Photo: Alison McAfee
Every beekeeper has wrestled with the Varroa destructor mite. Every beekeeper also knows we’re running out of weapons. 

With abounding resistance to conventional miticides, researchers at Monsanto are instead trying to use biotechnology to fight the mite – it could work, but is another mite-killing agent what the industry really needs?

Recently, I had the chance to listen to talks from two members of Monsanto’s bee research team: Jerry Hayes (America Bee Journal’s own respected author of The Classroom) and Alex Inberg. Hayes was a keynote speaker at the British Columbia Honey Producers’ Association annual general meeting in Kelowna last October, and Inberg spoke at the Entomological Society of America conference in Denver last November. Both highlighted how Monsanto is developing a new method of varroa control using a biotechnology technique called “RNA interference” (RNAi for short). By feeding honey bee colonies with syrup containing specific RNA molecules, they can suppress the growth of mite populations. It is not yet ready for market and it’s no silver bullet, but it’s an intriguing endeavor for many reasons (scientifically and otherwise).

RNAi took molecular biology by storm over the last few decades, earning its discoverers (Andrew Fire and Craig Mello) a Nobel Prize in 2006. In the late 1990’s, they found that injecting small worms (Caenorhabditis elegans) with a certain kind of RNA molecule could change the worms’ behavior, and deduced that this happened because they were disrupting the regular flow of genetic information. Cells contain DNA in the nucleus, which is normally transcribed into smaller molecules called messenger RNA (mRNA) that travel outside the nucleus to be translated into proteins (Figure 1A). DNA is double-stranded (think of the famous double helix structure) whereas mRNA is single-stranded (we call it the “sense” strand), but it too has the potential to form a double-strand if it’s given the right complimentary partner (the “antisense” strand). Fire and Mello found that when the sense and antisense RNA strands were combined to form double-stranded RNA and injected into the worm, the gene corresponding to the sense sequence (the one that’s needed to create a functional protein) was silenced and the worms began to twitch. So what was going on?

Figure 1. Schematic showing A) the regular flow of genetic information from DNA to messenger RNA (mRNA) to protein, and B) the RNA interference pathway, resulting in degradation of the messenger RNA. * Dicer cuts up each long double-stranded RNA molecule into shorter double-stranded RNA molecules with different sequences. These alternative sequences can also be incorporated into RISC, allowing many different parts of the mRNA molecule to be targeted and cleaved. This is why there is a low risk of mites evolving resistance to the RNAi treatment – many different mutations would have to occur to render the treatment ineffective.

The gene that Fire and Mello were targeting was called unc22, which is normally produced in muscle cells and is important for muscle contraction and motility. Under regular conditions, the unc22 DNA sequence is transcribed into the mRNA messenger and translated into the unc22 protein (the lack of italics here isn’t a mistake – it’s a weird convention in the field that gene names are italicized, and protein names aren’t). This is the normal flow of information for any gene.

Injecting double-stranded unc22 potently suppressed expression of that gene, causing the worms to lose muscle control (hence the twitching). After Fire and Mello’s seminal paper in 1998, many other researchers participated in elucidating the key molecular players (Figure 1B) – as it turns out, the double-stranded RNA is cut up into smaller pieces by a protein called, fittingly, “Dicer,” and incorporated into a bigger protein complex called “RISC.” RISC uses these bits of RNA like puzzle piece detectors to seek and destroy normal mRNA molecules that have the matching piece, or complimentary RNA sequence. When the normal mRNA molecule is destroyed, it can’t be translated into a protein and the function it normally serves is lost. So, when unc22 double-stranded RNA is injected into worms, the end result is suppression of the unc22 protein production and a subsequent loss of muscle function.

This technology hasn’t escaped Monsanto’s attention for applications in pest control. Except for prokaryotes, nearly every living thing has the molecular machinery RNAi requires, including many plants, insects, and mammals. However, this doesn’t mean that the RNA treatments are also universally toxic – that would be bad for business. Since the researchers control the double-stranded RNA sequence, they also control the genes they decide to shut down. To target one organism over another (say, varroa, but not honey bees), we just need to find an essential gene sequence that one species has, but the other doesn’t.

In the case of varroa, there’s a gene expressed in the foundress mite which she needs for reproduction. With RNAi, sometimes there can be surprising off-target effects, so it all requires rigorous testing, but we are pretty good at predicting what genes will and will not be affected.

“The RNA isn’t toxic to humans,” Hayes told the crowd at the Kelowna meeting. “You eat it – we all eat RNA in our regular food. And if you spill it on the ground, it quickly breaks down.” RNAi treatments are often marketed as “non-synthetic,” but that description is imprecise; the double-stranded RNA molecules are synthetic – large amounts of purified enzymes are used to synthesize it in controlled chemical reactions. The difference between RNAi and conventional treatments is in the molecule’s properties and the mechanism of action.

The formulation isn’t corrosive, like conventional acid varroa treatments, nor should it accumulate in the wax of the hives. Risk of developing resistance to RNAi is also lower than for small molecule pesticides like coumaphos and fluvalinate. Both conventional miticides bind to a specific spot on a protein (acetylcholinesterase – an enzyme that’s essential for nerve function), inhibiting its activity and preventing the nerve cells from communicating with each other. Just like how microbes evolve resistance to antibiotics, mites evolve resistance to miticides. All it takes is a mutation that alters the structure of acetylcholinesterase just enough so the miticide doesn’t bind, and resistance is born. RNAi, however, targets many parts of a gene sequence all at once, so chances of all the right mutations accumulating in one mite are extremely low. However, using this technology is not without challenges.

Hayes described how in their field trials for RNA-based varroa control, they’ve observed variable efficacy. That is, sometimes the RNA-laced syrup treatment works well, and sometimes it doesn’t. They aren’t yet sure why this is the case, but one possibility is that the bees disperse the RNA unevenly in the colony. “We think the RNAi is working on mites in the brood cells, rather than phoretic mites,” Hayes explains. “It probably depends on nurses taking up the RNA syrup mixture and transferring it to the brood.” It’s conceivable that this process could be influenced by other environmental conditions, such as whether the colonies are in a nectar flow or a dearth. Figuring out what causes the differences in efficacy will be an important hurdle to overcome if RNAi is actually going to be useful for beekeepers. But it’s not entirely clear that being useful to beekeepers is what’s motivating Monsanto.

In 2011, Monsanto acquired Beeologics – a biotech start-up that focused on using RNAi to control Israeli acute paralysis virus (IAPV) – which they’ve rebranded as BioDirect. Some of the work done by Beeologics has been influential in my own research, particularly a paper published by Wayne Hunter et al., which included Hayes as one of the co-authors. Published in the respected journal Plos Pathogens, they showed that after feeding colonies syrup containing double-stranded RNA that matched the IAPV sequence (which is itself a single-stranded RNA virus, much like mRNA), the bees were protected from virus infection. Remarkably, the data suggests that bees could retain this immunity for their whole adult life, possibly even passing it on to their brood. Beeologics named this RNA treatment “Remebee” (they seem to really like the bee puns) and even overcame one of the biggest hurdles in large-scale RNAi applications: fast, cheap double-stranded RNA synthesis. This large-scale RNA production technology is likely why Monsanto was interested in Beeologics – not for the virus treatment.

Alex Inberg’s talk was the last one I attended at the Entomological Society of America meeting in Denver, last November. There, he presented some of the data the Monsanto research team has acquired during their work to develop the RNA-based varroa control. “This isn’t meant to replace existing treatments,” he clarified. “It is meant to be integrated into existing IPM [integrated pest management] strategies.” In his experiments, he used colonies spanning the US from north to south and east to west, encompassing a wide range of climates and conditions. The results showed modest efficacy (the statistical cut-offs were more relaxed than what’s conventional) but were encouraging nonetheless.

With all the ambition for the varroa treatment, I began to wonder what ever became of the so-called Remebee formulation for IAPV, which years ago seemed so promising. So, in the post-talk question period, I asked. “Is Monsanto pursuing RNAi for virus treatments, and if so, what stage is it at in development?” Inberg’s response was that “honey bee virus control is not one of our priorities at this time.” Despite IAPV losing prevalence in recent years, this was still disappointing to me since the same technology is transferrable to controlling other, abundant viruses such as DWV (deformed wing virus). Furthermore, controlling the viruses achieves half the goal of controlling varroa. To me, this lack of interest in honey bee viruses somewhat confirms that Monsanto acquired Beeologics not to help protect honey bees, but to translate the same technology into control of crop pests, weeds and plant diseases, which is Monsanto’s usual modus operandi.

That’s not necessarily a bad thing, since RNAi could offer the same benefits in agriculture as it offers for honey bee colonies. It could even improve bee health – for both wild and managed species – indirectly through reduced usage of conventional, broad spectrum pesticides that persist in the environment. Peter Neumann, a researcher from the Institute of Bee Health, stated in my last American Bee Journal article, “Current pesticides use the shotgun approach. Instead, we need a sniper.” Since double-stranded RNA can be specifically designed to avoid harming beneficial insects, RNAi could be the sniper that agriculture needs, especially if its efficacy in the field can be improved. With an extremely short (think: hours to days) half-life, double-stranded RNA also poses a lower risk to the environment. But all this is beside my final musing, which is: Do we really need another miticide?

Certainly, there is high demand for more miticides – ones that aren’t (yet) overcome by resistance, ones that aren’t corrosive acids, and ones that don’t contaminate hive products with residues. Troy Anderson, from the University of Nebraska, presented work at the same conference as Inberg about developing more small molecule miticides that overcome one or more of the above concerns. His molecules – stilbene derivatives, for the chemistry enthusiasts – target a different mite protein, so varroa populations that are resistant to current miticides are still susceptible to the new ones. However, like virtually all lethal small molecules, resistance will develop eventually, just like we are seeing with fluvalinate, coumaphos and, possibly, Amitraz.

Although RNAi should avoid the resistance problem, it is still a “top-down” method of pest control (or middle-down, at best) – one that ideally should be left as one of the last resorts. If conventional miticides are “top-down” pest control, then what would be “bottom-up”? Methods that cause the bees to fight the mite themselves are bottom-up, such as selective breeding. Based on my November American Bee Journal article, it is probably apparent that I’m partial to this approach. Bottom-up methods usually involve working with the host, or the supply of sustenance for the pest (in this case, the bees that sustain varroa). Bottom-up methods also tend to be the long, hard road, whereas top-town methods kill the pest in a quick, easy fix.

Some might also say that bottom-up methods sound the best from behind a desk, but top-down methods sound the best from the field. A common balance needs to be struck between the desk and the field, such as making the bottom-up methods easier and more feasible. RNAi could help find that balance, but by putting more miticides on the market, we are also promoting the top-down method when the bottom-up strategy is more sustainable. True, in his presentation, Inberg clearly communicated that the RNA-based varroa treatment is not anticipated to be a silver bullet, and that it will work best if it’s incorporated into existing IPM. I don’t doubt those statements; however, IPM also tends to be harder to follow the more colonies you have to monitor.

Miticides certainly have their place in the varroa management scheme. We are not ready, yet, for all bee breeders to start the selective breeding process I discussed in November’s issue. We simply couldn’t handle the volume of samples to process if they did, and we’re not sure how well the strategy works outside of Canada (we’ll be testing that soon). And in the absence of selective breeding, naturally resistant populations cannot supply the industrial demand.

I consider current miticides to be placeholders – drugs that are buying us time to figure out more sustainable methods of pest control. Hopefully, new miticides are being developed because we need to buy more time, and not because people expect to find the magic varroa treatment. Whatever the motivation, I hope the prospects for new miticide development – RNA-based or otherwise – doesn’t distract us from having a go at the long, hard road.