R E S EARCH ART I C L E

CANCER

Dow

nloaded fro

ATRX loss promotes tumor growth and impairsnonhomologous end joining DNA repair in gliomaCarl Koschmann,1,2 Anda-Alexandra Calinescu,2 Felipe J. Nunez,2 Alan Mackay,3

Janet Fazal-Salom,3 Daniel Thomas,2 Flor Mendez,2 Neha Kamran,2 Marta Dzaman,2

Lakshman Mulpuri,2 Johnathon Krasinkiewicz,2 Robert Doherty,2 Rosemary Lemons,2

Jaqueline A. Brosnan-Cashman,4 Youping Li,2 Soyeon Roh,2 Lili Zhao,5 Henry Appelman,6

David Ferguson,6 Vera Gorbunova,7 Alan Meeker,8 Chris Jones,3

Pedro R. Lowenstein,2 Maria G. Castro2*

Recent work in human glioblastoma (GBM) has documented recurrent mutations in the histone chaperone proteinATRX. We developed an animal model of ATRX-deficient GBM and showed that loss of ATRX reduces median sur-vival and increases genetic instability. Further, analysis of genome-wide data for human gliomas showed that ATRXmutation is associated with increased mutation rate at the single-nucleotide variant (SNV) level. In mouse tumors,ATRX deficiency impairs nonhomologous end joining and increases sensitivity to DNA-damaging agents that inducedouble-stranded DNA breaks. We propose that ATRX loss results in a genetically unstable tumor, which is moreaggressive when left untreated but is more responsive to double-stranded DNA-damaging agents, resulting in im-proved overall survival.

m

by guest on O

ctober 14, 2020http://stm

.sciencemag.org/

INTRODUCTION

Glioblastoma (GBM) is a lethal primary brain tumorwith amedian sur-vival of less than 2 years. Recent work in human gliomas has documen-ted recurrentmutations in the histone chaperone protein ATRX.ATRXmutation in glioma is primarily seen in adolescents and young adults(ages 10 to 30 years) (1). In pediatric patients,ATRXwas reported to bemutated in 31%of patients with primaryGBM [WorldHealthOrganiza-tion (WHO) grade IV glioma], often with concurrent mutation of TP53andpointmutation of the gene encoding the histoneH3.3 variantH3F3A(1, 2). In adults (age >30 years),ATRX ismutated less frequently in primaryGBM but is frequently found in lower-grade (WHO grade II/III) andsecondaryGBM (2–4). Recent profiling of adult grade II and III gliomasrevealed that amajority (~75%)of the subtype of low-grade gliomas thatcarry TP53 and IDH1 mutations also harbor ATRX mutations, thusunderscoring their fundamental role in gliomagenesis (4).

ATRX mutation is seen in at least 15 types of human cancers, in-cluding neuroblastoma, osteosarcoma, and pancreatic neuroendocrinetumors (PanNETs) (5, 6). However, the role of ATRX in tumori-genesis remains largely unknown. The ATRX protein likely plays animportant epigenetic role, depositing histones at heterochromatin andtelomeric DNA (7, 8). Previous characterization of human GBM andPanNETs has shown an association between ATRX loss and the mainte-nance of telomere length by alternative lengthening of telomere (ALT),or nontelomerase-based, pathways (1, 8). Transgenic loss of ATRX ina mouse model is embryonic-lethal, and postnatal conditional loss ofATRX alone impairs cortex development without causing tumor for-

1Division of Pediatric Hematology-Oncology, Department of Pediatrics, University ofMichigan School of Medicine, Ann Arbor, MI 48109, USA. 2Departments of Neurosurgeryand Cell and Developmental Biology, University of Michigan School of Medicine, AnnArbor, MI 48109, USA. 3Divisions of Molecular Pathology and Cancer Therapeutics, Insti-tute of Cancer Research, London SM2 5NG, UK. 4Department of Pathology, Johns HopkinsUniversity, Baltimore, MD 21287, USA. 5Department of Biostatistics, University of MichiganSchool of Medicine, Ann Arbor, MI 48109, USA. 6Department of Pathology, University ofMichigan School of Medicine, Ann Arbor, MI 48109, USA. 7Department of Biology, Uni-versity of Rochester, Rochester, NY 14627, USA. 8Departments of Pathology and Urology,Johns Hopkins University, Baltimore, MD 21287, USA.*Corresponding author. E-mail: mariacas@med.umich.edu

www.Scie

mation (7). Mutations in ATRX result in loss of ATRX protein by im-munostaining and are thought to mediate loss of function (1).

The development of genetically engineered mouse models allowsfor the systematic evaluation of the contribution of specific genetic lesionsto glial tumor development. The Sleeping Beauty (SB) transposase systemis particularly well-suited to providing a platform for the rapid validationof proposed driver mutations (9). The SB system uses a synthetic plas-mid DNA encoding a transposase gene that inserts a desired transposonDNA element stably into genomic DNA. With various combinations ofhuman oncogenes and inhibitors of tumor suppressor genes, tumorsresembling human GBM can be reliably generated (10, 11).

We used the SB transposase system to create an animal model ofATRX-deficient GBM. Loss of ATRX accelerated tumor growth andreduced median survival, uncovering the impact of ATRX loss on gliomatumor proliferation. We show that ATRX loss causes genetic instabilityin mouse GBM, including both microsatellite instability (MSI) and im-paired telomere maintenance. In accordance with this, analysis of pub-licly available human glioma genome-wide data integrated frommultiple sequencing platforms showed that ATRXmutations are asso-ciated with increased mutation rate at the single-nucleotide variant(SNV) level, but not at the chromosomal/copy number level. We alsoshow that loss of ATRX results in impairment of nonhomologous endjoining (NHEJ) activity and is strongly correlated with loss of activated(phospho-) DNA-dependent protein kinase catalytic subunit (pDNA-PKcs) staining. By uncovering the connection of ATRX mutation andimpaired NHEJ, we provide a mechanism for genetic instability and anactionable therapeutic target for ATRX-deficient GBM. Together, these re-sults provide insights into the role of ATRXmutations in human glioma.

RESULTS

Generation of ATRX-deficient mouse GBM using SB modelTo assess the impact of ATRX loss on GBM, we developed an en-dogenous mouse model using the SB transposase system (10). We clonedan ATRX knockdown sequence (shATRX)] into a plasmid with flanking

nceTranslationalMedicine.org 2 March 2016 Vol 8 Issue 328 328ra28 1

R E S EARCH ART I C L E

sequences recognized by the SB transposase for insertion into thehost genomic DNA (Fig. 1A and fig. S1). We induced GBM in miceby injecting plasmids encoding (i) SB transposase/firefly luciferase,

www.Scie

(ii) shp53, and (iii) NRAS, with or without (iv) shATRX, into the lateralventricle of neonatal mice. Transfection efficiency and tumor growthwere monitored by in vivo imaging of luminescence (Fig. 1B). ATRX

by guest on October 14, 2020

http://stm.sciencem

ag.org/D

ownloaded from

ROI 1=2.262e+07

1 day after injection

C

35 days after injection

7 days after injection

*

GFP ATRX DAPI 60

ATRX

GFP

Nontransfected

Transfected

*

*

APGK

Sleeping Beauty transposase

pCAGGs Luciferase

NRAS pCAGGs

shp53 CMV GFP SV40

shATRX CMV GFP SV40

IR/DR sequence miR-30 sequence

Transfected

ATRX

GFP

B

E

7 days after injection

5 µm

GFP (tumor) Nestin GFP (tumor) OLIG2 GFP GFP (tumor) GFAP

GFP GFAP DAPI GFP Nestin DAPI GFP OLIG2 DAPI

GFP GFAP DAPI GFP Nestin DAPI GFP OLIG2 DAPI

7 days after injection

21 days after injection

Moribund mouse

GFAP Nestin OLIG2

20 µm

75 µm

50 µm

20 µm

50 µm

50 µm

20 µm

50 µm

50 µm

F

75 µm

D

40 µm

75 µm

40 µm

Fig. 1. SBmousemod-el represents ATRX-

deficient GBM. (A) Con-structs of SB plasmids,including the plasmidwith short hairpin againstATRX. Purple bracketsrepresent IR/DR (invertedrepeat/direct repeat)sequences; the area be-tween these IR/DR se-quences is recognizedby the SB transposasefor insertion into the hostgenomic DNA. miR-30sequences representthe 5′ and 3′ flankingsequences of the 300-nucleotide primary mi-croRNA molecule. PGK,phosphoglycerate kinase;CMV, cytomegalovirus.(B) Plasmid insertionand tumor growth aremonitored by in vivoluminescence. ROI, re-gion of interest. (C andD) Hematoxylin and eo-sin staining of an 8-day-old mouse brain (7 daysafter injection) (C), in-cluding inset of sub-ventricular zone liningthe lateral ventricle (D).(E) Immunofluorescenceof the region depictedin (D), with transfectedcells (GFP-positive, to theright of the dotted line)showing ATRX reduc-tion at 7 days after injec-tion. Asterisks indicate atumor cell that is positivefor expression of GFP andnegative for ATRX. DAPI,4′,6-diamidino-2-phenyl-indole. (F) IHC staining oftransfected cells/tumorsfor markers associatedwith neural stemcells, in-cludingGFAP,OLIG2, andNestin in mice injectedwithshp53/NRAS/shATRX.

nceTranslationalMedicine.org 2 March 2016 Vol 8 Issue 328 328ra28 2

R E S EARCH ART I C L E

http://stm.

Dow

nloaded from

loss was tested in the context of overexpression of the oncogene NRASand a short hairpin against p53, because ATRX mutation by itself isnot associated with cancer development in humans (12), and glial tu-mors with ATRX mutations almost always include TP53 mutations(1). The receptor tyrosine kinase (RTK)–RAS–phosphatidylinositol3-kinase (PI3K) pathway is mutated in a large percentage of adultand pediatric high-grade gliomas (1, 13, 14). Thus, many geneticallyengineered animal models of GBM have taken advantage of theglioma-promoting effects of NRAS, either by activating mutations orthrough loss of NF1 expression, which results in NRAS up-regulation(table S1).

After injection of plasmids (shp53, NRAS, and shATRX), the micewere euthanized at early time points to characterize the model. At 7 daysafter injection, transfected cells [green fluorescent protein (GFP)–positive] were found within 50 to 100 mm of the lateral ventricles andshowed early loss of ATRX expression by immunohistochemistry(IHC) (Fig. 1, C to E). To determine whether all three plasmids (shp53,NRAS, and shATRX) were being expressed in the same cells, we injectedmice with plasmids expressing single fluorescent markers: (i) shp53-GFP(green), (ii) NRAS-Katushka (red), and (iii) shATRX-noGFP [detectedby IHC with immunofluorescent blue secondary antibody (405 nm)].At 15 days after injection, cells showed evidence of cotransfection withall plasmids (fig. S2).

Under both experimental conditions (shp53/NRAS with or with-out shATRX), GFP-positive transfected cells coexpressed glial fibrillary

by guest on October 14, 2020

sciencemag.org/

acidic protein (GFAP) and Nestin, mar-kers of neural stem cells (Fig. 1F) (15),and not myosin VIIa, a marker of epen-dymal cells (fig. S3) (16). By day 15, wefound multiple clusters of proliferativecells, which were GFAP-negative by thistime point but remained OLIG2- andNestin-positive. This expression patternpersisted through late-stage tumors frommoribund animals (Fig. 1F and figs. S4to S7). Tumors in moribund animalsshowed strong phospho–extracellularsignal–regulated kinase (pERK) staining(fig. S6), which is downstream of RASand expressed in human GBM (17).

In the process of optimizing and char-acterizing this model, we established alarge cohort of mice injected with shp53/NRAS plasmids (n = 47). We then com-pared their survival with our experimentalgroup of mice injected with shp53/NRAS/shATRX plasmids (n = 19). The mediansurvival of mice injected with shp53/NRAS/shATRX was significantly decreased(69 days) compared to that of mice injectedwith shp53/NRAS (84 days, P = 0.0032)(Fig. 2A). All tumors (with or withoutshATRX) showed histological hallmarks ofhuman GBM, including pseudopalisadingnecrosis (Fig. 2B, black arrows). Loss ofATRX was localized only within tumorsgenerated by injection of the shATRX-expressing plasmid and not in the adjacent

www.Scie

normal cortex or choroid plexus (Fig. 2C). ATRX loss was evidentthroughout the entire tumor, and mice injected with shp53/NRAS/shATRX had larger tumors at earlier time points (Fig. 2C and tableS2). These data show that ATRX loss accelerates GBM tumor growthand reduces survival in mice bearing GBM.

MSI in ATRX-deficient GBMATRX encodes a protein with a DNA binding finger and a SWI2/SNF2-like adenosine triphosphatase (ATPase) motif, making it a mem-ber of a family of adenosine triphosphate (ATP)–dependent chromatin-associated proteins (18). Other members of this class participate inDNA damage repair (DDR) pathways, in particular nucleotide exci-sion repair and double-stranded break repair (19). Previous work hasshown that ATRX is recruited to sites of DNA damage (20). Thus, wehypothesized that ATRX may play a role in maintaining genetic sta-bility in GBM.

One type of genetic instability that is seen in human GBM is MSI.Microsatellites are repeated mononucleotide and dinucleotide sequencesthat are prone to error during DNA replication. Loss of DNAmismatchrepair results in variation of the length of microsatellite sequences insome human tumors, including GBM (21, 22). Both ATRX mutation(1) and MSI (21) are seen more frequently in younger patients withGBM. We used established primer sets for mouse microsatellite se-quences (23, 24) to assess our mouse GBMs and determine whetherATRX loss resulted in the presence of MSI.

D a y s a f t e r in je c t io n

Pe

rce

nt

su

rviv

al

0 5 0 1 0 0 1 5 0 2 0 0

0

5 0

1 0 0A B

p53/NRAS

Tumor

DA

PI A

TR

X 4

0

Tumor *

Choroid plexus

Choroid plexus

Sleeping Beauty transposase

DA

PI A

TR

X 4

0

E Sleeping Beauty

transposase

47 days after injection

NRAS shp53 NRAS shp53 shATRX

p53/NRAS p53/NRAS/shATRX C

Median survival time : p53/NRAS 84 days p53/NRAS/shATRX 69 days

P = 0.0032

129 days after injection

* 75 µm 75 µm

p53/NRAS (n = 47) p53/NRAS/shATRX (n = 19)

100 µm 25 µm

Fig. 2. ATRX loss decreases median survival in mice bearing SB-generated GBMs. (A) Kaplan-Meiersurvival curves of C57BL/6 mice bearing SB-induced tumors (P = 0.0032, Mantel log-rank test). (B) Repre-

sentative brain tumors with histologic hallmarks of GBM: pseudopalisading necrosis and nuclear atypia(black arrows and blue arrow, respectively). (C) Tumors with addition of shATRX plasmid are larger atearlier time points than tumors with shp53/NRAS alone and show ATRX loss throughout the tumor. Asteriskin the left panel indicates a region in a p53/NRAS tumor with positive expression of ATRX. Asterisk in theright panel shows a region in a p53/NRAS/shATRX tumor which is negative for ATRX.

nceTranslationalMedicine.org 2 March 2016 Vol 8 Issue 328 328ra28 3

R E S EARCH ART I C L E

by guest on October 14, 2020

http://stm.sciencem

ag.org/D

ownloaded from

With established MSI software para-meters, tumors were scored for differ-ences in predominant microsatellite[polymerase chain reaction (PCR) frag-ment] length between tumor and controltail DNA. Tumors with ATRX lossshowed greater rates of MSI using fourdistinct microsatellite markers (Fig. 3, Aand B, and table S3). Overall, ATRX lossincreased the rate of instability fivefold(P = 0.014; n = 48 total tumor versus con-trol DNA comparisons). The rate of MSIin ATRX-deficient mouse tumors (21%)was comparable to the rate of MSI inhuman pediatric GBM (19%) (21).

Increased single variant ratein human glial tumors withATRX mutationsThe finding of MSI in ATRX-deficientmouse GBM pointed to ATRX playing arole in maintaining genetic stability atthe sequence level. To validate this find-ing in human glioma, we evaluated wheth-er ATRX mutation was associated withincreased SNV rate in genome-wide datafrom multiple data sets. We retrievedmultiple publicly available genome-widedata sets available in the European Ge-nome Archive (EGA). We then integratedthese multiple sequencing platforms,along with additional pediatric high-gradeglioma samples (deposited at EGA acces-sion no. EGAS00001001436), to producefull somatic sequence and copy numberinformation on 293 pediatric high-gradeglioma samples (up to age 30 years), ofwhich 38 of 293 (13%) samples containedATRX mutations. We also retrieved andanalyzed 290GBMsamples (age>30 years)available from The Cancer Genome Atlas(TCGA) (10, 25). After analyzing our in-tegrated pediatric and adult data sets, wedid not observe any clustering of muta-tions in the SNF2-DNA binding region ofthe gene (fig. S8), as has been described byothers (1). A higher average SNV rate wasseen in ATRX-mutated pediatric high-grade glioma in all anatomical locations(P = 0.0038) and particularly in pediatricGBM (P = 0.0005), but not in adult GBM(P =0.71) (Fig. 3C).

ATRX mutation is frequently foundwith concurrent TP53 mutation in hu-man glioma. Because loss of p53 functionmay be permissive of genetic instabilityin tumor cells (26), we evaluated whetherour finding of increased somatic variant

**

**

% T

um

ors

wit

h M

SI

mB

at2

6

D7m

it91

D1m

it62

D6m

it59

To

tal

0

1 0

2 0

3 0

4 0

5 0

*

B

Microsatellite instability (mouse GBM)

A

Single-nucleotide variant (SNV) frequency (human glioma)

SNVs by ATRX statusC D

+2

p53/NRAS/shATRX

tumor DNA

Mouse tail DNA

D1Mit62

130128 129127126125

Adult

GBM

(n = 290)

Pediatric

GBM

(n = 128)

SN

Vs p

er

tum

or

Pediatric

HGG

(n = 293)

NS

p53/NRAS Tumor

p53/NRAS/shATRX Tumor

SNVs by ATRX and TP53 status

p53 wild type/ ATRX wild type

p53 mutated/ ATRX wild type ATRX wild type

ATRX mutated

n = 208 66 11 5 78 31 22

Pediatric GBM (n = 128)Adult GBM (n = 290)

SN

Vs p

er

tum

or

ATRX P = 0.074

p53 P = 0.10

ATRX P = 0.027

p53 P = 0.42

p53 mutated/ ATRX mutated

p53 wild type/ ATRX mutated

0

50

100

150

200

0

50

100

200

300

400

Fig. 3. ATRX loss increases MSI in mouse GBM and SNV frequency in human glioma. (A) Represent-ative microsatellite lengths from a tumor with MSI-high positivity (log2 ratio of >4 for predominant tumorsample allele versus control mouse tail DNA) using marker D1Mit62. Plot shows a shift in populations ofallele lengths in p53/NRAS/shATRX tumor (orange) compared to control mouse tail DNA (gray). Pre-dominant allele size in this example is two nucleotides longer (orange arrow) than control mouse tailDNA allele (gray arrow). (B) Comparison of the percentage of tumors with MSI positivity using four inde-pendent MSI primer sets (mBat26, D7mit91, D1mit62, and D6mit59) (n = 48 tumor versus control DNAcomparisons). Data sremeans ± SEM; *P< 0.05 using two-sided c2 analysis. (C) Analysis ofmatched humantumor/germline integrated sequencing data sets showing SNV frequency in tumors by ATRX mutationalstatus (GBM, WHO grade IV; pediatric GBM excludes diffuse intrinsic pontine glioma; HGG, high-gradeglioma, WHO grades III and IV). **P < 0.005 using unpairedMann-Whitney test. (D) Analysis of significanceof contribution to SNV rate by ATRX and TP53 mutational status, using a two-way analysis of variance(ANOVA) model in adult GBM (n = 290) and pediatric GBM (n = 128). Each data point represents an indi-vidual human tumor; line represents mean ± SEM.

www.ScienceTranslationalMedicine.org 2 March 2016 Vol 8 Issue 328 328ra28 4

R E S EARCH ART I C L E

Dow

nload

rate in ATRX-mutated tumors was confounded by TP53 mutation inhuman GBM. TP53 mutation was not a predictor of variant rate ineither data set, whereas ATRX was a predictor in pediatric nonbrain-stem GBM (P = 0.027 ANOVA), but not in adult GBM (Fig. 3Dand table S4). IDH1 mutation was also not a predictor of variant ratein pediatric or adult data sets, with ANOVA P values of 0.14 (adultGBM) and 0.19 (pediatric GBM) (table S4).

Association between ATRX loss and copy number alterationsPrevious analysis of pediatric GBM has shown an association betweentumors with concurrent mutations in H3F3A, TP53, and ATRX andcopy number alterations (1). ATRXmutations were not associated withtotal copy number alterations, nor gains or losses considered separately,in the pediatric data set, or with percentage genome fraction altered inthe adult glioma data set (Fig. 4A).

To further examine the impact of ATRX on stability at the chromo-somal level, we investigated whether ATRX loss was associated withkaryotypic changes in our mouse GBMs. We generated primary cellcultures from SB-generated mouse GBMs and demonstrated that tumor

www.Scie

by guest on October 14, 2020

http://stm.sciencem

ag.org/ed from

neurosphere cultures generated with p53/NRAS/shATRX tumorsshowed stable reduction of ATRX expression compared to p53/NRAScells (fig. S9). With a previously established method (27), karyotypesof multiple independent GBM cell cultures showed similar mean chro-mosomal counts and coefficients of variation between neurosphereswith and without shATRX (Fig. 4B).

Assessment for ALT in mouse GBMPrevious characterization of human GBM has shown an associationbetween mutations of ATRX and the maintenance of telomere lengthby ALT, or nontelomerase-based, pathways (1, 8). To establish a causallink, we assessed our mouse GBMs for evidence of impact on telomerelengthening. We isolated DNA from GFP-positive mouse GBM tumortissue (with or without shATRX) and noted no difference in telomerelengths in tumors with ATRX loss by established telomere quantitativePCR (qPCR) primers (fig. S10 and table S5) (28). We then surveyed DNAextracted from tumors and neurospheres for c-circle amplification, whichhas been shown to be a specific assay for ALT (29). C-circles were foundin a subset of tumor and neurosphere samples with ATRX loss, butnot in DNA extracted from normal mouse brains (fig. S11).

Recently, ultrabright spots seen on human tumor tissue by telo-meric fluorescence in situ hybridization (FISH) have been demon-strated as a sensitive indicator for the presence of ALT (8). Mosthuman PanNETs have been found to harbor both ATRX mutationsand ALT assessed by telomeric FISH (8). We therefore hybridizedFISH probes against human PanNETs to be used as positive controlsand noted distinct bright spots (Fig. 5A, white arrows) in a subset ofcells in PanNETs, which are consistent with tumor ALT positivity.

In our mouse tumors, we noted a population of tumor cells withhigher fluorescence signal in ATRX-deficient mouse GBM (Fig. 5A,white dotted circle) that was not seen in control tumors or striatalcells. To quantify the difference, we calculated corrected total cellfluorescence (CTCF) in tumor cells randomly chosen from multipletumors under each experimental condition. The mean CTCF fromp53/NRAS tumors was similar to that of mouse striatal cells but wasincreased in p53/NRAS/shATRX tumor cells (Fig. 5B). Again, we sawa distinct set of cells with higher CTCF in both p53/NRAS/shATRXtumors and human PanNETs (Fig. 5B, black dotted circles), consistentwith ALT physiology.

ATRX loss and impaired NHEJTo determine the mechanism by which ATRX loss affects genomicand telomeric stability, we assessed its impact on DDR pathways. Weused reporter plasmids previously designed to quantify homologousrecombination (HR) and NHEJ efficacy (30). Hepa 1-6 cells weretransfected with linearized reporters harboring a DNA-damagedGFP that can be restored by NHEJ or HR, depending on the plasmidsystem used. Cotransfection with shATRX reduced NHEJ function by50% (P = 0.0024), but HR activity remained unaffected (Fig. 6A) whenquantified by flow cytometry (percentage of GFP-positive cells aftertransfection with shATRX or shSCRAMBLE normalized to mean con-trol; Fig. 6B and fig. S12).

To confirm that the NHEJ pathway was impaired in vivo in mouseGBM with ATRX loss, we investigated whether key NHEJ pathwayproteins were affected. Previous studies have shown that the proteins,namely, Ku70 and Ku80, first bind to double-stranded DNA breaksand then recruit the DNA-PKcs. DNA-PKcs can then recruit andphosphorylate other NHEJ pathway proteins, as well as phosphorylate

Pediatric HGG (n = 293)

Copy number alteration frequency (human glioma)

p53/NRAS tumor cell culture

p53/NRAS/shATRX tumor cell culture

p53/NRAS tumor cell culture A

p53/NRAS/ shATRX

tumor cell culture B

Total

A

Chromosome count (mouse GBM cell culture) B

A B A B C

NS

Adult glioma (TCGA)

Total CNAs Gains Losses GBM (n = 290)

Gen

om

e fr

acti

on

alt

ered

Low-grade glioma (n = 289)

Co

py

nu

mb

er a

lter

atio

ns

Ch

rom

oso

me

cou

nt

/ cel

l

NS

NS NS

NS

NS

0

2 0

4 0

6 0

8 0

1 0 0

Fig. 4. ATRX loss does not cause structural/chromosomal alterations orchange chromosome count. (A) Analysis of chromosomal alterations, in-

cluding copy number alterations, in the integrated pediatric GBMdata sets,and percentage of genomic alterations in the adult TCGA data set show nodifference byATRXmutational status. Line representsmean± SEM; P≥ 0.05usingunpairedMann-Whitney test; NS, not significant. (B) Chromosomecountby metaphase preparation of independent GBM neurosphere cultures. Linerepresents mean ± SEM; P ≥ 0.05 using unpaired Mann-Whitney test. Bothgroups had similar coefficients of variation [31.4% in p53/NRAS tumor cells(n = 15) and 30.7% in p53/NRAS/shATRX tumor cells (n = 20)].

nceTranslationalMedicine.org 2 March 2016 Vol 8 Issue 328 328ra28 5

R E S EARCH ART I C L E

by guest on October 14, 2020

http://stm.sciencem

ag.org/D

ownloaded from

itself to promote its activity (pDNA-PKcs). These activated pathwayproteins then bridge broken ends to facilitate rejoining (30).

We stained mouse GBMs (with and without shATRX) and foundthat ATRX loss was strongly correlated with loss of pDNA-PKcsby immunofluorescence staining (Fig. 6C and table S6). Tumors withp53/NRAS (n = 3) alone showed robust pDNA-PKcs staining through-out the tumor, but it was almost absent in all tumors with p53/NRAS/shATRX (n=3).We observed staining for pDNA-PKcs in the nontumorstriatum of all animals in both groups (Fig. 6C). IHC for the NHEJDNA repair enzymes Ku70 and XRCC4 revealed no overt changes inexpression (fig. S13). Similar results were obtained for HR repair

www.ScienceTranslationalMedicine.org

pathway enzymes RAD51 and BRCA1(fig. S14) and base excision repair proteinPARP1 [poly(adenosine diphosphate–ribose) polymerase 1] (table S6). On thebasis of our finding of increased MSI inATRX-deficient mouse GBM, we alsosurveyed for mismatch repair proteinsknown to affect MSI (MLH1, MSH6, andPMS2) and found similar expression pat-terns in tumors with and without ATRXloss (fig. S15).

Sensitivity of ATRX-deficient GBMtumor cells to double-strandedDNA-damaging treatmentInhibition of NHEJ in cancer cells inducesan accumulation of double-stranded DNAbreaks, which increases susceptibility toradiation (31). Because our data show thatATRX loss impairs NHEJ, we hypothesizedthat ATRX-deficient GBM cells would haveincreased sensitivity to DNA-damagingagents that primarily induce double-stranded breaks. Indeed, ATRX-deficientGBM cells were more sensitive in vitro totreatment with radiation and previouslypublished doses of doxorubicin, irinotecan(SN-38), and topotecan (Fig. 7A) (32–35). Incontrast, treatment of GBM cells with agentsthat primarily induce single-stranded defects[CCNU (1-(2-chloroethyl)-3-cyclohexyl-1-nitrosourea) and temozolomide (32)] wasnot affected by ATRX status. To confirmour findings in vivo, we assessed biolumi-nescence in mice with GBMs treated withwhole-brain radiation. ATRX-deficient miceshowed reduced growth at days 4 and 10after radiation compared to controls (micewith shp53/NRAS) [Fig. 7B (treated) andfig. S16 (untreated)].

We found that ATRX-deficient tumorsshowed reduction in pDNA-PKcs at mul-tiple time points after radiation (fig. S17),whereas control tumors and nontumor brainshowed both pDNA-PKcs and ATRX ex-pression. In vitro, we found that ATRX-deficient tumor cells showed an increase

in gH2A.X expression, a sensitive marker for double-stranded DNAbreaks, 24 hours after treatment with doxorubicin (35), compared to con-trol tumor cells (fig. S18A). In vivo, we saw increased gH2A.X expres-sion in ATRX-deficient tumors 24 hours after radiation (single dose of6 Gy) (fig. S18B).

ATRX mutation and improved survival in treated human GBMRecent data have shown that adultswith treatedATRX-mutatedGBMhavea survival advantage (3, 4).Weused our integrated humangliomagenome-wide data set to confirm thatATRXmutation provides a survival advan-tage in pediatric high-grade gliomapatients aswell (P=0.0035, Fig. 8A).

0

2 .0 ×1 05

4 .0 ×1 05

6 .0 ×1 05

8 .0 ×1 05

Calculated total cell fluorescence (CTCF)

p53/NRAS/ shATRX

tumor cells

p53/NRAS tumor cells

Striatal cells

Human PanNET tumor cells

(ALT +)

*

NS

B

Human PanNET tumor

Mouse p53/NRAS tumor Mouse p53/NRAS/ shATRX tumor

A

Telomere FISH DAPI Telomere FISH DAPI

5 µm

Telomere FISH DAPI Telomere FISH DAPI

Striatum Tumor Tumor Tumor

Fig. 5. ATRX-deficient mouse GBMs display ALT. (A) Detection of ALT using telomeric FISH assayshowing characteristic ultrabright spots in human PanNETs (white arrows, positive control). A distinct

population of cells with increased telomere signal is seen in ATRX-deficient tumors (white dottedcircle). (B) CTCF in arbitrary units for telomeric FISH signal (data points represent individual cells fromthree tumors under each condition); black dotted circles denote cells qualitatively showing ultrabrightspots, consistent with ALT. Line represents mean ± SD; *P < 0.05 using unpaired Mann-Whitney test.

2 March 2016 Vol 8 Issue 328 328ra28 6

R E S EARCH ART I C L E

by guest on October 14, 2020

http://stm.sciencem

ag.org/D

ownloaded from

In contrast, our data indicate thatATRX loss in untreatedmice confers asurvival disadvantage (Fig. 2). Increased sensitivity of ATRX-deficientGBM to radiation and/or chemotherapy that induces double-strandedbreaks could explain this apparent discrepancy. Thus, we propose thatATRX loss causes impaired NHEJ and genetic instability in glioma,

www.ScienceTranslationalMedicine.org

which is more aggressive when untreatedbut is more responsive to double-strandedDNA-damaging therapy, ultimately result-ing in improved overall survival (Fig. 8B).

DISCUSSION

Recurrent mutations in ATRX point toits critical role in tumor progression in mul-tiple human cancers (5, 6). Despite this,little is known about the molecular mecha-nism by which it affects tumor develop-ment and growth. We used an animalmodel of ATRX-deficient GBM to uncoverthe impact of ATRX loss on glioma tu-mor proliferation and loss of geneticstability. We show that ATRX reductionimpairs NHEJ and pDNA-PKcs recruit-ment, thus providing a mechanism forgenetic instability and a molecular target.These results provide insights into theimpact of ATRXmutations in human gli-oma, and possibly other human tumors.

Our data link ATRX to regulation ofNHEJ. Other chromatin remodelers areknown to play important roles in provid-ing proteins access to damaged DNA andtelomeres (36). We hypothesize that re-duced ATRX causes conformational changesin heterochromatin, restricting the accessof NHEJ proteins such as DNA-PKcs todamaged DNA. Our finding of impairedNHEJ is consistent with previous research,which has demonstrated that ALT is depen-dent on HR of sister telomere chromatids(37). Our results suggest that a relativeincrease in HR in relation to NHEJ inATRX-deficient glioma may be supportiveof ALT (proposed schematic, fig. S19).

Our experimental data reproduceATRX loss resulting in ALT in an animalmodel.Human (non-GBM) immortalizedcell lines with ALT have been associatedwith genetic instability and altered DNAdamage response (38). Our data substan-tiate the connection between ATRX, ALT,and the DNA damage response. Our ani-malmodel showedALT inATRX-deficienttumors by telomere FISH and in a subsetof our ATRX-deficient tumors and tumorneurospheres by c-circle assay but did notdisplay elongated telomeres by telomere

qPCR. Overall, our results suggest that ATRX loss in GBM is capableof producing ALT but that additional factors and/or species differencesmay also play a role. Mouse telomeres are much longer than humantelomeres (28); thus, we hypothesize that this species difference couldinfluence ALT assays and ALT physiology.

NHEJ activity (% GFP-positive)

HR reporter NHEJ reporter

GFP (HR repair) GFP (NHEJ repair)

+ shSCRAMBLE + shSCRAMBLE

+ shATRX + shATRX

Control Control

GFP (HR repair) GFP (NHEJ repair)

GFP (HR repair) GFP (NHEJ repair)

**

pDNA-PKcs staining – NHEJ pathway (mouse GBM in vivo) C

DNA damage repair pathway reporter assay (in vitro)

A

NS

B

p53/NRAS p53/NRAS/shATRX

ATRX

DAPI pDNA-PKcs 60

Tumor Tumor

DAPI ATRX 20

Tumor Tumor

Striatum

Striatum

75 µm

20 µm 10 µm

75 µm

20 µm 10 µm

75 µm

DAPI pDNA-PKcs 60

DAPI pDNA-PKcs 60 DAPI pDNA-PKcs 60

C o n t r o l s h S C R A M B L E s h A T R X0

5 0

1 0 0

HR activity (% GFP-positive)

C o n t r o l s h S C R A M B L E s h A T R X0

5 0

1 0 0

Fig. 6. ATRX loss reduces NHEJ repair. (A) Reporter assay with GFP expression that is restored by NHEJor HR in the appropriate plasmids. Addition of shATRX impairs NHEJ (assay performed in triplicate). (B)

Flow cytometric quantification of HR andNHEJ activity as assessed by the percentage of GFP-positive cellsnormalized to control (differences in HR activity are nonsignificant). Line represents mean ± SD; **P < 0.005using unpaired t test. (C) Loss of ATRX in mouse GBM decreases pDNA-PKcs immunostaining (showingrepresentative results of threemouse tumors per condition). Dotted line represents distinction betweentumor and nontumor brain.

2 March 2016 Vol 8 Issue 328 328ra28 7

R E S EARCH ART I C L E

by guest on October 14, 2020

http://stm.sciencem

ag.org/D

ownloaded from

According to the mutator hypothesis of oncogenesis, early muta-tions in “caretaker genes” can drive further tumor development (39).Our data show that ATRX plays this role in glioma because it isrequired for NHEJ DNA repair. It is possible that the genetic in-stability in ATRX-deficient GBM drives proliferation by affecting cellcycle control or differentiation, as has been shown in other genetically

www.Scie

unstable tumor models (40, 41). Additionally, impaired apoptoticsignaling through defective DNA-PKcs phosphorylation (42) and/orconcurrent TP53 mutations could provide an additional proliferativeadvantage to ATRX-mutated tumors.

We show that ATRX is implicated in maintaining stability at thesequence level in our animal model and in our pediatric human data

[To p o te c a n ] µ M

Re

lati

ve

pro

life

rati

on

0 .0 0 0 1 0 .0 0 1 0 .0 1 0 .1 1 1 0 1 0 00 .0

0 .5

1 .0

[D o x o ru b ic in ] µMR

ela

tiv

e p

roli

fera

tio

n0 .0 0 1 0 .0 1 0 .1 1

0 .0

0 .5

1 .0

[C C N U ] µ M

Re

lati

ve

pro

life

rati

on

0 .1 1 1 0 1 0 00 .0

0 .5

1 .0

[Te m o zo lo m id e ] µ M

Re

lati

ve

pro

life

rati

on

0 .1 1 1 0 1 0 0 1 0 0 0 1 0 , 0 0 00 .0

0 .5

1 .0

Re

lati

ve

pro

life

rati

on

0 3 1 0 3 00 .0

0 .5

1 .0

Ra d ia tio n d o s e [G y ]

Radiation

** *

D a y s a f t e r r a d i a t i o n

Lu

min

es

ce

nc

e

(x1

07 p

ho

ton

s/s

)

B a s e lin e 4 1 0 1 8

0

1

2

3

4

In vitro A

In vivo

* *

Day of life 1 32 33

SB plasmids injected

(± shATRX)

Whole-brain radiation (6 Gy per day 2

doses)

10 days

2.50e6 4.32e6

8.51e6 1.14e7 3.31e6 (photons/s)

2.32e6

4 days pre-XRT

p53 NRAS

shATRX

p53 NRAS

(photons/s)

Double-stranded DNA damage Single-stranded DNA damage

B

Temozolomide

NS

NS

Irinotecan (SN-38)

Topotecan

*** ***

***

p53/NRAS GBM cell culture p53/NRAS/shATRX GBM cell culture

Doxorubicin

p53/NRAS GBM cell culture p53/NRAS/shATRX GBM cell culture

CCNU

[ I r i n o t e c a n ( S N-3 8 ) ] µ M

Re

lati

ve

pro

life

rati

on

0 .0 0 0 1 0 .0 0 1 0 .0 1 0 .1 1 1 0 1 0 00 .0

0 .5

1 .0

1 .5

p53/NRAS (n = 6 mice) p53/NRAS/shATRX (n = 6 mice)

Fig. 7. ATRX-deficient GBM cells are sensitive to double-stranded DNA-damaging treatments. (A) In vitro data showingproliferation ofmouseGBM

inhibitory concentration). (B) Schematic of whole-brain radiation for micewith GBM (with or without shATRX). Tumor growth assessed by in vivo lumi-

cell cultures (with or without shATRX) after exposure to escalating doses ofcytotoxic agents. ATRX-deficient tumor cells have reduced proliferation onlyafter exposure to agents that induce double-stranded breaks. Line representsmean± SEM; *P<0.05 , **P<0.01, ***P<0.001 using F test of logIC50 (median

nescence is reduced in ATRX-deficient tumors (representative mice and theirluminescence values are shown). Plot on right shows average tumor lumines-cence for all mice at early time points after radiation (n = 6 mice in eachgroup). Line represents mean ± SEM; *P < 0.05 using unpaired t test.

nceTranslationalMedicine.org 2 March 2016 Vol 8 Issue 328 328ra28 8

R E S EARCH ART I C L E

by guest on October 14, 2020

http://stm.sciencem

ag.org/D

ownloaded from

sets. Notably, in our adult human data sets, we did not see a differencein SNV rate between ATRX-mutated and nonmutated tumors, possiblymaking this finding most applicable to younger patients. The choice ofDDR pathway influences the quality of the repair and the introductionof new somatic mutations and chromosomal rearrangements (43). Re-cent work has highlighted that NHEJ is actually a combination of tworelated but distinct processes: (i) canonical NHEJ (C-NHEJ), which isthe traditional pathway involving repair of double-stranded breaksusing DNA-PKcs, and (ii) alternative NHEJ (A-NHEJ), which is driv-en by PARP1, uses microhomology-mediated end joining, and is as-sociated with deletions at repair junctions (44). Because our tumors(with or without ATRX loss) showed robust PARP1 staining (tableS6), it is possible that the reduction in C-NHEJ in ATRX-deficientglioma results in mutagenesis from use of the more error-prone A-NHEJ pathway. Increase in A-NHEJ in ATRX-deficient glioma couldpotentially explain the increase in point mutations that result in bothincreased somatic mutation rate and MSI (which is also the accumu-lation of single-stranded mutations), thus explaining the phenotypicfeatures seen in our animal and human data.

An important feature of endogenous animal models of glioma isthe choice of genetic drivers. One possible limitation of our modelis the use of NRAS mutation, which is only rarely mutated in humanGBM. Nevertheless, RAS provides a useful driver of GBM formation,which retains relevant histologic features of the human disease whileactivating the RTK-RAS-PI3K pathway, for which most GBMs con-tain signal alterations (13, 14). Although ATRXmutation co-occurs withIDH-R132H mutation in adult glioma, it almost never does in pediatricGBM and instead frequently co-occurs with histone H3.3 mutations.In both cases, it almost always co-occurs with TP53 mutation (1–4).This overlap highlights the fact that ATRX mutation can affect tumordevelopment with multiple other glioma drivers, as long as TP53signaling is impaired, a feature our model also encapsulates.

Our results have translational relevance in that we modeled ATRXloss in mice to identify its role in furthering GBM progression andtherapy response. We believe that this seemingly paradoxical role willassist in the design of future therapy for ATRX-mutated glioma. For

www.ScienceTranslationalMedicine.org

example, detection of ATRX mutationand/or loss of ATRX by immunostainingmay be evidence of a treatment-responsivesubtype of glioma that would encouragethe use of radiation and/or chemotherapyagents that induce double-strandedbreaks. Our data showing both a reduc-tion in NHEJ activity and an increase inresponse to double-stranded DNA-damaging agents in tumor cells withATRX loss are consistent with previousresearch showing that NHEJ inhibition re-duces double-stranded break repair andincreases sensitivity to radiation (31).Further preclinical studies could highlightregimens that more selectively target thedefects in NHEJ and double-strandedbreak repair in ATRX-deficient glioma.Our data raise the possibility that topo-isomerase inhibitors (topotecan or irino-tecan) might be clinically useful toexploit this defect.

These results lay a foundation for uncovering the molecular impactof ATRX mutations in the pathogenesis and response to therapeuticsin human GBM. Additionally, this mouse model provides a platformfor the development of targeted therapy for GBM patients harboringATRX mutations.

SUPPLEMENTARY MATERIALS

www.sciencetranslationalmedicine.org/cgi/content/full/8/328/328ra28/DC1Materials and MethodsFig. S1. SB-responsive shATRX plasmids are cloned to explore the impact of ATRX on GBMdevelopment.Fig. S2. Cells cotransfected with shp53, shATRX, and NRAS plasmids are characterized at 15days after injection.Fig. S3. SB-mediated transfected cells are distinct from ependymal cells in mice.Fig. S4. Characterization of p53/NRAS/shATRX mice 7 days after injection shows tumor cellsexpressing GFAP and Nestin.Fig. S5. p53/NRAS/shATRX mice at 21 days after injection show tumor cells expressing OLIG2 and Nestin.Fig. S6. Moribund p53/NRAS/shATRX mice show tumor cells expressing OLIG2, Nestin, and pERK.Fig. S7. p53/NRAS/shATRX mice show loss of GFAP expression by 15 days after injection.Fig. S8. ATRX mutations do not cluster in SNF2/helicase domain in human glioma.Fig. S9. Primary GBM cell cultures are generated from SB-induced mouse tumors.Fig. S10. Telomere length measured by qPCR is not different in tumors with or without ATRX.Fig. S11. ALT was assessed by c-circle assay with DNA from mouse tumors and neurospheres.Fig. S12. Cells transfected with shATRX plasmid show reduction in NHEJ by flow cytometry.Fig. S13. NHEJ pathway is characterized by IHC.Fig. S14. HR pathway is characterized by IHC.Fig. S15. Mismatch repair pathway is characterized by IHC.Fig. S16. SB mouse tumor growth (without radiation treatment) is characterized by luminescence.Fig. S17. Tumors with ATRX loss show reduction in pDNA-PKcs before and after radiation treatment.Fig. S18. Mouse tumors with ATRX loss show increased expression of gH2A.X in vitro and in vivo.Fig. S19. Detailed schematic shows proposed impact of ATRX loss on GBM progression.Table S1. Animal models of glioma all have RTK-RAS-PI3K alterations.Table S2. Mice injected with p53/NRAS/shATRX have faster-growing tumors.Table S3. MSI rate is increased in p53/NRAS/shATRX tumors.Table S4. IDH1 and TP53mutations do not alter SNV rate calculated by two-way ANOVA model.Table S5. Telomere qPCR average telomere length ratio worksheet and standard curve showsimilar results in all experimental groups.Table S6. Immunofluorescence analysis shows reduced pDNA-PKcs expression in ATRX-deficient mouseGBM.Table S7. Antibodies used for tissue analysis are summarized in table form.

Proposed mechanism of impact of ATRX loss on GBM progression

Double-stranded break

Impaired NHEJ, increased ALT, and gain of mutations

NO TREATMENT

TREATMENT

Tumor more responsive to double-stranded DNA-damaging

treatment (compared to GBM with intact ATRX)

B

Single-stranded mutations

0 2 0 4 0 6 0 8 0

0

5 0

1 0 0

Su r v iv a l ( m o n t h s )

Pe

rce

nt

su

rviv

al

Survival in pediatric high-grade glioma

(with treatment)

A

P = 0.0035

ATRX wild type (n = 255)ATRX mutated (n = 39)

Untreated tumor more aggressive (compared to GBM with intact ATRX)

KEY

Dead tumor cell

KEY

Viable tumor cell

DNA damage

Fig. 8. Pediatric patients with high-grade glioma and ATRX mutation survive longer. (A) Kaplan-Meier curve based on genome-wide data from multiple pediatric data sets (n = 293) showing the survival

benefit of ATRX mutation in treated patients. (B) Schematic of impact of ATRX loss on GBM.

2 March 2016 Vol 8 Issue 328 328ra28 9

R E S EARCH ART I C L E

by guest on October 14, 2020

http://stm.sciencem

ag.org/D

ownloaded from

REFERENCES AND NOTES

1. J. Schwartzentruber, A. Korshunov, X.-Y. Liu, D. T. W. Jones, E. Pfaff, K. Jacob, D. Sturm,A. M. Fontebasso, D.-A. K. Quang, M. Tönjes, V. Hovestadt, S. Albrecht, M. Kool, A. Nantel,C. Konermann, A. Lindroth, N. Jäger, T. Rausch, M. Ryzhova, J. O. Korbel, T. Hielscher, P. Hauser,M. Garami, A. Klekner, L. Bognar, M. Ebinger, M. U. Schuhmann, W. Scheurlen, A. Pekrun,M. C. Frühwald, W. Roggendorf, C. Kramm, M. Dürken, J. Atkinson, P. Lepage, A. Montpetit,M. Zakrzewska, K. Zakrzewski, P. P. Liberski, Z. Dong, P. Siegel, A. E. Kulozik, M. Zapatka,A. Guha, D. Malkin, J. Felsberg, G. Reifenberger, A. von Deimling, K. Ichimura, V. P. Collins,H. Witt, T. Milde, O. Witt, C. Zhang, P. Castelo-Branco, P. Lichter, D. Faury, U. Tabori, C. Plass,J. Majewski, S. M. Pfister, N. Jabado, Driver mutations in histone H3.3 and chromatin remodel-ling genes in paediatric glioblastoma. Nature 482, 226–231 (2012).

2. Y. Jiao, P. J. Killela, Z. J. Reitman, A. B. Rasheed, C. M. Heaphy, R. F. de Wilde, F. J. Rodriguez,S. Rosemberg, S. M. Oba-Shinjo, S. K. Nagahashi Marie, C. Bettegowda, N. Agrawal, E. Lipp,C. J. Pirozzi, G. Y. Lopez, Y. He, H. S. Friedman, A. H. Friedman, G. J. Riggins, M. Holdhoff,P. Burger, R. E. McLendon, D. D. Bigner, B. Vogelstein, A. K. Meeker, K. W. Kinzler, N. Papadopoulos,L. A. Diaz Jr., H. Yan, Frequent ATRX, CIC, FUBP1 and IDH1 mutations refine the classification ofmalignant gliomas. Oncotarget 3, 709–722 (2012).

3. H. Suzuki, K. Aoki, K. Chiba, Y. Sato, Y. Shiozawa, Y. Shiraishi, T. Shimamura, A. Niida, K. Motomura,F. Ohka, T. Yamamoto, K. Tanahashi, M. Ranjit, T. Wakabayashi, T. Yoshizato, K. Kataoka, K. Yoshida,Y. Nagata, A. Sato-Otsubo, H. Tanaka, M. Sanada, Y. Kondo, H. Nakamura, M. Mizoguchi, T. Abe,Y. Muragaki, R. Watanabe, I. Ito, S. Miyano, A. Natsume, S. Ogawa, Mutational landscape andclonal architecture in grade II and III gliomas. Nat. Genet. 47, 458–468 (2015).

4. J. E. Eckel-Passow, D. H. Lachance, A. M. Molinaro, K. M. Walsh, P. A. Decker, H. Sicotte,M. Pekmezci, T. Rice, M. L. Kosel, I. V. Smirnov, G. Sarkar, A. A. Caron, T. M. Kollmeyer, C. E. Praska,A. R. Chada, C. Halder, H. M. Hansen, L. S. McCoy, P. M. Bracci, R. Marshall, S. Zheng, G. F. Reis,A. R. Pico, B. P. O’Neill, J. C. Buckner, C. Giannini, J. T. Huse, A. Perry, T. Tihan, M. S. Berger, S. M. Chang,M. D. Prados, J. Wiemels, J. K. Wiencke, M. R. Wrensch, R. B. Jenkins, Glioma Groups Based on 1p/19q,IDH, and TERT promoter mutations in tumors. N. Engl. J. Med. 372, 2499–2508 (2015).

5. E. Cerami, J. Gao, U. Dogrusoz, B. E. Gross, S. O. Sumer, B. A. Aksoy, A. Jacobsen, C. J. Byrne,M. L. Heuer, E. Larsson, Y. Antipin, B. Reva, A. P. Goldberg, C. Sander, N. Schultz, The cBiocancer genomics portal: An open platform for exploring multidimensional cancer genomicsdata. Cancer Discov. 2, 401–404 (2012).

6. J. Gao, B. A. Aksoy, U. Dogrusoz, G. Dresdner, B. Gross, S. O. Sumer, Y. Sun, A. Jacobsen, R. Sinha,E. Larsson, E. Cerami, C. Sander, N. Schultz, Integrative analysis of complex cancer genomics andclinical profiles using the cBioPortal. Sci. Signal. 6, pl1 (2013).

7. N. G. Bérubé, M. Mangelsdorf, M. Jagla, J. Vanderluit, D. Garrick, R. J. Gibbons, D. R. Higgs,R. S. Slack, D. J. Picketts, The chromatin-remodeling protein ATRX is critical for neuronalsurvival during corticogenesis. J. Clin. Invest. 115, 258–267 (2005).

8. C. M. Heaphy, R. F. de Wilde, Y. Jiao, A. P. Klein, B. H. Edil, C. Shi, C. Bettegowda, F. J. Rodriguez,C. G. Eberhart, S. Hebbar, G. J. Offerhaus, R. McLendon, B. A. Rasheed, Y. He, H. Yan, D. D. Bigner,S. M. Oba-Shinjo, S. K. Marie, G. J. Riggins, K. W. Kinzler, B. Vogelstein, R. H. Hruban, A. Maitra,N. Papadopoulos, A. K. Meeker, Altered telomeres in tumors with ATRX and DAXX mutations.Science 333, 425 (2011).

9. S. L. Rankin, G. Zhu, S. J. Baker, Review: Insights gained from modelling high-grade gliomain the mouse. Neuropathol. Appl. Neurobiol. 38, 254–270 (2012).

10. S. M. Wiesner, S. A. Decker, J. D. Larson, K. Ericson, C. Forster, J. L. Gallardo, C. Long, Z. L. Demorest,E. A. Zamora, W. C. Low, K. SantaCruz, D. A. Largaespada, J. R. Ohlfest, De novo induction ofgenetically engineered brain tumors in mice using plasmid DNA. Cancer Res. 69, 431–439 (2009).

11. A.-A. Calinescu, F. J. Núñez, C. Koschmann, B. L. Kolb, P. R. Lowenstein, M. G. Castro, Transposonmediated integration of plasmid DNA into the subventricular zone of neonatal mice togenerate novel models of glioblastoma. J. Vis. Exp. e52443 (2015).

12. R. J. Gibbons, D. R. Higgs, Molecular-clinical spectrum of the ATR-X syndrome. Am. J. Med.Genet. 97, 204–212 (2000).

13. B. S. Paugh, C. Qu, C. Jones, Z. Liu, M. Adamowicz-Brice, J. Zhang, D. A. Bax, B. Coyle, J. Barrow,D. Hargrave, J. Lowe, A. Gajjar, W. Zhao, A. Broniscer, D. W. Ellison, R. G. Grundy, S. J. Baker,Integrated molecular genetic profiling of pediatric high-grade gliomas reveals key differenceswith the adult disease. J. Clin. Oncol. 28, 3061–3068 (2010).

14. Cancer Genome Atlas Research Network, Comprehensive genomic characterization defineshuman glioblastoma genes and core pathways. Nature 455, 1061–1068 (2008).

15. S. Alcantara Llaguno, J. Chen, C. H. Kwon, E. L. Jackson, Y. Li, D. K. Burns, A. Alvarez-Buylla,L. F. Parada, Malignant astrocytomas originate from neural stem/progenitor cells in a somatictumor suppressor mouse model. Cancer Cell 15, 45–56 (2009).

16. D. Wei, S. Levic, L. Nie, W.-Q. Gao, C. Petit, E. G. Jones, E. N. Yamoah, Cells of adult braingerminal zone have properties akin to hair cells and can be used to replace inner earsensory cells after damage. Proc. Natl. Acad. Sci. U.S.A. 105, 21000–21005 (2008).

17. Z. Sheng, L. Li, L. J. Zhu, T. W. Smith, A. Demers, A. H. Ross, R. P. Moser, M. R. Green, Agenome-wide RNA interference screen reveals an essential CREB3L2-ATF5-MCL1 survivalpathway in malignant glioma with therapeutic implications. Nat. Med. 16, 671–677 (2010).

18. Y. Xue, R. Gibbons, Z. Yan, D. Yang, T. L. McDowell, S. Sechi, J. Qin, S. Zhou, D. Higgs, W. Wang,The ATRX syndrome protein forms a chromatin-remodeling complex with Daxx and localizes

www.Scien

in promyelocytic leukemia nuclear bodies. Proc. Natl. Acad. Sci. U.S.A. 100, 10635–10640(2003).

19. M. A. Osley, T. Tsukuda, J. A. Nickoloff, ATP-dependent chromatin remodeling factors andDNA damage repair. Mutat. Res. 618, 65–80 (2007).

20. J. W.-C. Leung, G. Ghosal, W. Wang, X. Shen, J. Wang, L. Li, J. Chen, Alpha thalassemia/mental retardation syndrome X-linked gene product ATRX is required for proper replica-tion restart and cellular resistance to replication stress. J. Biol. Chem. 288, 6342–6350(2013).

21. M. Viana-Pereira, A. Lee, S. Popov, D. A. Bax, S. Al-Sarraj, L. R. Bridges, J. N. Stávale, D. Hargrave,C. Jones, R. M. Reis, Microsatellite instability in pediatric high grade glioma is associated withgenomic profile and differential target gene inactivation. PLOS One 6, e20588 (2011).

22. J. Haines, J. Bacher, M. Coster, R. Huiskamp, E. Meijne, M. Mancuso, S. Pazzaglia, S. Bouffler,Microsatellite instability in radiation-induced murine tumours; influence of tumour typeand radiation quality. Int. J. Radiat. Biol. 86, 555–568 (2010).

23. J. W. Bacher, W. M. Abdel Megid, M. G. Kent-First, R. B. Halberg, Use of mononucleotiderepeat markers for detection of microsatellite instability in mouse tumors. Mol. Carcinog.44, 285–292 (2005).

24. E. Vilar, C. M. Bartnik, S. L. Stenzel, L. Raskin, J. Ahn, V. Moreno, B. Mukherjee, M. D. Iniesta,M. A. Morgan, G. Rennert, S. B. Gruber,MRE11 deficiency increases sensitivity to poly(ADP-ribose)polymerase inhibition in microsatellite unstable colorectal cancers. Cancer Res. 71, 2632–2642(2011).

25. Cancer Genome Atlas Research Network, Comprehensive, integrative genomic analysis ofdiffuse lower-grade gliomas. N. Engl. J. Med. 372, 2481–2498 (2015).

26. Y. Wang, J. Yang, H. Zheng, G. J. Tomasek, P. Zhang, P. E. McKeever, E. Y.-H. P. Lee, Y. Zhu,Expression of mutant p53 proteins implicates a lineage relationship between neural stemcells and malignant astrocytic glioma in a murine model. Cancer Cell 15, 514–526 (2009).

27. D. O. Ferguson, J. M. Sekiguchi, S. Chang, K. M. Frank, Y. Gao, R. A. DePinho, F. W. Alt, Thenonhomologous end-joining pathway of DNA repair is required for genomic stability andthe suppression of translocations. Proc. Natl. Acad. Sci. U.S.A. 97, 6630–6633 (2000).

28. R. J. Callicott, J. E. Womack, Real-time PCR assay for measurement of mouse telomeres.Comp. Med. 56, 17–22 (2006).

29. J. D. Henson, Y. Cao, L. I. Huschtscha, A. C. Chang, A. Y. Au, H. A. Pickett, R. R. Reddel, DNAC-circles are specific and quantifiable markers of alternative-lengthening-of-telomeres activity.Nat. Biotechnol. 27, 1181–1185 (2009).

30. Z. Mao, Y. Jiang, X. Liu, A. Seluanov, V. Gorbunova, DNA repair by homologous recombination,but not by nonhomologous end joining, is elevated in breast cancer cells. Neoplasia 11,683–691 (2009).

31. A. Azad, S. Jackson, C. Cullinane, A. Natoli, P. M. Neilsen, D. F. Callen, S.-M. Maira, W. Hackl,G. A. McArthur, B. Solomon, Inhibition of DNA-dependent protein kinase induces acceleratedsenescence in irradiated human cancer cells. Mol. Cancer Res. 9, 1696–1707 (2011).

32. J. Plowman, W. R. Waud, A. D. Koutsoukos, L. V. Rubinstein, T. D. Moore, M. R. Grever,Preclinical antitumor activity of temozolomide in mice: Efficacy against human brain tumorxenografts and synergism with 1,3-bis(2-chloroethyl)-1-nitrosourea. Cancer Res. 54, 3793–3799(1994).

33. I. F. Pollack, M. Erff, D. Bom, T. G. Burke, J. T. Strode, D. P. Curran, Potent topoisomerase Iinhibition by novel silatecans eliminates glioma proliferation in vitro and in vivo. Cancer Res.59, 4898–4905 (1999).

34. J. N. Sarkaria, B. L. Carlson, M. A. Schroeder, P. Grogan, P. D. Brown, C. Giannini, K. V. Ballman,G. J. Kitange, A. Guha, A. Pandita, C. D. James, Use of an orthotopic xenograft model forassessing the effect of epidermal growth factor receptor amplification on glioblastoma radia-tion response. Clin. Cancer Res. 12, 2264–2271 (2006).

35. J. E. Wolff, T. Trilling, G. Mölenkamp, R. M. Egeler, H. Jürgens, Chemosensitivity of gliomacells in vitro: A meta analysis. J. Cancer Res. Clin. Oncol. 125, 481–486 (1999).

36. C. J. Lord, A. Ashworth, The DNA damage response and cancer therapy. Nature 481, 287–294(2012).

37. M. Ramamoorthy, S. Smith, Loss of ATRX suppresses resolution of telomere cohesion tocontrol recombination in ALT cancer cells. Cancer Cell 28, 357–369 (2015).

38. C. A. Lovejoy, W. Li, S. Reisenweber, S. Thongthip, J. Bruno, T. de Lange, S. De, J. H. J. Petrini,P. A. Sung, M. Jasin, J. Rosenbluh, Y. Zwang, B. A. Weir, C. Hatton, E. Ivanova, L. Macconaill,M. Hanna, W. C. Hahn, N. F. Lue, R. R. Reddel, Y. Jiao, K. Kinzler, B. Vogelstein, N. Papadopoulos,A. K. Meeker; ALT Starr Cancer Consortium, Loss of ATRX, genome instability, and an alteredDNA damage response are hallmarks of the alternative lengthening of telomeres pathway.PLOS Genet. 8, e1002772 (2012).

39. S. Negrini, V. G. Gorgoulis, T. D. Halazonetis, Genomic instability—An evolving hallmark ofcancer. Nat. Rev. Mol. Cell Biol. 11, 220–228 (2010).

40. M. H. Sherman, C. H. Bassing, M. A. Teitell, Regulation of cell differentiation by the DNAdamage response. Trends Cell Biol. 21, 312–319 (2011).

41. L. Schneider, S. Pellegatta, R. Favaro, F. Pisati, P. Roncaglia, G. Testa, S. K. Nicolis, G. Finocchiaro,F. d’Adda di Fagagna, DNA damage in mammalian neural stem cells leads to astrocytic dif-ferentiation mediated by BMP2 signaling through JAK-STAT. Stem Cell Reports 1, 123–138(2013).

ceTranslationalMedicine.org 2 March 2016 Vol 8 Issue 328 328ra28 10

R E S EARCH ART I C L E

Dow

nloaded f

42. S. Burma, D. J. Chen, Role of DNA–PK in the cellular response to DNA double-strand breaks.DNA Repair 3, 909–918 (2004).

43. W. P. Roos, B. Kaina, DNA damage-induced cell death by apoptosis. Trends Mol. Med. 12,440–450 (2006).

44. M. Bétermier, P. Bertrand, B. S. Lopez, Is non-homologous end-joining really an inherentlyerror-prone process? PLOS Genet. 10, e1004086 (2014).

Acknowledgments: We thank J. Ohlfest (University of Minnesota, deceased) for his support ofour implementation of the SB model. C.K. wishes to thank P. Robertson and H. Garton for theiracademic support. We gratefully acknowledge P. Jenkins and the Department of Neurosurgeryat the University of Michigan Medical School for their support of our work. We are also gratefulto K. Muraszko for her academic leadership and D. Tomford, S. Napolitan, M. Dahlgren, andC. Shaw for superb administrative support. This study makes use of data generated by the St. JudeChildren’s Research Hospital–Washington University Pediatric Cancer Genome Project, theprincipal investigator C. Hawkins and the Hospital for Sick Children, the McGill UniversityHealth Centre and the DKFZ (Deutsches Krebsforschungszentrum)–University of HeidelbergPediatric Brain Tumour Consortium, and the Institute of Cancer Research (ICR)–Institut GustavRoussy–Hospital Sant Joan de Déu collaborative group. Funding: This work was supported byNIH/National Institute of Neurological Disorders and Stroke (NINDS) grants 1RO1-NS 054193,1RO1-NS 061107, and 1RO1-NS082311 to P.R.L.; grants 1UO1-NS052465, 1RO1-NS 057711, and1RO1-NS074387 to M.G.C., and grant NIH/National Cancer Institute R01CA172380 to A. Meeker.C.K. was supported by the St. Baldrick’s Foundation Fellowship and the Alex’s Lemonade StandFoundation/Northwestern Mutual Young Investigator Grant. C.J., A. Mackay, and J.F.-S. acknowl-edge National Health Service funding to the National Institute for Health Research BiomedicalResearch Centre at The Royal Marsden and the ICR, and the INSTINCT network funded by The

www.Scien

Brain Tumour Charity, Great Ormond Street Hospital Children’s Charity, and Children with CancerUK. Author contributions: C.K. carried out the animal and in vitro studies and drafted themanuscript. A.-A.C., F.J.N., D.T., F.M., N.K., M.D., L.M., J.K., R.L., Y.L., and S.R. participated in theanimal and in vitro studies. L.Z. performed statistical analysis. A. Meeker and J.A.B.-C. assistedwith the design and interpretation of ALT studies (FISH and c-circle). H.A. provided humanPanNET samples. D.F. assisted with the design and interpretation of metaphase preparation/chromosome counting. A. Mackay, J.F.-S., and C.J. performed analysis of human pediatric gliomadata sets. V.G. assisted with the design and interpretation of DDR plasmid assays. M.G.C. and P.R.L.conceived and supervised the study, participated in its design and coordination, and helped todraft and edit the manuscript. All authors read and approved the final manuscript. Competinginterests: The authors declare that they have no competing interests. Data and materials avail-ability: We have deposited previously unpublished sequence data for additional pediatric high-grade glioma samples (C.J., EGAS00001001436).

Submitted 18 June 2015Accepted 25 January 2016Published 2 March 201610.1126/scitranslmed.aac8228

Citation: C. Koschmann, A.-A. Calinescu, F. J. Nunez, A. Mackay, J. Fazal-Salom, D. Thomas, F. Mendez,N. Kamran, M. Dzaman, L. Mulpuri, J. Krasinkiewicz, R. Doherty, R. Lemons, J. A. Brosnan-Cashman, Y. Li,S. Roh, L. Zhao, H. Appelman, D. Ferguson, V. Gorbunova, A. Meeker, C. Jones, P. R. Lowenstein,M. G. Castro, ATRX loss promotes tumor growth and impairs nonhomologous end joining DNArepair in glioma. Sci. Transl. Med. 8, 328ra28 (2016).

http://stm.sciencem

arom

ceTranslationalMedicine.org 2 March 2016 Vol 8 Issue 328 328ra28 11

by guest on October 14, 2020

g.org/

in gliomaATRX loss promotes tumor growth and impairs nonhomologous end joining DNA repair

Gorbunova, Alan Meeker, Chris Jones, Pedro R. Lowenstein and Maria G. CastroLemons, Jacqueline A. Brosnan-Cashman, Youping Li, Soyeon Roh, Lili Zhao, Henry Appelman, David Ferguson, VeraMendez, Neha Kamran, Marta Dzaman, Lakshman Mulpuri, Johnathon Krasinkiewicz, Robert Doherty, Rosemary Carl Koschmann, Anda-Alexandra Calinescu, Felipe J. Nunez, Alan Mackay, Janet Fazal-Salom, Daniel Thomas, Flor

DOI: 10.1126/scitranslmed.aac8228, 328ra28328ra28.8Sci Transl Med

susceptible to treatment.presence of ATRX mutation correlated with better outcomes in patients, because these tumors were morethat damage the DNA, such as radiation and some types of chemotherapy. Consistent with these findings, the

treatmentsquickly. However, because of their DNA repair defect, these tumors also proved to be more sensitive to impairs DNA repair, resulting in genetically unstable tumors that can accumulate oncogenic mutations moreaggressively than their counterparts with wild-type ATRX. The reason this happens is that the loss of ATRX

. developed a mouse model of ATRX-deficient glioma and discovered that these tumors grow moreet alATRX is a protein that is often mutated in glioma, a lethal and relatively common brain tumor. Koschmann

Aggressive gliomas' Achilles' heel

ARTICLE TOOLS http://stm.sciencemag.org/content/8/328/328ra28

MATERIALSSUPPLEMENTARY http://stm.sciencemag.org/content/suppl/2016/02/29/8.328.328ra28.DC1

CONTENTRELATED

http://stm.sciencemag.org/content/scitransmed/11/492/eaav4508.fullhttp://stm.sciencemag.org/content/scitransmed/11/479/eaaq1427.fullhttp://stke.sciencemag.org/content/sigtrans/11/540/eaat0229.fullhttp://stm.sciencemag.org/content/scitransmed/10/443/eaah6816.fullhttp://science.sciencemag.org/content/sci/352/6288/aad3018.fullhttp://science.sciencemag.org/content/sci/355/6320/40.fullhttp://stm.sciencemag.org/content/scitransmed/8/366/366ra161.fullhttp://stm.sciencemag.org/content/scitransmed/8/362/362ps17.fullhttp://stm.sciencemag.org/content/scitransmed/8/350/350fs14.fullhttp://stm.sciencemag.org/content/scitransmed/8/343/343re2.fullhttp://science.sciencemag.org/content/sci/352/6290/1178.fullhttp://stm.sciencemag.org/content/scitransmed/8/341/341ra75.fullhttp://stm.sciencemag.org/content/scitransmed/6/229/229ra42.fullhttp://stm.sciencemag.org/content/scitransmed/7/289/289ra86.fullhttp://stm.sciencemag.org/content/scitransmed/7/304/304ra143.fullhttp://stm.sciencemag.org/content/scitransmed/7/309/309ra163.full

REFERENCES

http://stm.sciencemag.org/content/8/328/328ra28#BIBLThis article cites 43 articles, 14 of which you can access for free

Terms of ServiceUse of this article is subject to the

registered trademark of AAAS. is aScience Translational MedicineScience, 1200 New York Avenue NW, Washington, DC 20005. The title

(ISSN 1946-6242) is published by the American Association for the Advancement ofScience Translational Medicine

Copyright © 2016, American Association for the Advancement of Science

by guest on October 14, 2020

http://stm.sciencem

ag.org/D

ownloaded from

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

registered trademark of AAAS. is aScience Translational MedicineScience, 1200 New York Avenue NW, Washington, DC 20005. The title

(ISSN 1946-6242) is published by the American Association for the Advancement ofScience Translational Medicine

Copyright © 2016, American Association for the Advancement of Science

by guest on October 14, 2020

http://stm.sciencem

ag.org/D

ownloaded from