Abstract
DNA damage bypass pathways promote the replication of damaged DNA when replication forks stall at sites of DNA damage. Template switching is a DNA damage bypass pathway in which fork-reversal helicases convert stalled replication forks into four-way DNA junctions called chicken foot intermediates, which are subsequently extended by replicative DNA polymerases. In yeast, fork-reversal is carried out by the Rad5 helicase using an unknown mechanism. To better understand the mechanism of Rad5 and its specificity for different fork DNA substrates, we used a FRET-based assay to observe fork reversal in real time. We examined the ability of Rad5 to bind and catalyze the reversal of various fork DNA substrates in the presence of short gaps in the leading or lagging strand as well as in the presence or absence of RPA and RNA primers in the lagging strand. We found that Rad5 preferentially reverses fork DNA substrates with short gaps (10 to 30 nt.) in the leading strand. Thus, Rad5 preferentially reverses fork DNA substrates that form chicken foot intermediates with 5’ overhangs that can be extended by replicative DNA polymerases during the subsequent steps of template switching.
Keywords: DNA repair, DNA replication, template switching, translesion synthesis, replication fork reversal
Introduction
The active sites of replicative DNA polymerases, such as DNA polymerase (pol) δ and pol ε, are unable to accommodate the distorted structures of DNA lesions in the template strand. Thus, template DNA damage blocks replicative DNA polymerases causing replication forks to stall, which can ultimately lead to genome instability and cell death [1–4]. Eukaryotes possess several DNA damage bypass pathways to overcome these replication blocks and reduce the likelihood of genome instability and cell death [5–8].
The two principal DNA damage bypass pathways in eukaryotes are translesion synthesis (TLS) and template switching. TLS requires specialized DNA polymerases, such as pol η and pol ι, to incorporate nucleotides opposite template DNA lesions [9–16]. Template switching requires fork-reversal helicases, such as Rad5 or HLTF, that convert the stalled replication fork into a chicken foot intermediate [17–25].
In yeast, Rad5 plays two important roles in template switching. First, it functions as an E3 ubiquitin ligase that facilitates the K63-linked polyubiquitylation of proliferating cell nuclear antigen (PCNA), the essential replication accessory factor that functions as a DNA sliding clamp during DNA replication, repair, and recombination [26–34]. The polyubiquitylation of PCNA initiates the template switching pathway through an unknown mechanism [26,29,33,34]. Second, Rad5 functions as a fork-reversal helicase that hydrolyses ATP and couples the energy of ATP hydrolysis with the reversal of stalled replication forks to form four-way chicken foot intermediates [18–20,35,36]. These chicken foot intermediates allow replicative polymerases to extend the stalled primer strands using the newly synthesized sister strands as templates. The mechanism of fork reversal by fork-reversal helicases remains poorly understood.
To better elucidate the role of Rad5 in fork reversal and its specificity for fork DNA substrates with different structures, we used an ensemble, Förster resonance energy transfer (FRET)-based assay to observe fork reversal in real time. We have examined the ability of Rad5 to bind and catalyze the reversal of various fork DNA substrates. We also examined the influence of RPA and RNA primers on the ability of Rad5 to reverse fork DNA substrates. We found that Rad5 preferentially reverses fork DNA substrates with short gaps (10 to 30 nucleotides) on the primer strand of the leading arm compared to substrates with short gaps on the lagging arm or substrates with no gaps. Thus, Rad5 preferentially reverses fork DNA substrates that form chicken foot intermediates with 5’ overhangs that can be directly extended by replicative DNA polymerases during template switching compared to substrates that form chicken foot intermediates with 3’ overhangs that cannot be extended.
Materials and Methods
Protein expression and purification.
S. cerevisiae Rad5 was codon optimized for bacterial expression with an N-terminal 6xHis tag and a C-terminal Twin-Strep tag as described previously [37]. Rad5 was overexpressed in BL21 Star (DE3) cells (ThermoFisher) and purified using a Strep-Tactin column (IBA Lifesciences) followed by a HiLoad Superdex 200 size-exclusion column (GE Healthcare) as described previously [37].
DNA substrates.
Oligodeoxynucleotides were purchased from Integrated DNA Technologies (see supplemental table 1). Fork DNA substrates were annealed in a two-step process. First, the parental and daughter strands of the leading arms were annealed by heating to 90°C for two minutes and slowly cooling to 65°C over 30 minutes. At the same time, the parental and daughter strands of the lagging arm were annealed in the same manner. Second, the annealed leading and lagging arms were then mixed at 65°C and annealed by slowly cooling to 20° over an hour. All annealing reactions were done in 10 mM TrisCl, pH 7 and 150 mM NaCl. The annealed DNA substrates were purified on a non-denaturing 7% polyacrylamide gel followed by electroelution.
Helicase activity assays.
All experiments were carried out at 25°C in a multi-channel Cary Eclipse Fluorescence Spectrophotometer (Agilent Technologies). The parental strand of the lagging arm was 5′ end-labeled with Cy5, and the parental strand on the leading arm was 3′ end-labeled with Cy3. Unless otherwise noted, helicase activity assays were carried out using 5 nM fluorescently labeled fork DNA substrate, 5 nM Rad5, and 2 mM ATP in 20 mM TrisCl, pH 7, 2 mM MgCl2, 0.1 mg/mL BSA, 1 mM DTT, and 10% glycerol. Unless otherwise noted, the fluorescently-labeled fork DNA substrate and the ATP were pre-incubated in the cuvette for 2 minutes. The reactions were initiated by the addition of Rad5, and the Cy3 and Cy5 fluorescence signal was monitored for ten to fifteen minutes. In experiments containing RPA, 5 nM RPA was added to the pre-incubated fork DNA substrate and ATP for five minutes prior to initiation with Rad5. The calculated FRET values were converted to nM product formed, and the rates of the reactions (nM of product per minute) were determined by linear regression using the initial linear region of each graph. All kinetics experiments were carried out at least three times with means and standard deviations reported (see Supplemental Figure S2 and Supplemental Table 2).
Gel-based DNA-binding assays.
Fork DNA substrates (20 nM) were incubated with various concentrations of either Rad5 (0 to 200 nM) or RPA (0 to 200 nM) for 10 minutes at 4°C in 20 mM TrisCl, pH 7, 2 mM MgCl2, 0.1 mg/mL BSA, 1 mM DTT, and 10% glycerol. Sample of each binding reaction (10 μL) were run on a non-denaturing 4 to 20% gradient polyacrylamide gel. Gels were imaged using a ChemiDoc MP Imaging System (BioRad) to visualize Cy3 and Cy5 fluorescence.
Mass Photometry.
Mass photometry experiments were performed using a Reyfen Two Mass Photometer (Refeyn Ltd, Oxford, UK). Microscope coverslips (High Precision Deckgläser No.1.5H, 24 × 50 mm rectangle, 170±5 μm thickness) were cleaned by serial rinsing with Milli-Q water and HPLC-grade isopropanol (Sigma Aldrich) followed by drying with a filtered air stream. Silicon gaskets (Grace Bio-Labs) to hold the sample drops were cleaned in the same procedure immediately prior to measurements. All microscope coverslips were poly-L-lysine treated immediately prior to measurements. All MP measurements were performed at room temperature using Dulbecco’s phosphate-buffered saline (DPBS) without calcium and magnesium (Thermo Fisher). The instrument was calibrated using a protein standard mixture: β-amylase (Sigma-Aldrich, 56, 112 and 224 kDa), and thyroglobulin (Sigma-Aldrich, 670 kDa). Before each measurement, 15 μL of DPBS buffer was placed in the well to find focus. The focus position was searched and locked using the default droplet-dilution autofocus function after which 5 μL of protein and DNA (12.5 nM final concentration of each) was added and pipetted up and down to briefly mix before movie acquisition was promptly started. Movies were acquired for 60 s (3000 frames) using AcquireMP (version 2.3.0; Refeyn Ltd) using standard settings. All movies were processed, analyzed using DiscoverMP (version 2.3.0; Refeyn Ltd).
Results
The FRET-based fork reversal assay.
To better understand the mechanism of Rad5-catalyzed fork reversal as well as the specificity of Rad5 for reversing various fork DNA substrates with different structures, we used an ensemble, FRET-based assay to observe fork reversal in real time. We placed Cy3 and Cy5 fluorescent probes at the 3′ and 5′ ends of the template strands of the leading arm and the lagging arm of the fork DNA substrate (Fig. 1A). These fork DNA substrates had low FRET values of ~0.25. In the presence of ATP, Rad5 reversed these substrates to form two duplex DNA products. One of these products had high FRET values of ~0.80. Thus, we were able to observe the conversion of substrate to product in real time by monitoring the increase in FRET.
Figure 1: The FRET-based fork reversal assay.
A, Diagram of the assay showing the conversion of the low-FRET fork DNA substrate to the high-FRET duplex DNA product. B, Plot showing the increase of FRET as a function of time with 5 nM Rad5 (black), 10 nM Rad5 (red), 20 nM Rad5 (blue), and 50 nM Rad5 (green) with the fork DNA substrate containing no gap. C, Close up of the initial linear region of panel B. The solid lines represent the best fit of the data with an initial rate of 0.13 nM/min for 5 nM Rad5 (black), 0.26 nM/min for 10 nM Rad5 (red), 0.46 nM/min for 20 nM Rad5 (blue), and 0.98 nM/min for 50 nM Rad5 (green). D, Plot of the initial rate of fork reversal as a function of Rad5 concentration.
We carried out fork reversal assays with Rad5 (5 nM) and the fork DNA substrate containing no gap (5 nM) at equal concentrations in the presence of ATP (2mM). The observed initial linear rate of fork reversal was 0.13 ± 0.02 nM/min (Fig. 1B and C). To determine whether this reaction occurred under steady state conditions or under single-turnover conditions, we repeated these experiments with different the concentrations of Rad5 (10, 20, and 50 nM). If the reactions occurred under single-turnover conditions, the observe rate of fork reverse would be constant. If the reactions occurred under steady state conditions, however, the observed rate would increase linearly with Rad5 concentration. In this case, the observed initial linear rate of fork reversal increases linearly with Rad5 concentration (Fig. 1D). This shows that the fork reversal reactions are occurring under steady state conditions and that the observed initial linear rates of fork reversal are the most appropriate way to compare these reactions.
Rad5 catalyzes fork reversal with different fork DNA substrates.
We next examined whether Rad5 preferentially reversed a fork DNA substrate containing a short gap (10 nt.) in the primer strand of the leading arm, a fork DNA substrate containing a short gap (10 nt.) in the primer strand of the lagging arm, or a fork DNA substrate containing no gaps (Fig. 2). The leading arm and the lagging arm of the fork DNA substrates are both 35 nt. long. Thus, the DNA substrate with no gaps contained 35 bp. of duplex DNA on both arms. The DNA substrates with 10-nt. gaps in the leading arm or lagging arm contained 35 bp. of duplex DNA on one arm and 25 bp. of duplex DNA on the other arm.
Figure 2: Rad5-catalyzed fork reversal with different fork DNA substrates.
A, Plot showing the increase of FRET as a function of time on a fork DNA substrate containing no gap. B, Plot showing the increase of FRET as a function of time on a fork DNA substrate containing a 10-nucleotide gap on the leading arm. C, Plot showing the increase of FRET as a function of time on a fork DNA substrate containing a 10-nucleotide gap on the lagging arm. D, Close up of the initial linear region of panel A. The solid line represents the best fit of the data with an initial rate of 0.13 nM/min. E, Close up of the initial linear region of panel B. The solid line represents the best fit of the data with an initial rate of 4.6 nM/min. F, Close up of the initial linear region of panel C. The solid line represents the best fit of the data with an initial rate of 1.9 nM/min. Mean and standard deviations of the initial rates are given in the text.
We first determined whether the order by which the fork DNA substrate, Rad5, and ATP are mixed affected the observed rate of fork reversal (Supplemental Fig. S1). When the DNA substrate containing a short gap (10 nt.) in the primer strand of the leading arm (5 nM) and the ATP (2 mM) were pre-incubated for 2 min. followed by the addition of Rad5 (5nM) to initiate the reaction, the observed initial rate of fork reversal was equal to 4.4 ± 0.6 nM/min. When the DNA substrate (5 nM) and Rad5 (5 nM) were pre-incubated for 2 min. followed by the addition of ATP (2 mM) to initiate the reaction, the observed rate was equal to 2.3 ± 0.2 nM/min. This slower rate of fork reversal suggests that Rad5 forms a non-productive complex on the fork DNA substrate in the absence of ATP that must be converted to a productive complex upon the addition of ATP. To avoid this complication, all subsequent experiments were performed with the DNA and ATP substrates pre-incubated.
The observed initial rate of reversal of the fork DNA substrate containing a 10 nt.-gap on the primer strand of the leading arm was equal to 4.4 ± 0.6 nM/min (Fig. 2 and Supplemental Fig. S2). Reversal of the substrate containing a 10 nt.-gap on the lagging arm was more than two-fold slower with an observed rate equal to 1.9 ± 0.1 nM/min. Reversal of the substrate containing no gaps on either strand was more than ten-fold slower with an observed rate equal to 0.11 ± 0.02 nM/min. This shows that Rad5 preferentially reverses fork DNA substrates containing short gaps in the primer strand of the leading arm.
Rad5 binding to different fork DNA substrates.
The differences in the observed rates of fork reversal with the various fork DNA substrates could be due to differences in the binding of Rad5 to these DNA substrates. To determine whether there were significant changes in the binding of Rad5 to the different fork DNA substrates, we used an electrophoretic mobility shift assay (Fig. 3A). We mixed various concentrations of Rad5 (0 to 200 nM) with 20 nM of the fork DNA substrate containing a short gap (10 nt.) in the primer strand of the leading arm, the fork DNA substrate containing a short gap (10 nt.) in the lagging arm, or the fork DNA substrate containing no gaps in the absence of ATP. We observed similar binding of Rad5 to all three fork DNA substrates.
Figure 3: Rad5 binding to different fork DNA substrates.
A, Gel showing Rad5 binding to fork DNA substrate containing no gap, a 10-nucleotide gap on the leading arm, or a 10-nucleotide gap on the lagging arm. B, Mass photometry of Rad5 incubated with the fork DNA substrate containing no gap. C, Mass photometry of Rad5 incubated with the fork DNA substrate containing a 10-nucleotide gap on the leading arm. D, Mass photometry of Rad5 incubated with the fork DNA substrate containing a 10-nucleotide gap on the lagging arm.
We next examined the stoichiometry of the Rad5-fork DNA substrate complex using mass photometry. In the case of the fork DNA substrate containing no gaps, the mass photometry histogram (with a bin size of 5 kDa) that was recorded one minute after dilution and loading on the mass photometer had three clear peaks with maximum values at 70, 130, and 190 kDa. These corresponded to the calculated masses of the fork DNA substrate, Rad5, and the Rad5-fork DNA substrate complex, which were 73 kDa, 138 kDa, and 211 kDa, respectively. (Control experiments using only Rad5 gave a single peak corresponding to a mass equal to 135 kDa.) Three minutes after dilution and loading on the mass photometer, the Rad5-fork DNA substrate complex dissociated to free Rad5 and fork DNA substrate. The lack of any larger mass species showed that predominant Rad5-fork DNA substrate complex contained one Rad5 and one fork DNA substrate.
In the case of the fork DNA substrate containing the 10-nt. gap on the primer strand of the leading arm or containing the 10nt. gap on the lagging arm, the mass photometry histogram (with a bin size of 5 kDa) that was recorded after one minute had three clear peaks with maximum values at 60, 135, and 195 kDa. These corresponded to the calculated masses of the fork DNA substrate, Rad5, and the Rad5-fork DNA substrate complex, which were 69 kDa, 138 kDa, and 207 kDa, respectively. Similar to what was observed with the fork DNA substrate containing no gaps, the Rad5-fork DNA substrate complexes dissociated to free Rad5 and fork DNA substrate after three minutes. Taken together, the stoichiometries of the Rad5-fork DNA substrate complexes were 1:1 with all three fork DNA substrates.
Effects of RPA on Rad5-catalyzed fork reversal.
Replication protein A (RPA; the eukaryotic single-stranded DNA-binding protein) is likely to bind to the short single-stranded gaps at stalled replication forks. To understand how the presence of RPA at stalled replication forks affects the ability of Rad5 to catalyze fork reversal, we determined the rates of Rad5-catalyzed fork reversal with the three fork DNA substrates with 5 nM RPA: one containing a 10-nt. gap on the leading arm, one containing a 10-nt. gap on the lagging arm, and one containing no gaps (Fig. 4). We performed electrophoretic mobility shift assays to show that RPA binds to fork DNA substrates with short gaps (Supplemental Fig. S3).
Figure 4: Rad5-catalyzed fork reversal in the presence of RPA.
A, Plot showing the increase of FRET as a function of time on a fork DNA substrate containing no gap in the presence of RPA. B, Plot showing the increase of FRET as a function of time on a fork DNA substrate containing a 10-nucleotide gap on the leading arm in the presence of RPA. C, Plot showing the increase of FRET as a function of time on a fork DNA substrate containing a 10-nucleotide gap on the lagging arm in the presence of RPA. D, Close up of the initial linear region of panel A. The solid line represents the best fit of the data with an initial rate of 0.50 nM/min. E, Close up of the initial linear region of panel B. The solid line represents the best fit of the data with an initial rate of 2.8 nM/min. F, Close up of the initial linear region of panel C. The solid line represents the best fit of the data with an initial rate of 1.4 nM/min. Mean and standard deviations of the initial rates are given in the text.
In the presence of RPA, Rad5 carried out fork reversal with the substrate containing no gaps with a rate equal to 0.47 ± 0.09 nM/min (Fig. 4 and Supplemental Fig. S2). This rate was approximately the same as rate measured in the absence of RPA. Rad5 carried out fork reversal with the substrate containing a gap in the leading arm with a rate equal to 2.0 ± 0.4 nM/min. This rate is two-fold slower than the rate measured in the absence of RPA. Rad5 carried out fork reversal with the substrate containing a gap on the lagging arm with a rate equal to 1.5 ± 0.1 nM/min. This rate was approximately the same as rate measured in the absence of RPA. These results show that RPA has little effect on the rates of fork reversal with the fork DNA substrates containing a gap in the lagging arm or containing no gaps. RPA, by contrast, does reduce the rate of fork reversal with fork DNA substrates containing a gap in the leading arm approximately two-fold.
Effects of RNA primers on Rad5-catalyzed fork reversal.
At stalled replication forks in cells, the 5′-end of the primer strand on the lagging arm will contain about 10 nt. of RNA. To understand how the presence 10 nt. of RNA on the 5′-end of the lagging arm primer strand affects the ability of Rad5 to catalyze fork reversal, we determined the rates of Rad5-catalyzed fork reversal with our three model fork DNA substrates, when 10 nt. of RNA was present on the lagging arm (Fig. 5). In the presence of these RNA primers, Rad5 carried out fork reversal with the substrate containing no gaps, containing a 10-nt. gap on the leading arm, and containing a 10-nt. gap on the lagging arm with rates equal to 0.33 ± 0.05 nM/min, 1.5 ± 0.2 nM/min, and 2.1 ± 0.5 nM/min, respectively (Fig. 5 and Supplemental Fig. S2). These results show that the presence of short RNA primers on the lagging arm has little effect on the rates of fork reversal with the fork DNA substrates containing a gap in the lagging arm or containing no gaps. The presence of RNA primers, by contrast, does reduce the rate of fork reversal with fork DNA substrates containing a gap in the leading arm by approximately three-fold.
Figure 5: Rad5-catalyzed fork reversal in the presence of an RNA primer on the lagging arm.
A, Plot showing the increase of FRET as a function of time on a fork DNA substrate containing no gap. B, Plot showing the increase of FRET as a function of time on a fork DNA substrate containing a 10-nucleotide gap on the leading arm. C, Plot showing the increase of FRET as a function of time on a fork DNA substrate containing a 10-nucleotide gap on the lagging arm. The position of the RNA on the primer is shown as orange and is indicated by the black dot. D, Close up of the initial linear region of panel A. The solid line represents the best fit of the data with an initial rate of 0.27 nM/min. E, Close up of the initial linear region of panel B. The solid line represents the best fit of the data with an initial rate of 1.4 nM/min. F, Close up of the initial linear region of panel C. The solid line represents the best fit of the data with an initial rate of 2.0 nM/min. Mean and standard deviations of the initial rates are given in the text.
Effects of 5′-phosphates on Rad5-catalyzed fork reversal.
All of the Rad5 activity experiments described above were done with OH groups at the 5′-end of the primer strand on the lagging arm. To determine the effects of a phosphate group at this position, we repeated all of the prior experiments with fork DNA substrates containing a 5′-phosphate on the primer strand of the lagging arm. In the presence of the 5′-phosphate, Rad5 carried out fork reversal with the substrate containing a 10-nt. gap on the leading arm with a rate equal to 1.7 ± 0.6 nM/min, which represents approximately a two-fold reduction in rate compared to the identical fork DNA substrate with a 5′-OH (Supplemental Fig. S2). Rad5 catalyzed fork reversal with all other fork DNA substrates with approximately the same rate with the 5′-phosphate and with the 5′-OH.
Effects of gap size on Rad5-catalyzed fork reversal.
The studies described above utilized fork DNA substrate with 10-nt. gaps on either the leading or lagging arms. To determine the effect of gap size on the ability of Rad5 to catalyze fork reversal, we carried out experiments with fork DNA substrates with 30-nt. gaps on either the leading or lagging arms. To accommodate these longer gaps, we needed to design new fork DNA substrates with longer leading and lagging strands. We placed Cy3 and Cy5 fluorescent probes at the 3′ and 5′ ends of the template and primer strands of the leading arm of the fork DNA substrate (Fig. 6). These fork DNA substrates had high FRET values of ~0.80. In the presence of ATP, Rad5 reversed these substrates to form two duplex DNA products. These products had low FRET values of ~0.25.
Figure 6: Rad5-catalyzed fork reversal in the presence of longer gaps.
A, Plot showing the decrease of FRET as a function of time on a fork DNA substrate containing a 30-nucleotide gap on the leading arm. B, Plot showing the decrease of FRET as a function of time on a fork DNA substrate containing a 30-nucleotide gap on the lagging arm. C, Close up of the initial linear region of panel A. The solid line represents the best fit of the data with an initial rate of 4.0 nM/min. D, Close up of the initial linear region of panel B. The solid line represents the best fit of the data with an initial rate of 0.87 nM/min. Mean and standard deviations of the initial rates are given in the text.
Rad5 carried out fork reversal with the substrate containing a 30-nt. gap on the leading arm with a rate equal to 4.0 ± 0.2 nM/min. Similarly, Rad5 carried out fork reversal with the substrate containing a 30-nt. gap on the lagging arm with a rate equal to 2.9 ± 0.5 nM/min. These results show that the presence of the longer, 30-nt. gaps on the leading arm or the lagging arm of fork DNA substrates has little effect on the rates of fork reversal relative to fork DNA substrates with shorter, 10-nt. gaps.
Finally, we examined whether the presence of the primer strand in either the leading arm or the lagging arm of the fork DNA substrates were necessary for Rad5 to carry out fork reversal. We prepared two fork DNA substrates: one that lacked the primer strand on the leading arm and one that lacked the primer strand on the lagging arm. Rad5 was unable to catalyze fork reversal of either of these fork DNA substrates (Fig. 7).
Figure 7: Rad5 requires daughter strand on both the leading and lagging arms of the fork to reverse fork DNA substrate.
A, Plot showing no change in FRET as a function of time on a fork DNA substrate lacking the daughter strand on the leading arm. B, Plot showing no change in FRET as a function of time on a fork DNA substrate lacking the daughter strand on the lagging arm.
Discussion
Template switching is one of several pathways that eukaryotic cells use to bypass DNA damage during DNA replication [5,6,18,23,24]. In yeast, the fork-reversal helicase Rad5 plays a central role in template switching [19,20], and in humans, HLTF and SHPRH, which are both homologs of Rad5, play similar roles in template switching [17,21,22]. When DNA damage blocks DNA synthesis on the leading strand, these fork-reversal helicases are thought to use their helicase activities to convert the stalled replication fork into a four-way DNA junction called the chicken foot intermediate [18,19]. The formation of the chicken foot intermediate essentially switches the template of the stalled leading strand to the newly synthesized lagging strand (the newly synthesized sister strand). These annealed sister strands comprise the “middle toe” of the chicken foot intermediate. Using the newly synthesized sister strand as a template, a replicative polymerase such as pol δ, catalyzes the extension of the primer strand of the middle toe. Ultimately, this leads to bypass of the DNA damage in the leading strand once the extended chicken foot intermediate is converted back to a replication fork. Given this model of template switching, the bypass of DNA damage blocking synthesis on the leading strand will be relatively straight forward. By contrast, the bypass of DNA damage blocking synthesis on the lagging strand will be far more complicated, if possible at all. In this discussion, we will confine ourselves to the mechanism of bypass of DNA damage blocking synthesis on the leading strand.
To better understand the mechanism of fork-reversal by Rad5, we determined the rates of Rad5-catalyzed fork reversal using a variety of different fork DNA substrates. The rates of fork-reversal that we have measured here ranged from 0.05 to 4.4 nM/min, which are slower than what is typically observed for DNA helicases. Given that the rate of fork-reversal approximately doubled when twice as much Rad5 was used shows that these assays were done under steady state conditions and that the observed rates do not directly correspond to the rate of the fork-reversal step itself. Instead, the observed rates most likely reflect a slow initiation step in the Rad5-fork DNA substrate complex prior to the fork-reversal step. Further support for this comes from the order-of-mixing experiments that we reported here. Fork reversal occurred nearly two-fold slower when Rad5 was pre-incubated with the fork DNA substrate followed by the addition of ATP compared to when the fork DNA substrate was pre-incubated with ATP followed by the addition of Rad5. These results also suggests that in the absence of ATP, Rad5 may form a transient, non-productive complex with the fork DNA substrate, which would explain the slower rate of initiation when Rad5 is pre-incubated with the fork DNA substrate.
We showed that the rate of fork reversal by Rad5 was the fastest when the fork DNA substrate containing a short gap in the primer strand of the leading arm. This rate was greater than two-fold faster than when the fork DNA substrate contained a short gap in the primer strand of the lagging arm and was more than ten-fold faster than when the fork DNA substrate contained no gaps. This suggests that the absence of gaps on either the leading or lagging arms may present a structural obstacle to Rad5 initiating fork reversal. The greater fork-reversal activity with the fork DNA substrate with the leading arm gap relative to the fork DNA substrate with the lagging arm gap, however, is interesting from a biological perspective. Consider fork reversal when the damage blocks synthesis on the leading strand. The DNA substrate can contain a gap in either the primer of the leading arm or lagging arm. Only fork DNA substrates with a gap in the primer strand of the leading arm will form chicken foot intermediates with 5’ overhangs that are capable of being directly extended by a replicative DNA polymerase (Fig. 8). Fork DNA substrates with a lagging strand gap will form chicken foot structures with 3’ overhangs that cannot be directly extended. Thus, the most favorable fork DNA substrate for Rad5 is the one that can subsequently be extended in the next step of the template switching pathway.
Figure 8: Model of the fork-unwinding and primer-extension steps of template switching.
A, Reversal of a fork DNA possessing a gap on the daughter strand of the leading arm results in formation of a chicken foot intermediate that can be extended by a DNA polymerase. B, Reversal of a fork DNA possessing a gap on the daughter strand of the lagging arm results in formation of a chicken foot intermediate that cannot be extended by a DNA polymerase. C, Possible mechanism of Rad5-catalyzed fork reversal. The helicase domain is shown in blue and HIRAN domain is shown orange.
RPA is likely to be present at stalled replication forks by binding any single-stranded gaps in the DNA [38–42]. Here we determined whether Rad5 was capable of carrying out fork reversal in the presence of RPA. We used fork DNA substrates with short (10 nt.) gaps and longer gaps (30 nt.). In all cases, Rad5 was capable of unwinding fork DNA substrates in the presence of RPA. On the preferred fork DNA substrate, the one with a short gap on the primer strand of the leading arm, the presence of RPA reduced the observed rate of fork reversal by approximately two-fold relative to the absence of RPA. On all other DNA substrates, Rad5 catalyzed fork reversal at almost the same rate in the presence and absence of RPA. Thus, Rad5 is not prevented from catalyzing fork reversal in the presence of RPA.
In cells, the 5′ end of the primer strand of the lagging arm is expected to contain a short RNA primer of approximately 10 nt. in length. Here we showed that Rad5 is in fact capable of catalyzing fork reversal in the presence of such RNA primers on the lagging arm of the fork DNA substrates. It is currently unclear whether the Rad5 helicase directly unwinds the two strands of the lagging arm or the two strands of the leading arm (see below). In either case, however, these results clearly show that Rad5 is able to disrupt both DNA-DNA and DNA-RNA double helices, either directly by unwinding the lagging arm or indirectly by unwinding the leading arm.
The precise mechanism of Rad-catalyzed fork reversal has remained elusive. Several pieces of evidence provided here and other pieces of evidence provided previously by us and by others collectively suggested a possible mechanism of Rad5. First, we showed here that Rad5 and the fork DNA substrate form a one-to-one complex. Second, Rad5 has a HIRAN domain (residues 170 to 293) and a helicase domain (residues 430 to 1169) that are separated by a long, flexible linker of approximately 140 residues [25,27,37]. The HIRAN domain of the human Rad5 homolog HLTF binds 3′ end of DNA substrates [17,20]. Presumably in the context of a fork DNA substrate, this would be the 3′ end of the primer strand on the leading arm. Third, amino acid substitutions in the HIRAN domain of HLTF substantially reduces the efficiency of fork reversal [43]. Fourth, Rad5 will only catalyze fork reversal if the leading and lagging arms are homologous [19], and as we showed here, primer strands on the leading arm and the lagging arm are both necessary for fork reversal. Finally, DNase I footprinting experiments have shown that HLTF interacts more with the lagging arm of the fork DNA substrate than with the leading arm [17].
Taken together, these finding are all consistent with a model in which the Rad5 helicase domain binds the lagging arm of the fork DNA substrate at the fork junction and starts translocating away from the fork junction on the lagging arm while unwinding the primer strand from the template strand (Fig. 8). The HIRAN domain, which is separated from the helicase domain by a flexible linker, engages the 3′ end of the primer strand of the leading arm and uses the energy of binding to melt it from the template strand in an ATP hydrolysis-independent manner. As the Rad5 helicase domain continues to translocate and unwind the lagging arm primer strand in an ATP hydrolysis-dependent manner, the displaced leading arm primer strand bound to the HIRAN domain can base pair with it. Once the two newly synthesized strands base pair, further unwinding of the lagging arm (as well as the annealing of the two template strands behind the translocating Rad5) will provide the directionality and the energy necessary to drive the indirect unwinding of the leading arm. Ultimately, this will result in fork reversal and chicken foot formation. While this mechanism is somewhat speculative, further biochemical and structural studies will be able to test specific aspects of this model of Rad5-catalyzed fork reversal.
Supplementary Material
Highlights.
Rad5 is a fork-reversal helicase required for template switching
We monitored the kinetics of fork reversal using a FRET-based assay
Rad5 preferentially reverses forks with gaps in the leading strand
Rad5 reverses forks in the presence of RPA or RNA primers in the lagging strand
The preferred forks are the ones that can be extended during template switching
Acknowledgments
The project described was supported by award number GM081433 to M.T.W. from the National Institute of General Medical Sciences (NIGMS) and award number CA232425 to M.S. from the National Cancer Institute (NCI). The content is solely the responsibility of the authors and does not necessarily represent the official views of NIGMS, NCI, or NIH. We acknowledge Dr. Zhen Xu and the University of Iowa Protein and Crystallography core for technical assistance. Finally, we acknowledge Dr. Lynne Dieckman, Dr. Brittany Ripley, Zach Frevert, and Devin Reusch for discussions.
Abbreviations
- FRET
Förster resonance energy transfer
- PCNA
proliferating cell nuclear antigen
- pol
DNA polymerase
- RPA
replication protein A
- TLS
translesion synthesis
Footnotes
Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- [1].Hanahan D, Weinberg RA, The Hallmarks of Cancer, Cell. 100 (2000) 57–70. 10.1016/S0092-8674(00)81683-9. [DOI] [PubMed] [Google Scholar]
- [2].Hanahan D, Weinberg RA, Hallmarks of Cancer: The Next Generation, Cell. 144 (2011) 646–674. 10.1016/j.cell.2011.02.013. [DOI] [PubMed] [Google Scholar]
- [3].Veltman JA, Brunner HG, De novo mutations in human genetic disease, Nat. Rev. Genet. 13 (2012) 565–575. 10.1038/nrg3241. [DOI] [PubMed] [Google Scholar]
- [4].López-Otín C, Blasco MA, Partridge L, Serrano M, Kroemer G, The hallmarks of aging., Cell. 153 (2013) 1194–217. 10.1016/j.cell.2013.05.039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Boiteux S, Jinks-Robertson S, DNA Repair Mechanisms and the Bypass of DNA Damage in Saccharomyces cerevisiae, Genetics. 193 (2013) 1025–1064. 10.1534/genetics.112.145219. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Marians KJ, Lesion Bypass and the Reactivation of Stalled Replication Forks., Annu. Rev. Biochem. 87 (2018) 217–238. 10.1146/annurev-biochem-062917-011921. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Hakem R, DNA-damage repair; the good, the bad, and the ugly, EMBO J. 27 (2008) 589–605. 10.1038/emboj.2008.15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Kondratick CM, Washington MT, Spies M, Making Choices: DNA Replication Fork Recovery Mechanisms., Semin. Cell Dev. Biol. 113 (2021) 27–37. 10.1016/j.semcdb.2020.10.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Sale JE, Lehmann AR, Woodgate R, Y-family DNA polymerases and their role in tolerance of cellular DNA damage, Nat. Rev. Mol. Cell Biol. 13 (2012) 141–152. 10.1038/nrm3289. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Pryor JM, Dieckman LM, Boehm EM, Washington MT, Eukaryotic Y-Family Polymerases: A Biochemical and Structural Perspective, in: Murakami KS, Trakselis MA (Eds.), Nucleic Acid Polym, Springer Berlin Heidelberg, Berlin, Heidelberg, 2014: pp. 85–108. 10.1007/978-3-642-39796-7_4. [DOI] [Google Scholar]
- [11].Yang W, Weng PJ, Gao Y, A new paradigm of DNA synthesis: three-metal-ion catalysis, Cell Biosci. 6 (2016) 51. 10.1186/s13578-016-0118-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Friedberg EC, Wagner R, Radman M, Specialized DNA polymerases, cellular survival, and the genesis of mutations., Science. 296 (2002) 1627–30. 10.1126/science.1070236. [DOI] [PubMed] [Google Scholar]
- [13].Prakash S, Prakash L, Translesion DNA synthesis in eukaryotes: A one- or two-polymerase affair, Genes Dev. 16 (2002) 1872–1883. 10.1101/gad.1009802. [DOI] [PubMed] [Google Scholar]
- [14].Prakash S, Johnson RE, Prakash L, EUKARYOTIC TRANSLESION SYNTHESIS DNA POLYMERASES: Specificity of Structure and Function, Annu. Rev. Biochem. 74 (2005) 317–353. 10.1146/annurev.biochem.74.082803.133250. [DOI] [PubMed] [Google Scholar]
- [15].Powers KT, Washington MT, Eukaryotic translesion synthesis: Choosing the right tool for the job, DNA Repair (Amst). 71 (2018) 127–134. 10.1016/j.dnarep.2018.08.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Yang W, Gao Y, Translesion and Repair DNA Polymerases: Diverse Structure and Mechanism, Annu. Rev. Biochem. 87 (2018) 239–261. 10.1146/annurev-biochem-062917-012405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Chavez DA, Greer BH, Eichman BF, The HIRAN domain of helicase-like transcription factor positions the DNA translocase motor to drive efficient DNA fork regression, J. Biol. Chem. 293 (2018) 8484–8494. 10.1074/jbc.RA118.002905. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Unk I, Hajdú I, Blastyák A, Haracska L, Role of yeast Rad5 and its human orthologs, HLTF and SHPRH in DNA damage tolerance, DNA Repair (Amst). 9 (2010) 257–267. 10.1016/j.dnarep.2009.12.013. [DOI] [PubMed] [Google Scholar]
- [19].Blastyák A, Pintér L, Unk I, Prakash L, Prakash S, Haracska L, Yeast Rad5 Protein Required for Postreplication Repair Has a DNA Helicase Activity Specific for Replication Fork Regression, Mol. Cell. 28 (2007) 167–175. 10.1016/j.molcel.2007.07.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Shin S, Hyun K, Kim J, Hohng S, ATP Binding to Rad5 Initiates Replication Fork Reversal by Inducing the Unwinding of the Leading Arm and the Formation of the Holliday Junction, Cell Rep. 23 (2018) 1831–1839. 10.1016/j.celrep.2018.04.029. [DOI] [PubMed] [Google Scholar]
- [21].Hishiki A, Hara K, Ikegaya Y, Yokoyama H, Shimizu T, Sato M, Hashimoto H, Structure of a Novel DNA-binding Domain of Helicase-like Transcription Factor (HLTF) and Its Functional Implication in DNA Damage Tolerance, J. Biol. Chem. 290 (2015) 13215–13223. 10.1074/jbc.M115.643643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Hishiki A, Sato M, Hashimoto H, Structure of HIRAN domain of human HLTF bound to duplex DNA provides structural basis for DNA unwinding to initiate replication fork regression, J. Biochem. 167 (2020) 597–602. 10.1093/jb/mvaa008. [DOI] [PubMed] [Google Scholar]
- [23].Xu X, Blackwell S, Lin A, Li F, Qin Z, Xiao W, Error-free DNA-damage tolerance in Saccharomyces cerevisiae., Mutat. Res. Rev. Mutat. Res. 764 (2015) 43–50. 10.1016/j.mrrev.2015.02.001. [DOI] [PubMed] [Google Scholar]
- [24].Poole LA, Cortez D, Functions of SMARCAL1, ZRANB3, and HLTF in maintaining genome stability, Crit. Rev. Biochem. Mol. Biol. 52 (2017) 696–714. 10.1080/10409238.2017.1380597. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Johnson RE, Henderson ST, Petes TD, Prakash S, Bankmann M, Prakash L, Saccharomyces cerevisiae RAD5-encoded DNA repair protein contains DNA helicase and zinc-binding sequence motifs and affects the stability of simple repetitive sequences in the genome., Mol. Cell. Biol. 12 (1992) 3807–18. 10.1128/mcb.12.9.3807-3818.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Hoege C, Pfander B, Moldovan G-L, Pyrowolakis G, Jentsch S, RAD6-dependent DNA repair is linked to modification of PCNA by ubiquitin and SUMO, Nature. 419 (2002) 135–141. 10.1038/nature00991. [DOI] [PubMed] [Google Scholar]
- [27].Gangavarapu V, Haracska L, Unk I, Johnson RE, Prakash S, Prakash L, Mms2-Ubc13-Dependent and -Independent Roles of Rad5 Ubiquitin Ligase in Postreplication Repair and Translesion DNA Synthesis in Saccharomyces cerevisiae, Mol. Cell. Biol. 26 (2006) 7783–7790. 10.1128/MCB.01260-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Carlile CM, Pickart CM, Matunis MJ, Cohen RE, Synthesis of Free and Proliferating Cell Nuclear Antigen-bound Polyubiquitin Chains by the RING E3 Ubiquitin Ligase Rad5, J. Biol. Chem. 284 (2009) 29326–29334. 10.1074/jbc.M109.043885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Moldovan GL, Pfander B, Jentsch S, PCNA, the Maestro of the Replication Fork, Cell. 129 (2007) 665–679. 10.1016/j.cell.2007.05.003. [DOI] [PubMed] [Google Scholar]
- [30].Ulrich HD, Jentsch S, Two RING finger proteins mediate cooperation between ubiquitin-conjugating enzymes in DNA repair., EMBO J. 19 (2000) 3388–97. 10.1093/emboj/19.13.3388. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [31].Stelter P, Ulrich HD, Control of spontaneous and damage-induced mutagenesis by SUMO and ubiquitin conjugation, Nature. 425 (2003) 188–191. 10.1038/nature01965. [DOI] [PubMed] [Google Scholar]
- [32].Bergink S, Jentsch S, Principles of ubiquitin and SUMO modifications in DNA repair, Nature. 458 (2009) 461–467. 10.1038/nature07963. [DOI] [PubMed] [Google Scholar]
- [33].Dieckman LM, Freudenthal BD, Washington MT, PCNA structure and function: insights from structures of PCNA complexes and post-translationally modified PCNA., Subcell. Biochem 62 (2012) 281–299. 10.1007/978-94-007-4572-8_15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [34].Boehm EM, Gildenberg MS, Washington MT, The Many Roles of PCNA in Eukaryotic DNA Replication, 1st ed., Elsevier Inc., 2016. 10.1016/bs.enz.2016.03.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].Johnson RE, Prakash S, Prakash L, Yeast DNA repair protein RAD5 that promotes instability of simple repetitive sequences is a DNA-dependent ATPase., J. Biol. Chem. 269 (1994) 28259–28262. 10.1016/S0021-9258(18)46922-0. [DOI] [PubMed] [Google Scholar]
- [36].Chen S, Davies AA, Ulrich HD, The RING finger ATPase Rad5p of Saccharomyces cerevisiae contributes to DNA double-strand break repair in a ubiquitin-independent manner, Nucleic Acids Res. 33 (2005) 5878–5886. 10.1093/nar/gki902. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [37].Gildenberg MS, Todd Washington M, Conformational flexibility of fork-remodeling helicase Rad5 shown by full-ensemble hybrid methods, PLoS One. 14 (2019) 1–16. 10.1371/journal.pone.0223875. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [38].Wold MS, REPLICATION PROTEIN A:A Heterotrimeric, Single-Stranded DNA-Binding Protein Required for Eukaryotic DNA Metabolism, Annu. Rev. Biochem. 66 (1997) 61–92. 10.1146/annurev.biochem.66.1.61. [DOI] [PubMed] [Google Scholar]
- [39].Caldwell CC, Spies M, Dynamic elements of replication protein A at the crossroads of DNA replication, recombination, and repair., Crit. Rev. Biochem. Mol. Biol. 55 (2020) 482–507. 10.1080/10409238.2020.1813070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [40].Prakash A, Borgstahl GEO, The structure and function of replication protein A in DNA replication., Subcell. Biochem 62 (2012) 171–196. 10.1007/978-94-007-4572-8_10. [DOI] [PubMed] [Google Scholar]
- [41].Bhat KP, Bétous R, Cortez D, High-affinity DNA-binding domains of replication protein A (RPA) direct SMARCAL1-dependent replication fork remodeling., J. Biol. Chem. 290 (2015) 4110–7. 10.1074/jbc.M114.627083. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [42].Machwe A, Lozada E, Wold MS, Li G-M, Orren DK, Molecular cooperation between the Werner syndrome protein and replication protein A in relation to replication fork blockage., J. Biol. Chem. 286 (2011) 3497–508. 10.1074/jbc.M110.105411. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [43].Kile AC, Chavez DA, Bacal J, Eldirany S, Korzhnev DM, Bezsonova I, Eichman BF, Cimprich KA, HLTF’s Ancient HIRAN Domain Binds 3′ DNA Ends to Drive Replication Fork Reversal, Mol. Cell. 58 (2015) 1090–1100. 10.1016/j.molcel.2015.05.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.