Current Biology 20, 1773–1778, October 12, 2010 ª2010 Elsevier Ltd All rights reserved
DOI 10.1016/j.cub.2010.08.056
Report Coupling of Apoptosis and L/R Patterning Controls Stepwise Organ Looping Magali Suzanne,1,2 Astrid G. Petzoldt,1 Pauline Spe´der,1 Jean-Baptiste Coutelis,1 Hermann Steller,2 and Ste´phane Noselli1,* 1University of Nice Sophia-Antipolis, CNRS, Institute of Developmental Biology and Cancer (IBDC), Parc Valrose, 06108 Nice Cedex 2, France 2Howard Hughes Medical Institute, The Rockefeller University, 1230 York Avenue, New York, NY 10065, USA
Summary Handed asymmetry in organ shape and positioning is a common feature among bilateria, yet little is known about the morphogenetic mechanisms underlying left-right (LR) organogenesis. We utilize the directional 360 clockwise rotation of genitalia in Drosophila to study LR-dependent organ looping. Using time-lapse imaging, we show that rotation of genitalia by 360 results from an additive process involving two ring-shaped domains, each undergoing 180 rotation. Our results show that the direction of rotation for each ring is autonomous and strictly depends on the LR determinant myosin ID (MyoID). Specific inactivation of MyoID in one domain causes rings to rotate in opposite directions and thereby cancels out the overall movement. We further reveal a specific pattern of apoptosis at the ring boundaries and show that local cell death is required for the movement of each domain, acting as a brake-releaser. These data indicate that organ looping can proceed through an incremental mechanism coupling LR determination and apoptosis. Furthermore, they suggest a model for the stepwise evolution of genitalia posture in Diptera, through the emergence and duplication of a 180 LR module. Results and Discussion Left-right (LR) asymmetric development is essential to the morphogenesis of many vital organs, such as the heart. Directional looping of LR organs is a complex morphogenetic process relying on proper coordination of early LR patterning events with late cell-tissue behaviors. In vertebrates, several developmental models have been proposed for gut coiling downstream of the Nodal-Pitx2 regulatory pathway, including intrinsic asymmetric elongation of the gut in Xenopus [1] or extrinsic force generation by mesenchymal tissue in Zebrafish [2, 3] and by dorsal mesentery in the chick and mouse embryos [4]. However, the cellular mechanisms underlying LR organ morphogenesis are mostly unknown. In Drosophila, directional clockwise (or dextral) rotation of the genital plate and gut has been shown only recently to be controlled by the LR determinant myosin ID (MyoID; Figure 1A). In myoID mutant flies, LR morphological markers are inverted, leading to counterclockwise (or sinistral) looping of the genital plate, spermiduct, gut, and testis [5–9]. This indicates that myoID is a unique situs inversus gene in Drosophila [10–12].
*Correspondence:
[email protected]
Intriguingly, the expression of MyoID is restricted to two rows of cells within the A8 segment of the genital disc (the analia and genitalia precursor), with one row of expression in the anterior compartment (A8a) and the other in the posterior compartment (A8p) (Figure 1A; for review see [13–15]. Removal of myoID function specifically in the A8 segment is sufficient to provoke the complete inversion of rotation (360 counterclockwise) of the genitalia and sinistral looping of the spermiduct to which it is attached. The A8 segment therefore represents a LR organizer controlling the directional rotation of the whole genitalia in Drosophila [9]. Because circumrotation (the process of 360 rotation) may result from a number of different morphogenetic processes, not deducible from the simple observation of the final adult phenotype, we developed a new and innocuous imaging method to follow the rotation in living pupae (Figure 1C; see Experimental Procedures). To be able to analyze the movement of distinct domains in live developing genitalia, we coupled time-lapse imaging with genital disc ‘‘painting’’ by expressing different fluorophores in various regions of the genitalia precursor. Analysis of wildtype live genitalia through this method revealed their spatial and temporal organization during rotation. We first determined that rotation begins at around 25 hr after puparium formation (APF) and lasts 12–15 hr (Figure 1B), consistent with a previous description made by Gleichauf [16]. At 25 hr APF, the genital disc is organized into concentric rings, which, from anterior to posterior, include an A8a ring, an A8p ring, and a large central disc composed of A9-A10 tissues (Figure 1A). The analysis of rotation in live pupae coupled to manual tracking allowed the identification of two distinct moving domains: a large posterior domain comprising A8p-A9-A10 (hereafter referred to as A8p for simplification, in red in Figures 2A and 2B) and a smaller anterior domain made of A8a (in green in Figures 2A and 2B). The A8p domain moves first and is followed by A8a, which starts moving later on (150 6 50 min after A8p) (Figures 2A and 2B and Movies S1 and S2, available online). During the entire process, cells from the abdomen, to which the genital disc is connected, remain immobile. The finding of two rotating domains, A8a and A8p, was unexpected. It reveals a complex rotational activity of the genitalia and rules out a simple model in which the genital plate would rotate by 360 as a whole. To further understand how rotation occurs, we performed timelapse imaging of the full, 15-hr-long rotation. This analysis revealed that each ring had a different rotational activity. When viewed from the posterior pole, the A8p ring undergoes 360 clockwise rotation, while the A8a ring makes a 180 clockwise rotation (Figure 3A and Movie S3).Whereas the rotation of the central part (A8p-A10) of the disc was inferred from the looping of the spermiduct around the gut, the 180 rotation of A8a was not predicted and could only be revealed by time-lapse analysis because this compartment solely gives rise to a tiny and colorless part of the cuticle [17]. Altogether, these in vivo analyses show that rotation of genitalia in D. melanogaster is a composite process involving two compartments of the A8 segment, A8a and A8p, each expressing a row of MyoID at its anterior boundary (Figure 1A) and having its own rotational behavior.
Current Biology Vol 20 No 19 1774
Figure 1. Genital Disc Organization and Timing of Genitalia Circumrotation (A) The upper panel shows a schematic front view of a pupal genital disc. Segments A8, A9, and A10 are organized into concentric rings. Green shows the A8 anterior compartment (A8a); red shows the A8 posterior compartment (A8p); black stripes show the MyoID expression in A8a and A8p. The lower panel shows the two rows of MyoID expression (green) with A8a and A8p outlined with white dotted lines; hhDsRed (red) marks the posterior compartments of all segments. (B) Distinct stages of genitalia rotation are shown (0 , 90 , 180 , 270 , and 360 ). Genitalia rotation takes place during pupal development and lasts 12–15 hr. The upper panel shows schematic representation of the progressive looping of the spermiduct (Sp) around the gut. The lower panel shows 3D reconstructions of confocal images from pupal genital discs at different angles (0 , 90 , 180 , 270 , and 360 ). Green shows armGFP; red shows hhDsRed. (C) Scheme of the mounting setup used for live imaging of the genital disc in pupae.
These findings raise the questions of the contribution of each of the two rings to the entire rotation and of how they interact during rotation. In order to address this question, we determined the intrinsic or real rotational activity of A8a and A8p. So far, each ring movement was analyzed relative to the same immobile referential: the abdomen. Although this referential allows the real movement of A8a to be determined (Figure 3b), it cannot be used to determine that of A8p, because A8p moves relatively to a mobile referential, i.e., A8a, to which it is attached (Figure 3A). To determine the real movement of A8p, it is thus essential to analyze its angular movement relative to A8a, in other words A8a contribution to motion must be subtracted from the apparent A8p movement. To do so, we have analyzed movies by setting A8a as a referential and by determining the angular movement of A8p. Reassessing A8p movement through this approach revealed that A8p rotates clockwise only by 180 relative to A8a (Figure 3B). The new angular velocity curve of A8p fits almost perfectly with that of A8a, indicating that both movements have similar features (Figure 3B). Importantly, these data also indicate that the observed 360 clockwise rotation is the result of a composite process involving two additive 180 clockwise components: an 180 rotation of the A8a relative to the abdomen and an 180 rotation of A8p relative to A8a. To further determine the autonomy of each ring relatively to the other, we dissected the role of the LR determinant MyoID in this process by specifically inactivating myoID in either A8a or A8p or in both. By convention, the presence or absence of myoID is represented by a + or – sign, respectively.
Accordingly, the wild-type context is noted ‘‘A8a+A8p+’’ and the myoID mutant ‘‘A8a–A8p–.’’ Upon specific inactivation of myoID in the A8a domain (A8a– A8p+ context), the adults showed an apparent ‘‘nonrotation phenotype’’ (0 , no spermiduct looping and genitalia correctly oriented). However, time-lapse imaging revealed that both rings were spinning, although in opposite directions: the A8a domain rotated counterclockwise by 180 (2180 ), whereas the A8p domain rotated clockwise by 180 (+180 , real movement, see Figure 3C and Movie S4). Reciprocally, the inactivation of myoID in the A8p domain (A8a+A8p– context) also led to an apparent nonrotation phenotype. In this context, the behavior of each domain was inverted compared to the previous condition, with the A8a domain rotating clockwise by 180 (+180 ) whereas the A8p domain rotated anticlockwise by 180 (2180 , relative or real movement; Figure 3D and Movie S4). In both cases, the movement of each ring is consistent with its myoID genotype and the ‘‘dextralizing’’ activity of this gene. The strict dependence on MyoID for the direction of the rotation is further confirmed in flies where both A8a and A8p were mutants for myoID (myoIDk1) (see Figure S1; Movie S5). The rotation is often incomplete in this genotype because of the hypomorphic nature of the myoIDk1 allele analyzed [9]; however, both domains show an anticlockwise movement. Therefore, in all genetic contexts analyzed, all parameters of the rotation remain unaffected except the direction of rotation, as illustrated by the perfect mirror image of the angular velocity curves (Figures 3B–3D). These experiments reveal that each ring adopts an independent 180 movement relative to more anterior structures
Apoptosis and L/R Patterning in Organ Looping 1775
Figure 2. Circumrotation Involves Two Distinct Rotating Domains (A and B) Still confocal images taken from AbdB > CD8-GFP; HhDsRed live pupae (Movies S1 and S2). CD8-GFP is shown in green; DsRed is shown in red. Upper panels show whole discs. Lower panels are close-ups of the region shown in dotted-lines. The right panel shows a summary of cell movements. The A8p/A8a boundary is indicated by a dotted line; tracking of A8p, A8a, and abdominal cells is shown in red, green, and blue, respectively. (A) Early A8p rotation. Note the directed clockwise movement of the A8p cells (indicated by white lines and arrows). At this stage, A8a cells show a stochastic movement (see Movie S1). (B) Early A8a rotation. The A8a cells begin to move clockwise whereas cells from the abdomen remain still (see Movie S2).
(A8a relative to the abdomen and A8p relative to A8a): clockwise in the presence of MyoID, anticlockwise in its absence. When both movements are unidirectional, the net rotation is circumrotation (6 360 ), whereas upon opposite movements of A8a and A8p, the net rotation is zero (0 ), leading to an apparent nonrotation phenotype. Therefore, the net rotation (or apparent rotation = R) can be modeled through a simple equation in which R equals the addition of A8a and A8p movements, with MyoID acting as a sign function (Figure 3E). We next wanted to characterize potential cellular mechanisms acting downstream of LR determination during genitalia rotation. In particular, we aimed at determining the cellular events responsible for uncoupling rings at the onset of their rotation. Initial insights came from blocking apoptosis, which leads to genitalia rotation defects ([18–20] and see Figure 4), but the role of apoptosis in the process is not completely understood. To determine the morphogenetic function of the apoptotic pathway during genitalia rotation, we first characterized the spatial and temporal requirements for apoptosis by analyzing the expression pattern of hid and reaper (rpr) in the genital disc, using two reporter lines. Both reporters were strongly expressed in the A9 and A10 segments. However, in the A8 segment, only hid expression is observed
(Figure S2, data not shown). This coincides with the phenotype of misrotated genitalia observed specifically when hid function is altered ([18] and Figure S2) but not in rpr mutants [21]. We then determined the pattern and timing of cell death in the genital disc. To do so, we followed nuclear fragmentation and used an in vivo reporter of caspase activation (the apoliner construct [22]). At the onset of rotation, a large number of apoptotic cells was detected on the most ventral part of the genital disc, first within the A8p ring bordering A8a, coinciding with the beginning of A8p movement. These data indicate an overlap between the apoptotic field and the domain of MyoID expression (Figures 4A and 4B; Movies S6 and S7). These results have been further confirmed by the detection of apoptotic cells by TUNEL staining of fixed pupal genital discs (Figure S2). Later on, a new wave of apoptosis was detected in the most anterior part of the A8a ring, at the junction between A8a and the abdomen (Figure 4B). In contrast, only marginal if any apoptosis was detected before and at the end of rotation (Figure 4B). Therefore, two waves of cell death are taking place in the A8 segment, coinciding spatially and temporally with the rotation of A8a and A8p rings. Given that rings are initially part of the same epithelium and move independently later, we reasoned that local cell death may be a mechanism to provide the degree of liberty
Current Biology Vol 20 No 19 1776
Figure 3. Autonomy and Additive Function of A8 Rings during Circumrotation Absolute (A) and relative (B) movements of A8a and A8p rings in an AbdB > CD8-GFP; hhDsRed pupae. (A) Still confocal images taken from Movie S3 showing full rotation of wild-type genitalia (A8a+A8p+ context). (B) Relative movement of A8p. Same Movie as in (A), with the A8a domain reoriented to serve as a referential, revealing the 180 rotation of A8p. CD8-GFP (green) marks the A8a domain, and DsRed (red) marks the A8p, A9p, and A10p domains. Dotted lines and arrows indicate the angular progression of A8a (in green) and A8p (in red). A8a and A8p domains are outlined (dotted white lines on the upper-left picture). The curves on the right describe the angular variation and the angular velocity of the A8p and A8a domains. (C and D) Relative movement of A8 rings in A8a2A8p+ (C) and A8a+A8p2 (D) contexts. The upper panel shows absolute movement with merged channels. The lower panels show relative movement with A8a serving as a reference. The curves on the right represent the angular variation of the A8p and A8a domains. Genotypes are as follows: UAS-CD8-GFP/+; AbdB > MyoIDRNAi/hhDsRed (C) and ptcDsRed/+; hh > MyoIDRNAi/UAS-CD8-GFP (D). Images are from Movie S4 (C) and Movie S5 (D). (D) Colors have been inverted in the acquisition software to keep the same color code (A8p in red, A8a in green). (E) Equation showing the additive and modular nature of circumrotation.
necessary for proper movement. To test this hypothesis, we inhibited cell death in each compartment separately by expressing the caspase inhibitor p35 [23]. Interestingly, inactivation of apoptosis in either A8p or A8a leads to a similar phenotype, with flies showing a high proportion of half-rotated genitalia (180 rotation, Figure 4C and Figure S2C; Movie S8), suggesting that rotation was blocked in the ring deficient for apoptosis. This has been further demonstrated by following the rotation process in vivo, when apoptosis is specifically blocked in the A8a. In this genetic context, the A8a ring stayed mostly still during the whole process, whereas A8p rotated normally. The resulting 180 rotation is thus exclusively due to the movement of one ring, i.e., A8p, in which apoptosis is unaffected (Figure 4D). Inhibiting apoptosis in both domains strongly aggravates the phenotype, with 40% of the flies showing nonrotated genitalia (0 ), suggestive of an additive phenotype. The rest of the population had 90 rotated genitalia, which may be due to incomplete inhibition of apoptosis (Figure 4C). Alternatively, it is possible that some rotation occurs without apoptosis thanks to tissue elasticity. In any case, our results indicate that cell death is required in each ring for separating them from the neighboring tissues and allowing their free rotation. Consistently, nuclei fragmentation and cell death occur normally in a myoID mutant background (data not shown). Because local cell death is not likely to
provide a direct force for rotation, we propose that it contributes to the release of rings from neighboring tissues. In this study, we reveal that organ looping can proceed through discrete steps, breaking down circumrotation into the simple building blocks of 180 each. The incremental nature of genitalia rotation is indeed based upon two 180 LR modules, sharing identical angular velocity and range as well as requirement for MyoID and apoptosis (Figures 3B–3D). Modularity in morphogenesis provides interesting control mechanisms (through addition or substraction) and therefore plasticity to the process, both at the organism level and during evolution. Entomologists have described different patterns of genitalia rotation in Diptera, ranging from 0 to 360 , that evolved together with changes in mating position [24–26]. Interestingly, in the Brachycera suborder, to which Drosophilidae belong, we notice that most ancestral species have a nonrotated genitalia (Stratiomyomorpha and Tabanomorpha), whereas 180 and 360 rotation have appeared progressively later in evolution (in Muscomorpha and Cyclorrhapha, respectively) (Figure S3). Together with this sequential organization of rotation amplitude in the phylogenetic tree, our data strongly support a model by which the 360 rotation observed in Brachycera (‘‘modern Diptera’’) would result from the emergence (transition from 0 to 180 ) and duplication (transition from 180 to 360 ) of a 180 L/R module (Figure S3), thus providing a simple additive
Apoptosis and L/R Patterning in Organ Looping 1777
Figure 4. Cell Death Pattern and Role in Circumrotation (A) Still confocal images taken from Movie S6. Wild-type rotation of hhDsRed genitalia is shown. Nuclear fragmentation and cell removal are followed by manual tracking in the A8p domain. Individual dying cells as well as the region showing the highest concentration of apoptosis are outlined in white. Lower panels are higher-magnification images of the same movie showing the nuclear fragmentation of a dying cell before its removal (outlined in red). (B) Still confocal images taken from Movie S7, showing wild-type rotation of myoID > Apoliner /armGFP genitalia. Cells in which caspases are activated show nuclear GFP staining (region showing highest concentration of cell fragmentation and removal is outlined in white). Note the presence of dying cells at the anterior boundaries of A8p and A8a, at the beginning of their respective movements. No cell death is observed when rotation is over. Lower panels are closeups of the region shown in dotted lines. (C) Histogram presenting the percentage of the different classes of genitalia rotation phenotypes observed in wild-type (black), AbdB > p35 (green), hh > p35 (red), or AbdB+hh > p35 (red and green stripes). (D) Still confocal images taken from Movie S8 showing the rotation of an AbdB > p35 genital disc. The curves on the right describe the angular variation and the angular velocity of the A8p and A8a domains (compare with wild-type rotation in Figure 3b). Note that the 180 rotation is due exclusively to the movement of A8p, whereas A8a stays mostly still during the whole process. (E) Model summarizing the succession of events taking place in each rotating module, including early LR determination and direction through MyoID (1) and late apoptosis at the boundary of each rotating ring (2). Note that the apoptotic domain is included in the MyoID expression domain.
model for both the origin of circumrotation and the evolution of genitalia rotation and mating position. However, it should be noted that alternative mechanisms may lead to a similar pattern of genitalia rotation among Diptera. The incremental model presented here also offers a solution to the apparent paradox of circumrotation and the question of its elusive utility, illustrated by the fact that both 360 rotation and the absence of rotation lead to the same final posture of genitalia. A facultative role of 360 rotation is further supported by the finding that D. melanogaster males with nonrotating genitalia (A8a–A8p+ or A8a+A8p–) are normally fertile (data not shown). An incremental origin of 360 rotation in which a second half-turn would be added to the existing 180 rotation would well explain this paradox. Thus, circumrotation can be viewed as recapitulating the evolutionary history of genitalia rotation in Brachycera, and its logic would reveal a case of
‘‘retrograde evolution,’’ in which duplication of a functional module is used to revoke a previous evolutionary step. Finally, our analysis of genitalia rotation highlights a new mechanism of morphogenesis relying on a combination of LR patterning and apoptosis (Figure 4E). In this process, we reveal a new role for apoptosis as a releasing mechanism allowing the sliding of two parts of an organ. It will be interesting to test in the future whether this releasing role of apoptosis is used more generally, in other morphogenetic movements requiring important cellular rearrangement. Experimental Procedures Genetics The following fly lines were used: AbdB-Gal4 is described in [27]. hhDsRed and ptcDsRed are from [28]. myoID RNAi and myoIDFK1 are described in [9].
Current Biology Vol 20 No 19 1778
Arm-GFP and UAS-CD8-GFP are from the Bloomington Stock Center. MyoID-Gal4 (NP1548) is from the Kyoto DGRC Stock Center. UAS-Apoliner is from [22] rpr-4kb-lacZ and hid-lacZ are described in FlyBase. Confocal Live Imaging Images were taken with a Zeiss LSM 510 META confocal microscope with 403 0.8 NA water immersion and 103 0.3 NA objectives and assembled with ImageJ 1.40g and Adobe Photoshop 7.0. Time-Lapse Experiments Male pupae were selected at prepupal stage (0–12 hr APF) and incubated at 25 C for 24 hr. The most posterior part of the pupae was removed in water by forceps. The pupae were mounted upside down in an agar plate. The preparation was then covered with water. Acquisitions by confocal microscopy were done every 5 min (Movies S1, S2, and S6) or every 15 min (Movies S3–S5, S7, and S8) during a 15 hr period, except for tracking experiments, where acquisitions were made during a 4 hr period. The tracking of individual cells was done with the ImageJ manual tracking plug-in (Movies S1 and S2). Immunostaining Genital discs were dissected in PBS and fixed for 1 hr in paraformaldehyde 4%. Antibody staining was then performed as previously described in Perez-Garijo et al. [29]. TUNEL Staining TUNEL staining was performed with the in situ cell death detection kit from Chemicon International (ApopTag Peroxidase In Situ Apoptosis Detection kit S71). Supplemental Information Supplemental Information includes three figures and eight movies and can be found with this article online at doi:10.1016/j.cub.2010.08.056. Acknowledgments We thank S. Schaub for help with live imaging and V. Orgozozo, C. Braendle, and members of the laboratory for discussions and critical reading of the manuscript. H.S. is an investigator of the Howard Hughes Medical Institute and part of this work was supported by National Institutes of Health grant R01 GM60124 (to H.S.). Work in the Noselli laboratory is supported by grants from the Centre National de la Recherche Scientifique (CNRS), French Ministry of Research, Association pour la Recherche sur le Cancer (ARC), Agence Nationale de la Recherche (ANR), and Fondation pour la Recherche Me´dicale (FRM). Received: July 29, 2010 Revised: August 25, 2010 Accepted: August 25, 2010 Published online: September 9, 2010 References 1. Muller, J.K., Prather, D.R., and Nascone-Yoder, N.M. (2003). Left-right asymmetric morphogenesis in the Xenopus digestive system. Dev. Dyn. 228, 672–682. 2. Horne-Badovinac, S., Lin, D., Waldron, S., Schwarz, M., Mbamalu, G., Pawson, T., Jan, Y., Stainier, D.Y., and Abdelilah-Seyfried, S. (2001). Positional cloning of heart and soul reveals multiple roles for PKC lambda in zebrafish organogenesis. Curr. Biol. 11, 1492–1502. 3. Horne-Badovinac, S., Rebagliati, M., and Stainier, D.Y. (2003). A cellular framework for gut-looping morphogenesis in zebrafish. Science 302, 662–665. 4. Davis, N.M., Kurpios, N.A., Sun, X., Gros, J., Martin, J.F., and Tabin, C.J. (2008). The chirality of gut rotation derives from left-right asymmetric changes in the architecture of the dorsal mesentery. Dev. Cell 15, 134–145. 5. Bate, M., and Martinez Arias, A. (1993). The Development of Drosophila melanogaster. In The Development of Drosophila melanogaster, Volume 2, M. Bate and A. Martinez Arias, eds. (New York: Cold Spring Harbor Laboratory Press), pp. 843–897. 6. Crampton, G.C. (1941). The terminal abdominal structures of male diptera. Psyche (Stuttg.) 48, 79–94.
7. Crampton, G.C. (1941). The external morphology of Diptera. Bull. Conn. Geol. Natur. Hist. 64, 10–165. 8. Hozumi, S., Maeda, R., Taniguchi, K., Kanai, M., Shirakabe, S., Sasamura, T., Speder, P., Noselli, S., Aigaki, T., Murakami, R., et al. (2006). An unconventional myosin in Drosophila reverses the default handedness in visceral organs. Nature 440, 798–802. 9. Speder, P., Adam, G., and Noselli, S. (2006). Type ID unconventional myosin controls left-right asymmetry in Drosophila. Nature 440, 803–807. 10. Coutelis, J.B., Petzoldt, A.G., Speder, P., Suzanne, M., and Noselli, S. (2008). Left-right asymmetry in Drosophila. Semin. Cell Dev. Biol. 19, 252–262. 11. Speder, P., and Noselli, S. (2007). Left-right asymmetry: Class I myosins show the direction. Curr. Opin. Cell Biol. 19, 82–87. 12. Speder, P., Petzoldt, A., Suzanne, M., and Noselli, S. (2007). Strategies to establish left/right asymmetry in vertebrates and invertebrates. Curr. Opin. Genet. Dev. 17, 351–358. 13. Estrada, B., Casares, F., and Sanchez-Herrero, E. (2003). Development of the genitalia in Drosophila melanogaster. Differentiation 71, 299–310. 14. Gorfinkiel, N., Sanchez, L., and Guerrero, I. (2003). Development of the Drosophila genital disc requires interactions between its segmental primordia. Development 130, 295–305. 15. Sanchez, L., and Guerrero, I. (2001). The development of the Drosophila genital disc. Bioessays 23, 698–707. 16. Gleichauf, R. (1936). Anatomie und Variabilita¨t des Geschlechtsapparates von Drosophila melanogaster (Meigen). Z. Wiss. Zool. 148, 1–66. 17. Keisman, E.L., Christiansen, A.E., and Baker, B.S. (2001). The sex determination gene doublesex regulates the A/P organizer to direct sex-specific patterns of growth in the Drosophila genital imaginal disc. Dev. Cell 1, 215–225. 18. Abbott, M.K., and Lengyel, J.A. (1991). Embryonic head involution and rotation of male terminalia require the Drosophila locus head involution defective. Genetics 129, 783–789. 19. Grether, M.E., Abrams, J.M., Agapite, J., White, K., and Steller, H. (1995). The head involution defective gene of Drosophila melanogaster functions in programmed cell death. Genes Dev. 9, 1694–1708. 20. Macias, A., Romero, N.M., Martin, F., Suarez, L., Rosa, A.L., and Morata, G. (2004). PVF1/PVR signaling and apoptosis promotes the rotation and dorsal closure of the Drosophila male terminalia. Int. J. Dev. Biol. 48, 1087–1094. 21. Peterson, C., Carney, G.E., Taylor, B.J., and White, K. (2002). reaper is required for neuroblast apoptosis during Drosophila development. Development 129, 1467–1476. 22. Bardet, P.L., Kolahgar, G., Mynett, A., Miguel-Aliaga, I., Briscoe, J., Meier, P., and Vincent, J.P. (2008). A fluorescent reporter of caspase activity for live imaging. Proc. Natl. Acad. Sci. USA 105, 13901–13905. 23. Hay, B.A., Wolff, T., and Rubin, G.M. (1994). Expression of baculovirus P35 prevents cell death in Drosophila. Development 120, 2121–2129. 24. Bickel, D. (1990). Sex with a twist in the tail: The common-or-garden housefly. NewScientist 1731, 24–27. 25. Griffiths, G.C. (1972). The Phylogenetic Classification of Diptera Cyclorrhapha (The Hague: W. Junk). 26. Huber, B.A., Sinclair, B.J., and Schmitt, M. (2007). The evolution of asymmetric genitalia in spiders and insects. Biol. Rev. Camb. Philos. Soc. 82, 647–698. 27. de Navas, L., Foronda, D., Suzanne, M., and Sanchez-Herrero, E. (2006). A simple and efficient method to identify replacements of P-lacZ by P-Gal4 lines allows obtaining Gal4 insertions in the bithorax complex of Drosophila. Mech. Dev. 123, 860–867. 28. Akimoto, A., Wada, H., and Hayashi, S. (2005). Enhancer trapping with a red fluorescent protein reporter in Drosophila. Dev. Dyn. 233, 993–997. 29. Perez-Garijo, A., Martin, F.A., and Morata, G. (2004). Caspase inhibition during apoptosis causes abnormal signalling and developmental aberrations in Drosophila. Development 131, 5591–5598.
