Tight regulation of transgene expression by tetracycline-dependent activator and repressor in brain.

Methods to temporally and spatially regulate gene mutations will provide a powerful strategy to investigate gene function in the brain. To develop these methods, we have established a tightly regulated system for transgene expression in the forebrain using both a tetracycline (Tc)-dependent transcription activator (rtTA) and a repressor (TetR-Kruppel-associated box). In this system, the repressor binds to the Tc-responsive element (TRE) in the absence of doxycycline (Dox), leading to the repression of leaky activation of TRE-mediated transcription caused by weak binding of rtTA to TRE. Upon Dox administration, only the activator binds to TRE and activates transcription. We tested this system in cultured cells by bicistronically expressing both the regulators using an internal ribosome entry site (IRES). In COS-1, HeLa and SHSY5Y cells, leaky transcription activation led by rtTA in the absence of Dox was repressed without decreasing the level of activated transcription in the presence of Dox. Using this system, transgenic mice were produced that express both the regulators using IRES in the forebrain under the control of the alphaCaMKII promoter and were bred with transgenic mice carrying the TRE-dependent reporter transgene. In reverse transcription-polymerase chain reaction and in situ hybridization analyses of the forebrain in adult double transgenic mice, the treatment of Dox induces reporter mRNA expression, which was not detected before the treatment and after the withdraw of Dox following the treatment. These results indicate that this system allows the tight regulation of transgene expression in a Dox-dependent fashion in the forebrain and will be useful in investigating gene function in the brain.

Tocopherol-associated protein is a ligand-dependent transcriptional activator.

Vitamin E is a term that encompasses a group of potent, lipid-soluble, chain-breaking antioxidants. Structural analysis reveals that molecules having vitamin E activity include four isomers (alpha, beta, gamma, and delta) of both tocopherols and tocotrienols. Alpha-tocopherol has been shown to have the highest biological vitamin E activity in mammalian tissues based on fetal resorption assays, and it reverses vitamin E deficiency symptoms. Although the molecular functions fulfilled specifically by alpha-tocopherol have yet to be fully described, it is unlikely that they are limited to general antioxidant functions. Here we show the functional characterization of alpha-tocopherol associated protein, TAP, which displays significant sequence similarity to the alpha-tocopherol transfer protein. Ligand competition analysis showed that recombinant TAP binds to alpha-tocopherol but not to other isomers of tocopherols. Using GFP fusion protein expression system, we observed that TAP translocates from cytosol to nuclei in alpha-tocopherol-dependent fashion. Transient transfection experiment showed that TAP activates transcription of the reporter gene in alpha-tocopherol-dependent manner. These results suggest that the biological function of alpha-tocopherol is not only as an antioxidant but also as a transcriptional regulator of gene expression via association with a transcription factor TAP.

Retinol equivalence of carotenoids can be evaluated by hepatic vitamin A content.

The present study demonstrates a new method to evaluate the bioavailability of carotenoids based on the calculation of the hepatic retinol contents. Weaning male rats of Wistar strain were divided into 5 groups. Each group respectively received retinol acetate (2000-10,000 IU per kg diet), alpha-carotene (2400-6000 micrograms per kg diet), beta-carotene (2400-6000 micrograms per kg diet), mixture of alpha- and beta-carotenes in the ratio of 1:2 (2400 and 4800 micrograms per kg dit), and palm-carotene oil (2400-6000 micrograms per kg diet). The derived retinol equivalences of each carotenoid calculated according to the hepatic retinol contents were almost constant regardless of the volume of respective intake (alpha-carotene: 1.25 micrograms per IU; beta-carotene: 0.59 microgram per IU; mixture of alpha- and beta-carotene in the ratio of 1:2: 0.96 microgram per IU; Palm-carotene oil: 1.23 micrograms per IU). The results suggest that the hepatic retinol contents can be used as a new measure to evaluate the vitamin A bioavailability of carotenoids.

Increased 9,13-di-cis-retinoic acid in rat hepatic fibrosis: implication for a potential link between retinoid loss and TGF-beta mediated fibrogenesis in vivo.

BACKGROUND/AIMS: During hepatic fibrosis, hepatic stellate cells (HSCs) transform into myofibroblastic cells and lose their intracellular droplets of retinyl esters, the storage form of vitamin A. Recently, we have demonstrated that 9,13-di-cis-retinoic acid (RA), a geometric isomer identified as a stable and major metabolite of vitamin A in circulation, stimulates the synthesis of plasminogen activator (PA) and induces PA/plasmin-dependent latent transforming growth factor (TGF)-beta activation in HSC cultures, probably via induction and activation of RA receptor (RAR) alpha. The aim of the present study was to address a potential link between the loss of retinyl esters to increased formation of RA(s), which might play a role in facilitating TGF-beta-mediated liver fibrogenesis in vivo. METHODS: We examined the effect of 9,13-di-cis-RA on transactivating activity of RARalpha in HeLa cells as well as its effect on PA- and TGF-beta-dependent collagen synthesis in rat and human HSC cultures. We measured the changes in 9,13-di-cis-RA levels both during activation of rat HSCs in vitro and during porcine serum-induced rat hepatic fibrosis in vivo and correlated this with RAR alpha/beta, PA, TGF-beta and type I procollagen mRNA expression in the fibrotic liver. RESULTS: 9,13-di-cis-RA transactivated RARalpha, and provoked PA/plasmin and TGF-beta-dependent procollagen synthesis in HSCs. 9,13-di-cis-RA levels were increased both in activated HSCs in vitro and in fibrotic liver accompanying the enhanced expression of RAR alpha/beta, PA, TGF-beta and procollagen in vivo. CONCLUSIONS: These findings suggest a potential link between 9,13-di-cis RA formation and hepatic fibrosis via formation of TGF-beta in vivo, and thus provide further insight into the biologic role of retinoids during hepatic fibrogenesis.

Tissue specific-distribution and metabolism of vitamin A are affected by dietary protein levels in rats.

The effect of dietary protein levels on tissue-specific distribution and metabolism of vitamin A was studied in rats given [15-14C] retinol (14C-ROH). The weanling rats were fed a low level vitamin A diet for 10 days, then rats (15 rats per group) were divided into 2 groups; one was given a 40% casein diet as a high protein diet (HP-diet), and the other a 5% casein diet as a low protein diet (LP-diet). After 10 days feeding on these diets, 14C-ROH (5 microCi/rat) was given to both groups, HP-diet and LP-diet, by intraperitoneal injection. The radioactivity in the exhalated gases, urine and feces was measured to estimate the rate of vitamin A metabolism. The tissue specific-distribution of ROH was studied in terms of the radioactivities of the ROH fractions separated by HPLC. The hepatic 14C-ROH content in the HP-diet group was lower than that in the LP-diet group at 24, 48, and 72 hours after administration of 14C-ROH. In contrast, 14C-ROH content in serum, spleen, pancreas, and small intestinal mucosa in the HP-diet group was higher than that in the LP-diet group. The radioactivity of the exhalated gas and feces was higher in the HP-diet group. These results suggest that metabolism of vitamin A is higher with intake of a HP-diet. Thus, dietary protein levels may affect tissue-specific distribution and metabolism of vitamin A, thereby modulating the actions of this vitamin.

Cloning of rabbit TR4 and its bone cell-specific activity to suppress estrogen receptor-mediated transactivation.

To clone a new nuclear receptor, we screened a rabbit heart complementary DNA (cDNA) library with degenerate oligonucleotide probes corresponding to the DNA-binding domain of nuclear receptors, which is highly conserved among receptors. One of the cDNA clones, clone 23, encodes a novel protein of 596 amino acids, and predicted molecular mass is 66 kDa. Homology search analysis identified this protein as rabbit TR4 (TR4-0). We also cloned the cDNA encoding a rabbit TR4 isoform (TR4-1), which lacks the putative C-terminal ligand-binding domain (350 amino acids) caused by a 23-bp exon deletion, which probably occurred during messenger RNA (mRNA) splicing. Northern blot analysis showed that TR4s are expressed with two kinds of mRNAs (9.0 kb and 2.8 kb), both of which are relatively abundant in brain, testis, and bone. RT-PCR analysis, using pairs of primers specific for each TR4, showed that both types of receptor express in various tissues. Furthermore, both are present in primary osteoblasts and bone marrow cells, though the mRNA levels of TR4-0 were much higher than those of TR4-1. A functional study, using a transient transfection assay, showed that both receptors suppressed retinoid X receptor (RXR)-retinoid acid receptor, RXR-TR, and RXR-VDR-mediated transactivation significantly in COS-1 and osteosarcoma cells (UMR-106, ROS17/2.8) and that TR4-0 was much more effective than TR4-1. Unexpectedly, we found that the TR4s effectively suppressed estrogen receptor-mediated transactivation in bone cells, but neither in kidney (COS-1) nor breast cancer cells (MCF-7, one of the major target cells of the estrogen action). Thus, the present study shows a novel property of the TR4 orphan receptor, acting as a bone cell-specific repressor in the estrogen receptor-mediated signaling pathway.

Cloning of chick cellular retinol-binding protein, type II and comparison to that of some mammals: expression of the gene at different developmental stages, and possible involvement of RXRs and PPAR.

We cloned chick cellular retinol-binding protein, type two (CRBP II) cDNA and compared it with those of some mammals. The deduced amino acid sequence showed that chick CRBP II was one amino acid greater in size than those of mammals, and the nucleotide sequence of chick CRBP II shared 72%-75% similarity with those of mammals. RNA blot hybridization analysis showed that CRBP II transcript of 0.7 kb was first detected in the duodenum of day-18 embryonic chick, and exhibited a rapid increase during 24 hr around the hatching. Northern blot hybridization also revealed that the transcripts of two types of retinoid X receptors (RXR alpha and RXR gamma) and peroxisome proliferator-activated receptor (PPAR) were expressed in the chick duodenum at hatching. The organ culture of day 16 embryonic chick duodenum showed that the addition of 9-cis retinoic acid in the medium caused a significant increase in CRBP II mRNA levels. In addition, arachidonic acid, from which putative ligands for PPAR were supposed to be generated, was accumulated around hatching in the duodenum. The results may suggest that the abrupt increase of the CRBP II gene expression in the chick duodenum around hatching may be related with RXRs and/or PPAR.

Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning.

1 alpha,25-Dihydroxyvitamin D3[1 alpha,25(OH)2D3], an active form of vitamin D, has roles in many biological phenomena such as calcium homeostasis and bone formation, which are thought to be mediated by the 1 alpha,25(OH)2D3 receptor (VDR), a member of the nuclear hormone receptor superfamily. However, the molecular basis for the actions of 1 alpha,25(OH)2D3 in bone formation, its role during development and VDR genetic polymorphisms for predicting bone mineral density are uncertain. To investigate the functional role of VDR, we generated mice deficient in VDR by gene targeting. We report here that in VDR null mutant mice, no defects in development and growth were observed before weaning, irrespective of reduced expression of vitamin D target genes. After weaning, however, mutants failed to thrive, with appearance of alopoecia, hypocalcaemia and infertility, and bone formation was severely impaired as a typical feature of vitamin D-dependent rickets type II (refs 8, 9). Unlike humans with this disease, most of the null mutant mice died within 15 weeks after birth, and uterine hypoplasia with impaired folliculogenesis was found in female reproductive organs. These defects, such as alopoecia and uterine hypoplasia, were not observed in vitamin D-deficient animals. The findings establish a critical role for VDR in growth, bone formation and female reproduction in the post-weaning stage.

Role of the C-termini of human and chicken plasma retinol-binding proteins.

The most prominent differences between mammalian and non-mammalian vertebrate retinol-binding proteins (RBP) are in the C-terminal sequences. We have cloned and sequenced the cDNA for chicken RBP. Transfected COS cells that transiently expressed mammalian (human) or non-mammalian (chicken) RBP were used to demonstrate that both proteins were able to bind retinol and human transthyretin. However, we observed an increased retinol-independent secretion in cells expressing chicken RBP and reduced ligand-dependent secretion compared to the human protein. It can therefore be concluded that the C-terminal amino acid tail which is missing in chicken RBP compared to human RBP might play a role in retention and ligand-induced secretion.

Interactions of transthyretin (TTR) and retinol-binding protein (RBP) in the uptake of retinol by primary rat hepatocytes.

The mechanism by which cells take up retinol from retinol-binding protein (RBP) and the role of the RBP-transthyretin (TTR) complex remain unclear. Here we report on retinol uptake through the RBP-TTR complex by primary cultured rat hepatocytes (parenchymal cells, PC) and nonparenchymal cells (NPC) following incubation with [3H]retinol-RBP or the [3H]retinol-RBP-TTR complex under several conditions. The cellular accumulation of retinol was time and temperature dependent in both PC and NPC. Analysis by HPLC showed that the incorporated [3H]retinol in NPC was mainly converted to retinyl ester, although in PC it remained mainly as unesterified retinol. However, the amount of retinol taken up from the RBP-TTR complex was nearly twofold greater than that from RBP alone. The uptake of [3H]retinol from protein-bound retinol was inhibited by an excess of either retinol-RBP or retinol-RBP-TTR complex. Moreover, retinol uptake through the RBP-TTR complex was inhibited by an excess of free TTR. From these results we postulate that TTR may take part as a positive regulator in the delivery of RBP-bound retinol from plasma, possibly by a membrane receptor, and that retinol uptake takes place preferentially from the RBP-TTR complex into both PC and NPC. The uptake of [3H]retinol (2 microM) by PC was saturated, whereas uptake by NPC was not. These results indicate that the physiological importance of TTR in retinol delivery may be especially important to vitamin A-storing stellate (Ito) cells in the NPC fraction.

Two forms of avian(chicken) TATA-binding protein mRNA generated by alternative polyadenylation.

We have isolated and sequenced the cDNA encoding avian(chicken) TATA-binding protein (cTBP). The cTBP protein shows a significant homology to those of the other species, and especially its C-terminal region (180 amino acid residues) is identical to those of the vertebrates. By Northern blot analysis, we found that two transcripts with about 2.1 kb (cTBP0) and 2.7 kb (cTBP1) were expressed in various chicken tissues, though only one type of the TBP transcript was reported in vertebrates. A primer extension study demonstrated a single transcription start site. The analysis of the genomic structure of cTBP with the sequences of the two types of cTBP cDNAs(cTBP0 and cTBP1) revealed that the alternative polyadenylation generates two transcripts with different 3'untranslated regions (3'UTRs), indicating a putative role of the different 3'UTRs on the stability of cTBP mRNAs.

ERC-55, a binding protein for the papilloma virus E6 oncoprotein, specifically interacts with vitamin D receptor among nuclear receptors.

VDR regulates gene expression in a ligand-dependent way by binding to cognate enhancer elements of target gene promoters. The ligand-dependent activation function, AF-2, of VDR is thought to require transcriptional co-activators/co-repressors together with basal transcriptional machinery. Using a yeast two hybrid system with VDR, we have isolated a mouse Ca(2+)-binding protein (designated as VAF1) specifically interacting in vivo and in vitro with VDR among nuclear receptors like RAR, RXR, ER and GR. VAF1 is a mouse homologue to human ERC-55, which has recently been shown to interact with human papillomavirus oncogenic protein, E6[1]. Unlike those of many previously identified co-activators, the VDR-VAF1 interaction was ligand-independent. Thus, VAF1 seems a putative VDR-specific cofactor modulating its function.

Steroid hormones differentially induce transcription of the chicken ovalbumin gene, but stabilize the mRNA with the same half-life.

The stabilization of chicken ovalbumin (OVA) mRNA by different classes of steroid hormones (estrogen, progesterone, glucocorticoid, and androgen) was studied in the oviducts of chicks treated with combinations of four steroids. The combination of estrogen with progesterone, glucocorticoid, or androgen enhanced the induction of the OVA gene more than did estrogen alone. Run-on analysis of the isolated oviduct nuclei to measure the transcription rate of the OVA gene showed that the enhanced induction of the OVA gene by the combined hormone treatments was partly caused by an increased rate of transcription. The half-life of OVA mRNA as determined using a transcription inhibitor (actinomycin D) was estimated to be about 24 h irrespective of the hormone treatment, though the half-life was about 6 h in the absence of hormones. These results suggested that the prolongation of the half-life of OVA mRNA by steroid hormones is constant irrespective of differential transcription rates of the OVA gene.

Bacterially expressed rat retinol-binding protein is functional for retinol and transthyretin bindings.

Retinol-binding protein (RBP) was expressed in Escherichia coli using the cDNA for rat RBP, and characterized. The expressed RBP was fused to maltose-binding protein (MBP) at the N-terminal end (MBP-RBP), and MBP was enzymatically removed from the MBP-RBP with proteinase factor Xa. The binding of retinol and transthyretin (TTR) to the recombinant RBP was monitored by means of gel filtration. The recombinant RBP specifically bound to retinol with an affinity similar to that of purified RBP from rat serum. Furthermore, the retinol-bound recombinant RBP formed hetero-complexes with TTR similar to RBP. Thus, the results showed that the recombinant RBP expressed in E. coli is as functional as serum RBP in terms of retinol and TTR bindings.

Intron retention generates a novel isoform of the murine vitamin D receptor that acts in a dominant negative way on the vitamin D signaling pathway.

We identified and characterized a novel rat vitamin D receptor isoform (rVDR1), which retains intron 8 of the canonical VDR (rVDR0) during alternative splicing. In this isoform protein directed by the stop codon in this newly identified exon, a part of the ligand binding domain (86 amino acids) is truncated at the C-terminal end but contains 19 extra amino acids. The rVDR1 transcript was expressed at a level 1/15 to 1/20 of that of rVDR0 in the kidney and intestine in adult rats but not in embryos. The recombinant rVDR1 protein showed no ligand binding activity. Homo- and heterodimers of the recombinant rVDR0 and rVDR1 proteins bound to a consensus vitamin D response element (VDRE) but not to consensus response elements for thyroid hormone and retinoic acid. However, unlike rVDR0, rVDR1 did not form a heterodimeric complex with RXR on the VDRE. A transient expression assay showed that this isoform acted as a dominant negative receptor against rVDR0 transactivation. Interestingly, the dominant negative activities of rVDR1 differed among VDREs. Thus, the present study indicates that this new VDR isoform negatively modulates the vitamin D signaling pathway, through a particular set of target genes.

Retinoic acid differentially up-regulates the gene expression of retinoic acid receptor alpha and gamma isoforms in embryo and adult rats.

The diverse biological effects of retinoic acid (RA) are exerted by its nuclear receptor-mediated gene expression. One of the two nuclear retinoic acid receptor subfamilies is composed of three subtypes of the all-transretinoic acid receptor (RAR-alpha, RAR-beta, and RAR-gamma). Furthermore, several isoforms are generated from each of three RARs by differential promoter usage and/or alternative splicing. It is thus thought that the developmental stage-specific actions of RA are modulated through the spatio-temporal expression of the subtype and isoforms of RARs. In this study, the auto-regulation of the RAR subtypes (RAR-alpha total, RAR-beta total and RAR-gamma total) and their major isoforms (RAR-alpha 1, alpha 2, RAR-beta 1, beta 2 and RAR-gamma 1, gamma 2) by RA was examined by means of Northern blotting in the 11.5 day embryo and maternal tissues by administering pregnant rats with an excess of all-trans RA. The expression of RAR-beta isoforms as well as the RAR-beta total was auto-regulated by RA in all maternal tissues and embryos examined. The gene expression of RAR-alpha 2, which was not affected by RA in the maternal tissues, was up-regulated in embryos, though there were no significant effects of RA on the levels of RAR-alpha 1 and the alpha total in the maternal tissues and the embryos. Likewise, RA did not affect the levels of RAR-gamma 1 and gamma total. However, unlike RAR-alpha 2, RAR-gamma 2 expression was up-regulated by RA only in the maternal tissues. Thus, these results indicates that two retinoic acid receptor isoforms (RAR-alpha 2 and RAR-gamma 2) are differentially auto-regulated in embryo and adult rats.

Steroid hormone-induced expression of the chicken ovalbumin gene and the levels of nuclear steroid hormone receptors in chick oviduct.

The induction of the chicken ovalbumin (OVA) gene by different classes of steroid hormones and the mRNA levels of estrogen (ER), progesterone (PR), and glucocorticoid (GR) receptors were studied in chick oviducts. Combined treatment with two hormones increased the induction of the OVA gene more than single treatment, when the levels of OVA mRNA were measured with Slot blot analysis. To discover the role of nuclear steroid hormone receptors as transcriptional factors in the OVA gene induction, we analyzed the levels of ER (with RT-PCR), PR, and GR mRNAs (with Northern blotting). The level of PR mRNA was increased only by estrogen, while no steroid hormone affected the levels of ER and GR mRNAs. Thus, these findings show that the levels of nuclear receptors do not reflect the OVA mRNA level in the oviduct of steroid hormone-treated chicks.

Vitamin A is involved in estrogen-induced cell proliferation but not in cytodifferentiation of the chicken oviduct.

We examined vitamin A-deficient chicks to determine whether vitamin A affects the estrogen-induced development of the chick oviduct. When oviduct development was stimulated for 5 days with the synthetic estrogen, diethylstilbestrol, the wet weight of the oviduct in vitamin A-deficient chicks was only half that in control chicks. The DNA content in this tissue showed that the decreased oviduct weight in the vitamin A-deficient chicks was caused by the decreased proliferation of oviduct cells. However, the estrogen-induced expression of the ovalbumin gene was not affected by the vitamin A deficiency, suggesting that estrogen-induced cytodifferentiation is not affected by vitamin A. To clarify the vitamin A action on estrogen-induced development in the oviduct, transcripts of nuclear estrogen receptor (ER) and all-trans-retinoic acid (RAR alpha, beta and gamma) receptors, which exert the effects of estrogen and vitamin A, were measured. The ER, RAR alpha and RAR beta genes, but not that of RAR gamma, were expressed during oviduct development, indicating that estrogen and vitamin A may control the expression of target genes through their cognate receptors. Thus, we have shown that vitamin A is involved in estrogen-induced cell proliferation but not in cytodifferentiation of the chicken oviduct.

Tissue-specific response of the human platelet-activating factor receptor gene to retinoic acid and thyroid hormone by alternative promoter usage.

We have studied the effects of retinoic acid (RA) and thyroid hormone (3,3',5-triiodothyronine; T3) on platelet-activating factor receptor (PAFR) gene expression in intact rats and the ability of two human PAFR gene promoters (PAFR promoters 1 and 2) to generate two transcripts (PAFR transcripts 1 and 2). Northern blotting showed that RA and T3 regulated PAFR gene expression only in rat tissues that express PAFR transcript 2. Functional analysis of the human PAFR promoter 2 revealed that responsiveness to RA and T3 was conferred through a 24-bp element [PAFR-hormone response element (HRE) located from -67 to -44 bp of the transcription start site, whereas PAFR promoter 1 did not respond to these hormones. The PAFR-HRE is composed of three direct repeated TGACCT-like hexamer motifs with 2-and 4-bp spaces, and the two upstream and two downstream motifs were identified as response elements for RA and T3. Thus, the PAF-PAFR pathway is regulated by the PAFR level altered by a tissue-specific response to RA and T3 through the PAFR-HRE of the PAFR promoter 2.

A synthetic oestrogen antagonist, tamoxifen, inhibits oestrogen-induced transcriptional, but not post-transcriptional, regulation of gene expression.

Oestrogen (E2) regulates the expression of its target genes at transcriptional and post-transcriptional levels. To clarify the mechanism of E2-induced post-transcriptional regulation, with attention to the involvement of the oestrogen receptor (ER), we studied the effect of tamoxifen (TAM), a synthetic E2 antagonist that inhibits ER-mediated transcription, on E2-induced transcriptional and post-transcriptional regulation of the chicken ovalbumin (OVA) gene in chick oviducts. Run-on analysis with oviduct nuclei isolated from E2-treated chicks showed that TAM treatment completely blocked E2-induced transcription of the OVA gene within 24 h without affecting ER gene expression. Likewise, the rate of transcription fell to below the limit of detection after E2 withdrawal from the chicks. Reflecting the transcription rate, OVA mRNA accumulated linearly in E2-treated chicks, and E2 withdrawal caused a rapid loss of OVA mRNA. However, in the chicks treated with TAM and E2, OVA mRNA was degraded slowly over 48 h with a half-life of 24 h, suggesting that TAM does not inhibit E2-induced mRNA stabilization. Moreover, E2-induced mRNA stabilization was observed even when transcription of the OVA gene was blocked by a transcription inhibitor. Western-blot analysis showed that the remaining OVA mRNA was translatable. Thus the present study indicates that E2 regulates expression of the OVA gene via distinct pathways at transcriptional and post-transcriptional levels.