help button home button Endocrine Society Molecular Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Molecular Endocrinology, doi:10.1210/me.2003-0346
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
18/7/1631    most recent
Author Manuscript (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow NURSA Molecule Pages Link
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Frankton, S.
Right arrow Articles by Williams, G. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Frankton, S.
Right arrow Articles by Williams, G. R.
Molecular Endocrinology 18 (7): 1631-1642
Copyright © 2004 by The Endocrine Society

Multiple Messenger Ribonucleic Acid Variants Regulate Cell-Specific Expression of Human Thyroid Hormone Receptor ß1

Sarah Frankton, Clare B. Harvey, Lisa M. Gleason, Abdessamad Fadel and Graham R. Williams

Molecular Endocrinology Group, Division of Medicine and Medical Research Council Clinical Sciences Centre, Faculty of Medicine, Imperial College London London W12 0NN, United Kingdom

Address all correspondence and requests for reprints to: Dr. Graham R. Williams, Molecular Endocrinology Group, Medical Research Council Clinical Sciences Centre, 5th Floor Clinical Research Building, Hammersmith Hospital, Du Cane Road, London W12 0NN, United Kingdom. E-mail: graham.williams{at}imperial.ac.uk.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 Note Added in Proof
 REFERENCES
 
Thyroid hormones are essential for development, growth, and metabolism and act via T3 receptors (TR) {alpha} and ß. The THRA and THRB genes have discrete physiological roles but their mRNAs are expressed widely in overlapping patterns. There is poor correlation between TR mRNA and protein, indicating that expression may be regulated by posttranscriptional mechanisms. Differences in the relative levels of expressed TR{alpha} and ß proteins have been suggested to modulate tissue T3 responsiveness. We determined the structure of the human THRB gene, cloned seven alternately spliced 5'-untranslated region (5'-UTR) TRß1 mRNAs, and identified five polyadenylation position elements in the 3'-UTR. At least six TRß1 mRNAs between 1.35 and 7.5 kb in length were expressed in discrete temporospatial patterns in fetal and adult human tissues. The 5'-UTRs contained up to seven upstream short open reading frames, which did not influence the structure of the TRß1 protein. In transfection studies, 5'-UTRs exerted cell-specific effects on mRNA expression but consistently reduced protein expression. Furthermore, each 5'-UTR strongly inhibited translation in vitro. Thus, developmental and tissue-specific expression of human thyroid hormone receptor ß1 5'-UTR mRNAs may regulate T3-responsiveness in target tissues by modulating TRß protein translation and thereby controlling the ratio of expressed TR{alpha} and -ß proteins.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 Note Added in Proof
 REFERENCES
 
THYROID HORMONES REGULATE crucial events during growth and development, and T3 is an important homeostatic regulator in adults. These actions are mediated by nuclear thyroid hormone receptors (TRs), which regulate T3-target gene transcription and are encoded by THRA and THRB genes (1). Both genes encode variant mRNAs that are expressed in overlapping temporospatial patterns (2, 3, 4, 5), although their levels correlate poorly with TR protein concentrations (1, 6). Analysis of TR-null mice has shown that TR{alpha} is required for postnatal growth, skeletal maturation, and cardiac function, whereas TRß is essential for development of the cochlea and retina and controls the hypothalamic-pituitary-thyroid feedback axis (7). In humans, dominant-negative mutant TRß proteins cause resistance to thyroid hormone (RTH) (8). Variable clinical features result from hypothyroidism in tissues such as pituitary and brain, with thyrotoxic signs evident in others including heart. Analysis of mice with a TRß RTH mutation (9) has shown that the abundance of mutant TRß in a particular tissue determines its phenotype. In liver and pituitary, where TRß levels are high, there is interference with T3-target gene expression, and tissue hypothyroidism results (10). In heart and bone, in which low mutant TRß levels are unable to interfere with TR{alpha}, tissue thyrotoxicosis reflects the high T3 and T4 concentrations in this model of RTH (10, 11). Recent studies have also shown that TRß1 undergoes diurnal regulation and is zonally expressed in liver (12), suggesting that T3 responses are fine tuned in individual organs by regulation of TRß.

The human THRB gene encodes ß1 and ß2 N-terminal variants, and the region containing common exons 5–10 has been characterized in detail (13). Specific exons encoding TRß1 and ß2 N-terminal domains have been identified, but the genomic structure of the 5'-region is less clear (Fig. 1Go) and has been complicated by nomenclature changes (13, 14, 15, 16). Surprisingly little is know about human TRß gene expression. The presence of N-terminal truncated TRß isoforms in Xenopus, chicken, and rat (3, 5, 17) and the demonstration of multiple 5'-untranslated region (5'-UTR) variants in Xenopus (2, 4) suggests that human THRB may be more complex than currently appreciated. TRß1 transcripts ranging between 1.0 and 10.0 kb have been identified by Northern blotting in a few studies (15, 18, 19, 20, 21, 22), but their structures and functions have not been determined. Transcript heterogeneity has been speculated to result from expression of mRNA variants, including heteronuclear RNA, or to reflect mRNA degradation or cross-hybridization of TRß probes with TR{alpha} mRNAs. We hypothesized that uncharacterized noncoding regions of human THRB contain sequences that regulate tissue T3 responsiveness by influencing TRß mRNA and protein expression. Thus, we determined the genomic structure of human THRB and characterized the expression, structure, and function of several new TRß 5'-UTR variants.



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 1. Structure of Human THRB

The four original TRß1-specific exons (boxes 1–4), single ß2-specific exon, and six exons common to ß1 and ß2 (boxes 5–10) are shown. Intron and exon sizes are indicated. Dotted introns between exons 1–3 indicate unknown sizes and locations. The size of exon 10 has not been determined. ATG and TAG in exons 3 and 10 delineate the ß1 coding region. Exons 1 and 2 and the first 44 bp of exon 3 make up the 5'-UTR; the remainder of exon 3 and exons 4–10 encode ß1 protein. Exons 5 and 6 encode the DNA-binding domain. Amino acid numbers and T3 binding domain are shown. In TRß2, exons 1–4 of ß1 are replaced by ß2 exon 1, which is transcribed from an alternate promoter.

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 Note Added in Proof
 REFERENCES
 
Diversity in the TRß1 5'-UTR Results from Alternate mRNA Splicing
Seven TRß1 5'-UTR mRNAs were identified in 5'-rapid amplification of cDNA ends (RACE) studies performed using cDNA from a variety of tissues. 1 No TRß2 variants or cDNAs similar to rat TRß3 and {Delta}ß3 were identified. It is important to note that cloning of distinct TRß1 5'-UTR variants from individual tissues by 5'-RACE does not necessarily imply their expression is restricted to a specific tissue (see Northern Blotting Studies below). cDNA clones A and B were isolated from brain, C and F were isolated from fetus, D came from placenta, E was derived from kidney, and G came from testis (Fig. 2Go). Five of these seven clones (A, B, D, F, and G) contained distinct arrangements of spliced exons, and two pairs of cDNAs (B and C and D and E) contained identical exons and differed only at their 5'-ends. Analysis of the structures of cloned TRß1 5'-UTR cDNAs indicated they result from alternative splicing of eight exons plus the first 44 bp of exon 3 (Fig. 3Go). The previously designated exon 1 comprised three exons that we rename exons 1a, 1b, and 1c. Two new exons, 1d and 1e, were identified, and we rename the previous exon 2 as exon 2a because two further exons, 2b and 2c, were identified. Clone A was identical to the previously published TRß1 5'-UTR (13, 15, 16) except it was 33 bp longer at the 5'-end. The donor and acceptor splice sites at intron/exon boundaries in each of clones A–G consisted of consensus sequences when analyzed using NetGene 2 v2.1 (23).



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 2. Ethidium Bromide-Stained Agarose Gel Electrophoretograms Showing Original cDNA Clones Obtained by 5'-RACE

The 613-bp clone A and 631-bp clone B were obtained from an adult brain cDNA library. The two bands labeled with asterisks represent 600-bp and 550-bp cDNAs obtained from brain, which were identical in sequence to clone A but contained 5'-ends that were 13 bp and 63 bp shorter than clone A. The 741-bp clone F and 597-bp clone C were obtained from a 7- to 8-wk-old whole fetus library. The 716-bp clone D and 790-bp clone G were from placenta and adult testis libraries, respectively, and the 707-bp clone E was from an adult kidney library. All unmarked 5'-RACE PCR products obtained from brain and fetus libraries were sequenced and found to be derived from clones that were unrelated to TRß. DNA marker (1.0 kb) ladders (bp) are shown on the right of each gel, and sizes of two marker bands are shown for each (note the left hand gel shows smaller 1.0-kb ladder fragments).

 


View larger version (22K):
[in this window]
[in a new window]
  		Fig. 3. Structure of Human TRß1 5'-UTR Experimentally Defined by 5'-RACE

The structure of the TRß1 5'-UTR consists of eight alternately spliced exons plus the first 44 bp of exon 3. Shading indicates how the newly defined 5'-UTR relates to the structure shown in Fig. 1Go. Structures of cloned 5'-UTR cDNAs A–G (see Fig. 2Go) and the tissues from which they originated are indicated; solid lines represent exons, dotted lines show splicing events, and the lengths of exon 1a or 1b at the 5'-end of each clone are indicated.

 
Structure of TRß1 5'-UTRs
The 5'-UTRs differed between 197 (clone C) and 390 bp (clone G) in length. The TRß1 open reading frame (ORF) initiates from two AUG codons (14) in exon 3, which were present in all mRNAs indicating that 5'-UTR variation does not influence protein structure. Sequence analysis revealed multiple upstream (u) AUGs and associated ORFs in each 5'-UTR (Fig. 4Go). There were four uAUGs in clones A, B, and C, five in D and E, six in G, and seven in F. The most favorable consensus sequence for translation initiation is (A/G)CCAUGG (24, 25). The second AUG of the TRß1 ORF lies within a reasonable context, but the first AUG is not favorable. Similarly, there was variation in the context of uAUGs, although the first uAUG in exon 1d of clone G was situated optimally. The uAUGs were associated with uORFs as follows; four in clones A, B, and C, five in D and E, six in G, and seven in F. Some ORFs overlapped (in A, B, C, F, and G), and they varied in length from two codons in all clones to one of 174 codons in clone F. Analysis of 5'-UTRs using Mfold v3.1 (26) revealed that clones G ({Delta}G = –129.0 kcal/mol) and A ({Delta}G = –68.6 kcal/mol) have the most and least stable secondary structures. The most stable hairpin loop identified was in clone C ({Delta}G = –26.0 kcal/mol). Predicted proteins arising from uORFs were analyzed using PROSITE (27). The 174-codon uORF in clone F shared 33% identity with amino acids 241–299 of the rat xanthine dehydrogenase/oxidase protein (GenBank accession no. P22985) but did not possess features of known functional domains. No homologies with protein or functional domain databases were identified for other uORFs.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 4. Structures of TRß1 5'-UTR cDNA Clones Obtained by 5'-RACE

Variants A–G are shown with exon structures and start and end nucleotides. Exons 2a and 3 are shaded to correspond with Figs. 1Go and 3Go. Each uAUG is indicated by a dotted vertical line, and positions of U nucleotides are shown below. Small arrows represent uORFs and the large arrow (coding sequence, cds) corresponds to the start of the TRß1 coding sequence, for which the two initiator codons are marked by asterisks. Small vertical lines within uORFs show positions of uAUGs, and numbered arrowheads represent ORF terminator positions.

 
Additional 5'-UTR Splicing Revealed by Database Searches
Clones A–G and exons 1d, 1e, 2b, and 2c were used to search expressed sequence tag (EST) databases. Ten further 5'-UTR mRNA variants were identified (Fig. 5Go). These confirmed our data by identifying exons 1a, 1b, 1c, 2a, 2b, and 2c and the use of identical splice sites, but no ESTs contained exon 1d or 1e. Our results were also extended. AW298109 contained 191 bp of exon 1a, extending the transcription start site by 23 bp relative to clone F and the identical clones BU157865 and BQ879477 contained 208 bp of exon 1a, thereby extending the exon 1a transcription start site by 40 bp. Three clones [M26747 (also designated X04707), AK096628, and AL713041] contained both exons 1b and 1c but originated from different transcription start sites in exon 1a. These findings were in contrast to those observed in brain cDNA A, in which transcription was initiated from exon 1b. In AK096628, exon 2c also lacked nucleotides 1–47, suggesting that additional alternative splicing events occur in the 5'-UTR that were not identified by our 5'-RACE experiments. Analysis with NetGene 2 v2.1 (23) identified a consensus splice site in exon 2c to support this possibility. Three ESTs (AL13041, R45891, and AW298109) were incomplete at their 3'-ends and terminated in either exon 4 or exon 2c.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 5. Structures of ESTs Identified by Interrogating Databases with TRß1 5'-UTR cDNA Sequences

Shading and depiction of exons, splicing events, and 5'-ends of each EST are as in Fig. 3Go. GenBank accession nos. for each EST are shown. The 3'-ends of R45891, AL713014, and AW298109 terminate in exon 2c or exon 4; the 3'-ends of AI650763, BF854291, and BE673162 terminate after exon 4 with a poly A sequence, but the 3'-ends of the remaining ESTs have not been identified.

 
Differing exon arrangements in ESTs also extended the complexity of 5'-UTR splicing. In BU157865 and BQ879477, splicing of exon 2c to exon 4 predicts translation of a TRß1 protein lacking the first 31 amino acids that originates from an in-frame AUG in exon 4. However, this cDNA was cloned from a pancreatic cancer cell line, raising the possibility that it results from aberrant splicing. Expression of aberrantly spliced TRß1 mRNAs that alter protein structure has been documented in human breast cancers and proposed as a contributor to malignant transformation (28). Finally, three ESTs (AI650763, BF854291, and BE673162) contained intron 4 sequence and terminated with poly A tails. These unusual transcripts might represent nonsense mRNAs that contribute to the overall regulation of TRß1 expression. Alternatively, they may be cloning artifacts that are not expressed in vivo or result from aberrant splicing.

Multiple Sites for Transcription Initiation in TRß1 5'-UTR Exons
The data summarized in Figs. 3–5GoGoGo suggest that TRß1 mRNAs may originate from multiple transcription initiation sites in exons 1a, 1b, and 2a. Nevertheless, this region of the THRB gene is GC rich, and it is possible that the apparent diversity of transcription initiation results from artifacts of RNA degradation or incomplete reverse transcription during cDNA synthesis in the generation of cDNA libraries. To address this, we employed the alternative experimental approach of inverse RT-PCR using either total RNA or poly A mRNA prepared from three tissues (brain, fetal liver, and placenta) as templates to map 5'-mRNA start sites in exon 1b (Fig. 6Go). Ten individual exon 1b mRNA start sites were identified in these studies. Two originating from brain or placenta were located between nucleotides 93–90 upstream of the boundary between exons 1b and 2a and were consistent with EST R45891 expressed in the ciliary body of the eye (Fig. 5Go). Three start sites identified in brain or fetal liver were clustered between nucleotides 86–82 and are consistent with clone C obtained from whole fetus (Fig. 3Go). An additional three start sites were clustered between nucleotides 41–39 and obtained from brain or fetal liver RNA, and two additional start sites at positions 58 and 32 were also identified in fetal liver. Together with data in Figs. 3Go and 5Go, this independent experimental evidence is highly suggestive that TRß1 mRNAs originate from multiple transcription initiation sites, at least in exon 1b. Nevertheless, it is always difficult to be absolutely certain that mRNA transcription start sites are truly authentic, especially in regions that are GC rich, such as the TRß1 5'-UTR.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 6. Independent Identification of 5'-UTR mRNA Start Sites in Exon 1b by Inverse RT-PCR

In panel (i) the primer in exon 2a used for reverse transcription is shown along with the cDNA product that contains exons 2a and 1b. A BstNI restriction site is shown in exon 1b (ii), and the positions of forward and reverse PCR primers are shown in relation to the BstN1 site and the boundary between exons 1b and 2a. Reverse transcribed cDNA was blunt ended using T4 DNA polymerase, ligated, and then digested with BstNI to generate the cDNA construct in (iii), which was used as a template for PCR amplification with primers FOR and REV. cDNA products were subcloned and sequenced to determine the TRß1 5'-UTR mRNA start sites in exon 1b. Ten identified cDNA sequences of differing lengths are shown in (iv), and these were obtained from human fetal liver (triangles), adult brain (circles), and placenta (star) RNA. The locations of 5'-UTR mRNA start sites are shown above the symbols and indicate the nucleotide position in exon 1b upstream of the boundary between exons 1b and 2a.

 
TRß 3'-UTR
Analysis of ESTs also provided information about the 3'-UTR. AK096628 contains 677 nucleotides after the stop codon, compared with the 28 bp up to the poly A tail in the original TRß1 cDNA (15). TRß1 sequences were used to interrogate the Unigene database. Overlapping cDNA clones extended the 3'-UTR by 4 kb and agreed with sequence obtained from the human chromosome 3 genomic contig clones RP11–81I24 and RP11–118N13, which include the genomic region containing the 3'-UTR, thus suggesting the 3'-UTR is derived from a single exon. We next interrogated EST databases with nucleotides 176808–183263 of AC093927, which includes the overlapping ESTs in exon 10 and extends 6456 bp downstream from the TRß stop codon. Seven EST clusters that terminate 1, 1.5, 2.4, 3, 4, 4.7, and 5.8 kb were identified by 158 BLAST (basic local alignment and search tool) hits after the stop codon (Fig. 7Go). Five 3'-end-processing poly (A) signal sequence position elements (PEs) were identified. Two consisted of a consensus AAUAAA sequence, and three contained the only common variant AUUAAA, which possesses 70% processing activity (29). Three PEs were located immediately 3' to clusters 5, 6, and 7. One mapped within cluster 3 and the other mapped to a region further 3' to cluster 7. This variation predicts that full-length TRß1 mRNAs with approximate sizes of 2.4, 4, 6, and 6.5 kb result from use of alternative 3'-UTR PEs and supports a previous study in which a longer, undefined form of the TRß1 3'-UTR was identified (13). The findings are also compatible with numerous TRß1 mRNAs identified in this study and previously (15, 18, 19, 20, 21, 22).



View larger version (37K):
[in this window]
[in a new window]
  		Fig. 7. TRß1 3'-UTR

BLAST search of EST databases with nucleotides 176808–183263 of AC093927. BLAST hits (158) are represented in relation to the TRß1 3'-UTR by horizontal bars, each representing an individual EST. These ESTs fall into seven main clusters that terminate at approximately 1, 1.5, 2.4, 3, 4, 4.7, and 5.8 kb after the TRß1 stop codon. The locations of five poly (A) position elements relative to the EST clusters are also shown. Their sequences and positions following the TRß1 stop codon are indicated. A representation of exon 10 is shown; gray shading indicates defined sequence and hatching demonstrates the extent of exon 10 deduced from overlapping ESTs.

 
Distribution of TRß1 mRNAs
TRß1 expression was assessed in 16 adult and four fetal tissues by Northern blotting. RNA was probed with the new 5'-UTR exons (1d, 1e, 2b, and 2c), exon 1c (only in clone A), 2a (all clones), and exon 7 (TRß coding), but hybridization with exon 1c failed to produce a signal. TRß1 mRNAs were expressed widely, although levels in adult lung, leukocyte, colon, thymus, and spleen were low (Fig. 8Go). At least six transcripts, ranging between 1.35 and 7.5 kb, were detected although transcripts of 2.4 and 5.0 kb were most widely expressed. Exon 1d, present in adult testis clone G, was expressed mainly as 2.4 and 5.0 kb mRNAs, but additional 2.0- and 3.0-kb mRNAs were seen in adult pancreas and fetal liver, respectively. Exon 1e, present in fetus clone F, was expressed at low levels in adult (particularly pancreas, skeletal muscle, liver, and whole brain) and fetal tissues mainly as 2.4- and 5.0-kb mRNAs but also as an additional 3.0-kb mRNA in fetal liver. Exon 2b, present in adult testis clone G, was expressed in all tissues as 2.4- and 5.0-kb transcripts, and a 3.0-kb mRNA was evident in placenta and fetal kidney, liver, and lung. In addition, the exon 2b probe revealed 2.0-kb mRNAs in skeletal muscle and fetal liver, and other mRNAs between 3.0 kb and 4.4 kb were seen in adult brain, testis, and fetal brain. Exon 2c, present in placenta and adult kidney clones D and E, was expressed mainly as a 2.4-kb mRNA in most adult tissues and fetal brain. An additional 7.5-kb mRNA was seen predominantly in skeletal muscle, small intestine, ovary, and prostate, and mRNAs between 1.35 and 2.6 kb were variably expressed in adult pancreas, skeletal muscle, lung, placenta, and ovary. Exon 2a, present in all 5'-UTRs, hybridized to a 5.0-kb mRNA in all tissues. Additional smaller transcripts were expressed in restricted tissues, and a 7.0-kb mRNA was evident in skeletal muscle. However, the 2.4-kb mRNA that hybridized to other probes in many tissues did not hybridize with exon 2a, indicating the 2.4-kb mRNA results from alternative splicing that skips this exon. The exon 7 probe also hybridized to a 5.0-kb transcript in most tissues, and hybridization to the widely expressed 2.4-kb mRNA was also seen at low levels. Smaller transcripts around 1.35 kb were expressed in adult pancreas and skeletal muscle and in fetal liver. The 3.0-kb mRNA that contained exons 1d, 1e, 2b, and 2a in varying patterns in placenta and fetal tissues was not observed in blots hybridized to exon 7, indicating that this transcript is unlikely to encode protein. Thus, multiple TRß1 mRNAs are expressed in specific patterns in human fetal and adult tissues.



View larger version (105K):
[in this window]
[in a new window]
  		Fig. 8. Distribution of TRß1 mRNAs

Northern blots of RNA from adult (left two blots) and fetal (right) tissues hybridized to exon 1d, 1e, 2a, 2b, 2c, and 7 and ß-actin probes. Arrowheads on the left of each panel indicate positions of the widely expressed 2.4- and 5.0-kb mRNAs. Stars to the right of panels indicate a 3.0-kb mRNA expressed in fetal tissues (exons 1d, 1e, 2b, and 2a) and placenta (exon 2b). Arrows to the left of exon 1d, 2b, and 2c blots indicate a 2.0-kb mRNA expressed in pancreas (exon 1d), in skeletal muscle, liver, heart, and fetal liver (exon 2b), and in kidney, skeletal muscle, and liver (exon 2c). The cross to the left of the exon 2c panels indicates a 7.5-kb mRNA seen in skeletal muscle, small intestine, ovary, and prostate. The x symbol to the left of the exon 2a panels indicates a 7.0-kb transcript in skeletal muscle. The circles to the left of the exon 2c, 2a, and 7 blots indicate smaller mRNAs less than 2.0 kb that are expressed in pancreas (exon 2c), in small intestine, prostate, and spleen (exon 2a) and in pancreas, skeletal muscle, and fetal liver (exon 7). Other mRNAs that are less distinct and expressed at low levels in some tissues (e.g. a 3.2-kb exon 2b transcript in whole brain and exon 2a and 2c transcripts of 4.0–4.4 kb in skeletal muscle) are also evident.

 
TRß1 5'-UTRs Regulate Gene Expression
The effects of each 5'-UTR on mRNA and protein expression were determined in monkey kidney COS-7 and human placental JEG-3 choriocarcinoma cells (Fig. 9Go). These immortalized cell lines can be transfected efficiently and reproducibly, and they originate from organs that express TR. In COS-7 cells, 5'-UTRs A and G did not alter luciferase mRNA, whereas clones B–F increased mRNA expression to a maximum of a 2.85-fold (5'-UTR D). Despite this, luciferase activity was reduced by each 5'-UTR, the least effect being seen for construct D. There was no correlation between the effects of 5'-UTRs on luciferase mRNA and activity (y = –0.827x + 219.181; r = 0.262, P > 0.05) suggesting that they exert independent effects on mRNA and protein expression in COS-7 cells. In JEG-3 cells, construct A reduced luciferase mRNA by 87% but did not alter activity. In contrast, 5'-UTRs B–G did not alter mRNA but reduced activity by 46–83%, the least and greatest effects being seen for 5'-UTRs D and F, respectively. There was a correlation between the effects of 5'-UTRs on luciferase mRNA and activity (y = –0.890x + 141.157; r = 0.582, P < 0.05) suggesting that they generally do not influence mRNA but inhibit protein expression in JEG-3 cells. Furthermore, correlation between the effects of 5'-UTRs on luciferase activity in COS-7 and JEG-3 cells (y = 0.448x + 30.688, r = 0.563, P < 0.05) supports the conclusion that they inhibit protein translation in both cell types. However, there was no correlation between the effects of 5'-UTRs on luciferase mRNA in each cell type (y = 0.304x + 145.590; r = 0.185; P > 0.05) suggesting that their effects on mRNA expression are cell specific. Thus, TRß1 5'-UTR variation results in cell-specific control of protein expression that is due to differential effects on mRNA expression and protein translatability. T3 (100 nM) did not influence the effect of any 5'-UTR on luciferase activity in either cell type (data not shown), indicating that the actions of TRß1 5'-UTRs are independent of hormone.



View larger version (55K):
[in this window]
[in a new window]
  		Fig. 9. Functional Analysis of 5'-UTRs

A, 5'-UTR pGL3 constructs (A–G) were transfected with a Renilla internal control into African green monkey kidney COS-7 cells. Luciferase mRNA expressed by each construct (normalized to Renilla mRNA) is shown relative to expression from a control plasmid that lacked a 5'-UTR (left). Luciferase activity for each construct (normalized to Renilla activity) is also shown (right). B, Luciferase mRNA and activity for each 5'-UTR in human placental JEG-3 choriocarcinoma cells. All results are shown from three independent experiments in the absence of T3 as mean % mRNA or activity ± SEM. Data were analyzed by ANOVA followed by Dunnett’s multiple comparison test; **, P < 0.01; *, P < 0.05 vs. control. LUC, Luciferase.

 
TRß1 5'-UTRs Inhibit Translation Directly
We next tested the effect of each 5'-UTR on translation efficiency (Fig. 10Go). 5'-UTR constructs were transcribed, and equal amounts of RNA were translated to generate protein that was quantified by luciferase assay. Each 5'-UTR inhibited luciferase activity by between 79% (5'-UTR A) and 98% (5'-UTR D), indicating direct inhibition of protein translation by a mechanism that is independent of mRNA concentration.



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 10. In Vitro Translation of 5'-UTRs

Luciferase activities (±SEM) from 5'-UTRs A–G after in vitro transcription-translation are shown relative to a control plasmid that lacks a 5'-UTR (Con). Data from three experiments (two concentrations of input RNA per experiment) were analyzed by ANOVA followed by Dunnett’s multiple comparison test; **, P < 0.01 vs. control. LUC, Luciferase.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 Note Added in Proof
 REFERENCES
 
We have characterized the human THRB gene. Four new exons and two splice sites in exon 1 were identified in the TRß1 5'-UTR. Along with exon 2a and the first 44 bp of exon 3, these exons were alternately spliced to form at least seven 5'-UTRs isolated from various tissues. Interrogation of EST databases revealed additional splicing in exon 2c, indicating that it may consist of 48- and 98-bp exons. No 5'-UTRs alter TRß1 protein structure as they all include exon 3, which contains the start of the TRß1 ORF. Thus, 5'-UTR variation of human TRß1 is similar to the Xenopus TR{alpha} and -ß (2, 17), human and mouse estrogen receptor-{alpha} (30, 31), and rat glucocorticoid receptor genes (32). However, the alternative use of eight 5'-UTR exons, numerous TRß1 transcription start sites, discrete TRß1 and ß2 coding exons that contain the start of the TRß1 and ß2 ORFs, and two AUG translation initiation sites for TRß1 contributes to variant expression from the THRB 5'-region and is combined with use of up to five alternative PEs in the 3'-region to give the human THRB gene a unique level of complexity.

At least six TRß1 mRNAs were expressed in tissue-specific patterns. Transcripts were expressed in complex patterns of varying intensity in Northern blots hybridized with six separate and specific probes. Differing intensities of hybridization of specific bands between blots (e.g. the low level of expression of 2.4-kb mRNA in exon 7 blots compared with levels of its expression in exon 2b blots) probably reflects differences in the efficiencies of hybridization and labeling of the individual exon-specific probes. Most of the expressed transcripts hybridized to an exon 7 probe, indicating their capacity to encode protein, although the smallest mRNAs around 1.35 kb may not have this capability as the TRß1 ORF spans 1383 bp. In addition, a 3.0-kb transcript in placenta and fetal tissues did not hybridize to exon 7, indicating that it may represent a noncoding mRNA. Comparison of 5'-UTR structures with their patterns of expression highlights the discrepancy that exon 2a (present in each 5'-UTR) did not hybridize to all transcripts (notably the widely expressed 2.4-kb mRNA) identified by exon 1d, 1e, 2b, 2c, and 7 probes. Similarly, the 1.35-kb mRNA identified by the exon 2a probe in small intestine and spleen did not hybridize with any other probe. Thus, additional TRß1 5'-UTR splicing events occur that have not been identified by these studies, a conclusion supported by examination of EST databases. Interrogation of EST databases also indicated alternative polyadenylation of TRß mRNAs and supported Northern blotting studies. The 6.0-kb difference between observed TRß1 mRNAs cannot result from variations in 5'-UTRs, which only differ in size by up to 193 bp. Thus, TRß1 expression is regulated by developmental and tissue-specific alternative 5'-UTR mRNA splicing and differential 3'-UTR polyadenylation. The TRß1 5'-UTR variants identified in this study inhibited protein expression in transfected cells and directly suppressed mRNA translation in vitro.

Regulation of translation initiation occurs via the 5'- and 3'-UTR or by alterations in the translation machinery. The 5'-UTR controls ribosome access, translation efficiency, and initiation codon selection and can direct translation via mRNA-binding proteins (33). Translation initiation occurs as a cap-dependent process at the 5'-end of the mRNA or by internal initiated mechanisms beginning at an internal ribosome entry site (IRES) in the 5'-UTR (33, 34). The cap-dependent process involves scanning of a 43S preinitiation complex along the 5'-UTR until it reaches an AUG codon in a favorable sequence context, where the 60S ribosomal subunit binds and translation initiates (33, 35). Most vertebrate mRNAs contain 5'-UTRs between 20–100 bp that facilitate recognition of the first AUG (24). However, 20–25% are longer and can regulate translation (24), whereas 5–10% possess AUGs that are ignored or inefficiently recognized (33, 35). Inefficient translation initiation is known as leaky scanning.

Interestingly, TR{alpha} protein expression in chicken erythroid cells also appears to be regulated by posttranscriptional and posttranslational mechanisms, although no evidence of alternative splicing of TR{alpha} mRNA has been described (36, 37). The full-length TR{alpha} transcript is translated into multiple N-terminal truncated protein variants by the use of in-frame alternate translation initiation codons in the TR{alpha} coding region that appear to be used with varying efficiency. A ribosomal scanning model in which 5'-AUG initiation sites are selectively bypassed has been hypothesized (37). The translated truncated receptors possess differing functions that may autoregulate the activity of full-length TR{alpha} protein (37) in a fashion similar to that proposed by us for the regulation of TRß3 activity by the N-terminal variant TR{Delta}ß3 in the rat (5).

In these studies, however, we show that the human TRß1 mRNA contains multiple uAUGs and uORFs in the 5'-UTR. These features are unusually common in 5'-UTRs of protooncogenes (24) including nuclear receptors. Long 5'-UTRs do not necessarily inhibit translation, providing that uAUGs and inhibitory mRNA secondary structures are not present (24). However, the presence of uAUGs and uORFs in TRß1 5'-UTRs is in keeping with their inhibition of translation (24, 35). The lack of significant secondary structure and absence of known protein binding domains within the 5'-UTRs indicate that the uAUGs and uORFs are likely to be primarily responsible for translation inhibition (33). This may occur by leaky scanning, the expression of inhibitory peptides from uORFs, or by a combination of both mechanisms, each of which has been reported for other nuclear receptors (31, 38, 39, 40, 41). Extensive mutation analysis of the uAUGs and uORFs within the 5'-UTRs will be required to determine the responsible mechanisms in TRß1. Furthermore, 5'-UTR A behaved differently to all others. This mRNA expressed at low levels in JEG-3 cells (in keeping with its lower secondary structure stability of –68.6 kcal/mol) but protein levels were no different than control, suggesting efficient translation. In contrast, in COS-7 cells 5'-UTR A did not inhibit mRNA expression but translation was suppressed. Clone A is the only 5'-UTR that contains exon 1c. Otherwise, its structure is similar to B and C, which differ only by 34 bp at the 5'-end and have identical properties in JEG-3 and COS-7 cells. This suggests that exon 1c, which does not include an uAUG or uORF, contains regulatory sequences that influence translation, such as a tissue-specific IRES (34). Many similar examples have been reported (http://www.rangueil.inserm.fr/IRESdatabase), but no IRES consensus has been identified, and this possibility requires investigation. An additional consideration is that we have been able to investigate the functional properties of TRß1 5'-UTRs only in the context of both a heterologous promoter and a reporter gene protein, although differences between constructs studied here were only due to the presence of different 5'-UTRs. Nevertheless, it is possible that native TRß1 promoter and protein sequences could influence the properties of 5'-UTRs in vivo. Studies to address this possibility directly will first require identification and characterization of the human TRß1 promoter.

The complexity of the human THRB gene and the temporospatial regulation of its expression are strong indicators that TRß1 protein concentrations are tightly controlled in T3 target tissues, although the relationships between TR mRNA and protein have proved difficult to characterize. Indeed, we have investigated the relationship between mRNA and protein expression in human tissues using a variety of available specific TR antibodies [polyclonal anti-TR antibodies PA1–211A and PA1–213A, and monoclonal anti-TRß antibody MA1–215 (Affinity Bioreagents) and polyclonal anti-TR{alpha} antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA)]. Although these antibodies are specific and can reliably detect in vitro translated TR proteins and overexpressed TR proteins in transfected cells, they lack sufficient sensitivity for the quantitative detection of endogenous TR proteins in human tissues (data not shown). Thus, it is not currently possible to determine the relationship between expressed endogenous TR mRNAs and proteins in human tissues. A detailed understanding of this relationship is of considerable future importance because studies of TR-null mice have shown that TR{alpha} and -ß have discrete developmental and physiological roles (7), and data from TRßPV mice show that the ratio of TR{alpha} and -ß proteins determines T3 responsiveness in individual tissues (10, 11). Our data suggest the likelihood that multiple human TRß1 mRNA variants control the level of expressed TRß1 protein in a tissue-specific manner by a complex array of posttranscriptional mechanisms. Such complex regulation of TR protein expression would ultimately be expected to regulate tissue T3 responsiveness. Elucidation of the molecular mechanisms underlying these processes will open new possibilities for tissue-specific modulation of T3 action in man.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
Note Added in Proof
 REFERENCES
 
5'-RACE
TRß1 5'-UTR cDNAs were isolated by 5'-RACE from adult brain, kidney, placenta, and testis and from 7–8 d fetus cDNA libraries (CLONTECH, Palo Alto, CA). PCR was performed using 0.2 µM AP1 primer (CLONTECH), 0.2 µM TRß1 exon 5 reverse primer (5'-CGTGACTTTGTCTATGACACATTTTCCT-3'), and 2.5 µl template (diluted 1:100 brain, placenta, testis; 1:200 kidney; 1:50 fetus) at 94 C for 30 sec and 68 C for 5 min (30 cycles). Nested PCR was performed with 2.5 µl cDNA diluted 1:50 in tricine-EDTA (10 mM tricine KOH, pH 9.2; 0.1 mM EDTA), 0.2 µM AP2 primer (CLONTECH), and TRß1 nested primer (5'-GATACAGCGGTAGTGATACCCGGTGGCT-3') at 94 C for 30 sec and 72 C for 3 min (five cycles); 94 C for 30 sec and 70 C for 3 min (five cycles); 94 C for 30 sec and 68 C for 3 min (20 cycles) with a final 3-min extension at 68 C.

Inverse RT-PCR
The 5'-boundary of exon A was mapped by inverse RT-PCR (see Fig. 6Go) as described elsewhere (5, 42). cDNA was synthesized from human placenta poly(A)+ RNA (1 µg) and from adult brain or fetal liver total RNA (10 µg) using 10 µM of an exon 2a reverse primer (5'-ATGTGTGGAACTTGGCACCCACGCA-3') with 5x buffer [250 mM Tris (pH 8.3), 30 mM MgCl2, 375 mM KCl], 1 nM (each) deoxynucleotide triphosphates and Moloney murine leukemia virus reverse transcriptase (20 U) in 10 µl at 42 C for 1 h. Second-strand cDNA was synthesized with 10 nM deoxynucleotide triphosphates, 5x buffer (250 mM Tris, pH 7.8; 50 mM MgCl2; 5 mM dithiothreitol; 5 mM ATP; 25% polyethylene glycol 8000) and 20x enzymes (Escherichia coli DNA polymerase I, 6 U/µl; DNA ligase, 1.2 U/µl; RNase H, 0.25 U/µl) in 80 µl at 16 C for 90 min (all reagents from CLONTECH). A 2-µl volume of T4 DNA polymerase was added for 45 min at 16 C to blunt the cDNA, which was self-ligated, cut with BstNI, amplified with forward (FOR: 5'-TCTGCGATTTCCTTCTGGTTGG-3') and reverse (REV: 5'-AGCCGGTTGGGTTGAGC-3') primers, in exons 2a and 1b, respectively. PCR products were subcloned, and 30 individual clones were obtained and sequenced.

TRß cDNA and Genomic Clones
IMAGE clones 2298537 (GenBank accession no. AI650763) and 1084814 (GenBank accession no. AA577807) were identified by interrogating EST databases with TRß1 exons 4–5 and obtained from the Human Genome Mapping Project Resource Centre. A chromosome 3 cosmid library, LL03NC01, was obtained and screened with exon 5 and exon 1a/1b/1c probes. Eight cosmids were identified with exon 5 (AC18C5, AC32A22, AC32I15, AC53B6, AC56C1, AC56E1, AC59B5, and AC119C8) and three with exons 1a/1b/1c (AC111C8, AC127E5, and AC128A2). They were obtained from Human Genome Mapping Project and analyzed by restriction mapping or sequencing. The human chromosome 3 genomic contig clones (RP11-118N13, GenBank accession no. AC098971; RP11-178K11, GenBank accession no. AC093574; RP11-489C3, GenBank accession no. AC099054; and RP11-81I24, GenBank accession no. AC093927) were identified by searching databases with exons 1d, 1e, 2b, and 2c. Analysis of THRB was facilitated using chromosome 3 reference contig NT_000003.5.

Northern Blots
Multiple tissue blots (CLONTECH) were hybridized with oligonucleotide (exons 1c and 1e) or cRNA (exons 1d, 2a, 2b, and 2c) probes and a ß-actin cDNA. Filters were also hybridized to a cRNA probe comprising nucleotides 867–978 in TRß1 exon 7 (GenBank accession no. X04707), which lacks homology with TR{alpha}. Filters probed with cRNAs or cDNAs were prehybridized (30 min at 68 C), hybridized (1 h at 68 C), and washed (0.1x sodium saline citrate/0.1% sodium dodecyl sulfate, 50 C, 2x 15 min). Filters probed with oligonucleotides were prehybridized (30 min at 37 C), hybridized (1 h at 37 C) and washed (0.1x sodium saline citrate/0.1% sodium dodecyl sulfate at room temperature, 2x 15 min).

5'-UTR Constructs
Each 5'-UTR was amplified with a unique forward primer containing a HindIII site and a common reverse primer containing an NcoI site. Products were cloned into pGEM-T and digested with HindIII and NcoI. Inserts were subcloned into pGL3 to obtain constructs in which 5'-UTRs were positioned between luciferase and the Simian virus 40 promoter for use in transfections. A probe for RNase protection analysis of luciferase consisted of 17 bp of 5'-UTR plus 150 bp of luciferase. This probe was transcribed with T7 polymerase to yield a 251-bp mRNA and protected fragments of 167 bp for constructs containing a 5'-UTR or 150 bp for the control lacking a 5'-UTR. A Renilla internal control probe was transcribed with SP6 polymerase to yield a 217-bp mRNA and protected fragment of 137 bp. Each pGL3 construct was digested with SacI and SpeI to release inserts containing the Simian virus 40 promoter and 5'-UTR plus luciferase. Inserts were subcloned into the SacI-SpeI site of pBluescriptSK+, which was digested with SacI and HindIII, blunt-ended with T4 DNA polymerase, and religated to obtain constructs, in which the 5'-UTR and luciferase were positioned downstream of the T7 promoter for in vitro transcriptions. In vitro transcription from these templates generates 5'-UTR-luciferase mRNAs that contain identical short 5'-ends derived from the pBluescriptSK+ vector. It is conceivable that additional vector sequence could contribute to the actions of each 5'-UTR, although functional differences between 5'-UTRs are unlikely to result from nonspecific effects of identical short 5'-ends introduced during vector construction.

Transfections
COS-7 and JEG-3 cells were maintained and transfected in six-well plates with 100 ng pRL-TK, 1 µg pGL3-hTRß15'-UTR, and 400 ng carrier DNA, as described previously (5). Cells were trypsinized 24 h later and reseeded into four wells of a 24-well plate, two containing DMEM plus charcoal-stripped fetal calf serum (DMEM/CSFCS) and two containing T3 (100 nM) in DMEM/CSFCS. Cells were treated for 48 h and extracts prepared to determine luciferase and Renilla activities. Duplicate transfections were performed for RNase protection analyses using Direct Protect Lysate RPA or RPA III kits (Ambion, Inc., Austin, TX) and 7,000–10,000 cpm of luciferase or 14,000 cpm of Renilla probe.

In Vitro Transcription-Translation
5'-UTR-pBluescript constructs were digested with PvuI and PvuII and transcribed with T7 polymerase in the presence of 2.5 µl [32P]UTP (800 Ci/mmol; Amersham Pharmacia Biotech, Arlington Heights, IL), and RNA concentrations were equalized. Each transcript was translated using rabbit reticulocyte lysate (Promega Corp., Madison, WI) in 25-µl reactions, and 10 µl were used in luciferase assays.


   Note Added in Proof
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 Note Added in Proof
REFERENCES
 
While this manuscript was in press, the identification of exon 1c was published (43).


   ACKNOWLEDGMENTS
 
We thank Samuel Refetoff for providing critical comments on the manuscript. We recently learned that investigators in Chicago and Milan have independently identified exon 1c (S. Refetoff, personal communication).


   FOOTNOTES
 
This work was supported by a Medical Research Council (MRC) Clinical Training Fellowship (to S.F.) and MRC Career Establishment Grant G9803002 to (G.R.W.).

Abbreviations: BLAST, Basic local alignment and search tool; EST, expressed sequence tag; IRES, internal ribosome entry site; ORF, open reading frame; PE, position element; RACE, rapid amplification of cDNA ends; RTH, resistance to thyroid hormone; TR, thyroid hormone receptor; uAUG, upstream AUG; uORF, upstream ORF; UTR, untranslated region.

1 These sequence data have been submitted to the GenBank databases under accession nos. AY286465–AY286471. Back

Received for publication September 10, 2003. Accepted for publication April 12, 2004.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 Note Added in Proof
 REFERENCES
 

  1. Yen PM 2001 Physiological and molecular basis of thyroid hormone action. Physiol Rev 81:1097–1142[Abstract/Free Full Text]
  2. Shi YB, Yaoita Y, Brown DD 1992 Genomic organization and alternative promoter usage of the two thyroid hormone receptor ß genes in Xenopus laevis. J Biol Chem 267:733–738[Abstract/Free Full Text]
  3. Forrest D, Sjoberg M, Vennstrom B 1990 Contrasting developmental and tissue-specific expression of {alpha} and ß thyroid hormone receptor genes. EMBO J 16:1519–1528[CrossRef]
  4. Yaoita Y, Brown DD 1990 A correlation of thyroid hormone receptor gene expression with amphibian metamorphosis. Genes Dev 4:1917–1924[Abstract/Free Full Text]
  5. Williams GR 2000 Cloning and characterization of two novel thyroid hormone receptor ß isoforms. Mol Cell Biol 20:8329–8342[Abstract/Free Full Text]
  6. Schwartz HL, Lazar MA, Oppenheimer JH 1994 Widespread distribution of immunoreactive thyroid hormone ß2 receptor (TRß2) in the nuclei of extrapituitary rat tissues. J Biol Chem 269:24777–24782[Abstract/Free Full Text]
  7. O’Shea PJ, Williams GR 2002 Insight into the physiological actions of thyroid hormone receptors from genetically modified mice. J Endocrinol 175:553–570[Abstract]
  8. Weiss RE, Refetoff S 2000 Resistance to thyroid hormone. Rev Endocr Metab Disord 1:97–108[CrossRef][Medline]
  9. Kaneshige M, Kaneshige K, Zhu X, Dace A, Garrett L, Carter TA, Kazlauskaite R, Pankratz DG, Wynshaw-Boris A, Refetoff S, Weintraub B, Willingham MC, Barlow C, Cheng S 2000 Mice with a targeted mutation in the thyroid hormone ß receptor gene exhibit impaired growth and resistance to thyroid hormone. Proc Natl Acad Sci USA 97:13209–13214[Abstract/Free Full Text]
  10. Zhang X-Y, Kaneshige M, Kamiya Y, Kaneshige K, McPhie P, Cheng S-Y 2002 Differential expression of thyroid hormone receptor isoforms dictates the dominant negative activity of mutant ß receptor. Mol Endocrinol 16:2077–2092[Abstract/Free Full Text]
  11. O’Shea PJ, Harvey CB, Suzuki H, Kaneshige M, Kaneshige K, Cheng S-y, Williams GR 2003 A thyrotoxic skeletal phenotype of advanced bone formation in mice with resistance to thyroid hormone. Mol Endocrinol 17:1410–1424[Abstract/Free Full Text]
  12. Zandieh Doulabi B, Platvoet-ter Schiphorst M, van Beeren HC, Labruyere WT, Lamers WH, Fliers E, Bakker O, Wiersinga WM 2002 TRß1 protein is preferentially expressed in the pericentral zone of rat liver and exhibits marked diurnal variation. Endocrinology 143:979–984[Abstract/Free Full Text]
  13. Sakurai A, Nakai A, DeGroot LJ 1990 Structural analysis of human thyroid hormone receptor ß gene. Mol Cell Endocrinol 71:83–91[CrossRef][Medline]
  14. Beck-Peccoz P, Chatterjee VK, Chin WW, DeGroot LJ, Jameson JL, Nakamura H, Refetoff S, Usala SJ, Weintraub BD 1994 Nomenclature of thyroid hormone receptor ß-gene mutations in resistance to thyroid hormone: consensus statement from the first workshop on thyroid hormone resistance, July 10–11, 1993, Cambridge, United Kingdom. J Clin Endocrinol Metab 78:990–993[CrossRef][Medline]
  15. Weinberger C, Thompson CC, Ong ES, Lebo R, Gruol DJ, Evans RM 1986 The c-erb-A gene encodes a thyroid hormone receptor. Nature 324:641–646[CrossRef][Medline]
  16. Sakurai A, Miyamoto T, DeGroot LJ 1992 Cloning and characterization of the human thyroid hormone receptor ß1 gene promoter. Biochem Biophys Res Commun 185:78–84[CrossRef][Medline]
  17. Yaoita Y, Shi YB, Brown DD 1990 Xenopus laevis {alpha} and ß thyroid hormone receptors. Proc Natl Acad Sci USA 87:7090–7094[Abstract/Free Full Text]
  18. Puzianowska-Kuznicka M, Nauman A, Madej A, Tanski Z, Cheng S-y, Nauman J 2000 Expression of thyroid hormone receptors is disturbed in human renal clear cell carcinoma. Cancer Lett 155:145–152[CrossRef][Medline]
  19. Shahrara S, Drvota V, Sylven C 1999 Organ specific expression of thyroid hormone receptor mRNA and protein in different human tissues. Biol Pharm Bull 22:1027–1033[Medline]
  20. Santos A, Freake HC, Rosenberg ME, Schwartz HL, Oppenheimer JH 1988 Triiodothyronine nuclear binding capacity in rat tissues correlates with a 6.0 kilobase (kb) and not a 2.6 kb messenger ribonucleic acid hybridization signal generated by a human c-erbA probe. Mol Endocrinol 2:992–998[Abstract]
  21. Sakurai A, Nakai A, DeGroot LJ 1989 Expression of three forms of thyroid hormone receptor in human tissues. Mol Endocrinol 3:392–399[Abstract]
  22. Markowitz S, Haut M, Stellato T, Gerbic C, Molkentin K 1989 Expression of the erb-ß class of thyroid hormone receptors is selectively lost in human colon carcinoma. J Clin Invest 84:1683–1687
  23. Brunak S, Engelbrecht J, Knudsen S 1991 Prediction of human mRNA donor and acceptor sites from the DNA sequence. J Mol Biol 220:49–65[CrossRef][Medline]
  24. Kozak M 1987 An analysis of 5'-noncoding sequences from 699 vertebrate messenger RNAs. Nucleic Acids Res 15:8125–8148[Abstract/Free Full Text]
  25. Kozak M 1987 At least six nucleotides preceding the AUG initiator codon enhance translation in mammalian cells. J Mol Biol 196:947–950[CrossRef][Medline]
  26. Mathews DH, Sabina J, Zuker M, Turner DH 1999 Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure. J Mol Biol 288:911–940[CrossRef][Medline]
  27. Bairoch A, Bucher P, Hofmann K 1997 The PROSITE database, its status in 1997. Nucleic Acids Res 25:217–221[Abstract/Free Full Text]
  28. Silva JE, Dominguez G, Gonzalez-Sancho JM, Garcia JM, Silva J, Garcia-Andrade C, Navarro A, Munoz A, Bonilla F 2002 Expression of thyroid hormone receptor/erbA genes is altered in human breast cancer. Oncogene 21:4307–4316[CrossRef][Medline]
  29. Graber JH, Cantor CR, Mohr SC, Smith TF 1999 In silico detection of control signals: mRNA 3'-end-processing sequences in diverse species. Proc Natl Acad Sci USA 96:14055–14060[Abstract/Free Full Text]
  30. Kos M, Reid G, Denger S, Gannon F 2001 Minireview: Genomic organization of the human ER{alpha} gene promoter region. Mol Endocrinol 15:2057–2063[Abstract/Free Full Text]
  31. Kos M, Denger S, Reid G, Gannon F 2002 Upstream open reading frames regulate the translation of the multiple mRNA variants of the estrogen receptor. J Biol Chem 277:37131–37138[Abstract/Free Full Text]
  32. McCormick JA, Lyons V, Jacobson D, Noble J, Diorio M, Nyirenda S, Weaver W, Ester JLW, Yau M, Meaney MJ, Seckl JR, Chapman KE 2000 5'-heterogeneity of glucocorticoid receptor messenger RNA is tissue specific: differential regulation of variant transcripts by early-life events. Mol Endocrinol 14:506–517[Abstract/Free Full Text]
  33. Gray NK, Wickens M 1998 Control of translation initiation in mammals. Annu Rev Cell Dev Biol 14:399–458[CrossRef][Medline]
  34. Vagner S, Galy B, Pyronnet S 2001 Irresistible IRES; attracting the translation machinery to internal ribosome entry sites. EMBO Rep 2:893–898[CrossRef][Medline]
  35. Kozak M 1989 The scanning model for translation: an update. J Cell Biol 108:229–241[Abstract/Free Full Text]
  36. Bigler J, Eisenman RN 1988 c-erb A encodes multiple proteins in chicken erythroid cells. Mol Cell Biol 8:4155–4161[Abstract/Free Full Text]
  37. Bigler J, Hokanson W, Eisenman RN 1992 Thyroid hormone receptor transcriptional activity is potentially autoregulated by truncated forms of the receptor. Mol Cell Biol 12:2406–2417[Abstract/Free Full Text]
  38. Diba F, Watson CS, Gametchu B 2001 5'UTR sequences of the glucocorticoid receptor 1A transcript encode a peptide associated with translational regulation of the glucocorticoid receptor. J Cell Biochem 81:149–161[CrossRef][Medline]
  39. Chen LI, Sommer KM, Swisshelm K 2002 Downstream codons in the retinoic acid receptor ß-2 and ß-4 mRNAs initiate translation of a protein isoform that disrupts retinoid-activated transcription. J Biol Chem 277:35411–35421[Abstract/Free Full Text]
  40. Reynolds K, Zimmer AM, Zimmer A 1996 Regulation of RAR ß 2 mRNA expression: evidence for an inhibitory peptide encoded in the 5'-untranslated region. J Cell Biol 134:827–835[Abstract/Free Full Text]
  41. Zimmer A, Zimmer AM, Reynolds K 1994 Tissue specific expression of the retinoic acid receptor-ß2: regulation by short open reading frames in the 5'-noncoding region. J Cell Biol 127:1111–1119[Abstract/Free Full Text]
  42. Ochman H, Gerber AS, Hartl DL 1988 Genetic applications of an inverse polymerase chain reaction. Genetics 120:621–623[Abstract/Free Full Text]
  43. Mannavola D, Moeller LC, Beck-Peccoz P, Persani L, Weiss RE, Refetoff S 2004 A novel splice variant involving the 5' untranslated region of thyroid hormone receptor ß1 (TRß1). J Endocrinol Invest 27:318–322[Medline]

NURSA Molecule Pages Link:

Nuclear Receptors:   TRβ
Ligands:   Thyroid hormone



This article has been cited by other articles:


Home page
 Mol. Endocrinol.Home page
I. Jones, L. Ng, H. Liu, and D. Forrest
An Intron Control Region Differentially Regulates Expression of Thyroid Hormone Receptor {beta}2 in the Cochlea, Pituitary, and Cone Photoreceptors
Mol. Endocrinol., May 1, 2007; 21(5): 1108 - 1119.
[Abstract] [Full Text] [PDF]

Home page
 EndocrinologyHome page
Y.-Y. Liu, R. S. Heymann, F. Moatamed, J. J. Schultz, D. Sobel, and G. A. Brent
A Mutant Thyroid Hormone Receptor {alpha} Antagonizes Peroxisome Proliferator-Activated Receptor {alpha} Signaling in Vivo and Impairs Fatty Acid Oxidation
Endocrinology, March 1, 2007; 148(3): 1206 - 1217.
[Abstract] [Full Text] [PDF]

Home page
 Mol Hum ReprodHome page
M. Khare, A. H. Taylor, J. C. Konje, and S. C. Bell
{Delta}9-Tetrahydrocannabinol inhibits cytotrophoblast cell proliferation and modulates gene transcription
Mol. Hum. Reprod., May 1, 2006; 12(5): 321 - 333.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
A. H. Conrad, Y. Zhang, A. R. Walker, L. A. Olberding, A. Hanzlick, A. J. Zimmer, R. Morffi, and G. W. Conrad
Thyroxine Affects Expression of KSPG-Related Genes, the Carbonic Anhydrase II Gene, and KS Sulfation in the Embryonic Chicken Cornea
Invest. Ophthalmol. Vis. Sci., January 1, 2006; 47(1): 120 - 132.
[Abstract] [Full Text] [PDF]

Home page
 EndocrinologyHome page
J. C. Barnard, A. J. Williams, B. Rabier, O. Chassande, J. Samarut, S.-y. Cheng, J. H. D. Bassett, and G. R. Williams
Thyroid Hormones Regulate Fibroblast Growth Factor Receptor Signaling during Chondrogenesis
Endocrinology, December 1, 2005; 146(12): 5568 - 5580.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
T. A. Hughes and H. J. M. Brady
Expression of axin2 Is Regulated by the Alternative 5'-Untranslated Regions of Its mRNA
J. Biol. Chem., March 4, 2005; 280(9): 8581 - 8588.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
18/7/1631    most recent
Author Manuscript (PDF)