Abstract
Hedgehog (Hh) is a member of a family of secreted proteins that direct patterning at multiple stages in both Drosophila and vertebrate development. During Drosophila embryogenesis, Hh protein is secreted by the cells of the posterior compartment of each segment. hh activates transcription of wingless (wg), gooseberry (gsb), and patched (ptc) in the cells immediately adjacent to Hh-secreting cells. Hh signaling is thought to involve the segment polarity gene cubitus interruptus (ci). ci encodes a zinc finger protein of the Gli family of sequence-specific DNA binding proteins. ci mRNA is expressed in all non-Hh expressing cells. Here we demonstrate ci activity is both necessary and sufficient to drive expression of Hh-responsive genes in the Drosophila embryos. We show that Ci is a sequence-specific DNA binding protein that drives transcription from the wg promoter in transiently transfected cells. We demonstrate that Ci binding sites in the wg promoter are necessary for this transcriptional activation. These data taken together provide strong evidence that Ci is a transcriptional effector of Hh signaling.
In Drosophila embryogenesis Hedgehog (Hh) plays an early role in segmentation and a later role in generating cell type diversity within each segment (reviewed in ref. 1). The Drosophila body is divided into parasegments, which are the basic units for patterning in the developing organism. As the embryo cellularizes, the parasegment borders form. A parasegment border is delineated by expression of two secreted proteins (reviewed in ref. 1). Wingless (Wg), a Wnt family member, is expressed in cells anterior to the parasegment border (2), and Hh is expressed by cells posterior to the parasegment border (3–5). During the first 2 h after gastrulation, secreted Hh protein activates a signaling pathway that results in transcription of wingless (wg), gooseberry (gsb), and patched (ptc) in cells that are competent to respond (6). Between 3 and 6 h after gastrulation, Hh takes on a new role. It participates in autoregulatory engrailed (en) expression (7) and it organizes the pattern of dorsal cuticle hairs (8). In its latter role, Hh has been proposed to act as a morphogen (8), though this remains to be demonstrated (9).
hh also specifies critical aspects of adult morphology in Drosophila. The cuticular structures of adult appendages derive from internal epithelial sacs within the larvae, the imaginal discs. The posterior compartment of each disc expresses Hh, and immediately adjacent anterior compartment cells respond to this Hh signal by induction of the secondary signals Wg or the transforming growth factor β family member, Decapentaplegic (Dpp). These signals then organize growth and patterning of both compartments (10–14). While the developing eye lacks distinct anterior and posterior compartments, Hh signaling is critical for organizing the highly ordered structure of the compound eye (reviewed in refs. 5 and 15–18).
Several components of the Hh signaling pathway have been identified in Drosophila. ptc is a negative regulator of Hh signaling (19). Genetic analysis suggests that Hh signaling either inactivates Ptc or overrides the ability of Ptc to repress transcription of Hh-responsive target genes. Because Ptc is a multiple membrane spanning cell surface protein (20, 21), it could be a Hh receptor (19). However, it cannot be the only Hh receptor, since Hh has effects in ptc null embryos (22). Two protein kinases are implicated in Hh signal transduction, fused as a positive regulator, and cAMP-dependent protein kinase A as a negative regulator (reviewed in ref. 23). How either of these fit into a signal transduction pathway remains to be determined.
In cubitus interruptus (ci) mutant embryos, transcription of Hh-responsive genes is lost (24, 25). Ci overexpression or misexpression drives transcription of Hh responsive target genes (26, 27). Ci is a member of the Gli family of sequence-specific DNA binding proteins (28, 29), a family characterized by five tandem zinc fingers with highly similar sequence. Based on the presence of these zinc fingers, Orenic et al. (28) proposed that Ci acts as a transcription factor. Alexandre et al. (27) showed that Ci can activate transcription in yeast, through a putative Ci binding site. These data suggest that Ci is a transcriptional effector of Hh signaling. However, direct Ci–DNA interaction has not been shown. Here we present genetic evidence confirming that Hh acts through Ci to activate wg and gsb transcription. We demonstrate that Ci has sequence-specific DNA binding activity and that direct interaction between Ci and sequences in wg promoter is required for transcriptional activation from this promoter. We conclude that Ci functions as a transcriptional activator in the Hh signal transduction pathway.
MATERIALS AND METHODS
Drosophila Stocks, Crosses, and Analysis.
Df(4)M62f flies were obtained from Robert Holmgren (Northwestern University), yw; Hs-hh M3/TM3 y+ Ser from Philip Ingham (Imperial Cancer Research Fund), and hhGS1, ru e/TM3 from Jym Mohler (Barnard College).
Hs-hh M3/+; Df(4)M62f/Df(4)M62f embryos were generated by mating Hs-hh M3/+; Df(4)M62f/+ males to Df(4)M62f/eyD females. Embryos from this cross were collected and heat-shock treated as in ref. 30.
hhGS1/hhGS1;ciD/ciD embryos were obtained from heterozygous hhGS1/TM3;ciD/+ parents. The stage 10/early stage 11 embryos were fixed, processed by in situ hybridization, and scored for wild type; ciD, hh, or hh; ciD patterns; and unscorable. Using the χ2 test, the observed distribution of phenotypes fit the expected 9:3:3:1 ratio with P = 0.8 for ptc, P = 0.44 for wg, and P > 0.9 for gsb.
The smo ciD phenotype was determined by collecting embryos from smo1 cn bw sp/Cyo; ciD/M63a parents overnight at 18°C. The embryos were fixed and assayed for wg, gsb, and ptc RNA patterns. The stage 10/early stage 11 embryos were scored for wild type; ciD, smo, or smo; ciD patterns; and unscorable. Using the χ2 test, the observed distribution of phenotypes fit the expected 9:3:3:1 ratio for wg, gsb, and ptc.
The fused (fu); ciD phenotype was determined by collecting embryos from fu513/fu513; ciD/M63a females crossed to fu513; ciD/M63a males. Stage 10 and early stage 11 embryos were scored for fu or fu; ciD phenotypes. Approximately 25% of the populations had a ciD like expression pattern for wg, gsb, and ptc.
In situ hybridizations utilized antisense-strand riboprobes, according to standard procedures (31, 32).
Recombinant DNAs.
To build wg-Luc, a 5-kb EcoRV fragment of wg promoter ligated into the SmaI site of pXP2Luc (33). To construct Δwg-Luc and Δwg*-Luc, a 1-kb XhoI fragment from pAlter (Promega) containing a 2-kb BamHI fragment of wg promoter was blunt-ended with Klenow and ligated into tk-Luc (33) at the SmaI site. wg-LucΔS was constructed by deletion of a 1-kb SalI fragment from the full-length wg-Luc construct. wg-LucΔB was constructed by removal of a 2-kb BamHI fragment from wg-Luc. mt-Ci was constructed by ligating a BglII/NotI fragment containing full-length Ci cDNA from NB40 (34) into pRmHa-1 (35) at the BamHI and HincII sites. The glutathione-S-transferase/Ci fusion protein (GST-Ci) was constructed by ligation of a EcoRV/HincII fragment of Ci cDNA into the SmaI site of pGEX-3X (36).
Transient Transfections.
Drosophila Schneider line 2 cells were transfected using the calcium phosphate precipitation method (37). Each transfection contained 50 ng wg-Luc reporter plasmid, variable amounts of expression plasmid mt-Ci as indicated, and pRmHa-1 plasmid in variable amounts to bring the total amount of expression plasmid to 1 μg. A total of 10 μg of DNA was used in each transfection. As an internal control 1 μg of Copia long terminal repeat (LTR)-lacZ [copia–β-galactosidase (β-gal), obtained from J. Manley, Columbia University] was added to each transfection. β-Gal assays were conducted as described in ref. 38. Luciferase assays were done as in ref. 39.
Transfections with the Δwg-Luc, Δwg*-Luc, wg-LucΔS, and wg-LucΔB reporter constructs contained 200 ng of each reporter. mt-Ci concentration was varied as specified. All transfections were performed in duplicate, and each experiment was conducted several times with similar results. Data is presented as the average of the duplicates.
Purification of GST-Ci and DNA Binding Assays.
GST-Ci was grown in HB101 Escherichia coli to saturation overnight. A 250 ml culture was grown to OD650 = 0.1–0.25 and induced by addition of 1 mM isopropyl β-d-thiogaloctoside, then grown overnight at 20°C. Growth at lower temperature reduced the formation of inclusion bodies. The cells were then harvested and lysed as described (40), and the protein was purified on glutathione Sepharose (Pharmacia). BSA was added to 1 mg/ml as carrier. Purified protein was concentrated and dialyzed into PBS with a Centricon 30 microconcentrator (Amicon).
BS-1 and BS-2 plasmids were obtained from David Zarkower (University of Minnesota). The binding site containing the fragment was excised with XbaI and labeled with [α-32P]dCTP by Klenow end filling (41). The labeled fragments were then purified on a native acrylamide gel. Electrophoretic mobility-shift assays (EMSAs) were performed in 1× EMSA buffer, 50 mM Hepes (pH 7.5), 10 μM ZnSO4, 50 mM KCl, 5 mM MgCl2, 1 mM DTT, and 20% glycerol, and resolved on 5% acrylamide native 1× TBE gels. Anti-GST antiserum was provided by Alex Franzusoff (University of Colorado Health Science Center).
Fragments containing mutant and nonmutant wg promoter sequences (fragments 1 and 2) were isolated by PCR. PCR fragments were end labeled with [γ-32P]ATP. Quantitation used a scanning densitomiter, Alpha Innotech model IS1000 digital imaging system (Alpha Innotech, San Leandro, CA).
Site-Directed Mutagenesis.
Mutagenesis of putative Ci binding sites in wg promoter/enhancer DNA used the altered sites in vitro mutagenesis kit according to manufacturers instructions (Promega). A 2-kb BamHI fragment of wg promoter was inserted into pAlter-1 phagemid vector. Single-stranded DNA was prepared using helper phage M13K07 or R408. Mutagenic primers were synthesized by Life Technologies. Primers used were Alt-1, TCGATGGCTACTCGTCGG; Alt-2, CGCCAGCGTAGACGATGAT; Alt-3, AAAGTGCGTAAACCATGTC; and Alt-4, AAAATGGACTTTCCCAGCGT. Each mutates one of four putative Ci binding sites in the distal 1 kb of wg promoter. Successful oligonucleotide mutagenesis was confirmed by sequencing of candidate clones using Sequenase kit (United States Biochemical).
RESULTS
hh Acts Through ci.
In the Drosophila embryo, for 2 h after gastrulation, Hh signaling maintains the expression of wg, gsb, and ptc (6, 42, 43). To demonstrate ci is required for transduction of the Hh signal, we examined expression of wg in ci null mutant embryos when Hh is ubiquitously expressed under control of a heat-shock promoter (Hs-hh). In Hs-hh embryos, wg is expressed ectopically in anteriorly expanded stripes (Fig. 1B) (13, 30). In ci− embryos Hs-hh does not induce ectopic expression of wg. Instead wg expression decays in the same manner seen in ci mutants (Fig. 1C). This conclusion is based on statistical analysis of mixed populations. In this experiment 25% of the embryos are ci null. If ci is required to transduce the Hh signal, then Hs-hh will have no effect in homozygous ci− embryos. If ci is not required, then only the homozygous ci− embryos not receiving Hs-hh (12.5%) will have the ci− phenotype. As shown in Fig. 1D, 24.5% of the population had the ci− phenotype. Similar results were obtained for gsb expression. The distribution of the three progeny classes for gsb expression is as follows: Hs-hh, 113 embryos; wild type, 104 embryos and ci−, 78 embryos—i.e., 26.4% had the ci− phenotype. Therefore ci is required for transduction of the hh signal.
ciD is a gain of function allele of ci (28) that produces an abnormally large protein at very low levels (44). In ciD embryos wg expression is initiated normally, but at stage 10 it diverges from the wild-type pattern to form stripes that are more broad and less regular (compare Fig. 2 G to A; also see ref. 6). In a similar manner, gsb expression also diverges from the wild-type pattern to form wider, less regular stripes (compare Fig. 2 H to B). At stage 11 ptc expression remains broad in ciD embryos, rather than narrowing to two single cell wide stripes at the edges of the anterior compartment (compare Fig. 2 I to C; also see ref. 6). The ciD phenotype with respect to wg and gsb expression is not fully penetrant because some segments show the wild-type expression pattern of these genes. This phenotype is radically different from that of ci null embryos and is also different from that of wild-type ci (compare Figs. 1 and 2).
There are at least three possible explanations for the unusual activity of ciD. ciD might increase the sensitivity of the cells to the Hh signal, ciD might alter distribution of the Hh signal, or ciD might substitute for Hh activity. To differentiate between these possibilities, we examined expression patterns of wg, gsb, and ptc in populations of embryos from hhGS1/TM3; ciD/M63a flies. hhGS1 is a strong allele of hh that can be considered a null (3). In hhGS1; ciD homozygous embryos ptc expression resembles that of ciD embryos. wg and gsb expression in the double mutant embryos shows the hh phenotype in some segments and the ciD phenotype in others. The partial restoration of wg and gsb expression is consistent with the poor penetrance of ciD. Apparently ciD can maintain expression of hh-responsive genes in the ectoderm in the absence of active Hh protein. This is consistent with the observation of Alexandre et al. (27) that overexpression of wild-type Ci can rescue expression of Hh-responsive genes in a hh mutant. We find that ci is necessary for Hh signaling and ciD can substitute for Hh protein in driving expression of Hh-responsive genes. Taken together, these two results confirm that hh acts through ci to activate transcription of its target genes.
Ci Acts Downstream of smo and fu.
If Ci is the transcriptional effector of Hh signaling then it should act downstream of all other known components of the pathway. Ci has already been shown to act downstream of ptc (24, 25, 45). We have tested epistasis of ci with two other members of the signaling pathway. The smoothened (smo) gene encodes a putative receptor for the Hh signal (46, 47). The fu gene encodes a Ser/Thr kinase that is necessary for Hh signal transduction (24, 43, 48). wg, gsb, and ptc expression patterns in smo;ciD and fu;ciD double mutants demonstrate that ciD is epistatic to both smo and fu. The phenotypes of these double mutants are similar to hhGS1;ciD double mutants (data not shown).
Ci Is a Sequence-Specific DNA Binding Protein.
ci encodes a zinc finger protein that is a member of the Gli family (29). The zinc finger domains of three of these, Gli, Gli3, and Tra-1, all have sequence-specific DNA binding activity. Furthermore, all three recognize the same 9 bp core consensus sequence: -TGGGTGGTC- (49–51). Ci has a high degree of sequence similarity to these proteins within the zinc finger domains and is identical to Gli in all amino acids that make base contacts, as determined by crystal structure (52). Therefore, we predicted that Ci and Gli would bind DNA with the same sequence specificity.
To demonstrate sequence specificity, we constructed a fusion protein in which the zinc finger domain of Ci is fused to GST (36). Purified GST-Ci fusion protein was used in EMSAs (Fig. 3). We found that purified GST-Ci bound to a 36-bp nucleotide that contained the 9-bp core consensus binding site (BS-1), but not to a control DNA (BS-2) containing a single G-A mutation that changes the core consensus binding site (data not shown). Competition with unlabeled BS-1 and unlabeled BS-2 showed markedly different effects. BS-1 competed efficiently for for binding to GST-Ci whereas BS-2 did not. Furthermore, an antibody that recognizes GST is able to supershift the complex, demonstrating that the complex is formed with GST-Ci/BS-1. Thus, the zinc finger domain of Ci binds DNA in a sequence-specific manner.
Ci Activates Transcription from the wg Promoter in Cultured Cells.
As shown above, Ci is a sequence-specific DNA binding protein that is necessary for wg expression. Therefore we predicted that Ci would be capable of activating transcription from wg promoter in transiently transfected Schneider line 2 (S2) cells. The reporter construct, wg-Luc, drives luciferase expression with a 5-kb fragment of wg promoter (Fig. 4A). In vivo this wg promoter fragment drives LacZ expression in a pattern that accurately reproduces most aspects of endogenous wg expression (D.L. and R.N., unpublished data). tk-Luc used as a control for promoter specificity, drives Luciferase expression off the thymidine kinase minimal promoter (33). All transfections included the copia–β-gal plasmid as an internal control for transfection efficiency.
wg-Luc expression but not tk-Luc expression is stimulated by coexpression of Ci (Fig. 4B). The effect of Ci on wg-Luc is bimodal. Low amounts of Ci stimulated wg-Luc to a maximum of 3.6-fold with 20 ng of mt-Ci. Higher concentrations of Ci decreased luciferase activity. This complex response to Ci dosage probably reflects interaction of distal activating elements and proximal inhibitory elements within the wg promoter; the inhibitory phase of the Ci response is lost when only a distal fragment of wg promoter is used (see below). This experiment demonstrates that Ci regulates gene expression from the wg promoter and that this effect is promoter-specific. It does not address whether Ci acts directly. In addition, wg-Luc transcripts retain 436 nt of wg 5′ untranslated region, leaving open the possibility that the effects of Ci on wg-Luc are posttranscriptional.
Ci-Responsive Elements Map to the Distal End of the wg Promoter.
To demonstrate direct transcriptional involvement of Ci, we mapped the Ci-responsive elements in wg-Luc. wg-LucΔS and wg-LucΔB, which respectively delete 1 kb and 2 kb from the distal end of wg-Luc (Fig. 4A), were tested for transcriptional activation by cotransfection of mt-Ci in S2 cells (Fig. 4C). wg-LucΔS showed a bimodal response to Ci similar to that of the intact wg promoter fragment, though the magnitude was significantly reduced. Maximal activation of only 1.7-fold was achieved with 200 ng mt-Ci while more mt-Ci decreased activation. wg-LucΔB showed poor transcriptional activation, only 1.4-fold activation at 300 ng mt-Ci. This suggests that the primary elements that are activated by Ci are in the distal regions of the promoter/enhancer.
To test this we used the distal 1 kb of wg promoter to drive Luciferase expression from the tk minimal promoter (Δwg-Luc; Fig. 4A). Δwg-Luc responded over a 100-fold range of increasing mt-Ci concentrations with maximal stimulation of 109-fold by 3,000 ng mt-Ci (Fig. 4D). This unambiguously maps Ci-activated elements of the wg promoter to the distal 1-kb fragment. Because Δwg-Luc is not inhibited by high levels of Ci, specific elements in the deleted proximal region probably mediate the Ci inhibition seen with the full-length promoter (Fig. 4B). Because no wg 5′ untranslated region is retained in Δwg-Luc, this eliminates concerns that the Ci effect is posttranscriptional. The difference in the magnitude of activation of Δwg-Luc versus wg-Luc (109-fold versus 3.6-fold) could be related to the much higher basal expression of wg-Luc relative to Δwg-Luc. The 4 kb deleted in Δwg-Luc might eliminate many Ci-independent activating elements. In addition, the tk minimal promoter might respond less well than the wg minimal promoter to enhancer elements retained in Δwg-Luc. While these issues are germane to an integrated view of wg promoter function, they are not critical to understanding whether Ci directly regulates transcription from the wg promoter.
Ci Binding Sites Mediate Transcriptional Activation of Δwg-Luc by Ci.
If Ci directly regulates transcription from wg promoter, then binding of Ci to sequences in wg promoter will be necessary for transcriptional activation. The wg promoter sequence has 10 possible Gli consensus binding sites, with 3 pairs of sites in the distal 1.2 kb (Fig. 4A). wg-LucΔS eliminated 2 of these pairs and wg-LucΔB eliminated all 3. Δwg-Luc included the distal 2 of these pairs of putative Ci binding sites.
To confirm direct interaction of Ci with its putative binding sites, DNA fragments containing each of the distal two pairs of putative Ci binding sites were analyzed by EMSA (Fig. 5). Fragment 1 corresponds to the region of wg promoter containing the first two putative Ci binding sites (−4652 to −4422), and fragment 2 contains the second pair of putative Ci binding sites (−4037 to −3787). GST-Ci was able to shift both fragments 1 and 2. This confirmed the presence of Ci binding sites in wg promoter fragments that mediate transcriptional activation.
To directly test whether the Ci binding sites in Δwg-Luc mediated its activation by Ci, we mutagenized the Ci binding sites in Δwg-Luc. In Δwg*-Luc the four putative Ci binding sites were mutagenized with the same G → A mutation that distinguishes BS-1 from BS-2 and eliminates Ci binding (Fig. 3). These mutations severely compromised activation by Ci in the transient transfection assay (Fig. 4D). The mutant Δwg*-Luc was activated only 9.1-fold with maximal Ci concentrations, compared with 109-fold for Δwg-Luc. That is, mutation of the Ci binding sites resulted in greater than 90% reduction in Ci-dependent transcriptional activation. An EMSA was used to confirm that the mutations in Δwg*-Luc compromised DNA binding (Fig. 5). Mutagenesis of fragment 2 completely eliminated the mobility shift by GST-Ci (Fig. 5B). Mutagenesis of fragment 1 reduced its efficiency of binding to GST-Ci by 7-fold (Fig. 5A), rather than completely eliminating the binding. The residual binding could be explained by weak Ci binding sites missed by sequence search or by residual binding at either of the two mutagenized sites. Nevertheless, the overall binding of Ci to Δwg*-Luc was greatly reduced by mutagenesis of four Ci binding sites. Reduced binding affinity correlates with greater than 10-fold difference in transcriptional activation (Fig. 4D).
In summary, these results demonstrate that Ci is capable of binding to wg promoter fragments, that mutagenesis of the putative Ci binding sites in these fragments greatly reduces Ci binding, and that the reduced binding correlates with greater than 90% reduction in transcriptional activation. This demonstrates that Ci binding sites profoundly influence transcriptional activation by Ci. In other words, Ci binding to DNA is necessary for Ci to mediate transcriptional activation. We conclude that Ci activates transcription by direct binding to its recognition sites wg promoter/enhancer.
DISCUSSION
The genetic data presented here show that ci is the last identified step in the Hh signal transduction pathway. First we demonstrated that ci is necessary for maintenance of Hh-responsive gene expression. Our use of a ci null allele provides rigorous support for previous studies which used hypomorphic alleles (24, 26, 44, 53), and further show that when ci is absent hh overexpression is unable to activate transcription of target genes. Second, we demonstrated that ci activity is sufficient to activate Hh-responsive gene expression in the absence of hh, smo, or fu. This supports and extends recent work by Alexandre et al. (27) who demonstrated ci overexpression rescues wg expression in hh mutants. Combined with previous work showing that ci is epistatic to ptc (24, 25, 45), this argues strongly that ci is the last step in the Hh signaling pathway before transcriptional activation.
This work demonstrates that Ci acts as a sequence-specific DNA binding protein. Orenic et al. (28) suggested DNA binding activity for Ci based on its zinc fingers. The sequence-specific DNA binding of other Gli family zinc finger domains supports this notion (49–51). Here we have demonstrated that the Ci zinc finger domain binds DNA with the same sequence specificity as other Gli zinc finger domains. We have identified the wg promoter as a target for Ci interaction and have identified Ci binding sites in the wg promoter. Most important, we have shown that direct interaction between Ci and sequences in the promoter/enhancer region of the wg gene is sufficient for Ci-mediated transcriptional activation. ptc, another target for Hh, can respond to Hh via a promoter fragment containing Gli consensus binding sites (6, 27). The genetic and molecular results taken together provide strong evidence that Ci is a transcription factor that mediates transcriptional activation of Hh-responsive genes.
How is Ci activated by Hh? Ci protein is expressed in a complimentary pattern to Hh in both embryos and imaginal discs (25, 54). Ci is activated in cells adjacent to Hh-expressing cells but not in cells at a distance from Hh-expressing cells. Because Ci mRNA levels are not modulated by the Hh signal (28, 55), ci must be posttranscriptionally activated by Hh. Ci protein levels are elevated in cells responding to the Hh signal (25, 26, 54) and Ci overexpression can drive expression of Hh-responsive genes in the absence of Hh (27), so Ci protein levels may play a role in response to the Hh signal. However, Ci protein must also have active and inactive states to account for the lack of activity in cells that do not receive the Hh signal but still express the modest levels of Ci protein.
Ci protein is predominantly cytoplasmic in all cells, regardless of whether they receive the Hh signal (25). Because Ci acts as a transcription factor, the small amount of nuclear Ci detected adjacent to Hh-secreting cells (25) must be sufficient to activate transcription of Hh responsive genes. Future studies must resolve the apparent paradox of why this transcription factor is predominantly cytoplasmically localized and how its subcellular localization relates to its activation.
Acknowledgments
We thank Aleem Siddiqui and Sandy Martin for critical comments on the manuscript, and Robert Holmgren, David Zarkower, Alex Franzusoff, and Victoria McKane for reagents. We would particularly like to thank Ben Perryman and Erik Bush for the mutagenesis kit. This work was supported by Grant GM45396 from the National Institutes of Health.
ABBREVIATIONS
- GST-Ci
glutathione-S-transferase/Ci fusion protein
- β-gal
β-galactosidase
- EMSA
electrophoretic mobility-shift assay
Footnotes
References
- 1.Perrimon N. Cell. 1994;76:781–784. doi: 10.1016/0092-8674(94)90351-4. [DOI] [PubMed] [Google Scholar]
- 2.Baker N E. EMBO J. 1987;6:1765–1773. doi: 10.1002/j.1460-2075.1987.tb02429.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Mohler J, Vani K. Development (Cambridge, UK) 1992;115:957–971. doi: 10.1242/dev.115.4.957. [DOI] [PubMed] [Google Scholar]
- 4.Tabata T, Eaton S, Kornberg T B. Genes Dev. 1992;6:2635–2645. doi: 10.1101/gad.6.12b.2635. [DOI] [PubMed] [Google Scholar]
- 5.Lee J J, von Kessler D P, Parks S, Beachy P A. Cell. 1992;71:33–50. doi: 10.1016/0092-8674(92)90264-d. [DOI] [PubMed] [Google Scholar]
- 6.Hidalgo A, Ingham P. Development (Cambridge, UK) 1990;110:291–301. doi: 10.1242/dev.110.1.291. [DOI] [PubMed] [Google Scholar]
- 7.Noordermeer J, Klingensmith J, Nusse R. Mech Dev. 1995;51:145–155. doi: 10.1016/0925-4773(95)00348-7. [DOI] [PubMed] [Google Scholar]
- 8.Heemskerk J, DiNardo S. Cell. 1994;76:449–460. doi: 10.1016/0092-8674(94)90110-4. [DOI] [PubMed] [Google Scholar]
- 9.Fietz M J, Jacinto A, Taylor A M, Alexandre C, Ingham P W. Curr Biol. 1995;6:643–650. doi: 10.1016/s0960-9822(95)00129-1. [DOI] [PubMed] [Google Scholar]
- 10.Basler K, Struhl G. Nature (London) 1994;368:208–214. doi: 10.1038/368208a0. [DOI] [PubMed] [Google Scholar]
- 11.Diaz-Benjumea F J, Cohen B, Cohen S M. Nature (London) 1994;327:175–179. doi: 10.1038/372175a0. [DOI] [PubMed] [Google Scholar]
- 12.Capdevila J, Guerrero I. EMBO J. 1994;13:4459–4468. doi: 10.1002/j.1460-2075.1994.tb06768.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tabata T, Kornberg T. Cell. 1994;76:89–102. doi: 10.1016/0092-8674(94)90175-9. [DOI] [PubMed] [Google Scholar]
- 14.Ingham P W, Fietz M J. Curr Biol. 1995;5:432–440. doi: 10.1016/s0960-9822(95)00084-4. [DOI] [PubMed] [Google Scholar]
- 15.Wolff T, Ready D F. In: The Development of Drosophila melanogaster. Bate M, Martinez-Arias A, editors. Plainview, NY: Cold Spring Harbor Lab. Press; 1993. pp. 1277–1325. [Google Scholar]
- 16.Ma C, Zhou Y, Beachy P A, Moses K. Cell. 1993;75:927–938. doi: 10.1016/0092-8674(93)90536-y. [DOI] [PubMed] [Google Scholar]
- 17.Ma C, Moses K. Development (Cambridge, UK) 1995;121:2279–2289. doi: 10.1242/dev.121.8.2279. [DOI] [PubMed] [Google Scholar]
- 18.Heberlein U, Singh C M, Luk A Y, Donohoe T J. Nature (London) 1995;373:709–711. doi: 10.1038/373709a0. [DOI] [PubMed] [Google Scholar]
- 19.Ingham P W, Taylor A M, Nakano Y. Nature (London) 1991;353:184–187. doi: 10.1038/353184a0. [DOI] [PubMed] [Google Scholar]
- 20.Hooper J E, Scott M P. Cell. 1989;59:751–765. doi: 10.1016/0092-8674(89)90021-4. [DOI] [PubMed] [Google Scholar]
- 21.Nakano Y, Guerrero I, Hidalgo A, Taylor A, Whittle J R S, Ingham P W. Nature (London) 1989;341:508–513. doi: 10.1038/341508a0. [DOI] [PubMed] [Google Scholar]
- 22.Bejsovec A, Wieschaus E. Development (Cambridge, UK) 1993;119:501–517. doi: 10.1242/dev.119.2.501. [DOI] [PubMed] [Google Scholar]
- 23.Perrimon N. Cell. 1995;80:517–520. doi: 10.1016/0092-8674(95)90503-0. [DOI] [PubMed] [Google Scholar]
- 24.Forbes, A. J., Nakano, Y., Taylor, A. M. & Ingham, P. W.(1993) Development (Cambridge, U.K.) Suppl., 115–124. [PubMed]
- 25.Motzny C K, Holmgren R. Mech Dev. 1995;52:137–150. doi: 10.1016/0925-4773(95)00397-j. [DOI] [PubMed] [Google Scholar]
- 26.Domínguez M, Brunner M, Hafen E, Basler K. Science. 1996;272:1621–1625. doi: 10.1126/science.272.5268.1621. [DOI] [PubMed] [Google Scholar]
- 27.Alexandre C, Jacinto A, Ingham P W. Genes Dev. 1996;10:2003–2013. doi: 10.1101/gad.10.16.2003. [DOI] [PubMed] [Google Scholar]
- 28.Orenic T, Slusarski C, Kroll K, Holmgren R. Genes Dev. 1990;4:1053–1067. doi: 10.1101/gad.4.6.1053. [DOI] [PubMed] [Google Scholar]
- 29.Kinzler K W, Ruppert J M, Bigner S H, Vogelstein B. Nature (London) 1988;332:371–374. doi: 10.1038/332371a0. [DOI] [PubMed] [Google Scholar]
- 30.Ingham P W. Nature (London) 1993;366:560–562. doi: 10.1038/366560a0. [DOI] [PubMed] [Google Scholar]
- 31.Tautz D, Pfeiffle C. Chromosoma. 1989;98:81–85. doi: 10.1007/BF00291041. [DOI] [PubMed] [Google Scholar]
- 32.Jiang J, Hoey T, Levine M. Genes Dev. 1991;5:265–277. doi: 10.1101/gad.5.2.265. [DOI] [PubMed] [Google Scholar]
- 33.Nordeen S K. BioTechniques. 1988;6:454–456. [PubMed] [Google Scholar]
- 34.Brown N H, Kafatos F C. J Mol Biol. 1988;203:425–437. doi: 10.1016/0022-2836(88)90010-1. [DOI] [PubMed] [Google Scholar]
- 35.Bunch T A, Grindblat Y, Goldstein L S B. Nucleic Acids Res. 1988;16:1043–1061. doi: 10.1093/nar/16.3.1043. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Smith D B, Johnson K S. Gene. 1988;67:31–40. doi: 10.1016/0378-1119(88)90005-4. [DOI] [PubMed] [Google Scholar]
- 37.Han K, Levine M S, Manley J L. Cell. 1989;56:573–583. doi: 10.1016/0092-8674(89)90580-1. [DOI] [PubMed] [Google Scholar]
- 38.Miller J H. Experiments in Molecular Genetics. Plainview, NY: Cold Spring Harbor Lab. Press; 1972. [Google Scholar]
- 39.de Wet J R, Wood K V, DeLuca M, Helinski D R, Subramani S. Mol Cell Biol. 1987;7:725–737. doi: 10.1128/mcb.7.2.725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Frangioni J V, Neel B G. Anal Biochem. 1993;210:179–187. doi: 10.1006/abio.1993.1170. [DOI] [PubMed] [Google Scholar]
- 41.Sambrook J, Fritsch E F, Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Plainview, NY: Cold Spring Harbor Lab. Press; 1989. [Google Scholar]
- 42.Hidalgo A. Mech Dev. 1991;35:77–87. doi: 10.1016/0925-4773(91)90059-f. [DOI] [PubMed] [Google Scholar]
- 43.Ingham P W, Hidalgo A. Development (Cambridge, UK) 1993;117:283–291. doi: 10.1242/dev.117.1.283. [DOI] [PubMed] [Google Scholar]
- 44.Slusarski D C, Motzny C K, Holmgren R. Genetics. 1995;139:229–240. doi: 10.1093/genetics/139.1.229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Hooper J E. Nature (London) 1994;372:461–464. doi: 10.1038/372461a0. [DOI] [PubMed] [Google Scholar]
- 46.Alcedo J, Ayzenzon M, Von Ohlen T, Noll M, Hooper J E. Cell. 1996;86:221–232. doi: 10.1016/s0092-8674(00)80094-x. [DOI] [PubMed] [Google Scholar]
- 47.van den Heuvel M, Ingham P W. Nature (London) 1996;382:547–551. doi: 10.1038/382547a0. [DOI] [PubMed] [Google Scholar]
- 48.Preat T, Therond P, Lamour-Isnard C, Limbourg-Bouchon B, Tricoire H, Erk I, Mariol M-C, Busson D. Nature (London) 1990;347:87–89. doi: 10.1038/347087a0. [DOI] [PubMed] [Google Scholar]
- 49.Kinzler K W, Vogelstein B. Mol Cell Biol. 1990;10:634–642. doi: 10.1128/mcb.10.2.634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Ruppert J M, Vogelstein B, Arheden K, Kinzler K W. Mol Cell Biol. 1990;10:5408–5415. doi: 10.1128/mcb.10.10.5408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Zarkower D, Hodgkin J. Nucleic Acids Res. 1993;21:3691–3698. doi: 10.1093/nar/21.16.3691. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Pavletich N P, Pabo C O. Science. 1993;261:1701–1707. doi: 10.1126/science.8378770. [DOI] [PubMed] [Google Scholar]
- 53.van den Heuvel, M., Klingensmith, J., Perrimon, N. & Nusse, R. (1993) Development (Cambridge, U.K.) Suppl., 105–114. [PubMed]
- 54.Johnson R L, Grenier J K, Scott M P. Development (Cambridge, UK) 1995;121:4161–4170. doi: 10.1242/dev.121.12.4161. [DOI] [PubMed] [Google Scholar]
- 55.Eaton S, Kornberg T B. Genes Dev. 1990;4:1068–1077. doi: 10.1101/gad.4.6.1068. [DOI] [PubMed] [Google Scholar]