1 INTRODUCTION

A central question in neuroscience is to understand how animal behaviors are generated and modulated in both genetic and neuronal levels. Over a 100 years of studies on the male courtship behavior in Drosophila melanogaster provide perhaps the best understanding of how regulatory genes (even a single gene) control the development of neuronal circuitry responsible for an innate behavior, and the function of that neuronal circuity in single-neuron resolution.1

Drosophila male courtship is one of the best studied innate behaviors. Wild-type males possess the ability to perform a multistep courtship ritual to females, a process mainly controlled by the fruitless (fru) gene and its expressing neurons.2-6 The most distal P1 promoter of the fru gene is dedicated to the male-specific Fru proteins (FruM), while other promotors from P2 to P4 are used to generate non-sex-specific Fru proteins (FruCom).3, 7-9 FruM is expressed in ~2000 neurons in the adult nervous system, forming a sex circuitry from sensory neurons to motor neurons.4, 5, 7, 10, 11 Many of the fruM-positive neurons exhibit sex difference in cell number and neuronal projection, and some of them are only present in either sex.11-19 It has been previously found that fruM is necessary and largely sufficient for building the potential of courtship behavior into a dedicated courtship circuity.5, 6, 20, 21 Recent findings show that FruM is not absolutely necessary for courtship as males that completely lack FruM function still acquire courtship through adult experience.22 Further studies show that the amount of FruM and the cells where FruM is expressed determine whether males show innate heterosexual, homosexual, bisexual behaviors or learned courtship behaviors,23 which are referred to as different courtship modes in this review. Importantly, adult social experience further modulates expression of FruM and therefore courtship behaviors.24, 25 Here we discuss the complex relationship between FruM and the generation and diversification of courtship behaviors, aiming to provide general implications on how innate behaviors are generated and modulated.

2 IS FRUM A SWITCH FOR SEX-SPECIFIC COURTSHIP?

For many years, FruM was considered to be a master switch controlling Drosophila male courtship as it is both necessary and largely sufficient for the behavior,4-6, 20, 21 but recent findings have challenged this idea.22, 23, 26 In previous findings, fruM mutant males exhibit strong male–male courtship or chaining behavior, with reduced or no female-directed courtship.1-3, 6, 8, 27-29 That males completely lack FruM function still perform intermale courtship behaviors seems contradictory to its necessary role for courtship. Later studies tackle this puzzle and show that FruM is indeed necessary for courtship in single-housed males without prior social experience, but males completely lack FruM function could acquire courtship through social experience, regardless of male-directed or female-directed courtship.22 Notably, sexual orientation in the learned courtship behavior by fruM mutant males depends largely on the sex or species of flies they were group-housed with, distinct from the wild-type courtship that is strictly female-directed under various housing conditions.22 Thus, although FruM is not absolutely necessary for the generation of male courtship behavior, it specifies a sex circuitry that readily and specifically responds to conspecific females and generates robust courtship.

The fruM-independent but experience-dependent courtship behavior depends on the expression of another gene in the sex determination pathway, doublesex (dsx), which encodes male- and female-specific Dsx proteins (DsxM in males and DsxF in females) and controls somatic sexual differentiation.22, 30, 31 Loss of DsxM in males reduces courtship level in general and disrupts one aspect of courtship song (sine song) in the presence of FruM.32-34 Furthermore, DsxM is both necessary and sufficient for the experience-dependent courtship acquisition in the absence of FruM.22 DsxM is expressed in ~900 neurons in the central nervous system (CNS), most of which co-express FruM.35-37 dsx-expressing neurons represent a core part of the sex circuitry as activation of all dsx neurons, just like activation of all fruM neurons, elicits all steps of courtship behaviors in solitary males.38 Interestingly, activation of all dsx neurons in fruM null mutant males is also sufficient to induce courtship behaviors, but most of which are not directed to courtship targets.38 These results show crucial role of DsxM in a core courtship circuitry (~900 neurons) for establishing the fruM-independent, but experience-dependent courtship behaviors (Figure 1A,B), while FruM function in a larger population of neurons (~2000 neurons), including both dsx-positive neurons and dsx-negative neurons, in particular many dsx-negative sensory neurons, enables innate courtship to appropriate targets (Figure 1C). Thus, although fruM is not an ideal switch that turns on/off the courtship behavior in flies, its expression in addition to dsx transforms the courtship mode from a time-consuming learning process to a hardwired innate and robust behavior.

Generation of male courtship behavior by male-specific DsxM and FruM. (A) Female nervous system that expresses DsxF in ~700 neurons does not have the potential for courtship behavior. (B) Nervous system expressing DsxM in ~900 neurons, regardless of chromosomal sex, has the potential for courtship behavior, which could only be induced through social experience.22 (C) Male nervous system that expresses DsxM in ~900 neurons and FruM in ~2000 neurons, which are largely overlapped, has the capacity of innate courtship behavior that does not require learning during adulthood 3 FRUM TUNES FUNCTIONAL FLEXIBILITY OF THE SEX CIRCUITRY

A striking phenotype on the fruM gene is that different fruM alleles show distinct courtship abnormalities, ranging from a complete loss of courtship, to changes in courtship intensity and/or sexual orientation.2, 6, 8, 9, 22, 27, 28, 39, 40 These diverse courtship phenotypes observed in flies with different fruM alleles are hard to interpret as the exact change of FruM expression in these fruM alleles is not clear. A recent study addresses this issue by comparing courtship behaviors in males with distinct fruM expression modes, such as with wild-type fruM, systemic low level of fruM, spatially low level of fruM, or completely without fruM.23 Dramatically different from fruM null mutant that could only acquire courtship through experience, males with fruM knocked down in all fruM neurons using RNA interference display a high level of male–male courtship (without prior experience) and constantly high level of male chaining, but rarely court females.23 These results indicate that the amount of FruM, not just its presence or absence, determines sexual orientation and strength of courtship behavior. Chen et al., further show that knocking down fruM in the brain results in reduced courtship toward females but does not change the heterosexual orientation; in contrast, knocking down fruM outside the brain results in equally intensive male–female and male–male courtship. These results show that both the amount and location of fruM expression determine whether males show innate heterosexual, homosexual, bisexual courtship behaviors, or could only acquire courtship through adult experience in the complete absence of fruM.23 These observations show striking flexibility of the fly sex circuitry by manipulating fruM expression (Figure 2), providing an ideal model to investigate how a molecularly defined neural circuit is modulated by the target molecule (FruM) to generate distinct behavioral modes.

FruM expression determines courtship modes. Males with wild-type FruM function show innate heterosexual courtship, which is largely unaffected by social experience. Males with systemic lower lever of FruM expression show innate homosexual courtship, and males with lower FruM expression specifically in brain show innate heterosexual courtship, but increase male–male courtship through social experience, while males with lower FruM expression outside brain show innate bisexual courtship. Males completely lack FruM function do not show innate courtship but could acquire courtship behavior through social experience

The fruM gene is largely conserved in insects and regulates male courtship behaviors from flies to cockroaches and mosquitoes.41-46 fruM expression within closely related Drosophila species exhibits a significant number of differences.44, 47, 48 For example, fruM expression in the lamina of the visual system is observed in males of D. virilis and D. suzukii, but not in D. melanogaster and other examined Drosophila species.47, 48 It has been previously found that a subset of fruM-expressing visual interneurons (M neurons) are male-specific in D. melanogaster5, 14; however, the number of these M neurons are ~2.5 times higher in D. subobscura than D. melanogaster, and some M neurons are even present in female D. subobscura.44 Furthermore, fruM expression in the visual system is also observed in female D. suzukii.47 Given the rapid change of fruM expression and the diversity of male courtship behaviors across Drosophila species,49, 50 it is conceivable that changes in the fruM regulatory regions during evolution would alter the functional mode of the sex circuitry to generate diverse courtship behaviors across species.

4 ON THE FUNCTION OF FRUM DURING DEVELOPMENT AND ADULTHOOD

The nervous system of holometabolous insects, including Drosophila, is largely reorganized during metamorphosis, during which the sex determination genes fru and dsx may function developmentally to build sexually dimorphic neuronal circuitries.31, 33, 51, 52 Indeed, FruM expression commences at the wandering third-instar larval stage, peaks at the pupal stage and thereafter declines but does not disappear at the adult stage.10 This temporal expression mode suggests that FruM may function mainly during development to build a functional sex circuitry for adult courtship behavior. Chen et al., investigate the temporal role of FruM function using genetic tools to knock down fruM in male flies at different developmental stages, and find that FruM is required during a specific developmental period (pupation), but not adulthood, for innate courtship toward females (Figure 3A).23 Males with fruM knocked down at the critical developmental period during pupation rarely court and none successfully mate with females. Consistent with these behavioral findings, knocking down fruM at the same developmental period, but not later stages, results in morphology abnormality in a subset of fruM-positive gustatory receptor neurons (GRNs) that innervate the ventral nerve cord (VNC).23, 53 Whether this stage-specific effect of FruM function applies to development of other sexually dimorphic fruM-positive neurons12-18 is not determined yet. It is likely that FruM expression during pupation is crucial for neuronal development and reconstruction of the fruM-expressing sex circuitry.

Distinct FruM function during development and adulthood. (A) FruM expression changes dynamically throughout development, and FruM expression during pupation is critical for reconstruction of the sex circuitry underlying the innate courtship behavior, while FruM expression during adulthood further modifies courtship, for example, inhibiting male–male courtship.23 (B) Group housing increases neuronal sensitivity of Or47b neurons to pheromones by maintaining FruM expression that further promotes neuronal sensitivity of Or47b neurons.25, 73 In contrast, group housing experience inhibits neuronal excitability of P1 neurons,62, 74 but whether FruM expression changes or functions during such process is not determined yet. In hypomorphic fruM mutant males, group housing increases P1 neuronal activity to visual cues,62 which is opposite to males that have regular FruM function. In fruM null males, we suspect that group housing experience would also enhance P1 neuronal activity to sensory cues similarly to that in hypomorphic fruM males

How does fruM regulate the reconstruction of sex circuitry during development? The function of fruM for neuronal development has been intensively studied. fruM encodes a set of transcription factors with the conserved BTB domain and zinc finger motifs.2, 3 Previous studies have shown that FruM forms a complex with the transcriptional cofactor Bonus (Bon), which recruits a chromatin regulator, Heterochromatin protein 1a (HP1a) or Histone deacetylase 1 (HDAC1), to orchestrate the expression of a series of target genes.53-60 A recent study also identifies Lola-Q, a transcription repressor, regulated by FruM and protected from proteasome-dependent degradation in males.61 Thus, FruM specifies neuronal development by recruiting chromatin factors and changing chromatin states, and also by turning the activity of the transcription repressor complex.51, 52

Although knocking down fruM at adult stage does not affect female-directed courtship, it significantly increases male–male chaining behaviors,23 suggesting a vital function of FruM during adulthood to suppress male–male courtship (Figure 3A). In addition, previous studies show that visual information alone is not sufficient to induce courtship in wild-type males,62, 63 but is capable to induce intense courtship in fruM hypomorphic mutant (frusat) males, also suggesting an inhibitory role of FruM function in courtship.62 Interestingly, group-housing significantly enhances courtship in frusat males62 as well as in fruM null males.22 Functional imaging experiments further show that moving spots acutely induce a robust Ca2+ rise in the male-specific courtship-promoting P1 neurons that have been intensively studied recently,14, 62-70 only in group-housed frusat males but not in single-housed frusat males or wild-type males regardless of housing condition.62 These studies raise questions on how social experience (e.g., group vs. single housing) and fruM expression (e.g., wild-type vs. hypomorphic fruM) interact and jointly modulate adult courtship. In this regard, Yamamoto and Kohatsu proposed two hypotheses: (1) the courtship circuitry established during development in fruM mutants may have a latent defect that manifests after social experience, resulting in hyper sensitivity to visual cues; (2) fruM mediates experience-dependent tuning of neural responsiveness in adults.26

Recent studies on fruM function in olfactory sensory neurons as well as central P1 neurons favor the latter hypothesis. Or47b is an olfactory receptor activated by fly-produced fatty-acid pheromones common to both sexes,71 and promotes courtship in mature males in an experience-dependent manner.25, 72 In particular, group-housing experience enhances sensitivity of Or47b-expressing neurons to fatty-acid pheromones and maintains fruM expression in these Or47b neurons, which further promotes pheromone sensitivity in these neurons, thus forming a positive feedback loop underlying the experience- and age-dependent modulation of courtship (Figure 3B).25, 73 However, such a positive feedback loop in which FruM is a key modulator to enhance Or47b neuronal sensitivity does not generally apply to other fruM-positive neurons. For example, group-housing experience enhances neuronal sensitivity of the central P1 neurons to visual cues in hypomorphic frusat males but not in wild-type males (Figure 3B).62 We suspect that group-housing experience would induce courtship in fruM null males similarly as in hypomorphic frusat males by increasing P1 neuronal activity (Figure 3B).22 In contrast, group-housing experience decreases P1 excitability and modifies courtship behaviors in wild-type males (Figure 3B).25, 62, 70, 74 Thus, FruM seems to oppositely modulate neuronal sensitivity of the sensory Or47b neurons and central P1 neurons in response to social experience. Such discrepancy may be because of the different nature of these neurons, as Or47b sensory neurons specifically respond to certain fatty-acid pheromones while P1 neurons integrate multiple sensory cues. Nevertheless, these findings support a regulatory role of fruM expression during adulthood to mediate neuronal excitability in an experience-dependent manner, in addition to its well-recognized role in circuit reconstruction during development.

5 ON THE REGULATION OF FRUM EXPRESSION

The sex-specific splicing of fruM to express FruM in males, and a non-functional truncated version in females, is determined by the sex determination pathway.2, 3, 7 In particular, Sex-lethal (Sxl) is only effectively expressed in females to direct female-specific splicing of the transformer (tra) pre-mRNA to form functional Tra protein and control female differentiation.75-77 Tra mediates the female-specific splicing of fru P1-derived transcript to generate a truncated protein that has no identified function yet, while the absence of Tra in males allows default splicing of fruM to yield the male-specific FruM protein.2, 3, 7 Transcripts from fru P2-P4 promotors are thought to be expressed commonly in two sexes independent of the sex determination pathway, but a recent study finds male-specific Fru expression in the gonad stem cell niches from the fru P4 promotor, which is regulated directly by DsxM but not Tra.78

In contrast to how fruM is sex-specifically spliced to generate the FruM protein by the sex determination pathway, the mechanism underlying the temporal and spatial expression of fruM is largely unknown. First, the cis-regulatory sequences controlling fruM expression is not clear, with very few reports on the function of upstream regulatory sequences of the fruM gene,79 despite that dozens of GAL4 lines generated using such sequences as enhancers.80, 81 Second, expression of fruM changes rapidly during development, but how such changes occur remains unknown. It has been recently reported that both Juvenile hormone (JH) and social experience modulate fruM expression in the Or47b olfactory receptor neurons in adult males.24, 25, 73 Hueston et al., find that fruGAL4 expression in Or47b olfactory neurons is sustained through the adult stage only when these neurons are functionally active, and further identify CamK, histone acetyl transferase p300 and CREB binding protein (CBP) in maintaining fruM expression during adulthood.73 CBP can directly bind the upstream element of the fru P1 promotor in adult males.82 Furthermore, JH signaling regulates neuronal sensitivity and fruM expression in these Or47b neurons, such that older males that have stronger JH signaling are more sensitive to fatty acids pheromones and competitive in mating with females compared with young males.25, 72 Zhao et al., expand these findings by showing that group-housing experience acts through p300/CBP signaling and increases active chromatin marks near fru P1 promotor, and age-dependent JH signaling acts through its receptors that bind directly to a conserved JH response element upstream of the fru P1 promotor. These two pathways integrate internal (age/sexual maturity) and external (housing experience/population density) signals and jointly regulate fruM expression in Or47b neurons during adulthood (Figure 4), allowing males to adjust their sensitivity to pheromone cues and sexual behaviors in accordance with their reproductive maturity and social environment.

Regulation of FruM expression by hormones and social experience. Pheromone stimulation during group housing experience acts on Or47b sensory neurons and induces CamKI and CBP signaling, while increased juvenile hormone in older males acts on its receptor Met. The two signalings jointly increase FruM expression in Or47b neurons and modulate adult courtship behavior accordingly

In mice, sexually dimorphic behaviors are mediated by distinct sex hormones secreted from gonads that act on their receptors common in both sexes.83 In contrast, sexually dimorphic behaviors are mainly controlled by sex specific FruM expression in flies. In light of the above-mentioned hormonal regulation of fruM expression during adulthood, we would like to speculate that the dynamic changes of fruM expression during development could also be regulated by hormonal signals, for example, juvenile hormone and ecdysone whose expression dynamically changes rapidly during development. Future studies will show whether, and to what extent, juvenile hormone and ecdysone modulate expression of sex determination genes and control sexual behaviors.

6 CONCLUSIONS AND PERSPECTIVES

One of the most significant findings with respect to courtship behavior is that a single gene (fruM) seems to be a switch of male courtship through its male-specific products.4-6, 20 Here we propose a refined view on how fruM contributes to the generation and diversification of courtship behavior. Nervous system without FruM expression still has the potential for courtship behavior, which could only be induced through social experience. Such experience-dependent courtship is time-consuming and flexible. On such basis, FruM expression transforms the time-consuming learned courtship to a robust innate behavior, presumably by the reconstruction of the sex circuitry during development. The amount of FruM expression and the place where FruM is expressed determine whether males are innately heterosexual, homosexual and bisexual. Thus, FruM expression tunes functional flexibility of the sex circuitry, and changes of FruM expression may generate diverse courtship behaviors among different individuals and species. Future studies should focus on how FruM expression is regulated during development and adulthood, and how FruM expression patterns determine functional modes of the sex circuitry.

The study on FruM and male courtship behavior has broad implications on how animal behaviors are generated and modulated. First, certain animal behaviors could only be acquired through adult experience, but the potential for such experience-dependent behavioral acquisition is specified by regulatory gene(s) (e.g., dsx) during development. Second, there are also dedicated regulatory genes that function during development to build a neuronal circuitry readily functional without any social experience (e.g., fruM). Third, even innate behaviors are modifiable through social experience, and such modification may be at least partially dependent on the expression changes of the regulatory gene that specifies the innate behavior. Finally, the regulatory gene may endow multiple functional modes of a neuronal circuitry through different expression levels and patterns, a mechanism that could contribute to behavioral diversity among different individuals and species.

ACKNOWLEDGMENTS

We thank members of the Pan Lab for their contributions to the original research and fruitful discussions, and Dr. Margaret Ho in ShanghaiTech University for valuable comments on the manuscript. This work was supported by the National Natural Science Foundation of China (31970943 and 31700905), and the Jiangsu Innovation and Entrepreneurship Team Program.

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

REFERENCES

1Yamamoto D, Koganezawa M. Genes and circuits of courtship behaviour in Drosophila males. Nat Rev Neurosci. 2013; 14(10): 681- 692. 2Ito H, Fujitani K, Usui K, Shimizu-Nishikawa K, Tanaka S, Yamamoto D. Sexual orientation in Drosophila is altered by the satori mutation in the sex-determination gene fruitless that encodes a zinc finger protein with a BTB domain. Proc Natl Acad Sci U S A. 1996; 93(18): 9687- 9692. 3Ryner LC, Goodwin SF, Castrillon DH, et al. Control of male sexual behavior and sexual orientation in Drosophila by the fruitless gene. Cell. 1996; 87(6): 1079- 1089. 4Stockinger P, Kvitsiani D, Rotkopf S, Tirian L, Dickson BJ. Neural circuitry that governs Drosophila male courtship behavior. Cell. 2005; 121(5): 795- 807. 5Manoli DS, Foss M, Villella A, Taylor BJ, Hall JC, Baker BS. Male-specific fruitless specifies the neural substrates of Drosophila courtship behaviour. Nature. 2005; 436(7049): 395- 400. 6Demir E, Dickson BJ. Fruitless splicing specifies male courtship behavior in Drosophila. Cell. 2005; 121(5): 785- 794. 7Usui-Aoki K, Ito H, Ui-Tei K, et al. Formation of the male-specific muscle in female Drosophila by ectopic fruitless expression. Nat Cell Biol. 2000; 2(8): 500- 506. 8Anand A, Villella A, Ryner LC, et al. Molecular genetic dissection of the sex-specific and vital functions of the Drosophila melanogaster sex determination gene fruitless. Genetics. 2001; 158(4): 1569- 1595. 9Goodwin SF, Taylor BJ, Villella A, et al. Aberrant splicing and altered spatial expression patterns in fruitless mutants of Drosophila melanogaster. Genetics. 2000; 154(2): 725- 745. 10Lee G, Foss M, Goodwin SF, Carlo T, Taylor BJ, Hall JC. Spatial, temporal, and sexually dimorphic expression patterns of the fruitless gene in the Drosophila central nervous system. J Neurobiol. 2000; 43(4): 404- 426. 11Yu JY, Kanai MI, Demir E, Jefferis GS, Dickson BJ. Cellular organization of the neural circuit that drives Drosophila courtship behavior. Curr Biol. 2010; 20(18): 1602- 1614. 12Cachero S, Ostrovsky AD, Yu JY, Dickson BJ, Jefferis GS. Sexual dimorphism in the fly brain. Curr Biol. 2010; 20(18): 1589- 1601. 13Kimura K, Ote M, Tazawa T, Yamamoto D. Fruitless specifies sexually dimorphic neural circuitry in the Drosophila brain. Nature. 2005; 438(7065): 229- 233. 14Kimura K, Hachiya T, Koganezawa M, Tazawa T, Yamamoto D. Fruitless and doublesex coordinate to generate male-specific neurons that can initiate courtship. Neuron. 2008; 59(5): 759- 769. 15Kohl J, Ostrovsky AD, Frechter S, Jefferis GS. A bidirectional circuit switch reroutes pheromone signals in male and female brains. Cell. 2013; 155(7):