Axial (HNF3β) and retinoic acid receptors are regulators of the zebrafish sonic hedgehog promoter
1997; Springer Nature; Volume: 16; Issue: 13 Linguagem: Inglês
10.1093/emboj/16.13.3955
ISSN1460-2075
AutoresBei‐En Chang, Patrick Blader, Nadine Fischer, Philip W. Ingham, Uwe Strähle,
Tópico(s)Epigenetics and DNA Methylation
ResumoArticle1 July 1997free access Axial (HNF3β) and retinoic acid receptors are regulators of the zebrafish sonic hedgehog promoter Bei-En Chang Bei-En Chang IGBMC, CNRS/INSERM/ULP, BP 163, 67404 Illkirch, France Search for more papers by this author Patrick Blader Patrick Blader IGBMC, CNRS/INSERM/ULP, BP 163, 67404 Illkirch, France Search for more papers by this author Nadine Fischer Nadine Fischer IGBMC, CNRS/INSERM/ULP, BP 163, 67404 Illkirch, France Search for more papers by this author Philip W. Ingham Philip W. Ingham ICRF, 44 Lincoln's Inn Fields, London, WC2A 3PX UK Search for more papers by this author Uwe Strähle Corresponding Author Uwe Strähle IGBMC, CNRS/INSERM/ULP, BP 163, 67404 Illkirch, France Search for more papers by this author Bei-En Chang Bei-En Chang IGBMC, CNRS/INSERM/ULP, BP 163, 67404 Illkirch, France Search for more papers by this author Patrick Blader Patrick Blader IGBMC, CNRS/INSERM/ULP, BP 163, 67404 Illkirch, France Search for more papers by this author Nadine Fischer Nadine Fischer IGBMC, CNRS/INSERM/ULP, BP 163, 67404 Illkirch, France Search for more papers by this author Philip W. Ingham Philip W. Ingham ICRF, 44 Lincoln's Inn Fields, London, WC2A 3PX UK Search for more papers by this author Uwe Strähle Corresponding Author Uwe Strähle IGBMC, CNRS/INSERM/ULP, BP 163, 67404 Illkirch, France Search for more papers by this author Author Information Bei-En Chang1, Patrick Blader1, Nadine Fischer1, Philip W. Ingham2 and Uwe Strähle 1 1IGBMC, CNRS/INSERM/ULP, BP 163, 67404 Illkirch, France 2ICRF, 44 Lincoln's Inn Fields, London, WC2A 3PX UK *Corresponding author. E-mail: [email protected] The EMBO Journal (1997)16:3955-3964https://doi.org/10.1093/emboj/16.13.3955 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The signalling molecule Sonic hedgehog is involved in a multitude of distinct patterning processes during vertebrate embryogenesis. In the nascent body axis of the zebrafish embryo, sonic hedgehog is co-expressed with axial (HNF3β in mammals), a transcription regulator of the winged helix family. We show here that misexpression of axial leads to ectopic activation of sonic hedgehog expression in the zebrafish, suggesting that axial is a regulator of sonic hedgehog transcription. The sonic hedgehog gene was cloned from zebrafish and its promoter was characterized with respect to activation by axial. Expression of axial or rat HNF3β in HeLa cells results in activation of co-transfected sonic hedgehog promoter–CAT fusion genes. This effect is mediated by two Axial (HNF3β) recognition sequences. We furthermore identified a retinoic acid response element (RARE) in the sonic hedgehog upstream region which can be bound by retinoic acid receptor (RAR) and retinoid X receptor (RXR) heterodimers in vitro and confers retinoic acid inducibility to the sonic hedgehog promoter in the HeLa cell system. Our results suggest that both Axial (HNF3β) and retinoic acid receptors are direct regulators of the sonic hedgehog gene. Introduction Cell–cell signalling plays a crucial role in the differentiation of specific cell types during development. Sonic hedgehog (Shh), one of several vertebrate members of the hedgehog family of signalling molecules, has been implicated in the induction of ventral cell fates in the neural tube, anteroposterior specification of the limb, patterning of the somites as well as establishment of asymmetry across the left–right axis (Echelard et al., 1993; Krauss et al., 1993; Riddle et al., 1993; Fan and Tessier-Lavigne, 1994; Johnson et al., 1994; Laufer et al., 1994; Roelink et al., 1994; Ekker et al., 1995; Ericson et al., 1995; Fan et al., 1995; Levin et al., 1995; Münsterberg et al., 1995; Chiang et al., 1996; Currie and Ingham, 1996). This multitude of distinct processes controlled by Shh during embryogenesis is reflected in both temporally and spatially restricted expression of the shh gene (Echelard et al., 1993; Krauss et al., 1993; Riddle et al., 1993; Roelink et al., 1994; Ekker et al., 1995). Several lines of evidence show that the spatially restricted expression is essential for orderly inductive patterning of embryonic tissues by Shh. Aberrant expression of shh in the neural tube, for example, results in ectopic activation of target genes at dorsal aspects of the neural tube (Echelard et al., 1993; Krauss et al., 1993; Barth and Wilson, 1995; Ekker et al., 1995; MacDonald et al., 1995; Concordet et al., 1996; Hauptmann and Gerster, 1996). Similarly, ectopic expression of shh in the anterior wing bud causes mirror image duplications in the wing (Riddle et al., 1993; Laufer et al., 1994). Also, the absolute levels of Shh protein secreted from a localized source may be crucial; high levels of Shh protein induce floor plate, whereas lower levels induce motor neurons in neural plate explants (Marti et al., 1995; Roelink et al., 1995). Little is known about the mechanisms underlying the spatially restricted expression of shh. In the developing body axis of vertebrate embryos, shh expression is tightly correlated with expression of HNF3β, a transcription regulator of the winged-helix class (Ang et al., 1993; Echelard et al., 1993; Krauss et al., 1993; Monahagan et al., 1993; Riddle et al., 1993; Ruiz i Altaba et al., 1993; Sasaki and Hogan, 1993; Strähle et al., 1993; Roelink et al., 1994; Ekker et al., 1995). Targeted mutation of HNF3β in mouse results in impaired notochord development and loss of shh expression (Ang and Rossant, 1994; Weinstein et al., 1994). Overexpression of HNF3β induces aberrant activation of shh in the neural tube of Xenopus and mouse embryos (Ruiz i Altaba et al., 1995a; Hynes et al., 1995). In the chicken wing bud and the zebrafish fin bud, administration of retinoic acid leads similarly to activation of shh, indicating retinoic acid receptors (RARs) as further potential regulators of shh expression (Riddle et al., 1993; Helms et al., 1994; Akimenko and Ekker, 1995; Stratford et al., 1996). It remains unclear, however, whether shh expression is regulated directly by HNF3β and RARs or whether the experimental manipulations of the embryos affect shh expression indirectly. In the zebrafish embryo, as in other vertebrate embryos, axial (axl), the zebrafish homologue of HNF3β, and shh are expressed in overlapping domains in the embryonic shield (zebrafish organizer), the notochord and the floor plate (Krauss et al., 1993; Strähle et al., 1993, 1996). Onset of axl expression in the embryonic shield precedes that of shh (Krauss et al., 1993; Strähle et al., 1993), consistent with the notion that shh expression might be regulated by axl. To test this, we misexpressed axl in the zebrafish embryo and show here that this results in ectopic activation of shh expression in a wide variety of different cells normally not expressing the gene. We cloned the shh gene from zebrafish and characterized its promoter in a cell system. Evidence is provided that Axl (HNF3β) and RARs are direct regulators of the shh promoter. Results Ectopic expression of axl activates shh transcription The spatial and temporal relationships between axl and shh in the developing axis of the zebrafish embryo suggest that axl may regulate shh transcription. To test this possibility, we constructed an expression plasmid (pCMVaxl) in which the axl cDNA is placed under the control of the cytomegalovirus (CMV) enhancer/promoter (Boshart et al., 1985). Plasmid DNA injected into 1–4 cell stage embryos is highly mosaically distributed (Westerfield et al., 1992) and, since the CMV control region is active in many different cell types in the zebrafish embryo (J.Wegner, U.Strähle and M.Westerfield, unpublished), such injection generates embryos mosaic for axl-expressing cells throughout the body. Injected embryos were fixed after either 12 or 24 h development and hybridized with a shh probe. Embryos at 12 h of development show strong ectopic expression of shh at lateral and ventral positions of the embryo in both the ectoderm and the mesendoderm (Figure 1A and B). At 24 h of development, ectopic shh expression is detectable in pCMVaxl-injected embryos (n = 94) in many different cell types (Figure 1C) that do not usually express shh (Krauss et al., 1993). These ectopic sites of expression include cells of the yolk sac (Figure 1D), the gut (Figure 1E), the dorsal aspects of the neural tube (Figure 1F) and the somites (Figure 1G). Embryos (n = 53) injected with CMVβ (MacGregor and Caskey, 1989), in which the axl coding region has been replaced by lacZ, did not show ectopic activation of shh expression (data not shown). Figure 1.Ectopic expression of axl leads to transcription of shh in a wide variety of cell types. Plasmids carrying the axl cDNA under control of the CMV enhancer/promoter (pCMVaxl) were injected into one of the blastomeres of cleavage stage embryos up to the 4-cell stage. Embryos were fixed either after 12 (A and B) or 24 h development (C–G) and hybridized to the shh probe. (A) Five-somite stage embryo injected with pCMVaxl. Strong activation of shh expression in lateral parts of the embryo (arrow) is evident in addition to normal expression along the dorsal midline (arrowhead). Dorsal view, anterior up. (B) Transverse section through the embryo shown in (A). Cells from ectodermal (thin short arrow) and mesodermal (thick arrow) origin express shh in injected embryos. The arrowhead points to shh expression in the notochord and neuroectoderm. Dorsal up. The position of the yolk is indicated by y. (C) pCMVaxl-injected embryo stained with the shh probe at 24 h of development. shh is normally expressed in the ventral diencephalon (arrowhead with dot) and in the ventral neural tube from the midbrain to the tail (arrowheads). The nascent notochord in the tail bud also expresses shh still at this stage in uninjected embryos (arrowhead with asterisk). In addition to the normal expression, numerous cells express shh ectopically (arrows point out some cells) in embryos injected with pCMVaxl. The orientation is anterior up, dorsal left. The yolk was removed. (D) Ectopic activation of shh in cells of the yolk sac (arrow). (E) Ectopic expression of shh in the endoderm (arrow heads). (Lateral view at the level of the anterior trunk, dorsal up; arrowhead, normal shh expression in neural tube.) (F) Ectopic activation of shh in the anterior dorsal neural tube. (G) Transverse section through the tail of a pCMVaxl-injected embryo showing ectopic activation of shh in a somitic cell (arrow). The arrowhead indicates normal expression in neural tube and notochord. Download figure Download PowerPoint To assess the distribution of misexpressed Axl with respect to ectopic activation of shh in injected embryos, the axl coding region was tagged with six copies of the myc epitope recognized by monoclonal antibody 9E10 (Evan et al., 1985). In 11 h old embryos (n = 54) double-labelled with antibody 9E10 and shh RNA probe, a large number of cells ectopically expressing shh showed co-expression of myc–axial. However, a substantial number of cells expressed 9E10 immunoreactivity but did not show detectable shh expression. Similar results were obtained when CMVβ was co-injected with pCMVaxl, the untagged version of the axl expression plasmid (data not shown). This indicates that shh cannot be activated in all cells misexpressing axl. A small fraction of cells was also found which showed shh expression but in which 9E10 immunoreactivity could not be detected. In summary, the ectopic activation of shh in a predominantly cell-autonomous fashion suggests that Axl may be a direct regulator of the shh gene. Cloning of the shh gene To investigate the molecular mechanism of the observed activation of shh by Axl in the embryo, we screened a zebrafish genomic library (Molven et al., 1991) with a shh cDNA fragment (Krauss et al., 1993) to isolate the shh promoter. Two clones (λshh5 and λshh7) were isolated and shown to overlap by restriction mapping (Figure 2). After subcloning into plasmid vectors, the exon–intron structure was established by sequence analysis using primers complementary to the cDNA sequence (Figure 2). Both phage inserts span the entire shh coding region. The mature shh mRNA is encoded by three exons, and the exon–intron junctions are conserved among zebrafish shh, mouse shh and Drosophila hedgehog (Lee et al., 1992; Mohler and Vani, 1992; Chiang et al., 1996). The start sites were mapped by RNase protection analysis and 5′ RACE. Two major transcription start sites were found, one 253 bp and the other 339 bp upstream of the translation initiation codon (Figure 3) corresponding roughly to the ends of previously isolated cDNA clones (Krauss et al., 1993; Roelink et al., 1994). The second start site will be referred to as position +1. Both start sites are used equally during development at least up to 48 h (data not shown). Figure 2.Genomic structure of the zebrafish shh gene. The exons (black boxes) covering the coding region are outlined (ATG, translation initiation codon; TGA, translation termination codon). Exon 1 comprises the 5′ non-coding sequence and the first 100 amino acids, exon 2 encodes amino acids 101–187 and exon 3 encodes amino acids 188–418. The two independently isolated phages (λshh5 and λshh7) from which the map was deduced are outlined. Restriction sites indicated with vertical arrows are: S, SalI; H, HindIII; X, XhoI. Bar, 1 kb. Download figure Download PowerPoint Figure 3.Sequence of the shh promoter. Transcription start sites and translation initiation codon are indicated by horizontal arrows and M, respectively. The HNF3β-binding sites (HNF3β1 and HNF3β2) which show strong homology to the HNF3β recognition sequence are located upstream of the start sites. The DR5 RARE and a putative TATA box are indicated. Download figure Download PowerPoint A 376 bp shh promoter fragment is sufficient for transactivation by zebrafish axl and rat HNF3β To test whether axl transactivates the shh promoter, a 2.5 kb SalI–XhoI fragment containing ∼2.2 kb of upstream sequence was cloned in front of the coding region of chloramphenicol acetyl transferase (CAT), and co-transfection experiments in HeLa cells were carried out. Co-transfection of the −2200shhCAT reporter construct with pCMVaxl but not with the control plasmid CMVβ led to a strong increase in CAT activity (Figure 4). Transfection of −563shhCAT containing 563 bp of shh promoter sequence elicited the same response, when co-transfected with pCMVaxl. Replacement of the shh sequences by the herpes simplex thymidine kinase promoter (−105 to +51; McKnight and Kingsbury, 1982) did not show transactivation by axl (data not shown), suggesting that the increase in CAT activity in response to axl is mediated by the shh sequences inserted upstream of the CAT coding region. To delineate the sequences required for transactivation by Axl more precisely, 5′ deletions of the shh upstream region were generated and tested by co-transfection with pCMVaxl or CMVβ. Plasmid −155shhCAT harbouring 376 bp of shh sequence shows the same response to co-expression of axl as the plasmids containing larger shh inserts (Figure 4). Thus, 376 bp of shh sequence comprising 155 bp upstream of the transcription start site is sufficient to mediate transactivation by Axl. Figure 4.Deletion analysis of the shh promoter. A 155 bp shh promoter sequence is sufficient for transactivation by Axl in HeLa cells. HeLa cells were transfected with shh–CAT fusion genes together with the axl expression plasmid pCMVaxl (Axl) or the control plasmid CMVβ (C). In CMVβ, the axl coding region has been replaced by lacZ. Plasmids −2200shhCAT and −563shhCAT contain the 2.5 kb SalI–XhoI and the 0.8 kb PstI–XhoI shh fragments in front of the coding region of CAT. The extent of shh sequence in 5′ deletion plasmids −469shhCAT, −364shhCAT, −267shhCAT and −155shhCAT is indicated by black bars in the schematic drawing of the constructs. The transcription start sites (horizontal arrows) and relevant restriction sites (XhoI, PstI, SalI) are indicated. CAT enzymatic activity was determined in extracts from transfected cells after 48 h. The bar diagram shows the average of three independent transfections done in parallel. Standard deviations are indicated. Relative CAT activity was calculated by normalizing the protein concentration and setting the CAT activity obtained from −2200shhCAT as 100%. Download figure Download PowerPoint Zebrafish axl is closely related to mammalian HNF3β, sharing 97% sequence homology in the winged-helix DNA-binding domain (Lai et al., 1991; Strähle et al., 1993). It is thus likely that HNF3β and Axl have similar DNA recognition sequences. To test whether HNF3β activates the zebrafish shh promoter, increasing amounts of CMVHNF3β coding for rat HNF3β under control of the CMV promoter/enhancer (Pani et al., 1992) were transfected into HeLa cells together with a constant amount of −155shhCAT. For comparison, increasing amounts of pCMVaxl were transfected with −155shhCAT in parallel. HNF3β (Figure 5A) increases expression of CAT activity from −155shhCAT similarly to Axl (Figure 5B). Thus, Axl and HNF3β appear also to be functionally equivalent and are likely to recognize similar or identical DNA-binding sequences. Figure 5.Rat HNF3β induces CAT expression from −155shhCAT in a manner similar to zebrafish Axl. Increasing amounts of CMVHNF3β (A) or pCMVaxl (B) were transfected into HeLa cells together with −155shhCAT. CAT enzymatic activity was determined by assaying 400 μg of extract protein. The autoradiographs were exposed for 24 h. Download figure Download PowerPoint The sequence of the promoter region (Figure 3) was searched for homologies to the HNF3β consensus sequence [5′-(G/A) (T/C) (C/A) A A (C/T) A-3′; (Overdier et al., 1994; Roux et al., 1995; Kaufmann and Knöchel, 1996)]. Two HNF3β-binding site homologies (HNF3β1 and 2) reside within the 155 bp upstream sequence mediating transactivation by Axl and HNF3β (Figure 3). Two closely spaced HNF3β-binding sites are required for activation of the shh promoter To test whether the putative HNF3β-binding sequences may be relevant to activation of the shh promoter, we first investigated whether they bind Axl and HNF3β protein by gel retardation analysis. An Axl–GST fusion protein encoding full-length Axl and a HNF3β–GST fusion protein, which represents a truncated version of rat HNF3β but which shows DNA-binding specificity identical to full-length protein (Overdier et al., 1994), were expressed in Escherichia coli and partially purified. Recombinant Axl–GST (Figure 6A, left panel) or HNF3β–GST (Figure 6B, left panel) were incubated with oligonucleotides covering the binding site homologies HNF3β1 and 2 (Figure 6, Oligo 1 and 2, respectively) or with an oligonucleotide (Figure 6, Oligo 3) containing the sequence from −143 to −112 of the shh upstream region. Both recombinant proteins specifically retarded oligonucleotides containing elements HNF3β1 and HNF3β2 (Figure 6A and B, left panels). Oligonucleotide 3 did not bind either fusion protein. Mutation of the HNF3β consensus sequences (Oligo 1′ and 2′ respectively; see Materials and methods) by a series of base changes abolishes binding of Axl–GST (Figure 6A, right panel) and HNF3β–GST (Figure 6B, right panel). Oligonucleotides containing HNF3β1 and HNF3β2 sites are bound specifically by proteins present in whole cell extract prepared from 24 h old zebrafish embryos (Figure 6C, left panel). Formation of a specific complex was abolished when HNF3β1 and HNF3β2 sites were mutated (Figure 6C, right panel), suggesting that proteins are present in whole cell extract that interact with the same sequence as Axl and HNF3β. Figure 6.Axl and HNF3β bind to shh promoter sequences. Gel retardation analysis was carried out with partially purified zebrafish Axl–GST (A), rat HNF3β–GST (B) or whole cell extract from 24 h old zebrafish embryos (C). Oligonucleotides 1 and 2 containing HNF3β consensus elements HNF3β1 (from −103 to −74) and HNF3β2 (from −126 to −97) formed specific complexes (S, arrow) while oligonucleotide 3 (from −143 to −112) was not shifted. Retarded bands were abolished by addition of a 100-fold excess of cold competitor oligonucleotide (+ lanes). Oligonucleotides containing a cluster of point mutations in the HNF3β1 (Oligo 1′) or HNF3β2 (Oligo 2′) sites did not form specific complexes. S, specific complex; NS, complex due to non-specific binding; P, unbound oligonucleotide; Oligo, oligonucleotides. Download figure Download PowerPoint Thus, elements HNF3β1 and 2 may be important for activation of the shh promoter by Axl and HNF3β. To assess this hypothesis, element HNF3β1 or element HNF3β2 on their own or in combination were mutated by the same cluster of nucleotide changes in the centre of each element which abolished protein binding in vitro. Mutation of either HNF3β1 (−563shhCAT-M1) or HNF3β2 (−563shhCAT-M2) led to a decrease in CAT activity by ∼50%, relative to −563shhCAT when transfected into HeLa cells together with pCMVaxl (Figure 7A). −563shhCAT-DM carrying mutations in both elements shows a 5-fold reduction of CAT activity in comparison with the wild-type construct −563shhCAT (Figure 7A). Identical results were obtained in transfection with CMVHNF3β; CAT activity was moderately decreased by each single mutation, while mutation of both elements led to a strong reduction of CAT activity (Figure 7B). The basal level of expression in the absence of Axl or HNF3β was not affected by the mutations. In summary, these data show that elements HNF3β1 and 2 mediate transactivation of the shh promoter by Axl and HNF3β. As a weak response to co-transfection of either pCMVaxl or CMVHNF3β remains in the double mutant, however, other sequences may marginally contribute to transactivation. Figure 7.Mutation of HNF3β-binding sites abolishes transactivation of the shh promoter by Axl and HNF3β. −563shhCAT fusion constructs in which the HNF3β-binding sites HNF3β1 (M1) or HNF3β2 (M2) alone or both together (DM) had been mutated were transfected into HeLa cells with either pCMVaxl (A, Axl) or with CMVHNF3β (B, HNF). In controls (C) pCMVaxl or CMVHNF3β were replaced by CMVβ. Bars represent the average of the results from three transfections carried out in parallel. CAT activity obtained from the wild-type promoter construct (−563shhCAT) was set to 100%, and results from mutant constructs were normalized accordingly. Standard deviations are indicated. Download figure Download PowerPoint Retinoic acid receptors activate the shh promoter In zebrafish embryos treated with retinoic acid (RA), shh expression is ectopically activated in the anterior fin buds and its expression in the neural tube of the tail is expanded (Akimenko and Ekker, 1995). Implantation of beads soaked in RA into the anterior limb bud of chicken embryos leads similarly to ectopic activation of shh expression (Riddle et al., 1993), suggesting that shh may be regulated by RARs. We noted a DR5-type retinoic acid response element (Figure 3; RARE, GGTTCATTAGCGGGTCA) between position −242 and −258 in the shh upstream region. To test whether this element confers RA inducibility to the shh promoter, we first tested whether RARs can bind to the element by gel retardation experiments. RARs bind to DR5-type target sequences as heterodimers between RARs and RXRs (Yu et al., 1991; Berrodin et al., 1992; Bugge et al., 1992; Kliewer et al., 1992; Marks et al., 1992; Zhang et al., 1992). An oligonucleotide harbouring the shh DR5 element was specifically retarded when incubated with a mixture of recombinant mouse RARα and RXRα (Figure 8A, Oligo 2, left panel). Oligonucleotides containing a single repeat of the hexamers AGGTCA or TGTTCA, found at position +36 and −478 in the shh promoter upstream region, respectively, did not show retardation when incubated with RARα/RXRα (Figure 8A, Oligo 1 and 3). Mutation of the RAR-binding site in oligonucleotide 2 prevented complex formation (Figure 8A, left panel, Oligo 2′). Oligonucleotide 2 is specifically bound by proteins present in whole cell extract prepared from 24 h old zebrafish embryos (Figure 8B, left panel) and binding of extract protein is abolished when the DR5 element is mutated (Figure 8B, right panel, Oligo 2′). Figure 8.RAR/RXR complexes bind to the DR5 RARE in the shh promoter. Oligonucleotides were incubated with purified recombinant mouse RARα /RXRα or whole cell extract prepared from 24 h old zebrafish embryos. Oligonucleotide 2 (from −265 to −236) harbouring the DR5 RARE of the shh promoter region showed specific complex formation (S, arrow) which was competed by a 100-fold excess of cold oligonucleotide 2 (+ lanes). Oligonucleotides 1 (from +31 to +60) and 3 (from −490 to −461) which contain an imperfect single repeat of the hexamer motif characteristic of RAREs did not form complexes with RARα/RXRα or proteins present in whole cell extract. Mutation of the RARE sequences by a cluster of point mutations (Oligo 2′) abolished specific complex formation of Oligo 2 with recombinant proteins (A, right panel) and proteins from whole cell extract (B, right panel). S, specific complex; NS, complex due to non-specific binding; P, unbound oligonucleotide; Oligo, oligonucleotide. Download figure Download PowerPoint To test whether the zebrafish shh promoter is inducible by RA, the −563shhCAT expression plasmid was co-transfected into HeLa cells with plasmid RARβ encoding human RARβ2 (Zelent et al., 1989). The HeLa cells employed express RXRs endogenously (Leid et al., 1992). Transfection of plasmid RARβ and administration of all-trans retinoic acid (t-RA, 10−7M) resulted reproducibly in a 2.5- to 4-fold increase in CAT activity in comparison with controls exposed to ethanol vehicle alone (Figure 9). This effect is dependent on co-transfection of plasmid RARβ as t-RA did not cause an increase in CAT activity in the absence of co-transfected RARβ. Administration of 9-cis RA (10−7M) that activates RXR did not induce an increase in CAT activity, suggesting that endogenous RXRs on their own cannot mediate the effect. Also, co-transfection of plasmid RXRα encoding human RXRα (Elder et al., 1992) did not result in an increase in CAT activity in the presence of 9-cis RA (Figure 9). Figure 9.The DR5 RARE renders the shh promoter inducible by retinoic acid. The wild-type −563shhCAT or a mutant derivative −563shhCAT-DR5-M, in which the DR5 RARE at position −258/−242 had been mutated by a cluster of nucleotide changes, were transfected into HeLa cells in combination with either plasmid RARβ, RXRα or pSG5 without receptor cDNA inserted. Cells were treated with either 10−7 M all-trans retinoic acid (t-RA), 10−7 M 9-cis retinoic acid (9-cis) or ethanol vehicle alone (−). Bars indicate the percentage conversion of chloramphenicol to its acetylated forms with 400 μg of protein extracted from cells 48 h after transfection. Bars represent the mean of three independent transfections carried out in parallel. Standard deviations are indicated. The experiment shown is representative of three experiments carried out independently. Download figure Download PowerPoint To test whether the induction of CAT activity by RARβ in the presence of t-RA is mediated by the DR5 RARE located at position −258/−242 of the shh upstream sequence, the element was mutated in −563shhCAT by the same series of nucleotide exchanges which abolished protein binding in vitro. The resulting plasmid −563shhCAT-DR5-M did not show a response to co-transfection of plasmid RARβ and administration of t-RA (Figure 9). In summary, these data suggest that the shh promoter can be regulated by RARs in response to t-RA. Discussion We present evidence that shh is responsive to misexpression of axl in a wide variety of distinct cell types, suggesting that axl is a crucial determinant of shh expression. By analysis of the shh promoter in HeLa cells, two Axl/HNF3β-binding sites were found to be required for transactivation of the shh promoter, strongly suggesting that Axl/HNF3β is a direct regulator of the shh gene. We furthermore identified a RARE in the shh promoter which is functional in the HeLa expression system, indicating that the shh promoter is under direct transcriptional control by RARs. Despite the fact that direct proof by mutation is lacking in the zebrafish embryo, several lines of evidence strongly argue in favour of the notion that axl indeed controls shh expression in vivo during development of the axis in the zebrafish embryo. First, axl expression precedes that of shh in the embryonic shield, and the two genes have overlapping expression patterns (Krauss et al., 1993; Strähle et al., 1993, 1996). Second, ectopic expression of axl leads to strong activation of shh in cells that normally do not express shh. Third, shh and axl expression are affected similarly in midline-defective mutant embryos (Krauss et al., 1993; Strähle et al., 1993, 1996; Talbot et al., 1995). In the mouse embryo, the effects of targeted mutation of HNF3β are in agreement with our findings in the zebrafish embryo. HNF3β−/− mouse embryos show loss of the notochord, lack of shh expression and impaired patterning of the ventral neural tube (Ang and Rossant, 1994; Weinstein et al., 1994). It is, however, not clear from these studies in the mouse whether loss of shh expression and disturbed neural patterning is a direct effect of the mutation in HNF3β or whether it is an indirect consequence of impai
Referência(s)