Mechanical properties of DNA-like polymers

Abstract

The molecular structure of the DNA double helix has been known for 60 years, but we remain surprisingly ignorant of the balance of forces that determine its mechanical properties. The DNA double helix is among the stiffest of all biopolymers, but neither theory nor experiment has provided a coherent understanding of the relative roles of attractive base stacking forces and repulsive electrostatic forces creating this stiffness. To gain insight, we have created a family of double-helical DNA-like polymers where one of the four normal bases is replaced with various cationic, anionic or neutral analogs. We apply DNA ligase-catalyzed cyclization kinetics experiments to measure the bending and twisting flexibilities of these polymers under low salt conditions. Interestingly, we show that these modifications alter DNA bending stiffness by only 20%, but have much stronger (5-fold) effects on twist flexibility. We suggest that rather than modifying DNA stiffness through a mechanism easily interpretable as electrostatic, the more dominant effect of neutral and charged base modifications is their ability to drive transitions to helical conformations different from canonical B-form DNA.

Figure 1.

Experimental design. (A). Three-dimensional structure of B-form DNA (26). The central stacked DNA base pairs and deoxyribose sugars are uncharged (cyan), while each phosphodiester linkage carries a negative charge (gray). Seven of the eight tested base modifications replace the methyl group (red) of thymine bases in the DNA major groove, while one modification replaces the N₂ proton (blue) of adenine bases in the DNA minor groove. (B). Chemical structure of an A·T base pair. Base modifications occur either at the 5 position of thymine (compare 1 with 2-8) or the 2 position of adenine (9). Based on the pK_a values of the isolated functional groups, modifications 4, 6, and 8 alter DNA charge at neutral pH. (C). DNA ligase-catalyzed cyclization kinetics experiments to analyze DNA bend and twist stiffness. End-labeled (black circle) linear DNA fragments (∼200 bp, top) are detected when they either cyclize (right) or multimerize (bottom). The readout of these experiments is a ring closure probability (J-factor), which can be interpreted using the WLC model to estimate DNA mechanical parameters.

Figure 2.

Characterization of DNA analogs. (A) PCR assays analyzed by 5% native polyacrylamide gel electrophoresis. Total PCR volume 100 µl: 20 ng 418-bp DNA template (pJ1506), 0.4 mM each LJM-3222 (5'-GTACGCAGT) and LJM-3223 (5'-TGTGAGTAGCTCACTCATAG), 0.2 mM each dNTP with indicated analog triphosphate (1–9) completely replacing appropriate dNTP, and 5 U DNA polymerase (indicated with plus symbol) with associated buffer and cycle conditions. Taq DNA polymerase (Taq) conditions: Taq DNA polymerase buffer with 100 mg/ml BSA and 2 mM MgCl; 98°C (3 min), 30 cycles of [94°C (30 s), 60°C (30 s), and 72°C (45 s)], 72°C (5 min). PrimeSTAR HS DNA polymerase (PS) conditions: PrimeSTAR GC buffer with 2 M betaine; 98°C (3 min), 30 cycles of [98°C (15 s), 60°C (5 s), and 72°C (45 s)], 72°C (5 min). Pwo SuperYield DNA Polymerase (Pwo) conditions: Pwo PCR buffer with GC-rich solution and 2 M betaine; 98°C (3 min), 30 cycles of [98°C (1 min), 60°C (2 min), and 72°C (8 min)], 72°C (5 min). Lane 1 is marker (M) DNA (100 bp DNA ladder, Invitrogen) with 400 - and 500-bp bands indicated. (B) Anion exchange chromatography of 98-bp DNA-like polymers (pJ1923). Following equilibration in 20 mM Tris–HCl, pH 8 (buffer A), samples were eluted over 25 min at a 1 ml/min flow rate in a linear gradient from 50 to 100% buffer B (buffer A plus 1 M NaCl). Eluent absorbance at 260 nm (milli-absorbance units) was monitored with elution time (min).

Figure 3.

CD spectroscopy. CD spectra of 417-bp DNA-like polymers (pJ1741) in 10 mM phosphate buffer, pH 7.0, containing 1 M NaCl showing ellipticity (Θ) as a function of wavelength (λ) were divided into four groups based on HDR [HDR = Θ(λ₂₉₀)/Θ(λ₂₀₁)] in the ranges 0 < HDR < 0.5 (A), −0.5 < HDR < 0 (B), −1.5 < HDR < −0.5 (C) and 0.5 < HDR < 1.5 (D).

Figure 4.

Example measurement of mechanical properties. (A) Cyclization time course for 207-bp DNA-like polymer 5 (pJ1744). DNA ligase-catalyzed cyclization reaction was performed at ∼22°C with 1 nM DNA restriction fragment, T4 DNA ligation buffer (50 mM Tris–HCl, pH 7.5, 10 mM MgCl₂, 1 mM ATP, 10 mM dithiothreitol) and a final concentration of 100 U/ml T4 DNA ligase. Aliquots (10 µl) were removed at 1–15 min time points, quenched by addition of EDTA to 20 mM and then analyzed by electrophoresis through 5% native polyacrylamide gels in 0.5× TBE buffer (50 mM Tris base, 55 mM boric acid and 1 mM EDTA, pH 8.3). Gel lanes contains Invitrogen 100 bp DNA ladder (M), linear monomer without ligase (0) and increasing 1-min time points of the ligation reaction (1–15) showing the evolution of linear monomer (M), linear dimer (D), circular monomer (C_M) and circular dimer (C_D). Nearest molecular weight bands are indicated. (B) Cyclization kinetics analysis for 207-bp DNA-like polymer 5 (pJ1744). Fitting of data in (A) determines the J-factor, as previously described (30) (see also Supplementary Data S3). (C) WLC analysis for DNA-like polymer 5. Fit of experimental J-factor data using the WLC model. (D). Monte Carlo estimation of uncertainty. Fit of simulated J-factor data based on (C) using the WLC model.

Figure 5.

Summary of WLC analysis. J-factor curves for natural and DNA-like polymers (A) and parameter distributions for P (B), P_t (related to by C = kTP_t) (C) and γ₀ (D).

Figure 6.

Relationship between experimentally determined parameters and HDR. T_m (A), P (B), γ₀ (C) and C (D) are plotted as a function of HDR. See Figure 3 legend for details.