Did Zika mutate and turn ugly?

Zika virus (ZIKV) has been around for a while, first isolated by serendipity in a sentinel monkey that developed fever during observation in 1947. It was then observed in the human population during the course of several sporadic foci in Africa and sometimes in Asia, but the infection was generally mild. Then, only a few years ago, ZIKV became a public health concern with severe cases of Guillain-Barre syndrome (GBS) in adults and congenital Zika syndrome (CZS) in newborns. It is unsurprising that the scientific community has since tried to understand whether over the years ZIKV may have mutated in a way that happened to aggravate its pathogenesis. But is this the case? Despite some interesting studies summarised by Rossi et al and by Pierson and Diamond, the aggravated pathogenesis that followed the most recent ZIKV epidemics remains baffling.

The ZIKV strains
The closest phylogenetic relative to ZIKV is Spondweni virus, which shares with ZIKV 68% aminoacid identity in the envelope protein E. By contrast, E identity between ZIKV and dengue is about 50%. There are two strains of ZIKV: the African strain, including the original monkey isolate MR766 and other isolates from Senegal and Uganda; and the Asian strain, including isolates from the island of Yap in Micronesia (affected by ZIKV in 2007), from Cambodia, and from the Philipines. The American isolates all derive from the Asian strain, and some consider them part of the same Asian group, given the high level of similarities. The African and Asian strains are altogether rather similar, differing by only 10% at nucleotide level. Differently from dengue virus, there is only one ZIKV serotype, as antibodies against one strain neutralise the other strain as well.

African vs Asian
Given that GBS and CZS were associated only with the most recent ZIKV epidemics (French Polynesia in 2013, Brazil and south America from 2014), it is logical to hypothesise that there may be differences between the African and the Asian strains and to compare the two. Unfortunately, the first problem is that the comparison is generally made between the most recently isolated Asian strains and the historical African isolates like MR766, a comparison that doesn’t take into account the temporal evolution of the African strain.
Nonetheless, some comparisons across strains have been made, and the results are rather counterintuitive as the virulence of the African strains both in vitro and in mice seems to be higher than for the Asian. Similarly, when looking at mosquito vector competence, the African strains infect better. One possible explanation is that maybe the African strain is indeed more pathogenic, and while the Asian strain only causes CZS, the African strain may actually cause foetal death, but in the absence of surveillance this phenomenon may have gone unobserved. An alternative (and possibly more likely) explanation may be that the African population may benefit from better protection due to a pre-existing immune response against other flaviviruses. Studies in non-human primates suggest that while ZIKV antibodies may enhance dengue pathogenesis via antibody-dependent enhancement, antibodies against dengue or other flaviviruses can be protective against ZIKV.

The S139N mutation
Whether the African population is better protected against ZIKV is still unclear, and doesn’t rule out that there may still be something different about the viruses. Since virulence and vector competence don’t quite explain the aggravated pathogenicity of the Asian strain, some studies have hypothesised that maybe some mutations in the Asian strains may have altered ZIKV tropism and neurovirulence. A comparison between three Asian ZIKV strains isolated between 2015 and 2016 and one Asian strain isolated in 2010 has shown aggravated neurovirulence in neonatal mice (more apoptosis, more viral replication, and more neuronal damage) in the more recent strains. Virus sequencing identified 7 mutations, but when individually engineered into the 2010 strains, only one, S139N, rescued the virulent phenotype. This residue sits in the structural envelope protein prM, and given the importance of envelope proteins in tropism and infectivity, a mutation in any of their region may justify downstream changes in pathogenesis. Intriguingly, this mutations seems to have been introduced in 2013 and has been maintained since. Unfortunately, no comparison with African strains has been made, and the mode of administration directly into the brain of embryonic litter mates make it hard to determine the real impact of this mutation in a more physiological transmission.

The A188V mutation
A comparison between a strain isolated in 2016 in Venezuela and one isolated in Cambodia in 2010 also suggested increased virus release and vector susceptibility for the more recent Venezuelan strain. Sequencing studies revealed a possible association with a mutation in the virus non-structural protein NS1, A188V. This mutation has also been shown to inhibit interferon type I synthesis by binding to TBK-1 and preventing its phosphorylation. This mutation seems to have emerged in South-East Asia between 2003 and 2007; however it is also found in the African lineages, possibly explaining its high infectivity of Aedes, but not its milder pathogenicity compared to the Asian isolates.
A second mutation, T233A, has been suggested to destabilise NS1 dimers in vitro, although its impact in vivo ad in pathogenesis has not been investigated.

Changes in the 3’UTR
Changes in the virus untranslated regions have also been suggested to impact virulence and ZIKV replication. The Musashi protein, for instance, has been shown to bind to the 3’ UTR just downstream of a region where some mutations have been identified. This may have repercussions both on viral replications, but also on the ability of this protein to aid neuronal differentiation.

So, was it a mutation?
Given the evidence collected so far, how likely it is that the aggravated pathogenicity observed since 2013 are due to changes in ZIKV sequence? All together, rather slim. While some mutations may have increased fitness, no clear difference has emerged to explain why all of a sudden ZIKV has become a global threat.
As it’s always the case in infection, the pathogen is not the only player and it is likely that while some minor changes may have occurred to the virus, also the different populations affected may have responded to ZIKV in a different ways. This could be due to either different immunological profiles, or to different underlying genetic factors. Another possibility is that the higher numbers of infections observed in French Polymesia and South America may have been necessary to identify a correlation between ZIKV and GBS/CZG, and the dramatic diffusion of this virus in these populations may have simply helped linking the virus with the associated breath of disease outcomes.

Whether it was the virus, the host, the environment, or a combination of all remains unclear, but if we needed any further evidence of the baffling complexity of flaviviruses, ZIKV, unfortunately, has provided some new.