Gastric Adenocarcinomas Express the …mct.aacrjournals.org/content/molcanther/15/5/1008.full.pdfLarge Molecule Therapeutics Gastric Adenocarcinomas Express the GlycosphingolipidGb

Large Molecule Therapeutics

Gastric Adenocarcinomas Express theGlycosphingolipidGb3/CD77:TargetingofGastricCancer Cells with Shiga Toxin B-SubunitPhilipp Emanuel Geyer1, MatthiasMaak1, Ulrich Nitsche1, Markus Perl1, Alexander Novotny1,Julia Slotta-Huspenina2, Estelle Dransart3,4,5, Anne Holtorf1, Ludger Johannes3,4,5, andKlaus-Peter Janssen1

Abstract

The B-subunit of the bacterial Shiga toxin (STxB), which isnontoxic and has low immunogenicity, can be used for tumortargeting of breast, colon, and pancreatic cancer. Here, we testedwhether human gastric cancers, which are among the mostaggressive tumor entities, express the cellular receptor of Shigatoxin, the glycosphingolipid globotriaosylceramide (Gb3/CD77).Themajority of cases showed an extensive staining for Gb3 (36/50cases, 72%), as evidenced on tissue sections of surgically resectedspecimen. Gb3 expression was detected independent of type(diffuse/intestinal), and was negatively correlated to increasingtumor–node–metastasis stages (P ¼ 0.0385), as well as withmarkers for senescence. Gb3 expression in nondiseased gastricmucosa was restricted to chief and parietal cells at the bottom ofthe gastric glands, and was not elevated in endoscopic samples of

gastritis (n¼ 10). Gb3 expression in established cell lines of gastriccarcinoma was heterogeneous, with 6 of 10 lines being positive,evidenced by flow cytometry. STxB was taken up rapidly by liveGb3-positive gastric cancer cells, following the intracellular ret-rograde transport route, avoiding lysosomes and rapidly reachingtheGolgi apparatus and the endoplasmic reticulum. Treatment ofthe Gb3-expressing gastric carcinoma cell line St3051 with STxBcoupled to SN38, the active metabolite of the topoisomerase typeI inhibitor irinotecan, resulted in>100-fold increased cytotoxicity,as compared with irinotecan alone. No cytotoxicity was observedon gastric cancer cell lines lacking Gb3 expression, demonstratingreceptor specificity of the STxB–SN38 compound. Thus, STxB is ahighly specific transport vehicle for cytotoxic agents in gastriccarcinoma. Mol Cancer Ther; 15(5); 1008–17. �2016 AACR.

IntroductionWith only 30% 5-year survival in advanced stages, gastric

cancers are among the most aggressive tumors (1). The mostfrequent type, arising from the glandular epithelium of thestomach mucosa, is gastric adenocarcinoma, which displayshighly heterogeneous histologic differentiation patterns (2). Thetwomajor types of gastric carcinoma are the intestinal and diffusetype (Lauren classification). The intestinal type frequently showsdisorganized tubular structures, whereas diffuse-type cancer cells

are scattered, mucus-secreting, and typically poorly differentiat-ed (3). Currently, surgical resection is the only option forcurative treatment and represents the standard therapy forlocoregional gastric carcinoma. However, surgery alone isrecommended only for early stages (1). Most tumors are locallyadvanced at diagnosis, requiring perioperative systemic thera-pies, which are associated with toxic side effects and relativelylow response rates (1). Therefore, targeted therapy of gastriccancer, by specific delivery of cytotoxic compounds to gastriccancer cells, is urgently required.

We and others showed previously that gastrointestinal tumors,like colorectal cancer and pancreatic ductal adenocarcinoma,show an overexpression of the neutral glycosphingolipid globo-triaosylceramide in tumor cells, but also in normal as well astumor-associated vasculature (Gala1-4Galb1-4Glcb1-1Cer, Gb3/CD77; refs. 4–8). The terminal disaccharide of this lipid is thereceptor of the bacterial Shiga toxin. Its nontoxic B-subunit (STxB)is a promising tool for targeted therapy, as STxB specifically bindsto tumor cells that express its cellular receptor Gb3 on the outerleaflet of the plasma membrane (9). Upon binding, STxB israpidly taken up, following the "retrograde transport" route fromthe endosome via the trans-Golgi network and the Golgi appa-ratus to the endoplasmic reticulum (10). Thereby, STxB bypassesthe late endocytic pathway and avoids lysosomal degradation.Shiga toxin is produced by Shigella dysenteriae (serotype I), andShiga-like toxins (or verotoxins) by enterohemorrhagic Escher-ichia coli strains (11, 12). Shiga toxin is comprised of one toxic A-subunit, and five nontoxic B-fragments that form the B-subunit(STxB), and serve as a highly specific delivery tool.One STxBbinds

1Department of Surgery, Klinikum rechts der Isar, Technische Uni-versit€at M€unchen, Munich,Germany. 2Institute of Pathology and Path-ological Anatomy,Technische Universit€atM€unchen, Munich,Germany.3Endocytic Trafficking and Intracellular Delivery team, Institut Curie,Paris, France. 4CNRS UMR3666, Paris, France. 5INSERM U1143, Paris,France.

Note: Supplementary data for this article are available at Molecular CancerTherapeutics Online (http://mct.aacrjournals.org/).

Current address for P.E. Geyer: Max-Planck Institute of Biochemistry, Martins-ried, Germany; and current address for M. Maak: Department of Surgery,Universit€atsklinikum Erlangen, Friedrich-Alexander-Universit€at Erlangen-N€urnberg, Erlangen, Germany.

Corresponding Author: Klaus-Peter Janssen, Department of Surgery,Klinikum rechts der Isar, TU M€unchen, Ismaninger Str. 22, Munich 81675,Germany. Phone: 49-89-4140-2066; Fax: 49-89-4140-6031; E-mail:[email protected]

doi: 10.1158/1535-7163.MCT-15-0633

�2016 American Association for Cancer Research.

MolecularCancerTherapeutics

Mol Cancer Ther; 15(5) May 20161008

on May 31, 2018. © 2016 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst January 29, 2016; DOI: 10.1158/1535-7163.MCT-15-0633

http://mct.aacrjournals.org/

to up to 15 Gb3 molecules (12, 13), which leads to lipid rearran-gements in membrane microdomains (14), tubular membraneinvagination, and eventually to intracellular uptake (15).

Evolutionarily acquired properties of STxB, like high receptorspecificity and affinity, fast intracellular uptake, high stabilityagainst proteolytic degradation (16) and low immunogenicityin humans (17,18), predestine STxB for biomedical applications.The decisive requirement for application of STxB is the cell-specific expression of Gb3 in tumor cells: Gb3 is expressed byBurkitt and centrofollicular lymphoma and solid tumors liketesticular seminoma, breast, ovary, pancreatic, and colon cancer(5,6, 8, 19–26).

The potential for diagnostic use of STxB conjugates was suc-cessfully demonstrated by endoscopic, as well as by PET andultrasound imaging in xenografts of human tumors (27), and ingenetic mouse models of digestive cancer (28). The coupling ofSTxB to SN38 (7-ethyl-10-hydroxycamptothecin), the activemetabolite of the topoisomerase I inhibitor irinotecan (CPT-11), was successfully shown (29). The STxB–SN38 conjugate hashighly increased cytotoxicity as compared with irinotecan, whentested on Gb3-expressing colorectal or pancreatic cancer cell lines(29). SN38 is orders of magnitude more active than irinotecan,but its hydrophobicity and insolubility in most physiologicsolvents preclude it from clinical application (30). The parentalcompound, irinotecan, has been reported for second-line chemo-therapy of gastric carcinoma (31, 32).

Here, we report Gb3 expression in patient samples and estab-lished cell lines of gastric carcinoma. Moreover, intracellularuptake kinetics of fluorescently labeled STxB were established forgastric carcinoma cell lines. A STxB–SN38 conjugate showed highreceptor specificity and over 100-fold greater cytotoxicity ongastric cancer cells, as compared with the clinically established

drug irinotecan. Taken together, our study describes for the firsttime a specific therapeutic targeting of gastric carcinoma cells withthe conjugate STxB–SN38.

Materials and MethodsPatient collective

Informed written consent had been obtained with priorapproval of the ethics committee of the Faculty of Medicine ofthe TUM (#1926/2007). Tissue samples were obtained surgicallyfrom 50 patients with gastric carcinoma admitted to our SurgicalDepartment in 1990 to 2005 (clinical data in SupplementaryTable S1). Endoscopic samples of ten tumor-free patients withtype B/C gastritis were collected (Supplementary Table S2), andtested for presence ofH. pylori by rapid urease test, as well as PCR.Tumors were classified according to the International UnionAgainst Cancer (AJCC/UICC), summarized in Table 1. Histolo-gy-guided sample selection was performed by a trained pathol-ogist to identify tumor samples, as well as controls for nondi-seased gastric mucosa, essentially as indicated in (33). Sampleswere snap frozen in liquid nitrogen immediately and stored at–80�C.

Reagents and antibodiesThe recombinant variant STxB/Cys was produced in endo-

toxin-free form as described previously (28). Antibodies andreagents were purchased from the indicated suppliers: anti-golgin p97 (Molecular Probes), anti-von Willebrand factor(Millipore), anti-pepsinogen A (Acris), anti-chromogranin A(Epitomics), DAPI [2-(4-Carbamimidoylphenyl)-1H-indol-6-carboximidamide], anti-Ki67, and anti-calnexin antibodies (Sig-ma-Aldrich), and BSA-Cy3 (Linscott). Secondary antibodies

Table 1. Correlation of Gb3 expression with clinicopathological data in gastric cancer

Gb3 expression scorea

Clinical/pathologic parameter No. of cases (%) 0 1 2 P

Laur�en classificationIntestinal type 25 (50%) 6 15 4Diffuse type 25 (50%) 8 14 3 nsb

Histologic typePapillary carcinoma 3 (6%) 1 2 0Tubular carcinoma 18 (36%) 4 10 4Mucinous carcinoma 2 (4%) 1 1 0Signet-ring carcinoma 18 (36%) 5 11 2Undifferentiated carcinoma 8 (18%) 3 4 1Unclassified carcinoma 1 (2%) 0 1 0 ns

Tumor gradingModerately diff. (G2) 11 (22%) 2 7 2Poorly diff. (G3) 35 (70%) 9 21 5Undifferentiated (G4) 4 (8%) 3 1 0 ns

TNM stage (AJCC/UICC)I/II 26 (52%) 6 13 7III/IV 24 (48%) 8 16 0 0.0385

Tumor stagepT1/2 18 (36%) 4 8 6pT3/4 32 (64%) 10 21 1 0.0267

Nodal statuspN0 21 (42%) 6 12 3pN1/2/3 29 (58%) 8 17 4 ns

Metastatic statuspM0 38 (76%) 9 22 7pM1 12 (24%) 5 7 0 ns

Mean postoperative survival (months) 35.8 36.0 35.6 nsaGb3 score: 0, no STxB staining; 1, weak to intermediate STxB staining; 2, strong STxB staining in major areas of the section.bns, not significant, pairwise comparison.

STxB-Based Targeting of Gastric Cancer

www.aacrjournals.org Mol Cancer Ther; 15(5) May 2016 1009




were purchased from Jackson Immunoresearch (West Grove),cell culture reagents were from Invitrogen. Irinotecan wasobtained from the Pharmacy of the Klinikum rechts der Isar.SN38 was coupled to STxB as described previously (29).

Immunofluorescence microscopyImmunostaining was performed as described previously

(28). Briefly, after paraformaldehyde-fixation, cells on coverslips were permeabilized with 0.1% Triton X-100, blocked withPBS containing 2% BSA, and primary antibodies in blockingbuffer were added before counterstaining with secondary anti-bodies. At this concentration, Triton X-100 did not alter the Gb3detection on cryosections. STxB/Cys was purified from bacteriaas described previously (34). Covalent coupling of STxB/Cys tothe fluorophore Cy3 (Amersham Biosciences) was carried outaccording to supplier's instructions. Gb3 was stained on 3%paraformaldehyde fixed cryosections by incubation with STxB-Cy3 for 30 minutes at a final concentration of 10 mg/mL in PBScontaining 0.2% BSA. For counterstaining against Ki67, 0.1%Triton X-100 was used for permeabilization. Immunofluores-cence staining was detected with a Zeiss Axiovert 200M micro-scope with an AxioCamMRmCCD camera (Zeiss). Images werecaptured using Axiovision software (version 4.8.2, Zeiss) andimage files were imported into Adobe Photoshop version12.0.4 software (Adobe). STxB staining on tissue sections wasquantified by three independent observers, blind to sampleidentity (P. Emanuel Geyer, M. Maak, K.-P. Janssen). A scoringsystem was established, ranging from no detectable STxB stain-ing (0 points), weak to intermediate staining (1 point), andstrong staining (2 points; Supplementary Fig. S1).

Cell culture and STxB uptake assaysCell lines were acquired from ATCC or DSMZ since 2013,

respectively, and periodically tested for mycoplasma infection.To avoid contamination and phenotype changes, cells were keptas frozen stocks and cultured consecutively for 4 weeksmaximumin DMEM with FCS (7%), 1% penicillin/streptomycin, and 1%glutamine. Gastric cancer cell lines of diffuse/poorly differentiat-ed type: AGS and KATO III, GP220MKN-45 NUGC4, cell lines ofintestinal type:MKN-7, NCI-N87, St 2957, St 3051, and St 23132,and HCT 116 colorectal cancer cells as described previously (8).Cell lines have been tested and authenticated by multiplex shorttandem repeat analysis (Promega), the last test has been per-formed in October 2015 at the Institute of Legal Medicine, LMUMunich, Germany. For detailed information on cell lines, pleaserefer to Supplementary Table S3. Cellular morphology andgrowth characteristics were analyzed regularly, and no phenotypechanges were observed through the study. STxB-Cy3, or BSA-Cy3as control, was added at final concentration of 2.5 mg/mL aspublished earlier (6).

Cytotoxicity assaysCell growth analysis was performed essentially as described

(Maak and colleagues, 2011), using the Cell Proliferation Kit II(XTT) from Roche Diagnostics. Briefly, cell lines were split into6-well plates (1 � 105 cells/well) and grown for 3 days.Treatment with irinotecan or STxB–SN38 was carried out for6 hours in cell medium. Cells were washed with DMEM, andsubstituted with fresh medium for an additional 48 hours. Cellswere then seeded onto 96-well plates (triplicates, 1 � 105 cellsper well) in medium containing 0.5% FCS. After 12 to 18 hours,

the XTT solution was added in medium with 3% FCS, and thecells were incubated for 30 hours. The amount of formazan wasquantified with a Mithras LB 940 microplate reader (BertholdTechnologies) at 450 to 500 nm. Mean values of at least threeindependent analyses are shown.

Senescence staining for cultured cells and cryosectionsSenescent cells were stained by enzymatic activity of senes-

cence-associated b-galactosidase with the PromoKine SenescenceDetection Kit (PromoCell). Tissue cryosections, or cells grown oncoverslips for various time points, were washed briefly with PBS,and fixed following themanufacturer's instruction for 15minutesat room temperature. The staining solution was added (940 mLstaining solution, 10 mL staining supplement, 50 mL 20mg/mL X-Gal in DMSO), and incubated for 48 hours at 37�C.

Flow cytometryA total of 2 � 106 cells were seeded in 10-cm cell culture

dishes, cultivated for 24 hours, harvested, and counted. Stain-ing was carried out for 5 � 105 cells in 1 mL medium, witheither 20 nmol/L STxB-Cy3, staining control with 20 nmol/LBSA-Cy3, or without staining, for 15 minutes at 37�C. Cellswere harvested by centrifugation, resuspended in 0.5 mL PBSwith 0.5% (v/v) BSA, 0.01% (w/v) NaN3, and passed through acell strainer immediately before flow cytometry using a FACS-Calibur device (BD Becton Dickinson) with CellQuest Prosoftware (BD Biosciences). Doublets, dead cells, and debriswere excluded, and histograms were analyzed with the softwareFlowJo 8.8.2 (Tree Star, Ashland). Histograms for BSA-Cy3staining were used to control for the amount of unspecificstaining for each assay and cell line.

Statistical analysisAnalyses were performed using SPSS (version 16.0), Graph Pad

Prism 5, and GraphPad InStat3 (Graph Pad Software). Data arepresented as mean � SD or, where specified, as median (range).Comparison of Gb3 expression in carcinoma and correspondingnondiseased tissue was done using paired t test. Influence of Gb3expression on survival time was analyzed by Cox regression.Mann–Whitney U tests were used for correlation of Gb3 expres-sion with histopathologic data. All statistical comparisons weredone at a 0.05 level of significance.

ResultsGb3 expression in nondiseased gastric tissue

To establish the expression pattern of Gb3 in normal stomachmucosa, we analyzed surgically resected gastric tissue (10 cases,two samples each, from antrum and corpus regions), fromtumor-free patients. Samples were stained with fluorescentlylabeled STxB, according to methods established previously (8).In all cases, STxB-labeling was restricted to the abluminal partsof gastric glands, and no major differences were observedbetween different parts of the stomach. On the basis of ana-tomic localization and position within the gastric mucosa,as well as by IHC, the Gb3-expressing cell types correspondedto both parietal and chief cells (Fig. 1). This was confirmedby comparing sections stained with hematoxylin and eosin(Fig. 1A), to parallel cryosections stained with fluorescentlylabeled STxB (Fig. 1B–E). Blood vessel endothelial cells, iden-tified by staining against von Willebrand factor, did not show

Geyer et al.

Mol Cancer Ther; 15(5) May 2016 Molecular Cancer Therapeutics1010




Gb3 expression (Fig 1B). Gb3 expression in chief cells wasshown by counterstaining of tissue sections with labeled STxBand specific antibodies against pepsinogen-A (Fig. 1C; double-positive cells in yellow). However, not all pepsinogen-A positivecells were positive for STxB-Cy3. A part of presumed chief cells,located at the very bottom of gastric glands, did not showdetectable Gb3 expression (Fig. 1C, inset). STxB-staining wasrestricted to a narrow zone located distally above these cells,consisting of double-positive cells stained both for pepsinogenA and STxB (i.e., chief cells), as well as Gb3-positive and pepsin-ogen A–negative cells (i.e., parietal cells; Fig. 1C). Cells in theisthmus, identified by the proliferation marker Ki67, were nega-tive for STxB (Fig. 1D). Only a fraction of gastric neuroendocrinecells, identified by an antibody against chromogranin A, werepositive for Gb3 (Fig. 1E). The foveola part of themucosa, stainedby lectinUEA,was also negative for STxB (Fig. 1F). Taken together,expression of Gb3 was detected in nondiseased stomach mucosa,anatomically restricted to a narrow band of cells at the bottom ofglands, most likely corresponding to a subset of chief and parietalcells.More superficial areas of the gastricmucosa, the proliferatingcompartment and stroma cells were negative for Gb3 expression.Cholesterol depletion following published protocols alloweddetection of Gb3 expression in blood vessel endothelia (Supple-mentary Fig. S1; ref. 35). After investigation andGb3 expression innondiseased gastric mucosa, endoscopically derived samples of10 patients with clinically confirmed type B/C gastritis wereanalyzed (Supplementary Fig. S1; Supplementary Table S2). Twopatients were diagnosed positive for H. pylori infection, but noneof the patients had neoplastic malignancies. In all cases, the STxBstaining pattern was confined to the anatomically restricted zoneobserved in nondiseased gastric mucosa, and no elevated Gb3expression was observed.

Gb3 is strongly expressed in the majority of gastric carcinomaNext, we analyzed surgically resected tissue samples of gastric

carcinoma (n¼ 50). The collective consisted of n¼ 25 cases of theintestinal type, and n ¼ 25 cases of the diffuse type, according tothe classification by Laur�en (ref. 3; details in Supplementary TableS1). Sampleswere studied in ablinded fashion, following a robustscoring system (Supplementary Fig. S1). The analysis revealedstrong Gb3 expression in bona fide cancer cells of intestinal(Fig. 2A), as well as diffuse-type carcinoma (Fig. 2B). 36 of 50(72%) human gastric carcinomas were classified as Gb3-expres-sing, and 7 cases showed very high Gb3 levels. In intestinal-typecarcinoma, cancer cells arranged in glandular structures showedSTxB-staining (Fig. 2A). In diffuse-type carcinoma, dedifferen-tiated and scattered cancer cells were positive for Gb3 (Fig. 2B).Gb3 was often expressed only in parts of the tumors. In rare cases,Gb3 staining was completely absent from tumor cells, and onlydetectable at low levels in blood vessel–associated endothelia(Fig. 2C). Cholesterol depletion revealed Gb3 expression in mosttumor-associated blood vessels, albeit at varying levels (Supple-mentary Fig. S1). For control, the scoring system was appliedto normal tissue from n ¼ 20 patients. Around 45% of normaltissues were devoid of STxB staining (9/20), as compared with28% STxB-negative tested carcinomas (14/50; c2 test for trend, df¼ 4.536;P¼ 0.0332). Both proliferating (Ki67-positive) aswell asnonproliferating cancer cells were positive for Gb3, in accordanceto findings for other tumor entities (Fig. 2D). Controls with BSA-Cy3 demonstrated the specificity of the STxB-staining (Fig. 2E).Next, Gb3 expressionwas analyzed according to clinicopathologicparameters (Fig. 2F and G; Table 1). The frequency of Gb3-expres-sing tumors did not differ significantly between intestinal ordiffuse-type tumors. Postoperative survival of patients withGb3-negative or Gb3-positive tumors was indistinguishable

Figure 1.Analysis of Gb3 expression in nondiseasedgastric mucosa (representative images). A,hematoxylin and eosin stained tissue section.Inset: basophilic chief cells (arrow), andacidophilic parietal cells (arrowhead).Consecutive sections were stained with STxB (inred: B, C, D, and E), nuclei stained with DAPI (inblue: B–F). B, counterstaining for blood vesselswith anti-vWF antibody (green). Gb3 wasexpressed in an anatomically restricted regioncomprising chief and parietal cells (inset), basedon localization and morphology. No expressionwas detected in blood vessels (vWF-stained). C,staining for the chief-cell marker pepsinogen-A(green) identifies a subset of chief cells asdouble-positive for STxB (overlay: yellow). D,staining of proliferating cells in the isthmus withanti-Ki67 antibody (green) shows overlap withnuclear staining (green-blue color, marked byarrows in inset), but not with STxB. E, stainingwith chromogranin-A for neuroendocrine cells(green), and STxB (red) reveals partial Gb3expression in this cell type (Gb3-positiveneuroendocrine cell: arrow; Gb3-negative cell:arrowhead). F, fluorescently labeled lectin UEA(green) stains the surface epithelium of thefoveola, clearly distinct from the STxB-positivearea on the parallel sections (B, C, D, E).F, enlargement, 100-fold; size bars, 100 mm.






(35.8 vs. 35.6 months, n.s.), and metastatic status (nodal ordistal) was not significantly correlated to Gb3 expression. How-ever, a significant reduction of Gb3 expression was observedwith increasing tumor progression. Gb3 levels were negativelycorrelated to increasing tumor–node–metastasis (TNM) stage(P ¼ 0.0385, stage I/II compared with metastasized stage III/IVtumors). Moreover, Gb3 levels were negatively correlated totumor size (T-category, P ¼ 0.0267; Table 1).

Gb3 expression in gastric carcinoma cell linesExpression of Gb3 was further confirmed in gastric cancer

cell lines (details in Supplementary Table S3). The number ofGb3-expressing cells was quantified by flow cytometry, and inde-pendently by fluorescencemicroscopy in cell lines from intestinal(n¼ 5) and diffuse-type (n¼ 5; Fig. 3). In the intestinal-type linesSt 23132, St 3051, St 2957 and the diffuse-type lineNUGC4,morethan half of all cells expressed Gb3 (Fig. 3A). Intermediate to lowGb3 expression was observed for NCI-N87, MKN-7, GP220, andMKN-45, whereas lines KATO III and AGS were completelynegative for Gb3. As control, uptake in colorectal cancer cells(CaCo2) was monitored (Supplementary Fig. S2). There was atrend toward higher Gb3 expression in gastric cancer cell linesfrom intestinal type, as opposed to diffuse-type cancer (P ¼0.06593). Of note, several cell lines withweakGb3 levels (GP220,MKN-7, and MKN-45) showed Gb3 expression, confined toisolated postmitotic cells (Supplementary Fig. S2).

Cancer cell senescence may be associated with downregulatedGb3 expression

Gb3 levels strongly decreased in St3051 cells during prolongedcell culture periods (96 hours), analyzed as fraction of STxB-positive cells by flow cytometry (Supplementary Fig. S3). Thedecrease in Gb3 levels was accompanied by a strong increase inactivity of senescence-associated b-galactosidase, which has notbeendescribed to enzymatically degradeGb3 (Supplementary Fig.S3). Essentially, the same observations were made for HCT116colon cancer cells. Addition of fresh serum or glucose to theculture medium at 96 hours did not lead to increased Gb3biosynthesis. However, complete replacement of culturemediuminduced a rapid increase in Gb3 levels. Moreover, Gb3 expressionand senescence markers were mutually exclusive in parallel tissuesections of gastric carcinoma, indicating that senescence, or loss ofproliferative capacity, may inhibit Gb3 synthesis in cancer cells(Supplementary Fig. S3).

STxB is taken up along the retrograde transport routeIntracellular transport of STxB was investigated by incubation

of Gb3-expressing gastric carcinoma cells with fluorescentlylabeled STxB. Incubation with STxB-Cy3 for 15 minutes at 4�Cresulted in cell surface binding, no intracellular uptake wasobserved (Fig. 4A). Upon incubation for 15 minutes at 37�C,STxB-Cy3 was largely taken up (Fig. 4B), colocalizing in part withthe Golgi marker golgin p97. After incubation for 1 hour at 37�C,

Figure 2.Gb3 expression in gastric cancer of intestinal (A, D, E), or diffuse type (B, C). Left row: hematoxylin and eosin staining, middle and right row: STxB (red), anti-vWFantibody (green), and nuclei (DAPI, blue) on consecutive sections, unless stated otherwise. A, intestinal-type lesion with carcinoma cells arranged inglandular structures. Middle (overview): Gb3 expression in cancer cells (arrow), no expression in tumor-associated blood vessels (arrowhead). Right: higherenlargement. B, gastric-type cancer showing partial Gb3 expression. C, example for lesion of gastric type essentially negative for Gb3. Right: minor STxB-staining inblood vessel endothelia. D, intestinal-type carcinoma. Proliferation marker anti-Ki67 (green) reveals no obvious correlation with STxB-staining (red). E, control forspecificity: intestinal-type carcinoma, stained either for STxB and anti-vWF (middle panel), or for fluorescently labeled BSA and anti-vWF (right panel). Leftand middle row, and control staining in E: 100-fold; size bars, 100 mm. Right row: 400-fold; size bars, 25 mm. F, quantitative examination of Gb3 expression. Nodifference was observed between the fraction of Gb3-expressing tumors of intestinal (n ¼ 25) or diffuse type (n ¼ 25). G, Gb3 expression is downregulated withincreasing TNM stage. Metastatic tumors (UICC-stage III and IV) showed significantly decreased Gb3 levels, compared with locally restricted lesions(IþII; P ¼ 0.0385).

Geyer et al.





complete internalization and perinuclear localization in andaround the Golgi was visible (Fig. 4C). Upon a 48-hour chaseperiod, after an uptake for 1 hour, STxB was still detectable inintracellular compartments marked by anti-golgin p97 (Fig. 4D).Thus, STxB was taken up rapidly, following the retrograde trans-port route. Importantly, binding and uptake of STxB to the cellsurface of gastric cancer cells was maintained even at low pH(Supplementary Fig. S4). In contrast, no uptake of STxB wasdetected in Gb3-negative AGS cells (Supplementary Fig. S5). Toexclude unspecificfluid-phase uptake effects, theGb3-positive cellline St3051 was incubated with fluorescently labeled BSA. Evenlargemolar excesses (1 mmol/L BSA-Cy3, in contrast to 20 nmol/Lof STxB-Cy3), did not lead to detectable staining for BSA-Cy3(Supplementary Fig. S5).

STxB–SN38 can be used for targeted therapy ongastric cancer cells

Uptake in gastric cancer cells was exploited for targeted therapywith a STxB conjugate with SN38, the active metabolite of irino-tecan. On the basis of our previous experiments, cell lines withhigh (St3051) or absent (AGS) Gb3 expression were chosen andincubated with different concentrations of STxB–SN38, and cyto-toxicity and specificity were compared with irinotecan with a cellproliferation assay (Fig. 5). Pronounced effects of STxB–SN38were observed for the Gb3-positive St3051 cells, with a halfmaximal inhibitory concentration (IC50) for STxB–SN38 of0.44 mmol/L, more than 140-fold more potent than irinotecan(IC50: 64 mmol/L; summarized in Supplementary Table S4). Thecytotoxic effects of STxB–SN38 clearly dependent on the presenceof the STxB-receptor Gb3: the cell line AGShad a similar sensitivitytoward the parental drug irinotecan, as compared with St3051cells, but AGS cells were far less sensitive toward STxB–SN38

(IC50: 21 mmol/L). In addition, nonvectorized SN38, which isprecluded from clinical use due to its toxicity, was tested. Asexpected, SN38 had far greater toxicity than irinotecan (by 2orders of magnitude) in all cell lines, but was in the same rangeof activity as STxB–SN38 (Supplementary Fig. S6).

DiscussionHere, we demonstrate the use of the nontoxic B-subunit of

bacterial Shiga toxin (STxB) for specific targeting of human gastriccancer, the fourthmost common cancerworldwide.Gastric cancertreatment relies on surgical resection and multimodal therapywith cytotoxic agents (1). However, the therapy response ingeneral is limited, despite pronounced toxicity of chemotherapy.Therefore, targeted approaches with higher efficiency and reducedside effects offer new intervention strategies for gastric cancer. Thenontoxic B-subunit of the bacterial Shiga toxin (STxB) may beused as an efficient and specific tool for tumor targeting. Thecellular STxB receptor, the neutral glycosphingolipid globotriao-sylceramide Gb3 (Gala1-4Galb1-4Glcb1-1Cer), or CD77, is over-expressed in colon carcinoma (6, 21), pancreatic (5, 8), and breastcancer (25).Moreover, STxB can be functionalized and coupled tospecific agents for noninvasive or endoscopic tumor imaging,which has successfully been demonstrated in preclinical models(28,36). In the current study, we analyzed whether STxB could beused for targeted therapy in gastric cancer.

However, the expression pattern and regulation of glycosphin-golipids in gastrointestinal pathologies are still far from beingunderstood (37). On the basis of the analysis of nondiseasedgastric mucosa, we found expression of Gb3 in an anatomicallywell-defined narrow region at the distal bottom of glands, corre-sponding to a part of the chief, parietal, and enteroendocrine cell

Figure 3.Gb3 expression in gastric carcinoma cell lines, quantified by flow cytometry and fluorescence microscopy. A, four cell lines expressed Gb3 strongly in >33% of cells,four lines at intermediate to low levels (between 15% and 1% of cells), two were negative. B, flow cytometry–based quantification of Gb3 expression inSt3051 cells (histogram: STxB-Cy3, red line; BSA-Cy3: blue; unstained: green). C, fluorescence microscopy, incubation of St3051 cells with STxB (red), nuclei (DAPI,blue). D, cell line AGS is Gb3-negative (histogram: STxB-Cy3, red line; BSA-Cy3, blue; unstained, green). E, fluorescence microscopy, incubation of AGS cellswith STxB (red), nuclei (DAPI, blue). C and E, enlargement, 200-fold; size bar, 50 mm.






populations, as judged by cellular localization, histomorphologicappearance and counterstaining against pepsinogen A and chro-mogranin A, respectively. Of note, the Gb3 synthase gene seems tobe dispensable for stomach development and function (38).Proliferating cells in the isthmus, lectin-marked surface epithelia,as well as mesenchymal stroma cells, did not show Gb3 expres-sion. Therefore, as opposed to intestinal epithelia, where Gb3expression is restricted to enteroendocrine cells andnot detectablein the bulk of epithelial cells (28), there is physiologic expressionof Gb3 in human nondiseased gastric mucosa. Several cell typesand tissues express Gb3, among them myeloid immune cells,which are insensitive to the toxin (39). Gb3 expression has beenfound in children in the kidney, in accordance with Shiga-toxinbeing the disease agent for the childhood hemolytic-uremicsyndrome (40). In adults, Gb3 expression in glomeruli is stronglyreduced to absent (41). Furthermore, endothelial cells have a highGb3 content, and can be further stimulated by proinflammatorycytokines to express increased levels of Gb3 (42). Preclinicalexperiments in pigs have shown that the gastric mucosa of thefundus and antrum contains considerably less Gb3 than otherinternal organs such as kidney, colon, liver, and spleen (43).

Next, we analyzed n ¼ 50 patients with gastric cancer forintratumoral Gb3 expression and distribution. We observed themajority of gastric carcinomas to express the glycosphingolipidGb3 (36/50 cases, 72%). The fraction of Gb3-positive gastriccancers is similar to data reported for pancreatic adenocarcinoma,where increased Gb3 levels were found in 62% to 78% of tumors,respectively (5, 8). No Gb3 was detected in 14 patients (28%), ascompared with normal tissue samples, where 45% of the samples(9/29) were devoid of Gb3 staining. Gb3 expression did not differbetween patients with chronic gastritis or noninflamed mucosa.Of note, the heterogeneous pattern of intratumoral Gb3 expres-sion resembled in size and extent to the Gb3 expressing gland-likestructures in colon carcinomaobserved previously (6). Expressionwas mainly confined to cancer cells in glandular structures in theintestinal-type lesions, but was also detected in tumor-associatedblood vessels from cancers of all differentiation types. However,no significant differences between Gb3 expression were observ-able between intestinal or diffuse-type carcinoma (3).

We observed a significant negative association of Gb3 expres-sion with increasing tumor TNM stage, as well as with the size ofthe primary lesion. This may indicate that Gb3-negative tumors

Figure 4.Uptakekinetics intracellular retrograde transport, after incubationof liveSt3051cellswithSTxB-Cy3 (red), afterfixation staining forgolgin p97 (green)andnuclei (DAPI,blue). A, incubation on ice for 15 minutes inhibits active uptake, only plasma membrane staining is observable. B, after 15 minutes at 37�C, STxB is alreadyaccumulating within the cell, partially colocalizing with golgin p97 at perinuclear areas (arrows). C, after 60 minutes at 37�C, most of intracellular STxB has reachedthe Golgi (arrows). D, after 48-hour chase, STxB can still be detected in perinuclear areas around the Golgi (arrows). Enlargement, 1,000-fold; size bar, 10 mm.

Geyer et al.





have more aggressive properties. However, the link between Gb3expression and pathologic parameters or prognosis is still largelyunknown. In pancreatic, ovarian, and mammary carcinomas, ahigher level of Gb3 expression was noticed in less differentiatedtissue (5, 19, 25). In accordance with our findings on gastriccancer, Gb3-positive mammary carcinomas show lower malig-nant behavior, as comparedwithGb3-negative tumors (25). Thereare only few reports on Gb3 expression in stomach cancer.However, it was previously shown that stroma fibroblasts inscirrhous carcinomas, a rare subgroup of diffuse-type cancerfeaturing important stromal components, express Gb3 (44).

Similar to the clinical samples, 8 of 10 gastric cancer cell linesexpressedGb3, as evidenced independently byflowcytometry andimmunofluorescence microscopy. As intracellular uptake of STxBis a prerequisite for targeted therapy, we tested its uptake andintracellular routing in gastric cancer cells. Shiga toxin is taken upby clathrin-dependent (45) and clathrin-independent endocyto-sis (15), followed by intracellular transport to the endoplasmicreticulum, along the retrograde route (10). As the A-subunit of theholotoxin induces cell death (46), we used only nontoxic STxB as

delivery tool. In accordance to earlier results for pancreatic orcolorectal cancer cells (6), we observed a rapid and completeuptake of STxB in the gastric cancer cell line St3051. Within 15minutes, STxB reached theGolgi apparatus. This shows for thefirsttime that uptake of STxB in gastric cancer cells follows theretrograde trafficking route, avoiding the degrading environmentof the lysosomes. Binding and intracellular uptake were main-tained at acidic pH, even though thismay not even be required forefficient tumor targeting within a gastric tumor mass beyond theacid-exposed surface.

In tissue sections of clinical samples, we frequently observed astriking heterogeneity of Gb3 expression. This might be explainedby the genetic heterogeneity inherent to tumorigenesis, or bydifferences in tumor cell differentiation. In breast cancer cells, afunctional implication of Gb3 expression for cancer stem cellproperties is controversially discussed (47, 48). However, thephysiologic or pathophysiologic regulation of Gb3 biosynthesisis not fully understood. In cell culture experiments, we observed astriking downregulation of Gb3 expression with increased cultureperiods, which was negatively correlated to increased cellular

Figure 5.Cell proliferation assays show that the compound STxB–SN38 has highly increased and receptor-specific efficacy in gastric cancer cells, compared with the clinicallyused drug irinotecan. A, effects of irinotecan are comparable in Gb3-positive (St3051) or Gb3-negative (AGS) cell lines. B, STxB–SN38 inhibits cell growth efficientlyin St3051 cells, whereas Gb3-negative AGS cells show only minor effects. C, quantitation of half-maximal inhibition (n ¼ 3 independent assays, in triplicates).The efficiency of STxB–SN38 is over 100-fold increased as compared with irinotecan, and specific for Gb3-positive cells.






senescence, as indicated by altered cell morphology and senes-cence-associated b-galactosidase activity (49). The term senes-cence has originally been coined for primary cell lines, but wasalso observed to be induced by oncogenic stress in cancer cells.

Depletion of glucose or serum components could be excludedas reason for the downregulated biosynthesis of Gb3, as onlycomplete replacement of the cell culture medium induced Gb3biosynthesis in long-term cultured cells, but not addition ofglucose or fresh serum. However, further nutrient limitations oraccumulation of catabolic end products in the culture mediummight participate to the observed transient loss of proliferativecapacity. Cellular senescence is thought to suppress tumorigen-esis, defined as stable loss of proliferative capacity. Senescence canbe induced by oncogenic stress, DNA damage, oxidative stress,and exposure to cytotoxic therapies. In accordance to in vitroresults, staining of parallel tissue sections of gastric cancer indi-cated that Gb3 was absent in senescent tissue areas that stainedpositive for senescence-associated b-galactosidase. Therefore, wepropose cellular senescence, induced by oncogenic stress orhypoxia, to potentially inhibit Gb3 expression in the tumorcontext.

As the majority of gastric carcinoma expressed Gb3, and gastriccancer cells showed uptake of STxB along the retrograde route, weinvestigated the use of functionalized STxB covalently coupled viaa disulfide-linker to SN38, the active metabolite of the cytotoxicdrug irinotecan. Hence, binding of STxB–SN38 on the target cellresults in a fast uptake of the STxB-drug conjugate, and in releaseof the drug, a topoisomerase type I inhibitor, within the targetcells. As expected, our data show that SN38, whose hydropho-bicity and toxicity limit its therapeutic efficacy, is orders ofmagnitude more effective than the clinically applied drug irino-tecan (50). Moreover, the STxB–SN38 conjugate has the advan-tage of high receptor specificity, demonstrated by the fact that Gb3-negative AGS cells shownodetectable toxicity at the dosages usedin our study, even though they are clearly sensitive to the type Itopoisomerase inhibitors. Thus, a targeted therapy approachbased on STxB–SN38 holds the potential to be far more efficientin tumor killing, with reduced toxic side effects. Several potentiallimitations need to be considered: Gb3 was often expressed onlyin parts of the tumors, which may lead to acquired therapyresistance by escaping tumor cells.However, our data indicate thatGb3-negative cells are frequently senescent, potentially reducingthe risk of selective escape. Moreover, Gb3 expression wasdecreased in more advanced tumor stages, which have the mosturgent need for improved therapies.However, Gb3 expressionwasnot fully absent frommore advanced tumors. In fact, two of three

advanced-stage cancers were Gb3-positive, indicating a need forpatient stratification for Gb3 expression. Lastly, some cells innormal gastric mucosa were also positive for Gb3. Therefore,targeted therapy may affect noncancer cells as well, which couldlead to transient tissue damage in the gastric mucosa that mightbe compensated by self-renewal of gastric epithelia. Importantly,Gb3-positive cells in normal gastric mucosa were exclusivelynondividing. Thus, by using a compound that affects proliferatingcells, such as SN38, only cells that are (a) fast dividing and (b)Gb3-positive would be targeted. In contrast, Gb3-positive but nondi-viding cells in gastric and renal epitheliumwould be spared. Thus,a STxB-based targeted therapy may be feasible for patients withGb3-expressing gastric carcinoma.

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

Authors' ContributionsConception and design: P. Emanuel Geyer, M. Maak, K.-P. JanssenDevelopment of methodology: P. Emanuel Geyer, M. Maak, K.-P. JanssenAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): P. Emanuel Geyer, M. Maak, U. Nitsche, M. Perl,A. Novotny, A. Holtorf, K.-P. JanssenAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): P. Emanuel Geyer, M. Maak, U. Nitsche, M. Perl,J. Slotta-Huspenina, A. Holtorf, K.-P. JanssenWriting, review, and/or revision of the manuscript: P. Emanuel Geyer,M. Maak, M. Maak, U. Nitsche, M. Perl, A. Novotny, K.-P. JanssenAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): P. Emanuel Geyer, M. Maak, J. Slotta-Huspe-nina, E. Dransart, L. JohannesStudy supervision: M. Maak, K.-P. Janssen

AcknowledgmentsThe authors thank Alexandra Gnann, Anja Conrad, and Widya Johannes for

excellent technical assistance, and Bernhard Holzmann for critical discussion.

Grant SupportK.-P. Janssen was supported by Deutsche Forschungsgemeinschaft, JA 1024/

3-1 and L. Johannes byHuman Frontier Science Program grant RGP0029-2014,and European Research Council advanced grants (project 340485). TheJohannes team ismember of Labex CelTisPhyBio (11-LBX-0038) and Idex ParisSciences et Lettres (ANR-10-IDEX-0001-02 PSL).

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received July 29, 2015; revised January 13, 2016; accepted January 15, 2016;published OnlineFirst January 29, 2016.

References1. Schirren R, ReimD, Novotny AR. Adjuvant and/or neoadjuvant therapy for

gastric cancer? A perspective review. Ther Adv Med Oncol 2015;7:39–48.2. Bertuccio P, Chatenoud L, Levi F, Praud D, Ferlay J, Negri E, et al. Recent

patterns in gastric cancer: a global overview. Int J Cancer 2009;125:666–73.3. Lauren P. The two histological main types of gastric carcinoma: diffuse and

so-called intestinal-type carcinoma. An attempt at a histo-clinical classifi-cation. Acta Pathol Microbiol Scand 1965;64:31–49.

4. Bauwens A, Betz J, Meisen I, Kemper B, Karch H, Muthing J. Facingglycosphingolipid-Shiga toxin interaction: dire straits for endothelial cellsof the human vasculature. Cell Mol Life Sci 2013;70:425–57.

5. Distler U, Souady J, Hulsewig M, Drmic-Hofman I, Haier J, Friedrich AW,et al. Shiga toxin receptorGb3Cer/CD77: tumor-association andpromisingtherapeutic target in pancreas and colon cancer. PLoS One 2009;4:e6813.

6. Falguieres T, Maak M, von Weyhern C, Sarr M, Sastre X, Poupon MF, et al.Human colorectal tumors and metastases express Gb3 and can be targetedby an intestinal pathogen-based delivery tool. Mol Cancer Ther 2008;7:2498–508.

7. Janssen KP, Alberici P, Fsihi H, Gaspar C, Breukel C, Franken P, et al.APC and oncogenic KRAS are synergistic in enhancing Wnt signaling inintestinal tumor formation and progression. Gastroenterology 2006;131:1096–109.

8. Maak M, Nitsche U, Keller L, Wolf P, Sarr M, Thiebaud M, et al. Tumor-specific targeting of pancreatic cancer with Shiga toxin B-subunit. MolCancer Ther 2011;10:1918–28.

9. Simons K, Ikonen E. Functional rafts in cell membranes. Nature 1997;387:569–72.

Geyer et al.





10. Johannes L, Popoff V. Tracing the retrograde route in protein trafficking.Cell 2008;135:1175–87.

11. O'Brien AD, Tesh VL, Donohue-Rolfe A, Jackson MP, Olsnes S,Sandvig K, et al. Shiga toxin: biochemistry, genetics, mode ofaction, and role in pathogenesis. Curr Topics Microbiol Immunol1992;180:65–94.

12. Jacewicz M, Clausen H, Nudelman E, Donohue-Rolfe A, Keusch GT.Pathogenesis of shigella diarrhea. XI. Isolation of a shigella toxin-bindingglycolipid from rabbit jejunum and HeLa cells and its identification asglobotriaosylceramide. J Exp Med 1986;163:1391–404.

13. Ling H, Boodhoo A, Hazes B, Cummings MD, Armstrong GD, Brunton JL,et al. Structure of the shiga-like toxin I B-pentamer complexed with ananalogue of its receptor Gb3. Biochemistry 1998;37:1777–88.

14. Solovyeva V, Johannes L, Simonsen AC. Shiga toxin induces membranereorganization and formation of long range lipid order. Soft Matter2015;11:186–92.

15. Romer W, Berland L, Chambon V, Gaus K, Windschiegl B, Tenza D, et al.Shiga toxin induces tubular membrane invaginations for its uptake intocells. Nature 2007;450:670–5.

16. Johannes L, Decaudin D. Protein toxins: intracellular trafficking for tar-geted therapy. Gene Ther 2005;12:1360–8.

17. Bast DJ, Brunton JL, Karmali MA, Richardson SE. Toxicity and immuno-genicity of a verotoxin 1 mutant with reduced globotriaosylceramidereceptor binding in rabbits. Infect Immun 1997;65:2019–28.

18. Levine MM, McEwen J, Losonsky G, Reymann M, Harari I, Brown JE, et al.Antibodies to shiga holotoxin and to two synthetic peptides of the Bsubunit in sera of patients with Shigella dysenteriae 1 dysentery. J ClinMicrobiol 1992;30:1636–41.

19. Arab S, Russel E, Chapman WB, Rosen B, Lingwood CA. Expression of theverotoxin receptor glycolipid, globotriaosylceramide, in ovarian hyperpla-sias. Oncol Res 1997;9:553–63.

20. Kalisiak A, Minniti JG, Oosterwijk E, Old LJ, Scheinberg DA. Neutralglycosphingolipid expression in B-cell neoplasms. Int J Cancer 1991;49:837–45.

21. Kovbasnjuk O. New insights into the role of Shiga toxins in intestinaldisease. Gastroenterology 2005;129:1354–5.

22. LaCasse EC, Bray MR, Patterson B, Lim WM, Perampalam S, Radvanyi LG,et al. Shiga-like toxin-1 receptor on human breast cancer, lymphoma, andmyeloma and absence from CD34(þ) hematopoietic stem cells: implica-tions for ex vivo tumor purging and autologous stem cell transplantation.Blood 1999;94:2901–10.

23. Murray LJ, Habeshaw JA, Wiels J, Greaves MF. Expression of Burkittlymphoma-associated antigen (defined by the monoclonal antibody38.13) on both normal andmalignant germinal-centre B cells. Int J Cancer1985;36:561–5.

24. Ohyama C, Fukushi Y, Satoh M, Saitoh S, Orikasa S, Nudelman E, et al.Changes in glycolipid expression in human testicular tumor. Int J Cancer1990;45:1040–4.

25. Stimmer L, Dehay S, Nemati F, Massonnet G, Richon S, Decaudin D, et al.Human breast cancer and lymph node metastases express Gb3 and can betargeted by STxB-vectorized chemotherapeutic compounds. BMC Cancer2014;14:916.

26. Storck W, Meisen I, Gianmoena K, Plager I, Kouzel IU, Bielaszewska M,et al. Shiga toxin glycosphingolipid receptor expression and toxin suscep-tibility of human pancreatic ductal adenocarcinomas of differing originand differentiation. Biol Chem 2012;393:785–99.

27. Couture O, Dransart E, Dehay S, Nemati F, Decaudin D, Johannes L, et al.Tumor delivery of ultrasound contrast agents using Shiga toxin B subunit.Mol Imaging 2011;10:135–43.

28. Janssen KP, Vignjevic D, Boisgard R, Falguieres T, Bousquet G,DecaudinD,et al. In vivo tumor targeting using a novel intestinal pathogen-baseddelivery approach. Cancer Res 2006;66:7230–6.

29. El Alaoui A, Schmidt F, Amessou M, Sarr M, Decaudin D, Florent JC, et al.Shiga toxin-mediated retrograde delivery of a topoisomerase I inhibitorprodrug. Angew Chem 2007;46:6469–72.

30. Ebrahimnejad P, Dinarvand R, Sajadi A, Jaafari MR, Nomani AR, Azizi E,et al. Preparation and in vitro evaluation of actively targetable nanoparticlesfor SN-38 delivery against HT-29 cell lines. Nanomedicine 2010;6:478–85.

31. Bugat R. Irinotecan in the treatment of gastric cancer. Ann Oncol 2003;14Suppl 2:ii37–40.

32. Thuss-Patience PC, Kretzschmar A, Bichev D, Deist T, Hinke A, BreithauptK, et al. Survival advantage for irinotecan versus best supportive care assecond-line chemotherapy in gastric cancer–a randomised phase III studyof the Arbeitsgemeinschaft Internistische Onkologie (AIO). Eur J Cancer2011;47:2306–14.

33. Zeestraten EC, Maak M, Shibayama M, Schuster T, Nitsche U, MatsushimaT, et al. Specific activity of cyclin-dependent kinase I is a new potentialpredictor of tumour recurrence in stage II colon cancer. Br J Cancer2012;106:133–40.

34. Mallard F, Johannes L. Shiga toxin B-subunit as a tool to study retrogradetransport. Methods Mol Med 2003;73:209–20.

35. Novak A, Binnington B, Ngan B, Chadwick K, Fleshner N, Lingwood CA.Cholesterol masks membrane glycosphingolipid tumor-associated anti-gens to reduce their immunodetection in human cancer biopsies. Glyco-biology 2013;23:1230–9.

36. Viel T, Dransart E, Nemati F, Henry E, Theze B, Decaudin D, et al. In vivotumor targeting by the B-subunit of shiga toxin. Mol Imaging 2008;7:239–47.

37. Lingwood CA.Glycolipid receptors for verotoxin and Helicobacter pylori:role in pathology. Biochim Biophys Acta 1999;1455:375–86.

38. Okuda T, TokudaN, Numata S, ItoM,OhtaM, Kawamura K, et al. Targeteddisruption of Gb3/CD77 synthase gene resulted in the complete deletionof globo-series glycosphingolipids and loss of sensitivity to verotoxins.J Biol Chem 2006;281:10230–5.

39. Falguieres T, Johannes L. Shiga toxin B-subunit binds to the chaperone BiPand the nucleolar protein B23. Biol Cell 2006;98:125–34.

40. Richardson SE, Karmali MA, Becker LE, Smith CR. The histopathology ofthe hemolytic uremic syndrome associated with verocytotoxin-producingEscherichia coli infections. Hum Pathol 1988;19:1102–8.

41. Lingwood CA. Verotoxin-binding in human renal sections. Nephron1994;66:21–8.

42. Louise CB, Obrig TG. Shiga toxin-associated hemolytic uremic syndrome:combined cytotoxic effects of shiga toxin and lipopolysaccharide (endo-toxin) on human vascular endothelial cells in vitro. Infect Immun1992;60:1536–43.

43. Boyd B, Tyrrell G, Maloney M, Gyles C, Brunton J, Lingwood C. Alterationof the glycolipid binding specificity of the pig edema toxin from globote-traosyl to globotriaosyl ceramide alters in vivo tissue targetting and results ina verotoxin 1-like disease in pigs. J Exp Med 1993;177:1745–53.

44. Sawada R, Hotta H, Chung YS, Sowa M, Tai T, Yano I. Globotriaosylceramide and globoside as major glycolipid components of fibroblasts inscirrhous gastric carcinoma tissues. Jpn J Cancer Res 1998;89:167–76.

45. Sandvig K, Olsnes S, Brown JE, Petersen OW, van Deurs B. Endocytosisfrom coated pits of Shiga toxin: a glycolipid-binding protein from Shigelladysenteriae 1. J Cell Biol 1989;108:1331–43.

46. Johannes L, Romer W. Shiga toxins–from cell biology to biomedicalapplications. Nat Rev Microbiol 2010;8:105–16.

47. Liang YJ,Ding Y, Levery SB, LobatonM,HandaK,Hakomori SI. Differentialexpression profiles of glycosphingolipids in humanbreast cancer stem cellsvs. cancer non-stem cells. Proc Natl Acad Sci U S A 2013;110:4968–73.

48. Gupta V, Bhinge KN, Hosain SB, Xiong K, Gu X, Shi R, et al. Ceramideglycosylation by glucosylceramide synthase selectivelymaintains the prop-erties of breast cancer stem cells. J Biol Chem 2012;287:37195–205.

49. Itahana K, Campisi J, Dimri GP. Methods to detect biomarkers of cellularsenescence: the senescence-associated beta-galactosidase assay. MethodsMol Biol 2007;371:21–31.

50. Buck A, Halbritter S, Spath C, Feuchtinger A, Aichler M, Zitzelsberger H,et al. Distribution andquantificationof irinotecan and its activemetaboliteSN-38 in colon cancer murine model systems using MALDI MSI. AnalBioanal Chem 2015;407:2107–16.






2016;15:1008-1017. Published OnlineFirst January 29, 2016.Mol Cancer Ther Philipp Emanuel Geyer, Matthias Maak, Ulrich Nitsche, et al. B-Subunit/CD77: Targeting of Gastric Cancer Cells with Shiga Toxin

3Gastric Adenocarcinomas Express the Glycosphingolipid Gb

Updated version

10.1158/1535-7163.MCT-15-0633doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2016/01/29/1535-7163.MCT-15-0633.DC1

Access the most recent supplemental material at:

Cited articles

http://mct.aacrjournals.org/content/15/5/1008.full#ref-list-1

This article cites 50 articles, 13 of which you can access for free at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://mct.aacrjournals.org/content/15/5/1008To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-15-0633

http://mct.aacrjournals.org/content/suppl/2016/01/29/1535-7163.MCT-15-0633.DC1

http://mct.aacrjournals.org/content/15/5/1008.full#ref-list-1

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

mailto:[email protected]

http://mct.aacrjournals.org/content/15/5/1008


Documents

Gastric Adenocarcinomas Express the …mct.aacrjournals.org/content/molcanther/15/5/1008.full.pdfLarge Molecule Therapeutics Gastric Adenocarcinomas Express the GlycosphingolipidGb