The abdominal-A (abd-A) gene, a member of the bithorax complex, is required for the correct identity of parasegments (PS) 7 through 13. Mutations in iab-4, one of the crs-regulatory regions of abd-A, transform epidermal structures of PS 9 and also cause loss of gonads in adult flies. Here, we describe a developmental and molecular analysis of the role of iab-4 functions in gonadal development. In flies homozygous for a strong iab-4 allele, gonadogenesis is not initiated in the embryo because the mesodermal cells fail to encapsulate the pole cells. Flies homozygous for weaker iab-4 mutations sometimes form ovaries. The ovary-oviduct junctions are abnormal, however, and egg transfer from the ovary to the uterus is blocked in the adult. To localize the sites that require iab-4 function, we have analyzed animals chimeric for the mutant and wild-type cells. These chimeras were generated by three kinds of transplantation experiments: pole cells, embryonic somatic nuclei or larval ovaries. Our results suggest that iab-4 is required in the somatic cells of the gonadal primordia, but not the germ line. In addition, the formation of functional ovary-oviduct junctions and egg transfer also requires iab-4 functions in the somatic cells of the ovary and in at least one additional somatic tissue.

The bithorax complex (BX-C) of Drosophila melanogaster contains three homeotic genes -Ultrabithorax, abdominal-A and Abdominal-B -that specify the identity of two thoracic and eight abdominal segments. abdominal-A (abd-A) is required for determining the correct identity of parasegments (PS) 7 through 13, corresponding to the abdominal segment (A)l posterior compartment through A8 anterior compartment (for review see Peifer et al., 1987 and Duncan, 1987). Mutations in abd-A not only cause homeotic transformations of epidermal structures, but also affect other tissues such as muscle, fat body, gonads and the CNS (Lewis, 1985; Karch et al., 1985; Busturia et al., 1989). It has been proposed that abd-A functions as a molecular switch, selecting different developmental pathways by regulating the expression of different sets of downstream or ‘realizator’ genes (Lewis, 1964; Garcia-Bellido, 1975). In support of this model, recent molecular analyses have shown that the abd-A encodes a homeodomain protein (Karch et al., 1990; Macias et al., 1990) which can function as a transcription factor both in tissue culture cells and in the fly (B. Appel and S. Sakonju, unpublished results). Yet the molecular events between the initial activation of abd-A and the development of its target tissues and organs remain uncharted. One way of approaching this problem is to focus on a particular morphogenetic pathway, identify the steps in the pathway and then determine which of these steps are disrupted by mutations within abd-A. We have chosen to examine the role of abd-A during the formation and subsequent development of the gonadal primordia.

Mutations in the abd-A domain fall into two classes: the abd-A alleles and the iab alleles. The abd-A – alleles disrupt the homeobox-containing transcription unit and are homozygous embryonic lethal (Busturia et al., 1989; Karch et al., 1990; B. Appel, M. Lamka and S. Sakonju, in preparation). They cause a strong transformation of PS7-9, and a weaker transformation of PS10-12, towards PS6. The iab –alleles disrupt the cisregulatory regions of abd-A and cause transformations in subsets of PS7-9 through altered expression of the abd-A gene (Busturia et al., 1989; Karch et al., 1990; Simon et al., 1990). In this work, we have focused on mutations that affect the iab-4 cis-regulatory region located about 20 kb upstream from the start site of abdA transcription unit. Flies carrying deletions or chromosomal breaks that separate the iab-4 region from the abd-A transcription unit are fully viable. In these flies, the fourth abdominal segment (A4) is transformed towards the third (Lewis 1978; 1985; Karch et al., 1985). In addition, Lewis and co-workers (Lewis 1978; 1985; Karch et al., 1985) have also noted that iab-4 mutations cause sterility through a loss of adult gonads.

Here we describe a detailed analysis of the role of iab-4 functions in gonadal development. The male and female reproductive systems have been studied extensively in Drosophila (for reviews see Lindsley and Tokuyasu, 1980; Mahowald and Kambysellis, 1980; Babcock 1971a, b). The two gonadal primordia, each composed of a cluster of 8 –10 germ cell precursors surrounded by a layer of mesodermal cells, are first formed during mid-embryogenesis in the ventral mesoderm A5 region (Poulson, 1950; Underwood et al., 1980; Hay et al., 1988; Lasko and Ashbumer, 1990). The somatic cells of the gonadal anlagen originate in the primordia at approximately the fourth abdominal segment (Szabad and Nöthiger, 1992); however, the segmental boundaries of the mesoderm, relative to the ectoderm, are shifted posteriorly by a half segment. Thus, by the time of gonadogenesis, these cells are located at the level of A5 (Tremml and Bienz, 1989).

In females, the presumptive ovaries continue to grow with little differentiation until the pupal stage. Then the mesodermally derived cells differentiate into a variety of cell types, including the epithelial cells of the ovarioles, the peritoneal sheath, the basal stalk cells and the posterior calx (King et al., 1968; King, 1970). Concurrently, the genital discs, derived from ectodermal cells of A8-A10, are also developing, forming the uterus, the main oviduct and two lateral oviducts, (Poulson, 1950; Babcock, 1971a). After pupariation as the lateral oviducts grow, they meet and fuse with ovaries, thus completing the internal duct system. Transfer of mature oocytes from ovaries through oviducts to the uterus in adult females requires the coordinated functions of several organs, including the nervous system, muscular network of the ovarian peritoneal sheath and contractile fibrillae in the epithelial sheath of the ovarioles (Bodenstein, 1950).

Development of testes proceeds at a faster rate (Cooper, 1950). By the second instar, the testes are much larger and more differentiated than their female counterparts. By the early stages of pupal development, the seminal vesicles (derived from the male genital disc) connect with the testes.

In this paper, we first identified the morphological processes during gonadogenesis that are affected by lesions in the iab-4 region of abd-A. By constructing chimeras, we then investigated which of the tissues that constitute the reproductive organs are affected by iab-4 mutations. Our results suggest that iab-4 functions are required in the somatic cells, but not in the germ cells, of the gonads. For the establishment of correct ovaryoviduct connections and egg transfer in females, iab-4 functions are required in the gonadal somatic cells as well as at an additional site(s) outside of the ovaries.

Antibody staining

For staining with anti-vasa antibody, which specifically labels pole cells (Hay et al., 1988), embryos were collected from an iab-4302TM3, ftz-lacZ stock. This TM3 balancer chromosome has an insertion of lacZ fused to the ftz promoter, allowing the identification of genotypes in embryos (constructed and kindly provided by Sarah Smolik-Utlaut). Homozygous embryos were identified, when double labeling with antibodies directed against vasa and β-galactosidase, as those not showing the ftz-lacZ pattern. Immunodetection methods were as described in Boulet et al. (1991). Briefly, embryos were dechorinated in a hypochlorite solution, washed and fixed in 4% paraformaldehyde in PBS/heptane (1:1) for 20 minutes. The aqueous layer was replaced with methanol and the embryos were shaken until the vitelline membrane was removed, and then washed several times in methanol and rehydrated in PBS. After washing several times in PBS-BT (0.1% BSA, 0.1% Triton X-100 in PBS), embryos were incubated overnight in primary antibody diluted in PBS-BT, and then washed multiple times in PBS-BT. Biotinylated secondary antibody binding (Vectastain ABC Kit;.Vectorlabs) and subsequent HRP immunochemistry was carried out as per manufacturer’s directions. After staining, the embryos were dehydrated in ethanol and mounted in methyl salicylate. Anti-vasa antibody was used at a 1:5 dilution, and anti-β-gal antibody (Cappel) at 1:2,000. The anti-vasa antibody was generously provided by Y. N. Jan.

The monoclonal antibody, DMabd-A.2, directed against abd-A protein was prepared and kindly provided by Diane Mattson and Ian Duncan (Kellerman et al., 1990). For the wild-type expression pattern, Canton-S embryos were collected and labeled with the antibody at a dilution of 1:2,000. Histochemical staining procedures were the same as described above for the anti-vasa antibody. To observe the abd-A expression pattern in an iab-4 mutant, embryos were collected from iab-r302/TM3, ftz-lacZ stock and the mutant embryos identified as described above.

Germline chimeras

Pole cell transplantations were performed as described by Lehmann and Nüsslein-Volhard (1987). Donor embryos (Table 1) were collected from the following cross: iab-4302/TM6, Tb X iab-4302TMS, Sb. Host female embryos were obtained by crossing ovoD1/Y males to +/+ (Ore-R) females. Fj ovoD/+ females do not deposit eggs due to germline defects (Busson et al., 1983; Komitopoulou et al., 1983). Host male embryos were obtained from crossing X AY y su(w°) males to +/+ (Ore-R) females. The F 1XO males are sterile due to germline defects (Marsh and Wieschaus, 1978). Both donor and host eggs were collected for one hour and allowed to age at 25°C for 1.5 hours. Embryos were then dechorinated in a sodium hypochlorite solution, washed, desiccated and covered with 10S Voltalef oil. The pole cells from the donor embryos were collected and transplanted into the posterior pole of the host embryos. Transplantation procedures were performed at 18°C. Surrogate embryos were allowed to develop at 25°C. After hatching, the genotype of the transplanted germline cells was determined by crossing the host flies to wild-type flies, and scoring the phenotype of the F 2 progeny. Genetic markers and standard chromosomes are as described in Lindsley and Grell (1968).

Table 1.

Pole cell transplantations (A) Female germ-line chimeras

Pole cell transplantations (A) Female germ-line chimeras
Pole cell transplantations (A) Female germ-line chimeras

Larval ovary transplantations

Ovary transplantations were carried out according to the procedures of Clancy and Beadle (1937). Larvae were raised on standard cornmeal yeast media. Wandering third instars were collected from the walls of the bottles. Host and donor larvae were washed in 70% EtOH, etherized and dissected in sterile PBS. Genotypes of the donor and host larvae are listed in Table 4. After implantation of the ovaries, chimeric animals were raised at 25°C and screened for fertility. Analogous experiments with male third instar larvae were not performed since, by third instar, the testes are much larger than ovaries, making transplantations unfeasible.

Nuclei transplantations

Nuclei transplantations were performed as described by Santamaria (1986). In the first experiment, donor embryos were derived from a cross between +/+ (Ore-R) females and Fs(3)Apc, mwh e/TM3, Sb Ser males. Fs(3)Apc is a dominant female sterile mutation that alters follicle cell functions leading to a specific mutant phenotype (Szabad and Hoffmann, 1989; Erdelyi and Szabad, 1989). In the second experiment, donor embryos were obtained from flies homozygous for two copies of a hsp70promoter-lacZ transgene (Bg 9.6.1; Lis et al., 1983). Host embryos were derived from iab-4302TM3, Sb stock. Nuclei were removed from donor embryos, 1-2 hours after egg laying, and then injected into 0-1 hour host embryos at 38% egg length, at the site of the presumptive A4 segment. After hatching, adults were screened for fertility by mating to Canton-S flies. Each experimental animal was heat-shocked at 37°C for 1 hour, allowed to recover for 30 minutes at 25°C, and then dissected. The abdominal cavity and reproductive organs were removed, fixed in 4% paraformaldehyde in PBS for 10 minutes, washed in PBS, stained for /3-galactosidase activity as described in Lis et al. (1983), and scored using a compound microscope.

Gonadal defects caused by iab-4 mutant alleles

Based on the penetrance of sterility, the iab-4 alleles can be placed into the phenotypic series: iab-4302, iab-4166> iab-44,5Db> iab-445 (Lewis, 1985). For example, >99% of males or females homozygous or hemizygous for a strong iab-4~ allele, iab-43, are sterile, while 83% of the iab-445 hemizygous males and 92% of the iab-445hemizygous females are affected. We began our analysis by comparing the reproductive organs dissected from wild-type adult flies with those from iab-4302/iab-4302 and iab-445/iab-4302 adults (Fig. 1). 84% of iab-4302/iab-4302 females and 91% of males lack gonads (Fig. 1B, E). The associated sexual organs, all derivatives of the genital discs, appear to be normal, with the exception of the lateral oviducts in females. While the lateral oviducts are always present, they often appear stunted (Fig. 1B) compared to those of the wild-type female (Fig. 1A). The male seminal vesicles, however, appear to be fully formed (Fig. 1E) as in the wild-type male (Fig. 1D). Dissection of iab-4302 and iab-4-166 hemizygotes gave similar results (data not shown).

Fig. 1.

Morphology of adult reproductive organs from wild-type and iab-4302 homozygous flies. Adult females (left) and males (right) were dissected and the reproductive tissues were stained with methylene blue and photographed. (A) Wildtype (Canton S) female. Black arrowhead points to an egg in the lateral oviduct, and white arrowhead to an egg in the uterus. (B) iab-4302homozygous female, neither of the two ovaries are present. (C) iab-445/ iab-4302female, only one ovary is formed. Arrowheads point to eggs lodged in the main oviduct and at the ovary-oviduct junction. (D) Wild-type (Canton S) male. (E) iab-4302homozygous male; the testes are not formed. (C) iab-445/ iab-4302male, only one testis is present, o, ovaries; d, main oviduct; I, lateral oviduct; u, uterus; t, testes; v, seminal vesicles; a, accessory glands; e, ejaculatory duct.

Fig. 1.

Morphology of adult reproductive organs from wild-type and iab-4302 homozygous flies. Adult females (left) and males (right) were dissected and the reproductive tissues were stained with methylene blue and photographed. (A) Wildtype (Canton S) female. Black arrowhead points to an egg in the lateral oviduct, and white arrowhead to an egg in the uterus. (B) iab-4302homozygous female, neither of the two ovaries are present. (C) iab-445/ iab-4302female, only one ovary is formed. Arrowheads point to eggs lodged in the main oviduct and at the ovary-oviduct junction. (D) Wild-type (Canton S) male. (E) iab-4302homozygous male; the testes are not formed. (C) iab-445/ iab-4302male, only one testis is present, o, ovaries; d, main oviduct; I, lateral oviduct; u, uterus; t, testes; v, seminal vesicles; a, accessory glands; e, ejaculatory duct.

Females carrying one copy of a weaker iab-4 allele usually form one or two ovaries. Their attachment to the lateral oviduct, however, is often defective (Fig. 1C). For example, about a half the expected number of ovaries were present in the twelve iab-445/iab-4302 females dissected; of those present, two thirds were not attached to the oviduct. The free-floating ovaries were often facing anterior, rather than the normal posterior direction. Both attached and free-floating ovaries were composed of 14-18 ovarioles (normal ovaries usually contain 15-20 ovarioles) and fully mature oocytes. Often the mutant ovaries contained more mature oocytes than wild-type ovaries (compare Fig. 1A and C). In two of the three cases where the ovaries were attached, partially degenerated eggs were observed in the lateral oviduct and the oviduct-ovary junctions were abnormal. No eggs were ever observed in the uterus, and all twelve females were sterile. It appears that the sterility is caused, at least in part, by the disruption of normal egg migration, perhaps due to an abnormal junction between the gonad and the genital duct or to defective neuronal or muscular functions. All other associated sexual organs derived from the genital discs appear to be normal.

Male iab-445/iab-4302 flies displayed an analogous phenotype: often only one testis was present (Fig. IF). Approximately 42% of iab-445/iab-4302 males dissected contained at least one pigmented, mature testis with sperm; yet only 22% were fertile. We suspect that sperm transfer has been affected. However, unlike previously reported cases of defective sperm transmission, the seminal vesicles in these males were not distended (Lindsley and Tokuyasu, 1980).

These results suggest that lesions in the iab-4 region of abd-A affect at least two steps in gonadal development. First, the formation and/or development of male and female gonads is disrupted. Second, attachment of the presumptive ovaries to the lateral oviducts is altered. Migration of germ cells from the gonads through the reproductive ducts may also be inhibited, although we have observed this only for females. These phenotypes could be the manifestation of a single primary defect, or several independent malfunctions.

Gonadal defects in the embryo

To determine when, and in which tissues, iab-4+is required for gonad formation and differentiation, we have examined the consequences of loss of iab-4 function at several stages in development. We began by monitoring gonadogenesis in both wild-type and iab-4embryos. To follow pole cell migration and subsequent formation of the gonadal primordia, embryos were collected at various times, fixed and then stained with a monoclonal antibody that recognizes the pole-cell-specific vasa antigen (Fig. 2). In wild-type embryos, the pole cells move dorsally into the posterior midgut primordium during gastrulation (stage 6 of Campos-Ortega and Hartenstein, 1985), pass through the posterior midgut walls and, by the germ-band-shortening stage, align along the wall of the body cavity adjacent to the primordia of abdominal segments 5 through 8 (Fig. 2A). By stage 14, the mesodermal cells encapsulate a cluster of pole cells thereby forming a rudimentary gonad in the ventral A5 region (Fig. 2B; Poulson, 1950; Underwood et al., 1980; Hay et al., 1988). Pole cells not enclosed within the gonadal primordia gradually die. By stage 16, the mesodermal sheath of the gonad consists of approximately 27-37 cells (Sonnenblick, 1941).

Fig. 2.

Migration of pole cells in wild-type and iab-4302 homozygous embryos. Optical section of wild-type stage 12 (A), wild-type stage 13 (B) and iab-4302homozygous stage 14 (C) embryos stained with anti-vasa antibody. Anterior is to the left. Arrowheads in C show pole cells that are not encapsulated in the mutant.

Fig. 2.

Migration of pole cells in wild-type and iab-4302 homozygous embryos. Optical section of wild-type stage 12 (A), wild-type stage 13 (B) and iab-4302homozygous stage 14 (C) embryos stained with anti-vasa antibody. Anterior is to the left. Arrowheads in C show pole cells that are not encapsulated in the mutant.

In iab-4302 embryos pole cell migration is unimpeded until late stage 13/early stage 14. The pole cells move into the ventral mesoderm along the body cavity, but the somatic cells do not form a sheath around them and subsequent development of the gonadal primordia is arrested (Fig. 2C). The germ cells remain dispersed throughout the mesoderm, and do not form a cluster in the A5 region; instead, they appear to die gradually. Our results indicate that iab-4302function is not necessary for germ cell migration to the ventral mesoderm, but is required for the formation of the mesodermal gonadal sheath.

abd-A expression in gonadal mesoderm

It is not known if iab-4 functions autonomously in the mesoderm; therefore it is of particular interest to ascertain if abd-A is expressed in the presumptive gonadal soma and if the abnormal behavior of the gonadal anlage in iab-4~_embryos correlates with a change in expression of abd-A in these cells. The pattern of abd-A protein expression in the developing embryo has been described previously (Macias et al., 1990; Karch et al., 1990). abd-A protein is found in a complex pattern in parasegments 7 through 13 in the nuclei of the epidermis, the nervous system, the mesoderm and the amnioserosa. Both the above groups have reported that the pattern of epidermal expression in iab-4 embryos is similar to wild type. Karch et al. (1990) have also found that the intersegmental neurons and the pericardial cells lack detectable expression in the mutant.

Since these studies were not focused on expression specifically during gonadogenesis, we have used a monoclonal antibody directed against abd-A protein to compare abd-A expression in wild-type and iab-4_embryos in the mesodermal gonadal anlagen just before and during gonad formation. In late stage 13 wild-type embryos, low levels of abd-A antigen are detected in the mesoderm in parasegments 8-12, the region where the pole cells are found just prior to gonad formation. By stage 15, the gonads are clearly visible as a dorsolateral cluster of cells at A5. Anti-abd-A antibody weakly stains the somatic cells that surround the distinctly large pole cells (Fig. 3). Intersegmental neurons stain strongly with anti-abd-A antibody at this and later stages (Fig. 3), but not at stages prior to the formation of the gonads. In iab-4?02 and iab-4?5DB embryos, the spatial distribution of the abd-A protein in PS7-13 mesoderm was indistinguishable from the wild type (data not shown). We have also found that the intersegmental neurons no longer express abd-A protein, confirming the earlier observation (Karch et al., 1990). Thus the mutant phenotype of the mesodermal cells cannot be readily correlated with a concomitant qualitative change in abd-A expression. It is possible, however, that we failed to detect subtle spatial or temporal changes.

Fig. 3.

Distribution of abd-A protein in a stage 13 Canton S embryo. The embryo was stained using monoclonal antibody, DMabd-A.2, as described in Materials and methods. The embryo has been dissected and flattened to reveal interior tissues and is shown anterior to the top. abd-A protein is detected in the somatic cells surrounding the gonad (go), the intersegmental neurons of A4-A7 (arrowheads), in the ventral cord (vc) and in epidermal cells overlaying the gonad and intersegmental neurons. The abd-A expression in the gonadal soma is distinct but weak and is not always clear in photos taken from a single plane of focus.

Fig. 3.

Distribution of abd-A protein in a stage 13 Canton S embryo. The embryo was stained using monoclonal antibody, DMabd-A.2, as described in Materials and methods. The embryo has been dissected and flattened to reveal interior tissues and is shown anterior to the top. abd-A protein is detected in the somatic cells surrounding the gonad (go), the intersegmental neurons of A4-A7 (arrowheads), in the ventral cord (vc) and in epidermal cells overlaying the gonad and intersegmental neurons. The abd-A expression in the gonadal soma is distinct but weak and is not always clear in photos taken from a single plane of focus.

Somatic versus germ-line requirements for iab-4 +function

The inability of iab-4302/iab-4302 embryos to form gonads could reflect a requirement for iab-4 function either in the somatic tissue, or the germ cell precursors, or both. Using the technique of pole cell transplantation, we have constructed and analyzed germ-line chimeras to examine if iab-4 function is needed in the germ line (Table 1). Pole cells from iab-4302/iab-402 embryos were transplanted into surrogate ovoD1/+ female or X/O male host embryos, which cannot make functional gametes of their own. When female iab-4?02 homozygous pole cells are transplanted into host female embryos (Table 1A), or male iab-4302homozygous pole cells are transplanted into host male embryos (Table IB), the chimeric embryos develop normally and mature into fertile adults. Therefore, normal development of ovaries and testes does not require iab-4+function in the germ cells.

Rescuing the iab-4 mutant phenotype by transplantation of somatic nuclei

Since iab-4 function is not required in the germ cells, it must be required in somatic cells for gonad formation. To determine which somatic cells require iab-4 function, we constructed and analyzed somatic chimeric animals. iab-4+ nuclei were transplanted into host iab-402embryos along the ventral midfine, in the presumptive mesoderm region, at 38% egg length, where cells of the fourth abdominal segment form (Campos-Ortega and Hartenstein, 1985). Donor cells were marked using two independent methods. In the first experiment, 50% of the donor embryos carried the follicle cells-specific marker mutation Ape (Szabad and Hoffmann, 1989; see Materials and methods). A total of 8 iab-4302homozygous females were recovered (Table 2). Four of these females did not have ovaries. One female developed an ovary and another female had two ovaries, with all the eggs and egg primordia showing the Ape mutant phenotype. These Ape egg-containing ovaries were attached to the oviduct, although eggs were deposited from only one of them. We take these females to be examples of rescue of the gonadless iab-4 302phenotype. The other two host females carrying non-Apc eggs most likely originated through leakiness of the iab-402 mutant phenotype since their ovaries were not attached to the oviduct (Table 2).

Table 2.

Rescue of the gonadless phenotype in iab-4 302homozygous female embryos by nuclei transplantations

Rescue of the gonadless phenotype in iab-4 302homozygous female embryos by nuclei transplantations
Rescue of the gonadless phenotype in iab-4 302homozygous female embryos by nuclei transplantations

In the second experiment, the iab-4+donor nuclei were marked with a hsp70 promoter-lacZ transgene. The chimeras were first tested for fertility, then they were dissected and their internal abdominal organs (including the gonads when present) were stained for β-galactosidase (β-gal) activity. Among 65 control hosts (iab-4302/TM3, Sb females), nine (14%) contained ovaries that stained for β-gal. It should be noted, however, that only varying fractions of the egg primordia stained blue in these control females. 22 iab-4302homozygous females were recovered (Table 2). Five of the 22 females developed one or two ovaries. In four of these females, the somatic cells of at least one of the ovaries were labeled with β-gal (Table 2). In these females, unlike the control females, the entire ovaries, including all the ovarioles, stained blue. All the bluestaining ovaries were also attached to the oviducts. Three iab-4302homozygous females had one ovary that did not stain with β-gal activity. These ovaries were not attached to the oviduct, and may have resulted from the leakiness of the iab-4 allele since approximately 13% of iab-4302homozygous females form unattached ovaries. The correlation between the appearance of ovaries that are connected to oviducts and the presence of iab-4+ cells in these gonads suggests that iab-4+function is required in these cells.

All four β-gal-labelled ovaries were attached to the oviduct and one female was able to deposit eggs. In the others, eggs were never transferred to the uterus. This observation suggests that iab-4+function is required not only for the formation of the gonadal soma but also for egg transfer at a site outside of the ovary. In the one egg-laying female, the ovary and oviducts stained positive for β-gal activity. The oviduct staining may have been a consequence of follicle cell debris, which often accumulates in the genital ducts.

Nuclei transplantation experiment was also carried out in males (Table 3). In this experiment, the donor iab-4+cells were labeled with hsp-lacZ transgene. After transplantation, testes developed in five (28%) of the 18 host iab-4302homozygous males, ß-gal-labelled testicular sheet cells were detected in one testis of each of two males, one of whom gave rise to offspring.

Table 3.

Rescue of the gonadless phenotype following transplantation of iab-4 + nuclei into iab-4 302 homozygous embryos

Rescue of the gonadless phenotype following transplantation of iab-4 + nuclei into iab-4 302 homozygous embryos
Rescue of the gonadless phenotype following transplantation of iab-4 + nuclei into iab-4 302 homozygous embryos

In the two cases where complete rescue was obtained, the somatic tissue of the gonads also contained iab-4+cells, strongly supporting the model whereby iab-4 function is necessary in the mesoderm. Yet, this may not be sufficient to obtain complete rescue of sterility. Three chimeric females and one chimeric male produced gonads that stain with β-gal, but they were not functional. These results, in combination with our analysis of adult phenotypes, suggests that formation of competent gonad-duct connections may require iab-4 function in additional somatic cells.

Formation of functional ovary-oviduct junctions

Because the genital disc and the gonads develop independently and establish connection only after pupariation, it is possible to transplant gonads of one genotype into larval hosts of a different genotype. We constructed ovarian chimeras to test which tissues require iab-4+function in order to form competent ovary-oviduct junctions. First, we implanted iab-4+arval ovaries into iab-4302or iab-443homozygous larvae lacking ovaries (Table 4). The chimeric animals were allowed to develop into adults, the females were tested for fertility, and then dissected. As shown in Table 4, iab-4+ ovaries grew and developed in iab-4302or iab-443 hosts. They appeared to undergo normal differentiation and attached to the lateral oviducts, yet all of the chimeric females were sterile. Stage 14 oocytes could be seen in the lateral oviducts, but never in the uterus. This sterility is not caused by the failure of the females to mate since sperm was found in the sperm receptacle (Table 4). Therefore, these results suggest that iab-4+function is required for egg transfer and/or oviposition.

Table 4.

Larval ovary transplantations

Larval ovary transplantations
Larval ovary transplantations

The transplantation operation itself does not interfere with the formation of the lateral oviducts, nor does it impair egg movement: when control iab-4+ovaries were transplanted into ovoD1/+ larvae (ovo D+ females do not form functional ovaries of their own), eight of nine chimeric larvae grew into fertile adults (Table 4). In only one case did the ovary fail to attach to the lateral oviduct. These results indicate that at least one aspect of ovarian activity -normal egg transfer from the lateral oviducts to the uterus -requires iab-4+ function in the somatic cells outside of the ovaries.

By performing the reciprocal experiment, we also tested whether formation of ovary-oviduct junctions requires iab-4+function in the somatic tissues of the ovaries. Since iab-443is a leaky allele, some iab-443 homozygous larvae contain gonads. When these iab-443ovaries are transplanted to an iab-4+host, they appear to develop normally (Table 4), but the chimeric ovaryoviduct junction is deformed. In a few cases, egg deposition appears to be normal, although in most cases the eggs never migrate to the uterus (Table 4). In these chimeric animals, the structures derived from the genital discs, as well as the surrounding muscle and nervous tissue, are iab-4+only the ovaries are deficient in iab-4 function. Together these results indicate that development of functional ovaries requires iab-4+ function in the somatic cells of the ovary and at a second site in the animal.

We have undertaken a developmental analysis of iab-4 mutants in order to identify the primary lesion(s) from which the gonadal defects arise. By analyzing the embryonic and adult gonadal phenotypes of iab-4>+females, we have found that iab-4+function is required for three steps: (1) encapsulation of pole cells by nearby mesodermal cells during embryogenesis, (2) formation of functional oviduct-ovary junctions during pupal development and (3) normal egg transfer through the oviduct in the adult. Disruption of egg transfer may be a consequence of abnormal ovary-oviduct junctions or may represent an independent defect. Male iab-4flies have the same embryonic defect as females and may have analogous pupal and adult defects; many iab-443 males form testes, yet are infertile, suggesting that sperm transfer is faulty.

We have shown by analyzing germline chimeras that iab-4 funtion is not needed in pole cells to form gonads, suggesting that iab-4 functions strictly in somatic cells. Our nuclei and larval ovary transplantation experiments confirm that the formation of ovaries requires iab-4+activity in the somatic cells of the gonads. In addition, to develop functional ovaries iab-4+activity is required at a second site. This site is apparently outside of the ovarian cells since wild-type ovaries cannot develop into functional adult ovaries when transplanted into iab-4_larvae. Previous gynandromorph analyses (Szabad and Fajszi, 1982) have identified three foci in the Drosophila female nervous system that control egg transfer, from the ovaries through the oviduct to the uterus, and egg deposition. These foci map to more posterior blastoderm regions than the adult brain. Normal egg laying requires communication between the ovaries, the oviducts and the uterus, and is governed by various hormonal and neuronal cues. A most likely site of the abd-A requirement outside of the ovaries is in the neuronal cells. These neuronal cells may develop from the neuroectodermal region at the level of A4 of the blastoderm located between the mesoderm and ectoderm. In this regard, it is interesting to note that one of the few disturbances of abd-A expression patterns in iab-4_ embryos, observed by Karch et al. (1990) and confirmed in our study, is the loss of expression in the intersegmental neurons. We did not detect abd-A expression in these neurons prior to the formation of gonads. Therefore, abd-A in the intersegmental neurons may contribute to the proper functioning of the gonads in later stages but perhaps not for the initial formation of the gonads. A second, less likely, possibility is that abd-A is also required for genital disc development. With the availability of antibodies directed against abd-A protein, it should now be possible to determine the expression patterns in the pupal and adult neurons, as well as the genital structures.

There is a long-standing question as to whether determination of the insect mesoderm is autonomous or if it is governed by the ectoderm (see Lawrence and Johnston, 1984, for a discussion). For example, it has been suggested that some segmental muscle patterns in Drosophila are determined by Abd-B activity in the associated neurons and not in the myoblasts themselves (Lawrence and Johnston, 1986). This led us to ask if abd-A expression in the mesodermal gonad anlage is required for formation of the gonadal primordia. Two fines of evidence suggest that it is acting autonomously. First, the gonadal primordia are formed before the nervous system is fully developed (stage 14). At this time, the mesodermal cells that become incorporated into the gonads are already expressing abd-A protein. Second, analyses of chimeric animals generated by nuclei and larval ovary transplantation experiments suggest that iab-4+function is required in the ovarian mesoderm. In the nuclei transplantation experiment in which the donor cells were labeled with the hsp70 promoter-lacZ construct, the entire follicle cell population stained for ß-galactosidase activity. Similarly, with the donor cells marked with an Ape mutation, all eggs showed the Ape phenotype. Had donor cells of the neuroectoderm origin induced mesodermal cells to initiate gonadogenesis, patches of the host (iab-4_) cells would have been expected in the ovaries. The absence of the host cells in the rescued ovaries, therefore, argues that iab-4+function is required autonomously in the mesodermal cells.

If iab-4+is acting autonomously in the mesodermal gonad anlage, why do we not detect a corresponding loss of abd-A expression in mutant embryos? Most likely explanation is that we did not detect subtle spatial or temporal differences in the pattern of expression. The iab-4 deletions affect cis-regulatory regions and are partial loss-of-function alleles which are not fully penetrant. Stronger mutations decrease the frequency with which gonads are formed, but not the amount of somatic tissue formed. In the rare iab-4302 escapers where an ovary is present, it contains the correct number of ovarioles, whereas other mutations (e.g. some Oscar_ alleles) are known to reduce the number of ovarioles formed (Szabad and Nöthiger, 1992). These observations are consistent with the idea that a threshold level of abd-A activity triggers a decision in the mesoderm to form an ovary in an ‘all or nothing’ manner. Perhaps small changes in the level of abd-A expression cause the cells to drop below the threshold.

A variety of lacZ enhancer trap lines have been obtained recently whose β-gal expression patterns are restricted to specific parts of the reproductive system (Fasano and Kerridge, 1988). It may be possible to identify genes whose spatial and temporal expression patterns correlate with the tissues affected by the iab-4 mutations. Such genes would be good candidates for the realizator genes acting downstream in the pathway. It is not known if abd-A regulates the same downstream targets in, for example, the mesodermal gonad anlage, the pupal ovaries and the associated neurons. Having the same ‘selector’ gene act at multiple steps and in different tissues may help to coordinate the development of complex organ systems.

We thank Ed Lewis for providing many of the stocks used in this work, John Lis for the hsp70-lacZ line, Y. N. Jan for antivasa antibody and Diane Mattson and Ian Duncan for anti-abd-A antibody. We also thank Anne Boulet, Brad Johnson and Rolf Nöthiger for helpful comments on the manuscript.

Babcock
,
M. B.
(
1971a
).
Oviduct development in Drosophila. I. Effects of Ovariectomy and genital disc transplantations
.
Wilhelm Roux’s Archiv. EntwMech. Org
.
167
,
1
23
.
Babcock
,
M. B.
(
1971b
).
Oviduct development in Drosophila. U. metamorphic events in normal and ovariectomized females
.
Wilhelm Roux’s Archiv. EntwMech. Org
.
167
,
1
23
.
Bodenstein
,
D.
(
1950
).
The postembryonic development of Drosophila
.
In The Biology of Drosophila
(ed.
M.
Demerec
), pp
275
367
.
New York
:
Wiley
.
Boulet
,
A. M.
,
Lloyd
,
A.
and
Sakonju
,
S.
(
1991
).
Molecular definition of the morphogenetic and regulatory functions and the cisregulatory elements of the Drosophila Abd-B homeotic gene
.
Development
111
,
393
405
.
Bussen
,
D.
,
Gans
,
M.
,
Komitopoulu
,
K.
and
Masson
,
M.
(
1983
).
Genetic analysis of three dominant female sterile mutations located on the X chromosome of Drosophila melanogaster
.
Genetics
105
,
309
325
.
Busturia
,
A.
,
Casanova
,
J.
,
Sanchez-Herrero
,
E.
,
Gonzalez
,
R.
and
Morata
,
G.
(
1989
).
Genetic structure of the abd-A gene of Drosophila
.
Development
107
,
575
583
.
Campos-Ortega
,
J. A.
and
Hartenstein
,
V.
(
1985
).
The Embryonic Development of Drosophila melanogaster
.
Berlin
:
Springer-Verlag
.
Clancy
,
C. W.
and
Beadle
,
G. W.
(
1937
).
Ovary transplants in Drosophila melanogaster
.
Biol. Bull. Mar. Biol. Labs. Woods Hole
72
,
47
56
.
Cooper
,
K. W.
(
1950
).
Normal spermatogenesis in Drosophila
.
In The Biology of Drosophila
(ed.
M.
Demerec
), pp
1
61
.
New York
:
Wiley
.
Duncan
,
I.
(
1987
).
The bithorax complex
,
Ann. Rev. Genet
.
21
,
285319
.
Erdelyi
,
M.
and
Szabad
,
J.
(
1989
).
Isolation and characterization of dominant female sterile mutations of Drosophila melanogaster. I. Mutations on the third chromosome
.
Genetics
122
,
111
127
.
Fasano
,
L. N.
and
Kerridge
,
S.
(
1988
).
Monitering positional information during oogenesis in adult Drosophila
.
Development
104
,
245
253
.
Garcia-Bellido
,
A.
(
1975
).
Genetic control of wing disc development
.
In Cell Patterning. Ciba Foundation Symposium
, pp.
161
182
.
Amsterdam
:
Elsevier
.
Hay
,
B.
,
Ackerman
,
L.
,
Barbel
,
S.
,
Jan
,
L. Y.
and
Jan
,
Y. N.
(
1988
).
Identification of a component of Drosophila polar granules
.
Development
103
,
625
640
.
Karch
,
F.
,
Bender
,
W.
and
Weiffenbach
,
B.
(
1990
).
abd-A expression in Drosophila embryos
.
Genes and Dev
.
4
,
1573
1587
.
Karch
,
F.
,
Weiffenbach
,
B.
,
Peifer
,
M.
,
Bender
,
W.
,
Duncan
,
I.
,
Celniker
,
S.
,
Crosby
,
M.
and
Lewis
,
E. B.
(
1985
).
The abdominal region of the bithorax complex
.
Cell
43
,
81
96
.
Kellerman
,
K. A.
,
Mattson
,
D. M.
and
Duncan
,
I.
(
1990
).
Mutations affecting the stability of the fushi tarazu protein of Drosophila
.
Genes and Dev
.
4
,
1936
1950
.
King
,
R. C.
(
1970
).
Ovarian Development in Drosophila melanogaster
.
New York
:
Academic Press
.
King
,
R. C.
,
Aggarwal
,
S. K.
and
Aggarwal
,
V.
(
1968
).
The development of the female Drosophila reproductive system
.
J. Morphol
.
124
,
143
166
.
Komitopoulou
,
K.
,
Gans
,
M.
,
Margaritis
,
L. M.
,
Kafatos
,
F. C.
and
Masson
,
M.
(
1983
).
Isolation and characterization of sex-linked female-sterile mutants in Drosophila melanogaster with special attention to egg shell mutants
.
Genetics
105
,
897
920
.
Lasko
,
P.
and
Ashburner
,
M.
(
1990
).
Posterior localization of vasa protein correlates with, but not sufficient for, pole cell development
.
Genes and Dev
.
4
,
905
921
.
Lawrence
,
P.
and
Johnson
,
P.
(
1984
).
The genetic specification of pattern in a Drosophila muscle
.
Cell
36
,
775
782
.
Lawrence
,
P.
and
Johnson
,
P.
(
1986
).
The muscle pattern of a segment of Drosophila may be determined by neurons and not by contributing myoblasts
.
Cell
45
,
505
513
.
Lehmann
,
R.
and
Nüsslein-Volhard
,
C.
(
1987
).
hunchback, a gene required for segmentation of an anterior and posterior region of the Drosophila embryo
.
Dev. Biol
.
119
, 402-407–152.
Lewis
,
E. B.
(
1964
).
Genetic control and regulation of developmental pathways
.
In The Role of Chromosomes in Development
, (ed.
M.
Locke
), pp
231
253
.
New York
:
Academic Press
.
Lewis
,
E. B.
(
1978
).
A gene complex controlling segmentation in Drosophila
.
Nature
276
,
565
570
.
Lewis
,
E. B.
(
1985
).
Regulation of the genes of the bithorax complex in Drosophila
.
Cold Spring Harbor Symp. Quant. Biol
.
50
,
155
164
.
Lindsley
,
D. L.
and
Grell
,
E. H.
(
1968
).
Genetic Variations of Drosophila melanogaster
.
Carnegie Inst. Wash. Pubi. 627
.
Lindsley
,
D. L.
and
Tokuyasu
,
K. T.
(
1980
).
Spermatogenesis
.
In The Genetics and Biology of Drosophila, vol. 2d (eds. M. Ashbumer and T. R. F. Wright)
, pp
225
294
.
London
:
Academic Press
.
Lis
,
J. T.
,
Simon
,
J.
and
Sutton
,
C.
(
1983
).
New heat shock puffs and β-galactosidase activity resulting from transformation of Drosophila with a hsp70-lacZ hybrid gene
.
Cell
35
,
403
410
.
Macias
,
A.
,
Casanova
,
J.
and
Morata
,
G.
(
1990
).
Expression and regulation of the abd-A gene of Drosophila
.
Development
11
,
11971207
.
Mabowald
,
A. P.
and
Kambysellls
,
M. P.
(
1980
).
Oogenesis
.
In The Genetics and Biology of Drosophila
, vol.
2d
(eds.
M.
Ashbumer
and
T. R. F.
Wright
), pp
141
224
.
London
:
Academic Press
.
Marsh
,
J. L.
and
Wieschaus
,
E.
(
1978
).
Is sex determination in the germ line and soma controlled by separate genetic mechanisms?
Nature
272
,
249
251
.
Peifer
,
M.
,
Karch
,
F.
and
Bender
,
W.
(
1987
).
The bithorax complex: control of segmental identity
.
Genes and Dev
.
1
,
892
98
.
Poulson
,
D. F.
(
1950
).
Histogenesis, organogenesis, and differentiation in the embryo of Drosophila melanogaster
.
In The Biology of Drosophila
(ed.
M.
Demerec
), pp
168
274
.
New York
:
Wiley
.
Santamaria
,
P.
(
1986
).
Injecting eggs
,
In Drosphila, a Practical Approach
. (
D.B.
Roberts
ed.)
IRL Press
,
Washington D.C
. pp
159
173
.
Simon
,
J.
,
Peifer
,
M.
,
Bender
,
W.
and
O’Connor
,
M.
(
1990
).
Regulatory elements of the bithorax complex that control expression along the antenor-postenor axis
.
EMBO J
.
9
,
39453956
.
Sonnenblick
,
B. P.
(
1941
).
Germ cell movements and sex differentiation of the gonads in the Drosophila embryo
.
Proc. Natl. Acad. Sci., USA
.
27
,
484
489
.
Szabad
,
J.
and
Fajszi
,
C.
(
1982
).
Control of female reproduction in Drosophila’, genetic dissection using gynandromorphs
.
Genetics
100
,
61
78
.
Szabad
,
J.
and
Hoffmann
,
G.
(
1989
).
Analysis of follicle-cell functions in Drosophila’, the Fs(3)Apc mutation and the development of chorionic appendages
.
Dev. Biol
.
131
,
1
10
.
Szabad
,
J.
and
Nöthiger
,
R.
(
1992
).
Gynandromorphs of Drosophila suggest one common primordium for the somatic cells of the male and female gonads in the region of the abdominal segments 4 and 5
.
Development
115
,
527
533
.
Tremml
,
G.
and
Bienz
,
M.
(
1989
).
Homeotic gene expression in the visceral mesoderm of Drosophila embryos
.
EMBO J
.
8
,
2677
2685
.
Underwood
,
E. M.
,
Caulton
,
J. H.
,
Allis
,
C. D.
and
Mahowald
,
A. P.
(
1980
).
Developmental fate of pole cells in Drosophila melanogaster
.
Dev. Biol
.
77
,
303
314
.