The evolutionary dynamics of DENV 4 genotype I over a 60-year period

Abstract

Dengue virus serotype 4 (DENV 4) has had a relatively low prevalence worldwide for decades; however, likely due to data paucity, no study has investigated the epidemiology and evolutionary dynamics of DENV 4 genotype I (DENV 4-I). This study aims to understand the diversity, epidemiology and dynamics of DENV 4-I. We collected 404 full length DENV4-1 envelope (E) gene sequences from 14 countries using two sources: Yunnan Province in China (15 strains during 2013-2016) and GenBank (489 strains up to 2018-01-11). Conducting phylogenetic and phylogeographical analyses, we estimated the virus spread, population dynamics, and selection pressures using different statistical analysis methods (substitution saturation, likelihood mapping, Bayesian coalescent inference, and maximum likelihood estimation). Our results show that during the last 60 years (1956-2016), DENV 4-I was present in mainland and maritime Southeast Asia, the Indian subcontinent, the southern provinces of China, parts of Brazil and Australia. The recent spread of DENV 4-I likely originated in the Philippines and later spread to Thailand. From Thailand, it spread to adjacent countries and eventually the Indian subcontinent. Apparently diverging around years 1957, 1963, 1976 and 1990, the different Clades (Clade I-V) were defined. The mean overall evolution rate of DENV 4-I was 9.74 (95% HPD: 8.68-10.82) × 10-4 nucleotide substitutions/site/year. The most recent common ancestor for DENV 4-I traces back to 1956. While the demographic history of DENV 4-I fluctuated, peaks appeared around 1982 and 2006. While purifying selection dominated the majority of E-gene evolution of DENV 4-I, positive selection characterized Clade III (Vietnam). DENV 4-I evolved in situ in Southeast Asia and the Indian subcontinent. Thailand and Indian acted as the main and secondary virus distribution hubs globally and regionally. Our phylogenetic analysis highlights the need for strengthened regional cooperation on surveillance and sharing of sample sequences to improve global dengue control and cross-border transmission prevention efforts.

Fig 1. The global distribution of DENV 4-I.

The number of unique strains (in bold, < 100% similarity in the same year) and the years of recorded transmission (in italic) in each location are shown in parenthesis after each location name.

Fig 2. The temporal distribution of DENV 4-I by location.

The red color dots represents the strains detected in mainland Southeast Asia and its adjacent provinces of China; similarly, blue depicts those found in maritime Southeast Asia; green those from the Indian subcontinent; and purple those few strains found in circulation elsewhere (Brazil and Australia). The diameter size of the circles corresponds to the number unique strains of DENV 4-I in a given year in a given location.

Fig 3. The Maximum Clade Credibility (MCC) tree summarized for DENV 4-I.

The colors of the branches correspond to their probable geographic location (see the legend). Circles and black dots indicated posterior probability support and ancestral location state probability of main node > = 0.85. For the key nodes, the median estimated divergence time and respective 95% HPD intervals were shown.

Fig 4. The global dissemination pathways of DENV 4-I.

Driven by the phylogeographical analysis of all included strains, red arrows indicate the probable direction of trans-border expansion of DENV4-1. The number of unique strains (<100% similarity in the same year) in each location is shown in parenthesis after each location name.

Fig 5

The demographical history of total DENV 4-I (A) and Clade I to V (B-F). The black line represents the median posterior value, and the blue area represents the 95% HPD intervals. The x-axis corresponds to time (years), while the y-axis is the product of the effective population size and the generation in years (N_eτ).