Mild mutations in the pan neural gene prospero affect male-specific behaviour in Drosophila melanogaster

Behavioural Processes 65 (2004) 7–13

Mild mutations in the pan neural gene prospero affect male-specific behaviour in Drosophila melanogaster Yaël Grosjean a , Mathilde Savy a , Julien Soichot a , Claude Everaerts a , Frank Cézilly b , Jean-François Ferveur a,∗ a b

CNRS-UMR 5548, Faculté des Sciences, Université de Bourgogne, 6 Blvd Gabriel, 21000 Dijon, France CNRS-UMR 5561, Faculté des Sciences, Université de Bourgogne, 6 Blvd Gabriel, 21000 Dijon, France Received 11 October 2002; received in revised form 16 December 2002; accepted 17 February 2003

Abstract The fruitfly Drosophila melanogaster is one of the most appropriate model organisms to study the genetics of behaviour. Here, we focus on prospero (pros), a key gene for the development of the nervous system which specifies multiple aspects from the early formation of the embryonic central nervous system to the formation of larval and adult sensory organs. We studied the effects on locomotion, courtship and mating behaviour of three mild pros mutations. These newly isolated pros mutations were induced after the incomplete excision of a transposable genomic element that, before excision, caused a lethal phenotype during larval development. Strikingly, these mutant strains, but not the strains with a clean excision, produced a high frequency of heterozygous flies, after more than 50 generations in the lab. We investigated the factors that could decrease the fitness of homozygotes relatively to heterozygous pros mutant flies. Flies of both genotypes had slightly different levels of fertility. More strikingly, homozygous mutant males had a lower sexual activity than heterozygous males and failed to mate in a competitive situation. No similar effect was detected in mutant females. These findings suggest that mild mutations in pros did not alter vital functions during development but drastically changed adult male behaviour and reproductive fitness. © 2003 Elsevier B.V. All rights reserved. Keywords: Drosophila; Reproductive fitness; Mating; Locomotion; prospero

1. Introduction Drosophila melanogaster is one of the best model organisms to study the genetic pathways and mechanisms underlying early development (NüssleinVolhard and Wieschaus, 1980). Among the genes that have a large impact on Drosophila development, we ∗ Corresponding author. Tel.: +33-380-39-37-82/62-19; fax: +33-380-39-6289. E-mail address: [email protected] (J.-F. Ferveur).

have focused on prospero (pros) because it is a key gene for the development of the nervous system. The pros gene specifies multiple aspects from the early formation of the embryonic central nervous system to the formation of larval and adult sensory organs (Doe et al., 1991; Vaessin et al., 1991). However, little is known about the effects of mild pros mutations on fly behaviour. We studied courtship and mating behaviours which are two of the many pleiotropic characters that pros could change. Transposable genomic elements (transposons) are versatile tools which have been used to unravel the

0376-6357/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0376-6357(03)00148-7

8

Y. Grosjean et al. / Behavioural Processes 65 (2004) 7–13

function of various genes (O’Kane and Gehring, 1987; Brand and Perrimon, 1993). We made use of a particular D. melanogaster variant strain (prosVoila1 or prosV1 ) which contains a PGal4 transposon inserted at 216 base pairs upstream of pros (Balakireva et al., 1998; Grosjean et al., 2001). New mutant alleles, also called excision alleles (prosVexc ), were created after the remobilization of the transposon. In some cases, the properly excised transposon allowed the rescue of the wild-type phenotype (Balakireva et al., 2000), indicating that the transposon was causing the mutant phenotype. However, in other cases, excision was incomplete and induced new prosV alleles in which homozygotes exhibited a lethality ranging from larval life to late imaginal life. A survey of 19 prosVexc alleles revealed that developmental lethality of homozygotes was correlated with the size of the inserted fragment (Grosjean et al., 2001). We studied some of the prosVexc alleles that produced viable adults of both sexes. Some of these alleles still have a tiny piece of DNA inserted, and are qualified as “mild” allele with regard to their effect on lethality. We found a difference between these mild alleles because some of them still produced heterozygous flies, after more than 50 generations. This difference between excision alleles could be caused by fitness difference between genotypes and in particular by increased fitness of heterozygous males. We therefore tested and compared several parameters involved in the reproduction of homo- and heterozygotes for various mild prosVexc alleles.

2. Materials and methods 2.1. Strains All D. melanogaster strains were maintained on standard cornmeal and yeast medium under a 12-h-dark/12-h-light cycle at 25 ◦ C. A description of the chromosomes and mutations used in this study can be found in Lindsley and Zimm (1992). The PGal4-prosV1 insertion line in the pros locus has been previously characterized for adult expression and behaviour (Balakireva et al., 1998), and for preimaginal expression and behavioural defects (Balakireva et al., 2000). Since prosV 1 is a recessive lethal allele, the prosV 1 chromosome was maintained over TM3, a bal-

ancer chromosome carrying the dominant mutations Stubble and Serrate (Sb, Ser). To generate derivative lines of prosV1 , we used the scheme described by Cooley et al. (1988) to mobilize the prosV 1 -PGal4 transposon. Excision of the transposon was performed by crossing prosV 1 /TM3 females with males from a jump starter strain which provided the P-element transposase (Robertson et al., 1988). Male progeny carrying the prosV 1 -PGal4 transposon and the transposase-producing 2–3 chromosome were crossed to white (w− ); +/TM3 females. In order to isolate flies without the P transposon, males of the next generation were scored for the loss of the mini-white+ gene. Each excision allele derived from the prosV 1 allele resulted from a unique excision event on the chromosome 3 that was subsequently maintained over a TM3 balancer chromosome. All strains were backcrossed to the control Cs strain in order to have a wild-type copy of the endogenous white (w+ ) gene. The variation of male courtship therefore cannot be related to the w− mutation. 2.2. Behavioural tests Courtship tests were carried out on 4-day-old female and male flies. After eclosion, females were kept in groups of five and males were kept alone. To compare male genotypes, a control female was individually aspirated into an observation chamber (2.8 cm diameter, 0.5 cm height). After 10 min, one or two (for no-choice or choice experiment, respectively) male(s) was(were) introduced. The courtship index value (CI) is the percentage of time that the subject male spends courting in a 10-min observation period. Courtship latency corresponds to the time at which the male shows the first stereotypical behaviour (tapping, wing vibrating, licking or attempted copulation) towards the female. For the comparison of female genotypes, two virgin females (one homozygote and one heterozygote of the same strain) were placed in an observation chamber, followed 10 min later by a single heterozygous male belonging to the same prosVexc strain. Male CI was measured toward either an intact or a decapitated female. Females were always decapitated 30 min prior to the experiment. Decapitation standardizes the duration of male courtship and thus

Y. Grosjean et al. / Behavioural Processes 65 (2004) 7–13

mostly measures female attractiveness (Ferveur et al., 1995). The CI towards an intact female measures both female attractiveness and receptivity. The comparison of CIs for decapitated versus intact female measure female receptivity, provided that headless and intact females induce different levels of male courtship. Each fly was tested only once. We only studied CI values >5; males with lower values were considered as “non-courters” and discarded. CI values were normally distributed and compared with a Student’s t-test. Latencies were log transformed before being compared with a Student’s t-test. Frequencies of courtship initiation and copulation were compared with a Chi-square test (with a Yates correction for one d.f.). Locomotor activity was measured in similar environmental conditions. We averaged the total number of lines drawn under the mating chamber crossed by the fly (locomotor activity units; Balakireva et al., 1998). For each experiment, four single flies were sequentially observed for five periods of 20 s, every minute, for 5 min. For each strain, locomotor activity was compared with a Mann–Whitney U test. The level of interstrain correlation between male locomotor activity and CI was determined with a Spearman rank test. 2.3. Fertility tests Single females were paired for 4 h with a single male. After that period, the male was discarded, and females were individually transfered to a fresh food vial every 5 days. For each mated female, viable adult progeny were counted 21 days after the first day of adult eclosion. We considered only females laying at least one fertile egg because we found no difference between genotypes for the frequency of non-laying females (6–11%). Data were compared using Kruskall–Wallis test and Mann–Whitney U test (Siegel and Castellan, 1988). When a significant difference was detected with a Kruskal–Wallis test, genotypes were compared two by two using a Mann–Whitney U test. As we performed three U tests, we corrected the significance levels according to the number of comparisons, and divided the level of significance by three (Holm, 1979). Therefore, the first threshold of significance (∗ P < 0.05) is reached with P < 0.0165.

9

3. Results 3.1. Heterozygous pros mutant flies are still present in some excision lines Molecular characterization of the DNA flanking the insertion was carried out in the seven alleles yielding viable and fertile homozygous flies. Three alleles had a clean excision of the transposon (including the DNA of prosV 14 allele which was sequenced) whereas the other four alleles (including prosV 28 , prosV 60 and prosV 65 ) still retained a small fragment of the transposon (size: 60–120 bp), mostly including the transposon feet (Grosjean et al., 2001; Grosjean, unpublished data). When looking at the progeny carrying some of these mutant alleles, more than 50 generations after their creation, we found that heterozygous flies of both sexes were still present at a high frequency (±65%) in prosV28 , prosV60 and prosV65 lines. In contrast, the clean excision allele prosV 14 produced only homozygotes. The high frequency of heterozygotes suggests that the chromosome carrying either pros mutant allele (prosV 28 , prosV 60 or prosV 65 ) causes a fitness difference between homo- and heterozygous flies. We investigated some of the major reproductive traits that could explain such a fitness difference between pros mutant genotypes. 3.2. Fertility Preliminary observations performed on intragenotype crosses (homo × homo; hetero × hetero) yielded abundant progeny for all prosV 28 , prosV 60 and prosV 65 mutant alleles and for the rescued prosV 14 allele (data not shown). The progeny of homo- and heterozygous females and males flies with the prosV 60 allele was measured in reciprocal crosses performed with the control Cs strain (Fig. 1). A Kruskall–Wallis test revealed differences between crosses (H7 = 72.44; P = 0.0001). However, only a small but significant difference for the number of progeny was found between homo- and heterozygous mutant females mated to Cs males (N1 = 14; N2 = 12; z = 2.78; P = 0.0055). Moreover, males of both mutant genotypes showed either no significant difference when mated to Cs females (N1 = 14; N2 = 15; z = 0.524; P = 0.60) or a slightly significant difference with

10

Y. Grosjean et al. / Behavioural Processes 65 (2004) 7–13

300

*

Adult progeny

250

200

* 150

n.s.

100

50

0

female male

Cs

60/TM3

60/60

Cs

Cs 60/TM3

60/TM3 60/60

60/TM3

60/60 60/60

Fig. 1. Number of viable adult progeny resulting from the cross between a single pair of flies of various genotypes. Data shown represent mean ± S.E.M. For each mated female, progeny was counted over the 21 days following the first day of adult eclosion. The genotypes of female and male parents are shown under the histogram. We checked for quantitative differences between 60/TM3 and 60/60 genotypes of either sex, when reciprocally paired with a Cs fly. The statistical differences between mutant genotypes are shown above the bars to be compared: ∗ P < 0.0175 (0.05:3; see Section 2); n.s.: not significant (N = 12–15).

homotypic females (N1 = 13; N2 = 15; z = 2.42; P = 0.0156). 3.3. Male courtship behaviour and locomotor activity Male courtship behaviour was measured for homoand heterozygous males paired with a single control virgin female. Two series of tests were carried out: (1) no-choice experiment with a single male of either genotype, and (2) a choice experiment with two males (one of each genotype). No-choice experiments were carried out with hetero- and homozygous males for the three prosV 28 , prosV 60 and prosV 65 mutant alleles and for the rescued prosV 14 allele (Table 1). Male CI and latency to initiate courtship were measured toward either an intact or a decapitated female. When paired with an intact female, all heterozygous mutant males showed a more vigourous courtship (e.g. a higher CI and a shorter latency) than homozygous males (for CIs: d.f. = 48–57; t = 2.85–5.29; P = 0.006–0.001).

However, no difference was noted between the two prosV 14 genotypes (for CIs: d.f. = 46; t = 0.55; P = 0.58), indicating that the transposon fragment still inserted in the DNA of the former strains is responsible for the behavioural difference. This result also indicates that the balancer chromosome, shared by all strains, has no significant effect on behaviour. When paired with a decapitated female, the difference between homo- and heterozygous males was not significant or weaker than with intact females (for CIs: d.f. = 50–55; t = 0.54–3.11; P = 0.59–0.002). This result means that males of all genotypes were more or less similarly aroused by their female partner. The genotype difference in male courtship intensity toward intact and decapitated female could be caused either by variation in female receptivity or in male activity, or both. Although we cannot formally exclude the first hypothesis, our results indicate a difference in male activity, because heterozygous males showed a higher locomotor activity which was highly correlated with their CI (Spearman correlation: n = 8;

Y. Grosjean et al. / Behavioural Processes 65 (2004) 7–13

11

Table 1 Locomotor activity, courtship index and mating frequency of homo- and heterozygous males with various prosV alleles Genotype (male)

Locomotor activity

Courtship with control females (Cs) Decapitated Latency (s)

Intact CI

Latency (s)

CI

Copulation/total

28/TM3 28/28

66.1 ± 4.6 33.2 ± 3.9∗∗∗

65 ± 13 71 ± 15 (n.s.)

71 ± 2 59 ± 5∗

61 ± 27 153 ± 18∗∗

70 ± 6 48 ± 6∗∗

20/30 9/29∗∗

60/TM3 60/60

61.2 ± 3.7 35.7 ± 2.8∗∗∗

55 ± 6 77 ± 13 (n.s.)

71 ± 2 65 ± 4 (n.s.)

41 ± 5 106 ± 20∗∗

73 ± 2 50 ± 4∗∗∗

24/34 16/28 (n.s.)

65/TM3 65/65

59.2 ± 3.8 24.2 ± 4.0∗∗∗

82 ± 17 140 ± 33 (n.s.)

69 ± 2 49 ± 5∗∗

89 ± 17 180 ± 39∗

68 ± 4 37 ± 6∗∗∗

17/28 8/22 (n.s.)

14/TM3 14/14

45.3 ± 3.7 37.6 ± 2.8 (n.s.)

46 ± 6 49 ± 8 (n.s.)

111 ± 24 97 ± 29 (n.s.)

51 ± 4 54 ± 6 (n.s.)

16/24 14/21 (n.s.)

68 ± 14 79 ± 9 (n.s.)

Data shown represent mean ± S.E.M. Locomotor activity was measured in L.A.U. (Balakireva et al., 1998) (N = 18–20). The courtship index (CI) represents the duration of male courtship during 10 min; courtship latency (measured in seconds) is the time necessary for the male to establish the first contact with the female. CI and courtship latency were measured toward both decapitated (N = 25–32) and intact females. The number of copulating pairs is shown relative to the total number of tests. The significance level of t-tests is indicated besides each set of data to be compared: ∗ P < 0.1; ∗∗ P < 0.01; ∗∗∗ P < 0.001; n.s.: not significant. CIs for heterozygous males carrying a wild-type chromosome 3 (from the Cs strain) were not significantly different from those for males carrying the TM3 balancer chromosome (data not shown).

Rs = 0.952; P = 0.0117). Furthermore, heterozygous males paired with intact females had a tendency to mate more often than homozygotes, although only prosV 28 allele showed a significant difference (χ2 = 7.32, P < 0.01). 3.4. Competition between heterozygous and homozygous males When two males (one of each genotype) were simultaneously paired with a single intact control female, the difference between male genotypes was very clear (Table 2). During the 10-min observation period, heterozygous males performed courtship much more frequently than homozygous males (χ2 = 71.3; P < 0.001). A similar difference was observed for the three mutant alleles (χ2 = 21.43–27.86; P < 0.001). Also, heterozygotes initiated courtship first more frequently than homozygotes (χ2 = 239.96; P < 0.001). Strikingly, only heterozygotes copulated within a 1-h period. 3.5. Female behaviour Courtship and mating behaviour of homo- and heterozygous intact females paired with a single heterozygous male was measured within each mutant strain

(data not shown). No difference between female genotypes was detected for male CI or for copulation frequency. This result indicates that the three mild pros mutations do not significantly affect female courtship and mating propensity. Furthermore, no difference for female locomotor activity was found between strains and genotypes (data not shown). It is possible that Table 2 Mating competition of two males with a single control Cs female prosV /TM3

vs.

prosV /prosV

120

∗∗∗

65

Initiate courtship (within 10 min) prosV28 line prosV60 line prosV65 line

45/45 40/40 35/35

First to initiate Copulation (within 1 h)

117 80

∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗

27/45 23/40 15/35 3 0

Two males, one heterozygous prosV /TM3 and one homozygous prosV /prosV , were simultaneously paired with a virgin female. The frequency of male initiating courtship (within the 10-min observation period) was noted for each genotype, and for each mutant allele. We also noted the genotype of the first male initiating courtship. The frequency of copulation was measured over a 1-h period. Male genotypes were distinguished by clipping the distal tip of the right wing, which was alternatively clipped for each genotype. Male frequencies were compared with a χ2 test with a Yates correction: ∗∗∗ P < 0.001.

12

Y. Grosjean et al. / Behavioural Processes 65 (2004) 7–13

some mild pros mutations affect subtle aspects of female behaviour, but this will require more intensive study.

4. Discussion Our data provides a link between the developmental and behavioural effects of the key-developmental gene pros. The prosV alleles studied here are probably mild hypomorph mutations because they did not affect vital functions during development in homozygous flies. The slight fertility decrease observed in homozygotes, relatively to heterozygotes, could be caused by disadvantageous recessive factors on the homozygous chromosome 3. The most striking difference between mutant genotypes concerned male adult behaviour. The phenotypes induced by these mild mutant alleles can be compared with those of other prosV alleles with a larger inserted transposon fragment. We consider the latter alleles as strong or medium hypomorphic pros mutations because they induced complete lethality between early larval stage and early imaginal life (<2 days at 25 ◦ C; Grosjean et al., 2001). In addition, these strong or moderate hypomorphic alleles induced defective larval locomotion and gustatory behaviour but this was not the case for the mild prosV alleles tested here (Grosjean et al., 2003). Although it is possible that the mating differences between male genotypes were a result of differential female receptivity, we believe that it was rather caused by increased courtship intensity which was correlated with increased locomotor activity in heterozygous mutant males. It is worth noting that prosV 28 , prosV 60 and prosV 65 heterozygous males showed a similarly vigorous heterosexual courtship and locomotor activity to that shown by heterozygous prosV 1 males (Balakireva et al., 1998). Comparison of the CI shown by these mutant males with the CIs of homo- and heterozygous males for the prosV 14 rescued allele indicates that the transposon fragment (still inserted in the former strains) is likely to be responsible for the increased activity of heterozygous mutants. The fact that prosV 28 , prosV 60 and prosV 65 homozygous males showed a CI similar to that of rescued prosV 14 (homo- and heterozygous) males indicates that heterozygous mutants produced an abnormaly strong CI. The fact that CI and locomotor activity in these mutant males and in

heterozygous prosV 1 males were similarly increased suggests that one copy of defective prosV mutation can cause these male-specific behavioural defects. We are currently trying to map the region of the nervous system that could be involved in controling these behaviours. The prosV 1 shows a strong expression in the larval and adult gustatory sensory organs and in the mushroom bodies (MBs; Balakireva et al., 1998, 2000). MBs of flies and of other insects have been shown to control male-specific behaviour and the level of locomotor activity (Wadepuhl and Huber, 1979; Ferveur et al., 1995). Finally, our results show that mild mutations of the pros gene, probably due to a slightly altered level of Pros protein, have an impact on differential reproduction between mutant genotypes. These results are important because they show that pros is involved in both the development of the nervous system, and in reproductive behaviour. These data could contribute to the unification of the genetics of behaviour (deBelle, 2002) because they support the hypothesis that naturally occuring behavioural variants can be caused by mild lesion in pleiotropic genes (Greenspan, 1997), as recently demonstrated by Toma et al. (2002). We also need to understand why these mild pros alleles had a male-specific behavioural effect specially on locomotor activity. Such male-specific effect could be related to the sexual dimorphism of adult locomotor activity whose nervous focus has been mapped in the neural region (pars intercerebralis) where prosV 1 is strongly expressed (Gatti et al., 2000).

Acknowledgements This research was largely supported by the Centre National de la Recherche Scientifique and by the Conseil Régional de Bourgogne. YG received a grant from the French Ministry of Research and Education. Fabien Lacaille and Angel Acebes are thanked for comments on the manuscript, and Laurence Dartevelle for technical assistance.

References Balakireva, M., Stocker, R.F., Gendre, N., Ferveur, J.-F., 1998. Voila, a new Drosophila courtship variant that affects the

Y. Grosjean et al. / Behavioural Processes 65 (2004) 7–13 nervous system: behavioral, neural, and genetic characterization. J. Neurosci. 18, 4335–4343. Balakireva, M., Gendre, N., Stocker, R.F., Ferveur, J.-F., 2000. The genetic variant Voila1 causes gustatory defects during Drosophila development. J. Neurosci. 20, 3425–3433. deBelle, J.S., 2002. Unifying the genetics of behaviour. Nat. Genet. 31, 329–330. Brand, A.H., Perrimon, N., 1993. Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118, 401–415. Doe, C.Q., Chu-Lagraff, Q., Wright, D.M., Scott, M.P., 1991. The prospero gene specifies cell fates in the Drosophila central nervous system. Cell 65, 451–464. Ferveur, J.-F., Störtkuhl, K.F., Stocker, R.F., Greenspan, R.J., 1995. Genetic feminization of brain structures and changed sexual orientation in male Drosophila. Science 267, 902–905. Gatti, S., Ferveur, J.-F., Martin, J.-R., 2000. Genetic identification of neurons controlling a sexually dimorphic behaviour. Curr. Biol. 10, 667–670. Greenspan, R.J., 1997. A kinder, gentler genetic analysis of behaviour: dissection gives way to modulation. Curr. Opin. Neurobiol. 7, 805–811. Grosjean, Y., Balakireva, M., Dartevelle, L., Ferveur, J.-F., 2001. PGal4 excision reveals the pleiotropic effects of Voila, a Drosophila locus that affects development and courtship behaviour. Genet. Res. 77, 239–250. Grosjean, Y., Lacaille, F., Acebes, A., Clemencet, J., Ferveur, J.-F., 2003. Taste, movement and death: varying effects of new

13

prospero mutants during Drosophila development. J. Neurobiol. 55, 1–13. Holm, S., 1979. A simple sequentially rejective multiple test procedure. Scan. J. Stat. 6, 65–70. Lindsley, D.L., Zimm, G.G., 1992. The Genome of Drosophila melanogater. Academic Press, San Diego. Nüsslein-Volhard, C., Wieschaus, E., 1980. Mutations affecting segment number and polarity in Drosophila. Nature 287, 795– 801. O’Kane, C.J., Gehring, W.J., 1987. Detection in situ of genomic regulatory elements in Drosophila. Proc. Natl. Acad. Sci. U.S.A. 84, 9123–9127. Robertson, H.M., Preston, C.R., Phillis, R.W., Johnson-Schiltz, D.M., Benz, W.K., Engels, W.R., 1988. A stable genomic source of P-element transposase in Drosophila melanogaster. Genetics 118, 461–470. Siegel, S., Castellan, N.J., 1988. Nonparametric Statistics for the Behavioural Sciences, 2nd ed. McGraw-Hill, New York. Toma, D.P., White, K.P., Hirsch, J., Greenspan, R.J., 2002. Identification of genes involved in Drosophila melanogaster geotaxis, a complex behavioural trait. Nat. Genet. 31, 349–353. Vaessin, H., Grell, E., Wolff, E., Bier, E., Jan, L.Y., Jan, Y.N., 1991. Prospero is expressed in neuronal precursors and encodes a nuclear protein that is involved in the control of axonal outgrowth in Drosophila. Cell 67, 941–953. Wadepuhl, M., Huber, F., 1979. Elicitation of singing and courtship movements by electrical stimulation of the brain of the grasshopper. Natürwissenschaften 66, 320.