In honey bee workers, aging depends on social behaviors rather than on chronological age. Here we show how worker-types with very different aging patterns can be obtained and analyzed for cellular senescence.
Societies of highly social animals feature vast lifespan differences between closely related individuals. Among social insects, the honey bee is the best established model to study how plasticity in lifespan and aging is explained by social factors.
The worker caste of honey bees includes nurse bees, which tend the brood, and forager bees, which collect nectar and pollen. Previous work has shown that brain functions and flight performance senesce more rapidly in foragers than in nurses. However, brain functions can recover, when foragers revert back to nursing tasks. Such patterns of accelerated and reversed functional senescence are linked to changed metabolic resource levels, to alterations in protein abundance and to immune function. Vitellogenin, a yolk protein with adapted functions in hormonal control and cellular defense, may serve as a major regulatory element in a network that controls the different aging dynamics in workers.
Here we describe how the emergence of nurses and foragers can be monitored, and manipulated, including the reversal from typically short-lived foragers into longer-lived nurses. Our representative results show how individuals with similar chronological age differentiate into foragers and nurse bees under experimental conditions. We exemplify how behavioral reversal from foragers back to nurses can be validated. Last, we show how different cellular senescence can be assessed by measuring the accumulation of lipofuscin, a universal biomarker of senescence.
For studying mechanisms that may link social influences and aging plasticity, this protocol provides a standardized tool set to acquire relevant sample material, and to improve data comparability among future studies.
The complex colony structures of highly social animals are maintained through the interaction of a reproductive caste, and a helper caste of typically non-reproducing workers with different social task behaviors. In the different workers, specific physiological adaptations enable distinct sib care behaviors, and are also linked to extreme lifespan differences. Honey bees and mole rats represent the best-developed animal models to study how sociality is linked to patterns of accelerated, negligible or reversed aging1-3.
In honey bee colonies, a single egg-laying queen is assisted by thousands of workers that tend the brood, forage for food, and engage in guarding, thermoregulation or hygienic behaviors4. Among these workers are the extremely short-lived foragers, nurse bees with intermediate, and winter (diutinus) bees with longest lifespans. Individuals, however, are not permanently bound to a certain worker-type, but display a flexible behavioral ontogeny: they change from one social task behavior to another (“temporal castes”). Callow bees can change to brood tending nurse bees, which eventually may change to outside foraging. However, callow nest bees can also transform into longest-lived winter bees, and short-lived foragers can even revert into typically longer-lived nurses. Workers with extreme (winter bees) and intermediate (nurse bees) lifespan have well-developed food production and storage organs with copious resources – as opposed to short-lived foragers (reviewed in1,5). However, that the regulation of individual lifespan goes beyond simple changes in an individual’s resource balance is suggested by research on a yolk protein, which has diverse adapted functions in the non-reproducing worker caste, such as jelly production6, hormonal control7, immune8 and anti-oxidant defense9.
Patterns of functional decline (senescence) mirror lifespan disparities among workers, as established for olfactory, and also for other brain or motor functions10-13. Specifically, the significant decline in learning function after only two weeks of foraging matches a similar mortality progression in foragers14, as opposed to the lack of detectable decline (negligible senescence) in long-lived winter bees15.
To identify the molecular fingerprints of flexible aging we adapted established experimental paradigms that allow for monitoring and manipulating aging-type transitions8,16,17. Experiment 1 details how to obtain samples in which the effects of chronological age and worker-type specific social behaviors on aging can be separated. Experiment 2 describes the reversal of foragers with accelerated into nurse bees with slowed aging dynamics. Experiment 3 provides an approach for probing effects of cellular senescence by anatomical quantification of an established biomarker for cellular aging (lipofuscin)18.
1. Decoupling Senescence from Chronological Age
This section describes the setup of double cohort colonies, which consist of a cohort of identified individuals that share the same chronological age (“single age cohort”) and a cohort of nest bees. Same aged individuals of the single age cohort will eventually separate into different worker-types with different aging dynamics – these are nurse bees with slowed and forager bees with accelerated functional decline. All procedures are described for one experimental colony. We advise, however, to perform experiments for at least two colony replicates so that colony effects can be controlled for (two-replicate-design).
2. Reversal of Workers with Rapid to Workers with Slowed Aging by Changing the Hive’s Demography
This section details how the reversal from workers with accelerated aging (foragers) to workers with slowed aging (nurse bees) is performed. Such behavioral reversal is induced, when foragers experience a lack of nurse bees, which normally engage in brood care. The reversion procedure will separate a single colony replicate into two hives: one hive with the nurse bee fraction (“nurse-derived”), and another one with the forager fraction (“forager-derived”). After successful reversal, possible symptoms of plastic and reversed aging can be studied in the single age cohort with reverted workers, continuing foragers, continuing nurse bees and newly recruited foragers. As before, identified bees of the single age cohort, not the cohort of unidentified nest bees, constitute the experimental focus group.
3. Analyzing Worker-type Specific Cellular Senescence Patterns by Quantification of Lipofuscin
Lipofuscin is a universal biomarker of cellular senescence. As an intrinsic accumulation product, lipofuscin’s specific autofluorescence (emissionmax = 530-650 nm) can be used for detection18.
Protocol sections 1 and 2 detail how test groups can be obtained to study attributes of accelerated, slowed and reversed aging in colonies with a single age cohort. To monitor worker-type differentiation that accompanies the normal ontogeny we assessed forager counts (“entrance counts”) for 6 colonies (Figure 1, compare section 1). The graphs show that considerable change from nurse to the forager state is typically not observed before individuals are more than 10 days old. Marked variability in forager counts was observed with regards to the timing of foraging onset among different colonies, and as a marked day-to-day variation within each colony. Apart from colony specific demographic factors, such as different brood load, much variability is explained by changing weather conditions (time points marked in red in Figure 1). Close monitoring of foraging dynamics therefore is advised to optimize marking and collection efforts during the experiment.
The reversed ontogeny (section 2) from foraging back to nursing tasks can be validated by inspecting brood combs that are introduced into the forager-derived colonies (see steps 2.2 and 2.3). For three replicates Figures 2A, C, and E show brood combs before introduction into forager-derived colonies. Figures 2B, D, and F show the respective combs after removal. Patches of newly capped brood, healthy larvae, and increased pollen storage around brood cells indicate that former foragers now had successfully performed typical nest, including nursing tasks.
Lipofuscin (section 3) is a highly conserved symptom of cellular senescence, and can be readily assessed for post-experimental analyses in the various bee tissues. Figure 3 contrasts lipofuscin accumulation, measured as granule size (Figure 3E), in the hypopharyngeal glands of age matched nurse and forager bees. The difference in chronological age between the two young and the two old groups was ≥17 days, with only one group (foragers) spending these ≥17 days with outside flight and food collection activities. Representative microscopic images (Figures 3A-D) show increased lipofuscin accumulation only for the group of older foragers after more than 17 days of foraging (Figure 3D), not for older nurse bees of similar chronological age (37-43 days; Figure 3B). A two-factorial ANOVA with the fixed main factors worker-type (foragers, nurses) and age difference (Δage ≥17 days) revealed significant effects for worker-type, age difference and the interaction between both factors (Ftype = 33.67, P<0.001; FΔage = 21.93, P<0.001; Ftype x Δage = 22.07, P<0.001). However, post-hoc tests showed significant effects only when contrasting older foragers (≥17 days of foraging) to younger foragers, or to both nurse groups (PF17d vs. F1d/N1d/N17d<0.001, Fisher’s LSD; Figure 3E). No difference was detected among the latter three groups, including chronologically young and old nurse groups (all tests with P >0.5, Fisher’s LSD; Figure 3E). This suggests that lipofuscin accumulation depends on forager specific activities (foraging age), rather than being function of chronological age only per se.
Figure 1. Worker-type differentiation during normal ontogeny. The graph displays entrance counts of foragers returning from foraging flights counted for 6 different colonies beginning 5 days after they were established (for details compare Protocol section 1.4). Considerable transition from nest to foraging activities was first observed when marked individuals of the single age cohort were about 10 days old. Varying slopes for the cumulative entrance counts indicate that the dynamics of the nurse bee to forager transition differ between colonies, and are affected by climatic factors. For example, on days with rain and less than two hours of foraging, the increase in entrance counts typically flattened out (data points in red).
Figure 2. Validating behavioral reversion. To test if foragers have successfully reverted to nursing tasks, we compared brood combs before they were introduced into forager-derived hives, and after they were removed from these hives. Representative images show brood combs before introduction (A, C, E) and after removal (B, D, F) from three different forager-derived hives, respectively. Brood care by previous forager bees is indicated by an increasing number of cells with capped brood (B, D, F; black arrow, inset in D), sustained survival of larvae in open cells (red arrow) and increased storage of pollen close to brood cells (white arrow). Note that forager-derived-colonies initially are typically less efficient in tending the brood than nurse-derived colonies. This can lead to higher larval mortality in the forager-derived-colonies. Pictures in B, D, F were taken 5, 4 and 7 days after brood combs were introduced into forager-derived colonies.
Figure 3. Accumulation of lipofuscin, a biomarker of cellular senescence, can indicate worker-type specific tissue deterioration. Representative microscopic images of hypopharyngeal glands in young (A) and old nurse bees (B), as well as in age-matched forager bees with ≥1 day (C), respectively ≥17 days foraging experience (scale bar in A = 20 μm). Lipofuscin accumulation was measured as granule size, and is given as medians and quartiles for N = 5 individuals for each age and worker-type (E). Foraging for 17 days resulted in significant lipofuscin accumulation, while the same period did not lead to lipofuscin changes in nurse bees (for statistics see Results).
We here adopt previously described approaches8,16,17,19,20, and integrate them into a single workflow that will facilitate studying flexible aging in honey bees. Our aim is to provide scientists that are novice to this field with a standardized tool set to obtain relevant sample material, and to improve experimental reproducibility among different research teams. While our procedures are simplified and do not require special equipment as in earlier descriptions (compare for example8), some measures of precaution are advised and are collected below.
Decoupling senescence from chronological age. A most critical aspect is to avoid false identification of forager bees during initial confirmation of foraging behaviors (2nd marking). Therefore, when foragers are to be monitored (“entrance counts”) or marked, do strictly avoid daily periods with orientation flights. During these periods many pre-foraging stage bees will depart from or enter the hive. These bees do not display typical physiological characteristics of mature foragers, but build up a spatial map of the hive surroundings by readily identifiable circular flight patterns21.
While most bees change to foraging with the age of 2 weeks and older, sporadic foraging is observed already at very young ages (Figure 1). Extremely precocious foragers typically develop directly from callow nest bees without having passed through the nurse stage. To not include individuals with such an aberrant ontogeny (compare22 and references therein), individuals that begin foraging with the age of 10 days or less are not considered for further analyses.
To further avoid overrepresentation of precocious foragers, we do not make use of classical “single cohort colonies” that consist only of the single age cohort17,23. Instead, when setting up colonies we add random nest bees (“nest bee cohort”) to the marked single age cohort (see steps 1.1 and 1.3). Since random nest bees are typically older, they can reduce the pressure on very young bees to develop into extremely precocious foragers17. Such double cohort colonies, therefore, may better resemble a natural hive demography with individuals that slowly progress from nursing to foraging.
When long-term worker specific adaptations are to be studied, collect all test groups outside foraging hours. This is advised to reduce bias by more acute metabolic adjustments due to recent locomotor activities, for example exhausting flight.
Reversal of workers with rapid to slowed aging by changing the hive’s demography. After foragers had flown back to the original location it is essential to move away the nurse-derived hive (>3 km). This is to avoid that pre-foraging stage bees are recruited and guided to the old location by other bees, respectively through pheromone communication24.
To further prevent any nurse or other pre-foraging individuals from entering the forager-derived hive, we advise keeping with the following rules: (I) Terminate the separation procedure before daily orientation flights begin. (II) Only attempt reversal on days when strong foraging activity is observed. (III) During and after the initial translocation of the original hive, avoid unnecessary agitation of bees, in particular do not open the hive.
In principle, more artificial setups that confine foragers in a nurse-deprived environment also may lead to reversion. However, such setups only have limited informative value as the forager-derived fraction experiences other stressful environments, thus precluding a direct comparison with control groups from nurse-derived colonies.
Confirming different senescence patterns by quantification of lipofuscin, a biomarker of cellular senescence. Here we exemplified lipofuscin assessment with images and statistical data of hypopharyngeal glands because lipofuscin is most easily detectable in this tissue. This, we believe, is important to help the inexperienced observer setting up the correct protocols for microscopic detection. However, unlike other tissues, hypopharyngeal glands do display significant apoptosis and necrosis during nurse to forager transition25. Such processes may interact with accumulation of the senescence marker, even though we did not detect increased levels of lipofuscin in young foragers that recently had changed from nursing tasks (Figures 3C, E). However, to assess senescence measures in other bee tissues, the microscopy-based methods described here can be easily adapted.
Alternatively, flow-cytometric approaches are less time consuming26. Microscopy-based analyses have the advantage that cellular aging symptoms can be assessed for different regions or even for cells within a single organs27. For studies in brain and other complex organs with spatial heterogeneity in cellular aging28, we therefore recommend the microscopy based approach.
The authors have nothing to disclose.
We thank Osman Kaftanoglu for helpful advice and assistance during filming. We would like to thank the anonymous reviewers for insightful comments. This work was supported by the Research Council of Norway (grants 180504, 191699, and 213976), Marie Curie/FP7 (project ref. 238665), the National Institute on Aging (grant NIA P01 AG22500), and the Pew Charitable Trusts.
Name of reagent | Company | Catalogue number | Comments |
Apifonda | Südzucker AG, Mannheim/Ochsenfurt, Germany | ||
paraformaldehyde | Sigma-Aldrich | 158127 | |
phosphate-buffered saline | Sigma-Aldrich | P4417 | |
Glycerol | Merck | 1.04094.1000 |