Genomic epidemiology of coxsackievirus A24 variant during the 2024 acute hemorrhagic conjunctivitis outbreak in Coastal Kenya

Fig 1. Temporal distribution of CVA24v positivity across the four surveillance platforms. (A) Temporal distribution of CVA24v positives at the community-based respiratory surveillance. (B) Temporal distribution of CVA24v positives in the in-patient pneumonia surveillance. (C) Temporal distribution of CVA24v positives in the out-patient acute respiratory infection surveillance. (D) Temporal distribution of CVA24v positives in the in-patient diarrheal etiology surveillance. The bar plot shows the percentage of samples positive for CVA24v, stratified by the detection or confirmatory methods used, while the black line represents the total number of samples tested.

Fig 2. Genetic diversity, root-to-tip divergence, and demographic reconstruction of global CVA24v strains sampled between 1970 and 2024 (A) ML phylogenetic tree of 194 whole-genome sequences selected from the global (129) and Kenya (n = 65) data sets. (B) ML phylogenetic tree of 80 whole-genome sequences from Kenya, Mayotte and Malawi. Tip labels are colored by country of sampling. (C) Root-to-tip regression showing accumulation of genetic divergence in CVA24v over time. The phylogeny was rooted using TreeTime. (D) ML skyline reconstruction of genetic diversity (presented as the effective population size, Ne) of CVA24v over time, assuming 50 generations per year. The solid line represents the median estimate of Ne, while the shaded area shows the approximate 95% confidence bounds (±2 standard deviations of the log-likelihood). (E) World map showing the distribution of countries that have deposited CVA24v WGS in GenBank as of 01 November 2025. Color intensity indicates the number of sequences contributed by each country (using a standard “hotter” scale for higher density). The map was generated in R. The base layer was obtained from the Humanitarian Data Exchange (Global Administrative…

Fig 3. Genetic diversity of globally circulating CVA24v. (A) ML phylogenetic tree of global CVA24v VP1 sequences sampled between 1970 and 2024, showing the clustering of major CVA24v genotypes. (B) ML phylogenetic tree of genotype IV CVA24v partial VP1 sequences, showing clustering into distinct sub-groups (GIV-C1 to GIV-C22) using a 0.05 nucleotide divergence threshold. The black dashed box indicate the East African sequences while the red dashed box indicates sequences responsible for the Asia outbreak. (C) Temporal distribution of CVA24v sub-groups detected between 2015–2024, classified using the established 0.05 VP1 pairwise divergence threshold for genotype IV (GIV) sub-group delineation.

Fig 4. Genetic diversity of Kenya, Malawi and Mayotte CVA24v sequences. (A) ML phylogenetic tree of the Kenya 2024 CVA24v whole-genome sequences. Tips are colored according to the major clade identified (C1/1/24 in blue, C1/2/24 in red and C1/3/24 in tan). Different shapes denote the country where the virus was sampled. (B) Pairwise SNP-distance distribution plots of C1/1/24, C1/2/24 and C1/3/23 sequences. (C) Temporal distribution of the C1/1/24 and C1/2/24 clusters in coastal Kenya.

Fig 5. Amino acid variability across the CVA24v genome. (A) Summary of amino acid changes across different genomic regions of the Kenya CVA24v, comparing the prototype strain (D90457.1/Singapore/1970). (B) Amino acid differences between sequences from Kenya, Malawi, and Mayotte, with the Kenyan sequence acting as the reference. (C) Global sequences showing variable amino acid positions and their distribution across coding regions. (D) Variable amino acid positions in the Kenyan sequences and their distribution across coding regions. Peaks represent regions with higher amino acid variability.

Fig 6. Nucleotide similarity between CVA24v and other species C enteroviruses. (A) Nucleotide similarity plots between Kenyan CVA24v genomes and other selected sequences from enterovirus C sequences obtained from GenBank as of 15th August 2025. The query sequence is one of the 245 from this study (accession number PX659173) and is compared with 2,418 Enterovirus Coxsackiepol sequences, including poliovirus (PV1, PV2, PV3) and non-polio EV-C reference sequences. (B) RDP5 BOOTSCAN analysis showing CVA24v as the major parent and PV2 as the closest possible minor parent. The analysis was based on a 200-nucleotide sliding window with a 20-nucleotide step size.

Fig 7. ML phylogenetic trees of CVA24v and other Enterovirus C (EV-C) genomes demonstrating the 3Dpol intra-species recombination event. (A) Phylogeny based on nucleotide positions 1–7400 (complete genome). (B) Phylogeny based on nucleotide positions 1–6080 (upstream of the inferred recombination breakpoint), showing that the 2024 East African CVA24v outbreak sequences cluster as a distinct monophyletic group within Genotype IV. (C) Phylogenetic based on nucleotide positions 6081–7400 (downstream of the inferred recombination breakpoint, corresponding to the 3Dpol region). A major topological incongruence is observed: the 2024 East African outbreak sequences and Uganda 2017 are phylogenetically distinct from other CVA24v lineages and instead cluster with recombinant West African PV2-related sequences. Bootstrap support values ≥99% are shown for internal nodes.