Prospects for probiotics in social bees

Abstract

Social corbiculate bees are major pollinators. They have characteristic bacterial microbiomes associated with their hives and their guts. In honeybees and bumblebees, worker guts contain a microbiome composed of distinctive bacterial taxa shown to benefit hosts. These benefits include stimulating immune and metabolic pathways, digesting or detoxifying food, and defending against pathogens and parasites. Stressors including toxins and poor nutrition disrupt the microbiome and increase susceptibility to opportunistic pathogens. Administering probiotic bacterial strains may improve the health of individual bees and of hives, and several commercial probiotics are available for bees. However, evidence for probiotic benefits is lacking or mixed. Most bacterial species used in commercial probiotics are not native to bee guts. We present new experimental results showing that cultured strains of native bee gut bacteria colonize robustly while bacteria in a commercial probiotic do not establish in bee guts. A defined community of native bee gut bacteria resembles unperturbed native gut communities in its activation of genes for immunity and metabolism in worker bees. Although many questions remain unanswered, the development of natural probiotics for honeybees, or for commercially managed bumblebees, is a promising direction for protecting the health of managed bee colonies. This article is part of the theme issue 'Natural processes influencing pollinator health: from chemistry to landscapes'.

Keywords: Gilliamella; Nosema; Snodgrassella; apiculture; foulbrood; microbiome engraftment.

Figure 1.

Microorganisms associated with honeybees and their hives. Distinct microbial communities occupy different niches, including a distinct community in hindguts of adult worker bees and a set of microbes exchanged between hive surfaces, food reserves, larval guts, queen guts and worker foreguts. These communities have been shown to confer benefits for bees and hives, but they also include pathogens of larvae and/or adults. Some of these microbes are acquired from or exchanged at environmental sources including nectar and plant surfaces.

Figure 2.

Establishment of commercial probiotic bacteria versus native bee gut bacteria 5 days post-inoculation. (a) Total bacterial load estimated by qPCR as 16S rRNA copies per 100 ng RNA. Abbreviations: NP, no probiotic; ProB, probiotic; DC, defined community of native bee gut bacteria; or ProB + DC, mixture of ProB plus DC. All conditions have similar bacterial loads (p > 0.05, Kruskal–Wallis). (b) Relative abundances of bacterial lineages as determined by 16S V4 sequencing. Abbreviations: A1, Bartonella apis; A2.1, Commensalibacter spp.; A2.2, Bombella apis; S, Snodgrassella alvi; B, Bifidobacterium spp.; F4, Bombilactobacillus spp.; F5, Lactobacillus nr. melliventris; L, environmental Lactobacilli; G, Gilliamella spp.; F, Frischella perrara; E, Enterobacteriaceae; O, other uncategorized. NP and ProB groups are dominated by Enterobacteriaceae, and both groups fed DC are dominated by core bee gut bacteria present in the DC. (c) Average percentages of bacterial taxa grouped by origin. BGM, taxa of bee gut microbiota; ENV, environmentally associated taxa; ProB, taxa in commercial probiotic mixture. Data are available online in the electronic supplementary material.

Figure 3.

Expression of genes underlying immunity and hormonal signalling in bees lacking a native microbiota (MD), colonized with a defined community of native bee gut strains (DC), or colonized with gut homogenate from a donor bee (GH). (a) Bacterial load estimated as bacterial 16S rRNA copies and (b–d) relative transcript levels of genes encoding abaecin (b), apidaecin (c) and hymenoptaecin (d) in bee abdomens across groups (n = 11 per group). (e) Bacterial 16S rRNA copies and (f–k) relative transcript levels of genes encoding defensin 2 (f), vitellogenin (g), insulin-like peptides 1 (h) and 2 (i), and insulin receptors 1 (j) and 2 (k) in whole bee bodies across groups (n = 10 per group). The linear regression ‘lm’ option in the pcr package in R was applied to estimate differences between groups. ***p < 0.001, **p < 0.01, *p < 0.05. Data are available online in the electronic supplementary material. (Online version in colour.)