The Neuroproteomic Basis of Enhanced Perception and Processing of Brood Signals That Trigger Increased Reproductive Investment in Honeybee (Apis mellifera) Workers

Abstract

The neuronal basis of complex social behavior is still poorly understood. In honeybees, reproductive investment decisions are made at the colony-level. Queens develop from female-destined larvae that receive alloparental care from nurse bees in the form of ad-libitum royal jelly (RJ) secretions. Typically, the number of raised new queens is limited but genetic breeding of "royal jelly bees" (RJBs) for enhanced RJ production over decades has led to a dramatic increase of reproductive investment in queens. Here, we compare RJBs to unselected Italian bees (ITBs) to investigate how their cognitive processing of larval signals in the mushroom bodies (MBs) and antennal lobes (ALs) may contribute to their behavioral differences. A cross-fostering experiment confirms that the RJB syndrome is mainly due to a shift in nurse bee alloparental care behavior. Using olfactory conditioning of the proboscis extension reflex, we show that the RJB nurses spontaneously respond more often to larval odors compared with ITB nurses but their subsequent learning occurs at similar rates. These phenotypic findings are corroborated by our demonstration that the proteome of the brain, particularly of the ALs differs between RJBs and ITBs. Notably, in the ALs of RJB newly emerged bees and nurses compared with ITBs, processes of energy and nutrient metabolism, signal transduction are up-regulated, priming the ALs for receiving and processing the brood signals from the antennae. Moreover, highly abundant major royal jelly proteins and hexamerins in RJBs compared with ITBs during early life when the nervous system still develops suggest crucial new neurobiological roles for these well-characterized proteins. Altogether, our findings reveal that RJBs have evolved a strong olfactory response to larvae, enabled by numerous neurophysiological adaptations that increase the nurse bees' alloparental care behavior.

Keywords: Apis mellifera; Label-free quantification; antennal lobes; brain proteome; mass spectrometry; mushroom bodies; neurobiology; quantification; reproductive investment; tissues.

Graphical abstract

Fig. 1

The workflow chart represents experimental procedure. Honeybee photos are provided by Professor Jianke Li. Brain photo is quoted from the book: Honeybee-Neurobiology-and-Behavior.

Fig. 2

Cross-fostering experiment between Italian bees (ITBs) and high RJ producing bees (RJBs) by grafting ITB and RJB larvae into colonies of the same or opposite stock.A, RJ in the queen cell cups of RJBs (panel a) and ITBs (panel b) 72h after larval grafting. B, Comparison of larval acceptance between ITB and RJB nurse bees with larvae of ITBs or RJBs (Mean±S.D., n = 15). C, Comparison of RJ production of per queen cell between ITB and RJB colonies with larvae of ITBs or RJBs (Mean±S.D., n = 15). “**” represents p < 0.01. IN: ITB nurse bees, RN: RJB nurse bees, IL: ITB larvae, RL: RJB larvae.

Fig. 3

Proboscis extension response (PER) of nurse bees of two honey bee strains (high royal jelly producing bees, RJBs) and Italian bees (ITBs) in response to brood odors. RJB and ITB nurses with differ in reproductive investment and this experiment showed a major difference in initial response and a minor difference in learning between the two bee stocks.A and B, show the experimental procedure of olfactory PER conditioning with larval odors. C, Percentage of individuals exhibiting conditioned PER to larval odor cues differed significantly between the nurse bees of ITBs and RJBs (Mean±S.D., “**” represents p < 0.01, “*” represents p < 0.05, n = 63 nurse bees for three replicates in total per stock). Spontaneous PER after stimulation with larval odors (trial 1) was significantly higher in RJBs than in ITBs, explaining all subsequent differences.

Fig. 4

Comparison of the proteome of antennal lobes (ALs) between newly emerged bees (NEB) from unselected Italian (ITBs, 3 pools each including 200 individual samples) and high royal jelly producing (RJBs, 3 pools each including 200 individual samples) honey bee stocks.A, Venn diagram performing the shared and unique protein groups identified in ITBs and RJBs (>97% are shared). B and C, The enriched functional classes, and pathways in quantitatively up-regulated proteins in the ALs of RJBs relative to ITBs (S0 = 0.1, FDR = 0.05). The percentage of genes/term represents the proportion of genes enriched in the respective functional group. Identical color summarizes bars from the same functional group. For details of the enrichment analysis results, see supplemental Table S4. “*” represents p < 0.05; “**” represents p < 0.01. D, Confirmation of different abundance of candidate proteins in the ALs of newly emerged bees of ITBs and RJBs by western blots. β-actin is used as a control. E, Normalized fold changes of selected protein abundances in the ALs of newly emerged RJBs (n = 3) compared with ITBs (n = 3), tested by western-blots.

Fig. 5

Volcano plot of differentially abundant proteins. Differentially expressed proteins were obtained considering p < 0.05, S0 = 0.1.A, Volcano plot of differentially abundant proteins from the antennal lobes (ALs) comparison of unselected Italian bees (ITBs) and high royal jelly producing bees (RJBs) in newly emerged bees (NEB). B, Volcano plot of differentially abundant proteins from the mushroom bodies (MBs) comparison of ITB and RJB newly emerged bees. C, Volcano plot of differentially abundant proteins from the ALs comparison of ITB and RJB nurse bees (NB). D, Volcano plot of differentially abundant proteins from the MBs comparison of ITB and RJB nurse bees.

Fig. 6

Comparison of proteome of mushroom bodies (MBs) between newly emerged bees (NEB) from unselected Italian (ITBs) and high royal jelly producing (RJBs) honeybee stocks.A, Venn diagram representing the shared and unique protein groups identified in ITBs and RJBs (>97% are shared). B, The enriched functional classes and pathways of quantitative comparison by up-regulated proteins in the MBs of RJBs relative to ITBs (S0 = 0.1, FDR = 0.05). The percentage of genes/term represents the proportion of genes enriched in the respective functional group. Identical color summarizes bars from the same functional group. For details of the enrichment analysis results, see supplemental Table S7. “*” represents p < 0.05; “**” represents p < 0.01. C, Confirmation of different abundance of candidate proteins in the MBs of newly emerged bees of ITBs and RJBs by western blots. β-actin is used as a control. D, Normalized fold changes of selected protein abundances in the MBs of RJBs (n = 3) newly emerged bees compared with ITBs (n = 3), tested by western-blots. E, Immunostaining of brain sections with antibodies of Hex110 of newly emerged bees. Control staining with 4', 6-diamidino-2-phenylindole (DAPI). The red fluorescence represents the respective proteins, stained with Cy3-conjugated antibodies. Areas of particularly pronounced quantitative differences between RJB (n = 3) and ITB (n = 3) samples are indicated by white arrows. The scale bars for whole brain sections represent 200 μm; the scale bars of the MBs and ALs represent 50 μm.

Fig. 7

Comparison of the proteome of antennal lobes (ALs) between nurse bees (NB) from unselected Italian (ITBs, 3 pools each including 200 individual samples) and high royal jelly producing (RJBs, 3 pools each including 200 individual samples) honey bee stocks. A, Venn diagram representing the shared and unique protein groups identified in ITBs and RJBs (>97% are shared). B, The enriched functional classes and pathways of quantitative comparison by up-regulated proteins in the ALs of RJB nurse bees relative to ITBs (S0 = 0.1, FDR = 0.05). The percentage of genes/term represents the proportion of genes enriched in the respective functional group. Identical color summarizes bars from the same functional group. For details of the enrichment analysis results, see supplemental Table S10. “*” represents p < 0.05; “**” represents p < 0.01. C, Normalized fold change of selected protein abundance of the ALs of RJB nurse bees (n = 3) compared with ITBs (n = 3), based on all western-blot data. D, Confirmation of different abundance of candidate proteins in the ALs of nurse bees of ITBs and RJBs by western blots. β-actin is used as a control.