Issue 99: Japanese Encephalitis part 3: in the mosquito and on the straight and narrow

Written by Nobel Laureate Professor Peter Doherty

All arboviruses infect humans (and/or other vertebrates) by the bite of one or other haematophagous (blood -feeding mosquito, tick or sandfly) arthropod (#97, #98). With Japanese encephalitis virus (JEV), the proboscis of a female Culex spp ‘flying needle’ mosquito inserts into the lumen of a small blood vessel. As that skin/blood vessel wall barrier is breached, two ‘pump organs’ in the mosquito’s head push saliva and gut fluids (which contain anti-clotting agents and JEV virions) into us while taking the blood meal needed to sustain the insect’s eggs. The innocuous males feed on nectar.

Once the virus is in the superhighway of our – or some other amplifying/maintaining vertebrate host – blood stream it goes everywhere. Virus entry and replication can occur in a variety of cells, including those around the injection site and vascular endothelium. Arboviruses that do a lot of damage to blood vessels cause extensive bleeding into tissues (haemorrhagic diathesis). Though this is not characteristic of JEV, the virus may ‘grow through’ the vessel wall and breach the blood-brain-barrier (#11, #35) to infect neurons, which generally can’t be replaced. The severity of the resultant encephalitis, which can vary a lot, will depend on the extent of nerve cell damage and where it occurs.

While this is happening, the infection triggers (as described extensively for SARS-CoV-2 in this series, #87, #88) an immediate innate immune response – with the release of cytokines and chemokines that can cause fever and malaise – and the very virus-specific B cell and T cell responses of adaptive immunity. One difference from respiratory infections like influenza and COVID-19 is that any lymphoid tissue, particularly the spleen that filters the blood, can potentially be involved from the outset. After seven to 10 days or so, as antigen-stimulated plasmablasts exit the lymphoid tissue to provide the immunoglobulin (Ig) secreting plasma cells that localise in bone marrow, secreted antibodies will quickly terminate the viremic phase. Then, with effector T cells eliminating the virus-producing cell ‘factories’, the process of recovery can begin (#21, #34, #96).

Both SARS-CoV-2 and JEV have single stranded, positive sense RNA genomes, with 30,000 and 10,000 base pairs respectively. As we are now all too familiar, the SARS-CoV-2 genome is subject to high levels of mutation, with changes in the spike protein progressively defeating vaccine protection (#93). When it comes to JEV, which has likely been circulating in Asia for centuries, there are five genotypes and the one vaccine works fine for all of them. The same is true for yellow fever virus (YFV), the prototype flavivirus. The ‘live attenuated’ 17D YFV vaccine strain that is in current use was first developed by Max Theiler in 1937, leading to his being awarded the 1951 Nobel Medicine Prize, the only one ever given directly for making an antiviral vaccine.

So, the CoVs mutate a lot but the flaviviruses don’t: is that just a basic difference for these virus families? It seems not. There’s another weird flavivirus that, while some closely related strains do circulate in bats and birds in nature, is not transmitted by an insect vector. That’s hepatitis C virus (HCV) which, though humans may have once been infected by mosquitoes, is now transmitted only by another type of needle, a hypodermic needle. There’s a good drug to treat HCV, but making a vaccine was impossible because, like HIV, HCV mutates at a very high rate within us.

What keeps JEV and YFV on the ‘straight and narrow’ in the genomic sense? While SARS-CoV-2 infects and multiplies only in cells of mammals (humans, bats, minks, ferrets, hamsters), JEV replicates in mammals, birds and mosquitoes, especially in the epithelium of the posterior mid-gut. The mozzie is not just a ‘flying needle’, it’s also a virus-producing factory! In fact, all the arboviruses infect, and replicate, in cells of their vector species. Furthermore, in experiments aimed at probing the question of over-wintering, lizards were shown to be infected by JEV-carrying mosquitoes, which became viremic and infected other mosquitoes that passed the disease to birds. I’m not a virologist, but I can’t think of another virus family (as distinct from bacteria) that can potentially be transferred naturally from cold to warm blooded vertebrates. Do we see few changes in arbovirus genomes because a mutation compromising any of the molecular mechanisms required to maintain this complex (for a virus) lifestyle would be lethal, at least for the insect transmission phase? Is that what happened way back with HCV?

The likely profile with JEV and the Murray Valley encephalitis (MVE) is that, unless the world warms to the extent that there is a year-round mosquito problem, we will see a repeat of the MVE scenario with sporadic incursions from the north in very wet years. There’s no case to start wiping out the local reptiles, or to be concerned about native mammals which, so far, have not been implicated as important maintenance or amplifying hosts for any Australian arbovirus. That leaves us with the protection of insect repellents, some excellent and durable vaccines, and environmental spraying. The latter, which can compromise food sources for insectivorous birds and bats is, of course, the least environmentally friendly of these alternatives.