InteractiveFly: GeneBrief

Gustatory receptor 68a: Biological Overview | Regulation | Developmental Biology | Effects of Mutation | References

BIOLOGICAL OVERVIEW

Reproduction in higher animals requires the efficient and accurate display of innate mating behaviors. In Drosophila, male courtship consists of a stereotypic sequence of behaviors involving multiple sensory modalities, such as vision, audition, and chemosensation. For example, taste bristles located in the male forelegs and the labial palps are thought to recognize nonvolatile pheromones secreted by the female. A putative pheromone receptor, GR68a, is expressed in chemosensory neurons of about 20 male-specific gustatory bristles in the forelegs. Gr68a expression is dependent on the sex determination gene doublesex, which controls many aspects of sexual differentiation and is necessary for normal courtship behavior. Tetanus toxin-mediated inactivation of Gr68a-expressing neurons or transgene-mediated RNA interference of Gr68a RNA leads to a significant reduction in male courtship performance, suggesting that GR68a protein is an essential component of pheromone-driven courtship behavior in Drosophila (Bray, 2003).

Courtship behaviors are highly diversified, innate behaviors essential for propagation in higher animals. In general, courtship is composed of a series of behavioral displays controlled by the CNS, modulated by the endocrine system, and triggered only by highly specific, external stimuli that emanate from the mating object (Bray, 2003).

In wild flies, mating occurs near feeding sites to which they are attracted by long-range olfactory cues. Mate recognition, courtship, and mating are then mediated by visual, auditory, and pheromone signals and displayed in a stereotypic sequence of behaviors particularly well defined in the male: first, the male orients toward and follows a female (1), taps her abdomen with his forelegs (2), and proceeds to generate a 'courtship song' by rapid wing vibrations (3). He then licks the female's genitalia (4), curls his abdomen to attempt mounting (5), and eventually succeeds in mounting and copulation with the female (6). These steps entail visual (1) and chemosensory (2 and 4) recognition of female features by the male, auditory reception of the male courtship song by the female (3), and somatosensory agility of both sexes (3, 5, and 6). Even though this behavior is displayed in a stereotypical, sequential manner, a male generally executes each step multiple times before proceeding to the next. Females do not display a distinct courting behavior, but mated females actively reject a new potential mate by walking away, kicking with her hind legs, flicking of her wings, and extension of her ovipositor (Bray, 2003).

Efficient performance of courtship is the major determinant of mating success and can be quantified in single pair mating experiments by measuring mating latency (time from a first encounter between a male and a female until copulation), nonmating frequency, or the courtship index (CI = the percentage of time a male performs any of the first five courtship steps during a mating experiment). All these parameters have proved valuable for the quantification of male mating performance (Bray, 2003).

Mating behavior and sexual differentiation are regulated by a genetic cascade of splicing factors including Transformer (Tra) and Transformer2 (Tra2) that control the sex-specific, alternative splicing of mRNAs encoding the transcription factors Doublesex (Dsx) and Fruitless (Fru). Dsx^{m (male)}, Dsx^{f (female)}, and Fru^m control the expression of numerous male and female effector genes responsible for differentiation and maintenance of sexual identity. Several dsx-dependent effectors have been identified and are found to be expressed in endocrine tissues of the adult or in the genital disc during differentiation of adult structures. Fru^m and Dsx^m are required, but neither alone is sufficient, for wild-type male courtship behavior, because males lacking Fru^m but expressing Dsx^m (X/Y; fru¹/ fru¹) and intersexes which lack Dsx^m but express Fru^m [X/Y dsx¹/Df(dsx)] display severely reduced courtship behavior. Additionally, fru males court both males and females indiscriminately (Bray, 2003).

Several lines of evidence suggest that pheromone-elicited mate recognition is mediated mainly through the contact chemosensory system. For example, males tap the pheromone-coated, female abdomen and genitalia with their forelegs and labial palps, respectively, both of which are covered with taste bristles. The taste bristles on the forelegs are also implicated in a sex-specific function due to a quantitative difference in their number between males (50) and females (37). Finally, the chemical properties of the known female pheromones, which are nonvolatile, long-chain hydrocarbons, further support a role for contact chemosensory neurons/receptors in male courtship behavior (Bray, 2003).

Taste bristles in the labial palps and legs are composed of two to four gustatory receptor neurons (GRNs). A family of about 70 G protein-coupled receptor (GPCR) genes has been identifed, members of which are expressed in small subsets of GRNs in all known taste organs including the labial palps and the forelegs (Clyne, 2000; Dunipace, 2001; Scott, 2001). Upon analysis of about one quarter of the ">Gustatory receptor genes, the expression and function of Gr68a, a Gr gene expressed in chemosensory neurons of about ten male-specific taste bristles in the foreleg, is described. It is proposed that GR68a recognizes a female pheromone(s) involved in the second step of the courtship display, which is essential for efficient execution of the entire courtship sequence and timely mating (Bray, 2003).

Thus, a set of about 20 neurons associated with male-specific taste bristles in the forelegs of Drosophila is crucially involved in pheromone recognition during male courtship behavior. These bristles are molecularly characterized by the expression of the proposed taste receptor GR68a. RNA-mediated repression of this gene shows that Gr68a is in fact directly involved in recognition of a female pheromone, providing a precedent for a sex-specific pheromone receptor with a defined function in courtship behavior (Bray, 2003).

In principle, courtship behaviors serve two purposes: to attract the attention of a mating partner and to identify the sex and mating status of a con-specific animal. The complex sequence of behaviors of male flies combines both these purposes and is critical in guiding the male in a coordinated fashion through the entire courtship ritual culminating in successful copulation. The perception of female pheromones during the second and fourth step of the sequence are crucial events of courtship and must be integrated with other sensory input, including visual cues (female-specific coloration of the abdomen) and behavioral responses of the female toward the male during the entire courtship (Bray, 2003).

The functional characterization of the Gr68a-expressing neurons associated with male-specific taste bristles of the forelegs provides an opportunity to dissect the male courtship behavior. When Gr68a-expressing neurons were functionally inactivated by coexpressing TNT, males show a significant reduction in courtship activity toward females or males with a female pheromone profile, but no increase of courtship toward other males. Therefore, these neurons mediate a stimulatory response of an attractive, female pheromone as opposed to a repressive response of an inhibitory male pheromone. The specific role for these neurons is associated with the tapping step during courtship (step 2), in which the male directly contacts the pheromone-coated abdomen of the female with the tarsi of his forelegs. By quantitatively analyzing individual courtship steps, males lacking functional Gr68a-expressing neurons were shown to stall during the second step. Moreover, the modest increase in initiation/orientation (step 1) suggests that these males 'start over' more often with the courtship sequence than males with intact Gr68a-expressing neurons. Knowing the identity of these neurons and the specific phenotype associated with their inactivation should provide future opportunities to address more complex questions pertinent to this intriguing behavior. For example, how is the pheromone input in step 2 integrated with visual information received during step 1 and additional pheromone input received in step 4? And how does this input affect the motorneuron output so strikingly displayed in steps 3 and 5? At least a partial answer to these questions will require the identification of the first- and second-order target neurons of the Gr68a-expressing sensory neurons, which eventually should become feasible using axonal, synaptic, and trans-synaptic marker proteins expressed under the control of the Gr68a promoter (Bray, 2003).

The courtship phenotype associated with inactivating Gr68a-expressing neurons is likely to be mediated by the GR68a receptor itself. Males in which GR68a expression is suppressed by RNAi have an increase in mating latency, fraction of nonmaters, and reduced courtship intensity, as was observed in males in which the Gr68a-expressing neurons were inactivated altogether. Moreover, the detailed courtship analysis reveals that these males also stall during the same step in the courtship sequence, with almost identical severity as males with inactivated Gr68a-expressing neurons, arguing for a major role of this receptor in recognition of a female pheromone. It is quite possible that Gr68a is the only Gr gene that is expressed in these neurons. Reported expression studies of about ten Gr genes (Dunipace, 2001; Scott, 2001) and ongoing studies of an additional ten Gr genes (N. Thorne and H. A., unpublished data cited in Bray, 2003) indicate that most Gr genes are expressed in distinct sets of gustatory neurons, and expression of different Gr genes are to a large extent nonoverlapping. In any case, even if Gr68a-expressing neurons coexpress another Gr gene, its transcripts are unlikely to be affected by the expression of ds_Gr68a RNA, because nucleotide sequence similarity between Gr68a and any other Gr gene is far too low to allow RNAi to occur (Bray, 2003).

Gr68a is the first putative, sex-specific, pheromone receptor gene with a defined function in courtship behavior. A gene cluster containing 16 putative pheromone receptors (V1Rs) expressed in the vomeronasal organ was recently reported to be required for normal mating behavior of mice (Del Punta, 2002). Female mice homozygous for this multigene knockout show reduced aggression toward invaders, and homozygous male mice show reduced sexual aggression toward both sexes. However, none of the V1Rs included in this deletion were reported to be sex specific, and it remains to be investigated whether specific behavioral phenotypes can be associated with individual V1R genes (Bray, 2003).

The 70 Drosophila Gr genes, which are distantly related to the olfactory receptor (Or) genes, encode a diverse family of G protein-coupled receptors that share between 15% and 80% sequence similarity. No other candidate chemosensory receptors have emerged from the complete genome sequence of Drosophila, suggesting that the GR proteins might accommodate the detection of all nonvolatile substrates to which Drosophila is responsive. In mammals, distinct groups of nonvolatile compounds are recognized by unrelated G protein-coupled receptors, encoded by four distinct gene families that are expressed in neurons of the vomeronasal organ or in taste cells of the tongue. For example, the taste cells in the tongue express two classes of receptors, the T1Rs and T2Rs; the T1Rs were shown to detect sweet-tasting substrates such as various sugars and many L-amino acids, whereas the much more numerous T2Rs appear to recognize the large spectrum of compounds perceived as bitter tasting to humans (Bray, 2003 and references therein).

The only known substrate for a GR protein, GR5a, is trehalose, which is an important food source for Drosophila melanogaster. The number of biologically relevant sugars is fairly small compared to the number of Gr genes, and hence, members of this protein family are likely to recognize other classes of substrates. It is suggested that GR68a is a receptor that detects a pheromone(s) of Drosophila melanogaster females, possibly a long-chain hydrocarbon such as the female-specific 7,11 heptacosadiene and 7,11 nonacosadiene, both of which have been strongly implicated in eliciting male courtship behavior (Coyne, 1995; Ferveur, 1996). Thus, it is proposed that the GR protein family recognizes a whole spectrum of nonvolatile, complex substrates (sugars, amino acids, alkaloids and bitter-tasting compounds, hydrocarbons, and possible other pheromones, etc.) to which Drosophila responds. For example, different sugars might bind to receptors of the GR5a subfamily, which include seven additional receptors encoded by Gr61a and Gr64a-f (members of a subfamily were defined by sharing at least 34% sequence similarity [Dunipace, 2001]). Similarly, proteins encoded by the Gr68a subfamily -- Gr2a, Gr32a and Gr39a.a-39a.d -- might interact with different pheromone components. Expression of Gr32a has been analyzed and found to be restricted to the labelum and the distal tip of the forelegs of both sexes (Scott, 2001). Preliminary experiments indicate that simultaneous inactivation of Gr68a- and Gr32a-expressing neurons by TNT compounds the courtship defect in males, resulting in about 80% nonmaters and almost 20 min latency time, suggesting that this receptor might function in the fourth step of the courtship sequence. Further analysis of Gr32a and the other members of the Gr68a subfamily using RNAi should reveal the specific roles, if any, that these genes have during male mating behavior (Bray, 2003 and references therein).

Finally, amino acids and various classes of bitter-tasting substrates might be recognized by receptors encoded by yet other Gr subfamilies. It is even conceivable that some GR proteins detect volatile molecules, as a few Gr genes were found to be expressed in olfactory neurons both in the larvae and the adult (Dunipace, 2001; Scott, 2001). Thus, the highly diverse GR proteins are likely to mediate a multitude of strikingly different behaviors including courtship and mating, feeding behavior, and avoidance behavior elicited both by soluble and volatile compounds (Bray, 2003 and references therein).

REGULATION

Transcriptional Regulation

If Gr68a encodes a male-specific pheromone receptor, it would be predicted that the sex determination genes, which control all aspects of sexual differentiation, would regulate its expression. Thus, Gr68a expression was investigated in chromosomally female (XX) flies that were sexually transformed into Ψ males by mutations in tra2 or dsx. Both types of Ψ males show the normal male expression pattern of the p[Gr68a]-Gal4 driver. Since sex-specific fru expression is directly controlled by Tra and Tra2, and hence, independent of dsx (i.e., XX; dsx Ψ males express no Fru^m), male-specific expression of Gr68a is fru independent. Thus, Gr68a is a dsx-dependent effector gene expressed in chemosensory neurons of taste bristles in the foreleg, which is consistent with a function for this gene in pheromone recognition (Bray, 2003).

Functional analysis of neurons expressing Gr68a

To determine whether Gr68a-expressing neurons function in pheromone recognition, neuronal transmission was inactivated using tetanus toxin light chain protein (TNT) and courtship and several other behaviors of such males was investigated. TNT, which cleaves the synaptic vesicle protein N-Syb, a protein essential for neurotransmitter release, has been widely used in Drosophila to inactivate various types of sensory neurons, including chemosensory neurons in developing and adult Drosophila. Gr68a-expressing neurons were inactivated in males using a UAS-tnt reporter gene expressed under the control of the p[Gr68a]-Gal4 driver (Sweeney, 1995). In addition, males were generated expressing an inactive TNT protein (TNTⁱⁿ) in these neurons (Sweeney, 1995), to control for nonspecific effects of overexpression of GAL4 and TNT. It should be noted that TNT does not kill the neurons in these males; coexpression of GFP persists for at least 18 days without apparent change in cell morphology or loss in fluorescence. Overall sensory perception was not affected since two courtship-unrelated behaviors, sugar recognition/sensitivity and gravity locomotion, were not affected in males without functional Gr68a-expressing neurons (Bray, 2003).

The function of Gr68a-expressing neurons was investigated to determine if they could act as putative pheromone receptor neurons. In principle, these neurons may be activated by a stimulatory female pheromone or an inhibitory male pheromone. Males lacking functional GRNs expressing a stimulatory pheromone receptor may fail to recognize a female and therefore exhibit reduced courtship; alternatively, absence of GRNs expressing an inhibitory male pheromone receptor may lead to anomalous male-to-male courtship. A stimulatory function of Gr68a-expressing neurons was investigated and the courtship of various males toward virgin females was tested in single pair mating experiments. Indeed, males expressing TNT in Gr68a-positive neurons performed poorly: first, an average of 41% of males failed to mate during the observation period (30 min), whereas less than 10% of four different classes of control males were nonmaters. In addition, the males that did mate showed a significant increase in mating latency, when compared with all control males (Bray, 2003).

To test whether the reduction in mating performance was specifically mediated by neurons expressing Gr68a, males were generated in which TNT was expressed in other sets of gustatory neurons, those expressing Gr66a or Gr22e. These two Gr genes are expressed in a similar or larger numbers of neurons than Gr68a; Gr66a is expressed in about 14-16 neurons in the labelum and in 2 neurons of each foreleg (Dunipace, 2001; Scott, 2001); Gr22e is expressed in almost 100 GRNs and is the most abundantly expressed Gr gene characterized thus far, with extensive expression in all chemosensory organs including the legs (Dunipace, 2001). It should be noted that Gr22e, Gr66a, and Gr68a are expressed in largely (and possibly entirely) nonoverlapping groups of neurons in the foreleg (Dunipace, 2001). Inactivation of Gr66a-expressing neurons results in no increases in latency time and no increase in fraction of nonmaters when compared to control males. Males lacking functional Gr22e-expressing neurons exhibited a slight increase in both latency time and represents only a fraction of nonmaters when compared to wild-type or the various control males; however, this increase is only about 1/3 of that observed in males lacking Gr68a-expressing neurons. These experiments show that neuron identity, rather than absolute number, is important for efficient male courtship behavior and identify Gr68a-expressing neurons as a set of gustatory neurons critically involved in this behavior (Bray, 2003).

To test whether the reduction in mating performance of males lacking functional Gr68a-expressing neurons is due to reduced courtship intensity, the courtship index (CI) was determined of several types of males. These experiments show that males lacking functional Gr68a-expressing neurons spend significantly less time courting a female (CI of 40 ± 3), when compared with both groups of control males (66 ± 3 and 80 ± 3, respectively). Thus, loss of mating efficiency in males lacking functional Gr68a-expressing neurons is caused, at least in part, by reduced courtship intensity (Bray, 2003).

In Drosophila, mating is essentially characterized by a 'first come first serve basis': once fertilized, females increase egg laying and exhibit a strong rejection behavior toward additional suitors and elicit less courtship activity for several days. Thus, males carrying out the courtship sequence more efficiently and flawlessly have a competitive mating advantage. In order to investigate the role of neurons expressing Gr68a in a more natural mating environment, three sets of competition experiments were carried out. In the first two experiments, a male lacking functional Gr68a-expressing neurons was forced to compete for a virgin female against either a wild-type male or a male expressing TNTⁱⁿ. As expected, males lacking functional Gr68a-expressing neurons were far less successful in their copulation attempts than the competing control males under these mating conditions (only once in 46 and 40 competition experiments, respectively). The third set of competition experiments pitted the two control types, wild-type males versus those expressing TNTⁱⁿ, against each other and showed that they had similar success rates (24 times and 16 times, respectively, in 40 experiments). Thus, these experiments establish that males lacking functional Gr68a-expressing GRNs in each of the two forelegs performed poorly in a controlled competitive mating environment, demonstrating an essential role for these neurons in efficient male courtship behavior (Bray, 2003).

The results described thus far suggest that Gr68a-expressing neurons are involved in the recognition of a stimulatory female pheromone. A priori, however, the possibility that these neurons might recognize an inhibitory pheromone present in males can not be excluded. To investigate this possibility, two sets of experiments were performed: (1) males lacking functional Gr68a-expressing neurons, wild-type males and males expressing the inactive tetanus toxin protein were paired with virgin control males and male-to-male courtship behavior was determined by measuring their CI. This analysis revealed that all tested males exhibited virtually no courtship activity toward other males, an observation that further supports a role for Gr68a-expressing neurons in the recognition of a stimulatory female pheromone. (2) To address the possibility that reduced courtship of males lacking functional Gr68a-expressing neurons could be attributed to their failure to recognize other female cues (visual, behavioral, etc.), the three types of males were subjected to single mating experiments with males exhibiting a female pheromone profile. These males (S12/14; UAS-tra), which are visually and behaviorally indistinguishable from wild-type males, express the Tra protein under the control of an inducible GAL4 protein in secretory cells including oenocytes. Selective expression of Tra in oenocytes results in the production of female pheromones in otherwise normal males, which elicit aggressive courtship from other males including wild-type males. When S12/14; UAS-tra males were used as mating partners for males expressing the active TNT protein in Gr68a-positive cells, a reduction was observed in the CI similar to that observed with female mating partners; similarly, the control males showed virtually as high a CI toward these 'pheromonally feminized' males as they did toward virgin females (Bray, 2003).

Taken together, these data strongly suggest that the Gr68a-expressing neurons in the male forelegs are necessary for the recognition of a female pheromone component (Bray, 2003).

Drosophila courtship consists of a sequence of behaviors, the proper order of which is crucial for efficient mating. To identify the specific step(s) in which Gr68a is involved, three individual courtship steps (1, 3, and 5 - see Biological Overview) were examined, all of which can be readily quantified. The two types of control males initiated courtship about once per minute, whereas males lacking functional Gr68a-expressing neurons had a modest, but significant, increase in initiation (1.4 times/min). Interestingly however, wing extension/vibration and attempted mating were 2- to 3-fold reduced in these males when compared to the two control males. Thus, the analysis shows that the neurons expressing GR68a are crucial after step 1 and before step 3, suggesting that males without proper pheromone input through the Gr68a-expressing neurons in the forelegs stall at the tapping step during male courtship (Bray, 2003).

DEVELOPMENTAL BIOLOGY

To better understand the possibly diverse functions of different GR proteins, the global expression profile of all Gr genes is being examined. The widely dispersed location of taste bristles, low expression level of Gr genes, and intractability of many taste organs (legs and wings) to in situ hybridization methods (Clyne, 2000; Dunipace, 2001; Scott, 2001) makes it necessary to utilize the GAL4/UAS system to determine the spatial distribution of GRNs expressing a given receptor. Specifically, the yeast transcriptional activator GAL4 is expressed under the control of the putative promoter of various Gr genes, and GAL4 activity is visualized with UAS-lacZ or UAS-gfp reporter genes (Dunipace, 2001; Scott, 2001). Initial investigations using this method revealed that most Gr genes are expressed in a small fraction of chemosensory neurons in spatially defined and distinct sets of GRNs in a subset of taste organs (labelum, pharyngeal sense organs, legs, and wings) (Dunipace, 2001; Scott, 2001; Bray, 2003 and references therein).

Upon analysis of about a quarter of the 70 Gr genes, a Gr gene, Gr68a, was identified exhibiting the hallmarks of a putative pheromone receptor. In adults, Gr68a is exclusively expressed in neurons of about ten male-specific taste bristles in the forelegs. No expression is observed in females or any other organ or structure of males. Identical β-gal or GFP expression patterns were observed with four independent transgenic p[Gr68a]-Gal4 driver lines, indicating that male-specific expression reflects an intrinsic property of the Gr68a promoter. To verify that the β-gal-positive cells are indeed neurons and not support cells associated with taste bristles, antibody staining was performed; β-gal immunoreactive cells were found to have the typical structure of sensory neurons and express ELAV protein, a pan-neuronal marker not expressed in other cell types. To verify that the Gr68a gene is expressed in one of the chemosensory neurons and not in the single mechanosensory neuron present in taste bristles, its expression was analyzed in a pox-neuro (poxn) mutant background. Poxn is necessary for specification of chemosensory neurons, and poxn mutant flies show a complete transformation of all chemosensory neurons into mechanosensory neurons. Indeed, the p[Gr68a]-Gal4 driver is not expressed in these flies: this confirms that Gr68a is expressed in chemosensory neurons in the male foreleg (Bray, 2003).

EFFECTS OF MUTATION

Expression studies of about a quarter of all Gr genes has revealed that, with the exception of Gr22e, most are expressed in a small number of GRNs (Dunipace, 2001; Scott, 2001). These studies suggest that a single GRN expresses a very small number -- possibly just one -- of the 70 different receptor genes. Hence, the phenotype observed in males lacking functional Gr68a-expressing neurons could be attributed solely to a second GR protein present in these cells. To address this possibility, RNA interference was employed to knock out/down the expression of Gr68a RNA/protein. Males were generated expressing a double-stranded Gr68a RNA (UAS-ds_Gr68a) under the control of the p[Gr68a]-Gal4 driver; a statistically significant reduction in mating performance was observed both with regard to the fraction of nonmaters and the increase in latency time. The more modest phenotype compared to males lacking functional Gr68a-expressing neurons might be explained by a temporal delay of ds_Gr68a RNA expression, which requires first the accumulation of GAL4 protein; hence some GR68a protein may be produced before ds_Gr68a RNA is transcribed to promote endogenous Gr68a RNA degradation. Alternatively, Gr68a RNA might not be efficiently degraded or a second Gr gene expressed in these neurons might partially substitute for Gr68a function. To investigate these possibilities, older males, aged for an additional 7-10 days, were subjected to the single mating assay. If protein turnover is the major cause for the difference between young males expressing TNT versus ds_Gr68a RNA, the phenotype should become more severe in older males expressing ds_Gr68a; if, however, incomplete knockdown or presence of a second receptor is the major cause for the weaker phenotype, no change in severity should be observed in older males. Single mating experiments reveal that older males indeed show a further reduction in courtship performance, reaching levels similar to males lacking Gr68a-expressing neurons altogether. The CI of these males toward virgin females was also significantly reduced to 50.8 ± 4, a value close to that observed in males expressing TNT in GR68a-positive neurons (40.0 ± 3). Importantly, age per se has no effect on courtship performance because older control males showed no reduction in mating efficiency and were indistinguishable in their performance from young control males (Bray, 2003).

Finally, individual courtship steps of ds_Gr68a RNA-expressing males were quantified in order to determine the courtship deficit more precisely. These males also show a significant decrease in executing later courtship steps, such as wing extension/vibration, mating attempts, as well as copulation, whereas courtship initiation was increased, just as observed in males expressing TNT. It is noted that no reduction in courtship performance was observed when ds_yellow RNA, which reproduces an exact phenocopy of null yellow mutations, was expressed in GR68a-positive neurons, showing that the phenotype was gene specific. Thus, these data strongly support the notion that GR68a expressed in male-specific neurons functions as the crucial receptor involved in the recognition of a female pheromone (Bray, 2003).

A Drosophila protein specific to pheromone-sensing gustatory hairs delays males' copulation attempts

In insects, increasing evidence suggests that small secreted pheromone binding proteins (PBPs) and odorant binding proteins (OBPs) are important for normal olfactory detection of airborne pheromones and odorants far from their source. In contrast, it is unknown whether extracellular ligand binding proteins participate in perception of less volatile chemicals, including many pheromones, that are detected by direct contact with chemosensory organs. CheB42a, a small Drosophila protein unrelated to known PBPs or OBPs, is expressed and likely secreted in only a small subset of gustatory sensilla on males' front legs, the site of gustatory perception of contact pheromones. CheB42a is expressed specifically in the sheath cells surrounding the taste neurons expressing Gr68a, a putative gustatory pheromone receptor for female cuticular hydrocarbons that stimulate male courtship. Surprisingly, however, CheB42a mutant males attempt to copulate with females earlier and more frequently than control males. Furthermore, CheB42a mutant males also attempt to copulate more frequently with other males that secrete female-specific cuticular hydrocarbon pheromones, but not with females lacking cuticular hydrocarbons. Together, these data indicate that CheB42a is required for a normal gustatory response to female cuticular hydrocarbon pheromones that modulate male courtship (Park, 2006).

The elaborate courtship ritual performed by Drosophila melanogaster males presents an ideal opportunity for studying the genetic determinants of chemosensory function because this ritual has been extensively documented, is readily quantified, and is modulated by both olfactory and gustatory pheromones. Male front legs were identified as the likely site for gustatory perception of pheromones because they have more taste hairs than their female counterparts and include taste neurons with characteristic contralateral projection patterns not found in females. In a molecular screen for genes that are expressed specifically in the front legs of males, a previously undescribed gene, CheB42a was identified (Xu, 2002). CheB42a codes for a small protein expressed and likely secreted specifically by non-neuronal cells associated with a small subset of taste hairs on the front legs of males. Furthermore, although CheB42a is not obviously similar to any known protein, it is related to eleven other CheB genes encoded by the Drosophila genome; several of these genes are also expressed in sexually dimorphic patterns on appendages with high concentrations of chemosensory hairs. These observations suggested that CheB42a and other CheB genes are involved in sex-specific chemosensory perception, perhaps of pheromones. Subsequently, Gr68a, a member of the family of gustatory receptor genes, was also shown to be expressed in a subset of taste hairs on the front legs of males. Excitingly, both inactivation of Gr68a-expressing neurons by targeted expression of neurotoxins and knockdown of the Gr68a mRNA result in a decreased male response to female courtship-activating contact pheromones, suggesting that Gr68a-expressing gustatory neurons detect these pheromones and that Gr68a is a pheromone receptor (Park, 2006).

What is the relationship between the subsets of CheB42a-expressing and Gr68a-expressing gustatory sensilla on male front legs? To answer this question, expression of the GFP (Green Fluorescent Protein) reporter protein was used under indirect control of the regulatory region of each gene through the Gal4/UAS system. Animals were compared that express GFP from a UAS-GFP transgene in the presence of (1) CheB42a-Gal4 (Xu, 2002), (2) Gr68a-Gal4, or (3) both Gal4 drivers. Gustatory sensilla can be distinguished from mechanosensory sensilla by their curved shaft, the absence of a bract at their bases, and their stereotypic positions on the tibia and on the five tarsal segments of a male's front legs. For males of all three genotypes, GFP expression is associated with ten distinct gustatory sensilla on the tibia and first, second, third, and fourth tarsal segments of the front legs of males, but never on the fifth tarsal segment. Furthermore, if the two Gal4 drivers resulted in GFP expression in overlapping but nonidentical sets of hairs, there would be fewer GFP-positive sensilla in at least one type of animal containing a single driver, relative to those possessing both drivers. For males of all three genotypes, it was found that, as previously reported for Gr68a-Gal4, not all animals express GFP in all ten sensilla. The small difference between the average numbers of GFP-expressing sensilla in animals of each of the three genotypes may therefore result from the slightly greater variability observed in GFP expression in the presence of CheB42a-Gal4 than in the presence of Gr68a-Gal4. More significantly, the maximum number of GFP-expressing sensilla on each tarsal segment is the same for animals of all three genotypes. Together, these data suggest that Gr68a and CheB42a are coexpressed in the same subset of gustatory sensilla on male front legs (Park, 2006).

Gustatory sensilla on the legs of Drosophila are complex structures that include four gustatory neurons, one mechanosensory neuron, and three types of highly differentiated non-neuronal cells: hair or trichogen cells, socket or tormogen cells, and sheath or thecogen cells. The membranes of sheath cells wrap around the cell bodies of neurons and are in turn surrounded by the membranes of hair and socket cells in an onion-like concentric pattern. Furthermore, the cell bodies of sheath cells and neurons are found deeper under the cuticle relative to those of the other two types of non-neuronal cells. Previous observations have shown that whereas Gr68a is expressed in gustatory neurons, CheB42a expression is restricted to non-neuronal cells of gustatory sensilla. To further characterize the cells that express the CheB42a protein, frozen sections of male front legs were double labeled by using a CheB42a antibody and the 22C10 monoclonal antibody, which recognizes a membrane-associated, neuronal-specific antigen. Although the two signals do not overlap, confirming that CheB42a-expressing cells are indeed not neurons, CheB42a is always found in close proximity to the neurons of gustatory sensilla. Furthermore, CheB42a often appears in a donut-like pattern that surrounds the 22C10-positive neuronal membranes. Direct comparison of CheB42a distribution with that of GFP expressed under control of the Gr68a promoter leads to similar conclusions. The highest level of staining with each antibody is found in two clearly distinct but neighboring cell bodies. Here also, CheB42a signal sometimes appears to surround the cell body of the GFP-expressing cell, suggesting that CheB42a-expressing non-neuronal cells are in close contact with, and partially surround, Gr68a-expressing neurons. To determine which of the three types of non-neuronal cells expresses CheB42a, its distribution was compared to that of another chemosensory protein, PBPRP2. PBPRP2 is an insect OBP family member that is expressed not only in olfactory appendages but also in many gustatory sensilla, including most, if not all, gustatory sensilla on the legs. Moreover, within taste hairs of the legs, PBPRP2 is expressed only in two types of non-neuronal cells: trichogen and internal tormogen cells, and is absent from neurons, external tormogen cells, and sheath cells. Comparison of CheB42a expression to the distribution of GFP driven by a pbprp2-Gal4 driver shows that the two proteins are expressed in adjacent but non-overlapping cells. Furthermore, PBPRP2-expressing cells are distal (closer to the tip of the leg) and nearer the cuticle relative to the CheB42a signal. Taken together, these results suggest that CheB42a is specifically expressed by the sheath cells that tightly surround the gustatory neurons that express Gr68a and likely detect female pheromones (Park, 2006).

Drosophila male courtship behavior involves an ordered series of simple behaviors: The male orients toward the female, taps her with his forelegs, generates a species-specific courtship song by vibrating one of its wings, licks the female's genitalia, and attempts to copulate. Both initiation of courtship behavior and progression to the late steps in this behavioral series involve perception of pheromones by gustatory organs, olfactory organs, or both. The close association between cells expressing CheB42a and Gr68a-expressing neurons suggests that CheB42a may also be required for a gustatory response to contact pheromones that modulate male courtship behavior. In a study of the neighboring ppk25 gene, Δ5-68, a CheB42a deletion that is henceforth referred to as CheB42a^Δ5-68, does not affect the total time homozygous mutant males spend courting females (Lin, 2005). Furthermore, CheB42a^Δ5-68 has no effect on several other behaviors unrelated to courtship; such behaviors include preening, walking, geotaxis, and gustatory response to sugars. Surprisingly, however, in a detailed analysis of the individual behavioral steps involved in male courtship behavior, it has since been found that males homozygous for CheB42a^Δ5-68 display a specific increase in the last step in this sequence: attempted copulation. CheB42a^Δ5-68 mutant males perform an average of eight attempted copulations in a 10 min observation period, whereas control males perform only three. This increased number results from a faster progression from initiation of courtship behavior to the first attempted copulation, as well as more frequent subsequent attempts. In contrast, CheB42a^Δ5-68 mutant males are not different from controls in the timing or frequency of earlier steps in the courtship sequence (lag to courtship initiation, tracking and following of the female, tapping, and wing vibration). Introduction of a transgene encoding CheB42a and 3.5 kb of upstream DNA with all known regulatory sequences, but no other identified gene, completely reverses the effect of CheB42a^Δ5-68 on both the accelerated progression to attempted copulation and its increased frequency, suggesting that both effects result from loss of CheB42a. To confirm that the CheB42a gene is responsible for this phenotype, males of two other genotypes that lack the endogenous CheB42a gene but carry one of two different transgenes were compared. The first transgene carries a genomic fragment that includes both CheB42a and the neighboring ppk25 gene (Lin, 2005). The second transgene is identical to the first except for a mutation in the ATG codon for CheB42a's initiator methionine that prevents its translation. Lack of CheB42a in a male doubles the number of attempted copulations as a result of an approximately 2-fold faster progression from initiation of courtship to the first attempted copulation, along with more frequent subsequent copulation attempts. Together, these data suggest that CheB42a is required for the gustatory response to pheromones that modulate the male courtship response (Park, 2006).

What are the pheromones involved in CheB42a-mediated inhibition of attempted copulation? Abundant, low-volatility, cuticular hydrocarbons are the only pheromones known to modulate the courtship behavior of Drosophila melanogaster males. Male courtship is stimulated by 7,11-(Z,Z) heptacosadiene and inhibited by 7-(Z) tricosene, the main hydrocarbons on the cuticle of females and males, respectively. However, females almost completely lacking cuticular hydrocarbons as a result of a pulse of traF (transformer^F) overexpression in the first few hours after eclosion are still actively courted by males, suggesting the existence of other, as yet unidentified stimulating female pheromones. Such females were used as sexual objects to test whether CheB42a is required for detection of the major hydrocarbons on the female cuticle. In contrast to their different responses to wild-type females, CheB42a and control males attempt to copulate with such heat-shocked hs-tra^F females with indistinguishable frequency and kinetics, suggesting that the behavioral effect of CheB42a depends on the presence of the major female cuticular hydrocarbons. In the same experiment, mutant males display earlier and more frequent copulation attempts toward genetically identical control hs-tra^F females that have not been heat-shocked (Park, 2006).

In an independent and complementary test of the involvement of CheB42a in detection of female cuticular hydrocarbons, the response of CheB42a mutant males and controls was examined toward males whose oenocytes have been feminized and thus produce female-specific cuticular hydrocarbons, in response to targeted expression of traF. As expected, such partially transformed males trigger significant courtship responses both from controls and CheB42a males. However, here again, mutant males outperform controls specifically in the kinetics and frequency of attempted copulations. Together, these data suggest that CheB42a is required for normal gustatory response to female cuticular hydrocarbon pheromones that modulate male behavior (Park, 2006).

What is the mechanism of CheB42a function? CheB42a and all other CheBs contain a predicted signal peptide at their amino terminus, suggesting that they may be secreted (Xu, 2002). Furthermore, the thecogen cells that express CheB42a secrete proteins into the internal cavity of the taste hair, within which pheromone molecules come in contact with the dendritic membranes of gustatory neurons. CheB42a may therefore be secreted into the aqueous medium that fills the internal hair cavity and interact directly with pheromones, gustatory pheromone receptors, or both (Park, 2006).

CheB42a may facilitate, or even be required for, the activation of a gustatory receptor by a cuticular hydrocarbon that delays attempted copulation, similar to the requirement for the LUSH OBP and two moth PBPs for the action potentials observed in olfactory neurons in response to volatile pheromones. Alternatively, CheB42a may reduce the effect of cuticular hydrocarbons that stimulate male courtship behavior. The latter model is supported by the expression of CheB42a in the same subset of taste hairs that house Gr68a-expressing taste neurons, which are required for response to stimulatory female pheromones and whose inactivation results in a much-reduced frequency of late steps but an increased frequency of early steps in the courtship sequence. Gustatory perception of the same pheromones that induce males to progress from early to late courtship steps may therefore be mediated by Gr68a-expressing gustatory neurons and be inhibited by CheB42a. For example, CheB42a may contribute to the removal or inactivation of pheromone molecules in the inner hair lumen; this removal or inactivation must take place to allow monitoring of changing pheromone concentrations in the external environment as a function of time (Park, 2006).

More generally, what could be the adaptive value of a protein that delays a male's attempts to copulate? Slower progression through the courtship sequence as a result of CheB42a modulation of pheromone perception, or through the action of other genes or neural circuits, may contribute to a male's ability to target its mating efforts toward an appropriate partner. Indeed, chemosensory perception of cuticular hydrocarbons in Drosophila plays a critical role in a male's ability to identify a sexual partner of the appropriate species and sex and even allows discrimination between virgin and mated females (Park, 2006).

This work indicates that CheB42a is required for a normal gustatory response to contact female cuticular hydrocarbon pheromones that modulate male courtship behavior and suggests that other CheBs expressed in sexually dimorphic patterns on appendages enriched in chemosensory organs may also modulate pheromone perception. Given the importance of contact pheromones in vertebrates, it will be of great interest to determine whether secreted putative ligand binding proteins expressed specifically in the main vertebrate pheromone-sensing organ, the vomeronasal organ, play similar roles (Park, 2006).

REFERENCES

Gustatory receptor 68a

Bray, S. and Amrein, H. (2003). A putative Drosophila pheromone receptor expressed in male-specific taste neurons is required for efficient courtship. Neuron 39: 1019-1029. 12971900

Clyne, P. J., Warr, C. G. and Carlson, J. R. (2000). Candidate taste receptors in Drosophila. Science 287: 1830-1834. 10710312

Coyne, J. A. and Oyama, R. (1995). Localization of pheromonal sexual dimorphism in Drosophila melanogaster and its effect on sexual isolation. Proc. Natl. Acad. Sci. 92: 9505-9509. 7568163

Del Punta, K., Leinders-Zufall, T., Rodriguez, I., Jukam, D., Wysocki, C. J., Ogawa, S., Zufall, F. and Mombaerts, P. (2002). Deficient pheromone responses in mice lacking a cluster of vomeronasal receptor genes. Nature 419: 70-74. 12214233

Dunipace, L., Meister, S., McNealy, C., and Amrein, H. (2001). Spatially restricted expression of candidate taste receptors in the Drosophila gustatory system. Curr. Biol. 11, 822-835. 11516643

Ferveur, J. F., Cobb, M., Boukella, H. and Jallon, J. M. (1996). World-wide variation in Drosophila melanogaster sex pheromone: behavioural effects, genetic bases and potential evolutionary consequences. Genetica 97: 73-80. 8851882

Lin, H., et al. (2005). A Drosophila DEG/ENaC channel subunit is required for male response to female pheromones. Proc. Natl. Acad. Sci. 102(36): 12831-6. 16129837

Park, S. K., et al. (2006). A Drosophila protein specific to pheromone-sensing gustatory hairs delays males' copulation attempts. Curr. Biol. 16(11): 1154-9. 16753571

Scott, K., Brady, R., Cravchik, A., Morozov, P., Rzhetsky, A., Zuker, C. and Axel, R. (2001). A chemosensory gene family encoding candidate gustatory and olfactory receptors in Drosophila. Cell 104: 661-673. 11257221

Sweeney, S. T., Broadie, K., Keane, J., Niemann, H., and O'Kane, C. J. (1995). Targeted expression of tetanus toxin light chain in Drosophila specifically eliminates synaptic transmission and causes behavioral defects. Neuron 14: 341-351. 7857643

Xu, A., et al. (2002). Novel genes expressed in subsets of chemosensory sensilla on the front legs of male Drosophila melanogaster. Cell Tissue Res. 307(3): 381-92. 11904775

date revised: 20 February 2007

Home page: The Interactive Fly © 2017 Thomas Brody, Ph.D.