Role of Sex Peptide in Drosophila Males

Abstract

Drosophila male sex peptide ACP70A is a small peptide mainly produced in the accessory glands. It elicits a high number of post-mating responses in mated females; yet its function in male physiology is not well known. Here, we explore its role in male sex behavior and pheromone biosynthesis, using males either mutant or RNAi knocked-down for Acp70A. Courtship was severely affected in both Acp70A mutants and Acp70A knocked-down males, with only 2% of the males succeeding copulation. Cuticular hydrocarbon amounts were moderately affected with 25% decrease in sp0 mutant (without Acp70A expression) and 10–22% increase in flies overexpressing Acp70A. Acp70A knock-down either ubiquitously or in the testes surprisingly resulted in an overproduction of hydrocarbons, whose amounts were double of the controls. We tested eight putative “off-target” genes but none of these led to an increase in hydrocarbon amounts. These results show that male courtship behavior is largely dependent on the presence of Acp70A and independent of cuticular hydrocarbons. The presence of potential “off-target” genes explaining the hydrocarbon phenotype is discussed.

1. Introduction

In most reproducing animals, including Drosophila, seminal fluid is transferred along with sperm to females during mating. These seminal fluid components have important effects on female behavior and physiology and have been extensively studied in Drosophila melanogaster [1, 2]. Most of these seminal proteins are synthesized in the accessory glands (AGs) and therefore, named ACcessory gland Proteins (ACPs). One well-characterized ACP, ACP70A (also called SP, sex peptide), plays a major role in eliciting postmating response: it modifies the female behavior, resulting in the rejection of courting males [3, 4]. It has a crucial role on female reproduction: it increases oogenesis [5], egg production [6] and egg-laying [3, 4]. It also induces dramatic effects on female nutrition: it increases food uptake [7], modifies food preference by altering nutrient balancing [8] and alters gut water absorption and intestinal transit [9]. The other physiological modifications are the inhibition of sleep [10] and the regulation of sperm release from the storage organs [11]. All these effects are caused by the binding of the C-terminal part of the sex peptide to a neuronal sex peptide receptor (SPR) in the female [12, 13, 14]. The central part of sex peptide elicits the expression of immune response genes [15], and the N-terminal part activates the corpora allata (CA), inducing increased synthesis of juvenile hormone (JH) [16], which triggers oogenesis and vitellogenic oocyte progression [5] and also leads to decreased pheromone biosynthesis [17].

Whereas, there are numerous studies on the role of male sex peptide on female physiology, there are no such studies concerning male physiology. As we observed that there was a defect in courtship behavior of sex peptide mutant males, we wanted to elucidate the possible roles of ACP70A in male behavior and physiology. In this study, we report clear defects in male sex behavior and moderate defects in hydrocarbon and pheromone synthesis concerning mutant males. Using sex peptide knocked-down males, we confirmed the control of sex peptide on male sex behavior. Conversely, ubiquitous expression of Acp70A-RNAi resulted in a twofold increase in cuticular hydrocarbon (CHC) amounts. We could exclude the role of eight off-targets in this CHC augmentation and localize this RNAi effect in the accessory glands (responsible for a 35% increase) and in the testes (responsible for the rest of the effect). The presence of sperm in the testes does not affect CHC biosynthesis.

2. Materials and methods

2.1. Drosophila strains and rearing

Three strains mutant for sex peptide were used:

• the deficiency Δ130/TM3 (covering the Acp70A gene);

• the point mutant sp0, produced by targeted mutagenesis by homologous recombination [4]. sp0 males were used balanced by TM3 (sp0/TM3: one copy of Acp70A is active) or homozygous (sp0/ sp0: no production of ACP70A), or crossed by Δ130/TM3 (sp0/Δ130: no production of ACP70A).

• DTA-E [18], which are sperm-less and lack ACPs produced from the main cells (96% of the accessory glands).

The laboratory wild-type Canton-S strain was also used as a control.

The following Gal-lines from the Bloomington Drosophila Stock Centre were used: daughterless (da)-Gal4, a ubiquitous driver; elav-Gal4, a driver expressed in the nervous system [19], dopa decarboxylase (ddc)-Gal4, expressed in epidermis and nervous system [20], 1407-Gal4 and PromE-Gal4, both expressed in pupal and adult oenocytes [21, 22], c564-Gal4, expressed in fat body [23], Acp26A-Gal4, expressed in accessory glands [3], svp-Gal80, which specifically blocks Gal4 activity in the oenocytes [24]. Using a UAS-GFP line, we could show that 1407-Gal4 was also expressed in testes and built a line with the following genotype: 1407-Gal4; svp-Gal80 that drives the expression only in the testes. Images were visualized and photographed on a Nikon eclipse E800 microscope with a Cool Snap camera.

A UAS-Acp70A line was generated in our laboratory and noted UAS-Acp70A+ [17]. The following UAS-RNAi-lines were obtained from the VDRC Stock Center and directed against: Acp70A, SP (109,175 KK); lamp1, CG3305, (7309 GD); dco, CG4379 (101,524 KK); rgk1, CG44011 (108,710 KK); CG5961 (100,023 KK); CG15128 (100238KK); CG9413 (108,867 KK); CG8315 (105,654 KK); tinc, CG31247 (101,175 KK).

Drivers were maintained as heterozygous over a Balancer (Cyo or TM3). In all RNAi knock-down (or overexpression) experiments, balanced gal4-driver females were crossed to UAS males. Balanced progeny was taken as the control of RNAi knocked-down (or overexpression) progeny.

Flies were grown at 25°C with 12/12 light–dark (LD) cycles, on standard cornmeal medium. They were separated by sex at emergence and kept sex-separated in groups of 10 in fresh food vials until testing (4 days after emergence).

2.2. Hydrocarbon analyses

CHCs were removed from single 4-day-old flies by washing them for 5 min in 100 μL heptane containing 500 ng hexacosane (n-C26) as an internal standard. The fly was then removed from the vial and 5 μL of each sample was injected into a Perichrom Pr200 gas chromatograph, with hydrogen as the carrier, using a split injector (split ratio 40:1). The oven temperature started at 180°C, ramped at 3°C/min to 300°C, for a total run of 40 min. The data were automatically computed and recorded using Winilab III software (version 04.06, Perichrom) as previously described [17]. As we did not observe significant variation in the CHC profiles, we only represented the total amount of CHCs as means ± SEM (n = 10 for all tests).

2.3. Analysis of Acp70A expression

Quantitative PCR was performed as described [25] using RNA TRIzol™ (Invitrogen) to extract RNAs from 10 adults for each sample. cDNAs were synthesized with SuperScript II, and PCR was perfumed with a LightCycler^® 480 SYBR Green I Master (Roche Applied Science). Primers for Acp70A (5′-ATTCTTGGTTCTCGTTTGCG-3′ and 5′-TAACATCTTCCACCCCAGG-3′) were used. To normalize mRNA amounts, we tested six different genes and used one gene, which was shown to be very stable in all samples: CG7598 (5′-AACGGATGTGGTGTTCGATT-3′ and 5′-TAATGCCATCCTTGGTGTGA-3′). Samples were performed in independent triplicates (each consisting of two technical replicates).

2.4. Mating experiments

A 4-day-old Canton-S female was introduced into the observation chamber, consisting of a watch glass (28-mm diameter and 5-mm internal height) placed on a glass plate and left for 2 min before the introduction of the male. The following parameters were recorded: lengths of courtship, first copulation attempt and copulation latency (time from introduction of the male into the observation chamber to courtship, first copulation attempt or copulation), percentages of courtship, first copulation attempt and copulation (percentages of males performing courtship, copulation attempt or copulation). The effects of genotypes were evaluated by Kruskal-Wallis tests (latencies) and χ²-tests (percentages of flies). N ≥ 50 for all tests.

3. Results

3.1. Expression of Acp70A in adult males

Acp70A expression was not significantly different in controls (Canton-S and Acp26A) and in sp0 mutants that possess a point mutation in the signal sequence. In contrast, Acp70A expression was dramatically inhibited in Acp26A > Acp70A-RNAi males (−99%) and higher expression was observed in Acp26A > Acp70A males (+63%) (Figure 1).

Figure 1.

Transcriptional expression of Acp70A in control, mutant, knocked-down or overexpressing male flies. Each bar represents mean ± SEM of three independent trials. *, ** and *** indicate significant differences (P = 0.05, 0.01 and 0.001, respectively).

3.2. Effect of Acp70A on male sex behavior

Males mutant for Acp70A (sp0/+ and sp0/sp0), overexpressing Acp70A (Acp26A > Acp70A) or RNAi knocked-down (Acp26A > Acp70A-RNAi and da > Acp70A-RNAi) were tested in face of wild-type females (Figure 2).

Figure 2.

Courtship and mating experiments in fly pairs composed of a wild-type (Canton-S) female and a male of a different genotype: percentages of males performing courtship (WB), copulation attempts (CA) and copulation (C) and time needed to initiate these tasks. Effect of the sp0 mutation and overexpression or RNAi knock-down of Acp70A in males (drivers Acp26A-Gal4 and da-Gal4). Each bar represents mean ± SEM of 50 trials. *, ** and *** indicate significant differences (P = 0.05, 0.01 and 0.001, respectively). N is indicated below each bar.

All the steps of courtship were affected in sp0 mutants: the number of heterozygous males that attempted or succeeded copulation decreased by 54 and 73%, respectively. The effect of the homozygous mutation was dramatic: sp0/sp0 males performing courtship (wing vibration) were 5 times fewer than heterozygous or control males, and out of the 50 homozygous males tested, only 3 attempted to copulate and 1 succeeded copulation. Time needed to perform these tasks was higher as well: homozygous sp0 males needed 5 times more than sp0/+ or wild-type males to initiate courtship and the time to attempt copulation was 1.7 and 2 times longer in heterozygous and homozygous mutants, compared to control males.

Overexpression of Acp70A in the accessory glands did not modify the proportion of males performing the different steps of courtship behavior. On the other hand, the time necessary to perform wing vibration was double.

Courtship of males knocked-down for Acp70A in accessory glands was also affected: the percentage of these males performing copulation attempts and copulation was, respectively, 30 and 41% lower when compared to control males. It took them 2 and 1.5 times longer to perform wing vibration and copulation attempts when compared to control males. When knock-down was induced ubiquitously (da > Acp70A-RNAi), the inhibition of courtship was more severe and similar to that observed in homozygous sp0 males.

These results show that there is a significant inhibition of courtship behavior in absence of sex peptide expression.

3.3. Effect of Acp70A on CHCs

Heterozygous sp0 male CHCs were not significantly different from wild-type ones (Figure 3). Conversely, homozygous sp0 males as well as males bearing one sp0 over a deficiency covering the entire Acp70A gene showed a 25% decrease in the total CHC amount. This result confirms that sp0 is a null mutant. Inversely, a ubiquitous overexpression of Acp70A led to a small but significant increase in CHCs (+22%).

Figure 3.

Cuticular hydrocarbon amounts in adult males either mutant for sp0 (left) or overexpressing Acp70A (right) under the Acp26A-gal4 driver. Each bar represents mean ± SEM (n = 10). * indicates significant differences (P = 0.05).

Acp70A ubiquitous knock-down (da > Acp70A-RNAi) was followed by a twofold increase in CHC amount (Figure 4). We thus wondered whether this increase could be due to off-target effect. The RNAi line was described as having no off-target sequence (no gene covering a 19-mers sequence of the RNAi sequence). We performed a Blast analysis with different 16-mers from the RNAi sequence and obtained eight putative off-target genes containing a stretch of coding sequence identical to at least 15-mers of Acp70A sequence (Figure 5). The RNAi of these genes was expressed ubiquitously to measure their effect on male CHCs. For five RNAi tested, we obtained no effect on CHCs and for three RNAi (directed against lamp1, rgk1 and tinc), there was a lower amount of CHCs (from −14 to 22%, depending on the RNAi) (Figure 6). In conclusion, the dramatic increase in CHC amount following ubiquitous Acp70A knock-down cannot be explained by an off-target effect due to these genes.

Figure 4.

Cuticular hydrocarbon amounts in adult males that were RNAi knocked-down for Acp70A in different tissues: ubiquitously (da), in the accessory glands (Acp26A), in fat body (c564), in epidermis (ddc), in nervous system (elav), in oenocytes (Prome), in oenocytes and testes (1407) and in testis (1407; svpgal80). Each bar represents mean ± SEM (n = 10). ** and *** indicate significant differences (P = 0.01 and 0.001, respectively).

Figure 5.

Putative off-target genes containing a stretch of coding sequence identical to at least 15-mers of Acp70A sequence.

Figure 6.

Cuticular hydrocarbon amounts in adult males knocked-down for putative off-target genes. Each bar represents mean ± SEM (n = 10). * and ** indicate significant differences (P = 0.05 and 0.01, respectively).

3.4. Characterization of the tissue involved in CHC control

In males, Acp70A is expressed at a very high level in accessory glands and at a moderate level in testis and carcass (8631, 100 and 95 arbitrary units, respectively; FlyAtlas).

We then wanted to determine the tissue responsible for this effect by targeting Acp70A-RNAi to various tissues. We confirmed the locations of expression of the different Gal4 lines used in this study and showed that 1407-Gal4 was additionally expressed in the testes (Figure 7).

Figure 7.

Photomicrographs showing GFP expression in male reproductive apparatus from 1407-Gal4; svp-Gal80. Fluorescence could be detected only in the testes. Scale bar: 0.5 mm.

No significant effect on CHCs was obtained when Acp70A RNAi was expressed in fat body (c564 > Acp70A RNAi), in epidermis (ddc > Acp70A RNAi) and in oenocytes (PromE > Acp70A RNAi). On the other hand, Acp70A knock-down in accessory glands led to a moderate (+35%) increase in CHC amount. CHC amount was multiplied by a factor of 2 in elav > Acp70A RNAi (nervous system) and a factor of 3 in 1407 > Acp70A RNAi (oenocytes + testes) and 1407; svp-Gal80 > Acp70A -RNAi (testes). This last result shows an essential role of the testes on CHC production (Figure 4).

3.5. CHC profile of the DTAE-line

The DTA-E line is characterized by the absence of accessory glands and some defects in testes, among them, a lack of sperm. DTA-E males were found to produce 1.4-fold more CHCs. We wanted to evaluate the effect of the absence of sperm on CHCs. Four elongase genes are essential to spermatozoid development and the lack of expression in testes leads to sterile males without sperm [26, 27, 28]. We knocked-down these genes in the testes, using the 1407-Gal4 line. We verified the absence of sperm in the RNAi males. None of these genes had any effect on male CHC production (Figure 8).

Figure 8.

Cuticular hydrocarbon amounts in adult males that do not produce sperm: either DTA-E or knocked-down for CG6821, CG17821, CG31141 and CG3971. Each bar represents mean ± SEM (n = 10). ** indicates significant difference (P = 0.01).

4. Discussion

4.1. Sex behavior

Ubiquitous overexpression of sex peptide had no significant effect on male sex behavior: the percentage of males performing the different steps of courtship (wing vibration, copulation attempts and copulation) was unchanged and only the time to begin courtship was lengthened. Conversely, sp0 males showed difficulties to court and the effect was dependent on the dose of the mutant allele: heterozygous sp0 males courted wild-type females the same way as wild-type males did but only a half of them attempted copulation and one-eighth succeeded to mate. The inhibition was more drastic in homozygous sp0 males, as less than one-fifth courted the females and only 2% succeeded to mate.

We tested the males that were RNAi knocked-down for sex peptide in the accessory glands. To target the expression in the accessory glands, we used the driver Acp26A-Gal4. Acp26A gene is almost exclusively expressed in the accessory glands (3589 and 97 units in the accessory glands and the testes, respectively; FlyAtlas). Courtship behavior of Acp26A > Acp70A RNAi males was affected, but less than that of sp0 males: they courted wild-type females the same way as wild-type males did, two-third knocked-down males attempted copulation and less than a half copulated. This result raised the question: does sp0 affect tissues other than accessory glands? When we ubiquitously expressed sex peptide RNAi, we obtained courtship results similar to those with sp0. Taken together, the results suggest a positive control of sex peptide on male courtship behavior. They also pose the problem of the reason of the absence of mating in sp0 and da > Acp70A RNAi males since Q-PCR results clearly show that the expression of Acp70A RNAi in accessory glands via Acp26A-Gal4 reduces Acp70A expression to only 1%.

4.2. Cuticular hydrocarbons

In the female, the transfer of ACP70A during mating induces a decrease in cuticular hydrocarbon amount. This decrease occurs 3 and 4 days after mating and might be due to the overproduction of juvenile hormone following mating, caused by the action of Acp70A on the corpora allata [17]. We therefore wondered whether Acp70A could regulate the production of hydrocarbons in the male. The sp0 mutation as well as Acp70A ubiquitous overexpression led to moderate effects on male CHC production: whereas, wild-type and sp0 heterozygous males had similar CHC amounts, there was a 25% decrease and a 10–22% increase in homozygous sp0 that do not produce ACP70A and da > Acp70A (overproduction of ACP70A) males, respectively. This result seems to be in favor of a positive regulation of sex peptide on CHC production.

The results concerning the effect of Acp70A RNAi on cuticular hydrocarbons were unexpected: a 35% increase occurred when Acp70A expression was inhibited in the accessory glands, using Acp26A-Gal4. Acp26A gene is mainly, but not exclusively, expressed in the accessory glands (3589 and 97 units in the accessory glands and the testes, respectively; FlyAtlas). Acp26A expression in the testes represents 2.7% of the expression in the accessory glands, similar to Acp70A (1.1%). Moreover, a ubiquitous Acp70A knock-down led to a twofold increase in CHC amount; we firstly ascribed this dramatic effect to the presence of possible off-targets of the RNAi.

ACP70A is a small peptide (55 amino acids, including the signal sequence). The nucleic sequence of Acp70A RNAi covers almost the totality of the coding sequence, and also includes the small intron. We found eight putative off-target genes, containing a stretch of coding sequence identical to at least 15-mers of Acp70A RNAi sequence. However, none of these putative off-target genes could be accountable for the dramatic CHC increase resulting in Acp70A RNAi expression.

4.3. Search of the tissue involved in the control on hydrocarbon production

We knocked-down sex peptide in different tissues and could demonstrate that neither the fat body, nor the oenocytes or the epidermis could be responsible for the large rising level of CHCs. On the other hand, sex peptide expression in the testes or in the nervous system led to a CHC increase similar to ubiquitous overexpression.

Sex peptide Acp70A is mainly expressed in the accessory glands, but some expression is also observed in the testes and the carcass (FlyAtlas). Inside the accessory glands, it is exclusively produced by the main cells (96% of the accessory glands) [29]. When we used the DTA-E line in which accessory gland main cell function was genetically disrupted [18], we obtained as well a large-fold increase in CHCs. DTA-E line was obtained after the introduction of diphtheria toxin subunit A (DTA) into the accessory glands via the promoter of Acp95EF [18]. ACP95EF is also a sex peptide produced in the accessory glands and transmitted to the female after mating. It has the same place of production as ACP70A; in the accessory glands, it is exclusively produced in the main cells [29]. Within the fly, it is mainly expressed in the accessory glands and marginally in the testes (787 and 62 arbitrary units, respectively; FlyAtlas). DTA-E males lack ACPs produced from the main cells but have normal secondary cells as well as ejaculatory bulb and duct [30]. DTA-E males are sterile and the block of spermatogenesis occurs at the primary spermatocyte stage [18]. The occurrence of a faint expression of this gene in the testes (FlyAtlas) could explain the lack of sperm. However, the lack of sperm is not directly responsible for the large increase in CHC amounts since flies that did not produce sperm after RNAi knock-down for different elongases involved in sperm production did not increase their CHC production.

The question is: why does DTA-E line show a similar male CHC phenotype to da > Acp70A-RNAi? In the former line, no off-target can be involved. An explanation could be that a “leakage” of the Acp95EF promoter has resulted in a lack of sperm and probably other defects [18]. In males that have been RNAi knocked-down ubiquitously, one may suppose the effect of unknown “off-target” genes that are essential to testis function. This might suggest a role (yet unknown) of the testes in the control of male hydrocarbons.

5. Conclusion

This study demonstrates a role of sex peptide on male courtship behavior. Moreover, the data with DTA-E and RNAi knocked-down flies show the importance of the integrity of the testes (not the sperm) in the control of CHCs.

Acknowledgments

We want to thank Dr. Jacques Montagne for helpful suggestions on the manuscript. Funding was provided by the French Ministry of Research and Education.