Image: Waldan Kwong
The honey bee gut is home to hundreds of millions of microbes – some of which help the host, and some of which destroy it.

Has a round of antibiotics ever given you an upset stomach or an uncomfortable yeast infection? These are common side effects, but they’re not caused by the drugs directly; rather, the antibiotics kill our friendly bacteria along with the ones giving us grief. They don’t discriminate. The niche that was once filled with harmless and beneficial bacteria are invaded by selfish microbes, causing the unpleasant symptoms as an indirect side effect of antibiotics. Eventually the good microbes recolonize the gut, returning it to the normal, balanced microbial community – or microbiome – it once was.

Honey bees are much the same. While estimates place the human gut microbiome at tens of trillions of microbial cells – maybe even 100 trillion, made up of up to 1,000 different species – honey bees are thought to have just one billion, composed of just 8–10 core species (Figure 1). Altogether, that’s still close – in sheer mass – to half the bee’s own brain, purely made up of bacteria and other microbes.

Figure 1. The gut microbiome of the adult honey bee. A. Schematic of the adult honey bee gut. B. Approximate order of abundance of the core gut microbiome bacteria in the hindgut. Adapted from Kwong et al. (2016).

Similar to our own microbiome, these bacteria have important roles such as helping to fend off parasites and breaking down complex molecules into nutrients. Modern research now shows that the gut microbiome is more critical to honey bee health than we had previously thought.

The importance of getting to know this simple microbial ecosystem wasn’t lost on Nancy Moran, who is a long-standing aphid researcher currently running a bustling lab at the University of Texas at Austin. She was always interested in this idea of ‘symbiosis’ – particularly, the intimate relationship between a host and its microbes – and how it evolved. Moran saw a great opportunity for studying the symbiosis within honey bees and began pioneering the field of honey bee microbiome research at a time when it was underappreciated. But with her leading the way, researchers quickly realized it was a gold mine.

What we learn about the honey bee gut bacteria will not only let us help our bees, it could also be applied to more complex systems, like that of our own gut. Moran was one of the first to begin studying how the gut microbiome established, and what having a healthy microbiome does for the host. From 2003 onward, experiment after experiment painted the picture of the honey bee gut microbiome as being a surprisingly simple and stable community – a prime system for scientific study. And as it turns out, the relationship between these microbes and their host is ancient.

In 2010, Moran and other researchers found that eusocial Bombus and Apis species have a very similar core microbiome (i.e. they are colonized by many of the same bacterial species), whereas solitary bees have very different, sporadic communities. This suggests that there could be a deep evolutionary link between sociality and microbes, where these stable microbial communities developed because colonies – with their continual food-sharing and fecal-oral contact – offer a steady, reliable transmission system. The common ancestors to eusocial bee species probably began acquiring their bacteria well before the Bombus and Apis lineages diverged. With so long to coexist, these microbes are also remarkably well-adapted to live in harmony, and the best example of this is the co-dependence of two main bacterial players.

Gilliamella apicola and Snodgrassella alvi are two of the most ubiquitous honey bee gut inhabitants. Waldan Kwong, who was a student in Moran’s lab at the time (now a post-doctoral researcher at the University of British Columbia), wanted to understand the relationship between these two types of bacteria. “Snodgrassella forms a tight layer on the inner lining of the gut, and Gilliamella forms another layer on top of it,” Kwong explains. They almost always colonize the gut in this pattern, and their metabolism is extremely well-adapted for this specific situation. “The gut wall is more aerobic [oxygen-rich], whereas closer to the lumen is more anaerobic [oxygen-poor]. So Snodgrassella gets most of its energy through aerobic respiration, whereas Gilliamella gets most of its energy through anaerobic fermentation.” These tendencies to metabolize molecules differently is more than just an adaptation to the demands of living in slightly different local environments. It’s built right in to their DNA.

Gilliamella turns simple carbohydrates into energy through glycolysis – a type of sugar metabolism that occurs in virtually all living things. Normally, when glycolysis is finished, the leftover energy-containing molecules are fed into a second energy-production pathway: the Krebs cycle. But Gilliamella is missing the Krebs cycle genes. Instead, Gilliamella essentially hands the leftover molecules to Snodgrassella, which does have the necessary genes for the Krebs cycle, but is missing the genes for glycolysis. It’s a metabolic match made in heaven.

Kwong was also interested in the bigger picture. What other bacterial genes are beneficial for living in the gut? He and his colleagues systematically deleted each gene and looked for how that affected gut colonization. “In total, we found about 400 genes that were beneficial, so I reconstructed them into their networks and pathways. They were mostly involved in metabolism and stress responses. The gut is a challenging environment to live in.” Indeed, at times the gut can resemble a battlefield, with the honey bee caught in the middle of multiple wars fought between good microbes, pathogens, and antibiotics.

One defense that honey bees have against intruders is a kind of molecular army – antimicrobial peptides – made up of miniature proteins that murder bacteria, and sometimes fungi or protozoans (complex microbes, like amoebas), by penetrating and disrupting their cell membranes or cell walls. And a healthy microbiome is necessary to activate these defenses.

“We knew that exposure to foreign bacteria stimulates pathways that lead to the production of antimicrobial peptides. But we didn’t know if the core microbiome bacteria would do this too.” Kwong found that when newly emerged workers – which lack any appreciable gut colonization – were inoculated with a normal microbiome (achieved by feeding them their sisters’ poop – yes, really – mixed with sugar syrup), the bees produced elevated amounts of two key antimicrobial peptides compared to controls (bees which were only fed syrup). So, gut colonization is important for activating this part of a honey bee’s immune system.

This seems like bad news for the beneficial gut bacteria, since they are also susceptible to “death by antimicrobial peptide,” but clearly, they can still colonize the gut to an extent. This immune stimulation appears to be a tool that allows honey bees to moderate the bacterial population – it doesn’t kill them all off, but it also doesn’t let them boom. By keeping the population from in check, everyone benefits in the long term: the microbes don’t overwhelm the host and lose their home, and the host maintains the proper balance of pH, nutrients, and waste in their gut. And for the honey bees, this immune stimulation also has another hidden benefit: defense against pathogens.

It’s important to keep the population of regular bacterial residents balanced, but controlling pathogenic bacteria is a matter of life or death. European foulbrood (EFB; Melissococcus plutonius) and American foulbrood (AFB; Paenibacillus larvae) are both detrimental bacteria that infiltrate the gut of larvae (Figure 2). Unlike in adults, the larval midgut is disconnected from the hindgut until just prior to pupation, so everyone – pathogens and bacterial residents alike – must battle for space in the same chamber.

Figure 2. Honey bee gut pathogens. A. Larva and adult gut pathogens. B. Approximate pattern of gut colonization over time by the core microbiome members (non-pathogenic bacteria). The larva gut is colonized gradually until the gut lining is shed at pupation. When the adult bee emerges, colonization commences once again, stabilizing after about 4 days. Adapted from Kwong et al. (2016).

If the EFB and AFB kill the larva, then the resident bacteria lose their home; of course, the larva itself is also fighting for its life, and thankfully it has a counter-attack. The same antimicrobial peptides that regulate the microbiome can also act against pathogens. When infected with EFB and AFB, larvae will (not surprisingly) produce even more of these peptides. But by then it is often too late.

We don’t know for sure, but if the resident microbiome primes the innate immune system in larvae like Kwong and Moran found for adult bees, they could help prevent brood diseases before they start. Unfortunately, EFB and AFB circumvent this defense by infecting very young larvae, which don’t yet have a well-established microbiome, and therefore haven’t fully mustered their antimicrobial army.

This difference in the larval and adult microbiomes could be one reason why adult bees, who have fully activated innate immune defenses, tend to harbor EFB, yet remain asymptomatic. A recent survey of apiaries across Canada, performed by the National Bee Diagnostic Center, found that a surprisingly high number of colonies (around 40%) had adult bees harboring the EFB bacterium, yet very few of the colonies showed signs of infection (Carlos Castillo, personal communication).

But just because a balanced microbiome stimulates production of beneficial antimicrobial peptides, this doesn’t mean that feeding bees probiotics – which claim to activate bees’ immune system or promote gut colonization – will have a positive impact on bee health. There has been very little research investigating whether probiotic formulae are beneficial for honey bees, and some of the research that has been done suggests that we should be extremely cautious when trying to manipulate the normal microbial community.

Ryan Schwarz, Nancy Moran, and Jay Evans showed that when they supplemented bees with a single member of the gut microbial community (our old friend Snodgrassella), rather than with a complete inoculum of a balanced community, this caused enough of a microbiome imbalance (or ‘dysbiosis’) to make the bees more susceptible to a troublesome trypanosomatid parasite. This was true even for hive-reared bees, which weren’t under any other unusual stress. In their own words, this research is a “cautionary tale to the arbitrary use of probiotics in animal health management.” Rigorous research must be conducted before we can use probiotics confidently and responsibly. Beekeepers may want to exercise some caution in believing all the claims made in probiotic marketing materials.

Regarding EFB and AFB, antibiotics (e.g. Oxytet® or Tylosin®) are one common treatment strategy that has been shown to be effective, depending on the type and stage of infection. Even asymptomatic colonies are sometimes treated (although usually not recommended) – for example, if you don’t trust the source of your package, want to prevent acquiring foulbrood from your neighbour, or are using the “shook swarm” method of getting your bees off their old AFB-contaminated comb. But more research by Moran’s group suggests that we may want to think twice about that, too.

Kasie Raymann, Zack Shaffer, and Nancy Moran found that tetracycline treatments can significantly reduce honey bee lifespan and drastically alter the gut microbiome. In cage trials, they treated adult workers with either sugar syrup or syrup laced with tetracycline, and measured the changes in the microbial community over time. They found that about half of the core gut species were significantly depreciated after 5 days of treatment – unsurprising, given that antibiotics are built to kill bacteria. But what does this mean for the bees’ health?

Ten days after antibiotic treatment, about half the treated bees were dead, compared to only one fifth of untreated bees. It’s not clear exactly why, but some of Moran’s other research suggests it could be due to nutrient cycling, since bees with a deficient microbiome gain about 20% less body weight post-emergence than bees with a healthy microbiome. They also produce about five times less vitellogenin – a protein linked with the increased longevity of winter bees. Adding to the injury, the microbiome deficiency causes our favourite molecular army – the antimicrobial peptides – to go down, leaving the bees in a state of immunosuppression, too.

Of course, this cage trial is an extreme scenario. When beekeepers apply antibiotic treatments, half their colony doesn’t die. And in many cases, the risk of letting a case of foulbrood run rampant is bigger than the risk of temporarily reducing the bees’ longevity. Moran and her colleagues used an antibiotic dose that’s comparable to what’s used in beekeeping, but they probably observed such extreme effects on longevity because laboratory cages stress out the bees and exacerbate the negative effects of the antibiotic. Indeed, Schwarz and his colleagues found that just the stress associated with living in lab cages made bees more susceptible to a trypanosomatid parasite than hive-reared bees, independent of any changes to the microbiome. Regardless of caveats associated with cage experiments, Raymann, Schaffer, and Moran’s work is still very useful information: it suggests that secondary stressors – say, Nosema, which tetracycline doesn’t kill – could be detrimental to a hive that’s already been given antibiotics.

Unlike EFB and AFB, Nosema is not a bacterial pathogen. Rather, it’s a ‘microsporidian,’ something that’s more akin to a fungus than a bacterium, and targets the adult bees. Nosema infects midgut epithelial cells, where it reproduces and amplifies, rather than in the lumen like other pathogenic and resident microbes. Since Nosema occupies such a different niche, why doesn’t it invade the hindgut epithelial cells too? Right now, the dominant hypothesis is that the epithelial cells in the midgut are the only ones that lack a cuticle layer, making it easier for the Nosema to penetrate and invade. There also happens to be no Gilliamella or Snodgrassella biofilm layer in the midgut, which could deny Nosema access, but hasn’t been experimentally shown. Either way, midgut cells are easy pickings for these microsporidia.

Since Nosema has to invade a host cell to reproduce, it’s kind of like a virus – it doesn’t have everything it needs to replicate on its own, so it hijacks the host’s resources instead. It’s sneaky like a virus in one other way, too: it shuts down the host’s defenses. Those antimicrobial peptides on which bees so often depend are weakened by Nosema, healthy microbiome or not. As you might expect, this weakening of antimicrobial defenses gets even worse if the resident microbiome has already been diminished by antibiotics, as researchers at the USDA Agricultural Research Service published last fall. There, Judy Chen, Jay Evans, and colleagues inoculated bees with either an antibiotic, Nosema, a mix of the two, or plain sugar water, and measured abundances of three major antimicrobial peptides. “Antimicrobial peptides can act against not only bacteria but also viruses, fungi or parasites,” Chen explains. All three of the peptides they measured were suppressed by antibiotic treatment, and antibiotics and Nosema together produced the strongest effect of all.

“The study used some really broad-spectrum antibiotics, not the typical beekeepers mix here in the US,” Evans – who co-authored the studies with Chen – cautions. But combined with what we know from Raymann’s experiments with tetracycline, we should probably still be using antibiotics more judiciously than current practices. “Routine prophylactical treatment of bee colonies with antibiotics is not an acceptable practice, and can lead to selection of drug resistant bacteria, fungal infections and antibiotic-associated complications,” says Chen.

Fumagillin, which is an antifungal agent used to treat Nosema itself, could be having negative effects as well. “Fumagillin increases energy demand, oxidative stress, susceptibility to diseases, and decreases other fitness traits,” Chen elaborates. “Application of antibiotics in bee colonies should be carefully considered, and possible negative side effects should be taken into account.”

It seems like the list of bee gut pathogens may never end, and there’s one more: chalkbrood. Unlike Nosema, the chalkbrood (Ascosphaera apis) fungus goes after the gut lumen of larvae – in direct competition with the resident microbes. As we’ve seen with EFB and AFB, the younger larvae tend to be more susceptible. Once infected, the larva activates production of the usual antimicrobial peptides, which chalkbrood can’t control like Nosema can – but in the process, the larva uses up its nutrient resources. As the fungus takes over, the larva also stops eating, and starves to death within a day or two. There is surprisingly little research investigating the relationships between chalkbrood, resident microbes, and antibiotics (although there has been some work on Megachile rotundata, the leaf-cutter bee) but my guess is that will quickly change as this field of research continues to take off.

We are still only beginning to understand the bee gut microbiome and its relationship with the host. Probiotics may prove to be a useful supplement in the future, but currently, research is lacking. Antibiotics, too, certainly have a time and a place, but recent data is urging us to be more cautious, especially regarding prophylactic treatment. We are not only beekeepers, we are also microbe-keepers – it’s becoming increasingly clear that we need to consider the impacts on the microbial community that conventional treatments, like antibiotics, may have. For honey bees, the outcome can be much worse than a yeast infection.

This article appeared in the April 2018 issue of American Bee Journal.

Sending thanks to Waldan Kwong for helping me edit this article.