Chang-Shung Tung

A Rigorous Basepair Oriented Description of DNA Structures

We propose new, rigorous definitions for (i) basepair fixed coordinate systems and (ii) the twist, tilt, and roll angles (called tau, t, rho) describing the relative orientation of adjacent basepairs and bases in a pair, in arbitrary DNA structures obtained from x-ray diffraction, 2D NMR, or energy calculations. In contrast to the corresponding angular parameters (tg, theta T, theta R) and coordinate systems introduced by Dickerson and co-workers and currently in use, our angular parameters and coordinate systems, together with a set of three displacement parameters, dx, dy, dz, provide a mathematically correct and general description of DNA conformations at the basepairs and/or base level. For instance, our description is applicable when the DNA structure considered is inherently curved, irregular, and/or does not possess dyad (or pseudodyad) axes. We develop a computationally convenient algorithm for rigorous DNA conformational analysis and apply it to some of the known crystal structures. We establish the connection to the currently used parameters and test the consistency and efficiency of our methodology by reconstructing the Dickerson B dodecamer using only the sequence and the set of parameters obtained from the atomic coordinates. The six parameter (tau, t, rho, dx, dy, dz) basepair level reconstruction is good but not perfect. Perfect reconstruction is obtained when one also considers each base in a basepair (consideration of propeller twist alone is not sufficient). The variation of the rigorous parameters proposed along the sequence is much larger, but their average values agree with fiber and solution data much better than in the case of the currently used set. The results of our analysis do not support Trifonovs AA.TT wedge model for DNA curvature but provide some evidence in favor of the Crothers junction-bend model. We point out some of the limitations of basepair level approaches when applied to DNA structure prediction and quantitative understanding of sequence-dependent variations in structure.

An Extension of the Rigorous Base-unit Oriented Description of Nucleic Acid Structures

Our proposed description for DNA base/base-pair structures (1), though rigorous, does not satisfy some of the requirements as established at the Cambridge Workshop (2). Here, we propose a revised description for base/base-unit structures of nucleic acids. This new description is as rigorous and satisfies all the requirements (2). Following the original approach, the moment-of-inertia frame is still the choice of the internal coordinate system for a base/base-unit. The revised description has the minimum number of parameters (i.e., six parameters per rigid body) in the set. Besides regular Watson-Crick type of helices (e.g., A-DNA, A-RNA, B-DNA, Z-DNA, etc.), the revised description also works for non-Watson-Crick, multiple stranded molecules (e.g., triplex, quadruplex, etc.) as well as parallel stranded molecules.

ENERGETICS OF THE HAIRPIN TO MISMATCHED DUPLEX TRANSITION OF D(GCCGCAGC) ON NACL SOLUTION

We report Potential of Mean Force studies to describe the relative thermodynamic stabilities of d(GCCGCAGC) in a mismatched duplex and a hairpin monomer conformation in NaCl solution. The PMF calculations are combined with previous molecular mechanics and normal mode analysis in order to estimate the role of different components of the free energy in determining the relative stability of the duplex and hairpin structures. The high entropy associated with the loop region and the lack of minor groove phosphate-phosphate interactions in the hairpin compete against the gain in enthalpic contribution to the free energy due to base pairing in the mismatched duplex. The combined free energy calculations show that the hairpin is the most stable conformation at low salt and that a hairpin to duplex transition takes place at approximately 0.47 M NaCl. In addition, we studied the hairpin to partially stacked single helical conformation equilibrium at low salt. We found a small variation in transition temperature in salt concentration, delta Tm/delta log10(cs) approximately 2-3 degrees K/decade, in contrast to the duplex to hairpin or duplex to partially stacked single helix transition where the transition temperature exhibited marked dependence on salt concentration. This is in qualitative agreement with experimental data. Based on the Potential of Mean Force free energy calculation, the order of relative stability of the three-conformations studied varies with salt concentration. We observed the following orders of stability: stacked single helix greater than hairpin greater than duplex for cs less than 0.77 M NaCl; single helix greater than duplex greater than hairpin for 0.77 less than Cs less than 2.1 M; and duplex greater than hairpin greater than single strand for cs greater than 2.1 M. From the calculated PMF free energy curves in the NaCl concentration range, 0.012 less than cs less than 5.0 M, we can assign upper and lower bounds for the non-ionic differences in free energy between the duplex, hairpin, and stacked single helical states (at standard conditions: cs = 1.0 M, T = 25 degrees C, and 1 M oligomer concentration). We found that for delta G duplex single helix = G duplex - 2 x G single helix less than -7.38 Kcal/mol, the single helix is the least stable state. For the duplex-to-hairpin free energy difference in the range, -1.87 less than delta G duplex-hairpin less than 0.03 Kcal/mol, there will always be a salt-induced hairpin-to-duplex transition for 0.01 less than cs less than 1.6 M NaCl. If delta G duplex-hairpin less than -1.87, the duplex is always more stable than the hairpin; and for delta G duplex-hairpin greater than Kcal/mol, the hairpin state is always more stable than the duplex, for all salt concentrations.

Rigorous description of DNA structures. II. On the computation of best axes, planes, and helices from atomic coordinates.

We show that diagonalization of the moments of inertia tensor of an arbitrary molecular structure automatically yields the best (least squares) points, line, and plane through the structure. Furthermore, we show how to compute best helices using atomic coordinates. These results supplement the general methodology for description of DNA structures proposed recently (D.M. Soumpasis and C.-S. Tung, J. Biomol. Struct. Dyn. 6, 397-420, 1988).

Structural Study of Homeodomain Protein-DNA Complexes Using a Homology Modeling Approach

The homeodomain is a conserved protein motif that binds to DNA and plays a central role in gene regulation. We use homeodomain as a model system to study the specific interactions between protein and DNA in a complex. Following the fundamental concept of homology modeling, we have developed an algorithm for predicting structures of both protein and DNA using the known structure of a similar complex as the template. The accuracies of the algorithm in predicting the complex structures are evaluated when two of the homeodomain protein-DNA complexes with known structures (antennapedia and MATalpha2) are selected as test systems. This algorithm allows structural studies of homeodomain binds to DNA with different sequences.

The Construction of DNA Helical Duplexes Along Prescribed 3-D Curves

Curved DNA structures can be either intrinsic or induced through interaction with proteins. Large DNA molecules can adopt complex curved shapes in space. Many curved DNA structures (e.g., Cro-operator, Cap-operator, nucleosomal DNA, etc.) are believed to have important biological functions. To model curved DNA molecules is a challenging task. In this work, we introduce a method for the computer generation of DNA structures having any prescribed 3-D shape in space. The approach is purely geometrical and highly efficient. This method is used successfully to construct an atomic level nucleosomal DNA that is consistent with the experimental data. The smallest closed DNA circle we are able to construct with the method is a 51 bp DNA duplex.

A Quantitative Measure of DNA Curvature Enabling the Comparison of Predicted Structures

A growing body of data indicates that the equilibrium structures of some DNA fragments are curved and that curvature is sequence-directed. We describe a quantitative measure of DNA curvature that can be used for evaluating and comparing current proposed models for the molecular basis of DNA curvature. We demonstrate that this measure, in conjunction with any given prediction model, enables both the comparison of experimental data to predictions and the scanning of nucleotide sequence databases for potential curved regions.

A GEOMETRICAL APPROACH IN FOLDING A PSEUDOKNOT MOTIF WITHIN THE E. COLI 16S-RNA

Except for tRNA, the tertiary structure of RNA molecules are very little known. The many possibilities in the arrangement of different helices in space and the flexibility in the single-stranded loops that connect the helical regions make the modeling of the tertiary structure of RNA molecule a very complex task. Here, we introduce an approach to fold RNA tertiary structure based only on the information of the secondary structure and the stereochemistry of the molecule. This approach was used to construct an atomic structure of a pseudoknot (bases 500-545) in the E. coli 16S RNA. The resulting structure is a closely packed molecule that is consistent with the predicted secondary structure and stereochemically feasible. This new approach is very general and easily adaptable. Experimental data (e.g., NMR, fluorescence energy transfer, etc.), as they become available, can be incorporated directly into the approach to improve the accuracy of the modeled structure.

Three-dimensional model of a selective theophylline-binding RNA molecule

A three‐dimensional (3D) model for an RNA molecule that selectively binds theophylline but not caffeine is proposed. This RNA, which was found using SELEX (Jenison et al., 1994), is 10 000 times more specific for theophylline (KD = 320 nM) than for caffeine (KD = 3.5 mM), although the two ligands are identical except for a methyl group substituted at N7 (present only in caffeine). The binding affinity for ten xanthine‐based ligands was used to derive a comparative molecular field analysis model (R2 = 0.93 for three components, with cross‐validated R2 of 0.73), using the SYBYL and GOLPE programs. A pharmacophoric map was generated to locate steric and electrostatic interactions between theophylline and the RNA binding site. This information was used to identify putative functional groups of the binding pocket and to generate distance constraints. On the basis of a model for the secondary structure (Jenison et al., 1994), the 3D structure of this RNA was then generated using the following method: each helical region of the RNA molecule was treated as a rigid body; single‐stranded loops with specific end‐to‐end distances were generated. The structures of RNA‐xanthine complexes were studied using a modified Monte Carlo algorithm. The detailed structure of an RNA–ligand complex model, as well as possible explanations for the theophylline selectivity are discussed.

Characterization of the distribution of potential short restriction fragments in nucleic acid sequence databases: Implications for an alternative to chemical synthesis of oligonucleotides

We have searched the GenBank nucleic acid sequence database for potential short restriction fragments. All possible oligonucleotides up to length five are found at least once flanked by known restriction recognition patterns. Thus, searches in the database for a specific sequence corresponding to a desired oligonucleotide would often point to one or more sources of short, retrievable fragments containing that sequence. These results underscore the potential of nucleic acid sequence databases in planning experiments.