Damage-dependent regulation of MUS81-EME1 byFanconi anemia complementation group A proteinAnaid Benitez1 Fenghua Yuan1 Satoshi Nakajima2 Leizhen Wei2 Liangyue Qian1

Richard Myers1 Jennifer J Hu3 Li Lan2 and Yanbin Zhang1

1Department of Biochemistry amp Molecular Biology University of Miami Miller School of Medicine Miami FL33136 USA 2Department of Microbiology amp Molecular Genetics University of Pittsburgh Pittsburgh PA 15213USA and 3Department of Epidemiology amp Public Health University of Miami Miller School of Medicine MiamiFL 33136 USA

Received July 19 2013 Revised September 25 2013 Accepted October 1 2013

ABSTRACT

MUS81-EME1 is a DNA endonuclease involved inreplication-coupled repair of DNA interstrand cross-links (ICLs) A prevalent hypothetical role of MUS81-EME1 in ICL repair is to unhook the damage by incisingthe leading strand at the 30 side of an ICL lesion In thisstudy we report that purified MUS81-EME1 incisesDNA at the 50 side of a psoralen ICL residing in forkstructures Intriguingly ICL repair protein Fanconianemia complementation group A protein (FANCA)greatly enhances MUS81-EME1-mediated ICLincision On the contrary FANCA exhibits a two-phase incision regulation when DNA is undamagedor the damage affects only one DNA strand Studiesusing truncated FANCA proteins indicate that both theN- and C-moieties of the protein are required for theincision regulation Using laser-induced psoralen ICLformation in cells we find that FANCA interacts withand recruits MUS81 to ICL lesions This report clarifiesthe incision specificity of MUS81-EME1 on ICLdamage and establishes that FANCA regulates theincision activity of MUS81-EME1 in a damage-depend-ent manner

INTRODUCTION

Interstrand cross-links (ICLs) covalently tether bothstrands of a DNA helix and block essential DNA trans-actions including replication and transcription DNA rep-lication is one of the critical factors to elicit repair of ICLs(1ndash4) Generally when an ICL blocks the replication ma-chinery a protein complex termed as the Fanconi anemiacore complex is recruited to stalled replication forks andmonoubiquitinates another two Fanconi anemia proteinsFANCD2 and FANCI This initiates a string of ICL

repair events including damage incision translesion syn-thesis and re-establishment of replication forks throughhomologous recombination (5ndash10)Fanconi anemia is a severe genetic disorder

characterized by bone marrow failure developmentaldefects chromosomal instability and predisposition tocancer Fanconi anemia cells are hypersensitive to DNAcross-linking compounds including mitomycin C cisplatinand diepoxybutane indicating that they are defective inrepairing ICLs (511ndash19) Thus far 15 distinct genes havebeen identified to cause the severe disease (17) Althoughdeficiency of each gene shows similar clinical and cellularphenotypes 60 of Fanconi anemia patients presenteddefective FANCA (915) indicating that this protein mayhave additional biological functions beyond the canonicalpathway through FANCI-FANCD2 monoubiquitinationIndividual components of the Fanconi anemia core

complex directly participate in ICL repair as well as inmaintenance of replication forks (20ndash25) FANCA is con-sidered a component of the Fanconi anemia core complex(including FANCA B C E F G L M and otherFanconi anemia associated proteins (FAAP)) FANCAhas been shown to have intrinsic affinity to nucleic acidsand has been found to be localized to chromatin in a rep-lication-dependent manner (26ndash28) FANCA deficientcells clearly showed lower incision of psoralen ICLscompared with wild-type cells and FANCB cellsindicating a specialized role for FANCA in ICL incision(29) Additionally using nuclear protein extracts andcomplementation analysis it was demonstrated thatFANCA is required for efficient incisions at the sites ofpsoralen-mediated ICLs (30) These data imply thatFANCA may function outside the Fanconi anemia corecomplex and directly participate in ICL incisionIt is well established that ICLs are incised in a replica-

tion-dependent manner (123132) Several prevalentmodels propose that two members of the Xeroderma

To whom correspondence should be addressed Tel +1 305 243 9237 Fax +1 305 243 3955 Email YZhang4medmiamiedu

Published online 28 October 2013 Nucleic Acids Research 2014 Vol 42 No 3 1671ndash1683doi101093nargkt975

The Author(s) 2013 Published by Oxford University PressThis is an Open Access article distributed under the terms of the Creative Commons Attribution License (httpcreativecommonsorglicensesby30) whichpermits unrestricted reuse distribution and reproduction in any medium provided the original work is properly cited

pigmentosum group F protein (XPF) family of DNAendonucleases XPF-ERCC1 and MUS81-EME1 partici-pate in replication-dependent ICL incision by cuttingDNA at the 50 and 30 sides of an ICL respectively (33ndash36) MUS81-EME1 cleaves 30 single-stranded DNA(ssDNA) branch and replication fork efficiently (37ndash47)making it a suitable candidate for ICL incision in replica-tion forks MUS81-EME1 promotes conversion ofICLs into double strand breaks (DSBs) in a replication-dependent manner (48) Intriguingly Kanaar and col-leagues also found that MUS81 is not involved in the gen-eration of DSBs from DNA damage that affects only onestrand of the DNA duplex (48) Collectively these resultsindicate that the structure-specific DNA endonucleaseMUS81-EME1 is specifically involved in incision ofICLs but not non-ICL DNA damage residing in a repli-cation fork However it remains unknown how MUS81-EME1 exactly incises the ICL-damaged replication forksand how the incision is regulated to promote ICL unhook-ing and how it avoids non-specific incision of undamagedor nonndashICL-damaged forksIn this study we investigated the ICL incision activity

of MUS81-EME1 using purified proteins and a definedsite-specific psoralen ICL substrate We report thatMUS81-EME1 incises leading strand at the 50 side of theICL More importantly we found that FANCA regulatesthe endonuclease activity of MUS81-EME1 in a damage-dependent manner

MATERIALS AND METHODS

Expression and purification of human MUS81-EME1and FANCA

Complementary DNAs for human MUS81 EME1 andFANCA were obtained by polymerase chain reaction amp-lification from a universal complementary DNA pool(BioChain Institute Inc) The full-length open readingframes were confirmed by sequencing and found to exactlymatchNCBIReference SequenceNM_025128NM_152463and NM_000135 respectively Co-expression of thehexahistidine-tagged EME1 and non-tagged MUS81 andoverexpression of non-tagged FANCA were achieved ininsect High Five cells using the Bac-to-Bac expressionsystem (Invitrogen Carlsbad CA) Expression ofMUS81-EME1 FANCA and its mutants was confirmedby western blot analysis using a Pierce ECL kit (PierceRockford CA) Antibody against MUS81 was purchasedfrom Novus Biologicals (Littleton CO) A monoclonalantibody against hexahistidine tag (GenScript PiscatawayNJ) was also used to confirm EME1 expression and subse-quent purification Antibody against FANCA was kindlyprovided by the Fanconi Anemia Research FundUpon expression of MUS81 and His6-EME1 in insect

cells the cells were homogenized in a protein extractionbuffer (20mM HepesndashKOH pH 75 05mM MgCl250mM NaCl 02M sucrose 5mM b-mercaptoethanolprotease inhibitors (05mM PMSF 03mgml benzamidinehydrochloride 05mgml of pepstatin A 05mgml ofleupeptin 05mgml of antipain)) by a Dounce homogenizerand 10 strokes on ice MUS81-EME1 were purified by

using a HiTrap chelating column charged with nickel aMono S andor a Superdex 200 gel filtration column (GEHealthcare Piscataway NJ) and by tracing MUS81-EME1protein through SDSndashPAGE and western blot Wild-type(WT)-FANCA and its mutants were purified using aprotocol described previously (28) Protein concentrationwas determined by the Coomassie (Bradford) ProteinAssay Reagent (Pierce Rockford CA) The purifiedproteins were stored in 80C in aliquots

Creation of psoralen ICLs

To create a site-specific DNA ICL a short oligo 50-GCTCGGTACCCGG with an internal psoralen modifiedT (underlined) was synthesized by Midland CertifiedReagent Company (Midland TX) After elongatingthrough ligation annealing with a partially complement-ing oligo exposing to ultraviolet A irradiation and purify-ing by denaturing gel (Supplementary Figure S2) weobtained an ICL-damaged splayed arm structure Thesplayed arm structure can be labeled at the 30 end on thetop leading strand through a32P incorporation by KlenowDNA polymerase (Supplementary Figure S3) Labeling ofthe 50 ends of the leading and lagging strands is donebefore creation of the ICL (Supplementary Figure S3)Through annealing with different oligos complementingto the leading or lagging strand or both 50 ssDNAbranch 30 ssDNA branch and static replication fork struc-tures were created (Supplementary Figure S2)

DNA incision assay

A total of 15 nM of labeled DNA substrate was incubatedwith indicated amount of proteins in a 10-ml reactionmixture with 25mM HepesndashKOH pH 76 1mM DTT3mM MgCl2 65 glycerol 120 mgml BSA and 100mMKCl After incubation for 20min at 37C the reaction wasterminated by adding 5 ml proteinase K (25mM EDTA067 SDS 150 mgml of proteinase K) and 10min incu-bation at 37C and by adding 15 ml sequencing dyeThe reaction mixture was resolved by running a 10denaturing sequencing gel DNA size markers wereprepared by labeling a mixture of defined oligos throughg-32P-adenosine triphosphate

RNA interference induction of ICL in living cells andconfocal microscopy

To study the interaction of FANCA and MUS81 inhuman cells we treated U2OS cells with ON-TARGETPlus SMART Pool siFANCA (Thermo ScientificDhamacon cat L-019283-00) and a non-targetingcontrol small interfering RNA (siRNA) (cat D-001810-01-05) using the Dhamafect transfection agent for 48 hAfter western blot verification of knockdown efficiency(Figure 6A) a GFP-tagged MUS81 construct pEGFP-N1-MUS81 (vectors are purchased from Clontech) wastransfected into the U2OS cells by Lipofectamine 24 hbefore drug treatment pEGFP-N1-FANCA was trans-fected into WT U2OS cells for monitoring status ofFANCA The Olympus FV1000 confocal microscopysystem was used (Cat F10PRDMYR-1 Olympus UPCIfacility) and FV1000 software was used for acquisition of

1672 Nucleic Acids Research 2014 Vol 42 No 3

images To create psoralen ICLs cells were pretreatedwith 100 nM of 8-methoxypsoralen (8-MOP) for 10minright before a 405 nm laser micro-irradiation The outputpower of the laser (original 50mW) passed through thelens is 5mWscan Laser light passed through a PLAPON60 oil lens (super chromatic abe corr obj W14NA FVCat FM1-U2B990) Cells were incubated at 37C on athermo-plate (MATS-U52RA26 for IX8171517050metal insert HQ control Cat OTH-I0126) in Opti-MEM during observation to avoid pH changes Theimages were taken 2min after laser treatment

RESULTS

MUS81-EME1 does not incise on the 30 side of apsoralen ICL

Although it was proposed that MUS81-EME1 unhooksDNA ICLs by incising the leading strand at the 30 sideof the damage (33ndash35) it had never been tested howMUS81-EME1 acts on ICL-damaged fork structuresTo evaluate how MUS81 incises ICL-damaged DNAwe co-overexpressed human MUS81 and hexahistidine-tagged EME1 in High Five insect cells and purifiedthe MUS81-EME1 complex to near homogeneity(Supplementary Figure S1) Using defined sequence wealso prepared a splayed arm structure with a psoralenICL located immediately at the junction site of ssDNAand double-stranded DNA (Supplementary Figure S2)The rationale for this design was based on recent obser-vations from Walterrsquos group (34) who reported that ICL

incision happens after the ICL damage is within 1 nt of thenascent strand endTo test whether MUS81-EME1 incises on the 30 side of

ICL damage as previously hypothesized we labeled theleading strand at the 30 end of the ICL-damaged splayedarm by a-32P incorporation (Supplementary Figure S3)By annealing with different oligos on leading andorlagging strands 50 ssDNA branch 30 ssDNA branchand static replication fork structures were created(Supplementary Figure S2) We next incubated thepurified MUS81-EME1 with the DNA structures withICL If MUS81-EME1 incised the top leading strand atthe junction site on the 30 side of the ICL a short DNAfragment would be expected Surprisingly no incisionproduct below 74 nt could be detected with increasingamount of MUS81-EME1 (Figure 1) Instead a muchlarger incision product band was observed with30 ssDNA branch and static replication fork structures(Figure 1 lanes 6ndash9) These data indicate that ourMUS81-EME1 does incise the ICL-damaged DNA andhas the same structure specificity for 30 ssDNA branchand replication fork as previously reported (37ndash46) butit does not incise leading strand at the junction site on the30 side of the ICL damage as previously proposed

MUS81-EME1 incises the leading strand at the 50 sideof the psoralen ICL

To determine the exact incision site we next labeled the50 ends of the leading and lagging strands separately(Supplementary Figure S3) Again incubation of

MUS81-EME1

P

PP

Lane 1 2 3 4 5 6 7 8 9

P

P

74-nt

0

Expected incisionProduct if at the 3rsquo side(35-nt)

148-nt

P

P

Figure 1 MUS81-EME1 incision on the 30-labeled ICL-containing DNA structures Titration of purified MUS81-EME1 (5ndash10 nM) on the 30-labeledpsoralen ICL-damaged DNA structures as shown on the top The schematic appearance of the products after incision were shown on the right Allreactions were performed in 10 ml of 25mM Hepes pH 76 100mM KCl 3mM MgCl2 1mM DTT 120 mgml BSA 65 Glycerol and 1 nM ofeach DNA substrate as indicated The reactions were performed at 37C for 20min and resolved on a 10 denaturing PAGE gel Letter P with acircle indicates a-32P labeling by Klenow DNA polymerase Asterisk (74 nt) indicates a decayed and uncross-linked species

Nucleic Acids Research 2014 Vol 42 No 3 1673

MUS81-EME1 with ICL substrates with 50 end labelingon the lagging strand showed that the endonuclease didnot react with splayed arm and 50 ssDNA branch struc-tures but it effectively incised ICL-damaged 30 ssDNAbranch and replication fork (Supplementary Figure S4Alanes 9ndash16) The incision was not on the splayed arm sideof the lagging strand because only a large incision productwas observed (Supplementary Figure S4A) As expectedMUS81-EME1 also did not incise undamaged DNA onthe lagging strand (Supplementary Figure S4B)Next MUS81-EME1 was titrated with ICL substrates

labeled at the 50 end on the top leading strand As shownin Figure 2 incision of the ICL-containing 30 ssDNAbranch and replication fork structures yielded a majorband at 34ndash35 nt position and some smaller minorbands (Figure 2A lanes 9ndash16) Because the distancebetween the 50 end 32P labeling and the junction site was39 nt the predominant site of MUS81-EME1-mediatedincision of ICL was calculated to be 4ndash5 nt away fromthe junction and located at the 50 side of the ICL MUS81-EME1 also incised undamaged controls at the similar sitewith slightly stronger activity (Figure 2 compare the re-maining substrate bands located on the top of lanes 9ndash16of B with those in lanes 9ndash16 of A) However no smallerminor bands were observed with the undamaged DNADuring our test we noticed a slight size difference

between the major incision product from ICL substratesand the one from undamaged controls (Figure 4A) Todefine the difference we analyzed the incision productsby extensively running the reaction mixtures from ICL-damaged and ICL-undamaged DNA next to each otherwith a size marker Figure 3 clearly showed that the majorincision product from the ICL-damaged DNA is 34 nt insize 1 nt shorter than the product from the undamagedcontrol This means MUS81-EME1 incises the psoralenICL-damaged DNA 5nt away from the fork junctionwhereas it cuts the undamaged DNA 4nt away (Figure 3)These data establish that MUS81-EME1 unequivocally

incises leading strand at the 50 side of the psoralen ICLwhen located at the junction site and the psoralen damageaffects the incision sites of MUS81-EME1

FANCA regulates MUS81-EME1-mediated DNAincision in a damage-dependent manner

Because FANCA interacts with DNA (28) and plays a rolein ICL incision (2930) we reasoned that FANCA maydirectly interact with ICL-unhooking DNA endonucleasessuch as MUS81-EME1 for more efficient ICL incision Totest this hypothesis purified FANCA was titrated in thepresence or absence of the psoralen ICL damage in thedefined in vitro incision assay with suboptimal amount ofMUS81-EME1 (15 nM) (Figure 4A and B)In the presence of ICL damage FANCA dramatically

stimulates ICL incision mediated by MUS81-EME1 upto 14-fold (Figure 4A and B ICL and Figure 4C)Intriguingly FANCA exerted a two-phase regulation ofMUS81-EME1 activity in the absence of the ICL damage(Figure 4A undamaged) In Phase I increasing FANCAconcentration up to 5 nM enhances MUS81-EME1incision activity of undamaged DNA although less

efficiently than its effect on ICL-damaged DNA(Figure 4C up to 3-fold) In Phase II increasingFANCA concentration from 10 to 20 nM inhibitsMUS81-EME1 activity on the undamaged DNAMUS81-EME1 incision activity was abrogated at 20 nMFANCA (Figure 4A and B) Using 20 nM of FANCAand 15 nM MUS81-EME1 we also performed a timecourse experiment to further examine the extent towhich FANCA stimulates or inhibits MUS81-EME1-mediated incision in the presence or absence of ICL(Supplementary Figure S5) The results clearly confirmedthat FANCA stimulates MUS81-EME1 incision on theICL-damaged DNA and under the same condition itinhibits MUS81-EME1 activity in the absence of ICLThese observations demonstrate that FANCA function-ally interacts with MUS81-EME1 to distinguish an ICLfrom undamaged DNA for damage-specific incision

It has been previously reported that MUS81 is involvedin ICL incision but spares DNA damage that affects onlyone DNA strand (48) Inhibition of MUS81-EME1activity by FANCA in the absence of DNA damage(Figure 4A and B) inspired us to hypothesize thatFANCA might also be involved in the regulation ofMUS81-EME1 activity on non-ICL DNA damage Totest this hypothesis we performed the same FANCA ti-tration using the psoralen thymine mono-adduct designedfor the creation of the ICL The psoralen mono-adduct isidentical in sequence and overall structure to the ICL onlyit was not exposed to ultraviolet A irradiation and conse-quently not allowed to form an ICL Surprisingly like onthe undamaged DNA FANCA stimulated or inhibitedMUS81-EME1-mediated incision of the psoralen mono-adduct in a two-phase concentration-dependent manner(Figure 4D compare the last lane with 20 nM ofFANCA to the second lane without FANCA)Furthermore experiments using a benzo(a)pyrenediolepoxide adducted deoxyguanosine a damage thataffects only one DNA strand (49) also showed thatFANCA can stimulate or suppress MUS81-EME1activity in a two-phase concentration-dependent manner(Figure 4E compare the last two lanes with 15 and 20 nMof FANCA respectively to the second lane withoutFANCA)

Overall we demonstrated that FANCA directly partici-pates in ICL-specific incision via the stimulation ofMUS81-EME1 activity Up- and downregulation ofMUS81-EME1 activity by FANCA provides an import-ant mechanism to uncover why it is possible for MUS81 tobe involved in the incision of ICLs but not DNA damagethat affects one DNA strand (48)

Both N- and C-terminals of FANCA are required for theregulation of MUS81-EME1

Because FANCA is a DNA binding protein we nextasked whether the damage-dependent regulation ofMUS81-EME1 by FANCA is caused by its affinity toDNA To test we created two truncation mutants ofFANCA Q772X and C772-1455 Q772X is a Fanconianemia disease-causing C-terminal truncation mutantC772-1455 is the complementing C-terminal fragment of

1674 Nucleic Acids Research 2014 Vol 42 No 3

Q772X (Supplementary Figure S1) The DNA bindingdomain of FANCA is located at the C-terminal C772-1455 fragment (28) Therefore it is conceivable to hy-pothesize that the C-terminal of FANCA confers theMUS81-EME1 regulating activity

Using 20 nM of protein where WT-FANCA stimulatesMUS81-EME1-mediated incision of ICL but inhibitsincision of undamaged DNA (Figure 4A and B) wefound that both mutants showed drastic reduction instimulating MUS81-EME1 in the presence of ICLcomparing with the WT protein (Figure 5A and B

compare lanes 5ndash6 with lane 4 in the ICL panel)Additionally different from WT-FANCA both theN- and C-terminals of FANCA did not inhibit MUS81-EME1 activity on undamaged DNA (Figure 5A and Bcompare lanes 5ndash6 with lane 4 in the undamaged panel)To further evaluate the functional application of the

FANCA mutants we created an incision discriminationfactor for measuring the ability of FANCA in regulatingMUS81-EME1-mediated incision of ICL damage versusundamaged DNA Figure 5C clearly shows that bothmutants in conjunction with MUS81-EME1 lost their

A P

MUS81-EME1 Mar

kerP P P

P

Mar

kerP P P

Lane 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

B

Lane 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

P

P

P

P

~34-35nt

148nt

99nt

99nt

~34-35nt

74-nt

MUS81-EME1

74-nt

Figure 2 MUS81-EME1 incision on DNA substrates with 50-end labeling on the leading strand (A) Psoralen ICL-damaged substrates(B) Undamaged structures Titration of purified MUS81-EME1 (3 nM 6 nM and 12 nM) on DNA substrates shown on the top of each gel Theschematic appearance of the products after incision was shown on the right Letter P with a circle indicates g-32P labeling by T4 polynucleotidekinase Asterisk (74 nt) indicates a decayed and uncross-linked species Arrows point to the incision sites and corresponding incision products

Nucleic Acids Research 2014 Vol 42 No 3 1675

ability to discriminate an ICL from undamaged DNAduring incision In summary these data suggest thatboth N- and C-terminals of FANCA are critical for theregulation

FANCA interacts with and recruits MUS81-EME1 to theICL damage site

To examine whether FANCA interacts with MUS81-EME1 in vivo we created site-specific ICL damages inliving human cells It has been reported that cellstreated with 8-MOP form ICL damages after light activa-tion (5051) To determine the dynamics of FANCA re-sponding to ICL damage we expressed Green fluorescentprotein (GFP)-FANCA in U2OS cells and visualized themthrough confocal microscopy As expected GFP-FANCA was expressed in both nucleus and cytoplasm(Figure 6B) The cells were then treated with either 8-MOP or low-energy 405 nm laser beam or a combinationof both FANCA did not respond to laser treatmentindicating minimum damage formation by the laseralone (Figure 6B) Next we pretreated the cells with 8-MOP and induced ICL formation through the laser beamirradiation Surprisingly only 2min after laser treatmentFANCA was efficiently recruited to the lesion created inthe path of the laser where ICLs formed (Figure 6B)

To investigate whether FANCA interacts with MUS81on ICL damages we then knocked down endogenousFANCA level by siRNA (Figure 6A) and transfectedthe knockdown U2OS cells with GFP-MUS81 beforetreatment with 8-MOP andor laser (Figure 6C) Asexpected GFP-MUS81 was only found in nucleus LikeFANCA GFP-MUS81 did not respond to laser treat-ment further proving that the laser produces minimaldamage (Fig 6C 405 nm laser panel) After inducingICL formation by both 8-MOP and laser we observedeffective MUS81 recruitment to the ICL damage siteswhen the cells were mock treated or treated with controlsiRNA However recruitment of MUS81 to the ICL sitediminished when FANCA was successfully knocked down(Figure 6C siFANCA panel) This result suggests thatrecruitment of MUS81 to ICL sites is dependent onFANCA and that FANCA functions upstream ofMUS81 in the ICL repair pathway

DISCUSSION

Heterodimeric DNA endonuclease MUS81-EME1 isknown to be involved in incision of the ICLs in mamma-lian cells (48) and has been proposed to cut the leadingstrand at a replication fork junction site at the 30 side of

GGTA

TGCAACATTTTGCTGCCGGTCACTTAAGCTCGAGCCAT3rsquo

3rsquo

5rsquo

CCGGTA

TGCAACATTTTGCTGCCGGTCACTTAAGCTCGAGCCAT3rsquo

3rsquo

5rsquo

MUS81EME1

ICL

Und

amag

edM

arke

r

99

71

60

50

434140393736353433

30

38

P148 mer

74 mer

P

MUS81EME1

5rsquo

TACGTTGTAAAACGACGGCCAGTGAATTCGAGCTCTC

P

Lane 1 2 3

35

34

5rsquo

TACGTTGTAAAACGACGGCCAGTGAATTCGAGCT

Figure 3 Psoralen ICL alters the MUS81-EME1 incision site Lane 1 32P labeled DNA markers shown in nucleotide on the left A total 3 nM ofpurified MUS81-EME1 was incubated with the undamaged (Lane 2) and ICL-damaged 30-ssDNA branch substrates (Lane 3) The sequence andschematic appearance of the products after incision were shown on the right A 74-mer indicated a decayed and uncross-linked species Letter P witha circle indicates 50-32P labeling

1676 Nucleic Acids Research 2014 Vol 42 No 3

ICL damage (33ndash36) Using a defined psoralen ICL sub-strate we have provided the first biochemical evidencethat human MUS81-EME1 does incise ICL-damagedfork structures However unlike the previously proposed30 side cleavage models MUS81-EME1 incises the leadingstrand at the 50 side of an ICL lesion This discrepancy wasstartling at first look Nevertheless further in-depthanalysis revealed that our results are compatible with thepreviously published incision behavior and substratespecificity of MUS81-EME1 Similar to previous

observations using yeast proteins our results showedthat purified human MUS81-EME1 incises 30 ssDNAbranch and replication fork structures 4 nt away fromthe junction site (3752) This special activity is likely tolend MUS81-EME1 an ability to incise at the 50 side ofDNA across the ICL damage Theoretically cleavageto the 30 side could happen only if the incision occurslong before the nascent strand reaches the ICL damagebecause of the incision activity of MUS81-EME1 on un-damaged DNA

FANCA

P P

Mar

ker

P

P

A

Undamaged Psoralen ICL

FANCA (nM)

MU

S81-

EME1

act

ivity

()

-20

-10

0

10

20

30

40

0 5 10 15 20

UndamagedICL

99

74

3534

Lane 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

B

C

P

D E FANCA

P

FANCA

Psoralen mono BPDE-dG

Sub

Sub

Sub

Sub

Incision () 0 12 21 22 41 20 11 2 Incision () 0 6 12 48 59 15 1 0

Phase I Phase II

0 0

0 0

Stim

ulat

ion

fold

P

PP

-202468

10121416

Undamaged

ICL

FANCA (nM)0 125 25 5 10 15 20

Figure 4 Effect of FANCA on MUS81-EME1-mediated DNA incision (A) Undamaged and psoralen ICL-damaged substrates were incubated with15 nM of MUS81-EME1 respectively and an increasing concentration of FANCA (0 125 25 5 10 15 and 20 nM) DNA markers are shown innucleotide on the left Same marker as described in Figure 3 was used (B) Quantitation of three independent experiments MUS81-EME1 activitywas calculated as percentage of incision products out of the input substrates The experiment without FANCA was normalized to 0 and used tocalibrate all other experiments with the indicated amount of FANCA Error bars standard deviation Dashed line Phase I and Phase II border(C) Fold changes of the MUS81-EME1 regulation by FANCA The experiment without FANCA was arbitrarily normalized to 0 for fold changeand used to calibrate all other experiments with the indicated amount of FANCA Error bars standard deviation Psoralen mono-adducted (D) anda benzo(a)pyrene diolepoxide deoxyguanosine adducted 30 ssDNA branch structures (E) were incubated with 15 nM of MUS81-EME1 and anincreasing concentration of FANCA (0 125 25 5 10 15 and 20 nM) The schematic appearance of the products after incision are shown on theright Letter P with a circle indicates 50-32P labeling

Nucleic Acids Research 2014 Vol 42 No 3 1677

Experimental evidence using Xenopus egg extracts andICL-damaged plasmids showed that incision of ICLshappens after the replication forks in opposite directionsconverge at the ICL site (34) However in some occa-sions replication forks may not converge at the ICL sitebecause of chromatin structures andor relative distancebetween replication origins and the ICL damage The ex-cellent work done by using ICL-damaged plasmids may

represent many but not necessarily all situations for rep-lication fork stalling In an occasion where the leadingstrand DNA synthesis pauses 24 nt upstream of theICL the DNA unwinding and lagging strand DNA syn-thesis may uncouple from the leading strand synthesisThe resulting DNA structure from this occasion resemblesthe ICL-damaged 30 flap structure we used in our studyNonetheless studies from McHughrsquos group provided

ICL

Undamaged

Dis

crim

inat

ion

Fact

or (

)

P

P

Sub

MU

S81-

EME1

FAN

CA-

WT

Q77

2X

C77

2-14

55

Mar

ker

A

C

Lane 1 2 3 4 5 6

B

Lane 1 2 3 4 5 6

P

P

0

20

40

60

80

100

120ICL damage

Rel

ativ

e In

cisi

on (

)

FAN

CA-

WT

Q77

2X

C77

2-14

55

MU

S81-

EME1

0

50

100

150

200

250

300

350Undamaged

Rel

ativ

e In

cisi

on (

)

FAN

CA-

WT

Q77

2X

C77

2-14

55

MU

S81-

EME1

100

39 810

20

40

60

80

100

120

FAN

CA-

WT

Q77

2X

C77

2-14

55

Figure 5 Effect of FANCA mutants on MUS81-EME1-mediated DNA incision (A) Psoralen ICL-damaged (top) and ICL-undamaged DNAsubstrates (bottom panel) were incubated with 15 nM of MUS81-EME1 and 20 nM of purified WT and mutant FANCA proteins as indicatedDNA markers are shown on the left The schematic appearance of the products after incision are shown on the right Letter P with a circle indicates50-32P labeling Asterisk indicates a decayed and uncross-linked species (B) Quantitation of three independent experiments Incision efficiency wasnormalized to WT-FANCA (arbitrarily assigned as 100) Error bars standard deviation (C) Ability to discriminate ICL from undamaged DNAby FANCA mutants The discrimination factor was calculated by dividing the relative incision rate of ICL of a protein in (B) by the relative incisionrate of undamaged DNA of the same protein The discrimination factor for FANCA was arbitrarily assigned as 100

1678 Nucleic Acids Research 2014 Vol 42 No 3

strong evidence that DNA exonuclease SNM1A collabor-ates with endonuclease XPF-ERCC1 to initiate ICL repair(53) MUS81-EME1 may only act as an alternative mech-anism when SNM1A and XPF-ERCC1 fail (3453)

The most important discovery of our studies isthat FANCA regulates MUS81-EME1-mediated DNAincision positively and negatively depending on thetype of DNA damage Enhancement of MUS81-EME1activity in the presence of the psoralen ICL damage helpsto repair the damage more effectively and suppression ofMUS81-EME1 in the absence of ICL damage helps in

protecting replication forks from being attacked due toother DNA damage affecting only one strand This inturn prevents the production of more deleteriousdamages such as DSBs Both events are beneficial forthe maintenance of replication forks It is conceivablethat the ICL damage is encountered and recognized bycomponents of the replication machinery The DNAbinding activity of FANCA may serve to verify thepresence of an ICL stalled replication fork If an ICL isconfirmed FANCA will recruit and activate MUS81-EME1 for efficient and precise ICL incision (Figures 6

8-MOPLaser +

+ ++

__

__

8-MOPLaser +

+ ++

__

__

GFP-MUS81

B

C

Mock

siCtrl

siFANCA

A

anti-FANCA

Actin

kcoM

lrtCi s

AC

NAFi s

GFP-FANCA

Figure 6 Interaction of FANCA and MUS81 on ICL damage in living cells (A) Western blot of FANCA knockdown U2OS cells weretransfected with an ON-TARGET Plus SMART Pool siFANCA a control siRNA (siCtrl) and dH2O (Mock) through lipofectamine Forty-eight hours later 50 mg of whole cell protein extract was prepared for western blot analysis using a FANCA-specific antibody Actin is the loadingcontrol (B) U2OS cells were transfected with GFP-FANCA and treated with 8-MOP andor a 405 nm laser beam as indicated Laser passing path isindicated by yellow arrows The panel is representative of 25 examined nuclei of each treatment 2525 showed the laser-induced FANCA stripes inthe presence of 8-MOP (C) Mock control siRNA and siFANCA treated U2OS cells were transfected with GFP-MUS81 and treated with 8-MOPandor a 405 nm laser beam as indicated Laser passing path is indicated by yellow arrows The panel is representative of 25 examined nuclei of eachtreatment 2525 showed the laser-induced MUS81 stripes in the presence of 8-MOP for the mock and control treatment 2525 of siFANCA nucleidid not show the laser-induced MUS81 stripes in the presence of 8-MOP

Nucleic Acids Research 2014 Vol 42 No 3 1679

and 7 ICL) If non-ICL damage is detected FANCA willprevent MUS81-EME1 from incising on the stalled repli-cation forks a mechanism that is yet to be elucidated(Figures 4 and 7 non-ICL) This working model reconcileswith our observations and explains why MUS81-EME1promotes ICL unhooking yet avoids non-specific incisionof undamaged or nonndashICL-damaged forks (48) Thisscenario is similar to the damage recognition and verifica-tion steps in nucleotide excision repair where Xerodermapigmentosum groupC (XPC) recognizesDNAdamage andXeroderma pigmentosum group A (XPA)-ReplicationProtein A (RPA) verifies before incision to prevent un-necessary incisions (54) This process definitely requiresboth the DNA binding activity at the C-terminal and anunidentified function at the N-terminal of FANCA (Figure5) Whether FANCA recognizes ICL and interacts withfactors other thanMUS81-EME1 for efficient ICL incisionneeds to be addressed in future studiesIt remains a mystery why low concentration of FANCA

stimulates MUS81-EME1 activity in the absence of DNAdamage although the stimulatory activity is lower than inthe presence of ICL (Figure 4) Because MUS81-EME1 isalso involved in resolution of Holliday junction which is alater step in the Fanconi anemia pathway of ICL repair(3655) we speculate that FANCA may additionally beinvolved in regulation of Holliday junction resolutioncatalyzed by MUS81-EME1 It would be interesting toaddress whether and how FANCA affects MUS81-EME1-mediated incision of Holliday junctions

FANCM has been considered an ICL damage recog-nition factor because it can stabilize and remodel stalledreplication forks thus it may provide temporal andspatial access for the damage to be repaired (5657)FANCM also appears to be required for assembly ofthe FA core complex onto chromatin and subsequentmonoubiquitination of the FANCIndashFANCD2 complex(83358ndash65) Controversially some other reportsdemonstrated that FANCM is not required for the for-mation of the eight-subunit core complex and FANCMnull cells are only partially defective in damage-inducedFANCD2 monoubiquitination (586667) Evidencefrom FANCM knockout mice further demonstratedthat FANCM may have a stimulatory but not essentialrole in monoubiquitinating FANCD2 (68) Furthermorea direct interacting partner for FANCM-FAAP24 in theFA core complex has not been identified thus faralthough FANCM-FAAP24 was originally identifiedthrough protein association in a FANCA-specificimmunoprecipitation assay (226069) AdditionallyFANCM cells are sensitive to camptothecin a topo-isomerase inhibitor Susceptibility to camptothecin is aunique feature identified only for downstream repairfactors such as FANCD1BRCA2 and FANCNPALB2 but not for components of the FA corecomplex (66) In summary these observations suggestthat FANCM may act downstream of FANCD2 andtherefore the upstream FA core complex may be re-cruited to DNA through other mechanisms One of

Blockage of replication

damage by FANCA

Endonucleaserecruitment

MUS81-EME1

ICL damage Endonucleaseinactivation

Non-ICL damage

ICL Non-ICL

MUS81-EME1

FANCA FANCA

Figure 7 A working model for the regulation of MUS81-EME1 activity by FANCA Blockage of DNA replication forks by an unknown damage(black box and question mark) initiates recruitment of FANCA (blue oval) FANCA recognizes ICL damage (red zigzag) and subsequently recruitsMUS81-EME1 to incise the leading strand at the 50 side of the ICL If the damage only affects one DNA strand (green block) or there is no damageFANCA will inactivate MUS81-EME1 to prevent unnecessary incisions Red arrow DNA incision red cross inactivation

1680 Nucleic Acids Research 2014 Vol 42 No 3

such mechanisms is likely through the DNA bindingactivity of FANCA (28)

FANCA also interacts and colocalizes with XPF-ERCC1 another important ICL unhooking DNA endo-nuclease (70ndash73) Both MUS81-EME1 and XPF-ERCC1interact with FANCPSLX4 a newly identified Fanconianemia protein The FANCP-SLX1ndashXPF-ERCC1ndashMUS81-EME1 tri-endonuclease complex has beendemonstrated to orchestrate nuclease actions during ICLincision and subsequent fork re-establishment throughhomologous recombination (6343674ndash79) It would beinteresting to examine whether FANCA also regulatesXPF-ERCC1 and the tri-endonuclease complex for effi-cient ICL incision as well as homologous recombinationusing our in vitro reconstitution system

In summary this report provides novel insight into theincision behavior of MUS81-EME1 on ICL damageand establishes that FANCA contributes to the mainten-ance of replication forks by directly regulating the incisionactivity of MUS81-EME1 in a damage-dependentmanner

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

The authors appreciate generous support from DrNicholas E Geacintov at New York University byproviding them a benzo(a)pyrene diolepoxide-dG adductand Dr Agata Smogorzewska at the RockefellerUniversity by providing them RA3087 FANCA null andcomplemented cell lines They are grateful to criticalcomments and discussion by Drs Murray P DeutscherPeggy Hsieh and Agata Smogorzewska

FUNDING

National Institutes of Health (NIH) [HL105631 to YZ]in part by [AG045545 to LL] University of PittsburghMedical Center CMRF (to LL) Funding for open accesscharge NIH

Conflict of interest statement None declared

REFERENCES

1 AkkariYM BatemanRL ReifsteckCA OlsonSB andGrompeM (2000) DNA replication is required to elicit cellularresponses to psoralen-induced DNA interstrand cross-links MolCell Biol 20 8283ndash8289

2 De SilvaIU McHughPJ ClingenPH and HartleyJA (2000)Defining the roles of nucleotide excision repair and recombinationin the repair of DNA interstrand cross-links in mammalian cellsMol Cell Biol 20 7980ndash7990

3 RaschleM KnipsheerP EnoiuM AngelovT SunJGriffithJD EllenbergerTE ScharerOD and WalterJC(2008) Mechanism of replication-coupled DNA interstrandcrosslink repair Cell 134 969ndash980

4 KnipscheerP RaschleM SmogorzewskaA EnoiuM HoTVScharerOD ElledgeSJ and WalterJC (2009) The Fanconi

anemia pathway promotes replication-dependent DNA interstrandcross-link repair Science 326 1698ndash1701

5 WangW (2007) Emergence of a DNA-damage response networkconsisting of Fanconi anaemia and BRCA proteins Nat RevGenet 8 735ndash748

6 KimY LachFP DesettyR HanenbergH AuerbachADand SmogorzewskaA (2011) Mutations of the SLX4 gene inFanconi anemia Nat Genet 43 142ndash146

7 KeeY and DrsquoAndreaAD (2010) Expanded roles of the Fanconianemia pathway in preserving genomic stability Genes Dev 241680ndash1694

8 SmogorzewskaA MatsuokaS VinciguerraPMcDonaldER 3rd HurovKE LuoJ BallifBA GygiSPHofmannK DrsquoAndreaAD et al (2007) Identification of theFANCI protein a monoubiquitinated FANCD2 paralog requiredfor DNA repair Cell 129 289ndash301

9 YuanF SongL QianL HuJJ and ZhangY (2010)Assembling an orchestra Fanconi anemia pathway of DNArepair Front Biosci 15 1131ndash1149

10 DronkertML and KanaarR (2001) Repair of DNA interstrandcross-links Mutat Res 486 217ndash247

11 MathewCG (2006) Fanconi anaemia genes and susceptibility tocancer Oncogene 25 5875ndash5884

12 LevitusM JoenjeH and de WinterJP (2006) The Fanconianemia pathway of genomic maintenance Cell Oncol 28 3ndash29

13 ThompsonLH HinzJM YamadaNA and JonesNJ (2005)How Fanconi anemia proteins promote the four Rs replicationrecombination repair and recovery Environ Mol Mutagen 45128ndash142

14 NiedernhoferLJ LalaiAS and HoeijmakersJH (2005)Fanconi anemia (cross)linked to DNA repair Cell 1231191ndash1198

15 AuerbachAD (2009) Fanconi anemia and its diagnosis MutatRes 668 4ndash10

16 PatelKJ and JoenjeH (2007) Fanconi anemia and DNAreplication repair DNA Repair (Amst) 6 885ndash890

17 KottemannMC and SmogorzewskaA (2013) Fanconi anaemiaand the repair of Watson and Crick DNA crosslinks Nature493 356ndash363

18 KeeY and DrsquoAndreaAD (2012) Molecular pathogenesis andclinical management of Fanconi anemia J Clin Invest 1223799ndash3806

19 SasakiMS (1975) Is Fanconirsquos anaemia defective in a processessential to the repair of DNA cross links Nature 257 501ndash503

20 HinzJM NhamPB UrbinSS JonesIM andThompsonLH (2007) Disparate contributions of the Fanconianemia pathway and homologous recombination in preventingspontaneous mutagenesis Nucleic Acids Res 35 3733ndash3740

21 SobeckA StoneS CostanzoV de GraafB ReuterT deWinterJ WallischM AkkariY OlsonS WangW et al(2006) Fanconi anemia proteins are required to preventaccumulation of replication-associated DNA double-strandbreaks Mol Cell Biol 26 425ndash437

22 ThompsonLH and HinzJM (2009) Cellular and molecularconsequences of defective Fanconi anemia proteins in replication-coupled DNA repair mechanistic insights Mutat Res 66854ndash72

23 WangLC StoneS HoatlinME and GautierJ (2008) Fanconianemia proteins stabilize replication forks DNA Repair (Amst)7 1973ndash1981

24 YuanF El HokayemJ ZhouW and ZhangY (2009) FANCIprotein binds to DNA and interacts with FANCD2 to recognizebranched structures J Biol Chem 284 24443ndash24452

25 XueY LiY GuoR LingC and WangW (2008) FANCM ofthe Fanconi anemia core complex is required for bothmonoubiquitination and DNA repair Hum Mol Genet 171641ndash1652

26 OtsukiT FurukawaY IkedaK EndoH YamashitaTShinoharaA IwamatsuA OzawaK and LiuJM (2001)Fanconi anemia protein FANCA associates with BRG1 acomponent of the human SWISNF complex Hum Mol Genet10 2651ndash2660

27 QiaoF MossA and KupferGM (2001) Fanconi anemiaproteins localize to chromatin and the nuclear matrix in a DNA

Nucleic Acids Research 2014 Vol 42 No 3 1681

damage- and cell cycle-regulated manner J Biol Chem 27623391ndash23396

28 YuanF QianL ZhaoX LiuJY SongL DrsquoUrsoG JainCand ZhangY (2012) Fanconi anemia complementation group A(FANCA) protein has intrinsic affinity for nucleic acids withpreference for single-stranded forms J Biol Chem 2874800ndash4807

29 PapadopouloD AverbeckD and MoustacchiE (1987) The fateof 8-methoxypsoralen-photoinduced DNA interstrand crosslinksin Fanconirsquos anemia cells of defined genetic complementationgroups Mutat Res 184 271ndash280

30 KumaresanKR and LambertMW (2000) Fanconi anemiacomplementation group A cells are defective in ability to produceincisions at sites of psoralen interstrand cross-linksCarcinogenesis 21 741ndash751

31 NiedernhoferLJ OdijkH BudzowskaM van DrunenEMaasA TheilAF de WitJ JaspersNG BeverlooHBHoeijmakersJH et al (2004) The structure-specific endonucleaseErcc1-Xpf is required to resolve DNA interstrand cross-link-induced double-strand breaks Mol Cell Biol 24 5776ndash5787

32 RothfussA and GrompeM (2004) Repair kinetics of genomicinterstrand DNA cross-links evidence for DNA double-strandbreak-dependent activation of the Fanconi anemiaBRCApathway Mol Cell Biol 24 123ndash134

33 CicciaA McDonaldN and WestSC (2008) Structural andfunctional relationships of the XPFMUS81 family of proteinsAnnu Rev Biochem 77 259ndash287

34 SengerovaB WangAT and McHughPJ (2011) Orchestratingthe nucleases involved in DNA interstrand cross-link (ICL)repair Cell Cycle 10 3999ndash4008

35 BhagwatN OlsenAL WangAT HanadaK StuckertPKanaarR DrsquoAndreaA NiedernhoferLJ and McHughPJ(2009) XPF-ERCC1 participates in the Fanconi anemia pathwayof cross-link repair Mol Cell Biol 29 6427ndash6437

36 KimY SpitzGS VeturiU LachFP AuerbachAD andSmogorzewskaA (2013) Regulation of multiple DNA repairpathways by the Fanconi anemia protein SLX4 Blood 12154ndash63

37 Bastin-ShanowerSA FrickeWM MullenJR and BrillSJ(2003) The mechanism of Mus81-Mms4 cleavage site selectiondistinguishes it from the homologous endonuclease Rad1-Rad10Mol Cell Biol 23 3487ndash3496

38 CicciaA ConstantinouA and WestSC (2003) Identificationand characterization of the human mus81-eme1 endonucleaseJ Biol Chem 278 25172ndash25178

39 GaillardPH NoguchiE ShanahanP and RussellP (2003) Theendogenous Mus81-Eme1 complex resolves Holliday junctions bya nick and counternick mechanism Mol Cell 12 747ndash759

40 OsmanF DixonJ DoeCL and WhitbyMC (2003)Generating crossovers by resolution of nicked Holliday junctionsa role for Mus81-Eme1 in meiosis Mol Cell 12 761ndash774

41 GaskellLJ OsmanF GilbertRJ and WhitbyMC (2007)Mus81 cleavage of Holliday junctions a failsafe for processingmeiotic recombination intermediates EMBO J 26 1891ndash1901

42 BoddyMN GaillardPH McDonaldWH ShanahanPYatesJR 3rd and RussellP (2001) Mus81-Eme1 are essentialcomponents of a Holliday junction resolvase Cell 107 537ndash548

43 ChenXB MelchionnaR DenisCM GaillardPH BlasinaAVan de WeyerI BoddyMN RussellP VialardJ andMcGowanCH (2001) Human Mus81-associated endonucleasecleaves Holliday junctions in vitro Mol Cell 8 1117ndash1127

44 ConstantinouA ChenXB McGowanCH and WestSC(2002) Holliday junction resolution in human cells two junctionendonucleases with distinct substrate specificities EMBO J 215577ndash5585

45 KaliramanV MullenJR FrickeWM Bastin-ShanowerSAand BrillSJ (2001) Functional overlap between Sgs1-Top3 andthe Mms4-Mus81 endonuclease Genes Dev 15 2730ndash2740

46 OgruncM and SancarA (2003) Identification andcharacterization of human MUS81-MMS4 structure-specificendonuclease J Biol Chem 278 21715ndash21720

47 DoeCL AhnJS DixonJ and WhitbyMC (2002) Mus81-Eme1 and Rqh1 involvement in processing stalled and collapsedreplication forks J Biol Chem 277 32753ndash32759

48 HanadaK BudzowskaM ModestiM MaasA WymanCEssersJ and KanaarR (2006) The structure-specificendonuclease Mus81-Eme1 promotes conversion of interstrandDNA crosslinks into double-strands breaks EMBO J 254921ndash4932

49 RechkoblitO ZhangY GuoD WangZ AminSKrzeminskyJ LounevaN and GeacintovNE (2002) trans-Lesion synthesis past bulky benzo[a]pyrene diol epoxide N2-dGand N6-dA lesions catalyzed by DNA bypass polymerasesJ Biol Chem 277 30488ndash30494

50 GoldRL AndersonRR NatoliVD and GangeRW (1988)An action spectrum for photoinduction of prolonged cutaneousphotosensitivity by topical 8-MOP J Invest Dermatol 90818ndash822

51 CruzLA GuechevaTN BonatoD and HenriquesJA (2012)Relationships between chromatin remodeling and DNA damagerepair induced by 8-methoxypsoralen and UVA in yeastSaccharomyces cerevisiae Genet Mol Biol 35 1052ndash1059

52 WhitbyMC OsmanF and DixonJ (2003) Cleavage of modelreplication forks by fission yeast Mus81-Eme1 and budding yeastMus81-Mms4 J Biol Chem 278 6928ndash6935

53 WangAT SengerovaB CattellE InagawaT HartleyJMKiakosK Burgess-BrownNA SwiftLP EnzlinJHSchofieldCJ et al (2011) Human SNM1A and XPF-ERCC1collaborate to initiate DNA interstrand cross-link repair GenesDev 25 1859ndash1870

54 LeeJH ParkCJ ArunkumarAI ChazinWJ and ChoiBS(2003) NMR study on the interaction between RPA and DNAdecamer containing cis-syn cyclobutane pyrimidine dimer in thepresence of XPA implication for damage verification and strand-specific dual incision in nucleotide excision repair Nucleic AcidsRes 31 4747ndash4754

55 HollingsworthNM and BrillSJ (2004) The Mus81 solution toresolution generating meiotic crossovers without Hollidayjunctions Genes Dev 18 117ndash125

56 GariK DecailletC DelannoyM WuL and ConstantinouA(2008) Remodeling of DNA replication structures by the branchpoint translocase FANCM Proc Natl Acad Sci USA 10516107ndash16112

57 GariK DecailletC StasiakAZ StasiakA andConstantinouA (2008) The Fanconi anemia protein FANCMcan promote branch migration of Holliday junctions andreplication forks Mol Cell 29 141ndash148

58 KimJM KeeY GurtanA and DrsquoAndreaAD (2008) Cellcycle-dependent chromatin loading of the Fanconi anemia corecomplex by FANCMFAAP24 Blood 111 5215ndash5222

59 MedhurstAL Laghmani elH SteltenpoolJ FerrerMFontaineC de GrootJ RooimansMA ScheperRJMeeteiAR WangW et al (2006) Evidence for subcomplexes inthe Fanconi anemia pathway Blood 108 2072ndash2080

60 MeeteiAR MedhurstAL LingC XueY SinghTR BierPSteltenpoolJ StoneS DokalI MathewCG et al (2005) Ahuman ortholog of archaeal DNA repair protein Hef is defectivein Fanconi anemia complementation group M Nat Genet 37958ndash963

61 LingC IshiaiM AliAM MedhurstAL NevelingKKalbR YanZ XueY OostraAB AuerbachAD et al(2007) FAAP100 is essential for activation of the Fanconianemia-associated DNA damage response pathway EMBO J 262104ndash2114

62 CicciaA LingC CoulthardR YanZ XueY MeeteiARLaghmani elH JoenjeH McDonaldN de WinterJP et al(2007) Identification of FAAP24 a Fanconi anemia core complexprotein that interacts with FANCM MolCell 25 331ndash343

63 MosedaleG NiedzwiedzW AlpiA PerrinaF Pereira-LealJB JohnsonM LangevinF PaceP and PatelKJ (2005)The vertebrate Hef ortholog is a component of the Fanconianemia tumor-suppressor pathway Nat Struct Mol Biol 12763ndash771

64 Garcia-HigueraI TaniguchiT GanesanS MeynMSTimmersC HejnaJ GrompeM and DrsquoAndreaAD (2001)Interaction of the Fanconi anemia proteins and BRCA1 in acommon pathway Mol Cell 7 249ndash262

1682 Nucleic Acids Research 2014 Vol 42 No 3

65 Montes de OcaR AndreassenPR MargossianSPGregoryRC TaniguchiT WangX HoughtalingS GrompeMand DrsquoAndreaAD (2005) Regulated interaction of the Fanconianemia protein FANCD2 with chromatin Blood 105 1003ndash1009

66 SinghTR BakkerST AgarwalS JansenM GrassmanEGodthelpBC AliAM DuCH RooimansMA FanQ et al(2009) Impaired FANCD2 monoubiquitination andhypersensitivity to camptothecin uniquely characterize Fanconianemia complementation group M Blood 114 174ndash180

67 RosadoIV NiedzwiedzW AlpiAF and PatelKJ (2009) TheWalker B motif in avian FANCM is required to limit sisterchromatid exchanges but is dispensable for DNA crosslink repairNucleic Acids Res 37 4360ndash4370

68 BakkerST van de VrugtHJ RooimansMA OostraABSteltenpoolJ Delzenne-GoetteE van der WalA van derValkM JoenjeH Te RieleH et al (2009) Fancm-deficient micereveal unique features of Fanconi anemia complementation groupM Hum Mol Genet 18 3484ndash3495

69 MeeteiAR SechiS WallischM YangD YoungMKJoenjeH HoatlinME and WangW (2003) A multiproteinnuclear complex connects Fanconi anemia and Bloom syndromeMol Cell Biol 23 3417ndash3426

70 SridharanD BrownM LambertWC McMahonLW andLambertMW (2003) Nonerythroid alphaII spectrin is requiredfor recruitment of FANCA and XPF to nuclear foci induced byDNA interstrand cross-links J Cell Sci 116 823ndash835

71 ReuterTY MedhurstAL WaisfiszQ ZhiY HerterichSHoehnH GrossHJ JoenjeH HoatlinME MathewCG et al(2003) Yeast two-hybrid screens imply involvement of Fanconianemia proteins in transcription regulation cell signaling oxidativemetabolism and cellular transport Exp Cell Res 289 211ndash221

72 FisherLA BesshoM and BesshoT (2008) Processing of apsoralen DNA interstrand cross-link by XPF-ERCC1 complexin vitro J Biol Chem 283 1275ndash1281

73 KuraokaI KobertzWR ArizaRR BiggerstaffMEssigmannJM and WoodRD (2000) Repair of an interstrandDNA cross-link initiated by ERCC1-XPF repairrecombinationnuclease J Biol Chem 275 26632ndash26636

74 AndersenSL BergstralhDT KohlKP LaRocqueJRMooreCB and SekelskyJ (2009) Drosophila MUS312 and thevertebrate ortholog BTBD12 interact with DNA structure-specificendonucleases in DNA repair and recombination Mol Cell 35128ndash135

75 CrossanGP van der WeydenL RosadoIV LangevinFGaillardPH McIntyreRE GallagherF KettunenMILewisDY BrindleK et al (2011) Disruption of mouse Slx4 aregulator of structure-specific nucleases phenocopies Fanconianemia Nat Genet 43 147ndash152

76 FekairiS ScaglioneS ChahwanC TaylorER TissierACoulonS DongMQ RuseC YatesJR 3rd RussellP et al(2009) Human SLX4 is a Holliday junction resolvase subunit thatbinds multiple DNA repairrecombination endonucleases Cell138 78ndash89

77 MunozIM HainK DeclaisAC GardinerM TohGWSanchez-PulidoL HeuckmannJM TothR MacartneyTEppinkB et al (2009) Coordination of structure-specificnucleases by human SLX4BTBD12 is required for DNA repairMol Cell 35 116ndash127

78 StoepkerC HainK SchusterB Hilhorst-HofsteeYRooimansMA SteltenpoolJ OostraAB EirichKKorthofET NieuwintAW et al (2011) SLX4 a coordinatorof structure-specific endonucleases is mutated in a new Fanconianemia subtype Nat Genet 43 138ndash141

79 SvendsenJM SmogorzewskaA SowaME OrsquoConnellBCGygiSP ElledgeSJ and HarperJW (2009) MammalianBTBD12SLX4 assembles a Holliday junction resolvase and isrequired for DNA repair Cell 138 63ndash77

Nucleic Acids Research 2014 Vol 42 No 3 1683

65 Montes de OcaR AndreassenPR MargossianSPGregoryRC TaniguchiT WangX HoughtalingS GrompeMand DrsquoAndreaAD (2005) Regulated interaction of the Fanconianemia protein FANCD2 with chromatin Blood 105 1003ndash1009

66 SinghTR BakkerST AgarwalS JansenM GrassmanEGodthelpBC AliAM DuCH RooimansMA FanQ et al(2009) Impaired FANCD2 monoubiquitination andhypersensitivity to camptothecin uniquely characterize Fanconianemia complementation group M Blood 114 174ndash180

67 RosadoIV NiedzwiedzW AlpiAF and PatelKJ (2009) TheWalker B motif in avian FANCM is required to limit sisterchromatid exchanges but is dispensable for DNA crosslink repairNucleic Acids Res 37 4360ndash4370

68 BakkerST van de VrugtHJ RooimansMA OostraABSteltenpoolJ Delzenne-GoetteE van der WalA van derValkM JoenjeH Te RieleH et al (2009) Fancm-deficient micereveal unique features of Fanconi anemia complementation groupM Hum Mol Genet 18 3484ndash3495

69 MeeteiAR SechiS WallischM YangD YoungMKJoenjeH HoatlinME and WangW (2003) A multiproteinnuclear complex connects Fanconi anemia and Bloom syndromeMol Cell Biol 23 3417ndash3426

70 SridharanD BrownM LambertWC McMahonLW andLambertMW (2003) Nonerythroid alphaII spectrin is requiredfor recruitment of FANCA and XPF to nuclear foci induced byDNA interstrand cross-links J Cell Sci 116 823ndash835

71 ReuterTY MedhurstAL WaisfiszQ ZhiY HerterichSHoehnH GrossHJ JoenjeH HoatlinME MathewCG et al(2003) Yeast two-hybrid screens imply involvement of Fanconianemia proteins in transcription regulation cell signaling oxidativemetabolism and cellular transport Exp Cell Res 289 211ndash221

72 FisherLA BesshoM and BesshoT (2008) Processing of apsoralen DNA interstrand cross-link by XPF-ERCC1 complexin vitro J Biol Chem 283 1275ndash1281

73 KuraokaI KobertzWR ArizaRR BiggerstaffMEssigmannJM and WoodRD (2000) Repair of an interstrandDNA cross-link initiated by ERCC1-XPF repairrecombinationnuclease J Biol Chem 275 26632ndash26636

74 AndersenSL BergstralhDT KohlKP LaRocqueJRMooreCB and SekelskyJ (2009) Drosophila MUS312 and thevertebrate ortholog BTBD12 interact with DNA structure-specificendonucleases in DNA repair and recombination Mol Cell 35128ndash135

75 CrossanGP van der WeydenL RosadoIV LangevinFGaillardPH McIntyreRE GallagherF KettunenMILewisDY BrindleK et al (2011) Disruption of mouse Slx4 aregulator of structure-specific nucleases phenocopies Fanconianemia Nat Genet 43 147ndash152

76 FekairiS ScaglioneS ChahwanC TaylorER TissierACoulonS DongMQ RuseC YatesJR 3rd RussellP et al(2009) Human SLX4 is a Holliday junction resolvase subunit thatbinds multiple DNA repairrecombination endonucleases Cell138 78ndash89

77 MunozIM HainK DeclaisAC GardinerM TohGWSanchez-PulidoL HeuckmannJM TothR MacartneyTEppinkB et al (2009) Coordination of structure-specificnucleases by human SLX4BTBD12 is required for DNA repairMol Cell 35 116ndash127

78 StoepkerC HainK SchusterB Hilhorst-HofsteeYRooimansMA SteltenpoolJ OostraAB EirichKKorthofET NieuwintAW et al (2011) SLX4 a coordinatorof structure-specific endonucleases is mutated in a new Fanconianemia subtype Nat Genet 43 138ndash141

79 SvendsenJM SmogorzewskaA SowaME OrsquoConnellBCGygiSP ElledgeSJ and HarperJW (2009) MammalianBTBD12SLX4 assembles a Holliday junction resolvase and isrequired for DNA repair Cell 138 63ndash77

Nucleic Acids Research 2014 Vol 42 No 3 1683