Background

DNA translocases and DNA helicases are structurally related and typically contain an ATP-binding core composed of two RecA-type lobes (Singleton et al., 2007) connected to accessory and family-specific domains that modulate the activity of the motor. While DNA helicases have the ability to processively move along single-stranded DNA or RNA, thereby unwinding the two strands, dsDNA translocases cannot, as these move on dsDNA instead with varied processivities. Altogether, ss/ds-DNA tracking ATPases are clustered into superfamilies (SF1-SF6) depending on their translocation mechanism, domain organization, and oligomerization state (Brosh and Matson 2020). Translocation requires multiple steps – at least two – first binding of the protein to the nucleic acid (loading) and second, sliding along it. The role of nucleotide in these processes is very much family-specific and does not always require ATP hydrolysis (Fairman-Williams et al., 2010). Therefore, translocases can have varied processivities and can operate using an active, inchworming mechanism or as Brownian motors (Briggs and Fischer 2014). Crystal structures of translocases mostly belonging to SF1 and SF2 (Subramanya et al., 1996; Korolev et al., 1997; Velankar et al., 1999; Tomko et al., 2007; Gu and Rice 2010), and, more recently, electron microscopy studies of chromatin remodelers (Yan et al., 2019; Han et al., 2020; Wagner et al., 2020), have contributed insightful mechanistic details. However, one must bear in mind that these provide still images of isolated steps in the catalytic cycle and rarely the complete trajectory of states underlying translocation.

Unlike assays for helicase activity, assays to monitor translocation on dsDNA are more challenging for the simple fact that no easily quantifiable product is formed. Some dsDNA translocases, such as select chromatin remodelers, have been tested in “cruciform extrusion” assays quantifying changes in superhelicity, which can be correlated with translocation (Havas et al., 2000). However, the most popular assay for dsDNA translocation remains, by far, the triplex-forming oligonucleotide (TFO) displacement assay (Firman and Szczelkun 2000).

As a general principle, a DNA triplex is formed by annealing a short TFO to a dsDNA substrate with a TFO binding site (TBS), and the structure is held together via Hoogsteen base-pairing (Gowers and Fox 1999). Triplex formation requires a low pH environment. Once the triplex is formed, it is no longer pH dependent, and the assay can be performed at neutral pH (Kopel et al., 1996) if Mg²⁺ is present (Singleton and Dervan 1993). The substrate can differ in length, depending on the protein to be tested and its processivity (if known), and the TFO is either labeled with ³²P or a fluorophore. The translocase will bind preferentially or exclusively to the double-stranded portion of the substrate, move along the DNA, and eventually dislodge the weakly bound TFO, which can then be detected. There are caveats to be considered when labeling the TFO either radioactively or fluorescently. In the former case, safety precautions regarding work with, storage, and disposal of radioactive materials have to be considered. On the other hand, fluorophores can interact non-specifically with proteins. Radioactivity-based assays are end-point and discontinuous based on detection of bound and free TFO on agarose or polyacrylamide gels (Firman and Szczelkun 2000; Smith et al., 2007). In contrast, the use of a fluorescently-labeled TFO can be coupled with either a gel-based end-point or continuous monitoring assays. The continuous assay exploits the attribute of fluorophores to change the intensity of the fluorescent signal when bound by the DNA duplex. In short, the signal will be “brighter” for the free TFO and “dimmer” when the TFO is incorporated within the triplex (McClelland et al., 2005). Although using a continuous assay has the advantage of monitoring translocation at every step, it requires access to a stop-flow fluorimeter.

In our assay, we combined elements of both approaches, e.g., the use of a fluorescently-labeled TFO and the ease of quantitation based on electrophoresis through standard commercially available gels. This makes this assay accessible for the majority of biochemical labs.

Special Considerations When Designing a TFO Displacement Assay

The key requirement for designing a successful TFO displacement assay is the sequence of the TFO – a single mismatch between TFO and dsDNA substrate can be destabilizing (Puri et al., 2004). Often, the following TFO sequence is used: TTTCTTTCTTCTTTCTTTTCTT. Next, the assembly of the triplex DNA has to be performed under low pH conditions (Maher et al., 1990; Plum et al., 1990). During the displacement assay, the pH can be raised to neutral as long as Mg²⁺ is present in the reaction buffer (Singleton and Dervan 1993). Depending on the protein of interest, the length of the dsDNA substrate can differ between short (in the range of 50 bp) and long (in the kb range). If the studied protein recognizes a specific DNA motif, it can be integrated into the substrate. Protein concentration is important, and 100-500 nM is a good starting point, but high protein concentrations can interfere with translocation. The assay can be performed at a range of temperatures, but temperatures above 40°C lead to significant spontaneous dissociation of the DNA triplex, and the timeframe of the experiment has to be adjusted. In case of contaminating DNAses, an unlabeled single-stranded oligonucleotide can be added to act as a nuclease trap as long as this has negligible affinity for the DNA translocase of interest