Ultrashort Nucleic Acid Duplexes Exhibit Long Wormlike Chain Behavior with Force-Dependent Edge Effects

Despite their importance in biology and use in nanotechnology, the elastic behavior of nucleic acids on “ultrashort” ($<15\mathrm{nt}$) length scales remains poorly understood. Here, we use optical tweezers combined with fluorescence imaging to observe directly the hybridization of oligonucleotides (7–12 nt) to a complementary strand under tension and to measure the difference in end-to-end extension between the single-stranded and duplex states. Data are consistent with long-polymer models at low forces ($<8\mathrm{pN}$) but smaller than predicted at higher forces ($>8\mathrm{pN}$), the result of the sequence-dependent duplex edge effects.

Figure 1

Measurement of extension difference between single- and double-stranded ultrashort oligonucleotides under force. (a) Schematic of the hybridization assay (not to scale). A DNA molecule containing a short, central ssDNA region containing a binding site (black) flanked by poly-dT spacers (red) and long dsDNA handles (blue) is attached to polystyrene beads (gray spheres) in optical traps (orange cones) and held at a constant force. A fluorescence excitation laser (green cone) is focused on the central ssDNA region. Short oligonucleotides (black) labeled with a Cy3 fluorophore at the 3’ end (green disk) bind and unbind to the complementary ssDNA sequence in the center of the tethered DNA. Binding and unbinding events are detected by the fluorescence emitted from the attached fluorophore and the simultaneous change in separation between the optical traps at constant force. (b) Representative time trace showing 10-nt DNA probes binding and unbinding a DNA construct held under constant force (12.4 pN). The extension difference between the single-stranded state and the double-stranded state, $\mathrm{\Delta }X$, is measured from the stepwise increase or decrease in the trap separation. (c) Histogram of recorded extension differences for binding and unbinding of the 10-nt probes using the hybridization assay.

Figure 2

Comparison of measured extension changes to long-polymer models. (a) Extension changes due to probe hybridization, scaled by probe length $\mathrm{\Delta }x$ (extension changes from both binding and unbinding events are combined for each data point; error bars denote standard error of the mean). The gray shaded region shows a force-dependent long-polymer model $\mathrm{\Delta }{x}_{\mathrm{model}}=x_{\text{ds}}(F)-x_{\text{ss}}(F)$ using the XWLC model for $x_{\text{ds}}(F)$ and the SLC model for $x_{\text{ss}}(F)$ (see Ref. [37]). Inset: The four oligonucleotide probes used in this study (bold), bound to their complementary sequences on the tethered DNA (not bold). GC pairs are highlighted. Each oligonucleotide has a Cy3 fluorophore conjugated to its 3’ end (green disks). (b) Extension changes due to hybridization of 9-nt probes to complementary sequences with varying numbers of spacers ($N_{\text{sp}}=2$, 1, 0), scaled by probe length. Inset: the 9-nt probe (bold) bound to the three DNA constructs used in these experiments (not bold). GC pairs are highlighted. The probes used for binding the 2- and 1-spacer constructs have a Cy3 fluorophore (green disks) conjugated to their 3’ ends, while the 9-nt probe used for binding the 0-spacer construct has a Cy3 fluorophore conjugated to an internal dT base to avoid steric clashes with the handles. (c) Extension changes due to hybridization of 10-nt probes of differing sequences, scaled by probe length. Inset: The three 10-nt probes used (bold), bound to their complementary sequences on the tethered DNA (not bold). GC pairs are highlighted. Each oligonucleotide has a Cy3 fluorophore conjugated to its 3’ end (green disks).

Figure 3

Deviation of measured extension changes from long-polymer model. (a) Schematic of the DNA constructs used in this study and modeling of extension changes. Top ($X_u$): DNA construct with no probe bound. The dotted lines indicate a variable number of spacers ($N_{\text{sp}}=2$, 1, 0) depending on the construct used. Bottom three ($X_b$): DNA constructs for varying number of spacers ($N_{\text{sp}}=2$, 1, 0) with probe bound. Blue: dsDNA handles. Green: Duplex edge regions. Red: ssDNA poly-dT spacers. Black: Probe binding site and probe duplex region. (b) Residuals from extension change data for different probe lengths and long-polymer model, $\mathrm{\Delta }X-\mathrm{\Delta }{X}_{\mathrm{model}}$, determined using optimal model parameters (see Ref. [37]). Magenta line: Fraying model of the duplex. Shaded area represents model over range of base-pairing energies for the different probes in Fig. 2; line represents model for average base-pairing energy. Black dotted line: Model for base fraying with additional force-dependent energy term. Shaded area represents model over range of base-pairing energies for the different probes in Fig. 2. (c) Residuals from extension change data of 9-nt probe binding to constructs with varying number of spacers ($N_{\text{sp}}=2$, 1, 0). (d) Additional force-dependent energy required to destabilize edge base pairs. Black dotted line: Phenomenological model for force dependence.