Double-strand break repair based on short-homology regions is suppressed under terminal deoxynucleotidyltransferase expression, as revealed by a novel vector system for analysing DNA repair by nonhomologous end joining.

We have constructed a novel, nonhomologous end-joining (NHEJ) assay vector (NAV), containing mKate2, Venus and ccdB genes. Cotransfection of NAV with a construct expressing the restriction enzyme I-SceI generated a double-strand break (DSB) in NAV that excised mKate2 and ccdB. Repair of this DSB produced an intact vector that expressed Venus, a green fluorescent protein. Because cells bearing the repaired NAV lacked the ccdB gene which slows cell proliferation, the cultures were enriched in cells containing repaired DSBs. DNA sequence analysis of the DSB junctions indicated that the repair was carried out mainly by using the closest homology sequence. Use of the NAV yielded rapid results within 3 days after transfection. We then used the NAV to analyse NHEJ in cells overexpressing terminal deoxynucleotidyltransferase (TdT). The results indicated that TdT suppresses DNA repair that is based on short (one- or two-base) homology regions, to efficiently add deoxynucleotides during VDJ recombination in lymphoid cells.

BTB-ZF Protein Znf131 Regulates Cell Growth of Developing and Mature T Cells.

Many members of the BTB-ZF family have been shown to play important roles in lymphocyte development and function. The role of zinc finger Znf131 (also known as Zbtb35) in T cell lineage was elucidated through the production of mice with floxed allele to disrupt at different stages of development. In this article, we present that Znf131 is critical for T cell development during double-negative to double-positive stage, with which significant cell expansion triggered by the pre-TCR signal is coupled. In mature T cells, Znf131 is required for the activation of effector genes, as well as robust proliferation induced upon TCR signal. One of the cyclin-dependent kinase inhibitors, p21(Cip1) encoded by cdkn1a gene, is one of the targets of Znf131. The regulation of T cell proliferation by Znf131 is in part attributed to its suppression on the expression of p21(Cip1).

Definition of the transcription factor TdIF1 consensus-binding sequence through genomewide mapping of its binding sites.

TdIF1 was originally identified as a protein that directly binds to terminal deoxynucleotidyltransferase, TdT. Through in vitro selection assays (SELEX), we recently showed that TdIF1 recognizes both AT-tract and a specific DNA sequence motif, 5'-TGCATG-3', and can up-regulate the expression of RAB20 through the latter motif. However, whether TdIF1 binds to these sequences in the cells has not been clear and its other target genes remain to be identified. Here, we determined in vivo TdIF1-binding sequences (TdIF1-invivoBMs) on the human chromosomes through ChIP-seq analyses. The result showed a 160-base pair cassette containing 'AT-tract~palindrome (inverted repeat)~AT-tract' as a likely target sequence of TdIF1. Interestingly, the core sequence of the palindrome in the TdIF1-invivoBMs shares significant similarity to the above 5'-TGCATG-3' motif determined by SELEX in vitro. Furthermore, spacer sequences between AT-tract and the palindrome contain many potential transcription factor binding sites. In luciferase assays, TdIF1 can up-regulate transcription activity of the promoters containing the TdIF1-invivoBM, and this effect is mainly through the palindrome. Clusters of this motif were found in the potential target genes. Gene ontology analysis and RT-qPCR showed the enrichment of some candidate targets of TdIF1 among the genes involved in the regulation of ossification. Potential modes of transcription activation by TdIF1 are discussed.

Truncated UDP-glucuronosyltransferase (UGT) from a Crigler-Najjar syndrome type II patient colocalizes with intact UGT in the endoplasmic reticulum.

Mutations in the gene encoding bilirubin UDP-glucuronosyltransferase (UGT1A1) are known to cause Crigler-Najjar syndrome type II (CN-II). We previously encountered a patient with a nonsense mutation (Q331X) on one allele and with no other mutations in the promoter region or other exons, and proposed that CN-II is inherited as a dominant trait due to the formation of a heterologous subunit structure comprised of the altered UGT1A1 gene product (UGT1A1-p.Q331X) and the intact UGT1A1. Here, we investigated the molecular basis of CN-II in this case by expressing UGT1A1-p.Q331X in cells. UGT1A1-p.Q331X overexpressed in Escherichia coli or mammalian cells directly bound or associated with intact UGT1A1 in vitro or in vivo, respectively. Intact UGT1A1 was observed as a dimer using atomic force microscopy. Fluorescent-tagged UGT1A1-p.Q331X and intact UGT1A1 were colocalized in 293T cells, and fluorescence recovery after photobleaching analysis showed that UGT1A1-p.Q331X was retained in the endoplasmic reticulum (ER) without rapid degradation. These findings support the idea that UGT1A1-p.Q331X and UGT1A1 form a dimer and provide an increased mechanistic understanding of CN-II.

Evaluation and identification of hepatitis B virus entry inhibitors using HepG2 cells overexpressing a membrane transporter NTCP.

Hepatitis B virus (HBV) entry has been analyzed using infection-susceptible cells, including primary human hepatocytes, primary tupaia hepatocytes, and HepaRG cells. Recently, the sodium taurocholate cotransporting polypeptide (NTCP) membrane transporter was reported as an HBV entry receptor. In this study, we established a strain of HepG2 cells engineered to overexpress the human NTCP gene (HepG2-hNTCP-C4 cells). HepG2-hNTCP-C4 cells were shown to be susceptible to infection by blood-borne and cell culture-derived HBV. HBV infection was facilitated by pretreating cells with 3% dimethyl sulfoxide permitting nearly 50% of the cells to be infected with HBV. Knockdown analysis suggested that HBV infection of HepG2-hNTCP-C4 cells was mediated by NTCP. HBV infection was blocked by an anti-HBV surface protein neutralizing antibody, by compounds known to inhibit NTCP transporter activity, and by cyclosporin A and its derivatives. The infection assay suggested that cyclosporin B was a more potent inhibitor of HBV entry than was cyclosporin A. Further chemical screening identified oxysterols, oxidized derivatives of cholesterol, as inhibitors of HBV infection. Thus, the HepG2-hNTCP-C4 cell line established in this study is a useful tool for the identification of inhibitors of HBV infection as well as for the analysis of the molecular mechanisms of HBV infection.

TdIF1 recognizes a specific DNA sequence through its Helix-Turn-Helix and AT-hook motifs to regulate gene transcription.

TdIF1 was originally identified as a protein that directly binds to DNA polymerase TdT. TdIF1 is also thought to function in transcription regulation, because it binds directly to the transcriptional factor TReP-132, and to histone deacetylases HDAC1 and HDAC2. Here we show that TdIF1 recognizes a specific DNA sequence and regulates gene transcription. By constructing TdIF1 mutants, we identify amino acid residues essential for its interaction with DNA. An in vitro DNA selection assay, SELEX, reveals that TdIF1 preferentially binds to the sequence 5'-GNTGCATG-3' following an AT-tract, through its Helix-Turn-Helix and AT-hook motifs. We show that four repeats of this recognition sequence allow TdIF1 to regulate gene transcription in a plasmid-based luciferase reporter assay. We demonstrate that TdIF1 associates with the RAB20 promoter, and RAB20 gene transcription is reduced in TdIF1-knocked-down cells, suggesting that TdIF1 stimulates RAB20 gene transcription.

BRCT domain of DNA polymerase µ has DNA-binding activity and promotes the DNA polymerization activity.

DNA polymerase µ (pol µ) catalyzes nonhomologous end-joining in DNA double-stranded break repair. Pol µ consists of an amino-terminal BRCA1 carboxyl-terminal homology (BRCT) domain and a pol ß-like region, which contains the catalytic site. By DNA cellulose column chromatography, using full-length pol µ and five different deletion mutants, we found that the amino-terminal region has double-stranded DNA (dsDNA)-binding activity. Pol µ without BRCT domain reduces the DNA polymerization activity when compared to full-length pol µ. Observation by atomic force microscopy showed that full-length pol µ binds to the ends and middle part of dsDNA. Pol µ lacking the amino-terminal region or with a mutation within the BRCT domain bound only to DNA ends, whereas the amino-terminal region with the BRCT domain bound to both the ends and the middle part of dsDNA (mpdDNA). Terminal deoxynucleotidyltransferase, which, like pol µ, belongs to the X family DNA polymerases, also bound to mpdDNA through its amino-terminal region.

Ubiquitylation of terminal deoxynucleotidyltransferase inhibits its activity.

Terminal deoxynucleotidyltransferase (TdT), which template-independently synthesizes DNA during V(D)J recombination in lymphoid cells, is ubiquitylated by a BPOZ-2/Cul3 complex, as the ubiquitin ligase, and then degraded by the 26 S proteasome. We show here that TdT is ubiquitylated by the Cul3-based ubiquitylation system in vitro. Because TdT could also be ubiquitylated in the absence of Cul/BPOZ-2, we determined that it could also be directly ubiquitylated by the E2 proteins UbcH5a/b/c and UbcH6, E3-independently. Furthermore, the ubiquitylated TdT inhibited its nucleotidyltransferase activity.

Heterocomplex formation by Arp4 and ß-actin is involved in the integrity of the Brg1 chromatin remodeling complex.

Although nuclear actin and Arps (actin-related proteins) are often identified as components of multi-protein chromatin-modifying enzyme complexes, such as chromatin remodeling and histone acetyltransferase (HAT) complexes, their molecular functions still remain largely elusive. Here, we investigated the role of human Arp4 (BAF53, also known as actin-like protein 6A) in Brg1-containing chromatin remodeling complexes. Depletion of Arp4 by RNA interference impaired the integrity of these complexes and accelerated the degradation of Brg1, indicating a crucial role in their maintenance, at least in certain human cell lines. We further found that Arp4 can form a heterocomplex with ß-actin. Based on structural similarities between conventional actin and Arp4, and the assumption that actin-Arp4 binding might mimic actin-actin binding, we introduced a series of mutations in Arp4 that might be expected to impair its interaction with ß-actin. Some of them indeed caused reduced binding to ß-actin. Interestingly, such mutant Arp4 proteins also showed reduced incorporation into Brg1 complexes, and their interaction with Myc-associated complexes as well as Tip60 HAT complexes were also impaired. Based on these findings, we propose that ß-actin-Arp4 complex formation might be a crucial feature in some chromatin-modifying enzyme complexes, such as the Brg1 complex.

TdIF2 is a nucleolar protein that promotes rRNA gene promoter activity.

Terminal deoxynucleotidyltransferase (TdT) interacting factor 2 (TdIF2) is an acidic protein that binds to TdT. TdIF2 binds to DNA and core histones and contains an acidic-amino acid-rich region in its C-terminus. It has therefore been suggested to function as a histone chaperone within the nucleus. TdIF2 localized within the nucleolus in HEK 293T cells, and its N-terminal (residues 1-234) and C-terminal (residues 606-756) regions were crucial for the nucleolar localization. A chromatin immunoprecipitation (ChIP) assay showed that TdIF2 associated with the promoter of human ribosomal RNA genes (hrDNAP), and an in vitro luciferase assay system showed that it promoted hrDNAP activity. Using the yeast two-hybrid system with TdIF2 as the bait, we isolated the cDNA encoding HIV Tat interactive protein 60 (Tip60), which has histone acetyltransferase (HAT) activity, as a TdIF2-binding protein. TdIF2 bound to Tip60 in vitro and in vivo, inhibited the Tip60 HAT activity in vitro and co-localized with Tip60 within the nucleolus. In addition, TdIF2 promotes upstream binding factor (UBF) acetylation in vivo. Thus, TdIF2 might promote hrDNAP activity by suppressing Tip60's HAT activity and promoting UBF acetylation.

TdT interacting factor 1 enhances TdT ubiquitylation through recruitment of BPOZ-2 into nucleus from cytoplasm.

We isolated human cDNA clone encoding Bood POZ containing gene type 2 (BPOZ-2) as a gene with a product that binds to TdT interacting factor 1 (TdIF1) using a yeast two-hybrid system. BPOZ-2 is an adaptor for E3 ligase CUL3 and participates in developmental processes. The binding between BPOZ-2 and TdIF1 was confirmed by GST pull-down and immunoprecipitation assays using specific antibodies against BPOZ-2 and TdIF1 in vitro and in vivo. Although when BPOZ-2 solely was expressed in COS7 cells, BPOZ-2 was observed mainly within the cytoplasm, co-transfection of pEGFP-BPOZ-2 and pDsRed-TdIF1 into COS7 cells resulted in co-localization of EGFP-BPOZ-2 and DsRed-TdIF1 within the nucleus. TdIF1 may recruit BPOZ-2 into the nucleus from the cytoplasm by directly binding to BPOZ-2. BPOZ-2 enhanced TdT ubiquitylation when TdIF1 was expressed together with BPOZ-2 in 293T cells, strongly suggesting that the recruitment of BPOZ-2 into the nucleus from the cytoplasm is significant for the TdT ubiquitylation within the nucleus.

BPOZ-2 directly binds to eEF1A1 to promote eEF1A1 ubiquitylation and degradation and prevent translation.

Bood POZ containing gene type 2 (BPOZ-2), which contains ankyrin repeats, NLS, BTB/POZ domains and LXXLL motifs, is an adaptor protein for the E3 ubiquitin ligase scaffold protein CUL3. We isolated a cDNA encoding eukaryotic elongation factor 1A1 (eEF1A1) as a BPOZ-2 binding protein by screening a human thymus cDNA library using a yeast two-hybrid system. eEF1A1 is essential for translation and is also involved in the 26S proteasome-dependent degradation of misfolded or unfolded proteins. The binding between BPOZ-2 and eEF1A1 was confirmed by pull-down and immunoprecipitation assays in vitro and in vivo, respectively. BPOZ-2 binds to eEF1A1 through the ankyrin repeats and both BTB/POZ domains in BPOZ-2 and Domains I and III in eEF1A1. BPOZ-2 and eEF1A1 over-expressed in HEK 293T cells co-localized as speckles within the cytoplasm. BPOZ-2 promoted eEF1A1 ubiquitylation and degradation, suggesting that eEF1A1 is a substrate of BPOZ-2. BPOZ-2 inhibited GTP binding to eEF1A1 and prevented translation in in vitro translation assay using rabbit reticulocytes.

Bood POZ containing gene type 2 is a human counterpart of yeast Btb3p and promotes the degradation of terminal deoxynucleotidyltransferase.

Bood POZ containing gene type 2 (BPOZ-2) is involved in the growth suppressive effect of the phosphatase and tensin homologue (PTEN). We showed that BPOZ-2 is a human counterpart of yeast Btb3p, which is a putative adaptor for Pcu3p-based ubiquitin ligase. BPOZ-2 bound to E3 ligase CUL3 in vitro and in vivo. BPOZ-2 itself was ubiquitinated through the CUL3-based E3 ligase mainly within the nucleus and degraded by the 26S proteasome. Although BPOZ-2 was mainly expressed within the cytoplasm, it accumulated within the nucleus in the presence of the specific 26S proteasome inhibitor MG132. BPOZ-2 may be recruited to the nucleus from the cytoplasm. Terminal deoxynucleotidyltransferase (TdT) was detected as a BPOZ-2-binding protein using a yeast two-hybrid system by screening a human thymus cDNA library. TdT, BPOZ-2, and CUL3 formed a ternary complex in vivo. TdT was ubiquitinated only within the nucleus and degraded by the 26S proteasome. The ubiqutination or degradation of TdT was markedly promoted by co-expression of BPOZ-2 and CUL3 or BPOZ-2 in 293T cells, respectively.

Direct binding of ligandin to uridine 5'-diphosphate glucuronosyltransferase 1A1.

AIM: Bilirubin, a final degradation product of heme produced mainly in the spleen, is carried to the liver through its binding to albumin in the blood circulation. After its transport to hepatocytes, ligandin (glutathione S-transferase; GST) carries bilirubin to the endoplasmic reticulum (ER). uridine 5'-diphosphate-glucuronosyltransferase 1A1 (UGT1A1) glucuronidates bilirubin for solubilization in the ER. METHODS: By GST pull-down and co-immunoprecipitation assays, GSTA2, a member of the alpha-class of GST, was observed to directly bind to UGT1A1 through the region present inside the ER. RESULTS: GSTA2 was detected in the microsomal fraction together with the cytosolic fraction after hepatocyte fractionation. CONCLUSION: These results strongly suggest that bilirubin is directly delivered to UGT1A1 from ligandin for glucuronidation.

Cell density-dependent regulation of tumor necrosis factor alpha gene expression in a human hepatoma cell line.

Human tumor necrosis factor alpha (TNFalpha) is a pro-inflammatory cytokine expressed in many cell types. Although the TNFalpha gene expression in human hepatocytes has been detected previously, its regulation is not well understood yet. In this study, we demonstrated that TNFalpha gene expression in human hepatoma cell line, huH2-2, was activated as a function of cell density. TNFalpha mRNA expression was low in the low-density culture, while significantly high expression was detected in the high-density culture. Moreover, stability of TNFalpha mRNA was not changed by cell density, eliminating a possibility of post-transcriptional regulation. Antibody neutralization against human TNFalpha had no significant effect on the TNFalpha mRNA expression. A cellular factor for the TNFalpha gene expression is suggested to be accumulated in the high-density cells. Data indicate that the level of TNFalpha gene transcription is elevated by a cellular factor in a cell density-dependent manner without influencing the TNFalpha secretion under the present cell-culture conditions used.

Identification of functional domains in TdIF1 and its inhibitory mechanism for TdT activity.

TdT interacting factor 1 (TdIF1) was identified as a protein that binds to terminal deoxynucleotidyltransferase (TdT) to negatively regulate TdT activity. TdT is a template-independent DNA polymerase that catalyzes the incorporation of deoxynucleotides to the 3'-hydroxyl end of DNA templates to increase the junctional diversity of immunoglobulin or T-cell receptor (TcR) genes. Here, using bioinformatics analysis, we identified the TdT binding, DNA binding and dimerization regions, and nuclear localization signal (NLS) in TdIF1. TdIF1 bound to double-stranded DNA (dsDNA) through three DNA binding regions: residues 1-75, the AT-hook-like motif (ALM) and the predicted helix-turn-helix (HTH) motif. ALM in TdIF1 preferentially bound to AT-rich DNA regions. NLS was of the bipartite type and overlapped ALM. TdIF1 bound to the Pol beta-like region in TdT and blocked TdT access to DNA ends. In the presence of dsDNA, however, TdIF1 bound to dsDNA to release TdT from the TdIF1/TdT complex and to exhibit TdT activity, implying that active TdT released microenvironmentally concentrates around AT-rich DNA to synthesize DNA.

Cholesterol hemisuccinate: a selective inhibitor of family X DNA polymerases.

Cholesterol hemisuccinate (compound 5), which consists of succinic acid esterified to the beta-hydroxyl group of cholesterol, selectively and strongly inhibited the activities of mammalian DNA polymerases (pols) such as pol beta, pol lambda, and terminal deoxynucleotidyltransferase (TdT), which are family X pols, in vitro, and the IC50 values were 2.9, 6.3, and 6.5 microM, respectively. The compound moderately suppressed the activities of other mammalian pols such as pol A (i.e., pol gamma), pol B (i.e., pols alpha, delta, and epsilon), and pol Y (i.e., pols iota, eta, and kappa) with 50% inhibition observed at concentrations of 131, 89.2-98.0, and 120-125 microM, respectively. The compound had no influence on the activities of plant pols alpha and beta, prokaryotic pols and other DNA metabolic enzymes tested. Since other cholesterol-related compounds such as cholesterol, cholesteryl chloride, cholesteryl bromide, cholesteryl acetate, and cholesteryl-5alpha, 6alpha-epoxide (compounds 1-4 and 6, respectively) did not influence the activities of any enzymes tested, the hemisuccinate group of compound 5 could be important for inhibition of the pol X family. Surface plasmon resonance analysis demonstrated that compound 5 bound selectively to the C-terminal 31 kDa domain of pol beta and pol lambda containing a pol beta-like region. On the basis of these results, the inhibitory mechanism of compound 5 on the pol X family was discussed.

Grb2 and Gads exhibit different interactions with CD28 and play distinct roles in CD28-mediated costimulation.

Although both CD28 and ICOS bind PI3K and provide stimulatory signal for T cell activation, unlike CD28, ICOS does not costimulate IL-2 secretion. CD28 binds both PI3K and Grb2, whereas ICOS binds only PI3K. We have generated an ICOS mutant, which can bind Grb2 by replacement of its PI3K binding motif YMFM with the CD28 YMNM motif, and shown that it induces significant activation of the IL-2 promoter. However, this mutant ICOS was insufficient to activate the NF-kappaB pathway. In this study, we show that Gads, but not Grb2, is essential for CD28-mediated NF-kappaB activation, and its binding to CD28 requires the whole CD28 cytoplasmic domain in addition to the YMNM motif. Mutagenesis experiments have indicated that mutations in the N-terminal and/or C-terminal PXXP motif(s) of CD28 significantly reduce their association with Gads, whereas their associations with Grb2 are maintained. They induced strong activity of the NFAT/AP-1 reporter comparable with the CD28 wild type, but weak activity of the NF-kappaB reporter. Grb2- and Gads-dominant-negative mutants had a strong effect on NFAT/AP-1 reporter, but only Gads-dominant-negative significantly inhibited NF-kappaB reporter. Our data suggest that, in addition to the PI3K binding motif, the PXXP motif in the CD28 cytoplasmic domain may also define a functional difference between the CD28- and ICOS-mediated costimulatory signals by binding to Gads.

Beta-sitosterol-3-O-beta-D-glucopyranoside: a eukaryotic DNA polymerase lambda inhibitor.

Beta-sitosterol-3-O-beta-D-glucopyranoside (compound 1), a steroidal glycoside isolated from onion (Allium cepa L.) selectively inhibited the activity of mammalian DNA polymerase lambda (pol lambda) in vitro. The compound did not influence the activities of replicative DNA polymerases such as alpha, delta and epsilon, but also showed no effect even on the activity of pol beta which is thought to have a very similar three-dimensional structure to the pol beta-like region of pol lambda. Since parts of compound 1 such as beta-sitosterol (compound 2) and D-glucose (compound 3) did not influence the activities of any enzymes tested, the converted structure of compounds 2 and 3 might be important for pol lambda inhibition. The inhibitory effect of compound 1 on both intact pol lambda (i.e. residues 1-575) and a truncated pol lambda lacking the N-terminal BRCA1 C-terminus (BRCT) domain (133-575, del-1 pol lambda) was dose-dependent, and 50% inhibition was observed at a concentration of 9.1 and 5.4 microM, respectively. The compound 1-induced inhibition of del-1 pol lambda activity was non-competitive with respect to both the DNA template-primer and the dNTP substrate. On the basis of these results, the pol lambda inhibitory mechanism of compound 1 is discussed.

Structural relationship of curcumin derivatives binding to the BRCT domain of human DNA polymerase lambda.

We previously reported that phenolic compounds, petasiphenol and curcumin (diferuloylmethane), were a selective inhibitor of DNA polymerase lambda (pol lambda) in vitro. The purpose of this study was to investigate the molecular structural relationship of curcumin and 13 chemically synthesized derivatives of curcumin. The inhibitory effect on pol lambda (full-length, i.e. intact pol lambda including the BRCA1 C- terminal [BRCT] domain) by some derivatives was stronger than that by curcumin, and monoacetylcurcumin (compound 13) was the strongest pol lambda inhibitor of all the compounds tested, achieving 50% inhibition at a concentration of 3.9 microm. The compound did not influence the activities of replicative pols such as alpha, delta, and epsilon. It had no effect on pol beta activity either, although the three-dimensional structure of pol beta is thought to be highly similar to that of pol lambda. Compound 13 did not inhibit the activity of the C-terminal catalytic domain of pol lambda including the pol beta-like core, in which the BRCT motif was deleted from its N-terminal region. MALDI-TOF MS analysis demonstrated that compound 13 bound selectively to the N-terminal domain of pol lambda, but did not bind to the C-terminal region. Based on these results, the pol lambda-inhibitory mechanism of compound 13 is discussed.