Two-dimensional 1H nuclear magnetic resonance studies on the gene V-encoded single-stranded DNA-binding protein of the filamentous bacteriophage IKe. I. Structure elucidation of the DNA-binding wing

Two-dimensional nuclear magnetic resonance techniques were used to obtain residue- and sequence-specific assignments in the 1H spectrum of the single-stranded DNA-binding protein encoded by gene V of the filamentous phage IKe (IKe GVP). The residue-specific assignments are based on the analysis of J-correlated spectra, i.e. correlated spectroscopy and homonuclear-Hartmann-Hahn total correlated spectroscopy. Complete assignments of side-chain spin systems, e.g. long side-chains, were, to a major part, derived from two-dimensional spectra obtained by means of the latter technique. Sequence-specific residue assignments were obtained for the two neighbouring residues V41 and Y42, and the amino acid sequence segment encompassing residues S17 through I29. The structure of this segment, a beta-loop, was deduced from the interresidue nuclear Overhauser effect pattern. Residues S17 through V19 and P26 through I29 form an anti-parallel beta-ladder segment, whereas residues Q21 to K25 constitute the loop region. The beta-loop is expected to project into the solution and is intimately involved in binding to single-stranded DNA; it is therefore designated the "DNA-binding wing". By analogy with the structure of the DNA-binding wing deduced from IKe GVP, a similar structure is proposed for the corresponding domain of the gene V protein encoded by the filamentous phage Ff for which, from X-ray diffraction studies, a three-dimensional structure has been deduced. Essential differences appear to exist between the DNA-binding domain in the X-ray structure and that proposed in this paper. Possible reasons for these differences are discussed.     

Identification of lysine residues at the binding site of bacteriophage-Pf1 DNA-binding protein.

The accessibility of NH2 groups in the DNA-binding protein of Pf1 bacteriophage has been investigated by differential chemical modification with the reagent ethyl acetimidate. The DNA-binding surface was mapped by identification of NH2 groups protected from modification when the protein is bound to bacteriophage-Pf1 DNA in the native nucleoprotein complex and when bound to the synthetic oligonucleotide d(GCGTTGCG). The ability of the modified protein to bind to DNA was monitored by fluorescence spectroscopy. Modification of the NH2 groups in the native nucleoprotein complex showed that seven out of the eight lysine residues present, and the N-terminus, were accessible to the reagent, and were not protected by DNA or by adjacent protein subunits. Modification of these residues did not inhibit the ability of the protein to bind DNA. Lysine-25 was identified by peptide mapping as being the major protected residue. Modification of this residue does abolish DNA-binding activity. Chemical modification of the accessible NH2 groups in the complex formed with the octanucleotide effectively abolishes binding to DNA. Peptide mapping established that, in this case, lysine-17 was the major protected residue. The differences observed in protection from acetimidation, and in the ability of the modified protein to bind DNA, indicate that the oligonucleotide mode of binding is not identical with that found in the native nucleoprotein complex with bacteriophage-Pf1 DNA.     

A consensus DNA-binding site for the androgen receptor.

We have used a DNA-binding site selection assay to determine a consensus binding sequence for the androgen receptor (AR). A purified fusion protein containing the AR DNA-binding domain was incubated with a pool of random sequence oligonucleotides, and complexes were isolated by gel mobility shift assays. Individually selected sites were characterised by nucleotide sequencing and compiled to give a consensus AR-binding element. This sequence is comprised of two 6-basepair (bp) asymmetrical elements separated by a 3-bp spacer, 5'-GGA/TACANNNTGTTCT-3', similar to that described for the glucocorticoid response element. Inspection of the consensus revealed a slight preference for G or A nucleotides at the +1 position in the spacer and for A and T nucleotides in the 3'-flanking region. Therefore, a series of oligonucleotides was designed in which the spacer and flanking nucleotides were changed to the least preferred sequence. Competition experiments with these oligonucleotides and the AR fusion protein indicated that an oligonucleotide with both the spacer and flanking sequences changed had greater than 3-fold less affinity than the consensus sequence. The functional activity of these oligonucleotides was also assessed by placing them up-stream of a reporter gene in a transient transfection assay and correlated with the affinity with which the AR fusion protein bound to DNA. Therefore, sequences surrounding the two 6-bp half-sites influence both the binding affinity for the receptor and the functional activity of the response element.     

Fluorescence and chemical studies on the interaction of Escherichia coli DNA-binding protein with single-stranded DNA.

Nanosecond and steady-state fluorescence spectoscopy were used to probe the environment of the tryptophan residues of Escherichia coli DNA-binding protein. A spectral shift and a change in quantum yield of the protein upon binding to DNA or oligonucleotides indicate that the tryptophan residues are near or at the DNA binding site. The observation of two excited-state lifetimes of the protein indicates that there is heterogeneity in the microenvironments of these tryptophan residues. The "short-lifetime" tryptophan residues are more sensitive to the interaction with DNA than the "long-lifetime" residues. The results of solute-perturbation studies with iodide or acrylamide indicate that there are tryptophan residues near the surface of the protein which are heterogeneous in their accessibility to these quenchers and that they become less accessible after DNA binding. Also, lysine residues of the protein have been shown to be essential to DNA binding by chemical-modification studies. Tyrosine, arginine, and cysteine residues appear not to be involved in this binding process. From studies of the decay of fluorescence anisotropy of the binding protein in the presence and absence of DNA, it has been concluded that (a) the tetrameric binding protein does not dissociate into subuniits upon binding to the oligonucleotide d(pT)16 and (b) the binding protein-fd DNA complex possesses "local flexibility" and, therefore, cannot be described as a continuous, rigid rod.     

Determination of the secondary structure in solution of the Escherichia coli DnaA DNA-binding domain

DnaA protein binds specifically to a group of binding sites collectively called as DnaA boxes within the bacterial replication origin to induce local unwinding of duplex DNA. The DNA-binding domain of DnaA, domain IV, comprises the C-terminal 94 amino acid residues of the protein. We overproduced and purified a protein containing only this domain plus a methionine residue. This protein was stable as a monomer and maintained DnaA box-specific binding activity. We then analyzed its solution structure by CD spectrum and heteronuclear multi-dimensional NMR experiments. We established extensive assignments of the 1H, 13C, and 15N nuclei, and revealed by obtaining combined analyses of chemical shift index and NOE connectivities that DnaA domain IV contains six alpha-helices and no beta-sheets, consistent with results of CD analysis. Mutations known to reduce DnaA box-binding activity were specifically located in or near two of the alpha-helices. These findings indicate that the DNA-binding fold of DnaA domain IV is unique among origin-binding proteins.     

The structure of the human retinoic acid receptor-beta DNA-binding domain determined by NMR.

The solution structure of the DNA-binding domain (DBD) of the human retinoic acid receptor-beta (hRAR-beta) has been determined by nuclear magnetic resonance (NMR) spectroscopy and distance geometry (DG). The assignments of 1H and 15N resonances were carried out by the use of 1H homonuclear and 15N-1H heteronuclear two- and three-dimensional NMR experiments. The structure of RAR DBD has been obtained on the basis of distance constrains derived from NMR experiments. The structure shows that two "zinc-finger" domains of the protein are followed by two perpendicular alpha-helices and a short beta-sheet near the N-terminus. Apolar residues in both helices form a hydrophobic core. Binding models of RAR DBD to its inverted and direct repeat response elements have been constructed based on this three-dimensional structure.     

Characterization of the DNA binding protein encoded by the N-specific filamentous Escherichia coli phage IKe. Binding properties of the protein and nucleotide sequence of the gene

A DNA binding protein encoded by the filamentous single-stranded DNA phage IKe has been isolated from IKe-infected Escherichia coli cells. Fluorescence and in vitro binding studies have shown that the protein binds co-operatively and with a high specificity to single-stranded but not to double-stranded DNA. From titration of the protein to poly(dA) it has been calculated that approximately four bases of the DNA are covered by one monomer of protein. These binding characteristics closely resemble those of gene V protein encoded by the F-specific filamentous phages M13 and fd. The nucleotide sequence of the gene specifying the IKe DNA binding protein has been established. When compared to the nucleotide sequence of gene V of phage M13 it shows an homology of 58%, indicating that these two phages are evolutionarily related. The IKe DNA binding protein is 88 amino acids long which is one amino acid residue larger than the gene V protein sequence. When the IKe DNA binding protein sequence is compared with that of gene V protein it was found that 39 amino acid residues have identical positions in both proteins. The positions of all five tyrosine residues, a number of which are known to be involved in DNA binding, are conserved. Secondary structure predictions indicate that the two proteins contain similar structural domains. It is proposed that the tyrosine residues which are involved in DNA binding are the ones in or next to a beta-turn, at positions 26, 41 and 56 in gene V protein and at positions 27, 42 and 57 in the IKe DNA binding protein.     

Effect of site-specific DNA methylation and mutagenesis on recognition by methylated DNA-binding protein from human placenta.

Methylated DNA-binding protein (MDBP) from human placenta is the first protein shown to bind specifically to certain DNA sequences only when they are methylated at cytosine residues. Among the sites recognized by MDBP is pB site 1, a pBR322-derived sequence which has a high affinity for MDBP when methylated at all CpG positions. We have substituted pB site 1 with 5-methyl-cytosine (m5C) residues at one to three of its CpG dinucleotides on one strand by the use of m5C-containing oligonucleotides. MDBP binds best when all three CpG dinucleotides in the region 5'-ATCGTCACGGCGAT-3' are methylated. Even more binding is obtained when both strands are methylated. Alteration of various residues in this binding site by oligonucleotide-directed mutagenesis decreased the binding. However, two mutations which increased the dyad symmetry of part of the binding site yielded ligands with a higher affinity for MDBP.     

DNA-binding specificity of the S8 homeodomain.

The murine S8 homeobox gene is expressed in a mesenchyme-specific pattern in embryos, as well as in mesodermal cell lines. The S8 homeodomain is overall similar to paired type homeodomains, but at position 50, which is crucial for specific DNA recognition, it contains a Gln, as is found in Antennapedia (Antp)-type homeodomains. We determined the DNA-binding specificity of the purified S8 homeodomain by in vitro selection of random oligonucleotides. The resulting 11-bp consensus binding site, ANC/TC/TAATTAA/GC resembles, but subtly differs from, the recognition sequences of Antp-type homeodomains. Equilibrium binding constants of down to 6.0 x 10(-10) M were found for binding of the S8 homeodomain to selected oligonucleotides. Using specific antibodies and an oligonucleotide containing an S8-site, we detected by band-shift two abundant DNA binding activities in mesodermal cell lines that correspond to S8 and two more that correspond to its close relative MHox. These S8 protein forms are differentially expressed in retinoic acid-treated P19 EC cells.     

A method for fabricating uni-dsDNA microarray chip for analyzing DNA-binding proteins

This paper describes an approach for preparing unimolecular double-stranded DNA (uni-dsDNA) microarray chip. In this method, the various target oligonucleotides containing a reverse complementary sequence at 5' end were firstly annealed to a same universal oligonucleotide with amino group at 5' end and immobilized on aldehyde-derivatized glass slide. An on-chip DNA polymerization reaction was then performed to elongate the universal oligonucleotides. After a denaturation and a followed intra-strand annealing, a hairpin structure was formed at the free 3' end of the immobilized oligonucleotides. Finally, another on-chip DNA polymerization was done to synthesize the uni-dsDNA microarray. Combining with a PCR amplification of chemically synthesized target oligonucleotides, this method was much cost-effective for production of the uni-dsDNA microarray. The uni-dsDNA microarray was verified applicable for detecting the presence and monitoring the DNA-binding activity of the sequence-specific DNA-binding proteins.     

A two-dimensional NMR study of the solution structure of a DNA dodecamer comprising the concensus sequence for the specific DNA-binding sites of the glucocorticoid receptor protein

A two-dimensional 500-MHz 1H-NMR study on the non-self-complementary double-stranded DNA dodecamer 5'd(C-C-A-G-A-A-C-A-G-T-G-G)5'd(C-C-A-C-T-G-T-T-C-T-G-G), is presented. This oligonucleotide contains the consensus octanucleotide sequence 5'd(A-G-A-A-C-A-G-T) for the specific DNA-binding sites of the glucocorticoid receptor protein [Payvar, F. et al. (1984) Cell 35, 381-392]. Using a combination of two-dimensional pure phase absorption nuclear Overhauser enhancement (NOESY) and homonuclear J-correlated (COSY) spectroscopy all non-exchangeable base (with the exception fo the adenine H2 protons), methyl and deoxyribose H1', H2', H2", H3' and H4' resonances are assigned unambiguously using a sequential resonance assignment strategy. From the relative intensities of the cross peaks in the pure phase absorption NOESY spectra at two mixing times it is shown that the dodecamer adopts a B-type conformation in solution.     

Sequential resonance assignments in 1H NMR spectra of oligonucleotides by two-dimensional NMR spectroscopy

A sequential assignment procedure is outlined, based on two-dimensional NOE ( NOESY ) and two-dimensional J-correlated spectroscopy ( COSY ), for assigning the nonexchangeable proton resonances in NMR spectra of oligonucleotides. As presented here the method is generally applicable to right-handed helical oligonucleotides of intermediate size. We applied it to a lac operator DNA fragment consisting of d( TGAGCGG ) and d( CCGCTCA ) and obtained complete assignments for the adenine H8, guanine H8, cytosine H6 and H5, thymine H6 and 5-methyl, and the deoxyribose H1', H2', H2", H3', and H4' resonances, as well as some H5', H5" (pairwise) assignments. These assignments are required for the analysis of two-dimensional NOE and J-coupling data in terms of the solution structure of oligonucleotides.     

Moloney murine leukemia virus integration protein produced in yeast binds specifically to viral att sites.

The integration protein (IN) of Moloney murine leukemia virus (MuLV), purified after being produced in yeast cells, has been analyzed for its ability to bind its putative viral substrates, the att sites. An electrophoretic mobility shift assay revealed that the Moloney MuLV IN protein binds synthetic oligonucleotides containing att sequences, with specificity towards its cognate (MuLV) sequences. The terminal 13 base pairs, which are identical at both ends of viral DNA, are sufficient for binding if present at the ends of oligonucleotide duplexes in the same orientation as in linear viral DNA. However, only weak binding was observed when the same sequences were positioned within a substrate in a manner simulating att junctions in circular viral DNA with two long terminal repeats. Binding to att sites in oligonucleotides simulating linear viral DNA was dependent on the presence of the highly conserved CA residues preceding the site for 3' processing (an IN-dependent reaction that removes two nucleotides from the 3' ends of linear viral DNA); mutation of CA to TG abolished binding, and a CA to TA change reduced affinity by at least 20-fold. Removal of either the terminal two base pairs from both ends of the oligonucleotide duplex or the terminal two nucleotides from the 3' ends of each strand did not affect binding. The removal of three 3' terminal nucleotides, however, abolished binding, suggesting an essential role for the A residue immediately upstream of the 3' processing site in the binding reaction. These results help define the sequence requirements for att site recognition by IN, explain the conservation of the subterminal CA dinucleotide, and provide a simple assay for sequence-specific IN activity.     

DNA–XPA interactions: a 31P NMR and molecular modeling study of dCCAATAACC association with the minimal DNA-binding domain (M98–F219) of the nucleotide excision repair protein XPA

Recent NMR-based, chemical shift mapping experiments with the minimal DNA-binding domain of XPA (XPA-MBD: M98–F219) suggest that a basic cleft located in the loop-rich subdomain plays a role in DNA-binding. Here, XPA–DNA interactions are further characterized by NMR spectroscopy from the vantage point of the DNA using a single-stranded DNA nonamer, dCCAATAACC (d9). Up to 2.5 molar equivalents of XPA-MBD was titrated into a solution of d9. A subset of ³¹P resonances of d9 were observed to broaden and/or shift providing direct evidence that XPA-MBD binds d9 by a mechanism that perturbs the phosphodiester backbone of d9. The interior five residues of d9 broadened and/or shifted before ³¹P resonances of phosphate groups at the termini, suggesting that when d9 is bound to XPA-MBD the internal residues assume a correlation time that is characteristic of the molecular weight of the complex while the residues at the termini undergo a fraying motion away from the surface of the protein on a timescale such that the line widths are more characteristic of the molecular weight of ssDNA. A molecular model of the XPA-MBD complex with d9 was calculated based on the ¹⁵N (XPA-MBD) and ³¹P (d9) chemical shift mapping studies and on the assumption that electrostatic interactions drive the complex formation. The model shows that a nine residue DNA oligomer fully covers the DNA-binding surface of XPA and that there may be an energetic advantage to binding DNA in the 3′→5′ direction rather than in the 5′→3′ direction (relative to XPA-MBD α-helix-3).     

1H-NMR studies on the gene-5-encoded single-stranded DNA binding protein of the filamentous bacteriophage IKe. General spectral and structural features

Under physiological conditions and at concentrations needed for NMR studies, severe aggregation of the gene-5 protein of the filamentous phage IKe occurs. Conditions are described for which well-resolved 1H-NMR spectra of the protein can be obtained. The aromatic part of the spectrum is analyzed by means of two-dimensional NMR techniques; a complete interpretation is presented. Oligonucleotide binding studies reveal that just one phenylalanyl residue and one tyrosyl residue are influenced by the binding of rAMP, (dA)2, (dA)3, (dA)4, (dA)6, d(pT)3 or (dT)4. Upon binding, the aromatic resonances of these amino acid residues are shifted upfield by about 0.4-0.5 ppm. NMR measurements at different pH values demonstrate that only one of the two histidyl residues is freely titratable. From CIDNP experiments it is concluded that three out of five tyrosyl residues are located at the surface of the protein. Measurements carried out as a function of protein concentration indicate the occurrence of specific protein-protein interactions between dimeric gene-5-protein molecules. The data obtained are compared with those available for the gene-5 protein of M13. It follows from the comparison that these proteins mimic each other in almost every respect.     

Study of the structure of the duplex (Phn-NH(CH2)2NH)pd(CCAAACA) .pd(TGTTTGGC) with covalently bound 10-(2-hydroxyethyl)phenazine in an aqueous solution by 2D-1H-NMR spectroscopy]

The spatial structure of duplex (Phn-NH(CH2)2NH)pd(CCAAACA).pd(TGTTTGGC) having a N-(2-oxyethyl)-phenazinium residue covalently linked with the 5'-terminal phosphate of the heptanucleotide was studied by means of one- and two-dimensional 1H-NMR spectroscopy. The resonances of phenazinium protons, ethylenediamine linker protons, as well as, oligonucleotide H5/H6/H8/CH3 base protons and H1',H2'a, H2'b, H3', H4' deoxyribose protons have been assigned by means of 1H-COSY, 1H-NOESY and 1H-13C-COSY. The presence of the phenazine residue in duplex causes an additional imino proton signal of the terminal (G-7).(C-1) base pair, suggesting a higher stability of the duplex (Phn-NH(CH2)2NH)pd(CCAAACA).pd(TGTTTGGC) as compared to the unmodified duplex pd(CCAAACA).pd(TGTTTGGC). Analysis of NOE interactions between protons of the dye and the oligonucleotides show the phenazinium polycyclic system to intercalate between G-7 and C-8 residues of the octanucleotide.     

Homo- and heteronuclear NMR studies of the human retinoic acid receptor beta DNA-binding domain: sequential assignments and identification of secondary structure elements.

An 80 amino acid polypeptide corresponding to the DNA-binding domain (DBD) of the human retinoic acid receptor beta (hRAR-beta) has been studied by 1H homonuclear and 15N-1H heteronuclear two- and three-dimensional (2D and 3D) NMR spectroscopy. The polypeptide has two putative zinc fingers homologous to those of the receptors for steroid and thyroid hormones and vitamin D3. The backbone 1H resonances as well as over 90% of the side-chain 1H resonances have been assigned by 1H homonuclear 2D techniques except for the three N-terminal residues. The assignments have been confirmed further by means of 15N-1H heteronuclear 3D techniques, which also yielded the assignments of the 15N resonances. Additionally, stereospecific assignments of methyl groups of five valine residues were made. Sequential and medium-range NOE connectivities indicate several elements of secondary structure including two alpha-helices consisting of residues E26-Q37 and Q61-E70, a short antiparallel beta-sheet consisting of residues P7-F9 and S23-C25, four turns consisting of residues P7-V10, I36-N39, D47-C50, and F69-G72, and several regions of extended peptide conformation. Similarly, two helices are found in the glucocorticoid receptor (GR) DBD in solution [H?rd et al. (1990) Science 249, 157-160] and in crystal [Luisi et al. (1991) Nature 352, 497-505], and in the estrogen receptor (ER) DBD in solution [Schwabe et al. (1990) Nature 348, 458-461], although the exact positions and sizes of the helices differ somewhat. Furthermore, long-range NOEs suggest the existence of a hydrophobic core formed by the two helices.