hnRNP U inhibits carboxy-terminal domain phosphorylation by TFIIH and represses RNA polymerase II elongation.

This study describes a potential new function of hnRNP U as an RNA polymerase (Pol II) elongation inhibitor. We demonstrated that a subfraction of human hnRNP U is associated with the Pol II holoenzyme in vivo and as such recruited to the promoter as part of the preinitiation complex. hnRNP U, however, appears to dissociate from the Pol II complex at the early stage of transcription and is therefore absent from the elongating Pol II complex. When tested in the human immunodeficiency virus type 1 transcription system, hnRNP U inhibits elongation rather than initiation of transcription by Pol II. This inhibition requires the carboxy-terminal domain (CTD) of Pol II. We showed that hnRNP U can bind TFIIH in vivo under certain conditions and inhibit TFIIH-mediated CTD phosphorylation in vitro. We find that the middle domain of hnRNP U is sufficient to mediate its Pol II association and its inhibition of TFIIH-mediated phosphorylation and Pol II elongation. The abilities of hnRNP U to inhibit TFIIH-mediated CTD phosphorylation and its Pol II association are necessary for hnRNP U to mediate the repression of Pol II elongation. Based on these observations, we suggest that a subfraction of hnRNP U, as a component of the Pol II holoenzyme, may downregulate TFIIH-mediated CTD phosphorylation in the basal transcription machinery and repress Pol II elongation. With such functions, hnRNP U might provide one of the mechanisms by which the CTD is maintained in an unphosphorylated state in the Pol II holoenzyme.

Structural and functional analysis of the rat malic enzyme gene promoter.

We have characterized sequences of genomic DNA 5' to the coding region of the rat malic enzyme gene. This sequence possesses neither TATA nor CCAAT sequences in their usual positions but is rich in GC residues. Sequences similar to those found in the regulatory regions of other genes are discussed. Deletion analyses have revealed that sequences +1 to -41 are sufficient to initiate expression, although inclusion of information up to -177 is necessary for maximal promoter activity.

Early postnatal isolation reduces dopamine levels, elevates dopamine turnover and specifically disrupts prepulse inhibition in Nurr1-null heterozygous mice.

Sensorimotor gating is a phenomenon that is linked with dopamine neurotransmission in limbic and cortical areas, and disruption of sensorimotor gating has been consistently demonstrated in schizophrenia patients. The nuclear receptor Nurr1 is essential for development of dopamine neurons and, using Nurr1-null heterozygous mice, has been found to be important for normal dopamine neurotransmission as null heterozygous mice have reduced limbic and cortical dopamine levels and elevated open-field locomotor activity. The current investigation compared sensorimotor gating, as measured by prepulse inhibition of the acoustic startle response, in Nurr1 wild-type and null heterozygous mice. When mice were weaned between 19 and 21 days of age either into isolation or groups of three to five and tested 12 weeks later, prepulse inhibition was elevated in group-raised null heterozygous mice and significantly disrupted in isolated null heterozygous mice as compared with isolation-raised wild-type mice and group-raised null heterozygous mice. Isolation had no effect on prepulse inhibition in wild-type mice. Isolation reduced tissue dopamine levels and elevated dopamine turnover in the nucleus accumbens and striatum in both wild-type and null heterozygous mice. In the prefrontal cortex, isolation reduced dopamine and 3,4-dihydroxyphenylacetic acid levels in null heterozygous as compared with isolation-raised wild-type mice, whereas no differences were observed between group-raised wild-type and null heterozygous mice. Neither the null heterozygous genotype nor isolation had any effect on basal or stress-induced corticosterone levels. These data suggest that the Nurr1 null heterozygous genotype predisposes these mice to isolation-induced disruption of prepulse inhibition that may be related to the interactions between intrinsic deficiencies in dopamine neurotransmission as a result of the null heterozygous genotype and isolation-induced changes in dopamine neurotransmission. Post-weaning isolation of Nurr1 null heterozygous mice provides a model to explore the interactions of genetic predisposition and environment/neurodevelopment on dopamine function that has important relevance to neuropsychiatric disorders.

Molecular size of the nuclear thyroid hormone receptor.

Among the previously reported putative nuclear thyroid hormone receptor forms having molecular masses of 56-59 kDa and 45-49 kDa, respectively, only the former can be the endogenous receptor. The latter must be a degradation product because it is virtually absent in rat liver nuclear extracts prepared in the presence of 20% glycerol and 5 mM Mg2+, which inhibit degradation. In the absence of glycerol, the receptor form of lower mass was present in large amounts in nuclear extracts. Sucrose could not replace glycerol as a protective agent, even in the presence of Mg2+. Thus, the endogenous nuclear thyroid hormone receptor appears to be labile under the experimental conditions used in preparing nuclear extracts. The molecular mass of the nuclear receptor was determined to be 57 kDa on the basis of SDS-polyacrylamide gel electrophoresis after photoaffinity labeling of nuclear proteins with (3,5-125I)-labeled thyroxine.

The effect of thyroid hormone on the chromatin structure and expression of the malic enzyme gene in hepatocytes.

Malic enzyme catalyzes the NADP-dependent oxidative decarboxylation of malate to pyruvate and carbon dioxide and is involved in lipogenesis. We have investigated the effect of thyroid hormone on the chromatin structure of the malic enzyme gene in rat liver. Hypersensitivity to DNase I in the immediate 5'-flanking region was altered by T3. T3 stimulation induced hypersensitive sites at -310 base pairs (bp) and -50 bp whereas a hypersensitive site at -170 bp was thyroid hormone independent. Hypersensitive sites identified in the 3'-flanking region showed no change with T3 stimulation. We further characterized expression of the malic enzyme gene as a function of thyroidal state by localizing malic enzyme mRNA in hepatocytes using in situ hybridization histochemistry. In hypothyroid and euthyroid states, two populations of hepatocytes were seen, some with malic enzyme message and others with no detectable message. These differences in malic enzyme gene expression were most evident between groups or regions of hepatocytes. After 10 days of thyroid hormone treatment all hepatocytes demonstrated malic enzyme message. The hypersensitivity results confirm that thyroid hormone stimulation of malic enzyme synthesis occurs in part at the level of transcription, and localization of malic enzyme gene expression suggests this stimulation is accompanied by recruitment of hepatocytes. Hepatocytes may be heterogeneous in their ability to respond to thyroid hormone.

Developmental regulation of the thyroid hormone receptor alpha 1 mRNA expression in the rat testis.

The multiplicity of thyroid hormone (TH) effects appears to be mediated by two TH receptors (THRs) encoded by two genes, alpha and beta, and, perhaps, by their various isoforms. The expression of THR beta is correlated with the presence of high affinity binding sites for TH, and all the mutations which cause the syndrome of generalized thyroid hormone resistance occur in THR beta. The function of THR alpha has not been clearly defined as yet. Another enigma in TH action is the effect on the testis. It has been shown that the testis of the adult rat does not respond to TH as measured by an increase in oxygen consumption. Furthermore, it has not been possible to demonstrate the presence of a nuclear high affinity binding site for TH in adult testis. To resolve these problems were measured the levels of THR alpha, its nonhormone binding variant, and THR beta mRNA in the testis at various stages of development. We discovered that the beta-message is absent at all times, whereas the alpha-message is expressed only from fetal through prepubertal stages and is absent in adult testis. THR alpha, but not the beta-mRNA, was detected in immature Sertoli cells in culture, and neither was found in adult Sertoli cell-enriched cultures. Furthermore, THR alpha and its variant mRNA was found, using in situ hybridization, in the seminiferous cords and seminiferous tubules of fetal and prepubertal testis.(ABSTRACT TRUNCATED AT 250 WORDS)

Thyrotropin regulation of malic enzyme in FRTL-5 rat thyroid cells.

TSH-induced increases in malic enzyme mRNA levels in FRTL-5 rat thyroid cells are paralleled by increases in malic enzyme activity and are mimicked by 8-bromo-cAMP. Apparent approximately 4 h after TSH challenge and maximal after 16 h, they decline by 24 h and are at basal levels by 48 h. The increase occurs in the absence of a measurable effect of TSH on DNA synthesis related to cell growth, since [3H] thymidine incorporation into DNA is still at basal levels 24 h after TSH challenge and is maximal only at 48 h. A protein(s) whose formation is inhibited by cycloheximide appears to be critical to the ability of TSH to increase malic enzyme mRNA levels. Thus, cycloheximide given 30 min before TSH prevents the hormone-induced increase in malic enzyme mRNA; also, when given 24 h after TSH, cycloheximide accelerates the loss of the TSH-induced increase in malic enzyme mRNA. In neither case does cycloheximide affect beta-actin mRNA levels. A second factor(s) whose formation is prevented by actinomycin-D appears to be important for the decrease in malic enzyme mRNA levels seen 24 and 48 h after TSH challenge. Thus, in experiments in which it is given 24 h after TSH, actinomycin-D preserves the hormone-induced increase in malic enzyme mRNA levels rather than accelerating the decrease, as does cycloheximide. In the same experiment, beta-actin mRNA levels decrease to less than 10-20% of control values over the same period; this factor also, therefore, appears to exhibit some degree of specificity.(ABSTRACT TRUNCATED AT 250 WORDS)

Thyrotropin increases malic enzyme messenger ribonucleic acid levels in rat FRTL-5 thyroid cells.

The addition of TSH to FRTL-5 thyroid cells induces a 7- to 8-fold increase in the steady state level of malic enzyme [L-malate-NADP+ oxidoreductase (decarboxylating); EC 1.1.1.40] mRNA, but does not alter beta-actin mRNA levels. Insulin alone or together with TSH has no effect on malic enzyme mRNA. The effect of TSH is not the result of thyroid hormone formation, since the addition of T3 in the presence or in the absence of TSH and the addition of 5% serum (which includes T3 and T4) have no effect. Forskolin (10(-6) M) reproduces the TSH effect, suggesting that cAMP is involved.

Cloning of the human thyrotropin beta-subunit gene and transient expression of biologically active human thyrotropin after gene transfection.

A 17 kilobase pair fragment of DNA containing the human TSH (hTSH) beta-subunit gene was isolated from a human leukocyte genomic library. Using a 621 base pair human CG alpha-subunit cDNA and a 2.0 kilobase pair genomic fragment of hTSH beta containing both coding exons, we constructed hCG alpha and hTSH beta expression vectors containing either the early promoter of simian virus 40 or the promoters of adeno-associated virus. Cotransfection of two adeno-associated virus vectors, each containing one subunit of hTSH, together with a plasmid containing the adenovirus VA RNA genes produced hTSH as well as free human alpha- and TSH beta-subunits in an adenovirus transformed human embryonal kidney cell line (293). The levels of protein expression in this system were 10- to 100-fold greater than that found in a simian virus transformed monkey kidney cell line (COS) using vectors containing the early promoter of simian virus 40. The hTSH synthesized in 293 cells was glycosylated as indicated by complete binding to concanavalin A-Sepharose but was larger in apparent molecular weight than a standard hTSH preparation on gel chromatography suggesting an altered glycosylation pattern. However, it was immunologically and biologically indistinguishable from two pituitary hTSH standards in an immunoradiometric and in vitro iodide trapping assay, respectively.

Identification of a nuclear protein from rat developing brain as heterodimerization partner with thyroid hormone receptor-beta.

Thyroid hormone receptors (TR) are ligand-activated transcription factors that modulate the expression of certain target genes in a developmental and tissue-specific manner. These specificities are determined by the tissue distribution of the TR isoforms alpha1 and beta1, the structure of the thyroid hormone response element (TRE) bound by the receptor, and heterodimerization partners. Among these, retinoid X receptors (RXR) have been recognized as the principal partners for TR. The present work reports the identification of a novel nuclear protein from 19-day-old embryonic rat brain that displays a distinct interaction pattern with TR isoforms at the level of the TRE of two genes known to be differentially expressed and regulated by thyroid hormone (T3): the ubiquitous malic enzyme and the brain-specific myelin basic protein. Electrophoretic gel mobility shift assays demonstrate that only TRbeta1 forms a specific complex with the rat brain nuclear factor on the myelin basic protein-TRE, but not on the malic enzyme-TRE. Thus, the interaction is selectively determined by both the receptor isoform and the structure of the TRE. The expression of this brain nuclear factor is restricted to the perinatal period, when myelination is sensitive to T3. Gel supershift assays with RXR-specific antibodies indicate that this factor is not one of the known RXR isoforms. However, it is most likely a new member of the RXR subfamily because it could be supershifted with an antibody raised against the highly conserved DNA-binding domain of RXRs.

Active repression by thyroid hormone receptor splicing variant alpha2 requires specific regulatory elements in the context of native triiodothyronine-regulated gene promoters.

Structural requirements for the inhibitory action of thyroid hormone receptor splicing variant alpha2 (TR alpha2) on T3/TRbeta1-mediated transactivation were investigated in native promoters of two T3-regulated genes: the brain-specific myelin basic protein (MBP) and the housekeeping malic enzyme (ME). T3/TRbeta1 transactivation of MBP256-chloramphenicol acetyl transferase (CAT) and ME315-CAT constructs was inhibited and unaffected by TR alpha2, respectively. In electrophoretic mobility shift assays, TR alpha2 bound MBP-thyroid response element (TRE) as a monomer but failed to interact with ME-TRE. Mutations of ME-TRE allowed TR alpha2 binding but not inhibition of T3/TRbeta1-mediated transactivation. In the context of the MBP promoter, replacement of MBP-TRE with ME-TRE or exchange of MBP TATA-like box with the ME GC-rich region spanning the transcription start site abolished TR alpha2 dominant negative action. Simultaneous introduction of both MBP-TRE and MBP TATA-like box in the context of ME promoter, however, triggered TR alpha2 inhibition of T3/TRbeta1 transactivation, indicating that these regulatory elements are necessary, but not individually sufficient, to mediate TR alpha2 dominant negative activity. Functional studies at low TR alpha2/TRbeta1 ratios revealed that binding to TRE facilitates TR alpha2 dominant negative action while prevention of DNA interaction by altering TR alpha2 P-box structure preserved TR alpha2 inhibitory effect, although with lower potency. In conclusion, the results suggest that, in native promoters of T3-regulated genes, a dual molecular mechanism, with DNA-binding dependent and DNA-binding independent components, underlies TR alpha2 dominant negative activity.

Apoptosis in the developing cerebellum of the thyroid hormone deficient rat.

The mechanism underlying transient reduction of cell number in the developing cerebellum have been studied for several decades. In this study we analyzed cell death by apoptosis in the developing cerebellum of euthyroid and hypothyroid rats. Results showed that in both groups the apoptotic activity is limited to the internal granular layer from postnatal (p) day 2 to day 12 in euthyroid animals, with the peak at 8 days. No apoptotic cells were detected in the cerebellum of 22 days old euthyroid rats. The level of apoptosis in the cerebellum of hypothyroid rats also reached a peak at 8 days but was four times higher than in control animals. Apoptosis in hypothyroid animals was also observed at p22 and corresponds to the value found in the time of the apoptotic peak in euthyroid cerebellum. At the age of 42 days, no apoptotic cells were found in the cerebellum of either group. Furthermore, it appears that the hormone also plays a role in the disappearance of the external germinal layer, since its presence is still apparent in 42 day old hypothyroid cerebellum. Hence, our results suggest that the deficiency of thyroid hormone (TH) not only increases, but also extends apoptosis during rat cerebellum development and affects the disappearance of the external germinal layer.

Distribution of messenger RNAs for the orphan nuclear receptors Nurr1 and Nur77 (NGFI-B) in adult rat brain using in situ hybridization.

Nurr1 and Nur77 (NGFI-B) are orphan nuclear receptors, belonging to the steroid/thyroid hormone receptor gene superfamily. They have conserved amino acid sequence in the zinc-finger DNA binding domains and similar COOH-terminal regions, but have no known ligands. However, different expression patterns during brain development and tissue distributions of these messenger RNAs imply that they might reflect a different transcriptional role in the brain. In this study, the regional and cellular expression of messenger RNAs encoding these two proteins in rat brain has been determined by in situ hybridization. Nurr1 messenger RNA is highly expressed in the piriform and entorhinal cortices, hippocampus, medial habenular and paraventricular thalamic nuclei. Moderate labeling was detected in layers II-V of most of the cerebral cortex, and in the dorsal lateral geniculate nucleus, substantia nigra (pars compacta and reticularis) and interpeduncular nucleus. No Nurr1 hybridization signal was seen in the rhombencephalon. In the cerebellum, Nurr1 messenger RNA is present in the internal granular cell layer and Purkinje cell layer. In contrast, Nur77 has a widespread distribution, with the highest level of expression in the cerebral cortex. Moderate hybridization signals were detected in the hippocampus, the lateral dorsal and posterior nuclei, reuniens thalamic nuclei, and paraventricular and supraoptic hypothalamic nuclei. In the rhombencephalon, higher signals were present in the medial and lateral vestibular, dorsal cochlear and facial, and raphe magnus nuclei. Nur77 signal was also detected in the nucleus of the spinal tract of the trigeminal nerve. In the cerebellum, Nur77 messenger RNA is highly expressed in the Purkinje cell layer and lateral deep nucleus of the cerebellum. Our results show that Nurr1 and Nur77 messenger RNAs have both overlapping and different distribution patterns within the brain, suggesting that they might regulate different sets of responsive genes.

Differential expression of tyrosine hydroxylase in catecholaminergic neurons of neonatal wild-type and Nurr1-deficient mice.

The orphan nuclear receptor Nurr1 is a transcription factor that belongs to the steroid/thyroid hormone receptor superfamily and is expressed in many regions of the brain. To determine the physiological role of Nurr1, we previously generated mice with a null mutation in the Nurr1 gene. Nurr1-null mice appear to develop normally but die within 12 h after birth. Subsequent analysis revealed the absence of neurotransmitter dopamine and tyrosine hydroxylase immunoreactivity in the central dopaminergic area of newborn pups. Herein, using in situ hybridization histochemistry, we show that Nurr1 is expressed only in subset of catecholamine producing neurons (A2 partly, A8-A10 and A11 catecholaminergic cell groups), and is excluded from the norepinephrine producing neurons (A1, A2, A5-A6 catecholaminergic cell groups). Nurr1 was not expressed in the dopamine synthesizing cell groups (A12-A16 catecholaminergic cell groups) of the diencephalon and the olfactory bulb. As previously shown and confirmed in this study, tyrosine hydroxylase immunoreactivity was absent in the substantia nigra and ventral tegmental area of Nurr1-deficient mice. However, the loss of Nurr1 expression in A2 and A11 dopaminergic neurons did not affect their tyrosine hydroxylase immunoreactivity. This study begins to dissect cues necessary for understanding the complex regulation of the catecholaminergic biosynthetic pathway with regard to local, chemical and developmental changes in the brain.

Nigrostriatal innervation is preserved in Nurr1-null mice, although dopaminergic neuron precursors are arrested from terminal differentiation.

Various factors, including the orphan nuclear receptor Nurr1, have been implicated in dopamine biosynthesis, but many of the specific events involved in this process have to be determined. Using genetic manipulations in mice, the obligatory role for Nurr1 in dopamine (DA) biosynthesis has been documented; however, the mechanism remains unclear. DA biosynthetic enzymes, transporters and receptors are absent in the substantia nigra (SN) and the ventral tegmental area (VTA) of Nurr1-null neonates. The current study establishes that the loss of Nurr1 function does not affect the normal ventralization of neuroepithelial cells to the ventral midbrain, their differentiation into neurons, and their topographical pattern in the SN and VTA. Futhermore, the absence of Nurr1 does not affect the survival of these DA precursor cells in the ventral midbrain, as determined by quantitative analysis of cells, expressing the general neuronal nuclear marker (NeuN) and the TUNEL assay for apoptosis. These neurons express cholecystokinin (CCK), a co-transmitter of dopaminergic neurons in this area. The untranslated exon 1-2 of the Nurr1 gene, which remains intact after homologous recombination, revealed the presence of dopaminergic precursors in the ventral midbrain of the Nurr1-null mice. In addition, these neurons establish their nigrostriatal projections, as shown by axonal transport of a fluorescent tracer, DiI. These results provide evidence that Nurr1 is essential for terminal differentiation of the dopaminergic neurons in the ventral midbrain but does not affect the early steps of their neurogenesis, migration, survival and striatal projections. Our findings suggest that activation of Nurr1 might be therapeutically useful in Parkinson's disease.

Nurr1-null heterozygous mice have reduced mesolimbic and mesocortical dopamine levels and increased stress-induced locomotor activity.

Nurr1, an orphan nuclear receptor, is essential for the differentiation of the midbrain dopamine (DA) neurons; however, its function in adult midbrain DA neurons has not been determined. The present study compared regional brain levels of catecholamines and spontaneous and pharmacologically induced locomotor behaviors between mice heterozygous for the Nurr1-null allele (+/-) and wild type (+/+) littermates. The Nurr1 +/- mice had significantly lower levels of DA in whole brain, midbrain, prefrontal cortex and nucleus accumbens, although no significant differences were observed in the striatum, olfactory bulb or hippocampus. Nurr1 +/- mice displayed significantly greater locomotor activity in a novel open field and after saline injection with no significant difference in activity after treatment with amphetamine (2.5 or 5.0 mg/kg) or MK 801 (0.2 or 0.4 mg/kg). A similar elevation in locomotor activity was observed in Nurr1 +/- mice at 35 days old as was found in 70 days old adults. These data demonstrate that the loss of a single Nurr1 allele results in reduced DA levels in mesolimbic and mesocortical pathways and increased locomotor activity in response to mild stress. The involvement of Nurr1 in DA neurotransmission and the implications for schizophrenia are discussed.

A divergent role of COOH-terminal domains in Nurr1 and Nur77 transactivation.

Orphan nuclear receptors such as Nurr1 and Nur77 have conserved amino acid sequences in the zinc finger DNA binding domains and similar COOH-terminal regions, but have no known ligands. These receptors can bind DNA sequences (response elements) as monomers and can also heterodimerize with the retinoid X receptor to activate transcription. We report here the identification and initial characterization of a novel COOH-terminal truncated isoform of Nurr1, Nurr1a. Internal splicing of Nurr1 generates a frameshift such that a stop codon is prematurely encoded resulting in a naturally occurring COOH-terminal truncation. Embryonic and postnatal mouse brain showed both Nurr1 and Nurr1a mRNAs expressed during development. To characterize essential COOH-terminal elements that may be deleted from Nurr1a and determine function in putative ligand binding, we created COOH-terminal deletion mutants. Nurr1, Nur77, and 3'-truncated mutants bind in gel mobility shift assays to the monomeric Nur77 response element (B1A-RE). However, in transient transfection assays, a truncation of as little as 15 Nurr1 COOH-terminal amino acids diminished transcriptional activation of B1A-thymidine kinase-chloramphenicol acetyltransferase reporter. This result was not seen for a similar Nur77 deletion mutant, Nur77-586. Unlike full-length Nurr1 and Nur77, transactivation by Nur77-586 was not augmented in response to the presence of retinoid-like receptor and 9-cis-retinoic acid. Thus, the interaction of putative ligand binding and transactivation for Nurr1 and Nur77 may function differently.

Molecular cloning of a cDNA sequence for rat malic enzyme. Direct evidence for induction in vivo of rat liver malic enzyme mRNA by thyroid hormone.

Malic enzyme (EC 1.1.1.40) mRNA was partially purified to 12% of total mRNA activity (greater than 150-fold enrichment) from 3-5-3'-triiodo-L-thyronine-carbohydrate-stimulated rat liver by polysome immunopurification followed by oligo(dT)-cellulose chromatography. This preparation was used as the template for synthesis of cDNA on a pBR322-SV40 hybrid plasmid DNA primer which was then circularized with linker DNA and used to transform Escherichia coli. The resulting transformants were screened by in situ differential hybridization using 32P-labeled poly(A+)RNA prepared from uninduced and 3-5-3'-triiodo-L-thyronine-carbohydrate-induced rat livers. Of 750 transformants screened, 6 were found to hybridize differentially; 1 of these, prME, contained an insert of about 1250 base pairs and hybrid-selected an mRNA which directed synthesis of malic enzyme in a cell-free translation system. Using this cDNA as a probe, we demonstrated that the level of malic enzyme mRNA after thyroid hormone treatment was markedly increased and the size of the major malic enzyme mRNA was shown by Northern analysis to be about 2700 nucleotides in length.

Regulation of expression of the alternative mRNAs of the rat alpha-thyroid hormone receptor gene.

The rat alpha-thyroid hormone receptor gene encodes through alternative splicing at least three protein isoforms with different functions, and three mRNA species (2.6, 5.4, 6.8 kilobase (kb) in size) are detected using alpha gene-specific probes (Mitsuhashi, T., Tennyson, G. E., Nikodem, V. M. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 5804-5808). In the present study, the identities of these mRNAs were analyzed by Northern analysis, and it was demonstrated that in rat brain the receptor protein is encoded by the minor 5.4- and 6.8-kb mRNAs and the variant proteins are encoded by the major 2.6-kb mRNA. Relative quantities of these mRNAs were determined by RNase protection assay, and the ratio of the receptor mRNAs to the variant mRNAs was estimated to be 1:6 in adult brain. The ratio between the mRNAs was regulated in both a tissue-specific and developmental stage-specific manner. The receptor mRNA levels were also regulated by the thyroid state of the animal showing an increased level in hypothyroid rat liver while those in brain were not affected. Analysis of the alpha-thyroid hormone receptor gene suggested that the choice between two poly-adenylation sites and subsequent RNA processing appear to generate the 3' heterogeneity of these alternative mRNAs.

Alternative splicing generates messages encoding rat c-erbA proteins that do not bind thyroid hormone.

The nucleotide and predicted amino acid sequences of two variant cDNAs [rat brain thyroid hormone receptor (rTR alpha) vI and vII], isolated from a rat brain cDNA library by using the Pst I fragment of v-erbA, showed virtual identity with the rat brain thyroid hormone receptor (rTR alpha) [Thompson, C. C., Weinberger, C., Lebo, R. & Evans, R. M. (1987) Science 237, 1610-1614] in the putative DNA binding domain and in the first 180 amino acids of the hormone binding domain but no similarity except for 5 amino acids at the extreme 3' end. Isolation and sequencing of the 3' end of the gene coding for rTR alpha, vI and vII mRNAs revealed that the 3' heterogeneity is due to alternative splicing of the primary transcripts of the same gene. RNA transfer blot analyses with probes unique to rTR alpha, rTR alpha vI, and rTR alpha vII showed that only the variant mRNAs are abundantly expressed in rat brain, contrary to the previously reported high-level expression of rTR alpha. Since in vitro translation products of rTR alpha vI and rTR alpha vII did not bind thyroid hormones specifically, our findings explain the discrepancy between the reported abundance of the receptor mRNA and the low receptor levels determined by ligand binding studies in rat brain. These variant mRNAs are also expressed in kidney, heart, spleen, and liver, albeit at lower levels. The presence of an intact DNA binding domain in rTR alpha vI and rTR alpha vII suggests that the variants might have modulating functions in thyroid hormone action.