neuroblastoma arginase activity creates an...

Microenvironment and Immunology

Neuroblastoma Arginase Activity Creates anImmunosuppressive Microenvironment ThatImpairs Autologous and Engineered ImmunityFrancis Mussai1, Sharon Egan2, Stuart Hunter1, Hannah Webber3, Jonathan Fisher4,Rachel Wheat1, Carmel McConville1, Yordan Sbirkov3, Kate Wheeler5, Gavin Bendle1,Kevin Petrie3, John Anderson4, Louis Chesler3, and Carmela De Santo1

Abstract

Neuroblastoma is the most common extracranial solidtumor of childhood, and survival remains poor for patientswith advanced disease. Novel immune therapies are currentlyin development, but clinical outcomes have not matchedpreclinical results. Here, we describe key mechanisms in whichneuroblastoma inhibits the immune response. We show thatmurine and human neuroblastoma tumor cells suppress T-cellproliferation through increased arginase activity. Arginase II isthe predominant isoform expressed and creates an arginine-deplete local and systemic microenvironment. Neuroblastomaarginase activity results in inhibition of myeloid cell activation

and suppression of bone marrow CD34þ progenitor prolifer-ation. Finally, we demonstrate that the arginase activity ofneuroblastoma impairs NY-ESO-1–specific T-cell receptor andGD2-specific chimeric antigen receptor–engineered T-cell pro-liferation and cytotoxicity. High arginase II expression corre-lates with poor survival for patients with neuroblastoma. Theresults support the hypothesis that neuroblastoma creates anarginase-dependent immunosuppressive microenvironment inboth the tumor and blood that leads to impaired immuno-surveillance and suboptimal efficacy of immunotherapeuticapproaches. Cancer Res; 75(15); 3043–53. �2015 AACR.

IntroductionNeuroblastoma is the most common extracranial malignancy

of childhood. Although the prognosis for low-risk neuroblastomahas improved, patients with high-risk disease have an extremelypoor survival despite intensive multimodal treatment, includingimmunotherapy (1).Neuroblastoma is associatedwith anuniqueinteraction with the immune system, clinically evidenced bypatients who develop paraneoplastic opsoclonus–myoclonussyndrome and patients whose tumors spontaneously regress(2, 3).

Over the last 10 years as the benefit of conventional ther-apies has been maximized, the focus has moved to enhancingan antineuroblastoma immune response. T cells are a majoreffector arm of the immune system and play a key role in the

recognition and targeting of cancer cells. Subsequently engi-neered chimeric antigen receptor (CAR) T cells against thepredominant neuroblastoma surface antigen GD2 demonstrat-ed antineuroblastoma cytotoxicity in vitro and in murine mod-els (4, 5). However, although preclinical studies demonstratethat T cells have the potential for antineuroblastoma activity,the clinical efficacy of immunotherapies has been controversial(6, 7). Immunotherapeutic approaches are reliant on an activeimmune system, therefore one likely hypothesis for theirfailure is that neuroblastoma creates an immunosuppressivemicroenvironment that inhibits autologous or adoptive immu-nity (8, 9).

The mechanisms underlying the immunosuppressive micro-environment in neuroblastoma are poorly understood. In thisstudy, we identify the key role of neuroblastoma arginase activityin inhibiting both autologous and engineered antineuroblastomaimmune responses.

Materials and MethodsNeuroblastoma patient samples

Blood and tumor samples were obtained from 26 patients withneuroblastoma treated at the Birmingham Children's Hospital,Children's Hospital Oxford, and Great Ormond Street Hospital(Supplementary Table S1). The samples were taken from patientswith newly diagnosed neuroblastoma, at the time of diagnosticbiopsy, before the start of treatment.

GD2þ tumor cell isolationFor isolation of GD2þ tumor cells from human and murine

tumors, tumorsweredigestedusingType II collagenase, labeledwithanti-GD2-PE antibody, and bound to anti-PE–coated magnetic

1School of Cancer Sciences, University of Birmingham, Birmingham,United Kingdom. 2School of Veterinary Medicine and Science, Univer-sityofNottingham,Nottingham,SuttonBonnington,UnitedKingdom.3Paediatric Solid Tumour Biology and Therapeutics, Institute of Can-cer Research, London, United Kingdom. 4Unit of Molecular Haematol-ogy and Cancer Biology, Institute of Child Health, University CollegeLondon, United Kingdom. 5Department of Paediatric Oncology, Chil-dren's Hospital Oxford, University of Oxford, John Radcliffe Hospital,Oxford, United Kingdom.

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

Corresponding Author: Francis Mussai, School of Cancer Sciences, Universityof Birmingham, Birmingham B15 2TT, United Kingdom. Phone: 0121-333-8234;Fax: 0121-333-8234; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-14-3443

�2015 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 3043

on December 26, 2019. © 2015 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst June 8, 2015; DOI: 10.1158/0008-5472.CAN-14-3443

http://cancerres.aacrjournals.org/

beads (Miltenyi). Cells were purified according to manufacturer'sinstructions (Miltenyi Biotec). Purity of GD2þ cells was >98% asconfirmed by flow cytometry.

Neuroblastoma murine mode immune characterizationAfter weaning, TH-MYCNmicewere palpated for intra-abdom-

inal tumors twice weekly. Mice with palpable tumors ranging insize between 5 and 20 mm in diameter were then humanlysacrificed. At sacrifice, unheparinized and heparinized wholeblood, as well as tumor tissue and spleen were obtained forfurther ex vivo analyses. Tumor tissue was processed as above.Spleensweremechanically digested andheparinizedwhole bloodwas lysedwith red blood cells lysis buffer (Qiagen). Tumor tissue,spleen, and blood cell suspensions were stained with anti-mouseLy6C, Ly6G, F480, CD3, andGD2 antibody (Biolegend) on ice for30minutes. The expression of these markers was assessed by flowcytometry.

Arginase enzyme activityThe activity of arginase II present within neuroblastoma cell

lines, sorted patient or murine cells, culture supernatants orplasma was determined by measuring the conversion of arginineinto urea, as previously described (10).

Monocyte polarization assayPeripheral blood was collected from healthy donors and

monocytes were separated using a Lymphoprep gradient andenriched by negative selection using a Monocyte Isolation Kit II(Miltenyi). Monocytes were cultured in the presence or absenceof neuroblastoma, neuroblastoma culture supernatants (50%of final volume), or patient plasma (50% of final volume)overnight, in low-adherent 24 plates (CoStar), and then treatedwith lipopolysaccharide (LPS; 10ng/mL, Sigma) for 24hours. Theculture supernatants were harvested and analyzed for cytokinerelease. Cells were harvested and stained with anti-CD86 (Biole-gend). Propidium iodide added to allow viable cells to be gated.The expression of CD86 on monocytes was assessed by flowcytometry.

Murinemyeloid cells were purified from spleens of healthy andtumor-bearing mice and from tumor tissue of neuroblastomamice using flow cytometric sorting. The cells were cultured inR10%or arginine-freemedium and treated with LPS for 24 hours.The culture supernatants were harvested and analyzed for cyto-kine release as before.

Statistical analysisAWilcoxon rank-sum test was used to determine the statistical

significance of the difference in unpaired observations betweentwo groups. All P values are 2-tailed and P < 0.05 was consideredto represent statistically significant events.

Study approvalIn accordance with the Declaration of Helsinki, patient samples

were obtained after written, informed consent prior to inclusion inthe study. Regional EthicsCommittee (RECNumber 10/H0501/39)and local hospital trust research approval for the study was grantedfor United Kingdom hospitals. The Institute of Cancer ResearchEthics Committee approved all animal protocols in this study.Procedures were carried out in accordance with UK Home OfficeGuidelines.

ResultsNeuroblastoma suppresses T-cell proliferation via arginase IIexpression and activity

T cells are the major effector arm of the endogenous anticancerimmune response. However, neuroblastoma is able to develop andmetastasize in patients, despite the immune system suggestingimpaired immunosurveillance. We investigated the effect of neuro-blastoma on T-cell immunity, using human cell lines as an estab-lished model of malignant neuroblastoma. Neuroblastoma celllines were cultured with T cells from healthy donors. T-cell prolif-eration was significantly suppressed by the presence of neuroblas-toma (Fig. 1A), even at low neuroblastoma: T-cell ratios. Using aTranswell assay, we identified that the T-cell suppression was non–contact-dependent (Fig. 1B). Cell line supernatantwas examined forimmunosuppressive molecules, and no detectable levels of IL10,IL4, and TGFb were identified (data not shown), consistent withprevious reports of an unknown alternative mechanism of immu-nosuppression mediated by neuroblastoma cells (9).

We have recently identified that another pediatric malignancy,acute myeloid leukemia (AML), creates an immunosuppressivemicroenvironment through altered arginase activity (10).Although AML is a hematologic malignancy and neuroblastomaa solid tumor, both cancers share a common association withpancytopenia, immune alteration, andbonemarrow infiltration atdiagnosis. Therefore, we hypothesized that neuroblastoma tumorcells similarly harness arginine depletion to suppress the immuneresponse. Arginine is a semiessential amino acid, that is metabo-lized in mammalian cells, principally by the enzymes arginase Iand II, and by inducible nitric oxide synthase (iNOS; ref. 11). Asarginase I and II and iNOSmay be responsible for inhibition of T-cell proliferation, we tested the functional relevance of eachenzyme (12). By measuring the conversion of arginine into urea,we demonstrated that neuroblastoma cells have a measurablearginase activity, albeit lower than AML cell lines (24 mU/106

cells NB vs. 67.5 mU/106 cells AML; Fig. 1C). No increasedproduction of reactive nitric oxide species was demonstrated (datanot shown). Using the small molecules NOHA and L-NMMA toinhibit arginase and iNOS, respectively, we demonstrated that theimmunosuppressive microenvironment is created by arginaseactivity alone (Fig. 1D). NOHA alone was able to rescue T-cellproliferation, with no effect from L-NMMA. Moreover, analysis ofneuroblastoma culture supernatants showed that the concentra-tion of arginine decreases significantly in the local microenviron-ment (Fig. 1E) and leads to downregulation of T-cell CD3z chain(Supplementary Fig. S1a). Addition of exogenous arginine led tomoderate rescue of T-cell proliferation (Supplementary Fig. S1b).Both RT-PCR and Western blotting identified that neuroblastomacell lines express arginase II as the main isoform (SupplementaryFig. S1c and S1d), with no measurable arginase I or iNOS proteinexpression (Supplementary Fig. S1e and S1f). The arginase activitycorrelated with the absolute arginase II intracellular concentrationof the various neuroblastoma cell lines analyzed (SKN-MC > LAN-1 > Kelly; Supplementary Fig. S1d). Silencing of arginase II led torestoration of T-cell proliferation, confirming the specific role ofarginase II (Supplementary Fig. S1g and S1h).

Neuroblastoma creates both a local and a systemic depletion inarginine, extending its immunosuppressive environment

Having established the ability of neuroblastoma cells to sup-press T-cell proliferation in our model, we investigated the role of

Mussai et al.

Cancer Res; 75(15) August 1, 2015 Cancer Research3044




arginase inhumanneuroblastoma samples.High arginase activity(mean, 1,100mU/106 cells) was confirmed in humanGD2þ cellsfrom patients' tumors at diagnosis, with rapid degradation ofarginine into urea (Fig. 2A). Monocytes and neutrophils fromhealthy donors were used as controls. GD2þ cells expressed botharginase isoform genes (Fig. 2B). However, immunohistochem-ical examination of diagnostic tumor biopsy samples, confirmedthe specific expression of arginase II withinGD2þ tumor cells (Fig.2C and Supplementary Fig. S2), with minimal staining for argi-nase I and iNOS enzymes. The findings are consistent with theestablished predominant expression of arginase II in healthyneuronal tissue (13). We performed T-cell proliferation assaysusing GD2þ tumor cells sorted from the neuroblastoma tumors.GD2þ cells showed a significant ability to suppress T-cell prolif-eration (Fig. 2D). Arginase inhibition with NOHA led to rescue ofT-cell proliferation consistent with our earlier findings (10). Nocorrelation was found between patient clinical characteristics andsuppressive activity, including patient age, stage, or Myc-N status.

Patients with advanced neuroblastoma can present with sys-temic immune alterations, including lymphopenia (14, 15). It haspreviously been shown that arginase enzymes can be released bytumor cells to enhance the area of immunosuppression (10). We

found no increase in arginase activity of culture supernatants(data not shown), indicating the enzyme was not released. How-ever, patients with neuroblastoma frequently present with largetumor masses. We hypothesized that neuroblastomamay cause asystemic depletion of arginine due to consumption by the tumormass, despite the failure to secrete free enzyme. Examination ofplasma from patients with neuroblastoma at diagnosis showed asignificant reduction in arginine concentrations in the blood,compared with healthy controls (Fig. 2E; healthy controls,135mmol/L vs. patients with neuroblastoma, 70 mmol/L; P ¼0.0057). No increased plasma arginase activity was seen com-pared with the healthy controls, confirming the absence of secret-ed free arginase enzyme from tumor cells (Supplementary Fig.S2b, P ¼ 0.7139). To assess the effect of low arginine on T-cellproliferation, donor T cells were cultured in the plasmaof patientswith neuroblastoma or healthy donors. We showed that the lowarginine in patient plasma significantly impaired T-cell prolifer-ation compared with T cells cultured in plasma from healthydonors (Supplementary Fig. S2c, P ¼ 0.045) and induces CD3zchain downregulation (Supplementary Fig. S2d). Consistent withpreviousfindings exogenous arginine induces significant rescue ofT-cell proliferation (P ¼ 0.0038, Supplementary Fig. S2e). Thus,

Figure 1.Arginine metabolism regulatesthe suppressive activity ofneuroblastoma. A, T-cell proliferationis suppressed by the presence ofdifferent ratios of neuroblastoma celllines. B, suppressive activity ofneuroblastoma cell lines is maintainedwhen T cells are separated by aTranswell. C, arginase activity fromneuroblastoma and AML cell linesmeasured by the conversion ofarginine into urea. D, T-cellproliferation is restored by theinhibition of neuroblastoma arginaseusing the specific inhibitor NOHA, butnot by the iNOS inhibitor L-NMMA. E,neuroblastoma cell lines significantlydeplete arginine from themicroenvironment. All data arerepresentative of three independentexperiments, Error bars, SD.

Neuroblastoma Arginase-Dependent Microenvironment

www.aacrjournals.org Cancer Res; 75(15) August 1, 2015 3045




our data suggest that neuroblastoma creates both a local and asystemic immunosuppressive microenvironment through thedepletion of arginine by tumor cells.

The neuroblastoma microenvironment leads to alteredimmune cell frequency and function

To understand the impact of neuroblastoma on immunepopulations in vivo, we first characterized the role of neuroblas-toma-derived arginase in a neuroblastoma murine model. TH-MYCN mice are transgenic mice providing the closest biologicmodel of human neuroblastoma development (16). In patients,the most common location for neuroblastoma primary tumors isthe adrenal gland. We compared the arginase activity of murineadrenals with murine neuroblastoma tumor s and identified thattumor cells have a significantly higher arginase activity (Fig. 3A,P ¼ 0.0004). Using a method described by Zhang and colleagues(17), extracellular fluid was isolated from neuroblastoma tumor

tissue and found to have significantly lower arginine concentra-tions than healthy adrenal tissue (Fig. 3B, P < 0.0001). Similar tohuman tissue, GD2þ cells isolated from murine tumors predom-inantly expressed arginase II; however, low expression of arginaseI is also found (Fig. 3C and Supplementary Fig. S3a). GD2þ cellssuppress T-cell proliferation in an arginase-dependent manner(Fig. 3D). We observed that mice with a larger tumor burden hadsignificantly lower serum arginine concentrations than mice withsmaller tumors (P¼ 0.007) or healthy controls (P¼ 0.0224; largetumors 98mmol/L vs. small tumors 233 mmol/L vs. controls222 mmol/L; Fig. 3E). Consistent with our human findings, noincrease in serum arginase activity was seen (Fig. 3F). Therefore,TH-MYCN mice demonstrate the features of the neuroblastomaarginase-dependentmicroenvironment found in humanpatients.

To understand the impact of neuroblastoma-derived argininedepletion on the immune system, we characterized key immunecompartments in the blood, spleen, and tumors of tumor-bearing

Figure 2.Human neuroblastoma creates alocal and systemic microenvironmentdeficient in arginine. A, GD2þ

neuroblastoma tumor cells havesignificant arginase activity,compared with nontumor cells. B,expression of arginase I, arginase II,and iNOS from GD2þ cells sorted fromthe tumors of 7 patients as examinedby quantitative PCR. Patients areidentified by unique symbols, whichare used consistently throughout thearticle. C, staining of sections fromneuroblastoma tumors at diagnosiswith DAPI alone (top), anti-humanGD2 (middle), and anti-arginase II(bottom). Representative sectionfrom 1 of 5 patients withneuroblastoma shown. D, GD2þ tumorcells suppress T-cell proliferation,which is restored by arginaseinhibition. Data are representativeof five independent experiments.Error bars, SD. E, plasma arginineis significantly lower in 22 newlydiagnosed patients withneuroblastoma compared with22 healthy donors.

Mussai et al.





mice. T-cell frequency in the spleens of tumor-bearing mice weresignificantly decreased (Supplementary Fig. S3b, bottom, P ¼0.03), and T cells represented only a minor population of cells(3.2%) within the tumor mass (Fig. 4A). Myeloid cells are asecond predominant immune population and have been showntopromoteneuroblastoma chemoresistance (18).Within tumors,myeloid populations were composed of F4/80þ macrophages(4.5%), Ly6Cþ (6.7%), and Ly6Gþ cells (2.7%; Fig. 4A). In ourmodel, we identified that tumor-bearing mice also had increasednumbers of F4/80þ myeloid cells in the blood compared withcontrols (Supplementary Fig. S3, top; F4/80, P ¼ 0.0049). In thespleen, all populations of myeloid cells were significantlyincreased compared with controls (Supplementary Fig. S3; F4/80þ, P ¼ 0.001; Ly6Gþ, P ¼ 0.0016; Ly6Cþ, P ¼ 0.04).

It is recognized that myeloid cell activation plays a key role incoordinating the anticancer, proinflammatory immuneresponse through cytokine release (19). To understand whetherthe neuroblastoma microenvironment affects the activation ofmyeloid cells, we isolated the 2 main populations of myeloidcells, Ly6Cþ or F4/80þ, from the spleen and tumors of tumor-bearing TH-MYCN mice and control mice, and tested theirresponse to the potent immune stimulator LPS. Both myeloidpopulations had significantly impaired TNFa release comparedwith those from healthy controls (Fig. 4B and C), comparativeto the effect of myeloid cells cultured in arginine-free condi-tions. No difference in IL10, IL12, or IL6 was identified suggest-ing an inactivation of myeloid function, as opposed to polar-ization (data not shown).

Figure 3The TH-MYC murine neuroblastomarecreates the arginase-dependentmicroenvironment. A, sorted GD2þ

cells from murine neuroblastomatumors have significantly higherarginase activity than adrenal tissue.B, arginine concentrations aresignificantly lower in the extracellularfluid of neuroblastomas than in healthyadrenals. C, expression of arginase I,arginase II, and iNOS of GD2þ cellssorted from murine neuroblastomatumors by quantitative PCR. D, GD2þ

tumor cells from murineneuroblastomas suppress T-cellproliferation, which is restored byarginase inhibition. E, plasma fromtumor-bearing mice with large(>1.5 cm3) and small (<1.5 cm3) tumorswas analyzed for arginineconcentration. Plasma arginine issignificantly lower in tumor-bearingmice with large tumors. F, no increasein plasma arginase activity of tumor-bearing mice compared with healthymice.






These murine studies identify the inhibitory role of the neu-roblastoma arginase microenvironment on myeloid cell activa-tion. To understandwhether the arginine-deficientmicroenviron-ment in human neuroblastoma can also inactivate human mye-loid cell function, CD14þ monocytes from healthy donors werecultured in the presence of neuroblastoma cell supernatant andstimulated overnight with LPS. Analogous to our murine study,neuroblastoma conditioned monocytes had decreased secretionof TNFa (Fig. 5A), and no increase in the T-cell activatory cytokineIL12 (Supplementary Fig. S4). Coculture with arginine-deficientmedia recapitulated these findings, confirming the specificity ofthe mechanism. Furthermore, no increase in the myeloid matu-ration marker CD86 was observed (Supplementary Fig. S4b).Similarly, no upregulation in the M2 polarizationmarker CD206was seen, indicating that the monocytes are inactivated but notpolarized (data not shown) by the low-arginine microenviron-ment of neuroblastoma.

It is well established that tumor-associated myeloid cells maysupport tumor pathogenesis by directly inhibiting T-cell function.Suppressive myeloid cells have been reported in neuroblastoma,although themechanism of their induction and their relevance inpatients has not been well-characterized (20). We co-culturedmonocytes conditioned with arginine-deplete neuroblastomamedia with T cells from healthy donors and found they are ableto significantly suppress T-cell proliferation (Fig. 5B). As neuro-blastoma creates a systemic arginine depletion, we also examinedthemyeloidpopulations in thebloodof newlydiagnosedpatients(Supplementary Fig. S4c). An increase in circulating CD14þ (P ¼0.034), but not CD15þ cells, was found compared with healthycontrols (Fig. 5C, top andmiddle). These CD14þmonocytes frompatients with neuroblastoma were capable of suppressing allo-

genic T-cell proliferation (Fig. 5D). Therefore, both the local andsystemic arginine depletion createdbyneuroblastomaaffect T-cellresponses not only directly but also through the induction ofimmunosuppressive myeloid cells.

Clinical impact of neuroblastoma arginase microenvironmentOur findings highlighted the key role of arginase activity by

neuroblastoma tumors. To understand the clinical significance ofour findings we interrogated the "R2: microarray analysis andvisualization platform." Analysis of primary patient samples atdiagnosis confirms the enzymearginase II is expressedby all stagesof neuroblastoma tumor, with the exception of stage 4s, at higherlevels than arginase I (Supplementary Fig. S5a). High arginase IIexpression correlates with a significantly worse overall survival(Supplementary Fig. S5b). The result suggests a key role forarginase II in neuroblastoma pathogenesis. With regard to mye-loid populations, patients with low intratumoral CD14 expres-sion, suggestive of inactivate myeloid cells, also do significantlyworse (Supplementary Fig. S5c).

T-cell–based immunotherapies commonly require harvestingof T cells frompatients for ex vivomanipulation (21). These studiesassume that T cells from patients are fully functional, but ourstudies suggest that these cells would not be in an optimalcondition. To test this concept, we isolated T cells from patientsat diagnosis and found they have reduced proliferative capacitycompared with those from healthy controls, suggesting they havebeen impaired by the immunosuppressive microenvironmentcreated by the developing neuroblastoma (Fig. 6A). This obser-vation complements ourfinding that T-cell frequency is also lowerin patients with neuroblastoma than in healthy controls (P ¼0.039; Fig. 5C, bottom). The cancer germline antigenNY-ESO-1 is

Figure 4.Neuroblastoma modulates myeloidcell populations in TH-MYC murinemodels. A, neuroblastoma tumorsfrommicewere disassociated, and thepercentage of CD3þ, Ly6Gþ, Ly6Cþ,and F4/80þ cells was determined byflow cytometry. B, F4/80þ cells fromtumors, spleens of tumor-bearingmice, or when cultured in arginine-deplete media had significantdecreases in LPS-driven TNFa release.All data are representative of threeindependent experiments. Error bars,SD. C, Ly6Cþ cells from tumors,spleens of tumor-bearing mice, orwhen cultured in arginine-depletemedia had significant decreases inLPS-driven TNFa release. All data arerepresentative of three independentexperiments. Error bars, SD.

Mussai et al.





an antigenic target on neuroblastoma and has been suggested as atarget for both endogenous immunity and novel T-cell therapies(22). To evaluate the impact of the arginase microenvironmenton engineered T-cell immunity, NY-ESO-1–specific T-cell receptor(TCR)–engineered T cells were conditioned with neuroblastomacell lines ex vivo. Neuroblastoma induced an arginase-dependentinhibition of proliferation by NY-ESO-1 TCR-engineered Tcells, which was rescued by the arginase inhibition (Fig. 6B).Extending these findings, we tested the impact of the neuroblas-toma microenvironment on engineered CAR T cells targetingGD2. We demonstrate that neuroblastoma leads to a significantimpairment in cytotoxicity and proliferation of CAR T cells(Fig. 6C and Supplementary Fig. S4d), with some rescue incytotoxicity when arginase is inhibited (Fig. 6D). These findingsprovide the first evidence for the role of neuroblastoma arginaseactivity in impairing antineuroblastoma T-cell therapies.

Patients with neuroblastoma, particularly those withadvanced disease, can present with cytopenias. However, themechanism by which neuroblastoma can impact on hemato-poietic progenitors has never been reported. As neuroblastomalowers plasma arginine, we hypothesized that the low-argininemicroenvironment could also inhibit the proliferation of bonemarrow hematopoietic progenitors. Human CD34þ hemato-poietic stem cells (HSC) have significantly impaired cell divi-sion, with no loss of CD34þ expression, in the presenceof neuroblastoma patient plasma (red histogram) comparedwith those cultured in plasma from healthy controls (bluehistogram; Fig. 6E). CD34þ cells cocultured with tumor cellsfrom patients or with arginine-deplete media had similarsignificant decreases in proliferation, compared with controls(Fig. 6F). Therefore, the arginase-dependent microenviron-ment of neuroblastoma not only suppresses T cells and

Figure 5.Neuroblastoma induces immunosuppressivemyeloid cells in patients. A, CD14þ cells cultured withneuroblastoma or arginine-deplete media hadsignificant decreases in LPS-driven TNFa release. Alldata are representative of three independentexperiments. Error bars, SD. B, neuroblastoma-conditioned monocytes suppress T-cell proliferation.Data are representative of three independentexperiments. C, percentages of CD11bþCD14þ (top),CD11bþCD15þ (middle), and CD3þ (bottom) in theblood of newly diagnosed patientswith neuroblastomacompared with healthy donors, as measured by flowcytometry. D, neuroblastoma patients' monocytessuppress T-cell proliferation.






myeloid cells but also inhibits the division and proliferation ofCD34þ HSCs.

DiscussionIn this study, we identify for the first time the ability of primary

human neuroblastoma cells to directly inhibit T-cell proliferationthrough high arginase activity. We showed that GD2þ neuroblas-toma cells from patients express predominantly the arginase IIisoform. Arginase I and II are localized to the cytosol and mito-chondria, respectively (23). In healthy states, arginase I is pre-dominantly expressed by hepatocytes, whereas arginase II haswider, tissue-specific expression (24). Both isoforms catalyze theconversion of arginine into ornithine and urea. Neuroblastoma isa malignancy of the sympathetic nervous system, arising fromneural crest progenitors that ordinarily develop into sympatheticganglia and adrenal medulla (25, 26). Arginase II is the principalisoform expressed in the nervous system, with little or absent

arginase I protein, consistent with our findings (27). Indeedarginine metabolism plays a key role in both healthy neuronalfunction and diseases of the nervous system and may contributeto the relative immunoprivileged niche that the nervous systemexists in (28).

Using the "R2:microarray analysis and visualization platform,"we confirmed that arginase II was expressed by all stages ofneuroblastoma suggestive of a fundamental role in neuroblasto-ma pathogenesis (29). Stage 4s neuroblastoma is associated withan excellent prognosis, and it is interesting that in this subtype,arginase II gene expression is not higher than arginase I. Althoughthe role of arginase I–expressingmyeloid-derived suppressor cells(MDSC) in altering T-cell responses in patients with cancer hasbeen well-established, arginase II in solid tumor immune biologyhas only received limited attention to date (30). We recentlyreported the role of arginase II in creating an immunosuppressivemicroenvironment in AML (10). Similarly, in prostate cancer,arginase II expression by both the tumor cells and cancer-

Figure 6.Neuroblastoma arginase activityimpair T-cell immunotherapies andhematopoietic stem cell division.A, T cells from patients withneuroblastoma at diagnosis haddecreased proliferative potentialcompared with those from healthydonors. B, neuroblastomaconditioning suppresses NY-ESO-1TCR-engineered T-cell proliferation.Data representative of threeindependent experiments. C,neuroblastoma conditioningsuppresses anti-GD2 CAR T-cellcytotoxicity against neuroblastoma.Data representative of twoindependent experiments.D, arginase inhibition duringneuroblastoma conditioning rescuesanti-GD2CART-cell cytotoxicity. Datarepresentative of two independentexperiments. E, neuroblastomapatient plasma suppresses humanCD34þ HSC proliferation. Data arerepresentative of five patientplasma experiments. Independentexperiments were performed on twoseparate occasions. F, CD34þ HSCproliferation is inhibited by theneuroblastoma low-argininemicroenvironment. Independentexperiments were performed on twoseparate occasions.

Mussai et al.





associatedfibroblasts serves to regulate immunosuppressive path-ways (31). The increased expression and immunosuppressivepotential of arginase II alone is likely to be tissue- and disease-specific (32).

We examined the impact of neuroblastoma arginase activityto understand the effect on both local and systemic immunities.We first showed that neuroblastoma tumors lower local argi-nine concentrations to levels that inactivate surrounding T andmyeloid cells. Increased arginase II expression correlates withdecreased T and myeloid cell tumor infiltration in head andneck squamous cell carcinoma, supporting our findings inneuroblastoma (33). Arginine is exchanged between the bloodand extracellular fluid to maintain homeostasis (34, 35). Weinvestigated plasma levels of arginine in newly diagnosedpatients with neuroblastoma and found them to be significant-ly lower than healthy controls. Similar changes were identifiedin mice with larger tumors. However, unlike AML, we showhere that neuroblastoma does not release free arginase IIenzyme. Therefore, uptake and catabolism by the tumor massmust presumably play a significant role in systemic depletion.Patients with high-stage neuroblastoma frequently present withsignificant cachexia (36). As tumor consumption of argininelowers systemic arginine concentrations, physiologic compen-sation through protein breakdown and arginine recycling viathe intestinal–renal axis may seek to maintain homeostasis(37). Corresponding decreases in plasma arginine have beendescribed in patients with renal cell carcinoma and cervicalcancer at diagnosis (38, 39).

Arginine is a semiessential amino acid required by healthytissues for a number of cell processes, including cell viability,proliferation, and protein synthesis. Thus, arginase activity likelyprovides 2 key roles in neuroblastoma pathogenesis—first main-taining normal cell metabolism but second contributing toimmune escape (40, 41). It is established that arginine depletionleads to T cell-cycle arrest, impaired proliferation, and reducedactivation (42, 43). This is consistent with our findings of bothlower T-cell numbers in the peripheral blood and their impairedproliferative capacity. The "R2: microarray analysis and visuali-zation platform" identifies that neuroblastoma tumors with lowCD3 expression have a worse overall survival, supporting thepathogenic role of arginine depletion in neuroblastoma tumors.

Monocytes and macrophages play a key role in anticancerimmunity through coordination of other immune effectors bycytokine secretion. Consistent with previous findings, we identifythat CD14þ myeloid cells are present within the tumor microen-vironment (20, 44). However, we describe for the first time thatneuroblastoma arginase activity impairs the ability of monocytesto respond to an inflammatory stimulus and decreases theirproduction of the immune activatory cytokine TNFa from mye-loid cells. A previous study noted an impaired TNFa-drivendendritic cell stimulation of T cells in neuroblastoma but nomechanism was found (9).

In both our murine model and in patients' blood, we observeda significant increase in the number of immunosuppressivemonocytes. It has been shown in xenografts that immunosup-pressive myeloid cells are associated with neuroblastoma but themechanism of cross-talk has not been identified (45). Here, weidentify that neuroblastoma arginase activity plays a key role inmodulating the CD14 immune phenotype. The clinical finding ofcytopenias in patients with neuroblastoma is also well-recog-nized. However, the mechanism of neuroblastoma-induced cyto-

penia is unclear, particularly in metastatic or end-stage diseasewhere the bone marrow is rarely significantly replaced by tumorcells. We show for the first time that the low-arginine microen-vironment of neuroblastoma suppresses CD34þ hematopoieticcell division and proliferation.

Immunotherapy requiring an active innate and adaptiveimmune response has become a major new approach in thetreatment of high-risk neuroblastoma. Despite a promisingpreclinical rationale, immunotherapy clinical trials in patientswith neuroblastoma have not resulted in significant improve-ments in overall survival (6). Studies show an unexplaineddecrease in adoptive T-cell numbers after administration, withno correlation between the dose of engineered cells adminis-tered, their numbers in peripheral blood or antitumor response(5). In xenograft studies of neuroblastoma cell vaccines, Tcells from tumor-bearing mice have defective antitumorimmune responses, both locally at the site of injection andsystemically, although no mechanism has been identified (46).Furthermore, immunotherapies that have activity alone in vitrooften require co-administration with cytokines to maximizetherapeutic effect in patients. Early-phase clinical trials usingautologous patient-derived dendritic cell vaccines illustratethat moderate benefit is only seen when co-administered withIL2 (47).

NY-ESO-1 is a cancer germline antigen known to be highlyimmunogenic in patients with melanoma and also is expressedby the majority of neuroblastomas. Although it has beendemonstrated through ex vivo manipulation that patients cangenerate a humoral and T-cell–specific response to this antigen,it is unclear why similar responses are not seen in patients (48).We show that neuroblastoma arginase activity significantlyimpairs antigen-specific T-cell proliferation by NY-ESO-1TCR-engineered T cells.

Following the identification of neuroblastoma surface gangli-oside D2 as a potential therapeutic target, CAR-engineered T cellsare an alternative therapeutic approach under development.However, similar to vaccine approaches, early-phase trial resultshave not matched preclinical results. We find that the arginase-dependent microenvironment created by neuroblastoma cansignificantly inhibit the proliferation and cytotoxicity of anti-GD2CAR T cells. This is the first report identifying a mechanism inwhich CAR T-cell activity can be impaired by tumor immuno-suppression. Interestingly, a previous report assessing CAR-engi-neered T-cell responses against a neuroblastoma cell line in vitroidentified a similar decrease in activity but nomechanism for thiswas identified (49). Most early trials of novel immune therapiesenroll refractory or relapsed patients who may have very signif-icant disease burden, due to treatment failure. It is thereforeunsurprising that the extent of the immunosuppressive microen-vironment both within the tumor mass and blood has the abilityto hamper adoptive T-cell responses.

Arginine supplementation has been administered to patients,in nontumor settings, to enhance immunity and tissue repair. In asubcutaneous xenograft model of neuroblastoma supplementedwith combination arginine and IL2, lymphokine-activated killercells demonstrated increased antineuroblastoma immunity andmice had prolonged survival (50). However, concerns that argi-nine supplementation may feed tumor growth would need to beevaluated further. Therapeutic targeting through arginase inhibi-tion is a more likely strategy. Although small molecules such asNOHAhave been shown to be toxic in vivo, other compounds that






act on arginase are already under preclinical evaluation. A recentstudy identified that the green tea derivative polyPhenon E canpromote antitumor immunity in a murine model of neuroblas-toma. Polyphenols have been shown to have activity againstarginase. It is possible that the activity previously reported mayhave been also due to blockade of arginase expressed by neuro-blastoma cells.

In conclusion, our findings provide evidence for the key role ofarginase activity in the creation of an immunosuppressive micro-environment in neuroblastoma and have significant clinicalimplications for T-cell immunotherapy approaches. Tumorescape from different arms of the immune response is likely tobe multifactorial, and control of arginase activity by tumor tissuefits rationallywith othermechanisms identified inmurinemodels(48). Targeting of arginase activity in neuroblastoma could pro-vide a new way of enhancing both autologous and therapeuticantineuroblastoma immunity.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception anddesign: F.Mussai, K. Petrie, J. Anderson, L. Chesler, C.De SantoDevelopment of methodology: F. Mussai, S. Egan, R. Wheat, L. Chesler,C. De Santo

Acquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): F. Mussai, S. Egan, S. Hunter, H. Webber, J. Fisher,R. Wheat, C. McConville, Y. Sbirkov, K. Wheeler, G. Bendle, J. Anderson,L. Chesler, C. De SantoAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): F. Mussai, S. Hunter, J. Fisher, R. Wheat, K. Petrie,L. Chesler, C. De SantoWriting, review, and/or revision of the manuscript: F. Mussai, S. Egan,H. Webber, J. Fisher, C. McConville, K. Wheeler, J. Anderson, L. Chesler,C. De SantoAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): F. Mussai, C. De SantoStudy supervision: F. Mussai, K. Petrie, L. Chesler, C. De Santo

AcknowledgmentsThe authors thank the patients and parents who contributed samples to the

study. Thank you to Jennie Godwin, Jane Cooper, and Cay Shakespeare forconsent and collection of patient samples.

Grant SupportThis work was supported by Cancer Research UK and the Birmingham

Children's Hospital Research Fund.The costs of publication of this articlewere defrayed inpart by the payment of

page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received November 25, 2014; revised May 8, 2015; accepted May 9, 2015;published OnlineFirst June 8, 2015.

References1. Maris J. The biologic basis for neuroblastoma heterogeneity and risk

stratification. Curr Opin Pediat 2005;17:7–13.2. Rudnick E, Khakoo Y, Antunes N, Seeger R, Brodeur G, Shimada H,

et al. Opsoclonus-myoclonus-ataxia syndrome in neuroblastoma:clinical outcome and anti-neuronal antibodies-a report fromthe Children's Cancer Group Study. Med Pediatr Oncol 2001;36:612–22.

3. Hero B, Simon T, Spitz R, Ernestus K, Gnekow A, Scheel-Walter H, et al.Localized infant neuroblastomas often show spontaneous regression:results of the prospective trials NB95-S and NB97. J Clin Oncol 2008;26:1504–10.

4. Mujoo K, Kipps T, Yang H, Cheresh D, Wargalla U, Sander D, et al.Functional properties and effect on growth suppression of human neuro-blastoma tumours by isotype switch variants of monoclonal antiganglio-side GD2 antibody 14.18. Cancer Res 1989;49:2857–61.

5. PuleM, Savoldo B,MyersG, Rossig C, Russell H,Doti G, et al. Virus-specificT cells engineered to coexpress tumour-specific receptors: persistence andanti-tumour activity in individuals with neuroblastoma. Nat Med 2008;14:1264–70.

6. Simon T, Hero B, Faldum A, Handgretinger R, Schrappe M, NierthammerD, et al. Consolidation treatment with chimeric anti-GD2-antibody ch1.18in children older than 1 year with metastatic neuroblastoma. J Clin Oncol2004;22:3549–57.

7. Yu AL, Gilman AL, Ozkaynak MF, London WB, Kreissman SG, Chen HX,et al. Anti-GD2 antibody with GM-CSF, interlukin-2, and isotretinoin forneuroblastoma. N Engl J Med 2010;363:1324–34.

8. Pistoia V, Morandi F, Bianchi G, Bezzolo A, Prigione I, Raffaghello L.Immunosuppressive microenvironment in neuroblastoma. Front Oncol2013;3:167.

9. Chen X, Doffek K, Sugg SL, Shilyansky J. Neuroblastoma cells inhibit theimmunostimulatory function of dendritic cells. J Pediatric Surg 2003;38:901–5.

10. Mussai F, De Santo C, Abu-Dayyeh I, Booth S,Quek L,McEwen-Smith RM ,et al. Acute Myeloid leukaemia creates an arginase dependent immuno-suppressive microenvironment. Blood 2013;122:749–58.

11. Grody W, Argyle C, Kern RM, Dizikes GJ, Spector EB, Strickland AD , et al.Differential expression of the two human arginase genes in hyperargini-nemia. J Clin Invest 1989;83:602–9.

12. Serafini P, Meckel K, Kelso M, Noonan K, Califano J, Koch W , et al.Phosphodiesterase-5-inhibition augments endogenous antitumor immu-nity by reducing myeloid-derived suppressor cell function. J Exp Med2006;27:2691–02.

13. Gotoh T, Araki M, Mori M. Chromosomal localization of the humanarginase II gene and tissue distribution of its mRNA. Biochem Biophys ResCommun 1997;233:487–91

14. Quinn JJ, Altman A. The multiple hematologic manifestations of neuro-blastoma. Am J Pediatr Hematol Oncol 1979;1:201–5.

15. Scott JP,Morgan E. Coagulopathy of disseminated neuroblastoma. J Pediat1983;103:219–222.

16. Terrile M, Bryan K, Vaughan L, Hallsworth A, Webber H, Chesler L, et al.miRNA expression profiling of the murine TH-MYCN neuroblastomamodel reveals similarities with human tumors and identifies novel can-didate miRNAs. PLoS One 2011;6:e28356.

17. Zhang Y, Xiang L, Hassan R, Pastan I. Immunotoxin and Taxol synergyresults from a decrease in shed mesothelin levels in the extracellular spaceof tumours. Proc Natl Acad Sci USA 2007;104:17099–104.

18. Song L, Asgharzadeh S, Salo J, Engell K, Hong-wei W, Sposto Ret al. Va24-invariant NKT cells mediate antitumor activity via killing of tumor-asso-ciated macrophages. J Clin Inv 2009;119:1524–36.

19. Liu D, Song,Wei J, Courney A, Gao X,Marinova E, et al. IL-15 protects NKTcells from inhibition by tumor-associated macrophages and enhancesantimetastatic activity. J Clin Invest 2012;12:2221–33.

20. Asgharzadeh S, Salo JA, Oberthuer A, Fischer M, Bethold F, Hadjidaniel M ,et al. Clinical significance of tumor-associated inflammatory cells inmetastatic neuroblastoma. J Clin Oncol 2012;30:3525–32.

21. Grupp S, Prak EL, Boyer J, McDonald K, Shusterman S, Thompson E , et al.Adoptive transfer of autologous T cells improves T-cell repertoire diversityand long-term B cell function in pediatric patients with neuroblastoma.Clin Cancer Res 2012;18:6732–41.

22. Rodolfo M, Luksh R, Stockert E, Chen YT, Collini P, Ranzani T , et al.Antigen-specific immunity in neuroblastoma patients: antibody and T-cellrecognition of NY-ESO-1 tumor antigen. Cancer Res 2003;63:6948–55.

23. JenkinsonC, GrodyW, CederbaumS. Comparative properties of arginases.Comp Biochem Physiol B Biochem Mol Biol 1996;114:107–32.

24. Morris SM Jr. Recent advances in argininemetabolism: roles and regulationof the arginases. Br J Pharm 2009;15:922–30.

Mussai et al.





25. Huber K. The sympathoadrenal cell lineage: specification, diversification,and new perspectives. Dev Biol 2006;298:335–43.

26. De Preter K, Vandesompele J, Heiann P, Yigit N, Beckman S, Schramm A ,et al. Human fetal neuroblast and neuroblastoma transcriptome analysisconfirms neuroblast origin and highlights neuroblastoma candidate genes.Genome Biol 2006;7:R84.

27. BraissantO,Gotoh T, LoupM,MoriM, BackmannC. L-arginine uptake, thecitrulline-NO cycle and arginase II in the rat brain: an in situ hybridisationstudy. Brain Res Mol Brain Res 1999;70:231–41.

28. Wiesinger H. Arginine metabolism and the synthesis of nitric oxide in thenervous system. Prog Neurobiol 2001;64:365–91.

29. R2:microarray analysis and visualization platform'. Available from: http://r2.amc.nl.

30. Mussai F, De Santo C, Cerundolo V. Interaction between iNKT cells andMDSCs in cancer patients: Evidence and Therapeutic Opportunities.J Immunother 2012;35:449–59.

31. Ino Y, Yamazaki-Itoh R, Oguro S, Shimada K, Kosuge T, Zavada J, et al.Arginase II expressed in cancer-associated fibroblasts indicates tissue hyp-oxia and predicts poor outcome in patients with pancreatic cancer. PLoSOne 2013;8:e55146.

32. RotondoR,Mastracci L, Piazza T, BarisioneG, FabbiM,CassanelloM , et al.Arginase 2 is expressed by human lung cancer, but it neither inducesimmune suppression, nor affects disease progression. Int J Cancer 2008;123:1108–16.

33. Bron L, Jandus C, Andrejevic-Blant S, Speiser DE, Monnier P, Romero P.Rivals JP. Prognostic value of arginase-II expression and regulatory T-cellinfiltration in head and neck squamous cell carcinoma. Int J Cancer 2013;132:E85–93.

34. Borsheim E, Kobayashi H, Traber DL, Wolfe RR. Compartmentaldistribution of amino acids during hemodialysis-induced hypoami-noacidemeina. Am J Physiology Endocrinol Metab 2006;290:E643–E652.

35. Lofberg E, Essen P, McNurlan M, Wernerman J, Garlick P, Anderstam B ,et al. Effect of hemodialysis on protein synthesis. Clin Nephrol 2000;54:284–94.

36. Green GJ, Weitzmann SS, Pencharz PB. Resting energy expenditure inchildren newly diagnosed with stage IV neuroblastoma. Pediatr Res 2008;63:332–36.

37. Buijs N, Luttikhold J, Houdijk AP, van Leeuwen PA. The role of a disturbedarginine/NO metabolism in the onset of cancer cachexia: a workinghypothesis. Curr Med Chem 2012;19:5278–86.

38. HasimA, Aili A,Maimaiti A,Mamtimin B, Abudula A, UpurH. Plasma-freeamino acid profiling of cervical cancer and cervical intraepithelial neopla-sia patients and its application for early detection. Mol Biol Rep.2013;40:5853–59.

39. Rodriguez P, Ernstoff M, Hernandez C, Atkins M, Zabaleta J, Sierra R , et al.Arginase I-producing myeloid derived suppressor cells in renal cell carci-noma are a subpopulation of activated granulocytes. Cancer Res 2009;69:1553–60.

40. Lin SE, Wu Fl, Wei MF, Shen LJ. Depletion of arginine by recombinantarginine deiminase induces nNOS-activated neurotoxicity in neuroblas-toma cells. Biomed Res Int 2014;2014:589424.

41. Morris SMJr. Arginine metabolism: boundaries of our knowledge. J Nutr2007;137:1602S–1609S.

42. Rodriguez PC, Quiceno DG, Ochoa AC. L-arginine availability regulates T-lymphocyte cell-cycle progression. Blood 2007;109:1568–73.

43. Zea AH, Rodriguez PC, Culotta KS, Hernandez CP, DeSalvo J, Ochoa JB ,et al. L-arginine modulates CD3zeta expression and T cell function inactivated human T lymphocytes. Cell Immunol. 2004;232:21–31.

44. Apps JR, Hasan F, Campus O, Behjati S, Jacques TS, Sebire N , et al. Theimmune environment of paediatric and solidmalignancies: evidence froman immunohistochemical study of clinical cases. Fetal Pediatr Pathol2013;32:298–07.

45. Santilli G, Piotrowska I, Cantilena SA, Chayka O, D'Alicarnasso M, Mor-genstern DA, et al. Polyphenon E enhances the antitumor immuneresponse in neuroblastoma by inactivating myeloid suppressor cells. ClinCancer Res 2013;19:1116–25.

46. Barr KM, JingW,HallettWH,Gershan JA, Johnson BD. Examining T cells atvaccine sites of tumor-bearing hosts provides insights to dysfunctionalT-cell immunity. J Immunother 2013;36:41–51.

47. Bowman L, GrossmannM, Rill D, BrownM, ZhongWY, Alexander B , et al.IL-2 adenovector transduced autologous tumor cells induce antitumorimmune responses in patients with neuroblastoma. Blood 1998;92:1941–49.

48. Rodolfo M, Luksch R, Stockert E, Chen YT, Collini P, Ranzani T , et al.Antigen-specific immunity in neuroblastoma patients: antibody and T-cellrecognition of NY-ESO-1 tumour antigen. Cancer Res 2003;63:6948–55.

49. Sun J, Dotti G, Huye LE, Foster AE, Savoldo B, Gramatges MM, et al. T cellsexpressing constitutively active Akt resist multiple tumor-associated inhib-itory mechanisms. Mol Ther 2010;18:2006–17.

50. LiebermanM,Nishioka K, RedmondH,Daly J. Enhancement of interlukin-2 immunotherapy with L-arginine. Ann Surg 1992;215:157–65.






2015;75:3043-3053. Published OnlineFirst June 8, 2015.Cancer Res Francis Mussai, Sharon Egan, Stuart Hunter, et al. ImmunityMicroenvironment That Impairs Autologous and Engineered Neuroblastoma Arginase Activity Creates an Immunosuppressive

Updated version

10.1158/0008-5472.CAN-14-3443doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2015/06/09/0008-5472.CAN-14-3443.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/75/15/3043.full#ref-list-1

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

Citing articles

http://cancerres.aacrjournals.org/content/75/15/3043.full#related-urls

This article has been cited by 3 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://cancerres.aacrjournals.org/content/75/15/3043To 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-14-3443

http://cancerres.aacrjournals.org/content/suppl/2015/06/09/0008-5472.CAN-14-3443.DC1

http://cancerres.aacrjournals.org/content/75/15/3043.full#ref-list-1

http://cancerres.aacrjournals.org/content/75/15/3043.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/75/15/3043


neuroblastoma arginase activity creates an...

Documents