posted Dec 14, 2014, 10:18 AM by Kevin Esvelt
updated Oct 25, 2016, 9:32 AM
Malaria is surely one of the most terrible diseases afflicting humanity. Even after an exceptionally successful effort to expand control and treatment measures, it currently kills over half a million people each year and sickens over two hundred million, crippling livelihoods and economies. On a typical day, over a thousand children will die of malarial fevers. Many thousands more will suffer permanent developmental damage due to anemia or cerebral infection.
We should certainly continue our current efforts to combat this terrible disease, but at the same time we must recognize their limits. Prevention and treatment cost resources. If our goal is to promote human well-being and community development, malaria control may presently be the most effective way to spend that money. Still, imagine what those resources could be spent on if people didn't have to deal with malaria: think of schools, hospitals, community centers, and more.
Sadly, our current tools have the proverbial snowball's chance in hell of leading to total eradication. Even worse, the Plasmodium parasites that cause malaria are continually evolving resistance to our best treatments, while the Anopheles mosquitoes that transmit the parasite from infected to uninfected people are evolving resistance to our best insecticides. They're even evolving behavioural changes enabling them to counter the bednets and indoor spraying that have been a cornerstone of recent successes: mosquitoes now bite progressively earlier in the evening, before people go to bed, and avoid resting indoors where they are vulnerable to insecticides. There is an urgent need to develop new control measures just to maintain the ground we've so laboriously won.
If our goal is total eradication – and there's no doubt it should be – we should look to historical successes for guidance. To date, we've managed to permanently eliminate only two diseases: smallpox, perhaps the most terrible plague ever to afflict humanity, and rinderpest, a disease of cattle. Both of these viruses were eradicated because scientists developed highly effective vaccines. By protecting a large enough fraction of the respective human and bovine populations, we decreased the basic reproductive number of the viruses for long enough to drive them extinct (save for highly controversial laboratory stocks).
Tremendous efforts have gone into attempts to create a similarly effective malaria vaccine. These efforts can and should continue. But so far they haven't borne fruit, because Plasmodium is a resourceful foe. Every parasite genome has a suite of roughly 60 var genes encoding variable surface antigens, meaning every parasite has a large number of “coats” it can don to evade recognition by the immune system. While it's true that people who are repeatedly exposed are eventually able to suppress most symptoms when infected (this is why malaria is particularly deadly to children who have not previously been exposed), precisely how we might induce this same tolerance with a vaccine remains a mystery. Historically, our best control measures have focused not on attacking the parasite, but on the mosquito vector.
Once found throughout most of the world and as far north as London, malaria has now been confined to the tropics and subtropics, primarily by controlling local populations of Anopheles mosquitoes. In the United States, a concerted effort involving insecticides, draining wetlands, and spraying oil on remaining swamps to kill mosquito larvae managed to locally eradicate the parasite. While still found in parts of Central and South America, the general consensus is that a concerted effort using available tools could eradicate the disease from the Americas. Southeast Asia, India, and especially Africa are another story. Part of the reason is that their mosquitoes are simply nastier. Any malaria-infected mosquito that bites an animal is a waste from the parasite's point of view, as animals aren't part of the chain of transmission (although they have their own forms of malaria, there aren't thought to be major animal reservoirs of Plasmodium falciparum or Plasmodium vivax, the most harmful human versions). That means the more a given mosquito species prefers eau de homo sapiens over the alternatives, the more effective it will be at transmitting malaria. American mosquitoes just aren't all that into humans compared to Eurasian and especially African species, some of which are nearly exclusive human-feeders.
This very reliance on a handful of mosquito species could well be Plasmodium's undoing, as we now possess a technology – the gene drive – that may allow us to alter wild populations, including mosquitoes. There are two ways we could use mosquito gene drives to tackle malaria.
First, we could alter mosquitoes so they can no longer transmit malaria. Scientists have already identified several naturally occurring alleles found in mosquitoes that reduce their ability to transmit Plasmodium, and also identified synthetic alterations such as antimalarial peptides that produce the same effect. We could build a gene drive to sweep these alleles and synthetic genes to fixation in wild mosquito populations, dramatically reducing their ability to spread the disease. Most of the driven naturally occurring alleles would be evolutionarily stable, while the synthetic additions might be slowly eliminated if they're costly to the mosquito, but we could always release mosquitoes carrying new drives with intact or newly discovered antimalarial genes. In a way, this would represent a new front in our existing struggle to come up with new antimalarial drugs and insecticides effective against mosquitoes.
But there is a distinct possibility that altering the mosquitoes won't be enough. Both major parasites have enormous effective population sizes and have already been exposed to the natural genetic changes that lead to resistance on the part of mosquitoes. Even if we drive many such changes to fixation, it's entirely possible that the parasite will evolve resistance to all of them. In that case we would buy ourselves a few years, saving several hundred thousand or perhaps a few million lives while preventing hundreds of millions of infections. But malaria would return. Could we come up with (and spread via subsequent drives) a new cocktail of genetic changes capable of blocking the parasite by then? Could this strategy hope to attain the holy grail of total malaria eradication worldwide?
Possibly not. And if the evolution of resistance does lead to a resurgence of malaria, there's a strong argument that we should follow up with a more permanent solution: extirpating the mosquito species that most readily transmit malaria.
There are two ways to suppress or extirpate an entire population of organisms, both outlined by Austin Burt of Imperial College London. The first is a “meiotic drive” that ensures most offspring will be male. For example, a meiotic drive might involve expressing a gene from the Y chromosome that shreds the X chromosome during spermatogenesis, causing most or all offspring to inherit and Y (and therefore become male in species where the Y confers maleness). Burt and coworkers even have a promising homing endonuclease that may eventually enable this approach if they can express it from the Y chromosome (although a study of this enzyme in yeast argues that resistance may evolve fairly quickly). Using an RNA-guided meiotic drive based on Cas9 to target many different sites would likely be more robust, although developing it would require repeating the optimization process Burt and colleagues applied to their homing endonuclease.
Burt's second proposed method of population suppression is still more elegant. Build an endonuclease drive that replaces a recessive gene required for fertility or viability. Ensure that the drive is copied only in the cells that give rise to sperm and eggs – the germline – after the target gene's function is required. Organisms that inherit one copy of the drive will be heterozygous in all of their tissues and therefore of nearly normal fitness. However, they will contribute the drive to (almost) all of their offspring. When two heterozygotes mate, the offspring will be either nonviable or sterile. This means that when the drive is rare, it will spread very rapidly through the population because most mating partners will be wild-type, and most offspring will therefore be heterozygotes. But once it becomes abundant, a higher and higher fraction of offspring will be reproductively nonviable, causing the population to crash. Burt's models show that several drives released in tandem could impose a “genetic load” sufficient to drive the population down to zero. No more malarial mosquitoes.
This couldn't be achieved before the development of Cas9-based genome editing simply because it was too difficult to target the requisite genes. And targeting them at just one site won't do, as any mutant gene that still functions yet can't be cut and replaced by the drive will always have fertile and viable offspring, thereby saving the population. But since Cas9 privileges design over evolution, we can use it to prevent this sort of evolutionary escape by cutting many different sites in each target gene. Genetic load drives will unquestionably be difficult to build because drive cutting and copying must occur at just the right time, with no active Cas9 and/or guide RNAs remaining by the moment of fertilization. This will take time and effort – and must be done not only in the main vector studied by most laboratories, Anopheles gambiae, but in all the major mosquito vectors of malaria. Absent a vaccine or alternative control method, perhaps 6-9 species in total would be targeted. Most of these haven't even been reared in captivity, much less rendered transgenic. Can it be done? Almost certainly, and perhaps within a decade if we get our act together. But it will require a concerted effort by many laboratories and dedicated funding to support them.
Of course, all of this assumes that gene drives are science as usual. They are anything but. The potential ecological consequences of spreading a particular alteration through an entire population of wild organisms must be rigorously evaluated before any potential release. That goes double for proposals to deliberately extirpate wild populations. We must also recognize that wild populations are part of the shared environmental commons. They must never be altered without transparent and broadly inclusive discussions informed by the best science available. While the moral case in favor of taking action against malaria is strong, this does not mean that future discussions will decide that gene drives targeting mosquitoes are worth the possible risks. No matter what is decided, for malaria or for other potential gene drive applications, it should go without saying that scientists must abide by the will of the people.
Tags: gene drives, malaria, technology