The evolutionary and molecular history of a chikungunya virus outbreak lineage

Abstract

In 2018-2019, Thailand experienced a nationwide spread of chikungunya virus (CHIKV), with approximately 15,000 confirmed cases of disease reported. Here, we investigated the evolutionary and molecular history of the East/Central/South African (ECSA) genotype to determine the origins of the 2018-2019 CHIKV outbreak in Thailand. This was done using newly sequenced clinical samples from travellers returning to Sweden from Thailand in late 2018 and early 2019 and previously published genome sequences. Our phylogeographic analysis showed that before the outbreak in Thailand, the Indian Ocean lineage (IOL) found within the ESCA, had evolved and circulated in East Africa, South Asia, and Southeast Asia for about 15 years. In the first half of 2017, an introduction occurred into Thailand from another South Asian country, most likely Bangladesh, which subsequently developed into a large outbreak in Thailand with export to neighbouring countries. Based on comparative phylogenetic analyses of the complete CHIKV genome and protein modelling, we identified several mutations in the E1/E2 spike complex, such as E1 K211E and E2 V264A, which are highly relevant as they may lead to changes in vector competence, transmission efficiency and pathogenicity of the virus. A number of mutations (E2 G205S, Nsp3 D372E, Nsp2 V793A), that emerged shortly before the outbreak of the virus in Thailand in 2018 may have altered antibody binding and recognition due to their position. This study not only improves our understanding of the factors contributing to the epidemic in Southeast Asia, but also has implications for the development of effective response strategies and the potential development of new vaccines.

Fig 1. Phylogenetic and Phylogeographic Analysis of the ECSA CHIKV samples included in the present study.

Phylogenetic tree of the CHIKV ECSA genotype. Strains are colour-coded by CHIKV lineage (where light green represents ECSA 1, red ECSA 2, and blue IOL). Strains sequenced in this study are indicated by an ^x. Black values at red circles indicate branching times, grey values display the mutational speed. Major (basal) branches with posterior probabilities of ≥0.95 are indicated by an *. Amino acid substitutions at different alphabetically characterized nodes are indicated in green boxes if affecting the entire IOL and in orange boxes if affecting only certain clades. Countries are colour-coded according to the phylogenetic tree, and suggested transmission events are shown with arrows on the global map. Uninterrupted lines on the map show data from our investigation, while dashed lines display previous assumptions of spread. The base layer of the map was retrieved from Wikimedia Commons and modified under a Creative Commons Attribution 4.0 International license (https://commons.wikimedia.org/wiki/File:Blank_World_Map_%28in_the_year_2024_and_with_borders_and_blue_oceans%29.png).

Fig 2. Emergent mutations in the Nsp1-Nsp2-Nsp4 replication complex.

A) Representative view of the CHIKV replication complex. Nsp1 is in grey, Nsp2 in yellow and Nsp4 in pink colour. Structural models for visualization were obtained from PDBid 7y38 [45] and 4ztb [46]. In all panels the mutations arising in the IOL are shown as purple sticks. Note that highlighted amino acids correspond to the position of the indicated mutations but not always to the actual amino acid involved in the mutation process. (see methods for more information) B) Close-up view, highlighting the mutations on Nsp1. The GTP and ATP which are cofactors for the Nsp1 are also shown as orange sticks. C) Model of Nsp2 with indicated mutations. The catalytic residues are shown as green sticks and the RNA fragment bound to the helicase domain is shown as orange sticks. D) Detailed view showcasing the mutations on Nsp4. E) Summary of all mutations found in the replication complex.

Fig 3. Emergent mutations in the E1-E2-E3 spike complex.

A) Representative view of the Spike trimer. Structural models for visualization were obtained from PDBid 6jo8 [47]. E1 is coloured cyan, E2 blue and E3 is green. The receptor Mxra8 is included to aid the visualization of the receptor binding interface and is coloured brown. The position of the membrane at the base of the spike is indicated. All emergent mutations are shown as sticks and coloured purple as in Fig 2. B) Close-up view, highlighting the mutations on E1. The domains I, II and III are indicated. C) enlarged model of E2 highlighting emergent mutations. Domains A, B and C are indicated. D) Close-up view, showcasing the mutations on E3. E) Summary of all mutations found in the spike complex. Asterix denotes mutations that are not part of the analysed structure and are therefore not visualized.

Fig 4. Emergent mutations in the Nsp3, 6K and CP proteins.

All mutations were mapped onto the ColabFold-predicted structural models. Positions of mutations are shown as purple sticks, and the overall colouring of the proteins is according to the pLDDT score indicating the confidence of the prediction. Blue colour signifies highest prediction confidence and orange lowest as shown by the legend. A) Emergent mutations in Nsp3 protein. Macro and Zinc-binding domains are indicated. B) Predicted structure of 6K protein with highlighted V6I mutation. C) Capsid protein model with highlighted mutations in the disordered N-terminal tail of the protein. D) Summary of all emergent mutations found in the Nsp3, 6K and capsid proteins.