Structure

Article

TheStructureof theXPF-ssDNAComplexUnderscoresthe Distinct Roles of the XPF and ERCC1 Helix-Hairpin-Helix Domains in ss/ds DNA RecognitionDevashish Das,1,3 Gert E. Folkers,1 Marc van Dijk,1 Nicolaas G.J. Jaspers,2 Jan H.J. Hoeijmakers,2 Robert Kaptein,1

and Rolf Boelens1,*1Bijvoet Center for Biomolecular Research, Utrecht University, Padualaan 8, 3584 CH Utrecht, The Netherlands2Department of Cell Biology and Genetics, Erasmus MC, P.O. Box 2040, 3000 CA Rotterdam, The Netherlands3Department of Radiology, Radboud University Nijmegen Medical Center, Geert Grooteplein 10, 6525 GA, Nijmegen, The Netherlands

*Correspondence: r.boelens@uu.nl

DOI 10.1016/j.str.2012.02.009

SUMMARY

Human XPF/ERCC1 is a structure-specific DNAendonuclease that nicks the damaged DNA strandat the 50 end during nucleotide excision repair.We determined the structure of the complex of theC-terminal domain of XPF with 10 nt ssDNA. A posi-tively charged region within the second helix of thefirst HhH motif contacts the ssDNA phosphate back-bone. One guanine base is flipped out of register andpositioned in a pocket contacting residues from bothHhH motifs of XPF. Comparison to other HhH-con-taining proteins indicates a one-residue deletionin the second HhH motif of XPF that has alteredthe hairpin conformation, thereby permitting ssDNAinteractions. Previous nuclear magnetic resonancestudies showed that ERCC1 in the XPF-ERCC1 het-erodimer can bind dsDNA. Combining the two obser-vations gives amodel that underscores the asymme-try of the human XPF/ERCC1 heterodimer in bindingat an ss/ds DNA junction.

INTRODUCTION

Ultraviolet (UV) radiation, chemical carcinogens, and cellular

metabolites are important contributors of DNA damage in cells

(Hoeijmakers, 2001). Of particular importance is the mammalian

nucleotide excision repair (NER) pathway and associated gene

products, which coordinate the elimination of UV-radiation-

induced DNA lesions, like CPDs and 6-4 photoproducts (de

Laat et al., 1999; Scharer, 2007). The NER pathway entails a

collective action of at least 25 proteins that are assembled at

the DNA lesion (de Laat et al., 1999; Riedl et al., 2003; Volker

et al., 2001). Two proteins, ERCC1 and XPF, form a very stable

complex. The role of the ERCC1/XPF complex is well character-

ized (Bessho et al., 1997; de Laat et al., 1998a, 1998c; Enzlin and

Scharer, 2002; Gaillard and Wood, 2001; Houtsmuller et al.,

1999; Park et al., 1995; Sijbers et al., 1996; Tsodikov et al.,

2005), demonstrating that the complex functions as a struc-

ture-specific DNA endonuclease that nicks at the 50 end of the

Structure 20

damaged DNA strand during nucleotide excision repair (de

Laat et al., 1999; Riedl et al., 2003; Scharer, 2007; Sijbers

et al., 1996), leading to the excision of 24 to 32 nucleotides after

concomitant incision 30 to the damage by XPG (de Laat et al.,

1999; Scharer, 2007). In NER, XPF/ERCC1 is the last factor to

be assembled at the 50 end of the DNA lesion (Houtsmuller

et al., 1999; Riedl et al., 2003; Volker et al., 2001), where it inter-

acts with other NER proteins (de Laat et al., 1998b; Orelli et al.,

2009; Tripsianes et al., 2007; Tsodikov et al., 2007; Winkler

et al., 2001). The crucial role of XPF/ERCC1 in NER follows

from observations that both ERCC1 (McWhir et al., 1993; Weeda

et al., 1997) and XPF (Tian et al., 2004) knockoutmice showmore

severe phenotypes compared to mice lacking other NER genes

(Melis et al., 2008). Along with the well-established role of XPF/

ERCC1 in the NER pathway, additional roles in DNA processing

have been described, including interstrand cross-link repair

(Kuraoka et al., 2000; Niedernhofer et al., 2004), telomere length

control (Munoz et al., 2005;Wu et al., 2008), homologous recom-

bination repair (Ahmad et al., 2008; Clingen et al., 2005), and

gene-targeting events (Al-Minawi et al., 2008; Niedernhofer

et al., 2001). The precise DNA recognition in these mechanisms

is structurally not understood.

The C-terminal HhH domains of the human XPF and ERCC1

proteins are required and sufficient for stable heterodimer forma-

tion (de Laat et al., 1998c). Structure determination of the

minimal dimerization domains of the two proteins showed the

presence of two canonical helix-hairpin-helix motifs (HhH)

related by a pseudo-2-fold symmetry axis (Tripsianes et al.,

2005; Tsodikov et al., 2005). We and others independently re-

ported that ERCC1 could only fold in the presence of XPF, the

latter acting as a scaffold (Choi et al., 2005; Tripsianes et al.,

2005). Whereas the C-terminal domain of ERCC1 was insoluble

in isolation, the C-terminal domain of XPFwas highly soluble. We

subsequently showed that this XPF domain forms a homodimer

that is significantly more stable than is the heterodimeric XPF/

ERCC1 complex. Nuclear magnetic resonance (NMR) structural

analysis provided a molecular explanation for the higher stability

in vitro (Das et al., 2008). NMR chemical shift perturbations found

by the addition of a hairpin substrate with 22 unpaired nucleo-

tides (H22) showed that the two hairpins of the HhH domain of

ERCC1 were primarily affected by the addition of DNA (Trip-

sianes et al., 2005). Using the structural homology with other

HhH domain proteins (Shao and Grishin, 2000), including RuvA

, 667–675, April 4, 2012 ª2012 Elsevier Ltd All rights reserved 667

Structure

Structure of the XPF-ssDNA Complex

bound to a Holliday junction (Ariyoshi et al., 2000) and the

archaeal homodimeric XPF bound to DNA (Newman et al.,

2005), a model was proposed for the ERCC1-dsDNA complex

with a clear role for the two ERCC1 hairpins (Tripsianes et al.,

2005). This mode of DNA binding by the C-terminal ERCC1

HhH domain fits well with the presence of two canonical HhH

motifs. However, the role of the XPF HhH motifs in DNA binding

was less clear. The loss of the canonical hairpin structure of XPF

led us to hypothesize that XPF would recognize a different DNA

structure.

Using mass spectrometry, NMR spectroscopy, and in vitro

DNA binding assays, we show here that the XPF homodimer

forms a stable complex with single-strand DNA. We determined

the solution structure of the human XPF homodimer in complex

with ssDNA which shows that the DNA forms a right-handed

structure that wraps around the protein mainly via interactions

along the ssDNA phosphate backbone. Interestingly one

guanine base is flipped out of register and is positioned in a

hydrophobic cavity formed between canonical HhH1 and the

distorted HhH2. Combined with the previously established

ERCC1-dsDNA model (Tripsianes et al., 2005), this allowed us

to build a model for the complex of ERCC1/XPF, which explains

the positioning of this endonuclease at the ss/dsDNA junction.

RESULTS

ssDNA Binding by the C-terminal XPF HhH HomodimerPreviously, we reported the DNA binding activity of the HhH

domains of the XPF/ERCC1 heterodimer using electrophoretic

mobility shift assay (EMSA) experiments and NMR spectroscopy

(Tripsianes et al., 2005). The NMR spectra of the XPF/ERCC1

heterodimer in complexes with a DNA hairpin revealed signifi-

cant chemical shift perturbations for residues of the HhH motifs

of the ERCC1 protein indicative for dsDNA binding, whereas

smaller but consistent shifts were found for the XPF residues.

To investigate the significance of these XPF chemical shifts

perturbations, we decided to further analyze the ability of the

C-terminal XPF homodimer, (XPF)2 to bind to various DNA

substrates.

First, we performed electrophoretic mobility shift assays using

the same probes for the (XPF)2 homodimer as used previously for

the XPF/ERCC1 heterodimer (Tripsianes et al., 2005). Similar to

the XPF/ERCC1 heterodimer, the (XPF)2 homodimer has a sub-

micromolar affinity for dsDNA, containing a bubble. However,

in contrast to the ERCC1/XPF heterodimer, which has only

weak affinity for ssDNA in electrophoretic mobility shift assays

(Tripsianes et al., 2005), we noted that the (XPF)2 homodimer

binds to ssDNA substrates. For a long 39 nt ssDNA fragment it

even binds with a micromolar range affinity (Figure S5 available

online). The ability of (XPF)2 to form a stable complex in the pres-

ence of an excess of 39 nt ssDNA was also established using

mass spectrometry. Next to the presence of some free (XPF)2,

matrix-assisted laser desorption/ionization time of flight mass

spectrometry experiments (MALDI-TOF-MS) revealed the pres-

ence of a major ion signal at m/z 15433.33 attributable to a 2:1

XPF/39 nt ssDNA complex (Figure 1A). These results show that

the (XPF)2 homodimers can form a stable complex with a single

long single-strand DNA fragment. The 2:1 stoichiometry can be

explained by the fact that the 39 nt ssDNA can loop around

668 Structure 20, 667–675, April 4, 2012 ª2012 Elsevier Ltd All rights

(XPF)2 and different parts of the 39 nt ssDNA fragment bind to

the two DNA binding sites of the dimer.

For the NMR structure determination of the XPF-ssDNA

complex, a 10 nt substrate, CAGTGGCTGA, was used. Though

this fragment has much weaker affinity (estimated Kd > 50 mM),

it leads as shown below in the NMR experiments to a 2:2

complex, where the symmetry of the (XPF)2 homodimer is main-

tained, which has considerable advances for the NMR analysis.

The formation of the complex could be easily followed by 31P

NMR. Relatively large 31P chemical shift changes are noted for

the DNAphosphate backbone upon addition of the (XPF)2 homo-

dimer. For further analysis, the 31P and 1H spectra of the free and

bound ssDNA fragment resonance assignments were required.

For this, we recorded two-dimensional (2D) 1H-1H NOESY and

the 2D 31P-1H HMQC spectra (Figures S1A, S1B, and S2A).

The obtained 1H chemical shifts for free and (XPF)2 bound

CAGTGGCTGA are shown in the Tables S1 and S2. The largest

chemical shift changes upon binding can be noted for nucleo-

tides 3–7, suggesting that this region of the ssDNA fragment is

involved in XPF binding (Table S3 and Figure 1C).

For XPF, DNA binding is also evident by comparing the 1H-15N

HSQC NMR spectra of free and DNA bound (XPF)2 homodimers

(Figure 2A). The 15N-1H HSQC spectra show chemical shift

changes for a small subset of peaks, demonstrating the speci-

ficity of the XPF homodimer for ssDNA. Since the spectra

show a single set of peaks during the titrations, the two 10 nt

ssDNA fragments should bind symmetrically to the homodimers

in a 2:2 stoichiometry. The chemical shift perturbations (CSP) are

most prominent for residues of the second helix of the canonical

HhHmotif, namely, Lys850, Asn851, Ser854, Leu855, Met856, His857,

His858, and Val859 (Figure 2B).

NMR titrations of the XPF/ERCC1 heterodimer with the same

10 nt ssDNA show chemical shift perturbations for similar resi-

dues as observed for XPF homodimers (Figures 2B and S3).

Based on this similarity, the complex of the homodimer (XPF)2with ssDNA is anticipated to be a good model for a structural

analysis of ssDNA binding by XPF in the XPF/ERCC1 hetero-

dimer. Compared to the heterodimer, the homodimer-ssDNA

complex has considerable advantages for the NMR spectral

analysis and structure because of the symmetry, resulting in

a lower number of signals and reduced spectral overlap.

Structure of the XPF-10 nt ssDNA ComplexThe structure of the (XPF)2 homodimer bound to 10 nucleotide

single-strand DNA (CAGTGGCTGA) was determined on the

basis of 3184 distance restraints and 178 dihedral angle

restraints. A summary of all structural and restraints statistics

is given in Table 1. The ensemble backbone rmsd for the 20

conformers is 0.57A for the well-folded part of the protein and

comparable to that of 0.39 A for the free structure (Das et al.,

2008). Figure 3A shows an overlay of the final 20 lowest-energy

conformers after water refinement.

The structure of the (XPF)2 bound to two single DNA strands

show a similar fold as free (XPF)2 (Das et al., 2008), with a back-

bone rmsd of �0.69 A between the two. For residues 845–861

and 8450–8610, which are close to ssDNA, we note minor

changes in the conformation of the side chains (rmsd �1.14 A).

In the XPF protein, the residues, which are part of the conserved

canonical and noncanonical HhH motif, form a cavity (Das et al.,

reserved

A

B

C

10.0 20.0 30.0 40.0

m/z

Inte

nsity

500.0

1000.0

1500.0

(ssDNA)(ssDNA)

2+

(Free XPF)2

+

+

11998.29

23995.09

15433.33

18881.18

(XPF-ssDNA Complex) +

30887.69

2 +(XPF-ssDNA Complex)

37792.43

Free DNA

31P

XPF-DNA complex

−0.7 −0.8 −0.9 −1.0 −1.1 −1.2 −1.3 −1.4 −1.5 ppm1.82.02.22.42.6

Nucleotides

1 2 3 4 5 6 7 8 9 10

5’C

A

G

T

G

G

C

T

G A

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

H C

SP

Sum

/ N

ato

ms

1

kDa

Figure 1. The C-terminal XPF HhH Homodimer

Binds ssDNA

(A) MALDI-TOF (MS) spectrum of the XPF-ssDNA

complex. The signals are assigned based on the expected

mass/charge ratio of the free and bound species.

(B) Overlay of 1D phosphorus spectra, free (red) and

bound (blue) states of the (XPF)2-(CAGTGGCTGA)2complex, indicating changes in the phosphate backbone

in the free and bound forms of the ssDNA.

(C) Plot of the 1H chemical shift differences of the10 nt

ssDNA fragment in free and bound forms. The chemical

shift differences were summed for all atoms of each

nucleotide of the10 nt ssDNA and normalized by dividing

by the number of atoms.

See Figure S1 and Tables S1, S2, and S3 for additional

NMR data and Figure S5 for an EMSA assay of ssDNA

binding by (XPF)2.

Structure

Structure of the XPF-ssDNA Complex

2008; Tripsianes et al., 2005). In the DNA-bound structure, the

base of Gua5 is positioned in this cavity and has a large interac-

tion surface area of �140A2 with the protein (Figure 3B). Altered

conformations are seen for the Lys850 and Asn879 side chains in

the free and DNA-bound XPF. This altered side-chain conforma-

tion of Lys850 allows for a hydrogen bond contact to the phos-

phate group of Cyt7. The slight reorientation of the Asn879 side

chain results in a hydrogen bond with Gua5, as is observed in

�30% of the conformers of the final ensemble.

Figure 3B shows a closer view of the ssDNA in the complex.

The DNA binding surface shows a buried surface area of

640A2 for each subunit. For nucleotides 2–7, which contact the

protein, the phosphate backbone is well defined with a rmsd of

1.18 A, whereas the nucleotides 8–10 have no well-defined

structure. The nucleotide 3–7 region adopts a right-handed

helical structure (Figure 3B) as supported by strong sequential

Structure 20, 667–675, April 4

protein-DNA NOEs between the H10 ribose

protons of Gua5 and Gua6 and side-chain

protons of Lys850 and Arg853, respectively (Fig-

ure S2B). Analysis of the DNA geometry shows

that the right-handed helical structure has

a rather extended conformation, which is quite

distinct from B-DNA (Table S4 and Figure S4).

This geometry cannot be easily adapted in

dsDNA, which can certainly contribute to the

preference of XPF for ssDNA.

Interactions at Interface of theXPF-ssDNA ComplexFigures 3B and 4A show how the ssDNA is posi-

tioned on the surface of XPF. The arrangement

of the ssDNA on the protein is a result of several

electrostatic contacts between protein and

DNA, mainly at the phosphate backbone. For

example, the side chain of Lys850 interacts

with the phosphate group of Cyt7, and the

side chain of Arg853 contacts the phosphate

groups of both Gua5 and Gua6 in �80% of

the conformers. A summary of the protein-

DNA contacts is shown in Figure 4B. Interest-

ingly, the Ser854 hydroxyl group interacts with

the oxygen O50 atom of Thy4. The hydrogen bonding of the OH

of Ser854 to DNA is confirmed by a slow exchange of the Ser

OH proton with water leading to an observable NOE between

the OH and the Ser854 CH2 protons. Without hydrogen bonding

stabilization, such OH protons are normally not observed in solu-

tion NMR spectra. Additionally, the side chain of His857 makes

a hydrogen bond interaction with Gua3, whereas the His858

side chain contacts the phosphate group of Gua3. Consistent

with the observed interaction between His857 side chain and

ssDNA, we find an NOE between the His857 HD1 proton and

the side-chain beta-protons only in the protein-DNA complex

(Figure S2C).

Most of the DNA bases are sticking outwards. A remarkable

aspect of this protein-DNA complex is, however, the position

of the guanine base of Gua5. This base is flipped out of register

with respect to the other bases and positioned in a pocket of

, 2012 ª2012 Elsevier Ltd All rights reserved 669

Table 1. Structural Statistics of the (XPF)2-(CAGTGGCTGA)2Conformers

rmsd (A) with respect to mean (backbone)a

(XPF)2 0.57

(XPF-ssDNA)2 1.17

No. of experimental restraintsb

DNA

Intramolecular DNA-DNA NOE 192

Protein-DNA

Total intermolecular protein-DNA 42

Homodimeric XPF2

Intraresidue NOE 664

Sequential NOEs (ji � jj = 1) 778

Medium-range NOEs (1 < ji � jj < 4) 739

Long-range NOEs (ji � jj > 4) 514

Interprotein 255

Total NOEs 3,184

Dihedral angle restraints 178

Restraint violations (A)

NOE distances with violations >0.5 A 4.3 ± 0.6

Dihedrals with violations >5� 0

rmsd for experimental restraints (A)

All distance restraints (A) 0.07 ± 0.005

Torsion angles (�) 0.46 ± 0.03

rmsd from idealized covalent geometry

Bonds (A) 0.004 ± 0.000

Angels (�) 0.7 ± 0.0

Improper (�) 0.4 ± 0.0

CNS energies after water refinement

Evdw (kcal/mol) �755.7 ± 19.5

Eelec (kcal/mol) �5708.5 ± 105.4

Ramachandran analysis

Residues in the favored regions (%) 93.3 ± 1.5

Residues in additional allowed regions (%) 4.8 ± 1.6

Residues in generously allowed regions (%) 1.50 ± 0.5

Residues in disallowed regions (%) 0.04 ± 0.2armsd values (in A) were calculated for XPF (835-894) and ssDNA(2-7).bNumber of restraints for the dimer, as used in the calculations.

A

B

8.2 8.0 7.8 7.6 7.4 7.2

122

120

118

116 I890 S875

Q885

K884 D839 F840

A867

A863

T892M844

L886

L855

L865

R853

L868

V859A883

E872

I876

1H

15N

8.4 8.2 8.0 7.8

115.5

115.0

114.5

114.0

113.5

H858

S854

M856

H857

1 H

Free XPF

XPF-DNA complexR853 side chain

CSP

Residue

1.4

1.2

1.0

0.8

0

0.2

0.4

0.6 > average + 2* Std.devi

> average

> average + Std.devi

ppm

missing

1.4-0.5

0.5-0.3

0.3-0.2

0.2-0.02

DSETLPESEKYNPGPQDFLLKMPGVNAKNCRSLMHHVKNIAELAALSQDELTSILGNAANKQLYDFIHTSFAEVVSKGKG

α β γ δ ε

βα

γ

ε

δ

C

N

Figure 2. NMR Chemical Shift Perturbations of the C-terminal XPF

HhH Homodimer upon Addition of ssDNA

(A) Part of a 1H-15N heteronuclear single quantum correlation (HSQC)

spectrum of the (XPF)2-ssDNA complex (blue) and of the free XPF protein (red).

The arrows indicate the observed changes in the titrations of XPF with ssDNA.

(B) Chemical shift perturbation (CSP) upon 10 nt ssDNA binding. CSP values

plotted on the lowest-energy structure of the XPF subunit. CSP values

for the amide proton (HN) and nitrogen (N) are calculated as CSP=ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffið53DppmHNÞ2 + ðDppmNÞ2

q, where the chemical shift difference (D ppmHN)

and (D ppmN) are the proton and nitrogen nuclei in part per million, respec-

tively, for the free and bound states of the protein. The color scheme is from

gray (CSP < 0.2) to red (CSP > 0.48). Black indicates residues for which the

CSP values are missing.

See Figure S3 and Table S4 for 1H-15N HSQC spectrum of the XPF-ERCC1

heterodimer binding to 10 nt ssDNA.

Structure

Structure of the XPF-ssDNA Complex

XPF, an observation that is clearly supported by the unambig-

uous protein-DNA NOE between the Asn879 side-chain atom

(HB) and �NH2 protons of the Gua5 in 2D 1H-1H NOESY and

three-dimensional (3D) 13C-edited-1H-1H NOESY spectra (Fig-

ure S2B). In the NMR structure, the position of the guanine

base within this pocket is very well-defined and is stabilized by

van der Waals contacts and a H-bond of the NH2 group of

Asn879 with the O6 atom of Gua5.

DISCUSSION

Structural Comparison to dsDNA Binding by HhH MotifsIn contrast to XPF, most proteins containing HhH motifs bind

dsDNA (Shao and Grishin, 2000). Several residues within the

two hairpins, as well as in the adjacent helices, contribute to

this dsDNA binding. For example, in the complex of the recom-

670 Structure 20, 667–675, April 4, 2012 ª2012 Elsevier Ltd All rights

bination protein RuvA with Holliday junction DNA, two basic resi-

dues, Lys119 and Arg123, located in the helices of the HhH motif,

bind to a negatively charged DNA phosphate oxygen atoms

(Ariyoshi et al., 2000). Similar protein-DNA contacts have been

reported for the archaeal XPF homodimers (Newman et al.,

2005), for UvrC (Singh et al., 2002) and XPF/ERCC1 (Tripsianes

et al., 2005). Interestingly, in the present XPF-ssDNA complex,

Lys850 and Arg853, which are both residues of XPF located in

equivalent positions as Lys119 and Arg123 of RuvA, contact

ssDNA. In addition, we observe a number of contacts from resi-

dues in the second helix of the first canonical HhH motif of XPF

(Figure 2B). No significant chemical shift perturbations were

observed for the hairpins of (XPF)2, indicating a distinct DNA

binding mode as compared to other HhHmotifs. Most molecular

reserved

Figure 3. Three-dimensional Structure of

the (XPF)2-(CAGTGGCTGA)2 Complex

(A) Stereoview of the final 20 conformers of the

ensemble. The protein subunits are colored blue,

and the DNA is colored green; the hairpins in both

protein subunits are in red.

(B) Surface representation of the XPF-CAGT

GGCTGA complex. The sugar phosphate back-

bone of the DNA is colored gray, the XPF subunit is

colored blue, and the surface is colored light gray.

See Figure S4 and Table S5 for the distribution of

the a (O30-P-O50-C50) and g (O50-C5-C40-C30)torsion angles for nucleotides 3–7 in the (XPF)2-

ssDNA complex.

Structure

Structure of the XPF-ssDNA Complex

contacts were mainly with the phosphate backbone of ssDNA in

an extended geometry. With dsDNA, these interactions would

lead to a disruption of the B-DNA geometry, and this may be

the explanation for the noted preference of XPF for ssDNA.

Structural Comparison to Protein-ssDNA ComplexesThere is a large set of proteins that form complexes with ssDNA

(Bochkarev et al., 1997; Braddock et al., 2002a, 2002b; Brand-

sen et al., 1997; Kerr et al., 2003; Mitton-Fry et al., 2002). In

most of these complexes, DNA bases that mediate protein-

DNA interactions point toward the protein. However, in an alter-

nate situation, DNA bases can also point away from the protein

surface such that the DNA phosphate groups mediate electro-

static contacts with the protein side chains (Werten and Moras,

2006). The structure of RecA-ssDNA complex is a good

example, where it has been noted that the three stacked DNA

bases are sandwiched between the RecA b hairpins with their

Watson-Crick edges solvent exposed and that one of the bases

contacts aliphatic groups of long-chain amino acids like Lys198,

Ile199, and Thr 208 (Chen et al., 2008). Notably, in the present

XPF-ssDNA complex, the phosphate backbone for nucleotide

2 to 7 of the 10 nt ssDNA is largely buried, whereas most bases

of the ssDNA are pointing away from the protein surface. The

ssDNA in the complex exists in a right-handed helical conforma-

tion. Such a right-handed helical ssDNA conformation has been

seen also for the hnRNPK-ssDNA complex (Braddock et al.,

2002a). Thus, the XPF surface presents a scaffold for single

strand DNA binding and the helical conformation of the bound

DNA is determined by the protein side chains, as seen before

in the structure of the CdC13-telomere DNA complex (Mitton-

Structure 20, 667–675, April 4, 2012

Fry et al., 2002). The ssDNA preference

of XPF is clearly distinct from the dsDNA

preference of ERCC1.

A Model for the Splayed Arm DNASubstrate Recognition by theFunctional, Full-Length XPF/ERCC1 HeterodimerIn this study, we determined solution

structure of XPF protein bound to 10 nt

ssDNA. NMR titration experiments also

showed that XPF when present in the

XPF/ERCC1 complex binds the 10 nt

ssDNA in a very similar way. Therefore,

the present structural data of the complex of the (XPF)2 homo-

dimer with ssDNA can serve as a model explaining how the

HhH domain of XPF binds ssDNA at a ss/ds DNA junction in

the ERCC1/XPF complex. Tsodikov et al. (2005) suggested

a model of XPF/ERCC1 binding to a splayed arm DNA substrate

to account for their biochemical and structural findings. In this

model, the HhH domain of XPF is in contact with a single

strand fragment of the splayed arm DNA, which is consistent

with our present structural findings. It has also been reported

that the ERCC1 HhH domain can interact with both dsDNA

and ssDNA (Tripsianes et al., 2005; Tsodikov et al., 2005).

Based on NMR titration data, we suggested that the conserved

HhH domain of the ERCC1 protein contact the phosphate back-

bone of the dsDNA in the minor groove as noted for the RuvA

(Ariyoshi et al., 2000) and archaeal XPF proteins (Newman

et al., 2005).

When combining the current structural data of the (XPF)2-

ssDNA complex with the previous findings of Newman et al.

(2005), Tsodikov et al. (2005, 2007), and Tripsianes et al. (2005,

2007), we can build a model for the complex of XPF/ERCC1,

including the nuclease domains at a splayed arm DNA substrate

(Figure 5). A major highlight of this model is that the XPF HhH

domain interacts with a defined 50/ 30 polarity with one of the

single strand arms of the splayed arm DNA. The ERCC1 HhH

domain interacts with the minor grove at the double stranded

region of a splayed arm duplex DNA. The XPF catalytic residues

on the nuclease domain are in proximity to the cleavage site in

the damaged DNA strand, in a position in accordance with

previous biochemical analysis of this DNA cleavage by XPF/

ERCC1 (de Laat et al., 1998a).

ª2012 Elsevier Ltd All rights reserved 671

A

B5‘

3‘

K850

R853

H858

H857

S854

1

2

3

4

5

6

7

8

9

10

N879

S875

I876

G878

L877

K850N848

N851

N879

H858

A2

G3

T4

G6R853 S854

H857

G5C7

C

5‘

3‘

Figure 4. The Interaction Interface of the XPF-CAGTGGCTGA

Complex

(A) Detailed structural view showing the protein DNA surface. The protein side

chains are colored yellow.

(B) Summary of the interactions between XPF and DNA. Residues mediating

phosphate interactions are shown by black arrows; a blue arrow indicates

nonbonding interactions involving protein residues and DNA. The purple

arrows indicate base hydrogen bond contact. The flipped nucleotide is en-

closed in the red box.

See Figure S2 for supporting NMR data.

ERCC1 Central

XPF Catalytic

5’

XPF HhH

ERCC1 HhH

ssDNA

h1

h2

Cleavage site

5’ 3’

5’

3’

Figure 5. A Model for XPF/ERCC1 Binding to a Splayed Arm DNA

Substrate

The ERCC1 HhH domain (red), containing two hairpins H1 and H2 (colored

magenta), contacts dsDNA. The XPF HhH domain (blue) is positioned on the

ssDNA (green), thereby stabilizing the ssDNA. The DNA strand colored brown

is the damaged strand to be cleaved by XPF nuclease domain (light blue). The

cleavage site marked by a thick black arrow is close to catalytic residues

(orange). Themodeled 6nts ssDNA (white) connects 30 end of the nondamaged

ssDNA strand (green) to the 50 end of the dsDNA (green). The ERCC1 central

domain (light yellow) can interact with XPA and ssDNA strand on the NER

complex.

Structure

Structure of the XPF-ssDNA Complex

Role of XPF in Positioning XPF/ERCC1 on the NERComplexXPF/ERCC1 cleaves DNA near the ss/ds DNA junction with

a defined polarity. Despite the available structures of archaeal

XPF bound to DNA, and various structures of parts of human

XPF and ERCC1, the mechanism underlying polarity remains

elusive (Tripsianes et al., 2005, 2007; Tsodikov et al., 2005,

2007). We propose that the substrate preference of the DNA

binding domains of ERCC1 and XPF could contribute to the

672 Structure 20, 667–675, April 4, 2012 ª2012 Elsevier Ltd All rights

correct positioning of the heterodimeric XPF/ERCC1 complex

with respect to the single-double strand junction (de Laat

et al., 1998a). This substrate preference well-clarifies the role

of XPF in positioning XPF/ERCC1 at the ss/ds DNA junction in

the NER complex.

EXPERIMENTAL PROCEDURES

Sample Preparation

Expression and purification of the C-terminal XPF823-905 as (XPF)2 homodimers

has been described before (Das et al., 2008). The protein fraction was concen-

trated to nearly 450 ml, with a final protein monomer concentration of �1 mM

using Amicon filters with a 3 kD cutoff. Using this protocol both 15N-labeled

and 13C/15N-labeled (XPF)2 samples were prepared.

The 10 nt sequence 50CAGTGGCTGA and the 39 nt sequence 50TGCG

AATTCATATGCAATATTCAGTGGCTGAGCTACTGG were purchased from

Operon Biotechnologies GmBH (Cologne, Germany).

Mass Spectrometry

The MALDI-TOF mass spectra were recorded on a Voyager-DE PRO mass

spectrometer (Applied Biosystems) with implemented delayed extraction

technique, equipped with a N2 laser (337 nm, 3 ns pulse width). The mass

spectra (positive-ion mode) were recorded in a linear mode at an accelerating

voltage of 25 kV using an extraction delay of 300 ns. The (XPF)2 homodimer

and 39 nt ssDNA samples (�0.6 ml) were mixed in a 1:1 ratio on the target plate

reserved

Structure

Structure of the XPF-ssDNA Complex

with the matrix solution consisting of a-cyano-4-4 hydroxycinnaminic acid

(10mg/ml) dissolved in acetonitrile/water/trifluoroacetic acid (50:49.7:0.3, v/v).

NMR Spectroscopy

NMR data were collected at 20.6�C on a Bruker DRX 600 MHz spectrometer

equipped with a z-gradient triple-resonance cryoprobe and a Bruker

Avance 900 MHz spectrometer equipped with a z-gradient triple-resonance

probe. The final NMR sample contained �1 mM (XPF)2-ssDNA. A set of triple

resonance experiments—HNCACB, CBCACONH, HNCO, HN(CA)CO, HNCA,

HN(CO)CA, HBHA(CBCACONH), HBHANH, HN(CA)HA, (H)CC(CO)NH-

TOCSY, H(CCCO)NH-TOCSY, (H)CCH-TOCSY, and H(C)CH-TOCSY—as

described previously (Das et al., 2008), were acquired at 600 MHz to reassign

the backbone and side chain 1H, 13C and 15N resonances for the complex.

A set of NOE spectra—3D-NOESY-(13C,1H)-HSQC, 3D-(13C)-HMQC-NO-

ESY-(15N,1H)-HSQC, 2D 15N-filtered/13C edited-(1H,1H)-NOESY (tmix =

80 ms, 100 ms), and 13C/15N double half filter (1H-1H)- NOESY (tmix = 80 ms,

100 ms)—were acquired at a Bruker DRX 600 MHz spectrometer using a

triple-resonance cryoprobe, and a 3D NOESY-(15N,1H)-HSQC (tmix = 80 ms,

100ms) was acquired at a Bruker Avance 600MHz spectrometer as described

previously (Das et al., 2008). 1H-decoupled one-dimensional (1D) 31P spectra

and 2D 1H-31P HMQC spectra of free and bound DNA were acquired at

a Bruker DRX 500 MHz spectrometer equipped with a QXI probe. 2D NOESY

(tmix = 80 ms, 100 ms, 150 ms, 200 ms) and 2D TOCSY (tmix = 20 ms, 40 ms,

70 ms) spectra on free DNA were acquired on a Bruker Avance 900 MHz

spectrometer. 2D-NOESY (13C/15N decoupled, 13C decoupled, coupled)

and 2D-TOCSY (15N/13C-filtered) were acquired on a �1mM (XPF)2-(CAGT

GGCTGA)2 sample in a buffer containing 2 mM NaPi (pH 5.2), �80 mM

NaCl, 95%/5% H2O/D2O, and a small amount of complex protease inhibitor

(Roche). All the spectra were processed using XWINNMR 3.5 (Bruker) and

analyzed as described previously (Das et al., 2008).

For the NMR titrations with DNA, we used 15N-labeled (XPF)2. The15N-1H

HSQC spectra at the different titration points were acquired and analyzed as

described previously (Tripsianes et al., 2005).

Structure Calculation

The structure of the (XPF)2-ssDNA complex was determined as follows and

essentially using a strategy as described before (Kopke Salinas et al., 2005;

Romanuka et al., 2009). First, the structure of the (XPF)2 homodimer part

without ssDNA was determined using distance restraints (based on 1H-1H

NOEs) and dihedral restraints (based on TALOS analysis of the backbone

chemical shifts) using the program CYANA/CANDID (Herrmann et al., 2002),

similarly as described for the free (XPF)2 homodimer (Das et al., 2008). Next,

the two 10 nt ssDNA fragments were docked onto the XPF homodimer using

a large set of unambiguous and ambiguous intermolecular protein-DNA NOE

restraints with the program CNS using an Aria protocol (Brunger et al., 1998)

as implemented in the Haddock 2.0 software suite (van Dijk et al., 2006; de

Vries et al., 2007). Finally, the conformers in the ensemble were refined with

CNS in explicit solvent and the conformers clustered. In this final calculation,

the unstructured C- and N-terminal residues of the protein, as evidenced

from our previous structural data and the absence of any nonsequential inter-

residue 1H-1H NOEs, were not included. A summary of the experimental

restraints used for the final structure calculations is given in Table 1. Additional

restraints for maintaining nucleic acid base planarity, the correct sugar pucker

conformation (C20-endo), and the phosphate backbone conformation were

used as described in van Dijk et al. (2006). The dihedral angles of the sugar-

phosphate backbone were obtained from a 2D 31P-1H HMQC spectrum (Fig-

ure S2A) and converted to restraints as described in Braddock et al. (2002a).

The error ranges attributed to these restraints were defined independently

for the four distinct docking stages described below.

(1) Preparation of the DNA starting structures: an extended ssDNA

structure generated using CNS was subjected to a single molecule

HADDOCK run using the intramolecular DNA NOEs. This procedure

generated 40 structures during the simulated annealing stage. All 40

structures were subjected to water refinement. The error range for

the sugar-phosphate backbone dihedral restraints was set to (ainp ±

50, binp ± 50, ginp ± 35, dinp ± 10, εinp ± 50, xinp ± 50). The 20 lowest-

Structure 20, 66

energy water-refinedDNA structures were selected for the calculations

of the (XPF)2-ssDNA complex.

(2) Sampling conformational space of the complex: using intramolecular

DNA/DNA NOE distance restraints and the ambiguous and unambig-

uous intermolecular protein/DNA NOE restraints 1000 docking solu-

tions of the complex were generated using rigid body docking.

From these the best 200 structures were selected according to the

HADDOCK score and subjected to simulated annealing using CNS,

as implemented in Haddock, followed by a gentle water refinement.

All residues of the ssDNA were set as fully flexible during the simulated

annealing stage of CNS. Because of the imposed DNA restraints, only

the C10- (N9/N1) bond angle between the nucleic acid base, the internal

sugar, and the sugar-phosphate backbone dihedral angles can adapt

at this stage. The error range of all sugar-phosphate backbone dihedral

restraints was set to a broad range, Xinp ± 100�. All parameters were as

described in van Dijk et al. (2006). Finally, the 200 solutions were clus-

tered using a 7.5 A rmsd cutoff on the pairwise backbone rmsd matrix

(Ca and P atoms). This resulted in a total of three clusters. The dielectric

constant (ε) as used in the rigid-body docking and simulated annealing

stage was set to 78. To reduce computational time the sampling in this

docking run was only directed toward one of the two symmetric XPF

binding sites (2-component docking).

(3) Restricting the conformational space. Sixteen DNA conformations ex-

tracted from the best cluster of stage 2 were used as input for the next

round of docking calculations. This ensemble of 16 DNA conformations

was extended to 20 by including the best solution from each of the two

other clusters (clusters 2 and 3) to minimize the risk of excluding

possible favorable conformations that ended up in the other clusters.

The docking proceeded in the samemanner as in stage 2, but the error

ranges for the phosphate backbone dihedral restraints were now

reduced to (ainp ± 80, binp ± 80, ginp ± 50, dinp ± 40, εinp ± 80, zinp ± 80).

(4) Final refinement. Using a duplicated set of protein/ssDNA NOEs,

ssDNA was docked on both sides of the homodimeric (XPF)2 using

a 3-component docking. For this round, five ssDNA conformations

extracted from the best solutions of stage 3 were used as starting

structures. Noncrystallographic symmetry restraints were used to

enforce a symmetrical orientation of all atoms of the two ssDNA mole-

cules in the complex as described in Kopke Salinas et al. (2005). The

structure calculation proceeded in the same manner as the previous

two stages, but the error range of the phosphate backbone dihedral

restraints was reduced to Xinp ± 50� for all dihedral angles, similarly

to Braddock et al. (2002a). The structures were clustered, and a final

ensemble containing the 20 lowest-energy structures was obtained.

The structures were analyzed using WHATCHECK (Hooft et al., 1996)

and PROCHECK (Laskowski et al., 1996).

Modeling of the XPF/ERCC1 DNA Complex

The model of the complex of XPF/ERCC1 with splayed arm DNA was made

using the program Pymol (http://www.pymol.org). The model is constructed

from the known structures of the human XPF-ERCC1 HhH domains (1z00),

the ERCC1 central domain (2jpd), the archaeal XPF homodimer bound to

dsDNA (2bgw), and the present structure of (XPF)2 bound to 10 nt ssDNA.

The position of dsDNA to ERCC1 is based upon the X-ray structure of A, pernix

XPF in complex with dsDNA duplex (PDB code 2bgw; Newman et al., 2005),

and the structure of free ERCC1/XPF (PDB code 1z00; Tripsianes et al.,

2005). The position of XPF HhH motifs on the ssDNA is established from the

present study. The gap from the 30 end of ssDNA to the 50 end of dsDNA

was filled with a 6 nt ssDNA sequence, whereas the alternate choice would

lead to a significant large nucleotide linker. Our choice for the model was the

shortest linker.

ACCESSION NUMBERS

Coordinates have been deposited in the Protein Data Bank (PDB) with the

accession code 2KN7, and the chemical shifts and structural restraints have

been deposited in the Biological Magnetic Resonance Bank with the entry

numbers 16449 and 18343.

7–675, April 4, 2012 ª2012 Elsevier Ltd All rights reserved 673

Structure

Structure of the XPF-ssDNA Complex

SUPPLEMENTAL INFORMATION

Supplemental Information includes five figures and five tables and can be

found with this article online at doi:10.1016/j.str.2012.02.009.

ACKNOWLEDGMENTS

The authors are grateful to Profs. A.M.J.J Bonvin, J. Romanuka, E. AB, and

K. Tripsianes for the assistance and discussions during the project. We thank

V. Blanchard and Prof. J.H. Kamerling for use of the MALDI-TOF mass spec-

trometer. This work was financially supported by the European Commission

(project 031220, Spine2-complexes), the Center for Biomedical Genetics,

and the Netherlands Organisation for Scientific Research, Chemical Sciences

(NWO-CW).

Received: October 24, 2011

Revised: January 22, 2012

Accepted: February 17, 2012

Published: April 3, 2012

REFERENCES

Ahmad, A., Robinson, A.R., Duensing, A., van Drunen, E., Beverloo, H.B.,

Weisberg, D.B., Hasty, P., Hoeijmakers, J.H., and Niedernhofer, L.J. (2008).

ERCC1-XPF endonuclease facilitates DNA double-strand break repair. Mol.

Cell. Biol. 28, 5082–5092.

Al-Minawi, A.Z., Saleh-Gohari, N., and Helleday, T. (2008). The ERCC1/XPF

endonuclease is required for efficient single-strand annealing and gene

conversion in mammalian cells. Nucleic Acids Res. 36, 1–9.

Ariyoshi, M., Nishino, T., Iwasaki, H., Shinagawa, H., and Morikawa, K. (2000).

Crystal structure of the holliday junction DNA in complex with a single RuvA

tetramer. Proc. Natl. Acad. Sci. USA 97, 8257–8262.

Bessho, T., Sancar, A., Thompson, L.H., and Thelen, M.P. (1997).

Reconstitution of human excision nuclease with recombinant XPF-ERCC1

complex. J. Biol. Chem. 272, 3833–3837.

Bochkarev, A., Pfuetzner, R.A., Edwards, A.M., and Frappier, L. (1997).

Structure of the single-stranded-DNA-binding domain of replication protein

A bound to DNA. Nature 385, 176–181.

Braddock, D.T., Baber, J.L., Levens, D., and Clore, G.M. (2002a). Molecular

basis of sequence-specific single-stranded DNA recognition by KH domains:

solution structure of a complex between hnRNP K KH3 and single-stranded

DNA. EMBO J. 21, 3476–3485.

Braddock, D.T., Louis, J.M., Baber, J.L., Levens, D., and Clore, G.M. (2002b).

Structure and dynamics of KH domains from FBP bound to single-stranded

DNA. Nature 415, 1051–1056.

Brandsen, J., Werten, S., van der Vliet, P.C., Meisterernst, M., Kroon, J., and

Gros, P. (1997). C-terminal domain of transcription cofactor PC4 reveals

dimeric ssDNA binding site. Nat. Struct. Biol. 4, 900–903.

Brunger, A.T., Adams, P.D., Clore, G.M., DeLano, W.L., Gros, P., Grosse-

Kunstleve, R.W., Jiang, J.S., Kuszewski, J., Nilges, M., Pannu, N.S., et al.

(1998). Crystallography & NMR system: A new software suite for macromolec-

ular structure determination. Acta Crystallogr. D Biol. Crystallogr. 54, 905–921.

Chen, Z., Yang, H., and Pavletich, N.P. (2008). Mechanism of homologous

recombination from the RecA-ssDNA/dsDNA structures. Nature 453, 489–4.

Choi, Y.J., Ryu, K.S., Ko, Y.M., Chae, Y.K., Pelton, J.G., Wemmer, D.E., and

Choi, B.S. (2005). Biophysical characterization of the interaction domains

and mapping of the contact residues in the XPF-ERCC1 complex. J. Biol.

Chem. 280, 28644–28652.

Clingen, P.H., De Silva, I.U., McHugh, P.J., Ghadessy, F.J., Tilby, M.J.,

Thurston, D.E., and Hartley, J.A. (2005). The XPF-ERCC1 endonuclease and

homologous recombination contribute to the repair of minor groove DNA inter-

strand crosslinks in mammalian cells produced by the pyrrolo[2,1-c][1,4]

benzodiazepine dimer SJG-136. Nucleic Acids Res. 33, 3283–3291.

674 Structure 20, 667–675, April 4, 2012 ª2012 Elsevier Ltd All rights

Das, D., Tripsianes, K., Jaspers, N.G., Hoeijmakers, J.H., Kaptein, R., Boelens,

R., and Folkers, G.E. (2008). The HhH domain of the human DNA repair protein

XPF forms stable homodimers. Proteins 70, 1551–1563.

de Laat, W.L., Appeldoorn, E., Jaspers, N.G., and Hoeijmakers, J.H. (1998a).

DNA structural elements required for ERCC1-XPF endonuclease activity.

J. Biol. Chem. 273, 7835–7842.

de Laat, W.L., Appeldoorn, E., Sugasawa, K., Weterings, E., Jaspers, N.G.,

and Hoeijmakers, J.H. (1998b). DNA-binding polarity of human replication

protein A positions nucleases in nucleotide excision repair. Genes Dev. 12,

2598–2609.

de Laat, W.L., Sijbers, A.M., Odijk, H., Jaspers, N.G., and Hoeijmakers, J.H.

(1998c). Mapping of interaction domains between human repair proteins

ERCC1 and XPF. Nucleic Acids Res. 26, 4146–4152.

de Laat, W.L., Jaspers, N.G.J., and Hoeijmakers, J.H.J. (1999). Molecular

mechanism of nucleotide excision repair. Genes Dev. 13, 768–785.

de Vries, S.J., van Dijk, A.D., Krzeminski, M., van Dijk, M., Thureau, A., Hsu, V.,

Wassenaar, T., and Bonvin, A.M. (2007). HADDOCK versus HADDOCK: new

features and performance of HADDOCK2.0 on the CAPRI targets. Proteins

69, 726–733.

Enzlin, J.H., and Scharer, O.D. (2002). The active site of the DNA repair endo-

nuclease XPF-ERCC1 forms a highly conserved nuclease motif. EMBO J. 21,

2045–2053.

Gaillard, P.H., and Wood, R.D. (2001). Activity of individual ERCC1 and XPF

subunits in DNA nucleotide excision repair. Nucleic Acids Res. 29, 872–879.

Herrmann, T., Guntert, P., and Wuthrich, K. (2002). Protein NMR structure

determination with automated NOE assignment using the new software

CANDID and the torsion angle dynamics algorithm DYANA. J. Mol. Biol. 319,

209–227.

Hoeijmakers, J.H.J. (2001). Genome maintenance mechanisms for preventing

cancer. Nature 411, 366–374.

Hooft, R.W., Vriend, G., Sander, C., and Abola, E.E. (1996). Errors in protein

structures. Nature 381, 272.

Houtsmuller, A.B., Rademakers, S., Nigg, A.L., Hoogstraten, D., Hoeijmakers,

J.H.J., and Vermeulen, W. (1999). Action of DNA repair endonuclease ERCC1/

XPF in living cells. Science 284, 958–961.

Kerr, I.D., Wadsworth, R.I., Cubeddu, L., Blankenfeldt, W., Naismith, J.H., and

White, M.F. (2003). Insights into ssDNA recognition by the OB fold from a

structural and thermodynamic study of Sulfolobus SSB protein. EMBO J. 22,

2561–2570.

Kuraoka, I., Kobertz, W.R., Ariza, R.R., Biggerstaff, M., Essigmann, J.M., and

Wood, R.D. (2000). Repair of an interstrand DNA cross-link initiated by

ERCC1-XPF repair/recombination nuclease. J. Biol. Chem. 275, 26632–

26636.

Laskowski, R.A., Rullmannn, J.A., MacArthur, M.W., Kaptein, R., and

Thornton, J.M. (1996). AQUA and PROCHECK-NMR: programs for checking

the quality of protein structures solved by NMR. J. Biomol. NMR 8, 477–486.

McWhir, J., Selfridge, J., Harrison, D.J., Squires, S., and Melton, D.W. (1993).

Mice with DNA repair gene (ERCC-1) deficiency have elevated levels of p53,

liver nuclear abnormalities and die before weaning. Nat. Genet. 5, 217–224.

Melis, J.P., Wijnhoven, S.W., Beems, R.B., Roodbergen, M., van den Berg, J.,

Moon, H., Friedberg, E., van der Horst, G.T., Hoeijmakers, J.H., Vijg, J., and

van Steeg, H. (2008). Mouse models for xeroderma pigmentosum group A

and group C show divergent cancer phenotypes. Cancer Res. 68, 1347–1353.

Mitton-Fry, R.M., Anderson, E.M., Hughes, T.R., Lundblad, V., and Wuttke,

D.S. (2002). Conserved structure for single-stranded telomeric DNA recogni-

tion. Science 296, 145–147.

Munoz, P., Blanco, R., Flores, J.M., and Blasco, M.A. (2005). XPF nuclease-

dependent telomere loss and increased DNA damage in mice overexpressing

TRF2 result in premature aging and cancer. Nat. Genet. 37, 1063–1071.

Newman, M., Murray-Rust, J., Lally, J., Rudolf, J., Fadden, A., Knowles, P.P.,

White, M.F., and McDonald, N.Q. (2005). Structure of an XPF endonuclease

with and without DNA suggests a model for substrate recognition. EMBO J.

24, 895–905.

reserved

Structure

Structure of the XPF-ssDNA Complex

Niedernhofer, L.J., Essers, J., Weeda, G., Beverloo, B., de Wit, J., Muijtjens,

M., Odijk, H., Hoeijmakers, J.H.J., and Kanaar, R. (2001). The structure-

specific endonuclease Ercc1-Xpf is required for targeted gene replacement

in embryonic stem cells. EMBO J. 20, 6540–6549.

Niedernhofer, L.J., Odijk, H., Budzowska, M., van Drunen, E., Maas, A., Theil,

A.F., de Wit, J., Jaspers, N.G.J., Beverloo, H.B., Hoeijmakers, J.H.J., and

Kanaar, R. (2004). The structure-specific endonuclease Ercc1-Xpf is required

to resolve DNA interstrand cross-link-induced double-strand breaks. Mol.

Cell. Biol. 24, 5776–5787.

Orelli, B., McClendon, T.B., Tsodikov, O.V., Ellenberger, T., Niedernhofer, L.J.,

and Scharer, O.D. (2009). The XPA-binding domain of ERCC1 is required for

nucleotide excision repair but not other DNA repair pathways. J. Biol. Chem.

285, 3705–3712.

Park, C.H., Bessho, T., Matsunaga, T., and Sancar, A. (1995). Purification and

characterization of the XPF-ERCC1 complex of human DNA repair excision

nuclease. J. Biol. Chem. 270, 22657–22660.

Riedl, T., Hanaoka, F., and Egly, J.M. (2003). The comings and goings of nucle-

otide excision repair factors on damaged DNA. EMBO J. 22, 5293–5303.

Romanuka, J., Folkers, G.E., Biris, N., Tishchenko, E., Wienk, H., Bonvin, A.M.,

Kaptein, R., and Boelens, R. (2009). Specificity and affinity of Lac repressor for

the auxiliary operators O2 and O3 are explained by the structures of their

protein-DNA complexes. J. Mol. Biol. 390, 478–489.

Kopke Salinas, R., Folkers, G.E., Bonvin, A.M.J.J., Das, D., Boelens, R., and

Kaptein, R. (2005). Altered specificity in DNA binding by the lac repressor:

a mutant lac headpiece that mimics the gal repressor. Chem. Bio. Chem. 6,

1628–1637.

Scharer, O.D. (2007). Achieving broad substrate specificity in damage recog-

nition by binding accessible nondamaged DNA. Mol. Cell 28, 184–186.

Shao, X., and Grishin, N.V. (2000). Common fold in helix-hairpin-helix proteins.

Nucleic Acids Res. 28, 2643–2650.

Sijbers, A.M., de Laat, W.L., Ariza, R.R., Biggerstaff, M., Wei, Y.F., Moggs,

J.G., Carter, K.C., Shell, B.K., Evans, E., de Jong, M.C., et al. (1996).

Xeroderma pigmentosum group F caused by a defect in a structure-specific

DNA repair endonuclease. Cell 86, 811–822.

Singh, S., Folkers, G.E., Bonvin, A.M.J.J., Boelens, R., Wechselberger, R.,

Niztayev, A., and Kaptein, R. (2002). Solution structure and DNA-binding prop-

erties of the C-terminal domain of UvrC from E.coli. EMBO J. 21, 6257–6266.

Structure 20

Tian, M., Shinkura, R., Shinkura, N., and Alt, F.W. (2004). Growth retardation,

early death, and DNA repair defects in mice deficient for the nucleotide exci-

sion repair enzyme XPF. Mol. Cell. Biol. 24, 1200–1205.

Tripsianes, K., Folkers, G., Ab, E., Das, D., Odijk, H., Jaspers, N.G.,

Hoeijmakers, J.H., Kaptein, R., and Boelens, R. (2005). The structure of the

human ERCC1/XPF interaction domains reveals a complementary role for

the two proteins in nucleotide excision repair. Structure 13, 1849–1858.

Tripsianes, K., Folkers, G.E., Zheng, C., Das, D., Grinstead, J.S., Kaptein, R.,

and Boelens, R. (2007). Analysis of the XPA and ssDNA-binding surfaces on

the central domain of human ERCC1 reveals evidence for subfunctionalization.

Nucleic Acids Res. 35, 5789–5798.

Tsodikov, O.V., Enzlin, J.H., Scharer, O.D., and Ellenberger, T. (2005). Crystal

structure and DNA binding functions of ERCC1, a subunit of the DNA struc-

ture-specific endonuclease XPF-ERCC1. Proc. Natl. Acad. Sci. USA 102,

11236–11241.

Tsodikov, O.V., Ivanov, D., Orelli, B., Staresincic, L., Shoshani, I., Oberman, R.,

Scharer, O.D., Wagner, G., and Ellenberger, T. (2007). Structural basis for the

recruitment of ERCC1-XPF to nucleotide excision repair complexes by XPA.

EMBO J. 26, 4768–4776.

van Dijk, M., van Dijk, A.D.J., Hsu, V., Boelens, R., and Bonvin, A.M.J.J. (2006).

Information-driven protein-DNA docking using HADDOCK: it is a matter of

flexibility. Nucleic Acids Res. 34, 3317–3325.

Volker, M., Mone, M.J., Karmakar, P., van Hoffen, A., Schul, W., Vermeulen,

W., Hoeijmakers, J.H.J., van Driel, R., van Zeeland, A.A., and Mullenders,

L.H. (2001). Sequential assembly of the nucleotide excision repair factors

in vivo. Mol. Cell 8, 213–224.

Weeda, G., Donker, I., de Wit, J., Morreau, H., Janssens, R., Vissers, C.J.,

Nigg, A., van Steeg, H., Bootsma, D., and Hoeijmakers, J.H.J. (1997).

Disruption of mouse ERCC1 results in a novel repair syndrome with growth

failure, nuclear abnormalities and senescence. Curr. Biol. 7, 427–439.

Werten, S., and Moras, D. (2006). A global transcription cofactor bound to

juxtaposed strands of unwound DNA. Nat. Struct. Mol. Biol. 13, 181–182.

Winkler, G.S., Sugasawa, K., Eker, A.P., de Laat, W.L., and Hoeijmakers,

J.H.J. (2001). Novel functional interactions between nucleotide excision DNA

repair proteins influencing the enzymatic activities of TFIIH, XPG, and

ERCC1-XPF. Biochemistry 40, 160–165.

Wu, Y., Mitchell, T.R., and Zhu, X.D. (2008). Human XPF controls TRF2 and

telomere length maintenance through distinctive mechanisms. Mech. Ageing

Dev. 129, 602–610.

, 667–675, April 4, 2012 ª2012 Elsevier Ltd All rights reserved 675