Cancer Stem Cells and Radiotherapy: New Insights Into = Tumor Radioresistance -- Diehn and Clarke 98 (24): 1755 -- JNCI Journal = of the National Cancer Institute

JNCI Journal of = the National=20 Cancer Institute 2006 98(24):1755-1757; doi:10.1093/jnci/djj505 =

This = Article

	Ex= tract=20
	FR= EE=20 Full Text (PDF)
	Sub= mit=20 a response
	Aler= t=20 me when this article is cited
	Alert=20 me when eLetters are posted
	Al= ert=20 me if a correction is posted

Services

	Email=20 this article to a friend
	Similar=20 articles in this journal
	Similar=20 articles in PubMed
	Alert me to=20 new issues of the journal
	Add=20 to My Personal Archive
	= Download=20 to citation manager
	Request=20 Permissions

Google = Scholar

	Articles by Diehn, M. =
	Articles by Clarke, M. F. =

PubMed

	PubMed=20 Citation
	Articles by Diehn, M. =
	Articles by Clarke, M. F.=20 =

=A9 The Author = 2006.=20 Published by Oxford University Press.

EDITORIAL

Cancer Stem Cells and Radiotherapy: New Insights Into Tumor = Radioresistance=20

Maximilian Diehn and Michael F.=20 Clarke

Affiliations of authors: Department of = Radiation=20 Oncology, Stanford University School of Medicine, Stanford, CA (MD); = Stanford=20 Institute for Stem Cell Biology and Regenerative Medicine, Palo Alto, CA = (MD,=20 MFC)

Correspondence: Correspondence to: Maximilian = Diehn, MD,=20 PhD, Department of Radiation Oncology, Stanford University School of = Medicine,=20 875 Blake Wilbur Dr., MC-5847, Stanford, CA 94305 (e-mail: diehn{at}stanford.edu ).

Solid tumors are histologically heterogeneous and include tumor⁼²⁰cells, stroma, inflammatory infiltrates, and vascular = structures.⁼²⁰Furthermore, it has long been recognized that subpopulations⁼of=20 cancer cells exist within the tumors that resemble the = developmental⁼²⁰hierarchy of the normal tissue from which the tumor arose. In⁼recent years, the cancer stem cell model of tumorigenesis has⁼received increasing attention. This model posits that tumors⁼²⁰are driven and maintained by a minority subpopulation of = cells⁼²⁰that have the capacity to self-renew (i.e., give rise to = progeny⁼²⁰with similar properties as themselves) and to generate the = more⁼²⁰differentiated progeny which make up the bulk of a tumor (= 1).⁼²⁰The former population has been termed cancer stem cells = (CSCs),⁼²⁰tumorigenic cancer cells, or tumor-initiating cells, by = various⁼²⁰investigators, to indicate that only they can give rise to = new⁼²⁰tumors when transplanted into immunodeficient animals.

Evidence for the existence of CSCs initially came from studies⁼of=20 acute myelogenous leukemia (AML), among which a subset of⁼leukemic=20 cells was identified that gave rise to AML in immunodeficient⁼mice=20 and that displayed a similar cell surface immunophenotypeas = normal=20 hematopoietic stem cells (= 2).=20 Subsequently, CSCs wereisolated from solid tumors, including = glioblastoma, medulloblastoma(= 3),=20 breast cancer (= 4),=20 melanoma (= 5),=20 and prostate cancer (= 6).⁼²⁰The existence of CSCs has profound implications for cancer = biology⁼²⁰and therapy because it is likely that eradication of CSCs is⁼²⁰the critical determinant in achieving cure. It has been = proposed⁼²⁰that CSCs may be particularly resistant to chemotherapy and⁼²⁰radiation therapy, although good evidence supporting this = notion⁼²⁰has been lacking. In this issue of the Journal, Phillips = et⁼²⁰al. (= 7)=20 report evidence that cells from breast cancer cell linesthat = resemble breast cancer=96initiating cells are radioresistant⁼compared=20 with the remainder of breast cancer cells. Similarly,a = recently=20 published report by Bao et al. (= 8)=20 suggests thatglioblastoma stem cells are radioresistant and = may=20 thereforecontribute to treatment failures.

Phillips et al. chose established breast cancer cell lines as⁼²⁰their model system. Specifically, they applied conditions = that⁼²⁰were previously used for culturing mammospheres from primary⁼²⁰breast cancer specimens (= 9)=20 to culture nonadherent cells thatthey isolated from two = adherent=20 breast cancer cell lines. Thesenonadherent subpopulations = could be=20 grown as spheroids in mammospheremedia and contained a = larger=20 fraction of cells than the adherentsubpopulation with the=20 CD24^=96/low/CD44⁺ phenotype, whichwas = previously=20 shown to identify breast tumor=96initiatingcells in primary = patient=20 specimens (= 4).=20 When the spheroids oradherent cultures were irradiated in = vitro, the=20 cells arisingfrom spheroids were more radioresistant, with = an=20 absolute differencein mean surviving fraction at 2 Gy of=20 approximately 20%. Importantly,both cell populations were = irradiated=20 as single-cell suspensions,thus removing the complicating = factor of=20 low oxygen tensionat the center of the spheroids. The = authors also=20 examined severalmolecular assays relevant to = radiosensitivity,=20 including levelsof reactive oxygen species and = phosphorylation of=20 histone H2AX,both of which were decreased in spheroid = cultures.=20 Furthermore,fractionated radiation appeared to increase the=20 percentage ofnonadherent = CD24^=96/low/CD44⁺ cells=20 in monolayer cultures,suggesting that the relative = radioresistance=20 of this subsetmay lead to their expansion during a course of = radiotherapy.Thus, the authors present evidence that = CSC-like cells=20 frombreast cancer cell lines appear to be = radioresistant.⁼²⁰

The findings of Phillips et al. are clearly provocative. = However,⁼²⁰close inspection of the study reveals several important = issues⁼²⁰that should be considered when interpreting CSC-related = studies.⁼²⁰First, as the authors point out, attempting to apply stem = cell⁼²⁰concepts initially described in primary tumors to cell lines⁼is=20 fraught with pitfalls. Because established cell lines have⁼been=20 passaged in artificial conditions for many generations,it is = possible that selection for growth in tissue culture has⁼blurred the=20 distinction between tumorigenic and nontumorigeniccells. = Indeed, a=20 recent study on CSC-like subpopulations inmurine ovarian = cancer cell=20 lines found that both the CSC-likeand non=96CSC-like = subpopulations=20 gave rise to tumors innude mice but that the non=96CSC-like = cells=20 formed tumorssomewhat more slowly than the CSC-like = population=20 (10).⁼²⁰

Second, culture conditions could contribute to functional = differences⁼²⁰between CSCs and non-CSCs found in vitro. The nonadherent = spheroids⁼²⁰in the study by Phillips et al. were grown in the presence of⁼epidermal growth factor and basic fibroblast growth factor = but⁼²⁰the adherent cells were not, and both growth factors have = been⁼²⁰shown previously to decrease radiosensitivity (11,12).=20 Accordingly,when these growth factors were added to the = adherent=20 cultures,their radioresistance was similar to that of the=20 nonadherentspheroids. Although the authors document a modest = increase inthe percentage of = CD24^=96/low/CD44⁺=20 cells in the monolayercultures after addition of the growth = factors=20 that may at leastpartially explain the increased = radioresistance,=20 the differentialculture conditions complicate interpretation = of=20 their findings.

Given the effects of selection for growth in tissue culture,⁼²⁰CSC-like subpopulations within established cell lines might⁼not=20 necessarily display the same cell surface immunophenotypesas = CSCs=20 within tumors. In general, the surface proteins thatare used = for=20 identifying CSCs have not been shown to be necessaryor = sufficient=20 for conferring stem cell=96like propertiesand thus must = currently be=20 viewed as purification markers withoutfunctional = implications.=20 Therefore, rather than simply ascribingCSC-like properties = to cells=20 of a given surface immunophenotype,it is crucial to show = that this=20 population has the functionalcharacteristics of CSCs. To = accurately=20 compare results fromdifferent studies, it is imperative to = show, in=20 each experimentalsystem, that purported CSC-like cells can=20 self-renew and giverise to more differentiated progeny and = that=20 non-CSCs lack theseproperties. Currently, this difference = can be=20 most convincinglydemonstrated by serial transplantation of = CSC-like=20 and non=96CSC-likecells in animal models (= 1).⁼²⁰

Another important point relevant to all CSC-related studies⁼²⁰concerns the purity of CSC populations. Given the number of⁼²⁰CSCs needed to initiate tumors in rodent models (generally in⁼the range of hundreds to thousands), it is possible that only⁼a=20 subset of cells within these populations will turn out tobe = true=20 CSCs. In the system used by Phillips et al., the spheroid⁼cultures=20 contained approximately 40% CD24^=96/low/CD44⁺ = cells⁼²⁰and thus may have contained an even smaller fraction of true⁼²⁰CSC-like cells. The presence of non-CSCs within the CSC = population⁼²⁰could cloud functional analyses because results might reflect⁼the properties of the contaminating cells (possibly early=20 progenitors)rather than those of CSCs. It will therefore be=20 important toreevaluate functional characteristics, such as=20 sensitivity totherapeutics, once purer populations of CSCs = can be=20 isolated.

Phillips et al. provide evidence that CSC-like cells may be⁼²⁰intrinsically radioresistant, a finding that could clearly = have⁼²⁰important clinical implications. However, it is possible that⁼in vivo microenvironmental effects either positively or = negatively⁼²⁰modulate this property. For example, if it turns out that = CSCs⁼²⁰tend to be located in hypoxic regions of tumors, this would⁼²⁰likely affect their sensitivity to radiotherapy via the = oxygen⁼²⁰enhancement ratio. Alternatively, if the cell cycle = distribution⁼²⁰of CSCs and non-CSCs in vivo differs (and one might well = expect⁼²⁰this to be true), cell cycle effects might contribute to in⁼²⁰vivo radiosensitivity. To address such questions, effects of⁼²⁰ionizing radiation on CSCs also should be examined in the = context⁼²⁰of whole tumors, ideally both in xenograft models and in = human⁼²⁰patients.

The contemporaneous study by Bao et al. both confirms and = extends⁼²⁰the findings of Phillips et al. These authors studied = xenografts⁼²⁰from primary glioblastoma multiforme specimens, thus = addressing⁼²⁰many of the aforementioned concerns regarding established = cell⁼²⁰lines. Prior work identified CD133⁺ cells as the=20 tumorigenicpopulation in primary glioblastoma multiforme = specimens=20 (= 3),⁼²⁰and Bao et al. found that these cells were radioresistant = compared⁼²⁰with CD133⁼⁹⁶ tumor cells. Importantly, they showed = that⁼²⁰CD133⁺ cells accumulated after irradiation both in = vitro=20 (i.e.,neurosphere cultures) and in vivo (i.e., xenografts).=20 Furthermore,they demonstrated that the modest enrichment of=20 CD133⁺ cellsafter irradiation has biologic = relevance by=20 showing that a slightincrease in the percentage of CD133+ = cells in=20 suspensions usedto initiate tumors dramatically increased = their=20 growth rate.Akin to the work by Phillips et al., Bao et al. = also=20 examinedmultiple molecular markers of radiation damage, = including=20 apoptosis,histone H2AX phosphorylation, and phosphorylation = of genes=20 involvedin DNA damage repair, and found that by each of = these=20 measures,CD133⁺ cells were more radioresistant. = They=20 concluded that CD133⁺cells can activate DNA = damage=20 checkpoint responses to a greaterdegree than = CD133⁼⁹⁶ cells=20 and thus repair DNA damage moreefficiently. Intriguingly, by = using=20 an inhibitor of the checkpointkinases Chk1 and Chk2, they = were able=20 to radiosensitize theCD133⁺ cells.

Taken together, the two studies suggest that radioresistance⁼may=20 be a general property of CSCs and that this may be due to⁼their=20 ability to more efficiently repair DNA than non-CSCs.As with = most=20 provocative studies, they raise more questionsthan they = answer: Are=20 CSCs of all tumor types radioresistant?Can the accumulation = of CSCs=20 after radiotherapy seen in vitroand in xenografts be = documented in=20 patients? Does the percentageof CSCs within human tumors = predict=20 radiosensitivity? Will CSCsbe similarly resistant to=20 chemotherapeutic agents? How aboutthe newer targeted therapy = agents?=20 Will it be possible to developradio- or chemosensitizers = that=20 preferentially sensitize CSCscompared with normal = tissues?⁼²⁰

The findings by Phillips et al. and Bao et al. have important⁼²⁰clinical implications. In light of these studies, it is even⁼²⁰clearer that identifying and characterizing CSCs for every = tumor⁼²⁰possible is of paramount importance and will likely lead to⁼new=20 therapeutic avenues. In the case of radiotherapy, it seems⁼possible=20 that response assays that capture the effects of radiationon = CSCs=20 rather than the bulk of a tumor could prove to be more⁼sensitive=20 predictors of treatment outcome than traditional measuresof=20 treatment response. Also, work on radiosensitizers should⁼begin to=20 focus on preferentially affecting CSCs compared withnormal = tissues=20 and normal tissue stem cells. With these studies,the often = quoted=20 notion that solid tumor CSCs may contributeto treatment = failure has=20 begun to be substantiated.

REFERENCES

Clarke MF, Dick JE, Dirks PB, Eaves CJ, Jamieson CH, = Jones DL,=20 et al. Cancer stem cells=97perspectives on current status and future = directions:=20 AACR workshop on cancer stem cells. Cancer Res 2006;66:9339=9644.[Free Full Text]
Bonnet D, Dick JE. Human acute myeloid leukemia is = organized as=20 a hierarchy that originates from a primitive hematopoietic cell. Nat = Med=20 1997;3:730=967.[CrossRef][ISI][Medline]
Singh SK, Hawkins C, Clarke ID, Squire JA, Bayani J, = Hide T, et=20 al. Identification of human brain tumour initiating cells. Nature=20 2004;432:396=96401.[CrossRef][ISI][Medline]
Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, = Clarke MF.=20 Prospective identification of tumorigenic breast cancer cells. Proc = Natl Acad=20 Sci U S A 2003;100:3983=968.[Abstract/Free Full Text]
Fang D, Nguyen TK, Leishear K, Finko R, Kulp AN, Hotz = S, et al.=20 A tumorigenic subpopulation with stem cell properties in melanomas. = Cancer Res=20 2005;65:9328=9637.[Abstract/Free Full Text]
Patrawala L, Calhoun T, Schneider-Broussard R, Li H, = Bhatia B,=20 Tang S, et al. Highly purified CD44+ prostate cancer cells from = xenograft=20 human tumors are enriched in tumorigenic and metastatic progenitor = cells.=20 Oncogene 2006;25:1696=96708.[CrossRef][ISI][Medline]
Phillips TM, McBride WH, Pajonk F. The response of=20 CD24^=96/low/CD44⁺ breast cancer=96initiating = cells to=20 radiation. J Natl Cancer Inst 2006;98:1777=9685.[Abstract/Free Full Text]
Bao S, Wu Q, McLendon RE, Hao Y, Shi Q, Hjelmeland AB, = et al.=20 Glioma stem cells promote radioresistance by preferential activation = of the=20 DNA damage response. Nature. 2006 Oct 18; [Epub ahead of print].
Ponti D, Costa A, Zaffaroni N, Pratesi G, Petrangolini = G,=20 Coradini D, et al. Isolation and in vitro propagation of tumorigenic = breast=20 cancer cells with stem/progenitor cell properties. Cancer Res = 2005;65:5506=9611.[Abstract/Free Full Text]
Szotek PP, Pieretti-Vanmarcke R, Masiakos PT, = Dinulescu DM,=20 Connolly D,Foster R, et al. Ovarian cancer side population defines = cells with=20 stem cell-like characteristics and Mullerian Inhibiting Substance=20 responsiveness. Proc Natl Acad Sci U S A 2006;103:11154=969.[Abstract/Free Full Text]
Wollman R, Yahalom J, Maxy R, Pinto J, Fuks Z. Effect = of=20 epidermal growth factor on the growth and radiation sensitivity of = human=20 breast cancer cells in vitro. Int J Radiat Oncol Biol Phys = 1994;30:91=968.[ISI][Medline]
Paris F, Fuks Z, Kang A, Capodieci P, Juan G, = Ehleiter D, et=20 al. Endothelial apoptosis as the primary lesion initiating intestinal=20 radiation damage in mice. Science 2001;293:293=967.[Abstract/Free Full Text]

This = Article

	Ex= tract=20
	FR= EE=20 Full Text (PDF)
	Sub= mit=20 a response
	Aler= t=20 me when this article is cited
	Alert=20 me when eLetters are posted
	Al= ert=20 me if a correction is posted

Services

	Email=20 this article to a friend
	Similar=20 articles in this journal
	Similar=20 articles in PubMed
	Alert me to=20 new issues of the journal
	Add=20 to My Personal Archive
	= Download=20 to citation manager
	Request=20 Permissions

Google = Scholar

	Articles by Diehn, M. =
	Articles by Clarke, M. F. =

PubMed

	PubMed=20 Citation
	Articles by Diehn, M. =
	Articles by Clarke, M. F.=20 =

JNCI Journal of the = National Cancer=20 Institute

EDITORIAL

Cancer Stem Cells and Radiotherapy: New Insights Into Tumor = Radioresistance=20