Maternal expression of the l(l)pole hole (l(l)ph) gene product is required for the development of the Drosophila embryo. When maternal l(l)ph+ activity is absent, alterations in the embryonic fate map occur as visualized by the expression of segmentation genes fushitarazu and engrailed. If both maternal and zygotic activity is absent, embryos degenerate around 7h of development. If only maternal activity is missing, embryos complete embryogenesis and show deletions of both anterior and posterior structures. Anteriorly, structures originating from labral and acron head regions are missing. Posteriorly, abdominal segments A8, 9 and 10, the telson and the proctodeum are missing. Similar pattern deletions are observed in embryos derived from the terminal class of female sterile mutations. Thus, the maternal 1(1)ph+ gene product is required for the establishment of cell identities at the anterior and posterior poles of the Drosophila embryo.

Pattern formation in the Drosophila embryo requires the regulation of maternal and zygotic gene activities to establish and maintain cellular identities within progressively smaller developmental fields. Initially, the interpretation of global embryonic polarity along the anterior-posterior and dorsal-ventral egg axes permits the establishment of specialized domains of zygotic gene expression (see reviews by Konrad et al. 1985; Mahowald & Hardy, 1985; Akam, 1987; Anderson, 1987; Nüsslein-Volhard et al. 1987; Perrimon & Mahowald, 1988; Ingham, 1988). It has been proposed that anterior-posterior polarity is determined by the generation of two gradients, each originating from opposite egg poles (Lehmann & Nüsslein-Volhard, 1986; Frohnhofer & Nüsslein-Volhard, 1986). The maternal bicoid gene product is responsible for the morphogenetic gradient originating at the anterior egg pole (Driever & Nüsslein-Volhard, 1988u,6). Maternal nanos* activity may serve as the posterior located morphogen (Nüsslein-Volhard et al. 1987; R. Lehmann, personal communication).

Independent of these gradients, activity encoded by the terminal class of maternal effect genes is required for the establishment of cell fates at the embryonic poles (Nüsslein-Volhard et al. 1982; Degelmann et al. 1986; Schupbach & Wieschaus, 1986; Nüsslein-Volhard et al. 1987). Thus, in embryos lacking maternal terminal group function, both head and tail structures are absent. A similar phenotype is observed when embryos lack zygotic tailless* expression (Stecker et al. 1988). For head differentiation, maternal bicoid* function is also required. Embryos developing without bicoid* product have an anterior duplication of posterior pattern. Therefore, establishment of terminal cell fates requires maternal terminal group and zygotic tailless gene expression. The distinction between head and tail determinative pathways requires bicoid* function. Recent genetic experiments suggest that the maternal terminal genes act through the zygotic tailless gene (Klinger et al. 1988).

The terminal group of female sterile mutations contains five members including fs(l)Nasrat (Degelmann et al. 1986), fs(l)pole hole1 (Perrimon et al. 1986), fs(2)trunk, fs(2)torso (Schupbach & Wieschaus, 1986), and fs(3)torso-like (Nüsslein-Volhard et al. 1987). All members of this class when defective in the female germline produce embryos showing deletions of anterior and posterior terminal cell fates (Mlodzik et al. 1987; Winslow et al. 1988). Presumably, these five maternal gene products function in a common pathway essential for the establishment of terminal cell identity. Another genetic locus, l(l)pole hole (l(l)ph), shows an equivalent maternal effect on embryonic development (Perrimon et al. 1985). Thus, it seems likely that the 1(1)ph gene product is also required for the expression of embryonic terminal cell fates. However, since l(l)ph+ activity is also necessary at other stages of development, the diverse nature of its function has obscured its maternal role.

The 1(1 )ph gene was first identified as a zygotic lethal mutation causing death at the larval-pupal transition stage in l(l)ph hemi- or homozygous progeny derived from heterozygous females (Perrimon et al. 1985). Profiferation of imaginal cells fails to occur in these third instar larvae resulting in small, undifferentiated imaginal discs. Thus, zygotic synthesis of l(l)ph+ product is essential for imaginal disc development, but it is not required for the growth of nondividing polyploid larval cells. Since all imaginal tissues are equally affected by l(l)ph mutations, it seems likely that expression of this gene plays a fundamental role in dividing larval cells.

The characterization of 1(1 )ph mutations using germline clonal analysis reveals an additional requirement for maternal l(l)ph expression during normal embryogenesis. Embryos derived from homozygous germline clones that have received a wild-type copy of the gene from their father phenotypically resemble embryos produced by the terminal class of female sterile mutations (Nüsslein-Volhard et al. 1982). They lack structures derived from abdominal segments 8, 9,10, and the telson. These embryos will be referred to as the ‘pole hole’ embryos throughout the text. Those embryos derived from homozygous germline clones that have not received a wild-type copy of the gene from their father have a poorly differentiated exoskeleton. Throughout the text we will refer to these as ‘null’ embryos.

In this analysis, the maternal and embryonic requirements for l(l)ph+ expression are described. We have analyzed the progression of development and the establishment of cellular identities in embryos derived from homozygous l(l)ph germline clones. Comparison of the maternal effect of l(l)ph with that produced by the terminal class of female sterile genes indicate that l(l)pole hole is a member of the terminal gene class.

Strains

In this analysis, we used amorphic mutations at the l(l)ph locus including l(l)phAls, Ipllph.0222 and l/llph1129. These alleles of 1(1 )ph give similar phenotypes and throughout the text the alleles will not be noted. The chromosomes carrying l(l)ph mutations were maintained as FM7c stocks. The two X-linked maternal effect loci fs(l)NasraPn and fs(l)pole hole1901 are maintained in stocks balanced with FM3 chromosomes. The autosomal maternal effect mutations: fs(2)trunkR153, cn bw sp/Cyo, fs(2)trunkHH28, cn bw sp/Cyo, fs(2)torsoWK3f cn bw sp/Cyo, fs(2)torsoHH, cn bw sp/Cyo and fs(3)torso-like0356/TM3, Sb were obtained from T. Schupbach.

The X-linked dominant female sterile mutation Fs(l)K1237 (Busson et al. 1983; Perrimon, 1984) is maintained as an attached-X stock: C(1)DX, y f/Y females crossed to Fs(l)K1237, v24/T males. The engrailed/ lacZ (en/lacZl) strain was obtained from C. Hama and T. Kornberg; the fushi-tarazu/lacZ (ftz/lacZ) strain from Y. Hiromi and W. Gehring (1987); and the even-skipped/lacZ (eve/lacZ) strain from P. MacDonald (Lawrence et al. 1987).

Descriptions of balancer chromosomes and mutations can be found in Lindsley & Grell (1968). All experiments were performed at 25°C on standard Drosophila medium.

Clonal analysis

Germline clones of l(l)ph mutations were produced by using the dominant female sterile technique (Perrimon, 1984; Perrimon et al. 1984). Briefly, virgin females heterozygous for l(l)ph/FM7c were mated to Fs(l)K1237 v/Y males. At the end of the first larval instar stage, progeny were irradiated at a constant dose of 1000 rads (Torrex 120D X-ray machine; 100kV, 5 mA, 3 mm aluminum filter). Mitotic recombination in the germline of l(l)ph/Fs(l)K1237 females was detected by individual inspection of ovary development. The frequency of females carrying germline clones homozygous for l(l)ph was about 5 %.

Embryo morphology

Cuticle preparations were prepared in Hoyers mountant as described by van der Meer (1977). Histological sections of embryos were prepared as described by Mahowald et al. (1979). Embryos were dehydrated and embedded in JB4 plastic (Polysciences). Serial 3 gm sections were cut using a Leitz 1516 microtome and stained with methylene blue. Slides were dried and mounted in Aquamount. Embryos were prepared for scanning electron microscopy as described by Turner & Mahowald (1976).

Observation of living embryos

Embryos were collected and observed through development as described by Wieschaus & Nüsslein-Volhard (1986).

Introduction of segmentation fusion genes

Females possessing homozygous germline clones for l(l)ph were crossed with males that carry either the ftz/lacZ, en/lacZ or eve/lacZ insertion. Similarly, in the case of the maternal effect lethal mutations, these transformants were introduced in mutant embryos by crossing flies homozygous for the maternal effect with males carrying the lacZ insertion. These promoter-fusion constructs show similar patterns of figalactosidase expression in embryos as observed by antibody staining of the native protein for ftz (Hiromi & Gehring, 1987), eve (Frasch et al. 1987; Lawrence et al. 1987) and en (DiNardo et al. 1985; Klingensmith et al. 1989; C. Hama & T. Kornberg, personal communication).

Immunohistochemistry

Hoechst 33258 staining of embryos was performed as described by Wieschaus & Nüsslein-Volhard (1986). Immunohistochemistry was performed as described in Smouse et al. (1988) for tubulin using an antibody obtained from T. Karr; for segmentation genes using a mouse anti-/3-galactosidase primary antibody from Promega-Biotec to detect /3-galactosidase from ftz/lacZ, en/lacZ, and eve/lacZ genes; and for the engrailed protein using an antibody obtained from S. DiNardo and P. O’Farrell. To examine the central and peripheral nervous systems of mutant embryos, we used polyclonal antibody against horseradish peroxidase (anti-HRP), which labels all central and peripheral nervous system cell bodies and axons (Jan & Jan, 1982), and the SOX2 monoclonal antibody, which recognizes the cell bodies and axons of the entire PNS and subset of CNS neurons (Goodman et al. 1984). The anti-HRP antiserum was from Cappel.

Early mitosis and cellular blastoderm formation is normal in both null and pole hole embryos

Nuclear cleavages and migration of nuclei to the egg cortex are normal in embryos derived from homozygous l(l)ph germline clones. As shown in Fig. 1, using Hoechst dye to stain nuclei (Fig. 1A) and a tubulin antibody to stain microtubules (Fig. 1B), embryos lacking maternal l(l)ph+ gene product develop to the blastoderm stage with no morphological defects. Cell formation at the periphery is normal in these embryos except at the posterior pole where a hole in the blastoderm cell layer below the pole cells is observed. In a previous study (Perrimon et al. 1985), we indicated that null embryos failed to complete blastoderm cellularization. For these embryos, the distribution of nuclei around the egg cortex was also abnormal. Following the analysis of considerably more embryos produced by germline clones, wre have found that the poorly cellularized blastoderm phenotype previously described actually depends upon the age of the mother rather than the genotype of the embryo. Among the embryos that have a normal morphology at the blastoderm stage, 50% develop into pole hole embryos and 50% into nulls. This is true for embryos derived from young mothers in which only 5 % of these eggs develop incomplete blastoderms, or embryos from old mothers in which a substantially greater number have early abnormalities. Thus the abnormal cellularization defect is related to the maternal rather than the zygotic genome.

Fig. 1.

Early mitosis and localization of nuclei in embryos derived from homozygous l(l)ph germline clones. (A) The distribution of blastoderm nuclei around the embryonic cortex, and the internal localization of yolk nuclei or vitellophages are normal in embryos derived from l(l)ph germline clones. At the posterior pole, the pole cells are visible. (B) A mitotic wave progressing normally from the posterior pole toward the middle of (see arrow) a l(l)ph germline clone embryo. Abbreviations: (bn) blastoderm nuclei; (pc) pole cells; (vg) vitellophages.

Fig. 1.

Early mitosis and localization of nuclei in embryos derived from homozygous l(l)ph germline clones. (A) The distribution of blastoderm nuclei around the embryonic cortex, and the internal localization of yolk nuclei or vitellophages are normal in embryos derived from l(l)ph germline clones. At the posterior pole, the pole cells are visible. (B) A mitotic wave progressing normally from the posterior pole toward the middle of (see arrow) a l(l)ph germline clone embryo. Abbreviations: (bn) blastoderm nuclei; (pc) pole cells; (vg) vitellophages.

The early pattern of gastrulation and morphogenetic movements

Gastrulation in wild-type embryos begins with the invagination of the midventral cells between 15 % -90 % egg length (0 % = posterior tip; 100 % = anterior tip) to form the embryonic mesoderm (Sonnenblick, 1950; Campos-Ortega & Hartenstein, 1985). Ventral furrow formation is normal in embryos derived from homozygous l(l)ph germline clones but extends posteriorly to approximately 5 % egg length. Both the cephalic furrow and the anterior midgut form correctly at 70 % and 85 % egg length, respectively, but formation of the posterior midgut invagination does not occur. In living embryos, germband extension is observed at 3-5 h of development for the majority of pole hole and null embryos. However, in 10% of the embryos, a twist along the dorsal-ventral axis occurs at this time, producing corkscrew-shaped embryos (Fig. 3B). Relaxation of these twists is observed in the pole hole embryos prior to dorsal closure (Fig. 3C). Twisting occurs in both pole hole and null embryos (Fig. 3F). At 4-5 h of development (stage 10), the invagination of cells to form the stomodeum is observed for most embryos and by 7-5 h (stage 11) segmental grooves are present in all embryos. Therefore, from the time of fertilization until the completion of germband extension there are no external morphological differences between pole hole and null embryos.

Distinct maternal and zygotic requirements for l(l)ph+activity

Two phenotypic classes of embryos are observed by 8h (stage 12) of development. At this time, approximately one-half the embryos undergo germband shortening and upon completing development show the terminal class phenotype (Figs 2D-F and 3C). For a small percentage of the embryos, the germband does not retract and U-shaped (Fig. 3D) pole hole embryos are formed.

Fig. 2.

Phenotypic analysis of pole hole and null embryos. (A) Cuticular preparation of a wild-type embryo showing 8 abdominal segments (A1-A8), cephalopharyngeal skeleton, and terminal posterior structures including the anal plate, telson and spiracles. (B) A parasagittal section of a 6-5 h (stage 11) wild-type embryo. The germband is fully extended bringing A9 in close proximity with the procephalic lobe. Anteriorly, the clypeolabrum and stomodeum have formed. Note there are small pockets of cell death that occur at this stage during wild-type development. Yolk fills the internal region of the embryo. (C) A parasagittal section of a 12 h (stage 15) wild-type embryo. The frontal sac has formed as the result of the anterior progression of the dorsal fold. Yolk is enclosed by the midgut. (D) Cuticular preparation of a pole hole embryo showing the terminal class phenotype. Only 7 abdominal segments have formed (A1-A7) and the cephalopharyngeal skeleton is truncated. (E) A parasagittal section of a 7h (stage 11) pole hole embryo. In the head and tail regions, a high level of cell death is observed. Derivatives of the proctodeum, including the hindgut and posterior midgut, are absent; however, the visceral mesoderm that normally joins these endodermal structures is present. Anteriorly, the stomodeum, which is slightly out of focus in this section, invaginates normally to form the pharynx and the oesophagus. Formation of the salivary glands appears normal. The terminal posterior segment is most likely A7. (F) A parasagittal section showing a 11 h (stage 14) pole hole embryo. The migration of the dorsal fold over the procephalic lobe to create the frontal sac will be complete at stage 15. The dorsal portion of the cephalopharyngeal skeleton is derived from the internalized procephalic lobe region and is abnormal in pole hole embryos. The cells that form the pharynx are also present, but differentiation of the cuticular posterior process of the pharynx is abnormal. The star (*) represents the anteriormost region of the central nervous system. Posteriorly, cell death is prominent, filling the space between the nerve cord and the epidermis. The periphery of the yolk sac has become cellularized. (G) A living null embryo after 20 h of development. All embryonic tissue is located at the anterior pole of the egg case while extruded yolk lies posteriorly. (H) A parasagittal section of a null embryo after 7h of development (approximately stage 11). The germ band is extended but will not retract. The stomodeum (not shown here) and salivary glands have formed. At this time cell death is pronounced in the head and throughout the central region of the germ band. (I) A parasagittal section of a null embryo after 12 h. Massive cell death and loss of cells is apparent in the head and tail regions (triangles). Cell death is also found the internal region of the surviving thoracic and anterior abdominal segments. Abbreviations: (ap) anal plate; (as) amnioserosa; (cd) cell death; (cl) clypeolabrum; (cs) cephalopharyngeal skeleton; (df) dorsal fold; (fs) frontal sac; (h) hindgut; (mg) midgut; (df) procephalic lobe; (ph) pharynx; (pm) posterior midgut; (sbg) suboesophageal ganglia; (sg) salivary gland; (sp) spiracles; (spg) supraoesophageal ganglia; (st) stomodeum; (te) telson; (vc) ventral nerve cord; (vm) visceral mesoderm; (y) yolk.

Fig. 2.

Phenotypic analysis of pole hole and null embryos. (A) Cuticular preparation of a wild-type embryo showing 8 abdominal segments (A1-A8), cephalopharyngeal skeleton, and terminal posterior structures including the anal plate, telson and spiracles. (B) A parasagittal section of a 6-5 h (stage 11) wild-type embryo. The germband is fully extended bringing A9 in close proximity with the procephalic lobe. Anteriorly, the clypeolabrum and stomodeum have formed. Note there are small pockets of cell death that occur at this stage during wild-type development. Yolk fills the internal region of the embryo. (C) A parasagittal section of a 12 h (stage 15) wild-type embryo. The frontal sac has formed as the result of the anterior progression of the dorsal fold. Yolk is enclosed by the midgut. (D) Cuticular preparation of a pole hole embryo showing the terminal class phenotype. Only 7 abdominal segments have formed (A1-A7) and the cephalopharyngeal skeleton is truncated. (E) A parasagittal section of a 7h (stage 11) pole hole embryo. In the head and tail regions, a high level of cell death is observed. Derivatives of the proctodeum, including the hindgut and posterior midgut, are absent; however, the visceral mesoderm that normally joins these endodermal structures is present. Anteriorly, the stomodeum, which is slightly out of focus in this section, invaginates normally to form the pharynx and the oesophagus. Formation of the salivary glands appears normal. The terminal posterior segment is most likely A7. (F) A parasagittal section showing a 11 h (stage 14) pole hole embryo. The migration of the dorsal fold over the procephalic lobe to create the frontal sac will be complete at stage 15. The dorsal portion of the cephalopharyngeal skeleton is derived from the internalized procephalic lobe region and is abnormal in pole hole embryos. The cells that form the pharynx are also present, but differentiation of the cuticular posterior process of the pharynx is abnormal. The star (*) represents the anteriormost region of the central nervous system. Posteriorly, cell death is prominent, filling the space between the nerve cord and the epidermis. The periphery of the yolk sac has become cellularized. (G) A living null embryo after 20 h of development. All embryonic tissue is located at the anterior pole of the egg case while extruded yolk lies posteriorly. (H) A parasagittal section of a null embryo after 7h of development (approximately stage 11). The germ band is extended but will not retract. The stomodeum (not shown here) and salivary glands have formed. At this time cell death is pronounced in the head and throughout the central region of the germ band. (I) A parasagittal section of a null embryo after 12 h. Massive cell death and loss of cells is apparent in the head and tail regions (triangles). Cell death is also found the internal region of the surviving thoracic and anterior abdominal segments. Abbreviations: (ap) anal plate; (as) amnioserosa; (cd) cell death; (cl) clypeolabrum; (cs) cephalopharyngeal skeleton; (df) dorsal fold; (fs) frontal sac; (h) hindgut; (mg) midgut; (df) procephalic lobe; (ph) pharynx; (pm) posterior midgut; (sbg) suboesophageal ganglia; (sg) salivary gland; (sp) spiracles; (spg) supraoesophageal ganglia; (st) stomodeum; (te) telson; (vc) ventral nerve cord; (vm) visceral mesoderm; (y) yolk.

Fig. 3.

Scanning electron micrograph analysis of pole hole and null embryos. (A and B) Embryos derived from l(l)ph germline clones at 6 and 7-5 h respectively. The distinction between pole hole and null embryos cannot be made based on external morphology at these times. The germband is extended in A and an example of a twisted embryo with external pole cells is shown in B. (C) An 11 h (stage 14) pole hole embryo showing seven abdominal segments, the germband is not fully retracted. (D) A U-shaped pole hole embryo at 20h after cuticle formation. Arrows delimit the head region containing noninvoluted gnathal segments. (E) A twisted null embryo after 7h. (F) A nontwisted null embryo after 15 h showing regions of massive cell death. Abbreviations: (as) amnioserosa; (cd) cell death; (cf) cephalic furrow; (pc) pole cells; (pl) procephalic lobe; (st) stomodeum; (tp) tracheal pit; (vf) ventral furrow.

Fig. 3.

Scanning electron micrograph analysis of pole hole and null embryos. (A and B) Embryos derived from l(l)ph germline clones at 6 and 7-5 h respectively. The distinction between pole hole and null embryos cannot be made based on external morphology at these times. The germband is extended in A and an example of a twisted embryo with external pole cells is shown in B. (C) An 11 h (stage 14) pole hole embryo showing seven abdominal segments, the germband is not fully retracted. (D) A U-shaped pole hole embryo at 20h after cuticle formation. Arrows delimit the head region containing noninvoluted gnathal segments. (E) A twisted null embryo after 7h. (F) A nontwisted null embryo after 15 h showing regions of massive cell death. Abbreviations: (as) amnioserosa; (cd) cell death; (cf) cephalic furrow; (pc) pole cells; (pl) procephalic lobe; (st) stomodeum; (tp) tracheal pit; (vf) ventral furrow.

Germband retraction does not occur in null embryos. Instead, development ceases at the elongated germband stage. As shown in Fig. 2G, the resulting embryos at 24 h consist of a ball of embryonic tissue at the anterior egg pole and an extruded mass of egg yolk at the posterior pole. Muscle contractions are observed in the anterior region of these living null embryos. However, the overall organization and differentiation of organ systems within the embryo is abnormal. This phenotype is the consequence of incomplete development and massive cell death (Fig. 3F). In the analysis of null tissue sections, cell death coincident with pycnotic nuclei is evident internally beginning at approximately 6-5 h. A tissue section of null embryo containing many pockets of cell death at 7 h is shown in Fig. 2H. After 12 h (Fig. 2I), the progression of cell death in the head and tail regions is so extreme that many cells are lost from the embryo (see triangles). In preparations of fixed embryos derived from l(l)ph germline clones, only fragments of null embryos containing the gnathal, thoracic and the anteriormost abdominal segments are recovered after 10 h. Cell death is also a consequence of wild-type development but is limited to small subepidermal clusters of necrotic cells that accumulate between 7 and 9 h of development (Fig. 2B). For pole hole embryos, many additional pycnotic nuclei are observed in the terminal embryonic regions (Fig. 2E and F).

Characterization of the fate map in pole hole and null embryos

Both maternal and zygotic gene products play a role in determining cellular identities at the time of blastoderm formation. In the mature embryo, alterations in head and tail regions are apparent in embryos lacking maternal l(l)ph+ product. To determine whether the establishment of segmental identities in these regions depends upon functional maternal l(l)ph+ activity, the expression patterns of the fushi-tarazu (ftz) and engrailed (en) segmentation genes were characterized in embryos derived from homozygous l(l)ph germline clones.

As shown in Fig. 4, at the blastoderm stage, instead of the seven alternating stripes of ftz expression observed in the wild-type embryo (Fig. 4A), only six regions of ftz activity were detected by /3-galactosidase expression in /tz/lacZ embryos lacking maternal l(l)ph+ function (Fig. 4B). For this embryo, the distance between the fifth and sixth stripe is greater than that of the normal embryo, and the sixth stripe is broader than the preceding anterior five stripes. Therefore, in the absence of maternal l(l)ph+ activity, cell fates in the posterior region of the embryo are changed. A deletion of terminal identities is observed and cells forming in this region assume more anterior fates. This altered segmental pattern is observed for all embryos derived from homozygous 1(1 )ph germline clones and at this stage pole hole and null embryos are phenotypically equivalent.

Fig. 4.

Characterization of ftz expression in embryos derived from l(l)ph germline clones. The pattern of ftz/lacZ expression is used as a marker to characterize the embryonic fate map and to identify the MP2 neuroblasts of the central nervous system. (A) and (C) Wild-type embryos at the blastoderm (3h) and extended germband stage (5 5 h), respectively, showing seven stripes of ftz expression. The center of each parasegmental stripe and nonstriped region defines the position where a segmental boundary will later form. At 5 h ftz expression is also detected in the MP2 neuroblasts (arrows) derived from the anterior compartment of each segment. (B) and (D) Embryos derived from l(l)ph germline clones showing only six regions of ftz expression. Note the expanded width of striped and nonstriped regions in the posterior of these embryos and the normal number of MP2 cells formed in the enlarged segments. Abbreviations: (PS) parasegment; (arrows) MP2 cells of the central nervous system.

Fig. 4.

Characterization of ftz expression in embryos derived from l(l)ph germline clones. The pattern of ftz/lacZ expression is used as a marker to characterize the embryonic fate map and to identify the MP2 neuroblasts of the central nervous system. (A) and (C) Wild-type embryos at the blastoderm (3h) and extended germband stage (5 5 h), respectively, showing seven stripes of ftz expression. The center of each parasegmental stripe and nonstriped region defines the position where a segmental boundary will later form. At 5 h ftz expression is also detected in the MP2 neuroblasts (arrows) derived from the anterior compartment of each segment. (B) and (D) Embryos derived from l(l)ph germline clones showing only six regions of ftz expression. Note the expanded width of striped and nonstriped regions in the posterior of these embryos and the normal number of MP2 cells formed in the enlarged segments. Abbreviations: (PS) parasegment; (arrows) MP2 cells of the central nervous system.

When engrailed (en) expression is compared between embryos derived from homozygous l(l)ph germline clones, variation in the striping pattern is observed. Cells that express en label the posterior compartment of the segmental unit. In early wild-type gastrulating embryos there are 15 stripes of en expression (Fig. 5A). Along the longitudinal embryonic axis the labial, 3 thoracic and 9 abdominal segments show three nonstaining nuclei and one en nucleus per segment (DiNardo et al. 1985). In addition, approximately 11 nonstaining cells lie posterior to the 9th abdominal en stripe at the blastoderm stage. For embryos derived from homozygous 1(1 )ph germline clones, only 12 or 13 stripes of en expression form (Fig. 5B). When cells were counted along the longitudinal axes of l(l)ph germline clone embryos, variation in the number of cells per segment in the region of expanded segmental identities were observed between embryos. Fig. 5C2 and 5C3 depicts schematically the variation in en expression between two embryos lacking maternal l(l)ph+ activity as compared to the wild-type expression pattern (Fig. 5C1).

Fig. 5.

Engrailed protein expression in embryos lacking maternal l(l)ph gene product. (A) An early gastrulating (3-5 h) wild-type embryo showing 15 stripes (the 2nd stripe lies in the groove of the cephalic furrow) of en expression. (B) A 3-5 h embryo derived from a l(l)ph germline clone with only 13 regions of en expression. The cell fates deleted include A8, A9 (interstripe and stripe 14 and 15), A10 and the telson (posterior nonstriped region). (Cl, C2, and C3) A representation of blastoderm cells (circles) beginning at the cephalic furrow and extending along the longitudinal axis to the posterior tip of the embryo; the cells whose nuclei contain the en protein are filled in. (Cl) For the wild-type embryo each segment is typically composed of three nonstaining (anterior compartment) and one en cell (posterior compartment). (C2 and C3) For embryos derived from l(l)ph germline clones the pattern of en expression in the posterior region is variable. Both the extent of segmental deletions and expansions vary between embryos. (D and E) A wild-type and l(l)ph germline clone embryo at 5·5−6 h (stage 11) with three head stripes of en expression (arrows). Abbreviations: (A) abdominal segment; (am) anterior midgut; (cf) cephalic furrow; (df) dorsal fold; (hg) hindgut; (pc) pole cells; (T) thoracic segment.

Fig. 5.

Engrailed protein expression in embryos lacking maternal l(l)ph gene product. (A) An early gastrulating (3-5 h) wild-type embryo showing 15 stripes (the 2nd stripe lies in the groove of the cephalic furrow) of en expression. (B) A 3-5 h embryo derived from a l(l)ph germline clone with only 13 regions of en expression. The cell fates deleted include A8, A9 (interstripe and stripe 14 and 15), A10 and the telson (posterior nonstriped region). (Cl, C2, and C3) A representation of blastoderm cells (circles) beginning at the cephalic furrow and extending along the longitudinal axis to the posterior tip of the embryo; the cells whose nuclei contain the en protein are filled in. (Cl) For the wild-type embryo each segment is typically composed of three nonstaining (anterior compartment) and one en cell (posterior compartment). (C2 and C3) For embryos derived from l(l)ph germline clones the pattern of en expression in the posterior region is variable. Both the extent of segmental deletions and expansions vary between embryos. (D and E) A wild-type and l(l)ph germline clone embryo at 5·5−6 h (stage 11) with three head stripes of en expression (arrows). Abbreviations: (A) abdominal segment; (am) anterior midgut; (cf) cephalic furrow; (df) dorsal fold; (hg) hindgut; (pc) pole cells; (T) thoracic segment.

Within a population of embryos derived from homozygous 1(1 )ph germline clones, many variations in segmental width for the region of expanded cell identities are observed. This variation is due to differences in cellular determination between embryos and not a secondary consequence of early cell death, since pycnotic nuclei are not observed until 6-5 h. For the 30 embryos derived from germline clones analyzed by counting the number of cells along the width of a segment, at 3-4 h the deleted pattern elements always included the en stripes 14-15 (posterior compartment of A8, and all of A9) and the posterior nonstaining region (A10, telson, and the endodermal derivatives of the proctodeum). Most often (60%) these embryos were also missing the eighth abdominal segment and the posterior compartment of A7 (Fig. 5E). Larger deletions were observed in 10% of the embryos. For the remaining embryos, 20% showed expression of a 13th en stripe (Fig. 5B) and 10 % contained nonstaining cells (A8 anterior) after the 13th stripe. The segmental identities typically expanded to fill this deleted region of the fate map (approximately 20%) were A5, A6 and A7. Occasionally, additional cells were observed in A3, and A4.

Characterization of the anterior defect in embryos lacking maternal l(l)ph+activity

The second region of the embryo affected by the loss of maternal l(l)ph+ product is the head. In wild-type extended germband embryos, the en antibody labels nuclei in three ectodermal regions on either side of the head, forming a characteristic striping pattern (Fig. 5D; also see DiNardo et al. 1985). A similar number and arrangement of en head stripes are present in embryos developing without maternal l(l)ph+ activity (Fig. 5E). This region of en labeling does not correspond to the head region affected by lack of maternal l(l)plD activity. To define the head defect and its relation to the embryonic fate map, the formation of cuticular head structures were analyzed in pole hole embryos. Also, the nervous systems of these embryos were examined using antibodies against horseradish peroxidase (HRP), which labels all central (CNS) and peripheral (PNS) nervous system cell bodies and axons (Jan & Jan, 1982), and SOX2, which recognizes the cell bodies and axons of the PNS and a subset of CNS neurons (Goodman et al. 1984). Based on the origin of the mutant structures identified, the regions of the head affected by the lack of maternal l(l)ph+ product can be defined and correlated to the embryonic fate map.

In cuticular preparations of mature pole hole embryos, several defects in the head skeleton are apparent. The cephalopharyngeal apparatus (Fig. 6A) is formed by the involution of head segments and the intemalmost elements are derived from regions of the head distal to the anterior pole (Campos-Ortega & Hartenstein, 1985; Jurgens et al. 1986). This apparatus is secreted by cells lining the atrium, pharynx and frontal sac. The truncated apparatus observed in pole hole embryos is thickened and collapsed (Fig. 6B). In detail, the dorsal bridge is absent and the lateral wall of the dorsal process reduced. In addition, the ventral process of the pharynx is abnormal. These structures of the mouth skeleton arise from the acron region of the fate map (Jurgens et al. 1986). The medial tooth derived from the labral head segment is also absent. Thus, the labrum and acron are two head regions affected by the absence of maternal l(l)ph+ product. The labrum occupies an anterior-dorsal position on the embryonic fate map while the adjacent acron region extends posteriorly and laterally; together these two regions define approximately the dorsal half of the procephalic region (Jurgens et al. 1986).

Fig. 6.

Head skeletal defects in pole hole embryos. Organization of the cephalopharyngeal skeleton in a wild-type (A) and a pole hole (B) embryo. The head skeleton is formed by the involution of the dorsal procephalic and labral regions to form the frontal sac; and the clypeolabrum and gnathal segments to form the atrium and pharynx of the foregut. In pole hole embryos the medial tooth, dorsal bridge, and ventral posterior process of the pharynx are absent. The H-piece appears shortened and the dorsal process is abnormal. The mouth hooks of maxillary origin are normal in pole hole embryos. Note in this preparation, both embryos are encased in their outer egg membrane and the micropyle is visible. Abbreviations-. (dp) dorsal process; (Hp) H-piece; (mh) mouth hooks; (mi) micropyle; (mto) medial tooth; (vpp) ventral posterior process.

Fig. 6.

Head skeletal defects in pole hole embryos. Organization of the cephalopharyngeal skeleton in a wild-type (A) and a pole hole (B) embryo. The head skeleton is formed by the involution of the dorsal procephalic and labral regions to form the frontal sac; and the clypeolabrum and gnathal segments to form the atrium and pharynx of the foregut. In pole hole embryos the medial tooth, dorsal bridge, and ventral posterior process of the pharynx are absent. The H-piece appears shortened and the dorsal process is abnormal. The mouth hooks of maxillary origin are normal in pole hole embryos. Note in this preparation, both embryos are encased in their outer egg membrane and the micropyle is visible. Abbreviations-. (dp) dorsal process; (Hp) H-piece; (mh) mouth hooks; (mi) micropyle; (mto) medial tooth; (vpp) ventral posterior process.

When the anterior nervous system in a pole hole embryo is examined, the structures affected by the lack of maternal l(l)ph+ function also originate from the acron and labrum regions. As shown in Fig. 7, much of the brain (supraoesophageal ganglia) is missing in the 16 h pole hole embryo (Fig. 7D) when compared to the wild-type (Fig. 7B). The brain lobes develop from neuroblasts originating from a medial-lateral position on either side of the acron. Another distinguishable structure derived from the acron are the optic lobes. The optic lobes arise as a single placode at a dorsal-posterior position of the acron that invaginates at 7h (stage 11) to join the developing supraoesophageal ganglia on its ventral side. The trajectory of Bolwig’s nerve serves as a good marker for the localization of the optic lobe anlagen in wild-type and pole hole embryos (Steller et al. 1987). Fig. 8A shows the extension of Bolwig’s nerve (arrows) in a wild-type embryo from cell bodies forming the larval photoreceptor cluster into the optic lobe region of the supraoesophageal ganglia. For pole hole embryos, however, the axons of Bolwig’s nerve project to a dorsal-posterior position of the procephalic lobe (Fig. 8B and 8C). In wild-type embryos, this region would normally give rise to the optic anlagen. Wandering of Bolwig’s nerve beyond this region is observed in some pole hole embryos at later stages of embryogenesis (data not shown). Thus, the brain and optic lobe anlagen, derivatives of the acron, are altered by the maternal effect of l(l)ph.

Fig. 7.

Pattern of the central nervous system in pole hole embryos. HRP antibody staining of the nervous system at 16 h in wild-type (A and B) and pole hole (C and D) embryos. A ventral optical section showing the ventral nerve cord in a wildtype (A) and pole hole embryo (C). Note the posterior deformation of the CNS ladder (arrow) observed after nerve cord condensation for many pole hole embryos. The anal pads stain with the HRP antibody in wild-type embryos. Dorsal optical sections of the anteriormost region of the CNS in wild-type (B) and pole hole (D) embryos. For the wild-type, the brain lobes or supraoesophageal ganglia are evident, but are absent in pole hole embryos (arrow). Abbreviations: (ap) anal pads; (spg) supraoesophageal ganglia; (vc) ventral nerve cord.

Fig. 7.

Pattern of the central nervous system in pole hole embryos. HRP antibody staining of the nervous system at 16 h in wild-type (A and B) and pole hole (C and D) embryos. A ventral optical section showing the ventral nerve cord in a wildtype (A) and pole hole embryo (C). Note the posterior deformation of the CNS ladder (arrow) observed after nerve cord condensation for many pole hole embryos. The anal pads stain with the HRP antibody in wild-type embryos. Dorsal optical sections of the anteriormost region of the CNS in wild-type (B) and pole hole (D) embryos. For the wild-type, the brain lobes or supraoesophageal ganglia are evident, but are absent in pole hole embryos (arrow). Abbreviations: (ap) anal pads; (spg) supraoesophageal ganglia; (vc) ventral nerve cord.

Fig. 8.

Pattern of the peripheral nervous system in pole hole embryos. Embryos are stained with the SOX2 antibody. (A) Lateral view of a 16 h wild-type embryo showing the trajectory of Bolwig’s nerve (arrows) into the optic lobe region of the supraoesophageal ganglia. The axons of Bolwig’s nerve (arrows) originate from the larval photoreceptor cluster and terminate on the optic lobe located on the ventral surface of the brain lobes. (B) The trajectory of Bolwig’s nerve (arrows) in a 14 h pole hole embryo. The brain lobes are missing in these embryos and Bolwig’s nerve projects to a dorsoposterior region of the procephalic epidermis that normally gives rise to the optic lobes (*). (C) Localization of larval photoreceptor cells and Bolwig’s nerve (arrows) in pole hole embryos at 11 h. (DI through E2) Horizontal views of 17h embryos showing the positions of two cephalopharyngeal sense organs: the hypophysis and epiphysis. For wild-type embryos, the hypophysis (DI) lies in an optical section ventral to the epiphysis (D2). For the pole hole embryo, the hypophysis is present (El), but in more dorsal optical sections the epiphysis (E2) is missing; note the absence of the supraoesophageal ganglia (*). (F and G) Lateral views of the abdominal PNS in wild-type and pole hole embryos, respectively (see also Ghysen et al. 1986). A characteristic pattern of sensory cells are present for each segment in wild-type embryos; these include cells of the dorsal cluster, lateral chordotonals, and ventral cell clusters. For pole hole embryos, no duplications of sensory organs are observed in the regions of expanded segmental identities. The posteriormost complete segment often shows a dorsolateral displacement of lateral chordotonals (arrow) and a fusion of cells in the dorsal cluster (not shown here). In the terminal region, the specialized sensory organs of A8 and A9 (see arrows in F) are absent in pole hole embryos (*). Abbreviations: (de) dorsal cluster; (ep) epiphysis; (hy) hypophysis; (Ich) lateral chordotonals; (Ipc) larval photoreceptor cells; (spg) supraoesophageal ganglia; (v and v’) ventral cell clusters.

Fig. 8.

Pattern of the peripheral nervous system in pole hole embryos. Embryos are stained with the SOX2 antibody. (A) Lateral view of a 16 h wild-type embryo showing the trajectory of Bolwig’s nerve (arrows) into the optic lobe region of the supraoesophageal ganglia. The axons of Bolwig’s nerve (arrows) originate from the larval photoreceptor cluster and terminate on the optic lobe located on the ventral surface of the brain lobes. (B) The trajectory of Bolwig’s nerve (arrows) in a 14 h pole hole embryo. The brain lobes are missing in these embryos and Bolwig’s nerve projects to a dorsoposterior region of the procephalic epidermis that normally gives rise to the optic lobes (*). (C) Localization of larval photoreceptor cells and Bolwig’s nerve (arrows) in pole hole embryos at 11 h. (DI through E2) Horizontal views of 17h embryos showing the positions of two cephalopharyngeal sense organs: the hypophysis and epiphysis. For wild-type embryos, the hypophysis (DI) lies in an optical section ventral to the epiphysis (D2). For the pole hole embryo, the hypophysis is present (El), but in more dorsal optical sections the epiphysis (E2) is missing; note the absence of the supraoesophageal ganglia (*). (F and G) Lateral views of the abdominal PNS in wild-type and pole hole embryos, respectively (see also Ghysen et al. 1986). A characteristic pattern of sensory cells are present for each segment in wild-type embryos; these include cells of the dorsal cluster, lateral chordotonals, and ventral cell clusters. For pole hole embryos, no duplications of sensory organs are observed in the regions of expanded segmental identities. The posteriormost complete segment often shows a dorsolateral displacement of lateral chordotonals (arrow) and a fusion of cells in the dorsal cluster (not shown here). In the terminal region, the specialized sensory organs of A8 and A9 (see arrows in F) are absent in pole hole embryos (*). Abbreviations: (de) dorsal cluster; (ep) epiphysis; (hy) hypophysis; (Ich) lateral chordotonals; (Ipc) larval photoreceptor cells; (spg) supraoesophageal ganglia; (v and v’) ventral cell clusters.

For the peripheral nervous system, the labral sense organ, originating from the labrum, is absent in pole hole embryos. The wild-type head skeleton contains two sense organs, the hypophysis derived from the labial head segment (Fig. 8D1) and the labral-derived sense organ, the epiphysis (Fig. 8D2). For embryos lacking maternal l(l)ph+ function, the hypophysis is present (Fig. 8E1); however, in more dorsal optical sections, the labral sense organ is missing (Fig. 8E2). Since the other sensory elements derived from the remaining head segments develop normally in pole hole embryos, the labrum and acron are the two areas of embryonic fate map in the head altered by the maternal effect of l(l)ph.

Consequences of the expansion of the posterior fate map in pole hole embryos

Since there are more ectodermal cells per segment in the tail region of 3 h pole hole embryos, this change may alter the subsequent pattern of the central nervous system derived from these enlarged segments. Briefly, neuroblasts leave the ectodermal layer at 4-6 h and are localized between the mesoderm and ectoderm. They undergo a series of divisions and then progeny neurons extend their axons in a specific manner to form the wildtype CNS ladder structure observed in Fig. 7A. The identification of the MP2 neuroblasts (Doe et al. 1988) serve as a good marker to analyze the pattern of segmental neuronal determination in pole hole embryos. First, the MP2 cells express the/rz gene and by 5h these cells are prominently stained in wild-type ftz/lacZ embryos (Fig. 4C). Secondly, since one set of MP2 cells arises from the anterior compartment of each abdominal segment, an alteration in the specification of MP2 neuroblasts in the expanded segmental fields would be easily detected. As shown in Fig. 4D for a pole hole embryo there is no duplication of MP2 neuroblasts in the region of expanded cell identities. It appears that the integrity of a single segment (and/or parasegment) is maintained in pole hole embryos and in the presence of excess ectodermal cells one set of MP2 neuroblasts are formed per segment. A deformation of the CNS ladder is observed for many pole hole embryos after nerve cord condensation in the region of expanded segmental identities (Fig. 7C). Usually, the anterior and posterior commissures are normal in the last complete terminal segment but breaks in the longitudinal axis of the CNS are observed (arrow). Perhaps, the presence of large numbers of dying cells interfere with the formation of axonal tracts in these regions. In the posteriormost region of pole hole embryos, ill-defined bundles of HRP staining cells are detected.

For the peripheral nervous system of pole hole embryos (Fig. 8G), the wild-type number of sensory cells per abdominal segment are observed in the enlarged segments but, in some cases, these elements are displaced laterally. In addition, fusion of the dorsal cell cluster is observed along the dorsal midline and the positions of the lateral chordotonal cells are shifted dorsally (Fig. 8G, arrow) in the posteriormost complete segment of pole hole embryos. For the terminalmost region, cells staining with SOX2 are observed (*), but these cells cannot be identified morphologically and they form several fused bundles at later stages of development.

Therefore, within the region of expanded segmental identities, the nervous system of pole hole embryos develops much like the more anterior unaffected segments. In the last incomplete segment at the posterior pole, the central and peripheral nervous systems are abnormal. In null embryos the development of the nervous system is severely affected by the lack of l(l)ph+ activity and only isolated clusters of HRP and SOX2 staining cells are detected; there is no organization of these cells to produce a ladder-like structure or to differentiate into morphologically identifiable sensory cells.

Comparison between the maternal effects o/l(l)ph and the terminal class of female sterile mutations

Table 1 lists the terminal class of female sterile mutations and the phenotypic changes associated with the lack of these maternal products during embryonic development (Degelmann et al. 1986; Schupbach & Wieschaus, 1986; Nüsslein-Volhard et al. 1987; Mlodzik et al. 1987; Winslow et al. 1988). The maternal effect of l(l)ph in germline clones is included for comparison; the pole hole phenotype shows the maternal requirement for l(l)ph while the null phenotype results from lack of both maternal and zygotic l(l)ph gene expression. For the genes fisted, lack of maternal function produces equivalent posterior alterations of the fate map and identical secondary effects for subsequent posterior CNS and PNS development as described here for l(l)ph. Generally for this group, lack of maternal function results in the deletion of A8 through the telson accompanied by the expansion of cell fates in A5 through A7. (In this study, slight variations in the number and/or position of en, ftz and/or eve striped and interstriped regions were also observed in the tails of progeny derived from this group.)

Table 1.

Comparison between the maternal effect of the terminal female sterile group and the l(l)ph mutation

Comparison between the maternal effect of the terminal female sterile group and the l(l)ph mutation
Comparison between the maternal effect of the terminal female sterile group and the l(l)ph mutation

In the anterior terminal region of progeny derived from members of the terminal group, differences in cellular determination are observed. A shift forward of the cephalic furrow and the anterior midgut invagination is shown for embryos derived from all members of this female sterile group except fs(l)Nasrat. Consequently, for these embryos instead of the 25−30 cells observed along the longitudinal axis in the heads of wild-type and/s(7)Nasrat embryos, only 20-25 cells are present for the progeny derived from fs(l)pole hole, fs(2)torso, fs(2)trunk and fs(3)torso-like homozygous mothers. In this regard, the maternal effect of l(l)ph is similar to that observed for fs(l)Nasrat and differs from the other mutations in this gene class. However, head structures derived from labrum and acron regions are missing or abnormal for embryos produced by all members of the female sterile terminal class group and no additional defects associated with other head segments are observed. When the pattern of en staining was compared between different members of this group some embryos derived from homozygous fs(2)torso mothers showed additional cells in the gnathal segments. Presumably, this increase in segmental width is related to the forward shift of the cephalic furrow.

There is also a difference in the percent of progeny that undergo twisting instead of extension of the germ band at 3-5 h of development. For example, greater than 90% of the progeny produced by homozygous fs(2)torso females twist at the time of germband extension while less than half of the embryos derived from homozygous fs(l)Nasrat females show this defect. Relaxation of these twists occurs prior to dorsal closure in a manner similar to that described for pole hole embryos. Thus, in this respect also the phenotypic affect of l(l)ph is equivalent to that of fs(l)Nasrat.

The l(l)pole hole gene plays an important role in the determination and differentiation of cells at many stages of development. Its activity is required in the unfertilized egg and/or in preblastoderm staged embryos for cellular determination in terminal regions of the embryo. When maternal l(l)ph+ activity is absent, the labral and acron portions of the head fail to differentiate and, in the tail, deletions include abdominal segments A8-A10 and the telson. The terminal class of female sterile mutations produce equivalent phenotypic changes on embryonic development (Schupbach & Wieschaus, 1986; Nüsslein-Volhard et al. 1987). Based on the similarity between the embryonic phenotypes produced by these mutations, the l(l)ph gene can be considered a member of the Drosophila terminal gene class.

In Drosophila embryos, overlapping maternal and zygotic genetic programs are required to ensure the establishment of wild-type segmental identities along the anterior-posterior embryonic axis. For the anterior half of the embryo, a gradient of bicoid protein, highest at the anterior embryonic pole, is responsible for specifying cellular identities in gnathal and thoracic segments in a concentration-dependent manner (Driever & Nüsslein-Volhard, 1988#). Also, bicoid+ activity is required to distinguish between head and tail pathways of differentiation in the terminal region of the head. Embryos that develop without maternal bicoid activity show an anterior duplication of the telson, loss of gnathal and thoracic segments, and spreading of terminal and abdominal cell fates into the deleted gnathal and thoracic domain of the embryo (Driever & Nüsslein-Volhard, 1988b). For embryos lacking maternal terminal class activity, a normal distribution of bicoid protein is observed (Driever & Nüsslein-Volhard, 1988b). These embryos show a deletion of terminal head structures accompanied by an anterior shift of the cephalic furrow and anterior midgut invagination (Schupbach & Wieschaus, 1986; Nüsslein-Volhard et al. 1987). Some embryos also show an enlargement of the thoracic segments (Winslow et al. 1988). However, for embryos lacking maternal fs(l)Nasrat or l(l)ph activity, spreading of cellular identities toward the anterior pole is not observed. Both the anterior midgut invagination and cephalic furrow form at their wild-type positions in these embryos but labral and acron head structures are missing or are abnormal. This difference in the maternal effect between members of the terminal group indicates that fs(l)Nasrat and/or l(l)ph gene activities may lie downstream or have a limited function in the terminal class pathway and thereby result in a less extreme embryonic phenotype.

Another aspect of the terminal phenotype, the twisting of the embryo at the time of germband extension, may occur as a direct result of the misplacement of the cephalic furrow during early gastrulation. Both l(l)ph and fs(l)Nasrat embryos show a decreased tendency to twist when compared to embryos derived from other terminal group genes. Placement of the cephalic furrow at approximately 70 % egg length may be necessary to generate the correct morphogenetic balance required for wild-type movement of the germband in a dorsal and then anterior direction. Twisting of terminal class embryos occurs when the germband begins to extend dorsally, then turns in a ventral direction. As a result of this continued ventral movement of the germband, the entire embryo becomes twisted. When the germband retracts, the twists are relaxed and the wild-type linear form of the embryo is restored.

In the posterior region of all terminal class embryos, cell fates normally found at 0-20% egg length are replaced by the posterior expansion of A5-A7 abdominal identities usually established at 40-20 % egg length (Degelmann et al. 1986; Mlodzik et al. 1987; Winslow et al. 1988). The assignment of segmental and compartmental cell identity within the expanded region of terminal group embryos appears to be somewhat random since the width and number of segments established is variable between embryos of the same maternal genotype. These slight variations in segmental patterning may result from establishment of abdominal cellular fates in a region containing a high concentration of posterior morphogen.

At the blastoderm stage, terminal class embryos show 2-3 times the wild-type number of cells contributing to the formation of the terminalmost complete abdominal segment (usually A6). However, at later developmental stages, a wild-type pattern in the PNS is observed within A6. Thus, a process of regulation must occur within the expanded segmental domains. Concurrent with nervous system differentiation, massive cell death is observed in the posterior of terminal class embryos. It seems likely that the mechanism used to generate a wild-type PNS pattern in the enlarged segments is via cell death. Perhaps, after all possible cell fates have been assumed, cells left with no identities to fulfill die. Cell death is also observed in the heads of terminal class embryos and is particularly striking in those derived from homozygous l(l)ph and fs(l)Nasrat eggs. Since there is little or no shift forward of the anterior fate map in these embryos, many more cells are formed in the head. Presumably, due to lack of terminal group activity, these cells cannot assume a viable cell state and die.

In conclusion, the maternal effect of l(l)ph is essentially identical to that produced by fs(l)Nasrat and similar to other members of the terminal class of female sterile mutations. However, the l(l)ph gene is unique when compared to the members of the terminal gene group because its expression is required zygotically as well as maternally.

We are grateful to T. Schupbach for the 2nd and 3rd chromosomal torso class female sterile stocks. We also thank C. Hama, T. Kornberg, Y. Hiromi, W. Gehring, P. MacDonald & G. Struhl for lacZ fusion stocks; and T. Karr, S. DiNardo and P. O’Farrell for antibodies. We thank K. Maier for technical assistance, and D. Smouse for helpful suggestions and critical comments on the manuscript. This work was supported by the Howard Hughes Medical Institute and NIH grant HD23684 to N.P. and HD17608 to A.P.M. L.A. was a postdoctoral fellow of T32HD07104-09 training program.

Akam
,
M.
(
1987
).
The molecular basis for metameric pattern in the Drosophila embryo
.
Development
101
,
1
22
.
Anderson
,
K. V.
(
1987
).
Dorsal-ventral embryonic pattern genes of Drosophila
.
Trends in Genetics.
Vol.
3
,
91
97
.
Busson
,
B.
,
Gans
,
M.
,
Komitopoulou
,
K.
&
Masson
,
M.
(
1983
).
Genetic analysis of three dominant female sterile mutations located on the Y-chromosome of Drosophila melanogaster
.
Genetics
105
,
309
325
.
Campos-Ortega
,
J. A.
&
Hartenstein
,
V.
(
1985
).
The Embryonic Development of Drosophila melanogaster
.
Springer-Verlag, New York/Berlin
.
Degelmann
,
A.
,
Hardy
,
P. A.
,
Perrimon
,
N.
&
Mahowald
,
A. P.
(
1986
).
Developmental analysis of the torso-like phenotype in Drosophila produced by a maternal-effect locus
.
Devi Biol.
115
,
479
489
.
Dinardo
,
S.
,
Kuner
,
J. M.
,
Theis
,
J.
&
O’farrell
,
P.
(
1985
).
Development of embryonic pattern in D. melanogaster as revealed by accumulation of the nuclear engrailed protein
.
Cell
43
,
59
69
.
Driever
,
W.
&
NÜsslein-Volhard
,
C.
(
1988a
).
A gradient of bicoid protein in Drosophila embryos
.
Cell
54
,
83
93
.
Driever
,
W.
&
NÜsslein-Volhard
,
C.
(
1988b
).
The bicoid protein determines position in the Drosophila embryo in a concentrationdependant manner
.
Cell
54
,
95
104
.
Doe
,
C. Q.
,
Hiromi
,
Y.
,
Gehring
,
XV. J.
&
Goodman
,
C. S.
(
1988
).
Expression and function of the segmentation gene fushi tarazu during Drosophila neurogenesis
.
Science
239
,
170
175
.
Frasch
,
M.
,
Hoey
,
T.
,
Rushlow
,
C.
,
Doyle
,
H.
&
Levine
,
M.
(
1987
).
Characterization and localization of the even-skippered protein of Drosophila
.
EMBO J.
6
,
749
759
.
Foe
,
V. A.
&
Alberts
,
B. M.
(
1983
).
Studies of nuclear and cytoplasmic behavior during the five mitotic cycles that precede gastrulation in Drosophila embryogenesis
.
J. Cell Sci.
61
,
31
70
.
Frohnhofer
,
H. G.
&
NÜsslein-Volhard
,
C.
(
1986
).
Organization of anterior pattern in the Drosophila embryo by the maternal gene bicoid
.
Nature, Land.
324
,
120
125
.
Ghysen
,
A.
,
Dambly-Chaudiere
,
C.
,
Aceves
,
E.
,
Jan
,
L. Y.
& JAN. Y. N
. (
1986
).
Sensory neurons and peripheral pathways in Drosophila embryos
.
Roux’s Arch, dev! Biol.
195
,
49
62
.
Goodman
,
C. S.
,
Bastiani
,
M. J.
,
Doe
,
C. Q.
,
Du Lac
,
S.
,
Helfand
,
S. L.
,
Kuwada
,
J. Y.
&
Thomas
,
J. B.
(
1984
).
Cell recognition during neuronal development
.
Science
125
,
1271
1279
.
Hiromi
,
Y.
&
Gehring
,
W. J.
(
1987
).
Regulation and function of the Drosophila segmentation gene fushi tarazu
.
Cell
50
,
963
974
.
Ingham
,
P. XV
. (
1988
).
The molecular genetics of embryonic pattern formation in Drosophila
.
Nature, Land.
335
,
25
33
.
Jan
,
L. Y.
&
Jan
,
Y. N.
(
1982
).
Antibodies to horseradish peroxidase as specific neuronal markers in Drosophila and grasshopper embryos
.
Proc. natn. Acad. Sci. U.S.A.
79
,
2700
2704
.
Jurgens
,
G.
,
Lehmann
,
R.
,
Schardin
,
M.
&
NÜsslein-Volhard
,
C.
(
1986
).
Segmental organization of the head in the embryo of Drosophila melanogaster
.
Roux’s Arch, devl Biol.
195
,
359
377
.
Jurgens
,
G.
(
1987
).
Segmental organization of the tail region in the embryo of Drosophila melanogaster
.
Roux’s Arch, devl Biol.
196
,
141
157
.
Klingensmith
,
J.
,
Noll
,
E.
&
Perrimon
,
N.
(
1989
).
The segment polarity phenotype of Drosophila involves differential tendencies toward transformation and cell death
.
Devl Biol, (in press)
.
Klinger
,
M.
,
Erdelyi
,
M.
,
Szabad
,
J.
&
NÜsslein-Volhard
,
C.
(
1988
).
Function of torso in determining the terminal anlagen of the Drosophila embryo
.
Nature, Land.
335
,
275
277
.
Konrad
,
K. D.
,
Engstrom
,
L.
,
Perrimon
,
N.
&
Mahowald
,
A. P.
(
1985
).
Genetic analysis of oogenesis and the role of maternal gene expression in early development
. In
Developmental Biology: A Comprehensive Synthesis, vol. 1. Oogenesis
(ed. by
L.
Brower
), pp.
577
617
. Plenum publisher.
Lawrence
,
P. A.
,
Johnston
,
P.
,
Macdonald
,
P.
&
Struhl
,
G.
(
1987
).
Borders of parasegments in Drosophila embryos are delimited by the fushi tarazu and even-skipped genes
.
Nature, Lond.
328
,
440
442
.
Lindsley
,
D. L.
&
Grell
,
E. H.
(
1968
).
Genetic variations of Drosophila melanogaster. Carnegie Institution of Washington Publ. No.
627
.
Lehmann
,
R.
&
NÜsslein-Volhard
,
C.
(
1986
).
Abdominal segmentation, pole cell formation, and embryonic polarity require the localized activity of oskar, a maternal gene in Drosophila
.
Cell
47
,
141
152
.
Lohs-Schardin
,
M.
,
Cremer
,
C.
&
NÜsslein-Volhard
,
C.
(
1979
).
A fate map for the larval epidermis of Drosophila melanogaster: Localized cuticle defects following irradiation of the blastoderm with an ultraviolet laser microbeam
.
Devi Biol.
73
,
239
255
.
Mahowald
,
A. P.
,
Caulton
,
J. H.
&
Gehring
,
W. J.
(
1979
).
Ultrastructural studies of oocytes and embryos derived from female flies carrying the grandchildless mutation in Drosophila subobscura
.
Devi Biol.
69
,
118
132
.
Mahowald
,
A. P.
&
Hardy
,
P. A.
(
1985
).
Genetics of Drosophila embryogenesis
.
Ann. Rev. Gen.
19
,
149
177
.
Mlodzjk
,
M.
,
Demontrion
,
C. M.
,
Hiromi
,
Y.
,
Krause
,
H. M.
&
Gehring
,
W. J.
(
1987
).
The influence on the blastoderm fate map of the maternal-effect genes that affect the antero-posterior pattern in Drosophila
.
Genes & Dev.
1
,
603
614
.
NÜsslein-Volhard
,
C.
,
Wieschaus
,
E.
&
Jurgens
,
G.
(
1982
).
Segmentierung. Drosophila Eine genetische analyse
. In
Verhandlungen der deutschen Zoologischen Gesellschaft,
pp.
91
104
. Gustav Fisher Verlag.
NÜsslein-Volhard
,
C.
,
Frohnhofer
,
H. G.
&
Lehmann
,
R.
(
1987
).
Determination of anteroposterior polarity in Drosophila
.
Science
238
,
1675
1681
.
Perrimon
,
N.
(
1984
).
Clonal analysis of dominant female sterile, germline-dependent mutations in Drosophila melanogaster
.
Genetics
108
,
927
939
.
Perrimon
,
N.
,
Engstrom
,
L.
&
Mahowald
,
A. P.
(
1984
).
The effect of zygotic lethal mutations on female germ-line functions in Drosophila
.
Devi Biol.
105
,
404
414
.
Perrimon
,
N.
,
Engstrom
,
L.
&
Mahowald
,
A. P.
(
1985
).
A pupal lethal mutation with a paternally influenced maternal effect on embryonic development in Drosophila melanogaster
.
Devi Biol.
110
,
480
491
.
Perrimon
,
N.
,
Mohler
,
D.
,
Engstrom
,
L.
&
Mahowald
,
A. P.
(
1986
).
X-linked female sterile loci in Drosophila melanogaster
.
Genetics
113
,
695
712
.
Perrimon
,
N.
&
Mahowald
,
A. P.
(
1988
).
Maternal contributions to early development in Drosophila. In Primers in Developmental Biology,
(ed.
G.
Malacinski
). pp.
305
328
. MacMillan publisher.
Schupbach
,
T.
&
Wieschaus
,
E.
(
1986
).
Maternal effect mutations affecting the segment pattern of Drosophila
.
Roux’s Arch devl. Biol.
195
,
302
307
.
Smouse
,
D. T.
,
Goodman
,
C.
,
Mahowald
,
A. P.
&
Perrimon
,
N.
(
1988
).
polyhomeotic: gene required for the embryonic development of axon pathways in the central nervous system of Drosophila
.
Genes & Dev,
2
,
830
842
.
Sonnenblick
,
B. P.
(
1950
).
The early embryology of Drosophila melanogaster. In Biology of Drosophila
(ed.
M.
Demerec
).
Wiley, New York
, pp.
62
167
.
Steller
,
H.
,
Fischbach
,
K.-F.
&
Rubin
,
G. M.
(
1987
).
disconnected: A locus required for neuronal pathway formation in the visual system of Drosophila
.
Cell
50
,
1139
1153
.
Strecker
,
T. R.
,
Merriam
,
J. R.
&
Lengyel
,
J. A.
(
1988
).
Graded requirements for the zygotic terminal gene, tailless, in the brain and tail region of the Drosophila embryo
.
Development
102
,
721
734
.
Technau
,
G. M.
&
Campos-Ortega
,
J. A.
(
1985
).
Fate-mapping in Drosophila melanogaster II. Injections of horseradish peroxidase in cells of the early gastrula stage
.
Roux’s Arch, devl Biol.
194
,
196
212
.
Turner
,
F. R.
&
Mahowald
,
A. P.
(
1976
).
Scanning electron microscopy of Drosophila embryogenesis. I. The structure of the egg envelopes and formation of the cellular blastoderm
.
Devl Biol.
50
,
95
108
.
Van Der Meer
,
J.
(
1977
).
Optical clean and permanent whole mount preparations for phase contrast microscopy of cuticular structures of insect larvae
.
Droso. Inf. Serv.
52
,
160
.
Wieschaus
,
E.
&
NÜsslein-Volhard
,
C.
(
1986
).
Looking at embryos. In Drosophila: a Practical Approach
edited by D. B. Roberts.
IRL Press Washington D.C
.
Winslow
,
G. M.
,
Carroll
,
S. B.
&
Scott
,
M. P.
(
1988
).
Maternal-effect genes that alter the fate map of the Drosophila blastoderm embryo
.
Devl Biol.
129
,
72
83
.