J Comp Physiol A (2017) 203:143–149 DOI 10.1007/s00359-017-1147-y ORIGINAL PAPER Sexual response of male Drosophila to honey bee queen mandibular pheromone: implications for genetic studies of social insects Justin R. Croft1 · Tom Liu1 · Alison L. Camiletti1 · Anne F. Simon1 · Graham J. Thompson1 Received: 11 August 2016 / Revised: 31 December 2016 / Accepted: 9 January 2017 / Published online: 1 February 2017 © Springer-Verlag Berlin Heidelberg 2017 Abstract Honey bees secrete a queen mandibular phero- Keywords Chemical ecology · Social pheromones · Apis mone that renders workers reproductively altruistic and mellifera · Fertility signals · Social evolution drones sexually attentive. This sex-specific function of QMP may have evolved from a sexually dimorphic sign- aling mechanism derived from pre-social ancestors. If so, Introduction there is potential for pre-social insects to respond to QMP, and in a manner that is comparable to its normal effect on The essence of a eusocial breeding system is a reproduc- workers and drones. Remarkably, QMP applied to female tive division of labour, in which one or more individuals Drosophila does induce worker-like qualities [Camiletti monopolize reproduction, while the remaining majority et al. (Entomol Exp Appl 147:262, 2013)], and we here acts as reproductively altruistic helpers (Crespi and Yanega extend this comparison to examine the effects of bee pher- 1995). This reproductive coordination is predicted from omone on male fruit flies. We find that male Drosophila inclusive fitness theory (reviewed in Bourke 2011; Mar- melanogaster consistently orient towards a source of queen shall 2015), but the precise mechanism governing social pheromone in a T-maze, suggesting a recruitment response task specialization varies among species (Beshers and comparable to the pheromone’s normal effect on drones. Fewell 2001). The European honeybee Apis mellifera main- Moreover, exposure to QMP renders male flies more sex- tains its division of labour through queen and brood phero- ually attentive; they display intensified pre-copulatory mones that signal queen fecundity to her worker daughters behavior towards conspecific females. We can inhibit this (Le Conte and Hefetz 2008; Kocher and Grozinger 2011). sexual effect through a loss-of-olfactory-function mutation, Queen mandibular pheromone (QMP) is likely an evolu- which suggests that the pheromone-responsive behavioral tionarily honest signal (Gadagkar 1997; Keller and Nonacs mechanism is olfactory-driven. These pheromone-induced 1993; Oi et al. 2015; Peso et al. 2015) to which the work- changes to male Drosophila behavior suggest that aspects ers have been indirectly selected to respond by shutting off of sexual signaling are conserved between these two dis- their ovaries and otherwise adopting reproductively altruis- tantly related taxa. Our results highlight a role for Drosoph- tic roles within the colony (Butler and Fairey 1963; Hoover ila as a genetically tractable pre-social model for studies of et al. 2003; Backx et al. 2012). social insect biology. While this sterility-inducing function of queen phero- mone is well understood (Mullen and Thompson 2015), it has a broader role within the honeybee society, and the Electronic supplementary material The online version of this function of QMP can vary with social context. For exam- article (doi:10.1007/s00359-017-1147-y) contains supplementary material, which is available to authorized users. ple, QMP can induce worker retinue formation, inhibit queen rearing and swarming, and excite congregating * Graham J. Thompson males (drones) to mate with virgin queens in flight (Sles- [email protected] sor et al. 2005; Free 1987). In this later context, QMP is 1 Biology Department, Western University, 1151 Richmond a mating attractant to which the males orient, and from Street, London, ON N6A 5B7, Canada which they anticipate sex (Gary 1962). As a consequence, a Vol.:(0123456789)1 3 144 J Comp Physiol A (2017) 203:143–149 QMP-emitting queen may receive sperm from more than a using established Drosophila choice and courtship dozen drones on a single mating flight (Tarpy and Nielsen assays, and do so using two control lines and one olfac- 2002). From these observations, it is clear that Apis mel- tory mutant genotype that we expect to differ in their lifera queen mandibular pheromone can coordinate more responsiveness to olfactory cues. than one aspect of reproductive behavior (Brockmann et al. 1998). Like other Hymenopteran queen pheromones (Van Oystaeyen et al. 2014), QMP may have evolved through modification of fertility signals already present in pre-social Methods ancestors (Chapuisat 2014; Peso et al. 2015; Oliveira et al. 2015). Unlike other Hymenoptera studied thus far, how- Fly rearing and pheromone treatment ever, Apis mellifera queen mandibular pheromone appears to be relatively derived in its blended chemical composi- We reared all strains of Drosophila melanogaster under tion of volatile carboxylic acids (9-ODA and 9-HDA), aro- standard conditions (25 °C, 60% humidity and a 12 h matics (HOB and HVA), and other compounds, as opposed light: dark cycle) in an insect growth chamber (Caron to a composition of structurally simple linear or branched Inc., Marietta, OH) on a standard cornmeal diet as alkanes typical of ancestral ants, bees, and wasps (reviewed described in Camiletti et al. (2013). We used the Ore- in Oi et al. 2015). gon-R (Ore-R, Bloomington Stock Center #2376) strain Despite the apparent complexity of QMP, there is of flies as control, for which females have been shown emerging evidence that the taxon-specific qualities of this to respond to QMP in a worker-like manner (Camiletti 1 pheromone are, nonetheless, effective at suppressing repro- et al. 2016). In addition, we included the Orco mutant 1118 1 duction via ovary de-activation in un-related insects. Cami- strain in our trials, outcrossed 6 times in w . Orco letti et al. (2013) showed, for the first time, that female is homozygous for loss-of-function alleles at the major fruit flies (Diptera) exposed to a synthetic blend of QMP olfactory co-factor Orco locus (formerly Or83b; Larsson respond in a manner comparable to queenright worker bees. et al. 2004) and is accordingly non-responsive to a wide That is, they develop smaller ovaries that contain fewer range of olfactory stimuli (Steck et al. 2012). For all gen- 1 mature eggs, as if they were partially sterile. Moreover, this otypes—that is, Ore-R, Orco , and its genetic background 1118 response to ovary-inhibiting pheromone has a genetic com- control w —we synchronized adult emergence by ponent: Drosophila sitter flies (forS) are more responsive housing (for 24 h) a small reproductive population (n = 30 to the ovary-inhibiting pheromone than are rovers (forR), males and n = 30 females) in collection cages (60 mm; suggesting that the foraging gene may in some way medi- Diamed, Mississauga, Canada) fitted with nutrient (grape ate this response (Camiletti et al. 2014). These findings are juice and agar) plates. We then collected and transferred intriguing in that they suggest an underlying pheromone- day-old larvae to fresh food vials (28.5 × 95 mm, VWR responsive mechanism that is genetically variable, yet suffi- International, Radnar, PA) at a density of n = 30 larvae ciently preserved in flies and potentially other invertebrates per vial. Finally, we collected same-age, same-sex adult (Nayer 1963; Carlisle and Butler 1956; Sannasi 1969). virgins at eclosion (within 1 h), and reared males for 3 There is, therefore, potential to exploit the Drosophila days until sexually mature. To prepare filter papers prior model to uncover conserved genes, pathways, and neural to treatment, we first warmed (50 °C in water bath; to ren- circuits involved in ovary-suppression and potentially other der it liquid from a soft wax) and diluted a 500 mg stock fertility-linked traits (Camiletti and Thompson 2016). of synthetic QMP (Contech Ltd, Victoria, Canada) with Motivated by the prospect of harnessing Drosophila absolute ethanol into working aliquots of ~13 Qeq units. gene-finding tools toward questions relevant to honeybee For reference, 1 Qeq is roughly the amount produced by behavior and biology, we seek to test the effect of syn- a single living honey bee queen in 24 h. With bioaccu- thetic QMP on male Drosophila. We postulate that males mulation, the pheromone can persist in wax, resin, and might show drone-like behavior towards QMP. Specifi- cuticles within a colony at multi-queen equivalent doses cally, we test if male Drosophila respond to biologically (Hoover et al. 2003; Naumann et al. 1991). We loaded relevant doses of QMP by orienting themselves towards one full aliquot in a final volume of 20 µl onto individual a source of queen bee pheromone—i.e., ‘the queen’, or if pieces (200 mm2) of filter paper (grade 413; VWR Inter- males anticipate sex by increased mating effort towards a national, Radnar PA) that we air dried for 5 h to ensure conspecific female, or both. If so, we infer that the fertil- that ethanol was fully evaporated. For controls, we sim- ity signal inherent within Apis mellifera queen mandibu- ply added the equivalent volume of ethanol (no QMP) to lar pheromone and the ability to perceive it is evolution- filter paper. Each trial was conducted between 12:00 P.M. arily conserved between species from at least one social and 2:00 P.M. across multiple days at 23–25 °C, with and one pre-social insect order. We test both predictions even light and humidity above 30%. 1 3 J Comp Physiol A (2017) 203:143–149 145 Testing male attraction to queen pheromone (orientation) for the first time. Second, we estimated over- in a T‑maze assay all mating effort—as measured by the graphical area under a time-by-intensity profile curve. Next, we estimated the To test if males are attracted to queen pheromone, we courtship index—simply, the proportion of time spent in used a T-maze apparatus (as described in Fernandez et al.