Kinetic partitioning modulates human telomere DNA G-quadruplex structural polymorphism

Abstract

Telomeres are specialized chromatin structures found at the end of chromosomes and are crucial to the maintenance of eukaryotic genome stability. Human telomere DNA is comprised of the repeating sequence (T2AG3)n, which is predominantly double-stranded but terminates with a 3' single-stranded tail. The guanine-rich tail can fold into secondary structures known as a G-quadruplexes (GQs) that may exist as a polymorphic mixture of anti-parallel, parallel, and several hybrid topological isomers. Using single-molecule Förster resonance energy transfer (smFRET), we have reconstructed distributions of telomere DNA GQ conformations generated by an in situ refolding protocol commonly employed in single-molecule studies of GQ structure, or using a slow cooling DNA annealing protocol typically used in the preparation of GQ samples for ensemble biophysical analyses. We find the choice of GQ folding protocol has a marked impact on the observed distributions of DNA conformations under otherwise identical buffer conditions. A detailed analysis of the kinetics of GQ folding over timescales ranging from minutes to hours revealed the distribution of GQ structures generated by in situ refolding gradually equilibrates to resemble the distribution generated by the slow cooling DNA annealing protocol. Interestingly, conditions of low ionic strength, which promote transient GQ unfolding, permit the fraction of folded DNA molecules to partition into a distribution that more closely approximates the thermodynamic folding equilibrium. Our results are consistent with a model in which kinetic partitioning occurs during in situ folding at room temperature in the presence of K(+) ions, producing a long-lived non-equilibrium distribution of GQ structures in which the parallel conformation predominates on the timescale of minutes. These results suggest that telomere DNA GQ folding kinetics, and not just thermodynamic stability, likely contributes to the physiological ensemble GQ structures.

Figure 1. Experimental setup for telomere DNA GQ single molecule FRET measurements.

(A) Planar telomere DNA G-quartet structure with coordinated monovalent cation. (B) Grey rectangles represent planar G-quartets and backbone polarity (5’à3’) is indicated by black arrowheads. Molecules are immobilized on a microscope slide and FRET is measured as the energy transfer between the donor (Cy3) and acceptor (Cy5) dye. (C) Human telomere G-quadruplex structures formed by Tel23 sequence.

Figure 2. Single-molecule FRET histograms for telomere DNA GQ constructs.

Gaussian fits to the data are shown in black. (A)Tel23 thermally annealed in 100 mM KCl. (B) Tel23 in situ refolded in 100 mM KCl.

Figure 3. Single-molecule FRET histograms of Tel23 in situ refolded in 100 mM KCl for the indicated period of time.

Figure 4. Single-molecule FRET histograms of Tel23 in situ refolded at the indicated KCl concentrations.

Figure 5. Structural assignments of distinct FRET states.

(A) CD spectra of Tel23 thermally annealed in 100 mM KCl (open circles), Tel23 in 100 mM NaCl (asterisks), and Hybrid Mutant in 100 mM KCl (closed circles). (B) The smFRET distribution of Tel23 thermally annealed in 100 mM NaCl fit with Gaussian functions. Gaussian fits to the data are shown in black. (C) The smFRET distribution of Hybrid Mutant thermally annealed in 100 mM KCl fit with Gaussian functions. Gaussian fits to the data are shown in black.

Figure 6. A qualitative energy landscape model for kinetic partitioning during telomere DNA GQ folding.

A lower energy barrier between the low-FRET (unfolded) and the high-FRET (folded) state creates a kinetic trap during the early stages of folding. To escape the kinetic trap, the molecule must unfold and then re-fold into one of the more energetically stable mid-FRET (folded) states. This process is facilitated during thermal annealing at higher temperatures, or during in situ folding by low ionic strength or prolonged incubation times.