Wolbachia vs. malaria – we may be the real winners

It is estimated that there are 220 million cases of malaria each year and a total of 3.3billion, half the world's population, at risk of the disease. The number at risk is continuing to rise as climate change extends the regions in which the vector for the malarial parasite, Anopheline mosquitoes, can live; making malaria one of the most pressing issues of current infectious disease control. Fortunately there are some preventative measures that can be taken to avoid Anopheles mosquitoes as they are night biting - it's fairly easy to sleep under a net in an area of risk. However, getting these nets to those in most need is a different issue entirely. Fortunately, pathogens such as Plasmodium species (the parasites that cause malaria) have a life cycle that involves stages in two different organisms, opening a whole new avenue for preventative strategies. If we can directly control the vector of disease, mosquitos, then this could theoretically block human infection since the malarial parasite cannot pass directly from human to human. Attempts have been made with the large-scale use of insecticides or oil in the lakes where eggs are laid by mosquitoes, but these approaches are costly and unsustainable. However, recent work has hinted at a new method of control that may well be able to make an unprecedented contribution to our fight against malaria, and moreover our fight against another mosquito borne disease, dengue fever.

Similarly to malaria, dengue fever is carried by mosquitoes, and over 2.5 billion people arethought to be at risk. There are two major, and important, differences between the two diseases. Firstly, malaria is caused by a parasite while a virus is responsible for dengue fever. Along with the different causative agents, the two diseases are transmitted by two different species of mosquito, with malaria being carried by Anopheles and dengue being carried by Aedes species. This has important implications for control of disease. Aedes mosquitoes are day biting, meaning that avoidance of the two species poses different challenges. While sleeping under mosquito nets may work for malaria, this will have little impact on control of dengue. However, the two diseases do have the important commonality that they are both completely reliant on mosquitoes for spread, and as such, cannot spread within a human population without their respective vectors.

In 2011 results were published from a study looking at the possibility of directly controlling the Aedes mosquito in an attempt to prevent human cases of dengue fever. A team in Australia led by Scott O'Neill showed that it is possible to substantially reduce the spread of dengue virus with the simple intervention of infecting Aedes mosquitoes with bacteria and releasing them into the wild. The bacterial infection responsible for this potentially remarkable breakthrough is from a species known as Wolbachia, which naturally infects many arthropods. 

Wolbachia have evolved to spread between arthropods and invade the population. The main way the bacteria are able to spread and become established is through an effect known as cytoplasmic incompatibility (CI). If you imagine a cell like a balloon filled with water, then cytoplasm is the water. Once inside the mosquitoes, Wolbachia are capable of infecting germ line cells, these are the sperm and eggs, and residing in the cytoplasm of these cells. Upon fertilization of an egg by a sperm there is fusion between the two cells causing mixing of the cytoplasm of each cell. If an infected male mates with a healthy female then the fused cell is destroyed by the presence of incoming bacteria; meaning there is no fertilization. However, if an infected male mates with an infected female there is no issue, the sperm and egg, each with bacteria in their cytoplasm, will fuse and there will be fertilization to give new offspring. Similarly, if a healthy male mates with an infected female there will be successful fertilization. This gives the bacteria a maternal inheritance pattern, as the female always needs to be infected. The newly produced offspring will all carry Wolbachia in their cells allowing invasion of the population by the bacteria. In essence the Wolbachia bacteria have developed a way to kill off any uninfected mosquitoes by making the adults sterile (for all intents and purposes). Only infected mosquitoes are ever born once CI has taken true effect within a population.

What makes Wolbachia infection even more interesting (other than simply as an evolutionary fascination) comes from the fact that infected Aedes mosquitoes are unable to carry enough dengue virus to effectively transmit it to humans. It isn't fully understood why yet, but it seems that this may be down to the mosquito mounting an immune response against the bacteria that causes collateral damage against the other microbe.

The effect of Wolbachia on dengue transmission is beginning to look like a truly viable option for control. However, dengue is just one of a whole host of mosquito borne disease; sitting at the top of the list for those most desired to be tackled is malaria. Much interest therefore stemmed from the dengue studies into how this approach could be used for the control of malaria. However, for a long time this proved elusive, that was until the last couple of weeks. A new study in China led by Zhiyong Xi has managed to find a species of Wolbachia capable of invading an Anopheles mosquito species, and suppressing the level of malaria within the mosquito.

A malaria causing plasmodium

This is a major breakthrough. Until this paper was published no Wolbachia species had been found that was capable of becoming established within any Anopheles species. What's more, the study showed that Plasmodium falciparum, which causes the most dangerous form of malaria, was affected by the presence of Wolbachia. The next stage will be to release infected mosquitoes into the wild and see if the bacteria can become established outside of the laboratory setting. If these Wolbachia infected mosquitoes can become established in the wild then we may well have a way to significantly reduce the spread of a disease that so many people are at risk from. Probably the best way to truly protect people from malaria will be to develop a vaccine, but until we manage that we need other strategies to tackle the infection. Our best strategy, at present, is the use of insecticides and nets. However, getting nets to the poorest areas of the world, which are often those most at risk, is not always an easy task. An easier goal may be to find a way to block mosquitoes from carrying malaria, making their taking of a blood meal essentially harmless. The discovery of a Wolbachia species able to establish within Anopheles population brings us on giant step closer to achieving this goal.

Mixed viruses with a mixed message

A little over a week ago I stumbled across a story in the journal Science describing work investigatingthe transmissibility of the influenza virus H5N1, commonly known as bird flu. Bird flu has sporadically made headlines ever since it was first detected in humans over fears that it could cause a serious pandemic. The fears are justified; pandemics are, for the most part, caused when the human population has no prior immunity to a virus. Traditionally, only H1-3 viruses have infected humans meaning there is no immunity to a H5 carrying virus, as we have never encountered it. However this fear has at times become somewhat out of hand; take for instance the hoopla over the two transmission studies on H5N1 at the start of last year, which I have previously covered on this blog (a two post article so here is link 1 and link 2). Those studies caused to a worldwide moratorium on influenza research, which has only recently been lifted. This most recent study is cut from a similar cloth and true to form, has been blown way out of proportion in the press (as an example). So, after a fairly long absence from writing my blog, I’m back to deconstruct what this headline making research actually did and explain why we don’t need to be as scared as it may seem.

An illustration of an
influenza viral particle

I’ll start with some background so that everyone is up to speed and can hopefully follow the discussion of the work in question. Everyone knows influenza as a virus that infects humans and causes seasonal outbreaks of the disease flu. However, influenza isn’t actually a human virus per se, it is more a virus of birds. On the surface of an influenza viral particle there are two proteins known as hemagglutinin (HA or H) and neuraminidase (NA or N) that we use to classify the virus. There are currently 17 known forms of HA and 9 forms of NA, however, only viruses with combinations of H1-3 and N1-2 are known to productively infect humans. All of the other HA and NA molecules combine in different ways to produce viruses that infect birds. The situation is changing slightly with the gradual emergence of H5N1 and, more recently, H7N9 viruses in human populations. However, to date, these viruses have shown very poor (if any) ability to spread between humans. While there have been many cases (particularly of H5N1) these have all be contracted directly from sick birds.

To add further detail about the influenza virus, it is a RNA virus with a segmented genome. In essence, a virus is simply a structure made of proteins and fats that encase genetic material. You can think of a virus as a vehicle ‘designed’ in the simplest possible way to carry its passengers (genetic material) into a cell. In the case of influenza, the passengers within the vehicle are 8 distinct RNA strands (RNA being the genetic material). These strands carry the instructions (genes) to make proteins that are used to produce new viruses, allowing infection to spread. Every new virus produced must contain a single copy of each of the distinct 8 strands of RNA so to produce a replica of the original infecting virus. Most viruses are not segmented in the way influenza is and just have one long strand of genetic material. There is a consequence to this segmentation, if more than one distinct influenza virus infects a cell at the same time, strands from one virus can be exchanged with strands from the other. This is a process known as reassortment and means a completely new virus is formed that is genetically different from either of the two original infecting viruses. This process of reassortment is the major cause for pandemic spread of influenza and is thus the topic of much research.

A basic diagram of reassortment

Currently H5N1 cannot transmit in an airborne form between humans, however H1N1 influenza viruses can (this was the virus responsible for the most recent pandemic in 2009 – Swine flu). If an H1N1 virus and an H5N1 virus infect the same cell at the same time there is a possibility that reassortment could occur. It is therefore possible that the H1N1 virus could swap some genes with H5N1 that confer the ability of airborne, human-to-human transmission. The reassortment produces a new virus, but as long as the HA gene does not exchange; it will still be an H5N1 virus. This is a very realistic possibility as pigs have been identified that have been infected with both H5N1 and H1N1 viruses.

This leads me nicely into the paper I wish to discuss which is investigating this very event. The work was conducted in China’s Harbin Veterinary Research Institute by a group of scientists lead by Chen Hualan. The group systematically produced every possible combination of H1N1 virus mixing with H5N1 virus. They started with an H5N1 virus with all its 8 strands of RNA, then artificially (not through infection of a cell) exchanged strand 1 of the H5N1 with that of an H1N1, giving a hybrid of H1N1 and H5N1. They did the same for every individual strand, and then moved on to produce combinations (eg. strand 1 and 2 of H1N1 with the rest being from H5N1, and so on). In total, they produced 127 H1N1/H5N1 hybrid viruses (no small task). Those proficient at maths may notice that 127 isn’t the total number of possibilities that could occur from the exchange of 8 gene segments between 2 viruses. However, since the team wanted to look at the spread of an H5N1 virus they needed to keep the segment that codes for the H5 gene meaning the maths becomes 2⁷ (minus the original H5N1 parental genome).

Having produced all of these hybrid viruses the team tested their pathogenicity in mice and subsequently looked at the transmissibility of the most dangerous ones. Obviously they cannot test them on humans, so instead animal models need to be used. The two most commonly used animal models for studying influenza transmission are ferrets and guinea pigs. The Chinese team opted for guinea pigs. Now just in case there was any doubt in your mind, guinea pigs are not the same as humans (the same is true for ferrets - which are also not the same as guinea pigs, just to clarify). These animals are MODELS, they allows us to get an idea of what may occur in humans and are the best we can do for the obvious reasons of not wanting to deliberately infect humans. The knowledge that is gained from the study of these models must always come with the caveat that humans are very different to guinea pigs and ferrets.

The genome segments of influenza

The team found that their parental H5N1 virus could transmit between guinea pigs by contact, but not through the air, where as the H1N1 parental virus was able to spread through the air (similar to the human scenario). Upon testing their selected viruses they made some interesting discoveries. Firstly, only 5 of their hybrid viruses had “highly efficient” transmission (comparable to that of the parental H1N1) with one addition hybrid capable of “efficient” transmission (not as good as H1N1 but higher than H5N1). Not only is it interesting that so few viruses were able to transmit, but an interesting theme emerged when assessing which viruses could transmit and which couldn’t. It was found that H5N1 viruses that receive the PA or the NS genes from H1N1 become transmissible. The hybrid viruses with PA or NS can receive other H1N1 genes and maintain their transmissibility, but if they receive either the NA or M genes the transmissibility is reduced. The PA and NS genes are therefore key determinants of transmission. The PA gene forms part of a protein complex that replicates the virus genome while the NS gene has two roles, firstly it plays a large role in evasion of the host immune response, and also plays a role in getting influenza genetic material out of the nucleus (where it is replicated) so that it can be packaged into new viruses. It is interesting that the NA gene of H1N1 causes a reduction in transmissibility of hybrid H5N1 viruses since both are considered to be N1. Clearly there are intrinsic differences between these two genes even though, in a broad sense, they are very similar.

I find this research very interesting from a scientific point of view, but the real motivation behind my desire to write this blog and explain the research was the response the work received in the press. Headlines proclaiming “killer flu” being produced in China or calling the work “appallingly irresponsible” are hyperbolic to say the least! There is certainly an element of playing with fire in terms of making H5N1 viruses transmissible, but the authors took the highest necessary precautions and made a point of explaining these in their published work. They used the second highest level of biological containment making it very hard for any of the virus to escape. There are only a handful of facilities equip from the highest level of containment and if all transmission studies were forced to use these nothing would ever get done. I direct you to a Wikipedia page should you be interested in the definition of these biosecurity levels.

Aside from the biosecurity aspects, the statements that these scientists were making “killer flu” really angered me. The press articles I have read seem to have conveniently missed a very important sentence written in the paper (twice) regarding the guinea pigs that became infected with the transmissible hybrid viruses, “NONE OF THE GUINEA PIGS DIED.”  Now as I have already mentioned, guinea pigs are models for humans and what is true for guinea pigs may not be true for humans. You could therefore throw that back at me and say these viruses could kill humans, which I cannot dispute (because we simply don’t know). But as far as the guinea pig model goes, these viruses are far from “killer.” The real point is that this study has given us some very interesting and useful insights into the necessary changes that may need to occur to allow H5N1 to become transmissible, and has hinted that with the ability of airborne transmission, some of the pathogenicity of the virus may be lost. With the knowledge gained from this paper we will be able have improved surveillance over cases of human H5N1 to look specifically for these changes allowing us to be properly prepared should H5N1 begin to spread amongst humans. Any prior warning, no matter how small, of a potential pandemic will allow us to stockpile the necessary drugs and formulate vaccines allowing us to have a rapid and effective response to stop the spread as soon as possible and potentially save countless lives. This paper has truly added useful knowledge to the influenza field, regardless of much that has been written about it maligning that fact.