Brain Tumor Stem = Cells

http://www.medscape.com/ =

To Print: Click = your browser's=20 PRINT button.

NOTE: To view = the article=20 with Web enhancements, go to:
http://www.medscape.c= om/viewarticle/540653

Brain Tumor Stem Cells: New Targets for = Clinical=20 Treatments?

Patrizia Tunici, Ph.D.; Dwain Irvin, Ph.D., = M.P.H.;=20 Gentao Liu, Ph.D.; Xiangpeng Yuan, Ph.D.; Zeng Zhaohui, Ph.D.; = Hiushan Ng,=20 M.S.; John S. Yu, M.D.

Neurosurg Focus. 2006;20(4):E27. = =A92006 American=20 Association of Neurological Surgeons

Posted 08/07/2006

Abstract and Introduction

Abstract

The observation of similarities between the self-renewal = mechanisms of=20 stem cells and cancer cells has led to the new concept of the = cancer stem=20 cell. In cases of leukemia, multiple myeloma, and breast cancer, = cells=20 with a high self-renewal potential have been identified. = Furthermore,=20 investigators have shown these cells' ability to drive the = formation and=20 growth of the tumor. Brain tumors have also been reported to = possess a=20 subpopulation of cancer stemlike cells that have the ability to=20 proliferate, self-renew, and be multipotent. When grafted into = mice, these=20 cells are also able to generate a tumor that recapitulates that of = the=20 patient from whom the cells were derived. The identification and=20 characterization of this new category of cells call for new = therapies=20 capable of selectively targeting and killing these multifaceted = cells.

Introduction

It was first reported in a study of leukemia that a population = of=20 neoplastic cells exhibit heterogeneity with respect to = proliferation and=20 differentiation.^[7] These stem-like cells, present = within the=20 leukemic population, possess a high capacity for self-renewal and=20 proliferation that is not present in the majority of leukemic=20 cells.^[6,24,50] An analysis of stem cells present in = leukemia=20 tumors demonstrated that their presence is necessary and = sufficient to=20 maintain the tumor population.^[36] Cancer stem cells = associated=20 with breast cancer^[2] and acute myelogenous=20 leukemia^[7] have been isolated and implanted into = experimental=20 animals to form new tumors; their success following implantation=20 demonstrates that these cells are sufficient to generate = tumors.

The study of normal neural stem cells can be applied to the = brain tumor=20 cell population by establishing a link between normal neurogenesis = and=20 brain tumorigenesis.^[34,40] Brain tumors are = phenotypically=20 heterogeneous because they are composed of cells expressing both = markers=20 of differentiation and nondifferentiation. Populations of = proliferating=20 tumor stem cells differentiate into more mature cell types that=20 characterize the tumor.^{[11,38,41,43]}

In 2003, Singh and colleagues^[44] reported the=20 identification and purification of cancer stem cells from human = brain=20 tumors of different phenotypes. These cells possess a marked = capacity for=20 proliferation, self-renewal, and differentiation. The authors = demonstrated=20 that BTSCs have a high self-renewal capacity, which correlates = with=20 increased malignancy (such as that seen in a medulloblastoma = compared with=20 a low-grade glioma). Brain tumor stem cells were isolated by = selecting=20 cells expressing the neural stem cell surface antigen CD133. The = authors=20 demonstrated that CD133⁺ cells can differentiate in = vitro into=20 tumor cells that phenotypically resemble the patient's tumor. = These tumor=20 stem cells represent a fraction of cells constituting the tumor = that are=20 identified by their CD133 expression. The CD133 marker is a 120-kD = five-transmembrane cell-surface protein. Previously known as a=20 hematopoietic stem cell marker,^[31] CD133 was recently = found to=20 be a marker for normal human neural stem cells as = well.^[51]=20 Singh and colleagues were able to confirm the stem cell activity = of=20 CD133⁺ tumor cells by plating BTSCs at limiting = dilutions.=20 These investigators demonstrated that the self-renewal capacity of = tumor=20 cells was only present in the fraction of cells that expressed=20 CD133⁺. In addition, they found that the = CD133⁺=20 population displayed a proliferative capacity not present in the=20 CD133⁺ fraction.

Singh and colleagues^[44] suggest that an analysis of = BTSCs=20 may also provide a novel means for testing new therapeutic = strategies for=20 brain tumors that focus on the eradication of the tumor by = targeting=20 BTSCs. In fact, the presence of BTSCs in all the tumors examined = may have=20 important implications for understanding the underlying mechanisms = of=20 brain tumor dissemination.

To be considered a BTSC, a cell must be able to do the = following: 1)=20 generate clonally derived cells that form neurospheres; 2) = self-renew and=20 proliferate; 3) differentiate and express markers typical of brain = cells=20 (that is, those markers associated with astrocytes, = oligodendrocytes, and=20 neurons); and 4) generate tumors after in vivo transplantation in = animal=20 models that resemble the original tumor in donor patients.

Singh and associates^[45] reported that only=20 CD133⁺ cells are able to generate tumors after having = been=20 grafted into the mouse brain, and these tumors resemble the donor=20 patient's original tumor. An injection of only 100 = CD133⁺ cells=20 into the NOD-SCID mouse brain led to the growth of a tumor that = could be=20 serially transplanted and was histologically identical to the = tumor=20 harbored by the patient from whom these cells were derived. In = contrast,=20 an injection of 10⁵ CD133⁺ cells did not = generate=20 tumor growth (Fig. 1).^[45] The cellular heterogeneity = of brain=20 tumors suggests that only a small fraction of cancer cells are = able to=20 regenerate a tumor and that targeting these cells could be an = innovative=20 approach to eradicating the = tumor.^[1]

Figure 1.
Flow chart. Brain tumor stem cells were obtained after = dissociation=20 of a solid tumor specimen and were separated by cell sorting for = the=20 CD133 surface antigen. The CD133⁺ cells are able to=20 proliferate, self renew, and grow as spheres. The = CD133⁼⁹⁶=20 cells lack the ability of self-renewal and are not tumorigenic = in=20 mice.

The mechanism by which normal neural stem cells become = malignant has=20 not yet been elucidated. Hemmati and associates^[21]=20 demonstrated that normal and tumor-derived spheres express = bmi-1,=20 even after the withdrawal of mitogen from the culture medium. The = gene=20 bmi-1 has been demonstrated to be important for = self-renewal of=20 both leukemic^[27] and normal = hematopoietic^[35] stem=20 cells. Thus, the presence and persistent expression of = bmi-1 in=20 tumor cells could indicate a greater capacity of these cells to=20 self-renew.

Other groups have reported the isolation of tumor neural stem = cells=20 from human adult GBMs; these cells have the capacity to establish, = sustain, and expand these tumors even after serial in vivo = transplantation=20 experiments.^[16] These cells also possess a = self-renewal=20 capacity, an extensive proliferation capacity, and multipotency. A = gene=20 expression profile was obtained for genes involved in the = regulation of=20 stem cells including the self-renewal homeobox gene=20 Emx2^[17] and tumor-related genes such as = PTEN,=20 p21, p27, p53, and = MDR1.^[16]

We previously reported that in adults, the GBM contains a = subpopulation=20 of cells that can self-renew and differentiate into mature cell = types,=20 recapitulating the complexity of a primary GBM.^[54] = These cells=20 have all the properties of a stem cell, including the ability to = form=20 neurospheres, express neural stem cell markers, and, when = differentiated,=20 form multilineage progeny. In addition, they can self-renew and=20 proliferate to generate subspheres and different types of progeny. = Spheres=20 generated at clonal density can reform spheres after induction of=20 differentiation and have genetic aberrations typical of brain = tumors (Fig.=20 2). This is an important point that distinguishes cancer stem = cells from=20 normal stem cells. In fact, there has been some evidence = suggesting that=20 CD133, the marker used by other researchers to separate cancer = stem cells=20 from the main tumor cell population, may not be the appropriate = marker for=20 distinguishing normal stem cells from cancer stem=20 cells.^[54]

Figure 2.
Flow chart demonstrating the establishment of a culture of=20 neurospheres. After enzymatic dissociation of the tumor sample, = the=20 monocellular suspension was plated in serum-containing medium. = After 24=20 to 48 hours, the medium was switched with medium containing = epidermal=20 growth factor (EGF) and basic fibroblast growth factor (bFGF);=20 neurospheres began to appear 5 days later. The floating = neurospheres=20 were removed and subcultured at clonal = density.

More remains to be elucidated on the role these cancer stem = cells play=20 in brain tumors. In particular, it has not yet been determined = what cell=20 is the origin of brain tumors. Two main hypotheses have been = proposed: 1)=20 brain tumors arise from the transformation of a normal stem cell = or=20 progenitor cell; and alternatively, 2) brain tumors arise from the = dedifferentiation of a mature brain cell in response to genetic=20 alterations.^[34,40] Many lines of evidence have = suggested that=20 brain tumors arise from normal stem cells that undergo genetic = mutations;=20 the similarities between normal stem cells and cancer stem cells = regarding=20 their self-renewal, cloning capacity, and multilineage = differentiation=20 corroborate this hypothesis.^[34] In addition, the = heterogeneity=20 of brain tumors, which are composed of cells expressing more than = one=20 neural lineage phenotype, suggests an origin in a multipotent = cell. It has=20 also been reported that mutations in the mechanisms that regulate = stem=20 cell self-renewal (the sonic hedgehog pathway) can cause brain=20 tumors.^[10,34] In vivo experiments have been performed = to=20 corroborate this hypothesis. Holland and = collaborators^[23]=20 demonstrated that undifferentiated neural precursors are more = sensitive to=20 transformation than differentiated cells. A recent review = indicates the=20 possibility that cancer stem cells originate from cell fusion or a = horizontal gene-transfer process.^[5] The authors of = that review=20 report that cell fusion between somatic cells and stem cells might = create=20 genetic instability and cause cellular aneuploidy. Another = mechanism=20 suggested by the authors includes the possibility of a horizontal = gene=20 transfer. Mutations in somatic cells could lead to apoptosis and = nuclear=20 fragmentation: the fragmented DNA may be phagocytosed by other = somatic=20 cells, leading to nuclear reprogramming and the formation of new = tumor=20 cells.^[5] The possibility that end-terminal cells=20 dedifferentiate into stemlike cells that acquire stemlike = properties has=20 not been formally ruled out.

The cancer stem cell hypothesis of the origin of tumor cells = has been=20 extended recently to a wide variety of tumors, including prostate=20 cancer,^[9,29,49] bone sarcoma,^[18] liver=20 cancer,^[3] and melanoma.^[7] Collins and=20 colleagues^[9] reported the isolation and = characterization of a=20 cancer stem cell population from human prostate tumors. These = cells have a=20 significant capacity to self-renew and are able to regenerate the=20 phenotypically mixed population of nonclonogenic cells, which = produces=20 products of differentiated cells such as androgen receptor and = prostatic=20 acid phosphatase. The cancer stem cells derived from prostatic = tumors have=20 a CD44⁺α2 β1^hiCD133⁺ = phenotype, and these=20 proteins were used to isolate cells from a series of prostatic=20 tumors.^[9] In bone sarcomas, other investigators also=20 identified a small population of cells with self-renewal=20 properties.^[18] These cells were shown to be capable of = forming=20 "sarcospheres"=97suspended spherical, clonal colonies that grow = under=20 serum-free conditions. These cells express activated STAT3 and = some=20 markers of embryonic stem cells, such as Oct-3/4 and Nanog, at a = higher=20 level of expression than that found in adherent cell cultures. A=20 subpopulation of sarcoma cells expresses surface markers of = mesenchymal=20 cells such as Stro-1, CD105, and CD44. This finding extends the = cancer=20 stem cell hypothesis to cover tumors of mesenchymal=20 origin.^[18]

Fang, et al.,^[15] reported the presence of a = subpopulation=20 of cells in melanomas (melanoma spheroid cells) which grow as = nonadherent=20 spheres. These cells can differentiate into multilineage cells, = such as=20 melanocytes, adipocytes, osteocytes, and chondrocytes. These = multipotent=20 melanoma spheroid cells persist after serial cloning passages in = vitro and=20 transplantation in vivo, and are more tumorigenic than adherent = cells=20 after they have been implanted into mice. The stem cell population = was=20 enriched in CD20⁺ cells, a marker of hematopoietic stem = cells.^[15] Cells that stained positively for this = marker were=20 separated using FACS techniques. Even if both CD20⁺ and = CD20⁼⁹⁶ fractions display a similar proliferation = capacity,=20 CD20⁺ cells form bigger spheres. In one melanoma cell = line=20 (WM115), the fraction of cells that stained positively for CD20 = was able=20 to produce spheres, whereas the CD20⁼⁹⁶ cells remained = single=20 cells. In addition, when subjected to differentiation conditions, = only the=20 CD20⁺ fraction was able to generate multilineage cells. = This=20 indicates that melanoma spheroid stemlike cells enriched with=20 CD20⁺ cells possess stem cell properties (the capacity = for=20 proliferation, self-renewal, and multipotent=20 differentiation).^[15]

Recently, it has been shown that ependymomas contain a rare = population=20 of cancer stem cells that display the radial glia phenotype and = exhibit=20 gene expression patterns that recapitulate those of radial glia = cells in=20 the corresponding region of the central nervous = system.^[48]=20 These cells are required and sufficient to propagate a tumor in=20 transplanted mice. Ependymomas arising from different parts of the = central=20 nervous system contain different chromosomal abnormalities and = possess a=20 distinct gene expression signature. In particular, the = Notch cell=20 signaling pathway and the Hox family (known to regulate = stem cell=20 self-renewal) are up-regulated in supratentorial and spinal = ependymomas,=20 respectively.^[48]

The fact that it is possible to obtain cells with stem cell = properties=20 from different types of tumors corroborates the hypothesis that a = small=20 population of highly potent cells is present, in various degrees, = in=20 almost all tumors. Moreover, these cells are the "motor" that = allows the=20 tumor to grow. The presence of this population of cells can = explain the=20 recurrence of some brain tumors after chemotherapy, radiation = therapy, and=20 other current treatments. In fact, a few of these cells are = sufficient to=20 give rise to a new recurrent tumor. Cancer stem cells share = another=20 property with normal stem cells: they are able to migrate to = different=20 areas of the brain, thereby forming tumors far from the site of = injection.=20 In an experiment conducted by grafting these cells into mouse = brains,=20 tumors developed in a similar manner to those in humans. They = mainly grew=20 at the site of the injection, but were able to infiltrate the = surrounding=20 normal brain and produce tumors that extended into the = contralateral=20 hemisphere. Studies on the in vitro and in vivo behaviors of these = cells=20 may be useful to target not only differentiated tumor cells but = also=20 cancer stem cells.

The Role Of Multiple Drug Transporters

The cancer stem cell hypothesis provides novel insights into = strategies=20 to treat brain tumors. Recent data on the role of MDR transporters = in the=20 protection and self-renewal of normal and cancer stem cells may be = used to=20 identify pathways that can be specifically targeted.^[1] = Since=20 the development of anticancer drug therapies, it has been reported = that=20 some cancer cells demonstrate resistance to chemotherapy through = the=20 action of a group of membrane proteins that are able to extrude = cytotoxic=20 molecules and maintain an intracellular drug concentration below a = cell-killing threshold. The discovery of the MDR family of genes=20 (MDR1 and MDR2) allows us to understand more fully = the=20 mechanisms involved in the various responses of cells to=20 treatment.^[12]

Multiple drug transporters belong to a superfamily of ABC = proteins=20 located in the plasma membrane of different cellular=20 organelles.^[19] These proteins mediate the transport of = various=20 molecules across the membrane and present a wide variety of = chemical=20 structures. In some instances, ABC proteins facilitate the = transport of=20 inorganic ions. Others pump organic molecules such as lipids, bile = acids,=20 glutathione, and short peptides. Most of the ABC proteins use = adenosine=20 triphosphate hydrolysis energy for this transport activity, while = others=20 form specific membrane channels.^[19]

All of the ABC proteins share the ability to recognize and = translocate=20 a wide number of molecules, and additionally each transporter can = efflux=20 specific compounds. The first member of this family to be = discovered was=20 MDR-1 (also known as ABCB1), but two other members have been = identified as=20 playing a role in the multidrug resistance of the tumors: the = multidrug=20 resistance protein 1 (also known as ABCC1) and BCRP (also known as = ABCG2).^[14,19]

The BCRP is a half transporter consisting of only one=20 nucleotide-binding domain and one membrane-spanning domain. It may = function as a homodimer or homotetramer (Fig. 3). Overexpression = of BCRP=20 is associated with high levels of resistance to a variety of = anticancer=20 agents, including anthracyclines, mitoxantrone, and camptothecins = by=20 enhancing drug efflux.^[13,30] The expression of BCRP = has been=20 detected in a wide variety of hematological malignancies and solid = tumors,=20 underlying its role in the clinical drug resistance of=20 tumors.^[39] The BCRP is highly expressed in the = placenta, the=20 apical membrane of the epithelium, the small intestine, the liver=20 canalicular membrane, and the luminal surface of the endothelial = cells of=20 human brain micro-vessels.^[13] Zhou and=20 coworkers^[55] demonstrated that ABCG2 is expressed in = stem=20 cells from a wide variety of sources. They found that the = expression of=20 ABCG2 is strictly correlated with SPs of cells from different = sources=20 (murine bone marrow, skeletal muscle, and cultured embryonic stem = cells).=20 The SP is defined as the population of cells able to efflux the = Hoechst=20 33342 dye; it is rich in highly repopulating cells.^[25] = The SP=20 was first demonstrated in hematopoietic stem cells of different = species=20 and in murine skeletal muscle and neurospheres derived from = embryonic=20 mice.^[8] An SP was also found in various cancer cell = lines=20 including the C6 rat glioma, MCF7 breast cancer, B104 = neuroblastoma, and=20 HeLa adenocarcinoma cell lines.^[26] In C6 cells, the SP = separated by FACS could generate both SP and non-SP cells in = culture and=20 form metastatic tumors in nude mice, but the non-SP cells could = not.=20 Therefore, the C6 SP contains cells with the characteristics of = both=20 normal stem cells and cancer stem cells. The persistence of SP = cells in=20 cancer cell lines cultured in serum-containing medium for decades = suggests=20 that the SP may be a general source of cancer stem=20 cells.^[26,42]

Figure 3.
A membrane model of BCRP. The breast cancer=13resistant = protein has a=20 nucleotide domain and a membrane domain with a six-transmembrane = a helix=20 and two or three N-glycosylation sites in the = extracellular=20 loops. The ABC domain is also indicated.

The correlation between SP cells and ABCG2-expressing cells = appears to=20 be controversial. Patrawala and colleagues^[37] reported = that SP=20 cells possess some intrinsic stem cell characteristics: they are = able to=20 generate both SP and non-SP cells, can be transplanted and = generate=20 tumors, express "stemness" genes such as Notch1 and=20 β-catenin, and demonstrate increased ABCG2 expression = compared with=20 non-SP cells. Nevertheless, although SP cells in vivo are more = tumorigenic=20 than non-SP cells, ABCG2⁺ purified cells display the = same=20 tumorigenicity as ABCG2⁼⁹⁶ cells. The authors also = demonstrated=20 that ABCG2⁺ cells form larger and more numerous clones = in=20 long-term clonal analyses and express several "stemness" = genes.

The SP cells also have been shown to be very heterogeneous. In = fact,=20 the SP detected in cancer stem cells contains a wide variety of = cells,=20 some of which express ABCG2. The presence of this subpopulation of = cells=20 can explain the increased expression of ABCG2 in the SP, but the = higher=20 tumorigenicity associated with the SP may also be conferred by the = combined effects of other subpopulations of cells present in the = SP. It=20 could be possible that other adenosine triphosphate=96binding = proteins, such=20 as MDR-1, MRP-1, and ABCA2, expressed in SP cells may play a role = in=20 conferring chemoresistance to this population of=20 cells.^[37]

How To Target Cancer Stem Cells

The presence and high expression of ABC transporter proteins in = cancer=20 stem cells may explain why common treatments, particularly = chemotherapy,=20 are not sufficient to kill the tumor. In fact, despite the = continuous=20 development of new chemotherapeutic agents, brain tumors remain = resistant=20 to therapy. Currently, the life span in patients harboring these = lesions=20 reaches 24 months in approximately 5% of cases. Therefore, novel = therapies=20 must be developed. In particular, immunotherapy with vaccination = of DCs=20 opens a new way to target infiltrating brain tumors. Dendritic = cells are=20 antigen-presenting cells that stimulate the na=EFve immune system = and play a=20 role in maintaining self-tolerance.^[47] Through the = expression=20 of high levels of MHC class I and class II molecules, adhesion and = costimulatory molecules (CD40, CD45, CD80, and CD86), and = stimulatory=20 cytokines (IL-12, IL-15, and IL-18), DCs lead to an efficient = priming of=20 cytotoxic T cells and CD4⁺ T helper cells, thus = inducing a=20 specific and therapeutic immune response.^[46] In fact,=20 cytotoxic T cells are believed to be the main cells involved in = tumor=20 rejection, because these cells can recognize antigens loaded onto = MHC=20 class I molecules, whereas CD4⁺ T cells are activated = after MHC=20 class II=96restricted presentation of exogenous antigens by DCs. = It has also=20 been reported that DCs can prime a CD8⁺ T = cell=96mediated=20 response against exogenous tumor antigens.^[22] = Dendritic cells=20 can be generated in culture both from peripheral blood = CD14⁺=20 monocytic precursors and from proliferating CD34⁺ = progenitor=20 cells by the addition of cytokines, including IL-4, = granulocyte=96macrophage=20 colony-stimulating factor, tumor necrosis factor=96α, = c-kit, and the=20 Flt-3 ligand.^[46]

In vaccination studies, DCs in culture were loaded with = different=20 tumor-derived substrates such as specific tumor-associated = peptides, tumor=20 RNA and complementary DNA, tumor cell lysate, or apoptotic tumor=20 cells.^[46]

Many animal studies and clinical trials have shown that DC=20 immunotherapy improves and prolongs the survival of tumor-bearing=20 experimental animals or patients.^{[4,20,28,32,33,52]} We=20 performed a Phase I clinical trial in which we administered = DC-based=20 vaccines to patients with gliomas (Fig. 4).^[53] We = pulsed DCs=20 with autologous tumor lysate and gave the vaccine to nine patients = with=20 recurrent GBM, three patients with recurrent anaplastic = astrocytoma, and=20 two patients with newly diagnosed tumors (one GBM and one = anaplastic=20 astrocytoma). The results we obtained showed that this tumor = lysate=96pulsed=20 DC vaccine is safe, with no evidence of side effect or autoimmune=20 toxicity, and can prolong the survival of patients (133 weeks = compared=20 with 30 weeks in control patients).

Figure 4.
A scheme of DC vaccination. After excision of the brain = tumor, BTSCs=20 were isolated and cultured. Dendritic cells were prepared from=20 peripheral blood after monocyte isolation, pulsed with the BTSC = lysate,=20 and injected into the patient. GM-CSF =3D = granulocyte=13macrophage=20 colony-stimulating factor; IL =3D = interleukin.

Future vaccination therapies may be directed toward cancer stem = cell=20 lysates to improve and ameliorate the antigen-presenting DC = response. In=20 this case, the activated immune system can directly kill tumor = stem cells=20 as well as tumor daughter cells. Preclinical studies are now = necessary to=20 test this hypothesis.

Conclusions

Although additional studies are necessary to provide us with a = better=20 understanding of the biology and behavior of tumor stemlike cells, = it is=20 evident that these cells represent a new target for future tumor=20 therapies.

References

Al-Hajj M, Becker MW, Wicha M, et al: Therapeutic = implications=20 of cancer stem cells. Curr Opin Genet Dev 14:43=9647, 2004
Al-Hajj M, Wicha MS, Benito-Hernandez A, et al: = Prospective=20 identification of tumorigenic breast cancer cells. Proc = Natl Acad=20 Sci U S A 100:3983=963988, 2003
Alison MR, Lovell MJ: Liver cancer: the role of stem = cells.=20 Cell Prolif 38:407=96421, 2005
Ashley DM, Faiola B, Nair S, et al: Bone = marrow-generated=20 dendritic cells pulsed with tumor extracts or tumor RNA = induces=20 antitumor immunity against central nervous system tumors. = J Exp=20 Med 186:1177=961182, 1997
Bjerkvig R, Tysnes BB, Aboody KS, et al: The origin of = the=20 cancer stem cell: current controversies and new insights. = Nat Rev=20 Cancer 5:899=96904, 2005
Bonnet D: Cancer stem cells: lessons from leukaemia. = Cell=20 Prolif 38:357=96361, 2005
Bonnet D, Dick JE: Human acute myeloid leukemia is = organized=20 as a hierarchy that originates from a primitive = hematopoietic=20 cell. Nat Med 3:730=96737, 1997
Bunting KD: ABC transporters as phenotypic markers and = functional regulators of stem cells. Stem Cells = 20:11=9620, 2002
Collins AT, Berry PA, Hyde C, et al: Prospective=20 identification of tumorigenic prostate cancer stem cells. = Cancer=20 Res 65: 10946=9610951, 2005
Dahmane N, S=E1nchez P, Gitton Y, et al: The sonic = hedgehog-gli=20 pathway regulates dorsal brain growth and tumorigenesis.=20 Development 128:5201=965212, 2001
Dirks PB: Brain tumor stem cells. Biol Blood Marrow = Transplant=20 11 (Suppl 2):12=9613, 2005
Donnenberg VS, Donnenberg AD: Multiple drug resistance = in=20 cancer revisited: the cancer stem cell hypothesis. J Clin=20 Pharmacol 45:872=96877, 2005
Doyle LA, Ross DD: Multidrug resistance mediated by = the breast=20 cancer resistance protein BCRP (ABCG2). Oncogene = 22:7340=967358,=20 2003
Ejendal KFK, Hrycyna CA: Multidrug resistance and = cancer: the=20 role of the human ABC Transporter ABCG2. Curr Protein Pept = Sci=20 3:503=96511, 2002
Fang D, Nguyen TK, Leishear K, et al: A tumorigenic=20 subpopulation with stem cell properties in melanomas. = Cancer Res=20 65:9328=969337, 2005
Galli R, Binda E, Orfanelli U, et al: Isolation and=20 characterization of tumorigenic, stem-like neural = precursors from=20 human glioblastoma. Cancer Res 64:7011=967021, 2004
Galli R, Fiocco R, De Filippis L, et al: Emx2 = regulates the=20 proliferation of stem cells of the adult mammalian central = nervous=20 system. Development 129:1633=961644, 2002
Gibbs CP, Kukekov VG, Reith JD, et al: Stem-like cells = in bone=20 sarcomas: implications for tumorigenesis. Neoplasia = 7:967=96976,=20 2005
Gottesman MM, Fojo T, Bates SE: Multidrug resistance = in=20 cancer: role of ATP-dependent transporters. Nat Rev Cancer = 2:48=9658, 2002
Heimberger AB, Crotty LE, Archer GE, et al: Bone = marrowderived=20 dendritic cells pulsed with tumor homogenate induce = immunity=20 against syngeneic intracerebral glioma. J Neuroimmunol = 103:16=9625,=20 2000
Hemmati HD, Nakano I, Lazareff JA, et al: Cancerous = stem cells=20 can arise from pediatric brain tumors. Proc Natl Acad Sci = U S A=20 100:15178=9615183, 2003
Hoffmann TK, Meidenbauer N, Dworacki G, et al: = Generation of=20 tumor-specific T lymphocytes by cross-priming with human = dendritic=20 cells ingesting apoptotic tumor cells. Cancer Res 60: = 3542=963549,=20 2000
Holland EC, Celestino J, Dai C, et al: Combined = activation of=20 Ras and Akt in neural progenitors induces glioblastoma = formation=20 in mice. Nat Genet 25:55=9657, 2000
Hope KJ, Jin L, Dick JE: Acute myeloid leukemia = originates=20 from a hierarchy of leukemic stem cell classes that differ = in=20 selfrenewal capacity. Nat Immunol 5:738=96743, 2004
Kim M, Morshead CM: Distinct populations of forebrain = neural=20 stem and progenitor cells can be isolated using = side-population=20 analysis. J Neurosci 23:10703=9610709, 2003
Kondo T, Setoguchi T, Taga T: Persistence of a small=20 subpopulation of cancer stem-like cells in the C6 glioma = cell=20 line. Proc Natl Acad Sci U S A 101:781=96786, 2004
Lessard J, Sauvageau G: Bmi-1 determines the = proliferative=20 capacity of normal and leukaemic stem cells. Nature = 423:255=96260,=20 2003
Liau LM, Black KL, Prins RM, et al: Treatment of = intracranial=20 gliomas with bone marrow-derived dendritic cells pulsed = with tumor=20 antigens. J Neurosurg 90:1115=961124, 1999
Maitland NJ, Collins A: A tumor stem cell hypothesis = for the=20 origins of prostate cancer. BJU Int 96:1219=961223, 2005
Mao Q, Unadkat JD: Role of the breast cancer = resistance=20 protein (ABCG2) in drug transport. AAPS J 7:E118=96E133, = 2005
Miraglia S, Godfrey W, Yin AH, et al: A novel=20 five-transmembrane hematopoietic stem cell antigen: = isolation,=20 characterization, and molecular cloning. Blood = 90:5013=965021, 1997
Ni HT, Spellman SR, Jean WC, et al: Immunization with=20 dendritic cells pulsed with tumor extract increases = survival of=20 mice bearing intracranial gliomas. J Neurooncol 51:1=969, = 2001
Okada H, Tahara H, Shurin MR, et al: Bone = marrow-derived=20 dendritic cells pulsed with a tumor-specific peptide = elicit=20 effective anti-tumor immunity against intracranial = neoplasms. Int=20 J Cancer 78:196=96201, 1998
Pardal R, Clarke MF, Morrison SJ: Applying the = principles of=20 stem-cell biology to cancer. Nat Rev Cancer 3:895=96902, = 2003
Park IK, Qian D, Kiel M, et al: Bmi-1 is required for=20 maintenance of adult self-renewing haematopoietic stem = cells.=20 Nature 423:302=96305, 2003
Passegu=E9 E, Jamieson CHM, Ailles LE, et al: Normal = and=20 leukemic hematopoiesis: are leukemias a stem cell disorder = or a=20 reacquisition of stem cell characteristics? Proc Natl Acad = Sci U S=20 A 100:11842=9611849, 2003
Patrawala L, Calhoun T, Schneider-Broussard R, et al: = Side=20 population is enriched in tumorigenic, stem-like cancer = cells,=20 whereas ABCG21 and ABCG22 cancer cells are similarly = tumorigenic.=20 Cancer Res 65:6207=966219, 2005
Pilkington GJ: Cancer stem cells in the mammalian = central=20 nervous system. Cell Prolif 38:423=96433, 2005
Plasschaert SLA, van der Kolk DM, de Bont ESJM, et al: = The=20 role of breast cancer resistance protein in acute = lymphoblastic=20 leukemia. Clin Cancer Res 9:5171=965177, 2003
Reya T, Morrison SJ, Clarke MF, et al: Stem cells, = cancer, and=20 cancer stem cells. Nature 414:105=96111, 2001
Sanai N, Alvarez-Buylla A, Berger MS: Neural stem = cells and=20 the origin of gliomas. N Engl J Med 353:811=96822, 2005
Setoguchi T, Taga T, Kondo T: Cancer stem cells = persist in=20 many cancer cell lines. Cell Cycle 3:414=96415, 2004
Singh SK, Clarke ID, Hide T, et al: Cancer stem cells = in=20 nervous system tumors. Oncogene 23:7267=967273, 2004
Singh SK, Clarke ID, Terasaki M, et al: Identification = of a=20 cancer stem cell in human brain tumors. Cancer Res = 63:5821=965828,=20 2003
Singh SK, Hawkins C, Clarke ID, et al: Identification = of human=20 brain tumor initiating cells. Nature 432:396=96401, 2004
Soling A, Rainov NG: Dendritic cell therapy of primary = brain=20 tumors. Mol Med 7:659=96667, 2001
Steinman RM, Turley S, Mellman I, et al: The induction = of=20 tolerance by dendritic cells that have captured apoptotic = cells. J=20 Exp Med 191:411=96416, 2000
Taylor MD, Poppleton H, Fuller C, et al: Radial glia = cells are=20 candidate stem cells of ependymoma. Cancer Cell = 8:323=96335, 2005
Tokar EJ, Ancrile BB, Cunha GR, et al: Stem/progenitor = and=20 intermediate cell types and the origin of human prostate = cancer.=20 Differentiation 73:463=96473, 2005
Tsai RYL: A molecular view of stem cell and cancer = cell=20 selfrenewal. Int J Biochem Cell Biol 36:684=96694, 2004
Uchida N, Buck DW, He D, et al: Direct isolation of = human=20 central nervous system stem cells. Proc Natl Acad Sci U S = A=20 97:14720=9614725, 2000
Yamanaka R, Zullo SA, Tanaka R, et al: Enhancement of=20 antitumor immune response in glioma models in mice by = genetically=20 modified dendritic cells pulsed with Semliki Forest = virus-mediated=20 complementary DNA. J Neurosurg 94:474=96481, 2001
Yu JS, Liu G, Ying H, et al: Vaccination with tumor=20 lysatepulsed dendritic cells elicits antigen-specific, = cytotoxic=20 T-cells in patients with malignant glioma. Cancer Res=20 64:4973=964979, 2004
Yuan X, Curtin J, Xiong Y, et al: Isolation of cancer = stem=20 cells from adult glioblastoma multiforme. Oncogene = 23:9392=969400,=20 2004
Zhou S, Schuetz JD, Bunting KD, et al: The ABC = transporter=20 Bcrp1/ABCG2 is expressed in a wide variety of stem cells = and is a=20 molecular determinant of the side-population phenotype. = Nat Med=20 7:1028=961034, = 2001

Abbreviation = Notes

ABC =3D adenosine triphosphate=96binding cassette; BCRP =3D = breast cancer=20 resistance protein; BTSC =3D brain tumor stem cell; DC =3D = dendritic cell;=20 FACS =3D fluorescence-activated cell sorting; GBM =3D glioblastoma = multiforme;=20 MDR =3D multiple drug resistance; MHC =3D major histocompatibility = complex; SP=20 =3D side population

Reprint Address

John S. Yu, M.D., Maxine Dunitz Neurosurgical Institute, Suite = 800E=20 East, 8631 West Third Street, Los Angeles, California 90048. = email: John.Yu@cshs.org

Patrizia Tunici, Ph.D., Dwain Irvin, Ph.D., = M.P.H.,=20 Gentao Liu, Ph.D., Xiangpeng Yuan, Ph.D., Zeng Zhaohui, Ph.D., = Hiushan Ng,=20 M.S., and John S. Yu, M.D., Maxine Dunitz Neurosurgical = Institute, Los Angeles, California

Financial Disclosure: Dr. Tunici is supported = by ARIN=20 (Associazione Italiana per le Ricerche Neurologiche, "Il Fondo di=20 Gio").