cancer research dna damage recognition via activated atm ... · therapeutics, targets, and chemical...

Ther

DNAPat

Sari JUlf-HåKimm

Abst

Intro

DNAintegrmultiprepairpermanature

AuthorHelsinkImagingBiomed3DeparInstitu5DeparUniversCentralCanceMarylaInstitute

Note:Resear

S. Jääm

Corresgy andof Med21231.jhmi.ed

doi: 10

©2010

Cance8630

Dow

Canceresearch

apeutics, Targets, and Chemical Biology

Damage Recognition via Activated ATM and p53

R

hway in Nonproliferating Human Prostate Tissue

äämaa1,3, Taija M. af Hällström1, Anna Sankila4, Ville Rantanen2, Hannu Koistinen5,
kan Stenman5, Zhewei Zhang7, Zhiming Yang7, Angelo M. De Marzo7, o Taari6, Mirja Ruutu6, Leif C. Andersson4,8, and Marikki Laiho1,7
ractDNA

modelslowlyby cytcontexmine aphospcells laylationseemstargetfindin

nent repland amo

s' Affiliationi and DeparUnit and Cicine and Gtment of Virte, Universtment of Clity CentralHospital, Hr, Johns Hnd; and 8Dt, Stockhol

Supplemench Online (h

aa and T.M

ponding AuMolecular Ricine, 1550Phone: 41u.

.1158/0008-

American A

r Res; 70(

nloaded

damage response (DDR) pathways have been extensively studied in cancer cell lines and mouses, but little is known about how DNA damage is recognized by different cell types in nonmalignant,replicating human tissues. Here, we assess, using ex vivo cultures of human prostate tissue, DDR causedotoxic drugs (camptothecin, doxorubicin, etoposide, and cisplatin) and ionizing radiation (IR) in thet of normal tissue architecture. Using specific markers for basal and luminal epithelial cells, we deter-nd quantify cell compartment–specific damage recognition. IR, doxorubicin, and etoposide induced thehorylation of H2A.X on Ser139 (γH2AX) and DNA damage foci formation. Surprisingly, luminal epithelialck the prominent γH2AX response after IR when compared with basal cells, although ATM phosphor-on Ser1981 and 53BP1 foci were clearly detectable in both cell types. The attenuated γH2AX responseto result from low levels of total H2A.X in the luminal cells. Marked increase in p53, a downstreamof the activated ATM pathway, was detected only in response to camptothecin and doxorubicin. Thesegs emphasize the diversity of pathways activated by DNA damage in slowly replicating tissues and revealxpected deviation in the prostate luminal compartment that may be relevant in prostate tumorigenesis.
an une
Detailed mapping of tissue and cell type differences in DDR will provide an outlook of relevant responses totherapeutic strategies. Cancer Res; 70(21); 8630–41. ©2010 AACR.

and dare actypesof centissuehuma

duction

damage response (DDR) pathways protect genomicity and form a barrier against cancer (1). DDR includesle mechanisms that detect damaged DNA, activate themachineries, delay cell cycle progression, and cause

icative arrest or activate apoptosis (2). Theunt of DNA damage, tissue-specific factors,

molecCell-bnomicmatiotissuesupprmammDNA dDN

rimenpair cradiatposedsenso[Ser198

sion bits ho(DNA-phospafter Dand hkinase

s: 1Molecular Cancer Biology Program, Biomedicumtment of Virology, Haartman Institute and 2Molecularomputational Systems Biology Laboratory, Institute ofenome-Scale Biology Program, University of Helsinki;ology and 4Department of Pathology and Haartmanity of Helsinki Hospital Laboratory Diagnostics;inical Chemistry, University of Helsinki and HelsinkiHospital; 6Department of Urology, Helsinki Universityelsinki, Finland; 7The Sidney Kimmel Comprehensiveopkins University School of Medicine, Baltimore,epartment of Pathology and Oncology, Karolinskam, Sweden

tary data for this article are available at Cancerttp://cancerres.aacrjournals.org/).

. af Hällström contributed equally to this work.

thor: Marikki Laiho, Department of Radiation Oncolo-adiation Sciences, Johns Hopkins University SchoolOrleans Street, CRB II, Room 444, Baltimore, MD0-502-9748; Fax: 443-502-2821; E-mail: mlaiho1@

5472.CAN-10-0937

ssociation for Cancer Research.

21) November 1, 2010

on May 2, 2017. © 20cancerres.aacrjournals.org from

evelopmental stage determine types of pathways thattivated (3–5). It is not known why some tissue and cellare more prone to tumorigenesis and why inactivationtral DDR pathways leads to cancer in only certains (6, 7). To better understand how DDR participates inn disease, it will be useful to delineate which signalingules participate in the responses in different tissues.ased assays have been most useful in studying how ge-integrity is maintained, but they do not provide infor-n on how heterotopic interactions or three-dimensionalenvironment affect DDR. Equally, despite that tumoressor and DDR pathways are generally conserved inals, mouse models do not recapitulate all aspects ofamage surveillance in human cells or tissues (8).A double-strand breaks (DSB) are considered most det-tal for genomic integrity, because their unsuccessful re-an lead to chromosomal aberrations. After ionizingion (IR), the DSB recognition and modification of ex-chromatin occurs within minutes, and one of the earlyrs, activated phosphoinositide 3-kinase (PI3K) ATM1 phosphorylated ATM (pATM)] is recruited to the le-y the Mre11-Rad50-Nbs1 (MRN) complex (9, 10). ATM,mologue ATR, and DNA-dependent protein kinasePK) are considered the primary kinases redundantlyhorylating histone variant H2A.X at its Ser139 residueSB and other types of DNA damage stress in mouse
uman cells (11, 12). Tissue specificity between thes has been shown to exist in vivo (13, 14). γH2AX is
10 American Association for Cancer Research.

http://cancerres.aacrjournals.org/

one ofconsidty (17,initialblastsproteifoci (1formcomplof DNaffectp53

cyclenescen(22). Iand pthymufor p5(13), aic epitIn agrculturthelialentialmechacontroIn t

responand IRsons,also wand tovate Aparedcells sand 53that Hlack opossibmulat

Mate

Ex vivPro

cystecHospiwith iby insNA_00Tablecoredthe cyKrumdCorp.)turedmente(20 nm

and 5mensslicesmentsDox

Aldricinto Dwas pand st137Cs

ImmuTiss

beddedeparby hea0.01 mbufferat +4°(1:50,γH2A(1:50,tomicCell SSignaltomicand C(RT),cal);Aldric(5-BrdSer198

(Asp17

antiboor Al(1:100nylatewere icountProbeMerckratorie

ImageTra

researcolorimagecenceCamand ZsingleMETAPlan-ADiodeand H

DNA Damage Response in Normal Prostate Tissue

www.a

Dow

the first proteins accumulating at DSB (15, 16) and isered essential for the maintenance of genomic integri-18). γH2AX has been shown to be dispensable for therecognition of DNA breaks in murine embryo fibro-(19) but is required for the retention of several DDRns, such as 53BP1, BRCA1, and MDC1, at DNA damage7, 18, 20). Together with MDC1, γH2AX forms a plat-onto which other DDR proteins, such as MRN-ATMex, 53BP1, and BRCA1, attach (11). The compositionA repair proteins at the damage site may vary andwhich DDR pathway is activated (21).tumor suppressor protein regulates key steps in cellcheckpoint activation, apoptosis, DNA repair, and se-ce and is frequently mutated in most cancer typesn response to IR, ATM phosphorylates p53 on Ser15,53 is stabilized at least in radiosensitive tissues likes and spleen (5, 23). However, ATM is not required3 induction in small intestine or epidermis in mousend we recently reported that human primary prostat-helial cells show attenuated p53 response to IR (24).eement, previous studies have shown that primaryes of human mammary epithelial cells, bronchial epi-cells, and human keratinocytes also exhibit a differ-p53 IR response (25). These findings indicate thatnisms may have evolved to maintain p53 under tightl in epithelial tissues.his study, we investigate how human prostate tissueds to DNA damage caused by chemotherapeutic drugs. We show that p53 is stabilized by topoisomerase poi-proteasomal and transcriptional stress in epithelium,ithout prior activation of ATM pathway, and that IRpoisomerase II poisons induce γH2AX foci and acti-TM in nonreplicating prostate cells. Surprisingly, com-with basal cells after DNA damage, luminal epithelialhow negligible levels of γH2AX foci, although pATMBP1 foci are clearly detectable in both cell types. Given2A.X is required for efficient DNA DSB repair and thef it leads to chromosomal instability (11, 17, 18), it isle that luminal epithelial cells are more prone to accu-ing DNA lesions.

rials and Methods

o tissue culture and treatmentsstate tissue was collected from 18 patients undergoingtomy or prostatectomy at Helsinki University Centraltal or at Johns Hopkins Medicine during 2007 to 2009nformed written consent of the patients and approvaltitutional review boards (390/E6/06, 371/E6/07, and015481). Clinical data are presented in SupplementaryS1. A cylinder of prostate tissue (8-mm diameter) wasout of the peripheral dorsal region of the prostate, andlinder was sliced at 300- to 500-μm thickness usingiek tissue slicer (Alabama Research and Development. The slices were placed onto titanium grids and cul-in serum-free RPMF-4A tissue culture medium supple-
d with growth factors and androgen analogue R1881ol/L) and incubated in a rotating platform at +37°C
ters, a40× an

acrjournals.org

on May 2, 2017. © 20cancerres.aacrjournals.org nloaded from

% CO2 (24). The histopathology of H&E-stained speci-was evaluated by A.S., and only nonmalignant tissuewere included in the study. All experimental treat-of the cultures were performed on days 1 to 2.orubicin, etoposide, camptothecin, cisplatin (Sigma-h), and MG132 (Biomol International LP) were dilutedMSO, aliquoted, and stored at −20°C. Actinomycin Durchased from Sigma, diluted into ethanol, aliquoted,ored at −20°C. Tissues were irradiated with calibratedγ-ray irradiator (BioBeam 8000, STS).

nohistochemistryues were fixed in 10% neutral buffered formalin, em-d in paraffin, and cut to 4 μm. The sections wereaffinized and rehydrated, and antigens were retrievedting the slides in a microwave oven (700–800 W) inol/L sodium citrate (pH 6.0) or Tris-EDTA (pH 9.0)up to 20 minutes. Sections were incubated overnightC with the following primary antibodies: p53, DO-7Dako) and 7F5 (1:50, Cell Signaling Technology);X, clone JBW301 (1:800, Upstate Biotechnology), 20E3Cell Signaling Technology), and pSer139 (1:400, Epi-s); histone H2A.X (1:200, Novus Biologicals and 1:50,ignaling Technology); p21 Waf1/Cip1 12D1 (1:200, Celling); p27/Kip1 (1:100, BD Biosciences and 1:200, Epi-s); p63 Ab-4 (1:100, Neomarkers); cytokeratin 5 (CK5)K14 [1:50, incubated 1 hour at room temperatureImgenex]; androgen receptor (AR; 1:50, Biocare Medi-α-smooth muscle actin (α-SMA) (1:1,000, Sigma-h); Ki67, MIB-1 (1:100, Dako); 5-bromo-2-deoxyuridineUrd; 1:100, Sigma); 53BP1 (1:200, Novus Biologicals);1 pATM (1:100, Cell Signaling); and cleaved caspase-35) 5A1E (1:100, Cell Signaling). Secondary fluorescentdies, antimouse- or antirabbit-conjugated Alexa-488exa-594, were purchased from Molecular Probes) and incubated at RT for 30 minutes. Secondary bioti-d antimouse or antirabbit antibodies (1:100, Dako)ncubated at RT for 30 minutes each. The tissues wereerstained with Hoechst 33342 (1 μg/mL; Moleculars) or with hematoxylin (Mayers hemalum solution,KGaA) and mounted with Vectashield (Vector Labo-s) or GVA mount (Invitrogen).

acquisition and analysisnsmitted light images were captured with Leica DM LBch microscope using N Plan objectives, Olympus DP50camera, and StudioLite 1.0 software. Fluorescences were acquired with Zeiss Axioplan 2 MOT epifluores-microscope, Zeiss Plan-Neofluar objectives, Zeiss Axio-HRm 14-bit grayscale charge-coupled device cameraeiss Axiovision 4.6 software. Confocal images weresections acquired sequentially using Zeiss LSM 510confocal microscope, 40× Plan-Neofluar DIC and 63×pochromat DIC oil objectives, LSM 510 point scanner,405 laser for Hoechst, Argon 488 laser for Alexa-488eNe1 543 laser for Alexa-594 label with appropriate fil-
nd LSM software release 3.2. The optical sections ford 63× images were <1 μm.
Cancer Res; 70(21) November 1, 2010 8631



Forcaptucenceeach tperforvelopeSegmecell nuand searea aexpresolding86.0 μThe

SD. Mproteisity innucleuof theSign

Basedfour dof meand ouplay hFor p5four ccells wcutoff(DO-7ationp53 (7cutoff

Resu

Ex vivnormPro

radicamaintex vivotissuesue sestandapathotissuewith Hterminmarkeboth istanda(Supping arp63 ancompacompacompa

sectio(Fig. 1In a

ref. 26plemeR1881sion awere csectiofirst 2Ki67-pInterethe riming, wexposAs

ductiosues.pointsincreacultur∼1,00fold lhealthin theFig. Sof thementapressiproducludemodeDNA-d

DivernonreWe

attenu(24). WtissueγH2Asal mato repbasedsarcoculturdoxorintrasinduceplemeysis, wbasedcells athe anhigh adense

Jäämaa et al.

Cance8632

Dow

quantification, two to six fields of each specimen werered using 40× objective on the wide-field epifluores-microscope. The total number of cells analyzed forreatment was on average 2,600. Image analysis wasmed using a Mathworks Matlab-based application de-d for the purpose and was designated as Cell Imagentation and Classifier (CISC). CISC was set to detectclei by thresholding DNA staining at several intensitiesarching for shapes fitting the expectations of nuclearnd roundness (Supplementary Fig. S1A). Roundness issed as the ratio (4π·area)/(perimeter length)2. Thresh-was set to identify objects of areas between 21.5 andm2 and roundness of >0.5.intensity of each protein was expressed as mean andargination was used for the outlier analysis. For nuclearns, margination was defined as the ratio of mean inten-the center of the nucleus and at the borders of thes. The border width was defined as 30% of the widthnucleus in question.al intensities were classified using k-means algorithm.on initial visual observation, we expected to retainifferent intensity classes for γH2AX in ascending orderan intensities: negative, weak positive, strong positive,tlier. The outlier group consists of dead cells that dis-igh-intensity nuclear staining (Supplementary Fig. S1).3 and p63, we obtained six, and for p27, we obtainedlasses representing negatively and positively stainedith different staining intensities. We determined thevalue of a positively stained cell for γH2AX, p53

), p27, and p63 as y > −0.62x + 1.3, wherein y = margin-and x = intensity mean (Supplementary Fig. S1D). ForF5), the cutoff value was y > −0.62x + 1.4, and for p21,was y > −1.33x + 1.6.

lts

o cultured prostate tissues retain viability andal cytoarchitecturestate tissue was collected from patients undergoingl prostatectomy or cystectomy (Fig. 1A), sliced, andained in culture up to 7 days. To confirm that thecultured tissues maintained viability and that overallmorphology was preserved in culture environment, tis-ctions cultured for 1, 2, and 7 days were stained usingrd H&E protocol and were evaluated by experiencedlogists (A.S., L.C.A., A.M.DM.; Fig. 1B). In addition, allsections included in the study were routinely stained&E and were confirmed to be nonmalignant. We de-ed the expression of established compartment-specificrs to ensure the viability of the three major cell typesn cultured prostate tissues (Fig. 1C) and, as control, inrd paraffin-embedded routine sections of the prostatelementary Fig. S2A). Conjugate controls for each stain-e provided in Supplementary Fig. S2B. Expression ofd cytokeratins 5 and 14 was detectable in the basal cellrtment, p27 and AR were expressed in the luminal cell
rtment, and α-SMA was expressed by stromal cells atrable levels to standard paraffin-embedded prostate
apoptthat t

r Res; 70(21) November 1, 2010


ns, although AR levels somewhat declined after a weekC; Supplementary Fig. S2A).dult prostate, the number of replicating cells is low (<1%;). Although the culture medium is serum free, it is sup-nted with mitogenic factors and androgen analoguethat could support replication. Based on Ki67 expres-nd 5-BrdUrd incorporation, only a few replicating cellsomparable with standard paraffin-embedded prostatens in the cultured prostate tissue sections during thedays (Fig. 1D; Supplementary Fig. S2C). The number ofositive cells increased somewhat after 7 days in culture.stingly, 5-BrdUrd–positive cells were found especially ins of the tissue sections with glands exposed to section-here epithelial cells seem to be recruited to cover theed surfaces (Supplementary Fig. S2C).a marker for luminal cell function, we tested for pro-n of prostate-specific antigen (PSA) by the cultured tis-We collected tissue culture medium at various timeand measured total PSA levels. On average, PSA levelssed to 1,700 and 2,900 μg/L after 24 and 48 hours ine, respectively (Supplementary Fig. S3A), which are0-fold higher than levels present in plasma and 1,000-ower than levels measured in the seminal fluid ofy men (27). There was a steady increase of PSAculture medium from 2 to 22 hours (Supplementary3B). PSA was also detected in the apical surfacesluminal cells in the cultured prostate tissues (Supple-ry Fig. S3C). Based on the intact cytoarchitecture, ex-on of several compartment-specific proteins, andction of an androgen-dependent marker, PSA, we con-that the ex vivo cultured prostate tissues are a tenablel for the analysis of human tissue level responses toamaging agents.

gent activation of γH2AX and p53 inplicating prostate tissuehave earlier shown that prostate epithelial cells haveated DNA damage checkpoint and p53 responsese therefore wanted to assess how the normal prostaterecognizes different forms of DNA damage. We usedX as a rapid marker for DDR and p53 as a more univer-rker for cellular stress. We selected incubation timesresent early and late responses of γH2AX and p53on DDR in LNCaP prostate cancer and U2-OS osteo-ma cells (Supplementary Fig. S4). We treated theed prostate tissues with IR; topoisomerase poisonsubicin, etoposide, and camptothecin; and DNAtrand crosslinker cisplatin, which are prominentrs of p53 and γH2AX in replicating cancer cells (Sup-ntary Fig. S4). To increase the robustness of the anal-e developed a quantitative image analysis applicationon k-means algorithm that allows identification ofnd their signal intensities (Supplementary Fig. S1). Inalyses of γH2AX, we excluded, as outliers, cells withnd even γH2AX staining and which displayed con-d chromatin and fragmented nuclei consistent with cell
osis (Supplementary Fig. S1; refs. 28, 29). To ascertainhe outlier cells were in fact apoptotic, we stained the
Cancer Research



tissuesmarkeconfiran apo

At 8a lesse

Figureexpressproteinof the emethodwere cuwere cuslices wand cuH&E stcell typCK4/15immunoof the texpressand 7 d


www.a

Dow

for γH2AX and cleaved caspase-3, a known apoptoticr. This showed colocalization of the two signals
ming that the γH2AX outlier analysis is suitable asptotic marker (Fig. 2D; Supplementary F
(Fig. 2

. Scale bars, 100 μm.

acrjournals.org


hours, γH2AX foci were robustly induced by IR and tor extent by doxorubicin, etoposide, and camptothecin
A and B). Cisplatin did not affect γH2AX. At 24 hours,
ig. S5). the intensity of γH2AX staining by IR, doxorubicin, and

1. Cultured prostate tissues retainion of compartment-specific markers and tissue architecture. A, outlinex vivo prostate tissue culture. Prostate tissue cores (Ø 8 mm)t to 300 to 500 μm, and the slicesltured. Following treatments, theere fixed, embedded into paraffin,

t into 4-μm sections. B, standardaining of cultured slices. C, prostatee–specific markers (p63, p27, AR,, and α-SMA) were detected byhistochemistry following culturingissues for 1 to 7 d. D, Ki67ion in tissue slices cultured for 1, 2,




etopodamawas ththe prmentabut nFig. Ssomeitive cthe p5lack o

γH2AespecγH2Afor p5tothecnot ethe trhours(Fig. 2

Figure(Doxo,after whfields wfor p53applicatwo-tainuclei,t test (*of γH2Awere counterstained with Hoechst 33342 (DNA). Confocal images were captured at 40× magnification. Scale bar, 50 μm. Insets from the indicatedregions are enlarged by 2.9-fold.

Jäämaa et al.

Cance8634

Dow

side was reduced, possibly due to repair of the DNAge (Fig. 2B; Supplementary Fig. S6). Camptothecine strongest and most consistent inducer of p53 inostate epithelium at 8 and 24 hours (Fig. 2; Supple-ry Fig. S6). Doxorubicin stabilized p53 in someot in all patient specimens (Fig. 2; Supplementary7). IR increased the number of p53-positive cells inpatient specimens, but compared with occasional pos-ells in the control, there was no significant increase in
3 intensity (Fig. 2B and C). Furthermore, there was af correlation in the induction of p53 and expression
relatio(data



X foci by the different treatments. This was evidentially following IR treatment where in abundantlyX-positive sections only a few cells stained weakly3 (Fig. 2A and B; data not shown). Conversely, camp-in, which most strongly induced p53 at 8 hours, didlicit a pronounced γH2AX foci formation. None ofeatments induced a robust apoptotic response at 24, despite the relatively high drug and IR doses usedD). Furthermore, at single cell level, there was no cor-

2. Divergent γH2AX and p53 responses to cytotoxic drugs and IR in human prostate tissues. A, tissue slices were treated with doxorubicin2 μmol/L), etoposide (Eto, 40 μmol/L), camptothecin (Cpt, 2 μmol/L), and cisplatin (Cis, 40 μmol/L) or exposed to IR (10 Gy) and incubated for 8 h,ich the slices were fixed and stained for p53 and γH2AX. Nuclei were counterstained with Hoechst 33342 (DNA). Confocal images of representativeere captured at 40× magnification. Scale bar, 50 μm. Insets from the indicated regions are enlarged by 3.5-fold. B, quantitative image analysisand γH2AX. Images were acquired from two to six fields of each specimen (n = 3–4 patients per treatment) and quantified using CISC imagingtion. The total number of cells analyzed for each treatment varied from 1,300 to 3,900 cells. Error bars, SD. P values were calculated using Student'sled t test (*, P < 0.05; **, P < 0.01; ***, P < 0.001) compared with control. C, relative intensity of p53 expression. The mean intensity of p53-positiveanalyzed as in B, was normalized by the mean intensity surrounding each nucleus. P values were calculated using Student's two-tailed, P < 0.05; **, P < 0.01; ***, P < 0.001) compared with control. D, tissues were treated as in A, incubated for 8 or 24 h and analyzed for the fractionX outliers (left). Immunostaining of control and camptothecin-treated tissues for cleaved caspase-3 (green) and γH2AX (red; right). Nuclei

n between p53 expression and the apoptotic responsenot shown).

Cancer Research



Thep53 byeringand thDNAshownnutlinport binhibiactivitp53 aFig. S

actinoepitheepitheand pTo

leadsstreamcin wfractioexpres

FigurecamptoNuclei3.8-foldthree p**, P <


www.a

Dow

apparent sensitivity of the prostate cells to stabilizedoxorubicin and camptothecin is interesting consid-that the prostate is a largely nonreplicating tissueat the topoisomerase poisons mainly act by causingDSBs in replicating cells (30–32). We have earlierthat p53 is induced in the prostate tissue by

-3, a negative regulator of MDM2, and a nuclear ex-locker leptomycin B (24). We hence tested whethertion of proteasome function or RNA polymerase Iy, known effectors of p53 pathway, would regulate
ctivity in prostate. As shown in Supplementary ly betw
atients per treatment (1,500–3,300 cells per treatment). Error bars, SD. P values0.01; ***, P < 0.001) compared with control. C, correlation analysis of p53 and p2

acrjournals.org


mycin D resulted in robust p53 accumulation inlial cells. These results indicate that the prostatelium displays expected responses to transcriptionalroteasomal stress.address whether the apparent lack of p53 stabilizationto a defect in the activation of a known p53 down-target, we assessed for regulation of p21. Camptothe-

as the most prominent inducer of p21 (Fig. 3). Then of p21-positive cells correlated quite well with p53sion (R2 = 0.8137), although the responses varied great-
een the patient specimens (Fig. 3A and C; Supplemen-
8, treatment of the prostate tissues with MG132 or tary Fig. S7).

3. p21 expression correlates with p53 responses. Tissue slices were treated with doxorubicin (Doxo, 2 μmol/L), etoposide (Eto, 40 μmol/L),thecin (Cpt, 2 μmol/L), cisplatin (Cis, 40 μmol/L), and IR (10 Gy) and incubated for 24 h, after which the slices were fixed and stained for p21 and p53.were counterstained with Hoechst 33342. Confocal images of representative fields were captured at 40× magnification. Scale bar, 50 μm. Insets,enlargement. B, quantitative image analysis for p21 using CISC imaging application from three to six fields of each specimen from two to

were calculated using Student's two-tailed t test (*, P < 0.05;1 expression in two patient specimens.




Figure(Cpt, 2with Hoenlarge(1,400–compa

Jäämaa et al.

Cance8636

Dow

4. p53 is stabilized in basal and luminal cells. A, prostate tissues were incubated with doxorubicin (Doxo, 2 μmol/L) and camptothecinμmol/L) or treated with IR (10 Gy) for 24 h, after which the tissues were fixed and stained for p53 and p63. Nuclei were counterstainedechst 33342 (DNA). Confocal images of representative fields were captured at 40× magnification. Scale bar, 50 μm. Insets, 3.3-foldment. B, p53 and p63 expression was quantified using CISC from three to four fields of each specimen from two patients per treatment
1,900 cells per treatment). Altogether, 1,800 basal and 4,950 nonbasal cells were analyzed. Error bars, SD. P values (**, P < 0.01) representrison of basal and nonbasal cells.
r Res; 70(21) November 1, 2010 Cancer Research

on May 2, 2017. © 2010 American Association for Cancer Research. cancerres.aacrjournals.org nloaded from


DispaepitheEpit

and inthe efγH2AXic resbasalfurthethe th

in baslizatiothe trSur

only lalthoubasalFig. S

FigureγH2AXspecimcompa2.5-fold ue. Sca


www.a

Dow

rate γH2AX responses to DSB in the prostatelial compartmentshelial and stromal cells are relatively radioresistant,general, rapidly replicating cells are more prone to

fects of cytotoxic drugs. Given the variability in theand p53 responses, we analyzed the cell type specif-

ponses by coimmunostaining γH2AX and p53 withand luminal markers, p63 and p27. The images were

enlargement. D, H2A.X and p27 staining in ex vivo cultured prostate tiss

r subjected to quantitative coexpression analyses. Ofree major cell types, p53 response was evident both

signalalso i

acrjournals.org


al and luminal epithelial cells (Fig. 4A and B). Stabi-n of p53 was not detected in stromal cells by any ofeatments.prisingly, following DSB inducing treatments, like IR,ow levels of γH2AX were detected in luminal cells,gh γH2AX foci were clearly visible in the adjacentand stromal cells (Fig. 5A and B; Supplementary9). At earlier time points (1–4 hours) when γH2AX

le bar, 50 μm. Insets, 2.8-fold enlargement.

5. Cell type–specific induction of γH2AX. A, tissue slices were treated with IR (10 Gy) for 8 h, after which the slices were fixed and stained forand p27. Scale bar, 50 μm. Insets, 2.8-fold enlargement. B, γH2AX and p27 expression was quantified using CISC from two to four fields of eachen from two patients per treatment (600–1,300 cells per treatment). Error bars, SD. P values (*, P < 0.05; **, P < 0.01; ***, P < 0.001) representrison of luminal and nonluminal cells. C, immunohistochemistry staining of H2A.X in paraffin-embedded prostate tissue. Scale bar, 100 μm. Inset,

was maximal, we could detect some γH2AX-signaln the luminal cells, although it was considerably




lowerS9B;staineculturwas eof H2histon(Fig. 5

pATMluminγH2

the re

nearlackwe wDDRtion.in boor cispATMand pFig. SPan

Figure(10 Gy)were co2.8-fold

Jäämaa et al.

Cance8638

Dow

than in basal cells and stroma (Supplementary Fig.data not shown). To investigate this further, wed standard paraffin-embedded prostate tissue anded tissue specimens to clarify whether total H2A.Xxpressed in luminal cells. Compared with the levelA.X present in the basal cells, the expression of thise variant in the luminal compartment was negligibleC and D).

and 53BP1 foci are sustained both in basal andal cells
AX is considered an essential scaffold protein fortention of DDR proteins on the chromatin at or
untreaand S

unterstained with Hoechst 33342 (DNA). Confocal images of representative fieldenlargement.



DNA DSB (11, 16, 17, 19, 20). Also, given that theof H2A.X leads to increased genomic instability,anted to assess the extent of activation of thepathway by analyzing pATM and 53BP1 foci forma-IR, doxorubicin, and etoposide induced pATM focith luminal and basal cells, whereas camptothecinplatin did not (Fig. 6A; Supplementary Fig. S10).foci colocalized with γH2AX foci in the basal cellsersisted even at 24 hours after IR (Supplementary10B).-nuclear 53BP1 was expressed in most luminal cells of
ted prostate tissue (Fig. 6B; Supplementary Figs. S1112). After IR, doxorubicin, and etoposide treatments,
6. DNA DSBs caused by IR lead to ATM and 53BP1 foci formation in both basal and luminal cells. A, prostate tissue slices were treated with IRand cultured for 8 h. Tissues were fixed and stained for ATM phosphorylated on Ser1981 and γH2AX (A) and 53BP1 and γH2AX (B). Nuclei

s were captured at 40× magnification. Scale bar, 50 μm. Insets,

Cancer Research



53BP1luminweredata nin damlittle γ

Discu

Theare bahumansurrogalternrespoprostaexperihumainterianalyswith Istressrobuscell cois actiings intinctmannepAT

and eting th(Supption wto thefoci fothe aThis rγH2Acandidrequirtransilationdosesor Mdetectdid nnonreapoptcells oand m(32, 3p53

sponswith ain noof apDNA-dantia

patienpathosomeinterinproposes (4scriptstabiliindicamal ccells pthat pa potobservcells fintensATMzationHis

mammbasallevelsculturmensγH2Aby allwereat eara lowuisiteareasH2A.Xand -iexpresof theaffectturedepigenentiatThe

instabonic clow leto lauLuminless scells awithby MYtheir vtion, acompled tothese(49).simila


www.a

Dow

accumulated into damage foci in basal, stromal, andal cells (Fig. 6B; Supplementary Fig. S11), and the focistill detectable by 24 hours (Supplementary Fig. S12;ot shown). Thus, the retention of pATM and 53BP1age foci of the prostate luminal cells seems to requireH2AX.

ssion

current concepts of prostate cancer biologysed on nonprimate animal models and few availableprostate cancer cell lines. These are unsatisfactory

ates for the human disease. We present here anative method to assay how different cell typesnd to DNA damage in ex vivo cultured humante tissues. The method provides a unique means tomentally test drugs and their combinations in liven normal and tumor tissues and to assess possiblendividual variations. Based on quantitative imagees for DDR proteins in prostate tissues treatedR and cytotoxic drugs, we show that the DNA damagees activate divergent responses. Notably, despitet activation of ATM kinase cascade, the luminalmpartment lacks γH2AX response, and p53 pathwayvated by only a subset of the treatments. These find-dicate that the DNA damage stresses activate a dis-set of responses in a cell compartment–dependentr.M and 53BP1 foci were detected in IR-, doxorubicin-,oposide-treated epithelium and stromal cells indicat-at the DDR upstream pathway was activated

lementary Table S2). However, γH2AX phosphoryla-as nonexistent in the luminal cells in responsese treatments. The finding is surprising as γH2AXrmation is considered essential for the assembly ofdaptor proteins 53BP1, MDC1, and BRCA1 (33).aises the question whether other factors besidesX stabilize damage foci onto chromatin. Possibleates could be H3 and H4 acetylation or methylationed for 53BP1 focus formation (34) or that even early,ent γH2AX response is sufficient to initiate accumu-of the downstream proteins. Additionally, higherof IR may bypass the requirement for γH2AX

DC1 (35, 36). γH2AX phosphorylation was alsoed in response to camptothecin treatment, whichot activate ATM. As the prostate represents aplicating tissue, it is plausible that, in addition tootic γH2AX signal, γH2AX activation in postmitoticccurs consequent to transcription-mediated lesionsay be activated by other PI3Ks like ATR or DNA-PK7, 38).response was launched most prominently in re-

e to camptothecin and doxorubicin and correlatedn increase in p21. The relevance of p21 inductionnproliferating cells is not evident. Given the lackparent apoptotic response to the relatively high
amaging treatments, p21 induction may reflect anpoptotic response (39). Doxorubicin displayed
they eof DN

acrjournals.org


t-to-patient variation in p53 stabilization in the non-logic glands, which could depend on the levels of topoi-rase II or other underlying causes resulting in thedividual variation (31). Nucleolar disruption has beensed as a common denominator for p53-inducing stres-0), which could explain why treatments causing tran-ional stress, like actinomycin D, led to prominent p53zation in prostate tissue. Interestingly, our results alsote that the epithelial cells are more prone than stro-ells to p53 stabilization. It is possible that in stromal53 stabilization is hindered by alternative signals or53 is transcriptionally silenced. IR, however, was notent stabilizer of p53 in the prostate. Whereas wee an increase in the fraction of p53-positive epithelialrom 3% to 9%, there is no significant difference in theity of the p53 signal. Lastly, our results indicate thatactivation alone is not sufficient for acute p53 stabili-in the prostate.

tone variant H2A.X comprises 10% of the total H2A inalian cells (11, 15). Unexpectedly, compared with

and stromal cells, luminal epithelial cells express lowof H2A.X. This difference was observed both in ex vivoed and standard paraffin-embedded prostate speci-in histologically normal glands. Consequently, theX response of the luminal cells was greatly diminishedDNA-damaging treatments. Following IR, faint fociobserved only with high doses of radiation (10 Gy)ly time points (<4 hours). This indicates that even iflevel of phosphorylation of H2A.X is attained, the req-expansion of the γH2AX signal over large chromatinis likely lacking in the luminal cells. Histone variantisoforms are transcribed in replication-dependent

ndependent manners (41), so it is plausible that itssion levels are determined by the differentiated statecells. Chromatin compaction has been indicated tothe extent of DDR and γH2AX foci formation in cul-cells (42–44), further suggesting that the chromatinetic state may determine the extent of DDR in differ-ed cells.lack or low levels of H2A.X predisposes to genomicility and repair defects in fibroblasts, mouse embry-ells, and lymphocytes (17, 18, 45, 46). Therefore, avel of H2A.X potentially renders the cells unablench a full DDR in the presence of DNA damage.al cells are postmitotic, which would make themusceptible to tumorigenesis. However, luminalre predisposed to inflammatory atrophy associatedincreased proliferation (47) and transformationC leading to Nkx3.1 downregulation (48), indicatingulnerability to cell stress and tumorigenesis. In addi-luminal stem cell capable of reconstituting luminal

artment was described in mice, and Pten deletionrapid formation of carcinomas, suggesting that

cells represent a population prone for transformationFurther studies are needed to assess whetherr luminal stem cells exist in humans and whether
xpress H2A.X and are capable of proficient repairA DSBs.



Disclo

No p

Ackn

We ttechnicaHelen FRepositDr. Karfacilitiediscussi

Refe1. Jac

and2. Ha

Mo3. Bo

expbe64

4. Fetar

5. Guto

6. Badig

7. Bamelostum

8. EfeATsup

9. Lethr55

10. BechrNa

11. BoCa

12. Cimrity

13. GuforCa

14. Kodeirra

15. Rodoser

16. Rodo199

17. Cemic

18. BasenPro

19. Cephbre

20. Pa

Jäämaa et al.

Cance8640

Dow

sure of Potential Conflicts of Interest Grant

AcadHällströS. Jäämdicum HMedicalhas bee

Theof pageaccorda

otential conflicts of interest were disclosed.

owledgments

hank Maija Salo, Ulla Kiiski, and Anni-Helena Sukupolvi for excellentl assistance, Kristiina Rantanen for assistance in patient recruitment,edor and the Brady Urological Research Institute Prostate Specimenory for specimen collection, Dr. Jorma Palvimo for providing R1881,
i Alitalo's laboratory for the use of their immunohistochemicals, and members of Laiho Laboratories in Helsinki and Hopkins for Rece
OnlineF

osphorylation is dispensable for the initial recognition of DNAaks. Nat Cell Biol 2003;5:675–9.ull TT, Rogakou EP, Yamazaki V, Kirchgessner CU, Gellert M,

Bofac

21. Xieasstr

22. Meres

23. Cana28

24. Kivproch72

25. Hares

26. Laisomo20

27. Liljca8:2

28. Mulatlian

29. Sophres

30. PoNa

31. Nitica

32. SomuDN

33. Stda(Am

34. Fitism

35. FeindNa

36. Lostada

37. Mocinind

38. FuH2



Support

emy of Finland grants 129699 (M. Laiho) and 126827 (T.M. afm) and Patrick C. Walsh Prostate Cancer Research Fund (M. Laiho).aa has received support from K. Albin Johansson Foundation, Biome-elsinki Foundation, Finnish-Norwegian Medical Foundation, FinnishFoundation, and Helsinki Biomedical Graduate School. V. Rantanenn supported by EU FP7/2007-2011 grant 201837.costs of publication of this article were defrayed in part by the paymentcharges. This article must therefore be hereby marked advertisement innce with 18 U.S.C. Section 1734 solely to indicate this fact.

ived 03/16/2010; revised 07/27/2010; accepted 08/08/2010; publishedirst 10/26/2010.
on.
renceskson SP, Bartek J. The DNA-damage response in human biologydisease. Nature 2009;461:1071–8.

rper JW, Elledge SJ. The DNA damage response: ten years after.l Cell 2007;28:739–45.uvard V, Zaitchouk T, Vacher M, et al. Tissue and cell-specificression of the p53-target genes: Bax, fas, mdm2 and waf1/p21,

fore and following ionising irradiation in mice. Oncogene 2000;19:9–60.i P, Bernhard EJ, El-Deiry WS. Tissue-specific induction of p53gets in vivo. Cancer Res 2002;62:7316–27.dkov AV, Komarova EA. The role of p53 in determining sensitivityradiotherapy. Nat Rev Cancer 2003;3:117–29.rlow C, Hirotsune S, Paylor R, et al. Atm-deficient mice: A para-m of ataxia telangiectasia. Cell 1996;86:159–71.rtkova J, Horejsi Z, Sehested M, et al. DNA damage responsediators MDC1 and 53BP1: Constitutive activation and aberrants in breast and lung cancer, but not in testicular germ cellours. Oncogene 2007;26:7414–22.yan A, Murga M, Martinez-Pastor B, et al. Limited role of murineM in oncogene-induced senescence and p53-dependent tumorpression. PLoS One 2009;4:e5475.

e JH, Paull TT. ATM activation by DNA double-strand breaksough the Mre11-Rad50-Nbs1 complex. Science 2005;308:1–4.rkovich E, Monnat RJ, Jr., Kastan MB. Roles of ATM and NBS1 inomatin structure modulation and DNA double-strand break repair.t Cell Biol 2007;9:683–90.nner WM, Redon CE, Dickey JS, et al. γH2AX and cancer. Nat Revncer 2008;8:957–67.prich KA, Cortez D. ATR: an essential regulator of genome integ-. Nat Rev Mol Cell Biol 2008;9:616–27.rley KE, Kemp CJ. Ataxia-telangiectasia mutated is not requiredp53 induction and apoptosis in irradiated epithelial tissues. Molncer Res 2007;5:1312–8.ike M, Sugasawa J, Yasuda M, Koike A. Tissue-specific DNA-PK-pendent H2AX phosphorylation and γ-H2AX elimination after X-diation in vivo. Biochem Biophys Res Commun 2008;376:52–5.gakou EP, Pilch DR, Orr AH, Ivanova VS, Bonner WM. DNAuble-stranded breaks induce histone H2AX phosphorylation onine 139. J Biol Chem 1998;273:5858–68.gakou EP, Boon C, Redon C, Bonner WM. Megabase chromatinmains involved in DNA double-strand breaks in vivo. J Cell Biol9;146:905–16.leste A, Petersen S, Romanienko PJ, et al. Genomic instability ine lacking histone H2AX. Science 2002;296:922–7.ssing CH, Chua KF, Sekiguchi J, et al. Increased ionizing radiationsitivity and genomic instability in the absence of histone H2AX.c Natl Acad Sci U S A 2002;99:8173–8.leste A, Fernandez-Capetillo O, Kruhlak MJ, et al. Histone H2AX

nner WM. A critical role for histone H2AX in recruitment of repairtors to nuclear foci after DNA damage. Curr Biol 2000;10:886–95.A, Hartlerode A, Stucki M, et al. Distinct roles of chromatin-

sociated proteins MDC1 and 53BP1 in mammalian double-and break repair. Mol Cell 2007;28:1045–57.ek DW. Tumour suppression by p53: a role for the DNA damageponse? Nat Rev Cancer 2009;9:714–23.nman CE, Lim DS, Cimprich KA, et al. Activation of the ATM ki-se by ionizing radiation and phosphorylation of p53. Science 1998;1:1677–9.iharju-af Hallstrom TM, Jaamaa S, Monkkonen M, et al. Humanstate epithelium lacks Wee1A-mediated DNA damage-inducedeckpoint enforcement. Proc Natl Acad Sci U S A 2007;104:11–6.llstrom TM, Laiho M. Genetic changes and DNA damageponses in the prostate. Prostate 2008;68:902–18.czko I, Hudson DL, Freeman A, Feneley MR, Masters JR. Compar-n of the zones of the human prostate with the seminal vesicle:rphology, immunohistochemistry, and cell kinetics. Prostate05;62:260–6.a H, Ulmert D, Vickers AJ. Prostate-specific antigen and prostatencer: prediction, detection and monitoring. Nat Rev Cancer 2008;68–78.kherjee B, Kessinger C, Kobayashi J, et al. DNA-PK phosphory-es histone H2AX during apoptotic DNA fragmentation in mamma-cells. DNA Repair (Amst) 2006;5:575–90.

lier S, Pommier Y. The apoptotic ring: a novel entity with phos-orylated histones H2AX and H2B, activated DNA. damageponse kinases. Cell Cycle 2009;8:1853–9.mmier Y. Topoisomerase I inhibitors: camptothecins and beyond.t Rev Cancer 2006;6:789–802.iss JL. DNA topoisomerase II and its growing repertoire of biolog-l functions. Nat Rev Cancer 2009;9:327–37.rdet O, Redon CE, Guirouilh-Barbat J, et al. Ataxia telangiectasiatated activation by transcription- and topoisomerase I-inducedA double-strand breaks. EMBO Rep 2009;10:887–93.ucki M, Jackson SP. γH2AX and MDC1: Anchoring the DNA-mage-response machinery to broken chromosomes. DNA Repairst) 2006;5:534–43.

zGerald JE, GrenonM, Lowndes NF. 53BP1: function and mechan-s of focal recruitment. Biochem Soc Trans 2009;37:897–904.rnandez-Capetillo O, Chen HT, Celeste A, et al. DNA damage-uced G2-M checkpoint activation by histone H2AX and 53BP1.t Cell Biol 2002;4:993–7.u Z, Minter-Dykhouse K, Franco S, et al. MDC1 maintains genomicbility by participating in the amplification of ATM-dependent DNAmage signals. Mol Cell 2006;21:187–200.rris EJ, Geller HM. Induction of neuronal apoptosis by camptothe-, an inhibitor of DNA topoisomerase-I: evidence for cell cycle-
ependent toxicity. J Cell Biol 1996;134:757–70.ruta T, Takemura H, Liao ZY, et al. Phosphorylation of histoneAX and activation of Mre11, Rad50, and Nbs1 in response to
Cancer Research



repma20

39. Abact

40. Ruof200

41. Maa cboAc

42. Mustr

43. CoγHPL

44. Aymoda

45. Cesuf20

46. Bade20

47. Deca

48. Iwatatep


www.a

Dow

lication-dependent DNA double-strand breaks induced bymmalian DNA topoisomerase I cleavage complexes. J Biol Chem03;278:20303–12.bas T, Dutta A. P21 in cancer: intricate networks and multipleivities. Nat Rev Cancer 2009;9:400–14.bbi CP, Milner J. Disruption of the nucleolus mediates stabilizationp53 in response to DNA damage and other stresses. EMBO J3;22:6068–77.nnironi C, Bonner WM, Hatch CL. H2A X a histone isoprotein withonserved C-terminal sequence, is encoded by a novel mRNA withth DNA replication type and polyA 3′ processing signals. Nucleicids Res 1989;17:9113–26.rga M, Jaco I, Fan Y, et al. Global chromatin compaction limits theength of the DNA damage response. J Cell Biol 2007;178:1101–8.
well IG, Sunter NJ, Singh PB, Austin CA, Durkacz BW, Tilby MJ.2AX foci form preferentially in euchromatin after ionising-radiation.oS One 2007;2:e1057.
49. Walial46

acrjournals.org


oub N, Jeyasekharan AD, Bernal JA, Venkitaraman AR. HP1-βbilization promotes chromatin changes that initiate the DNAmage response. Nature 2008;453:682–6.leste A, Difilippantonio S, Difilippantonio MJ, et al. H2AX haploin-ficiency modifies genomic stability and tumor susceptibility. Cell03;114:371–83.ssing CH, Suh H, Ferguson DO, et al. Histone H2AX: a dosage-pendent suppressor of oncogenic translocations and tumors. Cell03;114:359–70.Marzo AM, Platz EA, Sutcliffe S, et al. Inflammation in prostate

rcinogenesis. Nat Rev Cancer 2007;7:256–69.ta T, Schultz D, Hicks J, et al. MYC overexpression induces pros-ic intraepithelial neoplasia and loss of Nkx3.1 in mouse luminalithelial cells. PLoS One 2010;5:e9427.ng X, Kruithof-de Julio M, Economides KD, et al. A luminal epithe-
stem cell that is a cell of origin for prostate cancer. Nature 2009;1:495–500.



2010;70:8630-8641. Cancer Res Sari Jäämaa, Taija M. af Hällström, Anna Sankila, et al. in Nonproliferating Human Prostate TissueDNA Damage Recognition via Activated ATM and p53 Pathway

Updated version

http://cancerres.aacrjournals.org/content/70/21/8630

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2010/10/25/0008-5472.CAN-10-0937.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/70/21/8630.full.html#ref-list-1

This article cites 49 articles, 14 of which you can access for free at:

Citing articles

/content/70/21/8630.full.html#related-urls

This article has been cited by 4 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

[email protected] at

To request permission to re-use all or part of this article, contact the AACR Publications

on May 2, 2017. © 2010 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/content/70/21/8630

http://cancerres.aacrjournals.org/content/suppl/2010/10/25/0008-5472.CAN-10-0937.DC1

http://cancerres.aacrjournals.org/content/70/21/8630.full.html#ref-list-1

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

mailto:[email protected]


cancer research dna damage recognition via activated atm ... · therapeutics, targets, and chemical...

Documents