glioblastoma stem cells are regulated by interleukin-8 signaling...

Tumor and Stem Cell Biology

Glioblastoma Stem Cells Are Regulated by Interleukin-8Signaling in a Tumoral Perivascular Niche

David W. Infanger1, YouJin Cho1, Brina S. Lopez4, Sunish Mohanan2, S. Chris Liu1, Demirkan Gursel5,John A. Boockvar5, and Claudia Fischbach1,3

AbstractGlioblastoma multiforme contains a subpopulation of cancer stem–like cells (CSC) believed to underlie

tumorigenesis and therapeutic resistance. Recent studies have localized CSCs in this disease adjacent toendothelial cells (EC) in what has been termed a perivascular niche, spurring investigation into the role ofEC–CSC interactions in glioblastoma multiforme pathobiology. However, these studies have been limited by alack of in vitromodels of three-dimensional disease that can recapitulate the relevant conditions of the niche. Inthis study, we engineered a scaffold-based culture system enabling brain endothelial cells to form vascularnetworks. Using this system, we showed that vascular assembly induces CSCmaintenance and growth in vitro andaccelerates tumor growth in vivo through paracrine interleukin (IL)-8 signaling. Relative to conventionalmonolayers, endothelial cells cultured in this three-dimensional system not only secreted enhanced levels ofIL-8 but also induced CSCs to upregulate the IL-8 cognate receptors CXCR1 and CXCR2, which collectivelyenhanced CSC migration, growth, and stemness properties. CXCR2 silencing in CSCs abolished the tumor-promoting effects of endothelial cells in vivo, confirming a critical role for this signaling pathway in GMBpathogenesis. Together, our results reveal synergistic interactions between endothelial cells and CSCs thatpromote the malignant properties of CSCs in an IL-8–dependent manner. Furthermore, our findings underscorethe relevance of tissue-engineered cell culture platforms to fully analyze signaling mechanisms in the tumormicroenvironment. Cancer Res; 73(23); 1–11. �2013 AACR.

IntroductionGlioblastoma multiforme, a high-grade (IV) astrocytoma,

remains the most common and lethal primary brain tumorin adults with a median survival of 12 to 15 months (1).Clinical management of glioblastoma is largely palliative,due to poor efficacy of conventional therapeutics and reflex-ive secondary tumor formation following resection. Theidentification of glioblastoma–associated cancer stem cells(CSC) has fueled research into the contributing role ofthis cell type in glioblastoma pathogenesis and therapeuticresilience (2). Glioblastoma CSCs are capable of self-renewaland multilineage differentiation, enable tumorigenesis uponintracranial implantation within immunocompromised

rodents (3), and exhibit elevated therapeutic resistancerelative to bulk glioma cells (4). Therefore, elucidation ofthe molecular mechanisms underlying CSC maintenanceand proliferation in the tumor microenvironment offers newfocus toward the development of improved anti-glioblasto-ma therapies.

Cell–microenvironment interactions modulate glioblasto-ma pathogenesis, whereby perivascular localities are particu-larly important as they support the malignant behavior ofCSCs (5, 6). Specifically, association with endothelial cells (EC)or capillary structures supports maintenance of Nestinþ andCD133þ CSCs by sustaining their undifferentiated state andproliferation (5). Likewise, paracrine signaling by CSCsenhances the apoptotic threshold of endothelial cells (7),thereby preserving the perivascular microenvironment neces-sary for downstream processes involved in glioblastomapathology. Nevertheless, despite accumulating evidence con-firming the importance of CSC–EC signaling in glioblastomapathogenesis, the precise molecular mechanisms involved inCSC localization to—and function within—the perivascularniche remain poorly defined.

Interleukin (IL)-8 has received significant attention as apromigratory and proangiogenic stimulus in multiple can-cers but may regulate glioblastoma CSC functions as well(8). The effects of the �8-kDa chemokine are mediated viabinding to either of 2 related G-protein–coupled receptors,CXCR1 and CXCR2, whose expression varies dramatically bycell type and throughout pathogenesis (9). While initially

Authors' Affiliations: Departments of 1Biomedical Engineering and 2Com-parative Biomedical Sciences, 3Kavli Institute at Cornell for NanoscaleScience, Cornell University, Ithaca, New York; 4Department of VeterinaryMedicine, Colorado State University, Fort Collins, Colorado; and 5Labo-ratory for Translational Stem Cell Research, Weill Cornell Brain TumorCenter, Department ofNeurological Surgery,Weill CornellMedical College,New York

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Corresponding Author: Claudia Fischbach, Department of BiomedicalEngineering, Cornell University 157 Weill Hall, Ithaca, NY 14853. Phone:607-255-4547; Fax: 607-255-7330; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-13-1355

�2013 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org OF1

Research. on June 2, 2018. © 2013 American Association for Cancercancerres.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2013; DOI: 10.1158/0008-5472.CAN-13-1355

http://cancerres.aacrjournals.org/

identified as a monocyte-secreted chemotactic factor forneutro/basophils and T lymphocytes, IL-8 can similarlyfacilitate invasion of bulk glioma cells and has been asso-ciated with increased tumor grade in astrocytic neoplasms(10, 11). In addition to its direct proangiogenic effects, IL-8can promote glioblastoma tumor vascularization indirectlyby elevating the apoptotic threshold of endothelial cells aswell as promoting expression of matrix-remodeling enzymesnecessary for endothelial sprouting (9, 12). Although it hasbeen recently revealed that autocrine IL-8 signaling con-tributes to CSC self-renewal within other tumors of thebreast and liver (13, 14), its impacts on the behavior ofglioblastoma CSCs, signaling with endothelial cells of theperivascular niche, and participation in glioblastoma tumorgrowth have yet to be elucidated.

IL-8 expression and secretion are heavily influenced by3-dimensional (3D) cell–microenvironment interactions (15,16), underscoring the importance of physiologically relevant,3D culture models to study IL-8 signaling as it relates tohuman CSCs and endothelial cells throughout glioblastomapathogenesis. We have previously shown that tissue-engi-neered 3D culture systems mimic in vivo–like IL-8 expressionby recapitulating appropriate microenvironmental condi-tions including cell–cell and cell–extracellular matrix (ECM)interactions (15, 16). Here, we used porous, polymeric scaf-folds as platforms to investigate the role of paracrine endo-thelial cell signaling on patient-derived glioblastoma CSCpathogenesis in vitro and in vivo. Our findings indicate thatIL-8 serves as a critical mediator supporting CSC growthand migration toward endothelial cells, which may partiallyexplain their perivascular colocalization in the glioblastomatumor microenvironment, thereby offering promise as a noveltherapeutic target for the clinical management of the disease.Furthermore, these studies underscore the importance ofappropriate culture systems in studying tumor–stroma inter-actions and support that tissue-engineered models are suit-able to evaluate microenvironmentally regulated paracrinesignaling in vitro and in vivo.

Materials and MethodsCell culture

Immortalized human brain endothelial cells (hCMEC, pro-vided by Dr. Babette Weksler, Weill Cornell Medical College,New York, NY; ref. 17) were cultured on collagen-coated(1%) flasks in Clontecs EGM-2 media (Lonza). Patient-derivedglioblastoma multiforme stem cells (CSCs, provided by Dr.John Boockvar, Weill Cornell Medical College) were isolatedfrom glioblastoma tumor specimens (WHO grade IV) as de-scribed previously (18). Briefly, glioblastoma tissue was digest-ed in PBS containing papain (Worthington) and DNase 1(Sigma-Aldrich), then titrated via pipette, and filtered througha 70-mm sterile cell filter (BD Biosciences). Cells were resus-pended in stem cell media containing 1:1 Dulbecco's ModifiedEagle'sMedium:F12 (Gibco) plus basic fibroblast growth factorand EGF (each 20 ng/mL; Invitrogen) and 1� antibiotic/antimycotic (Gibco). An isolated cell was clonally expandedin nonadherent culture flasks and media changed every 48 to

72 hours, resulting in neurosphere aggregates. Two glioblas-toma multiforme CSC cell populations, labeled CSC 248(EGFRþ/PTENþ) and CSC 974 (EGFRþ/PTEN�), were used.Expression of stem cell markers Nestin, Sox2, Oct4, Musashi-1,GFAP, and Bim-1 and absence of Olig-1, Tuj-1, and NeuN wereverified via immunocytochemistry and tumorigenicity con-firmed via implantation into athymic mice (18).

Polymeric scaffold fabrication and cell seedingPorous, polymeric scaffolds were fabricated using a gas-

foaming, particulate leaching method (19). Briefly, poly(lac-tide-co-glycolide) (PLG) particles (ground and sieved to �250mm diameter; Lakeshore Biomaterials), PLG microspheres(formed via double emulsion, �5–50 mm diameter), and sodi-um chloride (sieved to�250–400mmdiameter; J.T. Baker)weremixed and pressed into matrices (8.5 mm diameter, 1 mmthick) at room temperature (Carver Press, Fred S. Carver).Polymer foaming was achieved by exposing matrices to high-pressure (800 psi) CO2 and releasing gas quickly using apressure vessel (Parr Instruments). Subsequently, sodiumchloride porogen was leached out with de-ionized water.Scaffolds were sanitized in 70% ethanol and washed in sterilePBS. 1.5E6 hCMECs, 1.0E5 CSCs, or both cell types (aforemen-tioned cell densities) were suspended in 30 mL of 1:1 (media:Matrigel; BD Biosciences) and slowly pipetted onto the scaf-fold. Following 30 minutes of incubation at 37�C (to allow cellpermeation), scaffolds were placed on ice until implantation.Alternatively, media were to scaffolds under dynamic condi-tions using an orbital shaker for in vitro experiments.

Animal studiesAnimal studies were conducted according to approved

protocols by the Cornell University Animal Care and UseCommittee. Male, 6- to 8-week-old, CB17 SCID mice (CharlesRiver Labs) were anesthetized and incisions made to thedorsal infrascapular skin. A subcutaneous pocket was cre-ated, irrigated with sterile PBS, cell-seeded PLG scaffolds(described above) inserted, and then sutured with 5-0Ethilon (Ethicon). Studies investigating in vivo polymer de-gradation used blank sanitized scaffolds. High-resolutionultrasound imaging was conducted weekly using the VEVO770 Imaging system and RMV 706 single-element transducer(Visualsonics). Mice were anesthetized (1.5% isoflurane) andimplantation site hair removed by chemical debridement(Nair, Church & Dwight Co.). Mice were placed prone on aheated stage and scaffolds imaged with semiautomated 3D,B-mode imaging at 40 MHz frequency. To calculate tumorvolume, cross-sectional areas of PLG scaffold þ tumor weredetermined and then integrated to measure total volume,using VEVO software (v. 3.0.0).

Immunostaining and histologyCSC neurospheres cultured in nonadherent flasks were

collected by centrifugation and embedded in OCT (Tissue-Tek) in minimal PBS following washing, fixation with 4%paraformaldehyde (PFA), and incubation in 20% sucrose/PBS.After cryosectioning (14 mm), immunostaining was conductedon Triton-X (VWR, 0.5%) permeabilized cells with antibodies

Infanger et al.

Cancer Res; 73(23) December 1, 2013 Cancer ResearchOF2




against human Sox-2 (Sigma), Oct-4 (Millipore), Nestin (Milli-pore), or control rabbit/mouse IgG (Invitrogen) at 1:200 dilu-tion. Secondary antibodies (1:500, anti-rabbit Alexafluor 488 oranti-mouse Alexafluor 546; Invitrogen) were diluted in PBScontaining 40,6-diamidino-2-phenylindole (DAPI; 1:5,000) fornuclear counterstain; imaging was conducted on a Zeiss LSM710 confocal microscope.For in vivo studies, tumors were removed and fixed over-

night in 4% PFA, then bifurcated and half submitted forparaffin sectioning (4 mm) and subsequent hematoxylin andeosin (H&E) staining; remaining half was immersed in 20%sucrose/PBS overnight, embedded in optimum cutting tem-perature (OCT), and cryosectioned (14 mm). Immunostainingwas conducted as above to detect stem cell marker levels; inaddition, species-specific ECmarker CD31 was probed (mouseanti-human, Invitrogen; rat anti-mouse, BD Pharmingen) at1:200 dilution, followed by secondary Alexafluor 546 (goat anti-mouse) or Alexafluor 647 (goat anti-rat) antibody at 1:500 (bothfrom Invitrogen). Sections were counterstained with DAPI(1:5,000) and imaged on a Zeiss LSM 710 confocal microscope.

Conditioned media preparationhCMEC-seeded PLG scaffolds were cultured for 3 days, after

which EGM-2 media were removed, scaffolds washed in sterilePBS, and basal EBM-2 media (sans growth supplements, with0.25% FBS and 0.1% penicillin/streptomycin) added. Mediawere collected at 24 hours and IL-8 ELISA (R&D Systems)conducted per manufacturer's instructions. Subsequently,media were concentrated 10� at 4�C using Amicon Ultrafree15 centrifugal filter units (3,000 MWCO; Millipore). Concen-trated media (termed "3D conditioned endothelial cell medi-um") was normalized to DNA content, as determined byfluorimetric DNA assay (Quantifluor Assay; Promega) of scaf-fold lysates in Caron's buffer. To generate 2D conditionedendothelial cell medium, hCMECs were cultured as subcon-fluent monolayers and media collected, concentrated, andnormalized to like DNA concentrations as above describedfor 3D conditioned media. Basal control medium was gener-ated by incubating basal EBM-2media for 24 hours at 37�C andconcentrating 10-fold as above described. Before use, condi-tioned media were diluted to 2� final concentration in stemcell medium and supplemented to CSC cultures for 3 days ofpreconditioning before subsequent analyses.Conditioned CSC medium was created by culturing CSCs in

nonadherent flasks until neurospheres of�100 mm formed, atwhich point media were replaced with fresh stem cell mediumlacking antibiotics. Media was then concentrated and normal-ized to DNA content as above described.

Transwell migration assayTranswell inserts (8 mm pore size; Falcon HTS FluroBlok

Inserts) were coated with 1% collagen I (BD Biosciences),washed, and placed in 24-well plates. Dissociated CSCs(10,000; with or without prior preconditioning in 3D condi-tioned endothelial cell media) were added to each Transwell in300 mL stem cell media and migration assessed followingaddition of recombinant human IL-8 or IgG (50 or 5 ng/mL;Genescript) to stem cell media in the lower compartment.

Alternatively, migration toward endothelial cell–secreted fac-tors was determined by adding 300 mL medium composed ofequal parts stem cell medium and either 10� 3D endothelialcell medium or 10� basal EC medium. To confirm the role ofIL-8 in CSC chemotaxis, antagonist antibodies to human IL-8(7.5 mg/mL), CXCR1 (2 mg/mL), or CXCR2 (2 mg/mL, all fromR&D Systems) were added before the start of the experiment(anti-IL-8 added to medium; anti-CXCR1/2 added to cells þmedium for 1 hour at room temperature before experiment).After 16 hours, Transwells were fixed in 4% PFA and DAPIstained (1:5,000). Transwell underside images were collected at10� on a Zeiss Observer Z.1 microscope fitted with an Axio-Cam MRN camera. For each of triplicate experiments, theaverage number of migrated cells from 7 random images perwell from 4 wells was calculated using ImageJ software (NIH).

Cell growth in response to IL-8 stimulationFreshly dissociated CSCs were placed in nonadherent plates

(30,000/12-well) in stem cell medium. IL-8 supplementation(50 ng/mL; Genescript) alone or with antibody antagonists toIL-8, CXCR1, or CXCR2 (dosages and pre-incubation as abovefor Transwell studies) was provided at the start of the exper-iment and at 24 hours. After 48 hours, cells were collected bybrief centrifugation and DNA assay conducted on Caron'sbuffer–treated cell lysates as above described. Data are repre-sented as normalized values relative to CSC cultures withoutIL-8 treatment.

Real-time reverse transcription PCR and short hairpinRNA–mediated CXCR2 gene knockdown

Following 3 days of preconditioning in 3D conditioned orbasal media, total RNA from CSC neurospheres was harvestedwith TRIzol (Invitrogen) according to manufacturer's instruc-tions and 1 mg reverse-transcribed to cDNA (qScript cDNAsupermix; Quanta BioSciences) using random hexamer andoligo(dt) primers. Real-time reverse transcription (RT)-PCRwas carried out on 25 ng template using SYBR green detection(Quanta Biosciences) and an Applied Biosystems 7500 System.Primer sequences (300 nmol/L; IDT Technologies) were asfollows: human CXCR1 (forward: 50tcaagtgccctctagctgtt30,reverse: 50gggctgtaatcttcatctgc30); human CXCR2 (forward:50agaagtttcgccatggactcctca30, reverse: 50agtggaagtgtgccctgaag-aaga30) and b-actin loading control (forward: 50aatgtggccga-ggactttgattgc30, reverse: 50aggatggcaagggacttcctgtaa30). Rela-tive quantification was conducted using the DDCt method (20).

To establish long-term knockdown of CXCR2 expression,lentiviruses encoding short hairpin RNA (shRNA) sequencestargeted against human CXCR2 or random shRNA sequence(MISSION shRNA lentiviral transduction particles; Sigma-Aldrich) were transduced at 5 multiplicity of infection to1E4 glioblastoma CSCs in a 12-well dish overnight, rinsed withsterile PBS, and transferred to stem cell medium. CXCR2knockdown of 70% (vs. control transduced) was confirmed byreal-time RT-PCR when 3 CXCR2-specific shRNA lentiviruseswere used concurrently (TRCN0000009136, TRCN0000009137,and TRCN0000009138; Sigma-Aldrich). In addition, CXCR2knockdown was confirmed at the protein level using Westernblot analysis. CSCs previously transduced with siRNA-control

Interleukin-8 Signaling on Glioblastoma Tumor Stem Cells

www.aacrjournals.org Cancer Res; 73(23) December 1, 2013 OF3




or siRNA-expressing lentiviruses were centrifuged and har-vested at 4�C in radioimmunoprecipitation buffer containingprotease and phosphatase inhibitors (Roche). Protein concen-trations were assessed by bicinchoninic acid assay (Pierce),50 mg separated on a pre-cast 10% PAGE (Bio-Rad), and thentransferred to a polyvinylidene difluoride membrane (Bio-Rad). Membranes were blocked with 5% non-fat dry milk for1 hour, then probed with mouse anti-human CXCR-2 (1:200;R&D Systems) and a-tubulin (1:5,000; Genscript) antibodiesfollowed by goat anti-mouse horseradish peroxidase (HRP)secondary antibodies. Densitometric analysis of bands wasconducted with ImageJ (NIH).

Statistical analysisAll experiments were carried out 3 separate times and data

reported as mean � SD. One-way ANOVA followed by theTukey post-test was conducted to conclude statistical signi-ficance, with a ¼ 0.05, using Prism 5 software (GraphPad).

ResultsIn vivo tumor formation by CSC-seeded 3D polymericscaffolds recapitulates clinical features of humanglioblastoma multiforme and is accelerated byendothelial cell signaling

To study the contributions of the 3D microenvironmentand endothelial cell signaling on CSC behavior, CSCs frompatient glioblastoma resections were used. These cells werecultured in nonadherent flasks under typical stem cellculture conditions over multiple passages and formed char-acteristic neurospheres that robustly expressed stem cellmarkers Sox2, Nestin, and Oct4 (Fig. 1A) and were capable ofmultilineage differentiation under serum-containing condi-tions (Supplementary Fig. S1). Using this undifferentiatedCSC population, we first wanted to confirm that coimplan-tation of human CSCs with human brain endothelial cells(hCMECs) within highly porous PLG scaffolds enhancesglioblastoma formation relative to that of CSCs alone. In-deed, longitudinal measurements of tumor and scaffoldvolumes via high-resolution ultrasound showed that CSCsimplanted together with hCMECs exhibited significantlyaccelerated tumor formation as compared with CSCs aloneby 8 to 11 weeks (Fig. 1B). As blank control scaffolds werelargely degraded (�80%) by 6 weeks, final tumor volumeswere due to differences in cell number rather than scaffoldcontributions (Fig. 1C). Furthermore, ultrasound analysisrevealed multiple regions of blood flow throughout the mass,suggesting that neoplasms derived from coimplantationwere highly vascularized (Supplementary Fig. S2A), whichis consistent with human glioblastoma multiforme.

Following explanation, tumors were subjected to histologicanalysis to confirm that the subcutaneous PLG implantationmodel mimicked bona fide glioblastoma multiforme. Evalua-tion of H&E-stained cross-sections showed that tumors con-tained extensive necrotic cores, which were circumscribed bylarge pseudopalisading nuclei (Fig. 1D), both considered clin-ical markers of human glioblastoma (21). Furthermore, fre-quent microvascular hyperplasia (Supplementary Fig. S2B),along with common detection of mitotic figures (Supplemen-

tary Fig. S2C) and diffuse tumor cell infiltration outside of theprimary tumor mass (Supplementary Fig. S2D), further sup-ported the diagnosis of a grade IV malignant glioma (22). Tofurther identify the fate of the implanted endothelial cells andassess whether CSCs maintained preferential perivascularcolocalization as reported in clinical specimens (5), we con-ducted immunohistochemical analysis of endothelial and CSCmarkers. Similar to previous results with human dermalmicrovascular endothelial cells (23), hCMECs formed capillarystructures that anastomosed with the mouse vasculature (Fig.1E, Supplementary Fig. S3A). Similarly, these vessels bornefrom coimplantation appeared larger than those observed intumors formed from by CSC implantation alone; however, theextent of tumor vascularization was similar between bothgroups (Supplementary Fig. S3B). Interestingly, Sox2þ (Fig.1F) and Oct4þ (Supplementary Fig. S3C) CSCs were detectedlargely in close or direct proximity to these capillary structuresand associated more frequently with portions of the vesselsthat contained human endothelial cells (Fig. 1F), suggestingthe establishment of perivascular niches supportive of CSCmaintenance in our model.

3D microenvironmental conditions stimulateendothelial cell secretion of paracrine factors thatsupport CSC functions in vitro

To specifically isolate the paracrine signaling effects ofhCMECs on CSC-mediated gliomagenesis in our model, wenext conducted in vitro studies in which hCMECs were cul-tured in PLG scaffolds to recapitulate their 3D interactionswithin the vasculature. Cells adhered readily within the poresof the scaffolds and arranged into capillary tube–like struc-tures that matured over sustained culture periods (Fig. 2A),showing the ability of our model to support 3D vasculogenicprocesses and microenvironmental conditions in vitro. Toevaluate whether scaffold culture and, thus, 3D microenviron-mental context, affected the ability of hCMECs to modulateCSC functions, we carried out conditionedmedia experiments.To this end, conditioned media from hCMECs cultured inconventional 2D culture or 3D PLG scaffolds were collectedand supplemented to CSC cultures (Fig. 2B). Paracrine factorsfrom 3D hCMECs were significantly more efficacious in retain-ing expression of CSC stem cell markers Nestin and Sox2 thanconditionedmedia from 2D hCMEC cultures or basal medium,where Sox2 expression was minimal and Nestin expressionnearly absent (Fig. 2C andD). Furthermore, 3D-derived hCMECsignaling factors also accelerated CSC growth in vitro. This wasevidenced by larger neurosphere size in cultures supplementedwith media from 3D-cultured hCMECs, which was furthercorroborated by an almost 3-fold increase in total DNA relativeto similar cultures fortified with conditioned media from 2DhCMEC cultures (Fig. 2E). In light of the heterogeneous stemcell marker expression, cultures supplemented with 2D con-ditioned media likely contain a mixed population of CSCs(Fig. 2C) making the differences in neurosphere area andDNA content seem even more dramatic. This was substanti-ated by the presence of frequent plate adhesion of glial-likecells in CSC cultures receiving 2D conditioned media, whichwas nearly absent in similar cultures fortified with 3D factors

Infanger et al.





(Supplementary Fig. S4). Collectively, these data suggest that3D microenvironmental conditions play a pivotal role inmediating paracrine signaling by hCMECs, which supportCSC maintenance and growth.

IL-8 regulates the migration and growth of glioblastomaCSCsThree-dimensional culture conditions can increase IL-8

expression and secretion in various 3D culture formats(15, 19, 24). Furthermore, IL-8 reportedly plays importantroles in glioblastoma angiogenesis (9) glioma, and tumor-associated endothelial cell motility (11, 25), and non-glio-

blastoma CSCs self-renewal and pluripotency (13, 14). There-fore, we next hypothesized that greater CSC growth andmaintenance in response to 3D hCMEC-conditioned mediamay be related to elevated IL-8 availability. Indeed, 3Dcultured hCMECs secreted significantly more IL-8 relativeto the same number of cells maintained in 2D monolayers(Fig. 3A). Interestingly, exposure of 3D hCMEC cultures tosoluble factors secreted by CSCs further enhanced IL-8secretion by these cells. Together, these data show a sup-portive role for 3D cell–microenvironment interactions onendothelial cell–derived IL-8 concentrations within the peri-vascular microenvironment (Fig. 3A).

Figure 1. Brain endothelial cells accelerate tumor formation by CSC-seeded 3D polymeric scaffolds in vivo. A, patient-derived CSCs form neurospheres whengrown in stem cell culture conditions and robustly express stem cell markers Sox2, Nestin, and Oct4. Nuclear counterstain with DAPI. Scale bars, 50 mm.B, coimplantation of hCMECswithCSCs in porous PLGscaffolds accelerated tumor growth relative toCSC implantation alone as calculated from3DB-modeultrasound images (inset) collected at weekly time points following implantation. Images show representative cross-sections of coimplanted andcontrol groups 8 weeks following subcutaneous implantation. ��, P < 0.001 versus hCMEC alone; #, P < 0.01 versus CSC alone; †, P < 0.05 versus CSC þhCMEC at 6 weeks. Scale bars, 1 mm. C, because of hydrolytic degradation, the volume of blank control scaffolds was reduced markedly by 6 weeks asdetermined via longitudinal ultrasound measurements and confirmed following scaffold explanation. Scale bars, 1 cm. D, H&E-stained cross-sections oftumors explanted after 8 weeks revealed clinical features of human glioblastoma multiforme, including large necrotic (N) regions surrounded bypseudopalisades (�; left) and vascular hypertrophy (VH; right). Asterisk in right panel denotesnon-hypertrophic vasculature for comparison. Scalebars, 500mm(left) and 100 mm (right). E, immunohistochemical analysis of coimplanted tumors using species-specific antibodies to the endothelial cell marker CD31andDAPI counterstain aswell asdetectionof blood cellswithin vessel lumens suggests that implantedhCMECs formedcapillary structures that anastomosedwith the mouse vasculature. Scale bar, 20 mm. F, undifferentiated CSCs colocalized with hCMECs as detected by coimmunostaining for Sox2 andhuman CD31. Scale bar, 50 mm.






To better define the direct contributions of IL-8 to CSCmigration and growth (i.e., 2 parameters regulating CSCrecruitment to and maintenance within the perivascularniche), we conducted studies with recombinant human IL-8.The chemotactic ability of IL-8 on CSCs was assessed with aTranswell migration assay using 2 separate patient-derivedCSC populations. In both cases, IL-8 increased migration ofCSCs 5- to 10-fold from control groups challenged with controlmedia (Fig. 3B), whichwas preserved to a lesser, yet significant,extent with a 10-fold reduction in IL-8 concentration (Supple-mentary Fig. S5A). To analyze whether IL-8 stimulation sim-ilarly accelerates CSC growth, patient-derived glioblastomaCSCswere cultured in typical stem cellmedium in the presenceor absence of human IL-8. Addition of IL-8 elicited a dose-dependent increase in neurosphere size of CSCs, which pla-teaued at 50 ng/mL (data not shown; Fig. 3C). To confirm thatthis effect was mediated via binding to the cognate IL-8

receptors CXCR1 and/or CXCR2 (9), CSCs were incubated withthe respective antagonist antibodies before IL-8 treatment.CXCR1 blockade abolished IL-8–induced proliferation andresulted in CSC growth, which matched that of control cul-tures, suggesting that this receptor is partly responsible for theproliferative effects of IL-8 on CSCs (Fig. 3C). Interestingly,selective impairment of CXCR2 binding more dramaticallyimpacted CSC growth: cell numbers dropped below controlsand even below cell numbers at the start of the experiment(data not shown). This indicates that (i) signaling via thisreceptor is necessary for normal CSC viability and (ii) CSCfunctions may partly depend on autocrine IL-8 signaling (Fig.3C). To confirm the functionalization of IL-8 on CSC growth,saturating levels of an antagonist IL-8 antibody were intro-duced into cultures that were concomitantly stimulated withIL-8. Again, CSC numbers decreased to levels observed withCXCR2 blockade, further supporting that IL-8 signaling is

Figure 2. 3D microenvironmentalconditions stimulate endothelialcell secretion of molecules thatsupport CSC functions in vitro.A, hCMECs seeded into PLGscaffolds readily adhered withinscaffold pores and assembled intocapillary-like structures by 3 days(middle, arrow) that matured intoendothelial networks with luminalarchitecture (�) by 2 weeks (right).DAPI counterstain. Scale bars, 20mm (phalloidin) and 50 mm (CD31).B, to evaluate microenvironment-dependent hCMEC paracrinesignaling on glioblastoma CSCs,conditioned media were collectedfrom 2D or 3D hCMEC (endothelialcell) cultures, normalized to DNAcontent, and supplemented toCSC cultures. C, representativeimages of CSC neurospheresfollowing 5 days of culture with 3D(left) or 2D (middle) conditionedhCMECmedia (or basal endothelialcell media, right) supplementationillustrates enhanced presence ofstem cell markers Nestin (red) andSox2 (green) in the presence of 3Dderived soluble factors. DAPIcounterstain. Scale bars, 20 mm.D, image analysis of Sox2- orNestin-positive cells from CSCcultures supplemented with mediafrom 2D or 3D endothelial cellsconditioned media or basal media.��, P < 0.01 or ��, P < 0.001 versus3D endothelial cell conditionedmedia. E, media collected from 3Dcultured hCMECs enhancedneurosphere size (left and insetimages) and DNA content (right)relative to media obtained from 2DhCMEC cultures. ��, P < 0.01versus 2D; ��,P<0.001 versus 2D.Scale bar, 200 mm.

Infanger et al.





necessary for both maintenance and proliferation of glioblas-toma multiforme CSCs (Fig. 3C).

Paracrine communication between brain endothelialcells and CSCs enhances IL-8 signaling and facilitatesCSC migration and growth in vitroNext, we sought to determine the functional consequence of

the abovementioned results by evaluatingwhether 3D-culturedhCMECs promote CSC migration and growth due to elevatedIL-8 signaling. Toward this end, we first broadly comparedCSC migration toward 3D hCMEC-conditioned media, IL-8–fortified, or control basal media using Transwell migrationassays. Our results indicated that paracrine factors secreted byscaffold-cultured hCMECs enhanced CSC migration in a sim-ilar manner as recombinant IL-8 (2.7- vs. 1.9-fold relative tocontrol conditions). To further determine whether chronicsignaling by endothelial cells as present in the brain perivas-cular microenvironment impacts the migratory potential ofCSCs, CSCs were preconditioned with 3D hCMEC-secretedfactors before migration experiments. While this pre-exposure

enhanced the overall migratory potential of CSCs only insigni-ficantly, a much more pronounced effect was detected for di-rected migration toward either 3D hCMEC-conditionedmedia or IL-8 (Fig. 3D). As these results suggested thathCMEC-derived paracrine signaling factors elevated IL-8 re-sponsiveness, we hypothesized that enhanced migration ofpreconditioned CSCs to IL-8 may be related to altered expres-sion of the cognate receptors CXCR1 and/or CXCR2. Indeed,quantitative PCR (qPCR) analysis of receptor expression formultiple patient-derived glioblastoma CSCs revealed that pre-conditioning of these cells with 3D hCMEC-conditioned mediaresulted in a 20- and 40-fold increase in CSCCXCR1 andCXCR2expression, respectively (Fig. 3E; CSC 248 population). Impor-tantly, preconditioning with 3D hCMEC-derived factors onlyaffected CSC expression of IL-8 receptors significantly; intrinsicexpression of IL-8 was also downregulated, but these levelswere statistically indifferent from 2D hCMEC-treated cells(Supplementary Fig. S5B). Taken together, these findingsshow that synergistic paracrine interactions between brainendothelial cells and glioblastomaCSCs culminate in enhanced

Figure 3. 3Dmicroenvironmental conditions enhancehCMEC-dependentCSCmigration andgrowth in an IL-8–dependentmanner. A, ELISA revealed that IL-8secretion fromhCMECs (normalized toDNAcontent) is enhanced inPLGscaffold cultures (3D) versusmonolayers (2D) and further increasedwhen3Dculturedcells are preconditioned with media collected from CSCs. ��, P < 0.01 or ��, P < 0.001 versus 2D; †, P < 0.05 versus 3D. B, Transwell migrationassays elucidated that IL-8 elevated migration of 2 separate CSC lines relative to basal control media. �, P < 0.05 or ��, P < 0.001 versus media. C, DNAanalysis 48 hours following IL-8 stimulation suggested that IL-8 enhanced CSC growth. Addition of CXCR1 antagonist antibodies abolished this effect, andantagonist antibodies to either IL-8 or CXCR2 further attenuated growth to levels below control. �, P < 0.05 versus basal; #, P < 0.05 or ��, P < 0.001versusCSCþ IL-8; †,P < 0.01 versus basal. D,migration of CSCs toward 3DhCMEC-secreted factors (3D endothelial cell cond) or IL-8was enhancedby pre-incubation of CSCs with media collected from 3D cultured hCMECs (3D endothelial cell conditioned) relative to basal control media. Effects were morepronounced for migration toward 3D endothelial cell media as compared with IL-8. �, P < 0.05 or ��, P < 0.001 versus basal-conditioned CSCs of samechemotax; #, P < 0.05 versus 3D endothelial cell conditioned CSCs against basal cond chemotax; †, P < 0.05 versus 3D endothelial cell conditionedCSCs against IL-8 chemotax. E, analysis of gene transcription via qPCR indicates that preconditioning of CSCs with conditioned media from 3DhCMEC cultures upregulated (vs. basal media preconditioned) both cognate IL-8 receptors CXCR1 and CXCR2. �, P < 0.05 versus basal-conditioned.F, functional impacts of paracrine factors from 3D hCMECs on CSC migration and growth rely upon IL-8 signaling: Transwell migration of 3D hCMECpreconditioned CSCs toward IL-8 is robustly impaired with the addition of antagonist antibodies to CXCR1, CXCR2, and IL-8 (left). Elevated CSCgrowth imbued by similar preconditioning is attenuated with the presence of CXCR1, CXCR2, or IL-8 antagonist antibodies (right). �, P < 0.05 or��, P < 0.01 versus control 3D hCMEC-preconditioned CSCs.






IL-8 signaling via cell-specific upregulation of both ligand (inhCMECs) and its cognate receptors (in CSCs).

While these findings support endothelial cell paracrinesignaling as an underlying cause of enhanced CSC invasion,they do not directly evaluate participatory roles of IL-8 orCXCR1/2 in this process. Therefore, to directly assess thecontributions of IL-8 stimulation and/or CXCR1/2 binding oncell migration, antagonist antibodies to each were incorporat-ed in repeat Transwell assays. Selective inhibition of IL-8 in 3Dconditioned hCMEC media, strongly decreased migration ofpreconditioned glioblastoma CSCs, thereby showing that sig-naling cascades initiated by this factor are pivotal to CSCchemotaxis (Fig. 3F). Likewise, when cells were treated withantagonists to either CXCR1 or CXCR2, CSC migration wassimilarly attenuated, establishing that signaling through bothreceptors functions to facilitate CSC migration. Interestingly,CXCR2 inhibition appeared to have a greater impact on CSCinvasion than did CXCR1 blockade, yet these results were notstatistically significant. Importantly, the impact of these antag-onist antibodies was specific to IL-8 signaling, as treatmentwith IgG control antibody elicited no effect on CSC migration(Supplementary Fig. S5C). These data implicate both receptorsin CSC migration toward endothelial cell–secreted IL-8 withinthe perivascular microenvironment.

Lastly, the functional significance of 3D hCMEC-mediatedCXCR1 andCXCR2upregulation onCSC growthwas evaluated.To this end, preconditioned CSCs were cultured in the pres-ence or absence of selective antagonists to IL-8 or its cognatereceptors. Similar to the migration findings above, blockade ofeither receptor significantly impaired the effects of precondi-tioning on CSC growth (Fig. 3F). Similarly, direct blockade ofIL-8 also impaired the effects of preconditioning with 3DhCMEC-derived factors on CSC proliferation, further implicat-ing that the stimulatory effects of paracrine endothelial cellsignaling on CSC growth are mediated through IL-8 binding toboth CXCR1 and CXCR2.

Inhibition of CXCR2 signaling by CSCs abolishes thestimulatory effects of endothelial cells on tumor growthin vivo

Thus far, these results provide corroborating evidence thatIL-8 signaling is a critical mediator of glioblastoma CSC

migration, maintenance, and growth. Therefore, to directlyassess whether IL-8 receptor signaling in glioblastoma CSCsunderlies tumor formation siRNA-based inhibition strategieswere used. Given the more pronounced effect of CXCR2 in IL-8–mediated CSC growth (Fig. 3C), siRNA sequences targetedagainst human CXCR2 were generated and incorporated intolentiviruses to stably knockdown expression of this receptor(Supplementary Fig. S6A). Cotransduction of 3 constructs inglioblastoma multiforme CSCs caused a robust (�80%, Sup-plementary Fig. S6B) knockdown of CXCR2 expression andprotein levels (�65%, Supplementary Fig. S6C) with no impacton expression of CXCR1 (data not shown). Next, transduced ornative cells were seeded alone or in combination with hCMECsin PLG scaffolds, whichwere then subcutaneously implanted insevere combined immunodeficient mice (SCID) mice. Coim-plantation of native CSCs with hCMECs recapitulated prom-inent tumor formation by 8 weeks, whose volume dramaticallyexceeded that of tumors borne by CSC implantation alone (Fig.4A and B). However, these effects were abolished with CXCR2knockdown, whereby the volume of tumors formed by trans-duced glioblastoma CSCs was half that of those generated bynative CSCs (Fig. 4B). Despite this difference in size, thesetumors were vascularized to a similar extent as those resultingfrom native CSC implantation (Supplementary Fig. S7), sup-porting that differences in tumor growth were not merely theresult of stunted tumor perfusion. Importantly, coimplanta-tion of hCMECs with CXCR2 knocked-down CSCs failed toreverse these effects or accelerate tumor formation (Fig. 4B),thus providing strong evidence that the promoting effects ofbrain endothelial cells on glioblastoma multiforme CSC–dependent tumor growth are mediated, in large part, throughCXCR2 signaling.

DiscussionDelineating the signaling pathways that regulate CSC main-

tenance, proliferation, and tumorigenicity in the perivascularniche has proved challenging due to a paucity of in vitromodels, which faithfully recapitulate 3D microenvironmentalconditions. In this study, we used a scaffold-based culturesystem that enables brain endothelial cells to form vascularnetworks, thereby initiating paracrine IL-8 signaling cascades,

Figure 4. Inhibition of CSC CXCR2signaling abolishes the stimulatoryeffects of hCMECs ontumorigenesis in vivo. A,photographs of explanted tumors.B, 3D in vivo ultrasound analysis oftumors 8 weeks followingsubcutaneous implantation of PLGscaffolds seeded with CSCs or sh-CXCR2 knocked down CSCs aloneor in combination with hCMECsconfirmed a robust increase intumor volume due to hCMECcoimplantation, which wasabolishedwith CXCR2 knockdown.Scale bar, 1 mm. ��,P < 0.001 or †,P < 0.01 versus CSC only.

Infanger et al.





which enhance CSC maintenance and growth in vitro andtumor formation in vivo.The detected upregulation of IL-8 in our studies may have

been caused by a variety of microenvironmental conditionsincluding altered dimensionality (15), direct cell–cell con-tact (26), and cell–ECM interactions (27). Furthermore,varied intraculture oxygen concentrations could have beeninvolved (28), whereby this response is regulated by changesin cell morphology that may result from varied culturedimensionality (29), as well as pathologic consequences ofcell death (e.g., necrosis; ref. 30). Moreover, reciprocalinteractions between various glioblastoma multiforme–associated cell types modulate this response. For example,glioblastoma–associated endothelial cells exhibit enhancedexpression of IL-8 (31). Accordingly, hCMECs pre-exposedto conditioned media from glioblastoma CSCs increased IL-8 production in our studies (Fig. 3A), suggesting that CSC-derived signaling is partly responsible for the altered geneexpression observed in glioblastoma–derived primary endo-thelial cells (31). It is unknown whether the concentrationsof IL-8 secreted by endothelial cells in our study are suffi-cient to effect CSC functions as we observed, given thatthese latter experiments were carried out using a consid-erably higher IL-8 stimulus. However, ECM in the tumormicroenvironment can function indirectly to enhance localIL-8 concentrations to levels that approach those used inour studies. For example, components of the ECM suchas heparin and heparan sulfate can efficiently bind IL-8 viaits heparin-binding domains (32), thereby creating signal-ing depots that contribute to enhanced cell chemotaxis(33). Future quantitative analysis of spatial differences inIL-8 from ex vivo tumor samples would help evaluate thispossibility.Culture context–driven endothelial upregulation of IL-8

directly affected CSC maintenance and growth, which wasfurther enhanced by CSC preconditioning with media from 3Dendothelial cell cultures. Specifically, chronic exposure of CSCsto endothelial cell signals as would occur in the perivascularmicroenvironment–upregulated IL-8 receptors CXCR1 andCXCR2, with functional consequences of enhanced CSCmigra-tion and growth. Interestingly, other interleukin receptors havepreviously been associated with glioblastoma CSCmalignancy.More specifically, glioblastoma CSCs express elevated levels ofIL-6 receptors IL6Ra and glycoprotein 130 (gp130) relative tonon–stem glioma cells, which promotes their self-renewal andtumorigenic capability (34). Given that the respective ligandsfor these receptors (i.e., IL-6 and IL-8) are regulated by the sametranscription factors (i.e., NF-kB and NF-IL6; ref. 35), it ispossible that IL-6 secretion by 3D cultured endothelial cellsmay partially contribute to altered CSC functions in ourstudies. Alternatively, IL-6 may serve as an upstream stimulusfor the increased transcription of IL-8 (36). Hence, it is con-ceivable that CSC-derived IL-6 not only promotes gliomagen-esis in an autocrine manner (34) but also by stimulating thetranscription of IL-8 by brain endothelial cells (Fig. 3A).We have shown that both cognate receptors CXCR1 and

CXCR2 participate in the growth- and migratory-inducingeffects of IL-8 on CSCs. These findings agree with observations

in human microvascular endothelial cells, where blockade ofeither receptor attenuated chemotaxis toward IL-8 (37). How-ever, with regard to cell growth, our data indicate a far greaterimpact of CXCR2 blockade, which resulted in declining cellcounts below control culture levels (Fig. 3C). This may bepartially explained by the broader range of signaling mole-cules, which can bind to this receptor (9). Alternatively, thespecific downstream effectors of CXCR2 may be critical toIL-8–induced CSC growth. Included in these are activation ofphosphoinositide 3-kinase (PI3K) and extracellular signal–regulated kinase (ERK)1/2 (38), which have been implicatedin the proliferation and differentiation of neural stem cells(39), respectively. Interestingly, PTEN has also been shownto be a potent regulator of PI3K signaling and has beenimplicated in the maintenance of various CSC types (40, 41).However, in our studies, both PTENþ and PTEN� CSCsresponded similarly to 3D endothelial cell conditionedmedia (Fig. 2C), suggesting that CXCR2-mediated signalingmay invoke similar signaling pathways as PTEN, possiblythrough PI3K activation.

While this study was specifically designed to better definethe role of endothelial cell–derived soluble factors on CSCfunctions, the presented culture model may be broadly appli-cable to other questions of glioblastoma perivascular nicheinteractions. For example, CSCs may promote tumor growthby modulating blood vessel functions through differentiationinto endothelial cells (42) or pericytes (43). Future studiesinvolving direct coculture of both cell types within the pre-sented scaffold system may help better define the under-lying molecular and cellular mechanisms and assess whetheraltered IL-8 signaling may contribute to these changes.Furthermore, the composition and mechanical properties ofendothelial cell–deposited ECM likely play key roles in guid-ing CSCs but were not considered in the current study.Endothelial cells assemble an ECM that is enriched in laminina2, a promoter of CSC malignancy (44). As both laminina2 and IL-8 inversely correlate with patient survival (44, 45)and are upregulated in glioblastoma–associated endothelialcells relative to normal brain endothelial cells (25, 44), it isprovocative to consider that a functional link between these 2factors exists. Finally, ECM mechanical properties criticallynot only impact glioma migration (46) and endothelial cellbehavior (47) but also amplify cellular response to solublesignals (48). Modulating the presentation of biomaterialswithin these cell culture platforms would allow evaluationof the collective contributions of these parameters on CSCbehavior, which would significantly advance our understand-ing of the physicochemical parameters influencing cell behav-ior in the perivascular niche. Finally, it is essential to realizethat endothelial cells are not the sole source of IL-8 in theglioblastoma microenvironment. For example, additionalcell types resident within the tumor parenchyma, includingastrocytes (49) and bulk glioma cells themselves (50), cancontribute to interstitial IL-8 levels. Future studies using the3D culture platform with these additional cell types wouldhelp elucidate whether paracrine factors secreted by thesecell types can similarly enhance CSC sensitivity to IL-8 viareceptor upregulation as endothelial cells do.






It is critical to note that while subcutaneous coimplantationof CSCs with endothelial cells mimicked certain glioblastoma–like characteristics in vivo, CSCs implanted alone failed torecapitulate advanced tumors within the time span of theexperiment (11 weeks). These findings appear in contrast withother animal models of glioblastoma, for example, intracranialCSC implantation, which show rapid tumor formation by thesecells, and would thus suggest that additional signaling para-meters not afforded by the scaffold model or subcutaneousenvironment are involved inCSC-mediated tumorigenesis (51).As it is possible that varied vascular density contribute to thesedifferences (52), future studies will be required to evaluate theimportance of IL-8 signaling in an orthotopic or transgenicmodel of glioblastoma.

The finding that inhibition of the CXCR2 signalingpathway in CSCs abolishes tumorigenesis in vivo has signif-icant implications for the clinical management of humanglioblastoma multiforme. It is important to note that addi-tional cytokines—CXCL2, CXCL3, and CXCL5, among others(53)—also bind CXCR2 and, as such, may participate inCSC-driven gliomagenesis, which has not been investigatedherein. Interestingly, the application of certain anticancertreatments such as photodynamic or temozolomide therapyin multiple glioma cell lines has been shown to stimulatetranscription of many of the CXCR2 agonists (e.g., CXCL2,CXCL3, IL-8; refs. 54, 55) as well as IL-6 (54), which could,in theory, support subsequent CSC growth and recurrenttumor formation that is nearly reflexive for human gliobla-stoma. In light of this, the development and delivery ofCXCR2 antagonists in combination with conventional ther-apeutic approaches may offer a significant advantage forclinical treatment of the disease.

In conclusion, we have developed a tissue-engineered cul-ture model to resolve the impact of 3D microenvironmentalconditions on endothelial cell–mediated paracrine signaling,which supports glioblastoma tumor initiation and progression.Our results revealed that perivascular niche–associated IL-8signaling between endothelial and CSCs may contribute topoor clinical prognosis. These findings not only warrant the

future study of anti-CXCR2 therapies for clinical managementof glioblastoma multiforme but also underscore the value ofphysiologically relevant models for the study of patient cancercell behavior.

Disclosure of Potential Conflicts of InterestThe content is solely the responsibility of the authors and does not necessarily

represent the official views of NCI or NIH. No potential conflicts of interest weredisclosed.

Authors' ContributionsConception and design: D.W. Infanger, S.C. Liu, J.A. Boockvar, C. FischbachDevelopment of methodology: D.W. Infanger, C. FischbachAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): D.W. Infanger, Y. Cho, B.S. Lopez, S. Mohanan,S.C. Liu, D. Gursel, C. FischbachAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): D.W. Infanger, S. Mohanan, S.C. Liu, C. FischbachWriting, review, and/or revision of the manuscript: D.W. Infanger,S. Mohanan, J.A. Boockvar, C. FischbachAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): D.W. InfangerStudy supervision: D.W. Infanger, C. FischbachGeneration of glioblastoma stem–like cells from patients and theircharacterization by self-renewal, expression of progenitor stem cellmarkers, in vitro and in vivo tumorigenesis properties: D. Gursel

AcknowledgmentsThe authors thank the contributions and creativity of Dr. Caroline Coffey

(posthumous), which provided the basis of many experiments outlined in thisstudy.

Grant SupportThiswork was supported by NIH/NCI (RC1 CA146065) and the Cornell Center

on the Microenvironment & Metastasis through Award Number NCI U54CA143876, and by NIH 5K08CA130985-05 (J.A. Boockvar). D.W. Infanger wasfunded through an American Brain Tumor Association (ABTA) Basic ResearchFellowship onbehalf of TeamWill Power and aNational Cancer Institute/CornellNanobiotechnology Center Young Investigator Award. Y. Chowas supported by aNYSTEM summer research fellowship, and B.S. Lopez was supported by an NIHT35AI007227 training grant.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received May 8, 2013; revised September 19, 2013; accepted September 26,2013; published OnlineFirst October 11, 2013.

References1. Buckner JC, Brown PD, O'Neill BP, Meyer FB, Wetmore CJ, Uhm

JH. Central nervous system tumors. Mayo Clin Proc 2007;82:1271–86.

2. Singh SK, Clarke ID, Terasaki M, Bonn VE, Hawkins C, Squire J, et al.Identification of a cancer stem cell in human brain tumors. Cancer Res2003;63:5821–8.

3. Singh SK, Hawkins C, Clarke ID, Squire JA, Bayani J, Hide T, et al.Identification of human brain tumour initiating cells. Nature 2004;432:396–401.

4. BaoS,WuQ,McLendonRE,HaoY,ShiQ,HjelmelandAB, et al. Gliomastem cells promote radioresistance by preferential activation of theDNA damage response. Nature 2006;444:756–60.

5. Calabrese C, Poppleton H, Kocak M, Hogg TL, Fuller C, Hamner B,et al. A perivascular niche for brain tumor stem cells. Cancer Cell2007;11:69–82.

6. Lathia JD, Heddleston JM, Venere M, Rich JN. Deadly teamwork:neural cancer stem cells and the tumor microenvironment. Cell StemCell 2011;8:482–5.

7. BrownCK,KhodarevNN, Yu J,Moo-Young T, Labay E,Darga TE, et al.Glioblastoma cells block radiation-induced programmed cell death ofendothelial cells. FEBS Lett 2004;565:167–70.

8. Zhu VF, Yang J, Lebrun DG, Li M. Understanding the role of cytokinesin glioblastoma multiforme pathogenesis. Cancer Lett 2012;316:139–50.

9. Brat DJ, Bellail AC, Van Meir EG. The role of interleukin-8 and itsreceptors in gliomagenesis and tumoral angiogenesis. Neuro Oncol2005;7:122–33.

10. VanMeir E, CeskaM, Effenberger F,Walz A, Grouzmann E, DesbailletsI, et al. Interleukin-8 is produced in neoplastic and infectious diseasesof the human central nervous system. Cancer Res 1992;52:4297–305.

11. Raychaudhuri B, Vogelbaum MA. IL-8 is a mediator of NF-kappaBinduced invasion by gliomas. J Neurooncol 2011;101:227–35.

12. Li A, Dubey S, Varney ML, Dave BJ, Singh RK. IL-8 directly enhancedendothelial cell survival, proliferation, and matrix metalloproteinasesproduction and regulated angiogenesis. J Immunol 2003;170:3369–76.

Infanger et al.





13. Ginestier C, Liu S, Diebel ME, Korkaya H, Luo M, Brown M, et al.CXCR1blockade selectively targets humanbreast cancer stem cells invitro and in xenografts. J Clin Invest 2010;120:485–97.

14. Tang KH,Ma S, Lee TK, Chan YP, Kwan PS, Tong CM, et al. CD133(þ)liver tumor-initiating cells promote tumor angiogenesis, growth, andself-renewal through neurotensin/interleukin-8/CXCL1 signaling.Hepatology 2012;55:807–20.

15. FischbachC,ChenR,Matsumoto T, Schmelzle T, Brugge JS, PolveriniPJ, et al. Engineering tumors with 3D scaffolds. Nat Methods 2007;4:855–60.

16. Verbridge SS, Choi N, Zheng Y, Brooks DJ, Stroock A, Fischbach-Teschl C. Oxygen-controlled 3-D cultures to analyze tumor angiogen-esis. Tissue Eng Part A 2010;16:2133–41.

17. Weksler BB, Subileau EA, Perriere N, Charneau P, Holloway K, Leve-que M, et al. Blood-brain barrier-specific properties of a human adultbrain endothelial cell line. FASEB J 2005;19:1872–4.

18. Howard BM, Gursel DB, Bleau AM, Beyene RT, Holland EC, BoockvarJA. EGFR signaling is differentially activated in patient-derived glio-blastoma stem cells. J Exp Ther Oncol 2010;8:247–60.

19. Pathi SP, Kowalczewski C, Tadipatri R, Fischbach C. A novel 3-Dmineralized tumor model to study breast cancer bone metastasis.PLoS One 2010;5:e8849.

20. Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the com-parative C(T) method. Nat Protoc 2008;3:1101–8.

21. Brat DJ, Prayson RA, Ryken TC, Olson JJ. Diagnosis of malignantglioma: role of neuropathology. J Neurooncol 2008;89:287–311.

22. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger PC, Jouvet A,et al. The 2007 WHO classification of tumours of the central nervoussystem. Acta Neuropathol 2007;114:97–109.

23. Nor JE,PetersMC,ChristensenJB,SutorikMM,LinnS,KhanMK, et al.Engineering and characterization of functional human microvessels inimmunodeficient mice. Lab Invest 2001;81:453–63.

24. Verbridge SS, Chandler EM, Fischbach C. Tissue-engineered three-dimensional tumor models to study tumor angiogenesis. Tissue EngPart A 2010;16:2147–52.

25. Charalambous C, Pen LB, Su YS, Milan J, Chen TC, Hofman FM.Interleukin-8 differentially regulatesmigration of tumor-associated andnormal human brain endothelial cells. Cancer Res 2005;65:10347–54.

26. Tan CP, Seo BR, Brooks DJ, Chandler EM, Craighead HG, FischbachC. Parylene peel-off arrays to probe the role of cell-cell interactions intumour angiogenesis. Integr Biol (Camb) 2009;1:587–94.

27. Fischbach C, Kong HJ, Hsiong SX, Evangelista MB, Yuen W, MooneyDJ. Cancer cell angiogenic capability is regulated by 3D culture andintegrin engagement. Proc Natl Acad Sci U S A 2009;106:399–404.

28. Desbaillets I, Diserens AC, de Tribolet N, Hamou MF, Van Meir EG.Regulation of interleukin-8 expression by reduced oxygen pressure inhuman glioblastoma. Oncogene 1999;18:1447–56.

29. Verbridge SS, Chakrabarti A, Delnero P, Kwee B, Varner JD, StroockAD, et al. Physicochemical regulation of endothelial sprouting in a 3Dmicrofluidic angiogenesis model. J Biomed Mater Res Part A 2013;101:2948–56.

30. Desbaillets I, Diserens AC, Tribolet N, Hamou MF, Van Meir EG.Upregulation of interleukin 8 by oxygen-deprived cells in glioblastomasuggests a role in leukocyte activation, chemotaxis, and angiogenesis.J Exp Med 1997;186:1201–12.

31. Charalambous C, Chen TC, Hofman FM. Characteristics of tumor-associated endothelial cells derived from glioblastoma multiforme.Neurosurg Focus 2006;20:E22.

32. Kuschert GS, Hoogewerf AJ, Proudfoot AE, Chung CW, Cooke RM,Hubbard RE, et al. Identification of a glycosaminoglycan bindingsurface on human interleukin-8. Biochemistry 1998;37:11193–201.

33. WebbLM, EhrengruberMU,Clark-Lewis I, BaggioliniM,Rot A.Bindingto heparan sulfate or heparin enhances neutrophil responses to inter-leukin 8. Proc Natl Acad Sci U S A 1993;90:7158–62.

34. WangH, Lathia JD,WuQ,Wang J, Li Z,Heddleston JM, et al. Targetinginterleukin 6 signaling suppresses glioma stem cell survival and tumorgrowth. Stem Cells 2009;27:2393–404.

35. Matsusaka T, Fujikawa K, Nishio Y, Mukaida N, Matsushima K,Kishimoto T, et al. Transcription factors NF-IL6 and NF-kappa Bsynergistically activate transcription of the inflammatory cytokines,

interleukin 6 and interleukin 8. Proc Natl Acad Sci U S A 1993;90:10193–7.

36. Okamura S, Fujiwara H, Yoneda M, Furutani A, Todo M, Ikai A, et al.Overexpression of IL-6 by gene transfer stimulates IL-8-mediatedinvasiveness of KYSE170 esophageal carcinoma cells. AnticancerRes 2013;33:1483–9.

37. SalcedoR,Resau JH,HalversonD,HudsonEA,DambachM,Powell D,et al. Differential expression and responsiveness of chemokine recep-tors (CXCR1-3) byhumanmicrovascular endothelial cells andumbilicalvein endothelial cells. FASEB J 2000;14:2055–64.

38. Sotsios Y, Ward SG. Phosphoinositide 3-kinase: a key biochemicalsignal for cell migration in response to chemokines. Immunol Rev2000;177:217–35.

39. MaDK, Ponnusamy K, SongMR, Ming GL, Song H. Molecular geneticanalysis of FGFR1 signalling reveals distinct roles of MAPK andPLCgamma1 activation for self-renewal of adult neural stem cells.Mol Brain 2009;2:16.

40. Boontheekul T, Hill EE, Kong HJ, Mooney DJ. Regulating myoblastphenotype through controlled gel stiffness and degradation. TissueEng 2007;13:1431–42.

41. HuangFF,WuDS, ZhangL,YuYH,YuanXY, LiWJ, et al. Inactivation ofPTEN increases ABCG2 expression and the side population throughthe PI3K/Akt pathway in adult acute leukemia. Cancer Lett2013;336:96–105.

42. Ricci-Vitiani L, Pallini R, Biffoni M, TodaroM, Invernici G, Cenci T, et al.Tumour vascularization via endothelial differentiation of glioblastomastem-like cells. Nature 2010;468:824–8.

43. Cheng L, Huang Z, Zhou W, Wu Q, Donnola S, Liu JK, et al. Glioblas-toma stem cells generate vascular pericytes to support vessel functionand tumor growth. Cell 2013;153:139–52.

44. Lathia JD, LiM,Hall PE,Gallagher J,Hale JS,WuQ, et al. Laminin alpha2 enables glioblastoma stem cell growth. Ann Neurol 2012;72:766–78.

45. Institute NC. REMBRANDT home page. [cited 2013 Apr 30]. Availablefrom: http://rembrandt.nci.nih.gov. Search string IL8 and Kaplan-Meier survival plot.

46. Pathak A, Kumar S. Independent regulation of tumor cell migration bymatrix stiffness and confinement. Proc Natl Acad Sci U S A 2012;109:10334–9.

47. Mammoto A, Connor KM, Mammoto T, Yung CW, Huh D, AdermanCM, et al. Amechanosensitive transcriptionalmechanism that controlsangiogenesis. Nature 2009;457:1103–8.

48. Chandler EM,SeoBR,Califano JP,AndresenEguiluzRC, Lee JS,YoonCJ, et al. Implanted adipose progenitor cells as physicochemicalregulators of breast cancer. Proc Natl Acad Sci U S A 2012;109:9786–91.

49. Shah A, Silverstein PS, Singh DP, Kumar A. Involvement of metabo-tropic glutamate receptor 5, AKT/PI3K signaling and NF-kappaBpathway in methamphetamine-mediated increase in IL-6 and IL-8expression in astrocytes. J Neuroinflammation 2012;9:52.

50. Dwyer J, Hebda JK, Le Guelte A, Galan-Moya EM, Smith SS, Azzi S,et al. Glioblastoma cell-secreted interleukin-8 induces brain endothe-lial cell permeability via CXCR2. PLoS One 2012;7:e45562.

51. Lathia JD, Gallagher J, Myers JT, Li M, Vasanji A, McLendon RE, et al.Direct in vivo evidence for tumor propagation by glioblastoma cancerstem cells. PLoS One 2011;6:e24807.

52. Fukumura D, Yuan F, Monsky WL, Chen Y, Jain RK. Effect of hostmicroenvironment on the microcirculation of human colon adenocar-cinoma. Am J Pathol 1997;151:679–88.

53. Ahuja SK, Murphy PM. The CXC chemokines growth-regulated onco-gene (GRO) alpha, GRObeta, GROgamma, neutrophil-activating pep-tide-2, and epithelial cell-derived neutrophil-activating peptide-78 arepotent agonists for the type B, but not the type A, human interleukin-8receptor. J Biol Chem 1996;271:20545–50.

54. Kammerer R, Buchner A, Palluch P, Pongratz T, Oboukhovskij K,Beyer W, et al. Induction of immune mediators in glioma andprostate cancer cells by non-lethal photodynamic therapy. PLoSOne 2011;6:e21834.

55. Bruyere C, Mijatovic T, Lonez C, Spiegl-Kreinecker S, Berger W, KastRE, et al. Temozolomide-induced modification of the CXC chemokinenetwork in experimental gliomas. Int J Oncol 2011;38:1453–64.






Published OnlineFirst October 11, 2013.Cancer Res David W. Infanger, YouJin Cho, Brina S. Lopez, et al. Signaling in a Tumoral Perivascular NicheGlioblastoma Stem Cells Are Regulated by Interleukin-8

Updated version

10.1158/0008-5472.CAN-13-1355doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2013/10/11/0008-5472.CAN-13-1355.DC1

Access the most recent supplemental material 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

Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://cancerres.aacrjournals.org/content/early/2013/11/21/0008-5472.CAN-13-1355To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-13-1355

http://cancerres.aacrjournals.org/content/suppl/2013/10/11/0008-5472.CAN-13-1355.DC1

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/early/2013/11/21/0008-5472.CAN-13-1355


glioblastoma stem cells are regulated by interleukin-8 signaling...

Documents