Key Results & Achievements

Even though ADE from a prior ZIKV infection followed by a DENV infection has been demonstrated (George et al. 2017) this does not infer that ADE from a prior DENV infection followed by ZIKV infection will occur. Macaque monkey models to date, despite their low sample size, failed to generate ZIKV after a previous DENV infection (McCracken, et al., 2017; Pantoja et al. 2017).

Work was undertaken to establish if ADE could occur in humans and to determine the molecular/immunological mechanism, first using bioinformatics modelling and secondly using identified in vitro or ex vivo using reverse genetics.

Identification of the major type-specific and cross-reactive B-cell epitopes on ZIKV

It was established that ZIKV has a new mechanism of T-cell immunoevasion from previous DENV exposure.  Major advances were made in identifying how Zika virus avoids the cross-protective T-cell responses from previous DENV infections, the mechanism being interspecific recombination. The work was performed by developing two original bioinformatics programs or data pipelines. Results showed that although ZIKV avoids the major cross-protective T-cell epitopes of DENV, it is rendered susceptible to the cross-protective T-cell response from West Nile virus (WNV) and Japanese encephalitis virus (JEV). This is a probable, core reason why Zika virus has not caused epidemics in South East Asia, whereas in regions such as the Pacific Islands and South America, where the presence of WNV and/or JEV is low or absent, there have been dramatic epidemics.

A new bioinformatics method was identified based on “Deep Learning” to identify neutralizing B-cell epitopes between ZIKV and DENV, which precisely correlated with E-protein dimeric epitopes (EDE) neutralizing mutations identified in vitro by Dejnirattisai et al (2015; 2016) in DENV and associated with ZIKV-DENV ADE. These findings are very important because they precisely identified common epitopes in the main envelope (E) glycoproteins of DENV, which is highly endemic throughout the tropical and subtropical regions of the world, and ZIKV, and against which the antibodies that define them may generate more severe DENV or ZIKV disease through ADE. 

As a very useful and important aid for our ZIKV ADE assays, in addition to using live ZIKV, a reverse genetic system was developed using Zika virus-like particles (VLPs) for both mammalian and insect expression that produce native-like ZIKV virions which lack their full genomes. This system was so successful that WP6 went on to develop a single plasmid (insect cell expression) which can express all ZIKV structural proteins without associated helper plasmids. The system enables bioinformatic predictions to be transformed into naturally functional VLPs to assess the molecular mechanisms of ADE and can help in the design of candidate vaccines that lack ADE potential in ZIKV and possibly other flaviviruses (see below). 

Preliminary results suggest human sera previously exposed to dengue virus infections generate a great diversity in ZIKV ADE responses. This distinguishes ADE in situ from previous primate in vivo models. This panel of human anti-dengue virus sera are being used to study the potentials of ZIKV ADE using both the Zika VLPs and wild-type ZIKV, as well as generate antibody-mediated selection pressure to study the mutations that increase ZIKV ADE in Fc-receptor bearing cells. In the latter case, two amino acid substitutions which have been predicted to destabilise the ZIKV E glycoprotein dimers have been generated and are likely to result in the unnatural exposure of additional epitopes responsible for ZIKV ADE. As such, the results are very important for the understanding for ZIKV ADE being generated by antibodies against other flaviviruses such as the highly endemic DENVs and their candidate vaccines.  We have also repeatedly immunised rabbits with the live attenuated tetravalent DENGvaxia (Sanofi Pasteur) vaccine to assess the ZIKV ADE potential after different timepoints as well as primate sera generated against the live attenuated yellow fever (17DD) vaccine as core to our Work Package 6 goals.   

Epitope Mapping

An extensive ZIKV T-cell epitope map was completed, which is notably distinct from related flavivirus species and the first paper ‘Recombination of B- and T-cell epitope-rich loci from Aedes- and Culex-borne flaviviruses shapes Zika virus epidemiology’ was published.

These results therefore greatly add to our understanding of ZIKV differences (T-cell epitopes) and similarities (B-cell epitopes) that may result in epitope DENV cross-protection (CD8+ T cells) or ADE (antibodies).T-cell selection pressure for DENV to become ZIKV was published in ‘Cutting Edge: Transcriptional Profiling Reveals Multifunctional and Cytotoxic Antiviral Responses of Zika Virus−Specific CD8 + T Cells’ and additional dating of these genetic recombination events are being applied.

Participating Organisations

Group leader: Prof. Andrew Falconar, Fundación Universidad del Norte

  • London School of Hygiene & Tropical Medicine
  • La Jolla Institute for Allergy and Immunology
  • The University of North Carolina at Chapel Hill