The first sign of metamerization in the Drosophila embryo is the striped expression of pair-rule genes such as fushi tarazu (ftz) and even-skipped (eve). Here we describe, at cellular resolution, the development of ftz and eve protein stripes in staged Drosophila embryos. They appear gradually, during the syncytial blastoderm stage and soon become asymmetric, the anterior margins of the stripes being sharply demarcated while the posterior borders are undefined. By the beginning of germ band elongation, the eve and ftz stripes have narrowed and become very intense at their anterior margins. The development of these stripes in hairy, runt, eve,ftz and engrailecT embryos is illustrated. In eve embryos, the ftz stripes remain symmetric and lack sharp borders. Our results support the hypothesis (Lawrence et al. Nature 328, 440-442, 1987) that individual cells are allocated to parasegments with respect to the anterior margins of the eve and ftz stripes.

When the Drosophila egg is laid it is already polarized in the anteroposterior axis, a polarity that becomes translated into a gradient of positional information (Sander, 1960; Driever & Nüsslein-Volhard, 1988). Eventually a series of genetic and epigenetic steps (for example, see reviews by Nüsslein-Volhard et al. 1987; Ingham, 1988) divide up the main part of the body, both ectoderm and mesoderm, into fourteen parasegments, each of which will construct a precisely defined part of the larva and adult (Martinez-Arias & Lawrence, 1985). As parasegments are defined, so is each individual cell irrevocably allocated to a compartment of cells that all share one genetically determined fate (Garcia-Bellido et al. 1979).

Of the number of segmentation genes classified by Nüsslein-Volhard & Wieschaus (1980), fushi tarazu (ftz) and even-skipped (eve) may be complementary because there is evidence that these genes act together to delimit parasegments - ftz+ being directly responsible for the anterior borders of the even-numbered parasegments and eve+ for the anterior borders of odd-numbered parasegments. Drawing these ‘lines’ through the blastoderm may allocate each and every cell in the trunk of the embryo to a specific parasegment (Lawrence et al. 1987).

Our purpose is to describe the pattern of eve and ftz expression in greater detail than hitherto. We use staged embryos and study the patterns of expression of both gene products at cellular resolution. We show that the eve and ftz stripes are asymmetric from midblastoderm, that is the anterior edges of the stripes are more definite than the posterior edges. Later, both eve and ftz stripes decay from the posterior edge, the anterior boundaries stabilize and sharpen as expression of the genes becomes locally more intense. By the time the germ band is extending, expression of both genes is confined to ragged rows of cells at the anterior borders of each parasegment. In embryos lacking the eve gene, the ftz stripes are strong and broad but lack asymmetry and do not sharpen. Such ftz stripes fail to make metameres suggesting that, in the wildtype, it is the sharp and persistent anterior edges that are the elements of the stripes responsible for this function.

Antibody labelling

We have used affinity-purified rabbit anti-eve and anti-ftz sera generously given by Manfred Frasch and Henry Krause (see Frasch et al. 1987; Krause et al. 1988), as well as anti-/Sgalactosidase and anti-engrailed monoclonal antibodies kindly sent by Chris Doe and Kevin Coleman, respectively. Eggs were dechorionated, fixed in 4% paraformaldehyde/ heptane, the vitelline membrane was removed in heptane/ methanol and the embryos stored in methanol. The anti-ftz serum gave better results if it was preabsorbed - it was diluted × 100, mixed with about 20% volume of eggs aged approx. 6–16 h, agitated overnight at 4°C and used at 1:5. The anti-eve serum was used at 1:20.

For double labelling (Lawrence et al. 1987), batches of embryos were treated overnight in anti-/u primary antibody in PBS buffer plus 0·1% BSA and 01% triton (‘BBT’). After washing in BBT and in BBT containing 2% goat serum they were treated for 2h in biotinylated goat anti-rabbit IgG (Vectorlabs, preabsorbed against Drosophila eggs of 0·16 h old and used at a total dilution of 1:500). After washing, the embryos were reacted with ABC reagent and then with DAB as instructed by Vectorlabs. This gave an orange-brown colour. The embryos were then washed in BBT and transferred to overnight treatment with anti-eve antibody and then with biotinylated goat anti-rabbit IgG as before. Finally, cobalt and nickel ions (0·03%) were added to the DAB reaction to give a dark grey colour (Adams, 1981).

For labelling of galactosidase, ftz and eve antibodies were applied together as the first step and, after reaction for both with DAB, Doe’s monoclonal antibody against β galactosidase was preabsorbed with 6–16 h embryos and used at 1:10, preabsorbed biotinylated horse anti-mouse IgG (Vectorlabs) was used at 1:500 and the embryos stained with DAB plus Co2+ and Ni2+. The same procedure was followed for studying en embryos, except that a monoclonal antibody against engrailed was used in the second step - the ascites fluid was preabsorbed against 6–16 h embryos and used at 1:130.

Mounting embryos

The stained embryos were washed, dehydrated and cleared in methyl salicylate. Individual embryos were placed on a slide between two no. 1 (thickness approx. 140 pm) coverslips in Araldite or Canada balsam and then roofed over with a no. 1) (approx. 170 μm) coverslip. Sideways movement of the roofing coverslip rolls the eggs to the required orientation. For detailed examination of the cells, the eggs were broken open with a fine needle along the midventral axis, the insides removed as far as possible with the needle and the pieces mounted flat. The roofing coverslip was supported either by no. 0 (approx. 100 μ m) coverslips, or by glass beads, diameter 40–75 μ m.

Photography

Photographs were taken under bright-held or DIC optics using Zeiss WL and Zeiss Axiophot microscopes. Ekta-chrome 64 and Fuji 50 were used, the latter gave somewhat greater colour contrast. For the wildtype series of pictures (Figs 36, 1518) all embryos were taken from one batch of double-stained wildtype embryos and all photographed and printed under standard conditions. Increases in colour intensity with age seen in the stained cells should therefore reflect a genuine increase in the amount of gene products.

Staging

We have restricted this study to cell cycle 14 and onward. Although ftz and eve are transcribed from about cell cycle 12 (Hafen et al. 1984; Macdonald et al. 1986) there is no clear striping of proteins until beyond cell cycle 14.

After cell cycle 14 there is a long interphase; living embryos can be staged by observing the plasma membranes (under DIC optics) which grow down between the peripheral nuclei, eventually cutting up the cortical cytoplasm into individual cells (Foe & Alberts, 1983; Edgar et al. 1987). This whole period (which lasts from some 2 h 30 min - 3 h 15 min after egg laying at 22°C) is called stage 5 (Bownes, 1975; Campos-Ortega & Hartenstein, 1985; Wieschaus & Nüsslein-Volhard, 1986) and we have subdivided it into 3 substages depending on the amount the membranes have grown in, each of about 15 min duration (see Fig. 1). Thus stage 5(1) indicates the first third of stage 5 and 5(3) the last. Gastrulation then follows and this is divided up into stages 6 and 7, each some 15 min long (op. cit. and own observations on living embryos). The germ band begins to extend (stage 8) and our present interest in the process ends with the loss oiftz and eve expression - this occurs as germ band extension is completed. A useful landmark is when the germ band is only partially extended, and when the expression of ftz and eve is still strong, we call this 8(1) and it is some 15 min after the end of stage 7 (i.e. at about 4 h after egg laying at 22°C). Thus, the whole process of stripe maturation is accomplished within about 90 min of development, which we divide into six stages, each of about 15 min (Fig. 1).

Fig. 1

A summary of the criteria used to stage the embryos (see text). For subdividing stage 5 the ingrowing membranes are observed, for stages 6 and 7 the ventral and cephalic furrows provide a guide, and for stage 8(1) the extent of germ band elongation is used (see methods). Timings are approximate (22°C).

Fig. 1

A summary of the criteria used to stage the embryos (see text). For subdividing stage 5 the ingrowing membranes are observed, for stages 6 and 7 the ventral and cephalic furrows provide a guide, and for stage 8(1) the extent of germ band elongation is used (see methods). Timings are approximate (22°C).

Quantification

Although we hope the pictures speak for themselves, we did try and quantify the degree of asymmetry in the stripes oiftz and eve expression. Each of a different batch of embryos was split down the dorsoventral midline, mounted and a specific region was studied under bright-field optics (× 63, 1·4NA lens). This included ftz stripes ps 4, 6, 8 and eve stripes ps 5, 7 and 9 (note stripes are numbered according to the parasegments to be defined by the anterior margin of the stripe). An area including these 6 stripes of about 15 cells across was studied for both left and right halves.

We first counted all the cells stained darkly for each antigen and then those that were palely stained, separately noting those that were at the anterior or the posterior edge of the stripes. Unstained cells were also counted (these could be compared with cells anterior to the first eve stripe which express neither eve nor ftz). The staining tor ftz was clearer than for eve so the ftz figures are more reliable. This could not be done before stage 5(2) as all the cells stain so weakly. Some raw data are shown in Table 1.

Table 1

The staining of a midlateral set of cells in presumptive parasegments 4-9 inclusive

The staining of a midlateral set of cells in presumptive parasegments 4-9 inclusive
The staining of a midlateral set of cells in presumptive parasegments 4-9 inclusive

β galactosidase expression

We used ftz-β gal (Hiromi et al. 1985) and eve-β gal constructs (Lawrence et al. 1987) to see if the native ftz and eve patterns of expression are identical, cell-by-cell, to that of the β galactosidase. Unfortunately, in these stocks, the β galactosidase does not stain sufficiently intensely until the beginning of germ band elongation. By that time the two gene products are coextensive at the anterior margins of the stripes, but β galactosidase (presumably because it accumulates over a longer period as it is much more persistent (compare Hiromi et al. 1985; Edgar et al. 1986)) is graded and is detected further back (Fig. 24 for ftz and/iz-/jgal, eve-)3gal not shown). The anterior margins of stripes of both β galactosidase (Lawrence et al. 1987) and the native ftz protein (Carroll et al. 1988b) are coextensive with the anterior margin of the engrailed stripe. These results confirm that the β galactosidase patte rreflects an integration over time of the pattern of expression of the native ftz and eve genes (Lawrence et al. 1987).

Fly stocks

For eve embryos, eveRI3/Balancer and Df(2R)eve121/ Balancer were crossed (eveR13 is a null mutation and Df(2R)eve’21 lacks the entire gene, Nüsslein-Volhard et al. 1984, 1985).

For hairy (h) embryos, h1L79k/Balancer and h’12/ Balancer were crossed (both these hairy alleles are considered to be nulls, Ingham et al. 1985a).

For runt embryos runLB5/Balancer and Df(l)B102, run/ Balancer were crossed (runLB5 is considered to be a null, Gergen & Wieschaus, 1986).

For ftz embryos, ftzW20/Balancer were crossed to Df(3R)4 Scb/Balancer (ftzWV) is considered to be a null allele (Weiner et al. 1984) and Df(3R)4 Scb lacks the entire gene). Embryos from these four crosses were collected and double-stained for ftz and eve as before. The eve embryos could be recognized as they lacked all eve antigen. The ftz embryos had a few scattered weakly stained brown cells, which were not simply background as they were only found in the domain where ftz is expressed (up to 70% egg length). The h and runt embryos were identified by their markedly abnormal staining patterns. abscissa: stage and estimated age in minutes.

As development proceeds the proportion of cells strongly expressing ftz and eve diminishes. Some of the cells at the edges of the stripes express ftz and eve only weakly, the majority of these are at the posterior edge, with a diminishing number at the anterior. The slope marked A indicates the proportion of all weakly stained cells that are located at the anterior edge of the stripe; it falls from 20 to 0% over the period studied. Compare with Figs 35 and 1518. See Table 1.

For en embryos, eggs from a Df(2R)enB/Balancer stock (see Gubb, 1985) were collected and labelled as described under antibody labelling.

For studying ftz-βgal (Hiromi et al. 1985) and eve-βgal (Lawrence et al. 1987) distribution, stocks carrying either one or both these constructs were used as described under antibody labelling above.

The wildtype pattern

The main results are the Figs 36 which show the refinement of the ftz and eve stripes from the beginning of the 14th interphase in the blastoderm to the middle of germ band extension. Low-power pictures of embryos are shown in Figs 1518, these give an overview of the process. Initially, the ftz (brown stain) and eve (grey stain) gene products are widely distributed but by stage 5(2) (Figs 3,15) weak stripes are visible. By 5(3)

(Figs 4,16) some unlabelled cells show between the stripes and, as the stripes intensify, these gaps widen. The stripes are asymmetric (sharper anterior boundaries, softer posterior ones) from stage 5(3) but this becomes more obvious as they mature. By the beginning of gastrulation the distribution of both ftz and eve appears graded, stain is more concentrated in cells anteriorly, less so posteriorly.

Sharpening of the different stripes does not occur at the same rate - for example, the eve stripe corresponding to parasegment (ps) 1 is relatively sharp, and the ftz stripe corresponding to ps4 relatively fuzzy - details of this sort can be seen in the figures and they are reproducible. Unstained cells begin to appear first in the posterior regions of the presumptive even-numbered parasegments about stage 5(3) (Figs 4,16) and our impression is that this is due to loss of eve label (Figs 15,16). By stage 8(1) extension of the germ band is associated with increasing cell cross-sectional area, and cell rearrangement (Figs 6,18). At this stage the parasegments may vary from 2 to 6 cells across.

We have monitored the proportion of cells expressing ftz or eve (Fig. 2). Table 1 (see Methods) also shows how the stripes sharpen, with the number of weakly staining cells at the posterior edge of the stripes always being far more than at the anterior edge; weakly stained anterior cells drop to none by stage 8(1)–see Fig. 2.

Fig. 2

Measures of stripe maturation in the wildtype, ordinate:% of total cells scored.

Fig. 2

Measures of stripe maturation in the wildtype, ordinate:% of total cells scored.

Mutant phenotypes

There are many aspects of the altered ftz and eve expression that could be discussed but as we suspect that the crucial aspect of these stripes in the wildtype is the sharp anterior edges (because they delimit the parasegment borders, Lawrence et al. 1987) we will look particularly at the presence or absence of such edges in the various mutant phenotypes.

Pattern in hairy

The hairy gene is required for proper expression of ftz and eve (Carroll & Scott, 1986; Howard & Ingham, 1986; Frasch & Levine, 1987; Carroll et al. 1988a). Figs 7,8,19 and 20 show expression of ftz and eve in hairy embryos at stages 5(2) and 6. Initially, the ftz and eve expression are strongly overlapping, contrasting with the wildtype where such overlap is weak and transient. As the pattern resolves, many more cells than normal express ftz and only a few come to express eve strongly.

From stage 6, it is clear there is almost no vestige of the eve stripes except for psi, while the ftz stripes (presumably those equivalent to ps2, 4, 6, 8, 10, 14 in the wildtype) have fairly well developed anterior edges - although, with the possible exception of ps8, they are not as sharp or as straight as the wildtype (Fig. 20). The ftz stripes do narrow and sharpen with time, although not as much as the wildtype.

Thus, apart from psi, which should develop almost normally, the hairy embryos should have only the anterior margins of even-numbered parasegments - and these should, with the exception of ps8, be less sharp and straight than in the wildtype.

Fig. 3

Lateral view of emerging parasegments (numbered) in middle of body of wildtype. All these preparations were made as comparable as possible; they are from one batch of embryos, mounted nearly in the same way, photographed under the same conditions (partial DIC optics) on one film and printed under identical conditions. Note the sharpening and intensification of both the ftz (brown) and eve (grey) stripes. Using the rule that the anterior margins of the even-numbered parasegments are delimited by ftz, and the odd-numbered by eve (Lawrence et al. 1987) the observer can allocate each and every cell to a parasegment by stage 8(1) and, with greater difficulty, even in stage 6. Note how the number of cells across the parasegment varies from 2-5. Anterior to left, ventral to bottom, all X1250. Numbers shown above and below indicate the parasegments whose anterior borders correspond to the anterior border of the corresponding ftz stripes. Fig. 3, stage 5(2); Fig. 4, stage 5(3); Fig. 5, stage 6; Fig. 6, stage 8(1).

Fig. 3

Lateral view of emerging parasegments (numbered) in middle of body of wildtype. All these preparations were made as comparable as possible; they are from one batch of embryos, mounted nearly in the same way, photographed under the same conditions (partial DIC optics) on one film and printed under identical conditions. Note the sharpening and intensification of both the ftz (brown) and eve (grey) stripes. Using the rule that the anterior margins of the even-numbered parasegments are delimited by ftz, and the odd-numbered by eve (Lawrence et al. 1987) the observer can allocate each and every cell to a parasegment by stage 8(1) and, with greater difficulty, even in stage 6. Note how the number of cells across the parasegment varies from 2-5. Anterior to left, ventral to bottom, all X1250. Numbers shown above and below indicate the parasegments whose anterior borders correspond to the anterior border of the corresponding ftz stripes. Fig. 3, stage 5(2); Fig. 4, stage 5(3); Fig. 5, stage 6; Fig. 6, stage 8(1).

Fig. 4

Lateral view of emerging parasegments (numbered) in middle of body of wildtype. All these preparations were made as comparable as possible; they are from one batch of embryos, mounted nearly in the same way, photographed under the same conditions (partial DIC optics) on one film and printed under identical conditions. Note the sharpening and intensification of both the ftz (brown) and eve (grey) stripes. Using the rule that the anterior margins of the even-numbered parasegments are delimited by ftz, and the odd-numbered by eve (Lawrence et al. 1987) the observer can allocate each and every cell to a parasegment by stage 8(1) and, with greater difficulty, even in stage 6. Note how the number of cells across the parasegment varies from 2-5. Anterior to left, ventral to bottom, all X1250. Numbers shown above and below indicate the parasegments whose anterior borders correspond to the anterior border of the corresponding ftz stripes. Fig. 3, stage 5(2); Fig. 4, stage 5(3); Fig. 5, stage 6; Fig. 6, stage 8(1).

Fig. 4

Lateral view of emerging parasegments (numbered) in middle of body of wildtype. All these preparations were made as comparable as possible; they are from one batch of embryos, mounted nearly in the same way, photographed under the same conditions (partial DIC optics) on one film and printed under identical conditions. Note the sharpening and intensification of both the ftz (brown) and eve (grey) stripes. Using the rule that the anterior margins of the even-numbered parasegments are delimited by ftz, and the odd-numbered by eve (Lawrence et al. 1987) the observer can allocate each and every cell to a parasegment by stage 8(1) and, with greater difficulty, even in stage 6. Note how the number of cells across the parasegment varies from 2-5. Anterior to left, ventral to bottom, all X1250. Numbers shown above and below indicate the parasegments whose anterior borders correspond to the anterior border of the corresponding ftz stripes. Fig. 3, stage 5(2); Fig. 4, stage 5(3); Fig. 5, stage 6; Fig. 6, stage 8(1).

Fig. 5

Lateral view of emerging parasegments (numbered) in middle of body of wildtype. All these preparations were made as comparable as possible; they are from one batch of embryos, mounted nearly in the same way, photographed under the same conditions (partial DIC optics) on one film and printed under identical conditions. Note the sharpening and intensification of both the ftz (brown) and eve (grey) stripes. Using the rule that the anterior margins of the even-numbered parasegments are delimited by ftz, and the odd-numbered by eve (Lawrence et al. 1987) the observer can allocate each and every cell to a parasegment by stage 8(1) and, with greater difficulty, even in stage 6. Note how the number of cells across the parasegment varies from 2-5. Anterior to left, ventral to bottom, all X1250. Numbers shown above and below indicate the parasegments whose anterior borders correspond to the anterior border of the corresponding ftz stripes. Fig. 3, stage 5(2); Fig. 4, stage 5(3); Fig. 5, stage 6; Fig. 6, stage 8(1).

Fig. 5

Lateral view of emerging parasegments (numbered) in middle of body of wildtype. All these preparations were made as comparable as possible; they are from one batch of embryos, mounted nearly in the same way, photographed under the same conditions (partial DIC optics) on one film and printed under identical conditions. Note the sharpening and intensification of both the ftz (brown) and eve (grey) stripes. Using the rule that the anterior margins of the even-numbered parasegments are delimited by ftz, and the odd-numbered by eve (Lawrence et al. 1987) the observer can allocate each and every cell to a parasegment by stage 8(1) and, with greater difficulty, even in stage 6. Note how the number of cells across the parasegment varies from 2-5. Anterior to left, ventral to bottom, all X1250. Numbers shown above and below indicate the parasegments whose anterior borders correspond to the anterior border of the corresponding ftz stripes. Fig. 3, stage 5(2); Fig. 4, stage 5(3); Fig. 5, stage 6; Fig. 6, stage 8(1).

Fig. 6

Lateral view of emerging parasegments (numbered) in middle of body of wildtype. All these preparations were made as comparable as possible; they are from one batch of embryos, mounted nearly in the same way, photographed under the same conditions (partial DIC optics) on one film and printed under identical conditions. Note the sharpening and intensification of both the ftz (brown) and eve (grey) stripes. Using the rule that the anterior margins of the even-numbered parasegments are delimited by ftz, and the odd-numbered by eve (Lawrence et al. 1987) the observer can allocate each and every cell to a parasegment by stage 8(1) and, with greater difficulty, even in stage 6. Note how the number of cells across the parasegment varies from 2-5. Anterior to left, ventral to bottom, all X1250. Numbers shown above and below indicate the parasegments whose anterior borders correspond to the anterior border of the corresponding ftz stripes. Fig. 3, stage 5(2); Fig. 4, stage 5(3); Fig. 5, stage 6; Fig. 6, stage 8(1).

Fig. 6

Lateral view of emerging parasegments (numbered) in middle of body of wildtype. All these preparations were made as comparable as possible; they are from one batch of embryos, mounted nearly in the same way, photographed under the same conditions (partial DIC optics) on one film and printed under identical conditions. Note the sharpening and intensification of both the ftz (brown) and eve (grey) stripes. Using the rule that the anterior margins of the even-numbered parasegments are delimited by ftz, and the odd-numbered by eve (Lawrence et al. 1987) the observer can allocate each and every cell to a parasegment by stage 8(1) and, with greater difficulty, even in stage 6. Note how the number of cells across the parasegment varies from 2-5. Anterior to left, ventral to bottom, all X1250. Numbers shown above and below indicate the parasegments whose anterior borders correspond to the anterior border of the corresponding ftz stripes. Fig. 3, stage 5(2); Fig. 4, stage 5(3); Fig. 5, stage 6; Fig. 6, stage 8(1).

Fig. 7

Midlateral view of hairy embryos, stained for ftz and eve as in Figs 36. Fig. 7 stage 5(2). Note that eve and ftz overlap considerably giving a grey-brown colour. Fig. 8, stage 6. Only few cells now express eve, while ftz expression has intensified but is patchy. Staining is best assessed by comparing nuclei stained for eve (e), ftz if) or both (e/) with the background colour seen in unstained nuclei (b). The grey-brown colour which is due to coexpression of eve and ftz is also clearly seen in Figs 19 and 20.

Fig. 7

Midlateral view of hairy embryos, stained for ftz and eve as in Figs 36. Fig. 7 stage 5(2). Note that eve and ftz overlap considerably giving a grey-brown colour. Fig. 8, stage 6. Only few cells now express eve, while ftz expression has intensified but is patchy. Staining is best assessed by comparing nuclei stained for eve (e), ftz if) or both (e/) with the background colour seen in unstained nuclei (b). The grey-brown colour which is due to coexpression of eve and ftz is also clearly seen in Figs 19 and 20.

Fig. 8

Midlateral view of hairy embryos, stained for ftz and eve as in Figs 36. Fig. 7 stage 5(2). Note that eve and ftz overlap considerably giving a grey-brown colour. Fig. 8, stage 6. Only few cells now express eve, while ftz expression has intensified but is patchy. Staining is best assessed by comparing nuclei stained for eve (e), ftz if) or both (e/) with the background colour seen in unstained nuclei (b). The grey-brown colour which is due to coexpression of eve and ftz is also clearly seen in Figs 19 and 20.

Fig. 8

Midlateral view of hairy embryos, stained for ftz and eve as in Figs 36. Fig. 7 stage 5(2). Note that eve and ftz overlap considerably giving a grey-brown colour. Fig. 8, stage 6. Only few cells now express eve, while ftz expression has intensified but is patchy. Staining is best assessed by comparing nuclei stained for eve (e), ftz if) or both (e/) with the background colour seen in unstained nuclei (b). The grey-brown colour which is due to coexpression of eve and ftz is also clearly seen in Figs 19 and 20.

Fig. 9

Similar view of runt embryos at stages 5(2) and 6. The ftz stripes narrow while eve expression becomes patchy (compare Figs 19 and 20). All X1250.

Fig. 9

Similar view of runt embryos at stages 5(2) and 6. The ftz stripes narrow while eve expression becomes patchy (compare Figs 19 and 20). All X1250.

Fig. 10

Similar view of runt embryos at stages 5(2) and 6. The ftz stripes narrow while eve expression becomes patchy (compare Figs 19 and 20). All X1250.

Fig. 10

Similar view of runt embryos at stages 5(2) and 6. The ftz stripes narrow while eve expression becomes patchy (compare Figs 19 and 20). All X1250.

Fig. 11

ftz expression in eve embryos (stages 5(3) and 7). Note that the stripes are broad and symmetric, they do not narrow or sharpen. The spacing of stripes varies from embryo to embryo (compare Fig. 23). X1250.

Fig. 11

ftz expression in eve embryos (stages 5(3) and 7). Note that the stripes are broad and symmetric, they do not narrow or sharpen. The spacing of stripes varies from embryo to embryo (compare Fig. 23). X1250.

Fig. 12

ftz expression in eve embryos (stages 5(3) and 7). Note that the stripes are broad and symmetric, they do not narrow or sharpen. The spacing of stripes varies from embryo to embryo (compare Fig. 23). X1250.

Fig. 12

ftz expression in eve embryos (stages 5(3) and 7). Note that the stripes are broad and symmetric, they do not narrow or sharpen. The spacing of stripes varies from embryo to embryo (compare Fig. 23). X1250.

Fig. 13

Young extending germ band 8(1) to show ftz stripes marking anterior margin of ps4, 6 and 8. These embryos have been stained for ftz (brown) and for engrailed (grey). Fig. 13 is en+ and 14 is en, the nascent engrailed stripe which is easy to see elsewhere in the embryo, is just detectable in Fig. 13 (arrows). We can detect no difference in the ftz stripes.

Fig. 13

Young extending germ band 8(1) to show ftz stripes marking anterior margin of ps4, 6 and 8. These embryos have been stained for ftz (brown) and for engrailed (grey). Fig. 13 is en+ and 14 is en, the nascent engrailed stripe which is easy to see elsewhere in the embryo, is just detectable in Fig. 13 (arrows). We can detect no difference in the ftz stripes.

Fig. 14

Young extending germ band 8(1) to show ftz stripes marking anterior margin of ps4, 6 and 8. These embryos have been stained for ftz (brown) and for engrailed (grey). Fig. 13 is en+ and 14 is en, the nascent engrailed stripe which is easy to see elsewhere in the embryo, is just detectable in Fig. 13 (arrows). We can detect no difference in the ftz stripes.

Fig. 14

Young extending germ band 8(1) to show ftz stripes marking anterior margin of ps4, 6 and 8. These embryos have been stained for ftz (brown) and for engrailed (grey). Fig. 13 is en+ and 14 is en, the nascent engrailed stripe which is easy to see elsewhere in the embryo, is just detectable in Fig. 13 (arrows). We can detect no difference in the ftz stripes.

Fig. 15

Wildtype half embryos stained for ftz and eve as before. Stages 5(2), 5(3), 6 and 8(1). Fig. 17 shows the ftz stripes numbered according to the parasegments their anterior margins delimit.

Fig. 15

Wildtype half embryos stained for ftz and eve as before. Stages 5(2), 5(3), 6 and 8(1). Fig. 17 shows the ftz stripes numbered according to the parasegments their anterior margins delimit.

Fig. 16

Wildtype half embryos stained for ftz and eve as before. Stages 5(2), 5(3), 6 and 8(1). Fig. 17 shows the ftz stripes numbered according to the parasegments their anterior margins delimit.

Fig. 16

Wildtype half embryos stained for ftz and eve as before. Stages 5(2), 5(3), 6 and 8(1). Fig. 17 shows the ftz stripes numbered according to the parasegments their anterior margins delimit.

Fig. 17

Wildtype half embryos stained for ftz and eve as before. Stages 5(2), 5(3), 6 and 8(1). Fig. 17 shows the ftz stripes numbered according to the parasegments their anterior margins delimit.

Fig. 17

Wildtype half embryos stained for ftz and eve as before. Stages 5(2), 5(3), 6 and 8(1). Fig. 17 shows the ftz stripes numbered according to the parasegments their anterior margins delimit.

Fig. 18

Wildtype half embryos stained for ftz and eve as before. Stages 5(2), 5(3), 6 and 8(1). Fig. 17 shows the ftz stripes numbered according to the parasegments their anterior margins delimit.

Fig. 18

Wildtype half embryos stained for ftz and eve as before. Stages 5(2), 5(3), 6 and 8(1). Fig. 17 shows the ftz stripes numbered according to the parasegments their anterior margins delimit.

Pattern in runt

The runt gene is also required for the proper expression of ftz and eve (Carroll & Scott, 1986; Frasch & Levine, 1987) and from the beginning the patterns are altered in a complex way (Figs 9,10,21,22). Initially the ftz stripes corresponding to ps4, 8, and 14 are well developed, but 2, 6, 10, 12 are fading away. Later (Figs 10,22) ftz expression becomes limited to a few irregular rows of cells and a variable pattern of eve stripes survives, with irregular boundaries, some of which (such as that of psl3) are quite sharp. Some of these stripes appear to be sharpening at both anterior and posterior edges, suggesting there is some local reversal of polarity. The cuticle pattern of runt embryos does show zones of opposing polarity (Nüsslein-Volhard & Wieschaus, 1980).

Pattern of ftz expression in eve

As described by Carroll & Scott (1986), it is surprising that the pattern of expression of ftz is not much altered in eve embryos, even though such embryos produce larvae which show little residual metamerization (Nüsslein-Volhard et al. 1984), and do not express most of the engrailed stripes (Harding et al. 1986). However, the ftz stripes are abnormal in that they remain broad and do not sharpen at their anterior edges; weakly stained cells persist at both margins and the stripes remain symmetric (Figs 11,12,23).

Fig. 19

hairy embryos stained for ftz and eve as before; stages 5(2) and 6. Arrows mark the presumed anterior margins of indicated parasegment borders.

Fig. 19

hairy embryos stained for ftz and eve as before; stages 5(2) and 6. Arrows mark the presumed anterior margins of indicated parasegment borders.

Fig. 20

hairy embryos stained for ftz and eve as before; stages 5(2) and 6. Arrows mark the presumed anterior margins of indicated parasegment borders.

Fig. 20

hairy embryos stained for ftz and eve as before; stages 5(2) and 6. Arrows mark the presumed anterior margins of indicated parasegment borders.

Fig. 21

runt embryos stained for ftz and eve as before; stages 5(2) and 6.

Fig. 21

runt embryos stained for ftz and eve as before; stages 5(2) and 6.

Fig. 22

runt embryos stained for ftz and eve as before; stages 5(2) and 6.

Fig. 22

runt embryos stained for ftz and eve as before; stages 5(2) and 6.

Fig. 23

ftz expression in an eve embryo, stage 7.

Fig. 23

ftz expression in an eve embryo, stage 7.

Fig. 24

Lateral view of embryo (stage 8(1)) stained for both ftz and eve (both brown) and then with β galactosidase (dark grey). The stock carried a ftz-PgA construct, and the picture purports to show cell-by-cell correspondence of the β galactosidase and ftz staining at the anterior margins of the even-numbered parasegments. This correspondence of staining, which is more easily seen down the microscope by focusing up and down, can be pictured by the sympathetic reader if he or she carefully compares the even-numbered stripes (ftz, brown nuclei plus grey cytoplasm) with the alternating ones (eve, brown nuclei only).

Fig. 24

Lateral view of embryo (stage 8(1)) stained for both ftz and eve (both brown) and then with β galactosidase (dark grey). The stock carried a ftz-PgA construct, and the picture purports to show cell-by-cell correspondence of the β galactosidase and ftz staining at the anterior margins of the even-numbered parasegments. This correspondence of staining, which is more easily seen down the microscope by focusing up and down, can be pictured by the sympathetic reader if he or she carefully compares the even-numbered stripes (ftz, brown nuclei plus grey cytoplasm) with the alternating ones (eve, brown nuclei only).

Fig. 25

Lateral view of wildtype and ftz embryos stained as before for ftz and eve (stage 7). In the ftz embryo (Fig. 26) scattered cells show ftz staining but they are not striped. The eve stripes are similar in both wildtype and ftz . All the 14 parasegments that are basic to the body plan (Martinez-Arias & Lawrence, 1985) are indicated in Fig. 25.

Fig. 25

Lateral view of wildtype and ftz embryos stained as before for ftz and eve (stage 7). In the ftz embryo (Fig. 26) scattered cells show ftz staining but they are not striped. The eve stripes are similar in both wildtype and ftz . All the 14 parasegments that are basic to the body plan (Martinez-Arias & Lawrence, 1985) are indicated in Fig. 25.

Fig. 26

Lateral view of wildtype and ftz embryos stained as before for ftz and eve (stage 7). In the ftz embryo (Fig. 26) scattered cells show ftz staining but they are not striped. The eve stripes are similar in both wildtype and ftz . All the 14 parasegments that are basic to the body plan (Martinez-Arias & Lawrence, 1985) are indicated in Fig. 25.

Fig. 26

Lateral view of wildtype and ftz embryos stained as before for ftz and eve (stage 7). In the ftz embryo (Fig. 26) scattered cells show ftz staining but they are not striped. The eve stripes are similar in both wildtype and ftz . All the 14 parasegments that are basic to the body plan (Martinez-Arias & Lawrence, 1985) are indicated in Fig. 25.

Pattern of eve expression in ftzembryos

Staged ftz embryos (distinguished from siblings because they failed to express ftz) appeared to express eve quite normally, with the eve stripes maturing as in wildtype (Figs 25,26).

Pattern of ftz expression in enembryos

As striped engrailed expression begins soon after that of ftz and eve (Weir & Kornberg, 1985; DiNardo et al. 1985) it seemed possible that the coexpression of engrailed with ftz and eve in the anterior parts of the parasegments (Lawrence et al. 1987; Carroll et al. 1988a) might be essential for the narrowing and shar-pening of the ftz and eve stripes. This was not the case for ftz however; en embryos, easily detected at stage 8(1) because they lacked any engrailed protein, had ftz stripes that were indistinguishable from their en+ siblings (Figs 1314). We did not test for effect of en on eve stripes.

The development of the ftz and eve stripes has been extensively studied by in situ hybridization (Hafen et al. 1984; Ingham et al. 19856; Martinez-Arias & Lawrence, 1985; Weir & Kornberg, 1985; Harding et al. 1986; Macdonald et al. 1986), by means of antibodies (Carroll & Scott, 1985; Frasch et al. 1987; Frasch & Levine, 1987; Carroll et al. 1988a; Frasch et al. 1988; Krause et al. 1988) and the use of reporter gene constructs (Hiromi et al. 1985; Lawrence et al. 1987). From these studies, only some of which approach cellular resolution, cellular interpretations have been made (Duncan, 1986; Gergen et al. 1986; O’Farrell & Scott, 1986; Akam, 1987; DiNardo & O’Farrell, 1987; Scott & Carroll, 1987; Frasch et al. 1988; Ingham, 1988; Ingham et al. 1988; Martinez-Arias et al. 1988; Weir et al. 1988). We have provided photographs of staged embryos that describe the process at cellular resolution, so that theoretical descriptions and models can be given a firmer basis. We have also described some mutant phenotypes.

It is clear from the photographs and from cell counts (Table 1, Fig. 2) that the eve and ftz protein stripes can be first resolved midway through the blastoderm stage (stage 5(2) - for stages see Methods). By stage 5(3) they are already somewhat asymmetric. The asymmetry increases as their anterior edges stabilize and sharpen by the beginning of gastrulation about 30 min later. The definitive anterior boundaries must become established around the time of cellularization because there are so few weakly stained anterior cells at that time (stage 5(3)-6). During gastrulation and thereafter, the stripes intensify at the anterior edge, and narrow from the posterior edge. This behaviour of the stripes was predicted from the anteroposterior gradient of intensity, and sharp and stable anterior edges of the stripes, of ftz-β galactosidase and eve-/5galactosidase (Lawrence et al. 1987). We also suggested that one main wildtype function of the eve gene is to delimit the anterior boundaries of the odd-numbered parasegments - that is, in effect, to allocate each and every cell near to the anterior edges of the eve stripes to a parasegment. We proposed that ftz does the same with respect to the anterior boundaries of the even-numbered parasegments, so that each parasegment becomes defined as those cells that lie between one delimiting ftz boundary and one delimiting eve boundary (Fig. 6; Lawrence et al. 1987; Lawrence, 1987, 1988).

This hypothesis can be contrasted with a class of hypotheses that consider ftz and eve to label those cells in the stripes and distinguish them from those not in the stripes. Such hypotheses are that frz, eve and other pairrule genes (such as paired, see Kilcherr et al. 1986) are expressed in stripes that overlap and that the combination of active and inactive genes allocates each cell to a particular state which defines its subsequent development (Gergen et al. 1986; Scott & O’Farrell, 1986; Akam, 1987; DiNardo & O’Farrell, 1987; Scott & Carroll, 1987; Carroll et al. 1988a; Frasch et al. 1988; Ingham, 1988; Ingham et al. 1988; Martinez-Arias et al. 1988; Weir et al. 1988). To take just one example from many, Ingham et al. (1988) have suggested that the wingless gene (which becomes expressed in cells near the posterior edge of each parasegment, Baker, 1987) may be activated in certain cells as, and because, they become free of eve and ftz gene products. As the eve and ftz stripes narrow further, Ingham et al. suggest that activation of other genes (such as odd-paired and paired) may stop wingless becoming expressed more and more anteriorly. Models of this type treat gene expression as under digital control (Scott & O’Farrell, 1986) and complex combinatorial and hierarchical models can be built (see Martinez-Arias et al. 1988; Ingham, 1988). However, these models describe events at the cellular level, but mostly depend on descriptions of gene expression at less than cellular resolution. Thus, they treat the stripes of gene expression as if they had sharp limits and as if all cells fall clearly into or out of the stripes - as we have shown, this is not the case. Further, by elaborating beyond the careful description of Carroll & Scott (1985) the myth has grown up that the ftz stripes and interstripes are initially four cells wide and then change to being three cells on, five cells off. This digital picture has been used to build models (op. cit.) that clash with our observations that the ftz and eve stripes are asymmetric, graded, and vary markedly in cellular width (see Figs 36).

It is our hypothesis that ftz and eve are the first genes, in development of the anteroposterior axis, to allocate cells to a parasegment, that is to a limited cell fate or ‘genetic address’ (Garcia-Bellido et al. 1979) that is stable through cell lineage. Once parasegment borders have been defined they must be maintained and they could limit gradients of positional information (for references see Wolpert, 1969; Lawrence, 1973). Under this hypothesis subsequent expression of genes such as engrailed and wingless would result from the interpretation of the gradients, in the same general way as the segments of Galleria become subdivided into several zones bearing oriented scales (Marcus, 1962; see Lawrence, 1973). Since our Fig. 6 and the pattern of /3gal expression (Fig. 24 and Lawrence et al. 1987) do suggest that ftz and eve expression become graded it is even possible that the concentrations of these gene products themselves are interpreted, directly or indirectly, by the cells (like the bicoid protein, Driever & Nüsslein-Volhard, 1988) - but many other interpretations are equally plausible.

From the theoretical bias outlined above, we can look at the results of ftz and eve expression in mutants. In h mutants, the pattern at gastrulation has resolved into broad ftz stripes with edges of variable sharpness (Figs 8,20). Nevertheless one could draw the anterior margins of ps2, 4 and 8 fairly well, while 6,10,12 and 14 would be more indefinite and uneven. Apart from the first eve stripe, which has a clear anterior edge, the others have disappeared - although there are some scattered cells expressing eve. This phenotype predicts the hairy cuticle phenotype well - we expect parasegments 1, 2, 4, 6, 8, 10, 12, and 14 to develop anterior borders with variable success, and, because of the lack of the anterior borders of odd-numbered parasegments, to be initially of double width. The cuticle phenotype should therefore resemble that of weak mutant alleles of eve - with the exception of parasegment 1 - and this is the case (Nüsslein-Volhard et al. 1982; Ingham et al. 1985a,b). Our results help understand how ‘the h phenotype appears to be a consequence of the change in domains of activity of ftz and other pair rule genes …’ (Howard & Ingham, 1986). Unfortunately, the wildtype function of hairy remains mysterious. Fig. 7 shows that, by stage 5(2), many cells express both ftz and eve simultaneously in hairy embryos (such strong overlap does not happen in the wildtype), suggesting that the hairy gene product is needed from the beginning for separating the emerging ftz and eve stripes.

The expression of eve and ftz in runt embryos is very abnormal from an early stage (Figs 9,10,21,22), and is completely different from hairy embryos. We do not understand it.

Embryos that are eve provide good evidence for our hypothesis. The ftz stripes in eve embryos are well developed but are symmetric and lack sharp bound-aries. Over most of the embryo the/tz stripes appear to be completely ineffective–they lead to no expression of engrailed (Harding et al. 1986; Frasch et al. 1988) and there is no other evidence of metamerization in eveembryos (Nüsslein-Volhard et al. 1984). How the eve gene product helps sharpen the anterior boundaries of ftz expression is unclear, but we conjecture that it is the lack of a sharp boundary that makes them ineffective.

The interdependence of eve and ftz is not reciprocal, for the boundaries of the eve stripes appear normal in ftz embryos (see Fig. 26); it therefore seems unlikely that the ftz and eve genes simply compete for territory by mutual repression. It is known that eve later becomes weakly expressed in the even-numbered parasegments (Macdonald et al. 1986; Frasch et al. 1987) and thus may act directly on the ftz-expressing cells there.

Attempts to model the development of eve and ftz stripes are now being made (Lacalli et al. 1988; Edgar et al. 1988), they treat the stripes as symmetric: if our hypothesis of the importance of asymmetry - the sharp and stable anterior border and the indefinite and unstable posterior border - is correct, new models will be needed.

We thank Kevin Coleman, Walter Gehring, Tom Kornberg and Henry Krause for generously providing antibodies prior to publication. Other antibodies were kindly given by Chris Doe and Manfred Frasch. Gary Struhl started us staining for antibodies in 1986 and has provided much needed encouragement since. We thank Phil Ingham, David Ish-Horowicz, Yash Hiromi, Paul Macdonald and Janni Niisslein-Volhard for fly stocks and Jonathan Hodgkin, Andrew Tomlinson and Michael Wilcox for advice on the manuscript. Thanks to Michaels Akam and Ashbumer, too.

Adams
,
J. C.
(
1981
).
Heavy metal intensification of DAB-based HRP reaction product
.
J. Histochem. Cytochem
.
29
,
775
.
Akam
,
M.
(
1987
).
The molecular basis for metameric pattern in the Drosophila embryo
.
Development
101
,
1
22
.
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
.
Bownes
,
M.
(
1975
).
A photographic study of development in the living embryo of Drosophila melanogaster
.
J. Embryol. exp. Morph
.
33
,
789
801
.
Campos-Ortega
,
J. A.
&
Hartenstein
,
V.
(
1985
).
The Embryonic Development of Drosophila melanogaster
.
Springer
,
Berlin
.
Carroll
,
S. B.
,
Dinardo
,
S.
,
O’Farrell
,
P. H.
,
White
,
R. A. H.
&
Scott
,
M. P.
(
1988a
).
Temporal and spatial relationships between segmentation and homeotic gene expression in Drosophila embryos: distributions of the fushi tarazu, engrailed, Sex combs reduced, Antennapedia, and Ultrabithorax proteins
.
Genes & Development
2
,
350
360
.
Carroll
,
S. B.
,
Laughon
,
A.
&
Thalley
,
B. S.
(
1988b
).
Expression, function, and regulation of the hairy segmentation protein in the Drosophila embryo
.
Genes & Development
2
,
883
890
.
Carroll
,
S. B.
&
Scott
,
M. P.
(
1985
).
Localization of the fushi tarazu protein during Drosophila embryogenesis
.
Cell
43
,
47
57
.
Carroll
,
S. B.
&
Scott
,
M. P.
(
1986
).
Zygotically active genes that affect the spatial expression of the fushi tarazu segmentation gene during early Drosophila embryogenesis
.
Cell
45
,
113
126
.
Dinardo
,
S.
,
Kuner
,
J. M.
,
Theis
,
J.
&
O’Farrell
,
P. H.
(
1985
).
Development of embryonic pattern in D. melanogaster as revealed by accumulation of the nuclear engrailed protein
.
Cell
43
,
59
69
.
Dinardo
,
S.
&
O’Farrell
,
P. H.
(
1987
).
Establishment and refinement of segmental pattern in the Drosophila embryo: spatial control of engrailed expression by pair-rule genes
.
Genes & Development
1
,
1212
1225
.
Driever
,
W.
&
Nüsslein-Volhard
,
C.
(
1988
).
The bicoid protein determines position in the Drosophila embryo in a concentration-dependent manner
.
Cell
54
,
95
104
.
Duncan
,
I. M.
(
1986
).
Control of bithorax complex functions by the segmentation gene fushi tarazu of D. melanogaster
.
Cell
47
,
297
309
.
Edgar
,
B. A.
,
Odell
,
G. M.
&
Schubiger
,
G.
(
1987
).
Cytoarchitecture and the patterning of fushi tarazu expression in the Drosophila blastoderm
.
Genes & Development
1
,
1226
1237
.
Edgar
,
B. A.
,
Odell
,
G. M.
&
Schubiger
,
G.
(
1988
).
A genetic switch, based on negative regulation, sharpens stripes in Drosophila embryos
.
Developmental Genetics (In press)
.
Edgar
,
B. A.
,
Weir
,
M. P.
,
Schubiger
,
G.
&
Kornberg
,
T.
(
1986
).
Repression and turnover pattern of fushi tarazu RNA in the early Drosophila embryo
.
Cell
47
,
747
754
.
Foe
,
V. E.
&
Alberts
,
B. M.
(
1983
).
Studies of nuclear and cytoplasmic behaviour during the five mitotic cycles that precede gastrulation in Drosophila embryogenesis
.
J. Cell Sei
.
61
,
31
70
.
Frasch
,
M.
,
Hoey
,
T.
,
Rushlow
,
C.
,
Doyle
,
H.
&
Levine
,
M.
(
1987
).
Characterization and localization of the even-skipped protein of Drosophila
.
EMBO J
.
6
,
749
759
.
Frasch
,
M.
&
Levine
,
M.
(
1987
).
Complementary patterns of even-skipped and fushi tarazu expression involve their differential regulation by a common set of segmentation genes in Drosophila
.
Genes & Development
1
,
981
995
.
Frasch
,
M.
,
Warrior
,
R.
,
Tugwood
,
J.
&
Levine
,
M.
(
1988
).
Molecular analysis of even-skipped mutants in Drosophila development
.
Genes & Development
2
,
1824
1838
.
Garcia-Bellido
,
A.
,
Lawrence
,
P. A.
&
Morata
,
G.
(
1979
).
Compartments in animal development
.
Sei. Am
.
241
,
102
110
.
Gergen
,
J. P.
,
Coulter
,
D.
&
Wieschaus
,
E.
(
1986
).
Segmental pattern and blastoderm cell identities
.
In Gametogenesis and the Early Embryo
, pp.
195
220
.
New York
:
Alan R. Liss
.
Gbrgen
,
J. P.
&
Wieschaus
,
E
(
1986
).
Dosage requirements for runt in the segmentation of Drosophila embryos
.
Cell
45
,
289
299
.
Gubb
,
D.
(
1985
).
Further studies on engrailed mutants in Drosophila melanogaster
.
Roux’s Arch, devl Biol
.
194
,
236
246
.
Hafen
,
E.
,
Kuroiwa
,
A.
&
Gehring
,
W. J.
(
1984
).
Spatial distribution of transcripts from the segmentation gene fushi tarazu during Drosophila embryonic development
.
Cell
37
,
833
841
.
Harding
,
K.
,
Rushlow
,
C.
,
Doyle
,
H. J.
,
Hoey
,
T.
&
Levine
,
M.
(
1986
).
Cross-regulatory interactions among pair-rule genes in Drosophila
.
Science
233
,
953
959
.
Hiromi
,
Y.
,
Kuroiwa
,
A.
&
Gehring
,
W. J.
(
1985
).
Control elements of the Drosophila segmentation gene fushi tarazu
.
Cell
43
,
603
613
.
Howard
,
K.
&
Ingham
,
P.
(
1986
).
Regulatory interactions between the segmentation genes fushi tarazu, hairy, and engrailed in the Drosophila blastoderm
.
Cell
44
,
949
957
.
Ingham
,
P. W.
(
1988
).
The molecular genetics of embryonic pattern formation in Drosophila
.
Nature. Lond
.
335
,
25
34
.
Ingham
,
P. W.
,
Baker
,
N. E.
&
Martinez-Arias
,
A.
(
1988
).
Regulation of segment polarity genes in the Drosophila blastoderm by fushi tarazu and even skipped
.
Nature, Lond
.
331
,
73
75
.
Ingham
,
P. W.
,
Howard
,
K. R.
&
Ish-Horowicz
,
D.
(
1985b
).
Transcription pattern of the Drosophila segmentation gene hairy
.
Nature, Lond
.
318
,
439
445
.
Ingham
,
P. W.
,
Pinchin
,
S. M.
,
Howard
,
K. R.
&
Ish-Horowicz
,
D.
(
1985a
).
Genetic analysis of the hairy locus in Drosophila melanogaster
.
Genetics
111
,
463
486
.
Kilcherr
,
F.
,
Baumgartner
,
S.
,
Bopp
,
D.
,
Frei
,
E.
&
Noll
,
M.
(
1986
).
Isolation of the paired gene of Drosophila and its spatial expression during early embryogenesis
.
Nature, Lond
.
321
,
493
499
.
Krause
,
H. M.
,
Klemenz
,
R.
&
Gehring
,
W. J.
(
1988
).
Expression, modification, and localization of the fushi tarazu protein in Drosophila embryos
.
Genes & Development
2
,
1021
1036
.
Lacalli
,
T. C.
,
Wilkinson
,
D. A.
&
Harrison
,
L. G.
(
1988
).
Theoretical aspects of stripe formation in relation to Drosophila segmentation
.
Development
103
,
105
113
.
Lawrence
,
P. A.
(
1973
).
The development of spatial patterns in the integument of insects
.
In Developmental Systems: Insects
, vol.
2
(ed.
S. J.
Counce
&
C. H.
Waddington
), pp.
157
209
.
London and New York
:
Academic Press
.
Lawrence
,
P. A.
(
1987
).
Pair rule genes; do they paint stripes or draw lines?
Cell
51
,
879
880
.
Lawrence
,
P. A.
(
1988
).
The present status of the parasegment
.
Symp. Brit. Soc. devl Biol. Development (In press)
.
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
.
MacDonald
,
P. M.
,
Ingham
,
P. W.
&
Struhl
,
G.
(
1986
).
Isolation, structure, and expression of even-skipped: A second pair-rule gene of Drosophila containing a homeo box
.
Cell
41
,
721
734
.
Marcus
,
W.
(
1962
).
Untersuchungen über die Polarit ä t der Rumpfhaut von Schmetterlingen
.
Wilhelm Roux Arch. EntwMech. Org
.
154
,
56
102
.
Martinez-Arias
,
A.
,
Baker
,
N. E.
&
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
.
Martinez-Arias
,
A.
&
Lawrence
,
P. A.
(
1985
).
Parasegments and compartments in the Drosophila embryo
.
Nature, Lond
.
313
,
639
642
.
Nüsslein-Volhard
,
C.
,
Frohnhöfer
,
H. G.
&
Lehmann
,
R.
(
1987
).
Determination of anteroposterior polarity in Drosophila
.
Science
238
,
1675
1681
.
Nüsslein-Volhard
,
C.
,
Kluding
,
H.
&
Jürgens
,
G.
(
1985
).
Genes affecting the segmental subdivision of the Drosophila embryo
.
Cold Spring Harbor Symp. quant. Biol
.
50
,
145
154
.
Nüsslein-Volhard
,
C.
&
Wieschaus
,
E.
(
1980
).
Mutations affecting segment number and polarity in Drosophila
.
Nature, Lond
.
287
,
795
801
.
Nüsslein-Volhard
,
C.
,
Wieschaus
,
E.
&
Jürgens
,
G.
(
1982
).
Segmentation in Drosophila, a genetic analysis
.
Verh. Dtsch. Zool. Ges
.
91
104
.
Nüsslein-Volhard
,
C.
,
Wieschaus
,
E.
&
Kluding
,
H.
(
1984
).
Mutations affecting the pattern of the larval cuticle in Drosophila melanogaster. 1. Zygotic loci on the second chromosome
.
Wilhelm Roux’s Arch, devl Biol
.
193
,
267
282
.
Sander
,
K.
(
1960
).
Analyse des ooplasmatischen Reaktionssystems von Euscelis plebejus Fall (Cicadina) durch Isolieren und Kombinieren von Keimteilen II. Mitt.: Die Differenzierungsleistungen nach Verlagern von Hinterpolmaterial
.
Wilhelm Roux Arch. EntwMech. Org
.
151
,
660
707
.
Scott
,
M. P.
&
Carroll
,
S. B.
(
1987
).
The segmentation and homeotic gene network in early Drosophila development
.
Cell
51
,
689
698
.
Scott
,
M. P.
&
O’Farrell
,
P. H.
(
1986
).
Spatial programming of gene expression in earlv Drosophila embryogenesis
.
A. Rev. Cell. Biol
.
2
,
49
80
.
Weiner
,
A. J.
,
Scott
,
M. P.
&
Kaufman
,
T. C.
(
1984
).
A molecular analysis of fushi tarazu, a gene in Drosophila melanogaster that encodes a product affecting segment number and cell fate
.
Cell
37
,
843
851
.
Weir
,
M. P.
,
Edgar
,
B. A.
,
Kornberg
,
T.
&
Schubiger
,
G.
(
1988
).
Spatial regulation of engrailed expression in the Drosophila embryo
.
Genes & Development
2
,
1194
1203
.
Weir
,
M. P.
&
Kornberg
,
T.
(
1985
).
Patterns of engrailed and fushi tarazu transcripts reveal novel intermediate stages in Drosophila segmentation
.
Nature, Lond
.
318
,
433
445
.
Wieschaus
,
E.
&
Nüsslein-Volhard
,
C.
(
1986
).
Looking at embryos
.
In Drosophila: A Practical Approach
(ed.
D. B.
Roberts
), pp.
199
228
.
Oxford
:
IRL Press
.
Wolpert
,
L.
(
1969
).
Positional information and the spatial pattern of cellular differentiation
.
J. theor. Biol
.
25
,
1
47
.