Monoclonal Antibodies Targeting LecLex-Related Glycans with Potent Antitumor Activity.

Published OnlineFirst March 16, 2015; DOI: 10.1158/1078-0432.CCR-14-3030

Cancer Therapy: Preclinical

Monoclonal Antibodies Targeting LecLex-Related Glycans with Potent Antitumor Activity

Clinical Cancer Research

Jia Xin Chua1, Mireille Vankemmelbeke1, Richard S. McIntosh1, Philip A. Clarke2, Robert Moss1, Tina Parsons1, Ian Spendlove1, Abid M. Zaitoun3, Srinivasan Madhusudan1, and Lindy G. Durrant1

Abstract Purpose: To produce antitumor monoclonal antibodies (mAbs) targeting glycans as they are aberrantly expressed in tumors and are coaccessory molecules for key survival pathways. Experimental Design: Two mAbs (FG88.2 and FG88.7) recognizing novel tumor-associated Lewis (Le) glycans were produced by immunizations with plasma membrane lipid extracts of the COLO205 cell line. Results: Glycan array analysis showed that both mAbs bound LecLex, di-Lea, and LeaLex, as well as Lea-containing glycans. These glycans are expressed on both lipids and proteins. Both mAbs showed strong tumor reactivity, binding to 71% (147 of 208) of colorectal, 81% (155 of 192) of pancreatic, 54% (52 of 96) of gastric, 23% (62 of 274) of non–small cell lung, and 31% (66 of 217) of ovarian tumor tissue in combination with a restricted normal tissue distribution. In colorectal cancer, high FG88 glycoepitope expression was significantly associated with poor survival.

Introduction Successful cancer immunotherapy is dependent on the generation of monoclonal antibodies (mAbs) with good specificity and potent killing. The complexity of the glycome and altered expression of glycosyl transferases associated with malignant transformation make cancer cell-associated carbohydrates excellent targets (1–4). Glycolipids are particularly attractive due to their dense cell-surface distribution, mobility, and association with membrane microdomains, all of which contribute to their participation in a wide range of cellular signaling and adhesion processes (4–6). Generating antiglycolipid antibodies, however, is a challenging task as they do

1 Division of Cancer and Stem Cells, School of Medicine, City Hospital Campus, University of Nottingham, Nottingham, United Kingdom. 2 Division of Cancer and Stem Cells, School of Medicine, Queens Medical Centre, University of Nottingham, Nottingham, United Kingdom. 3Section of Surgery, School of Medicine, Queens Medical Centre, University of Nottingham, Nottingham, United Kingdom.

Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/). Corresponding Author: Lindy G. Durrant, University of Nottingham, Academic Clinical Oncology, City Hospital Campus, Hucknall Road, Nottingham NG5 1PB, United Kingdom. Phone: 44-115-8231863; Fax: 44-115-8231863; E-mail: [email protected] doi: 10.1158/1078-0432.CCR-14-3030 2015 American Association for Cancer Research.

The mAbs demonstrated excellent antibody-dependent cellular cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC), in addition to direct tumor cell killing via a caspaseindependent mechanism. Scanning electron microscopy revealed antibody-induced pore formation. In addition, the mAbs internalized, colocalized with lysosomes, and delivered saporin that killed cells with subnanomolar potency. In vivo, the mAbs demonstrated potent antitumor efficacy in a metastatic colorectal tumor model, leading to significant long-term survival. Conclusions: The mAbs direct and immune-assisted tumor cell killing, pan-tumor reactivity, and potent in vivo antitumor efficacy indicate their potential as therapeutic agents for the treatment of multiple solid tumors. In addition, internalization of saporin conjugates and associated tumor cell killing suggests their potential as antibody drug carriers. Clin Cancer Res; 21(13); 2963–74. 2015 AACR.

not provide T-cell help and the mAbs are usually low-affinity IgMs. Lewis (Le) carbohydrate antigens are formed by the sequential addition of fucose onto oligosaccharide precursor chains on glycoproteins and glycolipids through the concerted action of a set of glycosyltransferases (7). Type I chains (containing Galb(1!3)GlcNAc) form Lea and Leb, whereas type II chains (containing Galb(1!4)GlcNAc) form Lex and Ley. Lea and Leb antigens are regarded as blood group antigens, yet many human cancers express Lea or Leb antigens regardless of Lewis blood group status (8–10). In addition, Lea and Leb antigen frequently coexist in human tumor cells (11). Dimeric Lea and Lex have also been identified as tumor-associated antigens in breast and gastrointestinal carcinomas (9, 10, 12). Only a limited number of mAbs recognizing glycans have been described (13, 14). GNX-8 (human IgG1) bound the extended type I chain epitope Leb-Lea (15) and FC-215 (murine IgM), an anti-Lex mAb that induced transient antitumor responses, but caused profound neutropenia in phase I trials (16, 17). NCC-ST-421 (ST-421, murine IgG3) recognized gastric cancerassociated dimeric Lea and demonstrated significant antitumor effects in a human tumor xenograft model (18). The murine IgM mAb 43-9F, targeted Lea/Lex epitopes, cross-reacting with simple and extended Lea epitopes, but had no in vivo antitumor reactivity (19). The mAb 504/4 (SC104, murine IgG1), recognized sialyl (S) Le-related glycans, induced antibody-dependent cellular cytotoxicity (ADCC) and CDC, as well as direct tumor cell death and importantly, demonstrated tumor growth inhibition in vivo (20).

www.aacrjournals.org

Downloaded from clincancerres.aacrjournals.org on August 1, 2015. © 2015 American Association for Cancer Research.

2963


Chua et al.

Translational Relevance Differentially expressed tumor-associated carbohydrate antigens constitute excellent targets for therapeutic antibody development. This report describes the discovery of two monoclonal antibodies (mAbs) that recognize extended nonsialylated Le-containing glycans on tumor glycolipid and glycoproteins with high functional affinity. The FG88 mAbs exhibit comprehensive tumor tissue reactivity combined with limited normal tissue distribution and excellent immunemediated and direct tumor cell killing via a unique mechanism that may elicit further immune-mediated tumor regression. Furthermore, the mAbs have independent prognostic value in patients with colorectal tumors. These favorable attributes, combined with their potent antitumor activity in a mouse xenograft model suggest promising clinical potential. In addition, the internalization efficiency of the mAbs and their lysosomal colocalization make them attractive candidates for drug conjugation. We anticipate that the excellent preclinical effectiveness of our murine mAbs will translate, following humanization, into therapeutic antitumor efficacy in the clinic.

It is currently in phase I/II clinical trials in gastrointestinal cancer (NCT01447732). More recently, human SLea mAbs were produced using a patient vaccination strategy that showed specific binding to SLea and exhibited ADCC, CDC, and antitumor activity in a xenograft model (21). In this study, we describe the characterization of two mAbs, FG88.2 and FG88.7, targeting the novel galb1-3GlcNacb13Galb1-4(Fuca1-3)GlcNAc (LecLex) glycan as well as di-Lea and LeaLex glyco-epitopes that are expressed by a wide range of tumors but have a restricted normal tissue expression. Both mAbs display effective immune-mediated and direct cytotoxicity that translated into potent in vivo antitumor activity. Furthermore, their efficient internalization suggests potential as antibody drug carriers (ADC).

Materials and Methods Materials, cells, and antibodies Colorectal (COLO205, HCT-15, and HT29), ovarian (OVCAR3), gastric (AGS), lung (DMS79), T-cell leukemia (Jurkat), and mouse (Balb/c) lymphoblastoid myeloma (NS0) cancer cell lines were all obtained from the ATCC. Colorectal cell line HCT-15HM2 is a high-metastasizing variant of HCT-15 (22). All cell lines were regularly authenticated using short tandem repeat profiling. mAbs, 7-Le and 225-Le, were from Abcam, anti-CD40 mAb from R&D Systems, anti-Fas mAb from Upstate (Millipore), anti-CEACAM mAb from eBioscience, OKT3 (anti-CD3), FG27.10 (anti-Ley), FG27.18 (anti-Ley/x), 791T/36 (anti-CD55), and 505/4 (SC104; anti-Sdi-Lea) were produced in house. Alphagalactosylceramide was from Alexis Biochemical, glycan–human serum albumin (HSA) conjugates from IsoSep AB and lipids from Sigma. Generation of mAbs Plasma membrane lipid extract [0.36 mole%] from 5 107 COLO205 cells incorporated into liposomes was used to immu-

2964 Clin Cancer Res; 21(13) July 1, 2015

nize BALB/c mice at two-weekly intervals over a 2-month period, with a-galactosylceramide and anti-CD40 mAb as adjuvants. Five days after the final immunization, the splenocytes were harvested and fused with NS0 myeloma cells. Stable clones were established by repeated limiting dilutions and mAbs purified using standard protein G affinity chromatography. Flow cytometry Cells (1 105) were incubated with primary mAbs at 4 C for 1 hour followed by FITC-conjugated secondary mAb and fixing, as described previously (23) and in Supplementary Materials and Methods. For the propidium iodide (PI) uptake, cells (5 104) were incubated with mAbs for 2 hours at 37 C followed by the addition of 1 mg of PI for 30 minutes. Cells were resuspended in PBS and run on a Beckman Coulter FC-500 with WinMDI 2.9 for analysis. Thin-layer chromatography analysis of glycolipid binding Glycolipid extract from 2 106 COLO205 cells (20) spotted onto Merck high-performance thin-layer chromatography (HPTLC) silica plates and developed twice in chloroform: methanol:H2O (60:30:5) followed by twice hexane:diethyl ether:acetic acid (80:20:1.5). The dried plates were sprayed with 0.1% (w/v) polyisobutylmethacrylate (Sigma) in acetone and blocked with PBS 2% (w/v) BSA (PBS/BSA) for 1 hour at room temperature. The plates were then incubated overnight at 4 C with primary mAbs followed by two 1-hour incubations with biotinylated anti-mouse IgG (Sigma) and IRDye 800CW streptavidin (LICOR Biosciences), respectively. Plates were airdried and lipid bands visualized using a LICOR Odyssey scanner. SDS-PAGE and Western blot analysis Briefly, 1 105 or 1 106 cell equivalents of cancer cell lysate, total lipid extract, and plasma membrane lipid extract were subjected to SDS-PAGE (4%–12% Bis–Tris NOVEX; Invitrogen), and transferred to Immobilon-FL PVDF membranes (EMDMillipore). Membranes were blocked for 1 hour followed by incubation with primary mAbs. mAb binding was detected using biotinylated anti-mouse IgG (Sigma) and visualized using IRDye 800CW streptavidin (LICOR Biosciences). Lewis antigen and saliva sandwich ELISA ELISA plates were coated overnight at 4 C and processed as described in Supplementary Materials and Methods. Glycan array analysis mAbs were screened for binding to 600 natural and synthetic glycans (core H group, version 5.1) by the Consortium for Functional Glycomics (CFG). Slides were incubated with 1 mg/ mL of antibody for 1 hour, before detection with Alexa Fluor 488– conjugated secondary mAb. Immunohistochemistry assessment Normal and tumor tissue binding was analyzed by immunohistochemistry (IHC) as described previously (20). Briefly, after antigen-retrieval, blocking of endogenous peroxidase activity and nonspecific binding sites, the sections were incubated with primary mAbs at room temperature for 1 hour. Primary mAb binding was detected by biotinylated secondary mAb (Vector Labs) followed by preformed




Antitumor Activity of Novel Glycan Monoclonal Antibodies

streptavidin-biotin/HRPO (Dako Ltd.) and 3,30 -diaminobenzidine as the substrate. Finally, sections were counterstained with hematoxylin. Staining was analyzed via New Viewer software 2010 and scored with a semiquantitative histologic score (H-score, 0–300) taking into account the staining intensity: negative (0), weak (1), medium (2), and strong (3) and the percentage of positive cells. The stained slides were analyzed by two independent observers and a consensus was agreed. Statistical analysis was performed using SPSS 13.0 (SPSS Inc.). Stratification cutoff points for the survival analysis were determined using X-Tile software (24) and P < 0.05 were considered significant.

Patient cohorts The study populations include cohorts of consecutive series of 462 archived colorectal cancer (25) specimens (1994–2000; median follow-up, 42 months; censored December 2003; patients with lymph node–positive disease routinely received adjuvant chemotherapy with 5-flurouracil/folinic acid), 350 ovarian cancer (26) samples (1982–1997; median follow-up, 192 months; censored November 2005; patients with stage II to IV disease received standard adjuvant chemotherapy that in later years was platinum based), 142 gastric cancer (27) samples (2001–2006; median follow-up, 66 months; censored Jan 2009; no chemotherapy), 68 pancreatic and

Figure 1. A, cell-surface binding of FG88 hybridoma supernatants. Binding was assessed by flow cytometry on COLO205 cells by (i) FG88.2 and (ii) FG88.7 hybridoma supernatants. mAb SC104 (iii) and mouse IgG (iv), both at 8 3.3 10 mol/L were positive and negative control, respectively. B, FG88 a, c, x mAbs recognize Le -containing glycans on glycolipid and glycoproteins. i, Western blot analysis 8 of FG88.2 mAb (3.3 10 mol/L). 5 Lane 1, COLO205 cell lysates (1 10 cells); lane 2, COLO205 plasma 6 membrane (1 10 cells); lane 3, 6 COLO205 total lipid extract (1 10 cells), and lane 4, COLO205 plasma 6 membrane lipid extract (1 10 cells). ii, HPTLC analysis of FG88.2 binding 8 (3.3 10 mol/L) to cancer cell 6 (2 10 cells) glycolipid extracts. Lane 1, AGS; lane 2, COLO205. C, ELISA analysis of binding of FG88 mAbs to Lewis glycan–HSA conjugates. The 7-Le, 225-Le, FG27.18, and FG27.10 mAbs were included as positive a b x y controls for Le , Le , Le , and Le , respectively. The negative control consisted of omission of the primary mAb. Error bars represent the mean STD of quadruplicate wells ( , P < 0.05; , P < 0.01; and , P < 0.001 versus control, ANOVA followed by the Bonferroni multiple comparisons test, GraphPad Prism 6. D, Glycan array screening of FG88.2 and FG88.7 (CFG, core H, version 5.1). The top five binding glycans are shown as a percentage of the highest binding glycan with glucosylamine (&), galactose (*), fucose (~), and mannose (*). Sp denotes the length of spacer between glycan and slide. The complete results for FG88.2 (42) and FG88.7 (43) have been deposited on the CFG website.


Clin Cancer Res; 21(13) July 1, 2015


2965


Chua et al.






120 billary/ampullary cancer (28) samples (1993–2010: median follow-up, 45 months; censored 2012; 25%–46% of patients received adjuvant chemotherapy with 5-flurouracil/folinic acid and gemcitabine), 220 non–small cell lung cancers (January 1996–July 2006; median follow-up, 36 months; censored May 2013; none of the patients received chemotherapy prior to surgery but 11 patients received radiotherapy and 9 patients received at least 1 cycle of adjuvant chemotherapy postsurgery) obtained from patients undergoing elective surgical resection of a histologically proven cancer at Nottingham or Derby University Hospitals (Derby, United Kingdom). No cases were excluded unless the relevant clinicopathologic material/data were unavailable. Cell viability analysis Cells (1 103/well) were allowed to adhere prior to incubation for 72 hours with increasing concentrations of mAbs in the presence or absence of 20 mmol/L of a pan-caspase inhibitor: carbobenzoxy-valyl-alanyl-aspartyl-[O-methyl]-fluoromethylketone (Z-FMK-VAD), after which cell viability was analyzed using either 3H-thymidine (0.5 mCi/well) during the final 24 hours and its incorporation measured using Microscint 0 liquid scintillant on a Topcount NXT (PerkinElmer) or a WST-8 assay (CCK-8; Sigma-Aldrich). Phase-contrast images of mAb-treated HCT-15 cells were obtained with 10 magnification using a Nikon Eclipse TS100. Confocal microscopy FG88 mAbs, labeled with Alexa Fluor-488 (495/519 nm) according to the manufacturer's protocol (Invitrogen), were added to HCT-15 cells (1.5 105) on coverslips and incubated for 1 hour at 37 C before removing excess mAb with HEPES buffer (Invitrogen). Hoechst 33258 (350/461 nm) nucleic acid stain (1 mg/mL; Invitrogen) and LysoTracker deep red (647/668 nm) lysosomal stain (50 nmol/L; Invitrogen) were added during the last 30 minutes, with Cell Mask Orange (554/567 nm) plasma membrane stain (2.5 mg/mL; Invitrogen) added during the final 10 minutes. Localization of the mAbs was visualized using confocal microscopy through a 63 1.4 NA oil objective (ZEISS AX10 Observer Z1; Carl Zeiss) with Zen 2009 image acquisition software. ADC assay ADC was evaluated by measuring the cytotoxicity of immune-complexed mAbs with a mouse Fab–ZAP secondary conjugate (Advanced Targeting Systems; ref. 29). Cells were plated in triplicates overnight into 96-well plates (2,000 cells, 90 mL/well). After preincubaton (30 minutes at room temperature) of a concentration range of FG88 mAbs with 50 ng of the

Fab–ZAP conjugate, 10 mL of conjugate or free mAb was added to the wells and incubated for 72 hours. Control wells, consisted of cells incubated without conjugate, incubated with secondary Fab–ZAP without primary mAb and incubated with a control mAb in the presence of Fab–ZAP. Cell viability was measured by 3H-thymidine incorporation during the final 24 hours. Results are expressed as a percentage of 3H-thymidine incorporation by cells incubated with conjugate compared with primary mAb only. ADCC and CDC ADCC and CDC were performed as described previously (23). 51 Cr-labeled target cells (5 103) were coincubated with 100 mL of peripheral blood mononuclear cells (PBMC), 10% (v/v) autologous serum or media alone or with mAbs at a range of concentrations (E:T ratio of 100:1). Spontaneous and maximum release [counts per minute (cpm)] was evaluated by incubating the labeled cells with medium or with 10% (v/v) Triton X-100, respectively. The mean percentage lysis was calculated as follows: mean % lysis ¼ (experimental cpm spontaneous cpm)/ maximum cpm spontaneous cpm) 100. Scanning electron microscopy HCT-15 cells (1 105) were grown on sterile coverslips for 24 hours prior to mAb (30 mg/mL) addition for 2 or 20 hours at 37 C. Controls included medium alone and 0.5% (v/v) hydrogen peroxide (H2O2; Sigma). Cells were washed with prewarmed 0.1 mol/L sodium cacodylate buffer (SDB) pH 7.4 and fixed with 12.5% (v/v) glutaraldehyde for 24 hours. Fixed cells were washed twice with SDB and postfixed with 1% (v/v) osmium tetroxide (pH 7.4) for 45 minutes. After a final wash with H2O, the cells were dehydrated in increasing concentrations of ethanol and exposed to critical point drying, before sputtering with gold, prior to scanning electron microscopy (SEM) analysis (JSM-840 SEM, JEOL). In vivo model The study was conducted under a UK Home Office Licence in accordance with National Cancer Research Institute, Laboratory Animal Science Association, and Federation of European Laboratory Animal Science Associations guidelines. Age-matched male MF-1 nude mice (n 8 for each treatment group; Harlan Laboratories) were implanted intraperitoneally with HCT-15HM2 DLuX cells and tumor establishment monitored by bioluminescent imaging. Mice (bioluminescent signal 1 107 p/s) were dosed intravenously (i.v.) biweekly with mAbs (0.1 mg) or vehicle (PBS, 100 mL) until day 120. Bioluminescent intensity (BLI) was evaluated weekly. Briefly, 60 mg/kg D-luciferin substrate was administered to anesthetized mice subcutaneously (s.c.), and BLI

Figure 2. 9 Human tissue distribution of FG88.2. A, FG88.2 (2 10 mol/L) tumor tissue binding. Colorectal, gastric, pancreatic, non–small cell lung, and ovarian tumor TMAs were stained with FG88.2 mAb as described in Materials and Methods. Representative images of (i) negative, (ii) weak, (iii) moderate, and (iv) strong FG88.2 staining are shown (magnification, 20). B, the Kaplan–Meier analysis of progression-free survival in the colorectal cancer TMA cohort showing the impact of 9 FG88.2 glyco-epitope expression. Significance was determined using the log-rank test. C, FG88.2 (2 10 mol/L) normal human tissue (AMSBIO) binding. Placenta, 1; bladder, 2; adipose, 3; brain, 4; esophagus, 5; colon, 6; cervix,7; diaphragm, 8; skin, 9; gall bladder, 10; heart, 11; ileum, 12; rectum, 13; skeletal muscle, 14; spleen, 15; stomach, 16; breast, 17; cerebellum, 18; thyroid, 19; uterus, 20; duodenum, 21; pancreas, 22; ovary, 23; tonsil, 24; jejunum, 25; kidney, 26; liver, 27; lung, 28; testis, 29; thymus, 30; and small intestine, 31 (magnification, 20). D, FG88.2 tumor cell binding by flow cytometry. i, FG88.2 binding in the concentration 12 7 7 range: 1.1 10 –2 10 mol/L, is presented as geometric mean (Gm). ii, flow cytometry histograms of isotype control or FG88.2 (2 10 mol/L) binding to a subset of cancer cells (COLO205, HCT-15, HT29, and OVCAR3) for SABC determination as described in Supplementary Materials and Methods.




2967


Chua et al.

readings were taken 15 minutes after substrate administration. In this study, weight loss alone was an unreliable indicator of clinical endpoint due to increasing tumor burden. Thus, mouse survival was determined by morbidity and bioluminescent assessment of tumor burden that provided tumor mass to body weight information. Data processing and integrity checking occurred in SPSS v21.0, statistical analysis using GraphPad Prism 6. Significant differences in tumor sizes were assessed by the Mann–Whitney U test and survival (Kaplan–Meier) analysis was done using the log-rank (Mantel–Cox) test.

Results Identification and characterization of two mAbs following COLO205 lipid immunizations Initial screening of hybridoma supernatants for COLO205 cell surface and lipid reactivity led to the identification of two mAbs, FG88.2 and FG88.7. Both mAbs were mouse IgG3 isotypes with k light chains and strong COLO205 cell-surface reactivity (Fig. 1A). FG88.2 recognized glyco-epitopes on glycolipids and glycoproteins (molecular weight ranging from 10 to >230 kDa) in lipid extracts and cell lysates from COLO205 (Fig. 1B, i). Confirmation of glycolipid reactivity came from HPTLC analysis in which FG88.2 reacted with distinct glycolipid species extracted from COLO205, but not AGS (Fig. 1B, ii). FG88.7 displayed similar glyco-epitope recognition (data not shown). DNA sequencing revealed that FG88.2 and FG88.7 belonged to the IGHV6-6 01 heavy chain and IGKV12-41 01 gene families where they have 10 and eight mutations from IGHV6-6 01, and 11 and 12 mutations from IGKV12-41 01, respectively. The nature and pattern of the mutations suggests affinity maturation. The two FG88 mAbs differ by only three residues, two in the CDR and one in the FR region. FG88 mAbs recognize Lec- and Lea-containing glycans with high functional affinity Preliminary characterization of the glycan specificity of both mAbs, using ELISA on Lewis blood group HSA conjugates, revealed a strong preference for the Lea–HSA conjugate with minimal cross-reactivity against Lex–HSA and no binding to Ley-, or Leb–HSA (Fig. 1C). Both FG88 mAbs were then screened in more detail for reactivity with 600 natural and synthetic glycans by the CFG. The mAbs showed a strong preference for extended type I Lea-related glycans with LecLex, LeaLex, and di-Lea being the strongest binders (Fig. 1D). In addition, the mAbs bound simple Lea on the array but not Lec or Lex. This was corroborated by competition experiments in which preincubation of both mAbs with the Lea–HSA conjugate, fully inhibited COLO205 binding (Supplementary Fig. S1A), with Lex–HSA being less effective. We next examined the Lea–HSA binding kinetics of our mAbs using SPR (Biacore X). Fitting of the binding curves revealed strong apparent functional affinity (Kd 1010 mol/L) with fast association (105 1/Ms) and slow dissociation (105 1/s) rates for both mAbs (Supplementary Table S1). This was in line with EC50 values from Lea–HSA ELISA (Supplementary Table S1 and Supplementary Fig. S1B). Cellular functional affinity was lower (Kd 109 mol/L; Supplementary Table S1 and Supplementary Fig. S1C), reflecting the complex mAb binding at the cell surface with mixed glycolipid/glycoprotein recognition and the possibility of secondary events (internalization or cell killing).


FG88 mAbs exhibit strong differential tumor/normal tissue reactivity, with colorectal staining being an independent prognostic marker, associated with reduced patient survival We assessed the tumor-binding potential of FG88.2 by IHC screening for binding to pancreatic, colorectal, gastric, non– small cell lung (NSCLC), and ovarian TMAs. FG88.2 mAb stained 81% (155 of 192) of pancreatic, 71% (147 of 208) of colorectal, 54% (52 of 96) of gastric, 23% (62 of 274) of NSCLC, and 31% (66 of 217) of ovarian tumor tissues and examples of FG88.2 staining level of tumor tissues are shown (Fig. 2A). No significant association between FG88.2 glyco-epitope expression and clinicopathologic variables was observed in any of the tumor TMAs. In the colorectal cancer cohort (25), however, FG88.2 glyco-epitope expression was significantly associated with Bcl-2 (x2 ¼ 6.607; P ¼ 0.013) and with fucosyl transferase 3 (FUT3) expression (x2 ¼ 26.712; P ¼ 0.031). The Kaplan–Meier analysis of disease-free survival of colorectal cancer patients revealed a significantly lower mean survival time in the high FG88.2 glyco-epitope expressing group [mean survival, 36 months (high; n ¼ 27) versus 79 months (low; n ¼ 184); P ¼ 0.01, log-rank test; Fig. 2B]. On multivariate analysis using Cox regression, high FG88.2 glycoepitope expression in colorectal cancer was a marker of poor prognosis that was independent of stage and vascular invasion (P < 0.001). The strong tumor reactivity of FG88.2 was compared with its normal human tissue distribution (Fig. 2C and Supplementary Table S2). FG88.2 did not bind most normal tissues, including lung, liver (parenchyma), heart, brain, and kidney. Positive normal tissue staining included columnar epithelium of gall bladder (weak to moderate), bile duct (moderate), thymus (weak), glandular epithelium of colon (moderate), squamous epithelium of tonsil (moderate), and pancreas (weak). Normal human tissue binding by FG88.7 was similar to FG88.2, with the exception of rectum (Supplementary Table S2). On a normal cynomegalous monkey (CN) TMA, FG88.2 exhibited weak staining of skin, colon, ovary, liver, and thymus combined with stronger staining of glandular epithelium of stomach and small intestine (Supplementary Fig. S2A). As previous studies had shown that Lex was expressed on neutrophils, the binding of the FG88 mAbs to granulocytes and polymorphonuclear cells (PBMCs) from normal donors was analyzed by FACS, where no binding was observed (Supplementary Fig. S2B). In addition, Lea and Leb antigens found in tissue secretions can adsorb to erythrocytes. We determined secretor status of 9 healthy human donors by saliva sandwich ELISA (Supplementary Fig. S2C), followed by binding analysis of the FG88 mAbs to erythrocytes from a Lea-positive donor. Neither FG88 mAb bound to erythrocytes (Supplementary Fig. S2D). As it is difficult to use primary tumors for cytotoxic experiments due to their high rate of spontaneous apoptosis, we screened cancer cell lines as models for FG88 mAb binding and killing. High FG88.2-binding cancer cells comprised COLO205 and HCT-15 [geometric mean (Gm) 1,000 at saturating mAb concentration]. The antigen density (SABC) was calculated to be 618,000 and 902,000 for HCT-15 and COLO205, respectively. Moderately binding cells (Gm 100) included HT29 and OVCAR3 (SABC: 88,000 and 23,000, respectively) and low binding cells (Gm < 100) encompassed AGS and DMS79 (SABC: 12,000 and 10,000, respectively; Fig. 2D).





FG88 mAbs mediate direct tumor cytotoxicity via a unique caspase-independent mechanism In order to ascertain whether cancer cell binding by the FG88 mAbs affected cell viability, we analyzed mAb-induced PI uptake that reflects compromised membrane integrity and ensuing loss of cell viability (Fig. 3A and Supplementary Fig. S3A). FG88.2 induced dose-dependent PI uptake, with strong binding cells, such as HCT-15 and COLO205, being more susceptible than the moderate to weak FG88-binding cells. In addition, proliferation analysis (WST8-based) of mAb-treated HCT-15 cells revealed a dose-dependent growth inhibition with an IC50 of 1.04 108 mol/L (Supplementary Fig. S3B).

Classical extrinsic apoptotic cell death involves activation of effector caspases. We thus evaluated caspase activity in FG88.2induced cytotoxicity via cellular proliferation analysis in the presence and absence of the cell permeant pan-caspase inhibitor Z-FMK-VAD. FG88.2 binding to HCT-15 cells resulted in a dosedependent decrease in cellular proliferation that was not affected by the presence of Z-FMK-VAD, suggesting no involvement of caspase activity (Fig. 3B). In contrast, Z-FMK-VAD prevented Fas mAb-mediated apoptosis of Jurkat cells under similar conditions (data not shown). Another hallmark of apoptotic cell death is endonuclease-induced DNA fragmentation. FG88.2 did not induce DNA fragmentation in HCT-15 cells, in contrast to the

Figure 3. Direct cytotoxic activity of FG88 mAbs. A, dose-dependent increase in PI uptake by a range of tumor cell lines as a result of FG88.2 binding (7.4 9 7 10 –2 10 mol/L). B, dose3 dependent growth inhibition ( Hthymidine incorporation) of HCT-15 8 10 cells by FG88.2 (2.2 10 –2.7 10 mol/L) in the presence or absence of Z-FMK-VAD (20 mmol/L). Nonlinear regression was performed using GraphPad Prism 6. C, phase-contrast imaging of FG88 mAb-treated HCT-15 cells. Images (magnification, 10) showing HCT-15 cells after incubation with FG88.2 and FG88.7, both at 7 1.3 10 mol/L, and RPMI medium (negative control). D, SEM analysis of FG88.2-induced ultrastructural cellular changes. HCT-15 cells were treated with 7 FG88.2 mAb (2 10 mol/L) for 2 or 20 hours. Medium alone (RPMI) and 0.5% (v/v) H2O2 were used as negative and positive controls, respectively. Arrows indicate mAb-induced pores.




2969


Chua et al.

anti-Fas mAb that resulted in Z-FMK-VAD–sensitive DNA fragmentation in Jurkat cells (Supplementary Fig. S3C). The absence of caspase involvement combined with a lack of DNA fragmentation suggests that FG88.2 induces a direct cytotoxic effect via a distinct nonapoptotic mechanism. Next, light microscopy and SEM were used to evaluate FG88.2induced ultrastructural changes. Light microscopy evaluation of mAb-treated HCT-15 cells showed evidence of monolayer disruption, cell rounding, and clumping within 24 hours of mAb addition and a decrease in cell numbers between 24 and 48 hours after mAb addition. This was maintained over the 72-hour incubation period (Fig. 3C). FG88.2-treated HCT-15 cellular aggregates displayed a loss of surface microvilli and the formation of membrane blebs and surface wrinkles (Fig. 3D). Importantly, cellsurface pore formation suggests a cell death mechanism reminiscent of oncosis. Heterogeneous pore sizes with diameters ranging from 0.2 to 1 mm were observed after 2 hours as well as 20 hours. FG88.7 displayed similar characteristics to FG88.2 with respect to cell binding and cytotoxicity (data not shown). FG88 mAbs internalize efficiently Confocal microscopy of Alexa Fluor 488–labeled FG88.2 binding to HCT-15 cells over a 2-hour period showed efficient internalization of a proportion of the mAbs (Fig. 4A). In addition, the FG88 membrane staining pattern suggests heterogeneous distribution of the glyco-epitope in the HCT-15 plasma membrane. Over time, internalized FG88 mAbs colocalized with lysosomal compartments (Fig. 4A). Similar results were obtained with FG88.7 (data not shown). Importantly, internalization was validated through toxicity of Fab–ZAP–FG88 immune complexes

containing saporin (29). Internalization of the Fab–ZAP–FG88.2 and Fab–ZAP–FG88.7 complexes, but not the Fab–ZAP alone or the Fab–ZAP preincubated with a control mAb (data not shown), led to a dose-dependent (IC50 1011 mol/L) decrease in cell viability of the high glyco-epitope–expressing HCT-15 cells (Fig. 4B). The moderately binding HT29 cells were more refractory. FG88 mAbs exhibit excellent immune-mediated cancer cell killing (ADCC and CDC) in vitro The ability of the FG88 mAbs to induce COLO205 tumor cell death in the presence of human PBMCs through ADCC was investigated. Both FG88 mAbs induced potent cell lysis of the high-binding COLO205 cells in a concentration-dependent manner with EC50 values of 109 mol/L and near 100% lysis at 7.4 109 mol/L (Fig. 5A). Next, we analyzed a range of tumor cell lines for their susceptibility to FG88-mediated ADCC. The FG88 mAbs significantly lysed the high glyco-epitope–expressing COLO205 and HCT-15 above the killing observed with PBMCs alone (Fig. 5B). The mAb 791T/36, a murine IgG2b that cannot bind human CD16 (30), showed no significant killing over the background observed with PBMCs alone, in the majority of cell lines. PBMC killing in the absence of FG88 mAbs was highest for cell lines lacking MHC-I, such as HCT-15 and AGS, and probably reflects NK killing. Noticeably, less immune-mediated killing was seen with the FG88 mAbs on the moderate to weak binding HT29 and DMS79 cells even at high mAb concentration 6.7 108 M. The low-binding OVCAR3 and AGS cells were refractory. In addition, the FG88 mAbs were very efficient at inducing CDC of high-binding HCT-15 cells in the presence of human

A

60 min

B

80 min

i

100 min

HCT15

ii

75 50

0

HT29 100

% Control

% Control

100

25

120 min

FG88.7 FG88.2 –14 –13 –12 –11 –10 –9 log conc (mol/L)

75 50 25 0

–8

FG88.7 FG88.2 –14 –13 –12 –11 –10 –9 log conc (mol/L)

–8

Figure 4. Cellular internalization of FG88 mAbs. A, time-lapse confocal microscopy of Alexa Fluor 488–labeled FG88 mAbs internalizing into live HCT-15 cells. Merged images taken every 20 minutes demonstrate FG88 internalization and colocalization with lysosomal compartments (arrows). The nucleus was labeled with Hoechst 33258, lysosomal compartments with LysoTracker Deep Red and the plasma membrane with CellMask Orange. Magnification, 60. B, targeted toxin (saporin) delivery by internalized FG88 mAbs in two cancer cell lines [HCT-15 (i) and HT29 (ii)]. The antiproliferative effect of internalized FG88 mAbs preincubated 3 with a saporin-linked anti-mouse IgG Fab fragment was evaluated using H-thymidine incorporation. Concentrations refer to the FG88 mAbs. Results (mean STD from three independent experiments) are presented as a percentage of proliferation of cells treated with primary mAbs only; normalized values are shown for HCT-15. Nonlinear regression was performed using GraphPad Prism 6.






Discussion We have generated two antiglycolipid IgG3 mAbs through immunization of mice with COLO205-derived membrane lipids that cross-react with glyco-epitopes expressed on a range of glycoproteins. Glycan array and ELISA analysis showed that the overall glycoreactivity of FG88.2 and FG88.7 was similar. Both mAbs showed a preference for extended type I chain nonsialylated Lea-containing carbohydrates with LecLex (galb1-3GLcNacb13Galb1-4(Fuca1-3)GlcNAc) not previously described as a target for cancer therapy. Circulating Lea-containing glycolipids can be adsorbed by erythrocytes depending on the secretor status of individuals. Importantly, we found no reactivity of the FG88 mAbs with erythrocytes from Lea-positive human donors, suggesting that the FG88 preference for more complex Lea-containing glycans precludes erythrocyte reactivity. By relating antigen density by IHC to the density required for cell killing, it was observed that tumors needed to stain moderately to strongly (2–3) to be targets for FG88.2. This would include 39% to 53% of gastrointestinal cancers. Importantly, the significant association of strong FG88.2 binding with poor outcome in our colorectal cancer cohort, independent of stage and vascular invasion, suggests that the most aggressive cancers would benefit from FG88.2 therapy. Earlier work has demonstrated the prognostic value of Lex, SLex, and SLea expression in colorectal carcinomas and the association of SLea expression with increased metastatic potential (31–34), but our study is the first to demonstrate independent prognostic

B

ADCC

100 ***


***

% Cell lysis

FG88.2 FG88.7 –10 –9 log conc (mol/L)

25

D

100

100

75 50 25

FG88.7 FG88.2

0 –11

–10 –9 log conc (mol/L)

–8

% Cell lysis

C

*** ***

0

–8

C O

–11

*****

A R 3

0

50

A G S

25

75

T2 9

50

FG88.2 FG88.7 791T/36 PBMC

*** ***

20 5 H C T15

75

LO

% Cell lysis

100

% Cell lysis

Figure 5. In vitro immune effector activity of FG88 mAbs. A, dose-dependent ADCC activity of FG88 mAbs on 51 COLO205 cells. Cr-labeled COLO205 cells were coincubated with increasing concentrations of FG88 mAbs and human PBMCs (E:T of 100:1). B, ADCC activity of FG88 mAbs on a panel of tumor cell lines. FG88 mAbs and control (791T/36) 8 were used at 6.7 10 mol/L. C, dose-dependent CDC activity of 51 FG88 mAbs on HCT-15 cells. The Crlabeled HCT-15 cells were incubated in the presence of increasing concentrations of FG88 mAbs in the presence of human serum. D, CDC activity of FG88 mAbs on a panel of tumor cell lines. FG88 mAbs and control (791T/36) were used 8 6.7 10 mol/L. Significance (B and D) versus PBMC or serum control, respectively, was established by ANOVA followed by the Bonferroni multiple comparisons test, GraphPad Prism 6.

H

A

O VC

Potent in vivo antitumor activity by the FG88 mAbs in a human hepatic metastasis xenograft model The mouse HCT-15HM2 DLuX human hepatic metastasis tumor model was used to investigate the antitumor activity of the FG88 mAbs. The HCT-15HM2 DLuX cell line is a bioluminescent variant of a liver metastasizing HCT-15 cell line. Carcinoma establishment, growth, and metastasis were assessed noninvasively via optical imaging. FG88 mAb treatment was initiated 10 days following tumor cell implantation and metastasis to the liver. The mAbs (100 mg) or vehicle (PBS) was administered i.v. twice weekly, for 120 days. FG88.2 and FG88.7 significantly reduced tumor growth compared with vehicle control, by day 59 (P ¼ 0.016, Mann–Whitney U test) and day 65 (P ¼ 0.046, Mann–Whitney U test), respectively (Fig. 6A). FG88.7 mAb treat-

ment led to a significant survival benefit [P ¼ 0.0037; HR, 3.06; 95% confidence interval (95% CI), 2.26–17.39, log-rank test] compared with vehicle (Fig. 6B).

D M S7 9

complement, with nanomolar EC50 and over 80% lysis at 7.4 109 mol/L mAb (Fig. 5C). The FG88 mAbs displayed significant CDC activity against COLO205 cells and to a lesser degree DMS79 cells (Fig. 5D). No or little CDC was seen on the low- to moderatebinding cell lines HT29, OVCAR3, and AGS (data not shown). In order to relate tumor staining by IHC to the antigen density requirement for cell killing, HCT-15 xenograft tumor tissue was stained with FG88.2 (data not shown). From this analysis, the antigen density required for cell killing by our mAbs was comparable with a score of 2 to 3 (moderate to strong) in the IHC analysis of the tumor tissues and accounts for 39% to 53% of gastrointestinal cancers (Supplementary Table S3). Some of the normal gastrointestinal tissues also showed weak to moderate staining but this is mainly apical cells and it is unclear if mAb will be accessible to these cells.

75

FG88.2 FG88.7 791T/36 Serum

*** ***

*** ***

50 25 0

*** ***

COLO205 HCT-15

DMS79



2971


Chua et al.

value for Lea/c expression in colorectal cancer as defined by our FG88.2 mAb. Some of the normal gastrointestinal tissues also showed weak to moderate staining but this is mainly apical cells and it is unclear if mab will be accessible to these cells. Indeed, SC104 mab showed similar staining but showed not toxicity in primate models and no toxicity in clinical trials (NCT01447732, SC104/CEP-37250/ KHK2804). Lewis antigens are only synthesized by primates due to FUT3/4 expression. In this context, it was of interest that there was a significant association between FG88.2 binding and FUT3 expression in the colorectal cancer cohort. Formal toxicity studies would thus need to be done in CN monkeys. Staining of a CN monkey normal TMA with FG88.2 showed a similar binding pattern to the normal human tissues. The FG88 mAbs exhibited a direct growth-inhibitory effect (IC50 108 mol/L) on high glyco-epitope–expressing tumor cells. This was not evidenced on low or moderately expressing tumor cells, suggesting a contribution of the high functional affinity with which the FG88 mAbs bind to tumor cell surfaces. The lack of cell killing of moderate to low glyco-epitope–expressing cells offers a further level of protection for normal tissues. The growth-inhibitory effect of the FG88 mAbs was characterized by cellular aggregation followed by an effect on cell membrane integrity and pore formation. More detailed characterization of the cytotoxic effect revealed absence of DNA fragmentation and limited involvement of effector caspases, suggesting a nonapoptotic mechanism. A number of other antiglycan mAbs have been shown to induce nonapoptotic oncosis-like cell death (35–38). Anti-CD20 mAbs in the treatment of B-cell lymphoma have been classified functionally in type I and II mAbs. Type II anti-CD20 mAbs, such as tositumomab, are effective at mediating direct cell killing in a caspase-independent manner, which was associated with homotypic cell adhesion and exhibits many similarities with the FG88 mAbs direct cell killing (39). In both cases, release of cellular content into the extracellular space via mAb-induced pore formation, may constitute a form of immunogenic cancer cell death (ICD) through the release of danger-associated molecular patterns ("DAMPS"; ref. 40). This ICD may reinitiate immune responses in the immunosuppressive tumor microenvironment.

A

In contrast, SC104, which recognizes sialylated Lea-related glycans, was also cytotoxic to high-binding cells, such as HCT-15 and COLO205, but with a higher IC50 (2 107 mol/L) and via classical apoptosis with evidence of caspase activation (20). Two independent approaches demonstrated efficient internalization and lysosomal localization of a proportion of the FG88 mAbs in a glyco-epitope density-dependent manner. This indicates their potential clinical usefulness for ADC development. Furthermore, a fraction of the FG88 mAbs remains at the cell surface where their slow dissociation kinetics enables excellent immune-effector functions (ADCC and CDC). Importantly, moderate- to weak-binding cells were less susceptible to ADCC and CDC, offering a further level of protection. Carbohydrate-recognizing mAbs generally exhibit weak glycan affinity. In contrast, the FG88 mAbs displayed strong functional affinity for a high-density Lea–HSA glycoconjugate but nanomolar functional affinity for tumor cell lines overexpressing the Lea -containing glyco-epitope such as COLO205 and HCT-15. This probably reflects the more complex binding behavior of the mAbs at the cell surface where the mixed glyco-epitope (glycolipid and glycoprotein) recognition, the fluid membrane, or the potential for internalization and cell killing can explain the higher cellsurface Kd. Preliminary studies suggest that the glycolipids are responsible for the direct cell killing and ADCC/CDC, requiring high levels of mAb, 108 and 109 mol/L, respectively. In contrast, internalization occurs predominantly via glycoproteins and at subnanomolar (1011 mol/L) concentrations of mAb. This suggests that targeting glycans shared between glycoproteins and glycolipids may be ideal for ADC and direct/immune effector cell killing, respectively; potentially debulking the tumor via ADC followed by removing residual disease with mAb alone. The excellent in vitro cytotoxicity of both mAbs translated in potent antitumor efficacy and significant survival improvement in a colorectal hepatic metastasis xenograft model in which mAb treatment was initiated 10 days after tumor initiation and development of liver metastasis. The FG88 mAbs cured both intraperitoneal disease and liver metastases in 30% of animals. Postmortem imaging analysis of those animals showed very limited evidence of tumor regrowth. Previous xenografts studies have

B

100 FG88.7 Vehicle

75

FG88.2 (1 mg/mL) FG88.7 (1 mg/mL) PBS vehicle

50 25 0 0 12 5 15 0 17 5 20 0 22 5 25 0 27 5 30 0

20 40 60 Days after tumor implantation

10

0

*

50 75

*

0 25

160,000 140,000 120,000 100,000 80,000 60,000 40,000 20,000 0

% Survival

Mean % growth

FG88.2

Day

Figure 6. In vivo activity of FG88 mAbs in the mouse colorectal hepatic metastasis model. A, reduction in percentage tumor growth (ventral bioluminescent signal) by FG88 mAbs. FG88.2 and FG88.7 significantly reduced tumor growth by day 59 (¼ 0.016) and day 65 (P ¼ 0.046), respectively, compared with the vehicle (PBS) control (Mann–Whitney U test). B, survival benefit due to antitumor activity of FG88.7 and FG88.2. The HCT-15HM2 DLuX mouse xenograft model was used to compare the antitumor effect of the FG88 mAbs to vehicle control (group size n > 8). MF-1 nude mice received mAbs (100 mg) i.v. biweekly, treatment was initiated 10 days after tumor inoculation and the final dose was administered on day 120 (arrow).






shown good antitumor efficacy when treatment was initiated at the time of tumor implantation or shortly afterwards (18, 41). No studies, to date, have shown tumor eradication of 10-day established tumors and liver metastases. SC104 inhibited tumor growth in vivo, but was more effective in the tumor prevention model and only worked in the therapeutic models in combination with 5-FU/leucovorin (21). The humanized SC104 (NCT01447732) currently in phase I/II trials (SC104/CEP37250/KHK2804) has been defucosylated for enhanced effector functions. Further studies have been initiated to evaluate the therapeutic potential of human IgG1 chimeric versions of the FG88 mAbs and the initial results are encouraging.

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): J.X. Chua, M. Vankemmelbeke, R.S. McIntosh, P.A. Clarke, L.G. Durrant Writing, review, and/or revision of the manuscript: J.X. Chua, M. Vankemmelbeke, R.S. McIntosh, T. Parsons, I. Spendlove, S. Madhusudan, L.G. Durrant Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J.X. Chua, R. Moss, A.M. Zaitoun, S. Madhusudan, L.G. Durrant Study supervision: I. Spendlove, L.G. Durrant

Disclosure of Potential Conflicts of Interest

Grant Support

No potential conflicts of interest were disclosed.

Authors' Contributions Conception and design: J.X. Chua, T. Parsons, I. Spendlove, L.G. Durrant Development of methodology: J.X. Chua, M. Vankemmelbeke, R.S. McIntosh, P.A. Clarke, R. Moss, T. Parsons, I. Spendlove, L.G. Durrant Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J.X. Chua, M. Vankemmelbeke, R.S. McIntosh, P.A. Clarke, R. Moss, A.M. Zaitoun

Acknowledgments The authors acknowledge the CFG for glycan array analysis and Tim Self (University of Nottingham) for confocal microscopy.

This work was supported by the University of Nottingham Strategic Development Fund (to L.G. Durrant). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received November 24, 2014; revised February 27, 2015; accepted February 28, 2015; published OnlineFirst March 16, 2015.

References 1. Christiansen MN, Chik J, Lee L, Anugraham M, Abrahams JL, Packer NH. Cell surface protein glycosylation in cancer. Proteomics 2014;14:525–46. 2. Dalziel M, Crispin M, Scanlan CN, Zitzmann N, Dwek RA. Emerging principles for the therapeutic exploitation of glycosylation. Science 2014; 343:1235681. 3. Daniotti JL, Vilcaes AA, Torres Demichelis V, Ruggiero FM, RodriguezWalker M. Glycosylation of glycolipids in cancer: basis for development of novel therapeutic approaches. Front Oncol 2013;3:306. 4. Hakomori S. Glycosylation defining cancer malignancy: new wine in an old bottle. Proc Natl Acad Sci U S A 2002;99:10231–3. 5. Fuster MM, Esko JD. The sweet and sour of cancer: glycans as novel therapeutic targets. Nat Rev Cancer 2005;5:526–42. 6. Hakomori SI, Cummings RD. Glycosylation effects on cancer development. Glycoconj J 2012;29:565–6. 7. Stanley P, Cummings RD. Structures common to different glycans. In:Varki A, Cummings RD, Esko JD, et al., editors. Essentials of glycobiology. Cold Spring Harbor Laboratory Press: Cold Spring Harbor, New York; 2009. 8. Hakomori S, Andrews HD. Sphingoglycolipids with Leb activity, and the co-presence of Lea-, Leb-glycolipids in human tumor tissue. Biochim Biophys Acta 1970;202:225–8. 9. Inagaki H, Sakamoto J, Nakazato H, Bishop AE, Yura J. Expression of Lewis (a), Lewis(b), and sialated Lewis(a) antigens in early and advanced human gastric cancers. J Surg Oncol 1990;44:208–13. 10. Sakamoto J, Furukawa K, Cordon-Cardo C, Yin BW, Rettig WJ, Oettgen HF, et al. Expression of Lewisa, Lewisb, X, and Y blood group antigens in human colonic tumors and normal tissue and in human tumor-derived cell lines. Cancer Res 1986;46:1553–61. 11. Blaszczyk M, Pak KY, Herlyn M, Sears HF, Steplewski Z. Characterization of Lewis antigens in normal colon and gastrointestinal adenocarcinomas. Proc Natl Acad Sci U S A 1985;82:3552–6. 12. Cazet A, Julien S, Bobowski M, Burchell J, Delannoy P. Tumourassociated carbohydrate antigens in breast cancer. Breast Cancer Res 2010;12:204. 13. Durrant LG, Noble P, Spendlove I. Immunology in the clinic review series; focus on cancer: glycolipids as targets for tumour immunotherapy. Clin Exp Immunol 2012;167:206–15. 14. Rabu C, McIntosh R, Jurasova Z, Durrant L. Glycans as targets for therapeutic antitumor antibodies. Future Oncol 2012;8:943–60. 15. Yang MC, Hsu FL, Handa K, Hakomori S, Lee MH, Liu LY, et al. Human monoclonal antibody GNX-8 directed to extended type 1 chain: specific binding to human colorectal cancer. Int J Cancer. 2009 Dec 21; doi:10.1002/ijc.25131.


16. Capurro M, Ballare C, Bover L, Portela P, Mordoh J. Differential lytic and agglutinating activity of the anti-Lewis(x) monoclonal antibody FC2.15 on human polymorphonuclear neutrophils and MCF-7 breast tumor cells. In vitro and ex vivo studies. Cancer Immunol Immunother 1999;48:100–8. 17. Ballare C, Portela P, Schiaffi J, Yomha R, Mordoh J. Reactivity of monoclonal antibody FC-2.15 against drug resistant breast cancer cells. Additive cytotoxicity of adriamycin and taxol with FC-2.15. Breast Cancer Res Treat 1998;47:163–70. 18. Watanabe M, Ohishi T, Kuzuoka M, Nudelman ED, Stroud MR, Kubota T, et al. In vitro and in vivo antitumor effects of murine monoclonal antibody NCC-ST-421 reacting with dimeric Le(a) (Le(a)/Le(a)) epitope. Cancer Res 1991;51:2199–204. 19. Stroud MR, Levery SB, Martensson S, Salyan ME, Clausen H, Hakomori S. Human tumor-associated Le(a)-Le(x) hybrid carbohydrate antigen IV3 (Gal beta 1–>3[Fuc alpha 1–>4]GlcNAc)III3FucnLc4 defined by monoclonal antibody 43-9F: enzymatic synthesis, structural characterization, and comparative reactivity with various antibodies. Biochemistry 1994;33: 10672–80. 20. Durrant LG, Harding SJ, Green NH, Buckberry LD, Parsons T. A new anticancer glycolipid monoclonal antibody, SC104, which directly induces tumor cell apoptosis. Cancer Res 2006;66:5901–9. 21. Sawada R, Sun SM, Wu X, Hong F, Ragupathi G, Livingston PO, et al. Human monoclonal antibodies to sialyl-Lewis (CA19.9) with potent CDC, ADCC, and antitumor activity. Clin Cancer Res 2011; 17:1024–32. 22. Watson SA, Morris TM, Crosbee DM, Hardcastle JD. A hepatic invasive human colorectal xenograft model. Eur J Cancer 1993;29A:1740–5. 23. Noble P, Spendlove I, Harding S, Parsons T, Durrant LG. Therapeutic targeting of Lewis(y) and Lewis(b) with a novel monoclonal antibody 692/ 29. PLoS One 2013;8:e54892. 24. Camp RL, Dolled-Filhart M, Rimm DL. X-tile: a new bio-informatics tool for biomarker assessment and outcome-based cut-point optimization. Clin Cancer Res 2004;10:7252–9. 25. Watson NF, Madjd Z, Scrimegour D, Spendlove I, Ellis IO, Scholefield JH, et al. Evidence that the p53 negative/Bcl-2 positive phenotype is an independent indicator of good prognosis in colorectal cancer: a tissue microarray study of 460 patients. World J Surg Oncol 2005; 3:47. 26. Duncan TJ, Rolland P, Deen S, Scott IV, Liu DT, Spendlove I, et al. Loss of IFN gamma receptor is an independent prognostic factor in ovarian cancer. Clin Cancer Res 2007;13:4139–45.



2973


Chua et al.

27. Abdel-Fatah T, Arora A, Gorguc I, Abbotts R, Beebeejaun S, Storr S, et al. Are DNA repair factors promising biomarkers for personalized therapy in gastric cancer? Antioxid Redox Signal 2013;18:2392–8. 28. Storr SJ, Zaitoun AM, Arora A, Durrant LG, Lobo DN, Madhusudan S, et al. Calpain system protein expression in carcinomas of the pancreas, bile duct and ampulla. BMC Cancer 2012;12:511. 29. Kohls MD, Lappi DA. Mab-ZAP: a tool for evaluating antibody efficacy for use in an immunotoxin. BioTechniques 2000;28:162–5. 30. Price MR, Pimm MV, Page CM, Armitage NC, Hardcastle JD, Baldwin RW. Immunolocalization of the murine monoclonal antibody, 791T/36 within primary human colorectal carcinomas and identification of the target antigen. Br J Cancer 1984;49:809–12. 31. Nakagoe T, Fukushima K, Nanashima A, Sawai T, Tsuji T, Jibiki M, et al. Expression of Lewis(a), sialyl Lewis(a), Lewis(x) and sialyl Lewis(x) antigens as prognostic factors in patients with colorectal cancer. Can J Gastroenterol 2000;14:753–60. 32. Kannagi R. Carbohydrate antigen sialyl Lewis a—its pathophysiological significance and induction mechanism in cancer progression. Chang Gung Med J 2007;30:189–209. 33. Weston BW, Hiller KM, Mayben JP, Manousos GA, Bendt KM, Liu R, et al. Expression of human alpha(1,3)fucosyltransferase antisense sequences inhibits selectin-mediated adhesion and liver metastasis of colon carcinoma cells. Cancer Res 1999;59:2127–35. 34. Inagaki Y, Gao J, Song P, Kokudo N, Nakata M, Tang W. Clinicopathological utility of sialoglycoconjugates in diagnosing and treating colorectal cancer. World J Gastroenterol 2014;20:6123–32. 35. Loo D, Pryer N, Young P, Liang T, Coberly S, King KL, et al. The glycotopespecific RAV12 monoclonal antibody induces oncosis in vitro and has antitumor activity against gastrointestinal adenocarcinoma tumor xenografts in vivo. Mol Cancer Ther 2007;6:856–65.


36. Tan HL, Fong WJ, Lee EH, Yap M, Choo A. mAb 84, a cytotoxic antibody that kills undifferentiated human embryonic stem cells via oncosis. Stem Cells 2009;27:1792–801. 37. Zhang C, Xu Y, Gu J, Schlossman SF. A cell surface receptor defined by a mAb mediates a unique type of cell death similar to oncosis. Proc Natl Acad Sci U S A 1998;95:6290–5. 38. Hernandez AM, Rodriguez N, Gonzalez JE, Reyes E, Rondon T, Grinan T, et al. Anti-NeuGcGM3 antibodies, actively elicited by idiotypic vaccination in nonsmall cell lung cancer patients, induce tumor cell death by an oncosis-like mechanism. J Immunol 2011; 186:3735–44. 39. Alduaij W, Ivanov A, Honeychurch J, Cheadle EJ, Potluri S, Lim SH, et al. Novel type II anti-CD20 monoclonal antibody (GA101) evokes homotypic adhesion and actin-dependent, lysosome-mediated cell death in B-cell malignancies. Blood 2011;117:4519–29. 40. Krysko O, Love Aaes T, Bachert C, Vandenabeele P, Krysko DV. Many faces of DAMPs in cancer therapy. Cell Death Dis 2013;4:e631. 41. Schreiber GJ, Hellstrom KE, Hellstrom I. An unmodified anticarcinoma antibody, BR96, localizes to and inhibits the outgrowth of human tumors in nude mice. Cancer Res 1992;52:3262–6. 42. Consortium for Functional Glycomics FGG. FG88.2 on core H, version 5.1 glycan array. [Internet]. Functional Glycomics Gateway; c2010. Available from: http://www.functionalglycomics.org/glycomics/HServlet? operation¼view&sideMenu¼no&psId¼primscreen_5844. Accessed April 8, 2015. 43. Consortium for Functional Glycomics FGG. FG88.7 on core H, version 5.1 glycan array. [Internet]. Functional Glycomics Gateway; c2010. Available from: http://www.functionalglycomics.org/glycomics/HServlet? operation¼view&sideMenu¼no&psId=primscreen_5847. Accessed April 8, 2015.




Monoclonal Antibodies Targeting LecLex-Related Glycans with Potent Antitumor Activity Jia Xin Chua, Mireille Vankemmelbeke, Richard S. McIntosh, et al. Clin Cancer Res 2015;21:2963-2974. Published OnlineFirst March 16, 2015.

Updated version Supplementary Material

Cited articles

E-mail alerts Reprints and Subscriptions Permissions

Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-14-3030 Access the most recent supplemental material at: http://clincancerres.aacrjournals.org/content/suppl/2015/03/18/1078-0432.CCR-14-3030.DC1.html

This article cites 39 articles, 15 of which you can access for free at: http://clincancerres.aacrjournals.org/content/21/13/2963.full.html#ref-list-1

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].


Characterization of a new dual-targeting fully human antibody with potent antitumor activity against nasopharyngeal carcinoma.

Affinity of monoclonal antibodies for Globo-series glycans.

Targeting immunosuppression by Tregs with monoclonal antibodies against GARP.

A DNA-binding Molecule Targeting the Adaptive Hypoxic Response in Multiple Myeloma Has Potent Antitumor Activity.

Inhibitors of SRC kinases impair antitumor activity of anti-CD20 monoclonal antibodies.

Potent neutralizing monoclonal antibodies against Ebola virus infection.

A Inhibit Catalytic Activity.

BCMab1, a monoclonal antibody against aberrantly glycosylated integrin α3β1, has potent antitumor activity of bladder cancer in vivo.

Novel quinonyl derivatives of muramyl dipeptide possessing potent antitumor activity.

Trial Watch: Tumor-targeting monoclonal antibodies in cancer therapy.

Macaque Monoclonal Antibodies Targeting Novel Conserved Epitopes within Filovirus Glycoprotein.

Immunological intervention with monoclonal antibodies.

Rhamnazin, a novel inhibitor of VEGFR2 signaling with potent antiangiogenic activity and antitumor efficacy.

Synthesis and biological evaluation of diarylthiazole derivatives as antimitotic and antivascular agents with potent antitumor activity.

A chemically defined trifunctional antibody-cytokine-drug conjugate with potent antitumor activity.

A Novel Time-Dependent CENP-E Inhibitor with Potent Antitumor Activity.

An antimesothelin-monomethyl auristatin e conjugate with potent antitumor activity in ovarian, pancreatic, and mesothelioma models.

Mechanistic characterization of a copper containing thiosemicarbazone with potent antitumor activity.

Selective ALK inhibitor alectinib with potent antitumor activity in models of crizotinib resistance.

Highly Selective and Potent Thiophenes as PI3K Inhibitors with Oral Antitumor Activity.

Characterization of LY2228820 dimesylate, a potent and selective inhibitor of p38 MAPK with antitumor activity.

Targeting Human β-Microglobulin with Monoclonal Antibodies in Multiple Myeloma - A Potential in Treatment.

Highly optimized DNA vaccine targeting human telomerase reverse transcriptase stimulates potent antitumor immunity.

mTOR Signaling Displays Potent Antitumor Efficacy against Nonfunctioning Pituitary Adenomas.