obtained from the lake itself in 2018 and 2019, were AFR10e strains (Figure 2 panel B). Epidemiologic studies identified cholera hotspots in the AGLR as a source of major countrywide outbreaks reaching the capital, Kinshasa, and the Atlantic coast, via the Congo River, in 2011, 2012, and 2016 (5). These outbreaks were caused primarily by clade AFR10d, which has a wider geographic distribution than clade AFR10e (Figure 2) (14).

One striking characteristic of AFR10e isolates was the presence of a mutation in the quinolone resistance–determining region of the topoisomerase IV subunit A gene, parC (S85L), that led to higher MIC values (0.25–0.38 mg/L) for ciprofloxacin. These isolates, which have a lower susceptibility to ciprofloxacin than wild-type populations, would be classified as either resistant (MIC 0.38 mg/L) or susceptible (MIC 0.25 mg/L) in accordance with clinical breakpoints published by the European Committee on Antimicrobial Susceptibility Testing (https://www.eucast.org/clinical_breakpoints) in 2022, as inferred from the study of Vibrio strains and not Enterobacteriaceae strains. This parC mutation, sporadically reported in other AFR10 clades (Table; Figure 1, panel B; Appendix 2 Table 1), has been a distinctive characteristic of AFR10e isolates since 2014 (Figure 2, panel B). It was the second mutation affecting susceptibility to quinolones and fluroquinolones to be found in this AFR10e clade; the first was a mutation in the DNA gyrase subunit A gene, gyrA (S83I), present in all AFR10 isolates. This additional mutation does not seem to be associated with the specific use of fluoroquinolones for treating cholera outbreaks, because the antimicrobial drugs used for first-line cholera control in DRC are tetracyclines and macrolides, to which AFR10e isolates remain susceptible. Instead, the mutation may result from widespread self-medication with antimicrobials, a common practice in many sub-Saharan Africa countries including DRC (15).

All 96 isolates analyzed had known mutations of the VC_0715 and VC_A0637 genes conferring nitrofuran resistance (Table), consistent with previous findings (2). The isolates also carried the SXT/R391 genomic element ICEVchInd5, encoding resistance to streptomycin (strAB), sulfonamides (sul2), chloramphenicol (floR), trimethoprim and the O/129 vibriostatic agent (dfrA1), and trimethoprim–sulfamethoxazole (sul2 and dfrA1), with concordance between the phenotypic and genotypic data (Table; Appendix 2 Table 1).

We found that the cholera outbreaks in the eastern part of DRC during 2001–2020 were caused by V. cholerae O1 sublineage AFR10, which was introduced into East Africa from South Asia in the late 1990s. The AFR13 sublineage was already reported in 2015 in Tanzania, including the city of Kigoma, was located on the shore of Lake Tanganyika but had not been detected in DRC as of 2022. The AFR10 isolates of this region belong principally to 2 clades, AFR10d (Inaba, ST515) and AFR10e (Ogawa, ST69). AFR10d was responsible for outbreaks reported in the western part of DRC in 2011–2017 and neighboring countries; AFR10e (Ogawa, ST69) was restricted to the Lake Tanganyika basin, in which reduced susceptibility to ciprofloxacin has been seen since 2014. Lake Tanganyika seems to serve as a transmission channel, favoring the establishment of AFR10e in local human populations. Further investigation, including studies of population movement, should reveal why AFR10e clade has remained within the Lake Tanganyika basin. The replacement of other clades by this antimicrobial-resistant clade in this area highlights the need for more systematic documentation of antimicrobial drug use and the implementation of adapted stewardship programs, particularly in outbreak responses. Overall, these findings highlight the need for continuous genomic surveillance and for coordinated communication between countries for effective interventions.