| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Developmental Neurobiology, Instituto Cajal (CSIC), Madrid, Spain
2 Laboratory of Regulation of Neuronal Development, Institute for Protein Research, Osaka University, Osaka, Japan
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Necdin; p75NTR; E2F-1
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
At least 10 new human genes encoding proteins that contain a Mage domain and that reside outside of the mage-A, -B, and -C clusters have been described so far. As a result, the Mage protein superfamily has recently been expanded (10) to include this type II family of Mage proteins, comprising: Mage-D1/NRAGE/Dlxin-1, Mage-D2, Mage-D3/Trophinin/Magphinin, Mage-E1/Mage-D4, Mage-E2, Mage-F1 Mage-G1/Necdin-like 2, Mage-H1, Mage-L2, and Necdin (5). These proteins contain a phylogenetically distinct MHD, although like type I Mage proteins, most of them are encoded by a single exon (5). This has led to the suggestion that the whole Mage superfamily has evolved in mammals by retrotransposition followed by gene duplication from an ancestral gene (10). Accordingly, Necdin is not present in marsupials, and it was probably acquired by retrotransposition during the recent assembly of the Prader-Willi/Angelman syndrome region on the chromosome 15q in humans, an event that occurred 105–180 million yr ago (51). The ancestral mage gene is probably an ortholog of the mage-D genes, which are the only mage mammalian genes that contain introns (10).
Type II Mage proteins are widely expressed in many embryonic and adult tissues, particularly in the nervous system (2, 6, 29, 46, 53). These proteins also interact with the p75 neurotrophin receptor (p75NTR) (5), which displays multiple functions in this tissue. To date these type II Mage proteins have been implicated in the regulation of cell cycle progression, cell differentiation, and apoptosis, acting as adaptors in multiple signal transduction pathways (54). The best characterized example of a mammalian type II Mage protein is Necdin, which was initially isolated from mouse embryonal carcinoma cells differentiated into neurons (42). The mouse necdin gene is predominantly expressed in postmitotic neurons (58), and when expressed ectopically, it suppresses proliferation (25) and triggers neuronal differentiation (31) in different cell lines. These latter effects seem to be mediated by the capacity of Necdin to interact with and block the transactivation domain of E2F-1 (56), a transcription factor necessary for G1/S phase progression that is capable of inducing apoptosis in postmitotic cells (22, 23). The absence of the necdin gene has been associated with the Prader-Willi syndrome, a neurogenetic disorder caused by the deletion of the 15q11-q13 segment of the paternal chromosome, which triggers mental retardation and other physiological alterations (45). Like Necdin, the type II Mage protein Mage-D1 is expressed in neurogenic areas of the developing rat and mouse nervous systems (29, 53), as well as in the mature rat brain (6). Mage-D1 can also suppress cell cycle progression, and it is able to promote p75NTR-dependent apoptosis in sympathoadrenal cells (53). In addition, p75NTR has been shown to sequester Necdin or Mage-G1 through its intracellular domain (p75ICD), thereby favoring the proapoptotic activity of E2F-1 in postmitotic neurons (33, 56).
While it is clear that type II Mage proteins are key elements in neurogenesis and proapoptotic signaling triggered by p75NTR, our understanding of their function in the developing nervous system remains poor. The analysis of these proteins in mammals is complicated because of the large number of related genes with possible redundant functions. Indeed, only minor defects in the development of the nervous system have been observed in null-mutant mice for the necdin gene. Depending on the genetic background, these mice may die in the neonatal period due to apparent respiratory insufficiency that can be explained by abnormal neuronal activity within the putative respiratory rhythm-generating center. Alternatively, they may be viable and fertile simply displaying some functional alterations in the hypothalamus, and changes of behavior reminiscent of Prader-Willi syndrome (21, 45, 52). Therefore, a model system with fewer Mage proteins would facilitate the analysis of Mage function in development and adulthood.
In this study, we show that in the genome of nonmammalian species only one mage gene can be detected. Thus, we cloned the chicken Mage protein, CMage, whose pattern of expression is similar to that of p75NTR in the developing retina, ventral spinal cord, and dorsal root ganglia. Moreover, CMage shows functional similarities to the type II Mage protein Necdin, indicating that the chick may be a useful model system to further characterize the signal transduction pathways used by Mage proteins in the absence of functional redundancy.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Primary antibodies.
The rabbit anti-p75NTR polyclonal antiserum against the cytoplasmic domain of human p75NTR (Promega, Madison, WI) was used at a dilution of 1:500 for immunocytochemistry and 1:5,000 for Western blot analysis. The rabbit polyclonal antiserum (9992) against the intracellular domain of p75NTR was kindly provided by Moses Chao (New York University, New York, NY), and it was used at a dilution 1:1,000 for immunohistochemistry. The mouse polyclonal antibody obtained by immunizing mice with a p75NTR receptor-Fc chimeric protein (11), kindly provided by Alfredo Rodríguez-Tébar (CABIMER, Seville, Spain), was used at a dilution 1:500 for immunohistochemistry. The NC243 antiserum, raised against the 243 COOH-terminal amino acids of mouse Necdin (46), was used at a dilution of 1:2,000 for immunocytochemistry and immunohistochemistry, and a dilution of 1:20,000 in Western blots. The anti-FLAG monoclonal antibody (MAb; Sigma, St. Louis, MO) was used at 20 µg/ml for immunocytochemistry, at 30 µg/ml for immunoprecipitations and at 0.12 µg/ml in Western blots. The anti E2F-1 MAb KH95 (Santa Cruz Biotechnology, Santa Cruz, CA) was used at 1:15,000 in Western blots. 5-Bromo-2'-deoxy-uridine (BrdU) was visualized with the G3G4 MAb (Developmental Studies Hybridoma Bank, Iowa City, IA) used at a dilution of 1:4,000. The mouse TuJ-1 MAb against neuron-specific ßIII tubulin (Chemicon) was used at 1:2,000 dilution. The anti-Islet-1 MAb 40.2D6 (Developmental Studies Hybridoma Bank) was diluted 1:200 for immunohistochemistry.
cDNA probes for Southern blot.
A 319-bp cDNA fragment (corresponding to bp 347–665 of clone ChEST965i23) was amplified by PCR from the pcDNA6-CMage-FLAG plasmid using specific oligonucleotides. This cDNA fragment, included in the region codifying for the MHD of CMage, was then labeled with digoxigenin-11-dUTP by random priming using DIG-High Prime (Roche, Basel, Switzerland) according to the manufacturer's instructions.
RNA probes for in situ hybridization.
Complementary RNA probes for cmage, corresponding to bp 17–730 of clone ChEST965i23 [National Center for Biotechnology Information (NCBI) accession number BX934453] were generated by RT-PCR using a Pyrococcus furiosus (Pfu) DNA polymerase (Biotools, Madrid, Spain) and from a cDNA template derived from E4 eye/tectum. This PCR fragment was cloned into the pGEM-T Easy vector (Promega), and digoxigenin-labeled antisense riboprobes were obtained from linearized plasmid templates using Sp6 RNA polymerase (Roche).
Plasmids.
The pRc/CMV-E2F-1 and the pRc/CMV-Necdin expression vector have been described previously (33, 56). The vector expressing p75ICD (pRc/CMV-p75ICD-HA) was a generous gift of Yves-A. Barde (University of Basel, Basel, Switzerland). Green fluorescent protein (GFP) was expressed in the DF-1 cells by using the pEGFP-N1 plasmid (BD Biosciences, San Jose, CA). The coding sequence of CMage-FLAG corresponded to bp 17–754 of clone ChEST965i23 and was amplified with Pfu DNA polymerase (Biotools) from cDNAs derived from E4 eye/tectum using the following oligonucleotides: upstream primer (CACAAGCTTATGTCTCAGAGGAAGCGCAGC); downstream primer (CTCGAATTCCTA CTTATCGTCGTCATCCTTGTAATCCGTGTGGCTTTGGCCTCG). The PCR amplification product of 786 bp contained a HindIII cleavage site at the 5'-end, and a FLAG tag sequence followed by a stop codon and an EcoRI restriction site at the 3'-end. This fragment was cut with HindIII and EcoRI, inserted in the HindIII/EcoRI site of pcDNA 6/V5-His-A vector (Invitrogen, Carlsbad, CA), and the integrity of the resulting expression vector (pcDNA6-CMage-FLAG) was confirmed by sequencing. The coding region of the cmage gene was cloned into the pGEM-T Easy vector (Promega) (see below) and referred to as pGEM-cmage. The pRFPRNAiC and pRFPRNAi Luciferase vectors (12) were provided by Stuart Wilson (University of Sheffield, Sheffield, UK). The pRFPRNAi CMage vector capable of suppressing cmage expression was constructed using the pRFPRNAiC plasmid following the procedures described previously (12). The target sequence used to interfere with the cmage mRNA corresponded to bp 168–189 of clone ChEST965i23. Similar results were obtained when the sequences corresponding to bp 174–195, 58–579, or 666–687 of clone ChEST965i23 were used (data not shown).
Southern blot analysis.
Genomic DNA from E5 chick embryos (10 µg) was digested with CaiI (Fermentas) and separated on agarose gels. As positive controls, EcoRI/HindIII-digested pcDNA-CMage-FLAG plasmid (0.77 ng) and CaiI-digested pGEM-cmage (0.23 ng) were used, these digestions yielding 0.1 ng of DNA corresponding to the cmage cDNA or the cmage gene sequences, respectively. The DNA was transferred onto Hybond-N+ nylon filters (GE Healthcare) in 10x SSC (1x SSC contains 15 mM sodium citrate, 150 M NaCl; pH 7.0), and UV-cross linked (UV Stratalinker 2400, Stratagene). The filters were hybridized overnight at 47°C with the digoxigenin-labeled probes (see above) using DIG High Prime DNA Labeling and Detection Starter Kit II (Roche) following the protocol recommended by the manufacturer. The filters were then washed at low stringency, twice in 2x SSC/0.1% SDS at room temperature for 5 min and once in 2x SSC/0.1% SDS at 57°C for 1 h. The probes were detected following the protocol described by the manufacturer and the filters were exposed to Hyperfilm ECL (GE Healthcare).
Cell culture.
DF-1 chicken fibroblast cells and N1E-115 neuroblastoma cells were maintained in Dulbecco's modified Eagle's medium (DMEM)/10% fetal calf serum (FCS) (Invitrogen) at 37°C in a water-saturated atmosphere containing 5% CO2. N1E-115 neuroblastoma cells and DF-1 cells were both seeded at 30,000 cells/cm2 and maintained for 24 h in DMEM/10% FCS. The cells were transfected with Lipofectamine 2000 (Invitrogen), and the transfected cells were induced to differentiate by adding 2% DMSO (Sigma) to the culture medium 24 h after transfection (30). The cells were then fixed or prepared for Western blot analysis 48 h after transfection. In experiments to quantify apoptosis, N1E-115 neuroblastoma cells were grown on 12-mm coverslips (Menzel-Gläser, Braunschweig, Germany) coated with 500 µg/ml poly(D-L)ornithine.
In vivo BrdU treatment.
Eggs were opened at their blunt end, and 40 µl of a solution containing 10 mg/ml BrdU (Roche) prepared in phosphate-buffered saline (PBS) was applied to the chorioallantoic membrane. Subsequently, the eggs were sealed and returned to the incubator. Embryos were killed 1 h after BrdU treatment.
In situ hybridization.
In situ hybridization was performed as described previously (44), performing all steps under RNase-free conditions. Chick embryos of the ages specified were fixed for 4–8h at 4°C in 4% paraformaldehyde (PFA), incubated overnight at 4°C in 100 mM sodium phosphate buffer containing 30% sucrose, and then embedded in the OCT compound Tissue-Tek (Sakura, Torrance, CA). Cryosections (12 µm) were collected on 3-aminopropyl-trimethoxysilane-coated slides (Sigma), postfixed for 15 min in 4% PFA, and then carbethoxylated for 30 min in PBS containing 0.1% active diethyl-pyrocarbonate (Sigma). Sections were equilibrated for 5 min in 5x SSC and then prehybridized for 2 h at 60°C in 5x SSC containing 50% formamide (Fluka, Seelze, Germany) and 50 µg/ml tRNA (Roche). Hybridization was performed at 60°C overnight in the same solution containing 400 ng/ml of digoxigenin-labeled probes. Prior to hybridization, the riboprobes were denatured for 5 min at 80°C and then cooled on ice. Following hybridization, the sections were washed three times for 1 h in 2x SSC at room temperature (RT), 2x SSC at 65°C, and 0.1x SSC at 65°C. Slides were then incubated for 1 h in 0.5% blocking reagent (Roche) prepared in 150 mM NaCl, 100 mM Tris·HCl pH 7.5 (NT), and the localization of the bound riboprobes was detected by incubating overnight at 4°C with an AP-coupled antidigoxigenin antibody (Roche) diluted 1:5,000 in 0.5% blocking reagent prepared in NT. The slides were then washed twice in NT and equilibrated in 50 mM MgCl2, 100 mM NaCl, 100 mM Tris·HCl pH 9.5 (MNT). The antibody was visualized using an alkaline phosphatase substrate (338 µg/ml nitro blue tetrazolium, 175 µg/ml 5-bromo-4-chloro-3-indolyl phosphate; Roche) in MNT buffer. Finally, the color reaction was stopped by washing the slides in a 10 mM Tris·HCl pH 8.0 solution containing 1 mM EDTA.
Immunohistochemistry.
Immunohistochemistry was performed as described previously (38). In brief, embryos were fixed for 4 h at 4°C in 4% PFA incubated overnight at 4°C in 100 mM sodium phosphate buffer containing 30% sucrose and embedded in the OCT compound Tissue-Tek (Sakura). Cryosections 12 µm thick were permeabilized for 30 min at RT in the presence of 0.1% Triton X-100 (Sigma) in PBS (PBS-T). The sections were blocked for 1 h with 10% normal goat serum (NGS) in PBS-T and subsequently incubated overnight at 4°C with the primary antibody in 1% NGS/PBS-T. After five washes with PBS-T, the sections were incubated for 1 h at RT with a Cy2-conjugated goat anti-rabbit IgG (H+L) antibody (Jackson Immunoresearch, West Grove, PA) diluted 1,000-fold or an Alexa Fluor 594 goat anti-mouse IgG (H+L) antibody (Molecular Probes) diluted 1:1,000. Sections were finally washed five times in PBS-T and once in PBS alone, and the labeled sections were mounted in 50% glycerol in PBS. The sections that were immunolabeled for BrdU were first subjected to DNA denaturation by incubating for 30 min with 2 N HCl/0.33 x PBS at RT, which was neutralized by 3 x 15-min washes with 0.1 M Na borate (pH 8.9) and two 5-min washes with PBS-T. Images were acquired with a Leica (Nussloch, Germany) TCF-4D confocal microscope and used directly to create the figures.
Coimmunoprecipitation.
To detect the E2F-1 complexes, combinations of pRc/CMV-Necdin, pcDNA6-CMage-FLAG, pRc/CMV-E2F-1, and pRc/CMV-p75ICD-HA plasmids were transfected in N1E-115 neuroblastoma cells using Lipofectamine 2000 (Invitrogen). These cells were grown in P60 petri dishes at an initial density of 6.0 x 104 cells/cm2, and they were induced to differentiate with 2% DMSO as described above. The cells were then lysed in a Potter microhomogenizer with 250 µl of lysis buffer containing: 20 mM Tris (Roche) pH 7.5, 100 mM NaCl (Merck, Darmstadt, Germany), 5 mM MgCl2 (Merck), 0.5% Triton X-100 (Sigma), 0.5 mM EDTA (Merck), 0.5 µg/ml DNase I (Roche), and 1x protease inhibitor mix (Roche). After incubating the lysates for 10 min at 4°C, they were centrifuged at 13,000 g for 10 min at 4°C, and 50 µl of the supernatant from each extract was mixed with 50 µl of 2x Laemmli's buffer and boiled for 5 min (Input samples). The rest of the supernatant was incubated with 30 µg/ml anti-FLAG-specific MAb for 2h at 4°C followed by incubation with 20 µl (bed-volume) of protein A/G Sepharose (Santa Cruz Biotechnology) for 1 h at 4°C. Immunoprecipitates were washed three times with 500 µl of lysis buffer, mixed with 20 µl of 2x Laemmli's buffer lacking 2-mercaptoethanol, and boiled for 5 min.
Western blot.
Total cell extracts from 1.5 x 105 cells, or 10-µl immunoprecipitates were separated by SDS PAGE on 11% acrylamide gels and transferred to Immun-Blot polyvinylidene difluoride membranes (Bio-Rad, Hercules, CA). The membranes were incubated for 1 h with 2% ECL Advance blocking agent (ECL Advanced Western Blotting Detection Kit) (GE Healthcare Europe, Munich, Germany) in PBS containing 0.1% Tween 20 (PBT) (Sigma), and incubated for 2 h at room temperature with the appropriate antisera in blocking buffer. After being washed the membranes five times in PBT, they were incubated for 1 h at room temperature with a peroxidase-conjugated Affinipure goat anti-rabbit IgG antibody diluted 1:1,660,000 (Jackson Immunoresearch) or a goat anti-mouse IgG horseradish peroxidase-conjugated antibody at 1:500,000 (Bio-Rad) in blocking buffer. Finally, they were washed again as above, and the protein bands were visualized using ECL Advanced Western Blotting Detection Kit (GE Healthcare Europe).
5'- and 3'-rapid amplification of cDNA ends.
5'- and 3'-rapid amplification of cDNA ends (RACE) was performed with the BD SMART RACE cDNA Amplification Kit (BD Biosciences) using cDNA obtained from whole E3 chick embryos following the manufacturer's instructions. This cDNA was amplified with the Advantage 2 PCR Enzyme System (BD Biosciences) using cmage-specific oligos corresponding to bp 480–507 of clone ChEST965i23. The sequences of these oligos were TATGGGAGTTCCTGCGCCGGCTCCGGGT for 3'-RACE and ACCCGGAGCCGGCGCAGGAACTCCCATA for 5'-RACE.
Genomic DNA amplification.
Amplification of the coding sequence of the cmage gene was performed with Pfu DNA polymerase (Biotools) (n = 2) or with CertAMP (Biotools) (n = 2) from E3 whole chick embryo genomic DNA using the oligonucleotides corresponding to the 5'- and 3'-ends of the coding region of cmage described above. The 5'-untranslated region (UTR) of cmage gene was amplified with Pfu DNA polymerase (Biotools) (n = 2) or with CertAMP (Biotools) (n = 2) from this same genomic DNA using the primers corresponding to bp 1–17 and bp 68–87 of the ChEST868j13 clone (NCBI accession number BU381672). The PCR amplified products were cloned into the pGEM-T Easy vector and sequenced.
Cell death analyses.
N1E115 cells (6 x 104) were transfected with the pEGFP-N1 expression vector (BD Biosciences), together with different combinations of pRc/CMV-p75ICD-HA, pRc/CMV-E2F-1, and pRc/CMV-Necdin, or pcDNA6-CMage-FLAG (1 µg each) using Lipofectamine 2000 (Invitrogen). The plasmid quantities were adjusted to 5 µg with pBlueScript (Stratagene, La Jolla, CA). Cells were induced to differentiate as described previously (33). To quantify apoptosis, the DNA of PFA-fixed N1E115 cells was labeled with 1 µg/ml bisbenzimide (Sigma), and the number of pyknotic nuclei was established. Cells were counted on a Nikon E80i microscope using an oil immersion x60 objective with phase contrast and epifluorescence illumination, and an average of 500 cells were analyzed per coverslip. The means ± SE from at least three independent experiments are shown, and the statistical differences were analyzed by Student's t-test.
Database searches.
Tblastn searches were performed in the nonredundant DNA database available at the NCBI database using the Mage homology domain of mouse Necdin (amino acids 116–280; NCBI accession number: BAA11183) or CMage (amino acids 57–221, Fig. 1C). We ran tblastn searches using the BLOSUM-62 substitution matrix and the default values for the gap costs (existence: 11; extension: 1). Blastn searches in the Biotechnology and Biological Sciences Research Council (BBSRC) ChickEST database (http://www.chick.umist.ac.uk) were performed using the ChEST965i23 sequence as the query. Blastn searches were run using the BLOSUM-62 substitution matrix and the default values for gap and nucleotide mismatching. Blastn searches in the chicken genome database from NCBI were performed using the sequence between bp 185–679 of ChEST965i23, encoding the cmage MHD, following default settings.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Part of the genomic sequence corresponding to the cmage gene, flanked by two gaps of uncertain length, is contained in a genomic contig with NCBI accession number NW_098113. From their sequence, five small exons can be deduced corresponding to bp 219–580 of the ChEST965i23 sequence (equivalent to bp 2738–2817, 2917–3011, 3127–3206, 3293–3335, and 3437–3500 from the genomic contig mentioned above). We have sequenced the whole region of the cmage gene encoding CMage, identifying four additional exons in the 5'-end and two in the 3'-end (NCBI accession number DQ983362). Therefore, from the abovementioned sequence 11 exons encompass the whole cmage gene, corresponding to bp 1–14, 167–210, 309–404, 597–660, 897–976, 1079–1173, 1278–1357, 1444–1486, 1588–1650, 1941–2055, and 2157–2219 (Fig. 2A), being the coding sequence of cmage comprising exons 2 through 11 (Fig. 2, A and B). Ten small introns (from 86 to 290 bp in size) distributed throughout the sequence were detected in the cmage gene, in accordance with the reduced size of the chicken genome known to be one-third that of a typical mammal (8). All the introns obey the GT-AG rule of splice junctions (Fig. 2B).
|
A hypothetical CMage protein (NCBI accession number XM_424083) was previously predicted by automated computational analysis of a genomic region contained in the genomic contig mentioned above. This hypothetical protein derives from part of intron 4 and exons 5 to 9 of the cmage gene sequence, followed by two exons included in the chicken ESTs ChEST561j8, ChEST911e13, and ChEST183i21. These latter EST sequences are almost identical to the chicken 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 4 mRNA (NCBI accession number: NM_001030584). Moreover, the mRNA encoding this hypothetical CMage protein cannot be detected in the BBSRC ChickEST database. Indeed, we were unable to detect this mRNA by 3'- or 5'-RACE from E3 whole chick embryo or stage 36 heart cDNA (data not shown). Together, these data do not support the existence of this hypothetical protein.
mage gene family comprises single members in nonmammalian species.
Although more than 10 different type II mage genes have been described in mammals (5), the genomes of Drosophila, teleost fish, and chicken (G. gallus) seem to contain only a single mage gene (7, 47, 50; this study). We therefore decided to verify whether this is also the case in other nonmammalian species. Using information available in the NCBI public database, we examined the existence of putative mage genes in the genome of birds (Taeniopygia guttata), amphibia (Xenopus tropicalis), fishes (Danio rerio and Tetraodon nigroviridis), echinoderma (Strongylocentrotus purpuratus), insects (Drosophila melanogaster, Drosophila yakuba, Apis mellifera, and Anopheles gambiae), plants (Oryza sativa and Arabidopsis thaliana), fungi (Cryptococcus neoformans), nematodes (Caenorhabditis elegans), and protists (Entamoeba histolytica: Table 1, Fig. 2A). These data indicate that the genomes from nonmammalian species contain only one mage gene (Fig. 2A), most probably the ortholog of the ancestral gene that gave rise to the Mage superfamily of proteins in mammals (Fig. 2D). Although it was proposed that the ancestral mage gene originally contained multiple exons (10), the coding sequences of the Entamoeba histolityca and D. melanogaster mage genes are contained in a single exon (Fig. 2A). Therefore, the ancestral mage gene was probably encoded by a single exon, and it has acquired introns during the course of evolution. Indeed, the number of introns in the mage genes in the different animal phyla seems to increase as they have evolved (Fig. 2A).
|
|
|
310 million yr ago, at a phylogenetic distance not previously covered by other genome sequences (8). Therefore, these organisms are theoretically the best model system to study Mage protein function in higher vertebrates. To further characterize CMage, we amplified its coding sequence from cDNAs obtained from E4 chick embryos using specific primers designed to add a FLAG tag in the COOH terminus of the molecule. This cDNA was subsequently cloned into a eukaryote expression vector. The flagged CMage protein was expressed in DF-1 cells, an immortalized chick fibroblast cell line, together with an RNA interference (RNAi) construct specific for either cmage or luciferase. Immunostaining performed with the NC243 anti-Necdin antiserum could be detected in most DF-1 cells cotransfected with cmage and luciferase-specific RNAi constructs, but not when cmage was coexpressed with the cmage-specific RNAi vector (Fig. 4A).
|
High levels of CMage can be detected in vivo in regions that accumulate p75NTR and are enriched in postmitotic neurons.
The expression of cmage was analyzed in the chick embryo by in situ hybridization with a cmage-specific probe and by immunohistochemistry using the NC243 antiserum. At E4, the earliest stage analyzed, cmage mRNA was particularly enriched in the neural epithelium (Fig. 5A). Accordingly, strong levels of cmage mRNA were detected at different developmental stages in specific regions of the developing nervous system with the cmage-specific antisense probe (Fig. 5, B, E, H, and K), but not with a control probe containing the sense sequence of cmage (Fig. 5B). This analysis revealed that cmage was expressed strongly in the dorsal root ganglia at all developmental stages analyzed and in the ventral horn of the spinal cord between embryonic day (E) 4 and E6. The expression of cmage subsequently decreased in the ventral horn of the spinal cord as development proceeded and was virtually absent at E11. In addition, sympathetic ganglia also strongly expressed cmage mRNA at E6–E11. The expression of CMage in these neural structures was also analyzed by immunohistochemistry in adjacent brachial sections, and as expected, the distribution of CMage protein was highly coincident with that of its mRNA (Fig. 5, C, F, I, and L). Thus, CMage was enriched in cells located in the ventral horn of the spinal cord, the number of these cells diminishing as development proceeded, and it also accumulated in the dorsal root ganglia at all developmental stages studied. CMage was also detected in the sympathetic ganglia (data not shown).
|
|
The areas where CMage were seen to be expressed are known to contain postmitotic neurons. To directly test whether CMage is mainly expressed by cells that cannot proliferate, chick embryos from different developmental stages were treated with BrdU for 1 h and then killed. Sections from these embryos were double immunostained with the NC243 antiserum and anti-BrdU. Most areas containing high levels of CMage immunolabeling excluded BrdU immunostaining (Fig. 7, A and C, and data not shown), indicating that the cells expressing CMage at high levels were postmitotic. Double labeling was also performed in these sections with the NC243 antiserum and the MAb TuJ-1, which specifically recognizes neurons (39). This analysis confirmed that the cells that express CMage at high levels in the dorsal root ganglia, the ventral horn of the spinal cord, and the retina are predominantly postmitotic neurons (Fig. 7, B, D, and E).
|
CMage can interact with both E2F-1 and p75ICD.
To test whether CMage shows functional similarities to the mammalian type II Mage proteins, we compared it with Necdin, used in this study as a prototypical member of this protein family. Necdin has been shown to maintain neurons in a postmitotic state by binding to the transactivation domain of the E2F-1 transcription factor, thereby blocking its function and mimicking the function of Rb (56). Hence, we examined whether CMage also interacts with E2F-1 in differentiating N1E-115 neuroblastoma cells. Full-length p75NTR has been shown to interact with Necdin and Mage-G1 through its intracellular domain in differentiating N1E-115 neuroblastoma cells, displacing the interaction of these proteins with E2F-1 (33). Therefore, we also assessed whether the presence of p75ICD can prevent CMage from interacting with E2F-1.
To address these issues, we performed coimmunoprecipitation assays on cell extracts from N1E-115 neuroblastoma cells transfected with E2F-1 and CMage (or Necdin) alone or in the presence of p75ICD and induced to differentiate (30). An interaction between Necdin or CMage and E2F-1 was readily detected in the absence of p75ICD expression (Fig. 8, A and B), demonstrating that like Necdin, CMage can interact with E2F-1. The expression of p75ICD in differentiating N1E-115 neurons impaired the interaction between CMage and E2F-1 while a strong interaction of CMage with p75ICD was observed (Fig. 8B) similar to that observed for the interaction between Necdin and E2F-1 (Fig. 8A). Thus, we conclude that CMage can interact with p75ICD like other type II Mage proteins [including Mage-D1 (53), Necdin (33, 57), Mage-G1 (33), and Mage-H1 (57)]. Importantly, this association impaired the interaction of CMage with E2F-1, as previously shown for Necdin and Mage-G1 (33).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Despite the presence of multiple exons in most nonmammalian mage genes, alternative splicing of exons is a rare event that usually takes place in noncoding regions. This indicates that only one Mage protein per genome is expressed in most nonmammalian species. Nevertheless, alternative forms of zebrafish mage-specific mRNA lacking the first exon can be found in the NCBI database. The analysis of the BBSRC ChickEST database indicated the existence of two main isoforms of cmage mRNA, similar to mage isoforms 1 and 3 from zebrafish, both encoding an identical form of CMage. cmage isoform 1 differs from isoform 2 in its 5'-UTR due to alternative splicing of intron 1 in the mature mRNA.
Structurally, most Mage proteins from nonmammalian species contain a conserved MHD in the center of the molecule. This domain has been conserved during evolution, and it is highly related to the MHD of the type II Mage proteins, in accordance with a previous study demonstrating that the Mage proteins from Drosophila and mammals are phylogenetically conserved (50). Most nonmammalian Mage proteins contain two small MHD flanking regions similar to human Necdin, Mage-G1, and Mage F1. Nevertheless, other mammalian Mage proteins have acquired additional domains in these flanking regions, as is the case of the Mage-D proteins that contain the so-called MHD2 and an "interspersed repeat domain" (5). Such motifs seem to have been acquired during the course of evolution of the mammalian genome as they are not present in nonmammalian Mage proteins.
CMage is a new member of the Mage family expressed in specific areas of the nervous system enriched in p75NTR.
In this study, we cloned the full-length coding sequence corresponding to the cmage gene, which contains eleven exons distributed over a 2.2-kb genomic fragment. In accordance with other nonmammalian species, cmage seems to be unique in the chicken genome since only a single band could be observed in low-stringency Southern blots of genomic DNA, and Blastn searches of the chicken genome did not yield any additional sequence to that of cmage. The cDNA cloned encodes CMage, which is structurally similar to the members of the type II Mage protein family.
CMage was specifically recognized by the NC243 antiserum directed against mouse Necdin, which detected a specific band of 28.5 kDa in extracts from DF-1 cells transiently transfected with CMage or from chick retinal cells and was able to immunostain CMage-expressing DF-1 cells but not control cells. Furthermore, immunostaining with NC243 in tissue sections yielded a similar pattern to that observed by in situ hybridization, again indicating that this antibody specifically recognizes CMage.
These experiments demonstrate that CMage was strongly expressed in developing areas enriched in projecting neurons such as the retinal ganglion cell layer, the ventral horn of the spinal cord, and dorsal root ganglia. Low levels of CMage expression were also detected in proliferating progenitor cells at the early stages of retinal development, in accordance with the finding that the Drosophila Mage protein may be expressed in mitotically active neural precursors such as neuroblasts and ganglion mother cells (47). Moreover, weak levels of this Mage protein were detected throughout the embryo, in accordance with the existence of ESTs from several specific tissues encoding CMage (Table 2). The areas where CMage was strongly expressed contain postmitotic neurons and express p75NTR, known to interact with type II Mage proteins through its intracellular domain (5). Indeed, p75NTR is expressed by chicken retinal ganglion cells (60), dorsal root ganglia (26), and motoneurons (26, 43), and we confirmed that p75NTR is also expressed in these regions and colocalizes with CMage in most cells.
CMage shows functional similarities with the type II Mage protein Necdin.
In this study we present evidence of functional similarities between CMage and Necdin, based on the capacity of both proteins to interact with p75NTR and E2F-1.
p75NTR is able to transduce proapoptotic signals in response to ligand binding (20), which can be linked to alterations in cell cycle progression (37), and type II Mage proteins are known to interact with the intracellular domain of p75NTR (33, 53, 57), thus constituting a potential link between p75NTR, cell cycle regulation, and apoptosis (37). To date, four different type II Mage proteins have been shown to interact with the intracellular domain of p75NTR in mammals: Necdin (33, 57), Mage-D1 (53), Mage-G1 (33), and Mage-H1 (57). Of these proteins, Necdin is predominantly expressed in postmitotic cells (2, 42, 58), and its pattern of expression in the developing mouse correlates with p75NTR expression (4), as occurs with CMage (this study). Through coimmunoprecipitation of total extracts from differentiating N1E-115 neuroblastoma cells cotransfected with p75ICD and CMage or Necdin, we demonstrated that the latter proteins are both able to interact with p75ICD. These results stress the functional conservation between CMage and type II Mage proteins.
Besides its capacity to interact with p75ICD, Necdin can suppress proliferation in several cell lines due to its capacity to repress the activity of the E2F-1 transcription factor (25, 56). Previous studies in differentiating N1E-115 neuroblastoma cells have demonstrated that full-length p75NTR can sequester Necdin and Mage-G1 at the cell membrane, preventing their interaction with E2F-1 (33). In this study, we have demonstrated that the binding of the p75ICD fragment to Necdin or CMage can prevent these latter proteins from establishing an inhibitory interaction with E2F-1.
The presence of functional E2F-1 in postmitotic neurons is known to trigger cell death through well-characterized mechanisms, including the stabilization of p53 levels or the induction of cdk1/cdc2 expression (22, 23), being the latter mechanism directly attenuated by Necdin (32). Furthermore, the activation of p75NTR by neurotrophins can provoke the release of p75ICD, which is linked to the induction of apoptosis (19, 28, 49). We have demonstrated that releasing E2F-1 from its interaction with CMage or Necdin in the presence of p75ICD is capable of facilitating apoptosis of differentiating N1E-115 neuroblastoma cells, as occurs when Necdin is sequestered by full-length p75NTR in the cell membrane (33). In vivo, the lack of Necdin has been shown to trigger sensory defects derived from a loss of dorsal root ganglion neurons, probably due to apoptosis (3). Whether or not the mechanism used by Necdin to prevent this loss of sensory neurons is based on a mechanism dependent on p75ICD/E2F-1 remains to be analyzed.
While Necdin and Mage-G1 have been reported to bind to p75NTR and E2F-1, other type II Mage proteins such as Mage-L2 do not bind to these proteins (33). This functional divergence raises the question as to whether the capacity of some Mage proteins to interact with p75NTR and E2F-1, preventing E2F-1 activity, was acquired during mammalian evolution. Conversely, this function may have existed in other phyla before the origin of mammals, and it was lost in some mammalian Mage members. In this study we demonstrate that this function seems to have arisen before mammals diverged from other phyla, as a nonmammalian Mage protein, CMage, is able to interact with p75NTR and E2F-1, and prevent the proapoptotic function of the latter.
In this study we have compared the function of CMage and Necdin. However, CMage appears to be more closely related to Mage-D1 in structural terms (Fig. 1B). Therefore, we cannot exclude that Mage-D1 may also be able to interact with E2F-1 and block its proapoptotic function, as previously shown for other type II Mage proteins (33). Mage-D1 has incorporated novel protein domains into its amino acid sequence during the course of mammalian evolution (5) and unlike Necdin and CMage, it is strongly expressed in proliferative neural subpopulations where p75NTR is absent (29, 53). These distinctive features of Mage-D1 indicate that this protein may have acquired additional functions to that of other type II Mage proteins during the course of mammalian evolution.
In conclusion, we believe that the functional similarity between CMage and Necdin coupled with the fact that only one mage gene is present in the chicken genome strengthens the idea that the chick represents a potentially useful model system to further analyze Mage protein function.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Present address for Z. Fernández-González: Dept. of Neurology, Univ. of Navarra, E-31008 Pamplona, Spain.
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: J. M. Frade, Instituto Cajal, CSIC, Avda Doctor Arce 37, E-28002 Madrid, Spain (e-mail: frade{at}cajal.csic.es).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
FLAG), and immunoblotted (WB) with anti-p75ICD (