1. Structural Biology and Molecular Biophysics

DNA translocation mechanism of an XPD family helicase

  1. Kaiying Cheng
  2. Dale B Wigley  Is a corresponding author
  1. Imperial College London, United Kingdom
https://doi.org/10.7554/eLife.42400
Share this article
Cite this article
  1. Kaiying Cheng
  2. Dale B Wigley
(2018)
DNA translocation mechanism of an XPD family helicase
eLife 7:e42400.
https://doi.org/10.7554/eLife.42400
  2. Download BibTeX
  3. Download .RIS

Abstract

The XPD family of helicases, that includes human disease-related FANCJ, DDX11 and RTEL1, are Superfamily two helicases that contain an iron-sulphur cluster domain, translocate on ssDNA in a 5’−3’ direction and play important roles in genome stability. Consequently, mutations in several of these family members in eukaryotes cause human diseases. Family members in bacteria, such as the DinG helicase from Escherichia coli, are also involved in DNA repair. Here we present crystal structures of complexes of DinG bound to single-stranded DNA (ssDNA) in the presence and absence of an ATP analogue (ADP•BeF3), that suggest a mechanism for 5’−3’ translocation along the ssDNA substrate. This proposed mechanism has implications for how those enzymes of the XPD family that recognise bulky DNA lesions might stall at these as the first step in initiating DNA repair. Biochemical data reveal roles for conserved residues that are mutated in human diseases.

https://doi.org/10.7554/eLife.42400.001

Introduction

DNA helicases are enzymes that separate nucleic acid duplexes into their component strands (Singleton et al., 2007). They have been divided into a number of superfamilies on the basis of sequence similarities (Gorbalenya and Koonin, 1993) that have also proved to reveal similarities in mechanism. Although most family members unwind nucleic acid duplexes some, such as chromatin remodellers (Clapier et al., 2017), instead translocate along nucleic acid duplexes without unwinding them and others translocate along single-stranded substrates to displace bound proteins (Jankowsky et al., 2001). The two largest helicase/translocase families are Superfamily 1 (SF1) and Superfamily 2 (SF2). The latter superfamily contains the canonical ‘DEAD’ box RNA helicases – a large group of proteins with roles in unwinding RNA substrates in processes such as transcription, splicing and ribosome assembly (Bourgeois et al., 2016). However, this family also contains DNA helicases with roles in a variety of different biological processes.

SF1 and SF2 members are generally monomeric enzymes with ATP-dependent translocation on single-stranded DNA (ssDNA) templates that has a characteristic polarity of either 5’−3’ or 3’−5’ (Singleton et al., 2007). However, many of these enzymes interact with other proteins in larger, multi subunit complexes (such as RecBCD that contains two SF1 helicases with opposite polarities (Wigley, 2013). Others clearly act cooperatively (Eoff and Raney, 2010). The crystal structures of several proteins have revealed how these enzymes interact with ssDNA and some studies have obtained multiple catalytic states of the proteins that have allowed deduction of their translocation mechanisms. The first enzyme for which this was achieved was the SF1 3’−5’ helicase PcrA (Subramanya et al., 1996; Velankar et al., 1999). Similar nucleotide-bound states of the SF2 3’−5’ helicase NS3 also revealed how this enzyme couples ATP hydrolysis to ssDNA (or ssRNA) translocation (Gu and Rice, 2010) by a mechanism that is distinct from SF1 enzymes like PcrA. For 5’−3’ translocation, structures of the SF1 enzyme RecD2 were also determined to provide yet another translocation mechanism (Saikrishnan et al., 2009). In all three of these example enzymes, although the mechanisms are distinct, the translocation step size that is deduced is one ATP per base. However, an obvious gap in our mechanistic understanding of SF1 and SF2 enzymes is that of a SF2 enzyme capable of 5’−3’ translocation.

XPD family enzymes are SF2 helicases with 5’−3’ directionality and some, such as XPD itself, recognise bulky DNA lesions (Voloshin et al., 2003; Wirth et al., 2016). XPD itself has a role in DNA repair in archaea and eukaryotes (Houten et al., 2016), although it is not clear that the functions are equivalent in these different kingdoms of life. XPD is a part of the TFIIH complex that does not have a direct equivalent in eubacteria although Escherichia coli contains homologues of XPD (DinG) and XPB (RadD), both of which are subunits of TFIIH in eukaryotes. DinG and RadD are implicated in DNA repair (Voloshin et al., 2003; Thakur et al., 2014; Chen et al., 2015) although their precise functions are poorly defined. Other eukaryotic XPD family members (e.g. RTEL1, FANCJ, DDX11) also have roles in DNA repair as well as other cellular processes including recovery of stalled replication forks and bypassing DNA lesions (Vannier et al., 2014; Bharti et al., 2016; Abe et al., 2018). Consequently, mutations in these genes are linked to cancer and other human diseases (Cerami et al., 2012; Suhasini and Brosh, 2013; Guo et al., 2016; Krassowski et al., 2018). XPD is the only family member for which there is any structural information and this has been limited to crystal structures (Liu et al., 2008; Fan et al., 2008; Wolski et al., 2008; Kuper et al., 2012; Constantinescu-Aruxandei et al., 2016) or cryoEM at medium resolution of the protein in the context of TFIIH (Greber et al., 2017; Schilbach et al., 2017). None is visualised with a ssDNA substrate contacting the entire ssDNA-binding site, or with any ATP analogue. Consequently, a molecular understanding of the mechanism of XPD family enzymes is currently lacking.

DinG is another member of the XPD family, that appears to have a poorly defined role in DNA repair (Voloshin et al., 2003). Additionally, it appears to have a role, by cooperation with UvrD and Rep helicases, both in removing R-loops and removing with RNA polymerase at blocked replication forks (Boubakri et al., 2010). DinG shows specificity for a number of unusual DNA structures such as G-quadruplexes as well as D-loops (Voloshin et al., 2003; Thakur et al., 2014; Voloshin and Camerini-Otero, 2007), a property that is also shared by some other XPD family members like FANCJ (Bharti et al., 2016). DinG has also been reported to have a role in a type I CRISPR system in Acidithiobacillus ferrooxidans (Makarova et al., 2011). These CRISPR-associated DinG proteins are unusual because they lack FeS clusters and contain an N-terminal extension that may be a nuclease domain (McRobbie et al., 2012). The nuclease activity is likely to play an important role in spacer acquisition since these CRISPR systems appear to lack Cas1 and Cas2 proteins.

Despite intensive effort, high resolution structural information about the molecular mechanism of XPD family enzymes has remained elusive. Here, we present the crystal structures of two complexes of E. coli DinG; a binary complex bound to a ssDNA substrate, and a ternary complex with ssDNA and an ATP analogue (ADP•BeF3), that provide information about how these enzymes translocate on DNA. As observed in crystal structures of other SF1 and SF2 helicases (Velankar et al., 1999; Gu and Rice, 2010; Saikrishnan et al., 2009; Sengoku et al., 2006), binding of nucleotide induces a conformational change that involves closure of a cleft between the two canonical helicase domains (Singleton et al., 2007). However, additional changes occur in the detailed interactions between the protein and the bound ssDNA that result in unidirectional translocation in a 5’−3’ direction by a mechanism that is supported by biochemical data. The structures also suggest a potential mechanism for how some members of the XPD family of enzymes might recognise and stall at bulky lesions. The structures represent an important advance in our understanding of the mechanism of XPD family enzymes as well as explaining how some of the mutations in these proteins cause deficiencies in their activities.

Results and discussion

Overall structure of DinG and comparison with XPD

DinG shares significant similarity to XPD at both the structural and sequence level (Liu et al., 2008; Fan et al., 2008; Wolski et al., 2008; Kuper et al., 2012; Constantinescu-Aruxandei et al., 2016) (Figure 1, Figure 1—figure supplement 1, Figure 1—figure supplement 2). The proteins comprise four domains; the helicase (HD1 and HD2), arch and FeS cluster domains (Figure 1, Figure 1—figure supplement 1 and Figure 1—figure supplement 2). The mean r.m.s. deviation of Cα positions for each of the helicase and arch domains of DinG, superimposed individually on equivalent domains from XPD structures, ranges from 1.8 to 3.3 Å, comparable to the same comparisons between different XPD crystal structures. The canonical SF2 helicase domains (HD1 and HD2) have the expected fold seen in other structures (Singleton et al., 2007), with an ATP-binding site located between them. The FeS domain, that is inserted within the HD1 sequence, contains an iron-sulphur (FeS) cluster at its centre (Figure 1 and Figure 1—figure supplement 3). Although the FeS domains in XPD and DinG are both all α-helical, are topologically related with respect to coordination of the FeS cluster and occupy very similar positions in the structure, they differ somewhat in fold and sequence (Figure 1—figure supplement 3). The arch domain sits above the other domains and contacts the FeS domain above the ssDNA binding surface. However, the similarity of the helicase and arch domains (that together comprise >85% of the protein), to those in XPD, suggest that DinG is a good structural model for other XPD family enzymes.

Figure 1 with 3 supplements see all
Download asset Open asset
Structures of the DinG/ssDNA binary and ternary complexes.

(A) The domain arrangement of DinG. The HD1, HD2, 4FeS and Arch domains are coloured in pink, lilac, wheat and green, respectively. Conserved motifs are labelled and listed. Residues labelled in blue participate in ATP hydrolysis, and those in green participate in DNA binding. (B) Cartoon view of the DinG binary complex with ssDNA coloured as in (A). (C) Cartoon view of the DinG ternary complex with ssDNA and ADP•BeF3, coloured as in (A). (D) Superposition of the two structures based on the HD1 domain to illustrate the conformational changes between the structures. The binary complex is in grey and the ternary complex is coloured by domains as in (A).

https://doi.org/10.7554/eLife.42400.002

Structure of the DinG binary complex with ssDNA

The DinG binary complex structure was solved at 2.5 Å resolution by two wavelength anomalous diffraction data based on selenomethionine-substituted protein but the final model was built and refined using data from a crystal of native protein. Crystallographic statistics are given in Table 1. Although a twelve base oligo-dT substrate was used in the crystallisation, only eleven bases are evident in the structure suggesting these likely comprise the entire ssDNA-binding site (Figure 2 and Figure 2—figure supplement 1).

Figure 2 with 5 supplements see all
Download asset Open asset
Details of the ssDNA-binding site.

(A) The DNA-binding site for the binary (above) and ternary (below) complexes. The side chains of residues participating in DNA binding are shown as sticks, coloured according to the domains from which they originate (HD1 in pink, HD2 in pale blue). The nucleotides of the bound ssDNA are rainbow coloured in the 5’ (red) to 3’ (blue) direction. The colour scheme for the bases is the same in all figures. (B) Cartoon representation of the same views detailing the contacts between the protein and DNA in each complex.

https://doi.org/10.7554/eLife.42400.006
Table 1
Statistics from crystallographic analysis.
https://doi.org/10.7554/eLife.42400.012
ComplexSe-DinG-DNADinG-DNADinG-DNA-ADPBeF3
PDB ID--6FWR6FWS
Data collection
SourceI04I04I04-1
Wavelength (Å)0.9795 (peak)0.9860 (remote)0.97950.9159
Resolution (Å)59–2.6 (2.64–2.6)56.2–3.4 (3.46–3.4)79.59–2.5 (2.54–2.5)63.37–2.5 (2.54–2.5)
Space groupP21212P21212P21212P212121
Cell dimensions: a, b, c99.11, 134.47, 5999.11, 134.47, 5998.97, 133.87, 58.86109.77, 119.75, 126.74
Obeservation274170 (12271)124833 (6193)166828 (8129)266197 (13044)
Unique reflections24842 (1169)11591 (559)27857 (1351)58152 (2827)
Rmerge (%)12.3 (51.7)15.5 (43.7)10.9 (50.3)11.9 (54.1)
I/σI25.8 (4.7)13.7 (4.9)12.3 (3.3)9.6 (2.6)
Completeness (%)99.2 (95.3)99.5 (96.7)100 (98)99.5 (98)
Redundancy1110.86.04.6
Refinement statistics
Resolution (Å)--79.95–2.5 (2.54–2.5)63.37–2.5 (2.54–2.5)
Rfactor (%)/Rfree (%)--21.08/24.5721.91/25.57
rmsd bonds (Å)/angles (°)--0.008/1.1340.008/1.080
Ramachandran plot: Favored (%)--97.496.2

  1. The numbers in parentheses refer to the last shell.

    Rfactor = Σ||F(obs)- F(calc)||/Σ|F(obs)|.

  2. Rfree = R factor calculated using 5.0% of the reflection data randomly chosen and omitted from the start of refinement.

Overall, the DNA is kinked but with many bases stacked upon one another along the strand in a conformation that is quite different to that observed previously in any other SF1 or SF2 protein (Figure 2—figure supplement 5). However, the bases contacting HD1 are similar to those in an equivalent position in NS3 except that in NS3 there is a tryptophan residue intercalated between the bases (Gu and Rice, 2010; Kim et al., 1998) that is not conserved in DinG. The bases spanning the HD1 domain, and crossing over to the HD2 domain, are stacked on one another until they encounter a proline-rich strand (P motif) on the HD2 domain, at which point the next base would clash with this strand if stacked on the preceding one so it is flipped up and over this strand. In all, three bases are contacted by the HD1 domain and seven bases contact the HD2 domain, with an additional base spanning the gap between the domains. Two structures of XPD proteins also have bound ssDNA (Kuper et al., 2012; Constantinescu-Aruxandei et al., 2016). However, despite being longer oligonucleotides, only four or five bases are ordered in these structures and they only make contacts with the HD2 domain. Furthermore, the details of the interactions in each structure are inconsistent and both structures differ from those we observe in DinG (Figure 2—figure supplement 3). The reasons for this are unclear although in one case the DNA was chemically crosslinked to the protein, which may have induced some distortions.

The bound ssDNA sits in a positively-charged groove across the ATPase domains in a similar location, and with the same directional polarity, seen in other SF1 and SF2 helicases (Singleton et al., 2007), and the contacts are mainly with the phosphodiester backbone (Figure 2). Although largely polar hydrogen bonding interactions, some exceptions include two hydrophobic stacking interactions (F615 and F638) with deoxyribose moieties that help to discriminate between DNA and RNA substrates (Figures 2 and 4), as observed previously in other helicases (Saikrishnan et al., 2009).

In addition to the DNA backbone contacts there are a few contacts with the DNA bases (Figure 2). Notably, there is an aromatic stacking interaction with Tyr636 intercalated between the first and second bases at the 5’ end of the ssDNA on the HD2 domain. Although a similar base intercalation interaction is seen in NS3 (Kim et al., 1998; Gu and Rice, 2010), this is at the opposite end of the bound DNA at the 3’ end and involves a tryptophan residue from the HD1 domain. The bases spanning the HD1 domain, and crossing over to the HD2 domain, are stacked on one another until they encounter the P motif (residues 614-AFPPID-619). The P motif is located in front of the spring-helix and contains several hydrophobic residues, often prolines, that would rigidify this region of the structure. The P motif in the taXPD structure (Kuper et al., 2012; Constantinescu-Aruxandei et al., 2016), also forms a narrow channel (Figure 2—figure supplement 3). The major protein-DNA interactions at the HD1/FeS domain interface are mediated by two arginine residues, R117 and R211 (Figure 1—figure supplement 3). Although the structures of the FeS domain differs between XPD and DinG, there are two basic residues (R88 and K170 in taXPD) that are conserved and interact with a sulphate ion that mimics the DNA phosphodiester backbone (Figure 1—figure supplement 3 and Figure 2—figure supplement 4).

Structure of DinG ternary complex with ssDNA and ADP•BeF3

The DinG ternary complex structure was solved at 2.5 Å resolution by molecular replacement using individual domains of the binary structure, and crystallographic statistics are given in Table 1. The domain architecture of the ternary complex is similar to the binary complex, although there are changes in the relative orientations of the domains (Figure 1). Due to these changes, the protein now interacts with only ten bases of ssDNA (Figure 2). Although there are still three bases contacting HD1 and seven contacting HD2, there is no longer a base spanning the gap between the domains indicating that one of the domains must have slid along the ssDNA onto this base. Since the enzyme has 5’−3’ directionality (Voloshin et al., 2003), this would suggest it is the HD2 domain that has done the sliding. Consistent with this notion, the base at position +5 (Figure 2) no longer stacks on base +6 (as seen in the binary complex) but instead would now clash with the P motif so becomes flipped up and over the strand to occupy the position taken by the +4 base in the binary complex. Similarly, the contacts with other bases spanning the HD2 domain are displaced by one base compared to the binary complex. Although many of the ssDNA contacts are similar in both complexes, there are additional contacts between the HD1/FeS domains and the ssDNA in the binary complex than in the ternary complex (Figure 2). By contrast, the number of contacts between ssDNA and the HD2 domain is greater in the ternary complex. The implications of these differences between the structures are discussed below in regard to the translocation mechanism. The principal difference between the structures of DinG in the presence and absence of ADP•BeF3 is a conformational change (Figure 1) similar to that observed for other SF1 and SF2 helicase family members (Singleton et al., 2007). As expected, there is an ADP•BeF3 moiety bound at the interface between the HD1 and HD2 domains (Figure 2—figure supplement 2 and the nucleotide is contacted by several residues in a manner seen previously in other helicases (Singleton et al., 2007). This conformational change results when the cleft between the two canonical helicase/translocase domains closes around the bound nucleotide. The consequences of this conformational change are discussed below.

A mechanism for translocation of ssDNA in a 5’−3’ direction

Translocation on ssDNA can be in either a 3’−5’ or 5’−3’ direction (Singleton et al., 2007). Mechanisms for 3’−5’ translocation have been described in structural detail for both SF1 (PcrA, (Velankar et al., 1999)) and SF2 (NS3, (Gu and Rice, 2010)) helicases. A mechanism for 5’−3’ translocation for a SF1 helicase has also been reported (RecD2, (Saikrishnan et al., 2009)). The mechanisms all suggest a step size of one base per ATP hydrolysed, which is consistent with biochemical measurements (Dillingham et al., 2000).

The principal differences between the structures we present here are changes in the interactions of the HD1 and HD2 domains that result in sliding of the HD2 domain along the bound ssDNA when ATP binding induces a conformational change (Figure 1 and 2). These changes allow us to propose a mechanism for ssDNA translocation by SF2 5’−3’ enzymes such as the XPD family enzymes (Figure 3 and Video 1). Upon initial binding to ssDNA, the protein makes interactions via a groove running across the surface of the ATPase (motor) domains. Upon ATP binding, a conformational change causes the HD2 motor domain to slide along ssDNA, while the HD1 domain retains a tight grip. When ATP is hydrolysed, the enzyme relaxes to the apo state allowing the HD1 domain to slide along ssDNA while the HD2 domain now grips tightly. As with translocation mechanisms for other SF1 and SF2 enzymes (Velankar et al., 1999; Gu and Rice, 2010; Saikrishnan et al., 2009), the key to the mechanism is the regulation of grip for each of the motor domains and how that changes in response to the different nucleotide-bound states of the enzyme. For DinG, in the binary complex, the HD1 domain has a tighter grip than in the ATP-bound (ternary) state. This is achieved by increasing the number of contacts with the residues towards the 3’-end of the ssDNA binding site (Figure 2). By gripping the ssDNA more tightly than the HD2 motor domain, as the conformational change induced by ATP binding closes the cleft between the motor domains, this causes the ssDNA to slide across the surface of the HD2 domain. In doing so, there is a steric clash between the base at position +5 and the rigid P motif that forces the base out of a stacked conformation with the adjacent base. After ATP hydrolysis, changes occur that cause HD2 to now have a tighter grip on the 5’ bases (again due to an increased number of interactions) and thus, as the conformational changes open the cleft between the motor domains, the ssDNA slides across the HD1 domain. The consequence of these two steps is translocation along the bound ssDNA in a 5’−3’ direction, first across the HD2 domain as ATP binds and then across the HD1 domain after ATP hydrolysis. We note, however, that this is a simple two-step model and we cannot preclude additional steps in the overall reaction cycle but this (likely over simplified) model allows a framework to explain directional translocation by these enzymes in the same way that similar structures have provided a framework to explain translocation by other SF1 and SF2 enzymes (Velankar et al., 1999; Gu and Rice, 2010; Saikrishnan et al., 2009).

Figure 3 with 1 supplement see all
Download asset Open asset
A proposed mechanism for 5’−3’ translocation.

Cartoons depicting the conformational changes and events during DNA translocation at (A) the domain level, (B) detailed interactions between the bases of the bound ssDNA with the HD1 +FeS and HD2 domains during a translocation cycle. Dotted lines shown contacts between the protein and DNA. (C) cartoon showing how the HD1 and HD2 domains alter grip on the bound ssDNA at different steps in the translocation. Further details are provided in the text.

https://doi.org/10.7554/eLife.42400.013
Video 1
Download asset
Conformational changes associated with ATP binding.

A morph between the two structures showing how ssDNA is translocated across the surface of the protein in response to ATP binding.

https://doi.org/10.7554/eLife.42400.015

Although this translocation mechanism is different to that proposed for other helicases, it retains a step size of one ATP per base translocated as deduced for other SF1 and SF2 family enzymes (Velankar et al., 1999; Gu and Rice, 2010; Saikrishnan et al., 2009; Qi et al., 2013). Previous studies have revealed mechanisms for both 3’−5’ and 5’−3’ translocation in SF1 helicases but just 3’−5’ translocation by SF2 enzymes. The work we present here now completes our picture for the mechanisms of translocation in each of the two polarities by the two main families of monomeric, single-strand translocases (Figure 3—figure supplement 1).

Biochemical data in support of the model

In order to test the validity of our model, we created a number of site-directed mutant proteins and assessed their biochemical activities (Figure 4 and Figure 4—figure supplement 1). The residues selected all play important roles in the interaction with nucleic acid and the translocation mechanism we propose. The mutant proteins all show defects in various aspects of their biochemical activities to differing degrees. All of the mutants showed defects in ssDNA binding although these were minor for the R117A and R211A mutations in the FeS domain (Figure 4). As a consequence, although we tried to compare ATPase rates, these are very low in the absence of ssDNA (Voloshin et al., 2003) and we were unable to saturate most of the mutants with ssDNA for the assays. As expected, the defects in interaction with ssDNA also resulted in at least 100-fold reduction in helicase activity (Figure 4). Helicase activity was severely impaired even for the R117A and R211A mutations that have lesser effects on ssDNA binding than the other mutations. These data confirm the importance of the residues we have highlighted in the interaction with the DNA substrate and consequently for the translocation mechanism. Consistent with the essential roles we propose, several of the corresponding residues in XPD, FANCJ, DDX11 or RTEL1 are reported to be disease-related (Figure 4). Indeed, mutations in these residues across the ssDNA-binding site in all three proteins are linked to human diseases (Suhasini and Brosh, 2013; Guo et al., 2016; Krassowski et al., 2018) and explain the molecular basis for the deficiencies in activity of these mutant proteins.

Figure 4 with 1 supplement see all
Download asset Open asset
Biochemical data for mutant proteins.

Residues contacting the 3’-end (A) or the 5’-end (B) of the bound ssDNA that were mutated. (C) Table showing ssDNA binding constants and ssDNA-dependent ATPase activity for wildtype DinG and mutants. The equivalent residues in T. acidophilus XPD (taXPD), human XPD (hsXPD), human DDX11 (hsDDX11), human FANCJ (hsFANCJ) and human RTEL1 (hsRTEL1) are listed and those highlighted in bold are directly linked to human disease. Residues in bold have been found to be mutated and confirmed important for protein function or diseased related (for human XPD, DDX11, FANCJ and RTEL1, the related disease is shown in brackets, ‘TTD’, Trichothiodystrophy, ‘XP’, Xeroderma Pigmentosum, ‘WBS’, Warsaw Breakage Syndrome, ‘UCEC’, Uterine Corpus Endometrial Carcinoma, ‘M’, Mixed cancer types, ‘BC’, Breast cancer, ‘BLCA’, Bladder Urothelial Carcinoma). This information was obtained from Krassowski et al. (2018) and Cerami et al. (2012). (D) Comparison of helicase activities of wildtype and mutant DinG proteins. Forked DNA substrate (50 nM) was incubated at various concentrations (0.1, 1 and 10 μM) of enzyme and the products were analysed on 12% native-PAGE. Heat-denatured substrate was run as a positive control.

https://doi.org/10.7554/eLife.42400.016

Implications for DNA damage recognition by XPD family helicases

The similarity of DinG to other XPD family members provides broader insight into the mechanism of this important class of proteins. Structural studies of XPD (Liu et al., 2008; Fan et al., 2008; Wolski et al., 2008; Kuper et al., 2012; Constantinescu-Aruxandei et al., 2016) have revealed detail about the overall structure of the protein and its interaction with ssDNA. However, only nucleotide-free states have been crystallised and also the observed interactions with bound ssDNA do not span the entire binding site and are not consistent between structures. Consequently, there is a limited understanding of how XPD interacts with DNA, the mechanism of translocation, how it unwinds DNA substrates or how it recognises DNA lesions.

The step size of one base also raises an intriguing possibility with regard to recognition of bulky DNA lesions by those members of the XPD family that recognise such lesions. A central component of the translocation is the requirement for single bases to flip out of a stacked arrangement as they collide with the rigid and hydrophobic P motif. One of the major lesions recognised by XPD is a thymidine dimer in which adjacent thymidine bases in one strand become crosslinked to one another. Such an arrangement now has consecutive bases locked together so would require them both to flip together with a step of two bases. Consequently, this would be blocked from flipping during translocation and this may constitute one aspect of the damage recognition process, causing the enzyme to stall at these lesions as a signal for repair to initiate. Indeed, consistent with this hypothesis, there is a large hydrophobic pocket on the protein surface in this location, particularly in the binary complex, that could accommodate a variety of bulky lesions such as thymidine dimers (Figures 2 and 5). Our structures show that the pocket dimensions alter during ATP binding and hydrolysis (Figure 5). If the pocket were occupied with a bulky lesion then domain closure would be prevented and hence ATPase would be blocked and translocation would stall (Figure 5—figure supplement 1). Although further work is required to confirm the effects of these changes on enzyme stalling, it has been shown that XPD stalls at bulky lesions on the translocating strand (Wirth et al., 2016).

Figure 5 with 1 supplement see all
Download asset Open asset
A pocket underneath the ssDNA binding site.

The molecular surface of (A) the binary complex and (B) the ternary complex, coloured by electrostatic charge (negative in red, positive in blue). In the binary complex, a large hydrophobic pocket is present between the ATPase domains, beneath the ssDNA binding site that becomes restricted when ATP binds to the enzyme. The open pocket in the binary complex could accommodate much larger DNA lesions (e.g. cyclopyrimidine dimers) that could not be accommodated in the ternary complex, suggesting that binding of large lesions in this pocket would prevent ATP-induced conformational changes and thereby block translocation. Protein contact potential was calculated in PyMOL and shown as a surface with negative potential coloured in red and positive in blue. The ssDNA strand was omitted from the calculations but is overlaid as sticks.

https://doi.org/10.7554/eLife.42400.018

Our data here expand upon our knowledge of the ssDNA binding site and a mechanism for translocation by these enzymes, as well as supporting a potential mechanism for recognition of base lesions. The roles of different residues in the ssDNA binding site have additional implications for known disease-causing mutations in XPD and other family members. Furthermore, we have mutated several residues in the ssDNA-binding site and several of these have equivalents in XPD, FANCJ, DDX11 or RTEL1 that are associated with human diseases. We can now ascribe a function to these residues to explain the basis of the malfunction of these proteins and expand our understanding of diseases related to mutations in these XPD family members.

Materials and methods

Key resources table
Reagent type
(species) or resource		DesignationSource or referenceIdentifiersAdditional information
Gene
(E. Coli, strain K12)		DinGNANCBI-GeneID: 945431
Strain, strain
background
(E. coli BL21 (DE3))		E. coli BL21 (DE3)NEBC2527H
Recombinant
DNA reagent		Plasmid pET28-HMT
-DinG (encoding full
length E.coli DinG, fused
with N-terminal 6 × His tag,
MBP-tag and a TEV
protease recognition site)		This study
Commercial
assay or kit		QuikChangeTM
Site-Directed
Mutagenesis Kit		Stratagene, La
Jolla, CA		La Jolla, CA
Chemical
compound, drug		PEG4000sigma
Chemical
compound, drug		PEG3350sigma
Chemical
compound, drug		BeCl3sigma
Chemical
compound, drug		phosphoenolpyruvatesigma
Chemical
compound, drug		pyruvate kinasesigma
Chemical
compound, drug		lactate
dehydrogenase		sigma
Chemical
compound, drug		ATPsigma
Chemical
compound, drug		ADPsigma
Software,
algorithm		COOT(Emsley et al., 2010)https://www2.mrc-lmb.cam.ac.uk/personal/pemsley/coot/
Software,
algorithm		Phenix(Adams et al., 2010)https://www.phenix-online.org/documentation/reference/real_space_refine.html
Software,
algorithm		Refmac(Murshudov et al., 1997)NA
Software,
algorithm		Xia2(Winter et al., 2013)NA
Software,
algorithm		PyMOLSchrodinger, LLChttps://pymol.org/2/
Software,
algorithm		NT Analysis
software (NanoTemper)		NAhttps://nanotempertech.com/monolith/

Cloning and site-directed mutagenesis

Request a detailed protocol

A list of primers used for cloning and mutagenesis is provided in Supplementary file 1. The full-length gene encoding E. coli DinG (residues 1-716aa) was amplified from E. coli genomic DNA by PCR and cloned into a modified pET28a expression vector, pET28-HMT, which contains a fused N-terminal 6 × His tag, a MBP-tag and a TEV protease recognition site (His-MBP-TEV). Site directed mutagenesis was performed with a QuikChangeTM Site-Directed Mutagenesis Kit from Stratagene (La Jolla, CA) as described previously (Cheng et al., 2016). The fidelity of the mutants was confirmed by sequencing. Primers used for cloning and mutageneses are listed in Supplementary file 1.

Protein preparation

Request a detailed protocol

In brief, transformed E. coli BL21 (DE3) clones were grown at 37°C in LB medium containing 50 μg/ml Kanamycin to an optical density at 600 nm of 0.6–0.8. Protein expression was induced at 30°C for 5 hr by adding isopropyl-β-D- thiogalactopyranoside (IPTG) to a final concentration of 0.8 mM. After harvesting, cells were re-suspended in lysis buffer (20 mM Tris (pH 8.0), 1 M NaCl, 0.5 mM TCEP and 5 mM imidazole), lysed by sonication and centrifuged at 18,000 × g for 60 min at 4°C. The supernatant was purified on a HisTrap HP column (GE Healthcare, Fairfield, CT), equilibrated with buffer A (20 mM Tris (pH 8.0), 1 M NaCl, 0.5 mM TCEP and 5 mM imidazole), washed with 30 mM imidazole and finally eluted with 300 mM imidazole. After TEV-tag-removal using TEV protease, the protein was dialysed into buffer B (20 mM Tris (pH 8.0), 250 mM NaCl and 0.5 mM TCEP) and re-loaded onto the HisTrap HP column (GE Healthcare) to remove the uncleaved protein. The flow-through fractions were collected and loaded onto a Heparin HP column (GE Healthcare) pre-equilibrated with buffer B. Fractions containing DinG protein was eluted with a linear gradient from 250 mM to 600 mM NaCl. The protein was finally purified by Superdex 200 10/300 GL column (GE Healthcare) with buffer C (20 mM Tris (pH 8.0), 300 mM NaCl and 0.5 mM TCEP). The mutant variants were expressed and purified in the same way. The concentration of proteins was estimated using the extinction coefficient of 78858 M–1 cm–1 at 278 nm, calculated from the amino acid sequence with the ProtParam tool (www.expasy.ch/tools/protparam.html).

Crystallization and structure determination

Request a detailed protocol

Crystallization trials were carried out by the sitting drop vapour diffusion method at 293 K. Freshly purified DinG was concentrated to ~4 mg/ml and centrifuged to remove insoluble material before crystallization. To make the DinG-ssDNA complex, protein was mixed with 12 base oligo-dT in a 1:1.2 molar ratio. To prepare the ternary DinG-DNA- ADP•BeF3 complex, 10 mM MgAc, 2 mM ADP, 5 mM NaF and 2 mM BeCl3 were added sequentially and the mixture was incubated on ice for 10 min before setting up the crystallization. Crystals of DinG-ssDNA (binary complex) were obtained in 0.25 M KCl, 0.1M HEPES (pH 7.8) and 22% PEG 3350. Crystals of DinG-DNA-ADP•BeF3 (ternary complex) were obtained in 0.1M MgCl2, 0.1M HEPES (pH 7.0), and 15% PEG4000. Cryo-freezing was achieved as described previously (Cheng et al., 2016) before flash freezing in liquid nitrogen. X-ray diffraction data were collected on beamlines I04 and I04-1 at the Diamond synchrotron X-ray source, and were integrated and scaled with the Xia2 system (Winter et al., 2013). For selenium-labeled crystals, data were collected at two wavelengths, 0.9795 (peak) and 0.9860 (low remote).

Initial phases for the binary complex structure were calculated from the MAD data using PHENIX (Adams et al., 2010), and the MAD data analysis is shown in Table S2. The phases were applied to native data (Table 1) and this map was used for model building and refinement. The ternary complex structure was determined by molecular replacement using separated DinG domains from the binary model as search models in CCP4, followed by rigid-body refinement by Refmac5 (Murshudov et al., 1997). Structures were refined using PHENIX (Adams et al., 2010) and interspersed with manual model building using COOT (Emsley et al., 2010). Later stages of refinement utilized TLS group anisotropic B-factor refinement. The refined binary complex model contained one DinG molecule and one ssDNA molecule in the asymmetric unit, while the ternary complex model contained two DinG complexes in the asymmetric unit. All residues are in the most favourable and allowed regions of the Ramachandran plot. Structure figures were rendered in PyMOL. The statistics for data collection and model refinement are listed in Table 1.

DNA binding assays

Request a detailed protocol

Microscale thermophoresis (MST) was used to determine ssDNA-binding constants for DinG or mutant variants (Willhoft et al., 2016). Interactions were measured on a temperature equilibrated Monolith NT.115 (NanoTemper), using a 5’FAM fluorescent labelled 14 base oligo-dT. For each set of measurements, a two fold serial-dilution of wild-type DinG or mutant variants concentration (starting at 80 μM) was prepared at 2x final in MST-Buffer (25 mM Tris (pH 8.0), 200 mM NaCl, 1 mM DTT, 5% Glycerol), before mixing with an equal volume of DNA substrate (also in MST-Buffer), to yield a final substrate concentration of 10 nM. Samples were equilibrated for 15 min before being loaded into Monolith NT.115 MST Premium Coated capillaries (NanoTemper). The system was given a further 15 min to re-equilibrate at 25°C before beginning each scan. The LED power was set at 20% and the MST power was set to 80% for all experiments. All other settings were left at the default values. Measurements were carried out in triplicates and with different protein batches. Kd fitting was performed on the NT Analysis software (NanoTemper).

Helicase assays

Request a detailed protocol

Substrate for helicase assays was created by annealing equal amounts of ssDNA 5’- TTTTTTTTTTTTTTAATGTATAATGCGAGCACTGCTACAGCACGG-3’ and 5’- FAM-CCGTGCTGTAGCAGTGCTCGCATTATACATTTTTTTTTTTTTTTT-3’. This forked DNA substrate contains 14 free bases on each tail, and 31 base pairs of duplex. Typically, 50 nM substrate was incubated with various concentrations (0.1, 1 and 10 μM) of DinG in a 10 μl reaction mixture containing 50 mM Tris (pH 8.0), 200 mM NaCl, 10 mM MgCl2, 10 μg/ml bovine serum albumin, 1 mM DDT and 5% glycerol. Unwinding reactions were initiated by adding 2.5 mM ATP, followed by incubation at 37°C for 20 min and finally stopped by the addition of protease K, SDS and EDTA to 0.1 mg/ml, 1% and 10 mM, respectively. Each set of experiments contained a ‘positive control’ in which the helicase substrate was heat-denatured by incubation at 95°C for 5 min followed by quick chilling on ice. Products were separated in 12% TBE-native polyacrylamide gels. The gels were scanned and quantified in fluorescence mode (FAM) on a BIO-RAD ChemiDocTM MP imaging system.

ATPase assays

Request a detailed protocol

ATPase activity of DinG was measured using a coupled assay that monitors ADP release as described previously (Willhoft et al., 2016). A final concentration of 100 μM NADH, 0.5 mM phosphoenolpyruvate, 100 U/ml pyruvate kinase (Sigma), 20 U/ml lactate dehydrogenase (Sigma) were used in all reactions in a final volume of 50 μl. Typical reactions were conducted using 20 nM purified DinG with or without 200 nM 14 base oligo-dT and 1 mM ATP in reaction buffer (50 mM Tris (pH 8.0), 100 mM NaCl, 10 mM MgCl2, 10 μg/ml bovine serum albumin, 1 mM DDT). ATP solutions were made with a ratio of 1:1:2 ATP:MgCl2:Tris base. Reactions were conducted by mixing all components immediately prior to transferring to an Optiplate-384 Black Opaque 384-well microplate (Perkin Elmer) that had been pre-incubated at 37°C. Reactions were initiated with the injection of ATP using the built-in reagent injectors, which were pre-filled with 10 × ATP solution. Reactions were monitored fluorescently using an excitation of 335 nm and an emission of 469 nm at 37°C with a CLARIOstar microplate reader (BMG Labtech). The measurements were carried out in triplicate and with different protein batches. All reaction rates were determined from the initial linear rate and reaction kinetics were analysed assuming a Michaelis–Menten model.

Data availability

Diffraction data have been deposited in PDB under the accession codes 6FWR and 6FWS.

The following data sets were generated
    1. Cheng K
    2. Wigley DB
    (2018) Protein Data Bank
    ID 6FWR. Diffraction data from DNA translocation mechanism of an XPD family helicase.
    http://www.rcsb.org/structure/6FWR
    1. Cheng K
    2. Wigley DB
    (2018) Protein Data Bank
    ID 6FWS. Diffraction data from DNA translocation mechanism of an XPD family helicase.
    http://www.rcsb.org/structure/6FWS

References

    1. Emsley P
    2. Lohkamp B
    3. Scott WG
    4. Cowtan K
    (2010) Features and development of Coot
    Acta Crystallographica. Section D, Biological Crystallography 66:486–501.
    https://doi.org/10.1107/S0907444910007493
    1. Winter G
    2. Lobley CM
    3. Prince SM
    (2013) Decision making in xia2
    Acta Crystallographica. Section D, Biological Crystallography 69:1260–1273.
    https://doi.org/10.1107/S0907444913015308

Decision letter

  1. James M Berger
    Reviewing Editor; Johns Hopkins University School of Medicine, United States
  2. John Kuriyan
    Senior Editor; University of California, Berkeley, United States

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

Thank you for submitting your article "DNA translocation mechanism of an XPD family helicase" for consideration by eLife. Your article has been reviewed by two peer reviewers, and the evaluation has been overseen by a Reviewing Editor and John Kuriyan as the Senior Editor. The following individual involved in review of your submission has agreed to reveal his identity: Kevin Raney (Reviewer #1).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

The present work describes the structure of the DinG helicase, a member of the XPD family of helicases, bound to a ssDNA dT fragment (12 bases) in the absence and presence of ADP·BeF3. The structure provides important insights on how DinG and other members of the XPD family helicases translocate on ssDNA. Although the observations are similar overall to previous studies made by Wigley et al. suggesting a 1 base pair step size, the molecular details for movement of the ssDNA through the active site differs for DinG compared to other SF1 and 3'-to-5' SF2 helicases. The authors also present a model for how XPD family helicases may recognize bulky lesions based on their structure. The mechanistical impact of the data is significant for the field, as it represents a sub-family that has thus far eluded description in terms of translocation mechanism.

Essential revisions:

The reviewers have raised several questions that must be addressed in a revised manuscript. Only the first point, below, requires new experimental work.

1) The biochemical section should be completed with ATPase studies for the different variants. It is acknowledged that DNA binding defects can interfere with such an analysis; however, the Arg117 and Arg211 variants should be analyzed in particular, since there are also comparative data available for XPD. DDX11 is not discussed here despite its presence in the analysis.

2) The manuscript needs a more in-depth analysis of the data in the context of the current literature on iron sulfur cluster-containing helicases of the XPD family. In the Abstract two other family members, RTEL1 and FANCJ, are described as other prominent members of this family. However, DDX11 is not mentioned, yet is added in a table where DinG variants were analyzed.

The work points out that DinG is also involved in DNA repair in E. coli; however, the role of DinG in repair is not explained. Several works (e.g., Boubakri et al., 2012; Hwang et al., 2012; Grodick et al., 2014) should be considered as input for DinG function in DNA repair. It could then also be appreciated that DinG function in repair is not at all comparable to XPD or FANCJ. In addition, helicases like DDX11 or RTEL1 are not repair helicases per se that have a special function in a repair pathway as compared to XPD or FANCJ. It is important that a differentiated Introduction be provided that describes the major players concisely but sufficiently to appreciate the roles they assume in genome maintenance.

Some specific points along this theme include:

a) In the subsection “Overall structure of DinG and comparison with XPD”, the differences in the FeS domains of XPD and DinG should be described rather than just mentioning that they are different; a possible influence on DNA binding should also be at least considered. The two arginines in DinG could be more similar to the situation in MutY rather than XPD. Is there anything useful to be gained from considering MutY?

b) In the second paragraph of the subsection “Structure of the DinG binary complex with ssDNA”, it is mentioned that are differences between the structures of XPD and DinG, but these are not described nor are superpositions shown. The corresponding figure in the supplement in which XPD within TFIIH is superimposed with DinG is not very instructive, since it is not mentioned in the text at this position (Figure 1—figure supplement 2) and it does not explain what the significance of the differences is.

c) In the last paragraph of the subsection “Structure of the DinG binary complex with ssDNA”, differences between the FeS domains of XPD and DinG are again mentioned; however, it is difficult to discern in the sequence alignment where the residues are and how the differences manifest on a structural level. A detailed figure illustrating the comparative analysis, perhaps showing a superposition, should be implemented. The figure in the supplement is not very descriptive and the different perspective suggests that there is no significant homology between the FeS domains. Is this the case? If so, it raises questions about the relevance of the findings here for XPD and possibly for other family members as well. Similarly, in XPD, R88 also directly interacts with the FeS cluster, which does not seem to be the case for DinG. Can this be attributed to DNA binding or is this is a real structural difference? The answer may have implications for the generality of some of the findings presented here.

3) A more detailed comparative analysis of the current work with previous translocation mechanisms for other SF1 and SF2 helicases is needed. For example, in the last paragraph of the subsection “A mechanism for translocation of ssDNA in a 5’-3’ direction”, it is proposed that the DinG translocation mechanism is different to other mechanisms, these differences are not elaborated. Similarly, it is stated that the work here completes the mechanisms of translocation for SF1 and SF2 helicases in all directionalities; however, a detailed comparison of the data to previous data is lacking. As a consequence, a general reader cannot appreciate these differences nor the uniqueness of the findings.

4) Many of the mutants discussed in Figure 4 are disease related and should be more thoroughly explained. It should be delineated in which disease specific mutations are implicated and the disease should be explained briefly but correctly. Not all of the mutations mentioned are implicated in cancer as stated in the subsection “Biochemical data in support of the model”, (R112 for instance causes TTD). Warsaw breakage syndrome (DDX11) is not associated with cancerous malignancies. It is important that the basis and possible outcome of every variant analyzed for the respective protein be discussed and explained in the correct context.

5) Regarding damage recognition, has it been shown that all family members of the XPD family come across bulky lesions in their respective pathways? For example, the thymidine dimer mentioned is a very specific nucleotide excision repair (NER) substrate. Since there is no mention of NER in the manuscript, what are the implications here? DDX11, FANCJ and RTEL1 likely do not come across such a lesion in vivo. In prokaryotes, the UvrABC system takes care of NER, so the likelihood that DinG might verify such a lesion is questionable. There are no biochemical data testing a role for the pocket in damage sensing; if there likewise is no sequence or structural conservation of pocket residues within XPD, then this speculation should probably be omitted.

References:

Grodick MA, Segal HM, Zwang TJ, Barton JK. DNA-Mediated Signaling by Proteins with 4Fe–4S Clusters Is Necessary for Genomic Integrity. J. Am. Chem. Soc., 2014, 136 (17), pp 6470–6478. DOI: 10.1021/ja501973c

Hwang J, Lee K, Phadtare S, Inouye M. Identification of Two DNA Helicases UvrD and DinG as Suppressors for Lethality Caused by Mutant cspA mRNAs. J Mol Microbiol Biotechnol, 2012, 22:135–146. DOI: 10.1159/000339832

https://doi.org/10.7554/eLife.42400.027

Author response

Essential revisions:

The reviewers have raised several questions that must be addressed in a revised manuscript. Only the first point, below, requires new experimental work.

1) The biochemical section should be completed with ATPase studies for the different variants. It is acknowledged that DNA binding defects can interfere with such an analysis; however, the Arg117 and Arg211 variants should be analyzed in particular, since there are also comparative data available for XPD. DDX11 is not discussed here despite its presence in the analysis.

As requested, we have added the ATPase activity data, notably for R117A and R211A (Figure 4) to the manuscript although we still feel their value is limited given the inability to saturate with ssDNA in most cases. We have corrected the disease references in the figure legend for all of the proteins involved including DDX1.

2) The manuscript needs a more in-depth analysis of the data in the context of the current literature on iron sulfur cluster-containing helicases of the XPD family. In the Abstract two other family members, RTEL1 and FANCJ, are described as other prominent members of this family. However, DDX11 is not mentioned, yet is added in a table where DinG variants were analyzed.

We have added more information about DDX11 in the Abstract and Introduction of the revised manuscript.

The work points out that DinG is also involved in DNA repair in E. coli; however, the role of DinG in repair is not explained. Several works (e.g., Boubakri et al., 2012; Hwang et al., 2012; Grodick et al., 2014) should be considered as input for DinG function in DNA repair. It could then also be appreciated that DinG function in repair is not at all comparable to XPD or FANCJ. In addition, helicases like DDX11 or RTEL1 are not repair helicases per se that have a special function in a repair pathway as compared to XPD or FANCJ. It is important that a differentiated Introduction be provided that describes the major players concisely but sufficiently to appreciate the roles they assume in genome maintenance.

Unfortunately, the precise role of DinG in bacteria is unclear other than a general involvement in DNA repair as judged by mainly genetic studies. Indeed, the name of the gene (DinG) is DNA damage Inducible as it was identified in a genetic screen for DNA damage repair along with several other “Din” genes (e.g. DinB – a bypass polymerase). Some recent studies do indicate a role in R-loop removal and we have now added these to the Introduction. However, we note that many helicases have multiple roles in cells (e.g. UvrD) that can have specific roles in DNA damage repair pathways with other proteins of that pathway as well as roles in replication fork repair or transcription. It may well be that DinG also has multiple roles in cells, but that remains unclear at present. We hope our revised Introduction makes that more explicit now.

Some specific points along this theme include:

a) In the subsection “Overall structure of DinG and comparison with XPD”, the differences in the FeS domains of XPD and DinG should be described rather than just mentioning that they are different; a possible influence on DNA binding should also be at least considered. The two arginines in DinG could be more similar to the situation in MutY rather than XPD. Is there anything useful to be gained from considering MutY?

We now provide additional details of the comparison between different FeS domains both in the revised figure (Figure 1—figure supplement 3) and in the revised manuscript. The main purpose of the comparison is the interaction with basic residues that occupy similar positions relative to the bound ssDNA despite coming from different parts of the FeS domains.

Since the structure of MutY is quite different to these enzymes, and the interaction is with dsDNA rather than ssDNA, we do not feel a comparison is useful.

b) In the second paragraph of the subsection “Structure of the DinG binary complex with ssDNA”, it is mentioned that are differences between the structures of XPD and DinG, but these are not described nor are superpositions shown. The corresponding figure in the supplement in which XPD within TFIIH is superimposed with DinG is not very instructive, since it is not mentioned in the text at this position (Figure 1—figure supplement 2) and it does not explain what the significance of the differences is.

In the second paragraph of the subsection “Structure of the DinG binary complex with ssDNA”, we are comparing the differences between previously published XPD-DNA complex structures (Kuper et al., 2012; Constantinescu-Aruxandei et al., 2016) which are not consistent with one another. No reference is made to Figure 1—figure supplement 2 at this point. However, we have added an additional panel to Figure 2—figure supplement 3 to show the differences between the two XPD-DNA complex structures (now panels B and C).

c) In the last paragraph of the subsection “Structure of the DinG binary complex with ssDNA”, differences between the FeS domains of XPD and DinG are again mentioned; however, it is difficult to discern in the sequence alignment where the residues are and how the differences manifest on a structural level. A detailed figure illustrating the comparative analysis, perhaps showing a superposition, should be implemented. The figure in the supplement is not very descriptive and the different perspective suggests that there is no significant homology between the FeS domains. Is this the case? If so, it raises questions about the relevance of the findings here for XPD and possibly for other family members as well. Similarly, in XPD, R88 also directly interacts with the FeS cluster, which does not seem to be the case for DinG. Can this be attributed to DNA binding or is this is a real structural difference? The answer may have implications for the generality of some of the findings presented here.

We have revised Figure 1—figure supplement 3 to compare the FeS domains of DinG and XPD. We superimpose structures of the FeS domains of other XPD family members (Thermoplasma acidophilum XPD, PDB ID: 4A15; Saccharomyces cerevisiae Rad3 (XPD), PDB ID: 5OQM; Sulfolobus acidocaldarius XPD, PDB ID: 3CRV; Homo sapiens XPD, PDB ID: 5OF4; and DinG in this study). The point we are trying to make here is that although the structures of the FeS domains differ, the basic residues that contact the bound ssDNA in DinG are conserved in space and these residues are conserved between other XPD family members that do have more similar FeS domains. This suggests they may play a similar role in interacting with ssDNA. The mutational data in Figure 4 confirm the role in binding ssDNA in DinG. These residues are of unknown function in XPD, RTEL1, FANCJ and DDX11 yet crop up as disease causing mutations in these proteins. In XPD, mutations at these residues affect DNA binding but by an unknown manner. Our structural model of DinG now suggests a potential role for these residues in other XPD family members which is why we felt this was a useful addition to the analysis.

Although true that R88 in taXPD interacts directly with the FeS cluser, that is not the case for other XPD structures. The bound ssDNA substrate in the taXPD complex is too short to reach to R88 so the interaction with the FeS cluster observed here (which is not seen in other XPD structures) may have little relevance.

3) A more detailed comparative analysis of the current work with previous translocation mechanisms for other SF1 and SF2 helicases is needed. For example, in the last paragraph of the subsection “A mechanism for translocation of ssDNA in a 5’-3’ direction”, it is proposed that the DinG translocation mechanism is different to other mechanisms, these differences are not elaborated. Similarly, it is stated that the work here completes the mechanisms of translocation for SF1 and SF2 helicases in all directionalities; however, a detailed comparison of the data to previous data is lacking. As a consequence, a general reader cannot appreciate these differences nor the uniqueness of the findings.

To address this point, we provide a new figure (Figure 3—figure supplement 1) to compare the translocation mechanisms for SF1 and SF2 helicases. Details of each mechanism would be much too complicated to present but the figure legend addresses the major differences between each mode of translocation.

4) Many of the mutants discussed in Figure 4 are disease related and should be more thoroughly explained. It should be delineated in which disease specific mutations are implicated and the disease should be explained briefly but correctly. Not all of the mutations mentioned are implicated in cancer as stated in the subsection “Biochemical data in support of the model”, (R112 for instance causes TTD). Warsaw breakage syndrome (DDX11) is not associated with cancerous malignancies. It is important that the basis and possible outcome of every variant analyzed for the respective protein be discussed and explained in the correct context.

As requested, we have corrected the references to the correct disease types for the point mutations in the table (Figure 4C) with references to appropriate databases for further information.

5) Regarding damage recognition, has it been shown that all family members of the XPD family come across bulky lesions in their respective pathways? For example, the thymidine dimer mentioned is a very specific nucleotide excision repair (NER) substrate. Since there is no mention of NER in the manuscript, what are the implications here? DDX11, FANCJ and RTEL1 likely do not come across such a lesion in vivo. In prokaryotes, the UvrABC system takes care of NER, so the likelihood that DinG might verify such a lesion is questionable. There are no biochemical data testing a role for the pocket in damage sensing; if there likewise is no sequence or structural conservation of pocket residues within XPD, then this speculation should probably be omitted.

We have been careful not to suggest that DinG plays the same role in bacteria as XPD in eukaryotes, nor indeed the same functions as other XPD family members (RTEL1, DDX11 and FANCJ) which likely differ from one another. However, the similarity between the structures of XPD (the only other family member for which there is currently structural information) suggests a commonality across all family members in the mechanism of 5’-3’ translocation on ssDNA, with potentially an additional similarity to some family members (e.g. XPD) in an ability to recognise bulky DNA lesions. The downstream repair pathways may be, and indeed are almost certainly, quite different in bacteria and eukaryotes and between different systems. Indeed, as the reviewers point out, the roles of different family members (c.f. XPD and RTEL1) are likely quite different even within eukaryotes. We have pointed out the potential for a mechanism of bulky DNA lesion recognition and response by XPD and DinG. However, this may not be required by other family members, and for clarity we have now stated that in the Introduction. By contrast, the mechanism for ssDNA translocation in a 5’-3’ direction, which is the main point of this study, is likely to be conserved.

https://doi.org/10.7554/eLife.42400.028

Article and author information

Author details

  1. Kaiying Cheng

    Section of Structural Biology, Department of Medicine, Imperial College London, London, United Kingdom
    Contribution
    Conceptualization, Investigation, Writing—original draft
    Competing interests
    No competing interests declared

  2. Dale B Wigley

    Section of Structural Biology, Department of Medicine, Imperial College London, London, United Kingdom
    Contribution
    Conceptualization, Funding acquisition, Writing—original draft, Writing—review and editing
    For correspondence
    d.wigley@imperial.ac.uk
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-0786-6726

Funding

Medical Research Council

  • Dale B Wigley

Cancer Research UK

  • Dale B Wigley

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

The work was funded by the Medical Research Council and Cancer Research UK (DBW). We thank Rhodri ML Morgan for assistance with data collection and Martin Wilkinson for advice during model building. We acknowledge the Diamond Light Source for access to beamlines I04 and I04-1. We acknowledge use of the crystallisation facility at Imperial College London, which is part-funded by the Wellcome Trust.

Senior Editor

  1. John Kuriyan, University of California, Berkeley, United States

Reviewing Editor

  1. James M Berger, Johns Hopkins University School of Medicine, United States

Publication history

  1. Received: September 27, 2018
  2. Accepted: December 5, 2018
  3. Accepted Manuscript published: December 6, 2018 (version 1)
  4. Version of Record published: December 19, 2018 (version 2)

Copyright

© 2018, Cheng and Wigley

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 2,558
    Page views
  • 443
    Downloads
  • 23
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Kaiying Cheng
  2. Dale B Wigley
(2018)
DNA translocation mechanism of an XPD family helicase
eLife 7:e42400.
https://doi.org/10.7554/eLife.42400

Categories and tags

Research organism

Further reading

    1. Structural Biology and Molecular Biophysics

    The prolactin receptor scaffolds Janus kinase 2 via co-structure formation with phosphoinositide-4,5-bisphosphate

    Raul Araya-Secchi, Katrine Bugge ... Birthe B Kragelund
    Research Article

    Class 1 cytokine receptors transmit signals through the membrane by a single transmembrane helix to an intrinsically disordered cytoplasmic domain that lacks kinase activity. While specific binding to phosphoinositides has been reported for the prolactin receptor (PRLR), the role of lipids in PRLR signalling is unclear. Using an integrative approach combining NMR spectroscopy, cellular signalling experiments, computational modelling and simulation, we demonstrate co-structure formation of the disordered intracellular domain of the human PRLR, the membrane constituent phosphoinositide-4,5-bisphosphate (PI(4,5)P2) and the FERM-SH2 domain of the Janus kinase 2 (JAK2). We find that the complex leads to accumulation of PI(4,5)P2 at the transmembrane helix interface and that mutation of residues identified to interact specifically with PI(4,5)P2 negatively affects PRLR-mediated activation of signal transducer and activator of transcription 5 (STAT5). Facilitated by co-structure formation, the membrane-proximal disordered region arranges into an extended structure. We suggest that the co-structure formed between PRLR, JAK2 and PI(4,5)P2 locks the juxtamembrane disordered domain of the PRLR in an extended structure, enabling signal relay from the extracellular to the intracellular domain upon ligand binding. We find that the co-structure exists in different states which we speculate could be relevant for turning signalling on and off. Similar co-structures may be relevant for other non-receptor tyrosine kinases and their receptors.

    1. Biochemistry and Chemical Biology
    2. Structural Biology and Molecular Biophysics

    The structural basis of the multi-step allosteric activation of Aurora B kinase

    Dario Segura-Peña, Oda Hovet ... Nikolina Sekulic
    Research Article

    Aurora B, together with IN-box, the C-terminal part of INCENP, forms an enzymatic complex that ensures faithful cell division. The [Aurora B/IN-box] complex is activated by autophosphorylation in the Aurora B activation loop and in IN-box, but it is not clear how these phosphorylations activate the enzyme. We used a combination of experimental and computational studies to investigate the effects of phosphorylation on the molecular dynamics and structure of [Aurora B/IN-box]. In addition, we generated partially phosphorylated intermediates to analyze the contribution of each phosphorylation independently. We found that the dynamics of Aurora and IN-box are interconnected, and IN-box plays both positive and negative regulatory roles depending on the phosphorylation status of the enzyme complex. Phosphorylation in the activation loop of Aurora B occurs intramolecularly and prepares the enzyme complex for activation, but two phosphorylated sites are synergistically responsible for full enzyme activity.

    1. Neuroscience
    2. Structural Biology and Molecular Biophysics

    Cannabidiol sensitizes TRPV2 channels to activation by 2-APB

    Aaron Gochman, Xiao-Feng Tan ... Andres Jara-Oseguera
    Research Article

    The cation-permeable TRPV2 channel is important for cardiac and immune cell function. Cannabidiol (CBD), a non-psychoactive cannabinoid of clinical relevance, is one of the few molecules known to activate TRPV2. Using the patch-clamp technique, we discover that CBD can sensitize current responses of the rat TRPV2 channel to the synthetic agonist 2-aminoethoxydiphenyl borate (2-APB) by over two orders of magnitude, without sensitizing channels to activation by moderate (40°C) heat. Using cryo-EM, we uncover a new small-molecule binding site in the pore domain of rTRPV2 in addition to a nearby CBD site that had already been reported. The TRPV1 and TRPV3 channels are also activated by 2-APB and CBD and share multiple conserved features with TRPV2, but we find that strong sensitization by CBD is only observed in TRPV3, while sensitization for TRPV1 is much weaker. Mutations at non-conserved positions between rTRPV2 and rTRPV1 in either the pore domain or the CBD sites failed to confer strong sensitization by CBD in mutant rTRPV1 channels. Together, our results indicate that CBD-dependent sensitization of rTRPV2 channels engages multiple channel regions, and that the difference in sensitization strength between rTRPV2 and rTRPV1 channels does not originate from amino acid sequence differences at the CBD binding site or the pore domain. The remarkably robust effect of CBD on TRPV2 and TRPV3 channels offers a promising new tool to both understand and overcome one of the major roadblocks in the study of these channels – their resilience to activation.