1 INTRODUCTION

The ecological success of eusocial insects is attributed to an organized and efficient division of labor (Oster & Wilson, 1978). Social insects solve complex problems with individual behaviors, resulting in emergent group properties (Hölldobler & Wilson, 2008). The numbers of workers performing a particular task are optimized by feedback loops to efficiently collect, process, and distribute resources among colony members (Fewell, 2003). Individual workers transition among various tasks during their lifetime, and exhibit a broad range of phenotypic plasticity. More simply, colony demography is socially regulated (Z.-Y. Huang & Robinson, 1996), allowing a proximate internal response to unpredictable external environments. Various worker tasks involve different physiological and behavioral demands, producing strong selection on social phenotypes. Social insects are well suited to the study of sociality and phenotypic plasticity because they represent a complex adaptive system or ‘superorganism” in which the functional parts can be manipulated and measured (Hölldobler & Wilson, 2008).
Honey bees are highly social insects that live in complex societies consisting of one reproductive queen and thousands of facultatively sterile workers. While the queen spends a preponderance of her life laying eggs, workers build and maintain all aspects of the hive. Under normal conditions, adult workers display age polyethism–performing tasks within the hive for the first 2–3 weeks before transitioning to outside tasks (Seeley, 1982). Specifically, young adults function as “nurse bees” that feed growing larvae. Pupae are capped with wax and isolated until they emerge from their cell and become the next cohort of nurse bees. Meanwhile, near the end of their lifespan, older nurses transition into foragers that then procure nectar (carbohydrates), pollen (protein and lipids), propolis (antimicrobial plant resins), and water (Seeley, 1982). This delicate balance of age-based division of labor requires a high level of colony coordination and integration. Despite this well-established pattern, adult workers can decouple age from behavioral task in response to social cues from other workers and pollination environment (Z. Y. Huang & Robinson, 1992). Thus, behavioral task is extremely plastic and nursing/foraging behaviors can be accelerated, slowed, or reversed (Robinson, 1992).
Phenotypic plasticity in honey bees workers is directly associated with the availability of nutrition and storage proteins, vitellogenin in particular (G. V. Amdam, Norberg, Hagen, & Omholt, 2003). Vitellogenin (vg) is a phospholipoglyco-protein evolved to serve many functions; as an antimicrobial, antioxidant, and to produce brood food in the in nurse worker head (hypopharyngeal) glands (Gro V. Amdam, 2011; Siri-Christine Seehuus, Norberg, Krekling, Fondrk, & Amdam, 2007). In contrast, foragers switch to a diet of simple sugars to support the metabolism associated with flight. This transition is associated with reduced lipid stores (Toth & Robinson, 2005), reduced Vg titers (Fluri, Lüscher, Wille, & Gerig, 1982), decreased nutritional status (Ament, Corona, Pollock, & Robinson, 2008), and differential gene expression (Ben-Shahar, Robichon, Sokolowski, & Robinson, 2002). These are functional differences that likely contribute to being a successful forager; a decrease in body mass and a proportional increase in flight capacity (Vance, Williams, Elekonich, & Roberts, 2009). However, leaving the relative safety of the hive is the riskiest time of an adult bee’s life. A recent study documented that 40% of bees die during the pre-foraging stage of life, a time where bees perform exploratory and learning orientation flights (Prado et al., 2020). Bees that survive face a constant increase in extrinsic mortality risk per unit time that increases to 100% after 18 days of foraging activity (Dukas, 2008), yet only ~20% of foragers will live past ten days of foraging (Visscher & Dukas, 1997). Therefore, the age a worker initiates foraging has a strong impact on lifespan and group contribution.
Foraging also has direct consequences for intrinsic senescence, including increased sensitivity to physiological stressors (Remolina, Hafez, Robinson, & Hughes, 2007) and a decrease in innate immune defenses (Gro V. Amdam et al., 2005; Gro V. Amdam, Simões, et al., 2004; Lourenço et al., 2019; Schmid, Brockmann, Pirk, Stanley, & Tautz, 2008). Foragers also show an increased susceptibility to oxidative stress (S.-C. Seehuus, Norberg, Gimsa, Krekling, & Amdam, 2006), including oxidative damage to the brain (Rueppell, Christine, Mulcrone, & Groves, 2007), and the body (Hsieh & Hsu, 2011). The accumulation of oxidative damage from reactive oxygen species (ROS) is proposed as the main cause of aging (Harman, 1956). Thus, a precocious transition to foraging is predicted to result in premature aging. Flight and the associated ROS accumulation from muscle usage and attrition may surpass the capacity for antioxidant enzymes to remove them. Indeed, the honey bee’s innate antioxidant enzymes; various superoxide dismutases, catalase, and glutathione S-transferase, reach their greater expression in older workers (Corona, Hughes, Weaver, & Robinson, 2005). While the physiology of behavioral plasticity and aging has been explored in honey bees, the role of the gut microbiome in this process is poorly known (Vonaesch, Anderson, & Sansonetti, 2018).
Given the strong selection acting on social phenotypes, here we test their physiological range in association to the gut microbiome. The honey bee gut microbiota is remarkably consistent and dominated by five omnipresent, highly co-evolved phylotypes representing >95% of bacterial cells (Kwong & Moran, 2016; Martinson, Moy, & Moran, 2012; Sabree, Hansen, & Moran, 2012). Recent work has revealed a strong association of the microbiome with worker physiology including the expression of insulin like peptides and vitellogenin (Engel et al., 2016; Kešnerová et al., 2017; Maes, Rodrigues, Oliver, Mott, & Anderson, 2016; Powell, Carver, Leonard, & Moran, 2021; Raymann, Shaffer, & Moran, 2017; Ricigliano et al., 2017; Zheng, Powell, Steele, Dietrich, & Moran, 2017). Although nurses and foragers share a core microbiota (Corby-Harris, Maes, & Anderson, 2014), microbial composition differs by behavioral task and may impact host physiology and health (Anderson et al., 2018). The worker gut microbiota is spatially organized, with the greatest abundance of bacteria found in the distal regions of the gut (Martinson et al., 2012). Specifically, the rectum comprises >90% of the bacterial cells in honey bees and is an incredibly stable niche (Jones et al., 2018; Kapheim et al., 2015; Kešnerová et al., 2020; Ludvigsen, Andersen, Hjeljord, & Rudi, 2020; Maes, Floyd, Mott, & Anderson, 2021). Therefore, we chose to focus on the midgut and ileum as these niches are more likely to reveal the limits of phenotypic plasticity in relation to the microbiome. The midgut functions in digestion and absorption of food, but it is also considered a less stable substrate for bacterial colonization because midgut epithelial cells secrete an envelope called the peritrophic membrane in early adult life (Engel & Moran, 2013b). The peritrophic membrane is continuously replaced as it is shed and serves to facilitate digestion, and as a protective barrier from damage by food particles and microbial invasion (Engel & Moran, 2013b). At the pylorus, the Malpighian tubules excrete nitrogenous waste and electrolytes from the hemolymph presenting a nutrient-rich environment for microbes (Cintra-Socolowski, Nocelli, Roat, Silva-Zacarin, & Malaspina, 2016; Cohen, Sawyer, Peterson, Dow, & Fox, 2020). Midgut function including the production of peritrophic membrane decreases significantly in foragers (Harwood & Amdam, 2021), but the impact on disease susceptibility and the gut microbiome are unknown.
It has been demonstrated in honey bees that individual worker behavior and physiology can be manipulated via the perturbation of social structure (Z.-Y. Huang & Robinson, 1996). Here we ask whether such changes are reflected in the gut microbiome. To test if gut microbiota differences are associated with behavioral task or age, we created “single-cohort colonies” (SCC) comprised of bees that were all the same age (Robinson, Page, Strambi, & Strambi, 1989). As a result, some proportion of the young bees in this colony will initiate foraging precociously to fill the missing ranks of normal-aged foragers. As the population ages, we assess differences in nurses and precocious foragers (PF) of the same age, monitoring gene expression related to immunity and oxidative stress. Likewise, we assessed protein oxidation in the fat body resulting from precocious foraging. We hypothesized that precocious foragers would accumulate oxidative damage at a faster rate than same aged nurse bees. We test two hypotheses: (1) honey bees experience differences in the midgut and ileum microbiota that reflect age and behavioral task and (2) oxidative damage and gene expression reflect differences within and among age and behavioral task.