| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 282, Issue 18, 13290-13302, May 4, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
3 Nicotinic Receptor Subunit Gene Promoter*
¶1¶12¶¶¶¶3
From the
Department of Pharmacology, School of Medicine, University of Milan, 20129 Milan, CNR-Institute of Neuroscience, Cellular and Molecular Pharmacology Section, 20129 Milan, ¶Center of Excellence on Neurodegenerative Diseases, University of Milan, 20129 Milan, and ||CNR-Institute of Chemistry and Molecular Recognition, 20131 Milan, Italy
Received for publication, September 6, 2006 , and in revised form, January 24, 2007.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-hydroxylase. The results of this study show that the human orthologue of PHOX2A is also capable of regulating the transcription of the human 
3 nicotinic acetylcholine receptor gene, which encodes the ligand-binding subunit of the ganglionic type nicotinic receptor. In particular, we demonstrated by chromatin immunoprecipitation and DNA pulldown assays that PHOX2A assembles on the SacI-NcoI region of 3 promoter and, by co-transfection experiments, that it exerts its transcriptional effects by acting through the 60-bp minimal promoter. PHOX2A does not seem to bind to DNA directly, and its DNA binding domain seems to be partially dispensable for the regulation of 3 gene transcription. However, as suggested by the findings of our co-immunoprecipitation assays, it may establish direct or indirect protein-protein interactions with Sp1, thus regulating the expression of 3 through a DNA-independent mechanism. As the 3 subunit is expressed in every terminally differentiated ganglionic cell, this is the first example of a "pan-autonomic" gene whose expression is regulated by PHOX2 proteins. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
PHOX2A and PHOX2B are paired-like homeodomain proteins that have been shown to play a pivotal role in the development of the three peripheral divisions of the autonomic nervous system (1). They are also expressed in all of the noradrenergic neurons of the brainstem, in some cranial sensory ganglia that participate in autonomic reflexes, and in a subset of cranial motor neurons (2). None of the components of the autonomic nervous system develop properly in Phox2b knock-out mice (3), whereas Phox2a null mutants show an apparently less severe phenotype that only involves the agenesis of the Locus coeruleus and atrophy of the cranial sensory ganglia (4); nevertheless, they do not feed and die on the day of birth. The different phenotypes of Phox2 knock-out mutants along with their asynchronous onset of expression during development underscore that the two factors are not functionally equivalent. This has been more directly demonstrated by reciprocal gene replacement experiments (5) that led to the conclusion that biochemical differences may be responsible for the specific function of each paralogue.
It is worth noting that PHOX2 proteins are also involved in the transcriptional control of the neurotransmitter phenotype (6); they play a fundamental role in the terminal differentiation of the orthosympathetic system as they regulate the gene expression of tyrosine hydroxylase and dopamine--hydroxylase (DBH),4 two limiting enzymes in catecholamine synthesis. In particular, it has been shown that PHOX2A can synergize with dHAND in the regulation of the DBH promoter (7, 8). Because of their function in the development of the entire autonomic nervous system, it might be expected that PHOX2 proteins also control the expression of specific genes of the parasympathetic and enteric branches or pan-autonomic genes common to all three autonomic divisions. As no information is yet available concerning this fundamental issue, we tested this hypothesis using the human 3 nicotinic receptor subunit gene as a possible target of PHOX2A regulation.
All post-ganglionic neurons respond to preganglionic stimulation by a fast excitatory post-synaptic potential, which triggers the initiation of the post-synaptic spike. The fast excitatory post-synaptic potential is because of the activation of neuronal nicotinic acetylcholine receptors (nAChRs) whose pharmacological blockade abolishes all autonomic nervous system activity (9).
Neuronal nAChRs form a family of acetylcholine-gated cation channels that are expressed in the autonomic and sensory ganglia, the adrenal medulla, and distinct areas of the central nervous system and have a quaternary structure consisting of five transmembrane subunits assembled around a central channel. Twelve distinct neuronal nicotinic subunits have been cloned so far and classified into two subfamilies of nine subunits (2–10) and three subunits (2–4) (10). The 3 subunit appears precociously during neural crest cell differentiation (11), and its expression becomes very abundant in all of the ganglionic neurons of the autonomic nervous system, where it assembles with 4 or 2 (and, in some receptor molecules, also with 5) to form the ganglionic type nAChR.
Many knock-out mice lacking the 3 subunit succumb early after birth, and most of the remaining die during the next 6–8 weeks because of the severe abnormalities of autonomic functions (12). Thus, the 3 subunit is a very typical pan-autonomic gene responsible for relevant functions of terminally differentiated ganglionic neurons.
The aim of this study was to investigate whether PHOX2A regulates the expression of the human 3 nAChR subunit gene, and the possible molecular mechanisms underlying this process.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Plasmid Construction—The 4.3 KN, the 1 KN, and the BglII-NcoI human 3 promoter reporter plasmids have been described previously by Fornasari et al. (13). The BsiWI-AccIII (60) 3 minimal promoter construct as well as the plasmids, BglII-NcoI mutA, BglII-NcoI mutB, and BglII-NcoI mutAB bearing mutations either in the A or B or both Sp1-binding sites have been described by Terzano et al. (14). The SacI-NcoI construct was obtained by digesting 0.35 BN plasmid (13) with SacI, followed by re-ligation. The 5 –240/+180 construct has been described by Flora et al. (15). The human PHOX2A cDNA was cloned by means of reverse transcription-PCR, using RNA purified from SH-SY5Y cells; polymerase Superscript II (Invitrogen) was used for reverse transcription, and Pfx proofreading polymerase (Invitrogen) was used for amplification. The oligonucleotides were designed on the basis of the sequence deposited at the NCBI with accession number NM_005169
[GenBank]
; the upper primer corresponded to nucleotides 1–25 (5'-CTT GCG TTG CAC CCG GGC TGA GTG C-3'); the lower primer was complementary to nucleotides 997–1027 (5'-GAT TGG TCT TCA GGG CGG GGC CGG GCT TC-3'). The PCR product was subcloned in the pGEM-T Easy vector (Promega), and completely sequenced on both strands. The cDNA was excised by means of EcoRI digestion and cloned into pcDNA3.0 vector (Invitrogen), and the construct was named PHOX2A/pcDNA3.
Recombinant PCR was used to introduce point mutations into the homeodomain of PHOX2A (16). The oligonucleotides used to generate the mutations were Ph2aR53W-UP, 5'-CCA GAA CCG CTG GGC CAA GTT CC-3'; Ph2aR53W-LOW, 5'-GGA ACT TGG CCC AGC GGT TCT GG-3'; Ph2aN51H-UP, 5'-TTC CAG CAC CGC CGG GCC AAG-3'; Ph2aN51H-LOW, 5'-CCC GGC GGT GCT GGA ACC AG-3'; the external primers were N-ATG-UP, 5'-GGG CCG ATG GAC TAC TCC TAC C-3' and Ph2aCT-LOW, 5'-TTG CGG CCG CCT AGA AGA GAT TGG TCT TCA GGG C-3' (the changed codons are in boldface and underlined). The resulting DNA was subcloned into the pGEM-T easy vector (Promega) and sequenced on both strands. The cDNA was excised by means of EcoRI digestion and cloned into pcDNA3.0 vector (Invitrogen), and the constructs were named PHOX2A-R53W/pcDNA3 and PHOX2A-N51H/pcDNA3.
The 4xTK-luc construct was obtained by cloning into the pGL3-basic vector a synthetic double-stranded oligonucleotide (see Fig. 6A for the sequence) containing four copies (three in the right orientation and one in the opposite orientation) of domain II of the DBH promoter, bearing a well described PHOX2A-binding site (17) upstream of the TK promoter.
Transfections and Luciferase Assays—The transfection and co-transfection experiments were performed by means of lipofection, as described by Flora et al. (18), using 4 x 104 HeLa cells, 5 x 104 DAOY cells, or 2 x 105 IMR32 cells. The luciferase assays were carried out using the dual luciferase reporter assay system (Promega, Madison, WI) as described previously (18, 19). All of the transfections were performed in duplicate, and each construct was tested in at least three independent experiments using different batches of the plasmid preparation.
Preparation of Synthetic Peptide and Antibody Production—A peptide corresponding to the human PHOX2A sequence, CKPGPALKTNLF, located in the C terminus of the protein was synthesized as described in Cargnin et al. (20). The affinity-purified polyclonal antibodies from chicken egg yolk were produced by Davids Biotechnologie (Ragensburg, Germany).
In Vitro Protein Expression—The in vitro expression of PHOX2A-wild type and PHOX2A-R53W was obtained by means of a commercial rabbit reticulocyte lysate system (TNT quick-coupled transcription/translation system, Promega, Madison, WI) as described previously (20).
Protein Preparation and Western Blot Analyses—Ten micrograms of nuclear extract, prepared as described in Terzano et al. (14), or 2 µl of in vitro expressed proteins were separated by means of 10% SDS-PAGE and transferred to nitrocellulose membranes (Schleicher & Schuell). The membranes were preincubated for 1 h with blocking buffer (5% nonfat dry milk, 20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.1% Tween 20), after which the primary antibodies were added at an appropriate dilution and incubated for 2 h before the secondary antibodies conjugated with horseradish peroxidase (Davids Biotechnologie) were added and incubated for 1 h. After appropriate washes, the bands were revealed using Super Signal West Dura (Pierce). Standard molecular weights (New England Biolabs, Inc., Beverly, MA) were loaded in parallel, and the relative mass of PHOX2A was calculated by means of computer-assisted analysis. Competition experiments were carried out using PHOX2A peptide in a 10:1 mass ratio with the corresponding antibodies.
Immunofluorescence—DAOY cells plated on 1.7 x 1.7-cm2 glass coverslips were grown to 50% confluency and transfected with PHOX2A/pcDNA3, PHOX2A-R53W/pcDNA3, and PHOX2A-N51H/pcDNA3 plasmids or mock-transfected with the empty vector. Immunofluorescence was performed as described in Cargnin et al. (20). The primary antibody was used at a dilution of 1:400 and revealed by means of an anti-chicken Texas Red-conjugated secondary antibody (Jackson ImmunoResearch, West Grove, PA). Competition assays using the specific peptide were performed as described above.
Electrophoretic Mobility Shift Assays (EMSAs)—The EMSAs were performed as described by Terzano et al. (14) and Cargnin et al. (20). The oligonucleotide PRS1 has been described by Kim et al. (17). All of the oligonucleotides were purchased from Invitrogen.
Chromatin Immunoprecipitation Assays (ChIP)—Chromatin immunoprecipitation was carried out as described previously (20). Chromatin was incubated overnight at 4 °C with 5 µgof each antibody (anti-PHOX2A (Davids Biotechnologie, Germany), anti-human Sp1, and anti-acetylated histone H4 (Upstate, Charlottesville, VA), and chicken preimmune IgY (Davids Biotechnologie)), and the immunocomplexes were collected on monoclonal anti-chicken IgY-agarose beads or protein G/agarose bead slurry (Invitrogen), preadsorbed with 20 µg/µl tRNA and 10 µg/µl salmon sperm DNA (Sigma). After washings and elution, the cross-linking was reversed by heating to 65 °C overnight, and the samples were purified on columns (High Pure PCR product purification kit, Roche Diagnostics). For the PCR detection of the immunoprecipitated chromatin, 5% of the purified DNA was used as a template with the following primers: ChIP[3prom]UP, 5'-CTC CTT CCT GGT GGT GGT GAC-3', and ChIP[3prom]LOW, 5'-GGG CTC CTC TCC GCT TGC-3' to amplify the –386/–127 region of the nAChR 3 subunit promoter (numbering according to Ref. 14); and ChIP[5prom]UP, 5'-CTC TGC TCC AGG GTC GCA C -3', and ChIP[5prom]LOW, 5'-GAG TGT GAG TCG TGA GAC AAA ACG-3', to amplify the 5 promoter. The DNA samples were heated to 95 °C for 2 min, followed by 47 cycles of heating at 95 °C for 30 s, annealing at 64 (3 promoter) or 61 °C (5 promoter) for 30 s, and extension at 72 °C for 30 s. Transient chromatin immunoprecipitation assays were carried out as described in Wells and Farnham (21). Briefly, IMR32 cells were transfected with 5 µg of different 3 promoter plasmids (BglII-NcoI, BglII-NcoI mutA, BglII-NcoI mutB, or BglII-NcoI mutAB constructs). After 24 h, chromatin was prepared for immunoprecipitation as described above. For the PCR detection of the immunoprecipitated chromatin, specific primers designed on plasmid backbone sequences were used as follows: GL2 primer, 5'-CTT TAT GTT TTT GGC GTC TTC C-3', and RV3 primer, 5'-CTA GCA AAA TAG GCT GTC CC-3'. The DNA samples were heated to 95 °C for 3 min, followed by 38 cycles of heating at 95 °C for 45 s, annealing at 60 °C for 30 s, and extension at 72 °C for 30 s.
DNA Pulldown Assays—Five picomoles of the SacI-NcoI or –240/+180 DNA fragments of the 3 and 5 regulatory regions were biotinylated by filling in and incubated with streptavidin-coated magnetic beads (Roche Diagnostics); 250 µg of IMR32 nuclear extracts were added and incubated on ice for 40 min in binding buffer (4% Ficoll, 20 mM Hepes, pH 7.9, 1 mM MgCl2, 0.5 mM dithiothreitol), 100 mM final salt concentration (NaCl + KCl), and 25 µg of poly[d(I-C)] as the competitor. After magnetic separation, the beads were extensively washed with a binding buffer containing 25% glycerol instead of Ficoll and resuspended in protein loading buffer. The samples were analyzed by Western blotting as described previously.
In Vitro DNase I Footprinting Assays—The 376-bp probe, containing the SacI-NcoI region of the 3 promoter, was obtained as described in Terzano et al. (14). To generate the probe corresponding to the multimerized domain II of the DBH promoter, we subcloned the double-stranded oligonucleotide shown in Fig. 6A into the KpnI-BglII sites of pSP73 vector (Promega). The plasmid was then digested with HindIII to label the bottom strand.
In each footprinting reaction, 2 fmol of probes (corresponding to 20,000–30,000 cpm) were incubated with 50 µgof nuclear extract, prepared as described previously (14), and the reactions were performed using a previously described procedure (14). DNase I (DNase I-RNase free; Roche Diagnostics) was used at concentrations of 0.05 units/µg DNA without extract and 1.0 unit/µg DNA in the presence of extract, and the samples were processed as described previously (14).
Co-immunoprecipitation Assays—IMR32 cells were grown to 70% confluency and harvested after 24 h, when the nuclear extract was prepared for immunoprecipitation. The nuclei were obtained as described previously (14). Nuclear proteins were extracted in lysis buffer containing non-ionic detergent (0.5% Triton X-100, 50 mM Tris-HCl, pH 8.0, 100 mM NaCl, 0.5 mM EDTA, 10 mM imidazole, 10% glycerol, 0.5 mM dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride, Sigma protease inhibitors mixture) and precleared using protein G/agarose bead slurry (Invitrogen) and rabbit preimmune IgG or chicken preimmune IgY (Santa Cruz Biotechnology). The precleared extracts were incubated overnight at 4 °C with 5 µg of each primary antibody (polyclonal chicken anti-PHOX2A antibody (Davids Biotechnologie), polyclonal rabbit anti-Sp1 antibody (Santa Cruz Biotechnology), and preimmune rabbit IgG or preimmune chicken IgY (Santa Cruz Biotechnology)), and the immunocomplexes were captured by protein G/agarose bead slurry (Invitrogen). Because of the poor binding of chicken antibodies to protein G, a bridging antibody (rabbit anti-chicken IgG, Upstate) was added to enhance immunocomplexes capture.
The beads were collected by centrifugation and gently washed and resuspended in sample loading buffer. The immunocomplexes were dissociated from the beads by boiling the samples, then separated by SDS-PAGE, and transferred on nitrocellulose membrane. Western blotting was performed as described previously using the primary anti-PHOX2A, anti-Sp1, and anti-CREB-1 antibodies (Santa Cruz Biotechnology).
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
3 Promoter—To evaluate whether PHOX2A plays a role in 3 expression, we performed co-transfection experiments in which a vector expressing PHOX2A was introduced into DAOY and HeLa cells (not shown), together with different reporter constructs bearing the luciferase reporter gene under the control of different parts of the 35'-regulatory region or different promoters used as positive and negative controls (Fig. 1A). The 4.3 KN construct containing the whole intergenic region between 4 and 3 (13, 14, 22) was stimulated by 
3-fold (Fig. 1B), to the same extent as the 4xTK plasmid (Fig. 1B), and in a range of transactivation that was comparable with that observed using the native DBH promoter (23). Progressive deletions at the 5' end of the 3 regulatory region caused fluctuations in the response to PHOX2A (Fig. 1B, see constructs 1 KN and SacI-NcoI) but did not prevent the transcription factor from exerting its regulatory functions. Furthermore and to our surprise, the BsiWI-AccIII construct containing the 60-bp 3 minimal promoter was the most responsive to PHOX2A (Fig. 1B).
As expected, the SV40 promoter was not transactivated by PHOX2A, thus indicating the specific nature of the transcriptional effects on the 3 promoter (Fig. 1B), and the 5 promoter construct (corresponding to the –1011/+180 plasmid described in Ref. 15) showed a negligible and not statistically significant transcriptional response to PHOX2A (Fig. 1B). This lack of effect paralleled our previous finding that PHOX2B is unable to transactivate this nAChR subunit gene (18). As these data suggested that PHOX2A could regulate 3 transcription by acting through its minimal gene promoter, we decided to carry out biochemical experiments to ascertain whether PHOX2A could actually assemble on the 3 promoter.
Characterization of a Polyclonal Antibody Directed against the Human PHOX2A Protein—To follow the expression of the human PHOX2A protein, we produced an anti-peptide polyclonal antibody in chicken eggs and then verified its specificity by means of various criteria and techniques. We first made a Western blot analysis using nuclear extracts prepared from IMR32 or HeLa cells in parallel with the PHOX2A protein obtained by means of in vitro transcription/translation. The in vitro expressed PHOX2A protein was recognized by the antibody as a doublet with apparent molecular masses of 38.1 and 35.7 kDa, whereas no signal was detectable using the unprogrammed lysate (Fig. 2A, lanes 1 and 3). The same doublet was also detected in the IMR32 and SH-SY5Y nuclear extracts (Fig. 2, A, lane 2, and B, lanes 1 and 2), whereas no band was observed in the HeLa cells, as expected (Fig. 2, A, lane 4, and B, lane 3). To corroborate these data, we carried out competition experiments involving the addition of an excess of the PHOX2A peptide used for immunization. As shown in Fig. 2A, the specific PHOX2A bands disappeared from the lanes containing the IMR32 nuclear extracts or the in vitro expressed PHOX2A protein (lanes 5 and 6). The antibody was also tested against nuclear extracts prepared from HeLa cells transfected with an expression vector bearing mouse (Fig. 2B, lane 5) or human PHOX2A cDNA (Fig. 2B, lane 6); in both cases, the antibody recognized two bands with apparent molecular masses of 35.7 and 38.1 kDa, whereas no bands were detectable in the mock-transfected HeLa cells (Fig. 2B, lane 4). The difference in the apparent molecular weights of the two forms of PHOX2A is compatible with the alternative use of two in-frame methionine residues surrounded by perfect Kozak sequences (24), which are conserved in both human and mouse orthologues (18, 25). Immunofluorescence was used to confirm that the examined antibody recognized PHOX2A in its native state. DAOY cells (which do not endogenously express the antigen) were transfected with PHOX2A/pcDNA3 and probed with the antibody. Fig. 2C shows the nuclei of transfected DAOY cells stained with DAPI (panel c) or revealed by the anti-PHOX2A antibody (panel d). The antibody labeled the nuclei of a few cells, as may be expected in a transient transfection experiment, but did not produce any signal when mock-transfected cells were probed (Fig. 2C, panels a and b). Similarly, no signal was detectable in the PHOX2A/pcDNA3-transfected cells if the antibody was previously incubated with an excess of the PHOX2A peptide (Fig. 2C, panels e and f).
It is well known that DBH is a PHOX2A target gene, and so we used the PRS1 oligonucleotide to verify whether the examined antibody was capable of supershifting complexes containing PHOX2A. PRS1 corresponds to a region of 22 nucleotides within domain IV of the DBH promoter (26), contains two ATTA core motifs for homeodomain proteins, and has been shown to bind PHOX2A (17, 26). Fig. 2D shows that a prominent retarded band formed in the presence of the IMR32 nuclear extract and that its formation was prevented by an excess of cold oligonucleotide (lanes 2 and 3). When the anti-PHOX2A antibody was added to the reaction, a supershifted complex was detected (Fig. 2D, lane 4). To confirm these data, we carried out EMSAs using the in vitro expressed PHOX2A protein, and we observed a specific retarded band, which was competed by an excess of cold oligonucleotide and supershifted by the anti-PHOX2A antibody (Fig. 2D, lanes 5–7). Although the sizes of the ultra-retarded bands were similar regardless of the source of the PHOX2A protein (Fig. 2D, compare lanes 4 and 7), the retarded bands obtained using the in vitro expressed protein migrated slightly faster (Fig. 2D, compare lanes 2 and 4 with lanes 5 and 7). This may have been due to an additional factor (not present in the lysate) that is capable of binding the PRS1 oligonucleotide, or it may reflect the presence of posttranslational modifications that could obviously only affect the native proteins obtained from the nuclear extracts. More interestingly, the supershifted bands seemed to be more intense than the corresponding retarded bands obtained in the absence of the antibody (Fig. 2D, compare lane 2 with lane 4 and lane 5 with lane 7), and the fact that this was constantly observed regardless of the source of the PHOX2A protein (also see Figs. 5A and 7C) suggests that our antibody could stabilize the interactions between the transcription factor and its cognate DNA-binding sites. A very similar result has also been previously obtained by Kim et al. (17), who used the same oligonucleotide and a nuclear extract from a human neuroblastoma cell line but a different antibody. No specific band was detectable in the unprogrammed lysate, regardless of the presence of the antibody, which alone did not have any effect (Fig. 2D, lanes 8–10).
|
3 Promoter by Means of ChIP Assays—ChIP assays were performed to verify whether PHOX2A could bind the 3 nAChR subunit gene promoter in vivo. The previously characterized anti-PHOX2A antibodies were used to immunoprecipitate chromatin from IMR32 cells, and the associated DNA fragments were amplified using primers flanking the BsiWI-AccIII region containing the human 3 minimal promoter (14). An expected band of 280 bp was observed, thus suggesting that PHOX2A assembles on the 3 promoter in vivo (Fig. 3A, lane 2).
Bands of the same size were also observed when we used antibodies directed against the acetylated form of histone H4 (Fig. 3A, lane 4), a well known marker of transcriptionally active promoters, and Sp1 (Fig. 3A, lane 5), a ubiquitous transcription factor involved in 3 expression (14, 27). No amplification products were observed using preimmune IgY immunoglobulins (Fig. 3A, lane 3).
We also tested the specificity of the bands obtained using the anti-PHOX2A and -Sp1 antibodies using chromatin prepared from HeLa cells, which do not express PHOX2A. No signals were detected with these antibodies or the anti-acetylated form of histone H4 antibody (Fig. 3A, lanes 7, 9, and 10).
Autonomic ganglia (and therefore human neuroblastoma cells (15)) also express the 5 nAChR subunit, which can assemble with the 3 and 4 subunits to form a specific subtype of ganglionic nicotinic receptor (12) and is located with these two genes in the same genomic cluster (22). However, unlike many nicotinic subunits, the expression of 5 is not restricted to the nervous system, but it also involves several non-neuronal tissues and cell lines (15). To verify whether PHOX2A could bind to the 5 promoter, we carried out ChIP assays using chromatin obtained from IMR32 and HeLa cells. Both cell lines gave the expected bands of 370 bp when we used antibodies directed against the acetylated form of histone H4 (Fig. 3B, lanes 4 and 9), or Sp1 (Fig. 3B, lanes 5 and 10), in agreement with the role of this transcription factor in the expression of 5 (28). On the contrary, no band was detectable when the chromatin was immunoprecipitated by the anti-PHOX2A antibody (Fig. 3B, lanes 2 and 7). This finding in relation to HeLa cells is in line with the notion that PHOX2A is not expressed in non-neuronal tissues, but the lack of any band with chromatin from IMR32 clearly confirmed that PHOX2A specifically regulates the 3 but not the 5 promoter, as already suggested by our previous data obtained from co-transfection experiments.
|
3 Promoter by Means of DNA Pulldown Assays—To confirm the ChIP data by means of an alternative and independent experimental approach, we set up a DNA pulldown assay in which the DNA regions corresponding to the SacI-NcoI 3 promoter and the –240/+180 5 promoter were biotinylated at the 5' end and incubated with IMR32 nuclear extracts. After incubation, the DNA-protein complexes were purified by means of magnetic separation with streptavidin-coated beads and loaded onto a SDS-PAGE, after which the proteins were then transferred to nitrocellulose and probed with anti-PHOX2A or anti-Sp1 antibodies. When the anti-PHOX2A antibody was used, the lane containing the nuclear proteins purified with the probe corresponding to the 3 regulatory region (Fig. 4A, lane 3) showed a major band with an apparent molecular mass of 38 kDa and a less intense band of 36 kDa, but no signal was detected with the nuclear proteins purified with the 5 probe (Fig. 4A, lane 2).
As Sp1 is a ubiquitous transcription factor that plays an essential role in the activity of both 3 and 5 promoters (14, 15, 28),5 we expected to detect it in the nuclear protein preparations obtained using both probes. We therefore probed the preparations shown in Fig. 4A with an antibody directed against Sp1, and we detected an intense band with the expected molecular mass of 106 kDa in both (Fig. 4B, lanes 2 and 3). When the streptavidin-coated beads were not previously conjugated with 3 or 5 DNA, but directly incubated with the IMR32 nuclear extract, very faint signals were observed with both PHOX2A and Sp1 antibodies (Fig. 4, A and B, lanes 1), thus indicating negligible nonspecific binding of the nuclear proteins to the beads. These data fully confirmed that PHOX2A specifically bound the 3 but not the 5 promoter.
|
3 Minimal Promoter and PHOX2A—Co-transfection experiments showed that the BsiWI-AccIII construct was efficiently transactivated by PHOX2A, thus strongly suggesting that it should directly act on the 60-bp minimal promoter. However, as no consensus sequence for homeodomain proteins was identified in the 60-bp DNA region, it seemed that PHOX2A could bind to a non-canonical site. To study the DNA-protein interactions between the 3 minimal promoter and PHOX2A, we carried out EMSA and DNase I footprinting experiments. For the EMSA and supershift experiments, we used the BsiWI-AccIII region and PRS1 oligonucleotide as probes. When the nuclear extract from IMR32 cells was incubated with a labeled PRS1 probe, two retarded bands formed that could be competed by a molar excess of cold oligonucleotide (Fig. 5A, lanes 3 and 9). In the presence of the anti-PHOX2A antibody, a very strong ultra-retarded band appeared (Fig. 5A, lane 4), the formation of which was not because of a nonspecific interaction between the probe and the antibody (Fig. 5A, lane 2). As discussed previously, we interpreted this as indicating that the antibodies directed against the C terminus of PHOX2A strongly stabilize its binding to DNA.
|
|
|
3 promoter as a probe. Fig. 6C shows the typical protection pattern produced by nuclear extracts obtained from neuroblastoma cells, i.e. two distinct protected areas (F1 and F2) encompassing the BsiWI-AccIII minimal promoter (Fig. 6C, compare lanes 1 and 3), as described previously (14). However, unlike the multimerized domain II of the DBH promoter, the digestion pattern was not modified by the anti-PHOX2A antibody (Fig. 6C, compare lanes 3 and 4). The digestion pattern of the SacI-NcoI probe was slightly different in DAOY cells, in which the F2 protection was virtually absent (Fig. 6C, lanes 5–8). As the formation of F2 strictly depends on AP2 (14), it is possible that this finding was because of the small amount of this transcription factor in medulloblastoma cells. The digestion pattern of the SacI-NcoI probe was not modified by the transfection of the cDNA encoding PHOX2A (Fig. 6C, compare lanes 5 and 6) or by the presence of the anti-PHOX2A antibody (Fig. 6C, lanes 6 and 8).
The results of the EMSAs and DNase I footprinting experiments suggest that there is not a non-canonical PHOX2A-binding site in the 3 minimal promoter and are also apparently in conflict with those of the ChIP and DNA pull-down experiments. However, it should be remembered that EMSAs and DNase I footprinting assays are less sensitive for detecting indirect than direct DNA-protein interactions, in which a given transcription factor directly binds to the DNA, whereas ChIP and DNA pulldown assays are particularly suitable for revealing the presence of a given transcription factor in the context of a transcriptional complex even if it is tethered to its target promoter by means of interactions with one or more proteins in direct contact with DNA. Taken together, all of our biochemical studies led to the suggestion that PHOX2A may indirectly associate with the 3 minimal promoter by interacting with an unknown transcription factor.
Transcriptional Effects of PHOX2A Mutants on the Activity of the 3 and 4xTK Promoters—To test the hypothesis that PHOX2A regulates 3 promoter activity without directly binding to DNA, we generated PHOX2A mutants in which a single amino acid substitution in the helix 3 of the homeodomain prevents it from binding to DNA; the R53W-PHOX2A mutant could not bind the DNA back-bone, whereas the N51H mutant was unable to contact the nitrogenous bases (29). To confirm that the mutants no longer bound DNA, we carried out EMSAs using in vitro expressed proteins and PRS1 as a probe. PHOX2A-R53W and PHOX2A-N51H could be expressed in vitro as efficiently as their wild-type counterpart (Fig. 7A, compare lanes 2 and 3 with lane 4). We had previously established that the in vitro expressed wild-type PHOX2A protein was capable of binding its cognate site (Fig. 2D), although the retarded band was hardly detectable in this set of experiments (Fig. 7D, lane 3); however a prominent ultra-retarded band could be obtained in the presence of the anti-PHOX2A antibody, which was competed by an excess of cold oligonucleotide (Fig. 7B, lanes 3–6). As expected, no band could be detected with the PHOX2A-R53W and PHOX2A-N51H mutated proteins even when the antibody was co-incubated (Fig. 7B, lanes 7–10), thus confirming that they lacked the ability to bind DNA.
We also verified that the mutated versions of PHOX2A could efficiently reach the nucleus by transfecting DAOY cells with the plasmids PHOX2A-R53W/pcDNA3 and PHOX2A-N51H/pcDNA3 and carrying out immunofluorescence using the anti-PHOX2A antibody. It was found that both mutants entered into the nucleus (Fig. 7C), although they were also partially located in the cytoplasm, which suggested that the PHOX2A DNA-binding domain may also be involved in nuclear import and/or retention. Finally, we carried out co-transfection experiments using 4xTK and SacI-NcoI as reporter constructs, and PHOX2A-R53W/pcDNA3 and PHOX2A-N51H/pcDNA3 as effector plasmids.
Both SacI-NcoI and 4xTK were transactivated by the wild-type PHOX2A in the same way (Fig. 1B), but their transcriptional responses diverged when they were co-transfected with the mutated versions of PHOX2A; the 4xTK construct did not respond to either, whereas SacI-NcoI retained the ability to be transactivated by both, although to a lesser extent (Fig. 7D). These observations strongly indicate that the DNA-binding domain of PHOX2A is partially dispensable for the transactivation of SacI-NcoI because the activity of the 3 promoter is regulated by PHOX2A through a mechanism that does not depend on DNA binding, whereas the 4xTK construct, which contains a well characterized DBH promoter multimerized DNA-binding site for PHOX2A, can only be transactivacted if the transcription factor is allowed to bind DNA directly.
Identification of the Molecular Interactor That Tethers PHOX2A to the 3 Promoter—Our data suggest that PHOX2A might regulate the 3 promoter by indirectly binding to the BsiWI-AccIII region through interactions with one or more transcription factors directly located on DNA. The BsiWI-AccIII region has been widely characterized and is known to contain two sites for Sp1 that are essential for promoter activity (14), and so we reasoned that one possible molecular partner of PHOX2A could be precisely Sp1. We tested this hypothesis by setting up a co-immunoprecipitation assay. The nuclear extracts were obtained from IMR32 cells, and the proteins were immunoprecipitated with the anti-Sp1 or the anti-PHOX2A antibody and analyzed by Western blotting, using the appropriate antibody. When the nuclear extracts were immunoprecipitated with the anti-Sp1 antibody, the precipitated complex also contained PHOX2A, which was easily revealed as a doublet by its specific antibody in Western blot experiments (Fig. 8A, lane 3); however, when the nuclear extracts were immunoprecipitated with a preimmune antibody, no immunoprecipitation of PHOX2A was detectable (Fig. 8A, lane 4). A reciprocal experimental strategy showed that a protein complex immunoprecipitated with the anti-PHOX2A antibody also contained Sp1 (Fig. 8A, lane 7), and once again, when the nuclear extracts were immunoprecipitated with a preimmune antibody, no immunoprecipitation of Sp1 could be detected (Fig. 8A, lane 8). In addition to using a preimmune antibody as a control, we also used an antibody directed against the human CREB-1 transcription factor to precipitate IMR32 nuclear extracts. The fact that no interaction with PHOX2A could be detected (Fig. 8B, lane 3) confirmed the specificity of the observed interactions between Sp1 and PHOX2A.
|
3 Promoter—To assess whether the two Sp1-binding sites, spanning the BsiWI-AccIII region, contribute equally to the formation of the PHOX2A containing complex on the 3 promoter, we performed a transient chromatin immunoprecipitation assay. Different 3 promoter plasmids carrying mutations either in the A or B or both Sp1-binding sites or no mutations (BglII-NcoI mutA, BglII-NcoI mutB, BglII-NcoI mutAB, and BglII-NcoI, respectively) (Fig. 9, A and B, left panel) were transiently transfected in IMR32, and chromatin was immunoprecipitated with anti-PHOX2A or with anti-Sp1 antibodies, as a control. To avoid the amplification of the endogenous promoter region, the immunoprecipitated fragments were detected by PCR using primers designed on sequences of the plasmid backbone, flanking the 3 minimal promoter region. As expected a band of 472 bp was amplified after immunoprecipitation with anti-PHOX2A and anti-Sp1 antibodies in the presence of the chromatin extracted from IMR32 cells transfected with the wild-type plasmid (BglII-NcoI) (Fig. 9B, lanes 2 and 4), thus confirming what was obtained with the endogenous promoter region (Fig. 3A). The mutation of the site A (BglII-NcoI mutA) did not apparently affect the binding of Sp1, which remained capable of binding to the site B (Fig. 9B, lane 4), but completely abolished the tethering of PHOX2A to the promoter region (BglII-NcoI mutA, Fig. 9B, lane 2). On the contrary, the mutation of the site B did not change the immunoprecipitation pattern with respect to the wild-type plasmid. This suggests the different ability of the two sites in tethering PHOX2A to the 3 promoter.
|
3 promoter was also confirmed by functional data in which it was shown that the BglII-NcoI mut B plasmid was fully transactivated by PHOX2A (data not shown). |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
3 nAChR subunit gene promoter. In particular, our findings based on in vivo and in vitro molecular criteria show that PHOX2A assembles on the SacI-NcoI region of the 3 promoter, and our functional results indicate that it exerts its transcriptional effects by acting through the 60-bp minimal promoter, despite the absence of any putative DNA-binding site for homeodomain proteins. This is not surprising in itself as it is well known that some transcription factors can also use non-canonical DNA sites; however, our EMSA and DNase I footprinting experiments did not reveal any direct in vitro interaction between PHOX2A and the 60-bp 3 minimal promoter, thus leading to the suggestion that PHOX2A does not directly bind any sequence in the promoter to regulate 3 gene transcription but interacts with one or more transcription factors that are in direct contact with DNA. PHOX2A has been widely characterized on the basis of its ability to regulate gene expression by directly binding the DNA consensus sequence common to homeodomain protein, i.e. the "ATTA" core sequence. No information is available concerning the possibility that PHOX2A regulates gene expression by means of DNA-independent mechanisms, but there are some published data showing that homeodomain-containing proteins can transactivate their target genes without directly contacting DNA. For example it has been shown that Pax6
PD (a Pax6 isoform that lacks the paired domain), Rax, Pbx1, HoxB1, Lhx2, and Chx10 are capable of superactivating a synthetic reporter promoter containing six consensus Pax6 paired domain binding sites upstream from the adenovirus E1b minimal promoter, and that this superactivation occurs without any direct DNA contact as a result of interactions between these homeodomain-containing proteins and Pax6, bound to its cognate cis-acting elements through its paired domain (30). We reasoned that this could also be true of PHOX2A, which could establish protein-protein interactions with factors already bound to the 60-bp 3 minimal promoter. One possible implication of this molecular mechanism might be that the ability of PHOX2A to bind DNA, which resides in the homeodomain, becomes completely or partially dispensable when it is involved in regulating the 3 promoter. When strategic point mutations were introduced into PHOX2A to prevent it from binding to DNA, it completely lacked the ability to transactivate a synthetic DBH promoter but still retained part of its ability to transactivate the 3 promoter.
However, these findings also indicate that the homeodomain is not completely dispensable for the transactivation of the 3 promoter, probably because it also participates in additional functions, such as the nuclear import, which is partially affected in the mutated version of PHOX2A. Interestingly, it has been shown that a nuclear localization signal can be contained in the helix 3 of some homeodomain proteins and that mutations in this subdomain can affect nuclear import (31).
We have shown in a previous study that the 60-bp 3 minimal promoter contains several Sp1 sites, which play fundamental roles in the activity of the promoter, as their mutation completely abolished all transcriptional activity (14 and data not shown), and so we reasoned that Sp1 might be one of the molecular partners of PHOX2A, and we tested the hypothesis by means of co-immunoprecipitation assays. Our data clearly indicate that Sp1 and PHOX2A can specifically interact with each other, although it remains to be understood whether this interaction is direct or mediated by additional factors insofar as in vitro expressed Sp1 and PHOX2A proteins failed to interact in the co-immunoprecipitation assays (data not shown), thus suggesting that an adaptor may be required to connect the factors. Alternatively, it is also possible that Sp1, PHOX2A, or both have to undergo appropriate post-translational modifications missing from the in vitro expression process before they can directly interact. For example, it has been described that PHOX2A can undergo phosphorylation, and it influences its DNA binding activity (32). Whatever the case, our findings indicate that Sp1 could be the factor that tethers PHOX2A to the 3 promoter. In fact, it has been shown very recently that Cdx2, a transcription factor that is a member of the caudal related homeobox gene family and mainly expressed in the intestine, regulates the PEPT1 gene by means of a DNA-independent molecular mechanism that involves Sp1 as a tethering factor (33). In general, this molecular mechanism leaves open the question as to how specifically these homeodomain-containing proteins transactivate the genes whose promoters are regulated by Sp1.
In many of our experiments we used the promoter regulating the expression of the 5 nicotinic subunit gene as a control. The 5 gene is located in the same genomic cluster with the 3 gene (22), and its gene product can assemble with 3 in the same receptor molecule, and in particular the human 5 promoter contains several Sp1 sites (15) which, according to the data obtained using the bovine orthologue (28), should be essential for the expression of the 5 subunit. However, we did not detect PHOX2A on the 5 promoter in either our ChIP or DNA pull-down experiments, nor could we detect any transactivation of the 5 promoter by PHOX2A on our co-transfection experiments. It therefore seems that PHOX2A discriminates the different promoters that bind and are regulated by Sp1, although the molecular mechanism by which it does so remains to be elucidated. One possible explanation may reside in the geometry of the Sp1 sites on the promoter; in the simplest model, PHOX2A may interact with one or more Sp1 proteins whose sites on DNA should be appropriately localized to favor adequate protein-protein interactions for stabilizing PHOX2A on the promoter itself.
To this goal, we tried to establish whether the two Sp1 sites in the 3 minimal promoter were equally relevant to the PHOX2A recruitment. Our transient ChIP assays clearly demonstrate that, although the integrity of the site A was absolutely required for the tethering of PHOX2A, the site B was completely dispensable and thus not involved in this function. It is also possible that additional mechanisms participate in the stabilization of the PHOX2A-Sp1 complex, thus lending specificity to the interaction.
We have shown previously (14) that the 60-bp 3 minimal promoter has ubiquitous activity that is redirected in such a way as to ensure that 3 gene expression is mainly confined to the neural phenotype. This tissue-specific profile seemed to be mainly achieved by means of the combined effects of relatively short upstream and downstream regions flanking the minimal promoter. Each region alone stimulates the activity of the minimal promoter in both neuronal and non-neuronal cells, but when combined, they synergistically activated the promoter only in neuronal cells. Taken together, these data suggest that the transcription factors bound to these flanking regions do not exert their tissue-specific effects directly but by cooperatively promoting the recruitment and stabilization of one or more additional neuro-specific factors on th
