| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Variant Involved in the Regulation of Myoblast Differentiation
Unité Mixte de Recherche-866 Différenciation Cellulaire et Croissance [Institut National de la Recherche Agronomique (INRA)-Université Montpellier II-Ecole Nationale Supérieure Agronomique de Montpellier], INRA, 34060 Montpellier Cedex 1, France
Address all correspondence and requests for reprints to: François Casas, Unité Mixte de Recherche-866 Différenciation Cellulaire et Croissance, Institut National de la Recherche Agronomique (INRA), 2 place Viala, 34060 Montpellier Cedex 1, France. E-mail: casasf{at}ensam.inra.fr.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
(NR1A1) and TRß (NR1A2) genes. Several TR variants have already been characterized, but only some of them display T3 binding activity. In this study, we have identified another transcript, TR-
E6, produced by alternative splicing with microexon 6b instead of exon 6. This splicing leads to the synthesis of a protein devoid of a hinge domain. The TR-E6 transcript is detected in all mouse tissues tested. Although TR-E6 did not bind DNA, its expression induced a TR1 sequestration in the cytoplasm. Functional studies demonstrated that TR-E6 inhibits the transcriptional activity of TR1 and retinoic X receptor-, but not of retinoic acid receptor-. We also found that TR-E6 efficiently decreased the ability of TR to inhibit MyoD transcriptional activity during myoblast proliferation. Consequently, when overexpressed in myoblasts, it stimulated terminal differentiation. We suggest that this novel TR variant may act as down regulator of overall T3 receptor activity, including its ability to repress MyoD transcriptional activity during myoblast proliferation. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and TRß genes (NR1A1 and NR1A2 according to nuclear hormone receptor nomenclature) (1, 2). Both TR loci are complex, and numerous TR proteins are produced by alternative promoter usage, alternative splicing, and the use of internal initiation codon. These processes lead to the synthesis of four nuclear receptors (TR1, TRß1, TRß2, and TRß3) (3, 4, 5, 6, 7), three mitochondrial receptors (p43, p28, and TRß0) (8, 9, 10), and five nonreceptor isoforms (TR2, TR3, TR1, TR2, and TRß3) (7, 11, 12, 13). Because of this complexity, numerous models of TR knockout have been developed to understand the specific function of each protein. Phenotype analyses have revealed significant differences, despite redundancy among the various TRs, in the observation that the loss of one variant can be overcome by the activity of other isoforms in many tissues (14). However, TR appeared to be crucial for postnatal development and cardiac function, whereas TRß mainly controls thyroid hormone levels, liver metabolism, and development of retinal and auditory functions (14).
In earlier studies we have shown that T3 stimulates myoblast differentiation through inhibition of activator protein 1 (AP1) activity (15). This T3 myogenic influence mediated by TR1 is only functional at a particular stage of myoblast progression in the myogenic program characterized by RXR expression. In addition, we provided evidence that TR1 is involved in a mechanism preserving the duration of myoblast proliferation by repressing MyoD (myogenic determination factor 1) and myogenin transcriptional activity independently of the presence of the hormone (16).
In this report, using a RT-PCR approach in murine C2C12 cells or in mouse liver, we have identified a new TR transcript produced by alternative splicing in which a microexon (exon 6b) is used instead of exon 6 (TR-E6). This splicing leads to generation of a TR1 protein without a hinge domain. Furthermore, this transcript is detected in various mouse tissues. We found that TR-E6 did not bind DNA but induced a TR1 sequestration in the cytoplasm. Functional studies demonstrated that this TR variant inhibits the transcriptional activity of TR1 and RXR, but not of RAR. Last, its overexpression stimulates QM7 myoblast differentiation by restoring MyoD transcriptional activity through inhibition of the functional TR1/MyoD interaction. These data suggest that TR-E6 might exert a regulatory role during mouse muscle development.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Gene variants, RT-PCR experiments were performed on total RNA extracts from murine C2C12 cells using a set of specific primers covering the TR1-coding region (Fig. 1A
). Using primers designed to hybridize to the extremities of the TR1 transcript, we detected a PCR product of the expected size (Fig. 1B). Surprisingly, a second PCR product, approximately 200 bp shorter, was systematically obtained in the same PCR reactions (Fig. 1B). To test the amplification specificity of the shorter product, we purified the two cDNAs to perform additional PCR experiments. Using primer pairs –4/1250, 107/1250, –4/927, 107/927, we systematically observed the difference of approximately 200 bp between the cDNA encoding the TR1 transcript and the shorter form (Fig. 1C). However, when the primer pairs 444/1250 included a primer hybridizing to a sequence corresponding to exon 6 (444), only the cDNA corresponding to the full-length TR1 was amplified (Fig. 1C). Lastly, the use of primers pairs 733/927 and 733/1250 gave rise to a PCR product of similar size for both purified cDNAs (Fig. 1C). This set of data clearly supported the possible existence of a new TR transcript lacking a sequence of about 200 bp, including a part of exon 6.
|
variant lacked the 368 to 608-bp sequence corresponding to the exon 6 encoding region (numbering based on mouse nucleotide sequence GenBank accession no. X07750). However, 10 nucleotides were substituted for the exon 6 sequence relative to the TR1 transcript (Fig. 2). Interestingly, these changes do not result in a frame shift introducing a premature stop codon but could lead to the synthesis of a TR1 protein without a hinge domain (TR-E6). To rule out the possibility that this transcript could be a peculiarity of the C2C12 cell line, similar experiments were conducted using total RNAs extracted from mouse liver. Identical results were obtained, thus leading to the conclusion that a novel TR transcript is expressed in mice.
|
-E6 gene from mouse liver and C2C12 cells.
The first interesting indication provided by the analysis was that 10 nucleotides substituting the sequence corresponding to exon 6 in the full-length transcript originate from microexon 6b, located between exons 6 and 7 (Fig. 3A). This microexon is located in the context of seven G-rich repeats, similar to that previously reported in the cardiac troponin T gene, acting as an intron splicing enhancer leading to the inclusion of a microexon in the corresponding transcript (17). This observation suggests that these G-rich repeats could be involved in the use of microexon 6b.
|
-E6 sequence (Fig. 3C
). Whereas a thymidine in the genomic sequence of microexon 6b did not appear in the transcript, an additional cytidine was inserted four nucleotides downstream. These changes were simultaneously observed in several sequence analyses performed in both mouse liver and C2C12 genomic sequences and TR-E6 transcripts, thus ruling out the possibility of an experimental artifact. Interestingly, both nucleotides changes occurred in a genomic region displaying a perfect palindromic structure possibly involved in the substitution (Fig. 3D). These data lead us to propose that an editing activity is involved in the synthesis of the TR-E6 transcript.
The TR-E6 Transcript Is Ubiquitously Expressed in Mouse Tissues
The expression pattern of the TR-E6 transcript was monitored in different mouse tissues in RT-PCR experiments using a specific primer hybridizing to microexon 6b (5'E6). We observed that TR-E6 expression was detected in all tissues tested: spleen, liver, kidney, heart, lung, muscle, and white adipose tissue and was at the brink of detection in brain (Fig. 4).
|
-E6 Encodes a TR1 Protein without Hinge Domain-E6 transcript to encode the predicted protein, in vitro and in vivo translation experiments were performed with both TR1 and TR-E6. First, we found that TR-E6 in vitro synthesis in rabbit reticulocyte lysate encodes only one protein with an apparent molecular mass of approximately 38 kDa in SDS-PAGE (Fig. 5A). Moreover, as expected, we found that this protein was not detected by an antibody raised against the hinge domain of TR1 (TR-144) (Fig. 5B). These observations confirmed that the alternative splicing does not result in a frame shift introducing a premature stop codon, but gives rise to a TR1 protein without amino acids 121–198 corresponding to the hinge domain. The hinge domain (D domain), located between the DNA-binding domain (DBD) and the ligand-binding domain, is involved in the nuclear localization of the receptor (18) and plays an important role in ligand and DNA binding activities (19, 20, 21).
|
-E6 protein, we performed transient transfection experiments in QM7 cells using vectors encoding TR-E6/enhanced green fluorescent protein (EGFP) and TR1-EGFP fusion proteins. As expected, TR1 was predominantly expressed in the nucleus (Fig. 5D). In contrast, we observed that TR-E6 was predominantly expressed in the cytoplasm with minor nuclear staining (Fig. 5C). Interestingly, coexpression of TR-E6 with TR1-EGFP induced a major cytoplasmic localization of the nuclear receptor (Fig. 5D). This observation indicates that the expression of the TR variant induced a cytoplasmic sequestration of the full-length receptor.
To test the DNA binding activity of TR-E6, we performed gel mobility shift assays with in vitro synthesized proteins, using a direct repeat (DR) spaced by four nucleotides (DR4). As previously described (22), we found that TR1 binds to a DR4 sequence as monomer and homodimer (Fig. 5E). In contrast, TR-E6 fails to bind this thyroid hormone response element and did not affect the binding pattern of TR1 to a DR4 (Fig. 5E). This observation is in agreement with previous data indicating that efficient TR1 DNA binding requires the D domain (20).
Functional Interactions of TR-E6 with TR, RXR, and RAR
To investigate the possibility that the TR-E6 protein could affect nuclear receptor activity, we performed transient transfection experiments in QM7 avian myoblasts using a CAT (chloramphenicol acetyl transferase) reporter plasmid driven by a palindromic T3-responsive element (inverted repeat: IR0), or a DR spaced by one (DR1) or four nucleotides (DR4). In agreement with previous data, TR1 expression induced a 3-fold stimulation of gene reporter expression through an IR0 in the presence of T3 (306 ± 46%; P < 0.05) (Fig. 6A). Coexpression of TR1 and RXR induced a 2.4-fold stimulation of gene reporter expression through a DR4 thyroid hormone response element in the presence of T3 (237 ± 33%, P < 0.01) (Fig. 6B). Though TR-E6 expression alone does not exert any significant influence by itself (Fig. 6, A and B), its coexpression abrogated TR1 transcriptional activity in the T3 presence through IR0 (306 ± 46% for TR1 vs. 116 ± 6% for TR1/TR-E6; P < 0.05) (Fig. 6A) or DR4 response elements (237 ± 33% for TR1/RXR vs. 124±16% for TR1/RXR/TR-E6; P < 0.01)(Fig. 6B). As previously described, liganded RXR displayed a strong transcriptional activity through a DR1 response element (2118 ± 280%; P < 0.01) (Fig. 6C). Again, TR-E6 was devoid of activity by itself, whereas its coexpression with RXR partially abrogated the ligand-dependent activity of this receptor (2118 ± 280% for RXR vs. 1093 ± 121% for RXR/TR-E6; P < 0.05) (Fig. 6C). In parallel experiments, we found that, in contrast, RAR activity is not affected by TR-E6 (Fig. 6D).
|
-E6 involved in the inhibition of full-length TR1 activity, we used TR-E6 mutants deleted of the DNA binding, activation function 2 (AF-2), or amino-terminal domains (Fig. 7A), in CAT assays experiments. Whereas deletion of the AF-2 domain did not alter the repressive influence of TR-E6 on the transcriptional activity of TR1/RXR through a DR4 (207 ± 23% for TR1/RXR vs. 118 ± 25% for TR1/RXR/TR-E6/AF2; P < 0.01), we found that the two other mutations fully abrogated the ability of this TR variant to repress TR1/RXR activity (Fig. 7B). Western blot experiments using an antiserum raised against the carboxy-terminal domain of TR1 (8) ruled out the possibility that this lack of repressive activity could result from the absence of expression of these two mutants (Fig. 7C). These data indicated that the DNA-binding and amino-terminal domains of TR-E6 are involved in the inhibition of TR1 activity.
|
-E6 Is Involved in the Regulation of Myoblast Differentiation-E6, initially cloned in C2C12 myoblasts, we stably overexpressed this protein in QM7 avian myoblasts using pIRES-TR-E6 vector. These cells were currently used in our team to study myoblast terminal differentiation. Control cells were obtained after transfection of the pIRES-EGFP empty vector. Experiments were carried out on polyclonal populations of transfected cells. TR-E6 overexpression was assessed by Northern blot and Western blot (Fig. 8, A and B).
|
-E6 on myoblast withdrawal from the cell cycle, a major event allowing terminal differentiation to occur, by determining the frequency of cells expressing connectin, an early postmitotic marker (23), at the induction of differentiation occurring at cell confluence. Whereas in proliferation no connectin-positive myoblasts were detected in control cells transfected with the empty vector, TR-E6 overexpression increased the frequency of connectin-expressing cells (Fig. 8C) and induced a precocious appearance of small myotubes [fusion index (see Materials and Methods): 5.8 ± 1.3% vs. 0% for control myoblasts; P < 0.05) (Fig. 8D)]. In addition, this acceleration of myoblast withdrawal from the cell cycle was detected independently of T3 presence (fusion index: 5.8 ± 1.3% vs. 5.8 ± 0.9% with T3) (Fig. 8D).
As previously reported (24), after induction of terminal differentiation, addition of T3 to the culture medium induced a nearly 2-fold stimulation of myoblast differentiation in control cells, reflected by an increased number of myotubes in the culture and a rise in the value of the fusion index (fusion index: 6 ± 0.8% vs. 12 ± 0.3% with T3; P < 0.05) (Fig. 8, C and D). In agreement with its influence on myoblast withdrawal from the cell cycle, TR-E6 overexpression induced up to a 4-fold increase in the value of the fusion index relative to control cells, well reflected by the appearance of numerous thick myotubes (fusion index 29 ± 4.4% vs. 6 ± 0.8%; P < 0.05) (Fig. 8D). As for the induction of irreversible cell cycle arrest, this influence was independent of T3 presence (fusion index 29 ± 4.4% vs. 24 ± 2.6%) (Fig. 8D).
In agreement with these data, we found that, in contrast to control cells, expression of the acetylcholine receptor -subunit (AchR), a valuable marker of myoblast differentiation (24), was detected in TR-E6 overexpressing myoblasts even during the proliferation period (Fig. 9A) (78 ± 25% for TR-E6 vs. 17 ± 11% for control; P < 0.05) (Fig. 9A). During differentiation, AchR expression was strongly stimulated by TR-E6 overexpression in a T3-independent manner (170 ± 21% for TR-E6 vs. 100% for control, without T3; P < 0.05), whereas the hormone increased the level of this transcript in control cells. As previously shown in QM7 myoblasts, T3 did not influence myogenin mRNA and protein levels (Fig. 9, A and B). However, TR-E6 overexpression induced a significant rise in these levels, independently of the T3 presence (Fig. 9, A and B).
|
-E6 Impairs the Ability of TR1 to Repress Chicken MyoD1 (CMD1) Transcriptional Activity-E6 could exert its myogenic influence by interfering with a T3-independent TR mechanism. We have previously found that fundamental mechanisms involved in the regulation of myoblast differentiation by T3 receptors occur through protein-protein interactions (15). In particular, we have recently shown that TR1 could control the duration of myoblast proliferation by inhibiting CMD1 (avian MyoD) transcriptional activity (16). Because the functionality of this mechanism does not require T3, we studied the influence of TR-E6 on this regulation.
In transient transfection experiments using a CAT reporter gene driven by a minimal myogenin promoter including two E boxes, CMD1 induced a 7-fold increase in CAT activity (773 ± 46%; P < 0.01), whereas TR1 and TR-E6 were devoid of any transcriptional activity in the presence or absence of T3 (Fig. 10A). We also confirmed that, in contrast to TR1-E6, coexpression of TR1 significantly impaired CMD1 transcriptional activity, as reported by Daury et al. (16) (773 ± 46% for CMD1 vs. 420 ± 42% for CMD1/TR1; P < 0.05) (Fig. 10A). However, coexpression of TR1-E6 abrogated the ability of the full-length receptor TR1 to inhibit CMD1 activity (773 ± 46% for CMD1 vs. 713±44% for CMD1/TR1/TR-E6) (Fig. 10A).
|
-E6 and its ability to inhibit TR1-induced MyoD repression, we used the TR-E6 mutants previously tested for their capacity to repress TR1 and RXR transcriptional activity. As TR-E6, we found that deletion of the AF-2 domain alters the repressive activity of TR1 on CMD1 transcriptional activity (638 ± 89% for CMD1/TR1 vs. 1051 ± 192% for CMD1/TR1 /TR-E6 vs. 1037 ± 186% for CMD1/TR1 /TR-E6/AF2; P < 0.05) (Fig. 10B). In contrast, the repression of CMD1 activity by TR1 was not affected by the expression of mutants deleted from the DNA-binding or the amino-terminal domain (638 ± 89% for CMD1/TR1 vs. 701 ± 114% for CMD1/TR1/TR-E6/Nter vs. 679 ± 94% for CMD1/TR1 /TR-E6/DBD) (Fig. 10B). In transient transfection experiments using wild-type or mutants of TR-E6, we observed that, as previously shown, addition of T3 to the culture medium induced a stimulation of myoblast differentiation in control cells (Fig. 10C). Moreover, TR-E6 overexpression induced a stronger increase of myoblast differentiation relative to control cells independently of T3 presence. However, in good correlation with data from CAT assays, deletion of DNA binding or amino-terminal domains did not influence differentiation relatively to control cells, whereas overexpression of TR-E6/AF-2 induced a stimulation of myoblast differentiation similar to that recorded for TR-E6 (Fig. 10C). These data demonstrated that the myogenic activity of TR-E6 is related to its ability to abrogate the repression of CMD1 activity by TR1. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
locus, TR-E6, resulting from an alternative splicing with microexon 6b instead of exon 6 and from RNA editing. This transcript, first identified in mouse liver and C2C12 cells, was found in all mouse tissues tested, indicating that this isoform is ubiquitously expressed.
Analysis of the genomic context leading to the generation of TR-E6 through use of a microexon points out striking similarities with that previously described for the chicken cardiac troponin T gene, giving rise to a transcript displaying a sequence corresponding to a microexon of six nucleotides (17). As in the TR locus for microexon 6b, this microexon is surrounded by short G-rich repeats acting as intron-splicing enhancers leading to its inclusion for mRNA synthesis (17). A reasonable speculation leads us to propose that this mechanism is probably involved in the skipping of exon 6 in the case of TR-E6.
In addition, we found that two nucleotides differed when we compared the genomic sequence of microexon 6b and the corresponding sequence in mRNA. These differences were systematically observed after several separate experiments, in C2C12 cells as well as in mouse liver, thus ruling out the possibility of artifacts linked to polymerase amplification fidelity or sequencing errors. Generally, editing processes lead to the substitution of a cytidine by a thymidine in the generated transcripts. Surprisingly, the editing process leading to the TR-E6 transcript involves both deletion of a thymidine and addition of a cytidine downstream. Although the mechanism involved in this process remains unclear, one possibility is that the perfect palindromic structure of the editing area could be an important feature for this microedition event. Although the exact mechanisms involving alternative splicing coupled to RNA editing is far from proven, a TR protein without a hinge domain is generated.
Previous studies have demonstrated that the hinge domain is required in ligand binding activity. In particular, Lin et al. (21) and Miyamoto et al. (20) demonstrated that deletion of the D domain fully abrogated the ligand binding activity of TR1 and TRß1. Moreover, a naturally occurring point mutation in the hinge domain of TRß1was reported to impair T3 binding activity (26, 27). Last, structural study of TR1 indicated that -helix H1 occurring in the hinge domain is required for hormone binding (28). All these data support the importance of the hinge region in hormone binding and strongly suggest that the absence of hinge domain impairs TR-E6 ligand binding.
The hinge domain is also important for TR1 DNA binding. Thus, Miyamoto et al. (20) showed that deletion of this domain fully abrogated the DNA binding activity of TR1, in agreement with structural studies indicating that residues of the hinge domain form a turn after the second helix of the DBD (A box) (29), which participates in formation of heterodimers TR/RXR, and an extended carboxy-terminal -helix, which binds in the minor groove of the DNA to allow binding (30). In line with these data, we found in gel mobility shift assay using in vitro synthesized proteins, that TR-E6 did not bind to DNA. Moreover, the TR variant did not affect the binding pattern of TR1 on a DR4.
In line with the previous observation that the hinge domain is also involved in the nuclear localization of the receptor (18), we found that, in contrast to the full-length receptor TR1 predominantly expressed in the nucleus, TR-E6 displayed a cytoplasmic localization. Interestingly, we also observed that the full-length receptor TR1 could be sequestered in the cytoplasm by TR-E6.
We found that TR-E6 represses the activity of TR1 and RXR. Although the mechanism mediating this inhibitory activity is still unclear, the cytoplasmic sequestration of TR1 by TR-E6 could be a first element. Another possibility could be that TR1-E6 exerts its influence by sequestration of coactivators normally recruited by TR1 or RXR. In this hypothesis, the observation that RAR activity was unaffected by TR-E6 could suggest that the main target of TR mutant is a coactivator that interacts with TR1 and RXR, but not with RAR. Moreover, interaction between TR1-E6 and the target factor may take place in the cytoplasm because TR-E6 is predominantly localized in this compartment, resulting in coactivator sequestration outside the nucleus. Similar hypotheses have been raised by Chassande et al. (13) for the TR1 isoform, which also antagonizes the transcriptional activation induced by TR1 and RXR but not by RAR, without any interaction with these receptors. However, identification of this factor will be necessary for further insight into the mechanisms of inhibition by TR-E6.
To identify the domains of TR-E6 involved in the inhibition of full-length TR1 activity, we used TR-E6 mutants deleted of the DNA binding, AF-2, or amino-terminal domains in CAT assay experiments. We found that deletion of the AF-2 domain did not abrogate the ability of TR-E6 to repress TR1 and RXR transcriptional activity through a DR4. In contrast, in the absence of the DNA-binding or amino-terminal domains, TR-E6 failed to block this transcriptional activity. These data demonstrated that the DNA-binding and amino-terminal domains are involved in the repressing activity of the TR variant.
To study the biological function of TR-E6, we stably overexpressed the protein in QM7 myoblasts. We found that TR-E6 overexpression accelerated myoblast withdrawal from the cell cycle and terminal differentiation, as reflected by fusion index values, and expression of muscle-specific proteins. One striking feature of this influence is that it occurred independently of T3 presence, in agreement with the absence of a hinge domain in the TR variant leading to an impairment to bind T3 (19, 20). In addition, this result rules out the possibility that the mechanism involved in this myogenic influence includes inhibition of the T3-dependent TR1 transcriptional activity demonstrated in this study. More recently, we have also characterized a new mechanism involving TR1 to maintain an optimal duration of the proliferation period, through T3-independent inhibition of MyoD transcriptional activity, a major actor involved in the induction of myoblast differentiation (16). Interestingly, in this study, we found that TR-E6 strongly inhibits this functional interaction, in agreement with an anticipated myoblast withdrawal from the cell cycle resulting from a restoration of MyoD transcriptional activity. Because MyoD is involved in a positive autoregulatory loop controlling its own transcription rate as well as myogenin expression (31), this possibility is also well supported by the observation indicating that myogenin mRNA levels are increased by TR-E6 overexpression.
In this study, we also found a correlation between the capacity of TR-E6 mutants to abrogate the ability of TR1 to inhibit MyoD activity and their capacity to stimulate myoblast differentiation. Indeed, the mutant deleted of AF-2 domain, which has kept the ability to abrogate the repression of MyoD activity by TR1, stimulates differentiation as well as TR-E6. In contrast, the two other mutants deleted of DNA binding or amino-terminal domains, which do not repress the TR1/MyoD functional interaction, did not influence myoblast differentiation. These data demonstrated that the ability of TR-E6 to stimulate differentiation is related to its capacity to repress TR1.
In conclusion, the activity of truncated proteins as potent inhibitors of their full-length counterparts has been established for several transcription factors. The inhibition of TR1 activity by TR-E6 is a novel example of this kind of regulation. In particular, by inhibiting the TR1/CMD1 interaction, TR1-E6 could help to restore MyoD activity and enables the induction of terminal differentiation. If such a mechanism occurs in vivo, this balance between TR1 and TR-E6 could be critical for correct muscle development by controlling the duration of myoblast proliferation. Lastly, the observation that TR-E6 was found to be expressed in numerous tissues such as brain, liver, or heart suggests that this function is not restricted to muscle.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-E6 expression by RT-PCR, total RNA was reverse transcribed using random hexamer primers and AMV reverse transcriptase (Promega Corp., Madison, WI). The resulting cDNA product was subsequently PCR-amplified (30 cycles of 45 sec at 94 C; 45 sec at 58 C; 45 sec at 72 C) with the following primers: 5'-cea156 (TGT GTC GTG TGT GGG GAC AA) and 3'-cea927 (ACT TCC GTG TCA TCC AGG TT). Half-nested PCR was then performed using the internal primers 5'-ceaE6 (GTG GGC ATG GGG ACT AGT) and 3'-cea927 in similar conditions.
Cloning and Construction of Recombinant Plasmids
The mouse TR1 cDNA containing the entire open reading frame and the mouse TR-E6 cDNA containing the shorter fragment, without exon 6, were produced by RT-PCR amplification of total RNAs collected from mouse liver or C2C12 cells using the following primers: 5'cea-4 (GTG AAT GGA ACA GAA GCC AAG CAA GGT) and 3'cea1250 (GCC GCC TGA GGC TTT AGA CTT CCT GAT). The following primers were also used to characterize mTR-E6: 5'-cea107 (AGC AGC ATG TCA GGG TAT ATC CCT), 5'-cea444 (AGG CGA AAG GAG GAG ATG ATT CGC TCA CT) and 5'-cea733 (GAA GAC CAG ATC ATC CTC CT). The amplified cDNAs were cloned with Promega pGEM-T easy vector. The PCR fragments were released by EcoRI digestion and inserted into the EcoRI site of the pSG5 vector, resulting in pSG5-mTR1 and pSG5-mTR-E6 constructs. For stable transfection, mTR-E6 was released by EcoRI digestion and inserted into the EcoRI site of the pIRES-EGFP vector, resulting in the pIRES-mTR-E6 construct. To study the subcellular localization of TR1 and TR-E6 proteins, we fused the cDNA sequence of the proteins to the EGFP gene. The recombinant DNA molecules were generated by PCR and inserted in the EcoRI site of the pSG5 vector (pSG5-TR1/EGFP, pSG5-TR-E6/EGFP). The TR-E6 mutants were generated by PCR using the following primer: 1) TR-E6/Nter: 5'-cea Nter (AGC AGC ATG TCA GGG TAT ATC CCT) and 3'-cea1250 (GCC GCC TGA GGC TTT AGA CTT CCT GAT); 2) TR-E6/DBD: 5'-cea-4 (GTG AAT GGA ACA GAA GCC AAG CAA GGT) and 3'-cea DBD (AGT CCC CAT GCC CTG CTC GTC TTT GTC AGG TAA CT) for the first fragment, 5'-cea DBD (AAAGAC GAG CAG GGC ATG GGG ACT AGT CCC CAG TC) and 3'-cea1250 for the second fragment; 3) TR-E6/AF-2: 5'-cea-4 and 3'-cea AF-2 (GCT TTA GAG CGG GGG GAA GAG TTC). For TR-E6/DBD, a second step of PCR using the two mutated fragments was performed. The amplified cDNAs were cloned with Promega pGEM-T easy vector. The PCR fragments were released by EcoRI digestion and inserted into the EcoRI site of the pSG5 vector, resulting in pSG5-TR-E6/Nter, pSG5-TR-E6/DBD, and pSG5-TR-E6/AF-2 constructs.
DNA Sequencing and Identification of a TR Microexon
Sequencing of the mouse TR-E6 cDNA from liver and C2C12 cells was performed with the Big Dye terminator kit from PerkinElmer Life Sciences (Norwalk, CT) following the supplier’s protocol. Genomic cDNA clones containing the microexon 6 were produced by amplification of mouse liver and C2C12 cell DNA (30 cycles of 45 sec at 94 C, 45 sec at 58 C, 1 min at 72 C) using the following primers: 5'-ceaE6 (ATG ATT CGA AGC GGG TGG CCA AA) and 3'-ceaE7 (CTC GGA GAA CAT GGG CAG TTT T). The genomic cDNA fragment was sequenced as previously described.
Cell Cultures and Transcriptional Activation Assays
Quail myoblasts of the QM7 cell line (33) were seeded at a plating density of 7000 cells/cm2. They were grown in Earle’s 199 medium supplemented with 0.2% tryptose phosphate broth, penicillin (100 IU/ml), and 10% of T3-depleted fetal calf serum. The serum was T3 depleted as previously described (34). Terminal differentiation was induced at cell confluence by lowering the medium’s fetal calf serum concentration to 0.5%.
Transient transfection assays were performed in QM7 cells using expression plasmids, CAT reporter, and pRSV-ß-galactosidase expression vector, as described in Cassar-Malek et al. (15). The following expression vectors were used: pSG5, pSG5-TR1, pSG5-TR-E6, pSG5-RXR, pSG5-RAR, pSG5-TR-E6/Nter, pSG5-TR-E6/DBD, pSG5-TR-E6/AF-2, and pRSV-CMD1. When indicated, 10–8 M T3, 5.10–8 M 9-cis-retinoic acid or 10–7 M all-trans retinoic acid was added to the culture medium. After normalization according to ß-galactosidase activity, all results were expressed as a percentage of the control value. Data are presented as the mean ± SEM of four separate experiments.
Cytoimmunofluorescence Studies
Myoblast differentiation was assessed by morphological changes and accumulation of connectin, a muscle-specific marker expressed just after myoblast withdrawal from the cell cycle (23). After methanol fixation and appropriate washings, cells were stained with a monoclonal antibody raised against connectin (kindly provided by Dr. F. Pons, Institut National de la Santé et de la Recherche Médicale, Montpellier, France) and a fluorescein-conjugated antibody raised against mouse Igs. Nuclei were stained with Hoechst 33258 (1 µg/ml). To quantify terminal differentiation, the fusion index was calculated by counting the total number of nuclei and the number of nuclei incorporated in myotubes (fusion index = number of nuclei in myotubes x 100/total number of nuclei).
The intracellular localization of mouse TR1 or TR-E6 was assessed in transient transfection experiments using QM7 cells and the fusion protein TR1/EGFP or TR-E6/EGFP.
Preparation of Protein Extracts and in Vitro Protein Synthesis
QM7 cells plated in coated dishes (7000 cells/cm2) were harvested in 1.5 ml of PBS and then centrifuged for 5 min at 12,000 x g. The pellet was resuspended and lysed in 100 µl Tris-Nonidet P-40 (0.7% Nonidet P-40; 10 mM Tris, pH 7.8). Proteins were synthesized in vitro using the rabbit reticulocyte lysate Transcription/Translation kit TNT (Promega) according to the manufacturer’s instructions. Reactions were performed using the pGEM-mTR1 and pGEM-mTR-E6 vectors.
Gel EMSAs
Gel EMSA experiments were performed according to Wrutniak et al. (8) using a 32P-labeled DR4 oligonucleotide as probe (TCAGGTCACAGGAGGTCA) and cold mTR1 and/or mTR-E6 synthesized in vitro using the rabbit reticulocyte lysate. TR binds to specific target sequences termed thyroid-response element (TRE). TRE consists of two half-sites. Bold letters indicate the two half-sites (5'-AGGTCA-3' consensus sequence).
Western Blot Analysis
Total cellular extracts (50 µg) were electrophoresed onto 10% SDS-PAGE gels and blotted onto polyvinylidine difluoride membranes. The presence of TR-E6 was assessed using RHTII antiserum (8). To detect the full-length receptor TR1, we also used a polyclonal antibody raised against the hinge domain of TR1 (TR-144, Interchim, Montluçon, France). The presence of myogenin was detected using a polyclonal antibody raised against chicken myogenin (kindly provided by Dr. B. M. Paterson, Laboratory of Biochemistry, National Cancer Institute, National Institutes of Health, Bethesda, MD) (35). Signals were revealed using a chemifluorescence detection procedure (enhanced chemifluorescence kit, Amersham Biosciences, Orsay, France) and analyzed with a PhosphorImager (Molecular Dynamics, Inc., Sunnyvale, CA).
Northern Blot Analysis
Total RNA (20 µg) was loaded in each lane of 1% formaldehyde-agarose gel. After migration, RNAs were transferred onto a nylon membrane. Membranes were hybridized with specific cDNA probes labeled using the megaprime DNA labeling system (Amersham Biosciences). A S26 ribosomal cDNA probe was used to provide an invariant control as previously described (36). Quantification was performed using a STORM PhosphorImager (Molecular Dynamics).
Statistical analyses were performed using Student’s paired t test.
|
FOOTNOTES |
|---|
First Published Online December 1, 2005
Abbreviations: AchR, Acetylcholine receptor -subunit; AF-2, activation function 2; CAT, chloramphenicol acetyltransferase; DBD, DNA-binding domain; DR, direct repeat; EGFP, enhanced green fluorescent protein; IR, inverted repeat; RAR, retinoic acid receptor; RXR, retinoic X receptor; TR, thyroid hormone receptor.
Received for publication February 4, 2005. Accepted for publication November 21, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1 is located in the mitochondrial matrix of rat liver. J Biol Chem 270:16347–16354 and ß thyroid hormone receptor genes. EMBO J 9:1519–1528[Medline]1 plays a direct role in regulation of mitochondrial RNA synthesis. Mol Cell Biol 19:7913–7924 locus that encode inhibitors of retinoic acid receptor- and triiodothyronine receptor activities. Mol Endocrinol 11:1278–1290
