Displacement and dissociation of oligonucleotides during DNA hairpin closure under strain

Abstract

The hybridization kinetic of an oligonucleotide to its template is a fundamental step in many biological processes such as replication arrest, CRISPR recognition, DNA sequencing, DNA origami, etc. Although single kinetic descriptions exist for special cases of this problem, there are no simple general prediction schemes. In this work, we have measured experimentally, with no fluorescent labelling, the displacement of an oligonucleotide from its substrate in two situations: one corresponding to oligonucleotide binding/unbinding on ssDNA and one in which the oligonucleotide is displaced by the refolding of a dsDNA fork. In this second situation, the fork is expelling the oligonucleotide thus significantly reducing its residence time. To account for our data in these two situations, we have constructed a mathematical model, based on the known nearest neighbour dinucleotide free energies, and provided a good estimate of the residence times of different oligonucleotides (DNA, RNA, LNA) of various lengths in different experimental conditions (force, temperature, buffer conditions, presence of mismatches, etc.). This study provides a foundation for the dynamics of oligonucleotide displacement, a process of importance in numerous biological and bioengineering contexts.

Figure 1.

Detection of oligonucleotide-induced blockages during re-hybridization. (A) Schematic representation of the experimental set-up (6). (B) Schemes of the experimental process. Five different extension levels are expected: (i) Z_open the fully unzipped hairpin at a force F_open large enough (>15 pN) to open it, (ii) Z_unzip, the transient extension of the unzipped hairpin at a force F_test too small (<12 pN) to prevent its re-hybridization, (iii) the partially rezipped hairpin blocked by an oligonucleotide at Z_block, the transition state, and (iv) the extension Z_zip of a fully folded hairpin. (C) Left panel: Experimental traces of an 83 bp hairpin recorded at F_open = 17.8 pN (orange) and F_test = 11.4 pN (blue; see force trace at the top), marked as in (B). The black curve corresponds to a 1 s average of the raw data. Right panel: Histogram of blockages. The black curve represents the histogram of the number of blockages per cycle at a given extension of the hairpin upon re-hybridization at F_test: ΔZ = Z_block – Z_zip in nm on the left scale and base pairs on the right scale, obtained from a single hairpin. Gaussian fits to the data are shown in red. The variance of these fits (σ ∼ 1 nm) defines the resolution of the apparatus. The roadblocks Z_block is observed at the expected positions (green dashed line). Notice that reducing the force from F_open to F_test results in a slight change in the extension of the ssDNA, due to its elastic properties (Figure 1C and Supplementary Figure S5). (D) The scheme of energy landscape. The labelled (i)–(iv) status are corresponding to the ones shown in (B).

Figure 2.

Influence of oligonucleotide length and temperature on the displacement time <τ_disp>. (A) The displacement time versus oligonucleotide length in the fork blocking assay. (Insert) The histogram of the blocking time τ_disp displays a single exponential distribution with characteristic displacement time <τ_disp>. τ_disp are obtained from ∼200-fold/unfold hairpin cycles on the same molecule. The line is the prediction from a model of strand displacement (see Eq. 1). Points show the experimental evolution of <τ_disp> versus F_test for oligonucleotides of length N (9 ≤ N ≤ 12). In a simple picture, <τ_disp> varies exponentially with F_test. As a result, for a given hairpin, <τ_disp> can only be measured in a narrow force range. Adapting F_test provides a way to study the hybridization over a large range of lengths, here N varies from 9 to 95 nt (Supplementary Figure S2). The coloured continuous lines correspond to the predictions of the model described in the text. (B) Displacement time for the 10 nt oligonucleotide at a force F_test = 8.5 pN as a function of the temperature. Decreasing the temperature increases the oligonucleotide stability as expected from the model (continuous line).

Figure 3.

Influence of nucleotide types on the displacement time <τ_disp>. (A, B) Increased binding stability induced by changing nucleotides from DNA to (A) RNA, (B) LNA. The continuous lines are predictions based on the model described in Methods: (A) for 11 nt RNA, the model fits the data by considering an increase of stability of ΔΔG = 1.1 kcal/mol for each the 11 nt DNA–RNA oligonucleotide, and the prediction made using dinucleotide energy from overestimating the experimental results. (B) For 10 nt oligonucleotides with 1 or 3 LNA bases, the model fits the data with the following sequence dependent LNA–DNA increases in stability ΔΔG(AC) = –0.5 (kcal/mol), ΔΔG(CA) = –0.1 (kcal/mol), ΔΔG(AG) = –0.45 (kcal/mol). The underlined base in the motif is a LNA (all full lines have been divided by 1.5 so that the pure DNA case fits properly). (C) Evolution of <τ_disp> versus mismatch position in an oligonucleotide having 12 nt. The blue points correspond to the original oligonucleotide without mismatches, and the other colours correspond to an oligonucleotide with a mismatch at the underlined position. The model fits the data with a pairing parameter due to mismatches caused by the substitution of an A with a T in the oligonucleotide ΔG(TC/TG) = 0.304 (kcal/mol), ΔG(TG/TC) = –0.289 (kcal/mol) and ΔG(CT/GT) = –0.289 (kcal/mol). In agreement with the model, the displacement time of the oligonucleotides ACAGCGTCCCGA and ACACCCTGCCGA with two mismatches are too short to be detected.

Figure 4.

The oligonucleotide blocking time τ_off in the loop blocking assay versus F_test. (A, B) Schemes of the experimental process (A) and the energy landscape (B): (i) the hairpin is open, allowing the oligonucleotide hybridization at the apex of the hairpin. When this hybridization occurs, it transiently prevents the hairpin refolding (ii) for a time τ_off. Two mechanisms are possible for the oligonucleotide release: at high force (iii) spontaneous detachment; at low force (iii) the hairpin refolds encircling the oligonucleotide, i.e. the two spontaneous dissociation pathways: ‘oligo’ and ‘apex’ mode. The final state (iv) corresponds to the hairpin refolded. (C) <τ_off> versus F_test for an 8 nt (blue) and 9 nt (red) oligonucleotide blocking in the hairpin loop in Passivation Buffer (T = 25°C with [NaCl] = 150 mM). (D) <τ_off> versus F_test for the 9 nt oligonucleotide blocking in the hairpin loop in various salt concentrations. For both (C) and (D), full lines are predictions using the model in the main text without fork pressure. The energy barrier determining <τ_off> is dominated by the pairing energies and depends on the force through the change of extension between ssDNA and dsDNA.

Figure 5.

The transition elements and how to escape from a closing-blocked condition (i). Forward step: (ii) opening of the terminal base pair of the oligonucleotide followed by (iii) closing of the terminal base pair of the hairpin fork. Backward step: (iv) opening of the hairpin fork followed by (v) closing of the oligonucleotide. Note that the opening of a base pair of the oligonucleotide from (ii) and of the hairpin from (iv) are also possible, but the direct closing of the oligonucleotide from (iii) or of the hairpin fork from (v) is not possible. g_ss(F_test), and g_ds(F_test) are respectively the single-strand and double-strand elastic energies at the pulling force F_test. g_o(n,n + 1) are the zero-force base pairing energies which depend on the sequence (red or green bases) and include base-stacking effects. The opening and closing rates in the transition matrix are proportional to the exponent of the negative energy costs associated with the transitions indicated by the arrows.

Figure 6.

Comparison of oligonucleotide binding kinetics with experimental measurements done by various groups. Dupuis et al.(32), Cisse et al.(13), Whitley et al.(33) and Rieu et al.(42). All these measurements (excepted Rieu) are made using fluorescence which is limited by photobleaching. As shown in the figure, our model has good agreement with all the experimental data across four groups, especially the duration of τ_off.