Measuring Excess Heat Capacities of Deoxyribonucleic Acid (DNA) Folding at the Single-Molecule Level

Abstract

Measurements of the thermodynamic properties of biomolecular folding (ΔG°, ΔH°, ΔS°, etc.) provide a wealth of information on the folding process and have long played a central role in biophysical investigation. In particular, the excess heat capacity of folding (ΔC P) is crucial, as typically measured in bulk ensemble studies by differential scanning calorimetry (DSC) and isothermal titration calorimetry (ITC). Here, we report the first measurements of ΔC P at the single-molecule level using the single-molecule fluorescence resonance energy transfer (smFRET) as well as the very first measurements of the heat capacity change associated with achieving the transition state (ΔC ‡ P) for nucleic acid folding. The deoxyribonucleic acid (DNA) hairpin used in these studies exhibits an excess heat capacity for hybridization (ΔC P = −340 ± 60 J/mol/K per base pair) consistent with the range of literature expectations (ΔC P = −100 to −420 J/mol/K per base pair). Furthermore, the measured activation heat capacities (ΔC ‡ P) for such hairpin unfolding are consistent with a folding transition state containing few fully formed base pairs, in agreement with prevailing models of DNA hybridization.

Molecular Latent Space Simulators for Distributed and Multimolecular Trajectories

“…This approach can potentially reduce the volume of training data and training time required to parametrize new systems in related Multi-LSS models. The ability to reuse components of a pretrained LSS model is particularly valuable for DNA systems where simulation data can suffer from from a strong imbalance in hybridized or dissociated data when the system is pushed away from the melting temperature, and the thermodynamics and kinetics are highly sensitive to changes in temperature and ion concentration. − To evaluate temperature dependence, we repeated our MD simulation protocol at 310, 315, 325, and 330 K. We projected these trajectories into the same latent space using SRV encoders trained on WT data and retrained independent MDN propagators to learn the dynamics at each temperature. We repeated this procedure over ion concentrations 25 mM, 50 mM, 200 mM, and 400 mM while holding temperature fixed at 320 K. Given that only the propagator was retrained, we conserved 50% of training cost compared to training the full LSS pipeline.…”

“…We next test whether our Multi-LSS transfer learning procedure reproduces meaningful kinetics across temperature and ion concentration. Kinetics can be expressed in terms of k on and k off rates which are frequently monitored in single-molecule studies , and nanotechology applications such as DNA-PAINT. , We calculated rates by collecting all hybridized/dissociated dwell times and fitting a single exponential distribution to determine k on and k off (Figure ). Uncertainties were determined via a bootstrapping procedure where dwell times were randomly sampled with replacement and the model refitted.…”

Computational remodeling of an enzyme conformational landscape for altered substrate selectivity

“…The resulting two curves were integrated to measure the relative populations of each conformational state. The ratio of the two populations were used to calculate equilibrium constants at each temperature (Supplementary Table 8 ) and fit using the Solver function of Excel to the nonlinear van’t Hoff equation 55 (Equation 1), where enthalpy and entropy are not assumed to be temperature-independent (i.e. ΔC p ≠ 0) to extract thermodynamic parameters of equilibrium.…”

DNA Calorimetric Force Spectroscopy at Single Base Pair Resolution

“…Over time, improvements in calorimetric measurements 8 pointed out the significant role ∆C p played in DNA hybridization. During the last decades, bulk [9][10][11][12][13][14][15] and single-molecule [16][17][18] experiments assessed the effects of temperature, yielding ∆C p values per bp spanning two orders of magnitude depending on the experimental condition, the technique used and the DNA sequence. Single-molecule methods such as atomic force microscopy 19,20 , FRET 21 , and optical tweezers 22,23 have now reached the maturity to address such challenges.…”