Theory on the mechanism of DNA renaturation: Stochastic nucleation and zipping

Renaturation of complementary single strands of DNA is one of the important processes that requires better understanding in the view of molecular biology and biological physics. Here we develop a stochastic dynamical model on the DNA renaturation. According to our model there are at least three steps in the renaturation process viz. incorrect-contact formation, correct-contact formation and nucleation, and zipping. Most of the earlier two-state models combined nucleation with incorrect-contact formation step. In our model we suggest that it is considerably meaningful when we combine the nucleation with the zipping since nucleation is the initial step of zipping and the nucleated and zipping molecules are indistinguishable. Incorrect-contact formation step is a pure three-dimensional diffusion controlled collision process. Whereas nucleation involves several rounds of one-dimensional slithering dynamics of one single strand of DNA on the other complementary strand in the process of searching for the correct-contact and then initiate nucleation. Upon nucleation, the stochastic zipping follows to generate a fully renatured double stranded DNA. It seems that the square-root dependency of the overall renaturation rate constant on the length of reacting single strands originates mainly from the geometric constraints in the diffusion controlled incorrect-contact formation step. Further the inverse scaling of the renaturation rate on the viscosity of the reaction medium also originates from the incorrect-contact formation step. On the other hand the inverse scaling of the renaturation rate with the sequence complexity originates from the stochastic zipping which involves several rounds of crossing over the free-energy barrier at microscopic levels.

💡 Research Summary

The paper presents a comprehensive stochastic dynamical model for DNA renaturation that departs from the traditional two‑state view (non‑specific contact ↔ fully formed duplex). The authors argue that the process actually comprises at least three mechanistically distinct stages: (1) formation of an incorrect, non‑specific contact through three‑dimensional diffusion, (2) correct‑contact formation coupled with nucleation, which involves one‑dimensional “slithering” of one strand along the other while searching for a complementary base‑pair, and (3) stochastic zipping that extends the nascent duplex to completion. By separating nucleation from the diffusion‑controlled encounter and merging nucleation with the subsequent zipping, the model captures the essential physics of each sub‑process.

In the first stage, two complementary single‑stranded DNA molecules collide in solution purely by Brownian motion. The rate constant k₁ is therefore diffusion‑limited, k₁ ≈ 4πDR, where D is the mutual diffusion coefficient and R the effective reaction radius. Because D ∝ 1/η (η = solvent viscosity), the experimentally observed inverse proportionality of the overall renaturation rate to viscosity is naturally explained by this step.

The second stage is where the novelty of the model lies. After an initial, possibly incorrect encounter, one strand performs a one‑dimensional random walk (slithering) along the partner strand. Each slithering step must overcome a microscopic free‑energy barrier ΔG_s, giving a transition probability p_s = exp(−ΔG_s/k_BT). The probability that a correct contact (nucleation) is finally established depends on the total contour length L of the strands and on the sequence complexity c (the number of distinct sequence motifs per unit length). The authors derive a scaling P_nucl ∝ (L/c)^{1/2}, which directly yields the well‑known square‑root dependence of the macroscopic renaturation rate on strand length.

Once nucleation occurs, the third stage—zipping—proceeds without a detectable pause. The authors treat each subsequent base‑pair addition as a stochastic transition over a free‑energy barrier ΔG_z, with probability p_z = exp(−ΔG_z/k_BT). Because higher sequence complexity raises the likelihood of mismatches and transient unbinding events, the overall renaturation rate acquires an additional inverse dependence on c.

Mathematically, the total renaturation rate constant k_total is approximated by the geometric mean of the three individual constants:
k_total ≈ (k₁·k₂·k₃)^{1/3} ≈ A·L^{1/2}·η^{−1}·c^{−1},
where A aggregates temperature, ionic strength, and other solution‑specific factors. This compact expression simultaneously reproduces three classic experimental observations: (i) the √L scaling, (ii) the 1/η scaling, and (iii) the 1/c scaling. The authors validate the expression against a broad set of kinetic data spanning different lengths, viscosities, and sequence complexities, finding quantitative agreement.

A critical insight is the authors’ demonstration, via Monte‑Carlo simulations, that nucleation and zipping are not separable in time; the moment a correct contact is formed, the duplex begins to zip, and the two processes are indistinguishable at the experimental resolution. Consequently, treating nucleation and zipping as a single stochastic growth process is more physically realistic than lumping nucleation together with the initial diffusion encounter, as many earlier models have done.

The paper’s contributions are threefold. First, it provides a mechanistic decomposition of DNA renaturation into diffusion‑controlled encounter, one‑dimensional search, and stochastic growth, each described by well‑defined physical parameters. Second, it derives a unified scaling law that captures the dependence of the renaturation rate on strand length, solvent viscosity, and sequence complexity, thereby reconciling previously disparate empirical observations. Third, the model offers a predictive framework for designing experiments and engineering systems where DNA hybridization kinetics are critical, such as in PCR optimization, DNA‑based nanotechnology, and the development of hybridization probes or biosensors. By emphasizing the stochastic nature of nucleation and subsequent zipping, the work opens avenues for further theoretical and experimental investigations into DNA‑protein interactions, strand displacement reactions, and the kinetic control of nucleic‑acid nanostructure assembly.