Uncovering pathways in DNA oligonucleotide hybridization via transition state analysis

Abstract

DNA hybridization plays a central role in biology and, increasingly, in materials science. Yet, there is no precedent for examining the pathways by which specific single-stranded DNA sequences interact to assemble into a double helix. A detailed model of DNA is adopted in this work to examine such pathways and to determine the role of sequence, if any, on DNA hybridization. Transition path sampling simulations reveal that DNA rehybridization is prompted by a distinct nucleation event involving molecular sites with approximately four bases pairing with partners slightly offset from those involved in ideal duplexation. Nucleation is promoted in regions with repetitive base pair sequence motifs, which yield multiple possibilities for finding complementary base partners. Repetitive sequences follow a nonspecific pathway to renaturation consistent with a molecular "slithering" mechanism, whereas random sequences favor a restrictive pathway involving the formation of key base pairs before renaturation fully ensues.

Fig. 1.

Probability distribution for ξ in the TSE, for n = 14,15 (top) and n = 30 (bottom). Data are shown for REP14, REP30 (solid lines); RAN15, RAN30 (dashed lines); and MIX30 (dot–dashed line). The REP systems have their respective standard deviations shown, which were determined by randomly pooling 50% of the ensemble and performing a dozen replicates. These errors are representative of the other systems.

Fig. 2.

Probability of base pair contacts, Π_ij, renormalized with respect to the highest value for each system. The ordinate axes denote the system, the values of nξ shown, and the sequence of the sense strand. The 5′ end of each strand is indicated by the nucleic base set in red. Ideal duplex base pairs fall along the diagonal (black line). The color scale (bottom) corresponds to renormalized values of Π_ij.

Fig. 3.

Identification of DNA base pairs (gray bands) with a renormalized value of Π_ij ≥ 0.8 in the TSE, as shown in Fig. 2 Left (systems are similarly arranged from top to bottom). The sense strand is the top sequence of each duplex, and the 5′ end is denoted by an oblique bar. Sites shown represent base moieties (backbone sites are not shown for clarity): A (magenta circles), C (green squares), G (orange squares), and T (blue circles).

Fig. 4.

Plots of Π_ij for all nξ, in which evaluations were performed at the committor probabilities, P_B, indicated on the ordinate axes. Shown are results for the REP30 (Left) and RAN30 (Right) systems. Each plot is labeled as in Fig. 2, with corresponding color scale. For a given system, plots are arranged such that P_B decreases from top to bottom. For P_B ≠ 0.5, the size of the TSE is of (500).

Fig. 5.

Plot of U(ξ) normalized by the inverse system thermal energy, β = (k_BT)⁻¹, and scaled by the total number of interaction sites in the system, N (systems are similarly arranged from top to bottom as in Fig. 2).