To understand the nature of the regulatory signals impinging on the second promoter of the Antennapedia gene (Antp P2), analysis of its expression in mutants and in inhibitory drug injected embryos has been carried out. The maternally-active gene osk is identified as one of two general repressors of P2 which prevent Antp transcription until division cycle 14. Products of the zygoticaliy-active segmentation genes ftz, hb, Kr, gt and kni then act as activators or repressors of Antp P2 in a combinatorial fashion. The timing of these events, and their positive versus negative nature, is critical for generating the expression patterns normal for Antp.

The ability to isolate mutations in Drosophila melanogaster has made this organism a central one in the study of biological mechanisms. When it became possible to clone genes specified by mutations without recourse to their protein products, the study of mutations which altered the morphology of the fly were considered especially attractive since they represented a means to investigate the molecular basis of how an organism is assembled during development.

One type of abnormal development is associated with mutations in homeotic genes. These aberrations lead to the replacement of one or more body structures with others characteristic of a different part of the body. Two of the Drosophila homeotic genes, Ultrabithorax and Antennapedia, were amongst the first to be cloned and undergo initial molecular characterization (Bender et al. 1983; Scott et al. 1983; Garber et al. 1983). The results of these efforts have led to the finding that homeotic genes, as well as a group of the loci responsible for formation of segmental patterning in the larval body, encode transcription factors (Scott and Weiner, 1984). For homeotic genes we still know little about their targets of action, except for cases in which they act on themselves (autoregulation) or on each other. Targets for the action of the segmentation genes that have been identified are each other and the homeotic genes. The work of many labs is leading to an elegant map of gene-gene interactions that describes how the early phases of Drosophila development are controlled (see Akam, 1987; Ingham, 1988).

Antp is a homeotic gene which helps govern the proper formation of the thorax segments in the larval and adult Drosophila (Wakimoto and Kaufman, 1981; Struhl, 1981). Protein products of the Antp gene contain homeodomains, indicating they function by binding DNA (Scott and Weiner, 1984; McGinnis et al. 1984). Antp is over 100kb long, and transcription is driven from two independently regulated promoters (Stroeher et al. 1986; Laughon et al. 1986; Jorgensen and Garber, 1987). The first promoter, Pl, is situated around Okb near the start of the first exon. The second, P2, is situated at about 65 kb and is near to the start of the third exon. Transcripts controlled by these two promoters are found over the course of embryo development in different but overlapping sets of cells (E. M. Jorgensen and R. L. Garber, manuscript in preparation). That they govern different functions is seen by the different phenotypes arising in mutants bearing lesions in one or the other Antp promoter (Abbott and Kaufman, 1986; E. M. Jorgensen and R. L. Garber, manuscript in preparation).

The work we report here is an investigation into the gene interactions necessary to activate transcription of the homeotic genes. We have focussed on the Antp gene promoter P2 with the desire to determine the factors responsible for its activity. In order to test whether information flows to Antp directly from the segmentation genes, Antp transcription was studied in mutants lacking one or more of these genes. Since the loss of any such gene ultimately leads to highly abnormal development of the embryo, our studies were performed at very early stages of development. In addition, we have used pharmacological approaches to ask general questions about the nature, the location and the timing of regulatory signals directed to Antp. The changes in Antp activity we observed indicate that Antp promoter P2 receives both negative and positive input from the products of segmentation genes, including maternal effect loci. In the end we point out differences between Antp and the segmentation gene fushi taraza that occur in the application of genetic control.

Embryo collection and preparation

Eggs were collected from three to ten day old Drosophila melanogaster flies onto food plates dabbed with yeast and acetic acid. Embryos were dechorionated and fixed as described by Jorgensen and Garber (manuscript in preparation).

Ectopic hb expression

The fly line HSHB-3 (Struhl, 1989) contains a transformed copy of the hunchback (hb) gene controlled by the heat shock gene hsp70 promoter, and it was used to produce hb expression throughout the embryo. Twenty minute collections of eggs were heat shocked for 15 min at 37°C during nuclear cleavage cycles 10–11. These were divided in two, with half incubated at 25 °C until cycle 14 and fixed for in situ hybridizations while the rest were incubated to the end of embryogenesis and used for cuticle preparations.

Microinjection

The procedures followed were those of Edgar and Schubiger (1986). Cycloheximide (Sigma) was injected into embryos at 1mg ml to yield a final intracellular concentration of approximately 20 μg ml −1.

In situ hybridization

Embryos were embedded in paraffin (Tissueprep2 from Fisher or Ameriffin from Scientific Products) and 2.5 /im sections were cut and treated as described in Jorgensen and Garber (manuscript in preparation). 35S-labelled antisense RNA probes, the radioactive in situ hybridization procedure and autoradiography were as described in Jorgensen and Garber (1987). The Antp probes, derived from exon 1 and exon 3 genomic sequences, hybridize uniquely to transcripts from promotor 1 or promotor 2, respectively (Jorgensen and Garber, 1987). The fushi tarazu (ftz) probe was a genomic sequence (a 2.1 kb EcoRI-Psri fragment; see Laughon and Scott, 1984).

Non-radioactive in situ hybridizations were as described in Tautz and Pfeifle (1989) with probe preparation by random primer labeling suggested by B. Edgar and C. Oh (unpublished results). Digoxygenin-labelled probe was synthesized using 50–150 ng of gel-purified DNA fragments and the Genius non-radioactive DNA labelling and detection system (Boehringer-Mannheim).

Nuclear staining

To determine the exact nuclear division stage of an embryo, the nuclear density was determined by staining embryos with the DNA-specific dye Hoechst 33258 at 10μgml−1 in PBS for three minutes.

Fly stocks

The alleles used in this study were ftzW20 (Wakimoto et al. 1984), gts11. Kr9 (Preiss et al. 1985), kiu11D4S and hb14f21 (Jürgens et al. 1984), nosL7 and osk,h6 (Lehmann and Nusslein-Volhard, 1986).

Construction of ftz osk double mutants

Female th st roe pposk166/TM3, Sb pF flies were crossed to Ki ftZW20/TM3, Sb pp males. The th st roe pposk166/ Ki ftzW20 female progeny were crossed to Kt ftzW2n/TM3. Sb p1 males. Ki Sb pp individuals were saved as putative double mutants and tested for both ftz and osk mutant phenotypes by examination of cuticle preparations.

The initial P2 pattern

The fertilized Drosophila embryo undergoes a series of 13 nuclear divisions before cellularization of the somatic cell line begins (Foe and Alberts, 1983). Prior to nuclear division cycle 14 there is no detectable transcription directed by the Antp promoter P2. By the end of cycle 14, however, Antp RNAs from P2 can be found in three spatial domains (Fig. 1). Domain 1 is a circumferential band of Antp transcripts in cells of parasegment 4 (Martinez-Arias and Lawrence, 1985). It appears about 30min after the last nuclear division. Domains 2 and 3 appear slightly later in cycle 14. Domain 2 lies within parasegment 6 (PS6), but it is restricted to the ventral portion of this band. Domain 3 is within PS14, but only in laterally-positioned cells. In addition, there is a low level of Antp P2 detectable throughout the area between PS4 and PS14.

Fig. 1.

Antp promoter P2 transcripts localized in a wildtype embryo. Three domains of expression are visible in parasegments 4, 6 and 14. Although this embryo has just begun germ band elongation, the pattern is characteristic of the late blastoderm stage. In situ hybridization is with a digoxygenin-labeled Antp exon 3 probe, which is specific for P2 RNAs.

Fig. 1.

Antp promoter P2 transcripts localized in a wildtype embryo. Three domains of expression are visible in parasegments 4, 6 and 14. Although this embryo has just begun germ band elongation, the pattern is characteristic of the late blastoderm stage. In situ hybridization is with a digoxygenin-labeled Antp exon 3 probe, which is specific for P2 RNAs.

Cycloheximide injection

To learn what roles the synthesis of new proteins play in the evolution of the Antp P2 expression pattern, an inhibitor of protein synthesis was injected into embryos around the time of blastoderm formation. Animals injected with cycloheximide prior to cycle 14 do not show’ P2 expression, whereas drug added during cycle 14 arrests the developing transcription pattern at the time of injection (Fig. 2). Thus embryos treated midway through cycle 14 and analyzed one hour later by in situ hybridization had the PS4 band, but they did not gain the PS6 and PS14 bands which would normally become visible within this time period. Embryos injected at the end of cycle 14, however, showed arrested development and had all three Antp P2 domains.

Fig. 2.

The effect of cycloheximide treatment on Antp P2 transcription in wild-type embryos. (A,C,E,G) Antp P2 RNA patterns. (B,D,F,H) nuclear staining with Hoechst dye to determine the exact stage of development. (A-D) control embryos. (E-H) cycloheximide-injected embryos. Cycloheximide injection of wild-type embryos leads to normal timing and pattern of Antp P2 activation. The signal is stronger, however, and reveals lower levels of transcription posterior to the main band in parasegment 4. Control embryos were fixed at the time experimental embryos were injected. After injection, experimental embryos were aged for 60 min prior to fixation (cycloheximide arrests the cell cycle at the time of injection). A 35S-labeled UTP, anti-sense RNA probe was used for in situ hybridization. Autoradiography was for 2 days.

Fig. 2.

The effect of cycloheximide treatment on Antp P2 transcription in wild-type embryos. (A,C,E,G) Antp P2 RNA patterns. (B,D,F,H) nuclear staining with Hoechst dye to determine the exact stage of development. (A-D) control embryos. (E-H) cycloheximide-injected embryos. Cycloheximide injection of wild-type embryos leads to normal timing and pattern of Antp P2 activation. The signal is stronger, however, and reveals lower levels of transcription posterior to the main band in parasegment 4. Control embryos were fixed at the time experimental embryos were injected. After injection, experimental embryos were aged for 60 min prior to fixation (cycloheximide arrests the cell cycle at the time of injection). A 35S-labeled UTP, anti-sense RNA probe was used for in situ hybridization. Autoradiography was for 2 days.

The development of the three domains of P2, therefore, requires the synthesis of one or more proteins during cellular blastoderm. Formally these behave as activators of the Antp promoter.

P2 expression in segmentation gene mutants

Identification of the genes whose products are essential for the early transcription pattern of Antp P2 has been approached by mapping the position of Antp RNAs within mutant embryos lacking the function of segmentation genes. Because the three domains of Antp P2 expression align with even-numbered parasegments, as does the pair-rule gene ftz, speculation was that ftz might play a role in controlling Antp expression. This has been supported by finding that the P2 pattern is eliminated in ftz embryos (Ingham and Martinez-Arias, 1986; E. M Jorgensen and R. L. Garber, unpublished results). It should be pointed out that later in embryogenesis Antp P2 expression still occurs in the same background, indicating that ftz plays an essential role for only some of the P2 signals (unpublished results).

The testing of other mutant genotypes showed that P2 domain 1 (in PS4) has a dependence on the gap gene hunchback (hb), while the ventrally-located domain 2 is removed when the gap gene Kriippel (Kr) is mutated (Irish et al. 1989; Fig. 3A and B). Thus it was concluded that ftz behaves as an activator of Antp P2 in three regions of the embryo at cellular blastoderm, sharing responsibility with either hb or Kr for the first two regions. An equivalent coactivator has not been identified for domain 3.

Fig. 3.

Altered Antp P2 expression patterns in segmentation gene mutants. The complexity of the initial Antp P2 pattern is reduced in embryos lacking the activity of the Antp activators hunchback (A) and Krúppel (B). while it is expanded in embryos lacking the Antp repressors giant (C) and knirps (D). A. B and D are side views. C is a ventral view showing the anterior expansion of Antp P2 into parasegments 2–3 and 12–13.

Fig. 3.

Altered Antp P2 expression patterns in segmentation gene mutants. The complexity of the initial Antp P2 pattern is reduced in embryos lacking the activity of the Antp activators hunchback (A) and Krúppel (B). while it is expanded in embryos lacking the Antp repressors giant (C) and knirps (D). A. B and D are side views. C is a ventral view showing the anterior expansion of Antp P2 into parasegments 2–3 and 12–13.

The above results clearly are not the whole story. The three known activators are found in stripes completely surrounding the embryo, yet as described above the Antp P2 pattern involves substantially fewer cells than expected in domains 2 and 3. Second, there are cells that don’t show Antp P2 transcripts but which appear to be equivalent to those in PS4 (PS2 has both ftz and hb active) and to those in PS6 (PS8 has both ftz and Kr present). With the following experiments we have begun to address some of these questions.

giant repression of Antp P2

One hypothesis for the failure of Antp to come on in all the ftz stripes is that repressive factors are also active. Candidates for such, as judged by their spatial expression patterns, are the genes giant (gt) and knirps (kni). gt expression as determined by molecular analysis and by the area altered in mutants includes PS2 (ftz stripe 1) and PS 12 (ftz stripe 6) (Mohler et al. 1989).

Antp localization in the allele gt*11 differs from wild-type by a general anterior expansion by two of the P2 expression domains (Fig. 3C). Domain 1, rather than being confined to PS4, now includes PS2-4 (ftz stripes 1 and 2 plus the interband they define). Domain 3, the P2 pattern in PS14, also expands in an anterior direction so that it includes PS12-14 (although PS13 is weak). Domain 2 does not appear to be changed.

gt, therefore, appears to repress Antp P2 directly. The loss of gt activity allows P2 to come on in regions of the embryo where the activation signals by ftz and hb were positive but the gt protein in wild-type animals normally over-ruled.

knirps

Proper formation of the abdomen of embryos requires the gap gene knirps (kni) (Nauber et al. 1988). The pattern of Antp P2 in embryos lacking knirps (kni) function suggests that its normal activity leads to a repression of P2 transcription in the middle of the animal (PS6-12). The loss of kni leads to expansion of domain 2 in a posterior direction (Fig. 3D). Although this domain is usually described as being ventrally restricted, non-radioactive in situ RNA localization reveals that Antp P2 is ventrally strong in PS6 but that it also extends more weakly and in a smaller number of cells in a ring around the embryo. In kni embryos, Antp domain 2 expands posteriorly back to PS12.

Thus, gt and kni separately act to keep Antp P2 expression out of the middle of the embryo. Loss of either of these genes results in an expansion of P2 positive cells, as long as the region does not contain activity of the other repressive function. However, because kni embryos also have Kr and hb distributions that are abnormal, and because both of these are direct Antp regulators, the effect of kni on Antp P2 could be indirect.

P2 expression in oskar and nanos

Although there are several steps in the regulatory hierarchy between the maternally-active body axis genes and the zygotically-active homeotic genes (Carroll et al. 1986; Frôhnhofer and Nüsslein-Volhard, 1986), we chose to investigate whether the latter could play a direct role in guiding Antp P2 expression. Our primary focus has been on the posterior group gene oskar (osk) (Lehmann and Nüsslein-Volhard, 1986). Eggs derived from osk mothers show P2 transcription starting during cycle 14, as occurs in wild-type embryos. The pattern, however, is a broad band ranging from 18-58% of egg length, the anterior border coinciding with that found in wild-type (Fig. 4A). A similar pattern for P2 is seen in embryos from another member of the posterior group, nanos (nos), although the two are not identical. Antp P2 transcription in embryos derived from nos mothers shows elevated expression in three bands of cells over a broad swath (Fig. 4B).

Fig. 4.

Antp P2 transcripts in the maternal effect mutants oskar (A) and nanos (B). While these genes are both involved in organizing the embryo’s posterior end, their loss affects the P2 promoter differently.

Fig. 4.

Antp P2 transcripts in the maternal effect mutants oskar (A) and nanos (B). While these genes are both involved in organizing the embryo’s posterior end, their loss affects the P2 promoter differently.

To learn whether the roles of repressor and activator proteins in these maternal effect mutants reflected that in wild-type or not, we injected cycloheximide into osk mutant embryos at different stages of early development (Fig. 5). When injections occurred during cycle 11 or earlier, no P2 RNAs are found (not shown). Injection during cycle 13 led to the substantial accumulation of transcripts one cycle precociously, and the pattern observed was the osk mutant one. Very low signal levels appear in cycle 12, indicating transcription begins even earlier.

Fig. 5.

The effect of cycloheximide treatment on Antp P2 transcription in osk embryos. In the osk mutant background, P2 transcripts are first detected during cycle 14 (A-C). When cycloheximide is injected into similar embryos, P2 RNAs are easily detected during cycle 13, and this promoter may be weakly active during cycle 12 (D-F).

Fig. 5.

The effect of cycloheximide treatment on Antp P2 transcription in osk embryos. In the osk mutant background, P2 transcripts are first detected during cycle 14 (A-C). When cycloheximide is injected into similar embryos, P2 RNAs are easily detected during cycle 13, and this promoter may be weakly active during cycle 12 (D-F).

Heat shock - hunchback

To determine whether the early activation on Antp P2 was due to the precocious expression of Antp activators, several experimental approaches were followed. The first concerned the distribution of the activator gene hunchback, hb, whose matemal RNA contributions are usually restricted to the anterior portion of the egg by the posterior group genes. It is known that the loss of osk maternal activity leads to a uniform distribution of hb RNA over the early embryo. Could this, in conjunction with the observed broad pattern of ftz, be leading to ectopic P2 activation in osk embryos? To investigate this, a transformed fly line bearing hb controlled by the hsp70 heat shock gene promoter (Struhl, 1989) was temperature induced, fixed and Antp P2 RNAs were localized by in situ hybridization. The effectiveness of the heat shock was assayed by splitting up the induced embryos, aging them appropriately, and then making larval cuticle preparations from one half of the animals. Both the mutant cuticles and an altered ftz expression pattern at the cellular blastoderm (only 4 stripes appear) indicated that the heat shock was effective (Fig. 6A and B). The resulting pattern for P2 RNAs, however, appeared wild-type in the hs-hb induced embryos (Fig. 6C). In a wild-type embryo, therefore, expression of hb product everywhere is not sufficient to activate Antp P2 ectopically.

Fig. 6.

The effect of ectopic hunchback expression on Antp P2 transcription. HSHB-3 embryos bear a hb coding region under the control of a heat shock promoter. Heat treatment clearly alters the cuticle pattern (A) and ftz transcription (B). but it causes little change in the Antp P2 pattern. These same features are shown in wild-type embryos for comparison (D-F).

Fig. 6.

The effect of ectopic hunchback expression on Antp P2 transcription. HSHB-3 embryos bear a hb coding region under the control of a heat shock promoter. Heat treatment clearly alters the cuticle pattern (A) and ftz transcription (B). but it causes little change in the Antp P2 pattern. These same features are shown in wild-type embryos for comparison (D-F).

ftz; osk embryos

A second direction we followed was to genetically construct an embryo where we could determine whether the P2 expression in osk embryos was dependent on ftz. Tliis was done by mating ftz osk/+ osk mothers to ftz +/+ + males. While all the embryos from this cross should have the osk phenotype, only a quarter are expected to be homozygous mutant for ftz. The result was that P2 localization in all the embryos appeared identical to that in the osk genotype, indicating that P2 transcription in this mutant background was independent of the ftz gene (not shown).

Transcription is missing from the early Drosophila embryo until nuclear division cycle 10 (Zalokar, 1976). At this time the genome of the zygote becomes transcriptionally active as judged by the appearance of new RNAs. Edgar et al. (1986, 1988) employed cycloheximide inhibition of new protein synthesis to learn about the activation of the segmentation genes. We have extended this approach by observing the properties of the homeotic gene Antennapedia.

Cycloheximide experiments

One of the genes Edgar followed was fushi tarazu (ftz), a member of the pair-rule gene class. Previous studies revealed that there is normally an evolution of the ftz RNA pattern during cycle 14, changing from a broad swath of cells actively transcribing ftz during early stages of the cycle to a set of seven broad stripes in mid cycle 14 (Weir and Kornberg, 1985). By late cycle 14 the stripes became further narrowed. Edgar et al. (1988) injected cycloheximide at different times during cycle 14 and analyzed ftz RNA localization an hour later (see diagram in Fig. 7A). They saw that ftz expression expanded in the injected embryos to a point of losing stripes and mimicking a pattern characteristic of ftz during earlier’development. The continued transcription suggested that the ftz gene remained active and the transcription apparatus functional, but that the repressive activities that normally generated the stripes and kept ftz off at the anterior and posterior ends of the embryo were mostly lost.

Fig. 7.

Diagram representing the different effects of cycloheximide treatment at different times during nuclear cycle 14 on the induction of transcription of ftz and Antp P2. (A) Cycloheximide inhibition of new protein synthesis causes ftz to be found in a broader pattern than normal of that stage. This suggests that while transcription continues, repressive functions that normally limit ftz destribution are lost. (B) Cycloheximide injection freezes the Antp P2 pattern at the time of treatment, indicating that developments of the new pattern elements requires new protein to be made, ftz and Antp P2 are therefore under very different control mechanisms.

Fig. 7.

Diagram representing the different effects of cycloheximide treatment at different times during nuclear cycle 14 on the induction of transcription of ftz and Antp P2. (A) Cycloheximide inhibition of new protein synthesis causes ftz to be found in a broader pattern than normal of that stage. This suggests that while transcription continues, repressive functions that normally limit ftz destribution are lost. (B) Cycloheximide injection freezes the Antp P2 pattern at the time of treatment, indicating that developments of the new pattern elements requires new protein to be made, ftz and Antp P2 are therefore under very different control mechanisms.

One interpretation is that transcription of the ftz gene normally begins by its general activation throughout the animal, except at the poles where repressors previously synthesized (perhaps maternally) keep/tz off. Either ftz requires no specific activators and is transcribed ‘passively’, or the activators are translated before cycle 14 (perhaps transcribed maternally) and are abundantly present. Later, as the products of other genes become functional and begin to repress ftz, the stripes appear. Cycloheximide cancels the effect of these repressors, perhaps due to their rapid turnover and need for continual synthesis, allowing ftz to come on essentially everywhere as is normal in early cycle 14.

Cycloheximide does not stop transcription. In fact, the in situ hybridization signals for all the probes that have been tested are especially strong in injected animals. This supports the notion that cycloheximide inhibits the elongation step of translation, leading to ribosome-laden mRNAs which are therefore more stable and accumulate as further transcription occurs. New transcription in injected embryos therefore does not require new proteins. As mentioned above, late in cycle 14 repression also requires no new protein synthesis. This could be due to the accumulation of sufficient mRNAs and protein to override the embryo’s treatment, or it could be due to the locking of the gene into active and inactive chromatin conformations that require no new proteins to be made. Both are clearly speculations at this time.

Antp P2 regulation

We have employed cycloheximide in order to gain more information about Antp expression, and our results differ significantly from those observed by Edgar et al. (1988) for ftz and several other segmentation genes (Fig. 7B). The drug injection freezes the existing Antp transcription pattern, and it does not expand as occurs for ftz. One reason for this could be that the Antp P2 promoter starts out with repressors bound. Despite the ability for transcription to take place on many other genes, Antp is kept silent until mid cycle 14. If the repressors were maternally-derived and were finally degraded by cycle 14, then Antp could come on everywhere. Instead, as the analysis of Antp P2 expression in mutants demonstrates, specific activators (ftz, hb and Kr ) are needed. These could be used to remove the retiring repressors or needed to attract transcription complex formation because the Antp promoter P2 is too weak on its own. Similar to ftz gene transcription, the Antp P2 pattern is stable once established. If ftz protein is needed to activate the P2 promoter, it is not required for long or the protein is very stable.

The point to be made is that the segmentation gene ftz starts off with general activation followed by specific repression some time later. Antp P2, on the other hand, starts out with repression which is later followed by specific activation. The same events - repression and activation - take place. But by specifying different patterns of timing, very specific regulation of transcription can take place.

Another aspect of timing is indicated by the analysis of Antp P2 transcription in the mutant oskar, with and without cycloheximide treatment. Comparing these two conditions has shown that both osk and some unidentified gene contribute products which inhibit transcription by P2 until mid cycle 14. Their action in this regard is redundant, because the loss of either alone does not change the time when Antp RNAs first appear. Only the loss of both leads to precocious P2 transcription in cycle 12. One reason this could be important to the developing embryo is suggested by the gradual evolution in the/tz stripes. If the need for Antp expression was restricted to several of the even-numbered parasegments and not more, then keeping P2 off until the ftz pattern had fully matured would ensure a localized Antp expression. The redundancy makes this commitment even stronger.

One aside can be added here concerning the results of comparing Antp P2 expression in osk and nos embryos at both blastoderm and gastrulation stages. Although the posterior group genes generally form a linear pathway of information flow, the finding that the P2 patterns are not identical for the two genotypes adds another example where the pathway is not simply linear.

Specific activators

For Antp P2, the action of a series of proteins is needed generate its complex transcription pattern (summarized in Fig. 8). One of the proteins whose synthesis is required during cycle 14 is ftz. The products of two other genes also behave as Antp P2 activators, hb and Kr, however, are different from ftz in that each is required for a specific portion of the Antp pattern. Since the second Antp domain appears later than the first, and it is dependent on Kr expression, new Kr synthesis must be required during cellular blastoderm for Antp activation.

Fig. 8.

Schematic diagram indicating activators and repression functions which are known to be used to regulate the initial Antp P2 transcription during embryogenesis. Two of the domains require the presence of both ftz and either hb or Kr to be formed. Early repression by the osk product and an unknown one act to keep P2 off until mid cycle 14. Once activated. P2 is restricted dorsally and ventrally in different regions of the embryo (see Fig. I), as well as anteriorly and posteriorly by gt and kni.

Fig. 8.

Schematic diagram indicating activators and repression functions which are known to be used to regulate the initial Antp P2 transcription during embryogenesis. Two of the domains require the presence of both ftz and either hb or Kr to be formed. Early repression by the osk product and an unknown one act to keep P2 off until mid cycle 14. Once activated. P2 is restricted dorsally and ventrally in different regions of the embryo (see Fig. I), as well as anteriorly and posteriorly by gt and kni.

Further refinement of the pattern of Antp P2 activation in the early embryo also requires several repressors. The anterior boundary of Antp P2 transcription is PS4. gt seems to define this limit, and when gt function is missing P2 extends forward to include PS2. If P2 expression requires both/tz and a co-activator, like hb, then it is surprising that P2s extension in gt mutants includes more than just the next anterior ftz stripe (PS2). Either this is a time when P2 can initiate transcription without the help offtz, or P2 is responding to the low levels of ftz protein transiently present between the late cycle 14 ftz stripes (Karr and Kornberg, 1989).

In a genetic background missing the maternal product oskar, and thus lacking a component of the posterior organizing function, Antp P2 transcription is spatially relaxed and includes a large portion of the embryo. Whether this indicates a direct effect of the osk product on the P2 promoter is not proven. However, P2 expression in osk is ftz-independent, as judged by identical P2 localization patterns in osk and osk; ftz embryos. In osk, therefore, misexpression of a prime activator such as hunchback (Irish et al. 1989) may not be the cause of the altered Antp pattern. Other genes need to be checked directly as ftz has. As mentioned at the beginning of the discussion, the need for activators could be to aid in removal of the general repressors initially keeping Antp P2 inactive. In the absence of at least one of these repressors, namely in osk mutants, the activator ftz is still made but the dependence of Antp P2 on ftz is lost. As we learn more, these activators may turn out instead to be called antirepressors.

Combinatorial regulation

Antp P2 responds to the presence of both activators and repressors. It appears that the combination offtz and hb or ftz and Kr are need to create Antp domains 1 and 2. It is interesting that ftz plus hb coexpression is not sufficient when the gt gene product is also present. Thus, the repressor activity dominates in this case.

We note that although the other regional repressor kni could repress Kr and thereby repress P2 expression in the middle of the animal, under the hs-hb experimental conditions there was ftz plus hb in the mid region of the embryo. The fact that new Aníp P2 transcription did not occur indicates that kni may repress Antp P2 directly.

We especially wish to thank Drs Bruce Edgar and Gerold Schubiger for their comments, suggestions and encouragement for carrying out these experiments. We are grateful for the fly strains made available to us by Gary Struhl, Eric Wieschaus, Ruth Lehmann, Landon Boring, and the national Drosophila stock centers at Bowling Green and Bloomington. Support for this work was provided by the National Science Foundation grant DMB-8803324.

Abbott
,
M. K.
and
Kaufman
,
T. C.
(
1986
).
The relationship between the functional complexity and the molecular organization of the Antennapedia locus of Drosophila melanogasier
.
Generics
114
,
919
942
.
Akam
,
M.
(
1987
).
The molecular basis for metameric pattern in the Drosophila embryo
.
Development
101
,
1
22
.
Bender
,
W.
,
Weiffenbach
,
B.
,
Karch
,
F.
,
Beachy
,
P. A.
,
Peifer
,
M.
,
Spierer
,
P.
,
Lewis
,
E. B.
and
Hogness
,
D. S.
(
1983
).
Molecular genetics of the bithorax complex in Drosophila melanogasier
.
Science
221
,
23
29
.
Carroll
,
S. B.
,
Winslow
,
G. M.
,
Schüpbach
,
T.
and
Scott
,
M. P.
(
1986
).
Maternal control of segmentation gene expression
.
Nature
323
,
278
280
.
Edgar
,
B. A.
,
Odell
,
G. M.
and
Schubiger
,
G.
(
1988
).
A genetic switch, based on negative regulation, sharpens stripes in Drosophila embryos
.
Devi Genetics
10
,
124
142
.
Edgar
,
B. A.
and
Schubiger
,
G.
(
1986
).
Parameters controlling transcriptional activation during early Drosophila development
.
Cell
44
,
871
877
.
Edgar
,
B. A.
,
Weir
,
M. P.
,
Schubiger
,
G.
and
Kornberg
,
T.
(
1986
).
Repression and turnover pattern fushi taraza in the early Drosophila embryo
Cell
47
,
747
754
.
Foe
,
V. E.
and
Alberts
,
B. M.
(
1983
).
Studies of nuclear and cytoplasmic behaviour during the five mitotic cycles that procede gastrulation in Drosophila embryogenesis
.
J. Cell Sci
.
61
,
31
70
.
Frôhnhofer
,
H. G.
and
Nüsslein-Volhard
,
C.
(
1986
).
Organization of the anterior pattern in the Drosophila embryo by the maternal gene bicoid
.
Nature
324
,
120
125
.
Garber
,
R. L.
,
Kuroiwa
,
A.
and
Gehring
,
W. J.
(
1983
).
Genomic and cDNA clones of the homeotic locus Antennapedia in Drosophila
.
EMBO J
.
2
,
2027
2036
.
Ingham
,
P. W
(
1988
).
The molecular genetics of embryonic pattern formation in Drosophila
.
Nature
335
,
25
34
.
Ingham
,
P. W.
and
Martinez-Arias
,
A.
(
1986
).
The correct activation of Antennapedia and bithorax complex genes requires the fushi tarazu gene
.
Nature
324
,
592
597
.
Irish
,
V. F.
,
Martinez-Arias
,
A.
and
Akam
,
M.
(
1989
).
Spatial regulation of the Antennapedia and Ultrabithorax homeotic genes during Drosophila early development
.
EMBO J
.
8
,
1527
1537
.
Jorgensen
,
E. M.
and
Garber
,
R. L.
(
1987
).
Function and misfunction of the two promoters of the Drosophila Antennapedia gene
.
Genes Dev
.
1
,
544
555
.
Jürgens
,
G.
,
Wieschaus Nüsslein-Volhard
,
C.
and
Kluding
,
H.
(
1984
).
Mutations affecting the pattern of the larval cuticle in Drosophila melanogasier. II. Zygotic loci on the third chromosome
.
Roux’s Arch. Devi Biol
.
193
,
283
295
.
Karr
,
T. L.
and
Kornberg
,
T. B.
(
1989
).
fushi taraza protein expression in the cellular blastoderm of Drosophila detected using a novel imaging technique
.
Development
106
,
95
103
.
Laughon
,
A.
,
Boulet
,
A. M.
,
Bermingham
,
J. R.
,
Laymon
,
R. A.
and
Scott
,
M. P.
(
1986
).
Structure of transcripts from the homeotic Antennapedia gene of Drosophila melanogasier. Two promoters control the major protein coding region
.
Molec. cell Biol
.
6
,
4676
4689
.
Laughon
,
A.
and
Scott
,
M. P.
(
1984
).
Sequence of a Drosophila segmentation gene: Protein structure homology with DNA-binding proteins
.
Nature
310
,
25
31
.
Lehmann
,
R.
and
Nüsslein-Volhard
,
C.
(
1986
).
Abdominal segmentation, pole cell formation, and embryonic polarity require localized activity of oskar, a maternal gene in Drosophila
.
Cell
47
,
141
152
Martinez-Arias
,
A.
and
Lawrence
,
P. A.
(
1985
).
Parasegments and compartments in the Drosophila embryo
.
Nature
313
,
639
642
.
Mcginnis
,
W.
,
Garber
,
R. L.
,
Wirz
,
J.
,
Kuroiwa
,
A.
and
Gehring
,
W. J.
(
1984
).
A homologous protein-coding sequence in Drosophila homeotic genes and its conservation in other metazoans
Cell
37
,
403
408
.
Mohler
,
J.
,
Eldon
,
E. D.
and
Pirrotta
,
V.
(
1989
).
A novel spatial transcription pattern associated with the segmentation gene, giant, of Drosophila
.
EMBO J
.
8
,
1539
1548
.
Nauber
,
U.
,
Pankratz
,
M. J.
,
Kienlin
,
A.
,
Seifert
,
E.
,
Klemm
,
U.
and
Jâckle
,
H.
(
1988
).
Abdominal segmentation of the Drosophila embryo requires a hormone receptor-like protein encoded by the gap gene knirps
.
Nature
336
,
489
492
.
Scott
,
M. P.
and
Weiner
,
A. J.
(
1984
).
Structural relationships among genes that control development: Sequence homology between the Antennapedia, Ultrabithorax, and fushi tarazu loci of Drosophila
.
Proc. natn. Acad. Sci U.S.A
.
81
,
4115
4119
.
Scott
,
M. P.
,
Weiner
,
A. J.
,
Hazelrigg
,
T. I.
,
Polisky
,
B. A.
,
Pirrotta
,
V.
,
Scalenghe
,
F.
and
Kaufman
,
T. C.
(
1983
).
The molecular organization of the Antennapedia locus of Drosophila
.
Cell
35
,
763
776
.
Stroeher
,
V. L.
,
Jorgensen
,
E. M.
and
Garber
,
R. L.
(
1986
).
Multiple transcripts from the Antennapedia gene of Drosophila melanogasier
.
Molec. cell Biol
.
6
,
4667
4675
.
Struhl
,
G.
(
1981
).
A homeotic mutation transforming leg to antenna in Drosophila
.
Nature
292
,
635
638
.
Struhl
,
G.
(
1989
).
Differing strategies for organizing anterior and posterior body patterns in Drosophila embryos
.
Nature
338
,
741
744
.
Tautz
,
D.
and
Pfeifle
,
C.
(
1989
).
A non-radioactive in situ hybridization method for the localization of specific RNAs in Drosophila embryos reveals translational control of the segmentation gene hunchback
.
Chromosoma
98
,
81
85
.
Wakimoto
,
B. T.
and
Kaufman
,
T. C.
(
1981
).
Analysis of larval segmentation in lethal genotypes associated with the Antennapedia gene complex in Drosophila melanogasier
.
Devi Biol
.
81
,
51
64
.
Wakimoto
,
B. T.
,
Turner
,
F. R.
and
Kaufman
,
T. C.
(
1984
).
Defects in embryogenesis associated with the Antennapedia gene complex of Drosophila melanogasier
.
Devi Biol
.
102
,
147
172
.
Weir
,
M. P.
and
Kornberg
,
T.
(
1985
).
Patterns of engrailed and fushi tarazu transcripts reveal novel intermediate stages in Drosophila segmentation
.
Nature
318
,
433
439
.
Zalokar
,
M.
(
1976
).
Autoradiographic study of protein and RNA formation during early developoment of Drosophila eggs
.
Devi Biol
49
,
425
437
.