Conjugation of an anti–transferrin receptor IgG3-avidin ... · Cell images were captured using a Zeiss Axiovert 40 CFL PlasDIC Inverted Microscope using a 20 objective (Mikron Instruments,

Conjugation of an anti–transferrin receptor IgG3-avidinfusion protein with biotinylated saporin results insignificant enhancement of its cytotoxicity againstmalignant hematopoietic cells

Tracy R. Daniels,1 Patrick P. Ng,5 Tracie Delgado,1

Maureen R. Lynch,2 Gary Schiller,2,3

Gustavo Helguera,1 and Manuel L. Penichet1,3,4

1Division of Surgical Oncology, Department of Surgery; 2Divisionof Hematology and Oncology, Department of Medicine; 3JonssonComprehensive Cancer Center; and 4Department of Microbiology,Immunology, and Molecular Genetics, David Geffen School ofMedicine, University of California, Los Angeles, California and5Department of Medicine, Oncology Division, Stanford UniversityMedical Center, Stanford, California

AbstractWe have previously developed an antibody fusion proteincomposed of a mouse/human chimeric IgG3 specific forthe human transferrin receptor genetically fused to avidin(anti-hTfR IgG3-Av) as a universal delivery system forcancer therapy. This fusion protein efficiently deliversbiotinylated FITC into cancer cells via TfR-mediatedendocytosis. In addition, anti-hTfR IgG3-Av alone exhibitsintrinsic cytotoxic activity and interferes with hTfRrecycling, leading to the rapid degradation of the TfR andlethal iron deprivation in certain malignant B-cell lines.We now report on the cytotoxic effects of a conjugatecomposed of anti-hTfR IgG3-Av and biotinylated saporin6 (b-SO6), a toxin derived from the plant Saponariaofficinalis that inhibits protein synthesis. Conjugation ofanti-hTfR IgG3-Av with b-SO6 enhances the cytotoxiceffect of the fusion protein in sensitive cells and alsoovercomes the resistance of malignant cells that show lowsensitivity to the fusion protein alone. Our results show

for the first time that loading anti-hTfR IgG3-Av with abiotinylated toxin enhances the cytotoxicity of the fusionprotein alone. These results suggest that anti-hTfR IgG3-Av has great potential as a therapeutic agent for a widerange of applications due to its intrinsic cytotoxic activityplus its ability to deliver biotinylated molecules into cancercells. [Mol Cancer Ther 2007;6(11):2995–3008]

IntroductionThe human transferrin receptor (hTfR) is a type IItransmembrane glycoprotein involved in cellular ironuptake, iron homeostasis, and the regulation of cell growthand proliferation (reviewed in ref. 1). The hTfR is anattractive target for the delivery of cytotoxic agents dueto its increased expression in malignant compared withnormal cells, its extracellular accessibility, and its constitu-tive endocytosis cycle (reviewed in refs. 1, 2). Our grouppreviously developed an antibody-avidin fusion proteintargeting the hTfR to serve as a universal vector for thedelivery of biotinylated therapeutic agents into cancer cells(3, 4). Anti-hTfR IgG3-Av consists of a mouse/humanchimeric IgG3 genetically fused to chicken avidin at thecarboxy terminus of the CH3 domain of the heavy chain(g3; Fig. 1A). Fast protein liquid chromatography analysisof the fusion protein suggests that the protein is a dimerin solution (ref. 3; Fig. 1B), a structure that is expected dueto the noncovalent tetrameric nature of avidin (5). We havepreviously shown that this antibody-avidin fusion proteinefficiently delivers a biotinylated fluorochrome (FITC) intocancer cells by receptor-mediated endocytosis (3). Surpris-ingly, anti-hTfR IgG3-Av was found to exhibit intrinsiccytotoxic activity in certain malignant B-cell lines andprimary myeloma cells (4). The antibody-avidin fusionprotein interferes with the normal cycling pathway of thehTfR and directs the hTfR into a LAMP-1–positivecompartment where it seems to be degraded. Therefore,in anti-hTfR IgG3-Av–treated cells, surface hTfR expres-sion decreases, resulting in lethal iron deprivation andapoptosis (4). In more recent studies, we have shown thatanti-hTfR IgG3-Av specifically targets cells expressinghTfR1, but not hTfR2, a second receptor that shares 45%identity and 66% similarity with the extracellular domain ofhTfR1 (1).6

The cytotoxic properties of anti-hTfR IgG3-Av canpotentially be enhanced by its conjugation to biotinylatedcytotoxic agents, including bacterial or plant toxins. Among

Received 5/14/07; revised 9/12/07; accepted 9/24/07.

Grant support: NIH/National Cancer Institute grants CA86915 andCA107023, NIH/National Cancer Institute research supplementCA107023-02S1, and the 2004 Brian D. Novis International MyelomaFoundation Senior Grant Award. The Minority Access to ResearchCareers, Center for Academic and Research Excellence, and SummerPrograms in Undergraduate Research programs at University of Californiaat Los Angeles provided generous support and funding.

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

Note: T.R. Daniels and P.P. Ng contributed equally to this work.

Requests for reprints: Manuel L. Penichet, Division of Surgical Oncology,Department of Surgery, University of California at Los Angeles,10833 Le Conte Avenue, CHS 54-140, Box 951782, Los Angeles,CA 90095-1782. Phone: 310-825-1304; Fax: 310-825-7575.E-mail: [email protected]

Copyright C 2007 American Association for Cancer Research.

doi:10.1158/1535-7163.MCT-07-0330 6 Rodriguez et al., J Controlled Release, in press.

2995

Mol Cancer Ther 2007;6(11). November 2007

Research. on January 13, 2020. © 2007 American Association for Cancermct.aacrjournals.org Downloaded from

http://mct.aacrjournals.org/

them, a promising compound is the plant toxin saporin,produced by Saponaria officinalis . Saporin is a member ofthe ribosome-inactivating protein family of toxins(reviewed in ref. 6). The combination of saporin withvarious delivery systems has shown, in vitro and in vivo,that saporin immunotoxins [including genetic fusionproteins (7–9) and chemical conjugates (8, 10–13)] areeffective antitumor agents. In fact, the potential anticancerbenefits of saporin immunotoxins has been shown inphase I clinical trials for the treatment of advancedrefractory Hodgkin’s disease (anti-CD30 murine monoclo-nal antibody chemically conjugated to saporin; ref. 14) andfor advanced-stage B-cell lymphoma (bispecific antibodiesspecific for saporin and CD22; ref. 15). Immunotoxinstargeting the hTfR, via chemical conjugation with trans-ferrin (16) or an anti-hTfR antibody (17), have also shownin vitro cytotoxic effects in malignant cells.

Although saporin immunotoxins have been successfulantitumor agents, the immunotoxins described above havesome disadvantages. Chemical conjugates have manydrawbacks, including a lack of homogeneity, whereas theuse of fusion proteins requires that a different protein becreated for each application. In addition, there can be adecrease or loss of activity of one or both partners. The anti-hTfR IgG3-Av fusion protein overcomes these limitationsand can be used as a universal delivery vector, whicheliminates the need to make specific constructs for eachapplication. Anti-hTfR IgG3-Av also has the advantage ofhaving multiple modes of antitumor activity (intrinsicproapoptotic, delivery of therapeutic agents, and possibleFc effector functions). We now report the effective deliveryinto malignant cells of an active toxin, biotinylated saporin6 (b-SO6), by anti-hTfR IgG3-Av. This report provides theinitial proof-of-principle that conjugation of anti-hTfRIgG3-Av with a biotinylated toxin enhances its cytotoxicityand also overcomes the resistance in malignant cells thatshow low sensitivity to anti-hTfR IgG3-Av. The mechanismof the cytotoxicity induced by the anti-hTfR IgG3-Av/b-SO6complex is also addressed.

Materials andMethodsAntibodies, Antibody Fusion Proteins, and Toxin

ConjugatesThe mouse/human chimeric antibodies specific for the

hapten dansyl [5-dimethylamino naphthalene-1-sulfonylchloride (anti-DNS IgG3); refs. 18, 19], the hTfR (anti-hTfRIgG3; ref. 3), and the antibody avidin fusion proteins [anti-DNS IgG3-Av (negative isotype control) and anti-hTfRIgG3-Av] have been described previously (3, 20). Briefly,chicken avidin was genetically fused to the CH3 domain ofanti-DNS IgG3. To construct the anti-hTfR IgG3-Av fusionprotein, the heavy and light chain variable regions of anti-DNS IgG3-Av were substituted for the variable regions ofthe murine monoclonal anti-human TfR IgG1 antibody128.1 (3, 21). Anti-DNS IgG3-Av and anti-hTfR IgG3-Avwere expressed in murine myeloma cells, which were thenexpanded in roller bottles (Fisher Scientific). The antibodyfusion proteins were purified from cell culture super-

natants using affinity chromatography (3, 22). All fractionswere tested for fusion protein integrity and purity in 5%phosphate gels under nonreducing conditions. Fractionscontaining purified protein were pooled and dialyzed intoprotein buffer [150 mmol/L NaCl, 50 mmol/L Tris-HCl(pH 7.8)]. Protein concentrations were determined bybicinchoninic acid–based protein assay (BCA ProteinAssay, Pierce Biotechnology, Inc.; refs. 3, 22). Antibodyfusion proteins were then stored at �80jC in snap-frozenaliquots.

Saporin belongs to a gene family that consists of manyisoforms. SO6 is isolated from the seed of the S. officinalisplant and has been shown to contain higher catalyticactivity than another isoform, saporin 5 (23). The customconjugate b-SO6 was purchased from Advanced TargetingSystems, which was purified from the seed of theplant as previously described (24, 25) and conjugated toBiotin-HPDP (Pierce Biotechnology, Inc.) according tomanufacturer’s instructions. This custom conjugate wasmanufactured with 1.1 moles of biotin per 1 mole ofsaporin to avoid the formation of aggregates whencomplexed to the avidin fusion proteins. Conjugation ofb-SO6 with the antibody-avidin fusion proteins wascarried out on ice for 30 min at a 1:1 molar ratio beforethe addition of cell culture medium.

Cell Lines and Primary CellsIM-9 (an EBV-transformed lymphoblastoid cell line

isolated from the peripheral blood of a patient with

Figure 1. Schematic diagram of anti-hTfR IgG3-Av alone and com-plexed to b-SO6. A, the antibody avidin fusion protein is composed of amouse/human chimeric IgG3 molecule genetically fused to avidin at itscarboxy terminus (CH3 domain). The extended hinge region of human IgG3provides flexibility to the molecule. B, the suggested dimeric structureof anti-hTfR IgG3-Av conjugated to b-SO6. Each avidin molecule istheoretically capable of binding one biotin; therefore, the fusion protein iscapable of binding up to four b-SO6 molecules.

Cytotoxicity of Anti-hTfR IgG3-Av/b-Saporin2996




multiple myeloma) and U266 (a lymphoblast cell lineisolated from the peripheral blood of a patient withmyeloma/plasmacytoma) were purchased from theAmerican Type Culture Collection. The cell lines weremaintained in RPMI 1640 (Invitrogen Corporation) supple-mented with 10% heat-inactivated fetal bovine serum(Atlanta Biologicals, Inc.) and grown in 5% CO2 at 37jC.Cell images were captured using a Zeiss Axiovert 40 CFLPlasDIC Inverted Microscope using a �20 objective(Mikron Instruments, Inc.) and a Canon PowerShot A620digital camera (Mikron Instruments).

Bone marrow aspirates from either a patient withmultiple myeloma or with plasma-cell leukemia, a moreaggressive variant of multiple myeloma (26), were obtainedwith informed consent following the standards of theinstitutional review board of University of California atLos Angeles. CD138+ cells were collected as describedpreviously using the EasySep Human CD138+ Selection kit(StemCell Technologies, Inc.; ref. 4).

Proliferation AssayThe effects of the various treatments on cell proliferation

were determined by the [3H]thymidine incorporation assayas previously described (4). Briefly, IM-9 or U266 wereseeded at a density of 104 cells per well (total volume perwell, 100 AL) in 96-well tissue culture plates the daytreatment began. Cells were incubated with the antibodies,antibody fusion proteins, and conjugates for the indicatedtimes in triplicate. [3H]thymidine (0.5 ACi per well, MPBiomedicals) was incubated with the cells for the final16 h of the treatment period for the cell lines and for thefinal 48 h of treatment for primary myeloma cells. Cellswere then harvested onto glass fiber filters (PrintedFiltermat A, Perkin-Elmer Life and Analytical Science,Inc.) by using a 11050 Micro Cell Harvester (Skatron).Betaplate Scint (Perkin-Elmer Life and Analytical Science)scintillation fluid was then added to a sample bag (Perkin-Elmer Life and Analytical Science) containing the fiber filterand radioactivity was counted in a 1205 Betaplate LiquidScintillation Counter (Perkin-Elmer Life and AnalyticalScience). For competition studies, cells were preincubatedwith excess anti-hTfR IgG3-Av fusion protein for 1 h beforethe addition of the toxin conjugate. For studies with ironsupplementation, 25 Amol/L ferric ammonium citrate(FAC; Sigma-Aldrich Co.) was incubated simultaneouslywith the antibodies and antibody fusion proteins for theduration of the treatment. Cells were then harvested andradioactivity was determined as described above.

Determination of Complex StabilityConjugation of b-SO6 with anti-hTfR IgG3-Av was

carried out on ice for 30 min at a 1:1 molar ratio. Thecomplexes were incubated in RPMI 1640 (containing 10%fetal bovine serum) in duplicates for 0, 1, 12, 24, 48, or144 h (6 days) at 37jC in 5% CO2. Each duplicate wasthen tested in triplicate (six total replicates) in 96-wellplates by incubation with IM-9 or U266 cells for 48 h. Theability of the complexes to inhibit proliferation wasmonitored by the [3H]thymidine incorporation assay asdescribed above.

Apoptosis AssayIM-9 and U266 cells (105 per well) were treated in 48-well

tissue culture plates (total volume per well, 1 mL) with10 nmol/L of the various antibodies, antibody fusionproteins, and toxin conjugates. At 24, 48, and 72 h, the cellswere washed in cold PBS and the levels of apoptosis weredetermined using the Vybrant Apoptosis Assay kit 2(Invitrogen) following procedures suggested by themanufacturer. This kit consists of the Annexin V AlexaFluor 488 conjugate and propidium iodide stain. Sampleswere analyzed on a BD-LSR Analytic Flow Cytometer (BDBiosciences). Ten thousand events were recorded for eachflow cytometry measurement. Data were analyzed usingthe WinMDI 2.8 software (The Scripps Research Institute).Inhibition of Protein SynthesisThe effect of the b-SO6 toxin on protein synthesis was

assessed by determining the amount of [3H]leucineincorporated into newly synthesized proteins. Cells wereseeded at a density of 104 per well (total volume per well,100 AL) in 96-well tissue culture plates the day treatmentsbegan. Cells were incubated with 1 nmol/L of the variousantibodies, antibody avidin fusion proteins, and toxinconjugates for the indicated times. [3H]leucine (1 ACi/well;MP Biomedicals) was added to cells for the final 16 h oftreatment. Cells were then harvested as described above.

CaspaseActivityAssaysIM-9 and U266 cells were seeded at a density of 104 per

well (total volume per well, 200 AL) of black, clear-bottomed 96-well tissue culture plates (Fisher Scientific).Cells were treated with 10 nmol/L of the various anti-bodies, antibody fusion proteins, and toxin conjugates forthe indicated time points. The plates were then centrifugedat 250 � g for 10 min in an Allegra X-15R Centrifuge(Beckman Coulter, Inc.). One hundred microliters of thesupernatant were carefully removed and the remainingcells were assayed for caspase activity using fluorogenicsubstrates in a one-step assay as described previously(4, 27). At each time point, 50 AL of one-step assay buffer[1.5% NP40, 0.3% CHAPS, 30% sucrose, 30 mmol/L MgCl2,150 mmol/L KCl, 450 mmol/L NaCl, 150 mmol/L HEPES,1.2 mmol/L EGTA, 30 mmol/L DTT, 3 mmol/L phenyl-methylsulfonyl fluoride (pH 7.4)] containing 150 Amol/Lof the fluorogenic substrate being tested were added toeach well and incubated for 1 h at 37jC in 5% CO2. Fluoro-genic substrates specific for caspase-2 (Ac-VDVAD-AMC),caspase-9 (Ac-LEHD-AMC), caspase-8 (Ac-IETD-AMC),caspase-3 and caspase-7 (Ac-DEVD-AMC), and caspase-3(Ac-DMQD-AMC; all from Axxora Life Sciences, Inc.) wereused. The plate was then read at excitation and emissionwavelengths of 380 and 460 nm, respectively, in a DTX880Multimode Detector (Beckman Coulter). Background fluo-rescence, measured in wells containing only mediumand one-step assay buffer with substrate (no cells), wassubtracted from each sample. To calculate the fold increasein activation of each caspase, the average relative fluores-cence intensity of the treated wells was divided by theaverage relative fluorescence intensity of the negativecontrol wells. For caspase inhibition studies, cells were

Molecular Cancer Therapeutics 2997








pretreated with 100 Amol/L Z-VAD-FMK methyl ester(BIOMOL International, L.P.) for 1 h before the additionof anti-hTfR IgG3-Av or anti-hTfR IgG3-Av/b-SO6. Cellswere treated with 100 ng/mL of anti-Fas (MBL Interna-tional Corporation) as a positive control for caspase-dependent cell death (28).

Human Colony-Forming AssayFrozen human bone marrow mononuclear cells were

purchased from StemCell Technologies and thawed follow-ing the manufacturer’s instructions. The mononuclear cellswere treated with buffer alone (negative control), 1 nmol/Lanti-hTFR IgG3-Av, or 1 nmol/L anti-hTFR IgG3-Av/b-SO6for 1 h in Iscove’s modified Dulbecco’s medium containing10% fetal bovine serum in 5% CO2 at 37jC. The cells werethen washed thrice in Iscove’s modified Dulbecco’s medi-um containing 2% fetal bovine serum (StemCell Technolo-gies) and plated in MethoCult medium as instructed bythe manufacturer. Cells were plated in quadruplicate (at adensity of 2 � 104 per 35-mm dish) in MethoCult GF H4434(‘‘Complete’’ Methylcellulose Medium with RecombinantCytokines and Erythropoietin). The cells were then incu-bated for 14 days at in 5% CO2 at 37jC. The numberof colony-forming unit–erythroid (mature erythroid pro-genitors), burst-forming unit–erythroid (more primitiveprogenitor than colony-forming unit –erythroid), andcolony-forming unit–granulocyte/macrophage was deter-mined using an Olympus CK2 inverted microscope andthe criteria defined by StemCell Technologies for eachcolony type. The average number of colonies was deter-mined from quadruplicate samples.

Flow CytometryCD34+ cells were purchased from StemCell Technologies.

These cells showed a 97% purity as reported on the technicaldata sheet. The CD34+ cells were thawed according to themanufacturer’s instructions. IM-9, U266, and CD34+ cells(2 � 105) were incubated for 15 min on ice with eitherphycoerythrin-conjugated mouse IgG2a n isotype controlor phycoerythrin-conjugated mouse anti-human CD71(hTfR) monoclonal antibodies (BD Biosciences). Cells werethen washed and analyzed on a FACScan flow cytometer(BD Biosciences). Data were analyzed using the CELLQuestsoftware (BD Biosciences).

Statistical AnalysisAll statistical analyses were done using Microsoft Excel

2000 SR-1 Standard. Significant differences were calculatedusing the Student’s t test (unpaired samples, two-tailed,unequal variance). P values <0.05 were considered to besignificant.

ResultsConjugationofAnti-hTfRIgG3-Avwithb-SO6Results

in Significant Enhancement of the Cytotoxic Effects ofthe Fusion Protein Alone

We previously showed that a panel of malignanthematopoietic cell lines shows various sensitivities toanti-hTfR IgG3-Av (4). Therefore, we chose one highlysensitive cell line (IM-9) and one cell line with lowsensitivity (U266) to use in our studies for targeted deliveryof b-SO6. As expected, the parental antibody anti-hTfRIgG3 was slightly inhibitory to the growth of IM-9 cells,whereas anti-hTfR IgG3-Av elicited significant growthinhibition to IM-9 but not to U266 cells (Fig. 2A; left panels).When cells were treated with anti-hTfR IgG3-Av/b-SO6(Fig. 2A), as low as 0.1 nmol/L of the complex stronglyinhibited the growth of both cell lines. The concentrationthat inhibited 90% of proliferation (IC90) was decreasedfrom 1 nmol/L for the fusion protein alone to 0.1 nmol/Lfor the complex in IM-9 cells. The complex overcame theresistance to the fusion protein in U266 cells with an IC90

of 1 nmol/L. The growth of cells treated with b-SO6alone, anti-DNS IgG3, anti-DNS IgG3-Av, or anti-DNSIgG3-Av/b-SO6 was not inhibited. In the time coursestudy, strong antiproliferative effects of anti-hTfR IgG3-Av/b-SO6 was observed at 24 h in IM-9 cells comparedwith 48 h in U266 cells (Fig. 2A, right panels). When primarymyeloma cells isolated from bone marrow aspirates oftwo patients were treated with anti-hTfR IgG3-Av alone oranti-hTfR IgG3-Av/b-SO6, the conjugate showed highergrowth-inhibitory effects than the antibody-avidin fusionprotein alone (Fig. 2B).

Morphologic studies were carried out to determine ifanti-hTfR IgG3-Av induced cell death or simply blockedproliferation. In IM-9 cells, anti-hTfR IgG3 showed thatsome cell death did occur (Fig. 2C, top). Cells treated withanti-hTfR IgG3-Av showed even more cell death thanthose treated with the parental antibody as expected fromour previous studies. U266 cells treated with anti-hTfRIgG3, anti-hTfR IgG3-Av, or anti-DNS IgG3-Av/b-SO6 didnot exhibit morphologic changes compared with cellstreated with buffer alone (Fig. 2C, bottom). However, inboth IM-9 and U266 cells treated with anti-hTfR IgG3-Av/b-SO6, the cell number and the morphology of theremaining cells show that the treatment resulted in themassive induction of cell death. These results indicate thatthe effects induced by anti-hTfR IgG3-Av/b-SO6 are notonly due to a block in proliferation, but also due to theinduction of cell death.

Figure 2. Antiproliferative effects of b-SO6 delivered by anti-hTfR IgG3-Av. A, IM-9 (top panels ) and U266 (bottom panels ) cells were treated with theindicated concentrations for 72 h (left panels) or 1 nmol/L (right panels ) of b-SO6, antibodies, antibody fusion proteins, or toxin conjugates for theindicated times. Proliferation was monitored by the [3H]thymidine incorporation assay. The rate of proliferation of treated cells is reported as a percentageof [3H]thymidine incorporated into buffer alone– treated cells. The results shown are in logarithmic scale and are representative of three independentexperiments. B, CD138+ cells isolated from bone marrow aspirates of two patients diagnosed with plasma cell leukemia (PCL ) and multiple myeloma(MM ) were treated with 100 nmol/L of anti-hTfR IgG3-Av or anti-hTfR IgG3-Av/b-SO6 for 72 h. Proliferation was monitored as indicated above. C, IM-9(top ) and U266 (bottom ) cells were treated with 10 nmol/L of antibody, antibody fusion protein, or toxin conjugates for 72 h. Representative differentialinterference contrast images from three independent experiments. D, the anti-hTfR IgG3-Av/b-SO6 complexes were incubated in RPMI at 37jC in 5% CO2

for the indicated times. The stability of the complexes was then tested by monitoring the inhibition of proliferation on IM-9 (left ) or U266 (right ) cells. Afterthe addition of the complexes, the cells were cultured for 48 h. Control cells were incubated with buffer alone for the 48-h incubation period. Proliferationwas monitored by the [3H]thymidine incorporation assay.





To study the stability of anti-hTfR IgG3-Av/b-SO6, thecomplex was incubated in RPMI medium containing 10%fetal bovine serum for various times at 37jC. The anti-proliferative effect of the complex was then monitored byincubation with IM-9 (Fig. 2D, left) or U266 (Fig. 2D, right)cells for 48 h. Anti-hTfR IgG3-Av/b-SO6 is highly stable asevidenced by its ability to block proliferation in both IM-9and U266 cells even after incubation in medium for anextended period. A slight loss of activity (10% in U266 cellsand 16% in IM-9 cells) was only observed after the 6-dayincubation period.

To further study the cell death induced by the anti-hTfRIgG3-Av/b-SO6 complex, we examined the change in cellmembrane composition and disruption by monitoringAnnexin V and propidium iodide staining. Apoptotic cellsare reported as both early (Annexin V+/propidium iodide�)and late apoptotic cells (Annexin V+/propidium iodide+).In IM-9 cells, anti-hTfR IgG3 induced apoptosis only at lowlevels (Fig. 3, left). Anti-hTfR IgG3-Av induced high levels

of apoptosis at 48 h with a total Annexin V+ of 60.5% thatincreased to 78.1% at 72 h. However, higher levels of apop-tosis were induced by anti-hTfR IgG3-Av/b-SO6. At 24 h,the total Annexin V+ was 32%, which increased to 78.6%at 48 h and further increased to 92% at 72 h. In U266 cells,only anti-hTfR IgG3-Av/b-SO6 showed strong induction ofapoptosis with a total of 39.7% Annexin V+ cells at 24 h,57.9% at 48 h, and 71.9% at 72 h (Fig. 3, right). It should alsobe noted that in both cell lines treated with anti-hTfR IgG3-Av/b-SO6, the population of cells in the lower left quadrantshows a slight shift to the right, suggesting that all of thecells are becoming Annexin V+. Anti-DNS IgG3-Av/b-SO6only induced a low level of apoptosis in both cell lines overtime. Anti-DNS IgG3, anti-DNS IgG3-Av, and b-SO6 alonedid not induce apoptosis in either cell line (data not shown).

Targeting of the hTfR Is Required forAnti-hTfR IgG3-Av/b-SO6^ InducedToxicity

The fact that only marginal cytotoxic effects wereobserved with a nontargeted anti-DNS IgG3-Av/b-SO6

Figure 3. Anti-hTfR IgG3-Av/b-SO6 induces apoptosis. IM-9 (left panels ) and U266 (right panels ) cells were incubated with 10 nmol/L of theantibodies, antibody avidin fusion proteins, and toxin conjugates for 24, 48, or 72 h. Cells were then washed, stained with Annexin V–Alexa Fluor 488and propidium iodide, and analyzed by flow cytometry. The percentage of cells located in each quadrant is shown in the corner.





complex strongly suggests that targeting of the hTfR isrequired for the toxic effects of the anti-hTfR IgG3-Av/b-SO6 complex. To confirm this, competition assays wereperformed in U266 cells, because IM-9 cells are sensitive toboth anti-hTfR IgG3 and anti-hTfR IgG3-Av treatmentalone. Pretreatment of U266 cells with free anti-hTfR IgG3-Av blocked the cytotoxicity induced by anti-hTfR IgG3-Av/b-SO6 (Fig. 4). Both 10-fold (P < 0.001) and 100-fold(P < 0.001) excess anti-hTfR IgG3-Av significantly blockedthe cytotoxic effects of the complex. One thousand–foldexcess anti-hTfR IgG3-Av almost completely blocked thecytotoxicity induced by 0.1 nmol/L anti-hTfR IgG3-Av/

b-SO6 (P < 0.01). These results are consistent withthe internalization of the anti-hTfR IgG3-Av/b-SO6 com-plex through the interaction of the antibody with the cellsurface TfR.

Treatment with Anti-hTfR IgG3-Av/b-SO6 Results inthe Inhibition of Protein Synthesis

Because the mechanism of cytotoxicity of saporin isthrough the inactivation of the ribosome, we evaluated theability of delivered b-SO6 to inhibit protein synthesis inboth IM-9 and U266 cells. The kinetics of protein synthesisinhibition were very similar to the kinetics of the anti-proliferative effects induced by treatment with the anti-hTfR IgG3-Av/b-SO6 complex. In IM-9 cells treated with1 nmol/L anti-hTfR IgG3-Av/b-SO6, protein synthesiswas completely blocked by 48 h (99% blockage, P < 0.001;Fig. 5A). Inhibition of protein synthesis occurred moreslowly in U266 cells (Fig. 5B); however, significantinhibition was also observed by 48 h. No inhibition ofprotein synthesis was observed in cells treated with thecontrol treatments. These results indicate that some or all ofthe b-SO6 delivered by anti-hTfR IgG3-Av remains activeafter its internalization.

The Cytotoxicity Induced by Anti-hTfR IgG3-Av/b-SO6 Is Not Due to Iron Deprivation

The cytotoxicity of anti-hTfR IgG3-Av in two highlysensitive cell lines, ARH-77 and IM-9, was previouslyshown to be due to iron deprivation and could be blockedby the addition of FAC (4). To determine whether irondeprivation plays a role in anti-hTfR IgG3-Av/b-SO6–induced cell death, we coincubated the complex with25 Amol/L FAC. As expected, FAC significantly blockedthe cell death induced by anti-hTfR IgG3-Av alone in IM-9cells (Fig. 6A). However, FAC was unable to inhibit theantiproliferative effects induced by the anti-hTfR IgG3-Av/b-SO6 complex in IM-9 or U266 (Fig. 6), indicating that thecytotoxicity of the complex is mediated by b-SO6.

Figure 4. The antiproliferative effects of anti-hTfR IgG3-Av/b-SO6 areblocked by the addition of excess free fusion protein. U266 cells werepreincubated for 1 h with 10-, 100-, or 1,000-fold excess anti-hTfR IgG3-Av then treated with either 1 or 0.1 nmol/L of the toxin conjugate anti-hTfR IgG3-Av/b-SO6 for 72 h. The antiproliferative effect was monitoredusing the [3H]thymidine incorporation assay. Data are presented as thepercentage of the mean of triplicate samples of [3H]thymidine incorporatedinto buffer alone–treated cells. Data are representative of two indepen-dent experiments. *, P < 0.001.

Figure 5. Protein synthesis is inhibited by saporin delivered to malignant B-cell lines by anti-hTfR IgG3-Av. IM-9 (A) and U266 (B) cells were treated with1 nmol/L of b-SO6, antibodies, antibody avidin fusion proteins, or toxin conjugates for the indicated times. [3H]leucine was added to the culture before cellswere harvested, and the level of [3H]leucine incorporation was measured as described in Materials and Methods. Points, mean of triplicate samplesexpressed as the percentage of the mean of triplicate samples of [3H]leucine incorporated into buffer alone–treated cells. The result shown isrepresentative of two independent experiments.





Anti-hTfR IgG3-Av/b-SO6 Activates Caspases but IsNot Caspase Dependent

We monitored the activation of several caspases todetermine the apoptotic pathway that was induced bythe anti-hTfR IgG3-Av/b-SO6 complex. Caspase-8 is a keyinitiator caspase for the death receptor (extrinsic) pathwayof apoptosis, whereas caspase 9 is the initiator of themitochondrial (intrinsic) pathway (29). Caspase-3 is a keyeffector caspase and is activated in both the intrinsic andextrinsic apoptotic pathways. Caspase-2 is also an initiatorcaspase and is activated in response to UV irradiation andDNA damage (29). In U266 cells, treatment with anti-hTfRIgG3-Av alone did not result in a significant increase incaspase activation (Fig. 7A, top left), as expected becausethese cells show resistance to the cytotoxic effects of the

fusion protein (Figs. 2 and 3; ref. 4). An increase inactivation of caspase-2, caspase-3, caspase-8, and caspase-9was observed at late time points in IM-9 cells treated withanti-hTfR IgG3-Av (Fig. 7A, top right), also confirmingprevious results that showed simultaneous activation ofcaspase-8, caspase-9, and caspase-3 in the sensitive ARH-77cell line (4). Anti-DNS IgG3-Av/b-SO6 induced low levels

Figure 7. Caspase activation in cells treated with anti-hTfR IgG3-Av/b-SO6. A, U266 and IM-9 cells were treated with 10 nmol/L anti-hTfR IgG3-Av alone, anti-DNS IgG3-Av/b-SO6, or anti-hTfR IgG3-Av/b-SO6, andincubated for various times. The cells were then lysed and caspaseactivation was determined using Ac-IETD-AMC (caspase-8 and caspase-10), Ac-LEHD-AMC (capase-9), Ac-DMQD-AMC (caspase-3), or Ac-VDVAD-AMC (caspase-2) fluorogenic substrates. Points, mean oftriplicate samples; representative of two independent trials. B, IM-9 andU266 cells were pretreated for 1 h with 100 Amol/L of the pan-caspaseinhibitor Z-VAD-FMK. Cells were then treated with 1 nmol/L anti-hTfRIgG3-Av/b-SO6 or 100 ng/mL of anti-Fas antibody as a positive controlfor 24, 48, or 72 h. Cell proliferation was then monitored using the[3H]thymidine incorporation assay. Points, mean of triplicate samples;representative of two independent trials.

Figure 6. The growth inhibition induced by anti-hTfR IgG3-Av/b-SO6 isnot due to iron deprivation. IM-9 (A) and U266 (B) cells were treated with1 nmol/L antibody, antibody avidin fusion protein, or toxin conjugates withor without 25 Amol/L FAC for 48 h. Growth inhibition was monitored usingthe [3H]thymidine incorporation assay. Columns, mean of triplicatesamples expressed as the percentage of the mean of triplicate samplesof [3H]thymidine incorporated into buffer alone–treated cells. The resultshown is representative of two independent experiments. *, P < 0.001.





of caspase activation at late time points in both cell lines(Fig. 7A, middle), which is consistent with the low level ofapoptotic induction observed in IM-9 and U266 at late timepoints (Fig. 3). High levels of caspase-2 and caspase-3 areobserved in both cells lines treated with anti-hTfR IgG3-Av/b-SO6 (Fig. 7A, bottom). This activation of caspase-2and caspase-3 occurs before, and at higher levels, thancaspase-8 and caspase-9. These results suggest that in IM-9cells treated with anti-hTfR IgG3-Av as well as in both IM-9and U266 cells treated with anti-hTfR IgG3-Av/b-SO6, analternate apoptotic pathway is induced in which caspase-2plays a key role.

To determine if caspase activation is a required event inanti-hTfR IgG3-Av/b-SO6–induced cell death, we treatedIM-9 and U266 cells with the pan caspase inhibitor Z-VAD-FMK for 1 h before the addition of the fusion protein aloneor the anti-hTfR IgG3-Av/b-SO6 complexes. Cells werealso treated with an anti-Fas antibody (with and withoutZ-VAD-FMK) as a positive control for caspase-dependentcell death (28). U266 cells are resistant to Fas-mediated celldeath (30), whereas IM-9 cells are sensitive (31). In bothIM-9 and U266 cells, Z-VAD-FMK could not block theinhibition of proliferation induced by anti-hTfR IgG3-Av/b-SO6 (Fig. 7B), although it completely blocked caspaseactivation in these cells (data not shown). However, inIM-9 cells, Z-VAD-FMK was able to restore proliferation inanti-Fas–treated cells. These results indicate that althoughcaspases are activated by anti-hTfR IgG3-Av/b-SO6 treat-ment, cytotoxicity occurs, at least in part, through acaspase-independent process.

Toxicity of Anti-hTfR IgG3-Av and Anti-hTfRIgG3-Av/b-SO6 to Hematopoietic Progenitors

Because cytotoxicity of both the anti-hTfR IgG3-Av andanti-hTfR IgG3-Av/b-SO6 are observed in malignanthematopoietic cells, normal hematopoietic cells (CD34+)may also be vulnerable to these cytotoxic effects. However,a limitation in studying the toxic effects on these progenitorcells is that it is difficult to culture these cells for extendedperiods in liquid medium. Therefore, it is preferred to testthe toxicity to CD34+ cells by the human progenitor colony-forming assay, which allows the cells to be cultured in asemisolid medium containing the required growth factorsand cytokines that allow for colony formation. CD34+ cellsfrom healthy donors were incubated with 1 nmol/L anti-hTfR IgG3-Av alone or complexed to b-SO6 for 1 h in liquid

medium. The CD34+ cells were then plated in methylcellu-lose medium and the number of colony-forming unit–erythroid, burst-forming unit–erythroid, and colony-formingunit–granulocyte/macrophage colonies was determinedafter 14 days. It is important to note that in IM-9 andU266 cells, treatment with 1 nmol/L anti-hTfR IgG3-Av/b-SO6 for 1 h followed by replacement of cell culturemedium and further culture for 48 h results in an inhibitionof proliferation similar to that of continuous 48 h treatmentwith anti-hTfR IgG3-Av/b-SO6 (data not shown). Anti-hTfR IgG3-Av alone had no effect on colony formation(Table 1). However, toxicity to all colony types wasobserved with anti-hTfR IgG3-Av/b-SO6 treatment (Table 1).Although colony formation was inhibited, formation wasnot completely blocked.

Table 1. Toxicity of 1 nmol/L anti-hTfR IgG3-Av alone andcomplexed to b-SO6 on human hematopoietic progenitor cells

CFU-E BFU-E CFU-GM

Buffer alone 21 F 7 19 F 4 48 F 4Anti-hTfR IgG3-Av 17 F 2 21 F 6 43 F 4Anti-hTfR IgG3-Av/b-SO6 2 F 1 3 F 1 12 F 4

NOTE: Each value represents the mean number of colonies F the SD ofquadruplicate samples.

Abbreviations: CFU-E, colony-forming unit – erythroid; BFU-E, burst-form-ing unit – erythroid; CFU-GM, colony-forming unit – granulocyte/macro-phage.

Figure 8. Cell surface expression of the TfR. IM-9 (top ), U266(middle), and CD34+ cells (bottom ) were incubated with an anti–hTfR-phycoerythrin (thick lines ) or an isotype-phycoerythrin control antibody(thin lines ) and analyzed by flow cytometry. Data are representative oftwo independent experiments.





Cell Surface Expression of the hTfRBecause we observed some toxicity in hematopoietic

progenitor cells treated with anti-hTfR IgG3-Av/b-SO6, wedecided to evaluate the level of hTfR expression on thesurface of CD34+ cells as well as IM-9 and U266 cells. Thisanalysis was conducted simultaneously in all three celltypes by flow cytometry. Both IM-9 and U266 cells showhigh hTfR expression (Fig. 8). CD34+ cells are a heteroge-neous population in which the majority of cells show hTfRexpression (Fig. 8). However, some cells show low or noexpression of the hTfR as evidenced by significant overlapwith cells stained with a control antibody.

DiscussionIn the present study, we provide the initial proof-of-principle that anti-hTfR IgG3-Av has dual functionalactivity due to its intrinsic cytotoxic activity and its abilityto deliver active biotinylated agents into targeted cells.Conjugation of anti-hTfR IgG3-Av with b-SO6 enhancesthe cytotoxic effects of anti-hTfR IgG3-Av in sensitive cells(IM-9) and can also overcome the resistance observed incells that show low sensitivity to the effects of anti-hTfRIgG3-Av alone (U266). Flow cytometry data showed thatboth IM-9 and U266 express high levels of surface TfR. IM-9expresses higher levels of the hTfR compared with U266cells; however, sensitivity to anti-hTfR IgG3-Av is notnecessarily dependent on hTfR expression levels. We havepreviously tested a panel of cell lines and hTfR expressiondoes not always correlate with sensitivity to anti-hTfRIgG3-Av.7 Our preliminary analysis of myeloma cells,isolated from the bone marrow aspirates of two patients,showed that anti-hTfR IgG3/b-SO6 also enhanced thecytotoxicity of the fusion protein alone in these primarytumor isolates. However, a larger patient population willneed to be tested to define the variability of these effects oncells from patients with multiple myeloma.

This study also showed the high stability of the anti-hTfRIgG3-Av/b-SO6 complexes. These complexes were incu-bated in RMPI 1640 without a loss of activity for up to2 days. Only a slight loss of activity was observed after a6-day incubation. Interestingly, RPMI contains 0.2 mg/L(820 nmol/L) of biotin. The mean concentration of biotinin normal human serum is 244 F 61 pmol/L (32). Thus,the biotin concentration in RPMI is over 3,000-fold higherthan that of normal human serum. Our results show thatthe complexes are stable even in such high concentrationsof free biotin.

We show that anti-hTfR IgG3-Av alone activated caspase-2,caspase-3, caspase-8, and caspase-9 at late time points inthe sensitive cell line IM-9. A previous study showed thatanti-hTfR IgG3-Av induced caspase-3, caspase-8, andcaspase-9 activities in ARH-77 (another EBV-transformedlymphoblastoid cell line; ref. 4). However, no majordifference was found in the timing of activation and the

identification of a specific pathway of cell death remainedelusive. Alternatively, multiple apoptotic pathways maybe simultaneously activated. This study suggests thatcaspase-2 may be a key player in anti-hTfR IgG3-Av–induced cell death. Anti-hTfR IgG3-Av/b-SO6 also inducesapoptosis and activates caspase-2, caspase-3, caspase-8, andcaspase-9 in both anti-hTfR IgG3-Av sensitive (IM-9) andresistant (U266) cell lines. However, caspase-2 is activatedfirst and at higher levels in IM-9 cells. Caspase-2 andcaspase-3 are activated with similar kinetics and at similarlevels in U266 cells. Activation of caspase-3 by saporin hasbeen previously shown (10). In IM-9 cells treated with anti-hTfR IgG3-Av/b-SO6, caspase activation occurs morerapidly than in cells treated with anti-hTfR IgG3-Av alone.Caspase-2 is a unique caspase that contains propertiesof both initiator and executioner caspases (reviewed inref. 33). Caspase-2 is activated in response to UV irra-diation, cytokine deprivation, growth factor withdrawal,and DNA-damaging agents such as etoposide and cisplatin(33). How caspase-2 is activated upon treatment with anti-hTfR IgG3-Av alone, or conjugated to b-SO6, remains to bedetermined. Z-VAD-FMK effectively blocked Fas-mediatedcell death; however, it was unable to block the cell deathinduced by anti-hTfR IgG3-Av/b-SO6. This suggests thatanti-hTfR IgG3-Av/b-SO6 induces, at least in part, acaspase-independent mode of cell death. Malignant cellsoften harbor defects in the apoptotic machinery that allowthem to escape the cell death induced by many therapeutics(34–36). However, therapies that activate alternative formsof cell death, including caspase-independent cell death,may help to overcome this problem. Anti-hTfR IgG3-Av/b-SO6 may activate multiple cell death pathways simulta-neously. Inhibition of caspase activation does not alwaysprovide protection from the death stimulus and oftenreveals or enhances an underlying caspase-independentcell death pathway (34). Induction of multiple cell deathpathways may provide an advantage for anti-hTfR IgG3-Av/b-SO6 treatment by making it more difficult for thetumor to escape the cell death induced by the treatment.

We have previously shown that anti-hTfR IgG3-Avdisrupts the normal recycling of the hTfR and forces itinto the lysosomal compartment where it is presumablyrapidly degraded (4). This observation raises a concernthat the biotinylated protein toxin may also be rapidlydegraded. However, there seems to be some leakage of thistransport pathway or an alternative trafficking mechanismthat enables enough of the saporin toxin to escape andreach the cytoplasm where it can effectively inhibit proteinsynthesis. In addition, saporin is highly stable and isresistant to denaturation, high temperature, and proteolysis(37). The mechanism of how saporin escapes the endocyticpathway and reaches the cytoplasm is not known. Furtherstudies are needed to determine how saporin reaches thecytoplasm of targeted cells. However, the successfuldelivery of b-SO6 suggests that other cytotoxic agents canalso be delivered into cancer cells by anti-hTfR IgG3-Av.

Another concern in targeting the hTfR is the potentialnonspecific toxicity due to the expression of the hTfR on a7 Unpublished results.





variety of normal tissues. We have recently shown thatanti-hTfR IgG3-Av targets hTfR1 and not hTfR2.6 Thepattern of tissue expression of the two proteins is different.hTfR2 expression is limited to hepatocytes and enterocytesof the small intestine (1). hTfR1 is expressed at low levelson most cell types and is expressed at higher levels on cellswith an elevated proliferative activity (such as the basalepidermis and intestinal epithelium) or cells that requirehigh amounts of iron (such as erythroid progenitors thatneed iron for heme synthesis; ref. 1). Pluripotent hemato-poietic progenitors do not express detectable levels of thehTfR (38–40). In fact, our flow cytometry data show thatthere is a population of CD34+ cells that express little orno TfR as evidenced by the significant overlap with thenegative control peak. A second peak is also present,indicating that there are CD34+ cells that express the TfR.This is consistent with previous studies that detect TfRexpression on committed progenitors (41, 42).

In the present studies, we have shown that treatmentwith anti-hTfR IgG3-Av alone showed no toxic effects onhematopoiesis. However, anti-hTfR IgG3-Av/b-SO6 wastoxic to all types of colonies tested. This is consistent withthe results observed using another immunotoxin consistingof SO6 chemically conjugated to the murine IgG1 anti-hTfR antibody B3/25 (43). This study showed toxicity tocolony-forming unit–granulocyte/macrophage and burst-forming–erythroid, with progenitors of the erythroidlineage being more sensitive to the effects of the immuno-toxin. This is also consistent with our data and is expectedbecause erythroid progenitors express high levels of thehTfR (1, 39). Saporin immunotoxins targeting B-cellantigens have also shown toxicity to hematopoieticprogenitors (44, 45). It is important to note that thecolony-forming assay tests toxicity on committed progen-itors of each lineage. Less mature progenitors (pluripotent)can be distinguished from these committed ones by cellsurface antigen expression, including the lack of hTfRexpression on the less mature progenitors (39, 46). Theseless mature progenitor cells occur in the bone marrow atvery low frequency, estimated to be 0.05% to 0.1% of totalnucleated bone marrow cells (47) or about one cell per2 � 104 human bone marrow cells (48), making it difficult totest toxicity on these cells. Although there is some toxicityto the committed progenitors, anti-hTfR IgG3-Av/b-SO6 isexpected to have little to no toxicity on the less matureprogenitors due to the lack of hTfR expression (38–40);thus, the committed progenitors could be repopulated.

The hTfR is dramatically increased on the surface ofmalignant cells (1), which suggests that these malignantcells may be more vulnerable than normal cells toimmunotoxin targeting through the hTfR. In fact, clinicaltrials on the efficacy of a hTfR-targeted conjugate consistingof transferrin and CRM107, a mutated form of the inhibitoror protein synthesis toxin secreted by the bacteriumCorynebacterium diphtheria , support this hypothesis. Localinjection of transferrin-CRM107 (TransMID, Xenova Groupplc) has shown dramatic effects against brain cancer inphase I and phase II clinical trials, with minimal side effects

(recently reviewed in ref. 2). A phase III clinical trial inpatients with glioblastoma multiforme is currently underway (National Cancer Institute Clinical trial identifierNCT00087230). Clinical trials have been conducted withanother immunotoxin targeting the hTfR consisting of the454A12 murine anti-hTfR IgG1 antibody chemically conju-gated to the catalytic subunit of the plant toxin ricin (RTA;refs. 49, 50). 454A12-RTA has been tested in phase I clinicaltrials for the treatment of leptomeningeal neoplasia (49).Although high concentrations of 454A12-RTA showedcentral nervous system toxicity, a therapeutic window didexist and tumoricidal concentrations were safely reachedby intraventricular injections of 454A12-RTA. Centralnervous system toxicity was also observed in ovariancancer patients treated i.p. with high doses of 454A12-RTA,whereas no toxicities were observed at low doses (50).Systemic injection of TfR-targeted conjugates have alsoshown in vivo efficacy for anticancer treatment (51–53).In fact, human transferrin conjugated to cisplatin showeddramatic effects in two of five patients with advancedstages of breast cancer (51).

To construct the antibody-avidin fusion protein, wedecided to use human IgG3 due to its unique extendedhinge region that provides spacing and flexibility (54),thereby facilitating simultaneous binding of antigen andbiotinylated compounds. In fact, IgG3 is the most flexiblehuman IgG (55). IgG3 is also effective in complementactivation (56) and it binds all three FcgRs (56, 57).Moreover, fusing proteins after the CH3 domain of humanIgG3 does not result in the loss of Fc effector functions (58).Further studies are needed do determine the activity andrelevance of the Fc effector functions of anti-hTfR IgG3-Av.A concern that arises with the potential use of anti-hTfRIgG3-Av/b-SO6 complex as a therapeutic agent is itspossible immunogenicity in patients. A phase I clinicaltrial conducted with a murine monoclonal anti-CD30antibody conjugated to SO6 for the treatment of refractoryHodgkin’s disease showed transient tumor regression (14).However, the development of human antibodies againstboth the antibody and the toxin portion of the moleculelimited the efficacy of the immunotoxin. Another phase Iclinical trial using an immunotoxin consisting of murinemonoclonal bispecific antibodies (Fab¶2) with one armdirected against SO6 and one at CD22 on the surface ofB cells for the treatment of B-cell lymphoma has also beenconducted (15). Five patients were included in the studyand all showed transient responses to the treatment. Onlyone of the five patients developed human anti-mouseantibodies to the antibody portion of the immunotoxin aswell as human antitoxin antibodies to SO6. This immuno-toxin showed less immunogenicity than the full murinemonoclonal antibody tested in the first clinical trial.Because our antibody fusion protein was constructedusing human Fc regions, it is expected to show not onlydecreased immunogenicity but also enhanced cytotoxicitydue to the activation of human effector functions andincreased half-life in circulation (58). The avidin portion ofthe molecule could also be immunogenic. Because oral





antigens are known to induce tolerance (59), avidin waschosen for this study because it is expected to be lessimmunogenic compared with the bacterial protein strep-tavidin due to the human consumption of chicken eggs.The plant toxin b-SO6 is expected to be immunogenic inhumans after multiple administrations. However, admin-istration of immunosuppressive agents before or incombination with immunotoxins is expected to preventthe induction of antitoxin antibodies (60). For example, itmay be possible to treat patients with rituximab (anti-CD20 chimeric antibody: Rituxan; Genentech, Inc., andIDEC Pharmaceutical Corporation), which is in clinicaluse for the treatment of B cell non–Hodgkin’s lymphoma,to knockout B cells and prevent a humoral response withthe subsequent anti-hTfR IgG3-Av/b-SO6 treatment (60).The B cells would then be repopulated upon treatmentcessation. It is also possible that immunogenicity would below or absent when patients have an impaired immuneresponse, such as cases of hematologic malignanciesincluding multiple myeloma (61) and lymphoma (60).

It is also possible that the immunogenicity of the anti-hTfR IgG3-Av/b-SO6 complex can be decreased by usingan alternative biotinylated cytotoxic protein from humanorigin, such as the human eosinophil-derived neurotoxin,which is a part of the RNase A superfamily (62) or humanperforin (63). It is important to note, however, that toxichuman proteins are in general several orders of magnitudeless toxic than traditional bacterial or plant toxins (60). It isalso possible to use anti-hTfR IgG3-Av in a gene therapyapproach by conjugating it to a biotinylated carrier, such asa liposome containing a therapeutic gene. This approach togene delivery through targeting of the TfR has been termed‘‘transferrinfection’’ (64). The saporin gene has already beenshown to be a suitable candidate gene for this purpose (65).Specificity could be further enhanced by delivery of thesaporin gene under the control of a cell-specific promoter,such as the immunoglobulin promoter to target malignantB cells. The p53 gene could also be delivered into tumorcells by targeting of the hTfR to restore expression of wild-type p53 (66). This could sensitize tumor cells to othertraditional therapies such as radiation (67). Malignant cellscould also be sensitized to cell death through the deliveryof antisense oligonucleotides. In fact, antisense oligonu-cleotides to the antiapoptotic protein Bcl-2 delivered bytransferrin-conjugated liposomes resulted in a 10-foldincrease in daunorubicin toxicity in vitro (68). This strategyhas also been used to deliver antisense oligonucleotidesto a class folate receptor, another transmembrane receptorinvolved in cell growth (69). Delivery of these oligonucleo-tides sensitized breast cancer cells to doxorubicin treatmentin vitro .

The present study provides the initial proof-of-principlethat the antibody fusion protein anti-hTfR IgG3-Av iscapable of a two-pronged attack through its intrinsiccytotoxic activity and its ability to deliver an activebiotinylated toxin into malignant B cells. We also showedthat the delivery of a cytotoxic agent by the fusion proteinovercomes resistance in cells that show low sensitivity to

the fusion protein alone. Anti-hTfR IgG3-Av may be usedas a universal vector to deliver a wide variety oftherapeutic molecules (including other toxins or chemo-therapeutic drugs) in addition to saporin, increasing itsversatility as a therapeutic agent. In addition, we anticipatethat the utility of this therapeutic will not be restricted tothe elimination of myeloma cells in vivo , but that it canalso be used for in vitro approaches, including the efficientpurging of myeloma cells for use in autologous transplan-tation in patients with multiple myeloma. We would like tostress that the effect of the results obtained from the presentstudies is not restricted to multiple myeloma. Similarapproaches can be applied to other hematopoietic malig-nancies as well as other tumor types.

Acknowledgments

We thank Bruck Habtemariam (University of California at Los AngelesDivision of Hematology and Oncology, Department of Medicine,Administrative Analyst for Regulatory Affairs) for the coordination andcollection of patient bone marrow biopsies, Dr. Sherie L. Morrison(University of California at Los Angeles Department of Microbiology,Immunology, and Molecular Genetics) for her support on this project, andDr. Otoniel Martinez-Maza (University of California at Los AngelesDepartment of Obstetrics and Gynecology and Department of Microbi-ology, Immunology, and Molecular Genetics) for his critical evaluation ofthe manuscript.

References

1. Daniels TR, Delgado T, Rodriguez JA, Helguera G, Penichet ML. Thetransferrin receptor part I: biology and targeting with cytotoxic antibodiesfor the treatment of cancer. Clin Immunol 2006;121:144–58.

2. Daniels TR, Delgado T, Helguera G, Penichet ML. The transferrinreceptor part II: targeted delivery of therapeutic agents into cancer cells.Clin Immunol 2006;121:159–76.

3. Ng PP, Dela Cruz JS, Sorour DN, et al. An anti-transferrin receptor-avidin fusion protein exhibits both strong proapoptotic activity and theability to deliver various molecules into cancer cells. Proc Natl Acad SciU S A 2002;99:10706–11.

4. Ng PP, Helguera G, Daniels TR, et al. Molecular events contributing tocell death in malignant human hematopoietic cells elicited by an IgG3-avidin fusion protein targeting the transferrin receptor. Blood 2006;108:2745–54.

5. Green NM. Avidin and streptavidin. Methods Enzymol 1990;184:51–67.

6. Stirpe F, Barbieri L, Battelli MG, Soria M, Lappi DA. Ribosome-inactivating proteins from plants: present status and future prospects.Biotechnology 1992;10:405–12.

7. Davol PA, Frackelton AR, Jr. Targeting human prostatic carcinomathrough basic fibroblast growth factor receptors in an animal model:characterizing and circumventing mechanisms of tumor resistance.Prostate 1999;40:178–91.

8. Davol P, Beitz JG, Mohler M, et al. Saporin toxins directed to basicfibroblast growth factor receptors effectively target human ovarianteratocarcinoma in an animal model. Cancer 1995;76:79–85.

9. Lappi DA, Ying W, Barthelemy I, et al. Expression and activities of arecombinant basic fibroblast growth factor-saporin fusion protein. J BiolChem 1994;269:12552–8.

10. Duxbury MS, Ito H, Ashley SW, Whang EE. CEACAM6 as a noveltarget for indirect type 1 immunotoxin-based therapy in pancreaticadenocarcinoma. Biochem Biophys Res Commun 2004;317:837–43.

11. Adam PJ, Terrett JA, Steers G, et al. CD70 (TNFSF7) is expressed athigh prevalence in renal cell carcinomas and is rapidly internalised onantibody binding. Br J Cancer 2006;95:298–306.

12. Pistillo MP, Tazzari PL, Palmisano GL, et al. CTLA-4 is not restricted tothe lymphoid cell lineage and can function as a target molecule forapoptosis induction of leukemic cells. Blood 2003;101:202–9.

13. Foehr ED, Lorente G, Kuo J, Ram R, Nikolich K, Urfer R. Targeting of





the receptor protein tyrosine phosphatase h with a monoclonal antibodydelays tumor growth in a glioblastoma model. Cancer Res 2006;66:2271–8.

14. Falini B, Bolognesi A, Flenghi L, et al. Response of refractoryHodgkin’s disease to monoclonal anti-CD30 immunotoxin. Lancet 1992;339:1195–6.

15. French RR, Bell AJ, Hamblin TJ, Tutt AL, Glennie MJ. Response ofB-cell lymphoma to a combination of bispecific antibodies and saporin.Leuk Res 1996;20:607–17.

16. Bergamaschi G, Cazzola M, Dezza L, Savino E, Consonni L, LappiD. Killing of K562 cells with conjugates between human transferrinand a ribosome-inactivating protein (SO-6). Br J Haematol 1988;68:379–84.

17. Gosselaar PH, van-Dijk AJ, de-Gast GC, et al. Transferrin toxin but nottransferrin receptor immunotoxin is influenced by free transferrin and ironsaturation. Eur J Clin Invest 2002;32 Suppl 1:61–9.

18. Coloma MJ. Production and characterization of novel recombinantantibodies. Microbiology and molecular genetics. Los Angeles: Universityof California at Los Angeles; 1997. p. 263.

19. Morrison SL, Wims L, Wallick S, Tan L, Oi VT. Genetically engineeredantibody molecules and their application. Ann N Y Acad Sci 1987;507:187–98.

20. Shin SU, Wu D, Ramanathan R, Pardridge WM, Morrison SL.Functional and pharmacokinetic properties of antibody-avidin fusionproteins. J Immunol 1997;158:4797–804.

21. White S, Taetle R, Seligman PA, Rutherford M, Trowbridge IS.Combinations of anti-transferrin receptor monoclonal antibodies inhibithuman tumor cell growth in vitro and in vivo : evidence for synergisticantiproliferative effects. Cancer Res 1990;50:6295–301.

22. Helguera G, Penichet ML. Antibody-cytokine fusion proteins for thetherapy of cancer. Methods Mol Med 2005;109:347–74.

23. Bagga S, Hosur MV, Batra JK. Cytotoxicity of ribosome-inactivatingprotein saporin is not mediated through a2-macroglobulin receptor. FEBSLett 2003;541:16–20.

24. Lappi DA, Esch FS, Barbieri L, Stirpe F, Soria M. Characterization of aSaponaria officinalis seed ribosome-inactivating protein: immunoreactiv-ity and sequence homologies. Biochem Biophys Res Commun 1985;129:934–42.

25. Stirpe F, Gasperi-Campani A, Barbieri L, Falasca A, Abbondanza A,Stevens WA. Ribosome-inactivating proteins from the seeds of Saponariaofficinalis L. (soapwort), of Agrostemma githago L. (corn cockle) and ofAsparagus officinalis L. (asparagus), and from the latex of Hura crepitansL. (sandbox tree). Biochem J 1983;216:617–25.

26. Hayman SR, Fonseca R. Plasma cell leukemia. Curr Treat OptionsOncol 2001;2:205–16.

27. Carrasco RA, Stamm NB, Patel BK. One-step cellular caspase-3/7assay. Biotechniques 2003;34:1064–7.

28. Ko SC, Johnson VL, Chow SC. Functional characterization of JurkatT cells rescued from CD95/Fas-induced apoptosis through the inhibition ofcaspases. Biochem Biophys Res Commun 2000;270:1009–15.

29. Fuentes-Prior P, Salvesen GS. The protein structures that shapecaspase activity, specificity, activation and inhibition. Biochem J 2004;384:201–32.

30. Westendorf JJ, Lammert JM, Jelinek DF. Expression and function ofFas (APO-1/CD95) in patient myeloma cells and myeloma cell lines. Blood1995;85:3566–76.

31. Kim KM, Lee K, Hong YS, Park HY. Fas-mediated apoptosis andexpression of related genes in human malignant hematopoietic cells. ExpMol Med 2000;32:246–54.

32. Mock DM, Lankford GL, Mock NI. Biotin accounts for only half ofthe total avidin-binding substances in human serum. J Nutr 1995;125:941–6.

33. Zhivotovsky B, Orrenius S. Caspase-2 function in response to DNAdamage. Biochem Biophys Res Commun 2005;331:859–67.

34. Jaattela M. Programmed cell death: many ways for cells to diedecently. Ann Med 2002;34:480–8.

35. Jaattela M. Multiple cell death pathways as regulators of tumourinitiation and progression. Oncogene 2004;23:2746–56.

36. Kroemer G, Martin SJ. Caspase-independent cell death. Nat Med2005;11:725–30.

37. Santanche S, Bellelli A, Brunori M. The unusual stability of saporin, a

candidate for the synthesis of immunotoxins. Biochem Biophys ResCommun 1997;234:129–32.

38. Gross S, Helm K, Gruntmeir JJ, Stillman WS, Pyatt DW, Irons RD.Characterization and phenotypic analysis of differentiating CD34+ humanbone marrow cells in liquid culture. Eur J Haematol 1997;59:318–26.

39. Lansdorp PM, Dragowska W. Long-term erythropoiesis from constantnumbers of CD34+ cells in serum-free cultures initiated with highlypurified progenitor cells from human bone marrow. J Exp Med 1992;175:1501–9.

40. Bender JG, Unverzagt K, Walker DE, et al. Phenotypic analysis andcharacterization of CD34+ cells from normal human bone marrow, cordblood, peripheral blood, and mobilized peripheral blood from patientsundergoing autologous stem cell transplantation. Clin Immunol Immuno-pathol 1994;70:10–8.

41. Taetle R. The role of transferrin receptors in hemopoietic cell growth.Exp Hematol 1990;18:360–5.

42. Sieff C, Bicknell D, Caine G, Robinson J, Lam G, Greaves MF.Changes in cell surface antigen expression during hemopoietic differenti-ation. Blood 1982;60:703–13.

43. Cazzola M, Bergamaschi G, Dezza L, et al. Cytotoxic activity of ananti-transferrin receptor immunotoxin on normal and leukemic humanhematopoietic progenitors. Cancer Res 1991;51:536–41.

44. Bolognesi A, Polito L, Tazzari PL, et al. In vitro anti-tumour activity ofanti-CD80 and anti-CD86 immunotoxins containing type 1 ribosome-inactivating proteins. Br J Haematol 2000;110:351–61.

45. Barbieri L, Dinota A, Gobbi M, et al. Immunotoxins containing saporin6 and monoclonal antibodies recognizing plasma cell-associated antigens:effects on target cells and on normal myeloid precursors (CFU-GM). Eur JHaematol 1989;42:238–45.

46. Andrews RG, Singer JW, Bernstein ID. Precursors of colony-formingcells in humans can be distinguished from colony-forming cells byexpression of the CD33 and CD34 antigens and light scatter properties.J Exp Med 1989;169:1721–31.

47. Paquette R, Dorshkind K. Optimizing hematopoietic recovery follow-ing bone marrow transplantation. J Clin Invest 2002;109:1527–8.

48. Sutherland HJ, Lansdorp PM, Henkelman DH, Eaves AC, Eaves CJ.Functional characterization of individual human hematopoietic stem cellscultured at limiting dilution on supportive marrow stromal layers. Proc NatlAcad Sci U S A 1990;87:3584–8.

49. Laske DW, Muraszko KM, Oldfield EH, et al. Intraventricularimmunotoxin therapy for leptomeningeal neoplasia. Neurosurgery 1997;41:1039–49; discussion 1049–51.

50. Bookman MA. Biological therapy of ovarian cancer: current directions.Semin Oncol 1998;25:381–96.

51. Elliott RL, Stjernholm R, Elliott MC. Preliminary evaluation of platinumtransferrin (MPTC-63) as a potential nontoxic treatment for breast cancer.Cancer Detect Prev 1988;12:469–80.

52. Mayers GL, Razq J, Abu-Hadid MM. Cytotoxic drug conjugates fortreatment of neoplastic disease. US Patent No. 5393737, February 28,1995.

53. Bjorn MJ, Groetsema G. Immunotoxins to the murine transferrinreceptor: intracavitary therapy of mice bearing syngeneic peritonealtumors. Cancer Res 1987;47:6639–45.

54. Phillips ML, Tao MH, Morrison SL, Schumaker VN. Human/mousechimeric monoclonal antibodies with human IgG1, IgG2, IgG3 and IgG4constant domains: electron microscopic and hydrodynamic characteriza-tion. Mol Immunol 1994;31:1201–10.

55. Dangl JL, Wensel TG, Morrison SL, Stryer L, Herzenberg LA, OiVT. Segmental flexibility and complement fixation of geneticallyengineered chimeric human, rabbit and mouse antibodies. EMBOJ 1988;7:1989–94.

56. TaoMH,Morrison SL. Studies of aglycosylated chimericmouse-humanIgG. Role of carbohydrate in the structure and effector functions mediatedby the human IgG constant region. J Immunol 1989;143:2595–601.

57. Janeway CA, Travers P, Walport M, Shlomchik M. The generation oflymphocyte antigen receptors. 5th ed. In: Immunobiology: the immunesystem in health and disease. New York: Garland Publishing; 2001.p. 123–54.

58. Penichet ML, Morrison SL. Design and engineering human forms ofmonoclonal antibodies. Drug Dev Res 2004;61:121–36.

59. Weiner HL. Oral tolerance. Proc Natl Acad Sci U S A 1994;91:10762–5.





60. Frankel AE. Reducing the immune response to immunotoxin. ClinCancer Res 2004;10:13–5.

61. Schiller G. Multiple Myeloma, heavy chain diseases, and Walden-strom’s macroglobulinemia. Vol. 1. Chapter 48. In: Bick R, editor.Hematology: clinical and laboratory practice. Mosby-Year Book Inc.;1993. p. 705–27.

62. Newton DL, Nicholls PJ, Rybak SM, Youle RJ. Expression andcharacterization of recombinant human eosinophil-derived neurotoxin andeosinophil-derived neurotoxin-anti-transferrin receptor sFv. J Biol Chem1994;269:26739–45.

63. Wan L, Zeng L, Chen L, et al. Expression, purification, and refolding ofa novel immunotoxin containing humanized single-chain fragment variableantibody against CTLA4 and the N-terminal fragment of human perforin.Protein Expr Purif 2006;48:307–13.

64. Wagner E, Cotten M, Foisner R, Birnstiel ML. Transferrin-polycation-DNA complexes: the effect of polycations on the structure of the complexand DNA delivery to cells. Proc Natl Acad Sci U S A 1991;88:4255–9.

65. Zarovni N, Vago R, Solda T, Monaco L, Fabbrini MS. Saporin as anovel suicide gene in anticancer gene therapy. Cancer Gene Ther 2007;14:165–73.

66. Xu L, Huang CC, Huang W, et al. Systemic tumor-targeted genedelivery by anti-transferrin receptor scFv-immunoliposomes. Mol CancerTher 2002;1:337–46.

67. Xu L, Pirollo KF, Tang WH, Rait A, Chang EH. Transferrin-liposome-mediated systemic p53 gene therapy in combination with radiation resultsin regression of human head and neck cancer xenografts. Hum Gene Ther1999;10:2941–52.

68. Chiu SJ, Liu S, Perrotti D, Marcucci G, Lee RJ. Efficient delivery of aBcl-2-specific antisense oligodeoxyribonucleotide (G3139) via transferrinreceptor-targeted liposomes. J Control Release 2006;112:199–207.

69. Jhaveri MS, Rait AS, Chung KN, Trepel JB, Chang EH. Antisenseoligonucleotides targeted to the human a folate receptor inhibit breastcancer cell growth and sensitize the cells to doxorubicin treatment. MolCancer Ther 2004;3:1505–12.





2007;6:2995-3008. Mol Cancer Ther Tracy R. Daniels, Patrick P. Ng, Tracie Delgado, et al. hematopoietic cellsenhancement of its cytotoxicity against malignantfusion protein with biotinylated saporin results in significant

transferrin receptor IgG3-avidin−Conjugation of an anti

Updated version

http://mct.aacrjournals.org/content/6/11/2995

Access the most recent version of this article at:

Cited articles

http://mct.aacrjournals.org/content/6/11/2995.full#ref-list-1

This article cites 65 articles, 23 of which you can access for free at:

Citing articles

http://mct.aacrjournals.org/content/6/11/2995.full#related-urls

This article has been cited by 2 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://mct.aacrjournals.org/content/6/11/2995To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/content/6/11/2995.full#ref-list-1

http://mct.aacrjournals.org/content/6/11/2995.full#related-urls

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

mailto:[email protected]



Documents

Conjugation of an anti–transferrin receptor IgG3-avidin ... · Cell images were captured using a Zeiss Axiovert 40 CFL PlasDIC Inverted Microscope using a 20 objective (Mikron Instruments,