Single-molecule fluorescence reveals the DNA unwinding mechanism of mitochondrial helicase TWINKLE and its interplay with single-stranded DNA-binding proteins

Abstract

The mitochondrial DNA helicase TWINKLE, a hexameric ring-shaped helicase, plays a crucial role in maintaining mitochondrial DNA integrity. TWINKLE translocates along one DNA strand, unwinding the duplex by excluding the complementary strand through coordinated ATP hydrolysis. However, the precise mechanisms underlying this process remain incompletely understood. In this study, we utilized single-molecule Förster Resonance Energy Transfer (smFRET) to investigate the mechanisms of TWINKLE-mediated DNA unwinding. Our results reveal that TWINKLE occasionally pauses during unwinding, with the rate of unwinding and the duration of pausing strongly influenced by ATP concentration, but not by the presence of DNA mismatches or mitochondrial single-stranded DNA-binding protein (mtSSB). These findings suggest that the pausing events primarily arise from stochastic ATP hydrolysis within the helicase subunits. DNA mismatches exacerbate TWINKLE's pausing and dissociation from DNA, thereby impairing DNA unwinding. In contrast, mtSSB significantly mitigates helicase dissociation by stabilizing TWINKLE-DNA interactions. This study provides novel insights into the functional dynamics of TWINKLE, highlighting the role of ATP hydrolysis in orchestrating single-stranded DNA translocation, the detrimental effects of DNA mismatches on DNA unwinding, and the critical role of mtSSB in supporting helicase function.

Graphical Abstract

Figure 1.

ATP-dependent TWINKLE-mediated dsDNA unwinding behaviors were investigated by smFRET experiments. (A) Schematic of the smFRET assay for TWINKLE. A forked dsDNA with two 15-nt ssDNA overhangs was anchored on the PEGylating slide and labeled with a pair of Cy3-Cy5 fluorophores separated by 20 bp. Upon injection of TWINKLE, dsDNA unwinding mediated by TWINKLE is indicated by an increase in FRET values. The FRET value histograms for forked dsDNA alone and after the addition of TWINKLE, both (i) in the absence of ATP and (ii) in the presence of ATP. (B) FRET value histograms for forked dsDNA in response to TWINKLE at variable ATP concentrations. (C) TWINKLE-mediated dsDNA unwinding scenarios in the presence of ATP. After the addition of TWINKLE in the presence of ATP, four scenarios are observed. (i) Continuous unwinding: continuous increase in FRET value from 0.07 ± 0.08 to 0.73 ± 0.12, indicating complete dsDNA unwinding; (ii) Stepwise unwinding: stepwise increase in FRET value time trajectory, indicating stepwise dsDNA unwinding; (iii) Stopped unwinding: a detectable pause in FRET value time trajectory until the end of acquisition, indicating that TWINKLE falls off during dsDNA unwinding process; (iv) Reinitiated unwinding: reinitiated increase in middle FRET value from 0.40 ± 0.06, indicating the reinitiation of TWINKLE-mediated dsDNA unwinding process, to 0.73 ± 0.12. The KV_Step curve was generated by applying a 10-point adjacent averaging window followed by Kalafut–Visscher step detection analysis. (D) Distribution probabilities of observed events acquired under variable ATP concentrations. Superscripts “int” indicate that these data were acquired with internal cyanide-labeled DNA substrates. (E) ATP-dependent dsDNA unwinding rate obtained in continuous dsDNA unwinding events and (F) waiting time (black square) obtained in stepwise dsDNA unwinding events and unwinding time (red square) obtained in continuous dsDNA unwinding events as a function of ATP concentration (solid square: external cyanide dye labeling; empty square: internal cyanide dye labeling). The experiments were conducted in the presence of a complete reaction buffer (40 mM HEPES, pH 8.0, 7% glycerol, 5 mM MgCl₂, 100 mM NaCl, 50 μg/mL BSA, 2 mM Trolox, 0.3 mg/mL glucose oxidase, 0.015 mg/mL catalase and 1× ATP regeneration system) with 100 nM TWINKLE (batch #1) hexamer at variable ATP concentrations ranging from 0.1 to 5.0 mM.

Figure 2.

TWINKLE concentration-dependent dsDNA unwinding behaviors investigated by smFRET experiments. (A) Schematic of the smFRET assay for TWINKLE concentration dependence. FRET value histograms for forked dsDNA in response to variable TWINKLE concentrations in the presence of 0.5 mM ATP. (B) Representative time traces for the above-mentioned four scenarios. The KV_Step curve was generated by applying a 10-point adjacent averaging window followed by Kalafut–Visscher step detection analysis. (C) Distribution probabilities of observed events at variable TWINKLE concentrations. Prebound indicates that these data were acquired with 0.5 mM ATP in the presence of 100 nM TWINKLE hexamer. (D) TWINKLE concentration-dependent dsDNA unwinding rate obtained in continuous dsDNA unwinding events and (E) waiting time (black square) obtained in stepwise dsDNA unwinding events and unwinding time (red square) obtained in continuous dsDNA unwinding events as a function of TWINKLE concentration (solid square: no wash experiment with extra unbound TWINKLE in the reaction chamber; empty square: prebound experiment without extra free TWINKLE in the reaction chamber). The experiments were conducted in the presence of a complete reaction buffer (40 mM HEPES, pH 8.0, 7% glycerol, 5 mM MgCl₂, 100 mM NaCl, 50 μg/mL BSA, 2 mM Trolox, 0.3 mg/mL glucose oxidase, 0.015 mg/mL catalase and 1× ATP regeneration system) with 0.5 mM ATP in variable TWINKLE (batch #1) hexamer concentrations ranging from 10 to 100 nM.

Figure 3.

The effect of internal mismatches on the degradation behaviors of TWINKLE. (A) Schematics of DNA constructs with consecutive (1M, 3M, and 5M) mismatches. FRET value histograms for forked dsDNA with consecutive mismatches in response to 100 nM TWINKLE in the presence of 0.5 mM ATP. (B) (i) Distribution probabilities of observed events taken from DNA constructs with various mismatch configurations outlined in panel (A) (ii) Average dsDNA unwinding rate acquired from continuous dsDNA unwinding events and (iii) waiting time (black square) obtained in stepwise dsDNA unwinding events and unwinding time (red square) obtained in continuous dsDNA unwinding events as a function of the number of consecutive mismatches with 100 nM TWINKLE in the presence of 0.5 mM ATP. (C) (i) Distribution probabilities of observed events, (ii) average dsDNA unwinding rate acquired from continuous dsDNA unwinding events, and (iii) waiting time (black square) obtained in stepwise dsDNA unwinding events and unwinding time (red square) obtained in continuous dsDNA unwinding events as a function of TWINKLE concentration on DNA substrates with five consecutive mismatches in the presence of 0.5 mM ATP with variable TWINKLE hexamer. (D) (i) Distribution probabilities of observed events, (ii) average dsDNA unwinding rate acquired from continuous dsDNA unwinding events, and (iii) waiting time (black square) obtained in stepwise dsDNA unwinding events and unwinding time (red square) obtained in continuous dsDNA unwinding events as a function of ATP concentration on DNA substrates with five consecutive mismatches in the presence of 100 nM TWINKLE with variable ATP. The experiments were conducted in the presence of a complete reaction buffer (40 mM HEPES, pH 8.0, 7% glycerol, 5 mM MgCl₂, 100 mM NaCl, 50 μg/mL BSA, 2 mM Trolox, 0.3 mg/mL glucose oxidase, 0.015 mg/mL catalase, and 1× ATP regeneration system). The TWINKLE hexamer used is batch #1.

Figure 4.

The reannealing behaviors observed during TWINKLE-mediated dsDNA unwinding process. (A) Representing reannealing time traces observed in our experimental condition. (B) Distribution probabilities of observed reannealing events as a function of (i) ATP concentration, (ii) TWINKLE concentration, or (iii) number of mismatches. The experiments were conducted in the presence of (40 mM HEPES, pH 8.0, 7% glycerol, 5 mM MgCl₂, 100 mM NaCl, 50 μg/mL BSA, 2 mM Trolox, 0.3 mg/mL glucose oxidase, 0.015 mg/mL catalase, and 1× ATP regeneration system) with indicated concentration TWINKLE (batch #1) and ATP on DNA substrate with or without mismatched segments. (C) Unwinding/reannealing rate (bp/s) obtained from complete unwinding and alternative unwinding/reannealing process.

Figure 5.

The influence of mtSSB on the TWINKLE-mediated dsDNA unwinding process. (A) Schematic representation and FRET histogram analysis of the smFRET assay using forked dsDNA substrates with either a 15-nt or 60-nt 5′-ssDNA overhang, in the absence or presence of mtSSB. (B) (i) Probability distribution of observed unwinding events, (ii) unwinding rates measured from continuous dsDNA unwinding events, and (iii) waiting time (black square) obtained in stepwise dsDNA unwinding events and unwinding time (red square) obtained in continuous dsDNA unwinding events under different experimental conditions. (C) (i) One representative reannealing time trace observed under our experimental conditions. (ii) Probability distribution of observed reannealing events across different experimental conditions. All experiments were conducted in a complete reaction buffer (40 mM HEPES, pH 8.0, 7% glycerol, 5 mM MgCl₂, 100 mM NaCl, 50 μg/mL BSA, 2 mM Trolox, 0.3 mg/mL glucose oxidase, 0.015 mg/mL catalase, and 1× ATP regeneration system) supplemented with 0.5 mM ATP and 100 nM TWINKLE (batch #1) hexamers.

Figure 6.

A kinetics model for TWINKLE-mediated dsDNA unwinding. (i) TWINKLE hexamer initially binds to the forked dsDNA in an open-ring conformation. (ii) Transition to a tightly bound closed-ring conformation occurs, followed by (iii) ATP hydrolysis near the DNA’s 5′ end, propelling (iv) movement of the TWINKLE hexamer in the 5′ to 3′ direction. (v) ATP hydrolysis at the DNA’s 3′ end or middle can induce transient unwinding pauses. (vi) The TWINKLE hexamer may dissociate from the dsDNA, causing unwinding to cease. (vii) Alternatively, TWINKLE may move backward, promoting DNA reannealing. The complete unwinding sequence follows the path (i)→(ii)→(iii)→(iv); the stepwise unwinding process follows the path (i)→(ii)→(v)→(iii)→(iv); the stopped unwinding process progresses as follows (i)→(ii)→(vi); and the reinitiated unwinding process progresses as follows (vi)→(i)→(ii)→(iii)→(iv). The transition to reannealing (vii) occurs stochastically during the aforementioned unwinding scenarios.