Background

Human mitochondria contain circular double-stranded DNA (16,569 bp) replicated by DNA polymerase Pol γ in concert with Twinkle helicase and mitochondrial single-stranded DNA binding protein (mtSSB) (Korhonen et al., 2004). Mutations in the genes encoding the core mitochondrial DNA (mtDNA) replication machinery have been associated with human diseases, including progressive external ophthalmoplegia, ataxia neuropathy syndromes, and other neurodegenerative diseases collectively known as “mitochondrial diseases” (Tyynismaa et al., 2005; Goffart et al., 2009; Ylikallio and Suomalainen, 2012; Bhatti et al., 2017; Suomalainen and Battersby 2018). The relationship between dysfunction in mtDNA replication and mitochondrial diseases makes it crucial to understand the molecular mechanism governing mtDNA replication. Although core mitochondrial proteins at the DNA replication fork have been identified (Bogenhagen et al., 2008), the structure and dynamics of mtDNA replication remain elusive. Most of the previous mitochondrial replication studies have been based on ensemble averaging methods using short DNA as template. The transient nature of the protein-DNA interactions presents significant challenges towards elucidating the underlying mechanisms using traditional methods. Using traditional methods, it has been shown that Twinkle is capable of unwinding dsDNA up to 20–55 bps both with and without mtSSB (Korhonen et al., 2004; Sen et al., 2012). To elucidate the mechanism of Twinkle helicase binding and unwinding DNA, we sought to uncover the intricate transient states, which may be obscured by ensemble averaging effects. To fill this knowledge gap, we developed protocols using Atomic Force Microscopy (AFM) imaging in air as well as in liquids that reveal a multi-faceted, dynamic structural description of DNA binding and unwinding by Twinkle-mtSSB at the single-molecule level. Studying unwinding reactions using AFM unveils unprecedented dynamic as well structural information of the unwinding reaction, helping us to provide a functional step model of Twinkle unwinding of the DNA. These protocols elucidated the real-time dynamics of Twinkle helicase loading onto DNA in an open ring configuration, switching to a closed ring state for DNA unwinding, and unloading from DNA as an open ring structure. These changes in the Twinkle helicase conformation dynamics are stochastic in nature and would have been lost in ensemble averaging traditional methods. Results from this study using this protocol showed that Twinkle helicase exists in various oligomeric states. Furthermore, time-lapse AFM imaging in liquids showed that Twinkle hexamers are formed through sequential recruitment of individual subunits, and these hexamers can exist and switch between both open and closed ring conformations. These imaging results support a model in which Twinkle helicase loads onto DNA in open-ring conformation and searches along DNA in a closed ring conformation. Also, the Twinkle helicase unwinds the DNA in open ring conformation and unloads from DNA upon reopening of the ring after performing DNA unwinding. Our AFM studies with a model circular dsDNA substrate containing 37-nt ssDNA gaps revealed that Twinkle is capable of unwinding thousands of base pairs when facilitated by mtSSB (~240 bp/min without mtSSB to ~1,265 bp/min with mtSSB) (Kaur et al., 2020). Here, we outline detailed protocols that can be adapted to study other DNA replication proteins in vitro using AFM imaging in air and time-lapse buffer imaging. While air imaging elucidates the spatial interaction, time-lapse buffer imaging provides dynamic information on protein-DNA interactions.