Cytogenetics Laboratory, Department of Zoology, Banaras Hindu University, Varanasi 221 005, India

*Corresponding author (Fax, 91-542-317457; Email, jkroy@banaras.ernet.in).

The segment polarity genes wingless (wg) and engrailed (en) have been shown to play important roles in pattern formation at different stages of Drosophila development in the thoracic imaginal discs. We have studied the patterns of expression of these genes in genital discs from wild type larvae, pupae and pharate adults and also from hetero-allelic mutant combinations of these genes. Our results suggest that these genes play vital roles in the normal development and differentiation of genital discs and gonads. In the absence of normal wg or en functions, the flies showed a complete lack of internal accessory reproductive organs and specific defects in the external genitalia. In addition, the testes in such males were small, rounded and with an abnormal cellular organization, although the ovaries in females appeared normal. Temperature shift experiments using the conditional mutant allele of wg, (wg^IL-114) indicated a requirement of wg signaling from second instar onwards for normal development and differentiation of the accessory reproductive organs. Using a heat-shock allele (Hs-wg) we also show that the spatially regulated expression of wg as a pre-requisite for normal development and differentiation. Based on the expression patterns of en and hedgehog (hh) we suggest that even in the genital disc development and differentiation the action of en is mediated through hh.

1. Introduction

Genetic and molecular analyses in Drosophila have revealed that the intercellular signaling molecule produced by the segment polarity gene, wingless (wg), exerts an organizing influence on surrounding tissues (for reviews see Klingensmith and Nusse 1994; Siegfried and Perrimon 1994). With its complex signaling activity, wg generates diverse cell fates: for instance in the embryo it is required at multiple points in ectodermal segmental patterning and also for partitioning of cells into imaginal disc primordia (Baker 1988a; Simcox et al 1989; Cohen 1990; Bejsovec and Martinez-Arias 1991; Bejsovec and Wieschaus 1993; Cohen et al 1993; Ingham and Hidalgo 1993). In leg discs, it is required for dorso-ventral and antero-posterior axial patterning while in the wing disc it is required for distinguishing the notum/wing subfields, is responsible for compartmentalization of the dorsal and ventral wing surfaces and for bristle patterning along the margin (Morata and Lawrence 1977; Baker 1988b; Campbell et al 1993; Phillips and Whittle 1993; Williams et al 1993; Couso et al 1994). On the other hand, the engrailed (en) gene product is vital for defining and maintaining the posterior identity of cells and their proliferation throughout development (Kornberg et al 1985; DiNardo et al 1985; Hidalgo 1994). The en and hh genes are expressed in posterior compartment cells (Tabata and Kornberg 1994). en is positively required for hh expression and it negatively regulates patched (ptc) and decapentaplegic (dpp) in the posterior compartment (Tabata et al 1992; Sanicola et al 1995). The organizer property of en is mediated through the hh gene (Heemskerk and DiNardo 1994; Zecca et al 1995).

The single bilaterally symmetrical genital disc in the Drosophila larva differentiates into the internal accessory reproductive organs and external genitalia and analia of the adult. Each genital disc possesses separate primordia for male and female genital structures and a common primordium for anal structures (Bryant and Hsei 1977; Bryant 1978; Littlefield and Bryant 1979). Depending upon the sexual genotype, the anal primordium develops either the male or the female set of anal plates. However, among the two genital primordia for male and female accessory reproductive organs, only one, either the male or the female, develops and the other remains as a repressed primordium (Nothiger et al 1977; Epper 1980; Epper and Nothiger 1982). The internal accessory reproductive organs in the male include a pair of vas deferens, paired seminal vesicles, an ejaculatory duct and an ejaculatory bulb, while the external genitalia consist of the penis apparatus and periphallic apparatus. In the female, a pair of short lateral oviducts, a common oviduct, an uterus, a seminal receptacle, paired spermathecae and paired paraovaria comprise the internal accessory reproductive organs while the external genitalia consist of the egg guide.

Studies on the genital disc of Drosophila with particular emphasis on the segmental and compartmental organization have shown that each of the three primordia, the female and male genitalia plus the analia, are composed of an anterior and a posterior compartments. Moreover genes that are known to be expressed in a compartment-specific manner in other discs [en, hh, ptc, dpp, cubitus interuptus (ci) and wg] are expressed in analogous patterns in each primordium of the genital disc (Freeland and Kuhn 1996; Casares et al 1997; Chen and Baker 1997). Specifically, en and ci are expressed in complementary domains, while ptc, dpp and wg are expressed along the border between the two domains. The mitotic clones of en and ci expression domain did not cross these domains suggesting that these are true genetic compartments (Chen and Baker 1997; Sánchez et al 1997). Based on these it was proposed that functions of these genes are conserved in the genital disc. In an earlier study we have analysed the organizing roles of the genes en, hh, wg and dpp in order to understand their role in organizing the unique bilateral symmetry and shown that the organizing activity of en in the genital disc programs development and differentiation along with symmetry by regulating the expression of hh. hh in turn induces wg and dpp, genes whose products act as secondary signaling molecules. The complementary patterrn of wg and dpp expression, which is needed for bilateral symmetry, is maintained by mutual repression (Emerald and Roy 1998).

Based on the functional analyses of the homeotic and segmental genes Casares et al (1997) proposed that there is absence of lineage restrictions in some part of the terminalia and suggested that the compartmental organization evolves, at least in part, as the disc grows. If this is right then the expression pattern of those genes, which play a vital role in development and differentiation should change as development progresses. In order to further understand this functional dynamics and also to understand the temporal and spatial requirements of different genes in the development and differentiation of the genital disc we have analysed in detail the expression patterns of two well known patterning genes, wg and en, in the genital disc and accessory reproductive organs during larval, pupal and adult stages of Drosophila. Hetero-allelic mutant combinations were also used to analyse the functions of these genes in detail. Our results suggest that wg and en are expressed in a dynamic pattern throughout development and differentiation of the genital disc. In the absence of these genes the internal accessory reproductive organs are completely absent and the external accessory reproductive organs show specific defects. In addition, although the ovaries in females appeared normal the testes in such males were small, rounded and with an abnormal cellular organization. Temperature shift experiments using the conditional mutant allele of wg (wg^IL-114) indicated a requirement of wg signaling from second instar onwards for normal development and differentiation of the accessory reproductive organs. Using a heat-shock allele (Hs-wg) we show that the spatially regulated expression of wg as a pre-requisite for normal development and differentiation. Based on the fact that hh expression is similar to en and ectopic expression of hh and en also produced a similar type of pattern reorganization we suggest that in the genital disc the action of en is mediated through hh.

2. Materials and methods

2.1 Fly stocks

Following fly stocks were used (for details of mutations not specifically referred to in the following, see Lindsley and Zimm 1992):

(i) Wild type Oregon R strain.

(ii) wg^CX3/SM6a-TM6B (Baker 1987; Neumann and Cohen 1996). wg^CX3 is a regulatory mutation (see Emerald and Roy 1997) caused by a 16 kb insertion of unknown DNA downstream of the transcription end point. SM6a-TM6B is the second chromosome translocated balancer, which is linked with dominant markers Tb and Hu (Couso et al 1994).

(iii) wg^IL-114/SM6a-TM6B (Nusslein-Volhard et al 1984; Couso et al 1994). wg^IL-114 is a temperature sensitive mutant allele of wg.

(iv) wg^hs.PN/TM3, Sb (Noordermeer et al 1992) wg^hs.PN is a germline-transformed line in which wg coding sequence is fused to a heat shock promoter (referred to as Hs-wg hereafter). TM3 is a third chromosome balancer carrying the Sb marker.

(v) CyO en¹¹/In(2)LR Gla (Kassis et al 1992). CyO en¹¹ is a P-element mediated germline transgenic line in which the P-element construct is inserted at the wg locus so that the lacZ expression parallels wg expression (this has been referred to as wg^lacZ in the following). In(2)LR Gla is a second chromosome balancer carrying the Gla marker.

(vi) en¹ (Eker 1929; Kornberg 1981) is a viable mutant allele of en.

(vii) Df(2)en^A (Gubb 1985), Df(2)en^B and Df(2)en³⁰ (Eberlein and Russel 1983) chromosomes have deletions in the en region. These chromosomes are maintained against the CyO balancer.

(viii) en^Xho25/SM6a-TM6B (Hama et al 1990; Kassis et al 1992). en^Xho25 is an enhancer trap line with the P-element construct inserted at the en locus so that the lacZ expression mimics en expression (referred to as en^lacZ hereafter); this line is embryonic lethal.

(ix) hh^rJ413/TM6B (M P Scott, unpublished). hh^rJ413 is a P-element mediated germline transformed line in which lacZ expression parallels hh expression (referred to as hh^lacZ hereafter). TM6B is a third chromosomal balancer linked with the markers Tb and Hu.

Appropriate crosses were made between the different stocks to obtain progeny flies of the desired genotypes (see § 3). All flies (stocks and progeny of the various crosses) were reared in uncrowded condition on standard yeast supplemented food at 22 ±  1° C.

2.2 Mounting of external genitalia

For examining the external accessory reproductive organs, adult flies of the desired genotypes were boiled in 10% KOH, dehydrated through ethanol grades, cleared in clove oil and finally the penis apparatus, periphallic apparatus and vaginal plates were dissected out and mounted in Euparal.

2.3 b-galactosidase activity staining of genital discs and adult accessory reproductive organs

The expression patterns of wg or en in the genital discs from wg^lacZ/In(2)LR Gla, wg^CX3/wg^lacZ, en^lacZ/CyO and en³⁰/en^lacZ late third instar larvae or from wg^lacZ/In(2)LR Gla and en^lacZ/CyO pupae of different stages were monitored by b-galactosidase staining following Mukherjee et al (1995). The expression of hh in genital discs of hh^lacZ/TM6B late third instar larvae was also likewise monitored by b-galactosidase staining. The tissues were dissected in Poels’ salt solution (Lakhotia and Mukherjee 1980), fixed in 2·5% gluteraldehyde in 50 mM sodium phosphate buffer (pH 8·0) for 10 min, washed three times 5 min each in 50 mM sodium phosphate buffer (pH 8·0) and finally incubated in X-gal solution for 4 h to overnight at 37° C. After the staining they were washed in 50 mM sodium phosphate buffer, mounted in 80% glycerol in PBS and photographed. Expression of wg, en and hh in the accessory reproductive organs from adult flies of the above genotypes was also monitored by b-galactosidase staining as above.

2.4 In situ hybridization to RNA in reproductive tissues

A digoxigenin labelled wg cDNA probe (Rijsewijk et al 1987) was used to detect wg transcripts in the larval and adult testis and ovary. The procedure for whole organ in situ hybridization was essentially as described by Kramer and Zipursky (1992) with slight modifications (Mukherjee et al 1995). The tissues were dissected in Poels’ salt solution, fixed in 4% paraformaldehyde in PBS (pH 7·0–7·5) for 15 min on ice, refixed in 4% paraformaldehyde with 0·6% Triton X-100 in PBS for 15 min at 24° C, washed several times in PBT (PBS with 0·1% tween-20), treated with proteinase K (10 mg/ml in PBT) for 4 min at room temperature, re-washed in PBT with 2 mg/ml glycine, in PBT alone, re-fixed in 4% paraformaldehyde + 0·2% glutaraldehyde in PBS for 15 min at room temperature, washed again several times in PBT, incubated in 1 : 1 mixture of PBT : hybridization buffer (50% deionized formamide, 5 × SSC, 200 mg/ml yeast tRNA, 100 mg/ml sonicated, boiled salmon sperm DNA, 0·1% tween-20) for 10 min followed by incubation in hybridization buffer for 10 min at room temperature and finally hybridized with a heat-denatured dig-labelled small-sized DNA probe at the final concentration of 50 ng/100 ml in hybridization buffer for 36 h at 48° C. Following hybridization, the tissues were washed in hybridization buffer, 1 : 1 PBT-hybridization buffer at 48° C for 30 min each and then in PBT for 10 h at 48° C and for overnight at 4° C, incubated in 1 : 1000 dilution of preabsorbed anti-Dig antibody (Boehringer and Mannheim, Germany) for 1 h, washed in PBT again, incubated in staining buffer containing Levamisol and processed for colour reaction adding BCIP and NBT. After appropriate colour development, the reaction was stopped by washing the tissues in PBT and they were mounted on a clean slide in 80% glycerol.

2.5 Immunostaining for Wg and Hh

Distribution of Wg or Hh proteins in the adult reproductive system was studied by immunostaining using appropriate antibodies essentially according to the protocol of Pattatucci and Kaufman (1992). The adult male and female reproductive systems of wild type individuals were dissected, fixed in 4% paraformaldehyde for 20 min at room temperature, washed in PBS + 0.2% Triton X-100, blocked in PBS containing 3% BSA, 0·1% normal goat serum for 2 h and incubated overnight either with 1 : 300 dilution of anti-Wg (van den Heuvel et al 1989) (kindly provided by Dr R Nusse) or with 1 : 200 dilution of anti-Hh antisera (Capdevila et al 1994) (kindly provided by Dr I Guerrero). Following this, the tissues were repeatedly washed in PBT, blocked again and were finally incubated in 1/500 dilution of anti-rabbit HRP (for Wg) or anti-rat HRP (for Hh) (Sigma) for 2 h at room temperature. After several washes in PBS + 0·2% Triton X-100, the colour was developed in 0·5 mg/ml diamino benzidine (Sigma) in PBS containing 0·06% H₂O₂.

2.6 Temperature shift experiments

In order to identify the stage of development at which Wg is required for differentiation of accessory reproductive organs, eggs from a cross between flies of genotype wg^IL-114/SM6a-TM6B and wg^CX3/SM6a-TM6B flies were collected at 25° C and on different days the progeny embryos, larvae or pupae were shifted to 17° C (permissive temperature). The emerging wg^IL-114/wg^CX3 (non-tubby, non-curly) flies were examined for their accessory reproductive organs.

To study the effect of ubiquitous expression of wg on the development and differentiation of accessory reproductive organs eggs were collected from Hs-wg/TM6B flies or from a cross-between wg^lacZ/Sco; Hs-wg/TM3 Sb and wg^CX3/ Sm6a-TM6B flies. The eggs were grown at 22° C but were given 37° C heat shock pulses for 30 min everyday till the adults emerged. At emergence, the wg^CX3/wg^lacZ, Hs-wg and Hs-wg/TM6B flies were analysed for the state of differentiation of the accessory reproductive organs.

2.7 Histological studies on testis and ovary

For histological studies, the reproductive organs (testis and ovary) from flies of the desired genotype were dissected in Poels’ salt solution and fixed in Bouin’s fixative. Five-micrometer thick paraffin sections were stained with Haematoxyline and Eosin and mounted in DPX.

3. Results

3.1 wg expression in gonads and genital discs

Expression of the lacZ reporter gene in wg^lacZ/In(2)LR Gla individuals was monitored by b-galactosidase staining to determine the patterns of expression of wg in the genital discs or gonads from late third instar larvae and different stages pupae and in the reproductive system of adult flies (table 1). In situ hybridization using a wg probe in the larval and adult gonads was also used to analyze the pattern of expression of this gene. The wg^lacZ transgene used in the present study is known to mimic wg expression at all stages of development (Kassis et al 1992). In the male genital discs from third instar larvae

(figures 1A, C) or from pupae (data not shown), wg is expressed in the presumptive adult genitalia including the anlagen of penis apparatus, ejaculatory bulb, ejaculatory duct and hindgut. In the larval and adult testis wg expression was restricted to the apical cyst progenitor cells (figures 2A, 3A, 4A, C). In the adult male internal accessory reproductive system, wg expression is confined to the ejaculatory bulb and ejaculatory duct (figure 3A). The genital discs from female third instar larvae showed wg expression in the presumptive adult genitalia (figure 1B, D) including the anlagen of the oviduct, spermathecae, uterus, vagina, vulva and hind gut. The expression of wg in the female genital discs at various stages of evagination in pupae (data not shown) and in the reproductive system in adult (figure 3B) was also restricted to the oviduct, spermathecae, uterus, vagina and vulva. Furthermore, as in the larval testis, in the larval ovary the expression of wg was restricted to a small area of cells located at the apical tip (figures 2B, 4B), while the adult ovary showed b-galactosidase staining only in a few cells in the germarium (figure 3B). The in situ hybridization signal in a majority of the tissues mentioned above was similar to that of b-galactosidase staining. However, in case of adult ovary, in addition to a few cells in the germarium, the nurse cells from stage nine onwards up to the end of oogenesis also showed wg hybridization (figure 4D). These changes in the patterns of wg expression at particular stage and time provide evidence that wg plays a dynamic role in the development and differentiation of the gonadal structures.

3.2 Requirement of wg in the development of genital discs

As mentioned above if wg is specifically needed in particular structures at a specified time then in the absence of normal wg function, development of these structures should be affected specifically. In order to understand further the role of wg we analysed morphology of the genital discs in the wg^CX3/wg^lacZ heteroallelic combination generated by crossing wg^CX3/SM6a-TM6B and wg^lacZ/SM6a-TM6B flies. The SM6a-TM6B balancer chromosome permitted an unambiguous identification of wg^CX3/wg^lacZ individuals. All the 100 male and 100 female larvae of the wg^CX3/wg^lacZ genotype examined for genital discs showed drastically altered morphology of the disc. The absence of normal wg function in the male genital disc resulted in the male genital primordium being reduced substantially (figure 1E). Likewise, in the female genital disc, the female genital primordium is also reduced (figure 1F, also see A and B for genital primordia). These results show that wg plays important role in the development of genital discs as previously reported for other thoracic imaginal discs.

3.3 Requirement of wg in accessory reproductive system differentiation

Since in the absence of normal wg function genital discs did not show normal morphology, the differentiation of accessory reproductive organs was followed until the emergence of adult flies in the wg^CX3/wg^lacZ hetero-allelic combination. Flies of this genotype were found to be viable but sterile. Adult flies (500 male and 500 female) of the wg^CX3/wg^lacZ genotype were dissected and their internal reproductive systems were examined. The defects were uniform and all of them showed a complete absence of the internal accessory reproductive organs. Interestingly, while the ovaries in females appeared normal, the testes in males were small, rounded structures (figure 3C, D). In the absence of internal accessory reproductive organs, the abnormal testes appeared to lie closely apposed to the gut (figure 3C). In females the ovaries were free in the abdominal cavity (figure 3D). Interestingly, all the flies of this genotype also lacked antennae (data not presented).

In contrast to the external genitalia of the wild type male (figure 5A, B, D, E) or female (figure 5C, F), the external genitalia of wg^CX3/wg^lacZ individuals exhibited specific defects in both the sexes. The periphallic apparatus in the male showed fusion of the clasper and

lateral plate of one side with the clasper and lateral plate of the other side leaving no opening to the exterior (figure 5G). The penis apparatus was severely reduced and only a portion of hypandrium was present (figure 5H). Similarly, in females, the vaginal plates of both sides were fused leaving no opening (figure 5I).

3.4 Time of requirement of wg for normal development and differentiation of accessory reproductive system

Having shown that the normal expression of wg is required for the proper development and differentiation of the accessory reproductive organs in Drosophila, we have analysed at what stage of development this signal is critical for the normal development and differentiation. For this we made use of the temperature sensitive allele of wg, wg^1L-114. The temperature sensitive allele wg^IL-114 behaves as a null allele at 25° C (restrictive temperature) but exhibits a wild type phenotype at 17° C (permissive temperature) (Couso et al 1994). Flies of wg^IL-114/SM6a-TM6B and wg^CX3/SM6a-TM6B genotypes were crossed and the eggs were collected at restrictive temperature. They were then shifted to permissive temperature at different stages of development. wg^CX3 is a pupal lethal allele and wg^IL-114/wg^CX3 individuals reared at restrictive temperature die at the late pupal/pharate adult stage of development. Interestingly, all those that were shifted to permissive temperature (17° C) before the end of first instar larval stage (i.e., up to 48 h after egg laying), survived to adulthood and showed normal differentiation of both internal and external accessory reproductive organs (figure 5J–L). In contrast, individuals that were shifted to permissive temperature after the beginning of the second instar stage (i.e., after 36 h of hatching) died at the late pupal/pharate adult stage and showed defective accessory reproductive organs. A few of the individuals that were shifted to permissive temperature soon after 24 h of hatching survived to the adult stage, but showed defective accessory reproductive organs. These results thus revealed a requirement for the normal wg product at least from the end of first instar or beginning of early second instar stage onwards for the normal development and differentiation of accessory reproductive organs.

3.5 Effects of ubiquitous wg expression on the accessory reproductive organs

The importance of a local gradient of Wg protein in determining cell fate in the genital disc was tested in the Hs-wg/TM6B transgenic line. Such individuals have two normal copies of wg and one additional copy under the control of the hsp70 promoter. These individuals were given heat shocks for 30 min at 37° C daily for the entire period of development, i.e., from the embryo until the emergence of adult fly. It was expected that in addition to the normal temporal and spatial wg expression pattern, these individuals would have had ubiquitous expression of wg under the heat shock promoter. Out of the 73 adult males examined, 17 showed a defective reproductive system. The testis of one side in these cases appeared morphologically normal and associated with the accessory organs while the testis of the other side was bloated and was not associated with the accessory organs. The other 56 individuals had morphologically normal testes. However irrespective of the morphology of the testis, the size of the accessory gland and the ejaculatory duct was drastically reduced in all individuals (see table 2).

It has been shown in the earlier section of the results that defects caused by the absence of normal function of wg could be removed using the temperature sensitive allele of wg and ubiquitous expression of wg caused defects in the accessory reproductive organs. This raised the possibility that regulated expression of wg is needed for the normal development and differentiation of accessory reproductive organs. To further confirm the spatially regulated expression of wg as a prerequisite for normal development and differentiation of the genital discs and accessory reproductive organs, we analysed wg^CX3/wg^lacZ, Hs-wg/TM3 Sb flies for the development of the reproductive system after ubiquitous induction of wg by daily pulses of heat shock (see § 2). All 55 males and 64 females examined, showed a lack of internal accessory reproductive organs and defective external genitalia as seen in wg^CX3/wg^lacZ individuals suggesting that spatially regulated expression of wg is a prerequisite for normal development and differentiation of the genital discs and accessory reproductive organs.

3.6 Histology of gonads

Histological studies on testes revealed that the general wild type organization of testes, i.e., spermatogonial cells at the distal end and mature sperm at the proximal end with the three distinct zones, namely, the zone of growth, zone of maturation and zone of transformation (figure 6A, B), is lost in wg^CX3/wg^lacZ heterozygous flies (figure 6D, E). Masses of unidentified cells were also seen at various sites (figure 6D, E). The internal organization of polytrophic ovaries in wg^CX3/wg^lacZ (figure 6F), however, was comparable to that in wild type (figure 6C).

3.7 Expression of en in genital discs and in gonads

The pattern of expression of en during development of the reproductive system was studied in en^lacZ/CyO individuals by b-galactosidase staining (table 1). The en^lacZ enhancer trap line used in the present study is known to mimic en expression at all stages of development (Kassis et al 1992). As reported by Hama et al (1990) and Freeland and Kuhn (1996), en is expressed bilaterally in a large region in male genital primordia of third instar larva, rim of posterior lobes, except part of the presumptive anal plates and hind gut (figure 1G). However, en staining is not seen in the larval testis (figure 2C). As the genital disc evaginated in the antero-posterior axis during pupal stage of development, en is expressed in the anlagen for the adult genitalia (data not shown). At late pupal stages of development when the male accessory reproductive system is formed, the staining is restricted to only half the length of the ejaculatory duct and in the entire ejaculatory bulb (figure 7A). Since in the ejaculatory duct en is expressed in only half the length, we assume this region corresponds to the posterior compartment. The pattern of expression of en in the larval or pupal stages of the female genital disc is seen in the genital primordia (figure 1H, data not shown for pupal stages) while no expression of en is seen in the larval ovary (figure 2D). In the adult female reproductive system, the terminal filament and cap cells of the ovarioles, the posterior half of the oviduct and uterus exhibited staining (figure 7B). The pattern of expression of en in adult male and female accessory reproductive organs indicated a gradual decrease in the proportion of en expressing cells with the progression of differentiation.

3.8 Requirement of en in the development of genital discs

To examine the role of en in the development of genital discs, individuals with hetero-allelic combinations of en [Df(2)en³⁰/en^lacZ or Df(2)en^A/Df(2)en^B] and individuals homozygous for the viable en¹ allele (en¹/en¹) were examined. In agreement with earlier report by Epper and Sanchez (1983), the different hetero-allelic combinations and homozygosity for the viable allele affected the male and female genital discs differentially. Larvae of
the desired hetero-allelic combinations were identified without ambiguity because of the tubby phenotype of SM6a-TM6B balancer-carrying genotypes (see § 2). In the Df(2)en³⁰/en^lacZ hetero-allelic combination, which was analysed in detail, most of the 100 male genital discs analysed male genital primordia were affected drastically (figure 1I). In the female genital discs the female genital primordia was also affected, but to a lesser extent
(figure 1J, also see A and B for genital primordia). Df(2)en^A/Df(2)en^B hetero-allelic combination also showed the same results.

3.9 Requirement of en in accessory reproductive system differentiation

Since the genital discs were defective in absence of normal en function, the state of differentiation of accessory reproductive organs in adult flies was also examined in Df(2)en³⁰/en^lacZ and Df(2)en^A/Df(2)en^B hetero-allelic combinations, and in the viable en¹/en¹ homozygous genotype. In en¹/en¹ individuals, the internal accessory reproductive organs were normal while the external accessory reproductive organs showed minor defects in the claspers and in the penis apparatus. On the other hand, all the Df(2)en30/en^lacZ (200 flies) or Df(2)en^A/Df(2)en^B (25 flies) showed defective internal accessory organs with differential expressivity. In extreme cases, the external morphology of the gonads of both the sexes were similar to that seen in wg^CX3/wg^lacZ individuals (small and rounded testes lying apposed to the digestive tract in males and apparently normal ovaries free in the abdominal cavity in females) (figure 7C, D).

In contrast to what is seen in the absence of normal function of wg in these flies the claspers and lateral plates of the same side in the periphallic apparatus were fused in all cases (figure 5M). The stout black bristles, characteristic of claspers, were absent. The penis apparatus in the male was represented by a rudimentary penis, penis mandle and portion of hypandrial process (figure 5N). In females, the vaginal plates did not exhibit any fusion, instead the spine bristles which stand in a row all along the vaginal plate were duplicated along the axis
(figure 5O).

3.10 Expression of hh in the genital discs and in gonads

In embryonic epidermal cells hh is known to be regulated by en and to be expressed in posterior compartments precisely coinciding with the expression of en (Tabata et al 1992). To determine if the pattern of hh expression is comparable to that of en expression in genital discs and gonads from larvae or adult reproductive systems, these organs were dissected from hh^lacZ/TM6B individuals and stained for b-galactosidase activity using the chromogenic substrate X-gal or by immunostaining in wild type adult reproductive system using the Hh antiserum. The hh^lacZ/TM6B transgenic line always showed a weak expression of b-galactosidase, hence b-galactosidase staining remained lighter in comparison to the wg^lacZ or en^lacZ transgenic lines. The expression of hh was noted in the presumptive adult genital primordia in the genital discs from larvae of both the sexes in a manner comparable to the expression of en in these tissues (figure 1K, L; table 1 and also see Freeland and Kuhn 1996). In larval (figure 2E) and adult (figure 8A) testes, staining was seen in somatic hub cells (Forbes et al 1996). The ejaculatory bulb also showed hh expression (figure 8A). In larval ovary, staining was seen in a band of cells almost in the middle region of the ovary (figure 2F), while in the adult female reproductive system, hh expressed in the specific somatic cells, terminal filament and cap cells of the ovarioles, in the seminal receptacle and in uterus (figure 8B).

We have also tested the effect of the heteroallelic combination, hh^9K/hh^lacZ, for differentiation of accessory reproductive system. As reported above for heteroallelic combination of en mutants, all hh^9K/hh^lacZ male flies showed defects in the external genitalia viz., fusion of claspers and lateral plates of same side while the females showed less pronounced effect (Emerald and Roy 1998).

4. Discussion

4.1 wg and en functions are required for normal development and differentiation of genital discs

Three different bands of wg and en expression were observed corresponding to the three primordia of both male and female discs, and by comparing this with other discs it was suggested that there are three anterior and three posterior compartments in third instar male and female genital disc (Chen and Baker 1997; Casares et al 1997). The present study was aimed to examine the roles of wg and en in the development and differentiation of genital discs in Drosophila. wg and en were found to be expressed in a regulated fashion in genital discs and in reproductive tissues throughout development. Experiments with the temperature sensitive allele (wg^IL–114) showed that wg expression was required at least from second instar stage onwards (24 h after hatching) for normal development and differentiation of reproductive organs. In wing development also, wg acts at second instar stage as the first gene in the cascade of four genes (wg-apterous-scalloped-vestigial) that control the dorso-ventral (D-V) pattern in wing (Diaz-Benjumea and Cohen 1993; Williams et al 1993). wg is known to have an important role in cell proliferation (Skear and Martinez Arias 1992; Phillips and Whittle 1993; Neumann and Cohen 1996) and to affect proliferation rate positively or negatively in a context dependent manner (Klingensmith and Nusse 1994). Since it is known that cell proliferation in all tissues other than the nervous system ceases by 12 h of embryonic development, and that genital discs reinitiate proliferation around 24–26 h after hatching (Madhavan and Schneiderman 1977), it is likely that the second stage of proliferation of genital discs is regulated by wg. Moreover, it has been suggested that a mutual repression by wg and dpp signaling systems in Drosophila genital discs generates a stable regulatory circuit by which each gene maintains its own expression in a spatially restricted domain (Emerald and Roy 1998). This may further explain the failure to recover the defects in accessory reproductive organs by ubiquitous expression of wg in the hetero-allelic combination of wg, wg^CX3/wg^lacZ while the temperature sensitive allele which expresses the protein as same as the expression of the endogenous gene when monitored with temperature was able to recover the defects.

Previous studies have shown that the female genitalia, male genitalia and analia were affected by en mutations (Epper and Sánchez 1983). It is also shown that the expressivity of different alleles of en was different in different combinations (Epper and Sánchez 1983). Further clones of en lethal mutations also showed alteration in the male genitalia and analia, but no clear transformations of posterior to anterior compartment (Lawrence and Struhl 1982). However the transformations obtained in the wing when both en and inv genes are absent are quite different from those observed when en alone is removed (Hidalgo 1994, Tabata et al 1995). Casares et al (1997) made clones deficient for en and inv and showed that in male genitalia claspers, penis apparatus and genital arch were affected. In contrast to the clonal analysis where only duplication of clasper teeth have been seen it is interesting to note the hetero-allelic combinations of en show the absence of clasper teeth and fusion of the lateral lobe and posterior lobe. However, in female they reported that the vaginal plate and T8 bristle were affected which is in agreement with our observations in the hetero-allelic combinations. This is further supported by the expression studies in the adult using en^lacZ reporter construct (Hama et al 1990) where posterior region of the genital arch and claspers were positive for en expression in males while T8 and vagina were shown to be positive in the females. Thus our studies while confirming other studies also finely map the regions (clasper teeth, lateral lobe and posterior lobe, vaginal plate and T8) which are regulated by en. Further it is interesting to note that the hetero-allelic combination, Df(2)en³⁰/en^lacZ, showed complete lack of internal accessory reproductive system, which is similar to that of wg^CX3/wg^lacZ individuals though these genes are shown to play organizing activities in two different axes in other discs. This indicates that signaling by wg and en for proper development of both the axes is important and that if there is alteration in any one the defect is same and it results in complete absence of development. This combination also showed specific defects in the external accessory reproductive organs.

4.2 wg and en play roles in dorso-ventral and antero-posterior axes, respectively

As the imaginal discs differentiate, the two-dimensional discs having dorso-ventral and antero-posterior axes, evert to form a three dimensional adult structure with additional proximo-distal axis. The function of wg in axial patterning of the limbs has been studied in detail and loss-of-function mutation analyses have shown that wg is required for the specification of ventral structures (Baker 1988a, b). In contrast, en has been proposed to be of paramount importance in the specification of posterior compartment identity (Morata and Lawrence 1975). Based on the pattern of expression of genes such as wg and dpp, the genital disc resembles that of the leg disc (Freeland and Kuhn 1996; Casares et al 1997). This gets further support from the fact that mutations in the trithorax (trx), absent, small or homeotic (1) (ash1) and absent, small or homeotic (2) (ash2) genes transform genitalia into legs (Ingham 1985; Shearn et al 1987). In a recent study we had shown that the unique bilateral symmetry of the genital disc is because of the complementary pattern of expression of wg and dpp and this expression is maintained by mutual repression. This is again similar to the leg (Emerald and Roy 1998). Our observation on external genitalia in wg^CX3/wg^lacZ individuals that the claspers and lateral plates of both discs were fused in the dorso-ventral axis, while in en³⁰/en^lacZ flies, the clasper and lateral plate were fused together on each side (fusion in antero-posterior axis), supports further that in these discs also, the wg acts in dorso-ventral axis while the en establishes antero-posterior axis. Therefore it is interesting that in spite of the fact that these two genes specify two different axes in post-embryonic development like that of leg, the altered function of these genes had same consequences in the development of internal reproductive organs. It is likely that the proper third axis formation i.e., evertion of the disc and differentiation of internal accessory organs, is possible only when both the primary axes are normal.

4.3 Development and differentiation of the genital discs require continuous activity of wg and en

A stepwise morphogenetic program results in the development of genital discs and differentiation of accessory reproductive organs (Epper 1983). The disc opens along the posterior margin and the dorsal and ventral epithelia evert and thereby reverse the anterior-posterior orientation (Epper 1983). A continuous requirement for wg activity throughout larval stages in patterning the ventral face of the leg has been demonstrated by Couso et al (1993). We have also observed the expression of wg or en in small subsets of cells and the expression of wg and en when they are abutting each other’s expression. Eventhough in the absence of any of the two genes the same defects were elicited. Further ectopic expression of en induces ectopic expression of hh which in turn induce the secondary signaling molecules dpp and wg and this is responsible for the normal development and differentiation of the genital discs (Emerald and Roy 1998). It is likely that interactions mediated by these genes play key roles in normal development and differentiation of genital disc. The fact that en and inv are required in the posterior compartment in the genital disc to repress dpp and activate hh (Chen and Baker 1997), further confirms that the interactions mediated by these genes are indeed responsible for the normal development and differentiation of the genital discs.

4.4 en action is mediated through hh

It has been shown earlier that in the wing the organizer property of en is mediated by hh gene during development (Zecca et al 1995). It has also been demonstrated that the expression of hh corresponds with the posterior compartment as defined by disc fate mapping and lineage analysis (Schubiger 1971; Bryant 1975; Steiner 1976) and by expression of en (Kornberg et al 1985; Hama et al 1990; Lee et al 1992). In the genital disc also hh has an expression pattern similar to that of en in both female and male genital discs and is complementary to that of Ci, which is expressed in the anterior compartment (Chen and Baker 1997). This is very similar to the other discs. Mohler (1988), using temperature shift analysis of a temperature sensitive allele of hh, demonstrated that there are two activity periods for hh, one is between 2·5 and 6 h of embryonic development and the other, a larval/pupal sensitive period from 4–7 days of development (post embryonic sensitive period) and the lack of hh expression in the post embryonic period resulted in defects in the genitalia. Clonal analysis also showed that the genitalia in the two sexes were affected differentially (Mohler 1988) which is similar to that seen in the en mutants in our study. Our demonstration that hh expression in the genital disc corresponds with the expression pattern of en, and the lack of internal accessory reproductive organs and defects in the external accessory reproductive organs in en mutants can be recovered by ectopic expression of hh gene product under inducible promoter (data not shown), suggests that at later stages of development also, hh mediates en action.

In brief, the present study has demonstrated that wg and en functions are required from early stages of larval development for normal differentiation of accessory reproductive organs and male gonad. Eventhough the consequences of abnormal function of any of these genes are same, they act in different axes i.e., wg acts in dorso-ventral axis, while en acts in the anterior-posterior. Further, as it is shown in embryos or in other discs, in genital disc also the role of en is mediated through hh.

Acknowledgements

We thank Drs S Chandrasekharan, S M Cohen, J P Couso, S DiNardo, L I Jr Held, D Kalderon, P A Lawrence, A Martinez-Arias and the Bloomington Stock Centre for fly stocks. We are grateful to Drs R Nusse, I Guerrero and F M Hoffmann for anti-Wg and anti-Hh antibodies and for the wg cDNA clone, respectively. We are very grateful to Prof. S C Lakhotia and Dr R Raman for critical suggestions. This work was supported by research fellowship from the Council of Scientific and Industrial Research, New Delhi, to BSE and by the Department of Science and Technology, New Delhi to JKR.

MS received 26 March 1998; accepted 13 May 1999