Combinatorial selection and delivery of thioaptamers.

Oligonucleotide-based agents are emerging as potential therapeutic agents that can be attractive alternatives for the small-molecule chemical drugs. Monothiophosphate-backbone-modified DNA aptamers (thioaptamers) that specifically and tightly bind to the RNase H domain of the HIV RT (reverse transcriptase) have been isolated from nucleic acid libraries using combinatorial selection methods. The selected thioaptamer inhibited RNase H activity of the HIV RT in in vitro studies. In cell cultures, the transfected thioaptamer markedly reduced HIV production in a dose-dependent manner. Gel electrophoretic mobility-shift assays and NMR spectroscopy showed that the selected thioaptamer binds to the isolated RNase H domain, but did not bind to a structurally similar RNase H from Escherichia coli. In cell cultures, the transfected thioaptamer showed a dose-dependent inhibition of HIV replication, with a maximal inhibition of 83%. Using various liposome-delivery agents, the DNA thioaptamer was transfected into HIV-infected astrocytoma adherent cells with greater than 70% efficiency.

Potential double-flipping mechanism by E. coli MutY.

To understand the structural basis of the recognition and removal of specific mismatched bases in double-stranded DNAs by the DNA repair glycosylase MutY, a series of structural and functional analyses have been conducted. MutY is a 39-kDa enzyme from Escherichia coli, which to date has been refractory to structural determination in its native, intact conformation. However, following limited proteolytic digestion, it was revealed that the MutY protein is composed of two modules, a 26-kDa domain that retains essential catalytic function (designated p26MutY) and a 13-kDa domain that is implicated in substrate specificity and catalytic efficiency. Several structures of the 26-kDa domain have been solved by X-ray crystallographic methods to a resolution of up to 1.2 A. The structure of a catalytically incompetent mutant of p26MutY complexed with an adenine in the substrate-binding pocket allowed us to propose a catalytic mechanism for MutY. Since reporting the structure of p26MutY, significant progress has been made in solving the solution structure of the noncatalytic C-terminal 13-kDa domain of MutY by NMR spectroscopy. The topology and secondary structure of this domain are very similar to that of MutT, a pyrophosphohydrolase. Molecular modeling techniques employed to integrate the two domains of MutY with DNA suggest that MutY can wrap around the DNA and initiate catalysis by potentially flipping adenine and 8-oxoguanine out of the DNA helix.

NMR evidence for syn-anti interconversion of a trans opened (10R)-dA adduct of benzo[a]pyrene (7S,8R)-diol (9R,10S)-epoxide in a DNA duplex.

2D NMR has been used to examine the structure and dynamics of a 12-mer DNA duplex, d(T(1)A(2)G(3)T(4)C(5)A(6)A(7)G(8)G(9)G(10)C(11)A(12))-d(T(13)G(14)C( 15)C(16)C(17)T(18)T(19)G(20)A(21)C(22)T(23)A(24)), containing a 10R adduct at dA(7) that corresponds to trans addition of the N(6)-amino group of dA(7) to (-)-(7S,8R,9R,10S)-7,8-dihydroxy-9, 10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene [(-)-(S,R,R,S)-BP DE-2]. This DNA duplex contains the base sequence for the major dA mutational hot spot in the HPRT gene when Chinese hamster V79 cells are given low doses of the highly carcinogenic (+)-(R,S,S,R)-BP DE-2 enantiomer. NOE data indicate that the hydrocarbon is intercalated on the 5'-side of the modified base as has been seen previously for other oligonucleotides containing BP DE-2 (10R)-dA adducts. 2D chemical exchange-only experiments indicate dynamic behavior near the intercalation site especially at the 10R adducted dA, such that this base interconverts between the normal anti conformation and a less populated syn conformation. Ab initio molecular orbital chemical shift calculations of nucleotide and dinucleotide fragments in the syn and anti conformations support these conclusions. Although this DNA duplex containing a 10R dA adduct exhibits conformational flexibility as described, it is nevertheless more conformationally stable than the corresponding 10S adducted duplex corresponding to trans opening of the carcinogenic isomer (+)-(R,S,S, R)-BP DE-2, which was too dynamic to permit NMR structure determination. UV and imino proton NMR spectral observations indicated pronounced differences between these two diastereomeric 12-mer duplexes, consistent with conformational disorder at the adduct site and/or an equilibrium with a nonintercalated orientation of the hydrocarbon in the duplex containing the 10S adduct. The existence of conformational flexibility around adducts may be related to the occurrence of multiple mutagenic outcomes resulting from a single DE adduct.

Synchronized adiabatic decoupling.

A new decoupling scheme termed "synchronized adiabatic decoupling" is developed for use in the indirectly detected dimension. After each increment, the decoupling sequence is replaced by another one with different period T or different initial period T(ini) so that sampling always occurs at the end of a complete decoupling period. The effects of J coupling are therefore completely averaged out for all data points. As a result, all decoupling sidebands disappear and the center band increases correspondingly. Since the synchronized adiabatic decoupling does not require conventional editing techniques to cancel the sidebands, it is useful in high-field gradient-enhanced multidimensional experiments with only a single scan per increment.

Solution conformation of a bulged adenosine base in an RNA duplex by relaxation matrix refinement.

Bulges are common structural motifs in RNA secondary structure and are thought to play important roles in RNA-protein and RNA-drug interactions. Adenosine bases are the most commonly occurring unpaired base in double helical RNA secondary structures. The solution conformation and dynamics of a 25-nucleotide RNA duplex containing an unpaired adenosine, r(GGCAGAGUGCCGC): r(GCGGCACCUGCC) have been studied by NMR spectroscopy and MORASS iterative relaxation matrix structural refinement. The results show that the bulged adenosine residue stacks into the RNA duplex with little perturbation around the bulged region. Most of the bases in the RNA duplex adopt C(3)'-endo conformation, exhibiting the N-type sugar pucker as found in the A form helices. The sugars of the bulged residue and the 5' flanking residue to it are found to exhibit C(2)'-endo conformation. None of the residues are in syn conformation.

Structural similarities between MutT and the C-terminal domain of MutY.

One of the functions of MutY from Escherchia coli is removal of adenine mispaired with 7,8-dihydro-8-oxoguanine (8-oxoG), a common lesion in oxidatively damaged DNA. MutY is composed of two domains: the larger N-terminal domain (p26) contains the catalytic properties of the enzyme while the C-terminal domain (p13) affects substrate recognition and enzyme turnover. On the basis of sequence analyses, it has been recently suggested that the C-terminal domain is distantly related to MutT, a dNTPase which hydrolyzes 8-oxo-dGTP [Noll et al. (1999) Biochemistry 38, 6374-6379]. We have studied the solution structure of the C-terminal domain of MutY by NMR and find striking similarity with the reported solution structure of MutT. Despite low sequence identity between the two proteins, they have similar secondary structure and topology. The C-terminal domain of MutY is composed of two alpha-helices and five beta-strands. The NOESY data indicate that the protein has two beta-sheets. MutT is also a mixed alpha/beta protein with two helices and two beta-sheets composed of five strands. The secondary structure elements are similarly arranged in the two proteins.

Adiabatic decoupling sidebands.

An analytical solution is given for amplitudes and phases of adiabatic decoupling sidebands as a function of spin inversion time tau. Since all the adiabatic decoupling phases theta(t, tau) refocus at two periods (2T) of the decoupling pulse, the sidebands are located at n/2T rather than at n/T as observed in other decoupling schemes. The real (R(n)(tau)) and imaginary (I(n)(tau)) amplitudes of the sidebands have symmetry R(n)(tau) = R(-n)(tau) and I(n)(tau) = -I(-n)(tau), forming a mirror image between the counterparts of the sidebands. When frequency sweep changes direction all I(n)(tau) are inverted while all R(n)(tau) remain unchanged, leading to pure absorption sidebands with two accumulations as demonstrated by Kupce and Freeman, and to an exchange of sidebands between counterparts. The sum of the real parts for sidebands n = 1 and 2 is almost a constant near on-resonance decoupling, and it increases substantially for large decoupling offsets. The phase defocusing can be minimized for all decoupling offsets by inserting an initial decoupling period with T(ini) = T/2, eliminating all sidebands located at n/2T (n = +/-1, +/-3, +/-5, ...).

Enhanced suppression of residual water in a "270" WET sequence.

In certain water suppression experiments, the residual water, which comes from a region away from the center of the RF coil and experiences a much smaller flip angle than the designed one, may appear. The residual water in the WET sequence can be reduced significantly by using a composite 90(x)( degrees )90(y)( degrees )90(-x)( degrees )90(-y)( degrees ) pulse, which de-excites molecules experiencing a small flip angle. The composite pulse, however, has two null excitation points near on resonance, causing a severe loss of spectrum intensity and baseline distortion toward the null points. Since the residual water experiences a very small flip angle, it can be treated as a linear spin system; i.e., the intensity of the residual water is proportional to the pulse strength and width. Based on this principle, the residual water can be reduced dramatically by replacing the 90 degrees pulse in the "270" WET sequence with a 270 degrees pulse for one out of every four scans, without noticeable loss of intensity and baseline distortion.

How Do Proteins Recognize DNA? Solution Structure and Local Conformational Dynamics of Lac Operators by 2D NMR.

Abstract The NMR structures of the symmetrical lac operator DNA fragment, d(TGTGAGCGCTCACA)(2) and it's mutant, d(TATGAGCGCTCATA)(2), were determined by the MORASS hybrid relaxation matrix/restrained molecular dynamics methodology. The (1)H chemical shifts of nearly all of the non-exchangeable protons were assigned using standard two-dimensional NMR techniques. Ultimately, 181 NOE volumes/strand were used in the final MORASS structural determination for each molecule. Both model built A- and B-form DNA starting geometries were used which converged to final structures giving 1.85Å and 1.32Å RMSD for the wild-type and mutant operators respectively. An excellent agreement between experimental NOESY data with that calculated from the final structures was achieved. The sequence dependence of the DNA backbone torsional angle conformational dynamics was further examined using trajectories from four 500 ps AMBER PMES molecular dynamics calculations performed on the final NMR structures. These are discussed as well as the experimental vs. calculated JH3'-P coupling constants and their relation to backbone dynamics.

Aptamers containing thymidine 3'-O-phosphorodithioates: synthesis and binding to nuclear factor-kappaB.

Aptamers targeting NF-kappaB containing thymidine 3'-O-phosphorodithioates in selected positions of an oligonucleotide duplex were synthesized. Binding affinities to NF-kappaB varied with the number and positions of the dithioate backbone substitutions. One of the aptamers showed specific binding to a single NF-kappaB dimer in cell culture extracts.

Hybrid-hybrid matrix structural refinement of a DNA three-way junction from 3D NOESY-NOESY.

Homonuclear 3D NOESY-NOESY has shown great promise for the structural refinement of large biomolecules. A computationally efficient hybrid-hybrid relaxation matrix refinement methodology, using 3D NOESY-NOESY data, was used to refine the structure of a DNA three-way junction having two unpaired bases at the branch point of the junction. The NMR data and the relaxation matrix refinement confirm that the DNA three-way junction exists in a folded conformation with two of the helical stems stacked upon each other. The third unstacked stem extends away from the junction, forming an acute angle (approximately 60 degrees) with the stacked stems. The two unpaired bases are stacked upon each other and are exposed to the solvent. Helical parameters for the bases in all three strands show slight deviations from typical values expected for right-handed B-form DNA. Inter-nucleotide imino-imino NOEs between the bases at the branch point of the junction show that the junction region is well defined. The helical stems show mobility (+/- 20 degrees) indicating dynamic processes around the junction region. The unstacked helical stem adjacent to the unpaired bases shows greater mobility compared to the other two stems. The results from this study indicate that the 3D hybrid-hybrid matrix MORASS refinement methodology, by combining the spectral dispersion of 3D NOESY-NOESY and the computational efficiency of 2D refinement programs, provides an accurate and robust means for structure determination of large biomolecules. Our results also indicate that the 3D MORASS method gives higher quality structures compared to the 2D complete relaxation matrix refinement method.

"BEST" homonuclear adiabatic decoupling for 13C- and 15N-double-labeled proteins.

The cyclic irradiation sidebands appearing in homonuclear adiabatic decoupling are calculated in detail, which reveals the origin of the antisymmetric sidebands. The sidebands can be inverted by inserting an initial decoupling with a different period, but the same f1rms as the main decoupling that is required for Bloch-Siegert shift compensation. The sidebands can be eliminated in a broad decoupling range by adding spectra of opposite sidebands. Based on this scheme, an offset-independent double-adiabatic decoupling, named Bloch-Siegert Shift Eliminated and Cyclic Sideband Trimmed Double-Adiabatic Decoupling, or "BEST" decoupling for short, is constructed, which not only compensates the Bloch-Siegert shift as shown earlier by Zhang and Gorenstein (1998) but also eliminates residual sidebands effectively.

Novel combinatorial selection of phosphorothioate oligonucleotide aptamers.

A novel combinatorial approach is described for construction and screening of enhanced nuclease-resistant phosphorothioate DNA "decoys" or "aptamers." Aptamers have been selected that bind tightly to the nuclear factor for human IL6 (NF-IL6), a basic leucine zipper transcription factor involved in the induction of acute-phase responsive and cytokine gene promotors in response to inflammation. Using a random combinatorial selection approach and dNTP(alphaS)'s in the PCR amplification, we can select specific thio-substituted agents which have the highest specificity in binding to target NF-IL6. Using a 22-nucleotide-long duplex random library, nanomolar binding, specific 22-mer thiophosphate backbone substitution sequences (at dA positions only) were selected. These show a different consensus sequence than normal phosphate backbone CCAAT/enhancer binding protein recognition sites for NF-IL6. Unlike the wild-type 10-mer sequences, which bind 1 protein dimer/duplex, these 22-mer thiophosphate aptamers bind with a stoichiometry of 2 dimers/duplex.

Bloch-Siegert shift compensated and cyclic irradiation sidebands eliminated, double-adiabatic homonuclear decoupling for 13C- and 15N-double-labeled proteins.

A Gaussian-shaped, offset-independent adiabatic decoupling is adopted to decouple 13CO from 13C alpha or vice versa for 13C- and 15N-double-labeled proteins, together with a compensating decoupling applied on the opposite side of the 13C alpha resonance frequency. In a quite broad range, the double-adiabatic decoupling eliminates efficiently the cyclic sidebands caused by direct irradiation of the adiabatic decoupling and reduces significantly the Bloch-Siegert shift. The remaining Bloch-Siegert shift, which is almost a linear function of offset, can be removed by a dilated evolution time. The decoupling sequence is also quite insensitive to the RF field intensity or inhomogeneity due to the reduced transverse components of RF field at 13C alpha, leading to an efficient decoupling even under unfavorable conditions.

Solution structure of the minor conformer of a DNA duplex containing a dG mismatch opposite a benzo[a]pyrene diol epoxide/dA adduct: glycosidic rotation from syn to anti at the modified deoxyadenosine.

Polycyclic aromatic hydrocarbons (PAHs) are widespread environmental contaminants whose metabolism in mammals results in deleterious cell transformation. Covalent modification of DNA by diol epoxides metabolically formed from PAHs such a benzo[a]pyrene (BaP) provides a mechanism for the genotoxicity, mutagenicity, and carcinogenicity of PAHs. We had previously reported NMR evidence for a minor conformer of the duplex d(G1G2T3C4A5*C6G7A8G9).d(C10T11C12G13G14G15A16C17C18) containing a dG14 mismatch opposite a dA5* residue modified at the exocyclic amino group by trans addition to (+)-(7R,8S,9S,10R)-7,8-dihydroxy-9,10-epoxy-7,8,9,10-tetrahydrobenzo[a] pyrene [Yeh, H.J.C., Sayer, J.M., Liu, X., Altieri, A.S., Byrd, R.A., Lashman, M.K., Yagi, H., Schurer, E.J., Gorenstein, D.G., & Jerina, D.M. (1995) Biochemistry 34, 13570-13581]. In the present work, we describe the structure of this minor conformer (ca. 17% of the total conformer population). This represents the first structural determination of a minor conformer of a carcinogen-lesion DNA adduct. Two-dimensional NOESY, ROESY, TOCSY, and exchange-only spectra at 750 MHz allowed nearly complete sequential assignment of both conformers. In the minor conformer, the adducted base assumes an anti-glycosidic torsion angle whereas in the major conformer it assumes an unusual syn-glycosidic torsion angle. The aromatic hydrocarbon in the minor conformer is intercalated between dG13 and dG14, preserving the energetically favorable stacking interactions found in the major conformer. The major structural differences between the two conformers appear to be near the lesion site as evidenced by the large chemical shift differences between major and minor conformer protons near the lesion site; away from this site, the chemical shifts of the major and minor conformer protons are nearly identical. Because any of the conformations of benzo[a]pyrene diol epoxide-modified DNA may contribute to tumorigenic activity, structural determination of all conformations is essential for the elucidation of the mechanism of cell transformation initiated by covalent modification of DNA by PAHs.

Amphiphilicity determines binding properties of three mitochondrial presequences to lipid surfaces.

The interactions of three peptides, which correspond to presequences that direct mitochondrial protein import, with model membrane systems were characterized using NMR, fluorescence, and circular dichroism spectroscopies. The positively charged peptides adopted an ordered secondary structure only when the negatively charged phospholipid, cardiolipin, was present in small unilamellar vesicles. Conversely, the peptides adopted an ordered secondary structure in the presence of micelles formed from both formally neutral and negatively charged detergents. The peptides had the same relative affinity for micelles and small unilamellar vesicles containing 20% cardiolipin. Amide proton exchange rates showed that the region of the helical structure which had the greatest hydrophobic moment interacted most readily with micelles. Therefore, it appears that a major determinant of binding to lipid surfaces is the ability of the peptide to attain the correct orientation of hydrophobic and hydrophilic groups. For the three peptides studied, affinity also correlated with the length of the helix, but not with hydrophobic surface area. In each case, the interacting segment of the peptide was toward the C-terminal end of the helix. Previous work has allowed us to postulate that the N-terminus of the presequence is vital for import [Wang, Y., & Weiner, H. (1993) J. Biol. Chem. 268, 4759-4765] and the C-terminal end is essential for membrane interaction [Karslake, C., Piotto, M., Pak, Y. K., Weiner, H., & Gorenstein, D. G. (1990) Biochemistry 29, 9872-9878]. On the basis of the data that are now available, it appears that the interaction with membrane surfaces may depend on the location of an amphiphilic region of the sequence that is near but not necessarily at the C-terminus.