Preventing Nonhomologous End Joining Suppresses DNA Repair Defects of Fanconi Anemia

Molecular Cell

Article

Preventing Nonhomologous End JoiningSuppresses DNA Repair Defects of Fanconi AnemiaAdele Adamo,1,6 Spencer J. Collis,2,3,6 Carrie A. Adelman,2 Nicola Silva,1,4,5 Zuzana Horejsi,2 Jordan D. Ward,2

Enrique Martinez-Perez,5 Simon J. Boulton,2,* and Adriana La Volpe1,*1Institute of Genetics and Biophysics ‘‘Adriano Buzzati-Traverso,’’ CNR, Via Pietro Castellino 111, 80131, Napoli, Italy2DNA Damage Response Laboratory, London Research Institute, Clare Hall, South Mimms, EN6 3LD, UK3Institute for Cancer Studies, University of Sheffield Medical School, Beech Hill Road, Sheffield, S10 2RX, UK4Department of Structural and Functional Biology, University of Naples, ‘‘Federico II,’’ Complesso di Monte S. Angelo, 80138, Napoli, Italy5Clinical Sciences Division, Imperial College, Du Cane Road, London, W12 0NN, UK6These authors contributed equally to this work*Correspondence: [email protected] (S.J.B.), [email protected] (A.L.V.)

DOI 10.1016/j.molcel.2010.06.026

SUMMARY

Fanconi anemia (FA) is a complex cancer suscepti-bility disorder associated with DNA repair defectsand infertility, yet the precise function of the FAproteins in genome maintenance remains unclear.Here we report that C. elegans FANCD2 (fcd-2) isdispensable for normal meiotic recombination butis required in crossover defective mutants to preventillegitimate repair of meiotic breaks by nonhomolo-gous end joining (NHEJ). In mitotic cells, we showthat DNA repair defects of C. elegans fcd-2 mutantsand FA-deficient human cells are significantly sup-pressed by eliminating NHEJ. Moreover, NHEJfactors are inappropriately recruited to sites of repli-cation stress in the absence of FANCD2. Our findingsare consistent with the interpretation that FA resultsfrom the promiscuous action of NHEJ during DNArepair. We propose that a critical function of the FApathway is to channel lesions into accurate, asopposed to error-prone, repair pathways.

INTRODUCTION

Fanconi anemia (FA) is a complex multigene disorder character-

ized by severe genome instability, congenital abnormalities,

acute myeloid leukemia, and/or bone marrow failure and cancer

predisposition (D’Andrea and Grompe, 2003; Fanconi, 1967;

Kennedy and D’Andrea, 2005). The hallmark of FA cells is exqui-

site sensitivity to interstrand crosslinking (ICL) agents, indicative

of a repair defect in response to agents that block the replication

fork. The FA pathway is composed of at least 13 proteins corre-

sponding to the complementation groups found mutated in FA

patients: FANCA, B, C, D1, D2, E, F, G, I, J, L, M, and N

(Moldovan and D’Andrea, 2009). In response to replication

stress, the FA core complex (comprised of FANCA, B, C, E, F,

G, L, and M) catalyzes the monoubiquitylation of the FANCD2/

FANCI heterodimer, which promotes its recruitment to damaged

replication forks. Once recruited, FANCD2/FANCI colocalizes

with BRCA2 (FANCD1), PALB2 (FANCN), and FANCJ in nuclear

repair foci (Kennedy and D’Andrea, 2005; Wang, 2007). Since

FANCJ, BRCA2, and PALB2 are essential for homologous

recombination (HR) (Boulton, 2006; D’Andrea and Grompe,

2003; Martin et al., 2005; Petalcorin et al., 2006), it has been

proposed that the FA pathway maintains genomic integrity by

facilitating HR-mediated repair, but how this occurs remains

unclear.

The nematode C. elegans has emerged as a powerful system

for the study of FA in the context of a whole organism (Youds

et al., 2009). Similar to their vertebrate counterparts, mutants

in C. elegans FCD-2 (FANCD2), FNCI-1 (FANCI), DOG-1

(FANCJ), and FCM-1 (FANCM) are exquisitely sensitive to ICL-

inducing agents and display chromosomal aberrations and

enhanced mutagenesis following treatment with these agents

(Collis et al., 2006; Muzzini et al., 2008; Youds et al., 2009).

At the molecular level, the C. elegans FCD-2 and FNCI-1 proteins

are monoubiquitylated in response to replication stress, and this

requires the replication stress checkpoint proteins ATL-1 (ATR),

CHK-1 (CHK1), and RPA-1, as well as the FA core complex

protein FCM-1 (Collis et al., 2006, 2007; Lee et al., 2010). Once

activated, FCD-2 is recruited to nuclear DNA repair foci, demon-

strating the conservation of the central elements of the FA

pathway (Collis et al., 2006; Lee et al., 2010).

The C. elegans germline also offers an excellent system to

study the repair of double strand breaks (DSBs), which are phys-

iologically generated by the topoisomerase-like protein SPO-11

during meiosis. Meiotic DSBs are preferentially repaired by HR

using a parental homolog to form interhomolog crossovers that

are essential for accurate chromosome segregation at the first

meiotic division (Page and Hawley, 2003). In wild-type C. ele-

gans, RAD-51 foci, which form at meiotic DSBs, arise during

the late zygotene and early pachytene stages, and then rapidly

decrease in number during pachytene stage as meiotic DSB

repair progresses (Colaiacovo et al., 2003; Rinaldo et al.,

2002). Both timely disappearance of RAD-51 foci at pachytene

and establishment of crossovers require the presence of the

mismatch repair related proteins MSH-4 and MSH-5 (Kelly

et al., 2000; Zalevsky et al., 1999) and the assembly of the syn-

aptonemal complex (SC), a proteinaceous structure that is

Molecular Cell 39, 25–35, July 9, 2010 ª2010 Elsevier Inc. 25

mailto:[email protected]

mailto:[email protected]

Molecular Cell

Eliminating NHEJ Suppresses Repair Defects of FA

required to maintain homologous chromosomes in close align-

ment (Colaiacovo et al., 2003). In C. elegans, meiotic DSBs are

not a prerequisite for SC assembly (Dernburg et al., 1998), and

axis morphogenesis and SC formation are not required for

loading of the RAD-51 strand-exchange protein onto DSBs

(Colaiacovo et al., 2003; Rinaldo et al., 2002). These attributes

of C. elegans meiosis allow factors that affect synapsis to be

distinguished from those required for meiotic DSB repair. For

instance, despite a failure to form crossovers, the lack of chro-

mosome fragments in the oocytes of msh-4/5 and SC mutants

(syp-1 or syp-2) demonstrates that these mutants retain the

ability to repair meiotic DSBs. The repair observed in msh-4/5

and syp-1/2 mutants is most likely achieved by HR repair

between sister chromatids. Indeed, recent studies have estab-

lished that C. elegans BRCA1 (BRC-1), while being dispensable

for crossover formation between homologs, is essential for inter-

sister HR repair of meiotic DSBs in SC mutants (Adamo et al.,

2008).

Prompted by the infertility of FA patients (Alter, 1993) and the

presence of meiotic defects in FANCD2 knockout mice

(Houghtaling et al., 2003), we have investigated the meiotic

role of C. elegans FCD-2. Whereas FCD-2 is dispensable for

crossover recombination in normal meiosis, loss of FCD-2 in

syp-2 (SC defective) or msh-4 (crossover defective) mutants re-

sulted in erroneous repair of meiotic DSBs, a phenotype that

could be suppressed by deleting the DNA ligase component

lig-4. Strikingly, eliminating LIG-4 activity also suppressed the

ICL-induced defects of fcd-2 mutants to levels indistinguishable

from wild-type. Similarly, deletion or inhibition of NHEJ sup-

pressed the ICL sensitivity of FA-deficient mammalian cells.

Consistent with a role for the FA pathway in ensuring that DNA

lesions are repaired with high fidelity, cells lacking FANCD2

exhibited promiscuous recruitment of DNA-PKcs to sites of repli-

cation stress. Our findings reveal that the FA pathway

suppresses illegitimate repair of meiotic DSBs and ICL lesions

by NHEJ and that inhibition of NHEJ suppresses several FA

phenotypes that stem from defects in DNA repair.

RESULTS

Crossovers Occur Normally but Meiotic RAD-51 FociAre Elevated in fcd-2 MutantsAlthough FA patients and mouse models of FA have reduced

fertility, and FANCD2 has been previously implicated in DNA

repair processes, upon initial examination, no apparent meiotic

phenotype is detectable in C. elegans fcd-2 mutants. Both the

incidence of males (0.04%), reflecting the frequency of X chro-

mosome nondisjunction during meiosis, and the levels of

embryonic lethality (0.51%), indicative of missegregation of the

autosomes, appear normal in fcd-2 mutants (Collis et al., 2006;

Polanowska et al., 2006; this study). Homolog pairing, synapsis,

and chiasmata formation also occur as normal in fcd-2 mutant,

indicating that FCD-2 is dispensable for crossing over (Figures

1A and 1B). To confirm this, we assayed the frequency of recom-

bination in the fcd-2 mutant in two different chromosomal inter-

vals: (1) dpy-5 to unc-29 on chromosome I (a region with a low

level of recombination of about 1 cM/Mb), and (2) unc-60 to

dpy-11 on chromosome V (an interval with a relatively high

26 Molecular Cell 39, 25–35, July 9, 2010 ª2010 Elsevier Inc.

recombination frequency of about 3.7 cM/Mb). In both of these

intervals the frequency of recombination observed in fcd-2

mutants (3.17% and 20.10%, respectively) was similar to that

observed in wild-type worms (Table S1). These data indicate

that fcd-2 is dispensable for crossover formation in meiosis.

Given the potential link between the FA pathway and HR-

mediated repair, we also analyzed fcd-2 mutant germlines for

HR repair events by immunostaining with anti-RAD-51 anti-

bodies. RAD-51 foci were elevated between the leptotene and

pachytene stages of meiotic prophase (Figure 2 and Table S2).

This result was confirmed in 11C3[fcd-2] worms (Figure S1),

which carry an extrachromosomal fcd-2 transgene array that

induces cosuppression of both itself and the endogenous fcd-

2 gene (Dernburg et al., 2000). In contrast, animals derived

from the 11C3 line that have lost the extrachromosomal array

(at a frequency of 43%) do not exhibit increased meiotic RAD-

51 foci (Figure S1). Analysis of RAD-51 foci in spo-11;fcd-2

double mutants, in which meiotic DSB formation is impaired,

revealed a very similar RAD-51 staining pattern to that of spo-

11 mutants. Thus, the increase in RAD-51 foci in fcd-2 mutants

might either represent an increase in DSB formation or most

likely reflects compromised repair of SPO-11-induced meiotic

DSBs, as previously suggested for other DNA repair genes

such as msh-4, msh-5, syp-2, or brc-1 (Colaiacovo et al., 2003,

Adamo et al., 2008). These results establish that loss of fcd-2

leads to accumulation of meiotic RAD-51 foci, which suggest

that FCD-2 is required for efficient conversion of HR intermedi-

ates into poststrand exchange products in meiosis.

FCD-2 Prevents Illegitimate Repair of Meiotic DSBsin Crossover Deficient MutantsThe accumulation of meiotic RAD-51 foci together with the wild-

type levels of crossovers in fcd-2 mutants is highly reminiscent of

the meiotic phenotype of brc-1 mutant (Adamo et al., 2008).

BRC-1 is required for the repair of meiotic DSBs using the sister

chromatid under circumstances where the homologs are

unavailable for repair, such as in SC mutants (syp-2) or in

mutants that lack the prointerhomolog crossover protein MSH-

4 (Adamo et al., 2008). Prompted by these observations, we

sought to determine if FCD-2 plays a similar role to BRC-1 in in-

tersister repair of meiotic DSBs. In wild-type diakinesis oocytes,

six DAPI-stained bodies, or bivalents (representing the six homo-

logs held together by chiasmata) are observed, whereas syp-2 or

him-14/MSH4 (hereafter called msh-4) mutant oocytes typically

contain 12 unpaired chromosomes (univalents) at diakinesis.

In contrast, cytological analysis revealed that the majority of

fcd-2; syp-2 double-mutant oocytes possessed between 6 and

12 DAPI-stained bodies (Figures 1A and 1B). This phenotype

contrasts with the chromosome fragmentation observed in

brc-1;syp-2 double mutants (Adamo et al., 2008). Furthermore,

fcd-2; msh-4 double mutant oocytes displayed between 4 and

12 DAPI-stained bodies (Figures 1A and 1B).

We reasoned that the presence of between 4 and 12 DAPI-

stained bodies in fcd-2; syp-2 and fcd-2;msh-4 double-mutant

oocytes may reflect noncovalent aggregation of nonhomologous

chromosomes, partial restoration of HR-mediated interhomolog

repair, or illegitimate repair between nonhomologous chromo-

somes. To discriminate between these possibilities we used

Figure 1. Lack of FCD-2 in Crossover-Deficient

Mutants Leads to Chromosome Associations

(A) Representative images of diakinesis oocytes of the

indicated genotype stained with DAPI. The number of

DAPI-stained bodies present is indicated in the bottom

right of each panel. Scale bar, 2 mm.

(B) Quantification of the number of DAPI-stained bodies in

diakinesis oocytes of the indicated genotype (see color

legend on top of the graph). The difference in DAPI-stained

bodies between fcd-2;syp-2 and syp-2, and between fcd-

2; msh-4, and msh-4 are statistically significant (Table S3).

Molecular Cell


fluorescent in situ hybridization (FISH) to label two different chro-

mosomes (III and V) in fcd-2;syp-2 and fcd-2;msh-4 germlines.

Staining of a single DAPI-stained body with both probes would

suggest the physical association of chromosome III and chromo-

some V. As expected, this event was never observed in wild-type

(45 nuclei), fcd-2 (75 nuclei), syp-2 (40 nuclei), or msh-4 (45 nuclei)

oocytes (Figure 3A). However, chromosomes III and V were

found in the same DAPI body in oocytes of both fcd-2; syp-2

(6/63) and fcd-2;msh-4 (1/56) double mutants (Figure 3A). These

Molecular C

data imply that when crossover formation is

impaired, FCD-2 prevents inappropriate en-

gagements between nonhomologous chromo-

somes at diakinesis.

FCD-2 Is Required to Prevent NHEJduring Meiotic DSB RepairThe presence of nonhomologous chromosome

associations in diakinesis oocytes of fcd-

2;syp-2 and fcd-2;msh-4 double mutants could

be readily explained if meiotic DSBs are re-

paired inappropriately by nonhomologous end

joining (NHEJ) in these mutants. To test this

hypothesis, we quantified the number of DAPI-

stained bodies in diakinesis oocytes of the triple

mutants lig-4;fcd-2;syp-2 and lig-4;fcd-2;msh-4

in which NHEJ is abolished by mutation of lig-4,

the ligase component that seals the break (Cle-

jan et al., 2006). Deletion of lig-4 in the fcd-

2;syp-2 and fcd-2;msh-4 backgrounds led to

a redistribution of the number of DAPI-stained

bodies at diakinesis from an average of 4 to 12

bodies in the fcd-2;syp-2 and fcd-2;msh-4

double mutants to predominantly 12 or more

than 12 DAPI bodies in both lig-4;fcd-2;syp-2

and lig-4;fcd-2;msh-4 triple mutants (Figure 3B

and Table S3). These results suggest that

NHEJ is responsible for the association events

arising between nonhomologous chromosomes

in the fcd-2;syp-2 and fcd-2;msh-4 mutants. It is

notable that chromosome fragmentation is also

observed in both triple mutants implying that in-

tersister repair is not completely restored by

blocking NHEJ. It is therefore possible that

FCD-2 is also required for efficient intersister

HR repair. Nevertheless, these data support

a role for FCD-2 in suppressing the promiscuous use of NHEJ

during meiotic DSB repair.

Eliminating LIG-4 Suppresses the ICL Repair Defectsof fcd-2 MutantsOur findings in meiosis raised the possibility that the FA pathway

may also prevent NHEJ during ICL repair in mitotic cells, which

we next sought to examine. Measurement of embryonic survival

following treatment with ICL-inducing agents confirmed previous

ell 39, 25–35, July 9, 2010 ª2010 Elsevier Inc. 27

Figure 2. fcd-2 Mutants Exhibit an Increase in spo-11-Dependent RAD-51 Foci

(A) Representative images of early pachytene nuclei (indicated by circles) stained with anti-RAD-51 antibodies (red) and DAPI (blue).

(B) Quantification of RAD-51 foci in germlines of the indicated genotypes. Statistical analyses are reported in Table S2.

Molecular Cell


observations that C. elegans fcd-2 mutants are hypersensitive to

cisplatin (CDDP), nitrogen mustard (HN2), and UVA-activated tri-

methylpsoralen (TMP-UVA; Figures 4A, S3A, and S3B). Remark-

ably, elimination of NHEJ in fcd-2 mutants by deletion of lig-4

rescued the CDDP sensitivity to levels indistinguishable from

wild-type (Figure 4A). Furthermore, lig-4;fcd-2 double mutants

exhibited near wild-type sensitivity following treatment with

HN2 or TMP-UVA (Figures S3A and S3B). Thus, eliminating

LIG-4 is sufficient to suppress the ICL sensitivity of fcd-2 mutants.

It had been previously reported that RAD-51 foci form at sites

of ICL lesions independently of FCD-2 (Collis et al., 2006; Lee

et al., 2010). Treatment with HN2 induces extensive accumula-

tion of RAD-51 foci in wild-type, lig-4, fcd-2, and lig-4;fcd-2

double mutants with almost 100% of mitotic nuclei showing

RAD-51 foci 12 hr post treatment (Figure 4B). Normally, as repair

proceeds, RAD-51 foci gradually disappear, and by 48 hr post-

treatment the number of foci has returned to levels approaching

the wild-type before treatment. In contrast, over 60% of nuclei in

the fcd-2 mutant still retain RAD-51 foci at the 48 hr time point,

consistent with a defect in ICL repair. Strikingly, eliminating

NHEJ by deletion of lig-4 suppressed the persistence of RAD-

51 foci in fcd-2 mutants (Figure 4B).

Persistence of DNA lesions in the C. elegans germline is suffi-

cient to trigger the DNA-damage checkpoint and induce


apoptosis (Gartner et al., 2000). Consistent with the presence

of unresolved RAD-51 foci, fcd-2 mutants displayed much higher

levels of apoptosis following CDDP treatment when compared

with wild-type or lig-4 mutant worms (Figure 4C). Eliminating

NHEJ by deletion of lig-4 in the fcd-2 mutant suppressed the

ICL-induced germ cell apoptosis to levels comparable to the

wild-type (Figure 4C). Treatment of C. elegans FA mutants with

ICL-inducing agents also leads to the accumulation of chromo-

somal abnormalities in embryos (Figure 4D), similar to the situa-

tion in vertebrate FA cells (Collis et al., 2006). In accordance with

our previous data, eliminating NHEJ in fcd-2 mutants by deletion

of lig-4 suppressed the frequency of chromosomal abnormalities

to near wild-type levels following treatment with TMP-UVA

(Figure 4D).

It remained possible that the rescue of ICL repair defects in the

fcd-2 mutant by deletion of lig-4 could reflect the use of alternate

error-prone repair pathways that, while sufficient to rescue

sensitivity, may lead to enhanced mutation frequencies. To

investigate this possibility, we determined whether eliminating

LIG-4 in fcd-2 mutants would suppress or enhance the

frequency of postembryonic developmental defects either

before or after treatment with ICL-inducing agents. Phenotypic

analysis of untreated fcd-2 mutants revealed a low level of

postembryonic developmental defects (25/3630) that was

Figure 3. LIG-4 Is Responsible for the Chromosome Associations

Observed in Crossover-Deficient Mutants Lacking FCD-2

(A) Diakinesis nuclei of the indicated genotypes labeled with two FISH probes

to chromosomes III (red) and V (green). DNA is shown in blue. Arrows in the

syp-2;fcd-2 and fcd-2;msh-4 panels indicate a single DAPI body that is stained

with both FISH probes, demonstrating a fusion event between chromosomes

III and V. Scale bar = 2 mm.

(B) Quantification of the number of DAPI-stained bodies in diakinesis oocytes

of the indicated genotype (see color legend on top of the graph). Statistical

analysis is shown in Table S3.

Molecular Cell


statistically different from wild-type (0/4587), lig-4 (0/3513), and

lig-4;fcd-2 double mutants (7/4133, fcd-2 versus fcd-2;lig-4,

p = 0.0007; Figure 4E and Table S4). Treatment with CDDP

induced postembryonic developmental defects in 31/236

(13.1%) of the surviving progeny of fcd-2 mutants. This pheno-

type was suppressed to near wild-type or lig-4 levels (9/470 =

1.9% and 12/575 = 2.09%, respectively) in the lig-4; fcd-2

double mutant (6/354 = 1.7%; Figure 4E and Table S4). These

results suggest that fcd-2 mutants exhibit increased levels of

spontaneous as well as ICL-induced developmental abnormali-

ties that are NHEJ dependent.

Blocking NHEJ Substantially Rescues the ICL RepairDefect of FANCD2-Deficient Human CellsWe next set out to determine if the ICL sensitivity of FANCD2-

deficient mammalian cells could also be suppressed by blocking

NHEJ. For these studies we exploited human MO59K (DNA-

PKcs proficient) and MO59J (DNA-PKcs deficient) cells

(Figure 5A), which were derived from a glioblastoma from a single

patient in which a reversion mutation restores the reading frame

and expression of DNA-PKcs in MO59K cells (Anderson et al.,

2001; Galloway and Allalunis-Turner, 2000; Lees-Miller et al.,

1995). DNA-PKcs is a PI3K-like kinase family member found in

vertebrates that is recruited to breaks and activated by the

KU70/KU80 heterodimer. These proteins, in conjunction with

Artemis, XRCC4, and DNA Ligase IV are required for NHEJ

(Ma et al., 2005). As expected, treatment of MO59K cells with

FANCD2 siRNAs resulted in sensitivity to mitomycin C (MMC;

Figures 5A and 5B) (Timmers et al., 2001). Similar to our results

in worms, however, DNA-PKcs-deficient MO59J cells depleted

for FANCD2 exhibited levels of survival after MMC treatment

comparable to controls (Figures 5A and 5B).

To further validate these results, we employed a potent and

specific DNA-PKcs inhibitor (PKi) NU7026 (Veuger et al., 2003).

While PKi-treated cells exhibited increased sensitivity to ionizing

radiation, as previously described (Veuger et al., 2003) (58%

survival after 2.5 Gy IR, versus 92% survival of untreated

controls), addition of PKi had no measurable impact on the

MMC sensitivity of control siRNA cells. HeLa cells treated with

FANCD2 siRNAs resulted in the expected hypersensitivity to

MMC, whereas blocking NHEJ by addition of PKi suppressed

the MMC sensitivity of FANCD2 depleted cells to levels compa-

rable to controls (Figures 5C and 5D).

We next investigated whether inhibition of NHEJ by another

means could suppress the ICL sensitivity of FANCD2-deficient

cells. For these studies we opted to treat patient-derived

PD733 cells that harbor a mutation in FANCD2 with siRNAs to

the KU80 subunit of the Ku heterodimer, which functions as

the DNA-binding component of DNA-PK. As expected, PD733

cells exhibited hypersensitivity to MMC when subjected to

control siRNA. In contrast, treatment of cells with Ku80 siRNA

suppresses the MMC sensitivity of PD733 cells (Figure S4A).

Collectively, these results indicate that depletion or inhibition of

Ku80 or DNA-PKcs is sufficient to substantially rescue the ICL

sensitivity of FANCD2-depleted human cells.

Blocking NHEJ Partially Rescues the ICL Repair Defectof FA Core Complex-Deficient CellsSince the FA core complex is responsible for the monoubiquity-

lation of FANCD2 and FANCI in response to replication stress,

we also wished to determine the effect of blocking NHEJ on


Figure 4. Deletion of lig-4 Suppresses the ICL-Induced Defects Observed in fcd-2 Mutants

(A) Embryonic survival at different time points after treatment with 180 mM CDDP (cis-diamminedichloridoplatinum-II) for the indicated genotype. Error bars repre-

sent standard deviation.

(B) Quantification of RAD-51 foci detected in the mitotic compartment of the germline at different time points after treatment with nitrogen mustard (HN2).

(C) Quantification of germline apoptosis in germlines of the indicated genotypes. Apoptosis was scored using the vital dye SYTO12 before and 48 hr post treat-

ment with 180 mM CDDP. Standard deviations were calculated from at least three independent experiments.

(D) Quantification of chromosome breakages and bridges occurring in embryos of the indicated genotypes after treatment with 10 mg/ml trimethylpsoralen (TMP)

and increasing doses of UVA (J).

(E) Graphical representation of the frequency of postembryonic developmental abnormalities in the indicated genotypes observed before and after treatment with

180 mM CDDP.

Molecular Cell


the ICL sensitivity of FA core complex mutants. For these

studies, we focused our attention on the FANCA and FANCC

subunits of the FA core complex. Treatment of HeLa cells with

FANCA siRNAs (Figure 6A) resulted in the expected sensitivity


to MMC, when compared to control (Figure 6B). Blocking

NHEJ by addition of PKi suppressed the MMC sensitivity of

FANCA siRNA-treated cells (Figure 6B). To substantiate these

findings we also analyzed the impact of blocking NHEJ in mouse

Figure 5. Inhibiting NHEJ Suppresses the

ICL Sensitivity of FANCD2-Deficient Human

Cells

(A) Western blots showing FANCD2 protein levels

in siRNA control (Con) and siRNA FancD2 (D2)-

treated MO59K and MO59J cells, DNA-PKcs

levels in untreated cells, and Actin-loading

controls.

(B) Sensitivity of control and FANCD2-deficient

MO59K and MO59J cells to increasing doses of

mitomycin C (MMC).

(C) Western blots showing FANCD2 protein levels

and Actin-loading controls in control and FANCD2

siRNA-treated HeLa cells.

(D) Sensitivity of control and FANCD2-deficient

HeLa cells incubated without or with the DNA-PK

inhibitor NU7026 (PKi) during MMC exposure.

The number of experimental repeats is shown

on each graph and the error bars represent the

standard deviations.

Figure 6. Inhibiting NHEJ Suppresses the ICL Sensitivity of FANCA- and FANCC-Deficient Cells

(A) Western blots showing FANCA and Actin loading controls in control and FANCA siRNA-treated HeLa cells.

(B) Sensitivity of control and FANCA-deficient HeLa cells incubated without or with the DNA-PK inhibitor NU7026 (PKi) during MMC exposure.

(C) Sensitivity of wild-type (WT) and FANCA-deficient mouse embryonic fibroblasts incubated without or with the DNA-PK inhibitor NU7026 (PKi) during MMC

exposure.

(D) Sensitivity of WT and FANCC-deficient mouse embryonic fibroblasts incubated without or with the DNA-PK inhibitor NU7026 (PKi) during MMC exposure. The

WT MEF data is the same for (C) and (D) as experiments were performed in parallel. The data for FancA and FancC are presented in separate panels (C and D) to

improve clarity. The number of experimental repeats is shown on each graph and the error bars represent the standard deviations.

Molecular Cell


embryonic fibroblasts (MEFs) derived from FancA�/�or FancC�/�

knockout mice (Chen et al., 1996; Cheng et al., 2000). Consistent

with our findings with FANCA-deficient HeLa cells, PKi treatment

of FancA�/� or FancC�/�knockout MEFs also conferred suppres-

sion of MMC sensitivity (Figures 6C and 6D). Furthermore,

treatment of cells with Ku80 siRNA led to suppression of MMC

sensitivity of the FA patient derived cell line PD331, which carry

a mutation in FancC (Figure S4B). Our results indicate that

blocking NHEJ can also suppress the ICL sensitivity of FA core

complex-deficient cells.

Inappropriate Accumulation of DNA-PKcs at Sitesof Replication Stress Occurs in the Absence of FANCD2Given that NHEJ factors have a very high affinity for DNA ends,

we reasoned that the FA pathway might be required to prevent

the inappropriate engagement of these factors with DSB inter-

mediates formed during the ICL repair process. To investigate

this possibility, we utilized an antibody raised against phosphor-

ylated serine 2056 (pS2056) on DNA-PKcs that is able to detect

DNA-PK at sites of DSBs by immunofluorescence. Importantly,

phosphorylation of DNA-PKcs on serine 2056 is induced in

response to DSBs generated by ionizing radiation, and is not

normally induced by hydroxyurea (HU) (Yajima et al., 2006). Im-

munostaining for pS2056 DNA-PKcs revealed low levels of chro-

matin-associated foci in untreated wild-type human foreskin

fibroblasts (HFF) and FANCD2-deficient PD20 cells (Figures 7A

and 7B). IR treatment led to the formation of multiple pS2056

DNA-PKcs foci in 100% of HFF and PD20 cells (Figures 7A

and 7B). Consistent with previous studies, pS2056 DNA-PKcs

foci were not significantly induced in wild-type cells following

treatment with HU (Yajima et al., 2006). Furthermore, pS2056

DNA-PKcs foci were only moderately induced following treat-

ment with MMC. In contrast, treatment of FANCD2-deficient

PD20 cells with either HU or MMC resulted in the accumulation

of multiple pS2056 DNA-PKcs foci in 60%–80% of cells,

depending on the treatment (Figures 7A and 7B). As this phos-

phorylation event on DNA-PKcs is cell-cycle regulated (Chen

et al., 2005), these differences could result from gross changes

in cell-cycle distribution between HFF and PD20 cells following

exposure to HU or MMC. However, fluorescence activated

cell-sorting analysis excluded this possibility as both the control

and FANCD2-deficient cells exhibited similar cell-cycle profiles

following each of the HU, MMC, and IR treatments (Figure S6).

Furthermore, the elevated levels of pS2056 foci observed in

HU or MMC-treated PD20 cells could not be explained by differ-

ences in the number of DSBs generated in response to these

agents as the levels of gH2AX foci were comparable between

control and FANCD2-deficient cells (Figure 7C). Similar results

were also obtained in HeLa cells following siRNA depletion of

FANCD2 (Figure S5). Collectively, these data indicate that

FANCD2 is required to prevent the inappropriate engagement

of DNA-PKcs with damaged/stalled replication forks.

DISCUSSION

Here we report that DNA repair defects observed in C. elegans

and mammalian cells deficient for FA components can be sup-

pressed by preventing repair by NHEJ. We observed that C. el-

egans FCD-2 is dispensable for normal meiotic crossover

recombination but is required in syp-2 and msh-4 mutants to

prevent erroneous repair of meiotic DSBs through NHEJ. Our

data suggest that FCD-2 functions to ensure that meiotic

DSBs are channeled into HR as opposed to NHEJ repair path-

ways. Since meiotic RAD-51 foci accumulate in the early phases

of meiosis in the fcd-2 mutant and intersister repair is not

completely restored in the triple mutants fcd-2;syp-2;lig-4 and

fcd-2;msh-4;lig-4, it would appear that not all of the roles of

FCD-2 can be rescued by eliminating NHEJ. In particular, we

cannot exclude the possibility that FCD-2 may also play a role

in promoting efficient DSB repair through intersister HR in


Figure 7. FANCD2 Prevents the Inappropriate Engagement of DNA-PKcs at Damaged Replication Forks

(A and B) Representative images (A) and quantification (B) of phospho-Ser2056 DNA-PKcs foci in untreated control (HFF) and FANCD2-deficient (PD20) cells, and

cells treated with hydroxurea (HU), mitomycin C (MMC) or ionizing radiation (IR).

(C) gH2AX foci in cells treated as in (A). Cells were treated with 3 mM HU, 80 ng/ml MMC, or 5Gy IR and fixed 2 hr, 16 hr, and 30 min later, respectively. Error bars in

(B) and (C) represent the standard deviation for three experiments.

(D) Model illustrating the function of the FA pathway in suppressing NHEJ during ICL repair. Black and red represent WT and FA-deficient cells, respectively. The

dashed red arrow represents competition between HR and NHEJ that arises in FA.

Molecular Cell


meiosis, a function that is unlikely to be rescued by blocking

NHEJ. A role during intersister HR would be consistent with

previous observations that Fancd2-deficient cells exhibit an

approximate 2-fold reduction in HR repair of a site specific

DSB in mammalian cells (Nakanishi et al., 2005).

More strikingly, we demonstrate that eliminating NHEJ leads

to substantial rescue of the ICL-induced repair defects of

fcd-2 mutants, including ICL sensitivity, persistent RAD-51 foci,

damage-induced chromosomal abnormalities, increased

apoptosis, and increased postembryonic developmental

defects. Currently it is unclear why inappropriate utilization of

NHEJ should lead to persistence of RAD-51 foci during ICL repair.

Understanding this problem is also compounded by the fact that

there is considerable uncertainty about the structure of the

replication fork and how DNA lesions alter the fork in vivo. Never-

theless, it is known that HR proteins are recruited to damaged

replication forks and it is inferred that there are likely to be at least

two substrates for binding of HR factors following processing of

an ICL lesion based on published work: (1) ssDNA gaps that are

bound by HR proteins in order to protect the nascent strand

from degradation, and (2) one-ended DSBs formed following


ICL incision, which may also load HR proteins following DSB pro-

cessing. Based on our findings we believe that FA may counteract

NHEJ from inappropriate engagement with one-ended DSBs but

not with ssDNA gaps; NHEJ requires a free end to load and does

not load onto gaps. HR proteins ‘‘protecting’’ ssDNA gaps at the

fork are likely to persist until the one-ended DSB is repaired by

NHEJ, which may be considerably slower than repair of the lesion

by HR in wild-type. Hence, RAD-51 foci associated with ssDNA

gaps could persist in fcd-2 mutants following ICL treatment,

and this is suppressed by blocking NHEJ.

Our observations in C. elegans are reinforced by results

showing that the ICL sensitivity of FANCD2-deficient human

cells is also rescued by blocking NHEJ, which we observed

through inhibition/deletion of DNA-PKcs or depletion of Ku80.

We also demonstrate that the ICL repair defect of FA core-

complex-deficient cells is suppressed by inhibition of NHEJ. It

is notable that the level of suppression of MMC sensitivity of

FA core complex and FANCD2-deficient cells conferred by

blocking NHEJ differs in that we consistently observed partial

and substantial levels of suppression for FA core complex and

FANCD2 deficiencies, respectively. This difference may reflect

Molecular Cell


the fact that the FA core complex may perform additional roles in

ICL repair independent of its role in regulating FANCD2/FANCI,

such as ubiquitylaton of other factors important for ICL repair

(Matsushita et al., 2005). It is notable that suppression of ICL

repair defects by inhibition of NHEJ is not applicable to all FA

genes as the ICL sensitivity of dog-1 mutants (C. elegans FANCJ)

(Youds et al., 2008) was neither rescued nor enhanced by the lig-

4 mutation (Figure S3C). This may reflect the fact that DOG-1 has

functions independent of the FA pathway during replication fork

repair (Youds et al., 2008). Finally, we demonstrate that DNA-

PKcs is promiscuously activated and inappropriately recruited

to sites of replication stress in the absence of FANCD2 in human

cells. Collectively, our results lead us to propose that DNA repair

defects in FA cells arise as a consequence of the misuse of NHEJ

during lesion repair in S phase and in meiosis. This hypothesis

provides a plausible explanation for many of the observed FA

phenotypes (DNA-repair defects and meiotic problems) and is

consistent with the suppression of these phenotypes upon elim-

ination of the NHEJ pathway. Such a scenario is also supported

by cytological analysis of radial chromosomes in FA cells; over

99% of these events arise as a result of translocations between

nonhomologous chromosomes, which is a signature of repair by

NHEJ (Newell et al., 2004).

Our results raise the possibility that FA proteins act to regulate

the choice of DSB repair pathway, and we propose two potential

models that are consistent with published observations and our

current data: (1) FA proteins directly antagonize NHEJ by block-

ing or displacing NHEJ factors from DSB ends, and/or (2) FA

proteins coordinate the processing of meiotic DSBs and ICL

lesions to favor HR and disfavor NHEJ. Studies of the ICL repair

process have established that DSB intermediates are formed at

the replication fork following ICL incision. DSB ends are also

believed to arise as a result of replication fork regression, which

could be promoted by the helicase/translocase activities of

FANCM and/or BLM (Gari et al., 2008). Given that NHEJ factors

are present constitutively throughout the cell cycle and possess

a high affinity for DSB ends, one possible explanation for our

data is that FA proteins bind to and protect DSB intermediates

from engagement and repair by NHEJ. Supporting this idea,

in vitro DNA-binding assays indicate that FANCD2 is able to

interact with a number of DNA substrates including forked and

linear molecules containing double-stranded DNA ends (Park

et al., 2005; Roques et al., 2009). The hypothesis that FANCD2

physically occludes access of NHEJ factors to repair intermedi-

ates is consistent with our findings that eliminating NHEJ is suffi-

cient to substantially rescue FANCD2 deficiency in most, if not

all, assays and that DNA-PKcs engages inappropriately with

sites of replication stress in the absence of FANCD2. Engage-

ment of NHEJ factors with DSB intermediates during S phase

may inhibit the processing of DSBs required for loading of HR

factors while also promoting illegitimate repair (Figure 7D).

Our results are also consistent with an alternate model in which

the FA proteins function to coordinate processing of ICL lesions/

DSB intermediates in a way that would favor HR and disfavor

NHEJ. Indeed, it is known that the FA core complex associates

with the Blooms and MRN complexes (Meetei et al., 2003;

Nakanishi et al., 2002; Pichierri et al., 2002; Roques et al.,

2009), which are both implicated in DSB processing (Gravel

et al., 2008; Mimitou and Symington, 2008; Zhu et al., 2008).

Therefore, it is possible that the FA pathway may coordinate

BLM- and/or MRN-dependent end processing of DSB intermedi-

ates arising at damaged replication forks and meiotic DSBs. It is

also possible that the FA pathway may recruit other DSB end-

processing activities to promote HR. Ultimately, processing of

the DSB to produce a 30 single-stranded overhang would provide

an optimal substrate for binding by HR factors, while at the same

time reducing the affinity of NHEJ factors for these DNA ends. A

role for FCD-2/FANCD2 in coordinating ICL processing is also

supported by recent observations that FANCD2/FANCI-depleted

frog egg extracts are deficient for the initial incision of an ICL

lesion present within a synthetic DNA substrate, and this can

be rescued by addition of ubiquitylated FANCD2/FANCI

(Knipscheer et al., 2009). An essential role for FANCD2 in ICL inci-

sion is potentially difficult to reconcile with our findings that the

phenotypes resulting from FCD-2/FANCD2 deficiency can be

rescued by eliminating NHEJ. However, it is entirely possible

that in vivo, alternative nucleases process the ICL lesion inappro-

priately in the absence of FANCD2/FANCI and this produces

a substrate for NHEJ. This scenario is consistent with previous

observations and our findings (Figures 4 and 6) that DSBs are

formed at wild-type levels at ICL lesions in vivo in the absence

of FANCD2 (Rothfuss and Grompe, 2004). Moreover, structure-

specific nucleases such as MUS81, XPF, and SNM1A (PSO2)

have all been implicated in ICL processing in vivo and could

process the ICL lesion inappropriately in the absence of FANCD2

(Ciccia et al., 2008; Li et al., 2005). Additionally, a recent study has

shown that the unhooking of an ICL by XPF/ERCC1 is necessary

for the stable association of FANCD2 with damaged chromatin

in vivo, which questions the requirement for FANCD2 in initial

ICL incision (Bhagwat et al., 2009). Nevertheless, our data indi-

cate that blocking NHEJ is sufficient to allow the lesion to be

repaired effectively in the absence of FANCD2, whether or not

the ICL is inappropriately processed at the incision step.

Finally, it is important to mention that our results showing

that deletion or inhibition of Ku80 or DNA-PKcs is sufficient to

rescue the ICL sensitivity of FANCD2-deficient human cells was

not observed in FANCD2/DNA-PKcs double knockout MEFs

(Houghtaling et al., 2005). The most likely reason for this difference

is that human DNA-PKcs possesses additional cellular functions

that are not shared with its counterparts in other vertebrates, or

alternatively inhibition of DNA-PKcs has a different effect to dele-

tion of this kinase (Ruis et al., 2008). Irrespective of the reason for

thisdifference, our discovery that the defects inFANCD2-deficient

human cells can be substantially suppressed by blocking NHEJ

raises the possibility that DNA-PKcs inhibitors could be used in

a subset of FA patients to delay the onset of solid tumors, which

remains the most significant health issue in adult FA patients.

EXPERIMENTAL PROCEDURES

Strains

Strains were cultured according to Wood (1988), experiments were performed

at 20�C. The following C. elegans strains were provided by the Caenorhabditis

Genetics Center: wild-type strain Bristol N2. AV106 spo-11(ok79)IV/ nT1[unc-

?(n754)let-?](IV;V). AV276 syp-2(ok307)V/nT1[unc-?(n754)let-? qls50](IV;V).

AV308 him-14(it21)/mnCI, DR102 dpy-5(e61) unc-29(e403)I. BC3217

unc-60(m35) dpy-11(e224)V; sDp30 (V;X). RB873 lig-4(ok716) III. VC13 dog-1


Molecular Cell


(gk10) I. FX1298 fcd-2(tm1298) IV mutant (Collis et al., 2006) was generated

and provided by Shoehi Mitani of the National Bioresource Project for the

nematode, Japan.

Immunolocalization of RAD-51

The anti-RAD-51 antibody was described in Colaiacovo et al. (2003). DAPI

staining, immunostaining, and analysis of meiotic nuclei were performed using

whole-mount preparations of dissected gonads as described in Colaiacovo

et al. (2003). Quantitative analysis of RAD-51 foci was performed on Z series

of images acquired using a Leica DM6000 fluorescence microscope, Leica

DC 350 FX camera under control of Leica LAS AF 6000 software. Optical

sections were collected at 0.25 mm increments and a minimum of 100 nuclei

per zone of the germline was scored for each genotype. Quantitative analyses

of germline RAD-51 foci (Table S2) were analyzed with student’s t test for inde-

pendent samples. Distributions of numbers of DAPI-stained bodies in the

different genotypes were analyzed with the c2 test (Table S3).

C. elegans Treatment with ICL-Inducing Agents

Young adult worms were picked and individually cloned on freshly seeded

plates containing 180 mM cis-diamminedichloroplatinum(II) (CDDP, Sigma).

Each worm was transferred onto a fresh plate every 24 hr, and the eggs laid

after 48 and 72 hr were scored. The progeny were scored for embryonic

lethality (unhatched eggs) and for organisms with evident postembryonic

developmental defects under a dissecting microscope. Untreated controls

were singled out and screened at the same time points. Conditions for HN2

and TMP-UVA treatments were performed as previously described (Collis

et al., 2006; Youds et al., 2008).

Cell Culture

All cell culture, RNAi conditions, whole-cell extracts, and immunoblotting were

carried out as previously described (Collis et al., 2007). For survival assays,

cells were treated with 0–80ng/ml mitomycin C (MMC; Sigma). Three to four

days after damage, survival was assessed using MTT (Sigma). Cells were

treated with 0.5 mM DNA-PK inhibitor (NU7026; Tocris BioScience) or vehicle

control for 1–2 hr before damage induction. FancA�/� and FancC�/� MEFs

were derived from crosses with FancA+/� FancC+/� double heterozygous

mice. The control MEFs were derived from a wild-type littermate resulting

from the same cross. Six replicas were used per condition for each experi-

ment; the average of four to six independent experiments is shown.

Immunostaining for Phospho-DNA-PKcs and gH2AX

Cells were seeded in duplicate on coverslips and treated 24 hr later with either

3 mM HU, 80 ng/ml MMC, or 5Gy IR and fixed 2 hr, 16hr, 30 min, and 30 mins

later respectively. Cells were then stained for phospho-DNA-PKcs (Abcam;

ab18192, 1:500) and imaged using a DeltaVision microscope as previously

described (Collis et al., 2007). The presence of phospho-DNA-PKcs foci was

scored in over 200 cells per coverslip and the percentage of positive (foci-con-

taining) cells was calculated for each experiment. gH2AX staining (Upstate;

05-636, 1:1000) was carried out on cells treated identically to those processed

for phosho-DNA-PKcs staining.

SUPPLEMENTAL INFORMATION

Supplemental Information includes six figures, four tables, and Supplemental

Experimental Procedures and can be found with this article online at

doi:10.1016/j.molcel.2010.06.026.

ACKNOWLEDGMENTS

The labs of S.J.B. and A.L.V. contributed equally to this work and the cofirst

and cocorresponding authors are positioned alphabetically. We would like to

thank Jillian Youds for critical reading of the manuscript. The lab of S.J.B. is

funded by Cancer Research UK, A.L.V. is funded by CNR and Regione Cam-

pania legge 5. N.S. was recipient of a predoctoral fellowship from MIUR (PhD

School Molecular Genetics and Medicine) and EMBO ST fellowship. The lab of

EM-P is funded by a BBSRC David Phillips Fellowship and the MRC. S.J.B. is

a Royal Society Wolfson Research Merit Award holder.


Received: February 16, 2010

Revised: May 13, 2010

Accepted: June 9, 2010

Published online: July 1, 2010

REFERENCES

Adamo, A., Montemauri, P., Silva, N., Ward, J.D., Boulton, S.J., and La Volpe,

A. (2008). BRC-1 acts in the inter-sister pathway of meiotic double-strand

break repair. EMBO Rep. 9, 287–292.

Alter, B.P. (1993). Fanconi’s anaemia and its variability. Br. J. Haematol. 85,

9–14.

Anderson, C.W., Dunn, J.J., Freimuth, P.I., Galloway, A.M., and Allalunis-

Turner, M.J. (2001). Frameshift mutation in PRKDC, the gene for DNA-PKcs,

in the DNA repair-defective, human, glioma-derived cell line M059J. Radiat.

Res. 156, 2–9.

Bhagwat, N., Olsen, A.L., Wang, A.T., Hanada, K., Stuckert, P., Kanaar, R.,

D’Andrea, A., Niedernhofer, L.J., and McHugh, P.J. (2009). XPF-ERCC1

participates in the Fanconi anemia pathway of cross-link repair. Mol. Cell.

Biol. 29, 6427–6437.

Boulton, S.J. (2006). Cellular functions of the BRCA tumour-suppressor

proteins. Biochem. Soc. Trans. 34, 633–645.

Chen, M., Tomkins, D.J., Auerbach, W., McKerlie, C., Youssoufian, H., Liu, L.,

Gan, O., Carreau, M., Auerbach, A., Groves, T., et al. (1996). Inactivation of Fac

in mice produces inducible chromosomal instability and reduced fertility remi-

niscent of Fanconi anaemia. Nat. Genet. 12, 448–451.

Chen, B.P., Chan, D.W., Kobayashi, J., Burma, S., Asaithamby, A., Moroto-

mi-Yano, K., Botvinick, E., Qin, J., and Chen, D.J. (2005). Cell cycle depen-

dence of DNA-dependent protein kinase phosphorylation in response to

DNA double strand breaks. J. Biol. Chem. 280, 14709–14715.

Cheng, N.C., van de Vrugt, H.J., van der Valk, M.A., Oostra, A.B., Krimpenfort,

P., de Vries, Y., Joenje, H., Berns, A., and Arwert, F. (2000). Mice with

a targeted disruption of the Fanconi anemia homolog Fanca. Hum. Mol. Genet.

9, 1805–1811.

Ciccia, A., McDonald, N., and West, S.C. (2008). Structural and functional rela-

tionships of the XPF/MUS81 family of proteins. Annu. Rev. Biochem. 77,

259–287.

Clejan, I., Boerckel, J., and Ahmed, S. (2006). Developmental modulation of

nonhomologous end joining in Caenorhabditis elegans. Genetics 173,

1301–1317.

Colaiacovo, M.P., MacQueen, A.J., Martinez-Perez, E., McDonald, K., Adamo,

A., La Volpe, A., and Villeneuve, A.M. (2003). Synaptonemal complex assembly

in C. elegans is dispensable for loading strand-exchange proteins but critical

for proper completion of recombination. Dev. Cell 5, 463–474.

Collis, S.J., Barber, L.J., Ward, J.D., Martin, J.S., and Boulton, S.J. (2006). C.

elegans FANCD2 responds to replication stress and functions in interstrand

cross-link repair. DNA Repair (Amst.) 5, 1398–1406.

Collis, S.J., Barber, L.J., Clark, A.J., Martin, J.S., Ward, J.D., and Boulton, S.J.

(2007). HCLK2 is essential for the mammalian S-phase checkpoint and

impacts on Chk1 stability. Nat. Cell Biol. 9, 391–401.

D’Andrea, A.D., and Grompe, M. (2003). The Fanconi anaemia/BRCA

pathway. Nat. Rev. Cancer 3, 23–34.

Dernburg, A.F., McDonald, K., Moulder, G., Barstead, R., Dresser, M., and

Villeneuve, A.M. (1998). Meiotic recombination in C. elegans initiates by

a conserved mechanism and is dispensable for homologous chromosome

synapsis. Cell 94, 387–398.

Dernburg, A.F., Zalevsky, J., Colaiacovo, M.P., and Villeneuve, A.M. (2000).

Transgene-mediated cosuppression in the C. elegans germ line. Genes Dev.

14, 1578–1583.

Fanconi, G. (1967). Familial constitutional panmyelocytopathy, Fanconi’s

anemia (F.A.). I. Clinical aspects. Semin. Hematol. 4, 233–240.

http://dx.doi.org/doi:10.1016/j.molcel.2010.06.026

Molecular Cell


Galloway, A.M., and Allalunis-Turner, J. (2000). cDNA expression array anal-

ysis of DNA repair genes in human glioma cells that lack or express DNA-

PK. Radiat. Res. 154, 609–615.

Gari, K., Decaillet, C., Delannoy, M., Wu, L., and Constantinou, A. (2008). Re-

modeling of DNA replication structures by the branch point translocase

FANCM. Proc. Natl. Acad. Sci. USA 105, 16107–16112.

Gartner, A., Milstein, S., Ahmed, S., Hodgkin, J., and Hengartner, M.O. (2000).

A conserved checkpoint pathway mediates DNA damage—induced apoptosis

and cell cycle arrest in C. elegans. Mol. Cell 5, 435–443.

Gravel, S., Chapman, J.R., Magill, C., and Jackson, S.P. (2008). DNA helicases

Sgs1 and BLM promote DNA double-strand break resection. Genes Dev. 22,

2767–2772.

Houghtaling, S., Timmers, C., Noll, M., Finegold, M.J., Jones, S.N., Meyn,

M.S., and Grompe, M. (2003). Epithelial cancer in Fanconi anemia comple-

mentation group D2 (Fancd2) knockout mice. Genes Dev. 17, 2021–2035.

Houghtaling, S., Newell, A., Akkari, Y., Taniguchi, T., Olson, S., and Grompe,

M. (2005). Fancd2 functions in a double strand break repair pathway that is

distinct from non-homologous end joining. Hum. Mol. Genet. 14, 3027–3033.

Kelly, K.O., Dernburg, A.F., Stanfield, G.M., and Villeneuve, A.M. (2000). Cae-

norhabditis elegans msh-5 is required for both normal and radiation-induced

meiotic crossing over but not for completion of meiosis. Genetics 156,

617–630.

Kennedy, R.D., and D’Andrea, A.D. (2005). The Fanconi Anemia/BRCA

pathway: new faces in the crowd. Genes Dev. 19, 2925–2940.

Knipscheer, P., Raschle, M., Smogorzewska, A., Enoiu, M., Ho, T.V., Scharer,

O.D., Elledge, S.J., and Walter, J.C. (2009). The Fanconi anemia pathway

promotes replication-dependent DNA interstrand cross-link repair. Science

326, 1698–1701.

Lee, K.Y., Chung, K.Y., and Koo, H.S. (2010). The involvement of FANCM,

FANCI, and checkpoint proteins in the interstrand DNA crosslink repair

pathway is conserved in C. elegans. DNA Repair (Amst.) 9, 374–382.

Lees-Miller, S.P., Godbout, R., Chan, D.W., Weinfeld, M., Day, R.S., 3rd,

Barron, G.M., and Allalunis-Turner, J. (1995). Absence of p350 subunit of

DNA-activated protein kinase from a radiosensitive human cell line. Science

267, 1183–1185.

Li, X., Hejna, J., and Moses, R.E. (2005). The yeast Snm1 protein is a DNA

50-exonuclease. DNA Repair (Amst.) 4, 163–170.

Ma, Y., Pannicke, U., Lu, H., Niewolik, D., Schwarz, K., and Lieber, M.R. (2005).

The DNA-dependent protein kinase catalytic subunit phosphorylation sites in

human Artemis. J. Biol. Chem. 280, 33839–33846.

Martin, J.S., Winkelmann, N., Petalcorin, M.I., McIlwraith, M.J., and Boulton,

S.J. (2005). RAD-51-dependent and -independent roles of a Caenorhabditis

elegans BRCA2-related protein during DNA double-strand break repair.

Mol. Cell. Biol. 25, 3127–3139.

Matsushita, N., Kitao, H., Ishiai, M., Nagashima, N., Hirano, S., Okawa, K.,

Ohta, T., Yu, D.S., McHugh, P.J., Hickson, I.D., et al. (2005). A FancD2-mono-

ubiquitin fusion reveals hidden functions of Fanconi anemia core complex in

DNA repair. Mol. Cell 19, 841–847.

Meetei, A.R., Sechi, S., Wallisch, M., Yang, D., Young, M.K., Joenje, H., Hoat-

lin, M.E., and Wang, W. (2003). A multiprotein nuclear complex connects

Fanconi anemia and Bloom syndrome. Mol. Cell. Biol. 23, 3417–3426.

Mimitou, E.P., and Symington, L.S. (2008). Sae2, Exo1 and Sgs1 collaborate in

DNA double-strand break processing. Nature 455, 770–774.

Moldovan, G.L., and D’Andrea, A.D. (2009). How the fanconi anemia pathway

guards the genome. Annu. Rev. Genet. 43, 223–249.

Muzzini, D.M., Plevani, P., Boulton, S.J., Cassata, G., and Marini, F. (2008).

Caenorhabditis elegans POLQ-1 and HEL-308 function in two distinct DNA

interstrand cross-link repair pathways. DNA Repair (Amst.) 7, 941–950.

Nakanishi, K., Taniguchi, T., Ranganathan, V., New, H.V., Moreau, L.A., Stot-

sky, M., Mathew, C.G., Kastan, M.B., Weaver, D.T., and D’Andrea, A.D. (2002).

Interaction of FANCD2 and NBS1 in the DNA damage response. Nat. Cell Biol.

4, 913–920.

Nakanishi, K., Yang, Y.G., Pierce, A.J., Taniguchi, T., Digweed, M., D’Andrea,

A.D., Wang, Z.Q., and Jasin, M. (2005). Human Fanconi anemia monoubiquiti-

nation pathway promotes homologous DNA repair. Proc. Natl. Acad. Sci. USA

102, 1110–1115.

Newell, A.E., Akkari, Y.M., Torimaru, Y., Rosenthal, A., Reifsteck, C.A., Cox, B.,

Grompe, M., and Olson, S.B. (2004). Interstrand crosslink-induced radials

form between non-homologous chromosomes, but are absent in sex chromo-

somes. DNA Repair (Amst.) 3, 535–542.

Page, S.L., and Hawley, R.S. (2003). Chromosome choreography: the meiotic

ballet. Science 301, 785–789.

Park, W.H., Margossian, S., Horwitz, A.A., Simons, A.M., D’Andrea, A.D., and

Parvin, J.D. (2005). Direct DNA binding activity of the Fanconi anemia D2

protein. J. Biol. Chem. 280, 23593–23598.

Petalcorin, M.I., Sandall, J., Wigley, D.B., and Boulton, S.J. (2006). CeBRC-2

stimulates D-loop formation by RAD-51 and promotes DNA single-strand

annealing. J. Mol. Biol. 361, 231–242.

Pichierri, P., Averbeck, D., and Rosselli, F. (2002). DNA cross-link-dependent

RAD50/MRE11/NBS1 subnuclear assembly requires the Fanconi anemia C

protein. Hum. Mol. Genet. 11, 2531–2546.

Polanowska, J., Martin, J.S., Garcia-Muse, T., Petalcorin, M.I., and Boulton,

S.J. (2006). A conserved pathway to activate BRCA1-dependent ubiquitylation

at DNA damage sites. EMBO J. 25, 2178–2188.

Rinaldo, C., Bazzicalupo, P., Ederle, S., Hilliard, M., and La Volpe, A. (2002).

Roles for Caenorhabditis elegans rad-51 in meiosis and in resistance to

ionizing radiation during development. Genetics 160, 471–479.

Roques, C., Coulombe, Y., Delannoy, M., Vignard, J., Grossi, S., Brodeur, I.,

Rodrigue, A., Gautier, J., Stasiak, A.Z., Stasiak, A., et al. (2009). MRE11-

RAD50-NBS1 is a critical regulator of FANCD2 stability and function during

DNA double-strand break repair. EMBO J. 28, 2400–2413.

Rothfuss, A., and Grompe, M. (2004). Repair kinetics of genomic interstrand

DNA cross-links: evidence for DNA double-strand break-dependent activation

of the Fanconi anemia/BRCA pathway. Mol. Cell. Biol. 24, 123–134.

Ruis, B.L., Fattah, K.R., and Hendrickson, E.A. (2008). The catalytic subunit of

DNA-dependent protein kinase regulates proliferation, telomere length, and

genomic stability in human somatic cells. Mol. Cell. Biol. 28, 6182–6195.

Timmers, C., Taniguchi, T., Hejna, J., Reifsteck, C., Lucas, L., Bruun, D.,

Thayer, M., Cox, B., Olson, S., D’Andrea, A.D., et al. (2001). Positional cloning

of a novel Fanconi anemia gene, FANCD2. Mol. Cell 7, 241–248.

Veuger, S.J., Curtin, N.J., Richardson, C.J., Smith, G.C., and Durkacz, B.W.

(2003). Radiosensitization and DNA repair inhibition by the combined use of

novel inhibitors of DNA-dependent protein kinase and poly(ADP-ribose) poly-

merase-1. Cancer Res. 63, 6008–6015.

Wang, W. (2007). Emergence of a DNA-damage response network consisting

of Fanconi anaemia and BRCA proteins. Nat. Rev. Genet. 8, 735–748.

Wood, W. (1988). The nematode Caenorhabditis elegans (Cold Spring Harbor,

NY: Cold Spring Harbor Laboratory Press).

Yajima, H., Lee, K.J., and Chen, B.P. (2006). ATR-dependent phosphorylation

of DNA-dependent protein kinase catalytic subunit in response to UV-induced

replication stress. Mol. Cell. Biol. 26, 7520–7528.

Youds, J.L., Barber, L.J., Ward, J.D., Collis, S.J., O’Neil, N.J., Boulton, S.J.,

and Rose, A.M. (2008). DOG-1 is the Caenorhabditis elegans BRIP1/FANCJ

homologue and functions in interstrand cross-link repair. Mol. Cell. Biol. 28,

1470–1479.

Youds, J.L., Barber, L.J., and Boulton, S.J. (2009). C. elegans: a model of Fan-

coni anemia and ICL repair. Mutat. Res. 668, 103–116.

Zalevsky, J., MacQueen, A.J., Duffy, J.B., Kemphues, K.J., and Villeneuve,

A.M. (1999). Crossing over during Caenorhabditis elegans meiosis requires

a conserved MutS-based pathway that is partially dispensable in budding

yeast. Genetics 153, 1271–1283.

Zhu, Z., Chung, W.H., Shim, E.Y., Lee, S.E., and Ira, G. (2008). Sgs1 helicase

and two nucleases Dna2 and Exo1 resect DNA double-strand break ends. Cell

134, 981–994.


Preventing Nonhomologous End Joining Suppresses DNA Repair Defects of Fanconi Anemia

Documents