a short treatment with galactomannan gm-ct-01 corrects ......2014/03/19 ·...

Cancer Therapy: Preclinical

A Short Treatment with Galactomannan GM-CT-01 Correctsthe Functions of Freshly Isolated Human Tumor–InfiltratingLymphocytes

Nathalie Demotte1,2, Ren�e Bigirimana1,2, Gr�egoire Wie€ers1,2, Vincent Stroobant1,2, Jean-Luc Squifflet3,Javier Carrasco6, Kris Thielemans5, Jean-Francois Baurain4, Patrick Van Der Smissen2,Pierre J. Courtoy2, and Pierre van der Bruggen1,2

AbstractPurpose: Several galectins are released by tumor cells and macrophages and accumulate in the tumor

microenvironment. Galectin-1 and -3 were found to bind to glycosylated receptors at the surface of tumor-

infiltrating lymphocytes (TIL), forming glycoprotein–galectin lattices that could reduce the motility and

therefore the functionality of surfacemolecules. In contrast to blood T cells, humanTIL showdefective IFN-gsecretion upon ex vivo stimulation.Wehave previously shown that extracellular galectin-3 participates in the

impairment of TIL functions. Indeed, disruption of glycoprotein–galectin-3 lattices using anti-galectin-3

antibodies, or N-acetyllactosamine as a competing sugar, boosted cytokine secretion by TIL. Here we have

tested a clinical grade galectin antagonist: GM-CT-01, a galactomannan obtained from guar gum reported to

be safe in more than 50 patients with cancer.

Experimental Design: TIL were isolated from human tumor ascites, treated for 2 to 20 hours with

galectin antagonists and tested for function.

Results: We found that GM-CT-01 boosts cytotoxicity of CD8þ TIL and their IFN-g secretion in a dose-

dependentmanner. Treating TIL obtained frompatientswith various cancers, during a fewhours, resulted in

an increased IFN-g secretion in up to 80% of the samples.

Conclusions: These observations pave the way for investigating the potential benefit of this galectin

antagonist in patients with cancer, alone or combined with cancer vaccination, in order to correct in vivo

impaired functions of TIL. Clin Cancer Res; 20(7); 1–11. �2014 AACR.

IntroductionAccumulation of human tumor–infiltrating lymphocytes

(TIL) is considered as a good prognostic factor (reviewed inWie€ers; ref. 1). Moreover, when gene expression was ana-lyzed in tumor samples of patients from 3 active immuni-zation clinical trials, a gene signature, including T-cell

markers, correlated with clinical responses (2–4). However,TIL freshly isolated—without in vitro expansion—from var-ious human tumor samples (melanomas, renal cell carci-nomas, ovarian, and pancreatic carcinomas) often proveddefective in lyzing relevant target cells and producing IFN-gupon stimulation. This contrasted with blood T cells iso-lated from the same patients, which were readily cytotoxicand showed robust IFN-g secretion (3, 5–9).

Several immunosuppressive mechanisms have been pro-posed to explain the impaired functions of freshly isolatedTIL. First, TIL can express inhibitory receptors that down-modulate T-cell activation upon antigen recognition, forexample PD1, KIR, BTLA, CTLA-4, Tim-3 (10–13). Blockingthese receptors with antibodies has been shown to prolongsurvival of T cells and to boost their proliferation uponactivation in vitro. Such antibodies have also shown theirefficacy in vivo, as an objective-response rate of 40% wasobserved in a trial where 53 patients with advanced mela-noma received combined injections of ipilimumab andnivolumab targeting CTLA-4 and PD-1, respectively (14).Second, soluble molecules present at the tumor site such asTGF-b and PGE2 can block T-cell function or activation (15,16). Third, enzymes such as IDO can deplete the tumormicroenvironment in tryptophane (17). Fourth, regulatory

Authors' Affiliations: 1Ludwig Institute for Cancer Research Brussels andWELBIO; 2de Duve Institute; Departments of 3Gynecology and 4Oncology,Cliniques Universitaires Saint-Luc, Universit�e catholique de Louvain; 5Lab-oratory of Molecular and Cellular Therapy, Department of Immunology-Physiology, Medical School of the Vrije Universiteit Brussel, Brussels; and6Department of Oncology, Grand Hopital de Charleroi, Charleroi, Belgium

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

Current address for G. Wie€ers: Department of Internal Medicine, Cliniquesuniversitaires Saint-Luc, Universit�e catholique de Louvain, 10 av. Hippo-crate, B-1200 Brussels, Belgium.

Corresponding Author: Pierre van der Bruggen, Ludwig Institute forCancer Research Brussels and de Duve Institute, Universit�e catholiquede Louvain, 74 avenue Hippocrate, UCL B1.74.03, B-1200 Brussels,Belgium. Phone: 32-2-7647431; Fax: 32-2-7629405; E-mail:[email protected]

doi: 10.1158/1078-0432.CCR-13-2459

�2014 American Association for Cancer Research.

ClinicalCancer

Research

www.aacrjournals.org OF1

Research. on August 1, 2021. © 2014 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

Published OnlineFirst February 13, 2014; DOI: 10.1158/1078-0432.CCR-13-2459

http://clincancerres.aacrjournals.org/

T cells (Treg) could downmodulate antitumor responses asshown in mice depleted in Treg, which were more prone toreject tumors (18–20).

Afifthpotentialmechanism is surface paralysis becauseofglycoprotein–galectin lattices. Galectins are lectins fre-quently secreted by tumor cells and macrophages (21,22), and extracellular galectins seem responsible for defi-cient T-cell functions (23). We have shown that TIL harborsurface galectin-1 and galectin-3, and that treating TIL withN-acetyllactosamine (LacNAc) or GCS-100, 2 galectinantagonists, boosts cytokine secretion (8, 24). It was alsoreported that extracellular galectin-1 and galectin-3 pro-mote apoptosis of T cells (25–29).

The identification of antigens recognized by T cells onhuman tumor cells has resulted in numerous clinical trialsinvolving vaccination of tumor-bearing patients withdefined tumor antigens (30). Although new generations ofvaccines might be more effective (31), so far only 5% of thevaccinated patients withmetastatic melanoma show a com-plete or partial clinical response (32, 33). The tumors of thepatients about to receive a vaccine already contain T cellsdirected against tumor antigens that are probably function-ally impaired (34–36). The low toxicity of therapeuticvaccination of cancer, provided the target antigen istumor-specific, justifies efforts to improve its efficacy. Apotential strategy could be to inject galectin antagonists soas to boost TIL function.

Because LacNAc has a very short half-life in vivo and GCS-100 was not accessible for a clinical trial, we searched foranother galectin antagonist available for clinical use. Wedecided to test GM-CT-01, a galactomannan of plant originthat was shown to bind to galectin-1 but at a site differentfrom the conventional galectin carbohydrate bindingdomain (37). No data have been reported yet about inter-actions of GM-CT-01 with galectin-3. GM-CT-01 (operation-

al name) or Davanat (trade name) is isolated from seeds ofthe cluster bean, Cyamopsis tetragonoloba, as guar gum, andsubjected to controlled partial chemical degradation. Its sizeis in the range of 50 kDa and its half-life is between 12 and 18hours as determined in Cynomolgus monkeys (P.G. Traber;personal communication). Injection of GM-CT-01 to tumor-bearing mice potentiated the antitumor activity of 5-fluoro-uracil (38). In tumor patients with colorectal carcinoma orcholangiocarcinoma, a phase I clinical trial was completedand 3 phase II open label clinical trials are ongoing. A total of57 patients have been injected and a reduction of 5-fluoro-uracil–related side effects was described (39, 40).

We here compared the effect of GM-CT-01 and of LacNAcon the cytotoxicity and cytokine secretion by CD8þ andCD4þ fresh human TIL. We also tested its therapeuticpotential in a murine tumor model.

Materials and MethodsCells, TIL, and reagents

Samples of patient-derived solid tumors, tumor ascites,and blood were collected after approval by the institutionalreview boards of all collaborating institutions. Ascites cellswere concentrated by centrifugation and either frozen at�80�C or resuspended in Iscove’s modified Dulbecco medi-um (IMDM; Life Technologies) supplemented with 0.24mmol/L L-asparagine, 0.55 mmol/L L-arginine, 1.5 mmol/LL-glutamine (AAG), 100 mg/mL streptomycin, 100 U/mLpenicillin, and30mg/mLgentamycin (Sigma-Aldrich; culturemedium), enriched by 2% human serum.Mononuclear cellswere isolated from blood of donors without cancer, moreprecisely patients with hemochromatosis, or from someascites samples by Lymphoprep (Axis-Shield PoCAS) andusually frozen at �80�C or in liquid nitrogen, in culturemedium containing 10% DMSO/45% human serum. Cellswere thawed for 2 hours in culture medium supplementedwith 2% to 10%human serum and 5U/mLDNase I (Sigma-Aldrich). T cells were isolated from bloodmononuclear cellsor ascites cells, by rosetting with sheep erythrocytes(BioM�erieux) previously treated with 2(2-aminoethyl) iso-thioureadihydrobromide (Sigma-Aldrich).CD8þ andCD4þ

T cells were isolated by a positive selection strategy usingimmunomagnetic beads (AutoMACS system, Miltenyi Bio-tec). Purity of CD4þ or CD8þ T-cell preparations wasbetween 90% and 95%. Autologous tumor cell lines wereestablished in our laboratory from total cells, or a CD2�

fraction, isolated from tumor samples. Cell lines were cul-tured in IMDM supplemented with AAG, 100 mg/mL strep-tomycin, 100 U/mL penicillin, 1/100 fungizone (Invitro-gen), and 30 mg/mL gentamycin (Sigma-Aldrich), and with10% FBS and ACL-4. Human recombinant interleukin (IL)-2was from Chiron Healthcare SAS, IL-7 from R&D Systems,and LacNAc from Carbosynth. GM-CT-01 was kindly pro-vided by Anatole Klyosov from Galectin Therapeutics.

T-cell functional assaysNonspecific T-cell stimulation was performed in round-

bottomed microwells with 10,000 T cells in 200 mL culturemedium containing IL-2 (6 IU/mL) as appropriate, with

Translational RelevanceThis study was prompted by the need for a galectin

antagonist approved for clinical use and able to correctthe functions of human tumor-infiltrating lymphocytes(TIL). We describe data obtained with GM-CT-01, aclinical grade galactomannan extracted from guar gumand reported to be safe in more than 50 patients withcancer. Treating TIL obtained from patients with variouscancers boosted IFN-g secretion upon ex vivo stimulationin �80% of the CD8þ TIL samples and �50% of theCD4þ TIL samples. Increased IFN-g secretion by CD8þ

TIL inducedbyGM-CT-01was concentration-dependentand correlated with cytotoxicity. It is remarkable, and sofar unique to galectin antagonists, that treating human Tcells ex vivo for only a few hours was sufficient to stronglyboost TIL functions. The responsiveness of the vastmajority of samples suggests that treatmentwith galectinantagonists, in particular GM-CT-01, could be effectivein patients with cancer to correct impaired TIL functions.

Demotte et al.

Clin Cancer Res; 20(7) April 1, 2014 Clinical Cancer ResearchOF2




3,000 beads coated with anti-CD3 and anti-CD28 antibo-dies (Dynabeads, Invitrogen) or with B-EBV cells previouslyincubated at 37�C for 60minutes with a cocktail of 1 mg/mLof superantigens SEA, SEB, and TSST-1 (Sigma), as indicat-ed. IFN-g secreted after overnight coculture was measuredby ELISA using Biosource Cytoset reagents (Invitrogen). Inthe cytotoxicity assay, the target cells were either murine P1.azar cells derived frommastocytoma cell line P815 (referredto as P815). P1.azar cells were obtained inour institute (41).Target cells were labeled for 1 hourwith 50mCi ofNa51CrO4

(Perkin Elmer LifeSciences), washed and incubated at roomtemperature (RT) for 15 minutes with 1 mg/mL anti-CD3mAb (OKT3, Mabtech). For the degranulation assay, CD8þ

T cells (75,000) were stimulated for 5 hours in IMDM 2%human serumAAG,with 150,000P815 cells incubatedwithanti-CD3 antibody or 75,000 CD3/CD28 beads. Brefeldin-A (GolgiPlug, BD) and FITC-coupled anti-CD107aþb(1/100, BD), or the control isotype [immunoglobulinG (IgG)1, BD],were also added. After 5hours of stimulationat 37�C, cells were washed, labeled at 4�C for 15 minuteswith anti-CD3.PerCP (1/40 SK1, BD) and anti-CD8.APC(1/40 RPA-T8, BD) antibodies, washed and fixed in PBS-PFA 1%. Cells were analyzed on a FACSCalibur (BD). Thepercentage of CD107þ cells was estimated for theCD3þCD8þ T cells. For the sorting of galectinhighLELhigh

cells, CD8þ T cells were labeled with 5 mg/mL biotinylatedrat monoclonal anti-galectin-3 antibody (M3/38, IgG2a,Biolegend) followed by neutravidin R-phycoerythrin con-jugate (1.25mg/mL; Invitrogen) combinedwithfluorescein-labeled lectin of Lycopersicon esculentum (LEL; 2 mg/mL;Vector). GalectinhighLELhigh cells and galectinlowLELlow cellswere sorted using a BD FACSAriaIII.

Detachment of galectinsCD8þ T cells, isolated from the ovarian carcinoma ascites

obtained frompatient LB3122,were incubated at 37�C for 2hours with LacNAc or GM-CT-01, washed, incubated at 4�Cfor 15 minutes with FcR Blocking Reagent (1/5; MiltenyiBiotec) diluted in PBS/BSA 0.2% (BSA, Sigma-Aldrich),washed again and incubated with 5 mg/mL of either bioti-nylated rat anti-galectin-3 antibody M3/38 or polyclonalrabbit anti-galectin-1 IgG (5 mg/mL; Abcam). Cells werewashed and incubated at 4�C for 15 minutes with eitherneutravidin R-phycoerythrin conjugate (1.25 mg/mL; Invi-trogen) or anti-rabbit Ig secondary antibody coupled toAlexa Fluor 488 (10 mg/mL; Invitrogen). Cells were alsolabeled with anti-CD3.PerCP (1/40; BD) and anti-CD8.APC (1/40; BD). After a final washing step, cells were fixedwith 2% formaldehyde in PBS and analyzed on aFACSCalibur.

Fluorescence resonance energy transfer microscopyCells were plated at 1 to 2 � 105/cm2 on poly-L-lysine

(Sigma-Aldrich)-coated glass coverslips and allowed tobind at RT for 7 minutes, then labeled for 30 minutes onice with anti-CD8a (UCHT-4, mouse IgG2a; Sigma-Aldrich) and anti-TCRb (IP-26, mouse IgG1; e-Bioscience)diluted in PBS/BSA 0.2%. After 3 washes in the same cold

buffer, cells were fixed at RT for 20 minutes with 4%formaldehyde and 0.1% glutaraldehyde (Sigma-Aldrich)in 0.1 mol/L phosphate buffer, washed twice in PBS/BSAand incubated with 10 mmol/L glycine in PBS for 10minutes, then incubated for 30 minutes on ice with ananti-IgG2a-Alexa Fluor 488 antibody [fluorescence reso-nance energy transfer (FRET) donor, green; Invitrogen],and an anti-IgG1-Alexa Fluor 568 antibody (FRET acceptor,red; Invitrogen) diluted in PBS/BSA. After 3 washes, cellswere fixed again and coverslips were mounted onto glassslides using Prolong Gold (Invitrogen). Images wereacquired with an LSM 510 laser scanning microscope andanalyzed by AIM Software (Zeiss). Imaging was performedwith a 488 nm line generated by an Ar laser (30 mW) and a561 nm line generated by a DPSS laser (10mW), both usedat 1%. Acceptor photobleaching was achieved using theDPSS laser at 100% with 100 iterations. Three images wererecorded before and after bleaching. To calculate theincrease in donor emission (indicative of FRET efficiency),3 regions of interest for each cell were chosen, bleached, andcompared with 3 control regions in the nonbleached area.Increase in donor emission was calculated as follows:%FRET efficiency ¼ [1 � (donor intensity before bleach-ing/donor intensity after bleaching)] � 100.

ResultsGM-CT-01 boosts IFN-g secretion by CD8þ TIL in aconcentration-dependent manner

CD8þT cellswere isolated fromhuman carcinomaascitesand incubated for 2 hourswith increasing concentrations ofGM-CT-01 orwith 5mmol/L LacNAc as a positive control. Tcells were next stimulated nonspecifically by beads coatedwith anti-CD3 and anti-CD28 antibodies (CD3/CD28beads) and, after 20 hours, IFN-g secretion was measuredin the culture supernatant. In this representative patient,preincubation of TIL with 0.3 mmol/L (�15 mg/mL) of GM-CT-01 boosted IFN-g secretion by more than 3-fold. Half-maximal effect was obtained at 0.4 mmol/L (�20 mg/mL;Fig. 1). For the experiments described below, GM-CT-01was used at concentrations in the micromolar range (either0.6 or 1.8 mmol/L �30 or 100 mg/mL), which yielded amaximal response equivalent to that of 5 mmol/L LacNAc,as we observed previously (8). GM-CT-01 was thus at least1,000 times more potent than LacNAc on molar basis.

No growth inhibition was observed when 2 T-cell cloneswere cultured for 7weeks in the presence of up to 2.6mmol/Lof GM-CT-01 (Supplementary Fig. S1). Two melanoma celllines were also cultured for 25 days with GM-CT-01. Growthinhibition was observed starting at day 7 for only one of themelanoma cell lines at the concentration of 2.6 mmol/L. Weconcluded that GM-CT-01 effects reported below are notaffected by toxicity.

GM-CT-01 boosts IFN-g secretion by CD8þ and CD4þ

TIL from patients with various cancersCD8þ TIL were isolated from ascites obtained from 27

patients bearing tumors of different histologic origins:

Galactomannan GM-CT-01 Corrects Dysfunctions of Human TIL

www.aacrjournals.org Clin Cancer Res; 20(7) April 1, 2014 OF3




melanoma, biliary tract, prostate, esophagus, liver, colon,pancreas, and ovary (Fig. 2). We also isolated CD4þ TILfrom 11 of the 27 ascites. T cells were incubated with GM-CT-01 or LacNAc for 2 or 20 hours and subsequentlystimulated overnight with CD3/CD28 beads. GM-CT-01boosted IFN-g secretion by more than 3-fold in 22 of theCD8þ TIL isolates (81%; Fig. 2A). Treatment with LacNAchad a similar effect, except for 2 patients that onlyresponded to LacNAc (Fig. 2A). Incubation of CD4þ TILin the presence of GM-CT-01 boosted IFN-g secretion bymore than 3-fold in 5 of the 11 cultures but had no effectfor the other isolates (Fig. 2B). GM-CT-01 treatment wasalso applied to T cells isolated from the blood of donorswithout cancer. CD8þ and CD4þ blood T cells wereincubated for 2 hours with GM-CT-01 before an overnightstimulation with CD3/CD28 beads (Fig. 2C). There wasno consistent effect of GM-CT-01 or LacNAc on IFN-gsecretion. We concluded that the galectin antagonist hadno effect on the IFN-g secretion by blood T lymphocytesfrom donors without cancer but induced a robustresponse in the majority of CD8þ TIL and in half ofCD4þ TIL, by rapid relaxation of galectin-sensitive repres-sion. In addition to IFN-g , the secretion of other cytokineswas also tested (Supplementary Table S1). In comparisonwith untreated cells, most samples of GM-CT-01- andLacNAc-treated cells also secreted more IL-2 and TNF-a,but we measured only very minor changes in IL-4 andIL-10.

GM-CT-01boosts IFN-g secretionbyCD8þTILobtainedfrom tumor ascites and solid tumor fromapatientwithovarian carcinoma

To further test this hypothesis, we searched for theopportunity to compare the effect ofGM-CT-01 andLacNAcon CD8þ and CD4þ T cells isolated from samples collectedon the sameday in the samepatient: the solid tumor, blood,and tumor ascites (Fig. 2D for IFN-g and SupplementaryTable S1 for other cytokines). IFN-g secretion by untreatedCD8þ and CD4þ TIL, whether from the solid tumor orascites, was low compared with IFN-g secretion by blood Tcells. The presence of GM-CT-01 boosted IFN-g secretion byCD8þ andCD4þ TIL bymore than 4-fold, whether TIL wereisolated from the solid tumor or from the ascites. LacNAchad a similar effect. CD8þ but not CD4þ blood T cellssecreted slightly more IFN-g in the presence of GM-CT-01.In patientswith ovarian carcinoma, as in patients with othertypes of cancer, elevated level of galectin-3 can be found inthe serum (42, 43). But serum galectins are possibly boundto serum glycoproteins and probably not capable to beloaded on circulating blood T cells in sufficient amount toblock their function.

GM-CT-01 boosts cytotoxicity of CD8þ TILBecause the specificity of CD8þ TIL was unknown, P815

cells coated with an anti-CD3 antibody were used as targetsin a cytotoxicity assay. For some TIL samples, we hadautologous tumor cell lines and used them also as targets.We tested CD8þ TIL isolated from 4 different ascites andCD8þ T cells isolated from the blood (PBL) of 3 differentdonorswithout cancer. Although the cytotoxicity of untreat-ed CD8þ TIL was minimal, treatment with GM-CT-01 orLacNAc triggered a significant cytotoxicity against bothP815 bearing anti-CD3 and autologous tumor cell targets(Fig. 3A). The cytotoxicity of CD8þ blood T cells, obtainedfrom donors without cancer, was �10-fold more potentthan that of TIL and was similar for treated and untreatedcells. Degranulation, a prerequisite for cytolysis, was alsomeasured. The percentage of degranulating cells was esti-mated by surface expression of CD107a and CD107b,which reside at rest in cytolytic granules and, upon activa-tion, are mobilized to the cell surface by exocytosis. Afterstimulation by P815 cells coveredwith anti-CD3 antibodiesor anti-CD3/CD28 beads, the percentage of CD107þ cellswas low in untreated TIL and strongly increased in TILtreated by GM-CT-01 or LacNAc, but high and equivalentfor treated and untreated blood CD8þ T cells (Fig. 3B andSupplementary Fig. S2). We concluded that GM-CT-01 wasas potent as LacNAc to restore cytotoxicity in fresh CD8þ

TIL.

The CD8þ TIL responding to galectin antagonistsharbor poly-LacNAc motifs and galectin-3

To explain how galectin antagonists boost TIL function,we hypothesized that (i) TIL had been recently activated bycontact with tumor antigen so that the N-glycans includelarger LacNAc oligomers—the natural galectin-3 ligands—on surface glycoproteins as compared with resting T cells

Figure 1. Concentration-dependent effect of GM-CT-01 on IFN-gsecretion by CD8þ TIL. CD8þ TIL were isolated from ascitesobtained from colon adenocarcinoma patient GHC9. CD8þ TIL(10,000 per microwell) were incubated for 2 hours in culture mediumsupplemented with the indicated concentration of GM-CT-01(filled circles), 5 mmol/L LacNAc (filled square), or left untreated(empty circle). T cells were subsequently stimulated overnight with3,000 CD3/CD28 beads. The concentration of IFN-g in the supernatantwas measured by ELISA. Values are means � SD of triplicatemicrowells. Secretion of IFN-g by nonstimulated CD8þ TIL, whethertreated or not, was <3 pg/mL of IFN-g . Similar results were obtained withCD8þ TIL samples isolated from 2 other patients.

Demotte et al.





Figure 2. Survey of the effect of GM-CT-01 on IFN-g secretion by TIL obtained from patients with different types of cancer. CD8þ (A) or CD4þ TIL(B)were isolated from tumor ascites. (C)CD8þorCD4þbloodTcells (PBL)were isolated fromblood samples obtained fromdonorswithout cancer. (D)CD8þorCD4þ T cells were isolated from a solid tumor, tumor ascites, or blood collected on the same day from patient with ovarian carcinoma LB3191.T cells (10,000 per microwell) were first incubated for 2 hours (patient code in bold) or 20 hours (patient code in italics) in culture medium supplementedwith either GM-CT-01 (0.6–1.8 mmol/L, black bars) or 5 mmol/L LacNAc (gray bars) or left untreated (white bars). T cells were then stimulated with3,000 CD3/CD28 beads in overnight culture as in Figure 1. IFN-g wasmeasured in the supernatant by ELISA. Values are means� SD of triplicate microwells.Secretion of IFN-g by nonstimulated CD8þ or CD4þ blood T cells, whether treated or not, was <15 pg/mL of IFN-g .






(44), and (ii) binding of extracellular galectin-3 on surfaceglycoproteins favors galectin-3-glycoprotein lattices, there-by reducing the motility of surface glycoproteins and con-sequently impairing TIL functions. This extracellular galec-tin-3 could be secreted, for example, by the activated T cellsor captured by contact with tumor cells or macrophagescovered by galectin-3.

We decided to test if TIL responding to galectin antago-nists were those harboring more LacNAc motifs and moregalectin-3. CD8 TIL were isolated from ascites obtainedfrom a patient with ovarian carcinoma, and double labeledwith an anti-galectin-3 antibody and with LEL, a lectin thatrecognizes LacNAc oligomers. Four subpopulations of TILexpressing different levels of galectin-3 and LEL were sortedby flow cytometry (Fig. 4). Cells from each subpopulationwere treated for 2 hours with GM-CT-01 or LacNAc, andstimulated overnight with CD3/CD28 beads. The top 8%ofthe TIL expressing galectin-3 and LEL ligandmotifs secretedalmost no IFN-g compared with galectin-3lowLELlow cells.Treatment with either GM-CT-01 or LacNAc had no effecton galectin-3lowLELlow TIL but boosted IFN-g secretion ofgalectin-3highLELhigh TIL up to the level of galectin-3low-

LELlow TIL (Fig. 4 for IFN-g and Supplementary Table S1 for

other cytokines). The next 4% of the TIL secreted interme-diate levels of IFN-g in the absence of a treatment with agalectin antagonist. We tentatively concluded that galectinantagonists boosted cytokine secretion specifically in galec-tin-3highLELhigh TIL.

GM-CT-01 does not detach galectins from cells butdisorganizes galectin–glycoprotein lattices

Because we previously observed that LacNAc detachesgalectin-3 from cells (8), we tested if GM-CT-01 has thesame effect in TIL, which usually express low levels ofgalectins at cell surface, and in a melanoma cell line forwhich galectin-3 surface expression is abundant (Fig. 5 andSupplementary Fig. S3). Cells were incubated for 2 hourswith GM-CT-01 or LacNAc, labeled with specific anti-galec-tin antibodies, and analyzed by flow cytometry for surfaceexpression of galectins. As previously reported, labeling ofTIL for galectin-3 or galectin-1 decreased after LacNActreatment (8), but we did not observe galectin detachmentafter GM-CT-01 treatment (Fig. 5). This difference was evenmore striking for melanoma cells (Supplementary Fig. S3).

To explain that GM-CT-01 treatment, which failed todetach galectins, nevertheless increased the ability of TIL

Figure 3. Effect of GM-CT-01 on CD8þ TIL cytotoxicity. CD8þ TIL were isolated from ascites from patients with an ovarian carcinoma (LB2991, LB3015,LB3059, LB3023, LB3082) or a pancreatic carcinoma (LB3033). CD8þ T cells were isolated from blood samples obtained from donors withoutcancer (PBL). CD8þ T cells were treated for 20 hours with 1.8 mmol/L GM-CT-01 or 5mmol/L LacNAc, and washed. A, CD8þ T cells were incubatedwith 51Cr-labeled target cells at the indicated effector-to-target (E:T) ratios. Target cells were P815 cells coated with an anti-CD3 mAb (1 mg/mL) or autologoustumor cell lines for patients LB3023, LB3082, and LB3033. Chromium release was measured after 4 hours. B, CD8þ T cells (75,000, treated or not) werestimulated with either 150,000 P815 cells previously incubated for 15 minutes with an anti-CD3 antibody, or with 75,000 CD3/CD28 beads. Brefeldin-Aand FITC-coupled anti-CD107a and anti-CD107b antibodieswere added simultaneously to the stimulators. After 5 hours of stimulation, cells were labeled forCD3 and CD8, and the percentage of CD3þCD8þ T cells with detectable surface CD107 was determined.

Demotte et al.





to secrete IFN-g we examined if GM-CT-01 treatment coulddisorganize galectin–glycoprotein lattices enough toimprove TIL function but not to result into galectin detach-ment. To this aim, we used a microscopy-based FRETapproach that probes close contact between TCR and CD8molecules at the surface of CD8þ TIL, hereafter referred to ascolocalization. Indeed, we had previously observed a poorcolocalization of TCR and CD8 at the surface of CD8þ TIL,as compared with CD8þ blood T cells. Moreover, treatingCD8þ TIL with LacNAc increased both IFN-g secretion andthe colocalization of TCR with coreceptor CD8 (8, 24).Here, CD8þ TIL freshly isolated from ascites were treatedovernight with GM-CT-01 or LacNAc, attached to cover-slips, and double-labeled with an anti-TCR-b antibodycoupled to an acceptor fluorochrome and with an anti-CD8-a antibody coupled to a donor fluorochrome. Upon

excitation at donor wavelength, energy can be transferredfrom the donor to the acceptor if the 2 fluorochromes arecloser than �10 nm. In these conditions, full acceptorbleaching abrogates the energy loss, thus increasing donoremission. Data are shown in Table 1 for 3 CD8þ TILsamples.

In untreated CD8þ TIL, no increase in donor emissionwas detected upon acceptor photobleaching, indicatingpoor TCR:CD8 colocalization (Table 1). Negative values

Figure 4. T cells responding to GM-CT-01 express poly-LacNAc motifsand are covered with galectin-3. CD8þ T cells were isolated fromascites of patient with ovarian carcinoma VUB190, labeled withfluorescein-labeled LEL and with an anti-galectin-3 antibody followed byneutravidin-PE. Four different subpopulations were sorted by flowcytometry. For each, 10,000Tcells (triplicates)were incubated for 2 hoursin culture medium supplemented with either 0.6 mmol/L GM-CT-01 or 5mmol/L LacNAc, then cultured overnight with 20,000 B-EBV cellspreviously incubated with a cocktail of superantigens and washed.Secretion of IFN-g in the supernatant wasmeasured by ELISA. Values aremeans � SD of triplicate microwells.

Figure 5. In contrast to LacNAc, GM-CT-01 does not detach galectin-1and galectin-3 from the TIL surface. CD8þ T cells, isolated fromthe ovarian carcinoma ascites obtained from patient LB3122,were incubated at 37�C for 2 hwith either LacNAc orGM-CT-01,washed,incubated at 4�C for 15 minutes with Fc block, washed again andincubated with 5 mg/mL of either biotinylated rat anti-galectin-3 antibodyor polyclonal anti-galectin-1 rabbit IgG.Cellswerewashedand incubatedat 4�C for 15 minutes with either neutravidin R-PE or anti-rabbit Igsecondary antibodies coupled toAlexa Fluor 488.Cellswere also labeledwith anti-CD3.PerCP and anti-CD8.APC. Cells shown in the figureare CD3þCD8þ cells.






Tab

le1.

FRETco

loca

lizationof

TCRan

dco

rece

ptorCD8

TIL

CD8þ

VUB15

5aTIL

CD8þ

GHC9

TIL

CD8þ

GHC8

Untreated

GM-C

T-01

LacN

Ac

Untreated

GM-C

T-01

LacN

Ac

Untreated

GM-C

T-01

LacN

Ac

Increa

sein

dono

rem

ission(%

)Cell1

bCell2

Cell1

Cell2

Cell1

Cell2

Cell1

Cell2

Cell1

Cell2

Cell1

Cell2

Cell1

Cell2

Cell1

Cell2

Cell1

Cell2

Bleac

hedregion

�18

�111

136

12�8

�14

185

35

�3�6

23

28

�23

139

98

1�6

1311

1212

�1�8

35

34

�6�4

1414

1010

�2�2

135

158

�2�4

93

13

Mea

nc�5

�7

12�

29�

2�5

�5

11�

59�

5�4

�3

4�

34�

2

Sec

retionofIFN-g

by10

4TIL

(pg/m

L)d

308�

3191

3�

133

862�

7319

2�3

34,43

2�

516

2,20

5�

178

274�

41,19

9�

2391

5�

133

aCD8þ

TILwereisolated

from

ascitesof

patientswith

carcinom

aan

dfroz

en.C

ellswerethaw

n,was

hed,a

ndincu

bated

overnigh

tat

37� C

inmed

ium

containing

10%

ascitic

fluidin

the

prese

nceof

1.8mm

ol/L

GM-C

T-01

.After

was

hing

,som

ece

llsweretested

forIFN

-gse

cretionan

dtheothe

rswereattach

edto

cove

rslip

s,labeled

onicewith

anti-TC

Ran

tibod

ies(FRET

acce

ptor),

andan

ti-CD8an

tibod

ies(FRETdon

or),fixe

dan

dstaine

dwith

seco

ndaryan

tibod

ies.

bFo

reac

hof

the2ce

llsthat

werean

alyzed

,in3region

sof

interest,5

imag

eswerereco

rded

beforebleac

hing

and5othe

rsafterb

leac

hing

oftheac

ceptorfluo

roch

rome.Th

reeco

ntrolreg

ions

wereno

tbleac

hed.Inc

reas

ein

don

orem

ission

was

calculated

asfollo

ws:

[1�

(don

orintens

itybe

fore

bleac

hing

/don

orintens

ityafterbleac

hing

)]�

100.

cMea

nincrea

seindon

orem

ission

(inpe

rcen

tage

)was

�4.6

�5.1forn

ontrea

tedTIL,

9.1�4.9(P

<0.00

01by

Man

n–Whitney

test)for

GM-C

T-01

-treated

TIL,

and7.3�4.1(P

<0.00

01)for

LacN

Ac-trea

tedTIL.

d10

4TILweredistributed

inmicrowells,treated

overnigh

twith

1.8mm

ol/L

GM-C

T-01

,and

cultu

redov

ernigh

twith

3,00

0CD3/CD28

bead

s.Th

eprese

nceof

IFN-g

inthesu

perna

tant

was

estim

ated

byELISAin

triplicates

.

Demotte et al.





indicate partial bleaching of the donor fluorochrome dur-ing the procedure. For 2 of the treated TIL, VUB155, andGHC9, there was a consistent strong increase in donoremission after photobleaching, indicating TCR:CD8 colo-calization upon LacNAc and GM-CT-01 treatment. Theresponse to treatments was mild with TIL GHC8. Althoughmodest at first sight, these increases are within the rangereported in the literature (8, 24). We conclude that theincreased TCR:CD8 colocalization observed after GM-CT-01 treatment is in agreement with the hypothesis that thistreatment disorganizes galectin–glycoprotein lattices with-out detaching galectins from the cell surface.

Efficacy of different galactomannans for boosting TILfunctionsGM-CT-01 is obtained by hydrolysis of guar gum, a

galactomannan extracted from guar beans (Cyamopsis tetra-glonoloba). The guar gumbackbone is a linear chain of b 1.4-linked mannose residues to which galactose residues are1.6-linked to mannose, forming short side-branches. Theguar gum used to obtain GM-CT-01 has a mannose-to-galactose ratio of about 1.7. Considering that a large num-ber of plant-derived galactomannans have the same man-nose backbone but different mannose-to-galactose ratios,we wondered if different galactomannans would be able toboost TIL function.The nonhydrolyzed guar gum was inefficient for boosting

TIL function (Supplementary Fig. S4A). We tested partiallyhydrolyzed guar gum generously provided by Galectin Ther-apeutics, the company producingGM-CT-01, and also testeda guar gum that we hydrolyzed using a protocol set up in thelaboratory (Supplementary Fig. S4B). It seemed that hydro-lyzed guar gum with fragments of about 44 to 50 kDa wereable to boost the function of TIL (Supplementary Fig. S4A).We subsequently tested a number of galactomannans

obtained from various plants, either before or after partialhydrolyzation. None of them, except guar gum in its hydro-lyzed version, was able to boost TIL function (Supplemen-tary Fig. S4C). We tentatively concluded that both the sizeand the structure of the galactomannan are important for itsability to boost TIL function.

DiscussionThis study was prompted by the need for a galectin

antagonist approved for clinical use and able to boost TILfunction ex vivo. We here reported that treating TIL obtainedfrom patients with various cancers with GM-CT-01, a galac-tomannan extracted from guar gum, boosted IFN-g secre-tion upon ex vivo stimulation in �80% of the CD8þ TILsamples and �50% of the CD4þ TIL samples. IncreasedIFN-g secretion by CD8þ TIL induced by GM-CT-01 wasconcentration-dependent and correlated with cytotoxicity.The efficacy of GM-CT-01 for boosting IFN-g secretionseems to be equivalent to that of 2 other galectin antago-nists, LacNAc and modified citrus pectin GCS-100 (8). Itremains remarkable, and so far unique to galectin antago-nists, that treating human T cells ex vivo for only a few hourswas sufficient to strongly increase IFN-g secretion. In con-

trast, treating T cells with antibodies specific for inhibitoryreceptors or with IDO inhibitors does not provide animmediate functional correction but instead results in anenhanced proliferation of T cells, yielding a few days later ahigher number of functional T cells (10, 45, 46).

The response to galectin antagonists in only half of theCD4þTIL samples couldbe explainedby the contaminationby regulatory T cells that participate inblocking the functionof effector CD4þ T cells. Testing this hypothesis wouldrequire Treg depletion from CD4þ TIL samples, but nospecific surface marker is available for full removal of Treg.Moreover, eliminating the CD25-positive cells would alsoeliminate activated TIL, which could be abundant in ascites.

How does GM-CT-01 trigger IFN-g secretion? Galectin-1and -3 are abundantly released by tumor cells and macro-phages, so as to reach nanomolar concentrations in ascitesand readily bind to TIL. Our working hypothesis is that ahigh percentage of isolated TIL have been recently activat-ed, and therefore harbor a glycome with many LacNAcmotifs, the natural ligands of galectin-1 and galectin-3(44). The abundance of LacNAc motifs and galectinswould favor the formation of galectin–glycoprotein latticesat the TIL surface and result in a decreased surface motilityof molecular actors of T cell activation. This hypothesis issupported by the lower secretion of IFN-g by galectin-3highLELhigh as compared with galectin-3lowLELlow TIL, andby the selective response of the galectin-3highLELhigh TIL toLacNAc or GM-CT-01. This indicates that upon activation,more galectin-3 binds to TIL expressing glycoproteinsbearing more LacNAc motifs and suggests that the dys-function is related to the presence of galectin-3. LacNAc isable to detach both galectin-1 and galectin-3. Our previousobservation that a galectin-3 antibody, which alsodetached galectin-3 from TIL surface, was able to boostIFN-g secretion by TIL as efficiently as LacNAc, indicatedthat detaching galectin-3 from TIL is sufficient to restorefunction, while not excluding a contribution of othergalectins, in particular galectin-1 (8).

How to reconcile this scenario with the failure of GM-CT-01 to detach galectin-1 and galectin-3 from the cell surface?First, LacNAcandGCS-100 interactwith galectinsbybindingto their carbohydrate recognition domains. They are truecompetitors, whereas GM-CT-01 interacts with a site ofgalectin-1 opposite to the carbohydrate recognition domainand thus acts as an allosteric antagonist (47). The galectin-1site interactingwithGM-CT-01 is conserved in the galectin-3protein (48). In line with these reports, 2 other studies haveshown that the binding of oligomannan to galectin-3 cannotbe completed by lactose (49, 50). Second, acute treatment ofTIL with GM-CT-01 resulted in increased TCR:CD8þ colo-calization based on FRET, such as LacNAc and GCS-100.Wethus tentatively conclude that GM-CT-01 interacts withgalectin-3 without causing its detachment from the TILsurface, but that this interaction results in disruption ofglycoprotein:galectin lattices, therefore restoring motility ofsurface receptors implicated in T-cell activation.

In tumor-bearing mice vaccinated with a tumor antigen,we have previously shown as preliminary data that






injections of modified citrus pectin GCS-100 led to tumorrejection in half of the mice (8). In the same experimentalsetting, GM-CT-01 failed to confer any tumor rejection(data not shown). However, the experiments with GCS-100 were not extended to test whether T cells from tumor-bearing mice were dysfunctional because of glycoprotein–galectin lattices andwhether GCS-100 treatments disruptedthese lattices. Noteworthymurine T cells, as compared withhuman T cells, seem to be poor in tri- and tetra-antennarypoly-LacNac glycans and therefore less susceptible to galec-tin-mediated dysfunction (44, 51, 52). GCS-100 could aswell have increased apoptosis of tumor cells or inhibitedneoangiogenesis and metastases formation, independentlyof an effect on the immune system as it was reported formodified citrus pectin in other experiments (53–55).

It remains remarkable that a short treatment of but a fewhours with GM-CT-01 was sufficient to boost, if not fullyrestore, humanTIL functional capacities.Moreover, respon-siveness of the vast majority of samples suggests that treat-ment with galectin antagonists, in particular GM-CT-01,could be effective in patientswith cancer to correct impairedTIL functions.

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

Authors' ContributionsConception and design: N. Demotte, K. Thielemans, P. van der BruggenDevelopment of methodology: N. Demotte, R. Bigirimana, V. Stroobant,P.J. Courtoy, P. van der Bruggen

Acquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.): N. Demotte, R. Bigirimana, V. Stroobant, J.-L.Squifflet, K. Thielemans, P. Van Der SmissenAnalysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis): N. Demotte, R. Bigirimana, G. Wie€ers,V. Stroobant, J.-F. Baurain, P. Van Der Smissen, P. van der BruggenWriting, review, and/or revision of the manuscript: N. Demotte,R. Bigirimana, G. Wie€ers, V. Stroobant, K. Thielemans, J.-F. Baurain, P.J.Courtoy, P. van der BruggenAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): N. Demotte, J. Carrasco, J.-F.BaurainStudy supervision: N. Demotte, P. van der Bruggen

AcknowledgmentsThe authors thank P.G. Traber and A. Klyosov for providing GM-CT-01

and various galactomannans. The authors also thank D. Godelaine and M.Gordon-Alonso for critical reading, E. Jacobs for technical help, N. Dauguetfor cell sorting, and N. Krack for editorial assistance.

Grant SupportThis work was supported by grant #2010-175 from the Fondation contre

le Cancer (Belgium) and by grants #3.4514.12 and #3.4543.12 from theFonds de la Recherche ScientifiqueM�edicale-FRSM (Belgium). R. Bigirimanais supported by fellowship #1.EO21.13 from the Fonds pour la Formation �ala Recherche dans l’Industrie et dans l’Agriculture-FRIA (Belgium). G.Wie€erswas supported by fellowship #1.1.109.10 from the Fonds de la RechercheScientifique-FNRS (Belgium).

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

Received September 11, 2013; revised January 9, 2014; accepted February2, 2014; published OnlineFirst February 13, 2014.

References1. Wie€ers G, Demotte N, Godelaine D, van der Bruggen P. Immune

suppression in tumors as a surmountable obstacle to clinical efficacyof cancer vaccines. Cancers 2011;3:2904–54.

2. Gajewski TF, Louahed J, Brichard VG. Gene signature in melanomaassociated with clinical activity: a potential clue to unlock cancerimmunotherapy. Cancer J 2010;16:399–403.

3. Ulloa-Montoya F, Louahed J, Dizier B, Gruselle O, Spiessens B,Lehmann FF, et al. Predictive gene signature in MAGE-A3 antigen-specific cancer immunotherapy. J Clin Oncol 2013;31:2388–95.

4. Gajewski T, Zha Y, Thurner B, Schuler G. Association of gene expres-sion profile in metastatic melanoma and survival to a dendritic cell-basedvaccine. JClinOncol (2009ASCOAnnualMeetingProceedings)2009;27 No.15S:9002.

5. Lee PP, Yee C, Savage PA, Fong L, Brockstedt D, Weber JS, et al.Characterization of circulating T cells specific for tumor-associatedantigens in melanoma patients. Nat Med 1999;5:677–85.

6. Guilloux Y, Viret C, Gervois N, Le Drean E, PandolfinoMC, Diez E, et al.Defective lymphokine production by most CD8þ and CD4þ tumor-specific T cell clones derived from human melanoma-infiltrating lym-phocytes in response to autologous tumor cells in vitro. Eur J Immunol1994;24:1966–73.

7. Van den Hove LE, Van Gool SW, Van Poppel H, Baert L, Coorevits L,Van Damme B, et al. Phenotype, cytokine production and cytolyticcapacity of fresh (uncultured) tumour-infiltrating T lymphocytes inhuman renal cell carcinoma. Clin Exp Immunol 1997;109:501–9.

8. Demotte N, Wie€ers G, Van Der Smissen P, Moser M, Schmidt CW,Thielemans K, et al. A galectin-3 ligand corrects the impaired functionof human CD4 and CD8 tumor-infiltrating lymphocytes and favorstumor rejection in mice. Cancer Res 2010;70:7476–88.

9. Zippelius A, Batard P, Rubio-Godoy V, Bioley G, Li�enard D, Lejeune F,et al. Effector function of human tumor-specific CD8 T cells in mela-

noma lesions: a state of local functional tolerance. Cancer Res 2004;64:2865–73.

10. Fourcade J, Sun Z, BenallaouaM,GuillaumeP, Luescher IF, Sander C,et al. Upregulation of Tim-3 and PD-1 expression is associated withtumor antigen-specific CD8þ T cell dysfunction in melanoma patients.J Exp Med 2010;207:2175–86.

11. Le Drean E, Vely F, Olcese L, Cambiaggi A, Guia S, Krystal G, et al.Inhibitionof antigen-inducedTcell response andantibody-inducedNKcell cytotoxicity by NKG2A: association of NKG2A with SHP-1 andSHP-2 protein-tyrosine phosphatases. Eur J Immunol 1998;28:264–76.

12. PhanGQ, Yang JC, Sherry RM,HwuP, Topalian SL, SchwartzentruberDJ, et al. Cancer regression and autoimmunity induced by cytotoxic Tlymphocyte-associated antigen 4 blockade in patients withmetastaticmelanoma. Proc Natl Acad Sci USA 2003;100:8372–7.

13. Sedy JR, Gavrieli M, Potter KG, Hurchla MA, Lindsley RC, Hildner K,et al. B and T lymphocyte attenuator regulates T cell activationthrough interaction with herpesvirus entry mediator. Nat Immunol2005;6:90–8.

14. Wolchok JD, Kluger H, Callahan MK, Postow MA, Rizvi NA, LesokhinAM, et al. Nivolumab plus ipilimumab in advancedmelanoma. N Engl JMed 2013;369:122–33.

15. Betz M, Fox BS. Prostaglandin E2 inhibits production of Th1 lympho-kines but not of Th2 lymphokines. J Immunol 1991;146:108–13.

16. Kehrl JH,RobertsAB,Wakefield LM, JakowlewS,SpornMB,Fauci AS.Transforming growth factor b is an important immunomodulatoryprotein for human B lymphocytes. J Immunol 1986;137:3855–60.

17. Uyttenhove C, Pilotte L, Theate I, Stroobant V, Colau D, Parmentier N,et al. Evidence for a tumoral immune resistance mechanism based ontryptophan degradation by indoleamine 2,3-dioxygenase. Nat Med2003;9:1269–74.


Demotte et al.




18. Stockis J, Fink W, Francois V, Connerotte T, De Smet C, Knoops L,et al. Comparison of stable human Treg and Th clones by transcrip-tional profiling. Eur J Immunol 2009;39:869–82.

19. Francois V, Ottaviani S, Renkvist N, Stockis J, Schuler G, ThielemansK, et al. TheCD4þT-cell responseofmelanomapatients to aMAGE-A3peptide vaccine involves potential regulatory T cells. Cancer Res2009;69:4335–45.

20. Tran DQ, Ramsey H, Shevach EM. Induction of FOXP3 expression innaive human CD4þFOXP3 T cells by T-cell receptor stimulation istransforming growth factor-beta dependent but does not confer aregulatory phenotype. Blood 2007;110:2983–90.

21. Liu FT, Hsu DK, Zuberi RI, Kuwabara I, Chi EY, Henderson WRJr.Expression and function of galectin-3, a b-galactoside-binding lectin,in human monocytes and macrophages. Am J Pathol 1995;147:1016–28.

22. van Stijn CM, van den Broek M, van de Weerd R, Visser M, Tasdelen I,Tefsen B, et al. Regulation of expression and secretion of galectin-3 inhuman monocyte-derived dendritic cells. Mol Immunol 2009;46:3292–9.

23. Rabinovich GA, Toscano MA. Turning 'sweet' on immunity: galectin-glycan interactions in immune tolerance and inflammation. Nat RevImmunol 2009;9:338–52.

24. Demotte N, Stroobant V, Courtoy PJ, Van Der Smissen P, Colau D,Luescher IF, et al. Restoring the association of the T cell receptor withCD8 reverses anergy in human tumor-infiltrating lymphocytes. Immu-nity 2008;28:414–24.

25. Pace KE, Lee C, Stewart PL, Baum LG. Restricted receptor segrega-tion into membrane microdomains occurs on human T cells duringapoptosis induced by galectin-1. J Immunol 1999;163:3801–11.

26. Fukumori T, Takenaka Y, Yoshii T, Kim HR, Hogan V, Inohara H, et al.CD29 and CD7 mediate galectin-3-induced type II T-cell apoptosis.Cancer Res 2003;63:8302–11.

27. RubinsteinN, AlvarezM, Zwirner NW, ToscanoMA, Ilarregui JM,BravoA, et al. Targeted inhibition of galectin-1 gene expression in tumor cellsresults in heightened T cell-mediated rejection: a potential mechanismof tumor-immune privilege. Cancer Cell 2004;5:241–51.

28. Stillman BN, Hsu DK, Pang M, Brewer CF, Johnson P, Liu FT, et al.Galectin-3 and galectin-1 bind distinct cell surface glycoprotein recep-tors to induce T cell death. J Immunol 2006;176:778–89.

29. Perillo NL, Pace KE, Seilhamer JJ, Baum LG. Apoptosis of T cellsmediated by galectin-1. Nature 1995;378:736–9.

30. van der Bruggen P, Traversari C, Chomez P, Lurquin C, De Plaen E,Van den Eynde B, et al. A gene encoding an antigen recognized bycytolytic T lymphocytes on a human melanoma. Science 1991;254:1643–7.

31. Wilgenhof S, Van Nuffel AM, Benteyn D, Corthals J, Aerts C, HeirmanC, et al. A phase IB study on intravenous synthetic mRNA electro-porated dendritic cell immunotherapy in pretreated advanced mela-noma patients. Ann Oncol 2013 July 31. [Epub ahead of print].

32. Boon T, Coulie PG, Van den Eynde B, van der Bruggen P. Human Tcell responses against melanoma. Annu Rev Immunol 2006;24:175–208.

33. Rosenberg SA, Yang JC, Restifo NP. Cancer immunotherapy: movingbeyond current vaccines. Nat Med 2004;10:909–15.

34. Lurquin C, Leth�e B, Corbi�ere V, Th�eate I, van Baren N, Coulie PG, et al.Contrasting frequencies of anti-tumor and anti-vaccine T cells inmetastases of a melanoma patient vaccinated with a MAGE tumorantigen. J Exp Med 2005;201:249–57.

35. Germeau C, MaW, Schiavetti F, Lurquin C, Henry E, Vigneron N, et al.High frequency of anti-tumor T cells in the blood of melanoma patientsbefore and after vaccination with tumor antigens. J Exp Med2005;201:241–8.

36. Carrasco J, Van Pel A, Neyns B, Leth�e B, Brasseur F, Renkvist N, et al.Vaccination of a melanoma patient with mature dendritic cells pulsedwith MAGE-3 peptides triggers the activity of nonvaccine anti-tumorcells. J Immunol 2008;180:3585–93.

37. Miller MC, Klyosov A, Mayo KH. The a-galactomannan Davanat bindsgalectin-1 at a site different from the conventional galectin carbohy-drate binding domain. Glycobiology 2009;19:1034–45.

38. Klyosov A, Zomer E, Platt D. DAVANAT� (GM-CT-01) and coloncancer: preclinical and clinical (phase I and II) studies. In: Klyosov AA,editor. Glycobiology and drug design (ACS symposium series, 1102).Washington, DC: American Chemical Society; 2012. p. 89–130.

39. KlyosovAA, Traber PG.Galectins in disease andpotential therapeuticsapproaches. In: Klyosov AA, Traber PG, editors. Galectins and diseaseimplications for targeted therapeutics (ACS symposium series, 1115).Washington, DC: American Chemical Society; 2012. p. 3–43.

40. Klyosov AA, Zomer E, Platt D. DAVANAT� and colon cancer: preclin-ical and clinical (phase I) studies. In:Klyosov AA, Witczak ZJ, Platt D,editors. Carbohydrate drug design (ACS symposium series, 932).Washington, DC: American Chemical Society; 2006. p. 67–104.

41. Uyttenhove C, Maryanski J, Boon T. Escape of mouse mastocytomaP815 after nearly complete rejection is due to antigen-loss variantsrather than immunosuppression. J Exp Med 1983;157:1040–52.

42. Iurisci I, Tinari N, Natoli C, Angelucci D, Cianchetti E, Iacobelli S.Concentrations of galectin-3 in the sera of normal controls and cancerpatients. Clin Cancer Res 2000;6:1389–93.

43. Eliaz I. TheRole of galectin-3 as amarker of cancer and inflammation ina stage IV ovarian cancer patient with underlying pro-inflammatorycomorbidities. Case Rep Oncol 2013;6:343–9.

44. Antonopoulos A, Demotte N, Stroobant V, Haslam SM, van der Brug-gen P, Dell A. Loss of effector function of human cytolytic T lympho-cytes is accompanied by major alterations in N- and O-glycosylation.J Biol Chem 2012;287:11240–51.

45. FourcadeJ,KudelaP,SunZ,ShenH,LandSR, LenznerD, et al. PD-1 isa regulator of NY-ESO-1-specific CD8þ T cell expansion in melanomapatients. J Immunol 2009;182:5240–9.

46. Wong RM, Scotland RR, Lau RL, Wang C, Korman AJ, Kast WM, et al.Programmed death-1 blockade enhances expansion and functionalcapacity of human melanoma antigen-specific CTLs. Int Immunol2007;19:1223–34.

47. Miller MC, Nesmelova IV, Platt D, Klyosov A, Mayo KH. The carbohy-drate-binding domain on galectin-1 is more extensive for a complexglycan than for simple saccharides: implications for galectin-glycaninteractions at the cell surface. Biochem J 2009;421:211–21.

48. ,Nesmelova IV, Dings RPM, Mayo KH. Understanding galectin struc-ture-function relationship to design effective antagonists. In:KlyosovAA, Witczak ZJ, Platt D, editors. Galectins. Hoboken, New Jersey:John Wiley & Sons; 2008. p. 33–69.

49. Fradin C, Poulain D, Jouault T. b-1,2-Linked oligomannosides fromCandida albicans bind to a 32-kDa macrophage membrane proteinhomologous to the mammalian lectin galectin-3. Infect Immun2000;68:4391–8.

50. Kohatsu L, Hsu DK, Jegalian AG, Liu FT, Baum LG. Galectin-3 inducesdeath of Candida species expressing specific b-1,2-linked mannans.J Immunol 2006;177:4718–26.

51. Comelli EM, Sutton-SmithM, Yan Q, AmadoM, PanicoM, Gilmartin T,et al. Activation of murine CD4þ and CD8þ T lymphocytes leads todramatic remodeling of N-linked glycans. J Immunol 2006;177:2431–40.

52. Antonopoulos A, North SJ, HaslamSM,Dell A. Glycosylation ofmouseand human immune cells: insights emerging from N-glycomics anal-yses. Biochem Soc Trans 2011;39:1334–40.

53. StreetlyMJ,Maharaj L, Joel S, Schey SA, Gribben JG, Cotter FE.GCS-100, a novel galectin-3 antagonist, modulates MCL-1, NOXA, and cellcycle to induce myeloma cell death. Blood 2010;115:3939–48.

54. Chauhan D, Li G, Podar K, Hideshima T, Neri P, He D, et al. A novelcarbohydrate-based therapeutic GCS-100 overcomes bortezomibresistance and enhances dexamethasone-induced apoptosis in mul-tiple myeloma cells. Cancer Res 2005;65:8350–8.

55. Glinsky VV, Raz A. Modified citrus pectin anti-metastatic properties:one bullet, multiple targets. Carbohydr Res 2009;344:1788–91.






Published OnlineFirst February 13, 2014.Clin Cancer Res Nathalie Demotte, René Bigirimana, Grégoire Wieërs, et al. Lymphocytes

Infiltrating−the Functions of Freshly Isolated Human Tumor A Short Treatment with Galactomannan GM-CT-01 Corrects

Updated version

10.1158/1078-0432.CCR-13-2459doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2014/02/14/1078-0432.CCR-13-2459.DC1Access the most recent supplemental material at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://clincancerres.aacrjournals.org/content/early/2014/03/19/1078-0432.CCR-13-2459To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-13-2459

http://clincancerres.aacrjournals.org/content/suppl/2014/02/14/1078-0432.CCR-13-2459.DC1

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/early/2014/03/19/1078-0432.CCR-13-2459


a short treatment with galactomannan gm-ct-01 corrects ......2014/03/19 ·...

Documents