Isolation of cDNA, chromosome mapping, and expression of the human TBP-like protein

TBP is an essential factor for eukaryotic transcription. In this study, we identified a human cDNA encoding 21-kDa TBP-like protein (TLP). The TLP ORF, carrying 186 amino acids, covered the entire 180 amino acids of the C-terminal conserved domain of human TBP with 39% identity and 76% similarity. FISH determined that human tlp gene was located at chromosome 6 region q22.1-22.3. Northern blot analysis demonstrated that TLP mRNAs were expressed in various human tissues ubiquitously. We found that the TLP proteins exist in multiple mammalian cells and chicken cells. Although the Drosophila TBP-related factor (TRF) is a neurogenesis-related transcription factor, expression of TLP was nearly constant throughout the neural differentiation of P19 cells. Unlike TRF, TLP did not bind to the TATA-box nor direct transcription initiation in vitro. Similarity between TRF and TLP was considerably lower (35 in alignment score) than that between Drosophila TBP and human TBP (88 in alignment score). Multiple amino acids critical for the TBP function were deleted or substituted in TLP. We suggest that TLP is not a bona fide vertebrate counterpart nor a direct descendant of TRF.

Vertebrate TBP-like protein (TLP/TRF2/TLF) stimulates TATA-less terminal deoxynucleotidyl transferase promoters in a transient reporter assay, and TFIIA-binding capacity of TLP is required for this function

The TBP-like protein (TLP/TRF2/TLF), which belongs to the TBP family of proteins, is present in all metazoan organisms. Although the human TLP has been reported to interfere with transcription from TATA-containing promoters, the transcription activation potential of TLP in higher animals is obscure. We previously demonstrated that artificially promoter-recruited TLP behaves like an unconventional transcriptional activator. In this study, we investigated the effects of TLP on TATA-less promoters of mouse and human terminal deoxynucleotidyl transferase (TdT) genes by transient reporter assays. As expected, TLP repressed both basal and activator-augmented transcription from the TATA-containing adenovirus major late promoter (MLP) and E1B promoter. On the other hand, however, TLP significantly stimulated both basal and activated transcription from TdT promoters. We investigated the strength of the promoters in chicken DT40 cells that lack the TLP gene. The MLP showed higher activity but the TdT promoter showed lower activity in TLP-null cells than in the wild-type cells. Moreover, ectopic expression of mouse TLP in the TLP-null cells considerably stimulated the TdT promoter. Insertion of a TATA element upstream from the TdT core promoter resulted in a loss of TLP-mediated activation. The mouse TLP was demonstrated to bind specifically to TFIIA with greater strength than TBP. We constructed mutated TLPs having amino acid substitutions that impair TFIIA binding. A representative TLP mutant lacking TFIIA-binding ability could not stimulate transcription from the TdT promoter, whereas that mutation suppressed TLP-mediated transcription repression of TATA promoters. The results of the present study suggest that the vertebrate TLP potentiates exogenous TATA-less promoters and that TFIIA plays an important role in the TLP function.

TBP-like protein (TLP/TLF/TRF2) artificially recruited to a promoter stimulates basal transcription in vivo

Metazoan genomes generally contain one TBP-related gene designated as TBP-like protein (TLP/TLF/TRF2). Although TLP is thought to work for transcriptional regulation, its natural function has not been clearly demonstrated. Here we describe the stimulation of transcription from TATA-containing and TATA-less class II promoters by artificially recruited mammalian TLP. TLP fused with Gal4 DNA-binding domain stimulated transcription when it was recruited at a proximal promoter. Compared to TBP, stimulation by TLP was less TATA-dependent. Slight truncation from each terminus of TLP destroyed this function drastically. Amino acid substitutions of TLP whose corresponding residues in TBP are crucial for its function resulted in the loss of function. Consequently, Gal4-fused TLP was demonstrated to exhibit ability of transcription activation irrespective of the type of promoter, the mechanism of which was thought to be similar to that of artificially recruited TBP. TLP is presumably able to behave as a transcriptional activator in cells.

Analysis of the chicken TBP-like protein(tlp) gene: evidence for a striking conservation of vertebrate TLPs and for a close relationship between vertebrate tbp and tlp genes

TLP (TBP-like protein), which is a new protein dis-covered by us, has a structure similar to that of the C-terminal conserved domain (CCD) of TBP, although its function has not yet been elucidated. We isolated cDNA and genomic DNA that encode chicken TLP (cTLP) and determined their structures. The predicted amino acid sequence of cTLP was 98 and 91% identical to that of its mammalian and Xenopus counterparts, respectively, and its translation product was ubiquitously observed in chicken tissues. FISH detection showed that chicken tlp and tbp genes were mapped at 3q2.6-2.8 and 3q2.4-2.6 of the same chromosome, respectively. Genome analysis revealed that the chicken tlp gene was spliced with five introns. Interestingly, the vertebrate tbp genes were also found to be split by five introns when we focused on the CCDs, and their splicing points were similar to those of tlp. On the contrary, another TBP-resembling gene of Drosophila, trf1, is split by only one intron, as is the Drosophila 's tbp gene. These results support our earlier assumption that vertebrate TLPs did not directly descend from Drosophila TRF1. On the basis of these results together with phylogenetical exam-ination, we speculate that tlp diverged from an ancestral tbp gene through a process of gene duplication and point mutations.

HP33: hepatocellular carcinoma-enriched 33-kDa protein with similarity to mitochondrial N-acyltransferase but localized in a microtubule-dependent manner at the centrosome

Using a new subtraction method and chemically induced rat hepatocellular carcinomas, we identified a hepatocellular carcinogenesis and hepatocyte proliferation-related gene designated hp33 that encoded a 33-kDa protein. The predicted protein was similar to the bovine aralkyl N-acyltransferase and arylacetyl N-acyltransferase. HP33 was restrictively expressed in the liver and kidney, and its gene expression was stimulated in the regenerating liver as well as in hepatocellular carcinoma. Interestingly, it was demonstrated in various hepatic cells that HP33 was localized in regions surrounding the centrosome, where mitochondria were not concentrated. Moreover, its centrosomal localization was evident in the interphase but not in the mitotic phase of the cell cycle. The centrosomal localization of HP33 was dependent on microtubules, and ectopically expressed HP33 was seen at centrosomes even in fibroblasts, which do not exhibit a typical staining pattern of HP33. The centrosomal localization of HP33 became invisible by nocodazole treatment, whereas the mitochondrial staining pattern was not affected by it. In vitro cosedimentation experiments using purified microtubules indicated that HP33 bound to MTs directly and that its MT-binding ability was dependent on the C-terminal basic domain of the protein. These results suggest that, different from early predictions based on its primary structure, HP33 has a growth- and carcinogenesis-related function that may be independent of mitochondrial function.

Loss of Raf-1-binding activity of v-Ha-Ras by the deletion of amino acid residues 64-72 and 143-151

In order to elucidate the molecular events in signal transduction, examination of the interaction between Ras and Raf-1 seems crucial. Many Raf-1 mutants have been investigated in terms of their binding activities to Ras, where only a few Ras mutants have been examined thus far. We have investigated the Raf-1-binding activities of v-Ha-Ras and 21 insertion/deletion mutants of this protein. The results show that the mutants have varying levels of Raf-1-binding activity that are related neither to their transforming activity nor to their guanine nucleotide-binding activity. Deletion in the effector domain of Ras did not completely abolish Raf-1-binding, whereas the deletion in amino acid residues 64-72 or 143-151 resulted in complete loss of Raf-1-binding activity.

Cloning of a novel rat gene, DB83, that encodes a putative membrane protein

Using a partial cDNA sequence and a 5'-RACE technique, we isolated a novel cDNA from rat liver referred to as DB83. DB83 had four hydrophobic trans-membrane domains and one N-myristoylation site as well as multiple possible phosphorylation sites. The db83 gene was highly expressed in the liver and significantly in brain, lungs and kidneys. We suggest that DB83 is a tissue-specific putative membrane protein. p6

Structure of rat gamma-tubulin and its binding to HP33

Gamma-tubulin is localized at the microtubule organizing center and is thought to participate in the organizing of the microtubule network. In this study, we isolated a cDNA of rat gamma-tubulin. The rat gamma-tubulin cDNA encoded 451 amino acids, the same number as that of its counterpart in other vertebrates, and its structure was found to be highly conserved in vertebrates. In a previous work, we identified HP33 (hepatocarcinogenesis- and hepatocellular proliferation-related 33-kDa protein) that was localized at the centrosome of hepatic cells and that exhibited MAP-like activity. In vitro GST pull-down assay using highly purified recombinant HP33 and bacterially expressed gamma-tubulin demonstrated that HP33 bound to gamma-tubulin directly. These results suggest that HP33 is localized at the centrosome via association with both the microtubule and its minus end-specific component, gamma-tubulin.

Expression of mRNA of chemokine receptor CXCR4 in feline mammary adenocarcinoma

The expression of mRNA of the chemokine receptor CXCR4 in 65 surgically resected mammary adenocarcinomas from cats was investigated by in situ hybridisation. No expression of the receptor's mRNA was detectable in the mammary tissue of healthy cats, but it was expressed in areas adjacent to necrosis, surrounding blood vessels and cells infiltrating the lymphatics of 47 (72.3 per cent) of the 65 samples. There was a significant relationship between lymphatic infiltration by neoplastic cells and the expression of the receptor's mRNA (P < 0.005), but there was no significant relationship between its expression and the one-year survival of the cats.

Overexpression of cysteine sulfinic acid decarboxylase stimulated by hepatocarcinogenesis results in autoantibody production in rats

We developed a novel and efficient cDNA subtraction method to isolate rat hepatocellular carcinoma (HCC)-related genes. cDNAs from Solt-Farber procedure-driven HCCs were synthesized on Latex beads. The subtraction was accomplished by a simple centrifugation, PCR amplification, and dot blot screening. Among 2000 clones from the subtracted cDNA library, one clone with a full-length HCC-related cDNA was eventually obtained. Sequence analysis of this clone showed it to exhibit 90 and 60% similarity with the rat cysteine sulfinic acid decarboxylase (CSAD) and mammalian glutamic acid decarboxylases (GAD), respectively. Differences between our sequence data on CSAD and those reported previously were observed at two positions, which arose from a single amino acid substitution and frame shift mutation. The CSAD expression was restricted to the liver and kidney of rats. During hepatocarcinogenesis, expression of the CSAD mRNA and its protein was stimulated in the precancerous liver and maintained its high expression afterward. Interestingly, a high level of anti-CSAD autoantibody was detected in the HCC-bearing rats. The titer of anti-CSAD autoantibodies in these rats was 30-200 times higher than that in normal rats. The anti-CSAD autoantibody appeared in the precancerous state and was maintained afterward, and its pattern of appearance was similar to that of CSAD mRNAs and proteins. Thus, we propose that the high-titer CSAD autoantibody resulted from increased CSAD gene expression in the liver due to stimulation by the HCC. These results remind us of human autoimmune diseases including insulin-dependent diabetes mellitus and stiff-man syndrome, which are caused by autoantibodies against GAD.

Gene expression in hepatomas

Gene expression changes in accordance with cell growth, differentiation and carcinogenesis. To elucidate the molecular mechanisms for hepatocarcinogenesis as well as maintenance of normal hepatocytes, it is important to identify the genes that have altered expression with carcinogenesis. We established a new and efficient cDNA subtraction method via two cDNA populations. By using this method along with rat hepatomas made by the Solt-Farber protocol, we identified a number of genes, some of which are activated in hepatocellular carcinoma (HCC). These genes include ones which code for a transcription factor and a metabolic enzyme. One particular gene can be used as a tumour marker. Our method is beneficial for the isolation of a wide range of HCC-related genes in rats which, in turn, enables easy identification of their human counterparts. In this review, we describe details of our method and the isolated genes. We also briefly describe transcription factors in the liver.