Dengue virus evolution drives multiple disease outcomes

From the Bedford lab, Vaccine and Infectious Disease Division

Mosquito-borne Dengue virus (DENV) constitutes a large global health burden in Southeast Asia and South America. The virus, which causes Dengue hemorrhagic fever, consists of four genetically distinct serotypes. While primary infection with any DENV serotype is usually mild, subsequent infections with heterotypic DENV lead to varying degrees of disease severity. Although many people asymptomatically experience multiple infections, some develop severe disease due to antibody dependent enhancement (ADE), whereupon reinfection leads to cross-reactive but non-neutralizing antibodies that bind the virus and exacerbate infection. One potential explanation for the inconsistency in disease outcomes is that DENV may contain overlooked antigenetic diversity within a given serotype, leading to various levels of either protection from or enhancement of disease upon reinfection with another DENV clade. To interrogate these antigenic relationships between DENV serotypes, Sidney Bell, formerly of the Bedford lab (Vaccine and Infectious Disease Division), analyzed DENV sequence diversity and serological assay data and recently published this work in eLife.

Antigenic variation between viral strains can be measured experimentally with in vitro neutralization assays, which quantify how well serum raised against one virus can neutralize another virus. The Bedford lab revisited previously published DENV neutralization data to assess the antigenic diversity within and between DENV serotypes. Previous comparisons revealed that although DENV viruses within a serotype are more antigenically related than are heterotypic DENV viruses, homotypic viruses contain more antigenic variation than previously appreciated, supporting their hypothesis. To further define antigenic evolution in DENV, Bell and colleagues employed an established model to link genetic and antigenic evolution that was developed by the Bedford lab to study influenza evolutionary dynamics. They hypothesized that antigenic change is due to underlying viral genetic evolution. To test this, they fed the model known serum neutralization data, allowing the model to calibrate normalized antigenic distances to observed genetic mutations. The model attributed specific mutations to antigenic change, suggesting that evolutionary processes drive antigenic variation.

Phylogeny of dengue virus sequences and normalized antigenic distances.
Phylogeny of dengue virus sequences and normalized antigenic distances. Image from publication.

Once establishing the link between genetic and antigenic evolution, they then used the fitted model to predict antigenic distances between sets of untested DENV viruses by comparing genetic discrepancies between viral pairs. The model found that among this larger viral population, inter-serotype antigenic diversity remained greater than intra-serotype diversity, but that at the genotype level, viruses within a serotype were, in fact, moderately varied. These findings suggest that antigenic variation exists on a more granular scale within DENV and that ongoing viral evolution drives these distances, challenging the assumption that homotypic DENV are uniform. After establishing that previously unappreciated genetic variation within serotypes may contribute to a range of ability to provide protection from subsequent infections, Bell and colleagues sought to define how antigenic diversity drives population dynamics within circulating DENV. The relative frequency at which any given DENV virus serotype is currently circulating is driven by viral fitness: over time, more people acquire more immunity to the dominant strain, creating selective pressure for escape variants with novel antigenic sequences to spread. The composition of the viral population can be estimated by sequencing historical viral samples, where antigenically distinct, fitter viruses can be expected to increase in frequency over time. To test this hypothesis, the authors examined dengue virus samples collected at three month intervals (collected from 1970 to 2015) from Southeast Asia. Using observed relative frequencies of each virus at a given timepoint, the authors built a model to infer the growth rate, or success, of a serotype and found that viral fitness is a major driver of viral population dynamics. They also modeled how intra-serotype genetic variations drive population growth. Although their earlier discoveries demonstrated that viruses within a serotype varied more at the genotype level than previously thought, the model predicted that variation at the genotype level was not enough to drive large-scale DENV population composition.

This work helps to explain the inconsistency in antibody cross-neutralization between DENV serotypes, as inter-serotype, and even intra-serotype DENV strains may differ enough to produce cross-reactive but non-neutralizing antibodies, leading to ADE. Additionally, they show that human population immunity at the serotype level, which drives viral evolution and DENV population dynamics. Bedford concludes that “it’s clear that immune responses to antigenically variable pathogens such as dengue and influenza may shape viral evolution. It’s been interesting to use virus evolution as a lens to better understand immunity. Dengue is particularly complex in this regard, but these sorts of evolutionary approaches are beginning to shed light on questions long outstanding in the field.”

Bell S, Katzelnick L, Beford T. 2019. Dengue genetic divergence generates within-serotype antigenic variation, but serotypes dominate evolutionary dynamics. eLife. pii: e42496. doi: 10.7554/eLife.42496. [Epub ahead of print].