R. Berger and Cristina VossMartin Frank, Jeffrey C. Gildersleeve, Martin Helene Bayer, Katharina Essig, Sven Stanzel,  Cellular TargetingProperties Reveals a Novel Mechanism for Evaluation of Riproximin BindingGlycobiology and Extracellular Matrices:

doi: 10.1074/jbc.M112.368548 originally published online August 7, 20122012, 287:35873-35886.J. Biol. Chem. 

  10.1074/jbc.M112.368548Access the most updated version of this article at doi:

  .JBC Affinity SitesFind articles, minireviews, Reflections and Classics on similar topics on the

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

Supplemental material:

  http://www.jbc.org/content/suppl/2012/08/07/M112.368548.DC1.html

  http://www.jbc.org/content/287/43/35873.full.html#ref-list-1

This article cites 59 references, 19 of which can be accessed free at

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

by guest on February 4, 2014

http://ww

w.jbc.org/

Dow

nloaded from

Evaluation of Riproximin Binding Properties Reveals a NovelMechanism for Cellular Targeting*□S

Received for publication, April 3, 2012, and in revised form, August 5, 2012 Published, JBC Papers in Press, August 7, 2012, DOI 10.1074/jbc.M112.368548

Helene Bayer‡1, Katharina Essig‡, Sven Stanzel§, Martin Frank¶, Jeffrey C. Gildersleeve�, Martin R. Berger‡,and Cristina Voss**2

From the ‡Toxicology and Chemotherapy Unit, German Cancer Research Center, Im Neuenheimer Feld 581, 69120 Heidelberg,Germany, the §Department of Biostatistics, German Cancer Research Center, Im Neuenheimer Feld 581, 69120 Heidelberg,Germany, ¶BIOGNOS AB, Generatorsgatan 1, 41705 Goeteborg, Sweden, the �Chemical Biology Laboratory, National CancerInstitute, Frederick, Maryland 21702, and the **Department of Biochemistry, Heidelberg Pharma GmbH, Schriesheimer Strasse 101,68526 Ladenburg, Germany

Background: Riproximin is a cytotoxic lectin from Ximenia americana showing tumor selectivity.Results: Riproximin selectively binds to two types of glycoconjugates present on glycoproteins, cross-linking them by its twobinding sites.Conclusion: The biologic activity of riproximin is determined by specific and dynamic interactions with multivalent, cancer-related glycan targets.Significance: The selectivity of riproximin for cancer cells relies on its unique targeting mechanism.

Riproximin is a cytotoxic type II ribosome-inactivating pro-tein showing high selectivity for tumor cell lines. Its binding tocell surface glycans is crucial for subsequent internalization andcytotoxicity. In this paper, we describe a unique mechanism ofinteraction and discuss its implications for the cellular targetingand cytotoxicity of riproximin. On a carbohydrate microarray,riproximin specifically bound to two types of asialo-glycans,namely to bi- and triantennary complex N-glycan structures(NA2/NA3) and to repetitive N-acetyl-D-galactosamine(GalNAc), the so-called clustered Tn antigen, a cancer-specificO-glycan on mucins. Two glycoproteins showing high riproxi-min binding, the NA3-presenting asialofetuin and the clusteredTn-rich asialo-bovine submaxillary mucin, were subsequentlychosen as model glycoproteins to mimic the binding interac-tions of riproximinwith the two types of glycans. ELISAanalyseswere used to relate the two binding specificities of riproximin toits two sugar binding sites. The ability of riproximin to cross-link the two model proteins revealed that binding of the twotypes of glycoconjugates occurs within different binding sites.The biological implications of these binding properties wereanalyzed in cellular assays. The cytotoxicity of riproximin wasfound to depend on its specific and concomitant interactionwith the two glycoconjugates as well as on dynamic avidityeffects typical for lectins binding to multivalent glycoproteins.The presence of definite, cancer-related structures on the cellsto be targeted determines the therapeutic potency of riproxi-min. Due to its cross-linking ability, riproximin is expected toshow a high degree of specificity for cells exposing both NA2/NA3 and clustered Tn structures.

Riproximin is a lectin with potent antineoplastic activity invitro and in vivo (1, 2). It was originally identified as the activecomponent of a powdered plant material used in African tradi-tional medicine for treating cancer and has subsequently beenisolated from the fruit kernels of Ximenia americana. Riproxi-min is selectively cytotoxic to cancer cell lines with IC50 in thenanomolar range. For example, the breast cancer MCF7 cellsshowed�500-fold higher sensitivity than the non-tumorigenicbreast epithelium MCF10A cells (3). This remarkable potencyand selectivity prompted us to examine the molecular mecha-nisms of cell targeting in more detail.Riproximin belongs to type II of the cytotoxic ribosome-in-

activating proteins (RIPs)3 (4). Several members of this familywere investigated as potential antineoplastic agents, mostprominently ricin. The mechanism of action of the RIP familyof proteins involves two key steps represented by the A- andB-chains of the protein. First, the B-chain, a lectin, binds to cellsurface glycans, resulting in internalization of the RIP. Withinthe cell, the A-chain, an rRNA N-glycosidase, depurinates the28S RNA, leading to transcriptional arrest and eventually celldeath (5). Recently, it was shown that the mode of action alsoinvolves the induction of the unfolded protein response (6).The cellular patterns of cytotoxicity and in vivo toxicity of a

particular RIP are primarily determined by its binding andinternalization efficiency (7). Several types of biomolecules,such as glycoproteins, glycosphingolipids, proteoglycans, orglycosylphosphatidylinositol-linked proteins, contribute to a

* This work was supported, in whole or in part, by the National Institutes ofHealth, NCI, Intramural Research Program.

□S This article contains supplemental Tables 1–3 and Figs. 1– 6.1 Supported by German Federal Ministry of Economics and Technology

Grants Pro Inno II KF0425101UL6 and ZIM KF2301002AJ0.2 To whom correspondence should be addressed. Tel.: 49-6203-1009-29; Fax:

49-6203-1009-764; E-mail: c.voss@hdpharma.com.

3 The abbreviations used are: RIP, ribosome-inactivating protein; aBSM,asialo-bovine submaxillary mucin; aOSM, asialo-ovine submaxillary mucin;ASF, asialofetuin; BSM, bovine submaxillary mucin; Fet, fetuin; LacNAc,N-acetyl-D-lactosamine; NA, sialyl-free complex N-glycan; NA2, bianten-nary complex N-glycan; NA3, triantennary complex N-glycan; OSM, ovinesubmaxillary mucin; Tn, single O-linked GalNAc; Tn3, three consecutiveO-linked GalNAc; MTT, 3-(4,5-dimethythiazol-2-yl)-2,5-diphenyltetrazo-lium bromide; CG, cell growth; GP, glycoprotein; Rpx, riproximin; IRC, inhi-bition of riproximin cytotoxicity; MUC, mucin; CEA, carcinoembryonicantigen.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 287, NO. 43, pp. 35873–35886, October 19, 2012© 2012 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A.

OCTOBER 19, 2012 • VOLUME 287 • NUMBER 43 JOURNAL OF BIOLOGICAL CHEMISTRY 35873

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

cell surface glycosylation pattern (8, 9) and are potentialriproximin targets.Investigations of the glycan structures from malignant tis-

sues showed that the tumor-associated glycosylation signifi-cantly differs from that of normal tissues. Mucins showingaberrant O-glycosylation are a typical feature of epithelia-de-rived cancer (8). N-Glycans are significantly altered in cancertoo (10). The presence of immature (11) or highly branchedcore-fucosylated structures (12–15) has been related to varioustypes of tumors.The first attempts to use ricin as a cancer therapeutic were

based on its higher cytotoxicity in transformed cells (16), amechanism that today would be called “targeted toward can-cer-related glycostructures.” Its development, however, faileddue to unexpectedly high toxicity (17). Riproximin showedpotent antineoplastic activity that might be derived from itsglycan binding profile. The aim of this study was to identify theglycans that function as binding receptors for riproximin andare thus responsible for its tumor-specific cytotoxicity.Using a glycan microarray, riproximin was found to specifi-

cally bind to two types of glycans, theN-glycan structuresNA2/NA3 and the Tn antigen, a prominent cancer-relatedO-glycan.The mechanism of binding of lectins to multivalent globularand linear glycoproteins was recently elucidated (18). Both theinteraction of Gal-binding galectins with multiple NA3 struc-tures on ASF (19) and that of GalNAc-binding soybean agglu-tinin with Tn onTn-rich porcine submaxillarymucin (20) weredescribed as depending on bind-and-jump and negative coop-erativity effects. Using similar model glycoproteins, the mech-anism of riproximin interaction with its glycotargets and impli-cations for its cytotoxicity were investigated.

EXPERIMENTAL PROCEDURES

RiproximinPurification andLabeling—Riproximinwas puri-fied from Ximenia americana fruit kernels as described before(3). In short, the purification procedure included an initialaqueous extraction of proteins from the crude kernel material,removal of lipids with chloroform, and subsequent chromato-graphic purification on a strong anion exchange resin andlactosyl-Sepharose.For detection, riproximin was fluorescently labeled with an

amine-reactive, N-hydroxysuccinimide ester-activated dye(DyLight 549, Pierce) as described by the manufacturer. Pro-tein-containing fractions were pooled, washed, and concen-trated (molecular weight cut-off, 10,000) in 20 mM Tris-HClbuffer, pH 7.5, with 200 mM NaCl.Fluorescently labeled riproximinwas additionally purified on

a lactosyl-Sepharose column (Lactosyl Sepharose 4 Fast Flow,GEHealthcare) as described before (3) to exclude all conjugateswith blocked binding sites. The eluate fractions were concen-trated (molecular weight cut-off 10,000), and the protein con-tent was determined by the absorbance at 280 nm.Integrity and biological activity of the labeled riproximin

were controlled by SDS-PAGE and by a cell viability assay withHeLa cells, respectively (see below). The dye payloadwas deter-mined from the UV-visible spectrum.

Binding Analysis of Riproximin in a Carbohydrate Micro-array—The carbohydrate microarrays were prepared asdescribed previously, with the following modifications (21, 22).The array contained 157 components, including 97 differentdefined synthetic carbohydrates as bovine serum albumin(BSA) or human serum albumin conjugates, 28 synthetic glyco-peptides, and 32 natural glycoproteins. For a list of the carbo-hydrate structures, see supplemental Table 1. Samples wereprinted in duplicate on SuperEpoxy 2 Protein glass slides(TeleChem International, Inc., Sunnyvale, CA) using a Bioro-botics MicroGrid II microarrayer (Genomic Solutions, AnnArbor, MI) fitted with Stealth pins (Telechem International;catalog no. SMP3, which produce �100-�m spots). Relativehumidity within the printing chamber was maintained at�50%. Due to the small footprint of a single array (4.5 � 4.5mm), 16 copies of the full array could be printed onto each glassslide. Printed glass slides were stored at �20 °C until use.Riproximin binding was evaluated using minor modifica-

tions of the previously reported protocol (21, 23). Briefly, slideswere fitted with a 16-well module (Grace Bio-Labs) to physi-cally separate the 16 printed copies of the array. Each well wasblocked with 3% BSA (200 �l/well; immunoglobulin-free BSA;Sigma-Aldrich) in PBS for 1 h. DyLight-labeled riproximin wasdiluted 1:25 in 3% BSA in PBS and then incubated on the array(75 �l/well) for 2 h at room temperature in the dark. Afterwashing six times with PBS, the well module was disassembled,and the slide was incubated in PBS for 5 min. The slide wascentrifuged at 453 � g for 5 min and then scanned using aGenePix Scanner 4000A (Molecular Devices Corp., UnionCity,CA). Slides were scanned at 10 �m resolution, and image anal-ysis was carried out with GenePix Pro 6.0 analysis software(Molecular Devices Corp.). The fluorescent spots were definedas circular regions of interest with a diameter of 100 �m.Regions of interest were allowed to be adapted to the actualfeature size by�30�m.After local background subtraction, themedian pixel intensity of regions of interest was used to obtaina single fluorescence value for each spot. Themean value of thetwo replicate spots was used as the final value for each arraycomponent. To investigate the competitive binding of riproxi-min with Tn3 structures, riproximin was incubated with 60�g/ml Tn3-BSA (15 Tn3 molecules/BSA molecule) in 3% BSAin PBS for 1 h prior to the incubation of the mixture on thecarbohydrate microarray.Desialylation of Bovine Submaxillary Mucin—Asialo-bovine

submaxillary mucin (aBSM) was prepared by incubating 9mg/ml BSM with 28 milliunits/mg neuraminidase (RocheApplied Science) in 50mM sodiumacetate buffer, pH5.0, for 3 hat 37 °C. As a control, the same amount of BSM was incubatedin buffer without neuraminidase. The degree of desialylationwas monitored by dot blot analysis using biotinylated wheatgerm agglutinin (Vector Laboratories, Peterborough, UK). Forcell culture experiments, aBSM and BSM control samples werefilter-sterilized (0.45 �m). BSM sample concentrations weredetermined using the Glycoprotein Carbohydrate EstimationKit (Thermo Scientific, Pierce) and a BSM standard curve. TheBSM sample concentration was also used to estimate theaBSM concentrations because the desialylation procedure

Binding Mechanism of Riproximin

35874 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 43 • OCTOBER 19, 2012

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

created additional reducing ends, which interfered with themeasurement.Enzymatic Deglycosylation of Asialofetuin—To deglycosylate

ASF, 6 �g/�l glycoprotein was incubated for 5 min at 100 °C indenaturing buffer containing 50 mM sodium phosphate buffer,pH 7.5, 0.1% SDS, and 50mM �-mercaptoethanol. The mixturewas cooled on ice, and 0.75% (v/v) Triton X-100 as well as 1 �lof N-glycosidase F (5000 units/ml; Sigma-Aldrich) were addedsubsequently. The control reaction contained no N-glycosi-dase. Reaction mixtures were incubated overnight at 37 °C.Protein deglycosylation was monitored by SDS-PAGE.Dot Blot Analysis—Glycoproteins were serially diluted in

PBS, and 1 �l of each dilution was spotted onto a nitrocellulosetransfermembrane (Whatman Protran, Dassel, Germany). Themembranewas dried at room temperature and blockedwith 5%BSA. The membrane was probed with riproximin (30 �g/ml in5% BSA) followed by an anti-riproximin monoclonal mouseantibody (mRpx-Ab 62). HRP-linked, human preadsorbed goatanti-mouse antibody (SantaCruz Biotechnology, Inc.) was usedfor detection.Isolation of N-Glycans from Glycoproteins—The isolation of

N-glycans from asialofetuin and fetuin was performed accord-ing to Karg et al. (24) with some modifications. 100 mg/mlglycoprotein was digested with 110milliunits/�l pepsin (RocheApplied Science) in 10 mM HCl for 48 h at 37 °C. Pepsin wasinactivated by raising the pH above 5.0 with NaOH. The sam-ples were buffered with sodium acetate buffer, pH 5.2 (finalconcentration 100 mM). Peptide-N-glycosidase F (Sigma-Al-drich) was subsequently added, and the sample was incubatedfor 24 h at 37 °C. Free N-glycans and smaller peptides wereseparated by ultrafiltration on membranes with a molecularweight cut-off of 30,000.To remove the peptides, the glycans were additionally puri-

fied on C18 resin (Waters Corp., Milford, MA) spin columns,which were prepared using 30 �l of resin per column pre-washed with ethanol followed by H2O. The N-glycan/peptidemixtures were added to the spin columns, incubated for 5 min,and centrifuged. The eluates were subsequently desalted on acation exchange resin (AG 50W-X8, hydrogen form, 100–200mesh; Bio-Rad). For this step, the resin was washed with H2Obefore transferring 0.6 ml to an empty spin column and dryingthe resin by centrifugation. The N-glycan- containing eluateswere loaded onto the resin, incubated at room temperature for10 min, and eluted by centrifugation. The N-glycan sampleswere dried in a SpeedVac and stored at 4 °C.Enzyme-linked Immunosorbent Assay (ELISA)—Glycopro-

tein binding assayswere performedusing Immuno96MicroW-ell solid plates (Maxisorp, NUNC, Langenselbold, Germany)coated with 1 �g/ml ASF or aBSM in PBS, respectively. Afterblocking with 5% BSA in PBS, a 1-h preincubated mixture ofriproximin and serially diluted glycoprotein was added.Riproximin was detected using the monoclonal riproximinantibody Rpx-mAb 62, followed by an HRP-coupled, humanpreadsorbed anti-mouse antibody (Santa Cruz Biotechnology,Inc.). Binding was colorimetrically measured using 3,3�,5,5�-tetramethylbenzidine. When ASF binding was assessed, 5%milk in PBS was used for blocking, and an ovine anti-bovine

asialofetuin IgG (AbD Serotec, Düsseldorf, Germany) followedby HRP-linked anti-sheep antibody was used for detection.Binding signals of 2–3 replicates were averaged and plotted.

Concentration-response curves were fitted using the log-logis-ticmodel. To directly refer the potential inhibition effects of theglycoproteins to the relative number of carbohydrate struc-tures, the number ofN-acetyl-D-lactosamine (LacNAc) and Tn(GalNAc) residues per ASF and aBSMwas estimated as follows.For ASF, a molecular mass of 48 kDa and three NA3s resultingin nine terminal LacNAc residueswere used for calculation. ForaBSM, a molecular mass of 400 kDa was assumed (25, 26). 920GalNAc residues/molecule were estimated for aBSM based ondata available for porcine submaxillary mucin, which possessesa molecular mass of 106 Da with �2300 GalNAc residues/mol-ecule (20).Desialylation of Cells withNeuraminidase—For desialylation

of the cell surface, MDA-MB-231, MCF7, or HeLa cells wereseeded into microplates and allowed to settle down overnight.Neuraminidase (Roche Applied Science) was diluted intomedium without FCS to a concentration of 1 milliunit in 50 �l.FCS-containing medium was removed from the cells andreplaced by the neuraminidase-containing FCS-free medium.Control cells received FCS free medium without neuramini-dase. The plates were incubated for 1 h at 37 °C. Subsequently,neuraminidase-containing and control media were replaced byfresh FCS-containing medium, and the cells were treated withserial dilutions of riproximin. Cell viability was tested with the3-(4,5-dimethythiazol-2-yl)-2,5-diphenyltetrazolium bromide(MTT) assay after 72 h of incubation (see below). Three inde-pendent experiments were performed for cells in which aneffect of neuraminidase exposure was observed on riproximincytotoxicity.For the analysis, neuraminidase-pretreated cells (Neu�/

Rpx) and control cells (Neu�/Rpx) were analyzed as two sepa-rate groups. For each experiment, antiproliferative activity wasmeasured in triplicate for each riproximin concentration aswell as for cells treated with solvent control. Observed antipro-liferative activity response values were normalized tomean val-ues of solvent control cells and averaged per riproximin con-centration. For each treatment group, the four-parameter log-logistic model (27) was fitted to the averaged normalizedantiproliferative activity values. From each of the two fittedantiproliferative activity curves, the IC50 value, defining theconcentration that produces 50% of the responsive maximalcytotoxic effect of riproximin, was estimated.The ratio of the two IC50 estimates (i.e. (IC50 Neu�/Rpx)/

(IC50 Neu�/Rpx)) and the corresponding 95% confidenceinterval were computed to assess statistical significance. TheIC50 shift was considered significant when the IC50 ratio explic-itly differed from a value of 1 and the respective 95% confidenceinterval did not include the value 1.Competitive MTT Cell Viability Assay—The cytotoxic activ-

ity of riproximinwas assessed using theMTT cell viability assayonHeLa,MCF7, andMDA-MB-231 tumor cell lines. Cells werepropagated in a humid atmosphere containing 5%CO2 at 37 °Cin medium that was supplemented with 10% FCS, 2 mM L-glu-tamine, 100 units/ml penicillin, and 100 �g/ml streptomycin.For the assay, the cells were seeded intomicroplates (2500 cells/

Binding Mechanism of Riproximin

OCTOBER 19, 2012 • VOLUME 287 • NUMBER 43 JOURNAL OF BIOLOGICAL CHEMISTRY 35875

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

well for HeLa, 3000 cells/well for MDA-MB-231, 3500 cells/well for MCF7) and allowed to settle down overnight. To ana-lyze the influence of single glycoproteins, serial dilutions of theglycoproteins ASF, Fet, aBSM, and BSMwere added to the cellsprior to the addition of riproximin in a concentration corre-sponding to its IC50 (HeLa, 0.14 ng/ml; MCF7, 0.10 ng/ml;MDA-MB-231, 0.75 ng/ml). Plates were incubated for 72 h, andthe cell growth (CG) was determined by the MTT method.To investigate the inhibitory effect of the glycoprotein-de-

rived N-glycans alone, the HeLa cells were treated with N-gly-cans obtained from ASF and Fet by peptide-N-glycosidase Ftreatment (see above). The glycan effect on the cytotoxicity ofriproximin was correlated to the originating protein amountthat had been used for deglycosylation.For data analysis, the inhibitory effect of the single glycopro-

teins (GP) on riproximin (Rpx) cytotoxicity (inhibition ofriproximin cytotoxicity (IRC) (%)) was calculated as (CGGP/Rpx �CGRpx)/(100 � CGRpx) � 100. Both values, the CGGP/Rpx (aver-age cell growth with glycoprotein and riproximin) and CGRpx(average cell growth with riproximin alone), were normalizedto the respective control without riproximin treatment. Com-puted IRC values were averaged for every glycoprotein concen-tration. Inverse cell viability curves were determined for eachsingle glycoprotein by fitting the four-parameter log-logisticmodel (27). The IRC50 value determined from the curve depictsthe glycoprotein concentration that inhibits 50% of the riproxi-min cytotoxicity. Accordingly, IRC25 and IRC75 values describethe concentrations inhibiting 25 and 75% of the riproximincytotoxicity.Statistical two-way analysis of variance with interaction was

conducted to assess significance. The effects of the independentfactors “group” (i.e. native versus analogical desialylated glyco-protein) and “concentration” (i.e. the concentration of the gly-coprotein) as well as the interaction effect between group andconcentration on calculated average IRC values (dependentmodel variable) were studied by performing correspondingglobal F-tests. Analysis of variance was carried out on a signif-icance level of 5% (i.e. p values ofp� 0.05 obtained in the F-testswere regarded as statistically significant).Combination Cell Viability Assays—For the investigation of

ASF and aBSM in combination, different concentrations ofthese glycoproteins were combined according to the inhibitingeffect of the single proteins. Concentrations corresponding tothe IRC25, IRC50, and IRC75 of ASF were combined with aBSMconcentrations corresponding to IRC75, IRC50, and IRC25,respectively, resulting in the initial concentration for each com-bination. Serial dilutions of all three combinations were addedto the cells prior to the treatment of the cells with riproximin, asdescribed above.Four MTT viability experiments were performed for each of

the three combinations. Data analysis was performed as for theexperiments with the single glycoproteins (see above). To cir-cumvent potential problems due to negative or�100% IRC, themodel parameters of the lower and upper curve asymptoteswere constrained to be �0 and �100, respectively. Combina-tion indices (CIs) were used to assess the effect of each combi-nation of ASF and aBSM on riproximin cytotoxicity. CIs andthe corresponding pointwise 95% confidence bounds were

computed for theoretical IRC values ranging between 0.01 and0.99 (i.e. between 1 and 99%), with a step size of � 0.005between two neighboring theoretical IRC values, as describedby Lee and Kong (27). For each of the three examined combi-nations of ASF and aBSM, estimated CIs and their 95% confi-dence bounds were plotted against the theoretical IRC values.Combination effects of ASF/aBSM combinations were charac-terized as synergistic for CI 1, as additive for CI 1, and asantagonistic for CI � 1.Statistical Software—Data analysis of the carbohydrate

microarray, ELISA, and cell viability experiments was per-formed with Microsoft Excel. For further analysis, the opensource statistical software environment R, version 2.8.0 (avail-able from the R Project Web site) was used. Specifically, the Rapplication package “drc” was applied for fitting cell viabilitycurves and for estimation of IC50 values from the calculatedcurves, as well as for computation of the corresponding IC50ratio with its 95% confidence interval. The same package wasused for fitting inverse cell viability curves and for estimation ofIRC25, IRC50, and IRC75 values from the computed curves.Combination indices and according pointwise 95% confidencebounds were computed using a specific R function that waswritten and provided online by Lee and Kong (27).

RESULTS

Carbohydrate Microarray Analysis—Fluorophore-labeledriproximin was investigated on a glycan microarray with 97carbohydrates, 28 glycoproteins, and 32 glycopeptides. Glycanstructureswith riproximinbinding signalsof�100relative fluo-rescence units (RFU) are shown in Fig. 1A. For fluorescenceintensity data of all 157 glycan structures, see supplementalTable 2. From these carbohydrates, riproximin bound selec-tively to NA2, NA3, and Tn3 structures. The NA structuresdesignate biantennary (NA2) and triantennary (NA3) complexN-glycans with terminal Gal�1–4GlcNAc- structures. TheTn3 glycan consists of three consecutive O-linked GalNAc�-serine residues that form a short mucin-like polypeptide back-bone, a type of structure also referred to as clustered Tn. Thebinding signal of riproximin toNA3was the strongest (�20,000RFU), followed by NA2 (�10,000 RFU). The binding signal ofriproximin to Tn3 was about 5-fold lower than to NA3 but stillsignificant. Binding was strongest at the highest Tn3 density(Ac-Tn3-03 Ac-Tn3-15 Ac-Tn3-27).From the group of the glycoproteins, riproximin showed

strong binding toASF (�15,000 RFU), a glycoproteinwithNA2or NA3 N-glycan structures and weaker binding (5000 RFU)to the desialylated glycoproteins bovine (aBSM) and ovine(aOSM) submaxillary mucin, both glycoproteins with massiveO-glycosylation and a high proportion of Tn antigens. Riproxi-min also showed significant binding to the carcinoembryonicantigen (CEA) that is reported to contain N-glycan structures(Fig. 1A).Taken together, the riproximin binding profile on the array

was very narrow and confined to two types of glycans, theNA2/NA3 structures from the group of complex N-glycans and theO-glycan Tn3 structures representing the mucin-type clus-tered Tn.

Binding Mechanism of Riproximin

35876 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 43 • OCTOBER 19, 2012

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

To further investigate the relationship between the two bind-ing activities, competitive binding experiments were per-formed. The preincubation of riproximin with Tn3 linked toBSA led to a decrease of riproximin binding to O-glycan-richstructures, such as Tn3 or the mucins aBSM and aOSM. Incontrast, theN-glycan structures NA2 and NA3 and the glyco-protein CEA showed no decrease in the binding signal. How-ever, the signal of ASF declined to �50%, suggesting the pres-ence of a second riproximin binding sitewith lower affinity (Fig.1A).Binding Analyses—For validation of the microarray binding

results, selected glycoproteins were investigated by dot blotanalysis. ASF, fetuin (Fet), N-deglycosylated ASF, aBSM, andBSMwere immobilized on a nitrocellulosemembrane and sub-sequently probed with riproximin. As expected, riproximinshoweddistinctly stronger binding to the desialylated glycopro-teins than to their native forms. N-Deglycosylation of ASFresulted in a significant loss of riproximin binding (Fig. 1B),confirming that riproximin binding is confined to the sugarcomponent of this glycoprotein. The two glycoproteins ASFand aBSMwere thus chosen as model proteins to represent thetwo types of glycans NA3 and clustered Tn for further charac-terization of the riproximin binding mechanism and its biolog-ical significance.The binding affinity of riproximin for the twomodel proteins

was measured using the microscale thermophoresis method.ForASF, twoKd values of 11 nM and 7�Mwere determined. Thesecond affinity constant is in line with the assumption thatriproximin also binds to a non-NA3 glycoconjugate on the ASF

molecule. Riproximin bound to aBSM with a Kd of �50 �M.Two similarKd values of 48 nM and 1.1�Mwere also detected ina preliminary isothermal titration calorimetry experiment,when riproximin was titrated with ASF, and the two-site bind-ingmodel was applied for the analysis of the data. However, dueto the high microdiversity of the binding partners, the interac-tionsmeasured in this experiment turned out to be too complexto be covered by the available models, and no reasonable stoi-chiometry data were obtained. The complete thermophoresisand isothermal titration calorimetry data are presented in sup-plemental Figs. 2 and 3, respectively.Competitive Binding Analyses of ASF and aBSM in ELISA—

To further characterize the binding of riproximin to the twoclasses of glycans, ELISA analyses were performedwith riproxi-min alone or under competitive conditions. On an ASF coat,the preincubation of riproximin with ASF or aBSM resulted ina decrease of the signal with increasing ASF or aBSM concen-trations (Fig. 2A). On each ASF molecule, three NA3s and thusnine LacNAc structures have been reported. Mucins such asaBSM, however, show a much higher GalNAc density of�1000/molecule. The competitive effect of both glycoproteinsin the ELISA was therefore also related to the absolute numberof the respective carbohydrate structure, LacNAc for ASF andGalNAc for aBSM (Fig. 2B). The resulting curves indicate thatthe aBSM competition was only effective at very high Tn con-centrations, as expected from the higher affinity of riproximinfor LacNAc structures. On aBSM coat, however, preincubationwith ASF did not reduce the riproximin signal, despite thehigher affinity of riproximin for ASF (Fig. 2C).

FIGURE 1. Binding profile of riproximin. A, fluorophore-labeled Rpx (27 �g/ml) was applied to the carbohydrate microarray. For the competition experiment,riproximin was preincubated with 60 �g/ml BSA carrying 15 Tn3 moieties (Rpx � Tn3). Carbohydrate structures with signals �100 RFU are shown. Forcarbohydrate abbreviations and the numerical data of all 157 glycan structures, see supplemental Tables 1 and 2, respectively. The number beside thecarbohydrate abbreviation refers to the average number of carbohydrates per molecule of BSA (e.g. Ac-Tn3-27 has 27 Tn3s/BSA molecule). Riproximinsignificantly bound to the bi- and triantennary structures NA3 and NA2 as well as to Tn3 structures at high glycan density. Within the group of the glycoproteins,riproximin significantly bound to ASF, CEA, aBSM, and aOSM. B, dot blots comparing the binding of riproximin to serial dilutions of desialylated or deglycosy-lated glycoproteins versus their unprocessed counterparts. ASF deg, N-deglycosylated ASF. N-Deglycosylation of ASF significantly reduced its Rpx bindingcapacity. Error bars, S.D.

Binding Mechanism of Riproximin

OCTOBER 19, 2012 • VOLUME 287 • NUMBER 43 JOURNAL OF BIOLOGICAL CHEMISTRY 35877

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

To test whether the lower signal at high aBSM concentra-tionswas not an experimental artifact, plateswere either coatedwithASF followed by incubationwith different riproximin con-centrations or directly coated with different riproximin con-centrations under identical conditions. Upon aBSM incuba-tion, detection of riproximin decreased with rising aBSMconcentrations for both ASF-bound and directly coatedriproximin but more steeply for the ASF coat (Fig. 2D). More-over, aBSM abolished the riproximin signal only for ASF boundriproximin completely.A riproximin signal increased to more than 100% was

detected on the ASF coat at low aBSM concentrations (Fig. 2,Aand D). This finding indicates that riproximin may have cross-linked riproximin-loaded aBSM to the ASF coat. To furtherinvestigate cross-linking of ASF and aBSM by riproximin, anantibody against ASF was used. Coated aBSM was incubatedwith a serial dilution of riproximin in the presence of 10 or 50�g/ml ASF. The results showed significant binding of ASF toaBSM that was dependent on the riproximin but not the ASFconcentration (Fig. 2E). As shown in Fig. 2C, aBSM cross-link-ing to itself was not observed. The riproximin signal remained

�100% even when lower aBSM concentrations were tested(data not shown). Conversely and consistent with the secondbinding constant observed for ASF, riproximin showed someASF cross-linking to itself at high riproximin/ASF ratios (Fig.2F).Influence of Different Glycoproteins and Sugar Moieties on

Riproximin Cytotoxicity—To investigate the biological signifi-cance of riproximin binding to the identified glycostructures,the chosenmodel glycoproteins ASF and aBSM and their sialy-lated counterparts Fet and BSM were tested in three differenttumor cell lines (HeLa, MCF7, and MDA-MB-231) to evaluatetheir effect on riproximin cytotoxicity.The neutral glycoproteins ASF and aBSM inhibited riproxi-

min cytotoxicity significantly more strongly than their sialy-lated counterparts BSM and Fet in all three cell lines tested (allp 0.01; F-test for factor “group”). In addition, the statisticalanalysis revealed that this effect was concentration-dependent,becausewith increasing concentrations, the difference betweenthe inhibitory effect of the neutral and sialylated glycoproteinsalso increased (all p 0.01; F-tests for factor “concentration”and for the interaction between “group” and “concentration”).

FIGURE 2. ELISA binding analysis. A, plates coated with ASF were incubated with riproximin/ASF or riproximin/aBSM mixtures. Riproximin was detected byRpx antibody. B, binding curves from A were correlated to the numbers of glycans present on each glycoprotein (LacNAc for ASF and GalNAc for aBSM). C, platescoated with aBSM were incubated with riproximin/ASF or riproximin/aBSM mixtures. Riproximin was detected by Rpx antibody. D, plates were either coatedwith ASF and incubated with 5 �g/ml riproximin (ASF coat � Rpx (5)) or directly coated with 5 �g/ml riproximin (Rpx (5) coat). Serial dilutions of aBSM weresubsequently applied. Riproximin was detected by Rpx antibody. E, plates coated with aBSM were incubated with serial dilutions of riproximin in the presenceof 10 or 50 �g/ml ASF. Binding of ASF was detected by an anti-bovine ASF IgG. F, plates coated with decreasing ASF concentrations were incubated with 50�g/ml ASF alone or mixed with 10 or 30 �g/ml riproximin. Binding of ASF was detected by an anti-bovine ASF IgG. Each plot shows one representativeexperiment. Data are represented as mean � S.D. (error bars) from 2–3 replicates. Negative values due to background subtraction in plot D were set to zero.Concentration-response curves were fitted using the log-logistic model.

Binding Mechanism of Riproximin

35878 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 43 • OCTOBER 19, 2012

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

The relative inhibition of riproximin cytotoxicity achieved byASF or aBSM, however, varied between the cell lines. AlthoughASF and aBSMwere both able to completely abolish riproximincytotoxicity in HeLa cells (Fig. 3, A and B), only 70% maximalinhibition could be achieved in MCF7 cells (Fig. 3, D and E). Aparticularly intriguing finding was that the inhibitory effect ofaBSM decreased at high aBSM concentrations in MDA-MB-231 cells (Fig. 3G).

Cytotoxic activity of riproximin was also reduced in HeLacells following co-treatment withN-glycans that were obtainedby deglycosylation of ASF with peptide-N-glycosidase F. Thisfinding demonstrated that the inhibitory effects of glycopro-teins on riproximin cytotoxicity were glycan- but not core pro-tein-dependent (Fig. 3C).Influence of the Sialyl Caps on Riproximin Cytotoxicity—To

investigate the relationship between riproximin cytotoxicity

FIGURE 3. Cytotoxicity experiments. ASF versus Fet (A, D, and F) and aBSM versus BSM (B, E, and G) were added in combination with riproximin to HeLa (A andB), MCF7 (D and E), and MDA-MB-231 (F and G) cells and incubated for 72 h at 37 °C. The asialo-glycoproteins showed a significantly higher degree of inhibitionof riproximin cytotoxicity. C, N-glycans (N-Gly) derived from ASF or Fet were used to compete for riproximin cytotoxicity in HeLa cells as in A. The effect of theN-glycans was related to the original protein amount that had been deglycosylated. H, neuraminidase-pretreated (Neu�/Rpx) and control (Neu�/Rpx)MDA-MB-231 cells were treated with riproximin for 72 h at 37 °C. Vertical lines represent estimated IC50 values. Cell viability was determined by the MTT assay.For each graph, mean values from 2– 4 independent experiments were used and fitted using the four-parameter log-logistic model. For the inhibition effect ofaBSM/BSM on MDA-MB-231, a five-parametric Brain-Cousens hormesis model was used for fitting (61).

Binding Mechanism of Riproximin

OCTOBER 19, 2012 • VOLUME 287 • NUMBER 43 JOURNAL OF BIOLOGICAL CHEMISTRY 35879

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

and the abundance of sialyl groups on the cell surface, HeLa,MCF7, and MDA-MB-231 cells were treated with neuramini-dase and subsequently exposed to riproximin in a cellular via-bility assay. A detectable increase in riproximin cytotoxicitywas observed only in the MDA-MB-231 cells. For untreatedMDA-MB-231 cells, an IC50 value of 0.22 ng/ml (95% confi-dence interval, 0.192–0.250) was estimated for riproximin.After treatment with neuraminidase, the IC50 value decreasedsignificantly (IC50 0.08 ng/ml, 95% confidence interval,0.074–0.094). The ratio of the IC50 values with and withoutneuraminidase treatment of 2.64 (95% confidence interval,2.256–3.081) demonstrated that the enzymatic removal of thesialic acid caps from glycans on the cell surface led to signifi-cantly enhanced cytotoxicity of riproximin in MDA-MB-231cells (Fig. 3H). In contrast, neuraminidase treatment did notinfluence the cytotoxic potency of riproximin inMCF7 orHeLacells (data not shown). It is conceivable that the removal ofadditional sialic acids in the most sensitive, sialyl-poor MCF7cells could not further increase its riproximin response. Fur-thermore, the fact that neuraminidase treatment did not influ-ence the sensitivity of HeLa cells would indicate a low cell sur-face sialyl content.Competitive Influence of ASF and aBSM on Riproximin

Activity—To investigate the functional effects of the interac-tions between riproximin and the two different types of glycansthat were found to be its binding partners, the model proteinsASF and aBSM were added in combination to compete withriproximin in cytotoxicity assays. Because of the high inhibitoryeffect of both glycoproteins, HeLa cells were chosen for thecompetitive investigation. ASF and aBSM concentrations werecombined in three different proportions (ASF/aBSM of IRC25/IRC75, IRC50/IRC50, and IRC75/IRC25) according to the degreeof riproximin cytotoxicity inhibition each of themhad achievedalone (IRC values in Table 1). The three different combinationswere serially diluted, resulting in three concentration ranges(supplemental Table 3) and added in combinationwith riproxi-min to HeLa cells. The respective inverse cell viability curvesdepicting the cytotoxicity inhibition effects of the three combi-nations are shown in Fig. 4, A–C. The theoretical IRC valuescalculated for each combination model were examined foradditive, synergistic, or antagonistic effects (Fig. 4, D–F).

Combinations of the IRC25 of ASF with the IRC75 of aBSMcontained a lowerASF (i.e.NA-type active component) propor-tion and showed an additive effect that did not depend on theoverall cytotoxicity inhibition (Fig. 4D). The other combina-tions (ASF with IRC50 � aBSM with IRC50 as well as ASF withIRC75 � aBSM with IRC25) showed an effect that was depend-ent on the overall cytotoxicity inhibition degree achieved (Fig.4, E and F). At low riproximin cytotoxicity inhibition,27% forASF50/aBSM50 or 6% for ASF75/aBSM25, both combina-tions showed a synergistic effect. For inhibition degrees of26–79% (ASF50/aBSM50) or 6–71% (ASF75/aBSM25), anadditive effect was observed. At overall inhibition rates over79%, the combination effect turned into an antagonistic one(Fig. 4, E and F).In summary, the mode of inhibition of riproximin cytotoxic-

ity was dependent on the proportion of each glycoprotein (i.e.glycan type in the combination) as well as on the cumulatedglycoprotein amount. For most combinations, the effect wasadditive, which implies that ASF and aBSMwere able to inhibitriproximin in an independent manner.

DISCUSSION

NA2/NA3 and Cancer-related, Clustered Tn GlycostructuresAre Specifically Bound by Riproximin—Riproximin, a type IIRIP, was purified as a galactose/lactose-specific protein (3). Atfirst, the lactose binding and elution suggested broad galactosespecificity similar to that of ricin and abrin (28, 29), both verytoxic RIPs that renderRicinus communis andAbrus precatoriusseeds lethal on ingestion. However, the low peroral toxicity ofriproximin indicated that its binding profilewould be narrower.Indeed, analysis in a carbohydrate microarray revealed thatriproximin preferentially binds to two groups of glycoconju-gates, the branched bi- and triantennary N-glycan structuresNA2 and NA3 and the O-glycan structure known as clusteredTn, especially when present at high density. The terminal resi-due of NA2/NA3 structures is a Gal� that is connected to aGlcNAc of each antenna. Clustered Tn consists of a series ofsingle GalNAc� residues (single Tn) bound to adjacent aminoacids in protein regions rich in Ser/Thr repeats. They are typicalfor the extracellular domains of mucins. Tn and particularlyclustered Tn is an established tumor-specific antigen presenton many adenocarcinomas (30–32).Riproximin specificity for the clustered Tn antigen is

remarkable and could explain its tumor specificity describedpreviously (1, 2). Prominent targets for riproximin on cancercells could therefore be the cancer-associated mucins (MUCs).MUC 6, for example, was shown to be responsible for the highdensity of clustered Tn on the surface of the highly sensitiveMCF7 cells (33). Cancer-specific MUC 1 andMUC 2 were alsodescribed as Tn-rich glycoproteins (34, 35). However, riproxi-min showed its highest binding for bi- and triantennary com-plex N-glycans (NA2 and NA3 structures). No direct linkbetween the presence of NA2 or NA3 and cancer has beenestablished so far. However, N-glycans from cancer cells showhigher branching, resulting in a higher NA3 density (15, 36, 37)that would provide additional riproximin binding sites. On theother hand, the possibility cannot be excluded that these struc-

TABLE 1IRC concentration values as calculated from the fitted inverse cellviability curves and estimation of the related carbohydrateconcentrationShown is inhibition of riproximin cytotoxicity by 25% (IRC25), 50% (IRC50), or 75%(IRC75).

HeLa MCF7 MDA-MB-231IRC25 IRC50 IRC75 IRC50 IRC50

ASFProtein (�g/ml) 16 74 340 �1000a 59Protein (nM)b 333 1540 7080 19,900 1200LacNAc (�M)c 3 14 64 179 11

aBSMProtein (�g/ml) 0.1 0.4 1.6 6 0.3Protein (nM)b 0.3 1 4 14 0.8GalNAc (�M)c 0.3 0.9 3.7 13 0.7

a The real relative IRC50 value is �1000 �g/ml.b Molar concentrations were calculated using molecular masses of 48 kDa for ASFand 400 kDa for aBSM.

c The amount of the carbohydrate residues was estimated based on nine terminalLacNAc residues for an ASF and 920 GalNAc residues for an aBSMmolecule.

Binding Mechanism of Riproximin

35880 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 43 • OCTOBER 19, 2012

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

tures are also present on normal cells because 17% of the IgGmolecules in human serum are NA2-glycosylated (10).Riproximin Is Specific for Asialo-glycostructures—Riproxi-

min showed a clear preference for desialylated glycan struc-tures. Most cancer cells and cancer-related glycoproteins showabnormal sialylation, but its investigation revealed controver-sial results. For many tumors, increased sialyl-Lewis X expres-sion was described (15, 38–40). Conversely, N-glycans of theprostate-specific antigen showed a tumor-dependent decreaseof sialylation (41–43). Although cancer derivedO-glycans wereoften described as highly sialylated, they also contain the sialyl-free, cancer-specific Tn and T antigens (44). The sialidase Neu3, which is located in the plasma membrane and leads to sialyl-cap removal, is up-regulated during carcinogenesis (45).An increase in riproximin cytotoxicity was observed upon

neuraminidase treatment for the strongly sialylatedMDA-MB-231 cells but not for the sialyl-poor MCF7 (46), the most sensi-tive cell line within the panel tested (2). This experiment dem-onstrated a direct relationship between the biological activity ofriproximin and the extent of sialylation found on cell surfaces.Riproximin Displays a Narrow Binding Profile—Riproximin

demonstrated remarkable selectivity on the glycan array.Despite the presence ofmany other sialyl-free glycan structureswith terminal Gal� or GalNAc� on the array, such as LNnT

(Gal�1–4GlcNAc�1–3Gal�-) or the tumor-related Adi(GalNAc�1–3 Gal�-) (47), Forssman antigen (GalNAc�1–3GalNAc�-) (48), and TF antigen (Gal�1–3GalNAc�-) (49), nobinding of riproximin to these structures was detected. In com-parison, other commonly studied Gal- and GalNAc-bindinglectins, such as R. communis agglutinin, soybean agglutinin,Helix pomatia agglutinin, jacalin, and Bauhinia purpurea lec-tin, recognize awide range of glycans on the array. For example,the Gal-binding type II RIP R. communis agglutinin signifi-cantly bound to almost any structure with a terminal LacNAcand lactose (50). Thus, riproximin showed a very narrow bind-ing profile that was dependent not only on the nature of theterminal sugars but also on their amount and constellation.The preference of riproximin for trimeric structures, such as

NA3 and Tn3, was also remarkable. The 3-fold sugar specificitymight reflect the structure of the typical RIP B-chain bindingdomains, which are both trimers of an ancient lectin motif.However, molecular modeling of riproximin B-chain interac-tion with NA3 structures revealed that the terminal Gal resi-dues on NA3 cannot span the distance between two sub-domains but could interact with other aromatic residues thatare frequently present in the closer neighborhood (supplemen-tal Figs. 5 and 6). Amultivalent binding of riproximin toNA2 orNA3 structures could explain the observed high specificity,

FIGURE 4. Analysis of the combined competitive inhibitory effect of ASF and aBSM on the cytotoxicity of riproximin. ASF and aBSM were applied in acellular viability assay with riproximin in HeLa cells. Three glycoprotein combinations with the following ASF/aBSM active component proportions wereanalyzed: IRC25/IRC75 (A and D), IRC50/IRC50 (B and E), and IRC75/IRC25 (C and F). The cells were incubated with riproximin and a serially diluted glycoproteincombination for 72 h at 37 °C. Cell viability was determined by the MTT assay. For each combination of ASF/aBSM, average IRC values of four independentexperiments and the respective inverse cell viability curves were fitted by using the four-parameter log-logistic model (A–C). The combination effect of theglycoproteins was characterized by the CIs. For all theoretical IRC values of each ASF/aBSM combination, CI values and the respective confidence bounds werecalculated and plotted (D–F). Combination effects were characterized as synergistic for CI 1, as additive for CI 1, and as antagonistic for CI � 1.

Binding Mechanism of Riproximin

OCTOBER 19, 2012 • VOLUME 287 • NUMBER 43 JOURNAL OF BIOLOGICAL CHEMISTRY 35881

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

which is unusual for a lectin and comparable with that of amonoclonal antibody.On the other hand, the possibility cannot be completely ruled

out that the stronger binding signals on NA3 result from stoi-chiometric and/or bind-and-jump effects. Binding of two orthree ricin B-chains to a single NA3 structure has beendescribed (51). However, the fact that the presence of an addi-tional terminal Gal� in NA4 structures did not improve butinstead significantly reduced the binding signal contradicts thishypothesis.Glycoproteins as Binding Counterparts of Riproximin—Ac-

cordingly, the NA2/NA3-containing glycoproteins ASF (52)and CEA (53) as well as the Tn-rich glycoproteins aOSM andaBSM (54) on the array showed strong riproximin binding. ASFand aBSM were thus chosen as model proteins to mimic theeffects of the NA2/NA3 and Tn-structures, respectively, in cel-lular experiments.Both asialo-glycoproteins showed significant inhibition of

riproximin cytotoxicity in HeLa, MCF7, and MDA-MB-231cells. Moreover, deglycosylation of ASF resulted in loss of itsriproximin binding, whereas the ASF-derivedN-glycans signif-icantly inhibited riproximin cytotoxicity. These findings dem-onstrate that specific binding of riproximin to NA2/NA3and/or Tn glycostructures on the cellular surface is a prerequi-site for cytotoxicity.The degree of inhibition, however, strongly depended on the

cell line (i.e. the cell surface glycosylation). For example, �10-fold higher glycoprotein concentrations were required in theparticularly Tn-rich MCF7 cells (33) to reduce riproximin sen-sitivity by 50% as comparedwithHeLa cells, althoughHeLa andMCF7 are equally sensitive to riproximin. It must therefore beassumed that the abundance and surface distribution of glyco-structures on a particular cell play a crucial role in its sensitivityto riproximin. Moreover, it cannot be excluded that additionalriproximin glycotargets exist on the cell surface, which have notyet been identified.Two Different Binding Sites Are Associated with Different

Specificities of Riproximin—The specific binding of riproximinto both Gal� within NA2/NA3 and GalNAc� within Tn struc-tures related to the presence of two binding domains in theB-chain of riproximin. The structures of the clustered Tn anti-gen and NA2/NA3 are significantly different, and it is uncom-mon for a lectin or monoclonal antibody to bind both. Forexample, Tn-specific antibodies have never shown binding toNA2/NA3 (55). The very narrow binding profile of riproximinmakes the recognition of both clustered Tn and NA2/NA3 inthe same binding site very unlikely. We thus hypothesized thatriproximin binds the two sugar structures with different bind-ing sites corresponding to the two sugar binding sites of theB-chain.To evaluate this hypothesis, competitive analyses were per-

formed. In array experiments, preincubation of riproximinwithTn3 decreased the binding of riproximin to immobilized Tn3structures and aBSM but did not influence the NA binding.However, because the NA affinity of riproximin is significantlyhigher than its Tn3 affinity, it could not be excluded that theTn3 concentrations used in the experiment were too low toaffect NA binding.

The strongest argument supporting the existence of two dif-ferent binding specificities is based on the finding that riproxi-min was able to cross-link the two proteins ASF (NA2/NA3structures) and aBSM (Tn structures). Moreover, even at lowaBSM concentrations, no cross-linking of aBSM with itselfcould be detected. However, at very lowASF/riproximin ratios,a significant amount of ASF cross-linking was observed, whichprobably results from the interaction of the riproximin Gal-NAc-binding site with an additional, probable non-NA3 sugarof ASF. This finding is consistent with several other observa-tions regarding the interaction of riproximin with ASF, includ-ing the partial competition of ASF binding by Tn3 structuresobserved in the array, as well as the second affinity constantdetermined for riproximin and ASF. For fetuin, the presence ofan O-linked glycoconjugate has been described (56), and it isconceivable that the desialylated structure would be present onASF.Affinity of Riproximin for Glycostructures Depends onAvidity

Effects—In the microarray experiments, riproximin showed adistinctly stronger binding signal to NA2/NA3 than to Tn gly-costructures. In microscale thermophoresis experiments, thehigh affinity of riproximin for ASF was �1000-fold higher thanfor aBSM. In the ELISA experiments, however, aBSM was ableto competitively inhibit the binding of riproximin to ASF athigh concentrations but not vice versa. Moreover, in cellularexperiments, aBSM was the stronger cytotoxicity competitor.Even when the high number of Tn structures on aBSM wasconsidered, the concentration needed to achieve the same inhi-bition was lower for the aBSM-derived Tn than for the ASF-derived LacNAc. These observations demonstrate that theinteraction of riproximin with a particular glycostructure is notconfined to a strict lock-and-key correlation but is stronglyinfluenced by dynamic interactions similar to those describedby Dam and Brewer (18).Riproximin primarily interacts with terminal Gal and Gal-

NAc, respectively, via its two binding sites, with low but signif-icant affinity. The purification of riproximin on Gal-exposingresins was based on this primary affinity. Sequence analysis andmolecular modeling data revealed that the typical type II RIPGal-binding activity is retained for both �1 and �2 subdomainsof the riproximin B-chain (2). However, the highly specificinteractions of riproximinwith glycoconjugates presenting twoor three moieties of the preferred terminal sugars Gal (NA2/NA3) or GalNAc (Tn3) point toward additional glycan-proteininteractions.When the interactions of riproximin with each of the pro-

teins ASF and aBSM is considered, the main features of themechanism of lectins binding to multivalent targets apply (19,20). On multivalent structures, several equivalent epitopes areavailable for the so-called bind-and-jump effects; the lectin canbe recaptured by each of the remaining free epitopes, leading todecreased dissociation. On the other hand, negative coopera-tivity results from decreasing functional valence when morelectin molecules occupy additional epitopes. The affinity ofeach consecutive binding step fractionally decreases due to thedecreased number of free epitopes available for recapturing.Based on the hypothesis of the two binding sites specific for

NA2/NA3 and Tn3, respectively, riproximin would be able to

Binding Mechanism of Riproximin

35882 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 43 • OCTOBER 19, 2012

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

interact with three NA3 structures of ASF and with �1000 Tnclusters of aBSM, respectively. There is probably a second,O-linked glycan of ASF that would be able to interact with theTn3-binding site of riproximin. These interactions and theireffect on riproximin binding and detection in ELISA experi-ments are schematically depicted in Fig. 5, A–D. In particular,the high epitope density on aBSM is expected to allow broaddynamic interactions with the Tn3-binding site of riproximin.The increasing affinity of riproximin for BSA bearing Tn3 at

increasing density (Tn3-03Tn3-15Tn3-27) and its highaffinity for mucins bearing up to thousands of Tn per moleculereflect the bind-and-jump effects thatwere described byDam et

al. (57) for the interaction of soybean agglutinin with asialo-porcine submaxillary mucin. Together with the high Tn con-centration of aBSM, this dynamic increase in affinity explainsthe high competitive potency of aBSM in ELISA or cell experi-ments. On the other hand, because the NA2/NA3 structuresare only present at comparatively low density, they did not pro-vide enough binding sites for a detectable bind-and-jump-re-lated increase in specificity.The simultaneous interaction of riproximin with both pro-

teins therefore strongly depends on the relative amounts ofriproximin, NA2/NA3, and Tn3 epitopes. In the presence ofboth structures, riproximin cross-linked the NA2/NA3 and

FIGURE 5. Schematic of riproximin interactions with its two binding counterparts. A, schematic representation of the riproximin molecule showing its twodifferent binding sites (BS) and its possible interactions with ASF and aBSM, respectively, as described by Dam et al. (19) for galectin and ASF and later by Damet al. (20) for soybean agglutinin and asialo-porcine submaxillary mucin. At low riproximin concentrations, several binding epitopes are available on eachglycoprotein molecule, and an increased affinity would be observed due to bind-and-jump entropic effects. At high riproximin concentrations, most bindingsites on the glycoproteins have already been occupied. Due to negative cooperativity effects, the binding affinity is lower for each subsequent binding step.B–D, schematic representation of riproximin interacting with both of its counterparts as observed in ELISA experiments. On coated ASF at low aBSM concen-trations, cross-linking of aBSM to the plate occurs. Due to the high riproximin concentrations in the mixture, several other riproximin molecules are bound tothe cross-linked aBSM molecule, resulting in more detectable riproximin (B). At higher aBSM concentrations, competition and masking effects predominate (C).Riproximin cross-links ASF to the coated aBSM (D). E, proposed model for the selective targeting of NA3- and Tn3-presenting tumor cells. The cooperativebinding to the two different glycoconjugates confers riproximin an increased affinity for cells exposing both NA2/NA3 and cancer-related clustered Tnstructures, resulting in selectivity over cells presenting NA2/NA3 structures only. Moreover, cross-linking of the two structures on the cell surface would resultin increased internalization efficiency and thus cytotoxicity.

Binding Mechanism of Riproximin

OCTOBER 19, 2012 • VOLUME 287 • NUMBER 43 JOURNAL OF BIOLOGICAL CHEMISTRY 35883

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

Tn3 epitopes via its two binding sites (Fig. 5D). It is very likelythat cross-linking via riproximin also occurs on the cell surface.Cross-linking is known to be important for internalization ofseveral cell surface receptors (58, 59). Because ASF alone wasable to inhibit the cytotoxicity of riproximin by up to 50% inMCF7 cells that are particularly Tn-rich (33) but have 5%NA2/NA3 (60), it is very probable that the cross-linking isalso important for the internalization and cytotoxicity ofriproximin.In the presence of high aBSM concentrations, however, the

interaction of riproximin with the high density of Tn antigensmasked its epitopes. A large cell surface mucin molecule mighttherefore be able to recruit riproximin and concomitantlymaskits second, NA-specific binding site, thereby inhibiting itscross-linking potential (Fig. 5C).Riproximin Cytotoxicity Was Influenced by Its Dynamic

Interaction with ASF and aBSM—To investigate the interde-pendence of binding, cross-linking and cytotoxicity within thecomplex environment of a cell, viability experiments were per-formed in which both ASF (NA3) and aBSM (clustered Tn)were allowed to compete for riproximin binding and thusinhibit its toxicity. The results of these experiments revealedthat the interaction patterns of the mixtures containing a highaBSM proportion clearly differed from those of mixtures con-taining equal active amounts of aBSM and ASF or low aBSMproportions. Moreover, the combinatory effects of the lattertwo combinations were strongly dependent on the total glyco-protein concentration. The interactions expected to be respon-sible for these effects are schematically depicted in supplemen-tal Fig. 4.Mixtures with a high proportion of aBSM exhibited a broad

additivity, thus supporting the hypothesis that both bindingsites of riproximin are necessary for its cytotoxic activity. Inthese mixtures, the riproximin/aBSMmolecular ratio is partic-ularly low. Increased internal diffusion and bind-and-jumpeffects lead to an enhanced aBSM affinity, resulting in riproxi-min sequestration and masking of its ASF binding site. Verylittle cross-linking occurs under these circumstances, so thatthe entire ASF fraction would be available to bind to freeriproximin, leading to an additive effect.Conversely, mixtures containing �50% of aBSM showed

synergistic effects at low global concentrations that turned intoadditivity and eventually antagonism at high concentrations.Due to the higher riproximin/aBSM ratio, more riproximinmolecules are available to bind to the same aBSM molecule.Fewer Tn3 binding sites are available for bind-and-jumpeffects, leading to a lower apparent affinity and lessmasking. Athigh overall glycoprotein concentrations, ASF would bind tounmasked riproximin molecules that are also bound to aBSM;cross-linking would occur. Because a significant amount of gly-coproteins bind the same riproximin molecule, antagonismwould be observed. At low global concentration of both glyco-proteins, the chance for cross-links decreases, so that ASF andaBSM would bind different riproximin molecules. The syner-gistic effects observed under these circumstances strongly sug-gest that blocking of a single riproximin binding site signifi-cantly interferes with the activity of riproximin. A schematic ofthe interactions of riproximin with the two glycoproteins in

each of the three different cases discussed is presented in sup-plemental Fig. 4. Overall, these findings indicate that cross-linking is part of the mechanism responsible for riproximininternalization and cytotoxicity (Fig. 5E).In summary, even in the cellular context of a single cell type,

the interaction pattern of riproximin with its glycotargets wasvery complex and depended on dynamic effects. At lowriproximin concentrations, the internal diffusion along largeglycoproteins predominated. At high concentrations, nega-tive cooperativity opposed the bind-and-jump effects andfavored cross-linking. Cellular experiments revealed thatdespite the lower affinity of aBSM to riproximin in directbinding experiments, in cellular assays it had the strongerimpact on the cytotoxicity of riproximin.Conclusions—In summary, the antineoplastic-active type II

RIP riproximin was shown to specifically bind to two types ofglycostructures, theN-linked NA2/NA3 and theO-linked clus-tered Tn tumor-specific antigen. The two specificities wererelated to the two binding sites present on the riproximinB-chain. The sugar interactions of riproximin were shown tocombine high specificity with dynamic interactions that aretypical for lectins interacting with multivalent binding targets.Understanding the mechanism of riproximin targeting is

particularly important because its therapeutic potency stronglydepends on the presence of definite cancer-related structureson the cells to be targeted. Cross-linking of the two structuresNA2/NA3 and Tn3 confer on riproximin an enhanced selectiv-ity for cells exposing both structures. However, a relationshipbetween the concomitant occurrence of these two structuresand cancer has not yet been established. The ideal riproximintarget cell would contain surface glycostructures with high Tndensities and sialyl-freeNA3 structures.On the other hand, dueto the broad range of dynamic interactions riproximin can getinvolved in, it is conceivable that the biological activity ofriproximin might be modulated by the addition of particularglycan structures, such as glycopolymers with high riproximinbinding capacity but low affinity. Such a polymer could func-tion as a “carrier” for riproximin, preventing it from binding tolow affinity structures and delivering it specifically to tumorcells, where it would find high avidity targets.Because the field of glycobiology is rapidly expanding, the

availability of synthetic complex glycan structures and thedevelopment of novel tools will support further investigations.The detailed characterization of the binding properties ofriproximin reported here provides amodel for functional lectinstudies. Moreover, riproximin can be included in the panel oflectins with well characterized properties that currently find abroad usage in glycobiology.

Acknowledgments—We thank NanoTemper Technologies GmbH forkindly providing the Monolith instrument for the affinity measure-ment and the calculations of the affinity constants. We also thank Dr.Vladimir Rybin (Protein Expression and Purification Core Facility,European Molecular Biology Laboratory) for support in performingthe isothermal titration calorimetry experiment.

Binding Mechanism of Riproximin

35884 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 43 • OCTOBER 19, 2012

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

REFERENCES1. Voss, C., Eyol, E., and Berger, M. R. (2006) Identification of potent anti-

cancer activity in Ximenia americana aqueous extracts used by Africantraditional medicine. Toxicol. Appl. Pharmacol. 211, 177–187

2. Voss, C., Eyol, E., Frank,M., von der Lieth, C.W., and Berger,M. R. (2006)Identification and characterization of riproximin, a new type II ribosome-inactivating protein with antineoplastic activity fromXimenia americana.FASEB J. 20, 1194–1196

3. Bayer, H., Ey, N., Wattenberg, A., Voss, C., and Berger, M. R. (2011) Pro-tein Expr. Purif. 82, 97–105

4. Stirpe, F., and Battelli, M. G. (2006) Ribosome-inactivating proteins. Pro-gress and problems. Cell Mol. Life Sci. 63, 1850–1866

5. Endo, Y., and Tsurugi, K. (1988) The RNA N-glycosidase activity of ricinA-chain. The characteristics of the enzymatic activity of ricin A-chainwith ribosomes and with rRNA. J. Biol. Chem. 263, 8735–8739

6. Horrix, C., Raviv, Z., Flescher, E., Voss, C., and Berger, M. R. (2011) Plantribosome-inactivating proteins type II induce the unfolded protein re-sponse in human cancer cells. Cell Mol. Life Sci. 68, 1269–1281

7. Girbes, T., Ferreras, J. M., Arias, F. J., Munoz, R., Iglesias, R., Jimenez, P.,Rojo, M. A., Arias, Y., Perez, Y., Benitez, J., Sanchez, D., and Gayoso, M. J.(2003) Non-toxic type 2 ribosome-inactivating proteins (RIPs) from Sam-bucus. Occurrence, cellular and molecular activities, and potential uses.Cell Mol. Biol. 49, 537–545

8. Reis, C. A., Osorio, H., Silva, L., Gomes, C., and David, L. (2010) Altera-tions in glycosylation as biomarkers for cancer detection. J. Clin. Pathol.63, 322–329

9. Varki, A., Cummings, R. D., Esko, J. D., Freeze, H. H., Stanley, P., Bertozzi,C. R., Hart, G. W., and Etzler, M. E. (2009) Essentials of Glycobiology, 2ndEd., pp. 1–23, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

10. Arnold, J. N., Saldova, R., Hamid, U.M., and Rudd, P.M. (2008) Evaluationof the serum N-linked glycome for the diagnosis of cancer and chronicinflammation. Proteomics. 8, 3284–3293

11. de Leoz,M. L., Young, L. J., An, H. J., Kronewitter, S. R., Kim, J.,Miyamoto,S., Borowsky, A. D., Chew, H. K., and Lebrilla, C. B. (2011) Mol. CellProteomics 10,M110.002717

12. Alley, W. R., Jr., Vasseur, J. A., Goetz, J. A., Svoboda, M., Mann, B. F.,Matei, D. E., Menning, N., Hussein, A., Mechref, Y., and Novotny, M. V.(2012) N-Linked glycan structures and their expressions change in theblood sera of ovarian cancer patients. J. Proteome. Res. 11, 2282–2300

13. de Leoz, M. L., An, H. J., Kronewitter, S., Kim, J., Beecroft, S., Vinall, R.,Miyamoto, S., de Vere White, R., Lam, K. S., and Lebrilla, C. (2008) Gly-comic approach for potential biomarkers on prostate cancer. Profiling ofN-linked glycans in human sera and pRNS cell lines. Dis. Markers 25,243–258

14. Fang, M., Dewaele, S., Zhao, Y. P., Starkel, P., Vanhooren, V., Chen, Y. M.,Ji, X., Luo,M., Sun, B.M., Horsmans, Y., Dell, A., Haslam, S.M., Grassi, P.,Libert, C., Gao, C. F., and Chen, C. C. (2010) SerumN-glycome biomarkerfor monitoring development of DENA-induced hepatocellular carcinomain rat.Mol. Cancer 9, 215

15. Machado, E., Kandzia, S., Carilho, R., Altevogt, P., Conradt, H. S., andCosta, J. (2011) N-Glycosylation of total cellular glycoproteins from thehuman ovarian carcinoma SKOV3 cell line and of recombinantly ex-pressed human erythropoietin. Glycobiology 21, 376–386

16. Lin, J. Y., Tserng, K. Y., Chen, C. C., Lin, L. T., andTung, T. C. (1970)Abrinand ricin. New anti-tumor substances. Nature 227, 292–293

17. Fodstad,O., Kvalheim,G., Godal, A., Lotsberg, J., Aamdal, S., Høst, H., andPihl, A. (1984) Phase I study of the plant protein ricin. Cancer Res. 44,862–865

18. Dam, T. K., and Brewer, C. F. (2010) Multivalent lectin-carbohydrate in-teractions energetics andmechanisms of binding.Adv. Carbohydr. Chem.Biochem. 63, 139–164

19. Dam, T. K., Gabius, H. J., Andre, S., Kaltner, H., Lensch, M., and Brewer,C. F. (2005) Galectins bind to the multivalent glycoprotein asialofetuinwith enhanced affinities and a gradient of decreasing binding constants.Biochemistry 44, 12564–12571

20. Dam, T. K., Gerken, T. A., Cavada, B. S., Nascimento, K. S., Moura, T. R.,and Brewer, C. F. (2007) Binding studies of �-GalNAc-specific lectins to

the alpha-GalNAc (Tn-antigen) form of porcine submaxillary mucin andits smaller fragments. J. Biol. Chem. 282, 28256–28263

21. Oyelaran, O., Li, Q., Farnsworth, D., and Gildersleeve, J. C. (2009) Mi-croarrays with varying carbohydrate density reveal distinct subpopula-tions of serum antibodies. J. Proteome. Res. 8, 3529–3538

22. Wang, D., Liu, S., Trummer, B. J., Deng, C., and Wang, A. (2002) Carbo-hydratemicroarrays for the recognition of cross-reactivemolecularmark-ers of microbes and host cells. Nat. Biotechnol. 20, 275–281

23. Campbell, C. T., Zhang, Y., and Gildersleeve, J. C. (2010) Curr. ProtocolsChem. Biol. 2, 37–53

24. Karg, S. R., Frey, A. D., Ferrara, C., Streich, D. K., Umana, P., and Kallio,P. T. (2009) A small-scale method for the preparation of plant N-linkedglycans from soluble proteins for analysis by MALDI-TOF mass spec-trometry. Plant Physiol. Biochem. 47, 160–166

25. Downs, F., and Pigman, W. (1969) Preparation of glycopeptides from bo-vine submaxillary mucin by chemical degradation. Biochemistry 8,1760–1766

26. Tettamanti, G., and Pigman, W. (1968) Purification and characterizationof bovine and ovine submaxillary mucins. Arch. Biochem. Biophys. 124,41–50

27. Lee, J. J., andKong,M. (2009) Confidence intervals of interaction index forassessing multiple drug interaction. Stat. Biopharm. Res. 1, 4–17

28. Sujatha,M. S., and Balaji, P. V. (2004) Identification of common structuralfeatures of binding sites in galactose-specific proteins. Proteins 55, 44–65

29. Wu, J. H., Singh, T., Herp, A., andWu, A. M. (2006) Carbohydrate recog-nition factors of the lectin domains present in the Ricinus communis toxicprotein (ricin). Biochimie 88, 201–217

30. Byrd, J. C., and Bresalier, R. S. (2004) Mucins and mucin binding proteinsin colorectal cancer. Cancer Metastasis Rev. 23, 77–99

31. Ju, T., Otto, V. I., and Cummings, R. D. (2011) The Tn antigen. Structuralsimplicity and biological complexity. Angew. Chem. Int. Ed Engl. 50,1770–1791

32. Springer, G. F. (1984) T and Tn, general carcinoma autoantigens. Science224, 1198–1206

33. Freire, T., Bay, S., von Mensdorff-Pouilly, S., and Osinaga, E. (2005) Mo-lecular basis of incomplete O-glycan synthesis in MCF-7 breast cancercells. Putative role of MUC6 in Tn antigen expression. Cancer Res. 65,7880–7887

34. Inoue, M., Takahashi, S., Yamashina, I., Kaibori, M., Okumura, T., Ka-miyama, Y., Vichier-Guerre, S., Cantacuzene, D., and Nakada, H. (2001)High density O-glycosylation of the MUC2 tandem repeat unit by N-acetylgalactosaminyltransferase-3 in colonic adenocarcinoma extracts.Cancer Res. 61, 950–956

35. Lloyd, K. O., Burchell, J., Kudryashov, V., Yin, B. W., and Taylor-Papad-imitriou, J. (1996) Comparison of O-linked carbohydrate chains inMUC-1 mucin from normal breast epithelial cell lines and breast carci-noma cell lines. Demonstration of simpler and fewer glycan chains intumor cells. J. Biol. Chem. 271, 33325–33334

36. Abbott, K. L., Nairn, A. V., Hall, E. M., Horton, M. B., McDonald, J. F.,Moremen, K. W., Dinulescu, D. M., and Pierce, M. (2008) Focused gly-comic analysis of the N-linked glycan biosynthetic pathway in ovariancancer. Proteomics. 8, 3210–3220

37. Nan, B. C., Shao, D. M., Chen, H. L., Huang, Y., Gu, J. X., Zhang, Y. B., andWu, Z. G. (1998) Alteration of N-acetylglucosaminyltransferases in pan-creatic carcinoma. Glycoconj. J. 15, 1033–1037

38. Saldova, R., Reuben, J. M., Abd Hamid, U. M., Rudd, P. M., and Cristo-fanilli, M. (2011) Levels of specific serumN-glycans identify breast cancerpatients with higher circulating tumor cell counts. Ann. Oncol. 22,1113–1119

39. Telford, J. E., Doherty, M. A., Tharmalingam, T., and Rudd, P. M. (2011)Discovering new clinical markers in the field of glycomics. Biochem. Soc.Trans. 39, 327–330

40. Vercoutter-Edouart, A. S., Slomianny, M. C., Dekeyzer-Beseme, O.,Haeuw, J. F., and Michalski, J. C. (2008) Glycoproteomics and glycomicsinvestigation of membrane N-glycosylproteins from human colon carci-noma cells. Proteomics 8, 3236–3256

41. Peracaula, R., Tabares, G., Royle, L., Harvey, D. J., Dwek, R. A., Rudd, P.M.,and de Llorens, R. (2003) Altered glycosylation pattern allows the distinc-

Binding Mechanism of Riproximin

OCTOBER 19, 2012 • VOLUME 287 • NUMBER 43 JOURNAL OF BIOLOGICAL CHEMISTRY 35885

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from

tion between prostate-specific antigen (PSA) from normal and tumor or-igins. Glycobiology 13, 457–470

42. Sarrats, A., Saldova, R., Comet, J., O’Donoghue, N., de Llorens, R., Rudd,P. M., and Peracaula, R. (2010) Glycan characterization of PSA 2-DE sub-forms from serum and seminal plasma. OMICS. 14, 465–474

43. Tabares, G., Radcliffe, C. M., Barrabes, S., Ramırez, M., Aleixandre, R. N.,Hoesel, W., Dwek, R. A., Rudd, P. M., Peracaula, R., and de, L. R. (2006)Different glycan structures in prostate-specific antigen from prostate can-cer sera in relation to seminal plasma PSA. Glycobiology 16, 132–145

44. Brockhausen, I. (2006) Mucin-type O-glycans in human colon and breastcancer. Glycodynamics and functions. EMBO Rep. 7, 599–604

45. Kakugawa, Y.,Wada, T., Yamaguchi, K., Yamanami, H., Ouchi, K., Sato, I.,andMiyagi, T. (2002)Up-regulation of plasmamembrane-associated gan-glioside sialidase (Neu3) in human colon cancer and its involvement inapoptosis suppression. Proc. Natl. Acad. Sci. U.S.A. 99, 10718–10723

46. Storr, S. J., Royle, L., Chapman, C. J., Hamid, U. M., Robertson, J. F.,Murray, A., Dwek, R. A., and Rudd, P. M. (2008) The O-Linked glycosyl-ation of secretory/shed MUC1 from an advanced breast cancer patient’sserum. Glycobiology 18, 456–462

47. Li, Q., Anver, M. R., Li, Z., Butcher, D. O., and Gildersleeve, J. C. (2010)GalNAc�1–3Gal, a new prognosticmarker for cervical cancer. Int. J. Can-cer 126, 459–468

48. Hakomori, S. (1984) Tumor-associated carbohydrate antigens.Annu. Rev.Immunol. 2, 103–126

49. Kumar, S. R., Sauter, E. R., Quinn, T. P., andDeutscher, S. L. (2005) Thom-sen-Friedenreich and Tn antigens in nipple fluid. Carbohydrate biomark-ers for breast cancer detection. Clin. Cancer Res. 11, 6868–6871

50. Manimala, J. C., Roach, T. A., Li, Z., and Gildersleeve, J. C. (2006) High-throughput carbohydrate microarray analysis of 24 lectins. Angew. Chem.Int. Ed. Engl. 45, 3607–3610

51. Bhattacharyya, L., and Brewer, C. F. (1988) Binding and precipitation oflectins from Erythrina indica and Ricinus communis (agglutinin I) withsynthetic cluster glycosides. Arch. Biochem. Biophys. 262, 605–608

52. Green, E. D., Adelt, G., Baenziger, J. U., Wilson, S., and Van Halbeek, H.(1988) The asparagine-linked oligosaccharides on bovine fetuin. Struc-tural analysis ofN-glycanase-released oligosaccharides by 500-megahertz1H NMR spectroscopy. J. Biol. Chem. 263, 18253–18268

53. Chandrasekaran, E. V., Davila, M., Nixon, D. W., Goldfarb, M., and Men-dicino, J. (1983) Isolation and structures of the oligosaccharide units ofcarcinoembryonic antigen. J. Biol. Chem. 258, 7213–7222

54. Tsuji, T., and Osawa, T. (1986) Carbohydrate structures of bovine sub-maxillary mucin. Carbohydr. Res. 151, 391–402

55. Li, Q., Rodriguez, L. G., Farnsworth, D. F., and Gildersleeve, J. C. (2010)Effects of hapten density on the induced antibody repertoire. Chembi-ochem 11, 1686–1691

56. Nilsson, B., Norden, N. E., and Svensson, S. (1979) Structural studies onthe carbohydrate portion of fetuin. J. Biol. Chem. 254, 4545–4553

57. Dam, T. K., Gerken, T. A., and Brewer, C. F. (2009) Thermodynamics ofmultivalent carbohydrate-lectin cross-linking interactions. Importance ofentropy in the bind and jump mechanism. Biochemistry 48, 3822–3827

58. De Meyts P. (2008) The insulin receptor. A prototype for dimeric, allos-teric membrane receptors? Trends Biochem. Sci. 33, 376–384

59. Strålfors, P. (2012) Caveolins and caveolae, roles in insulin signaling anddiabetes. Adv. Exp. Med. Biol. 729, 111–126

60. Lattova, E., Tomanek, B., Bartusik, D., and Perreault, H. (2010) N-Gly-comic changes in human breast carcinomaMCF-7 and T-lymphoblastoidcells after treatment with herceptin and herceptin/Lipoplex. J. Proteome.Res. 9, 1533–1540

61. Brain, P., and Cousens, R. (1989) An equation to describe dose-responseswhere there is stimulation of growth at low doses.Weed Res. 29, 93–96

Binding Mechanism of Riproximin

35886 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 43 • OCTOBER 19, 2012

by guest on February 4, 2014http://w

ww

.jbc.org/D

ownloaded from