Inhibition of CD47 Effectively Targets

Published OnlineFirst February 23, 2015; DOI: 10.1158/1078-0432.CCR-14-1399

Cancer Therapy: Preclinical Clinical Cancer Research Inhibition of CD47 Effectively Targets Pancreatic Cancer Stem Cells via Dual Mechanisms Michele Ciofﬁ1, Sara Trabulo1,2, Manuel Hidalgo3, Eithne Costello4, William Greenhalf4, Mert Erkan5,6, Joerg Kleeff5, Bruno Sainz Jr1, and Christopher Heeschen1,2

Abstract

Purpose: Pancreatic ductal adenocarcinoma (PDAC) is a Results: CD47 was highly expressed on CSCs, but not on other cancer of the exocrine pancreas with unmet medical need and nonmalignant cells in the pancreas. Targeting CD47 efﬁciently is strongly promoted by tumor-associated macrophages enhanced phagocytosis of a representative set of primary human (TAM). The presence of TAMs is associated with poor clinical pancreatic cancer (stem) cells and, even more intriguingly, also outcome, and their overall role, therefore, appears to be directly induced their apoptosis in the absence of macrophages protumorigenic. The "don't eat me" signal CD47 on cancer during long-term inhibition of CD47. In patient-derived xeno- cells communicates to the signal regulatory protein-a on graft models, CD47 targeting alone did not result in relevant macrophages and prevents their phagocytosis. Thus, inhibition slowing of tumor growth, but the addition of gemcitabine or of CD47 may offer a new opportunity to turn TAMs against Abraxane resulted in sustained tumor regression and prevention PDAC cells, including cancer stem cells (CSC), as the exclu- of disease relapse long after discontinuation of treatment. sively tumorigenic population. Conclusions: These data are consistent with efﬁcient in vivo Experimental Design: We studied in vitro and in vivo the effects targeting of CSCs, and strongly suggest that CD47 inhibition of CD47 inhibition on CSCs using a large set of primary pancreatic could be a novel adjuvant treatment strategy for PDAC indepen- cancer (stem) cells as well as xenografts of primary human PDAC dent of underlying and highly variable driver mutations. Clin tissue. Cancer Res; 1–13. 2015 AACR.

Introduction (7), convincing evidence has emerged for several solid tumors indicating that like adult tissues, tumors are sustained and pro- Pancreatic ductal adenocarcinoma (PDAC) remains one of the moted by cells that exhibit features of stem cells, including most devastating cancers with a 5-year survival rate of less than 5% unlimited self-renewal (8). We and others have provided conclu- (1). Despite expanding research activities, there has been little sive evidence for a hierarchical organization in human PDAC and, therapeutic progress toward improving patients' long-term surviv- even more importantly, demonstrated that pancreatic CSCs, at the al. Gemcitabine (2), FOLFIRINOX (3), and more recently the apex of the hierarchy, have exclusive tumorigenic and metastatic addition of nab-paclitaxel (Abraxane; ref. 4) are able to moderately potential and are inherently resistant to chemotherapy (9–11). extend median survival, but eventually the vast majority of Indeed, the survival of such resistant CSCs during chemotherapy, patients still succumb from progressive disease. Therefore, devel- despite initial tumor regression, represents a plausible explanation oping new and more effective anti-PDAC treatments represent an for the later fatal relapse of disease in most patients (3, 4). urgent and unmet medical need (5, 6). Because the cancer stem cell Studies based on the inhibition of regulatory pathways that are (CSC) hypothesis was functionally validated for leukemia in 1994 crucially relevant for the self-renewal capacity of CSCs are promising (12, 13); however, the overly heterogeneous genetic back- ground of PDAC may render larger populations of cells resistant 1Stem Cells and Cancer Group, Molecular Pathology Programme, Spanish National Cancer Research Centre (CNIO), Madrid, Spain. to the targeting of single pathways. Consequently, we asked 2Centre for Stem Cells in Cancer and Ageing, Barts Cancer Institute, whether targeting pancreatic CSCs with broader immune-based A CR-UK Centre of Excellence, Queen Mary University of London, therapeutic approaches could represent a more viable and potent United Kingdom. 3Gastrointestinal Cancer Clinical Research Unit, Clin- ical Research Programme, CNIO, Madrid, Spain. 4Liverpool Cancer alternative for eliminating these highly tumorigenic and chemore- Research UK Centre, University of Liverpool, Liverpool, United King- sistant cells. Macrophages play crucial roles in adaptive and innate dom. 5Department of Surgery, Technical University Munich, Munich, immunity. In PDAC, tumor-associated macrophages (TAM) rep- 6 Germany. Koc University School of Medicine, Instanbul, Turkey. resent the major immune cell type present in the PDAC tumor Note: Supplementary data for this article are available at Clinical Cancer microenvironment (14), and these cells are believed to drive Research Online (http://clincancerres.aacrjournals.org/). cancer progression, presumably via promoting cancer cell prolif- Corresponding Author: Christopher Heeschen, Centre for Stem Cells in Cancer eration, tumor angiogenesis, extracellular matrix breakdown, and and Ageing, Barts Cancer Institute, Queen Mary University of London, Charter- subsequently tumor invasion and metastasis (15, 16). In addi- house Square, London EC1M 6BQ, United Kingdom. Phone: 44-0-20-7882-8201; tion, CD47, a transmembrane protein expressed on many cancer Fax: 44-0-20-7882-3885; E-mail: [email protected]. cells, serves as a ligand to signal regulatory protein-a (SIRPa), doi: 10.1158/1078-0432.CCR-14-1399 a molecule expressed on macrophages (17), resulting in the 2015 American Association for Cancer Research. inhibition of phagocytosis by macrophages through a signaling

www.aacrjournals.org OF1

Ciofﬁ et al.

picked, pooled, and further expanded to heterogeneous primary Translational Relevance cancer cell cultures (AAU77G and CHX6). Pancreatic ductal adenocarcinoma (PDAC) remains one of Human peripheral blood–derived mononuclear cells were the most devastating cancers, and very few new treatments obtained from healthy donors with informed consent. Mono- have revealed meaningful improvements in patient survival cyte-derived macrophage cultures were established in IMDM over the past decades. On the basis of our previous work supplemented with 10% human AB serum as previously demonstrating the existence of cancer stem cells (CSC) in described (25). Sixty ng/mL GM-CSF or M-CSF (R&D Systems) pancreatic cancer and their strong resistance to standard che- was added to the cultures to generate M1 and M2 monocyte– motherapy, we now provide multiple lines of functional and derived macrophages, respectively (26). Murine monocytes were mechanistic evidence for a treatment regimen, including inhi- isolated from mechanically disrupted spleens, passed through a bition of CD47 targeting both CSCs as well as their more 40-mm mesh ﬁlter, and differentiated into macrophages under differentiated progenies. Therefore, this new therapeutic strat- adherent conditions on non-tissue culture-treated 100-mm dishes egy should be further explored in the clinical setting as its in RPMI supplemented with 10% FBS and 10 ng/mL of murine M- success bears the potential to improve the poor prognosis of CSF (PeproTech). To generate M1- and M2-polarized murine patients with PDAC. macrophages, 10 ng/mL of IFNg (PeproTech), and LPS (Sigma; M1) or 10 ng/mL IL4 (M2; PeproTech) were added to the cultures.

Sphere formation assay 4 cascade mediated via phosphorylation of the immunoreceptor Spheres were generated by culturing 2 10 pancreatic cancer tyrosine-based inhibitory motif present on the cytoplasmic tail of cells in suspension in serum-free DMEM/F12 supplemented with – SIRPa (18). B27 (1:50; Invitrogen), 20 ng/mL bFGF, and 50 U/mL penicillin Previous work in preclinical models of bladder cancer, leuke- streptomycin for a total of 7 days, allowing spheres to reach a size mia, and lymphoma demonstrated that inhibiting the interaction of >75 mm. For serial passaging, 7-day-old spheres were retained between CD47 and SIRPa using anti-CD47 mAbs allows for using 40-mm cell strainers, dissociated into single cells, and then increased phagocytosis of cancer cells in vitro and decreased tumor recultured for 7 additional days as previously described (13). burden in vivo (19–21). Recently, CD47 was shown to be preferentially expressed in liver CSCs and inhibition of CD47 sup- RNA preparation and quantitative real-time PCR pressed growth of hepatocellular carcinoma xenografts and had a Total RNAs from human primary pancreatic cancer cells and chemosensitizing effect (22), suggesting that CD47 may also be a spheres were extracted with TRizol (Life Technologies) according promising therapeutic target for hepatocellular CSCs. Here, we to the manufacturer's instructions. One microgram of total RNA now demonstrate that CD47 is also highly expressed on pancre- was used for cDNA synthesis with SuperScript II reverse transcrip- atic cancer (stem) cells and that anti-CD47 mAbs did not only tase (LifeTechnologies) and random hexamers. Quantitative real- enable phagocytosis of these cells by macrophages, but also time PCR was performed using SYBR Green PCR master mix directly induced apoptosis of the cancer (stem) cells, while (LifeTechnologies), according to the manufacturer's instructions. exerting no effect on nonmalignant cells. Although CD47 target- Primers sequences used are: ing as a single treatment strategy was not effective in vivo, the combination with chemotherapy resulted in long-lasting tumor ACTIN: Forward—GCGAGCACAGAGCCTCGCCTT regression. Thus, our results demonstrate that targeting CD47 on Reverse—CATCATCCATGGTGAGCTGGCGG PDAC cells changes the behavior of resident TAMs to inhibit, CD47: Forward—GCGATTGGATTAACCTCCTTCGTCA rather than promote tumor growth, and thus may evolve as a Reverse—CCATGCATTGGTATACACGCCGC potent addition to our still sparse armamentarium against PDAC.

Materials and Methods Flow cytometry 6 Primary human and mouse pancreatic cancer cells and Cells were adjusted to a concentration of 10 cells/mL in sorting macrophages buffer [1X PBS; 3% FBS (v/v); 3 mmol/L EDTA (v/v)] before Human PDAC tissues were obtained with written informed analysis or sorting with a FACS Canto II or FACS Inﬂux instru- consent from all patients and expanded in vivo as patient-derived ment, respectively (BD Biosciences). To identify distinct cancer xenografts (PDX), as previously described (12). For in vitro studies, (stem) cells, the following antibodies were used: anti–CD133/1- PDX tissue fragments were minced, enzymatically digested with APC or PE (Miltenyi Biotec); CD47-APC, CXCR4-APC, SSEA-1- collagenase (STEMCELL Technologies) for 90 minutes at 37C APC, or appropriate isotype-matched control antibodies (all from (12) and after centrifugation for 5 minutes at 1,200 rpm cell BD Biosciences). DAPI was used for exclusion of dead cells. Data pellets were resuspended and cultured in RPMI supplemented were analyzed with FlowJo 9.2 software (Tree Star). For the with 10% FBS and 50 U/mL penicillin–streptomycin. These assessment of apoptosis, cells were incubated with the DAPI and primary cultures were used in vitro only until passage 10. Murine Annexin V FITC Staining Kit (BD Biosciences) according to the þ PDAC cells were derived from the K-Ras /LSL-G12D;Trp53LSL- manufacturer's instructions. R172H;PDX1-Cre mice (KPC) as a model of advanced PDAC (23). KPC-derived tumors were minced, mechanically (gentle- Antibody preparation MACS Dissociator; Miltenyi) and enzymatically dissociated with The anti-hCD47 (B6H12) hybridoma was obtained from the collagenase (STEMCELL Technologies), and subsequently cul- ATCC. Hybridoma cells were cultured using previously described tured in vitro as previously detailed (24). Epithelial clones were conditions (27) and antibodies were puriﬁed by protein G.

OF2 Clin Cancer Res; 2015 Clinical Cancer Research

CD47 Targets Pancreatic Cancer Stem Cells

AB 4 P < 0.001

3 10x 20x 10x 20x 10x 20x P < 0.05 Normal pancreasPancreatitis PDAC

10x 20x 10x 20x 10x 20x

PDAC Metastasis Metastasis Mean intensity relative 1

10x 20x 10x 20x 10x 20x 0 Normal Pancreatitis PDAC Metastasis Xenograft Xenograft Xenograft

CDAdherent Spheres 80 * Nontumor PDAC 100% *

80% 60 *

60% 40

40%

20 20% CD47 mRNA expression (absolute values) (absolute expression CD47 mRNA 0 0% CD47 surface expression (% cells) of positive (% expression CD47 surface Norm1 Norm2 PSC JH029 247 198 253 215 354 163 185 A6L A6L 185 354 215 253 *P < 0.05 E A6L 185 354 100% 0.62 4.27 0.618 2.95 9.52 23.4 5 5 5 10 87.3% 10 82.7% 10 71.1%

4 4 10 10 104 3 3 3 3 D1 C 3 3 3 80% A: 10 10 10 PE- PE-A: CD133 PE-A: CD1 cells + 2 2 10 10 102

0 0 0 17.1 78 29.2 67.2 24.6 42.5

2 3 4 5 2 3 4 5 010 10 10 10 010 10 10 10 0102 103 104 105 60% APC-A: CD47 APC-A: CD47 APC-A: CD47

CD133 215 253

1.35 4.91 0.289 12.3 5 5 10 78.4% 10 97.7% 40%

4 104 10 3 3

3 103 10 -A: CD1 PE PE-A: CD133 20% 2 CD133 in CD47 expression 102 10

0 0 27.1 66.6 2.96 84.5 2 3 4 5 0102 103 104 105 010 10 10 10 APC-A: CD47 APC-A: CD47 0% CD47 A6L 185 354 215 253

Figure 1. CD47 is expressed in pancreatic cancer (stem) cells. A, quantiﬁcation of CD47 expression in primary patient TMA–containing cores of normal pancreas, pancreatitis, PDAC, and metastases. Shown are the mean relative intensity values of CD47 staining within each core. (Continued on the following page.)

www.aacrjournals.org Clin Cancer Res; 2015 OF3

Ciofﬁ et al.

In vitro phagocytosis assay tumors were established mice were randomized to the respec- For in vitro phagocytosis analysis, 5 104 monocyte-derived tive treatment groups. Gemcitabine was administered twice a macrophages were plated in each well of 24-well tissue-culture week (125 mg/kg i.p.), Abraxane was administered every plates and labeled with PKH26 according to the manufacturer's 4 days (50 mg/kg i.v.), and anti-CD47 was administered daily instructions (Sigma). Macrophages were incubated in serum-free (500 mg/mouse i.p.; ref. 20). medium for 2 hours before adding 2 105 GFP-labeled live cancer cells. Anti-CD47 (B6H12) antibody (10 mg/mL) or IgG1 Statistical analyses control antibody was added and incubated for 2 hours at 37 C. Results for continuous variables are presented as means SD Macrophages were repeatedly washed and subsequently imaged unless stated otherwise and signiﬁcance was determined using the using an inverted microscope (Leica DMI6000B). The phagocytic Mann–Whitney test. All analyses were performed using SPSS 22.0 þ index was calculated as the number of phagocytosed GFP cells (SPSS). per 100 macrophages.

Immunohistochemistry Results For histopathologic analysis, FFPE blocks were serially sec- CD47 is expressed at higher levels in PDAC compared with tioned (3-mm thick) and stained with hematoxylin and eosin normal pancreatic tissue (H&E). Additional serial sections were used for immunohisto- We evaluated the level of CD47 expression by immunohisto- chemical (IHC) studies with anti-CD47 antibody (0.2 mg/mL; chemical analysis of paraffin sections of tissue microarrays con- Abcam ab3283). Antigens were visualized using 3,3-diaminoben- taining primary human tissues from "normal" adjacent non- þ zidine tetrahydrochloride plus (DAB ). Counterstaining was per- tumor pancreatic tissue, pancreatitis, PDAC, and regional lymph, formed with hematoxylin. Histologic quantification of digitalized and liver metastases. CD47 expression was significantly over- slides was performed using Pannoramic Viewer (3DHistech). expressed in primary PDAC tumors (P < 0.001) and metastasis (P < 0.05) versus pancreatitis and normal pancreatic tissue (Fig. Tissue microarrays 1A). Importantly, although CD47 was still detectable in normal Four human tissue microarrays (TMA) containing quadrupli- (non-cancer) pancreatic tissue, the level of expression was signif- cate 1-mm cores from selected areas of paraffin-embedded pan- icantly lower compared with PDAC, patient-derived PDAC xeno- creatic surgical specimens, including ducts, acini, pancreatitis, graft and metastasis samples, where CD47 expression was PDAC, and PDAC metastasis were constructed. A total of 42 markedly stronger, but restricted to epithelial cancer cells and tumors were included. Two xenograft TMAs containing quadru- absent in the stroma (Fig. 1B and Supplementary Fig. S1A). We plicate 1-mm cores from selected tumor areas of 56 paraffin- next analyzed independent large collections of primary tissues for embedded human PDACs grafted in nude mice were also con- the expression of CD47. The results indicated that CD47 expres- structed. The use of human tissue samples for the construction of sion varies considerably within tissues (Supplementary Fig. S1B) the TMAs was approved by the Ethics Committee of the Hospital and between patients with about one third of the patients bearing de Madrid Norte Sanchinarro. All sections were assessed and low to undetectable levels of CD47. Patients represented on each scored by an in-house pathologist (Maria Lozano). of the two independent sets of TMAs were dichotomized according to low to undetectable CD47 expression (CD47 negative) and In vivo tumorigenicity assay intermediate to high CD47 expression (CD47 positive); however, Primary pancreatic cells were treated in vitro with anti-CD47 and no association between CD47 expression and outcome could be serial dilutions of single-cells were resuspended in Matrigel (BD identified (Supplementary Fig. S1C). Biosciences) and s.c. injected into female 6- to 8-week-old NU- We next determined CD47 mRNA expression in a set of nine nu Foxn1 nude mice (Harlan Laboratories). In some experiments, primary patient-derived pancreatic cancer cell cultures, two nor- macrophages were depleted with the following treatment sched- mal pancreas samples, and primary pancreatic stellate cells (PSC). ule: 200 ml of clodronate was injected i.v. twice a week. Tumor We observed low to undetectable levels of CD47 mRNA expres- formation was evaluated after 2 months. Mice were housed accord- sion in normal tissue and in PSCs as compared with PDAC cells. ing to institutional guidelines and all experiments were approved The latter could be subdivided into three groups based on their by the local Animal Experimental Ethics Committee of the Insti- CD47 mRNA expression: low (JH029 and 247), medium (198, tuto de Salud Carlos III (PA 34-2012) and performed in accordance 253, 215, and 354), and high (163, 185, and A6L; Fig. 1C). To with the guidelines for Ethical Conduct in the Care and Use of confirm the expression of CD47 at the protein level, flow-cyto- Animals as stated in The International Guiding Principles for metry analysis was performed on both adherent cells and sphere- Biomedical Research involving Animals, developed by the Council derived cells, the latter of which are enriched in CSCs (13). We for International Organizations of Medical Sciences. observed relatively homogenous expression of CD47 in differentiated cells (ranging from 40% to 57%) whereas in sphere culture- In vivo treatment enriched CSCs, the surface expression of CD47 was higher, Primary tumor tissue pieces of approximately 2 mm3 were although more variable (ranging from 54% to 85%), suggesting nu implanted s.c. into the flanks of NU-Foxn1 nude mice, and once an enrichment of CD47 in CSCs (Fig. 1D). We next assessed the

(Continued.) B, representative pictures of CD47-stained TMA cores, including FFPE section of human-derived xenografts. C, RT-qPCR analysis of CD47 in normal pancreas samples, PSC cells, and several primary human pancreatic cancer cultures. b-Actin was used as a normalization control. D, flow-cytometry analysis of CD47 cell surface expression comparing adherent cells and sphere-derived cells. E, flow-cytometry analysis of CD47 and CD133 expressions on sphere- þ derived cells (left) and quantification of CD133 cells also expressing CD47 (right).

OF4 Clin Cancer Res; 2015 Clinical Cancer Research

CD47 Targets Pancreatic Cancer Stem Cells

AB 1,000 CD47– CD47+ CD47– CD47+ **

800 185 215

CD47+ 600 ** ** 215 400 Number of spheres

354 200 354

CD47+ 0 185 215 354 ** P < 0.01

C Tumorigenicity Tumorigenicity (# tumors / # injections) (# tumors / # injections) Enrich- 104 103 102 P value 104 103 102 CSC Freq ment

CD47– 4/4 2/4 1/4 - 215 3x 0.130 CD47 0.02% CD47+ 4/4 3/4 3/4 354 CD47– 3/4 2/4 0/4 + 354 n.a. <0.0001 CD47 100% CD47+ 4/4 4/4 4/4

D 185 215 250 250 *

200 200

* * 150 150 * * 100 100 Number of spheres

50 50

0 0 *P < 0.05 CD133– CD133+ CD133– CD133+ CD47– CD47+ CD47– CD47+ CD47– CD47+ CD47– CD47+

E Tumorigenicity (# tumors / # injections) Enrich- 103 102 P value ment

CD47– CD133– 1/4 0/4 2x n.s. CD47– CD133+ 2/4 0/4 215 264x <0.0001 CD47+ CD133– 4/4 2/4 <2x n.s. CD47+ CD133+ 4/4 4/4

www.aacrjournals.org Clin Cancer Res; 2015 OF5

Ciofﬁ et al.

þ þ percentage of CD47 cells within the CD133 [a well-established independent of the method of macrophage polarization (Sup- pancreatic CSC marker (9)] subpopulation, and as shown in Fig. plementary Fig. S2A). þ 1E, the majority of CD133 cells expressed CD47 albeit with We next attempted to mimic the tumor in vivo micro- different percentages (ranging from 77% to 97%). environment conditions by polarizing macrophage cultures toward an "M1" phenotype with GM-CSF and an "M2" phe- þ Pancreatic CSCs are mostly confined to CD47 cells notype with M-CSF, respectively (26), or by exposing them to Because our data suggested that CD47 is preferentially CSC-conditioned media from primary cultures of PDAC þ expressed in pancreatic CSC (i.e., CD133 cells), we aimed to spheres (29). We first confirmed that GM-CSF–treated macro- þ assess whether CD47 cells were more "stem-like." We first phages possessed a classic M1 circular morphology, whereas M- FACSorted primary pancreatic cancer cells for CD47 (Fig. 2A) CSF-treated and CSC-conditioned macrophages both showed a and then determined their self-renewal capacity using sphere more elongated shape typical of M2-polarized macrophages þ formation as a readout. We observed that CD47 cells isolated (Fig. 3C, top). In the absence of anti–CD47-blocking antibo- from 185, 215, and 354 primary cells formed significantly more dies, CSC-conditioned "M2" macrophages had the lowest and larger spheres compared with CD47 cells (Fig. 2B), suggest- phagocytic index levels compared with the other macrophage þ ing that CD47 cells are indeed enriched in CSCs. However, to subtypes (0.9 vs. 2.8 for M2-polarized macrophages), consis- obtain conclusive evidence for the latter, we performed in vivo tent with a truly protective and protumorigenic role for these limiting dilution tumorigenicity assays. Ten weeks after injection, macrophages. Treatment with the blocking mAb B6H12.2, þ CD47 cells had formed more tumors, indicating that CSCs are however, significantly enhanced phagocytosis of cancer cells þ mostly contained in the CD47 cell population (Fig. 2C). To across all macrophage subtypes, with a more pronounced further evaluate the function of CD47 in the CSCs context, we increase for M2 (6.7-fold increase) and CSC-conditioned sorted four populations based on CD133 and CD47 expressions (13-fold increase) macrophages (Fig. 3C, bottom). Important- þ þ and observed that CD47 CD133 cells possessed the highest ly, primary monocyte–derived murine macrophages, regardless sphere formation capacity (Fig. 2D). Building upon the latter, we of their initial polarization, were also capable of phagocytosing sorted cells for both CD133 and CD47 and injected them into human PDAC cells when CD47 was blocked (Fig. 3D). þ nude mice depleted for macrophage by means of treatment with Importantly, we evaluated the percentage of CD133 CSCs clodronate. Macrophage depletion was performed to more defin- following anti-CD47 treatment and observed a significant reduc- itively demonstrate that the tumorigenic capacities observed in tion compared with isotype-treated cells (Fig. 3E). Moreover, we vivo were indeed due to differences in functional CSC content and observed a consistent and significant reduction in the sphere not related to the inherent resistance of a cell to macrophage formation capacity of surviving/nonphagocytosed cells, indicat- phagocytosis based on cell surface CD47 expression. Not only did ing that anti-hCD47 treatment indeed eliminated the CSCs pool we confirm that in vivo tumorigenicity is indeed mostly confined (Fig. 3F). Using a more stringent ratio of macrophage:cancer cells þ to the CD47 population, as previously seen (Fig. 2C), but we also (1:1), we observed similar effects in terms of phagocytosis as well þ þ þ show that CD47 CD133 cells are more highly enriched for as significant reduction in sphere formation and CD133 CSC CSCs, and thus more tumorigenic, whereas CD47 CD133 cells content (Supplementary Fig. S2B–S2D). In contrast, for nontrans- bear the least tumorigenic potential (Fig. 2E). Taken together, formed cells no significant induction of phagocytosis was found these data suggest that targeting CD47 should achieve a major (Supplementary Fig. S2B, right), which might be attributed to the reduction in CSC activity. lack of "eat-me" signals on these cells. Finally, to validate the specificity of anti–CD47-induced phagocytosis, we tested a dif- Anti-CD47 treatment enables phagocytosis of pancreatic ferent antibody that also binds a large fraction of PDAC cells (i.e., CSCs anti-CD44). However, treatment with anti-CD44 did not induce It has been previously demonstrated that blocking CD47- phagocytosis whereas in the same experiments treatment with mediated SIRPa signaling using targeted mAbs induces phago- anti-CD47 showed strong induction of phagocytosis (Supple- cytosis of leukemia, lymphoma, and bladder cancer cells by mentary Fig. S2F). These data demonstrate that induction of human and mouse macrophages (19, 21, 28). Using primary phagocytosis by anti-CD47 is likely independent of FcR stimu- PDAC cells stably infected with a lentivirus-expressing GFP lation of macrophages. It is worth noting, however, that blocking (green) and PKH26 dye (red)–labeled primary human mono- CD47 using a Fab molecule would be necessary to definitively cyte-derived macrophages isolated from healthy donors (ratio demonstrate that Fc receptor is not required. Nonetheless, the data cancer cells:macrophages 4:1), we show that in contrast with cells suggest that CD47 is a legitimate therapeutic target for PDAC. treated with an isotype-matched mouse IgG control antibody, primary PDAC cells treated with the blocking anti-human CD47 Anti-CD47 treatment induces apoptosis of pancreatic CSCs (hCD47) mAb B6H12.2 were efficiently phagocytosed by macro- Antibodies directed against CD47 have also been shown to phages. This effect was observed for adherent cells, which mainly directly induce apoptosis of several hematopoietic malignancies þ þ contain non-CSC, and sphere-derived cells or CD47 CD133 (30–32). We therefore incubated sphere-derived cells with anti- sorted cells, which are enriched for CSCs (Fig. 3A and B) and was hCD47 mAb B6H12.2, but in the absence of macrophages, and

Figure 2. þ Pancreatic CSCs are mostly confined to CD47 cells. A, representative flow-cytometry plots of CD47 staining showing the gating strategy for sorting. B, representative images of spheres (left) and quantification of spheres (right) in 185, 354, and 215 primary pancreatic cancer cells sorted for CD47. C, In vivo tumorigenicity of FACSorted 185, 354, and 215 primary pancreatic tumor cells for CD47. D, sphere formation capacity for cells FACSorted for CD47 and CD133. E, in vivo tumorigenicity of cells FACSorted for CD47 and CD133 injected in mice depleted for macrophages by treatment with clodronate (twice a week).

OF6 Clin Cancer Res; 2015 Clinical Cancer Research

CD47 Targets Pancreatic Cancer Stem Cells

A IgG1 isotype Anti-CD47 B

IgG1-Iso Anti-CD47

185 215

40 40 * * IgG1-Iso Anti-CD47 60 60 ** 35 * 35

** * 50 50 30 30 ** * 25 25 40 40

20 20 30 30

Phagocytic index 15 15

Phagocytic index Phagocytic 20 20 10 10

10 10 5 5

0 0 0 0 215 354 215 354 CD47+CD133– CD47+CD133+ CD47+CD133– CD47+CD133+ Adherent Spheres * P < 0.05; ** P < 0.01 * P < 0.05

C Human macrophages EF 250K 250K 31.2% 6.1% 200K 200K

150K 150K 31.2 6.05 SSC A 100K 100K

30 50K 50K IgG1-Iso Anti-CD47

* 0 0 0102 103 104 105 0102 103 104 105 ** CD133 20 IgG1-Iso Anti-CD47 IgG1-Iso Anti-CD47 200 ** 120 * 10

Phagocytic index 100 150 0 Unpolarized M1 M2 CSC media 80 * * Mouse macrophages * D 60 100 * IgG1-Iso Anti-CD47 40 ** 40 ** Number of spheres 30 50 20 ** Fold change of CD133 expression 20 ** 0 0 354 215 354 215

Phagocytic index 10 * P < 0.05 * P < 0.05 ** P < 0.01 ** P < 0.01

0 * P < 0.05 Unpolarized M1 M2 ** P < 0.01

Figure 3. Anti-CD47 enables phagocytosis of pancreatic CSCs. A, representative confocal images (top) and phagocytic index (bottom) of human peripheral blood (PB)– derived macrophages (red) phagocytosing patient–derived CSCs (green) in the presence of blocking anti-CD47 mAb (B6H12) or IgG1 isotype control Ab. (Continued on the following page.)

www.aacrjournals.org Clin Cancer Res; 2015 OF7

Ciofﬁ et al.

A HPDE HFF PSC Early apoptosis Late apoptosis 1.5 1.5 3.7% 0.98% 1.1% Iso

1 1 3.2% 5.4% 1.5%

Annexin V DAPI 2.69% 1.65% 1.74% 0.5 0.5 Anti-CD47 Anti-CD47 Fold change in AnnV+DAPI- expression Fold change in AnnV+DAPI- Fold change in AnnV+DAPI+ expression Fold change in AnnV+DAPI+ 3.9% 6.7% 2.2% 0 0 Iso Anti-CD47 Iso Anti-CD47 B A6L 185 354 Early apoptosis Late apoptosis 4.5 4.5 9% 7.6% 11.6% * Iso

3 3 * 4.6% 19% 15%

Annexin V DAPI 13.6% 19.6% 28.5% 1.5 1.5 Anti-CD47 Fold change in AnnV+DAPI- expression Fold change in AnnV+DAPI- 31% 45% 35% expression Fold change in AnnV+DAPI+ 0 0 Iso Anti-CD47 Iso Anti-CD47 * P < 0.05

Figure 4. Anti-CD47 treatment directly induces apoptosis of pancreatic CSCs. A, ﬂow-cytometry analysis of apoptosis as determined by Annexin V/DAPI staining, in nontransformed human cells; and B, several primary human pancreatic sphere–derived cell cultures after treatment for 12 hours with anti-CD47 mAb (B6H12) or IgG1 isotype control mAb.

subsequently assessed apoptosis 2 and 12 hours after treatment phagocytosis whereas the second being an apparent PDAC- by Annexin V staining. Although we observed no induction of specific elimination of CSCs via direct induction of apoptosis apoptosis in nontransformed human cells (Fig. 4A) or in without involvement of macrophages. primary murine PDAC tumor cells (Supplementary Fig. S3A), we did detect a significant increase in apoptotic cells across Anti-CD47 treatment inhibits in vivo tumorigenicity and tumor several primary human PDAC cell lines following treatment progression, preventing relapse with the anti-CD47 antibody compared with IgG mAb-treated To test the efficiency of CSCs elimination in vitro,wetested controls (Fig. 4B). Importantly, no apoptosis was observed in the ability of surviving/nonphagocytosed cells after anti-CD47 any of the samples tested following 2 hours of treatment treatment to form tumors in vivo. We observed a significant (Supplementary Fig. S3B). Thus, because phagocytosis of CSCs reductioninthetumorigenicityofanti–CD47-treated cells by macrophages was detected as early as 15 minutes after compared with isotype-treated cells (Fig. 5A), and the few incubation with the anti-CD47 antibody (data not shown), tumors that formed from the anti-CD47–treated cultures were we identify two distinct mechanisms of action, the first being significantly smaller in size compared with isotype control

(Continued.) B, the phagocytic index of macrophages phagocytosing human PDAC cells FACSorted for CD47 and CD133 in the presence of blocking anti-CD47 mAb or IgG1 isotype control Ab. C, phagocytic index of human unpolarized, M1, M2, and CSC media polarized macrophages. D, the phagocytic index of murine M1- and M2-polarized macrophages in the presence of blocking anti-CD47 mAb or IgG1 isotype control Ab. E, ﬂow-cytometry analysis of CD133 cell surface expression on surviving cells following incubation with primary human macrophages and treatment with anti-CD47 mAb (B6H12) or IgG1 isotype control mAb. F, sphere formation quantiﬁcation of cells after treatment with anti-CD47 mAbs, compared with IgG1 control treated cells.

OF8 Clin Cancer Res; 2015 Clinical Cancer Research

CD47 Targets Pancreatic Cancer Stem Cells

ABIgG1-Iso Anti-CD47 IgG1-Iso Anti-CD47 CIgG1-Iso Anti-CD47 in 354 Anti-CD47 in 215 120 2.0 120

100 100

1.5

80 80

60 1.0 60 * * 40 * Tumor weight (g) * 40 0.5

20 20 Surface expression (% of IgG1-Iso) Tumorigenicity (tumor take rate in %) in rate take (tumor Tumorigenicity

0 * P < 0.05 0.0 * P < 0.05 0 354 215 354 215 EPCAMEpCAM CD133 SSEA1 D Experimental setup:

Day 01428 35 56 70 100

Abraxane Gemcitabine Implantation Anti-CD47 Assessment of CSC content of xenografts Control * P < 0.01 vs. single treatment 1,000 ** P < 0.01 vs. control

Control Gemcitabine 800

Anti-CD47 )

3 Anti-CD47

600 Gemcitabine Gemcitabine Abraxane + Anti-CD47 * Gemcitabine 400 ** + Anti-CD47

Tumor volume (mm volume Tumor * ** Abraxane 200 Abraxane + Anti-CD47 Abraxane + Anti-CD47

0 d0d7d14d21d28d35d42d49d56d63d70d77d84d91d100 Time (days) E F

60 250K Control GEM Anti-CD47 CD133 surface expression (%) * P < 0.05 37.8% 200K

150K 37.8 50 SSC-A Control 100K HE

50K

0 0102 103 104 105 40 APC-A: CD133 250K 48.2%

200K

CK19 * 150K 48.2

30 SSC-A ABX 100K GEM+Anti-CD47 ABX ABX+Anti-CD47 * * 50K

0 20 0102 103 104 105 APC-A: CD133 250K CD133 expression (%) CD133 expression

HE 20.0%

200K

150K 10 20 ABX SSC-A 100K +Anti-CD47

50K CK19

0 0 2 3 4 5 Control GEMAnti-CD47 GEM ABX ABX 010 10 10 10 +Anti-CD47 +Anti-CD47 CD133

www.aacrjournals.org Clin Cancer Res; 2015 OF9

Ciofﬁ et al.

tumors (Fig. 5B). In addition, although a similar amount of of target cells or (ii) the alternatively activated (M2) macrophage EPCAM expression was observed across all samples regardless that is involved in type II helper T-cell processes, such as wound of the treatment, anti–CD47-treated tumors contained a sig- healing and humoral immunity. In cancer, protumorigenic M2 nificantly lower percentage of cells expressing the CSC surface TAMs enhance neoplasia via matrix remodeling, angiogenesis and markersCD133andSSEA1(Fig.5C).Again,whenweuseda the secretion of protumor growth factors, such as TGFb (34). In more stringent ratio of macrophage:cancer cells (1:1), we contrast, M1 macrophages are believed to inhibit tumor growth observed that tumorigenicity following injection of surviving via antitumor-adaptive immunity mechanisms that include cells was essentially abrogated (Supplementary Fig. S2E). phagocytosis. The latter, however, mainly depends on macro- Encouraged by these promising in vivo tumorigenicity data, we phage recognition of prophagocytic ("eat me") signals on target next performed in vivo therapeutic intervention studies with cells, but can be inhibited by simultaneous expression of anti- human-derived PDAC xenografts expressing intermediate levels phagocytic ("don't eat me") signals, such as CD47. In the context of CD47. Once tumors had formed (100 mm3), mice were of cancer, CD47 has been found to be strongly overexpressed on randomized to one of the following six treatment groups: Diluent different tumor cells, conferring an anti-phagocytic benefitto control; gemcitabine (biweekly 125 mg/kg i.p.) from day 14 to 56; these cells (35, 36). Importantly, inhibition of CD47 using mAbs Abraxane (every 4 days 50 mg/kg i.v.) from day 14 to 28; anti- efficiently induces phagocytosis of cancer cells by macrophages in CD47 (daily 500 mg/mouse i.p.) from day 14 to 35; gemcitabine þ experimental models of leukemia, lymphoma, and bladder car- anti-CD47; and Abraxane þ anti-CD47. Interestingly, for both cinomas (19–21); however, the relevance of this molecule and its used PDX models no significant differences were observed for therapeutic targeting in PDAC (stem) cells had yet to be studied. chemotherapy and anti-CD47 single treatments; however, tumors Herein, we show that CD47 is overexpressed in the majority, treated with a combination of chemotherapy þ anti-CD47 were but not in all primary PDAC patient samples tested using multiple significantly reduced compared with control tumors and single tissue microarrays (>150 patients represented). Although CD47 treatment tumors. Specifically, for PDAC-185, treatment with was significantly overexpressed in about two thirds of neoplastic Abraxane plus anti-CD47 significantly stalled tumor growth. tissues, we did not observe a correlation between high CD47 Although mice previously treated with Abraxane alone showed protein expression and poor clinical outcome (Supplementary similar initial response, tumors eventually relapsed. No relapse Fig. S1C), which is in contrast with what has been shown for other (i.e., de novo growth of tumors); however, was observed when cancers including, AML, HCC, glioma, and ovarian cancer (19– mice were treated with both Abraxane and anti-CD47 (Fig. 5D). 21, 36). The analysis of tissue microarrays may not be sufficient to PDAC-185 tumors in the gemcitabine þ anti-CD47 treatment capture the general CD47 expression profile for each individual group had to be harvested early due to ulcerations. Of note, tumor. Indeed, we observed significant variation between differ- tumors did not differ in gross morphology, as assessed by H&E ent cores that were available from the same patients (Supplemen- and CK19 immunohistochemistry (Fig. 5E), and were also tary Fig. S1A and S1B). Thus, analysis of complete sections of similarly vascularized and contained M2 macrophages (Sup- primary patient PDAC samples is likely needed before a correl- plementary Fig. S4A). Importantly, flow-cytometry analysis of ative connection can be definitively determined for PDAC. Nota- digested tumors showed a significant decrease in the percentage bly, although CD47 was not "clinically predictive" in our tissue of cells expressing the CSC marker CD133 only in mice that had microarrays, we did note that CD47 expression increased in received anti-CD47 treatment (Fig. 5F), suggesting that anti- sphere-derived CSC-enriched cultures, was expressed at even þ þ CD47 treatment effectively targeted the CSC population. In the higher levels in CD133 cells, and CD47 PDAC cells exhibited PDAC-215 PDX model, similar treatment benefits were higher self-renewal and tumorigenic properties compared with observed when anti-CD47 was combined with a chemothera- CD47 cells. Thus, like leukemia, bladder cancer, and HCC (19– peutic, although in this PDX model gemcitabine was more 21), CD47 expression is strongly expressed on pancreatic CSCs; effectivethanAbraxanewhenusedincombinationwithanti- however, it is likely not a suitable surrogate CSC marker as its CD47 as evidenced by the lack of tumor relapse during long- strong expression on a large fraction of non-CSCs limits the level þ term follow-up (Supplementary Fig. S4B and S4C). of enrichment for CSCs in CD47 cells, and thus would require the use of other CSC markers (e.g., CD133) in combination. Using a neutralizing antibody for CD47, we found that inhibit- Discussion ing the anti-phagocytic function of CD47 allowed for both human Macrophages can undergo specific differentiation/polarization and mouse macrophages to phagocytose CSCs cells in vitro, and depending on the environment and surrounding cellular context. the non-phagocytosed surviving PDAC cells exhibited significant- Two distinct states of macrophage polarization have been defined ly reduced expression of CSC markers and functional phenotypes, (33): (i) the classically activated (M1) macrophage that plays an such as self-renewal and in vivo tumorigenicity. Importantly, this important proinflammatory effector role in TH1 cellular immune phenotype was independent of the polarization state of the responses, including the secretion of cytokines and phagocytosis macrophage, as M1 and M2 macrophages were equally capable

Figure 5. Anti-CD47 treatment inhibits in vivo tumorigenicity and tumor progression, preventing relapse. A, in vivo tumorigenicity; B, tumor weight; and C, flow-cytometry analysis for EPCAM, CD133, and SSEA1 for primary pancreatic sphere–derived cells after treatment with anti-CD47 mAb (B6H12) or IgG1 isotype control mAb. D, experimental setup for in vivo treatment (top left) and effects of allocated treatment regiments in 185 tissue xenografts transplanted in immunocompromised mice (bottom left). The mean tumor volume is given; n ¼ 6 tumors per group. Representative images of tumors extracted from mice at the end of the respective observation period (right). E, H&E staining and IHC analysis of CK19 expression in paraffin section from the tumors. F, flow-cytometry analysis of CD133 cell surface expression in cells isolated from tumors of mice treated as indicated.

OF10 Clin Cancer Res; 2015 Clinical Cancer Research

CD47 Targets Pancreatic Cancer Stem Cells

of phagocytosing PDAC cells treated with anti-CD47 mAbs com- Is CD47 targeting in PDAC suitable and ready for further pared with unpolarized macrophages. In the context of the tumor clinical exploration? Although our in vivo study provides proof- microenvironment, M1 macrophages infiltrate the tumor during of-concept for CD47 targeting in PDAC, indeed several open immune surveillance, but once recruited into tumor sites, M1 questions remain to be addressed in further preclinical studies, macrophages can differentiate into M2 macrophages upon expo- but could not be tackled in the present studies based on the sure to cytokines released by tumor cells and tumor stromal cells currently prohibitive costs of low scale antibody production. (e.g., TGFb, IL4, IL13, and IL10; ref. 33). Therefore, our finding Once high-affinity clinical grade antibodies or high-affinity SIRPa that MCSF-polarized M2 macrophages as well as CSC-condi- monomers are available in larger amounts (39), it should be tioned M2-like macrophages were able to phagocytose PDAC determined whether the abundant stroma in PDAC tumors cells treated with anti-CD47 mAbs highlights the potential of M2 represents a relevant physical barrier for the antibodies to reach macrophages, the predominant macrophage subtype present with the cancer cells. Therefore, it should be tested whether coadmin- the PDAC tumor microenvironment, as biologic tools to target istration of a stroma targeting agents leads to better response rates CSCs and their more differentiated non-CSCs progenies. for CD47 antibody treatment. However, a cautionary note comes From a therapeutic perspective, we additionally show in two from recent studies demonstrating that stroma targeting alone xenotransplantation models of PDAC that treatment with mAbs could result in adverse outcomes. Specifically, mouse studies for CD47 in combination with gemcitabine or Abraxane signif- demonstrated that loss of stroma leads to dedifferentiation of icantly reduced primary tumor growth. Specifically, we found that cancer cells rendering them more aggressive (40, 41). In addition, anti-CD47 therapy alone only marginally reduced the size and a recent clinical trial on hedgehog pathway inhibition was pre- rate of tumor growth, which again contrasts with previous find- maturely stopped on the basis of excessive death rates in the ings for other epithelial cancers (19–21), and may be attributed to treatment group. Whether enhanced delivery of CD47 antibodies the more aggressive growth nature of PDAC, which cannot be to stroma-depleted tumors and subsequently enhanced treatment completely controlled by phagocytic macrophages. Alternatively, response will outweigh these putative adverse effects of stroma- untreated tumors with their dense stroma may represent too targeting remains to be determined in carefully designed preclin- strong a barrier for the CD47 antibodies to reach the cancer ical studies. Second, we observed considerable variation in CD47 cells. Importantly, however, in mice treated with gemcitabine or expression across a large panel of primary PDAC samples. Spe- Abraxane, the addition of anti-CD47 therapy resulted in efficient cifically, 10% of patients showed no detectable or very low levels growth control of tumors and prevented relapse after discontin- of CD47 staining. Those patients may not gain significant ther- uation of treatment. The later was particularly apparent when apeutic benefit from anti-CD47 treatment. Such stratification Abraxane was used in combination with anti-CD47. Specifically, could be based on CD47 expression on circulating tumor cells tumors in mice treated with both Abraxane and anti-CD47 mAbs as these have been shown to also express CD47 (42). Third, it diminished in size, such that one tumor was completely elimi- seems reasonable to explore the possibility of combining anti- nated, and the remaining tumors failed to relapse as compared CD47 mAb therapy with treatments that either target TAM recruit- with mice treated with Abraxane alone where relapse was evident ment (e.g., anti-CSF1 therapy) or their polarization toward M2 in all mice by day 77. Regarding the tumor CSC content, we macrophages (e.g., anti-TGFb therapy). observed that only anti-CD47 therapy was able to reduce the In conclusion, we have found that CD47 is expressed on þ percentage of CD133 cells in the tumor, which confirms our primary PDAC cells and we have demonstrated that inhibiting in vitro results and indicates that anti-CD47 mAbs preferentially CD47 function using mAbs is an effective method of treating target CSCs. Taken together, these results strongly suggest that PDAC in vitro and in vivo, thereby forming the rationale for anti-CD47 therapy could be an effective means of treating primary evaluating the clinical efficacy of anti-CD47 therapy in more PDAC tumors, but combination with other anticancer therapeutic comprehensive preclinical studies, which may eventually lead to agents, such as Abraxane, is needed. first trials in human patients with PDAC. Although further mech- Although the main acute mechanism of action of anti-CD47 anistic studies are still needed to determine how anti-CD47 therapy relies on macrophage-mediated phagocytosis of CSCs, we treatment reduces tumor growth (i.e., phagocytosis and/or apo- did observe that long-term treatment of PDAC cells with anti- ptosis), the data presented herein add to the growing repertoire of CD47 mAbs induced a prominent and cancer cell–specific induc- tumors that can be potentially treated with anti–CD47 mAb- tion of apoptosis. It has previously been shown that ligation of based therapies. CD47 triggers caspase-independent programmed cell death in normal and leukemic cells (31, 32, 37); thus, in addition to Disclosure of Potential Conflicts of Interest blocking the antiphagocytic CD47 ligand on PDAC cells, anti- No potential conflicts of interest were disclosed. CD47 mAbs may also function to eliminate CSCs via a separate fi apoptotic-speci c mechanism of action. Additional studies are Authors' Contributions still needed to resolve whether anti-CD47 therapy induces apo- fi in vivo Conception and design: M. Ciof , M. Hidalgo, B. Sainz Jr, C. Heeschen ptosis of tumor cells . In addition, it is important to note that Development of methodology: M. Cioffi, B. Sainz Jr during cell death, prophagocytic ("eat me") signals such as Acquisition of data (provided animals, acquired and managed patients, calreticulin are shuttled to the cell membrane (38). Thus, it is provided facilities, etc.): M. Cioffi, S. Trabulo, M. Hidalgo, E. Costello, W. also logical to hypothesize that anti-CD47 mAb-induced apopto- Greenhalf, M. Erkan, J. Kleeff, B. Sainz Jr sis may also facilitate macrophage-mediated apoptosis via upre- Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M. Cioffi, S. Trabulo, E. Costello, M. Erkan, B. Sainz Jr, gulation of the prophagocytic ("eat me") signal calreticulin. C. Heeschen Therefore, anti-CD47 therapy may have multiple mechanisms of Writing, review, and/or revision of the manuscript: M. Cioffi, M. Hidalgo, M. action, each of which likely facilitates macrophage phagocytosis Erkan, J. Kleeff, B. Sainz Jr, C. Heeschen of CSCs. Study supervision: B. Sainz Jr

www.aacrjournals.org Clin Cancer Res; 2015 OF11

Ciofﬁ et al.

Acknowledgments nacionalizacion de la IþD, Subprogramma: FCCI 2009 (PLE2009-0105; both ﬁ The authors thank Magdalena Choda for excellent technical assistance. They Ministerio de Ciencia e Innovacion, Spain), all to C. Heeschen. M. Ciof is also thank Maria Lozano for assistance with the PDAC TMA analyses. supported by the La Caixa Predoctoral Fellowship Program. The costs of publication of this article were defrayed in part by the Grant Support payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate The research was supported by the ERC Advanced Investigator Grant this fact. (Pa-CSC 233460), European Community's Seventh Framework Programme (FP7/2007-2013) under grant agreement n256974, the Subdireccion General de Evaluacion y Fomento de la Investigacion, Fondo de Investigacion Sanitaria Received June 4, 2014; revised February 3, 2015; accepted February 10, 2015; (PS09/02129 & PI12/02643), and the Programa Nacional de Inter- published OnlineFirst February 23, 2015.

References 1. Paulson AS, Tran Cao HS, Tempero MA, Lowy AM. Therapeutic advances in 20. Willingham SB, Volkmer JP, Gentles AJ, Sahoo D, Dalerba P, Mitra SS, et al. pancreatic cancer. Gastroenterology 2013;144:1316–26. The CD47-signal regulatory protein alpha (SIRPa) interaction is a thera- 2. Burris HA III, Moore MJ, Andersen J, Green MR, Rothenberg ML, Modiano peutic target for human solid tumors. Proc Natl Acad Sci U S A MR, et al. Improvements in survival and clinical benefit with gemcitabine as 2012;109:6662–7. first-line therapy for patients with advanced pancreas cancer: a randomized 21. Chan KS, Espinosa I, Chao M, Wong D, Ailles L, Diehn M, et al. Identi- trial. J Clin Oncol 1997;15:2403–13. fi cation, molecular characterization, clinical prognosis, and therapeutic 3. Conroy T, Desseigne F, Ychou M, Bouche O, Guimbaud R, Becouarn Y, et al. targeting of human bladder tumor-initiating cells. Proc Natl Acad Sci U S A FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer. N Engl J 2009;106:14016–21. Med 2011;364:1817–25. 22. Lee TK, Cheung VC, Lu P, Lau EY, Ma S, Tang KH, et al. Blockade of CD47 4. Von Hoff DD, Ervin T, Arena FP, Chiorean EG, Infante J, Moore M, et al. mediated CTSS-PAR2 signaling provides a therapeutic target for hepato- Increased survival in pancreatic cancer with nab-paclitaxel plus gemcita- cellular carcinoma. Hepatology 2014;60:179–91. bine. N Engl J Med 2013;369:1691–703. 23. Hingorani SR, Wang L, Multani AS, Combs C, Deramaudt TB, Hruban RH, 5. Rothenberg ML, Moore MJ, Cripps MC, Andersen JS, Portenoy RK, Burris et al. Trp53R172H and KrasG12D cooperate to promote chromosomal HA III, et al. A phase II trial of gemcitabine in patients with 5-FU-refractory instability and widely metastatic pancreatic ductal adenocarcinoma in pancreas cancer. Ann Oncol 1996;7:347–53. mice. Cancer Cell 2005;7:469–83. 6. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin 24. Hermann PC, Sancho P, Canamero M, Martinelli P, Madriles F, Michl P, 2012;62:10–29. et al. Nicotine promotes initiation and progression of KRAS-induced 7. Lapidot T, Sirard C, Vormoor J, Murdoch B, Hoang T, Caceres-Cortes J, et al. pancreatic cancer via Gata6-dependent dedifferentiation of acinar cells in A cell initiating human acute myeloid leukaemia after transplantation into mice. Gastroenterology 2014;147:1119–33. SCID mice. Nature 1994;367:645–8. 25. Chao MP, Alizadeh AA, Tang C, Myklebust JH, Varghese B, Gill S, et al. Anti- 8. Clarke MF, Dick JE, Dirks PB, Eaves CJ, Jamieson CH, Jones DL, et al. Cancer CD47 antibody synergizes with rituximab to promote phagocytosis and stem cells–perspectives on current status and future directions: AACR eradicate non-Hodgkin lymphoma. Cell 2010;142:699–713. Workshop on cancer stem cells. Cancer Res 2006;66:9339–44. 26. Sierra-Filardi E, Puig-Kroger A, Blanco FJ, Nieto C, Bragado R, Palomero MI, 9. Hermann PC, Huber SL, Herrler T, Aicher A, Ellwart JW, Guba M, et al. et al. Activin A skews macrophage polarization by promoting a proin- Distinct populations of cancer stem cells determine tumor growth and flammatory phenotype and inhibiting the acquisition of anti-inflamma- metastatic activity in human pancreatic cancer. Cell Stem Cell 2007;1: tory macrophage markers. Blood 2011;117:5092–101. 313–23. 27. Lindberg FP, Lublin DM, Telen MJ, Veile RA, Miller YE, Donis-Keller H, 10. Li C, Heidt DG, Dalerba P, Burant CF, Zhang L, Adsay V, et al. Identification et al. Rh-related antigen CD47 is the signal-transducer integrin-associated of pancreatic cancer stem cells. Cancer Res 2007;67:1030–7. protein. J Biol Chem 1994;269:1567–70. 11. Miranda-Lorenzo I, Dorado J, Lonardo E, Alcala S, Serrano AG, Clausell- 28. Jaiswal S, Jamieson CH, Pang WW, Park CY, Chao MP, Majeti R, et al. CD47 Tormos J, et al. Intracellular autofluorescence: a biomarker for epithelial is upregulated on circulating hematopoietic stem cells and leukemia cells cancer stem cells. Nat Methods 2014;11:1161–9. to avoid phagocytosis. Cell 2009;138:271–85. 12. Mueller MT, Hermann PC, Witthauer J, Rubio-Viqueira B, Leicht SF, 29. Sainz B Jr, Moreno B, Tatari M, Heeschen C, Guerra S. ISG15 is a critical Huber S, et al. Combined targeted treatment to eliminate tumorigenic microenvironmental factor for pancreatic cancer stem cells. Cancer Res cancer stem cells in human pancreatic cancer. Gastroenterology 2009; 2014;74:7309–20. 137:1102–13. 30. Kikuchi Y, Uno S, Kinoshita Y, Yoshimura Y, Iida S, Wakahara Y, et al. 13. Lonardo E, Hermann PC, Mueller MT, Huber S, Balic A, Miranda-Lorenzo I, Apoptosis inducing bivalent single-chain antibody fragments against et al. Nodal/Activin signaling drives self-renewal and tumorigenicity of CD47 showed antitumor potency for multiple myeloma. Leuk Res 2005; pancreatic cancer stem cells and provides a target for combined drug 29:445–50. therapy. Cell Stem Cell 2011;9:433–46. 31. Mateo V, Lagneaux L, Bron D, Biron G, Armant M, Delespesse G, et al. CD47 14. Kurahara H, Shinchi H, Mataki Y, Maemura K, Noma H, Kubo F, et al. ligation induces caspase-independent cell death in chronic lymphocytic Significance of M2-polarized tumor-associated macrophage in pancreatic leukemia. Nat Med 1999;5:1277–84. cancer. J Surg Res 2011;167:e211–9. 32. Bras M, Yuste VJ, Roue G, Barbier S, Sancho P, Virely C, et al. Drp1 mediates 15. Pylayeva-Gupta Y, Lee KE, Hajdu CH, Miller G, Bar-Sagi D. Oncogenic Kras- caspase-independent type III cell death in normal and leukemic cells. Mol induced GM-CSF production promotes the development of pancreatic Cell Biol 2007;27:7073–88. neoplasia. Cancer Cell 2012;21:836–47. 33. Biswas SK, Mantovani A. Macrophage plasticity and interaction with 16. Qian BZ, Pollard JW. Macrophage diversity enhances tumor progression lymphocyte subsets: cancer as a paradigm. Nat Immunol 2010;11: and metastasis. Cell 2010;141:39–51. 889–96. 17. Brown EJ, Frazier WA. Integrin-associated protein (CD47) and its ligands. 34. Flavell RA, Sanjabi S, Wrzesinski SH, Licona-Limon P. The polarization of Trends Cell Biol 2001;11:130–5. immune cells in the tumour environment by TGFbeta. Nat Rev Immunol 18. Matozaki T, Murata Y, Okazawa H, Ohnishi H. Functions and molecular 2010;10:554–67. mechanisms of the CD47-SIRPalpha signalling pathway. Trends Cell Biol 35. Chao MP, Weissman IL, Majeti R. The CD47-SIRPalpha pathway in cancer 2009;19:72–80. immune evasion and potential therapeutic implications. Curr Opin Immu- 19. Majeti R, Chao MP, Alizadeh AA, Pang WW, Jaiswal S, Gibbs KD Jr, et al. nol 2012;24:225–32. CD47 is an adverse prognostic factor and therapeutic antibody target on 36. Chao MP, Majeti R, Weissman IL. Programmed cell removal: a new obstacle human acute myeloid leukemia stem cells. Cell 2009;138:286–99. in the road to developing cancer. Nat Rev Cancer 2012;12:58–67.

OF12 Clin Cancer Res; 2015 Clinical Cancer Research

CD47 Targets Pancreatic Cancer Stem Cells

37. Pettersen RD, Hestdal K, Olafsen MK, Lie SO, Lindberg FP. CD47 signals T 40. Ozdemir BC, Pentcheva-Hoang T, Carstens JL, Zheng X, Wu CC, Simpson cell death. J Immunol 1999;162:7031–40. TR, et al. Depletion of carcinoma-associated fibroblasts and fibrosis 38. Buoncervello M, Borghi P, Romagnoli G, Spadaro F, Belardelli F, Toschi induces immunosuppression and accelerates pancreas cancer with reduced E, et al. Apicidin and docetaxel combination treatment drives CTCFL survival. Cancer Cell 2014;25:719–34. expression and HMGB1 release acting as potential antitumor immune 41. Rhim AD, Oberstein PE, Thomas DH, Mirek ET, Palermo CF, Sastra SA, response inducers in metastatic breast cancer cells. Neoplasia 2012; et al. Stromal elements act to restrain, rather than support, pancreatic ductal 14:855–67. adenocarcinoma. Cancer Cell 2014;25:735–47. 39. Weiskopf K, Ring AM, Schnorr PJ, Volkmer JP, Volkmer AK, Weissman IL, 42. Baccelli I, Schneeweiss A, Riethdorf S, Stenzinger A, Schillert A, Vogel V, et al. Improving macrophage responses to therapeutic antibodies by et al. Identification of a population of blood circulating tumor cells from molecular engineering of SIRPalpha variants. Oncoimmunology 2013;2: breast cancer patients that initiates metastasis in a xenograft assay. Nat e25773. Biotechnol 2013;31:539–44.

www.aacrjournals.org Clin Cancer Res; 2015 OF13

Inhibition of CD47 Effectively Targets Pancreatic Cancer Stem Cells via Dual Mechanisms

Michele Cioffi, Sara Trabulo, Manuel Hidalgo, et al.

Clin Cancer Res Published OnlineFirst February 23, 2015.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-14-1399

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2015/02/25/1078-0432.CCR-14-1399.DC1

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/early/2015/04/08/1078-0432.CCR-14-1399. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.