Based on the expression pattern of the segment polarity genes engrailed and wingless during the embryonic development of the larval head, we found evidence that the head of Drosophila consists of remnants of seven segments (4 pregnathal and 3 gnathal) all of which contribute cells to neuromeres in the central nervous system. Until completion of germ band retraction, the four pregnathal segment remnants and their corresponding neuromeres become arranged in an S-shape. We discuss published evidence for seven head segments and morphogenetic movements during head formation in various insects (and crustaceans).

In Drosophila, the embryonic development of the larval trunk is much better known than the development of the larval head. This is especially true for the corresponding regions of the central nervous system (CNS). Main reasons for this are that (1) the presumed underlying segmental pattern in the brain is highly derived compared to the ventral nerve cord (Hartenstein and Campos-Ortega, 1984), and (2) landmarks in the embryonic brain are scarce.

The number of segments that contribute to the head and brain of insects has been intensely discussed by morphologists (for reviews see Sharov, 1966; Rempel, 1975) using mainly coelomic cavities, neuromeres and appendages as criteria. Investigators of Drosophila generally prefer Rempel’s view of six head segments (3 pregnathal and 3 gnathal; Jürgens et al. 1986; Cohen and Jürgens, 1991; Diederich et al. 1991) but the question to whether there is a fourth pregnathal segment seems to be unsettled (see below).

Fate mapping data show that until germ band retraction massive morphogenetic movements lead to an S-shaped deflection of epidermal head structures (Technau and Campos-Ortega, 1985; Fig. 1). These data also reveal that similar movements affect the formation of the brain, but they do not further elucidate the segmental organisation of the brain.

Fig. 1.

Fate-mapping data after Technau and Campos-Ortega (1985). The symbols in (A) mark the sites of HRP injections at the beginning of gastrulation. (B) Summarizes the distribution of labeled cells after germ band retraction (stage 12/13) in the peripheral ectoderm, and (C-F) show the distribution of markings in the CNS. The lines in B ligate markings that derive from injections at 0, 25, 50, 75 and 100% VD, respectively. Note the S-shaped distribution of labelled cells at stage 12/13 that results from a linear row of injections. % EL, % egg length; % VD, % ventrodorsal diameter.

Fig. 1.

Fate-mapping data after Technau and Campos-Ortega (1985). The symbols in (A) mark the sites of HRP injections at the beginning of gastrulation. (B) Summarizes the distribution of labeled cells after germ band retraction (stage 12/13) in the peripheral ectoderm, and (C-F) show the distribution of markings in the CNS. The lines in B ligate markings that derive from injections at 0, 25, 50, 75 and 100% VD, respectively. Note the S-shaped distribution of labelled cells at stage 12/13 that results from a linear row of injections. % EL, % egg length; % VD, % ventrodorsal diameter.

To obtain evidence concerning the position, orientation, and number of the brain neuromeres we studied the expression pattern of the segment polarity genes engrailed (en) and wingless (wg) during embryonic head development. In the trunk segments, en is expressed in posterior compartments and wg in anterior compartments, thus defining the polarity of segments on a molecular level (Baker, 1987; Martinez Arias et al. 1988; van den Heuvel et al. 1989). Our detailed description of the dynamic pattern of en and wg expression in the developing head argues for 4 pregnathal (and 3 gnathal) segments with their corresponding neuromeres. Until completion of germ band retraction these segments become arranged in a refolded band which is in good agreement with the fate mapping data (Technau and Campos-Ortega, 1985; Jürgens et al. 1986). Published data from a variety of insects (and crustaceans) give further support for 4 pregnathal segments and the occurrence of similar morphogenetic movements in other Mandibulata.

Fly stocks

In our experiments we used a wild-type Oregon R stock. For some of the double labelling experiments we used the transformant line ryxho25 (Hama et al. 1990), which carries an engrailed-lacZ fusion gene inserted at the engrailed locus, the stock wg1L114 (Nüsslein-Vollhard et al. 1984) which carries a temperature-sensitive allele affecting the secretion of the wg antigen (Gonzalez et al. 1991), and the transformant lines (CyO, wgen11/sli and wg17en40cP1/CyO) with lacZ insertions in the wingless gene (for which we are indebted to N. Perrimon, Harvard).

Antibodies

We used the monoclonal antibody 4D9 (Patel et al. 1989) which recognizes the antigens of the closely related engrailed and invected genes (1/3 dilution; in the following we refer to both as engrailed staining), a polyclonal anti-β-gal (Cappel, 1/7000 dilution, preabsorbed), an anti-wingless antiserum (González et al. 1991; 1/2000 dilution, preabsorbed) and a monoclonal anti-BrdU antibody (Becton & Dickinson, 1/100 dilution). Peroxidase-coupled goat-anti-mouse (IgG), goat-anti-rat (IgG), and goat-antirabbit (IgG) were used as secondary antibodies (Dianova, 1/500 dilution). For detection of the anti-wingless antibodies we used a biotinilated goat-anti-rat antibody with the Vectastain-ABC-Kit Elite (Camon).

Staining procedures

Embryos were dechorionated with 50% commercial bleach for 4 minutes and briefly washed in water. Excess liquid was blotted away and the embryos were subsequently fixed in a 1:1 mixture of n-heptane and 4% formaldehyde or paraformaldehyde in PEM buffer (0.1 M PIPES, 2 mM MgSO4, 1 mM EGTA, pH 6.9) for 25 minutes on a shaker. The aqueous solution was then replaced by methanol. After 15 seconds of vigorous shaking the devitellinized embryos had sunk to the bottom. The embryos were washed three times with methanol, three times for 20 minutes with PBT (0.13 M NaCl, 7 mM Na2HPO4, 3 mM NaH2PO4, 0.1% Triton X-100) and blocked for 30 minutes with 10% goat or lamb serum in PBT. Incubation with the first antibody was carried out overnight at 4°C, followed by three washes with PBT and the incubation with the secondary antibody for 3 hours at room temperature. After three more washes (10 minutes each) the samples were stained for HRP with diaminobenzidine (0.1% DAB, 0.02% H2O2 in 0.2 M phosphate buffer) for 10 minutes, dehydrated in an ethanol series and mounted in Araldite (Serva). Following a suggestion of E. R. Macagno (Columbia University, New York), embryos and larval brains were sucked in borosilicate capillaries (Hilgenberg) with an inner diameter of 0.2 mm. This allowed the rotation of the samples under the compound microscope.

For the BrdU experiments, freshly hatched larvae were fed on a diet containing 0.1% BrdU (stock solution 33 mM in 40% ethanol; Truman and Bate, 1988) for 1-to-4 hours. The samples were fixed with paraformaldehyde and stained first with antiengrailed or anti-β-gal antibodies. After the DAB reaction and two washes in PBT, the larval brains were treated with 2 N HCl in PBT for 1 hour, blocked for 30 minutes and stained with the antiBrdU antibody as described above. We added 0.08% NiCl2 to the DAB-solution which resulted in an almost black staining that was easily distinguished from the normal brown stain.

Embryos of the wg1L114 stock were raised at 27°C, fixed in paraformaldehyde, and stained first for en and second for wg antigens. Embryos of the transformant lines carrying a lacZ insertion at the wg locus were fixed in formaldehyde before double staining.

Note on figures

In all figures we used the following symbols to identify certain groups of en-expressing cells: ▸ en intercalary spot; ▹ en secondary head spot; → en antennal stripe; ▹ en in anterior dorsal hemispheres; ▸ en antennal spot; ▹ en clypeolabral spot; en head spot; > en cells of uncertain origin.

Figs 210 each show different views of the same individual. Anterior is to the left or to the top. T1 marks the first thoracic segment or neuromere.

Fig. 2.

Wild-type embryos at stage 6/7 (A) and 8 (B-D) stained with anti-en antibody. Anterior to the left, lateral views. At stage 6/7 the most anterior stripe corresponds to the mandibular segment. At stage 8 the en antennal stripe has formed (arrow in C and D) at the posterior margin of the procephalic lobe. T1 marks the first thoracic segment. Scale bar, 20 μm.

Fig. 2.

Wild-type embryos at stage 6/7 (A) and 8 (B-D) stained with anti-en antibody. Anterior to the left, lateral views. At stage 6/7 the most anterior stripe corresponds to the mandibular segment. At stage 8 the en antennal stripe has formed (arrow in C and D) at the posterior margin of the procephalic lobe. T1 marks the first thoracic segment. Scale bar, 20 μm.

engrailed (en) expression in the head

We found five “centers” of en expression anterior to the mandibular segment which appear in the following sequence from stage 8 on (stages after Campos-Ortega and Hartenstein, 1985): the “en antennal stripe”, the “en head spot”, the “en intercalary spot”, the “en expression in the anterior dorsal hemispheres” and the “en expression in the clypeolabrum”. Subsequently, the en antennal stripe and the en head spot split in two groups of cells, thus generating in total 7 en spots (Figs 210; Table 1). In this way the “en antennal spot” separates from the en antennal stripe and the “en secondary head spot” from the en head spot. For the terminology of spots we followed Cohen and Jürgens (1990); new names were only introduced for spots not mentioned by these authors.

Table 1.

Expression of en/inv

Expression of en/inv
Expression of en/inv

The en antennal stripe appears at stage 8 (Fig. 2). These cells remain largely in the peripheral ectoderm and can be traced until stage 17 (Figs 3-10). They migrate towards the dorsal organ and contribute axons to the antennal nerve (Fig. 11). The most anterior cells of this stripe (corresponding to the ventral cells of the trunk stripes) delaminate during stage 10/11 (Fig. 4, and Marcos González Gaitán, personal communication) to become part of the brain and can be traced until early larval stages (Fig. 11). We refer to these cells as en antennal spot (Fig. 4). At stage 12 these cells stain only faintly, but expression again becomes more prominent from stage 13 on (Figs 6-10). Thus, we cannot confirm the observation of Diederich et al. (1991) that “all of the en-expressing cell-clusters of the preoral head, with the exception of those associated with the clypeolabral lobe, appear to delaminate from the ectodermal surface”. Our data also suggest that the en-expressing cells of the en antennal spot are identical with those that Diederich et al. (1991, in their Fig. 1B) attributed to the labral neuromere (see also Discussion).

Fig. 3.

Wild-type embryos at stage 9 (A–E) and 10 (F–J) stained with anti-en antibody. Anterior to the left or to the top (D, E, H, I, J). Lateral views (A, B, F, G) are focused on the peripheral ectoderm in (B) and (G). Dorsal views are shown at the level of the en head spot (D, H), the en antennal stripe (E, I) and the en intercalary spot (J). At stage 9 the en head spot (short arrow) shows up and at stage 10 the en intercalary spot (arrowhead). In (A) and (F) the arrows indicate the position of the en head spot and the en antennal stripe (long arrow) that are both out of focus. T1 marks the first thoracic segment. Scale bar, 20 μm.

Fig. 3.

Wild-type embryos at stage 9 (A–E) and 10 (F–J) stained with anti-en antibody. Anterior to the left or to the top (D, E, H, I, J). Lateral views (A, B, F, G) are focused on the peripheral ectoderm in (B) and (G). Dorsal views are shown at the level of the en head spot (D, H), the en antennal stripe (E, I) and the en intercalary spot (J). At stage 9 the en head spot (short arrow) shows up and at stage 10 the en intercalary spot (arrowhead). In (A) and (F) the arrows indicate the position of the en head spot and the en antennal stripe (long arrow) that are both out of focus. T1 marks the first thoracic segment. Scale bar, 20 μm.

Fig. 4.

Wild-type embryos at stage 11 stained with anti-en antibody. Anterior to the left or to the top (C–E). (A) and (B), lateral views; (CD), dorsal views at the level of the en head spot (arrow in C), the en antennal spot (triangle in D), and en intercalary spot (arrowhead in E). en has become visible in the clypeolabrum (hollow triangle in A) and in the anterior dorsal hemispheres (filled triangle in C). Cells at the anterior end of the en antennal stripe have segregated (triangle in B, D). Note the position of the en head spot, the en antennal spot, and the en intercalary spot with respect to the mandibular segment (B). T1 marks the first thoracic segment, Md the mandibular neuromere. Scale bar, 20 μm.

Fig. 4.

Wild-type embryos at stage 11 stained with anti-en antibody. Anterior to the left or to the top (C–E). (A) and (B), lateral views; (CD), dorsal views at the level of the en head spot (arrow in C), the en antennal spot (triangle in D), and en intercalary spot (arrowhead in E). en has become visible in the clypeolabrum (hollow triangle in A) and in the anterior dorsal hemispheres (filled triangle in C). Cells at the anterior end of the en antennal stripe have segregated (triangle in B, D). Note the position of the en head spot, the en antennal spot, and the en intercalary spot with respect to the mandibular segment (B). T1 marks the first thoracic segment, Md the mandibular neuromere. Scale bar, 20 μm.

The en head spot (visible from stage 9 on; Fig. 3) becomes completely integrated into the CNS. At stage 13 it is a stripe that crosses each brain hemisphere transversely at its ventral side (Figs 5, 6). The two outermost cells of each stripe are larger than the inner ones. Until the end of embryogenesis en expression in the en head spot becomes restricted to only two cells (Fig. 10). At stage 13/14, a group of cells detaches from the en head spot and we refer to them as en secondary head spot (Figs 7, 8, 9). This splitting has not been noticed by Diederich et al. (1991).

Fig. 5.

Wild-type embryo at stage 12 stained with anti-en antibody. Anterior to the left. (B), (D), (F) and (H) show dorsal views at the level of en expression in the dorsal hemispheres (triangle), the en antennal stripe (long arrow), the en head spot (short arrow), and the en intercalary spot (arrowhead), respectively. The same symbols are used in the lateral views (A, C, E, G). The hollow triangle in (A) points to the en expression in the clypeolabrum. Note that the en head spot lies at the ventral side of the brain hemisphere and that the en intercalary spot now lies at the posterior margin of both connectives between the supraesophageal and the subesophageal ganglia. T1 marks the first thoracic neuromere. Scale bar, 20 μm.

Fig. 5.

Wild-type embryo at stage 12 stained with anti-en antibody. Anterior to the left. (B), (D), (F) and (H) show dorsal views at the level of en expression in the dorsal hemispheres (triangle), the en antennal stripe (long arrow), the en head spot (short arrow), and the en intercalary spot (arrowhead), respectively. The same symbols are used in the lateral views (A, C, E, G). The hollow triangle in (A) points to the en expression in the clypeolabrum. Note that the en head spot lies at the ventral side of the brain hemisphere and that the en intercalary spot now lies at the posterior margin of both connectives between the supraesophageal and the subesophageal ganglia. T1 marks the first thoracic neuromere. Scale bar, 20 μm.

Fig. 6.

Wild-type embryo at stage 13 stained with anti-en antibody. Anterior to the left. (B), (D), (F) and (H) show dorsal views. en head spot: small arrow in (A), (C), (D), (E) and (G); en expression in the anterior dorsal hemispheres: triangle in (B); en antennal spot: small triangle in (C) and (F); en antennal stripe: long arrow in (E) and (F); en expression in the clypeolabrum: open triangle in (G) and (H); en intercalary spot: arrowhead in (G) and (H). Note that the en head spot now forms a stripe crossing each hemisphere at its ventral side. The two outermost cells of each stripe are larger than the inner ones. T1 marks the first thoracic segment. Scale bar, 20 μm.

Fig. 6.

Wild-type embryo at stage 13 stained with anti-en antibody. Anterior to the left. (B), (D), (F) and (H) show dorsal views. en head spot: small arrow in (A), (C), (D), (E) and (G); en expression in the anterior dorsal hemispheres: triangle in (B); en antennal spot: small triangle in (C) and (F); en antennal stripe: long arrow in (E) and (F); en expression in the clypeolabrum: open triangle in (G) and (H); en intercalary spot: arrowhead in (G) and (H). Note that the en head spot now forms a stripe crossing each hemisphere at its ventral side. The two outermost cells of each stripe are larger than the inner ones. T1 marks the first thoracic segment. Scale bar, 20 μm.

Fig. 7.

Wild-type embryo at stage 14 stained with anti-en antibody. Anterior to the left. (C) and (D) show dorsal views at the level of the en head spot and of the en antennal stripe, respectively. The small hollow triangles in (C) point to cells of the en secondary head spot which separate from the en head spot cells at stage 13/14 and are not visible as a separate spot at earlier stages. Other symbols as in Fig. 6. Note that the cells of the en intercalary spot are now in contact with en-expressing cells of the mandibular neuromere which makes it difficult to distinguish between both groups. The en antennal spot has enlarged compared to the preceding stage. T1 marks the first thoracic neuromere. Scale bar, 20 μm.

Fig. 7.

Wild-type embryo at stage 14 stained with anti-en antibody. Anterior to the left. (C) and (D) show dorsal views at the level of the en head spot and of the en antennal stripe, respectively. The small hollow triangles in (C) point to cells of the en secondary head spot which separate from the en head spot cells at stage 13/14 and are not visible as a separate spot at earlier stages. Other symbols as in Fig. 6. Note that the cells of the en intercalary spot are now in contact with en-expressing cells of the mandibular neuromere which makes it difficult to distinguish between both groups. The en antennal spot has enlarged compared to the preceding stage. T1 marks the first thoracic neuromere. Scale bar, 20 μm.

Fig. 8.

Wild-type embryo at stage 15 stained with anti-en antibody. Anterior to the left. (B) and (D) show dorsal views at the level of the dorsal hemispheres and the en head spot, respectively. Symbols as in Figs 6 and 7. T1 marks the first thoracic neuromere (A) or segment (B). Scale bar, 20 μm.

Fig. 8.

Wild-type embryo at stage 15 stained with anti-en antibody. Anterior to the left. (B) and (D) show dorsal views at the level of the dorsal hemispheres and the en head spot, respectively. Symbols as in Figs 6 and 7. T1 marks the first thoracic neuromere (A) or segment (B). Scale bar, 20 μm.

Fig. 9.

Wild-type embryo at stage 16 stained with anti-en antibody. Anterior to the left. Dorsal views are shown at the level of the dorsal hemispheres (B), the en head spot (D) and the connectives (F). The angle in (E) and (F) marks an en-expressing cell of unknown origin at the posterior margin of each connective. Other symbols as in Figs 6 and 7. T1 marks the first thoracic neuromere (A) or segment (B). In several cases we observed a splitting of the en antennal spot in two groups of cells at this stage as shown in (C). Scale bar, 20 μm.

Fig. 9.

Wild-type embryo at stage 16 stained with anti-en antibody. Anterior to the left. Dorsal views are shown at the level of the dorsal hemispheres (B), the en head spot (D) and the connectives (F). The angle in (E) and (F) marks an en-expressing cell of unknown origin at the posterior margin of each connective. Other symbols as in Figs 6 and 7. T1 marks the first thoracic neuromere (A) or segment (B). In several cases we observed a splitting of the en antennal spot in two groups of cells at this stage as shown in (C). Scale bar, 20 μm.

Fig. 10.

Wild-type embryo at stage 17 (A–D) and hatching larva (E–H) stained with anti-en antibody. Anterior to the left. Lateral views and dorsal views (D, G) at the level of the en head spot. The thin arrow in (B) points to the dorsal organ that stains unspecifically. Other symbols as in Figs 6 and 7. Note that the en headspot reduces to two cells on each side (F, G) in the hatching larva. The cells of the en antennal stripe contribute to the dorsal organ (arrow in B, see also Figs 8C and 11). T1 marks the first thoracic neuromere. Scale bar, 20 μm.

Fig. 10.

Wild-type embryo at stage 17 (A–D) and hatching larva (E–H) stained with anti-en antibody. Anterior to the left. Lateral views and dorsal views (D, G) at the level of the en head spot. The thin arrow in (B) points to the dorsal organ that stains unspecifically. Other symbols as in Figs 6 and 7. Note that the en headspot reduces to two cells on each side (F, G) in the hatching larva. The cells of the en antennal stripe contribute to the dorsal organ (arrow in B, see also Figs 8C and 11). T1 marks the first thoracic neuromere. Scale bar, 20 μm.

The en intercalary spot (visible from stage 10 onwards; Fig. 3) also completely segregates into the CNS. At stage 12 these cells are located at the posterior margin of each connective between the supraand the subesophageal ganglion. At stage 14 the en intercalary spot touches enexpressing cells of the mandibular neuromere and at stage 16 the cells of the en intercalary spot switch off expression of the en antigen.

From stage 11 onwards we noticed en expression in the area of the dorsal clypeolabral suture. This staining is “fuzzy” but becomes more distinct at later stages. We refer to this staining which has not been mentioned by other authors as en expression in the anterior dorsal hemispheres (Figs 49).

en expression in the clypeolabrum starts at early stage 11 (Figs 410). These cells do not contribute to the CNS (see also Diederich et al. 1991). The clypeolabral expression is median and can be slightly dumb-bell shaped at stage 11 but is unpaired, in contrast to the patches of en expression mentioned above. At later states it is found in the roof of the pharynx. We did not observe a bilateral splitting of this patch in two spots as mentioned by Diederich et al. (1991) for stage 15.

All patches of en expression migrate and undergo changes in size (cell number) and shape (arrangement of stained cells) as documented in detail in Figs 210 and in Table 1.

For the following reasons we think that the traced movements of the en-expressing cells are real and do not simply reflect turning on and off of the en gene. (1) We analysed many more intermediate stages than those shown in Figs 2-10. (2) When cells turn off en expression towards the end of embryogenesis, weak staining can often still be recognized for some time. When using an en-lacZ strain we could follow cells that turned off en expression (and their descendants) even longer because of the perdurance of the lacZ product in the cellular cytoplasm (Monsma et al. 1988; Vincent and O’Farrell, 1992; Martinez Arias and Technau, unpublished data; see also Fig. 11 and Discussion). In both cases these cells followed the inferred movement of the enexpressing cells. (3) The traced morphogenetic movements are in good agreement with fate-mapping data (Fig. 1; Technau and Campos-Ortega, 1985; Jürgens et al. 1986).

Fig. 11.

Brain of transformant en-lacZ larva (ryxho25) two hours after hatching, stained with anti-β-gal antibody. Lateral view, anterior to the left. The arrow points to the en head spot cells and their descendants and the triangle to the cells of the en antennal spot and their descendants. Note staining of the antennal nerve (an) due to en-positive cells in the dorsal organ. Scale bar, 13 μm.

Fig. 11.

Brain of transformant en-lacZ larva (ryxho25) two hours after hatching, stained with anti-β-gal antibody. Lateral view, anterior to the left. The arrow points to the en head spot cells and their descendants and the triangle to the cells of the en antennal spot and their descendants. Note staining of the antennal nerve (an) due to en-positive cells in the dorsal organ. Scale bar, 13 μm.

wingless (wg) expression in the head

To find out whether patches of en expression in the pregnathal head can be attributed to segments, and eventually to learn about their orientation, we double stained wg1L114 (Fig. 12) and transformant wg lacZ (CyO, wgen11/sli and CyO/wg17en40cP1; data not shown) embryos against en and wg or β-galactosidase (lacZ-product). In heterozygous wg1L114 embryos raised at non-permissive temperatures (25°C or higher) the secretion of the wg protein is partly interrupted (González et al. 1991). The accumulation of wg protein inside the cells thus allows for a stronger signal when staining with antiserum. The embryos shown in Fig. 12 are phenotypically wild-type, as can be seen from the normal en expression, but are probably heterozygous for the wg1L114 allele because of the stronger wg signal compared to homozygous wild-type embryos. Fig. 13 schematically shows en and wg expression at stages 10/11 and late stage 12.

Fig. 12. Embryos heterozygous for the mutation wg1L114 (see text) double stained against en and wg. Anterior to the left. (A) Stage 10, only the head is shown from the latero-ventral side. en and wg expression in the ocular segment are out of focus. Mx, Md, Ic, An, Oc, and Lr: en and wg to maxillar, mandibular, intercalary, antennal, ocular, and labral segment, respectively. The arrowhead points to the en intercalary spot. The en expression in the anterior dorsal hemispheres is not yet visible at this stage (see text). Bar, 20 μm. (B, C) Early stage 12, lateral and dorsal view. Note the weak en expression in the anterior dorsal hemispheres which becomes more prominent in following stages (see also Figs 4, 5, 7, 8). The large arrow points to the wg head blob, other symbols as in Figs 6 and 7. Bar, 50 μm.

Fig. 13.

Scheme of the Drosophila head indicating expression of en (hatched) and wg (crosshatched) at embryonic stage 10/11 (A) and late stage 12 (B). The dotted line indicates the alignment of segments. Note that at stage 10/11 every patch of en expression has its wg counterpart. In (B) expression of en and wg in the peripheral ectoderm of the antennal and gnathal segments has been omitted. An, to antennal segment; ant s, antennal spot; a s, antennal stripe; cl, clypeolabral; D, deutocerebrum; d h, dorsal hemisphere; fg, foregut; hb, head blob; hs, head spot; Ic, to intercalary segment; ic s, intercalary spot; La, to labial segment; Lr, to labral segment; Mx, maxillary segment; Oc, to ocular segment; ol, optic lobe; P, protocerebrum; T, tritocerebrum.

Fig. 13.

Scheme of the Drosophila head indicating expression of en (hatched) and wg (crosshatched) at embryonic stage 10/11 (A) and late stage 12 (B). The dotted line indicates the alignment of segments. Note that at stage 10/11 every patch of en expression has its wg counterpart. In (B) expression of en and wg in the peripheral ectoderm of the antennal and gnathal segments has been omitted. An, to antennal segment; ant s, antennal spot; a s, antennal stripe; cl, clypeolabral; D, deutocerebrum; d h, dorsal hemisphere; fg, foregut; hb, head blob; hs, head spot; Ic, to intercalary segment; ic s, intercalary spot; La, to labial segment; Lr, to labral segment; Mx, maxillary segment; Oc, to ocular segment; ol, optic lobe; P, protocerebrum; T, tritocerebrum.

At stage 10/11, the en intercalary spot, the en antennal stripe, the en head spot and the en expression in the anterior dorsal hemispheres, have their wg counterpart anterior to them (see Discussion) when using spatial vicinity as a criteria (Figs 12, 13; see also Cohen and Jürgens, 1991). The wg labral spot (wg lr) remains in the peripheral ectoderm of the clypeolabrum. The wg head blob (wg hb) becomes incorporated in the brain but cells of the optic anlagen seem to be devoid of wg expression. The wg antennal stripe (wg a s) is limited to the peripheral ectoderm. However, until stage 10 it is in contact with the wg head blob (Baker, 1988, van den Heuvel et al. 1989). Expression in the wg intercalary spot (wg Ic) is turned off at stage 11 (Baker, 1988; van den Heuvel et al. 1989), but staining embryos of transformant lines carrying a lacZ insertion at the wg locus (CyO, wgen11/sli and wg17en40cP1/CyO) reveals that these cells segregate into the CNS (data not shown). There is also a median spot of wg expression in the foregut (Baker, 1988; van den Heuvel et al. 1989) which probably corresponds to the median en clypeolabral spot (Figs 12, 13). The gap between this wg expression and the median en clypeolabral spot increases during development. This might be due to cell proliferation and intercalation in the course of morphogenetic movements during stomodeum formation (Hartenstein and Campos-Ortega, 1985; Technau and Campos-Ortega, 1985).

Cells at the ventro-anterior side of the en head spot contribute to the antennal lobe cortex

In the brain of the freshly hatched larvae, five proliferating cells can be marked by BrdU. The fates of these five neuroblasts and their descendants are known (Ito and Hotta, 1992; Prokop et al., 1991). Four of them contribute to the mushroom body cortex. The fifth neuroblast and its descendants which lie laterally become incorporated in the cortex of the antennal lobe and hence belong to the deutocerebrum. This group of cells lies in the area of the en head spot. To find out if it is (partly) identical with the en head spot, we fed freshly hatched larvae for 1-to-4 hours on a diet containing BrdU, which becomes incorporated in the DNA of the five proliferating neuroblasts and their descendants in each brain hemisphere. We then double stained larval brains against en and BrdU (wild-type) or anti-β-gal and BrdU (ryxho25). This procedure revealed that the fifth (lateral) BrdU-labeled neuroblast is not identical with the cells of the en head spot or their descendants but is in direct contact with these cells (Fig. 14; n = 10, no exceptions). The descendants of en-expressing cells are recognizable due to the perdurance of the lacZ product in the transformant en-lacZ line (Monsma et al. 1988; Vincent and O’Farrell, 1992; Martinez Arias and Technau, unpublished data). As expected, the patterns of en expression in the en head spot as revealed by anti-β-gal and anti-en are very similar until stage 13, because only then do some cells of the en head spot turn off en expression so that more cells are stained by anti-β-gal in ryxho25 than by 4D9 in wild-type embryos at older stages.

Fig. 14. CNS of larvae fed on a BrdU diet for 3.5 and 1.5 hours after hatching. Lateral view, anterior to the left. (A) Wild-type 3.5 hours after hatching, double stained against en (brown) and BrdU (black); (B) en-lacZ ryxho25, 1.5 hours after hatching stained against β-gal (brown) and BrdU (black). The BrdU-marked lateral neuroblast (asterisk) and its descendants are later found in the cortex of the antennal lobe (Ito and Hotta, 1992; unpublished data). The arrow points to the en head spot. It is larger in (B) because of the perdurance of the lacZ product. Note the halo (white arrows) separating the BrdU-marked cells from the en head spot cells and their descendants (black arrow). br, brain; vn, ventral nerve cord. Scale bar, 20 μm.

The position of the BrdU-marked cells is always at the anterior ventral side of the en head spot cells (Fig. 14). In embryos stained with anti-β-gal and anti-BrdU (Fig. 14B) the BrdU-marked cells exhibit a halo that separates them from the en head spot. This suggests that the BrdU-marked cells do not express lacZ and, therefore, might not derive from the en head spot cells.

Our detailed description of the dynamic pattern of expression of en and wg contributes towards clarification of the segmental organization of the head. Furthermore, it provides useful reference points for future investigations of brain development.

The number and arrangement of neuromere remnants in the supraesophageal ganglion of Drosophila

We studied the spatial and temporal expression pattern of en (see also DiNardo et al. 1985 and Diederich et al. 1991) and wg (see also Baker, 1988 and van den Heuvel, 1989) in relation to each other during embryonic development of the head to obtain evidence concerning the number, location and orientation of brain neuromeres and their corresponding segments in the peripheral ectoderm. As shown below, our data support the theory that there are four pregnathal head segments which are aligned in an S-shape. These conclusions differ from those drawn by Diederich et al. (1991) who, based on the expression of en, argue for only three linearly aligned pregnathal segments (for discrepancies between their observations and ours, see first chapter of Results).

Four of five “centers” of en expression in the pregnathal head contribute cells to the brain. These are: the en intercalary spot, the en antennal stripe which contributes the cells of the en antennal spot, the en head spot, and the en expression in the anterior dorsal hemispheres (Fig. 13). These patches are good candidates for segmental markers because normally, a segment contributes a neuromere to the central nervous system.

We attribute these patches of en expression to four pregnathal neuromeres as follows.

The en intercalary spot is attributed to the intercalary neuromere because of its position just anterior to the mandibular neuromere.

The en antennal spot is attributed to the antennal segment for two reasons. (1) At stage 10/11, these cells segregate from the anterior end (corresponding to the ventral side of trunk stripes) of the en antennal stripe which contributes cells to the dorsal organ. (2) In buttonhead mutants (btd XG and btd XO; Jürgens et al. 1984) which lack wg and en expression in the mandibular, the intercalary, and the antennal segments (Cohen and Jürgens, 1990), the en antennal spot does not show up, either (Schmidt-Ott and González Gaitán, unpublished observation).

The en head spot is attributed to the ocular segment (to which the optic lobe fuses secondarily; Campos-Ortega and Hartenstein, 1985; own observations). We believe that the cells of the en secondary head spot also belong to the ocular segment because they detach from the en head spot at stage 13/14.

en expression in the anterior dorsal hemispheres is attributed to the labral neuromere as it has its wg counterpart on the clypeolabrum.

At stage 12/13 the four pregnathal neuromeres and their epidermal counterparts are arranged in an S-shape (Fig. 13). Several lines of evidence support this assumption. (1) The suggested alignment of segments is in excellent agreement with published fate-mapping data (Fig. 1; Technau and Campos-Ortega, 1985; Jürgens et al. 1986). Fig. 1A shows sites of HRP injection at the onset of gastrulation. After staining these embryos at stage 12/13, the markings were found in the peripheral ectoderm (Fig. 1B) and/or in the brain (Fig. 1C-F), revealing the S-shaped arrangement of cells that had been in a line at the onset of gastrulation (Technau and Campos-Ortega, 1985). Our observation that cells of the en head spot (ocular neuromere) lie in the immediate vicinity of cells that contribute to the cortex of the antennal lobe (Fig. 14) is also consistent with an S-shaped arrangement of the brain neuromeres: these presumptive cells of the cortex of the antennal lobe lie ventro-anterior (i.e. ventro-posterior with respect to the polarity of the ocular neuromere!) to the en head spot cells, and they do not express lacZ which suggests that these cells do not derive from the en head spot cells (Fig. 14B). We suggest that the segmental border runs just between both groups. In fact, in the adult brain the BrdU-marked cells shown in Fig.14 lie laterally in the antennal lobe cortex towards the cortex of the optic lobe (Ito and Hotta, 1992; Prokop and Technau, unpublished data). (2) At stage 13 the en head spot forms a horizontal stripe which transversely crosses each hemisphere on its ventral side (Figs 6A, D; 5E). Analogous to the situation in the trunc, we would expect that this stripe of cells marks the posterior segment border of the neuromere. This is only the case if the neuromeres are arranged as shown in Fig. 13B. (3) When assuming an Sshaped arrangement of head segment remnants (Fig. 13), wg counterparts to the en intercalary spot, the en antennal stripe and the en head spot, and en expression in the anterior dorsal hemispheres are located anterior to each en patch analogous to the situation in the trunk (Baker, 1987; Martinez Arias et al. 1988; van den Heuvel et al. 1989). The corresponding wg expression is also a good confirmation that the mentioned en spots mark different segments (neuromeres). (4) Four pregnathal head segments and their Sshaped arrangement are also in very good agreement with ectopic en expression in the head of patched null mutants (González Gaitán and Schmidt-Ott, unpublished data).

Expression of en and wg in the foregut

en expression in the clypeolabrum is median and unpaired. During further embryonic development this patch becomes part of the dorsal pharynx. wg expression is found in the invaginating stomodeum (Baker, 1988; van den Heuvel et al. 1989). We think that this “pair” of en and wg expression does not mark a true segment because these cells do not contribute to the CNS and because en and wg are expressed in an analogous manner in the non segmental hindgut (DiNardo et al. 1985; Baker, 1988; van den Heuvel et al. 1989).

The number of segments in the insect head

Our data suggest that 4 pregnathal segment remnants exist in the Drosophila head. How does this assumption fit with comparative data? Two lines of evidence support the assumption of 4 pregnathal segments in the Mandibulata. (1) In several species of insects and in a crustacean, vestiges of two pairs of coelomic cavities have been found anterior to the antennal segment (Wiesmann, 1926,1; Miller, 1940; Nair, 1949; Sharov, 1966; Rohrschneider, 1968). (2) Wada (1966a,b,c) found seven “morphogenetic units” in addition to the stomodeum in the head of the Tachycines embryo (Saltatoria) that he revealed by injuring the early embryos and analysing the malformations (doubling or absence of defined structures) of the hatching larvae. His first (i.e. most anterior in the fate map) unit contains the clypeolabrum and the frons, the second is formed by the eyes and neighbouring epidermal structures, the mushroom bodies, the central body and the lobi optici of the brain. The following five units refer to neuronal and epidermal structures of the antennal, the intercalary, and the three gnathal segments, respectively. Wada (1966a,b) did not find a part of the brain affected together with the first unit but this might be due to the lack of a known morphological marker. Based on these arguments (en and wg in Drosophila, teratology and coeloms in different insects and a crustacean) we suggest seven head segments in the basic body plan of insects (and other Mandibulata): the labral, the ocular, the antennal, the intercalary and the three gnathal segments.

Morphogenetic movements in the head of insects and crustaceans

Morphogenetic movements described for the Drosophila head might exist in a very similar manner in other insects and in crustaceans (Sharov, 1966). In Lepisma (Apterygota, Insecta) and Gammarus (Amphipoda, Crustacea) embryos, the labral segment shifts ventrally towards posterior during development (Weygoldt, 1958; Sharov, 1966). This leads to a recess in the segmental band (Sharov, 1966, his Fig. 63) of these hypognathous Mandibulata. In most insects and crustaceans the hypognathous state independently changed into a prognathous one which in insects brought the eyes into a more posterodorsal position (for review see Sharov, 1966). The resulting S-shaped arrangement of head segments seems to match well with the situation in the Drosophila embryo at the stage of germ band retraction. Therefore, this arrangement of head segments might be typical for embryos of prognathous mandibulates in general.

We thank Klaus Sander (Freiburg), Andreas Prokop (Mainz), Marcos Gonzalez Gaitan (Göttingen) and Kei Ito (Mainz) for reading an earlier version of the manuscript and improving it by helpful discussions. Marcos Gonzalez Gaitan kindly permitted us to mention unpublished results. For the anti-engrailed antibody (4D9) we are grateful to Harald Saumweber (Köln) and Christian Klämbt (Köln). The wingless antiserum was a gift by Alfonso Martinez Arias (Cambridge, UK) who also sent us the mutant lines wg1L114 and ryxho25. Norbert Perrimon (Boston) kindly sent us the wglacZ lines CyO, wgen11/sli and wg17en40cP1/CyO that were generated in his laboratory. For help with the photographs and the drawings we thank Ursula Böll, Erika Jost and Regina Gramsch. This work was supported by a grant from the Deutsche Forschungsgemeinschaft to G.M.T. (DFG, Te 130/3-2).

Baker
,
N. E.
(
1987
).
Molecular cloning of sequences from wingless, a segment polarity gene in Drosophila: the spatial distribution of a transcript in embryos
.
EMBO J
.
6
,
1765
1773
.
Baker
,
N. E.
(
1988
).
Localization of transcripts from the wingless gene in whole Drosophila embryos
.
Development
103
,
289
298
.
Campos-Ortega
,
J. A.
and
Hartenstein
,
V.
(
1985
)
The Embryonic Development of Drosophila melanogaster.Berlin: Springer
.
Cohen
,
S.
and
Jürgens
,
G.
(
1990
).
Mediation of Drosophila head development by gap-like segmentation genes
.
Nature
346
,
482
485
.
Cohen
,
S.
and
Jürgens
,
G.
(
1991
).
Drosophila headlines
.
Trends Genet
.
7
,
267
272
.
Diederich
,
R. J.
,
Pattatucci
,
A. M.
and
Kaufman
,
T. C.
(
1991
).
Development and evolutionary implications of labial, Deformed and engrailed expression in the Drosophila head
.
Development
113
,
273
281
.
DiNardo
,
S.
,
Kurner
,
J. M.
,
Theis
,
J.
and
O’Farrell
,
P. H.
(
1985
).
Development of embryonic pattern in Drosophila melanogaster as revealed by accumulation of the nuclear engrailed protein
.
Cell
43
,
59
69
.
Gonzalez
,
F.
,
Swales
,
L.
,
Bejsovec
,
A.
,
Skaer
,
H.
and
Martinez Arias
,
A.
(
1991
).
Secretion and movement of wingless protein in the epidermis of the Drosophila embryo
.
Mechanisms of Development
35
,
43
54
.
Hama
,
C.
,
Ali
,
Z.
and
Kornberg
,
T. B.
(
1990
).
Region-specific recombination and expression are directed by portions of the Drosophila engrailed promotor
.
Genes Dev
.
4
,
1079
1093
.
Hartenstein
,
V.
and
Campos-Ortega
,
J. A.
(
1984
).
Early neurogenesis in wild-type Drosophila melanogaster
.
Roux’s Arch. Dev. Biol
.
193
,
308325
.
Hartenstein
,
V.
and
Campos-Ortega
,
J. A.
(
1985
).
Fate-mapping in wildtype Drosophila melanogaster
.
Roux’s Arch. Dev. Biol
.
194
,
181
195
.
Ito
,
K.
and
Hotta
,
Y.
(
1992
).
Proliferation pattern of postembryonic neuroblasts in the brain of Drosophila melanogaster
.
Dev. Biol
.
149
,
134148
.
Jürgens
,
G.
,
Wieschaus
,
E.
,
Nüsslein-Vollhard
,
C.
and
Kluding
,
M.
(
1984
).
Mutations affecting the pattern of the larval cuticle in Drosophila melanogaster. II. Zygotic loci on the third chromosome
.
Roux’s Arch. Dev. Biol
.
193
,
283
295
.
Jürgens
,
G.
,
Lehmann
,
R.
,
Schardin
,
M.
and
Nüsslein-Vollhard
,
C.
(
1986
).
Segmental organisation of the head in the embryo of Drosophila melanogaster
.
Roux’s Arch. Dev. Biol
.
195
,
359
377
.
Martinez Arias
,
A.
,
Baker
,
N. E.
and
Ingham
,
P. W.
(
1988
).
Role of segment polarity genes in the definition and maintenance of cell states in the Drosophila embryo
.
Development
103
,
157
170
.
Miller
,
A.
(
1940
).
Embryonic membranes, yolk cells, and morphogenesis of the stonefly Pteronarcys proteus Newman
.
Annals of the Entomological Society of America
33
,
437
477
.
Monsma
,
A. S.
,
Ard
,
R.
and
Lis
,
J. T.
(
1988
).
Localized heat-shock induction in Drosophila melanogaster
.
J. Exp. Zool
.
247
,
279
284
.
Nair
,
B.
(
1949
).
The embryology of Cardina laevis Heller
.
Proc. Indian Acad. Sci. (B)
29
,
211
288
.
Nüsslein-Vollhard
,
C.
,
Wieschaus
,
E.
and
Kluding
,
H.
(
1984
).
Mutations affecting the pattern of the cuticle in Drosophila melanogaster I. Zygotic loci on the second chromosome
.
Roux’s Arch. Dev. Biol
.
192
,
267
282
.
Patel
,
N. H.
,
Martin-Blanco
,
E.
,
Coleman
,
K. G.
,
Poole
,
S. J.
,
Ellis
,
M. C.
,
Kornberg
,
T. B.
and
Goodman
,
C. S.
(
1989
).
Expression of engrailed proteins in arthropods, annelids and chordates
.
Cell
58
,
955968
.
Prokop
,
A.
,
Barleben
,
F.
,
Fischbach
,
K.-F.
and
Technau
,
G. M.
(
1991
).
The mushroom body defect gene affects proliferation behaviour of postembryonic neuroblasts
.
In Synapse, Transmission, Modulation
, (ed.
N.
Elsner
,
H.
Penzlin
), p.
544
Stuttgart: Georg Thieme.
Rempel
,
J. G.
(
1975
).
The evolution of the insect head: an endless dispute
.
Questiones Entomologicae
11
,
7
25
.
Rohrschneider
,
I.
(
1968
).
Beiträge zur Entwicklung des Vorderkopfes und der Mundregion von Periplaneta americana
.
Zool. Jb. Anat
.
85
,
537
578
.
Scholl
,
G.
(
1969
).
Die Embryonalentwicklung des Kopfes und Prothorax von Carausius morosus Br. (Insecta, Phasmida)
.
Z. Morph. Tiere
65
,
1
142
.
Sharov
,
A. G.
(
1966
).
Basic arthropodan stock with special reference to insects
, pp. 162f and 179.
Oxford: Pergamon
.
Technau
,
G. M.
and
Campos-Ortega
,
J. A.
(
1985
).
Fate-mapping in wildtype Drosophila melanogaster. II. Injections of horseradish peroxidase in cells of the early gastrula stage
.
Roux’s Arch. Dev. Biol
.
194
,
196
212
.
Truman
,
J. W.
and
Bate
,
M.
(
1988
).
Spatial and temporal patterns of neurogenesis in the central nervous system of Drosophila melanogaster
.
Dev. Biol
.
125
,
145
157
.
van den Heuvel
,
M.
,
Nusse
,
R.
,
Johnston
,
P.
and
Lawrence
,
P. A.
(
1989
).
Distribution of the wingless gene product in Drosophila embryos: A protein involved in cell-cell communication
.
Cell
59
,
739
749
.
Vincent
,
J.-P.
and
O’Farrell
,
H. O.
(
1992
).
The state of engrailed expression is not clonally transmitted during early Drosophila development
.
Cell
68
,
923
931
.
Wada
,
S.
(
1966a
).
Analyse der Kopf-Hals-Region von Tachycines (Saltatoria) in morphogenetische Einheiten. I. Mitteilung: Anatomische Befunde im schlüpfreifen und im imaginalen Normalzustand
.
Zool. Jb. Anat
.
83
,
185
234
.
Wada
,
S.
(
1966b
).
Analyse der Kopf-Hals-Region von Tachycines (Saltatoria) in morphogenetische Einheiten. II. Mitteilung: Experimentell-teratologische Befunde am Kopfskelett mit Berücksichtigung des zentralen Nervensystems
.
Zool. Jb. Anat
.
83
,
235326
.
Wada
,
S.
(
1966c
).
Topographie der Anlagenkomplexe der Cephalregion von Tachycines (Saltatoria) beim Keimstreif
.
Natuwissenschaften
53
,
414
.
Weygoldt
,
P.
(
1958
).
Die Embryonalentwicklung des Amphipoden Gammarus pulex pulex (L
.).
Zool. Jb. Anat
.
77
,
51
110
.
Wiesmann
,
R.
(
1926
).
Entwicklung und Organogenese der Coelomblasen von Carausius morosus Br
.
In Zur Kenntnis der Anatomie und Entwicklungsbiologie der Stabheuschrecke Carausius morosus Br
., (eds.
H.
Leuzinger
,
R.
Wiesmann
,
F. E.
Lehmann
), pp.
123
328
.
Jena
:
Gustav Fischer
.
1

In contrast to Wiesmann, Scholl (1969) believes that in Carausius (Phasmida) the second (optic) pair of coelomic cavities in front of the antennal segment derives from the intercalary coelom.