A helix-turn-helix structure unit in human centromere protein B (CENP-B)
1998; Springer Nature; Volume: 17; Issue: 3 Linguagem: Inglês
10.1093/emboj/17.3.827
ISSN1460-2075
Autores Tópico(s)Plant nutrient uptake and metabolism
ResumoArticle1 February 1998free access A helix–turn–helix structure unit in human centromere protein B (CENP-B) Junji Iwahara Junji Iwahara Department of Biophysics and Biochemistry, Graduate School of Science, University of Tokyo, Bunkyo-ku, Tokyo, 113 Japan Cellular Signaling Laboratory, Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-01 Japan Search for more papers by this author Takanori Kigawa Takanori Kigawa Cellular Signaling Laboratory, Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-01 Japan Search for more papers by this author Katsumi Kitagawa Katsumi Kitagawa Department of Molecular Biology, School of Science, Nagoya University, Chikusa-ku, Nagoya, 464-01 Japan Search for more papers by this author Hiroshi Masumoto Hiroshi Masumoto Department of Molecular Biology, School of Science, Nagoya University, Chikusa-ku, Nagoya, 464-01 Japan Search for more papers by this author Tuneko Okazaki Tuneko Okazaki Department of Molecular Biology, School of Science, Nagoya University, Chikusa-ku, Nagoya, 464-01 Japan Present address: Institute for Comprehensive Medical Science, Fujita Health University, School of Medicine, Toyoake, Aichi, 470-11 Japan Search for more papers by this author Shigeyuki Yokoyama Corresponding Author Shigeyuki Yokoyama Department of Biophysics and Biochemistry, Graduate School of Science, University of Tokyo, Bunkyo-ku, Tokyo, 113 Japan Cellular Signaling Laboratory, Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-01 Japan Search for more papers by this author Junji Iwahara Junji Iwahara Department of Biophysics and Biochemistry, Graduate School of Science, University of Tokyo, Bunkyo-ku, Tokyo, 113 Japan Cellular Signaling Laboratory, Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-01 Japan Search for more papers by this author Takanori Kigawa Takanori Kigawa Cellular Signaling Laboratory, Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-01 Japan Search for more papers by this author Katsumi Kitagawa Katsumi Kitagawa Department of Molecular Biology, School of Science, Nagoya University, Chikusa-ku, Nagoya, 464-01 Japan Search for more papers by this author Hiroshi Masumoto Hiroshi Masumoto Department of Molecular Biology, School of Science, Nagoya University, Chikusa-ku, Nagoya, 464-01 Japan Search for more papers by this author Tuneko Okazaki Tuneko Okazaki Department of Molecular Biology, School of Science, Nagoya University, Chikusa-ku, Nagoya, 464-01 Japan Present address: Institute for Comprehensive Medical Science, Fujita Health University, School of Medicine, Toyoake, Aichi, 470-11 Japan Search for more papers by this author Shigeyuki Yokoyama Corresponding Author Shigeyuki Yokoyama Department of Biophysics and Biochemistry, Graduate School of Science, University of Tokyo, Bunkyo-ku, Tokyo, 113 Japan Cellular Signaling Laboratory, Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-01 Japan Search for more papers by this author Author Information Junji Iwahara1,2, Takanori Kigawa2, Katsumi Kitagawa3, Hiroshi Masumoto3, Tuneko Okazaki3,4 and Shigeyuki Yokoyama 1,2 1Department of Biophysics and Biochemistry, Graduate School of Science, University of Tokyo, Bunkyo-ku, Tokyo, 113 Japan 2Cellular Signaling Laboratory, Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama, 351-01 Japan 3Department of Molecular Biology, School of Science, Nagoya University, Chikusa-ku, Nagoya, 464-01 Japan 4Present address: Institute for Comprehensive Medical Science, Fujita Health University, School of Medicine, Toyoake, Aichi, 470-11 Japan *Corresponding author. E-mail: [email protected] The EMBO Journal (1998)17:827-837https://doi.org/10.1093/emboj/17.3.827 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info CENP-B has been suggested to organize arrays of centromere satellite DNA into a higher order structure which then directs centromere formation and kinetochore assembly in mammalian chromosomes. The N-terminal portion of CENP-B is a 15 kDa DNA binding domain (DBD) consisting of two repeating units, RP1 and RP2. The DBD specifically binds to the CENP-B box sequence (17 bp) in centromere DNA. We determined the solution structure of human CENP-B DBD RP1 by multi-dimensional 1H, 13C and 15N NMR methods. The CENP-B DBD RP1 structure consists of four helices and has a helix–turn–helix structure. The overall folding is similar to those of some other eukaryotic DBDs, although significant sequence homology with these proteins was not found. The DBD of yeast RAP1, a telomere binding protein, is most similar to CENP-B DBD RP1. We studied the interaction between CENP-B DBD RP1 and the CENP-B box by the use of NMR chemical shift perturbation. The results suggest that CENP-B DBD RP1 interacts with one of the essential regions of the CENP-B box DNA, mainly at the N-terminal basic region, the N-terminal portion of helix 2 and helix 3. Introduction The centromere plays a major role in segregation of eukaryotic chromosomes in mitosis and meiosis, by serving as the site for kinetochore assembly and sister chromatid attachment (Clarke, 1990; Willard, 1990). In addition, the centromere is considered to regulate the cell cycle checkpoint for the metaphase–anaphase transition (Nicklas et al., 1995; Rieder et al., 1995). To understand these functions of the centromere at the molecular level, the DNAs and the proteins in centromeres have been characterized for a variety of eukaryotes (Pluta et al., 1995). In human cells several proteins localized in the centromere have been identified with centromere-specific autoantibodies (Earnshaw and Rothfield, 1985). CENP-A (17 kDa), CENP-B (80 kDa) and CENP-C (140 kDa) are such proteins with DNA binding activities (Masumoto et al., 1989; Palmer et al., 1991; Sugimoto et al., 1994; Sullivan et al., 1994). Thus far a sequence-specific DNA binding activity has been found only for CENP-B (Muro et al., 1992). In the CENP-B amino acid sequence deduced from the cloned gene (Sullivan and Glass, 1991) there is a DNA binding domain (DBD) within the N-terminal 125 residues (Yoda et al., 1992). This domain of CENP-B binds with high affinity to a 17 bp sequence, the CENP-B box, as shown in Figure 1A (Masumoto et al., 1989; Muro et al., 1992). Three regions, consisting of 4, 1 and 4 bp (Figure 1A), form the essential core recognition sequence for CENP-B to bind to the CENP-B box sequence (Masumoto et al., 1993; Yoda et al., 1996). The CENP-B box sequence frequently exists in human α-satellite (alphoid) DNA, the human centromere-specific repeating DNA family composed of 171 bp monomer units with chromosome-specific sequence variations (Willard, 1990; Yoda and Okazaki, 1997). The distribution of CENP-B boxes was studied in human chromosome 21 and it was found that CENP-B boxes are regularly spaced over a 1.3 Mbp region in one of the two adjacent α-satellite DNA arrays (Ikeno et al., 1994). CENP-B has a dimerization domain in the C-terminal 59 residues, which is separate from the DBD (Kitagawa et al., 1995), and the CENP-B dimer forms either a complex containing two DNA molecules with a CENP-B box (Muro et al., 1992) or a loop structure on a DNA strand containing two CENP-B boxes (K.Yoda, A.Okuda, A.Kikuchi and T.Okazaki, submitted for publication). These results suggested that the function of CENP-B in vivo may be to organize the long centromeric satellite arrays with frequent CENP-B boxes into a higher order chromatin structure by juxtaposing pairs of CENP-B box sequences and that this structure may then become the foundation for the centromere/kinetochore structure and activity (Muro et al., 1992; Ikeno et al., 1994; Yoda et al., 1996). Figure 1.(A) Human CENP-B box sequence. The base pairs recognized by CENP-B, elucidated by in vitro binding experiments using CENP-B and variants of the CENP-B box (Masumoto et al., 1993) are boxed. U and Y represent purine (A or G) and pyrimidine (C or T) respectively. (B) Base sequence alignment of CENP-B binding sites among human, Mus musculus, Mus caroli and African green monkey (AGM). The CENP-B binding sites of Mus musculus and Mus caroli are present in the centromere-specific repeating sequences, which are composed of monomer units of 120 and 79 bp respectively (Masumoto et al., 1989; Kipling et al., 1995). The CENP-B binding sites of African green monkey were found in the α-satellite DNA (Yoda et al., 1996), but the number of CENP-B binding sites in the α-satellite DNA is much smaller in African green monkey than in humans (Goldberg et al., 1996). (C) Amino acid sequence alignment between CENP-B DBD RP1 and RP2. Identical residues are boxed and similar residues are bridged with bars. Download figure Download PowerPoint CENP-B genes have also been cloned from mouse, hamster and African green monkey (Sullivan and Glass, 1991; Bejarano and Valdivia, 1996; Yoda et al., 1996). The amino acid sequences of human CENP-B and the mammalian homologs are highly conserved, particularly in the DBD (100% identity) (Figure 1C) and the dimerization domain (98% identity). The 9 base pairs (sites 1–3 in Figure 1A) recognized by human CENP-B are conserved in the other mammalian CENP-B binding sites studied thus far (Figure 1B) (Muro et al., 1992; Kipling et al., 1995; Yoda et al., 1996), although the frequency of occurence of CENP-B binding sites in centromeres is quite different among species (Goldberg et al., 1996; Romanova et al., 1996). Therefore, the mechanisms of DNA binding and dimerization of CENP-B are probably conserved among mammals. Furthermore, two CENP-B homologs (Abp1/Cbp1 and Cbh) have been identified in the fission yeast Schizosaccharomyces pombe (Murakami et al., 1996; Lee et al., 1997) and their binding sites have been found in the K repeat and/or central core sequence in the essential centromere DNA (Halverson et al., 1997; Lee et al., 1997; Ngan and Clarke, 1997). The CENP-B DBD was proposed to have a helix–loop–helix structure consisting of ∼60 residues at its N-terminal region, on the basis of the low sequence similarity with helix–loop–helix proteins such as Myc and MyoD (Sullivan and Glass, 1991). On the other hand, the helix–loop–helix DNA binding motif mediates dimerization (Ferre et al., 1993; Ma et al., 1994), while the CENP-B DBD binds to the CENP-B box sequence as a monomer (Yoda et al., 1992). Recently it was suggested, based on the amino acid sequence, that the CENP-B DBD consists of two repeating units (RP1 and RP2), as shown in Figure 1C (Suzuki and Brenner, 1995; Suzuki et al., 1995). CENP-B DBD RP1 mostly corresponds to the region that was suggested to have sequence similarity with the helix–loop–helix proteins. To our knowledge there have been no reports about the tertiary structure of any protein involved in centromere function. In order to gain insight into the molecular mechanism of the interaction between CENP-B and CENP–B box DNA at atomic resolution we have begun to study the CENP-B DBD. In this paper we report the tertiary structure of CENP-B DBD RP1, as determined by heteronuclear multidimensional NMR. Results and discussion The first repeat (RP1) of the CENP-B DBD was expressed under control of the T7 promoter in Escherichia coli as a polypeptide with 14 N-terminal vector-derived residues, including a 6× His tag (MRGSHHHHHHGMAS), followed by the CENP-B DBD RP1 moiety (amino acid residues 1–56; Figure 1C). About 6 mg/l culture purified CENP-B DBD RP1 was obtained using M9 minimal medium. The circular dichroism (CD) data showed that CENP-B DBD RP1 is predominantly α-helical; the α–helix content at 20°C was estimated to be 54%. In addition, CENP-B DBD RP1 exhibited cooperative thermal denaturation, centered at 58°C, as monitored by CD at 222 nm under conditions of 20 mM potassium phosphate, 400 mM Na2SO4, 0.25 mM EDTA and 0.01% NaN3 at pH 6.0 (data not shown). These results indicate that CENP-B DBD RP1 itself is a well-structured unit. Resonance assignments NMR measurements for resonance assignments and structure determination of CENP-B DBD RP1 were performed under conditions of 20 mM potassium phosphate, 400 mM Na2SO4, 0.25 mM EDTA and 0.01% NaN3 at pH 6.0. Figure 2A shows the 2D 1H-15N HSQC spectrum of CENP-B DBD RP1. First, ∼80% of the 1H and 15N resonances of CENP-B DBD RP1 were assigned by use of 3D 1H-15N NOESY-HSQC, 3D 1H-15N TOCSY-HSQC, 3D HNHB, 2D 1H-15N HSQC, 2D NOESY, 2D TOCSY and 2D DQF-COSY spectra. We completed the resonance assignments using the 3D CT-HNCA, 3D CT-HN(CO)CA, 3D HCCH-TOCSY and 3D 1H-13C NOESY-HSQC spectra. Figure 2B shows an example of sequential assignment along the Leu42–Leu53 backbone using the 3D CT-HNCA spectrum. Figure 2.(A) The 1H-15N HSQC spectrum of CENP-B DBD RP1. The negatively numbered residues are from the N-terminal vector-derived region. The cross-peaks due to sidechain resonances are indicated by sc. The spectral width in the 15N dimension was 775 Hz and aliased cross-peaks are displayed without sign discrimination. (B) Sequential resonance assignment for Leu42–Leu53 in the 3D CT-HNCA spectrum of CENP-B DBD RP1. The connectivities of the inter-residue cross-peaks are indicated with arrows. The sample used for the NMR measurements was CENP-B DBD RP1 (1.5 mM) in 20 mM potassium phosphate buffer, pH 6.0, containing 400 mM Na2SO4, 0.25 mM EDTA and 0.01% NaN3. Download figure Download PowerPoint Structure determination The 14 vector-derived residues exhibited only a few inter-residue and negative {1H}-15N NOEs, indicating that the region was unstructured. Therefore, structure determination was performed only for the 56 CENP-B-derived residues. The secondary structure elements of CENP-B DBD RP1 are shown in Figure 3, together with a summary of the sequential and medium range NOEs, the 3JHNHα coupling constants, observation of HN–H2O cross-peaks, 13Cα chemical shift indices and {1H}-15N NOE values. For calculation of tertiary structure inter-proton NOE information was obtained from the 3D 1H-15N NOESY-HSQC, 3D 1H-13C NOESY-HSQC and 2D homonuclear NOESY spectra, with 662 restraints (466 inter-residue distance restraints, 160 intra-residue distance restraints and 36 dihedral angle restraints) being used. Superimposition of the final family of 20 refined structures is shown in Figure 4A. The backbone root mean square (r.m.s.) deviation of the family was 0.48 ± 0.17 Å. Other statistics of these structures are listed in Table I. Figure 4B shows overall folding of the energy minimized average structure of CENP-B DBD RP1. Figure 3.The secondary structure elements of CENP-B DBD RP1 and a summary of the sequential and medium range NOEs, 3JHNHα coupling constants, presence of HN-H2O cross-peaks, 13Cα chemical shift indices and {1H}-15N NOE values. Δ, a 3JHNHα coupling constant >7 Hz. Residues that exhibited a cross-peak between HN and H2O resonances in the 3D 1H-15N NOESY-HSQC spectrum are indicated by ♦. Dotted lines indicate the NOE conectivities in which one resonance overlapped with another. The * in the bar graph of the {1H}-15N NOE values indicates a proline residue. Download figure Download PowerPoint Figure 4.Solution structure of CENP-B DBD RP1. (A) Superimposition of the backbones of 20 NMR-derived structures of CENP-B DBD RP1 (stereoview). (B) Ribbon representation of the energy-minimized average structure derived from the 20 refined CENP-B DBD RP1 structures, as viewed from the same angle as in (A). Helix 1, blue; helix 2, green; helix 3, light green; helix 4, orange. (C) Hydrophobic contacts between helices. The N-terminal nine residues are omitted for simplicity. The colors of the helices are the same as in (B). The molecular graphics package MOLSCRIPT (Kraulis, 1991) was used. Download figure Download PowerPoint Table 1. Structural statistics for the 20 refined structures of CENP-B DBD RP1 X-PLOR energies (kcal/mol) Evan der Waalsa 26.9 ± 5.1 ELennard–Jonesb −118.7 ± 29.4 r.m.s. deviation from experimental restraints Distance (Å) 0.037 ± 0.003 Dihedral angle (degree) 0.29 ± 0.25 Deviations from ideal covalent geometry used within X-PLOR Bonds (Å) 0.0034 ± 0.0002 Angles (degree) 0.61 ± 0.02 Impropers (degree) 0.45 ± 0.02 a The quadratic van der Waals term was calculated with a force constant of 4 kcal/mol/Å and the van der Waals radii were set to 0.75 times the standard value used in CHARMM. b Lennard–Jones potential calculated using the CHARMM empirical energy function. Solution structure of CENP-B DBD RP1 Four α-helices (helix 1, Phe10–Asn23; helix 2, Lys28–Phe35; helix 3, Ser40–Asn48; helix 4, Lys49–Ala54) exist in the solution structure of CENP-B DBD RP1. A canonical helix–turn–helix (HTH) structure is observed in the region from helix 2 to helix 3. This result is roughly consistent with the prediction by Suzuki et al. (1995). CENP-B DBD RP1 does not exhibit any structural similarity to the helix–loop–helix domain of MyoD complexed with DNA (Ma et al., 1994), in spite of the partial sequence similarity (Sullivan and Glass, 1991). The HTH region (from helix 2 to helix 3) and helix 1 of CENP-B DBD RP1 form a hydrophobic core with the sidechains of Ile16, Val20, Ile31, Phe35 and Leu42 (Figure 4C). The Thr9-Phe10-Arg11-Glu12 sequence at the N–terminus of helix 1 forms a helix-stabilizing T-X-X-E type 'capping box' (Harper and Rose, 1993; Gronenborn and Clore, 1994) with a hydrogen bond between the amide group of Thr9 and the γ-carboxyl group of Glu12, which is reflected in the characteristic pattern of 13Cα chemical shifts (Figure 3) and NOEs. The sidechain of Leu8 forms a hydrophobic interaction with those of Lys13, Ile37 and Thr41. Consequently the region from Arg5 to Thr9 is structured near the turn between helix 2 and helix 3 (Figure 4B) and therefore corresponds to the 'N-terminal arm' in other eukaryotic DNA binding proteins with a HTH motif (Qian et al., 1989; Hirsch and Aggarwal, 1995; König et al., 1996). The tertiary structure from the N-terminus to helix 3 of CENP-B DBD RP1 is quite similar to those of other eukaryotic HTH domains, such as homeodomains, the Myb DBD repeating units (Ogata et al., 1992, 1994) and the RAP1 DBDs (König et al., 1996). Figure 5 shows superimposition of the structures of these DBDs on that of CENP-B DBD RP1. The HTH region (helix 2, the turn and helix 3) of CENP-B DBD RP1 is best superimposed on that of the fushi tarazu homeodomain (Qian et al., 1994); the r.m.s. deviation value of the Cα atoms is as small as 1.4 Å (Figure 5A). However, the orientation of helix 1 relative to the HTH region of CENP-B DBD RP1 is different from that of the fushi tarazu homeodomain (Figure 5A). Setting aside the turn regions, the relative orientations between helices 1, 2 and 3 of CENP-B DBD RP1 are more similar to those of RAP1 and the Myb DBD repeating units than to that of the fushi tarazu homeodomain (Figure 5B and C). In particular, the structure of RAP1 DBD domain 1 superimposes well upon that of CENP-B DBD RP1, with an r.m.s. deviation value of 2.1 Å for the Cα atoms of the three helices. Figure 5.Comparison of the structure of CENP-B DBD RP1 with those of the fushi tarazu homeodomain, Myb DBD R2 and RAP1 DBD domain 1. (A) Superimposition of CENP-B DBD RP1 (yellow) on the fushi tarazu homeodomain (magenta). The Cα atoms of the HTH region (from helix 2 to helix 3) were fitted between the two molecules. (B) Superimposition of CENP-B DBD RP1 (yellow) on Myb DBD R2 (blue). The Cα atoms of helices 1–3 were fitted. (C) Superimposition of CENP-B DBD RP1 (yellow) on RAP1 DBD domain 1 (green). The Cα atoms in helices 1–3 were fitted. These pictures were generated with the MidasPlus system (Ferrin et al., 1988; Huang et al., 1991). Download figure Download PowerPoint Figure 6.(A) Mapping of CENP-B DBD RP1 residues that exhibited chemical shift perturbations in the HSQC spectra upon addition of a 1.5 molar equivalent of CB21 DNA. The results are shown on ribbon representations of the CENP-B DBD RP1 solution structure. The colors used are blue for residues with 0.25 p.p.m. < |Δδ15N| < 0.50 p.p.m. or 0.05 p.p.m. < |Δδ1HN| < 0.10 p.p.m. and magenta for residues with 0.5 p.p.m. < |Δδ15N| or 0.1 p.p.m. < |Δδ1HN|, where Δδ15N and Δδ1HN represent the 15N and 1H chemical shift differences between the absence and presence of CB21 DNA. Residues with HSQC cross-peaks that became too broad to observe upon addition of CB21 DNA are colored red. This drawing was generated with the MidasPlus system (Ferrin et al., 1988; Huang et al., 1991). (B) Surface electrostatic potential of CENP-B DBD RP1, with the peptide backbone viewed inside. Positively and negatively charged areas are colored blue and red respectively. Orientation of the molecule is as in (A). This figure was produced by the program GRASP (Nicholls et al., 1991; Nicholls, 1993). Download figure Download PowerPoint Helix 4 of CENP-B DBD RP1 Helix 4 is characteristic of CENP-B DBD RP1. Helices 3 and 4 are directly connected in the amino acid sequence and may be regarded as one long helix kinked at the N–Cα bond of Asn48 (with a dihedral angle φ = −97° for the minimized average structure). Long range 1H–1H NOEs between helices 1 and 4 were observed for several residues; in particular the aromatic ring protons of Phe10 (helix 1) gave 13 NOEs to the backbone and sidechain protons of Ile52 and Leu53 (helix 4). Note that helix 1 also makes a hydrophobic contact with helix 3, as described above. Therefore, hydrophobic interaction of helix 1 with helices 3 and 4 defines their relative locations, with an angle of 129° between the helix axes (Figure 4C). The anomalous structure at Asn48, at the junction of helices 3 and 4, was also supported by two independent sources of experimental data that were not included as constraints in the structure determination. First, in the region from Ser40 to Ala54 only Asn48 exhibited a 3JHNHα value >8 Hz, whereas other residues exhibited 3JHNHα values 0.05 p.p.m. (Arg5, Leu8, Asp25, Lys28, Ser43, Ile45, Leu46, Lys47 and Leu53). The 12 DNA-perturbed residues are mapped on the tertiary structure of CENP-B DBD RP1 in Figure 6A. It is likely that these residues of CENP-B DBD RP1 are located either in the DNA binding interface or in regions of DNA-induced conformation change. The most probable example of the latter case is Ile45, which is nearly completely buried inside the molecule (Figure 4C), while other DNA-perturbed residues are much more exposed. All of these DNA-perturbed residues, except for Leu53 in helix 4, are localized in three regions, the N-terminal arm and the N-terminal portions of helix 2 and helix 3. The same three regions of the homeodomains, the Myb repeats and the RAP1 domains, whose HTH folds superimpose well onto that of CENP-B DBD RP1, are actually in contact with DNA in the complex structures (Kissinger et al., 1990; Billeter et al., 1993; Ogata et al., 1994; König et al., 1996). In each case helix 3 contacts the major groove of the double-stranded DNA and the positively charged amino acid residues interact with the phosphate groups. In fact, in the electrostatic potential profile of CENP-B DBD RP1 (Figure 6B) several positively charged areas surround helix 3. Therefore, it is likely that helix 3 contacts the major groove of the DNA, like other HTH DBDs. Figure 7.Chemical shift changes in the base pair imino protons of CB21 DNA upon addition of CENP-B DBD RP1. Each value is the absolute difference between in the absence and presence of a 1.33 molar equivalent of CENP-B DBD RP1. The imino proton resonances that were too broad to observe are indicated by asterisks. The NMR samples were measured in 20 mM potassium phosphate buffer, pH 6.0, containing 250 mM Na2SO4, 0.25 mM EDTA and 0.01% NaN3. Download figure Download PowerPoint Base pairs recognized by CENP-B DBD RP1 The chemical shift changes of the base paired imino proton resonances of the CB21 DNA (Figure 7) were analyzed upon addition of CENP-B DBD RP1. Eighteen imino proton resonances were assigned sequentially (Wüthrich, 1986) by 2D NOESY, whereas imino proton resonances of base pairs 1, 20 and 21 (G:C, T:A and T:A respectively) were not observed. Chemical shifts of the assigned imino proton resonances of the CB21 DNA in the presence and absence of CENP-B DBD RP1 were compared (Figure 7). Proton chemical shift changes >0.05 p.p.m. were observed for base pairs in the TCGTT sequence. This region overlaps with 'site 1' (TTCG), one of the essential sites in the CENP-B box for CENP-B binding (Figure 7). In contrast, such large chemical shift changes were not observed for sites 2 and 3 (Figure 7). Therefore, CENP-B DBD RP1 is concluded to interact with the 4 bp of site 1 in the CENP-B box sequence. Note that other HTH structures similar to CENP-B DBD RP1 can recognize up to 5 bp (Kissinger et al., 1990; Billeter et al., 1993; Ogata et al., 1994; König et al., 1996). Consequently, sites 2 and 3 are considered to be recognized by the rest of the CENP-B DBD, including RP2. A model of association between CENP-B and the CENP-B box In the present study we found that CENP-B DBD RP1 has an HTH structure, which is a typical structure found in DNA binding proteins. The distribution of positively charged areas surrounding helix 3 and the chemical shift perturbations of the amide 1H and 15N resonances, as mentioned above, suggest that helix 3 of CENP-B DBD RP1 contacts the major groove side of B-form DNA, like other HTH DBDs, such as homeodomains, Myb and RAP1. As judged from the imino proton chemical shift perturbations, CENP-B DBD RP1 is considered to recognize site 1 (TTCG) of the CENP-B box sequence. The other repeat, RP2, of the CENP-B DBD is quite homologous to RP1 (Figure 1C) and may therefore include an HTH structure similar to that of RP1. Thus CENP-B DBD RP2 is likely to bind with site 2 (A) and site 3 (CGGG) of the CENP-B box (Figure 1A), possibly with the N-terminal 'linker' region and with the following HTH structure respectively (Suzuki et al., 1995). On these assumptions we tried to build a preliminary docking model of the complex of the entire CENP-B DBD (RP1–RP2) and CENP-B box DNA (sites 1–3) in the B-form. Two opposite orientations of helix 3 relative to site 1 were possible; the vector of helix 3 (from the N- to the C-terminus) lies either in the same direction as that of the major goove vector from sites 1 to 3 (Figure
Referência(s)