Target search on a dynamic DNA molecule

We study a protein-DNA target search model with explicit DNA dynamics applicable to in vitro experiments. We show that the DNA dynamics plays a crucial role for the effectiveness of protein “jumps” between sites distant along the DNA contour but close in 3D space. A strongly binding protein that searches by 1D sliding and jumping alone, explores the search space less redundantly when the DNA dynamics is fast on the timescale of protein jumps than in the opposite “frozen DNA” limit. We characterize the crossover between these limits using simulations and scaling theory. We also rationalize the slow exploration in the frozen limit as a subtle interplay between long jumps and long trapping times of the protein in “islands” within random DNA configurations in solution.

💡 Research Summary

In this paper the authors investigate how the dynamics of a DNA polymer affect the efficiency of a protein searching for a specific target site along the DNA. Building on the classic “facilitated diffusion” picture—where a protein alternates between three‑dimensional diffusion in solution and one‑dimensional sliding along the DNA—they focus on the role of intersegment transfer, or “jumping”, which allows a protein to move from one DNA segment to another that is spatially close but far apart along the contour.

To make the problem tractable they model the DNA as a self‑avoiding‑free chain of L = 5000 segments on a simple cubic lattice. The chain evolves via a kinetic Monte‑Carlo scheme that implements Rouse dynamics (k_D sets the rate of segment moves such as kinks, crankshafts, etc.). A protein is represented as a point particle that diffuses along the contour at rate k_p and may instantly jump to any other segment that occupies the same lattice site, also at rate k_p. The key dimensionless parameter is the ratio k = k_D/k_p. In the limit k → 0 the DNA is frozen (quenched), while k → ∞ corresponds to an annealed DNA that completely reshuffles after each jump.

Simulations reveal three distinct regimes. In the frozen limit the protein’s probability distribution along the contour spreads only diffusively, Λ(t) ∝ t^{1/2}, despite the presence of long‑range jumps; the heavy‑tailed jump length distribution does not translate into super‑diffusive motion because the same links are repeatedly used, creating strong temporal and spatial correlations. In the annealed limit, successive jumps are uncorrelated; the jump‑length distribution follows |s‑s′|^{‑3/2} for an ideal chain, and the width grows super‑diffusively, Λ(t) ∝ t^{α} with α≈1.7, until it saturates at half the chain length. For intermediate k values a crossover occurs: the crossover time τ_c grows with k as τ_c ∝ k^{1/3}. This scaling follows from comparing the equilibration length of a DNA blob, ℓ ∼ (k_D τ)^{1/2}, with the distance explored by the protein, Λ(τ) ∼ (k_p τ)^{α}. When ℓ≈Λ the DNA connectivity changes fast enough to decorrelate successive jumps, and the dynamics switches from super‑diffusive to quasi‑diffusive.

To understand why the frozen case is so slow, the authors analyze the static network of contacts (“links”) formed by DNA loops. Visualizing each link as an arc on the contour reveals clusters of densely interconnected arcs, which they term “islands”. Islands contain many internal links but are separated from each other by regions without links. Within an island the protein rapidly randomizes its position, and the probability of exiting to either side is roughly equal. However, moving from one island to the next requires sliding, which is much slower. The size distribution of islands follows p(s) ∝ s^{‑3/2}, the same exponent as the link length distribution, and the average trapping time in an island scales as τ ∝ s^{3/2}.

Mapping the problem onto a one‑dimensional walk on an “island space” with traps, the authors derive a phase diagram in terms of two exponents: μ (governing the island‑size distribution) and κ (governing the scaling of trapping time with island size). The typical time to cross n islands is T ∝ n^{max(1+κ/μ, 2)}. The total contour length covered after crossing n islands scales as S ∝ n^{1/μ} for μ<1. Combining these relations yields the transport law along the DNA. Remarkably, for the DNA model the exponents satisfy μ+κ = 2, which places the system exactly on the line where T ∝ S^{2}, i.e., quasi‑diffusive behavior. Thus, in the frozen limit the long jumps are precisely compensated by long trapping times in islands, leading to an overall diffusion‑like spread despite the presence of super‑diffusive steps.

The authors estimate that for a 43 µm DNA fragment the relaxation time is ≈30 s, while protein jumps occur on the millisecond scale. Consequently, most realistic experimental conditions lie in the crossover regime rather than the pure annealed or quenched limits. They suggest that protein jumping can efficiently reduce the redundancy of 1D sliding only when DNA dynamics is sufficiently fast compared to the jump timescale, or when many proteins search in parallel.

In conclusion, the study demonstrates that DNA dynamics critically modulates the effectiveness of intersegment transfer. Fast DNA motion decorrelates jumps and yields super‑diffusive exploration, whereas static DNA generates island‑based traps that render the search quasi‑diffusive. These findings highlight the importance of considering polymer dynamics in models of target search and provide a framework for interpreting single‑molecule experiments on protein‑DNA interactions.