A thyroid hormone receptor that is required for the development of green cone photoreceptors.

Color vision is facilitated by distinct populations of cone photoreceptors in the retina. In rodents, cones expressing different opsin photopigments are sensitive to middle (M, 'green') and short (S, 'blue') wavelengths, and are differentially distributed across the retina. The mechanisms that control which opsin is expressed in a particular cone are poorly understood, but previous in vitro studies implicated thyroid hormone in cone differentiation. Thyroid hormone receptor beta 2 (TR beta 2) is a ligand-activated transcription factor that is expressed in the outer nuclear layer of the embryonic retina. Here we delete Thrb (encoding Tr beta 2) in mice, causing the selective loss of M-cones and a concomitant increase in S-opsin immunoreactive cones. Moreover, the gradient of cone distribution is disturbed, with S-cones becoming widespread across the retina. The results indicate that cone photoreceptors throughout the retina have the potential to follow a default S-cone pathway and reveal an essential role for Tr beta 2 in the commitment to an M-cone identity. Our findings raise the possibility that Thrb mutations may be associated with human cone disorders.

The two thyroid hormone receptor genes have opposite effects on estrogen-stimulated sex behaviors.

The two genes coding for thyroid hormone receptors (TR) alpha 1 and beta have opposite effects on female sex behaviors. Deletion of TRalpha 1 reduced them, whereas deletion of TRbeta actually increased them. These results could not be attributed to altered levels of hormones in the blood, general alterations in estrogen responsiveness or altered general activity. Instead, they indicate a previously unknown molecular mechanism upon which the two TR genes exert opposite influences.

Maternal thyroid hormone is required for parvalbumin neurone development in the anterior hypothalamic area.

Thyroid hormone (TH) is crucial for brain development and function. This becomes most evident in untreated congenital hypothyroidism, leading to irreversible mental retardation. Likewise, maternal hypothyroxinaemia, a lack of TH during pregnancy, is associated with neurological dysfunction in the offspring, such as autism and reduced intellectual capacity. In the brain, TH acts mainly through TH receptor α1 (TRα1). Consequently, mice heterozygous for a dominant-negative mutation in TRα1 display profound neuroanatomical abnormalities including deranged development of parvalbumin neurones. However, the exact timing and orchestration of TH signalling during parvalbumin neurone development remains elusive. In the present study, we dissect the development of parvalbumin neurones in the anterior hypothalamic area (AHA) in male mice using different mouse models with impaired pre- and postnatal TH signalling in combination with bromodeoxyuridine birth dating and immunohistochemistry. Our data reveal that hypothalamic parvalbumin neurones are born at embryonic day 12 and are first detected in the AHA at postnatal day 8, reaching their full population number at P13. Interestingly, they do not require TH postnatally because their development is not impaired in mice with impaired TH signalling after birth. By contrast, however, these neurones crucially depend on TH through TRα1 signalling in the second half of pregnancy, when the hormone is almost exclusively provided by the mother. For the first time, our findings directly link a maternal hormone to a neuroanatomical substrate in the foetal brain, and underline the importance of proper TH signalling during pregnancy for offspring mental health. Given the role of hypothalamic parvalbumin neurones in the central control of blood pressure, the present study advocates the inclusion of cardiovascular parameters in the current discussion on possible TH substitution in maternal hypothyroxinaemia.

Ligand-dependent and -independent transactivation by thyroid hormone receptor beta 2 is determined by the structure of the hormone response element.

Chicken thyroid hormone receptor beta 2 (cTR beta 2) is likely to serve specific functions in gene regulation since it possesses a unique N-terminal domain and is expressed in very few tissues. We demonstrate here that TR beta 2 exhibits distinct transactivation properties which are dependent on the availability of ligand and on the structure of the hormone response element. First, a strong ligand-independent transactivation was observed with hormone response elements composed of direct repeats and everted repeats. Second, TR beta 2 was induced by triiodothyronine to transactivate more efficiently than TR beta 0 on palindromic and everted-repeat types of hormone response elements. However, coexpression of the retinoid X receptor reduced the strong transactivation by TR beta 2 but not by TR beta 0 via palindromic response elements, suggesting that TR beta 2 can transactivate as a homodimer. Finally, the N terminus of TR beta 2 contains two distinct transactivation regions rich in tyrosines, which are essential for transactivation. Our results thus show that the activity of the novel transactivating region of TR beta 2 is dependent on the organization of the half-sites in the response element.

Two different mRNAs are transcribed from a single genomic locus encoding the chicken erythrocyte anion transport proteins (band 3).

The chicken erythrocyte anion transport protein (band 3 of the erythrocyte cytoskeleton) is a central component taking part in two widely divergent functions of erythroid cells; it is a primary determinant of cytoskeletal architecture and responsible for electroneutral Cl-/HCO3- exchange across the plasma membrane. To analyze interesting aspects of the developmental regulation of this gene, we have cloned the cDNA and genomic counterparts of the erythroid-specific anion transport protein. We show that a single genetic locus for band 3 encodes two different erythroid cell-specific mRNAs, with different translational initiation sites, which predict polypeptides of sizes very close to those observed in vivo. In vitro translation and immune precipitation of synthetic mRNA derived from one putative fully encoding cDNA clone demonstrate that this clone gives rise to a protein which is identical in size and antigenicity to bona fide chicken erythroid band 3.

Chicken epidermal growth factor (EGF) receptor: cDNA cloning, expression in mouse cells, and differential binding of EGF and transforming growth factor alpha.

The primary structure of the chicken epidermal growth factor (EGF) receptor was deduced from the sequence of a cDNA clone containing the complete coding sequence and shown to be highly homologous to the human EGF receptor. NIH-3T3 cells devoid of endogenous EGF receptor were transfected with the appropriate cDNA constructs and shown to express either chicken or human EGF receptors. Like the human EGF receptor, the chicken EGF receptor is a glycoprotein with an apparent molecular weight of 170,000. Murine EGF bound to the chicken receptor with approximately 100-fold lower affinity than to the human receptor molecule. Surprisingly, human transforming growth factor alpha (TGF-alpha) bound equally well or even better to the chicken EGF receptor than to the human EGF receptor. Moreover, TGF-alpha stimulated DNA synthesis 100-fold better than did EGF in NIH 3T3 cells that expressed the chicken EGF receptor. The differential binding and potency of mammalian EGF and TGF-alpha by the avian EGF receptor contrasts with the similar affinities of the mammalian receptor for the two growth factors.

Anchorage-independent growth of v-myc-transformed Balb/c 3T3 cells is promoted by platelet-derived growth factor or co-transformation by other oncogenes.

Murine fibroblasts transformed by the myc oncogene have a reduced growth factor dependence for both anchorage-dependent and anchorage-independent proliferation. Here we show that v-myc-transformed Balb/c 3T3 cells require, in addition to insulin, only platelet-derived growth factor (PDGF) or epidermal growth factor (EGF) for anchorage-dependent growth in serum-free media. PDGF, however, cannot efficiently be substituted by EGF for anchorage-independent growth. The results suggest that constitutive v-myc expression reduces cellular growth factor requirements by non-autocrine mechanisms for proliferation in monolayer cultures. In contrast, v-myc-transformed cells require plasma components and growth factors of the 'competence' type for anchorage-independent growth. We also demonstrate that the requirement for PDGF by the myc-transformed cells can be abrogated by v-K-ras, v-src and v-fos but not the v-raf oncogene. The results demonstrate that oncogenes can cooperate in the expression of the transformed phenotype by also drastically reducing cellular growth factor requirements.

The transforming activity of the chicken c-myc gene can be potentiated by mutations.

It was previously demonstrated that four different avian v-myc oncogenes harbor several point mutations. At least one of these leads to an amino acid substitution located in the proximity of position 61 in the second exon, whereas additional substitutions are found in exon 3. We have investigated whether these mutations affect the transforming activity of myc. By constructing avian retroviral genomes expressing hybrid gag-myc oncogenes, in which all or parts of the v-myc domains were replaced by corresponding parts of c-myc, we show here that a substitution of threonine 61 of c-myc for a methionine (as in v-mycmc29) significantly enhances the fibroblast transforming capacity of the recombinant oncogene. However, such a hybrid v/c-myc gene is still several fold less active than the v-mycmc29 oncogene. We have also expressed c-myc from subgenomic retroviral mRNAs: in these constructions the AUG of gag in the RNA leader sequence is in the same reading frame as that of c-myc, apparently leading to the production of a myc protein with 11 N-terminal amino acids encoded by gag and non-coding c-myc sequences. These myc proteins also transform chicken embryo fibroblasts, albeit with a lower efficiency than v-myc, again suggesting that mutations can increase the transforming capacity of myc.

A choice between transcriptional enhancement and repression by the v-erbA oncoprotein governed by one nucleotide in a thyroid hormone responsive half site.

The v-erbA oncoprotein (P75gag-v-erbA) can repress thyroid hormone receptor induced transcriptional activation of target genes. A central question is how hormone responsive elements in a target gene determine the transcriptional regulation mediated by P75gag-v-erbA. We addressed this with receptors chimeric between P75gag-v-erbA and thyroid hormone receptor (TR) by testing their regulatory activities on thyroid hormone response elements (TREs) differing in the sequence of the consensus core recognition motif AGGTCA. We report here that enhances, TR dependent transcriptional activation is conferred by P75gag-v-erbA when the thymidine in the half site recognition motif is exchanged for an adenosine. The enhancement was independent of the DNA binding region of P75gag-v-erbA, whereas increased expression of corepressor abolished the enhancing effect. The data indicate that the enhancement results from an impaired DNA binding by the oncoprotein combined with an effective scavenging of corepressors. Our data thus suggest the P75gag-v-erbA indirectly can contribute to enhancement of thyroid hormone induced gene expression.

The oncoprotein P75gag-v-erbA represses thyroid hormone induced transcription only via response elements containing palindromic half-sites.

The v-erbA oncoprotein P75gag-v-erbA, derived from the thyroid hormone receptor alpha (TR alpha), functions as a transdominant transcriptional repressor. The mechanism by which P75gag-v-erbA acts is however poorly characterized. Here, we show that repression of TR alpha mediated transcription by P75gag-v-erbA in transformed erythroblasts is dependent on the structure of the thyroid hormone response element to which it binds. A very efficient repression was seen with hormone response elements having half-sites organized as everted repeats (ER), whereas repression was inefficient with directly repeated half-sites (DR). Promoters containing half-sites organized as an inverted palindrome (IR) gave an intermediate repression. Although P75gag-v-erbA failed to associate with the ligand binding domain of retinoid X (RXR) receptor in a two-hybrid test, the oncoprotein in nuclear extracts from transformed cells heterodimerised quantitatively with RXR upon binding to response elements of the DR type. On the other hand, both RXR/P75gag-v-erb heterodimers and other types of dimers formed on ER elements. P75gag-v-erbA also failed to bind to elements that contained only one half-site in vivo and in vitro. The data demonstrate that P75gag-v-erbA represses gene expression efficiently as a dimer, and suggest that thyroid hormone responsive genes that may be targets for the action of the oncoprotein are repressed most efficiently if they contain elements of the ER type.

Requirement for the C-terminal domain of the v-erbA oncogene protein for biological function and transcriptional repression.

In contrast to the normal thyroid hormone receptor, the v-erbA product fails to bind hormone due to mutations in the C-terminal ligand binding domain and thus appears to represent a hormone-independent, oncogenic transcription factor. Therefore, we asked whether or not the C-terminal domain of v-erbA is required for its biological activity and putative transcriptional control functions by analysing mutants with altered C-termini. A v-erbA protein truncated in the C-terminal domain lacked detectable biological activity in transformed erythroblasts and its transcriptional repression function with respect to the band 3 gene was abolished. The protein displayed a nuclear location and could still bind to DNA, indicating that the N-terminal region retained DNA-binding activity but was insufficient to produce characteristic v-erbA changes in erythroblasts. Another biologically defective v-erbA variant with a small frameshift towards the extreme C-terminus also failed to repress band 3, indicating a requirement for a specific C-terminal structure in repression. However, this mutant retained partial biological activity, stimulating erythroblasts to grow at a higher rate than cells containing a completely inactive, deleted v-erbA gene. The results demonstrate that the mutated hormone-binding domain, in addition to the DNA-binding region, is critical for v-erbA biological and transcriptional control functions.

A selectable temperature-sensitive v-src Moloney retrovirus.

We have cloned the v-src gene of ts339B77 RSV into a new Moloney-based retroviral expression vector (YN). In the resulting construct (ts339YNsrc), the ts339 gene is transcribed from the viral LTR, while a selectable resitance marker, the Tn5 neomycin phosphotransferase (neor) gene, is transcribed from an internal thymidine kinase (Tk) promoter. G418 resistance and focus formation were induced in NIH3T3 cells at comparable efficiencies within the permissive temperature range (33-37 degrees C). At 39 degrees, on the other hand, focus induction was reduced 15-fold with no corresponding decrease in expression of G418 resistance. In cells infected with ts339YNsrc, phosphoproteins were elevated and similar in pattern on SDS PAGE regardless of whether the cells were grown at the permissive or restrictive temperature. The ts339YNsrc virus will be useful for the study of effects of v-src expression, and may also be of help in identifying relevant substrates of the v-src product.

Defective v-erbB genes can be complemented by v-erbA in erythroblast and fibroblast transformation.

We have introduced 3'-terminal deletions of increasing size in the v-erbB oncogene and analysed the effects of these mutations on the transformation of fibroblasts and erythroblasts. The results show that the transforming activity of the mutants is gradually diminished, and completely abolished in those mutants that do not produce stable v-erbB proteins. The capacity to transform erythroblasts is lost before fibroblast transformation is severely affected, suggesting that a larger part of the C-terminal domain is required for mitogenic signalling in erythroid cells than in fibroblasts. In addition, the v-erbA oncogene was found to cooperate with v-erbB not only in erythroblast but also in fibroblast transformation, inducing a fully transformed phenotype in fibroblasts partially transformed by a mutant erbB oncogene.

The v-erb A oncogene causes repression of erythrocyte-specific genes and an immature, aberrant differentiation phenotype in normal erythroid progenitors.

We have compared the effects of the v-erb A oncogene on proliferation and differentiation of normal erythroid progenitors with those of tyrosine kinase oncogenes, e.g. v-sea. For this, a v-erb A retrovirus containing the neomycin resistance gene as a selectable marker or, alternatively, a v-erb A-ts v-sea retrovirus were used to infect normal bone marrow cells. V-erb A induced the outgrowth of immature, erythropoietin(EPO)-dependent erythroid cells from infected bone marrow which ceased to proliferate and disintegrated after 9 to 18 divisions. In contrast, ts-v-sea erythroblasts grew for the expected 25 to 40 population doublings in the absence of EPO. Transcription of the erythrocyte genes carbonic anhydrase II and erythrocyte anion transporter was significantly inhibited in v-erb A infected erythroblasts, indicating that v-erb A alone was sufficient for the repression of the erythrocyte-specific genes observed in AEV-transformed leukemic cells. A detailed analysis of the differentiation phenotype induced by v-erb A in erythroblasts (in the presence or absence of a temperature-inactivated ts sea oncogene) indicates that v-erb A-erythroblasts express a partially mature, aberrant phenotype characterized by the coexpression of mature and immature differentiation antigens. This phenotype clearly differs from that induced by tyrosine kinase oncogenes in erythroid cells.

The E5 gene of bovine papillomavirus type 1 is sufficient for complete oncogenic transformation of mouse fibroblasts.

The transforming 69% fragment of the bovine papillomavirus type-1 genome was inserted into a retrovirus vector, which also expresses G418 resistance, and the resulting construct was used for transfection of psi 2 cells. C127 cells infected with virus-containing medium from G418 resistant psi 2 clones were selected for G418 resistance and/or transformation. G418 resistant cells contained invariably a 4.4 kb provirus. The transformed cells, in contrast, contained either a 2.8 kb or a 5.4 kb provirus. Cells containing the 2.8 or the 5.4 kb proviruses were fully transformed according to several criteria: they had a transformed morphology, formed colonies in soft agar, contained disarranged F-actin cables and induced tumors when injected subcutaneously into nude mice. A molecular analysis of the 2.8 kb provirus showed that it contained only one complete BPV-1 gene, namely E5. Cells transformed by the 2.8 kb provirus contained colinear RNAs as well as subgenomic mRNAs, composed of a retrovirus leader sequence connected to an exon starting at nucleotide 3605 in the BPV-1 sequence. The only papillomavirus protein that can be expressed from these mRNAs is the E5 protein. The results suggest that the 44 amino acid long membrane protein which is encoded by the E5 gene of BPV-1, confers a fully transformed phenotype to immortalized mouse cells in the absence of other viral gene products or papillomavirus regulatory elements.

Chromosomal assignment of five cancer-associated rat genes: two thyroid hormone receptor (ERBA) genes, two ERBB genes and the retinoblastoma gene.

Using a panel of somatic cell hybrids that segregate rat chromosomes, the localization of five cancer-related rat genes was determined: (i) two thyroid receptor genes, THRA1/ERBA1 and THRB/ERBA2 on chromosomes 10 and 15 respectively, (ii) two ERBB genes, namely the epidermal growth factor gene (EGFR, also called ERBB1) and the ERBB2 gene (also designated neu) on chromosomes 14 and 10 respectively, and (iii) the retinoblastoma gene, RB1, on chromosome 15. The THRA1/ERBA1 and ERBB2/neu genes are thus included in a synteny group, conserved on rat chromosomes 10 and human chromosome arm 17q.

Retroviral capture of c-erbB proto-oncogene sequences: rapid evolution of distinct viral genomes carrying mutant v-erbB genes with different transforming capacities.

The evolution of oncogene-transducing retroviruses was followed by studying the genomes of five new, erbB carrying retroviruses. These viruses, isolated from cells of one chicken infected with Rous Associated virus 1 (RAV-1), had captured c-erbB sequences as a consequence of RAV-1 integration into the host genome. Their genome structures were distinct; however, their v-erbB genes had sustained identical 5' and 3' deletions and the v-erbB-env junctions were identical at the nucleotide level. The results therefore strongly suggest that all five viruses originate from the same capture event. Sequence analyses of the v-erbB genes from three of these viruses revealed that one of them had undergone no further mutation and lacked detectable capacity to transform cells, therefore probably representing an 'early' form of transducing virus. The two other v-erbB genes contained distinct mutations and differed in their potential to induce fibroblast- and erythroblast transformation; they therefore probably represent later derivatives of the virus that captured the erbB oncogene. The data suggest that the initial retrovirus rapidly underwent many alterations after capture of c-erbB sequences, already in the RAV-1 infected bird as well as during subsequent in vitro isolation procedures. The changes involve both major rearrangements of the genome as well as point mutations that activated the erbB oncogene.

Retardation of cochlear maturation and impaired hair cell function caused by deletion of all known thyroid hormone receptors.

The deafness caused by early onset hypothyroidism indicates that thyroid hormone is essential for the development of hearing. We investigated the underlying roles of the TRalpha1 and TRbeta thyroid hormone receptors in the auditory system using receptor-deficient mice. TRalpha1 and TRbeta, which act as hormone-activated transcription factors, are encoded by the Thra and Thrb genes, respectively, and both are expressed in the developing cochlea. TRbeta is required for hearing because TRbeta-deficient (Thrb(tm1/tm1)) mice have a defective auditory-evoked brainstem response and retarded expression of a potassium current (I(K,f)) in the cochlear inner hair cells. Here, we show that although TRalpha1 is individually dispensable, TRalpha1 and TRbeta synergistically control an extended array of functions in postnatal cochlear development. Compared with Thrb(tm1/tm1) mice, the deletion of all TRs in Thra(tm1/tm1)Thrb(tm1/tm1) mice produces exacerbated and novel phenotypes, including delayed differentiation of the sensory epithelium, malformation of the tectorial membrane, impairment of electromechanical transduction in outer hair cells, and a low endocochlear potential. The induction of I(K,f) in inner hair cells was not markedly more retarded than in Thrb(tm1/tm1) mice, suggesting that this feature of hair cell maturation is primarily TRbeta-dependent. These results indicate that distinct pathways mediated by TRbeta alone or by TRbeta and TRalpha1 together facilitate control over an extended range of functions during the maturation of the cochlea.

Requirements for repression of retinoid X receptor by the oncoprotein P75gag-v-erbA and the thyroid hormone receptors.

The oncogenic counterpart of thyroid hormone receptor-alpha (TRRalpha), denoted P75gag-v-erbA, has served as a paradigm for the ability of TRs to repress basal levels of transcription. We show here that the retinoid X receptor (RXR), when activated by its specific ligand SR11237, is repressed by both the normal TRalpha and the P75gag-v-erbA. The repression caused by the two proteins is distinct and dependent on both the cell type and the hormone-response element through which RXR acts. In HeLa cells only TR repressed efficiently through the palindromic 2xIR0 element, whereas the proteins were equally efficient in JEG cells. This demonstrates that proteins distinct in the two cell types mediate the repression. RXR-dependent induction via the natural response element of the cellular retinol-binding protein (CRBPII) gene was likewise (> or = 50%) repressed by TR, whereas P75gag-v-erbA did not repress during the same conditions. Furthermore, P75gag-v-erbA and its variants v-erbAtd359 (lacking repressing activity on TR) and v-erbAr12 (a highly active repressor of TR) efficiently repressed induction by a hybrid protein consisting of the DNA- binding domain of Gal4 and the ligand-binding region of RXR. The viral proteins did not, however, associate with RXR unless the two partners were allowed to heterodimerize upon binding to a specific response element, such as the 2xIR0 element or that of the CRBPII gene. In conclusion, we suggest that the efficient repression seen with the the 2xIR0 element is due to heterodimerization of TR or the viral oncoproteins with RXR and a concomitant inhibition of binding of the RXR-specific ligand that results in an inability of RXR to attract a cell type-specific cofactor. In addition, the data suggest that the interaction between RXR and P75gag-v-erbA on the CRBPII element is too weak to inhibit RXR from binding a ligand and therefore also to repress.

The adenovirus E1A protein is a potent coactivator for thyroid hormone receptors.

The thyroid hormone receptors interact with several different cofactors when activating transciption. In this study, we show that the adenovirus E1A oncoprotein functions as a strong coactivator for the thyroid hormone receptor (TR), and that TR and E1A synergistically activate transcription via direct (DR4) or palindromic (IRO) hormone-responsive sites. Cotransfection experiments using different isoforms of the chicken TR and E1A show synergistic, ligand-enhanced transactivation. This transactivation is accomplished through a direct, ligand-independent interaction between TR and E1A. The interaction domains in TR are localized to the DNA-binding domain and to the carboxy-terminal part of the ligand-binding domain. In E1A, the regions of interactions are localized to the conserved regions 1 and 3. Both of these domains in E1A are required for a 40-fold enhancement of TR-mediated activation in transfection experiments. Taken together, we show that E1A strongly enhances transcriptional activation, which suggests that it serves as a bridging factor between the receptor and other components of the transcription machinery.