The fushi tarazu pair-rule gene is required for the formation of alternating parasegmental boundaries in the Drosophila embryo. fushi tarazu encodes a homeodomain protein necessary for transcription of the engrailed gene in even-numbered parasegments. Here we report that, within an engrailed enhancer, adjacent and conserved binding sites for the Fushi tarazu protein and a cofactor are each necessary, and together sufficient, for transcriptional activation. Footprinting shows that the cofactor site can be bound specifically by Ftz-F1, a member of the nuclear receptor superfamily. Ftz-F1 and the Fushi tarazu homeodomain bind the sites with 4-to 8-fold cooperativity, suggesting that direct contact between the two proteins may contribute to target recognition. Even parasegmental reporter expression is dependent on Fushi tarazu and maternal Ftz-F1, suggesting that these two proteins are indeed the factors that act upon the two sites in embryos. The two adjacent binding sites are also required for continued activity of the engrailed enhancer after Fushi tarazu protein is no longer detectable, including the period when engrailed, and the enhancer, become dependent upon wingless. We also report the existence of a separate negative regulatory element that apparently responds to odd-skipped.

The Drosophila segment polarity gene engrailed (en) plays a critical role in the generation and maintenance of parasegments, the fundamental units of the insect body plan. Expression of en and wingless (wg) in adjacent stripes establishes and maintains the parasegmental boundary (Martinez Arias et al., 1988; DiNardo et al., 1988; Heemskerk et al., 1991; Vincent and O’Farrell, 1992) and organizes patterning across each segment (Bejsovec and Martinez-Arias, 1991; Dougan and DiNardo, 1992; Heemskerk and DiNardo, 1994). Initiation, refinement and maintenance of the en expression pattern is controlled separately. Initiation is in response to patterns of pair-rule gene expression (Howard and Ingham, 1986; DiNardo and O’Farrell, 1987; Lawrence et al., 1987; Ingham et al., 1988; Manoukian and Krause, 1993; Fujioka et al., 1995). As pair-rule expression subsides, the parasegmental boundary is refined by interactions between wg- and en-expressing cells (Martinez Arias et al., 1988; DiNardo et al., 1988; Heemskerk et al., 1991; Vincent and O’Farrell, 1992; Fujioka et al., 1995). Later still, maintenance of the pattern becomes dependent upon repression by Polycomb group genes (Moazed and O’Farrell, 1992).

Models for pair-rule regulation of en involve overlapping patterns of putative activators and repressors (DiNardo and O’Farrell, 1987; Ingham et al., 1988; Manoukian and Krause, 1993; DiNardo et al., 1994; Mullen and DiNardo, 1995; Fujioka et al., 1995). In odd parasegments, graded expression of even-skipped (eve) establishes the en stripes by setting the boundaries of the activator paired (prd) and the repressors runt and sloppy paired (Fujioka et al., 1996).Expression of en in even parasegments results from activation by fushi tarazu (ftz). Only the most anterior cells of each ftz stripe express en and this restriction is dependent upon odd-skipped (odd) and naked (nkd) (DiNardo and O’Farrell, 1987; Benedyk et al., 1994; Mullen and DiNardo, 1995). Repression of en by odd may be partially indirect since odd reduces the level of ftz expression in the posterior two thirds of each ftz stripe (Mullen and DiNardo, 1995).

Surprisingly, the Ftz homeodomain is not required for activation of en, for autoactivation of ftz, or for repression of wg (Fitzpatrick et al., 1992; Copeland et al., 1996; Hyduk and Percival-Smith, 1996). The Prd protein has been implicated in the homeodomain-independent repression of wg by Ftz: with or without its homeodomain, Ftz is unable to repress residual wg expression in prd mutant embryos and the N-terminal half of the Ftz protein (i.e., lacking the homeodomain) binds Prd in vitro (Copeland et al., 1996). However, Prd is not required for ftz-dependent activation of en in even-parasegmental stripes since these stripes are still present in prd mutants (DiNardo and O’Farrell, 1987).

Although autoactivation of ftz can occur independently of the homedomain (Fitzpatrick et al., 1992), analysis of the autoregulatory ftz enhancer element has shown that Ftz can also activate by binding directly to the autoregulatory enhancer by means of its homeodomain (Schier and Gehring, 1992). This suggests that homeodomain-dependent and -independent mechanisms of transcriptional activation are at least partially redundant. Consistent with this model, a Ftz-dependent enhancer contained within the first en intron contains multiple homeodomain binding sites (Kassis et al., 1989; Kassis, 1990). However, homeodomain-DNA contact is clearly insufficient for normal target recognition since tandem arrays of consensus binding sites fail to mediate a Ftz-dependent pattern of reporter expression in embryos (Vincent et al., 1990; Nelson and Laughon, 1993; Zeng et al., 1994). It is therefore likely that Ftz requires a cofactor for target recognition, analogous to the way in which Extradenticle (Exd)/Pbx proteins act as cofactors for Hox proteins (Peifer and Wieschaus, 1990; Rauskolb and Weischaus, 1994; van Dijk and Murre, 1994; Chan et al., 1994; Neuteboom et al., 1995; Chang et al., 1995, 1996).

If Ftz utilizes its homeodomain to bind en regulatory sequences, homeodomain binding sites should be required for activation. Sequences required in addition to homeodomain binding sites would be candidates for cofactor-binding sites and could be tested for their ability to facilitate target recognition by Ftz. If Ftz functions primarily through protein-protein interactions, the DNA sequences required for Ftz activation would be expected to contain binding sites for the transcription factor(s) with which Ftz interacts, and need not contain Ftz-binding sites.

The approximately 1 kb Ftz-responsive en enhancer provided a reasonable starting point for identifying sequences that are necessary for target recognition and activation by Ftz. We found that binding sites for Ftz and for Ftz-F1, contained within a 37 bp block of conserved sequence, were essential for activity of the en enhancer. When multimerized, this element directed Ftz-dependent reporter transcription in embryos. The native enhancer and the multimerized conserved block are dependent upon Ftz and maternal Ftz-F1 for the ability to activate transcription of pair-rule stripes. Our results support the homeodomain-DNA contact model and identify Ftz-F1 as a putative Ftz cofactor. The results also provide evidence that negative regulation of en in even parasegments is at least partially dependent upon direct interaction of a repressor with sequences located within the intron.

Deletion analysis

Fragments of the en intron were generated by cleavage at unique restriction sites or by exonuclease III digestion (Erase-a-base, Promega Corp.). These were ligated directionally into the EcoRI and BamHI sites of CaSpeR such that the intron sequences were located 140 bp from the mini-white start of transcription. The mini-white gene was used as a reporter to detect en enhancer activity, as previously described (Kassis, 1990). This avoids the problem of targeted transposon insertion associated with en promoter (Kassis et al., 1992). In each case, the patterns shown or described are representative of three or more independent transformant lines. white transcripts were detected in embryos by in situ hybridization using a digoxigeninlabeled antisense white riboprobe (Tautz and Pfeifle, 1989). Hybridizations and washes were done at 57°C. The 4 bp substitutions introduced into hd1 and nhr1 (shown in Fig. 1) were generated by PCR using primers containing the basepair substitutions and confirmed by DNA sequencing. The base pair substitutions created HindIII and EcoRV sites within hd1 and nhr1, respectively, which were used to reconstitute the intact B-I fragment. 3× C was generated by ligation of a double-stranded synthetic oligonucleotide corresponding to the block C sequence shown at the top of panel A. The oligonucleotides had additional half-sites for BamHI and BglII at opposite ends, such that digestion with these enzymes after blunt end ligation yielded only tandemly arranged multimers.

Fig. 1.

Ftz- and Ftz-F1-binding sites within the en intron. A map at the top shows open boxes lettered A-P indicating the location of conserved sequence blocks from the sequence of Kassis et al. (1986), and arrows indicating the location of Ftz homeodomain binding sites identified by DNase I footprinting (see Fig. 5 for methods). Solid arrows indicate binding sites that most closely match the Antp class consensus (Laughon, 1991; Ekker et al., 1994), open arrows indicate sites that match the Cad-binding site consensus (Dearolf et al., 1989). Ovals mark sequences similar to the Ftz-F1-binding site consensus (Ueda and Hirose, 1991). Sequences of these sites are listed below the map grouped as hd, cad or nhr (nuclear hormone receptor) sites in the order that they appear from left to right on the map. Binding site consensus sequences for the Ftz homeodomain, for Cad and for Ftz-F1 are also shown and mismatches with these are shaded.

Fig. 1.

Ftz- and Ftz-F1-binding sites within the en intron. A map at the top shows open boxes lettered A-P indicating the location of conserved sequence blocks from the sequence of Kassis et al. (1986), and arrows indicating the location of Ftz homeodomain binding sites identified by DNase I footprinting (see Fig. 5 for methods). Solid arrows indicate binding sites that most closely match the Antp class consensus (Laughon, 1991; Ekker et al., 1994), open arrows indicate sites that match the Cad-binding site consensus (Dearolf et al., 1989). Ovals mark sequences similar to the Ftz-F1-binding site consensus (Ueda and Hirose, 1991). Sequences of these sites are listed below the map grouped as hd, cad or nhr (nuclear hormone receptor) sites in the order that they appear from left to right on the map. Binding site consensus sequences for the Ftz homeodomain, for Cad and for Ftz-F1 are also shown and mismatches with these are shaded.

Reporter constructs were examined in Df(1)w67c23 backgrounds that were otherwise wild type or contained the following mutations: ftz11, Df(3L)Cat; Ftz-F1209, odd5, and wgl-17 (Lindsley and Zimm, 1992; Guichet et al., 1997).

DNase I footprinting

The Ftz homeodomain (FtzHD2) was expressed in E. coli and purified as previously described (Florence et al., 1991). A partial α -Ftz-F1 cDNA (lacking the amino terminal 191 codons) was provided by C. Wu and used to over-express Ftz-F1 under T7 control in E. coli. FtzF1 containing extracts for footprinting experiments were prepared as follows (all steps ice cold): Cells from a liter culture were lysed in 3 ml of 6.4 M guanidine-HCl, 200 mM Tris/50 mM glycine (pH 9.5), 20 mM ZnSO4, 40 mM DTT. An equal volume of glycerol was added and the lysate was dialyzed for 8 hours against 100 volumes of 50 mM Tris/12.5 mM glycine (pH 9.5), 1 M guanidine HCl, 1 M NaCl, 10 mM ZnSO4, 1 mM DTT and 50% glycerol. A second 8-hour dialysis was done against 25 mM Tris/6 mM glycine (pH 9.5), 200 mM NaCl, 10 mM ZnSO4, 1 mM DTT, 0.5 mM PMSF and 50% glycerol. A third 8-hour dialysis was done against 25 mM Tris (pH 7.5), 100 mM NaCl, 10 mM ZnSO4, 0.5 mM EDTA and 50% glycerol, followed by centrifugation in a microfuge and precipitation of the supernatant with two volumes of 80% saturated ammonium sulfate. After centrifugation, the resulting precipitate was resuspended in the pH 7.5 dialysis buffer, minus glycerol and briefly centrifuged. The supernatant was dialyzed against the pH 7.5 buffer (with glycerol) and stored at −20°C. DHR39 was expressed in E. coli as a fusion to the carboxy-terminus of maltose-binding protein, as described by Ohno and Petkovich (1992). The resulting fusion protein contained amino acids 294-808 of DHR39 was prepared according to the same protocol described above for unfused Ftz-F1.

For footprinting, the StyI 5′ end was labeled with γ -[32P] ATP using polynucleotide kinase, cut with EcoRI and gel purified. Labeled DNA (<3.5× 10− 10M) was incubated with the FtzHD2 protein and/or Ftz-F1 in 25 mM Tris (pH 7.5), 50 mM NaCl, 10 mM ZnSO4, 1 mM MgCl2, 0.1 mM DTT and 10% glycerol. After a 4 hour incubation on ice, one-tenth volume of 200 μg/ml sonicated herring sperm DNA was added immediately prior to addition of an equal volume of a 1:3 dilution of DNaseI (1U/ml, Promega Corp.). After 30 seconds, the reaction was stopped with one-tenth volume 10% SDS, 50 mM EDTA, 1 mg/ml sonicated herring sperm DNA and immediately extracted with an equal volume of phenol. Samples were then precipitated, washed with 70% ethanol, resuspended in 5 μl standard formamide/dye mix and fractionated on a denaturing 6% polyacrylamide/4 M urea gel.

Elements required for Ftz activation of the en intron enhancer

Enhancers that mediate Ftz-dependent even-parasegmental expression of en have been mapped upstream of the promoter (DiNardo et al., 1988) and in the first intron (Kassis, 1990). The intron sequence is about 30% conserved between D. melanogaster and D. virilis (Kassis et al., 1986), most of which falls into 16 discrete blocks (Fig. 1). By footprinting, we identified ten Ftz homeodomain binding sites within the enhancer (Fig. 1), nine of which lie within conserved blocks. Seven of these resemble the CC/AATTA consensus for Ftz homeodomain binding sites (Florence et al., 1991) and Antp class homeodomain binding sites (Laughon, 1991; Ekker et al., 1994). These sites are referred to as ‘hd’ sites. The other three match the CATAAA consensus for the Caudal (Cad) homeodomain (Dearolf et al., 1989) and are referred to as ‘cad’ sites. The set of footprints identified with the Ftz homeodomain overlaps the footprints identified using the En protein (Kassis et al., 1989).

To identify which, if any, of these Ftz-binding sites were required for Ftz-dependent activation, portions of the intron were tested for the ability to drive expression of a reporter gene mRNA in embryos, as detected by in situ hybridization (Figs 2, 3). Enhancer activity was not detectably altered by a deletion (construct A-I) that removed four Ftz-binding sites from the 3′ end of the intron. Deletions removing blocks A, B and the 5′ 14 bp of block C had no effect on the spatial pattern of reporter expression. For these constructs, and those described below, double labeling to detect endogenous en transcripts was used to determine the registration of reporter stripes. Unless noted otherwise, reporter patterns were coincident with endogenous en.

Fig. 2.

en intron DNA fragments tested for embryonic enhancer activity. Conserved sequence blocks, Ftz-F1 sites and Ftz-binding sites are indicated as in Fig. 1. The sequence of block C triplicated in the 3× C construct is shown expanded at the top. The cad1, nhr1 and hd1 sequences are B-I indicated with open arrows, and the 4 bp substitutions introduced into nhr1 and hd1 are indicated with solid arrows. Shaded bases are not conserved in D. virilis. The patterns of w+ reporter expression observed during the Ftz-dependent phase (stage 5-8/9) and during the later Wg-dependent phase (stage 8/9-10) are indicated alongside each intron fragment. Expression patterns are noted as follows: en, en-like; ftz, ftz-like; mesoderm only; dorsal, dorsal half of the embryo.

Fig. 2.

en intron DNA fragments tested for embryonic enhancer activity. Conserved sequence blocks, Ftz-F1 sites and Ftz-binding sites are indicated as in Fig. 1. The sequence of block C triplicated in the 3× C construct is shown expanded at the top. The cad1, nhr1 and hd1 sequences are B-I indicated with open arrows, and the 4 bp substitutions introduced into nhr1 and hd1 are indicated with solid arrows. Shaded bases are not conserved in D. virilis. The patterns of w+ reporter expression observed during the Ftz-dependent phase (stage 5-8/9) and during the later Wg-dependent phase (stage 8/9-10) are indicated alongside each intron fragment. Expression patterns are noted as follows: en, en-like; ftz, ftz-like; mesoderm only; dorsal, dorsal half of the embryo.

Fig. 3.

w+ reporter expression during the Ftz-dependent phase (stage 5-8/9). Shown are the stage 7-8 patterns of w+ transcripts in embryos containing the corresponding pattern of endogenous en transcripts. A-I resembles the normal en pattern in even parasegments; B-I expression is weaker and delayed in comparison with A-I; 3′C-I is similar to B-I; D-I and B-I constructs containing 4 bp substitutions in the hd1 or nhr1 sites fail to express the reporter, even as late as stage 9 (stage embryos 9 shown for hd1 and nhr1); B-F shows a widened three-cell-wide pattern of stripes (stage 7/8 in upper panel, stage 9 in lower panel) which is similar to even parasegmental en expression in an odd mutant (Benedyk et al., 1994). B-E fails to express the reporter; 3× C shows six 3-to 4-cell-wide stripes (corresponding to Ftz stripes 2-7) and weak parasegment 2 expression (Ftz stripe 1). Even parasegmental expression of A-I, B-I and 3× C is abolished in a ftz mutant background (data not shown).

Fig. 3.

w+ reporter expression during the Ftz-dependent phase (stage 5-8/9). Shown are the stage 7-8 patterns of w+ transcripts in embryos containing the corresponding pattern of endogenous en transcripts. A-I resembles the normal en pattern in even parasegments; B-I expression is weaker and delayed in comparison with A-I; 3′C-I is similar to B-I; D-I and B-I constructs containing 4 bp substitutions in the hd1 or nhr1 sites fail to express the reporter, even as late as stage 9 (stage embryos 9 shown for hd1 and nhr1); B-F shows a widened three-cell-wide pattern of stripes (stage 7/8 in upper panel, stage 9 in lower panel) which is similar to even parasegmental en expression in an odd mutant (Benedyk et al., 1994). B-E fails to express the reporter; 3× C shows six 3-to 4-cell-wide stripes (corresponding to Ftz stripes 2-7) and weak parasegment 2 expression (Ftz stripe 1). Even parasegmental expression of A-I, B-I and 3× C is abolished in a ftz mutant background (data not shown).

Removal of block A (construct B-I) weakened and delayed the appearance of the early stripe pattern by 15-20 minutes (i.e., stripes appeared at the end of stage 7). Additional truncation of the en intron, removing block B (construct C-I) or the 5′ 14 bp of block C, had no additional effect on the level or timing of early reporter expression. Removal of the remainder of block C (construct D-I) resulted in the loss of all Ftz-dependent activity. The deleted region contains one conserved binding site for Ftz (hd1) and a conserved sequence (nhr1) resembling the binding site consensus for Ftz-F1, a member of the nuclear receptor superfamily (Ueda et al., 1990; Lavorgna et al., 1991, 1993; Ueda and Hirose, 1991; Horner et al., 1995). Indeed, DNaseI footprinting determined that Ftz-F1 bound this site (as well second site nhr2, see below). Clustered mutations were introduced into each site in construct B-I (hd1 and nhr1) and tested for effects on reporter expression. In each case, reporter expression was reduced to background level. A cross of B-I into a ftz mutant background showed that expression was fully dependent upon Ftz (data not shown). Crosses of A-I males to females homozygous for a maternal effect Ftz-F1 mutation that eliminates expression of the α -Ftz-F1 isoform (Guichet et al., 1997) showed that reporter expression was also fully dependent upon maternally derived α -Ftz-F1 (Fig. 4). These results suggest that Ftz and the α -Ftz-F1 act through their respective binding sites in block C to activate en transcription in even parasegments.

Fig. 4.

w+ reporter expression of A-I and 3× C is abolished in Ftz-F1 mutant embryos. A-I expression in wild-type (A,B) and in Ftz-F1 mutant embryos (C,D). A and C, stage 6; B and D, stage 9. In the embryos C and D, the reporter is not expressed. 3× C expression in wild-type (E,F) and in Ftz-F1 mutant embryos (G,H). E and G, stage 6; F and H, stage 9. In the embryos G and H, the reporter is not expressed. A,B,D-H, lateral view; C, dorsal view.

Fig. 4.

w+ reporter expression of A-I and 3× C is abolished in Ftz-F1 mutant embryos. A-I expression in wild-type (A,B) and in Ftz-F1 mutant embryos (C,D). A and C, stage 6; B and D, stage 9. In the embryos C and D, the reporter is not expressed. 3× C expression in wild-type (E,F) and in Ftz-F1 mutant embryos (G,H). E and G, stage 6; F and H, stage 9. In the embryos G and H, the reporter is not expressed. A,B,D-H, lateral view; C, dorsal view.

Removal of sequences from the 3′ end of the intron up to block G (construct B-F) altered the pattern by widening the stripes to 2-3 cells rather than the normal one-cell width (Fig. 3) and reduced the level of reporter expression, perhaps due to the loss of three Ftz-binding sites in block I. Removal of block F by a 17 bp truncation of B-F (construct B-E) completely eliminated reporter expression prior to stage 11 (Fig. 3). Block F lacks detectable Ftz-binding sites but contains a second FtzF1-binding site, nhr2 (Fig. 2). Thus there are two Ftz-F1binding sites essential for activation by Ftz, one adjacent to the essential Ftz-binding site, hd1, and one located approximately 200 bp distant.

A small element is sufficient for activation by Ftz and Ftz-F1

To determine whether the combination of Ftz- and Ftz-F1binding sites is sufficient for Ftz-dependent activation, a construct containing three copies of block C (3× C) was built and tested. During stages 5-7, the 3× C construct showed a pattern of seven Ftz-like, 3-to 4-cell-wide transverse stripes (Fig. 3, bottom panel). Double-labeling to simultaneously detect endogenous en transcripts showed indirectly that reporter expression coincided with that of ftz. During late stage 7 and stage 8, the first stripe (parasegment 2) faded. The subsequent 6-stripe pattern was similar to the pattern reported for the Ftz-dependent pbx enhancer (Muller and Bienz, 1992). Crossing of 3× C into a ftz mutant background showed that expression was fully dependent upon Ftz. Crosses of 3× C males to females homozygous for a maternal effect Ftz-F1 mutation showed that expression was also fully dependent upon maternally derived Ftz-F1 (Fig. 4). Thus the Ftz and FtzF1 sites contained in block C are necessary and, when multimerized, sufficient for Ftz activation. Since multimerization of homeodomain binding sites alone failed to direct stripe formation in embryos (Vincent et al., 1990; Nelson and Laughon, 1993; Zeng et al., 1994), it is clear that the Ftz-F1binding sites are crucial for Ftz activity.

Cooperative binding of the Ftz homeodomain and Ftz-F1

We tested whether binding of Ftz-F1 to nhr1 could increase the affinity of the Ftz homeodomain for hd1 in vitro. Addition of Ftz-F1 protein reduced the concentration of Ftz homeodomain necessary to footprint hd1 by 4-to 8-fold (compare hd1 footprints in lanes 4 and 10 of Fig. 5), indicating that bound FtzF1 stabilizes binding of the Ftz homeodomain to the adjacent site. Similarly, the binding of Ftz-F1 to nhr1 was enhanced 4to 8-fold by the addition of Ftz homeodomain. The proteins fail to bind specifically to the corresponding mutated sites (these are the same mutations, shown in Fig. 2, that eliminated reporter expression in embryos). Approximately the same level of cooperative binding was obtained in a footprinting experiment using Ftz-F1 and full-length Ftz (data not shown). These results suggest that Ftz-F1 may enhance the regulatory specificity of Ftz by directly contacting the Ftz homeodomain. A closely related Drosophila monomer nuclear receptor, DHR39 (also referred to as Ftz-F1β ; Ohno and Petkovich, 1992; Ayer et al., 1993; Ohno et al., 1994; Horner et al., 1995), also binds to nhr1, but without effect on Ftz binding to hd1 (data not shown). In addition, block F, which contains a second Ftz-F1 site, is also required for enhancer activity. This second Ftz-F1 site (nhr2) is 50 bp from the nearest Ftz-binding site (cad2) and binding of Ftz-F1 to this site does not alter the affinity of the Ftz homeodomain for hd1 (or cad2).

Fig. 5.

DNase I protection of block C by the Ftz homeodomain and Ftz-F1. Ftz homeodomain and N-terminally truncated Ftz-F1 (Lavorgna et al., 1991) purified from E. coli were used to protect a StyI-EcoRI fragment, end-labeled at the StyI site located 41 bp 5′ of block C. Footprints are labeled on the right-hand side of the sequencing gel. Sequential two-fold increases in Ftz homeodomain or Ftz-F1 concentration are indicated by the ramps at the top of the gel. Lanes 1-5 and 10-14 contain 4× 10− 9M to 6× 10− 8 M Ftz homeodomain. Lanes 6-9 contain 1:240-to 1:30-fold dilutions of Ftz-F1-containing extracts. No protein was added for lane 1 and G/A marks the G+A cleavage ladder. The hd1 footprint is apparent at a 4to 8-fold lower homeodomain concentration in the presence of FtzF1 (compare lanes 5 and 11). Also, the nhr1 footprint intensifies as the Ftz homeodomain concentration increases. This result has been repeated twice each with two preparations of Ftz-F1 protein. hd8 and nhr3 are overlapping low affinity Ftz and Ftz-F1-binding sites that lie outside of conserved block C.

Fig. 5.

DNase I protection of block C by the Ftz homeodomain and Ftz-F1. Ftz homeodomain and N-terminally truncated Ftz-F1 (Lavorgna et al., 1991) purified from E. coli were used to protect a StyI-EcoRI fragment, end-labeled at the StyI site located 41 bp 5′ of block C. Footprints are labeled on the right-hand side of the sequencing gel. Sequential two-fold increases in Ftz homeodomain or Ftz-F1 concentration are indicated by the ramps at the top of the gel. Lanes 1-5 and 10-14 contain 4× 10− 9M to 6× 10− 8 M Ftz homeodomain. Lanes 6-9 contain 1:240-to 1:30-fold dilutions of Ftz-F1-containing extracts. No protein was added for lane 1 and G/A marks the G+A cleavage ladder. The hd1 footprint is apparent at a 4to 8-fold lower homeodomain concentration in the presence of FtzF1 (compare lanes 5 and 11). Also, the nhr1 footprint intensifies as the Ftz homeodomain concentration increases. This result has been repeated twice each with two preparations of Ftz-F1 protein. hd8 and nhr3 are overlapping low affinity Ftz and Ftz-F1-binding sites that lie outside of conserved block C.

Elements affecting expression during the wgdependent phase

Reporter expression continues to be affected by blocks A, B, C, F and G-I as Ftz expression declines and disappears during stages 7-10 (Fig. 6). Both the hd1- and nhr1-binding sites affect this later expression. A-I is the smallest construct that produces a normal en pattern during these stages. A-I embryos continue to express even parasegment stripes at high levels into stage 11. Odd parasegment stripes appear later than the even stripes and the two become approximately equal in intensity by late stage 10 (∼5 hours AEL). In wg mutant embryos, A-I expression begins to disappear near the end of stage 8, paralleling the requirement of wg for continued expression of en as Ftz and other pair-rule proteins disappear from the embryo (Heemskerk et al., 1991). Thus, as shown previously for the entire intron (Kassis, 1990), A-I contains elements that depend on wg signaling for transcriptional activation. Although blocks J-P appear dispensable for both regulation by Ftz and by Wg, these conserved sequences could perform redundant functions or contribute to regulation during later stages of development.

Fig. 6.

w+reporter expression during the Wg-dependent phase. Shown are mid-to late-stage-9 patterns of w+transcripts in embryos containing the indicated reporter construct, except for the panel labeled ‘en’, which shows the corresponding pattern of endogenous en transcripts. A-I now shows expression in both even and odd parasegments; B-I expression is weaker than A-I; 3′C-I and B-I containing the 4 bp substitution in the nhr1 site express mesodermal stripes (about 10% of B-I(hd1) embryos show a faint B-I pattern at late stage 9); D-I, B-E and B-I containing the 4 bp substitution in the hd1 site fail to express the reporter; B-F expression fades ventrally at stage 9 and laterally shows an irregular pattern of expression between the strong even parasegmental stripes, and an increase in stripe intensity in the anterior-toposterior direction; 3× C is similar to B-F except that expression is limited to even parasegments.

Fig. 6.

w+reporter expression during the Wg-dependent phase. Shown are mid-to late-stage-9 patterns of w+transcripts in embryos containing the indicated reporter construct, except for the panel labeled ‘en’, which shows the corresponding pattern of endogenous en transcripts. A-I now shows expression in both even and odd parasegments; B-I expression is weaker than A-I; 3′C-I and B-I containing the 4 bp substitution in the nhr1 site express mesodermal stripes (about 10% of B-I(hd1) embryos show a faint B-I pattern at late stage 9); D-I, B-E and B-I containing the 4 bp substitution in the hd1 site fail to express the reporter; B-F expression fades ventrally at stage 9 and laterally shows an irregular pattern of expression between the strong even parasegmental stripes, and an increase in stripe intensity in the anterior-toposterior direction; 3× C is similar to B-F except that expression is limited to even parasegments.

The expression of construct B-I was almost identical to that of A-I, but weaker in all reporter lines examined, suggesting that, as for early, Ftz-dependent expression, block A strengthens, but is not absolutely required for, later enhancer activity. By the beginning of stage 9, both construct C-I and 3′C-I showed little odd parasegment staining in the ectoderm except at the very dorsal edge of the embryo, and even parasegment expression had begun to fade (3′C-I in Fig. 6). Reporter RNA visible at this time was restricted mainly to the mesoderm. By the end of stage 9, a low level of ectodermal expression in the odd parasegments was evident in some embryos. Although even parasegmental expression had faded considerably, it was still stronger than the weak odd parasegmental expression. Taken together, these results suggest that the region containing block B responds to a factor that is required for, or strongly enhances, ectodermal expression of en during the Wgdependent phase.

Constructs D-I and B-I[hd1] showed no expression during stages 9-10. Construct B-I[nhr1] showed no ectodermal expression during this interval. These results show that the hd1 and nhr1 sequences are required for continued enhancer activity in the ectoderm during the Wg-dependent phase. Ftz dependent activation of en extends only until stage 9 indicating another factor may act on hd1 at later stages, or Ftz dependent early expression may be required in some other way for late expression, possibly to prevent the onset of repression.

Construct B-F displayed a distinctive pattern of stage 9-10 expression. During stage 9, as Ftz-dependent activation ended, expression in the ventral half of the embryo disappeared (Fig.6). Odd-parasegment stripes appeared at this time but similarly only in the dorsal half of the embryo. In addition, reporter expression began to appear between the odd and even stripes. B-F expression was strongest in PS8-15, weaker in PS6-7, weakest in PS4-5 and absent is PS1-3. The loss of expression ventrally is very similar to loss of en expression in extradenticle (exd) mutant embryos (Peifer and Wieschaus, 1990), suggesting the Exd homeodomain protein may interact with sequences contained within the G-I interval. Further truncation,removing nhr2 (construct B-E), resulted in loss of detectable enhancer activity, consistent with involvement of Ftz-F1 in continued expression during the Wgdependent phase. Interestingly, the 3× C construct shows continued expression until late stage 10, well after the time when Ftz disappears (Carroll and Scott, 1985; Krause et al., 1988). During stage 9, 3× C stripes become restricted to the dorsal half of the embryo.

Cofactors have been shown to play an important role in target recognition by several homeodomain proteins. Based on deletion analysis and footprinting of the ftz-UAS enhancer, it has been suggested that Ftz-F1 might function as a cofactor for Ftz (Schier and Gehring, 1993; Han et al., 1993). Matches to the FtzF1 consensus also have been noted in other Ftzdependent enhancers (Schier and Gehring, 1993). Recently, a P element insertion that disrupts α -Ftz-F1 function has been identified and shown to confer a recessive maternal effect pair-rule phenotype that resembles the ftz phenotype (Guichet et al.,1997). The resulting embryos fail to express en in even parasegments, while the blastoderm pattern of ftz mRNA and protein is normal.

Our results show that a Ftz-F1-binding site is necessary and sufficient in combination with a Ftzbinding site for Ftz-dependent enhancer activity. We also found that Ftz and maternally derived α -Ftz-F1 are necessary for transcriptional activation by the enhancer and by a multimerized segment containing the Ftz- and Ftz-F1-binding sites. Cooperative binding of the Ftz homeodomain and Ftz-F1 to conserved binding sites in vitro indicate that Ftz-F1 may increase the target specificity of Ftz by means of a direct interaction between the two proteins. Taken together, these results suggest that α-Ftz-F1 functions as a Ftz cofactor. Interestingly, early embryonic expression of en is normal in embryos homozygous for Df (3L)Cat, which removes Ftz-F1 (A. L., unpublished results). This confirms that only the maternally expressed α-Ftz-F1 isoform, and not the zygotically expressed β-Ftz-F1, is required for en activation (Guichet et al., 1997). While Ftz-F1 has been implicated in both positive and negative regulation of the ftz gene via dual regulatory elements (Dearolf et al., 1989; Topol et al., 1991; Tsai and Gergen, 1995), our results point only to a positive role in the transcription of en. Although our results indicate that Ftz and Ftz-F1 act directly upon en, we cannot strictly rule out the possibility that other regulatory proteins might act as intermediates.

Cooperative binding of Ftz-F1 and the Ftz homeodomain to block C in vitro suggests that cofactor function is a result of direct contact between Ftz-F1 and the Ftz homeodomain, and that homeodomain-DNA interactions contribute to regulation. Guichet et al. (1997) report that α-Ftz-F1 binds to immobilized Ftz regardless of whether the homeodomain is present. Our failure to detect a difference in the interaction of Ftz-F1 with the Ftz homeodomain versus full-length Ftz may have been due to the use in our experiments of a truncated α -Ftz-F1 protein that lacked 191 N-terminal amino acids [i.e., approximately half of the α -Ftz-F1 isoform-specific region that is generated by alternative splicing (Lavorgna et al., 1993)]. While cooperative binding of N-terminally truncated Ftz-F1 and the Ftz homeodomain may be limited to adjacent sites, the two full-length proteins may be capable of interacting while bound to distant sites, providing an explanation for the essential Ftz-F1binding site (nhr2) that did not mediate cooperative binding in our experiments.

The results of Guichet et al. (1997) are consistent with experiments demonstrating that the endogenous en and wg genes respond to Ftz protein that lacks a homeodomain (Fitzpatrick et al., 1992; Copeland et al., 1996; Hyduk and Percival-Smith, 1996). In light of these facts, our results present a paradox: the Ftz homeodomain is dispensible for Ftz activation of endogenous genes, but we have shown that a Ftz homeodomain binding site is required for activity of the en intron enhancer. A possible explanation lies in the suggestion (Hyduk and Percival-Smith, 1996) that homeodomain deficient Ftz would still be able to contact DNA at a homeodomain binding site indirectly through a homeodomain-containing cofactor. A candidate is the Prd homeodomain protein, which interacts directly with the N terminus of Ftz (Copeland et al., 1996). Prd is also required for FtzΔ homeodomainmediated repression of wg (Copeland et al., 1996). We propose an expansion of the Hyduk and Percival-Smith model for ‘Ftz homeodomain-independent’ regulation by speculating that, without its homeodomain, Ftz is still able to bind Ftz-F1 and Prd. These interactions allow cooperative binding of Prd (via its homeodomain) and Ftz-F1 to sites normally occupied by Ftz and Ftz-F1, thus maintaining target specificity. This model does not exclude a role for the Ftz homeodomain; by providing an additional means of contacting DNA, the Ftz homeodomain may contribute to the stability of enhancer-bound multiprotein complexes without being absolutely required for activation of transcription. The ability of the Ftz homeodomain to function in the regulation of target genes is supported by experiments in which its substitution for the En homeodomain generated a chimeric En-FtzHD-En protein capable of repressing genes that are normally activated by Ftz (John et al., 1995).

The observed elimination of later stage en intron reporter expression by mutation of either a Ftz- or a Ftz-F1-binding site might be due to a need for initiation of a transcriptionally active state by Ftz in order for transcription to continue after Ftz has disappeared. If this is the case, any cis regulatory element that affects Ftz-dependent activation would alter later expression as well, in agreement with our results. Alternatively, or in addition, a second homeodomain protein (or proteins) might be required for continued enhancer activity during the Wg-dependent phase. In this respect, it is interesting to note that the minimal Ftz-responsive 3× C reporter construct continues to be expressed in dorsal even-parasegmental stripes after Ftz protein is no longer detectable in the ectoderm.

Regulation by Ftz has been the main focus of this report, but several additional elements uncovered in our analysis deserve brief discussion. Block A affects expression level but not pattern, and we note that it contains an 8-out-of-9 bp match to the consensus binding site of GLI, a human protein that is closely related to Opa (Benedyk et al., 1994). Opa has been shown to up-regulate en expression through repression of odd and the activation wg (Benedyk et al., 1994). The presence of a putative Opa-binding site in block A suggests that Opa may also exert a direct effect on en. The segment polarity gene cubitus interruptus (ci) (Orenic et al., 1990) encodes a second GLI-related protein, but ci is not required for the normal embryonic pattern of en (Hidalgo, 1991).

The en intron enhancer responds to Ftz in only the anterior-most cell column of each Ftz stripe. Deletion of the G-I interval widens the pattern of reporter expression to all Ftz-expressing cells, paralleling the patterns of en or of full-length en intron-reporter constructs in odd mutant embryos (DiNardo and O’Farrell, 1987; Kassis, 1990; Benedyk et al., 1994; data not shown). These results suggest that Odd directly represses en by binding to one or more sites within the 300 bp G-I interval. The separation of repressing from activating elements is consistent with repression by ‘quenching’, which may be a more general mechanism than repression by direct competition for binding to DNA (Johnson, 1995; Gray and Levine, 1996). Recently, nkd has been shown to contribute to repression of en in even parasegments (Mullen and DiNardo, 1995), leaving open the possibility that the G-I interval could respond to nkd as well.

Because deletion of block C or block F eliminated detectable reporter expression, our analysis does not reveal whether blocks D and E contribute to enhancer activity. However, we note that block D contains a striking match (11 out of 11 bp for D. virilis, 8 out of 11 bp for D. melanogaster) to the binding site consensus for the mammalian transcription factors LEF-1 and TCF-1. LEF-1 and XTcf-3, members of the LEF-1/TCF-1 subgroup of HMG domain proteins, interact with β catenin and have been implicated as a nuclear effectors of Wnt signaling (Behrens et al., 1996; Molenaar et al., 1996), suggesting that a homologous Drosophila protein(s) regulates transcription in response to wg. It is striking that an enhancer as large and structurally complex as the B-I segment of the en intron [eight conserved blocks totaling about 200 bp (Kassis et al., 1986)] is critically dependent on single binding sites for two factors. In contrast, redundancy among Ftz-binding sites has been well documented for the ftz-UAS autoregulatory element (Schier et al., 1992; 1993; Han et al., 1993). Indeed, it is possible that the block C binding sites may be dispensable in the context of the complete intron, with its additional conserved blocks and Ftzbinding sites. Nonetheless, analysis of the B-I interval has demonstrated the important role played by a cofactor in facilitating Ftz-dependent transcriptional activation.

We thank Judy Kassis, Carl Wu and Martin Petkovich for generously providing plasmids and fly strains. For communicating unpublished data, we thank John Copeland, Henry Krause, Judy Kassis and Mary Whitely. We are also indebted to Sean Carroll, Henry Krause and Bill McGinnis for comments on the manuscript. This work was supported by the American Cancer Society and the Milwaukee Foundation (A. L.), the EMBL (A. E. and A. G.) and a fellowship from the Ministere de la Recherche et du Travail (A. G.). This is Laboratory of Genetics publication number 3408.

Ayer
,
S.
,
Walker
,
N.
,
Mosammaparast
,
M.
,
Nelson
,
J. P.
,
Shilo
,
B.
and
Benyajati
,
C.
(
1993
).
Activation and repression of Drosophila alcohol dehydrogenase distal transcription by two steroid hormone receptor superfamily members binding to a common response element
.
Nucl. Acid Res.
21
,
1619
1627
.
Behrens
,
J.
,
von Kries
,
J. P.
,
Kuhl
,
M.
,
Bruhn
,
L.
,
Wedlich
,
D.
,
Grosschedle
,
R.
and
Birchmeier
,
W.
(
1996
).
Functional interaction of β catenin with the transcription factor LEF-1
.
Nature
382
,
638
642
.
Benedyk
,
M. J.
,
Mullen
,
J. R.
and
DiNardo
,
S.
(
1994
).
odd-paired: a zinc finger pair-rule protein required for the timely activation of engrailed and wingless in Drosophila embryos
.
Genes Dev.
8
,
105
117
.
Bejsovec
,
A.
and
Martinez Arias,
A.
(
1991
).
Roles of wingless in patterning the larval epidermis of Drosophila
.
Development
113
,
471
485
.
Carroll
,
S. B.
and
Scott
,
M. P.
(
1985
).
Localization of fushi tarazu protein during embryogenesis
.
Cell
43
,
47
57
.
Chan
,
S.-K.
,
Jaffe
,
L.
,
Capovilla
,
M.
,
Botas
,
J.
and
Mann
,
R. S.
(
1994
).
The DNA binding specificity of Ultrabithorax is modulated by cooperative interactions with Extradenticle, another homeoprotein
.
Cell
78
,
603
615
.
Chang,,
C.-P.
,
Shen
,
W.-F.
,
Rozenfeld
,
S.
,
Lwarence
,
H. J.
,
Largman
,
C.
and
Cleary
,
M. L.
(
1995
)
Pbx proteins display hexapeptide-dependent cooperative DNA binding with a subset of Hox proteins
.
Genes Dev.
9
,
663
674
.
Chang
,
C. P.
,
Brocchieri
,
L.
,
Shen
,
W. F.
,
Largman
,
C.
and
Cleary
,
M. L.
(
1996
)
PBX modulation of Hox homeodomain amino-terminal arms establishes different DNA-binding specificities across the Hox locus
.
Mol. Cell. Biol.
16
,
1734
1745
.
Copeland
,
J. W. R.
,
Nasiiadka
,
A.
,
Dietrich
,
B. H.
and
Krause
,
H. M.
(
1996
).
Patterning of the Drosophila embryo by a homeodomain-deleted Ftz polypeptide
.
Nature
379
,
162
165
.
Coulter
,
D. E.
,
Swaykus
,
E. A.
,
Beran-Koehn
,
M. A.
,
Goldberg
,
D.
,
Wieschaus
,
E.
and
Schedl
,
P.
(
1990
).
Molecular analysis of odd-skipped, a zinc finger encoding segmentation gene with a novel pair-rule expression pattern
.
EMBO J
.
8
,
3795
3804
.
Dearolf
,
C. R.
,
Topol
,
J.
and
Parker
,
C. S.
(
1989
).
The caudal gene product is a direct activator of fushi tarazu transcription during Drosophila embryogenesis
.
Nature
341
,
340
343
.
DiNardo
,
S.
and
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 Dev
.
1
,
1212
25
.
DiNardo
,
S.
,
Sher
,
E.
,
Heemskerk-Jongens
,
J.
,
Kassis
,
J. A.
and
O’Farrell
,
P. H.
(
1988
).
Two-tiered regulation of spatially patterned engrailed gene expression during Drosophila embryogenesis
.
Nature
332
,
604
9
.
DiNardo
,
S.
,
Heemskerk
,
J.
,
Dugan
,
S.
and
O’Farrell
,
P. H.
(
1994
).
The making of a maggot: patterning the Drosophila embryonic epidermis
.
Curr. Opin. Genet. Dev
.
4
,
529
534
.
Dougan
,
S.
and
DiNardo
,
S.
(
1992
).
Drosophila wingless generates cell type diversity among engrailed expressing cells
.
Nature
360
,
347
349
.
Ekker
,
S. C.
,
Jackson
,
D. G.
,
von Kessler
,
D. P.
,
Sun
,
B. I.
,
Young
,
K. E.
and
Beachy
,
P. A.
(
1994
).
The degree of variation in DNA sequence recognition among four Drosophila homeotic proteins. EMBO J
.
13
,
3551
3560
.
Fitzpatrick
,
V. D.
,
Percival-Smith
,
A.
,
Ingles
,
C. J.
and
Krause
,
H.
(
1992
).
Homeodomain-independent activity of the fushi tarazu polypeptide in Drosophila embryos
.
Nature
356
,
610
612
.
Florence
,
B.
,
Handrow
,
R.
and
Laughon
,
A.
(
1991
).
DNA-binding specificity of the fushi tarazu homeodomain
.
Mol. Cell. Biol.
11
,
3613
23
.
Fujioka
,
M.
,
Jaynes
,
J. B.
and
Goto
,
T.
(
1995
).
Early even-skipped stripes at as morphogenetic gradients at the single cell level to establish engrailed expression
.
Development
121
,
4371
4382
.
Gray
,
S.
and
Levine
,
M.
(
1996
).
Short-range transcriptional repressors mediate both quenching and direct repression within complex loci in Drosophila
.
Genes Dev
.
10
,
700
710
.
Guichet
,
A.
,
Copeland
,
J. W. R.
,
Erdelyi
,
M.
,
Zavorszky
,
P.
,
Krause
,
H. M.
and
Ephrussi
,
A. M.
(
1997
)
Nuclear receptor homologue Ftz-F1 and homeodomain protein Ftz are mutally dependent cofactors
.
Nature, in press
.
Han
,
W.
,
Yu
,
Y.
,
Altan
,
N.
and
Pick
,
L.
(
1993
).
Multiple proteins interact with the fushi tarazu proximal enhancer
.
Mol. Cell. Biol
.
13
,
5549
5559
.
Heemskerk
,
J.
,
DiNardo
,
S.
,
Kostriken
,
R.
and
O’Farrell
,
P. H.
(
1991
).
Multiple modes of engrailed regulation in the progression towards cell fate determination
.
Nature
352
,
404
410
.
Heemskerk
,
J.
and
DiNardo
,
S.
(
1994
).
Drosophila hedgehog acts as a morphogen in cellular patterning
.
Cell
76
,
449
460
.
Hidalgo
,
A.
(
1991
)
Interactions between segment polarity genes and the generation of the segmental pattern
in
Drosophila. Mech. Dev.
35
,
77
87
.
Horner
,
M. A.
,
Chen
,
T.
and
Thummel
,
C. S.
(
1995
).
Ecdysteroid regulation and DNA binding properties of Drosophila nuclear hormone receptor superfamily members
.
Dev. Biol
.
168
,
490
502
.
Howard
,
K.
and
Ingham
,
P.
(
1986
).
Regulatory interactions between the segmentation genes fushi tarazu, hairy, and engrailed in the Drosophila blastoderm
.
Cell
44
,
949
957
.
Hyduk
,
D.
and
Percival-Smith
,
A.
(
1996
).
Genetic characterization of the homeodomain-independent activity of the Drosophila fushi tarazu gene product
.
Genetics
142
,
481
492
.
Ingham
,
P.
,
Baker
,
N. E.
and
Martinez Arias,
A.
(
1988
).
Regulation of segment polarity genes in the Drosophila blastoderm by fushi tarazu and even skipped
.
Nature
331
,
73
5
.
John
,
A.
,
Smith
,
S. T.
and
Jaynes
,
J. B.
(
1995
)
Inserting the Ftz homeodomain into engrailed creates a dominant transcriptional repressor that specifically turns off Ftz target gene in vivo
.
Development
121
,
1801
1813
.
Johnson
,
A.
(
1995
)
The price of repression
.
Cell
81
,
655
658
.
Kassis
,
J. A.
(
1990
).
Spatial and temporal control elements of the Drosophila engrailed gene. Genes Dev
.
4
,
433
43
.
Kassis
,
J. A.
,
Poole
,
S. J.
,
Wright,
D.K.
andO’Farrell
P. H.
(
1986
).
Sequence conservation in the protein coding and intron regions of the engrailed transcription unit
.
EMBO J
.
56
,
3583
3589
.
Kassis
,
J. A.
,
Desplan
,
C.
,
Wright
,
D. K.
and
O’Farrell
,
P. H.
(
1989
).
Evolutionary conservation of homeodomain-binding sites and other sequences upstream and within the major transcription unit of the Drosophila segmentation gene engrailed
.
Mol. Cell. Biol
.
9
,
4304
4311
.
Kassis
,
J. A.
,
Noll
,
E.
,
VanSickle
,
E. P.
,
Odenwald
,
W. F.
and
Perrimon
,
N.
(
1992
).
Altering the insertional specificity of a Drosophila transposable element
.
Proc. Nat. Acad. Sci. USA
89
,
191
1923
.
Krause
,
H. M.
,
Klemenz
,
R.
and
Gehring
,
W. J.
(
1988
).
Expression, modification, and localization of the fushi tarazu protein in Drosophila embryos
.
Genes Dev
.
2
,
1021
1036
.
Laughon
,
A.
(
1991
).
DNA binding specificity of homeodomains
.
Biochemistry
30
,
11357
11367
.
Lavorgna
,
G.
,
Ueda,
H.
Clos,
J.
and
Wu
,
C.
(
1991
).
Ftz-F1, a steroid hormone receptor-like protein implicated in the activation of fushi tarazu
.
Science
252
,
848
51
.
Lavorgna
,
G.
,
Karim
,
F. D.
,
Thummel
,
C. S.
and
Wu
,
C.
(
1993
).
Potential role for a FTZ-F1 steroid receptor superfamily member in the control of Drosophila metamorphosis
.
Proc. Nat. Acad. Sci. USA
90
,
3004
3008
.
Lawrence
,
P. A.
,
Johnson
,
P.
,
MacDonald
,
P.
and
Struhl
,
G.
(
1987
).
Borders of parasegments in Drosophila embryos are delimited by the fushi tarazu and even-skipped genes
.
Nature
328
,
440
442
.
Lindsley
,
D. L.
and
Zimm
,
G. G.
(
1992
)
The Genome of Drosophila melanogaster
.
Academic Press
,
San Diego
.
Manoukian
,
A. M.
and
Krause
,
H. M.
(
1993
).
Control of segmental asymmetry in Drosophila embryos
.
Development
118
,
785
796
.
Martinez Arias
,
A.
,
Baker
,
N. E.
and
Ingham
,
P.
(
1988
).
Role of segment polarity genes in the definition and maintenance of cell states in the Drosophila embryo
.
Development
103
,
157
170
.
Moazed
,
D.
and
O’Farrell
,
P. H.
(
1992
).
Maintenance of the engrailed expression pattern by Polycomb group genes in Drosophila
.
Development
116
,
805
810
Molenaar
,
M.
,
can de Wettering,
M.
,
Oosterwegel
,
M.
,
Peterson-Maduro
,
J.
,
Godsave
,
S.
,
Korinek
,
V.
,
Roose
,
J.
,
Destree
,
O.
and
Clevers
,
H.
(
1996
).
Xtcf-3 transcription factor mediates β-catenin-induced axis formation in Xenopus embryos
.
Cell
86
,
391
399
.
Mullen
,
J. R.
and
DiNardo
,
S.
(
1995
).
Establishing parasegments in Drosophila embryos: roles of the odd-skipped and naked genes
.
Dev. Biol.
169
,
295
308
.
Muller
,
J.
and
Bienz
,
M.
(
1992
).
Sharp anterior boundary of homeotic gene expression conferred by the fushi tarazu protein
.
EMBO J.
11
,
3653
61
.
Nelson
,
H. B.
and
Laughon
,
A.
(
1993
).
Drosophila glia architecture and development: analysis using a collection of new cell-specific markers. Roux’s Arch. Dev. Biol
.
202
,
341
354
.
Neuteboom
,
S. T. C.
,
Peltenburg
,
T. C.
,
van Dijk
,
M. A.
and
Murre
,
C.
(
1995
)
The hexapeptide LFPWMR in Hoxb-8 is required for cooperative DNA binding with Pbx1 and Pbx2 proteins
.
Proc. Nat. Acad. Sci. USA
92
,
9166
9170
.
Ohno
,
C. K.
and
Petkovich
,
M.
(
1992
).
Ftz-F1β, a novel member of the Drosophila nuclear receptor family
.
Mech. Dev
.
40
,
13
24
.
Ohno
,
C. K.
,
Ueda
,
H.
and
Petkovich
,
M.
(
1994
).
The Drosophila nuclear receptors Ftz-F1α and Ftz-F1β compete as monomers for binding to a site in the fushi tarazu gene
.
Mol. Cell. Biol.
14
,
3166
3175
.
Orenic
,
T.
,
Slusarski
,
D. C.
,
Kroll
,
K. L.
and
Holmgren
,
R. A.
(
1990
)
Cloning and characterization of the segment polarity gene cubitus interruptus Dominant of
Drosophila. Genes Dev.
4
,
1053
1067
.
Peifer
,
M.
and
Wieschaus
,
E.
(
1990
).
Mutations in the Drosophila gene extradenticle affect the way specific homeodomain proteins regulate segmental identity
.
Genes Dev
.
4
,
1209
1223
.
Rauskolb
,
C.
and
Wieschaus
,
E.
(
1994
).
Coordinate regulation of downstream genes by extradenticle and the homeotic selector proteins
.
EMBO J
.
13
,
3561
3569
.
Schier
,
A. F.
, and
Gehring
,
W. J.
(
1992
).
Direct homeodomain-DNA interaction in the autoregulation of the fushi tarazu gene. Nature
356
,
804807
.
Schier
,
A. F.
, and
Gehring
,
W. J.
(
1993
).
Analysis of a fushi tarazu autoregulatory element: multiple sequence elements contribute to enhancer activity
.
EMBO J
.
12
,
1111
9
.
Tautz
,
D.
and
Pfeifle
,
C.
(
1989
).
A non-radioactive in situ hybridization method for the localization of specific mRNAs in Drosophila embryos reveals translational control of the segmentation gene hunchback
.
Chromosoma
98
,
81
85
.
Topol
,
J.
,
Dearolf
,
C. R.
,
Prakash
,
K.
and
Parker
,
C. S.
(
1991
).
Synthetic oligonucleotides recreate Drosophila fushi tarazu zebra-stripe expression
.
Genes Dev
.
5
,
855
867
.
Tsi
,
C.
and
Gergen
,
P.
(
1995
).
Pair-rule expression of the Drosophila fushi tarazu gene: a nuclear receptor response element mediates the opposing regulatory effects of runt and hairy
.
Development
121
,
453
462
.
Ueda
,
H.
,
Sonoda
,
S.
,
Brown
,
J. L.
,
Scott
,
M. P.
and
Wu
,
C.
(
1990
).
A sequence-specific DNA-binding protein that activates fushi tarazu segmentation gene expression
.
Genes Dev.
4
,
624
635
.
Ueda
,
H.
and
Hirose
,
S.
(
1991
).
Defining the sequence recognized with Bm FTZ-F1, a sequence specific DNA binding factor in the silkworm, Bombyx mori, as revealed by direct sequencing of bound oligonucleotides and gel mobility shift competition analysis
.
Nucleic Acids Res.
19
,
3689
3693
.
van Dijk
,
M. A.
and
Murre
,
C.
(
1994
).
Extradenticle raises the DNA binding specificity of homeotic selector gene products
.
Cell
78
,
617
624
.
Vershon
,
A. K.
and
Johnson
,
A. D.
(
1993
).
A short, disordered protein region mediates interactions between the homeodomain of the yeast α 2 protein and the MCM1 protein
.
Cell
72
,
105
112
.
Vincent
,
J. P.
,
Kassis
,
J. A.
and
O’Farrell
,
P. H.
(
1990
).
A synthetic homeodomain binding site acts as a cell type specific, promoter specific enhancer in Drosophila embryos
.
EMBO J
.
9
,
2573
2578
.
Vincent
,
J. P.
and
O’Farrell
,
P. H.
(
1992
).
The state of engrailed expression is not clonally transmitted during early Drosophila development
.
Cell
68
,
923
931
.
Zeng
,
C.
,
Pinsonneault
,
J.
,
Gellon
,
G.
,
McGinnis
,
N.
and
McGinnis
,
W.
(
1994
).
Deformed protein binding sites and cofactor binding sites are required for the function of a small segment-specific regulatory element in Drosophila embryos
.
EMBO J
.
13
,
2362
2377
.