Session S8 - Nucleic Acids II.
ORAL session, Wednesday afternoon, March 24
510A, Palais des Congres

[S8.001] Computing Optimal Binding of Two Nucleic Acid Chains

We present an algorithm to calculate the optimal binding conformation and free energy of two RNA molecules, one or both oligomeric. This algorithm has applications to modeling DNA microarrays, RNA splice-site recognitions, and other anti-sense problems. While other recent algorithms perform the same calculation in time proportional to the sum of the lengths cubed O((N_1+N_2)^3), our \sc bindigo algorithm scales as the product of the sequence lengths O(N_1 \cdot N_2). To demonstrate its speed and utility, we use \sc bindigo to screen vast sequence data to investigate the binding proclivities of U1 snRNA to donor splice sites.

[S8.002] Thermodynamic Modeling of Donor Splice Site Recognition in pre-mRNA

When eukaryotic genes are edited by the spliceosome, the first step in intron recognition is the binding of a U1 snRNA with the donor (5') splice site. We model this interaction thermodynamically to identify splice sites. Applied to a set of 65 annotated genes, our Finding with Binding method achieves a significant separation between real and false sites. Analyzing binding patterns allows us to discard a large number of decoy sites. Our results improve statistics-based methods for donor site recognition, demonstrating the promise of physical modeling to find functional elements in the genome.

[S8.003] Isomers of the DNA bases and their possible role in base pair mismatch

We have found stable isomers of the two DNA purines, adenine and guanine, in global searches using a force field potential and ab initio local relaxations. The stability characteristics were investigated by searching for the lowest energy connecting saddles, and by using Car-Parrinello molecular dynamics. Expected equilibrium concentrations and possible non-equilibrium production mechanisms are discussed. It is suggested that these isomers may play a role in base-pair miss-match leading to point mutation of the gnome, and in a number of other important biological processes based on the purine structures.

[S8.004] Statistical mechanics of base stacking and pairing in DNA melting

We introduce a new statistical mechanics model which fits a wide range of experimental data on DNA melting. Base stacking and pairing are explicitly introduced as distinct degrees of freedom, within an Ising and Zipper model approach. Unlike previous approaches, this simple model describes thermal denaturation of DNA secondary structure for both single stranded (ss) and double stranded (ds) DNA in the whole experimentally accessible temperature range. We present new experimental measurements which allow a stringent comparison between model and experiment.

[S8.005] Molecular dynamics simulations of the d(CCAACGTTGG)_2 decamer in crystal environment: comparison of atom-centered charge, extra-point and polarizable force fields.

Molecular dynamics simulations of the DNA duplex d(CCAACGTTGG)_2 were used to study the relationship between DNA sequence and structure. Three different force fields were used: a traditional description based on atomic point charges, a polarizable force field and an ``extra-point" force field (with additional charges on extra-nuclear sites). It is found that in crystal environment all the force fields reproduce fairly well the sequence-dependent features of the experimental structure. The polarizable force fields, however, outperforms the other two, pointing out to the need of the inclusion of polarization for accurate descriptions of DNA.

[S8.006] Excluded volume effects in unzipping DNA with a force

A double stranded DNA molecule when pulled with a force acting on one end of the molecule can become either partially or completely unzipped depending on the magnitude of the force F. For a random DNA sequence, the number M of unzipped base pairs goes as M~(F-Fc)-2 and diverges at the critical force Fc with an exponent c=2. We find that when excluded volume effect is taken into account for the unzipped part of the DNA, the exponent c=2 is not changed but the critical force Fc is changed. The force versus temperature phase diagram depends on only two parameters in the model, the persistence length and the denaturation temperature. Furthermore a scaling form of the phase diagram can be found. This scaling form is parameter independent and depends only on the spatial dimension. It applies to all DNA molecules and should provide a useful framework for comparison with experiments.

[S8.007] Diffusion-limited looping of biopolymers: theory and biological implications

We show that Kramers rate theory gives a straightforward, accurate estimate of the closing time \tau_c of a biopolymer. The calculation is valid in cases of physical interest and reveals how the time scales of chain relaxation and closing are intertwined, illuminating an apparent conflict between two ways of calculating \tau_c in the flexible limit. For semiflexible polymers, the closing time is shortest for chains that are 3-4 times the persistence length L_p. We explain briefly the biological implications of this result.

[S8.008] A Gibbs sampler for motif detection in phylogenetically close sequences

Genes are regulated by transcription factors that bind to DNA upstream of genes and recognize short conserved ``motifs'' in a random intergenic ``background''. Motif-finders such as the Gibbs sampler compare the probability of these short sequences being represented by ``weight matrices'' to the probability of their arising from the background ``null model'', and explore this space (analogous to a free-energy landscape). But closely related species may show conservation not because of functional sites but simply because they have not had sufficient time to diverge, so conventional methods will fail. We introduce a new Gibbs sampler algorithm that accounts for common ancestry when searching for motifs, while requiring minimal ``prior'' assumptions on the number and types of motifs, assessing the significance of detected motifs by ``tracking'' clusters that stay together. We apply this scheme to motif detection in sporulation-cycle genes in the yeast S. cerevisiae, using recent sequences of other closely-related Saccharomyces species.

[S8.009] Bioinformatics in the Thermodynamic Limit: Applications to pre-mRNA Splice Site Detection

In the thermodynamic limit, physical systems exhibit predictable behavior. Traditional bioinformatics methods of matching patterns from a few cases can be superseded by alternate techniques when the number of instances approaches the thermodynamic limit. By enhancing an already large dataset, we find that we can differentiate real signals from decoys with superior accuracy. We apply our framework to the problem of precursor mRNA splice site detection in human sequences.

[S8.010] Differences in Muon and Muonium Trapping in A- and B-form DNA and in Individual Nucleic Acids: an \textitab initio study

Muon Spin Relaxation (\muSR) experiments in A- and B-form DNA have shown evidence(E. Torikai et al.), Hyperfine Interactions 138, 509 (2001). for an enhanced electron mobility in the more closely-packed A-form. Besides dynamic effects (electronic diffusion) that could cause the observed difference in muon spin relaxation, one also needs to consider static effects (hyperfine fields) that could alter the relaxation rate of the muon. We are therefore investigating the (static) trapping properties of muon (\mu^+) and muonium (\mu^+e^-) in A-form DNA, B-form DNA and, for comparison, also in the four individual nucleic acids, using the first-principles Hartree-Fock method with MP2. Our aim is to understand how the different structural geometries of the three systems and the respective hydration shells of A- and B-form DNA influence the hyperfine interaction of trapped muonium.

[S8.011] Fabrication of Electrically Addressable Nanopore Arrays for DNA Translocation Studies

We will present a novel method of fabricating the nanopore arrays in the range of 10\~20 nm size utilizing Si processing. Using e-beam lithography and anisotropic wet etching with KOH, we make a series of V-grooves on the top of the wafer and one at the right angle on the bottom surface. This opens up a series of nanopores. The unique feature of electrically addressable nanopore arrays (EANA) allows us to control the ionic solution flow and detect DNA translocations in each pore independently. Preliminary results of application of EANA device for studying biomolecules will be shown. This work is supported by NSF-NER-0304325.

[S8.012] Self Assembly of Nanotubes from DNA Tiles

Short sequences of DNA can be designed to self-assemble into extended structures based on Watson-Crick pairing rules. The most generic design scheme makes use of building blocks, or “tiles”, of three or more strands that hybridize into a core of cross-linked double helices with single-stranded sticky ends. We use tiles comprised of 5 strands in a DAE-type double-crossover motif to create tubular polymers that are ~10 nm in diameter, up to 100 µm in length and correspondingly stiff (persistence length ~ 5 µm).

As an artificial biopolymer these DNA nanotubes present a uniquely accessible system in which to study and manipulate self assembly. Here we investigate the stability and length distribution of our nanotubes. We present an understanding of the assembly mechanism derived from fluorescence video microscopy. Length distributions and single nanotube tracking provide direct evidence for rapid nucleation and growth followed by end-to-end annealing. Defects that arise during annealing are identified as a major limiting factor in nanotube stiffness.

[S8.013] Reversible aggregation and Phase Transition in the DNA-linked gold nanoparticle Systems

We have performed a series of experiments to study the aggregation and phase transtion of DNA-linked gold nanoparticle systems. Colloidal gold particles with diameter ranging from 10 to 40 nm are conjugated with thiol-modified single-stranded DNA. Based on the hybridization between complementary DNA single strands, three different types of DNA/nanoparticle networks can be obtained: (i) the ABL system; (ii) the AB system; and (iii) the AAL system. We study the particle size dependence of the aggregation and phase transition and compare the properties among different systems.

[S8.014] Phase Transition of DNA Coated Nanogold Networks

Melting and hybridization of DNA-coated gold nanoparticle networks are investigated with optical absorption spectroscopy and tansmission electron microscopy. Single-stranded DNA-coated nanogold are linked with complementary, single-stranded linker DNA to form particle networks. Network formation results in a solution color change, which can be used for DNA detection. Compared to free DNA, networked DNA-nanoparticle systems result in a sharp melting transition. Melting curves calculated from percolation theory agree with our experimental results(1).

(1) C.-H. Kiang, ``Phase Transition of DNA-Linked Gold Nanoparticles,'' Physica A, 321 (2003) 164--169.

[S8.015] HMGB1 proteins strongly alter the flexibility of single DNA molecules

The High Mobility Group B (HMGB) proteins are known to bind non-sequence specifically to DNA. These abundant non-histone proteins are believed to recognize and stabilize certain bent DNA motifs, possibly reorganizing the nucleosome structure as a prelude to transcription. The active site of the protein, termed the HMG box, is known to bind in the DNA minor groove, slightly intercalating between the bases and producing a strong bend in the backbone. We are interested in the ability of a single HMG box domain from HMGB1 to increase the apparent flexibility of DNA. Here, we use an optical tweezers instrument to measure the forces required to stretch single DNA molecules. By fitting the measured force-extension data to models for polymer elasticity, parameters describing DNA flexibility, including its contour length and persistence length, are revealed. In the presence of nanomolar concentrations of an isolated HMG box from HMGB1, DNA is observed to show a marked decrease in its persistence length from 50 to 10 nanometers.

[S8.016] Effects of kinks on DNA elasticity.

We study the elastic response of a worm-like polymer chain with reversible kink-like structural defects. This is a generic model for (i) dsDNA with sharp bends induced by binding of certain proteins, and (ii) effects of the trans-gauche rotations in the backbone of ssDNA. The problem is solved analytically by extending the standard analogy to the quantum rotator and using the variational method. In the small-force regime, we find that the persistence length of the chain is renormalized due to the presence of kinks, while the response to the strong stretching is determined solely by the bare persistence length. In the limit of the sufficiently rigid chain, the extinction of kinks occurs as a sharp crossover at a certain critical force. This constant-force plateau in the stretching curve is a manifestation of the coexistence of the kinked and the kink-free phases (in 1D). The associated "phase diagram" of the system is calculated.

Part S of program listing