Sequence requirement for specific interaction of an enhancer binding protein (EBP1) with DNA.

Short DNA sequence motifs have been identified in viral and cellular enhancers which represent the binding sites for a variety of trans- acting factors. One such HeLa cell factor, EBP1, has been purified and shown to bind to sequences in the SV40 enhancer. The PRDII element in the human beta-interferon gene regulatory element (IRE) shows strong sequence similarity to the EBP1 binding site in the SV40 enhancer. We demonstrate here that EBP1 binds to its sites in the SV40 enhancer and IRE in a similar manner, making base specific contacts over one complete turn of the DNA double helix. Mutational analysis of the EBP1 sites in the IRE and SV40 enhancer has identified the DNA sequence requirements necessary for specific EBP1/DNA complex formation. In addition, 34 DNA sequences related to the EBP1 binding site were analysed for their ability to bind EBP1. Sequences constituting high affinity binding sites possess the sequence 5'-GG(N)6CC-3'. Single base pair changes in the region between the conserved Gs and Cs can generally be tolerated although it is clear that these intervening bases contribute to binding affinity. Mutations in the recognition site which could lead to gross structural changes in the DNA abolish EBP1 binding.     

Stereoselective interaction with chiral phosphorothioates at the central DNA kink of the EcoRI endonuclease-GAATTC complex.

We have probed the contacts between EcoRI endonuclease and the central phosphate of its recognition site GAApTTC, using synthetic oligonucleotides containing single stereospecific Rp- or Sp-phosphorothioates (Ps). These substitutions produce subtle stereospecific effects on EcoRI endonuclease binding and cleavage. An Sp-Ps substitution in one strand of the DNA duplex improves binding free energy by -1.5 kcal/mol, whereas the Rp-Ps substitution has an unfavorable effect (+0.3 kcal/mol) on binding free energy. These effects derive principally from changes in the first order rate constants for dissociation of the enzyme-DNA complexes. The first order rate constants for strand scission are also affected, in that a strand containing Sp-Ps substitution is cleaved 2 to 3 times more rapidly than a strand containing a normal prochiral phosphate, whereas a strand containing Rp-Ps substitution is cleaved about 3 times slower than normal. As a result, single-strand substitutions produce pronounced asymmetry in the rates of cleavage of the two DNA strands, and this effect is exaggerated in an Rp,Sp-heteroduplex. Ethylation-interference footprinting indicates that none of the Ps substitutions cause any major change in contacts between endonuclease and DNA phosphates. When an Sp-Ps localizes P = O in the DNA major groove, a hydrogen-bonding interaction with the backbone amide-NH of Gly116 of the endonuclease is improved relative to that with a prochiral phosphate having intermediate P-O bond order and delocalized charge.     

Structure of a triple helical DNA with a triplex-duplex junction

Extended purine sequences on a DNA strand can lead to the formation of triplex DNA in which the third strand runs parallel to the purine strand. Triplex DNA structures have been proposed to play a role in gene expression and recombination and also have potential application as antisense inhibitors of gene expression. Triplex structures have been studied in solution by NMR, but have hitherto resisted attempts at crystallization. Here, we report a novel design of DNA sequences, which allows the first crystallographic study of DNA segment containing triplexes and its junction with a duplex. In the 1.8 A resolution structure, the sugar-phosphate backbone of the third strand is parallel to the purine-rich strand. The bases of the third strand associate with the Watson and Crick duplex via Hoogsteen-type interactions, resulting in three consecutive C(+).GC, BU.ABU (BU = 5-bromouracil), and C(+).GC triplets. The overall conformation of the DNA triplex has some similarity to the B-form, but is distinct from both A- and B-forms. There are large changes in the phosphate backbone torsion angles (particularly gamma) of the purine strand, probably due to the electrostatic interactions between the phosphate groups and the protonated cytosine. These changes narrow the minor groove width of the purine-Hoogsteen strands and may represent sequence-specific structural variations of the DNA triplex.     

Remote phosphate contacts trigger assembly of the active site of DNA topoisomerase IB

Vaccinia topoisomerase IB forms a covalent DNA-(3'-phosphotyrosyl)-enzyme intermediate at its target site 5'-CCCTTp downward arrow in duplex DNA. The contributions of backbone electrostatics and individual phosphate oxygens to the transesterification reaction were probed by introducing 22 single Rp and Sp methylphosphonate diastereomers at 11 positions flanking the cleavage site. Methyl groups at eight positions (four on the scissile strand and four on the nonscissile strand) inhibited the rate of single-turnover cleavage by factors of 50-50,000. Stereospecific interference was observed at several phosphates, thereby distinguishing simple electrostatic contributions from putative specific polar contacts to either the pro-Sp or pro-Rp oxygens. The functionally relevant phosphate oxygens are located on the minor groove face of the helix on which the scissile phosphodiester resides. Our findings, combined with available crystal structures of vaccinia and human topoisomerase IB, show how specific phosphate contacts remote from where chemistry occurs are critical for assembly of the active site.     

Crystal structure of the chromodomain helicase DNA-binding protein 1 (Chd1) DNA-binding domain in complex with DNA

Chromatin remodelers are ATP-dependent machines that dynamically alter the chromatin packaging of eukaryotic genomes by assembling, sliding, and displacing nucleosomes. The Chd1 chromatin remodeler possesses a C-terminal DNA-binding domain that is required for efficient nucleosome sliding and believed to be essential for sensing the length of DNA flanking the nucleosome core. The structure of the Chd1 DNA-binding domain was recently shown to consist of a SANT and SLIDE domain, analogous to the DNA-binding domain of the ISWI family, yet the details of how Chd1 recognized DNA were not known. Here we present the crystal structure of the Saccharomyces cerevisiae Chd1 DNA-binding domain in complex with a DNA duplex. The bound DNA duplex is straight, consistent with the preference exhibited by the Chd1 DNA-binding domain for extranucleosomal DNA. Comparison of this structure with the recently solved ISW1a DNA-binding domain bound to DNA reveals that DNA lays across each protein at a distinct angle, yet contacts similar surfaces on the SANT and SLIDE domains. In contrast to the minor groove binding seen for Isw1 and predicted for Chd1, the SLIDE domain of the Chd1 DNA-binding domain contacts the DNA major groove. The majority of direct contacts with the phosphate backbone occur only on one DNA strand, suggesting that Chd1 may not strongly discriminate between major and minor grooves.     

A novel type of interaction between cruciform DNA and a cruciform binding protein from HeLa cells.

We recently identified and enriched a protein (CBP) from HeLa cells with binding specificity for cruciform-containing DNA. We have now studied the interaction of CBP with stable cruciform DNA molecules containing the 27 bp palindrome of SV40 on one strand and an unrelated 26 bp palindrome on the other strand by hydroxyl radical footprinting. The CBP-DNA interaction is localized to the four-way junction at the base of the cruciforms. CBP appears to interact with the elbows of the junctions in an asymmetric fashion. Upon CBP binding, structural distortions were observed in the cruciform stems and in a DNA region adjacent to the junction. These features distinguish CBP from other cruciform binding proteins, which bind symmetrically and display exclusively either contacts with the DNA backbone or structural alterations in the DNA.     

Missing-base and ethylation interference footprinting of P1 plasmid replication initiator.

RepA, the replication initiator protein of plasmid P1, binds to specific 19 bp sequences on the plasmid DNA. Earlier footprinting studies with dimethylsulfate identified the guanines that contact RepA through the major groove of DNA. In this study, base elimination was used to identify the contribution of all four bases to the binding reaction. Depurination and depyrimidation of any base in the neighborhood of the contacting guanines was found to decrease RepA binding. These results are consistent with the notion that RepA contacts bases of two consecutive major grooves on the same face of DNA. We also observed that depurination but not methylation of three guanines (G3, G8 and G9) affected binding. We identified the DNA phosphate groups (3 in the top strand, one of which mapped between G8 and G9, and 4 in the bottom strand, one of which was adjacent to C3) that strongly interfered with RepA binding upon ethylation. These results indicate that certain bases (e.g. G3, G8 and G9) may not contact RepA directly but contribute to base and backbone contacts by maintaining proper structure of the binding site.     

Structure of homeodomain-leucine zipper/DNA complexes studied using hydroxyl radical cleavage of DNA and methylation interference

Homeodomain-leucine zipper (HD-Zip) proteins, unlike most homeodomain proteins, bind a pseudopalindromic DNA sequence as dimers. We have investigated the structure of the DNA complexes formed by two HD-Zip proteins with different nucleotide preferences at the central position of the binding site using footprinting and interference methods. The results indicate that the respective complexes are not symmetric, with the strand bearing a central purine (top strand) showing higher protection around the central region and the bottom strand protected toward the 3' end. Binding to a sequence with a nonpreferred central base pair produces a decrease in protection in either the top or the bottom strand, depending upon the protein. Modeling studies derived from the complex formed by the monomeric Antennapedia homeodomain with DNA indicate that in the HD-Zip/DNA complex the recognition helix of one of the monomers is displaced within the major groove respective to the other one. This monomer seems to lose contacts with a part of the recognition sequence upon binding to the nonpreferred site. The results show that the structure of the complex formed by HD-Zip proteins with DNA is dependent upon both protein intrinsic characteristics and the nucleotides present at the central position of the recognition sequence.     

Base mutation analysis of topoisomerase II-idarubicin-DNA ternary complex formation. Evidence for enzyme subunit cooperativity in DNA cleavage.

Antitumor drugs, such as anthracyclines, interfere with mammalian DNA topoisomerase II by forming a ternary complex, DNA-drug-enzyme, in which DNA strands are cleaved and covalently linked to the enzyme. In this work, a synthetic 36-bp DNA oligomer derived from SV40 and mutated variants were used to determine the effects of base mutations on DNA cleavage levels produced by murine topoisomerase II with and without idarubicin. Although site competition could affect cleavage levels, mutation effects were rather similar among several cleavage sites. The major sequence determinants of topoisomerase II DNA cleavage without drugs are up to five base pairs apart from the strand cut, suggesting that DNA protein contacts involving these bases are particularly critical for DNA site recognition. Cleavage sites with adenines at positions -1 were detected without idarubicin only under conditions favouring enzyme binding to DNA, showing that these sites are low affinity sites for topoisomerase II DNA cleavage and/or binding. Moreover, the results indicated that the sequence 5'-(A)TA/(A)-3' (the slash indicates the cleaved bond, parenthesis indicate conditioned preference) from -3 to +1 positions constitutes the complete base sequence preferred by anthracyclines. An important finding was that mutations that improve the fit to the above consensus on one strand can also increase cleavage on the opposite strand, suggesting that a drug molecule may effectively interact with one enzyme subunit only and trap the whole dimeric enzyme. These findings documented that DNA recognition by topoisomerase II may occur at one or the other strand, and not necessarily at both of them, and that the two subunits can act cooperatively to cleave a double helix.     

Uranyl photofootprinting of triple helical DNA.

Two triple helix structures (15-mers containing only T.A-T triplets or containing mixed T.A-T and C.G-C triplets) have been studied by uranyl mediated DNA photocleavage to probe the accessibility of the phosphates of the DNA backbone. Whereas the phosphates of the pyrimidine strand are at least as accessible as in double stranded DNA, in the phosphates of the purine strand are partly shielded and more so at the 5'-end of the strand. With the homo A/T target increased cleavage is observed towards the 3'-end on the pyrimidine strand. These results show that the third strand is asymmetrically positioned along the groove with the tightest triple strand double strand interactions at the 5'-end of the third strand. The results also indicate that homo-A versus mixed A/G 'Hoogsteen-triple helices' have different structures.     

Substrate recognition and induced DNA deformation by transposase at the target-capture stage of Tn10 transposition

The bacterial transposon Tn10 inserts preferentially into sites that conform to a 9 bp consensus sequence: 5' NGCTNAGCN 3'. However, this sequence is not on its own sufficient to confer target specificity as the base-pairs flanking this sequence also contribute significantly to target-site selection. We have performed a series of "contact-probing experiments" to define directly the protein-DNA interactions that govern target-site selection in the Tn10 system. The HisG1 hotspot for Tn10 insertion was the main focus here. We infer that there is a rather broad zone ( approximately 24 bp) of contact between transposase and target DNA in the target-capture complex. This includes base-specific contacts at all of the purine residues in the consensus positions of the target core and primarily backbone contacts out to 7-8 bp in the two flanking regions immediately adjacent to the core. Also, highly localized sites of chemical hypersensitivity are identified that reveal symmetrically disposed deformations in DNA structure in the target-capture complex. Furthermore, the level of strand transfer is shown to be reduced by phosphorothioate substitution of phosphate groups at or close to the sites of target DNA deformation. Interestingly, for one particular target DNA, a mutant form of HisG1 called MutF, the above phosphorothioate inhibition of strand transfer is suppressed by replacing Mg(2+) with Mn(2+). Based on these results a model for sequence-specific target capture is proposed which attempts to define possible relationships between transposase interactions with the target core and flanking sequences, transposase-induced DNA deformation of the target site and divalent metal ion binding to the target-capture complex.     

5' --> 3' molecular polarity of human replication protein A (hRPA) binding to pseudo-origin DNA substrates

Human replication protein A (hRPA) was previously seen to efficiently bind a 48 bp simian virus 40 (SV40) "pseudo-origin" (PO) substrate that mimics a DNA structure found within the SV40 T antigen-origin (ori) complex. To understand the role of hRPA during the initiation of replication, we examined the PO sequence and structure requirements for hRPA interaction. Binding and unwinding were found to be most efficient when both strands of the central 8 nt single-stranded DNA (ssDNA) bubble region contained a polypyrimidine structure, with these activities proportionately reduced when the bubble region was replaced with a purine tract on one or both strands. Examination of the importance of the two duplex flanks indicates that the early gene side contains a DNA structural feature located one duplex turn from the bubble whose mutation significantly affects the affinity of hRPA for the substrate. When present in the context of ori, mutation of this sequence was seen to have significant effects on SV40 DNA replication in vitro and on the denaturation of ori, indicating that origin activity can be modulated by cis-acting elements which alter the hRPA binding affinity. Use of fork and overhang substrates containing 8 nt pyrimidine or purine arms demonstrates that hRPA binding to DNA involves a particular molecular polarity in which initial hRPA binding occurs on the 5' side of a ssDNA substrate, and then extends in the 3' direction to create a stably bound hRPA. These data have implications on the mechanism of the initiation of eukaryotic DNA replication as well as on the sites of nascent strand synthesis within the origin.     

Reaction of the antitumor antibiotic CC-1065 with DNA. Location of the site of thermally induced strand breakage and analysis of DNA sequence specificity

CC-1065 is a unique antitumor antibiotic produced by Streptomyces zelensis. The potent cytotoxic effects of this drug are thought to be due to its ability to form a covalent adduct with DNA through N3 of adenine. Thermal treatment of CC-1065-DNA adducts leads to DNA strand breakage. We have shown that the CC-1065 structural modification of DNA that leads to DNA strand breakage is related to the primary alkylation site on DNA. The thermally induced DNA strand breakage occurs between the deoxyribose at the adenine covalent binding site and the phosphate on the 3' side. No residual modification of DNA is detected on the opposite strand around the CC-1065 lesion. Using the early promoter element of SV40 DNA as a target, we have examined the DNA sequence specificity of CC-1065. A consensus sequence analysis of CC-1065 binding sites on DNA reveals two distinct classes of sequences for which CC-1065 is highly specific, i.e., 5'PuNTTA and 5'AAAAA. The orientation of the DNA sequence specificity relative to the covalent binding site provides a basis for predicting the polarity of drug binding in the minor groove. Stereo drawings of the CC-1065-DNA adduct are proposed that are predictive of features of the CC-1065-DNA adduct elucidated in this investigation.     

Effect of phosphorothioate substitutions on DNA cleavage by Escherichia coli DNA topoisomerase I

To evaluate the structural influence of the DNA phosphate backbone on the activity of Escherichia coli DNA topoisomerase I, modified forms of oligonucleotide dA(7) were synthesized with a chiral phosphorothioate replacing the non-bridging oxygens at each position along the backbone. A deoxy-iodo-uracil replaced the 5'-base to crosslink the oligonucleotides by ultraviolet (UV) and assess binding affinity. At the scissile phosphate there was little effect on the cleavage rate. At the +1 phosphate, the rectus phosphorus (Rp)-thio-substitution reduced the rate of cleavage by a factor of 10. At the +3 and -2 positions from the scissile bond, the Rp-isomer was cleaved at a faster rate than the sinister phosphorus (Sp)-isomer. The results demonstrate the importance of backbone contacts between DNA substrate and E. coli topoisomerase I.