HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE ADVANCED SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Kubota, T. Articles by Koiwai, O. Search for Related Content PubMed Articles by Kubota, T. Articles by Koiwai, O.

Department of Applied Biological Science, Faculty of Science and Technology, Tokyo University of Science, Noda, Chiba 278-8510, Japan

	Abstract

Top Abstract Introduction Results Discussion Experimental procedures References

 
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 ß-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.

	Introduction

Top Abstract Introduction Results Discussion Experimental procedures References

 
TdT interacting factor 1 (TdIF1; Dnttip1) was first identified as a TdT-binding protein by screening a human thymus cDNA library using a yeast two-hybrid system and their association in bovine thymus was detected in vivo (Yamashita et al. 2001). TdIF1 is a 37-kDa DNA-binding protein that has a high degree of amino acid sequence homology with the C-terminal region of oncofetal protein p65, which has been reported to belong to a nuclear receptor superfamily (Hanausek et al. 1996). TdIF1 contains two regions similar to the AT-hook motif, which is generally composed of a positively charged stretch of nine amino acids containing the invariant core peptide sequence RGRP flanked by basic residues (Aravind & Landsman 1998). The AT-hook motif, which was first described in the high-mobility-group non-histone chromosomal protein HMGA, binds to AT tracts in the minor groove of DNA (Solomon et al. 1986; Reeves & Nissen 1990). Recently, in addition to terminal deoxynucleotidyltransferase (TdT; Dntt), TdIF1 has also been reported to bind to TReP-132, which functions as a transcriptional co-activator of steroidogenic factor 1 to induce the P450scc gene expression in steroid-hormone-producing cells (Gizard et al. 2001; Fujisaki et al. 2006). Although both TdIF1 and TReP-132 bind to TdT, these three proteins do not form a ternary complex (Fujisaki et al. 2006).

TdT is a template-independent DNA polymerase that catalyzes the polymerization of deoxynucleotides to the 3'-OH end of DNA in a distributive manner. The repertoires of immunoglobulin (Ig) or T-cell receptor (TcR) genes are enhanced with TdT by adding the extra nucleotides (N nucleotides) to the junction between V and D, D and J, or V and J DNA segments during V(D)J recombination (Gilfillan et al. 1993; Komori et al. 1993). TdT belongs to the family X DNA polymerases, a subgroup of an ancient nucleotidyltransferase superfamily, which is defined by amino acid sequence homologies in their nucleotide-binding domains and active-site motifs (Pol X motifs) (Holm & Sander 1995). DNA polymerase ß, µ and (Pol ß, Pol µ and Pol , respectively) also belong to the family X DNA polymerases (Aoufouchi et al. 2000; Dominguez et al. 2000; Nagasawa et al. 2000; Shimazaki et al. 2002), the C-termini of which have a Pol ß-like region, which contains a DNA-binding motif and a Pol X motif as a catalytic site. The DNA-binding motif consists of two helix-hairpin-helix (HhH) motifs. A crystal structure analysis of murine TdT showed that TdT binds to the 3'-OH end of single-stranded DNA (ssDNA) through the second HhH motif (Delarue et al. 2002). Similarly to the Pol ß-like region, TdT, Pol µ and Pol have a nuclear localization signal (NLS) motif and a BRCA1 C-terminal domain (BRCT domain). The BRCT domain is considered to mediate protein–protein or protein–DNA interactions in DNA repair or cell cycle checkpoint regulation in DNA damage (Huyton et al. 2000). TdT binds to Ku70/Ku86 through its BRCT domain (Mahajan et al. 1999; Purugganan et al. 2001). Recently, we have found that, in addition to Ku70/Ku86, TdIF1, TdIF2, PCNA or TReP-132 binds to TdT in vitro and in vivo to negatively regulate TdT activity in vitro (Ibe et al. 2001; Yamashita et al. 2001; Fujita et al. 2003; Fujisaki et al. 2006).

Here, we performed a structural analysis of TdIF1 to identify the functional domains for nuclear localization, DNA binding, TdT binding and dimerization, and predicted these domains on a 3D structure model of TdIF1. In addition to the structural analysis, we studied the regulatory mechanism of TdIF1 for TdT activity in vitro.

	Results

Top Abstract Introduction Results Discussion Experimental procedures References

 
Multiple sequence alignments of TdIF1 orthologues

We isolated human cDNA encoding TdIF1 as a TdT-binding protein using a yeast two-hybrid system (Yamashita et al. 2001). Here, we initially searched for TdIF1 orthologue amino acid sequences from the NCBI non-redundant database using PSI-BLAST search. Proteins homologous to human TdIF1 were found in different species from nematodes, for example, Caenorhabditis elegans, to higher vertebrates. Figure 1 shows the multiple amino acid sequence alignments of TdIF1 orthologues generated using the CLUSTALW program. Identical and similar amino acids are shown in black and gray, respectively. The regions corresponding to human TdIF1 residues 66–86 and 194–297 are highly conserved in all the species represented. Human TdIF1 has a high degree of amino acid sequence homology to vertebrate TdIF1s (97% similar and 96% identical to mouse TdIF1, 78% similar and 69% identical to Xenopus laevis TdIF1, 76% similar and 63% identical to pufferfish Tetraodon nigroviridis TdIF1), and is also similar to non-vertebrate TdIF1s (purple sea urchin Strongylocentrotus purpuratus TdIF1 (53%), Drosophila melanogaster TdIF1 (52%) and C. elegans TdIF1 (46%)). Note that TdIF1 orthologues are found even in non-vertebrates, which do not undergo V(D)J recombination, suggesting that TdIF1 regulates the DNA polymerization activities of the family X DNA polymerases other than TdT, or is generally involved in non-homologous end joining. Alternatively, based on the finding that TdIF1 binds to the transcriptional co-activator TReP-132 (Fujisaki et al. 2006), TdIF1 could function as a transcriptional regulator.

View larger version (62K): [in this window] [in a new window]   Figure 1  Alignment of TdIF1 orthologues and secondary-structure prediction of human TdIF1. Homo sapiens (human), Pan troglodytes (chimpanzee), Pongo pygmaeus (orangutan), Mus musculus (mouse), Rattus norvegicus (rat), Bos taurus (bovine), Canis familiaris (dog), Xenopus laevis (frog), Tetraodon nigroviridis (pufferfish), Danio rerio (zebrafish), Strongylocentrotus purpuratus (purple sea urchin), Drosophila melanogaster (fly) and Caenorhabditis elegans (nematodes) TdIF1s were aligned, with identical and similar amino acids shown in black and gray, respectively. The secondary structure predicted using the PSIPRED server is displayed on the top of the alignment. The -helix and ß-strand are indicated by the column and thick arrow, respectively. The asterisks on the top of the alignment indicate the predicted disordered region. The numbers on the top of the alignment indicate the number of amino acid residues of human TdIF1. ALM and HTH indicate the AT-hook-like motif and the predicted helix-turn-helix motif, respectively. The underline indicates the AT-hook motif in C. elegans TdIF1. Residues 304–319 of D. rerio TdIF1 and residues 1–385 and 774–883 of D. melanogaster TdIF1 are not displayed.

TdIF1 preferentially binds to AT-rich DNA through AT-hook-like motif (ALM)

TdIF1 exhibits DNA binding activity and contains two ALMs (ALM1, 145-GIKRGRQAE-153; ALM2, 165-KKRKGRPPG-173) (Yamashita et al. 2001). Thus, we attempted to determine whether TdIF1 binds to AT-rich DNA through ALMs by GST pull-out assay (Fig. 2A(i)) using GST-TdIF1 and pcDNA3.1 digested with HaeIII (pcDNA/HaeIII). As shown in Fig. 2A(ii), TdIF1 strongly bound to DNA fragments of 317, 455, 692 and 752 bp. By DNA sequence analysis, the DNA fragments were found to be AT-rich (Fig. 2A(iv)). We confirmed that GST-TdIF1 was coupled to glutathione beads by Western blotting as shown in Fig. 2A(iii) (lane 1). Next, we carried out electrophoretic mobility shift assay (EMSA) to confirm that TdIF1 directly binds to oligo(dA)-oligo(dT) (AT-DNA) using ³²P-labeled AT-DNA. As shown in Fig. 2B (lane 2), TdIF1 bound to AT-DNA. The bands of TdIF1/DNA complexes decreased in size with increasing amount of cold AT-DNA (lanes 3 and 4), but not with increasing amount of cold oligo(dG)-oligo(dC) (GC-DNA) (lanes 5 and 6), demonstrating that TdIF1 specifically binds to AT-DNA. To determine whether ALM1 or ALM2 confers DNA binding activity, we constructed vectors encoding GST-ALM1 (ALM1 peptide: 145-GIKRGRQAEEECAH-158) and GST-ALM2 (ALM2 peptide: 159-RGSPLPKKRKGRPPG-173), expressed them in Escherichia coli, and performed GST pull-out assay. We confirmed that GST-ALM1 and GST-ALM2 were coupled to glutathione beads by Western blotting (Fig. 2A(iii) (lanes 2 and 3)). As shown in Fig. 2A(ii) (lanes 12 and 13), only ALM2 confers DNA binding activity. The consensus amino acid sequence of the core motif in the AT-hook is RGRP, whereas the amino acid sequences of the core motif in ALM1 and ALM2 are RGRQ and KGRP, respectively. Because the fourth P is considered to be crucial in stabilizing the structure of the AT-hook (Aravind & Landsman 1998), the reason ALM1 did not bind to AT-rich DNA might be the replacement of the fourth P by Q.

View larger version (25K): [in this window] [in a new window]   Figure 2  TdIF1 preferentially binds to AT-rich dsDNA through ALM2. (A) (i) Schematic flowchart of GST pull-out assay. (ii) DNA fragments bound to GST-TdIF1 were sequentially eluted with buffer A containing 100, 150, 200, 250, 300, 350 or 600 mM NaCl (lanes 3–9, respectively). DNA fragments bound to GST-ALM1 (lane 12) and GST-ALM2 (lane 13) peptides were eluted with buffer A containing 350 mM NaCl. Each eluate was analyzed by 5% PAGE followed by silver staining. Lanes 1 and 10 contained a 200-bp DNA ladder marker. Lanes 2 and 11 contained 1/5 the amount of pcDNA/HaeIII used in the reaction. (iii) GST-fused proteins coupled to beads were eluted with SDS sample buffer. One microgram out of the eluate (4 µg) was analyzed by SDS-PAGE and Western blotting with a monoclonal anti-GST antibody. The arrow indicates full-length GST-TdIF1. (iv) Content of consecutive A or T sequences in each DNA fragment. Six and more consecutive A or T nucleotides were counted. Asterisks indicate AT-rich DNA fragments, which strongly bound to TdIF1. (B) EMSA. The reaction mixture contains 100 ng of His-TdIF1 (lanes 2–6) and 10 000 cpm of 32P-labeled 32 bp AT-DNA. As competitors, 10 and 50 ng of cold AT-DNA (lanes 3 and 4, respectively), and 10 and 50 ng of cold GC-DNA (lanes 5 and 6, respectively) were added to the reaction mixture. The black arrow indicates the TdIF1/DNA complex. (C) Schematic view of DNA substrate and AFM images of TdIF1/DNA complex. The black arrows show the length of the fragment and the position of consecutive A or T sequences from the nearest end. The left and right black boxes in the diagram indicate 5'-TAATTTTTTTTATTTAT and 5'-TTTTTT, respectively. TdIF1 on DNA is indicated by the white arrow. Bars, 100 nm.

TdIF1 binds to DNA through three DNA binding domains

In general, the AT-hook motif is considered to be an auxiliary protein motif cooperating with other DNA binding activities (Aravind & Landsman 1998). Therefore, to identify other DNA binding regions in TdIF1, we first performed GST pull-out assay by constructing eight TdIF1 deletion mutants based on the secondary structure of TdIF1 (Fig. 3A). As shown in Fig. 3B, 1–183 with ALM2 and 184–329 without ALM2 bound to DNA, indicating that 184–329 includes DNA binding regions. To confine DNA binding regions in 184–329, the DNA binding activities of 184–218, 218–329 and 184–243 were examined by GST pull-out assay. As shown in Fig. 3B, 184–243 conferred DNA binding activity, whereas 184–218 and 218–329 did not, indicating that a DNA binding region exists around residue 218. On the basis of the secondary structure, this region is expected to form a helix-turn-helix (HTH) structure (Fig. 3A PSIPRED). A prediction server for HTH DNA-binding motif also showed that the region between residues 216 and 237 can form an HTH structure. Next, we determined whether 1–75 and 1–158 confer DNA binding activity. As shown in Fig. 3B(i), unexpectedly, both 1–75 and 1–158 were found to weakly bind to DNA. Thus, we determined whether 74–158 confers DNA binding activity by constructing DGD74–183 with a mis-sense mutation in ALM2. As shown in Fig. 3B(i) (74–183 and DGD74–183), no DNA binding activity was detected, indicating that 74–158 does not confer DNA binding activity. Namely, ALM2 confers a unique DNA binding activity in 74–183. We confirmed that GST-fused proteins were coupled to glutathione beads by Western blotting as shown in Fig. 3B(ii). Taken together, TdIF1 is considered to bind to DNA mainly through ALM2 and the predicted HTH motif and through residues 1–75 that confers a weak DNA binding activity.

View larger version (21K): [in this window] [in a new window]   Figure 3  TdIF1 contains three DNA binding regions. (A) Schematic representation of TdIF1 deletion mutants and their DNA binding activities. The predicted secondary structure of TdIF1 is shown on the top of the schematic structures with the predicted domains: N- and C-terminal domains and loop structure. The -helices and ß-strand are indicated by the columns and black rectangle, respectively. DNA BA indicates DNA binding activity. (++), high affinity (DNA fragments were eluted with a buffer containing more than 250 mM NaCl); (+), low affinity (150–200 mM NaCl); (–), no activity. DNA-BR indicates the DNA binding region. (B) (i) DNA binding activities of TdIF1 deletion mutants were detected by GST pull-out assay. DNA fragments bound to TdIF1 deletion mutants were sequentially eluted with buffer A containing 150, 200, 250 or 300 mM NaCl. Each eluate was analyzed by 5% PAGE, followed by silver staining. Lane M contained a 200-bp DNA ladder marker. Lane In contained 1/5 the amount of pcDNA/HaeIII used in the reaction. (ii) GST-fused proteins coupled to beads were eluted with SDS sample buffer. One microgram out of the eluate (4 µg) was analyzed by SDS-PAGE and Western blotting with a monoclonal anti-GST antibody.

TdIF1 exhibits ssDNA binding activity, as determined using a ssDNA-cellulose column (Yamashita et al. 2001). TdT adds nucleotides to the 3'-OH end of ssDNA. Therefore, TdIF1 and TdT may competitively bind to ssDNA, thereby inhibiting TdT activity in the presence of TdIF1. To test this possibility, we first examined the DNA binding activity for ssDNA by competition assay using ³²P-labeled dsAT-DNA. Labeled dsAT-DNA was added to a reaction mixture containing cold ss-oligo(dA) or ss-oligo(dT) and TdIF1. As shown in Fig. 4A(i), TdIF1 bound to dsAT-DNA (lanes 3 and 4), indicating that TdIF1 does not bind to ssDNA. In contrast to ss-oligo(dA) or ss-oligo(dT), ss-oligo(dG) or ss-oligo(dC) abolished and suppressed the binding between dsAT-DNA and TdIF1, respectively. The difference among the effects of the oligonucleotides on the binding to dsAT-DNA is probably due to the type of DNA structure used, because ss-oligo(dG) or ss-oligo(dC) is considered to form a secondary structure (Sen & Gilbert 1990; Leroy et al. 1994). To confirm our speculation, we determined whether ss-oligo(dG) or ss-oligo(dC) can form a secondary structure by agarose gel electrophoresis. As shown in Fig. 4A(ii) (lanes G and C), the bands of ss-oligo(dG) or ss-oligo(dC) forming secondary structures are strong and faint, respectively. The reason TdIF1 bound to ssDNA using a ssDNA-cellulose column may be that dG- or dC-rich regions in ssDNA formed dsDNA and TdIF1 bound to the dsDNA formed. Thus, our EMSA results show that TdIF1 does not bind to ssDNAs except ssDNA forming a secondary structure.

View larger version (61K): [in this window] [in a new window]   Figure 4  TdIF1 binds to neither ssDNA nor DNA ends. (A) (i) ssDNA binding activity of TdIF1. EMSA was carried out using 100 ng of His-TdIF1 (lanes 2–8) and 10 000 c.p.m. of 32P-labeled dsAT-DNA. His-TdIF1 was incubated with 100 ng of ss-oligo(dA) (lane 3), ss-oligo(dT) (lane 4), ss-oligo(dG) (lane 5) or ss-oligo(dC) (lane 6) or 200 ng of cold dsAT-DNA (lane 7) or cold dsGC-DNA (lane 8) for 10 min prior to the addition of labeled DNA. The asterisk indicates the position of the loading well at the top of the gel. (ii) Detection of ss-oligonucleotide-formed secondary structure. Five hundred nanograms each of ss-oligo(dA) (lane A), ss-oligo(dT) (lane T), ss-oligo(dG) (lane G) and ss-oligo(dC) (lane C) was applied to a 1.5% agarose gel, and the gel was stained with ethidium bromide. Lane M contained the /HindIII marker. (B) (i) TdIF1 does not bind to DNA ends. The DNA end binding activities of TdIF1 and Ku70/Ku86 were assayed by EMSA. All reaction mixtures in the left panel contained 300 nM biotin-labeled 32-bp DNA duplex with a 3’ (4-base) overhang. His-TdIF1 (250 nM) (lanes 2–6) was included in the reaction mixtures containing biotin-labeled DNA and 3.75 ng of pcDNA-HA/EcoRI (lane 3), pcDNA-HA/Sau3AI (lane 4), pcDNA-HA/NlaIII (lane 5) or pcDNA-HA/HaeIII (lane 6). All reaction mixtures in the right panel contained 10 nM biotin-labeled DNA. Ku70/Ku86 (13 nM) (lanes 8–10) was included in the reaction mixtures containing biotin-labeled DNA and 75 ng of pcDNA-HA/EcoRI (lane 9) or pcDNA-HA/Sau3AI (lane 10). (ii) The DNA fragments produced with EcoRI (lane 2), HaeIII (lane 3), NlaIII (lane 4) or Sau3AI (lane 5) were analyzed by 5% PAGE, followed by silver staining. Lane 1 contained a 200-bp DNA ladder marker. (iii) Properties of dsDNAs used in the reactions. RE indicates restriction enzyme.

TdIF1 is recruited into nucleus through basic bipartite NLS overlapping with ALM2

Because TdT synthesizes the N region during V(D)J recombination within the nucleus, the association between TdT and TdIF1 should be observed within the nucleus. In fact, TdIF1 solely was observed within the nucleus (Yamashita et al. 2001), and TdIF1 and TdT were co-localized within the nucleus when EGFP-TdT and DsRed-TdIF1 were co-expressed in COS7 cells (Fujisaki et al. 2006). From these observations and speculation, we expected that NLS must be present in TdIF1 and attempted to identify the region of NLS. We first observed the localization of endogenous TdIF1 using an anti-TdIF1 antibody by immunostaining. As shown in Fig. 5A, endogenous TdIF1 localized within the nucleus. Thus, we constructed eight vectors expressing EGFP-fused TdIF1 deletion mutants, transfected them into HeLa cells, and observed the localizations of the expressed mutants under a fluorescence microscope. As shown in Fig. 5B(ii), full-length EGFP-TdIF1 and del 2 localized exclusively within the nucleus. In contrast, del 1, which contains the predicted NLS (43-KHRQVQRRGR-52, as predicted by PREDICTNLS), distributed diffusively within the entire cell, indicating that the predicted NLS does not function as NLS. Although del 3 and del 5 localized within the nucleus and cytoplasm, del 4, del 6, del 7 and del 8, which contain ALM2, localized within the nucleus, strongly suggesting that the NLS is present in the region between residues 159 and 173, which overlaps with the basic amino acid region in ALM2. To further define NLS, we replaced the basic amino acid residues in ALM2 with Ala or Asp (Fig. 5C(i)), and then transfected the mutants into HeLa cells to determine localizations of the mutants. As shown in Fig. 5C(ii), two amino acid substitutions (M1 and M3) were not sufficient to inhibit the recruitment of TdIF1 from the cytoplasm to the nucleus. Additional mutations (M2 and M4) partially inhibited the recruitment of TdIF1 into the nucleus. To completely abolish the function of NLS in TdIF1, we constructed a vector expressing mutant M5, in which all the basic amino acid residues in ALM2 were replaced with Ala or Asp. However, M5 was also observed within the entire cell. Proteins smaller than 25–40 kDa can passively diffuse through aqueous channels (~10 nm in diameter) around the perimeter of a nuclear pore complex (Yoneda et al. 1999; Tran & Wente 2006). Because the estimated molecular mass of EGFP-TdIF1 is 64 kDa, EGFP-TdIF1 is too large to diffuse through the nuclear pore complex. Thus, we suspected that other amino acid residues involved in the import of TdIF1 to the nucleus exist around ALM2. From the results showing that del 8 localized within the cytoplasm as well as in the nucleus when compared with del 7 (Fig. 5B(ii)), we expected that there is a region involved in the nuclear recruitment between residues 145 and 158. Typical NLS motifs, monopartite or bipartite NLS, consist of 5–20 residues including lysine and arginine residues (Dingwall & Laskey 1991). Bipartite NLS usually consists of two interdependent positively charged clusters, which are separated by a mutation-tolerant linker region of 10–12 amino acids, as found in the NLS of nucleoplasmin (Robbins et al. 1991). The consensus sequence in bipartite NLS is 2 K/R-(10–12 aa)-3 K/R. Thus, we presumed that TdIF1 contains NLS as bipartite NLS and attempted to determine whether the conserved basic amino acid residues 147K and 148R correspond to the first cluster of bipartite NLS. We changed 147K and 148R in M5 to A (M6) and expressed M6 in HeLa cells. As shown in Fig. 5C(ii), M6 localized exclusively within the cytoplasm. Thus, we identified the NLS between residues 147 and 170 as the bipartite NLS of TdIF1. Note that the linker region (16 aa) of the bipartite NLS of TdIF1 is longer than that of the consensus sequence (10–12 aa).

View larger version (50K): [in this window] [in a new window]   Figure 5  The NLS in TdIF1 is of the bipartite type and overlaps with ALM2. (A) Localization of endogenous TdIF1 in HeLa cells. For endogenous TdIF1 visualization, HeLa cells were immunostained with a polyclonal anti-TdIF1 antibody and an Alexa Fluor 488-conjugated secondary antibody. DNA was stained with Hoechst 33258. (B) (i) Schematic structure of EGFP-fused TdIF1 deletion mutants and their localizations (Local). N and C indicate nuclear and cytoplasmic localizations, respectively. (ii) HeLa cells expressing EGFP-TdIF1 mutants were observed by fluorescence microscopy. (C) (i) Amino acid sequences of residues 145–173 in TdIF1 mutants. Letters in italics indicate ALM2. (ii) HeLa cells expressing EGFP-TdIF1WT or mutants were observed by fluorescence microscopy. (D) (i) TdIF1 is recruited into the nucleus independently of TdT. HeLa cells expressing EGFP-TdT and DsRed-TdIF1 (a), EGFP-TdT124 and DsRed-TdIF1 (b) or EGFP-TdT and DsRed-TdIF1 M6 (c) were observed under a fluorescence microscope. (ii) Detection of the activities of binding between TdT and TdIF1 M6 or between TdT124 and TdIF1 by GST pull-down assay in vitro. One microgram of purified His-TdT was mixed with GST (lane 1), GST-TdIF1 (lane 2) or GST-TdIF1 M6 (lane 3). The lysate of E. coli expressing His-TdT124 was mixed with GST (lane 5) or GST-TdIF1 (lane 6). Bound proteins were analyzed by Western blotting with anti-His (upper panel) and anti-GST (lower panel) antibodies. Lane 4 in the upper panel contained 1/10 (100 ng) the amount of purified His-TdT used in the reaction. Lane 7 in the upper panel contained 1 µg of E. coli lysate expressing His-TdT124. Lanes 4 and 7 in the lower panel are blanks.

Because TdIF1 binds to TdT in vivo (Yamashita et al. 2001), TdIF1 may be recruited into the nucleus together with TdT. To test this possibility, we constructed a vector expressing EGFP-TdT124, which does not contain NLS and the BRCT domain, and then transfected pEGFP-TdT124 and pDsRed-TdIF1 into HeLa cells to observe the localizations of EGFP-TdT124 and DsRed-TdIF1. As shown in Fig. 5D(i)(b), although TdT124 can bind to TdIF1 in vitro (Fig. 5D(ii)), EGFP-TdT124 and DsRed-TdIF1 localized exclusively within the cytoplasm and nucleus, respectively, indicating that TdIF1 is recruited into the nucleus independently of TdT. Furthermore, as shown in Fig. 5D(i)(c), EGFP-TdT and DsRed-TdIF1 M6, which has TdT binding activity, localized exclusively within the nucleus and cytoplasm, respectively. These results indicate that TdIF1 is recruited into the nucleus independently of TdT and vice versa. Namely, TdIF1 is considered to associate with TdT after each protein is recruited into the nucleus.

TdT binding region in TdIF1

We found that TdIF1 binds to TdT through the entire C-terminal 360 amino acids (Pol ß-like region) in TdT using a yeast two-hybrid system (Yamashita et al. 2001). Next, to identify the TdT binding region in TdIF1, TdT binding activities were examined using 1–183 and 184–329 (Fig. 6A) by GST pull-down assay. As shown in Fig. 6B (lanes 3 and 7), TdT bound to both TdIF1 deletion mutants. To further confine the TdT binding region in TdIF1, the TdT binding activities of five truncated mutants, 1–149, 1–75, 74–149, 184–218 and 218–329, were examined. As shown in Fig. 6B (lanes 4–6), 1–149, 1–75 or 74–149 in the N-terminal half region bound, weakly bound, and did not bind to TdT, respectively. 184–218 or 218–329 in the C-terminal region did not bind and faintly bound to TdT, respectively, strongly suggesting that the junctional region at residue 218 is essential in TdT binding. Therefore, we examined TdT binding activity by constructing 184–265. As shown in Fig. 6B (lane 10), 184–265 bound to TdT, indicating that the region around residue 218 is a TdT binding region. Finally, to determine whether the loop region confers TdT binding activity, we performed GST pull-down assay using 74–183. As shown in Fig. 6B (lane 14), TdT bound to 74–183. Taken together, TdIF1 binds to TdT through the three regions: residues 1–75 and 150–183, and around residue 218 in TdIF1.

View larger version (50K): [in this window] [in a new window]   Figure 6  TdT binding and dimerization regions in TdIF1. (A) Schematic structures and TdT binding activities of TdIF1 deletion mutants. TdT BA, TdT binding activity. Dimer A, dimerization activity. (++), positive interaction; (+), weak interaction; (+/–), slight interaction; (–), no interaction. DNA-BR, TdT-BR, NLS and Dimer-R indicate the regions for the DNA binding, TdT binding, nuclear localization and dimerization, respectively. (B) Detection of direct binding between GST-TdIF1 deletion mutants and His-TdT by GST pull-down assay. One microgram of His-TdT was mixed with GST-fused proteins bound to glutathione Sepharose beads. Bound proteins were analyzed by Western blotting with an anti-GST or anti-His antibody. The arrow indicates full-length GST-TdIF1. Lanes 11 and 15 contained 1/10 (100 ng) the amount of His-TdT used in the reaction. (C) Dimerization region in TdIF1. Detection of dimerization between GST-TdIF1 deletion mutants and His-TdIF1 by GST pull-down assay. GST-fused proteins bound to glutathione Sepharose beads were mixed with 0.5 µg of His-TdIF1. Bound proteins were analyzed by Western blotting with anti-GST or anti-His antibody. Lane 12 contained 1/5 (100 ng) the amount of His-TdIF1 used in the reaction. (D) (i) TdIF1 binds to the family X DNA polymerases. Schematic structures of the family X DNA polymerases. NLS, BRCT, HhH and the pol X motif indicate the nuclear localization signal, BRCA1 C-terminal domain, helix-hairpin-helix motif, and active center of DNA polymerization of the family X polymerases, respectively. (ii) and (iii) Detection of the activity of binding between GST or GST-TdIF1 and His-tagged family X polymerases by GST pull-down assay. GST bound to glutathione Sepharose beads were mixed with 1 µg of His-Pol ß (lane 2 in (ii), lane 1 in (iii)), His-Pol µ (lane 5 in (ii), lane 3 in (iii)), His-Pol (lane 8 in (ii), lane 5 in (iii)) or His-TdT (lane 11 in (ii), lane 7 in (iii)). GST-TdIF1 bound to glutathione Sepharose beads were mixed with 1 µg of His-Pol ß (lane 3 in (ii), lane 2 in (iii)), His-Pol µ (lane 6 in (ii), lane 4 in (iii)), His-Pol (lane 9 in (ii), lane 6 in (iii)) or His-TdT (lane 12 in (ii), lane 8 in (iii)). Bound proteins were analyzed by Western blotting with an anti-Pol ß, anti-Pol µ, anti-Pol , or anti-TdT antibody in (ii) or an anti-GST antibody in (iii). Lanes 1, 4, 7 and 10 in (ii) contained 1/20 (50 ng) the amount of purified His-Pol ß, His-Pol µ, His-Pol and His-TdT used in the reaction, respectively.

We showed that TdIF1 is eluted at monomeric and dimeric positions by gel filtration using a Superose 6 column (Yamashita et al. 2001). The cDNA that encodes TdIF1 was also isolated using a yeast two-hybrid system with TdIF1 as bait (data not shown). From these results, we presumed that TdIF1 forms a dimer and attempted to identify the dimerization region in TdIF1 by GST pull-down assay using nine deletion mutants (Fig. 6A). As shown in Fig. 6C, the regions corresponding to residues 1–75, 150–183 and 218–265 were found to confer dimerization activity.

TdIF1 binds to family X DNA polymerases in vitro

TdIF1 binds to the Pol ß-like region (residues 150–509) in TdT (Yamashita et al. 2001). The C-terminals of all the family X DNA polymerases TdT, Pol ß, Pol µ and Pol contain Pol ß-like regions (Fig. 6D(i)), which share high degree of amino acid sequence homologies and are predicted to form similar 3D structures. We then determined whether TdIF1 binds to the family X DNA polymerases by GST pull-down assay in vitro. As shown in Fig. 6D(ii) and (iii), TdIF1 bound to all the family X DNA polymerases, suggesting that TdIF1 recognizes the structure of the Pol ß-like region to regulate the activities of the family X DNA polymerases.

TdIF1 prevents TdT from accessing ssDNA

We showed that TdIF1 inhibits TdT activity using dsDNA or ssDNA as a primer in vitro (Fujisaki et al. 2006). Next, to elucidate the inhibitory mechanism, we initially investigated the dose dependence of TdIF1 on TdT activity using ssDNA or dsDNA as a primer. As shown in Fig. 7A(i) and (ii), TdT activity decreased to 10% or 5% of its maximum with increasing amount of TdIF1, when 10 pmol of TdIF1 was added to the reaction mixtures containing ssDNA and dsDNA, respectively. Furthermore, we determined whether TdIF1 inhibits TdT activity by primer extension assay. As shown in Fig. 7A(iii) (lanes 4 and 8), TdIF1 inhibited TdT activity. Because TdIF1 binds to the Pol ß-like region, which contains the DNA binding region and catalytic site, in TdT (Yamashita et al. 2001, Fig. 5D(ii) (lane 6)) and does not bind to ssDNA (Fig. 4A), TdIF1 may mask the DNA binding region or catalytic site in TdT. Before we determined whether TdIF1 actually inhibits the DNA binding activity of TdT, we first investigated the DNA binding activity of TdT or TdIF1 using ssDNA or dsDNA. TdT or TdIF1 was mixed with biotinylated ssDNA or dsDNA coupled to streptavidin–agarose. As shown in Fig. 7B(i), TdT bound to both ssDNA and dsDNA with a 3'-protrusion, whereas TdIF1 bound to only dsDNA. Thus, we determined whether TdIF1 inhibits the DNA binding activity of TdT using ssDNA. As shown in Fig. 7B(ii), the amount of TdT bound to ssDNA decreased in the presence of TdIF1 (lanes 3 and 4). Although TdIF1 did not bind to ssDNA, it inhibited the ssDNA binding activity of TdT. Furthermore, we determined whether TdIF1 inhibits ssDNA binding activity of TdT by EMSA. As shown in Fig. 7B(iii), the amount of TdT/ssDNA complex decreased in the presence of TdIF1 (lane 5). These results strongly suggest that TdIF1 masks the DNA binding region in TdT to inhibit TdT activity.

View larger version (49K): [in this window] [in a new window]   Figure 7  TdIF1 negatively regulates TdT activity. (A) TdT activity was assayed in the presence of various amounts of BSA (diamond) or TdIF1 (square) using oligo(dT)16 (i) or dsDNA with a 3'-protrusion (ii) as a primer. Various amounts (0, 1, 2, 4, 6, 8 and 10 pmol) of BSA or TdIF1 were used for each assay. (iii) Primer extension assay. TdT activity was assayed by extension of biotinylated 32 mer ssDNA or dsDNA in the presence or absence of TdIF1 as described in Experimental procedures. (B) TdIF1 prevents TdT from accessing ssDNA. (i) ssDNA and dsDNA binding activities of TdT or TdIF1. Two hundred nanograms of His-TdT was incubated with streptavidin–agarose (lane 2), biotinylated ssDNA coupled with streptavidin–agarose (lane 3), or biotinylated dsDNA with a 3'-protrusion coupled with streptavidin–agarose (lane 4). Two hundred nanograms of His-TdIF1 was incubated with streptavidin–agarose (lane 6), biotinylated ssDNA coupled with streptavidin–agarose (lane 7), or biotinylated dsDNA with a 3'-protrusion coupled with streptavidin–agarose (lane 8). After washing, bound proteins were analyzed by Western blotting using an anti-His antibody. Lane 1 contained 40 ng of His-TdT. Lane 5 contained 40 ng of His-TdIF1. (ii) ssDNA binding activity of TdT in the presence of TdIF1. Two hundred nanograms of His-TdT was incubated with streptavidin–agarose (lane 2) or biotinylated ssDNA coupled with streptavidin–agarose (lanes 3 and 4) in the absence (lane 3) or presence (lanes 2 and 4) of 200 ng of His-TdIF1. After washing, bound proteins were analyzed by Western blotting using an anti-His antibody. Lane 1 contained 40 ng each of His-TdT and His-TdIF1. (iii) EMSA. The reaction mixtures contain 400 nM bio-ssDNA (lanes 1–5), 1.6 µM His-TdT (lanes 3–5) and/or 200 nM (lane 4) or 800 nM (lanes 2 and 5) His-TdIF1. (C) dsDNA competes with TdT for TdIF1. GST-TdIF1 bound to beads was incubated with 2 µg of His-TdT. After washing, 0.25 (lanes 1 and 3) or 1 µg (lane 4) of pcDNA/HaeIII was added to the reaction mixtures and incubated with the TdIF1/TdT complex bound to beads. Bound His-TdT and DNA fragments were analyzed using anti-His antibody and silver staining, respectively. Lane 5 in the upper panel contained 100 ng of purified His-TdT. Lane M in the lower panel contained a 200-bp DNA ladder marker. Lane In in the lower panel contained 50 ng of pcDNA/HaeIII. (D) Recovery of TdT activity with addition of dsDNA to reaction mixture containing TdIF1. TdT activity was assayed in the presence of BSA and oligo(dT)16 (diamond), TdIF1 and oligo(dT)16 (square), BSA (open circle) or TdIF1 (cross). Various amounts (0, 3.9, 6.5, 13 and 26 ng) of pcDNA-HA/EcoRI were added to the reaction mixtures.

Because the dsDNA and TdT binding regions in TdIF1 overlap, we expected that dsDNA competes with TdT for the binding site in TdIF1. Thus, we performed a competition experiment with GST pull-down assay using dsDNA with a blunt end. As shown in Fig. 7C, the amount of TdT bound to TdIF1 decreased with increasing amount of dsDNA, indicating that dsDNA competes with TdT for the binding site in TdIF1. Namely, TdIF1, TdT and dsDNA do not form a ternary complex.

On the basis of the results above, when TdIF1 in the TdIF1/TdT complex binds to dsDNA, TdT is expected to be released from the TdIF1/TdT complex and consequently the released TdT exhibits its activity in the presence of ssDNA. Thus, we assayed TdT activity using dsDNA fragments generated by digesting pcDNA-HA with EcoRI in the presence of oligo(dT)₁₆ as ssDNA. Many TdIF1-binding sites were present in the dsDNA fragments, whereas only two 3' recessed ends were present in DNA fragments, which can hardly be used as primers for TdT. As shown in Fig. 7D, TdT activity recovered to 70% of its maximum with increasing amount of pcDNA-HA/EcoRI. Trace TdT activity using dsDNA as primers was detected in the absence of oligo(dT)₁₆. These findings show that the recovery of TdT activity is not caused by an increase in the number of DNA ends but by the release of TdT from the TdIF1/TdT complex, as expected. As shown in Fig. 7A(ii), when we used dsDNA with a 3'-protrusion as primers, no decrease in TdT activity was observed until more than 2 pmol of TdIF1 was added to the reaction mixture, whereas when we used ssDNA a decrease in TdT activity was observed with increasing amount of TdIF1 (Fig. 7A(i)), strongly suggesting that TdT released from the TdIF1/TdT complex adds deoxynucleotides to the ssDNA protruding from dsDNA to compensate for the decrease in TdT activity induced by TdIF1. These are very well consistent with our results showing that TdIF1 in the TdIF1/TdT complex binds to dsDNA to release TdT and the released TdT adds deoxynucleotides to ssDNA.

	Discussion

Top Abstract Introduction Results Discussion Experimental procedures References

 
When we first isolated the TdIF1 gene, we speculated that TdIF1 belongs to the nuclear receptor superfamily because of its high degree of amino acid sequence homology to p65. However, in this study, we demonstrated that TdIF1 binds to DNA mainly through ALM2 and the predicted HTH motif, which are not found in the nuclear receptor superfamily. p65 was originally reported as the member of a nuclear receptor superfamily by its high degree of homology to the DNA binding domain consisting of two C4-type Zn fingers found in an estrogen receptor (Hanausek et al. 1996). Because nuclear receptors are evolutionarily considered to be of common ancestry (Aranda & Pascual 2001), p65 should have a similar structure to members of the nuclear receptor superfamily. However, p65 share no homology to the ligand binding domain (AF-2) or the signature motif ((FW)AKxhPxFxxLxxxDQxxLL; h, hydrophobic residue; x, any residue) characteristically found in the nuclear receptor (Burke et al. 1996). Therefore, p65 is considered not to belong to the nuclear receptor superfamily. Because no proteins homologous to TdIF1 were found in the database other than p65, we conclude that TdIF1 is a novel nuclear protein, which does not belong to the nuclear receptor superfamily.

We identified the DNA binding, TdT binding and dimerization regions, and NLS in TdIF1. Here, we attempted to construct a possible 3D structure of TdIF1. Because a comparative modeling program was not available for constructing the 3D structure of TdIF1, we adopted a ROBETTA server, which can predict 3D structures based only on amino acid sequences (Kim et al. 2004). Thus, we constructed a conceivable model based on our findings and plausible hypotheses that: (i) ALM2 forms a loop structure, like an AT-hook motif (Huth et al. 1997); (ii) ALM2 and NLS localize on the surface of TdIF1, to which DNA and importin can bind, respectively; (iii) the predicted HTH motif forms a helix-turn-helix structure and extends from the surface of TdIF1 to fit into the major groove of DNA; (iv) the N- and C-terminal half regions are connected by a loop; and (v) TdIF1 binds to DNA through residues 1–75, ALM2 and the predicted HTH motif. We separately displayed the DNA binding, TdT binding and dimerization regions, and NLS on the 3D structure model in Supplementary Fig. S1. Note that such functional regions localize on the same surface in TdIF1, suggesting that each function is ingeniously regulated.

From results of our identification of the TdT binding region and 3D structure prediction analysis, TdIF1 is suggested to bind to TdT through the broad surface area in TdIF1. TdIF1 binds to TdT through its entire Pol ß-like region (Yamashita et al. 2001). Furthermore, TdIF1 prevents TdT from accessing DNA. From these findings, we propose that TdIF1 masks the DNA binding region in TdT to inhibit TdT activity. Because all the functional regions for TdT binding, DNA binding, dimerization and nuclear localization are predicted to localize in the same surface area in TdIF1, TdIF1 should exquisitely exhibit each activity in vivo. ALM2, which is one of the three DNA binding domains found, partly overlaps with NLS. Because importin , which usually recognizes NLS to transfer NLS-containing proteins from the cytoplasm to the nucleus, dissociates from such proteins after nuclear recruitment (Yoneda et al. 1999), importin should have no effect on the DNA binding activity of TdIF1 within the nucleus. That is, to ensure that the DNA binding activity is exhibited only within the nucleus, NLS overlaps with the DNA binding region. Cokol et al. reported that 90% of the proteins identified, which contain both NLS and the DNA binding region, have NLS overlapping with the DNA binding region (Cokol et al. 2000): for instance, bZip in c-Jun and c-Fos (Roux et al. 1990; Mikaelian et al. 1993); bHLH in MyoD and c-Myc (Dang & Lee 1988; Vandromme et al. 1995); the homeodomain in Oct-6, Pit-1 and TTF-1 (Theill et al. 1989; Sock et al. 1996; Christophe-Hobertus et al. 1999); the HMG box in SRY and Sox9 (Sudbeck & Scherer 1997); the paired domain in Pax8 (Poleev et al. 1997); Ets in Ets1 (Boulukos et al. 1989); and the Zn2Cys6 binuclear cluster in AlcR and Gal4 (Nelson & Silver 1989; Nikolaev et al. 2003). Thus, the structure that enables NLS to overlap with the DNA binding region is considered to be widely adopted by nuclear proteins. We also presume that importin proteins bind to the NLS of nuclear proteins within the cytoplasm and consequently mask the DNA binding region to prevent them from binding to exotic DNA, which could be introduced into the cytoplasm during viral infection.

Interestingly, as shown in Fig. 1, no amino acid sequences of ALM2 and NLS in D. melanogaster TdIF1 (DmTdIF1) are conserved. DmTdIF1 is predicted to be a fusion protein of snurportin 1 (SPN) and TdIF1. Because SPN facilitates the nuclear recruitment of UsnRNPs by binding to importin ß (Strasser et al. 2005), DmTdIF1 should be recruited into the nucleus by binding to importin ß through the SPN region in DmTdIF1. This may be why no region corresponding to NLS in DmTdIF1 is conserved. Because no ALM2 in DmTdIF1 is conserved, DmTdIF1 would not link the DNA binding activity to the AT-rich DNA region.

TdIF1 preferentially binds to AT-rich dsDNA (Fig. 2). AT-rich DNA regions are found in matrix attachment regions (MARs), which are considered to organize chromatin to a topological loop-structure by anchoring DNA to non-histone proteins in the nuclear matrix (Mirkovich et al. 1984; Cockerill & Garrard 1986). MAR is directly linked to biological activities such as replication (Vaughn et al. 1990), cell-type-specific transcription (Forrester et al. 1994), demethylation (Kirillov et al. 1996) and chromatin accessibility (Jenuwein et al. 1997). Note that MAR is found at a much higher frequency in Ig V loci than in other regions of the genome (Goebel et al. 2002) and contributes to V(D)J recombination or Ig gene transcription (Jenuwein et al. 1993, 1997; Hale & Garrard 1998; Yi et al. 1999). In addition to Ig V loci, for example, human IgH V1-69, AT-rich DNAs are also found in the non-coding regions of TcR gene loci, for example, human TcR V12-2 or TcR JP (Supplementary Fig. S2). Thus, we can expect that TdIF1 binds to AT-rich DNA regions in Ig gene loci to remodel the chromatin structure and regulate V(D)J recombination. The results showing that HMGA with AT-hooks strongly binds to MAR (Zhao et al. 1993) support our speculation. Here, on the basis of our results, it is possible to speculate for the function of TdIF1 in N-region synthesis as follows: TdIF1 in the TdIF1/TdT complex, which is an inactive form of TdT, binds to AT-rich DNAs present in the non-coding regions of Ig or TcR genes, to release TdT. Active TdT released microenvironmentally accumulates at coding ends and distributively adds deoxynucleotides to the junctions between V and D, D and J, or V and J DNA segments.

Proteins that have both AT-hook motifs and HTH-related DNA-binding domains such as the ETS domain (Kopp et al. 2007), homeodomain (Barthelemy et al. 1996), HTH-like domain (Singh et al. 2006) or pipsqueak domain (Lours et al. 2003), have been shown to be involved in transcription. TdIF1 has also been shown to bind to the transcriptional co-activator TReP-132 (Fujisaki et al. 2006), which regulates P450scc or p21^WAF1/Cip1 and p27^Kip1 gene expressions in adrenal or HeLa cells (Gizard et al. 2002; Gizard et al. 2005), respectively. Thus, we can expect that TdIF1 plays a role in transcription other than the regulatory function of TdIF1 for TdT in N-region synthesis. TdIF1 may bind to AT-rich DNAs in the promoters of Ig or TcR genes to regulate ‘germline’ transcription, which synthesizes a primary transcript from unrearranged Ig or TcR gene segments to increase the accessibility of recombinase RAG1/RAG2 to recombination signal sequences by remodeling the chromatin structure (Whitehurst et al. 1999; Oettinger 2004). Note also that although TdIF1 is mainly expressed in the thymus, it is ubiquitously detected in many organs (Fujisaki et al. 2006). TdIF1 should be involved in transcription in non-lymphoid cells. In particular, in adrenal cells, TdIF1 may work synergistically with TReP-132 as a co-activator to enhance steroidogenic factor 1 activity. We are in the process of elucidating the biological function of TdIF1 in N-region synthesis and determining whether TdIF1 is actually involved in germline transcription by producing knockout mice of the TdIF1 gene.

In conclusion, we have shown that TdIF1 binds to TdT and dsDNA through residues 1–75, ALM2 and the predicted HTH motif. NLS in TdIF1 was of the bipartite type and overlapped ALM2. We also showed that TdIF1 blocks TdT access to DNA ends to inhibit TdT activity in vitro.

	Experimental procedures

Top Abstract Introduction Results Discussion Experimental procedures References

 
Construction of expression plasmids

TdT, Pol ß, Pol µ, or Pol and TdIF1 cDNAs were subcloned into the pET28 (Novagen) and pPROEX1 (Life Technologies, Gaithersburg, MD) vectors to express His-tagged proteins in E. coli, respectively (Ibe et al. 2001; Yamashita et al. 2001; Shimazaki et al. 2002). To determine the regions for TdT binding, DNA binding, dimerization and nuclear localization in TdIF1, TdIF1 deletion mutants were subcloned into the pGEX-4T (GE Healthcare UK Ltd., Buckinghamshire, England), pEGFP or pDsRed (Clontech Laboratories, Inc., Mountain View, CA) vectors. To construct the pGEX-ALM1 peptide and the pGEX-ALM2 peptide, the annealed oligonucleotides (ALM1 oligo: 5'-GATCAAGCGTGGCCGTCAGGCAGAAGAAGAATGTGCCCATC-3' and 5'-TCGAGATGGGCACATTCTTCTTCTGCCTGACGGCCACGCTT-3', and ALM2 oligo: 5'-GATCCCCCCTTAAAAAGAGGAAAGGACGGCCTCCTGGA-3' and 5'-TCGATCCAGGAGGCCGTCCTTTCCTCTTTTTAGGAAGGGGG-3') were inserted into the pGEX-5X1 and pGEX-4T1, respectively. Site-directed mutagenesis of TdIF1 was performed by PCR using a Quick Change site directed mutagenesis kit (Stratagene, Inc., La Jolla, CA). Pairs of mutagenic primers were used for the TdIF1 mutant M1 (5'-CCCTTCCTAAAGCGGCGAAAGGACGGC-3' and 5'-GCCGTCCTTTCGCCGCTTTAGGAAGGG-3'), M2 (5'-GCCCCCTTCCTGCAGCGGCGAAAGGACGGCC-3' and 5'-GGCCGTCCTTTCGCCGCTGCAGGAAGGGGGC-3'), M3 (5'-CCTAAAAAGAGGGACGGAGACCCTCCTGGACAC-3' and 5'-GTGTCCAGGAGGGTCTCCGTCCCTCTTTTTAGG-3'), M4 (5'-CCCTTCCTAAAGCGGCGGACGGAGACC-3' and 5'-GGTCTCCGTCCGCCGCTTTAGGAAGGG-3'), M5 (5'-GCCCCCTTCCTGCAGCGGCGGACGGAGACCC-3' and 5'-GGGTCTCCGTCCGCCGCTGCAGGAAGGGGGC-3') and M6 (5'-GAGCTTCCAGGAATAGCGGCTGGCCGTCAGGCAG-3' and 5'-CTGCCTGACGGCCAGCCGCTATTCCTGGAAGCTC-3'). The substituted bases are shown in italics. The nucleotide sequences were determined by the dideoxy termination method. The cDNA fragments containing each mutation were subcloned into the pEGFP, pDsRed or pGEX vectors.

Expression and purification of proteins

Plasmids encoding the full-length and truncated TdIF1 were expressed in E. coli strain BL21 (DE3). Cells were cultured with a rotary shaker (Taiteck, Ltd., Tokyo, Japan) at 37 °C until 0.6 OD₆₀₀ and GST- or His-tagged proteins were induced with IPTG (final concentration; 0.4 mM) for 3 h at 37 °C. GST-TdIF1 1–75 and His-TdT were induced with IPTG (final concentration; 0.1 mM) for 12 h at 22 °C. After centrifugation (3000 rpm), the cells were washed with PBS and stored at –80 °C. For the purification of His-TdIF1 and His-TdT, the cells were lysed with buffer I (20 mM Tris–HCl, pH 8.0 at 4 °C, 300 mM NaCl, 5 mM ß-mercaptoethanol, 0.1% TritonX-100, 10% glycerol, 10 mM imidazole) containing 1 mM PMSF, 3 µg/mL pepstatin A and 3 µg/mL leupeptin. Soluble lysate was loaded onto Ni-NTA resin (1 mL) (Sigma, Inc., St. Louis, MO) column pre-equilibrated with buffer I. The column was washed with 25 vol. of buffer I and with 5 vol. of buffer II (20 mM Tris–HCl, pH 8.0 at 4 °C, 0.6 M NaCl, 5 mM ß-mercaptoethanol, 0.1% TritonX-100, 10% glycerol, 20 mM imidazole). His-tagged proteins bound to the column were eluted with buffer III (20 mM Tris–HCl, pH 8.0 at 4 °C, 300 mM NaCl, 5 mM ß-mercaptoethanol, 0.1% TritonX-100, 10% glycerol, 250 mM imidazole). The eluted sample (3 mL) was loaded onto HiTrap Heparin column (GE Healthcare). His-tagged proteins bound to the column were eluted with a 0.3–2.0 M NaCl gradient. His-TdIF1 was concentrated using Amicon Ultra-4 30 000 MWCO (MILLIPORE, Inc., Bedford, MA) and buffer was exchanged to buffer IV (20 mM Tris–HCl, pH 7.5 at 4 °C, 100 mM NaCl, 5 mM ß-mercaptoethanol, 0.1% TritonX-100, 50% glycerol). After purification using the Heparin column, the His-TdT fractions were loaded onto a Sephacryl-S200 gel filtration column (GE Healthcare) equilibrated with buffer V (50 mM Tris–HCl, pH 7.5, 150 mM NaCl, 5 mM ß-mercaptoethanol, 10% glycerol, 0.25% Tween20). The eluted sample was dialyzed against buffer VI (50 mM Tris–HCl, pH 7.5, 100 mM NaCl, 1 mM DTT, 50% glycerol, 0.25% Tween 20). Ku proteins were purified as described (Yumoto et al. 1998).

GST pull-out assay

pcDNA3.1(+) plasmid (Invitrogen Corp., Carlsbad, CA) was digested with HaeIII. The resulting DNA fragments (0.25 µg) were mixed with GST-TdIF1 deletion mutants bound to glutathione Sepharose 4B beads (GE healthcare) in buffer A (50 mM Tris–HCl, pH 7.4, 5 mM ß-mercaptoethanol, 0.25% Tween 20, 10% glycerol) containing 50 mM NaCl and 200 µg/mL BSA and incubated at 4 °C for 90 min. Beads were pelleted and washed 3 times in buffer A containing 100 mM NaCl to remove all the unbound DNA fragments. DNA fragments bound to the beads were sequentially eluted with buffer A containing 100–600 mM NaCl and the eluates were analyzed on a 5% non-denaturing polyacrylamide gel, followed by silver staining. Proteins bound to the beads were extracted with SDS sample buffer (50 mM Tris–HCl, pH 6.8, 6% ß-mercaptoethanol, 2% SDS, 10% glycerol, 0.002% bromophenol blue) and the eluates were analyzed on a 10% SDS-PAGE, followed by Western blotting with a monoclonal anti-GST antibody as described (Yamashita et al. 2001).

GST pull-down assay

GST or GST-TdIF1 deletion mutants bound to glutathione Sepharose beads were incubated with 1 µg each of His-TdT, His-TdIF1, His-Pol ß, His-Pol µ or His-Pol in buffer B (50 mM Tris–HCl, pH 7.5, 1 mM MgCl₂, 5 mM ß-mercaptoethanol, 10% glycerol, 0.1% Triton X-100) containing 150 mM NaCl and 400 µg/mL BSA for 60 min at 4 °C, and washed 5 times with buffer B containing 150 mM NaCl. His-tagged proteins were eluted with buffer B containing 1 M NaCl, and then GST or GST-TdIF1 deletion mutants were eluted with SDS sample buffer. These eluates were analyzed by SDS-PAGE and Western blotting with an anti-GST, anti-His (QIAGEN K.K., Tokyo, Japan), anti-TdT, anti-Pol ß (Kamiya Biomedical Co., Seattle, WA), anti-Pol µ or anti-Pol antibody.

RI-EMSA

Synthetic deoxyoligonucleotides were: oligo(dA), 5'-GACTAAAAAAAAAAAAAAAAAAAAAAAACTGA-3'; oligo(dT), 5'-TCAGTTTTTTTTTTTTTTTTTTTTTTTTAGTC-3'; oligo(dG), 5'-GACTGGGGGGGGGGGGGGGGGGGGGGGGCTGA-3'; oligo(dC), 5'-TCAGCCCCCCCCCCCCCCCCCCCCCCCCAGTC-3'. AT-DNA and GC-DNA were prepared by mixing the two strands with equimolar ratios in buffer (20 mM Tris–HCl, pH 7.5, 50 mM NaCl and 10 mM MgCl₂), heating at 95 °C for 5 min and slowly cooling. AT-DNA (10 pmol) was 5'-end-labeled with T4 polynucleotide kinase (NIPPON GENE, Co., Ltd., Toyama, Japan) and [-³²P]ATP (GE Healthcare) in 10-µL reaction mixtures containing 50 mM imidazole–HCl, pH 6.4, 18 mM MgCl₂, 5 mM DTT, 6% PEG6000, 40 µCi [-³²P]ATP, and 10 units of kinase at 37 °C for 30 min. Labeled DNAs were purified using Sephadex G-50 column (GE Healthcare).

Binding reactions were carried out at 4 °C in a mixture containing 10 mM Tris–HCl, pH 7.5, 50 mM NaCl, 1 mM EDTA, 1 mM DTT, 5% glycerol, 10 000 c.p.m. of ³²P-labeled AT-DNA, competitors and 100 ng of His-TdIF1 in a reaction volume of 20 µL. After incubation for 30 min, the reaction mixtures were loaded onto a pre-electrophoresed 6% polyacrylamide gel. The electrophoresis was carried out at 4 °C and 120 V in 0.5x TBE (45 mM Tris–HCl, pH 8.3, 45 mM boric acid, 0.5 mM EDTA). The labeled DNA was detected by autoradiography.

Non-RI EMSA

A biotin-labeled 32 bp double-stranded DNA (dsDNA) with 3' (4-base) overhang (bio-dsDNA) was prepared by annealing 5' biotinylated oligonucleotide (bio-ssDNA; biotin-5'-CGGTATTAAAAATTACTGACAATATTAAATTAGATC-3') to unlabeled oligonucleotide (5'-TAATTTAATATTGTCAGTAATTTTTAATACCG-3'). Purified His-TdIF1 or Ku proteins were incubated with bio-dsDNA for 30 min on ice in the presence of pcDNA-HA digested with EcoRI, HaeIII, NlaIII or Sau3AI in a final volume of 10 µL in binding buffer (10 mM Tris–HCl, pH 7.5, 50 mM NaCl, 1 mM EDTA, 1 mM DTT, 5% glycerol). Purified His-TdT and/or His-TdIF1 were incubated with bio-ssDNA for 30 min on ice in binding buffer. Samples were subjected to electrophoresis at 4 °C for 60 min at 100 V using a 6% polyacrylamide gel in 0.5x TBE buffer. Samples were electrophoretically transferred to positively charged nylon membrane Hybond-N+ (GE healthcare) at 190 mA for 30 min. Transferred DNA was cross-linked to membrane for 5 min on a UV transilluminator equipped with 312 nm bulbs. After blocking, the membrane was washed 3 times with TTBS (10 mM Tris–HCl, pH 7.4, 150 mM NaCl, 0.05% Tween 20), and then incubated with Streptavidin–HRP (1 : 4000 dilution) (GE healthcare) in TTBS for 60 min at room temperature (R.T.). After washing the membrane 4 times with TTBS, bio-dsDNA was detected using ECL plus reagent (GE healthcare).

AFM

The region between 1731 and 2717 bp, which contains SV40 early promoter and a part of neomycin resistance gene, in the pcDNA3.1+ vector was amplified by PCR using primers (5'-CGAGCCGGCCTGTGGAATGTGTGTCAG-3', italics indicate NaeI site; 5'-TAAGGTACCCGGCAAGCAGGCATCGCC-3', italics indicate KpnI site). PCR products were cloned into the pGEM-T vector (Promega Corp., Madison, WI). The plasmids generated were digested with NaeI and KpnI and then the DNA fragments produced were purified and used in AFM analysis. Protein/DNA complexes were formed by incubating 120 ng of His-TdIF1 with 5 ng of DNA for 10 min at R.T. in binding buffer (20 mM Tris–HCl, pH 7.5, 10 mM NaCl, 5 mM MgCl₂) in a total volume of 50 µL. The reaction was deposited onto freshly cleaved mica at R.T. After a 10-min incubation, the mica surface was rinsed with dH₂O, and dried under a stream of nitrogen. The images were captured in air with NanoWizard (JPK Instruments AG, Berlin, Germany) microscope in intermittent contact mode. Pointprobe non-contact mode monolithic silicon probes (NanoWorld AG, Neuchatel, Switzerland) with resonant frequencies 278.5 kHz were used. Images were collected at speed of 1 Hz.

Fluorescence microscopy

HeLa cells were transiently transfected using 4 µL of Lipofectamine (Invitrogen), 3 µL of Plus reagent (Invitrogen) and 1.0 µg of plasmid DNA. Transfected cells were grown on coverslips for 36 h at 37 °C at 5% CO₂. Transfected cells were fixed in 4% paraformaldehyde (PFA) in PBS for 10 min at R.T. and washed twice with 1 mL of PBS. DNA was counterstained with 2 µg/mL Hoechst 33258 for 20 min at R.T. Cells were washed 3 times with 1 mL PBS, mounted in 50% glycerol in PBS and observed under fluorescence microscope (Axiovert 200, Carl Zeiss AG, Oberkochen, Germany).

For visualization of endogenous TdIF1, HeLa cells were fixed in 4% PFA in PBS for 10 min at R.T., permeabilized in 0.1% TritonX-100 in PBS for 10 min at R.T., and incubated with 1% BSA in PBS for 30 min at R.T. Cells were washed and reacted with 5 µg/mL rabbit polyclonal anti-TdIF1 antibody for 60 min at R.T., followed by 5 µg/mL Alexa Fluor 488 conjugated goat anti-rabbit IgG antibody (Molecular Probes, Inc., Eugene, OR) for 60 min at R.T. DNA was counterstained with 2 µg/mL Hoechst 33258 for 20 min at R.T.

The structure prediction

TdIF1 orthologues were searched using the PSI-BLAST <http://www.ncbi.nlm.nih.gov/BLAST/> and aligned using the CLUSTAL W <http://clustalw.genome.ad.jp/>. The secondary structure, 3D structure, NLS, disordered region and HTH motif of TdIF1 were predicted using the PSIPRED <http://bioinf.cs.ucl.ac.uk/psipred/>, ROBETTA server <http://robetta.bakerlab.org/>, PREDICTNLS server <http://cubic.bioc.columbia.edu/services/predictNLS/>, DISOPRED2 <http://bioinf.cs.ucl.ac.uk/disopred/index.html> and HTH prediction server <http://pbil.ibcp.fr/htm/index.php>, respectively.

Competition between TdT and DNA for TdIF1

GST or GST-TdIF1 bound to glutathione Sepharose beads was incubated with 1 µg of His-TdT in buffer B containing 150 mM NaCl and 200 µg/mL BSA for 60 min at 4 °C, washed once with buffer B containing 150 mM NaCl, incubated with 0.25 or 1 µg of pcDNA/HaeIII for 60 min at 4 °C, and washed 3 times with buffer B containing 150 mM NaCl. His-TdT and DNA fragments were eluted with buffer B containing 1 M NaCl, and then GST or GST-TdIF1 was eluted with SDS sample buffer. These eluted proteins were analyzed by SDS-PAGE and Western blotting with an anti-His antibody. Eluted DNA fragments were treated with PCI, precipitated in ethanol, and then analyzed by 5% PAGE and silver staining.

Pull-down assay using streptavidin–agarose

Two hundred nanograms of His-TdT and/or 200 ng of His-TdIF1 were incubated with 20 pmol of bio-ssDNA or bio-dsDNA coupled with 5 µL of streptavidin–agarose (Sigma) in buffer C (50 mM Tris–HCl, pH 7.4, 1 mM MgCl₂, 10% glycerol, 0.1% TritonX-100) containing 100 mM NaCl for 60 min at 4 °C. After washing with buffer C containing 100 mM NaCl, bound proteins were eluted with buffer C containing 1 M NaCl. The eluates were analyzed by SDS-PAGE and Western blotting with an anti-His antibody.

TdT activity assay

TdT activity was assayed in the reaction mixture A containing 50 mM Tris–HCl (pH 7.5), 100 mM NaCl, 0.5 mM MnCl₂, 1 mM DTT, 200 µg/mL BSA, 0.2 µCi [³H]dTTP (61.1 Ci/mmol, Moravek Biochemicals, Inc., Brea, CA), 80 µM dTTP and proteins in a reaction volume of 25 µL. After preincubation of 1 pmol of His-TdT and 0–10 pmol of BSA or His-TdIF1 for 30 min on ice, the mixture was incubated in the reaction mixture A containing 1 pmol of oligo(dT)₁₆ or dsDNA at 37 °C for 60 min. For recovery assay, after preincubation of 1 pmol of His-TdT and 10 pmol of BSA or His-TdIF1 for 30 min on ice, 3.9, 6.5, 13 or 26 ng of pcDNA-HA/EcoRI was added to the reaction mixture and incubated for 30 min on ice. Then, the mixture was incubated in the reaction mixture A containing 1 pmol of oligo(dT)₁₆ or not at 37 °C for 60 min. The radioactivity was counted with a liquid scintillation counter (PerkinElmer Inc., Waltham, MA) as described (Yamashita et al. 2001).

TdT activity was also assayed by primer extension assay. After preincubation of 1 pmol of His-TdT and 1 or 10 pmol of His-TdIF1 for 30 min on ice, the mixture was incubated in the reaction mixture B containing 50 mM Tris–HCl (pH 7.5), 100 mM NaCl, 0.5 mM MnCl₂, 1 mM DTT, 200 µg/mL BSA, 100 µM dNTPs and 40 nM bio-ssDNA or bio-dsDNA in a reaction volume of 25 µL at 37 °C for 10 min. Reactions were stopped by addition of stop solution (95% formamide, 20 mM EDTA, 0.025% xylene cyanol and 0.025% bromophenol blue), heated at 95 °C for 3 min, and analyzed by gel electrophoresis using a 20% denaturing polyacrylamide gel. After blotting on Hybond-N+ membrane, biotinylated DNAs were detected as described above.

	Acknowledgements

 
The authors thank Nobuhiro Saito, Yukiko Mizuguchi and Yasuko Ichikawa (TOYO Corporation, Tokyo, Japan) for capturing the images with AFM.

	Footnotes

 
Communicated by: Fumio Hanaoka

^* Correspondence: E-mail: koiwai{at}rs.noda.tus.ac.jp

	References

Top Abstract Introduction Results Discussion Experimental procedures References

 
Aoufouchi, S., Flatter, E., Dahan, A., Faili, A., Bertocci, B., Storck, S., Delbos, F., Cocea, L., Gupta, N., Weill, J.C. & Reynaud C.A. (2000) Two novel human and mouse DNA polymerases of the polX family. Nucleic Acids Res. 28, 3684–3693.[Abstract/Free Full Text]

Aranda, A. & Pascual, A. (2001) Nuclear hormone receptors and gene expression. Physiol. Rev. 81, 1269–1304.[Abstract/Free Full Text]

Aravind, L. & Landsman, D. (1998) AT-hook motifs identified in a wide variety of DNA-binding proteins. Nucleic Acids Res. 26, 4413–4421.[Abstract/Free Full Text]

Barthelemy, I., Carramolino, L., Gutierrez, J., Barbero, J.L., Marquez, G. & Zaballos, A. (1996) zhx-1: a novel mouse homeodomain protein containing two zinc-fingers and five homeodomains. Biochem. Biophys. Res. Commun. 224, 870–876.[CrossRef][Medline]

Boulukos, K.E., Pognonec, P., Rabault, B., Begue, A. & Ghysdael, J. (1989) Definition of an Ets1 protein domain required for nuclear localization in cells and DNA-binding activity in vitro. Mol. Cell. Biol. 9, 5718–5721.[Abstract/Free Full Text]

Burke, L., Downes, M., Carozzi, A., Giguère, V. & Muscat, G.E.O. (1996) Transcriptional repression by the orphan steroid receptor RVR/Rev-erbß is dependent on the signature motif and helix 5 in the E region: functional evidence for a biological role of RVR in myogenesis. Nucleic Acids Res. 24, 3481–3489.[Abstract/Free Full Text]

Christophe-Hobertus, C., Duquesne, V., Pichon, B., Roger, P.P. & Christophe, D. (1999) Critical residues of the homeodomain involved in contacting DNA bases also specify the nuclear accumulation of thyroid transcription factor-1. Eur. J. Biochem. 265, 491–497.[Medline]

Cockerill, P.N. & Garrard, W.T. (1986) Chromosomal loop anchorage of the immunoglobulin gene occurs next to the enhancer in a region containing topoisomerase II sites. Cell 44, 273–282.[CrossRef][Medline]

Cokol, M., Nair, R. & Rost, B. (2000) Finding nuclear localization signals. EMBO Rep. 1, 411–415.[CrossRef][Medline]

Dang, C.V. & Lee, W.M. (1988) Identification of the human c-myc protein nuclear translocation signal. Mol. Cell. Biol. 8, 4048–4054.[Abstract/Free Full Text]

Delarue, M., Boulé, J.B., Lescar, J., Expert-Bezançon, N., Jourdan, N., Sukumar, N., Rougen, F. & Papanicolaou, C. (2002) Crystal structures of a template-independent DNA polymerase: murine terminal deoxynucleotidyltransferase. EMBO J. 21, 427–439.[CrossRef][Medline]

Dingwall, C. & Laskey, R.A. (1991) Nuclear targeting sequences: a consensus? Trends Biol. Sci. 16, 178–181.

Dominguez, O., Ruiz, J.F., de Lera, T.L., Garcia-Diaz, M., Gonzalez, M.A., Kirchhoff, T., Martinez-A, C., Bernad, A. & Blanco, L. (2000) DNA polymerase µ (Pol µ), homologous to TdT, could act as a DNA mutator in eukaryotic cells. EMBO J. 19, 1731–1742.[CrossRef][Medline]

Forrester, W.C., van Genderen, C., Jenuwein, T. & Grosschedl, R. (1994) Dependence of enhancer-mediated transcription of the immunoglobulin µ gene on nuclear matrix attachment regions. Science 265, 1221–1225.[Abstract/Free Full Text]

Fujisaki, S., Sato, A., Toyomoto, T., Hayano, T., Sugai, M., Kubota, T. & Koiwai, O. (2006) Direct binding of TReP-132 with TdT results in reduction of TdT activity. Genes Cells 11, 47–57.[Abstract/Free Full Text]

Fujita, K., Shimazaki, N., Ohta, Y., Kubota, T., Ibe, S., Toji, S., Tamai, K., Fujisaki, S., Hayano, T. & Koiwai, O. (2003) Terminal deoxynucleotidyltransferase forms a ternary complex with a novel chromatin remodeling protein with 82 kDa and core histone. Genes Cells 8, 559–571.[Abstract]

Gilfillan, S., Dierich, A., Lemeur, M., Benoist, C. & Mathis, D. (1993) Mice lacking TdT: mature animals with an immature lymphocyte repertoire. Science 261, 1175–1178.[Abstract/Free Full Text]

Gizard, F., Lavallée, B., DeWitte, F. & Hum, D.W. (2001) A novel zinc finger protein TReP-132 interacts with CBP/p300 to regulate human CYP11A1 gene expression. J. Biol. Chem. 276, 33881–33892.[Abstract/Free Full Text]

Gizard, F., Lavallée, B., DeWitte, F., Teissier, E., Staels, B. & Hum, D.W. (2002) The transcriptional regulating protein of 132 kDa (TReP-132) enhances P450scc gene transcription through interaction with steroidogenic factor-1 in human adrenal cells. J. Biol. Chem. 277, 39144–39155.[Abstract/Free Full Text]

Gizard, F., Robillard, R., Barbier, O., Quatannens, B., Faucompré, A., Révillion, F., Peyrat, J.P., Staels, B. & Hum, D.W. (2005) TReP-132 controls cell proliferation by regulating the expression of the cyclin-dependent kinase inhibitors p21^WAF/Cip1 and p27^Kip1. Mol. Cell. Biol. 25, 4335–4348.[Abstract/Free Full Text]

Goebel, P., Montalbano, A., Ayers, N., Kompfner, E., Dickinson, L., Webb, C.F. & Feeney, A.J. (2002) High frequency of matrix attachment regions and cut-like protein x/CCAAT-displacement protein and B cell regulator of IgH transcription binding sites flanking Ig V region genes. J. Immunol. 169, 2477–2487.[Abstract/Free Full Text]

Hale, M.A. & Garrard, W.T. (1998) A targeted immunoglobulin gene containing a deletion of the nuclear matrix association region exhibits spontaneous hyper-recombination in pre-B cells. Mol. Immunol. 35, 609–620.[CrossRef][Medline]

Hanausek, M., Szemraji, J., Adams, A.K. & Walaszek, Z. (1996) The oncofetal protein p65: a new member of the steroid thyroid receptor superfamily. Cancer Detect. Prev. 20, 94–102.[Medline]

Holm, L. & Sander, C. (1995) DNA polymerase ß belongs to an ancient nucleotidyltransferase superfamily. Trends Biochem. Sci. 20, 345–347.[CrossRef][Medline]

Huth, J.R., Bewley, C.A., Ni