the dose-dependent tumor targeting of antibody ifn-g fusion … · 2014. 5. 2. · tumor therapy...

Research Article

The Dose-Dependent Tumor Targeting of Antibody–IFN-gFusion Proteins Reveals an Unexpected Receptor-TrappingMechanism In Vivo

Teresa Hemmerle and Dario Neri

AbstractCytokines often display substantial toxicities at low concentrations, preventing their escalation for

therapeutic treatment of cancer. Fusion proteins comprising cytokines and recombinant antibodies mayimprove the anticancer activity of proinflammatory cytokines. Murine IFN-g was appended in the diabodyformat at the C-terminus of the F8 antibody, generating the F8–IFN-g fusion protein. The F8 antibody isspecific for the extra-domain A (EDA) of fibronectin, a tumor-associated antigen that is expressed in thevasculature and stroma of almost all tumor types. Tumor-targeting properties were measured in vivo using aradioiodinated preparation of the fusion protein. Therapy experiments were performed in three syngeneicmurine models of cancer [F9 teratocarcinoma, WEHI-164 fibrosarcoma, and Lewis lung carcinoma (LLC)].F8–IFN-g retained the biologic activity of both the antibody and the cytokine moiety in vitro, but, unlike theparental F8 antibody, it did not preferentially localize to the tumors in vivo. However, when unlabeled F8–IFN-g was administered before radioiodinated F8–IFN-g , a selective accumulation at the tumor site wasobserved. F8–IFN-g showed dose-dependent anticancer activity with a clear superiority over untargetedrecombinant IFN-g . The anticancer activity was potentiated by combining with F8–IL-4 without additionaltoxicities, whereas combination of F8–IFN-g with F8–TNF was lethal in all mice. Unlike other antibody–cytokine fusions, the use of IFN-g as payload for anticancer therapy is associated with a receptor-trappingmechanism, which can be overcome by the administration of a sufficiently large amount of the fusion proteinwithout any detectable toxicity at the doses used. Cancer Immunol Res; 2(6); 1–9. �2014 AACR.

IntroductionImmunocytokines are fusion proteins consisting of a cyto-

kine and a recombinant antibody. They represent a novel classof "armed" antibodies with considerable anticancer potential(1–4). Indeed, antibodies capable of selective accumulation atthe tumor site may act as delivery vehicles and may substan-tially increase the therapeutic index of proinflammatorycytokines.Various cytokines have been fused to the C-terminal extrem-

ity of full immunoglobulin G (IgG) antibodies, leading toproducts with considerable antitumor activity in mouse mod-els of cancer, and therefore have progressed to clinical studies(1). We and others have fused cytokines to recombinantantibody fragments (e.g., scFv fragments and diabodies) devoid

of the Fc portion to generate proteins that do not activatecomplement or bind to Fc receptors (1). We have focused onthe use of antibodies that recognize tumor-associated antigens(TAA) found in the subendothelial extracellular matrix of solidtumors and lymphomas (5–8). In particular, we have used theF8 and L19 antibodies, specific to the alternatively splicedextra-domain A (EDA) and B (EDB) of fibronectin, respectively,for the construction and in vivo testing of several immunocy-tokines. These oncofetal isoforms of fibronectin are virtuallyundetectable in normal adult organs (except the placenta andendometrium during the proliferative phase; ref. 9), while theyare expressed abundantly in the neovasculature and stroma ofvirtually all aggressive tumors in mice and humans (7, 10–13).Some proinflammatory cytokines [e.g., interleukin (IL)-2, IL-12,and TNF] that were fused to L19 or F8 have exhibited impres-sive anticancer activity and selective uptake at the tumor site(11–20). Other cytokines have shown limitations either intumor targeting or in therapy experiments (e.g., IL-7, IL-17,IL-15, and IL-18; refs. 21–23), indicating that the immunocy-tokine format and the choice of payload have to be evaluatedfrom case to case. At one extreme, anti-inflammatory cyto-kines (such as IL-10) can be used as fusion partners withdisease-homing antibodies with no inhibition of tumorgrowth (E. Trachsel and D. Neri; unpublished data) but witha substantial inhibition of the autoimmune and/or inflam-matory conditions (9, 24).

Authors' Affiliation: Department of Chemistry and Applied Biosciences,Swiss Federal Institute of Technology (ETH Z€urich), Z€urich, Switzerland

Note: Supplementary data for this article are available at Cancer Immu-nology Research Online (http://cancerimmunolres.aacrjournals.org/).

Corresponding Author: Dario Neri, Department of Chemistry and AppliedBiosciences, Swiss Federal Institute of Technology (ETH Z€urich), Wolf-gang-Pauli-Strasse, 10, Zurich 8093, Switzerland. Phone: 41-44-633-74-01; Fax: 41-44-633-13-58; E-mail: [email protected]

doi: 10.1158/2326-6066.CIR-13-0182

�2014 American Association for Cancer Research.

CancerImmunology

Research

www.aacrjournals.org OF1

Research. on August 16, 2021. © 2014 American Association for Cancercancerimmunolres.aacrjournals.org Downloaded from

Published OnlineFirst March 24, 2014; DOI: 10.1158/2326-6066.CIR-13-0182

http://cancerimmunolres.aacrjournals.org/

A comprehensive analysis of antibody payloads, whichcompletely abrogate the in vivo accumulation of the parentalantibody at the tumor site, revealed a number of possiblemechanisms. For example, a fusion of the L19 antibody withhighly chargedmolecules [such as calmodulin, TAT peptide, ormurine VEGF-164 (but not the less-charged VEGF-120 iso-form)] abolished tumor targeting in vivo. Similarly, excessiveglycosylation (25) or too large molecular weight of the fusionprotein (17) can lead to immunocytokines with targetingpreference of the parental antibody in vitro, but they do notaccumulate at the tumor site in vivo.

IFN-g has long been considered as a payload for immuno-cytokine development (26, 27), as high concentrations of thisprotein at the tumor site can mediate a potent influx andactivation of leukocytes (28). The expression of IFN-g–basedimmunocytokines is complicated by the presence of manycysteine residues, which lead to disulfide-linked high-molec-ular-weight aggregates. We have shown that the deletion ofcysteine residues or their mutation to serines leads to theproduction of immunocytokines with acceptable pharmaceu-tical properties (29). One of these mutants, L19-IFN-gmut4,was found to exhibit measurable antitumor activity, whichcould be potentiated by combination with L19–IL-2. Thetumor-targeting activity of L19-IFN-gmut4 was substantiallyhigher in mice with deletion of the gene encoding the receptorfor IFN-g than in wild-type mice (29).

In this study, we investigated novel IFN-g–based immuno-cytokines, using the F8 antibody, whose cognate antigen is theEDA domain of fibronectin, a TAA highly expressed in murine(30, 31) and human tumors (11–13). Furthermore, comparedwith the work of Ebbinghaus and colleagues (29), we exploreddifferent strategies for the fusion of IFN-g , which resulted in theproduction and purification of immunocytokines with phar-maceutically acceptable profiles in SDS-PAGE, gel-filtration,surface plasmon resonance (SPR) analysis, and cytokine activ-ity assays.

We found that F8–IFN-g did not exhibit the expectedpreferential localization on tumors in vivo. However, when20 mg of unlabeled F8–IFN-g was administered before theintravenous injection of radioiodinated F8–IFN-g , a selectiveaccumulation of the immunocytokine at the tumor site wasobserved (tumor:blood ratio ¼ 20 at 24 hours), indicating thepresence of a receptor-trapping mechanism that could besaturated. The biodistribution information was used for ther-apy experiments at relatively high doses (200 mg every 3 days),demonstrating potent antitumor activity with no detectabletoxicity for immunocytokine F8–IFN-g . The anticancer activityof F8–IFN-g was superior to that of KSF–IFN-g , a fusion proteinconsisting of the anti-hen egg lysozyme KSF antibody, whichserved as a negative control in this study.

Materials and MethodsCell lines and tumor models

Chinese hamster ovary (CHO) cells, WEHI-279 cells, andmurine tumor cell lines F9 teratocarcinoma, WEHI-164 sar-coma, and Lewis lung carcinoma (LLC) were purchased fromAmerican Type Culture Collection-LGC (ATCC-LGC). All cells

were cultured according to the supplier's protocol, and noadditional authentication was performed. All animal experi-ments were performed under a project license granted by theVeterin€aramt des Kantons Z€urich, Switzerland (42/2012), andanimals were sacrificed when the tumor volume reached amaximum of 2,000 mm3. Female 129/SvEv and male DBA/1mice were obtained from Charles River Laboratories. FemaleBalb/c and C57BL/6J mice were obtained from Janvier. Fortherapy studies and biodistribution experiments, mice wereimplanted subcutaneously in the flank with 25 � 106 (F9), 3 �106 (WEHI-164), or 1 � 106 (LLC) tumor cells.

Cloning of IFN-g–based antibody fusion proteinsThe fusion proteins F8–IFN-g and KSF–IFN-g contain the

antibody F8 (specific to the alternatively spliced EDA domainof fibronectin; ref. 8) or the antibody KSF (specific to egglysozyme) in the diabody format sequentially fused to murineIFN-g (aa 23-155, Cys155 to Ser155; cDNA from Source Bio-Science) by a 9-amino acid (aa) linker. Genes encoding the F8antibody, the KSF antibody, and murine IFN-g were PCRamplified, PCR assembled, and cloned into the mammaliancell expression vector pcDNA3.1(þ) (Invitrogen) by a HindIII/NotI restriction site (New England BioLabs; for nucleotide andamino acid sequence, see Supplementary Data S1).

Expression and in vitro characterization of IFN-g–basedimmunocytokines

Fusion proteins were expressed by the clonal, stably trans-fected CHO cell line. G418-resistant clones were screened forprotein expression by ELISA using recombinant EDA or henegg lysozyme. For protein production, cells were adapted forsuspension growth, and cell culture supernatant was purifiedto homogeneity by protein A chromatography (GE Healthcare;ref. 32). The purified proteins were analyzed by SDS-PAGE(NuPage 4%–12% Bis–Tris Gel, MOPS running buffer; Invitro-gen), size exclusion chromatography (Superdex200 10/300GL;GE Healthcare), and SPR analysis (BIAcore; GE Healthcare) onan EDA-coated sensor chip (8, 33). Cytokine activity wasanalyzed using a cytostasis assay with WEHI-279 lymphomacells (20,000 cells/well). Cells were incubated in triplicates for72 hours with different concentrations of F8–IFN-g , KSF–IFN-g , or recombinant murine IFN-g (produced in Escherichia coli;Merck Millipore), and cell viability was determined with Cell-Titer AQueous One Solution (Promega).

Immunofluorescence on tumor sectionsFreshly frozen cryostat sections (10 mm) of untreated

tumors were fixed in ice-cold acetone and stained withbiotinylated SIP(F8), SIP(L19), SIP(KSF) or F8–IFN-g , andrat anti-CD31 (BD Biosciences). Streptavidin–Alexa Fluor488 (Biospa) and donkey anti-rat IgG Alexa Fluor 594 wereused for detection. Slides were mounted with fluorescentmounting medium (Dako) and analyzed with an Axioskop2mot plus microscope (Zeiss).

Quantitative biodistribution experimentsTo assess the targeting activity, F8–IFN-g and KSF–IFN-g

proteins were labeled with 125-iodine as described previously

Hemmerle and Neri

Cancer Immunol Res; 2(6) June 2014 Cancer Immunology ResearchOF2




(23), and the radioiodinated fusion proteins were injectedintravenously into the tail vein of tumor-bearing mice. Micewere sacrificed after 1 or 24 hours, and the organs wereexcised and weighed. Radioactivity content was measuredusing a Packard Cobra g counter and expressed as apercentage of injected dose per gram of tissue (%ID/g� SE).

Blood binding assayFresh murine blood obtained from DBA/1 mice was incu-

bated for 10 minutes with the radioiodinated protein prepara-tions at different concentrations in lithium heparin containingMicrotainer tubes (BD Biosciences). Blood cells and liquidcontentwere separated by centrifugation (3minutes; 3,000� g)and the radioactivity was measured. Results are expressed as apercentage of radioactivity dose.

Tumor therapy studiesFor the assessment of antitumor activity, mice were

randomly grouped (n ¼ 5) when tumors reached the sizeof 100 mm3 and treatment was started by intravenousinjection into the lateral tail vein according to therapyschedule. The combination agents F8–TNF and F8–IL-4 aredescribed elsewhere (30, 31). Mice were monitored daily andtumor volume was measured with a digital caliper. Tumorvolume was calculated using the formula: volume ¼ 0.5 �length � width2.

Immunofluorescence analysis of tumor-infiltrating cellsFor ex vivodetectionof targeting,micewere treated according

to the therapy schedule (3 injections, every 72 hours, 200mg) andtumorswere excised 1 day after the final injection. Tumorswereembedded in optimum cutting temperature (OCT) medium(Thermo Scientific) and 10-mm cryostat sections were stainedusing the rat anti-murine IFN-g antibodies (eBioscience) andanti-rat Alexa Fluor 488–coupled secondary antibodies (Invitro-gen). For the assessment of tumor-infiltrating immune cells,sections were stained using rat anti-CD45 (leukocytes; BDBiosciences), rat anti-CD4 (CD4þ T cells; BioXCell), rat anti-CD8 (CD8þ T cells; BioXCell), rat anti-F4/80 (macrophages;Abcam), rabbit anti-Asialo GM1 [natural killer (NK) cells; WakoPureChemical Industries], rat anti-CD45R (B cells; eBioscience),rat anti-Foxp3 (eBioscience), and rabbit anti-CD25 (Santa CruzBiotechnology) antibodies, detected with Alexa Fluor 488–cou-pled secondary antibodies (Invitrogen). For vascular costaining,goat or rat anti-CD31 (Santa Cruz Biotechnology; eBioscience)and anti-goat or anti-rat IgGAlexa Fluor 594–coupled secondaryantibodies (Invitrogen) were used. Slides were mounted withfluorescent mounting medium (Dako) and analyzed with anAxioskop2 mot plus microscope (Zeiss).

Statistical analysisDifferences in targeting, tumor volume, and survival data

were compared using the repeated-measures (mixed-model)

Figure 1. Cloning, expression, andin vitro characterization of thenoncovalent dimer F8–IFN-g .A, schematic representation ofplasmid map. mIFN-g , murine IFN-g . B, schematic representation ofthe domain assembly of fusionprotein; VH, variable domain ofheavy chain; VL, variable domain oflight chain. C, SDS-PAGEof affinitypurified F8–IFN-g (predictedmolecular weight monomer: 41kDa; two glycosylation forms).M, molecular weight marker; Red,reducing conditions; Non,nonreducing conditions. D,cytostasis assay of F8–IFN-g ,KSF–IFN-g , and commercialrecombinant murine IFN-g usingWEHI-279 lymphoma cells (20,000cells/well); IC50 (rIFN-g ), 1.9� 10�8

mol/L; IC50 (F8–IFN-g ), 1.3 � 10�13

mol/L; IC50 (KSF–IFN-g ), 5.3 �10�14 mol/L. E, size exclusionchromatography of F8–IFN-g(predicted molecular weight dimer82 kDa). 1, Thyroglobulin 669 kDa;2, bovine serum albumin (BSA) 67kDa; 3, b-lactoglobulin 35 kDa. F,SPR (BIAcore) profile of F8–IFN-gon an EDA-coated sensor chip(apparent KD, 1 nmol/L).

The Tumor-Targeting Properties of F8–IFN-g

www.aacrjournals.org Cancer Immunol Res; 2(6) June 2014 OF3




ANOVA analysis and Mantel–Cox test, respectively, of Graph-Pad Prism.

ResultsCloning, production, and characterization of fusionproteins

The immunocytokines F8–IFN-g (specific to the EDAdomain of fibronectin, a marker for tumor angiogenesis) andKSF–IFN-g (specific to hen egg lysozyme) were cloned andexpressed by stable transfection in CHO cells. The immuno-cytokine comprised the antibody in a stable noncovalenthomodimeric diabody format (i.e., 5-amino acid linker betweenVHandVLdomains), with the IFN-gmoiety appended to theC-terminal extremity via a flexible 15-amino acid linker (Fig. 1Aand B). F8–IFN-g and KSF–IFN-g were purified from cell

supernatant by protein A chromatography, yielding a proteinpreparation (Fig. 1C and E) containing different glycosylationforms similar to the naturally occurring IFN-g (34), as indicatedby the twomolecular weight bands in SDS-PAGE (Fig. 1C). Theproduct retained high affinity for the cognate antigen, asrevealed by SPR analysis (Fig. 1F), and displayed a higherbiologic cytokine activity than the IFN-g from recombinantE. coli, as measured in a cytostasis assay with WEHI-279 cells(Fig. 1D).

Biodistribution studies and single-agent therapeuticactivity

To study the tumor-targeting properties and the therapeuticactivity of immunocytokines F8–IFN-g compared with thenegative control KSF–IFN-g , we used three murine modelsof cancer, grafted onto immunocompetent mice. The F9

Figure 2. Characterization of tumor-targeting performance byimmunofluorescence analysis andquantitative biodistributionstudies. A, freshly frozen untreatedtumor sections were stained withbiotinylated SIP(KSF) (specific toegg lysozyme, as negative control),SIP(F8) (specific to EDA), and F8–IFN-g (Alexa Fluor 488; green) andanti-CD31 antibody (Alexa Fluor594; red). B, immunofluorescenceanalysis of tumors treated withPBS, KSF–IFN-g , or F8–IFN-g (totalof 600 mg protein, given in 3intravenous injections, every 72hours). Cryostat sections of tumorswere stained with anti-IFN-gantibody (Alexa Fluor 488; green)and anti-CD31 antibody (AlexaFluor 594; red). Magnification,�20; scale bar, 100 mm. C,biodistribution data afterintravenous injection of either 15mg radioiodinated SIP(F8), 15 mgradioiodinated F8–IFN-g , or 20 mgunlabeled F8–IFN-g followed 10minutes later by the injection of 15mg radioiodinated F8–IFN-g . Micewere sacrificed after 24 hours.Organs were excised andradioactivity counted, expressingresults as %ID/g � SE. D, ex vivoblood binding assay by incubationof fresh murine blood withradioiodinated F8–IFN-g . E,biodistribution experiment with 15mg radioiodinated SIP(F8), 15 mgradioiodinated F8–IFN-g , or 20 mgunlabeled F8–IFN-g and F8-B7.2,respectively, followed 10 minuteslater by the injection of 15 mgradioiodinated F8–IFN-g . Micewere sacrificed 1 hour afterintravenous injection.

Hemmerle and Neri





teratocarcinoma grows in Sv129 mice, the LLC grows inC57BL/6 mice, and the WEHI-164 fibrosarcoma grows inBalb/c mice. Figure 2A shows that EDA is strongly expressedin the vasculature and in the interstitiumof LLC andWEHI-164tumors, while antigen expression is preferentially foundaround the angiogenic blood vessels of F9 tumors. Followingintravenous injection of 200 mg of either F8-IFN-g or KSF–IFN-g , only the F8–IFN-g immunocytokine could be detected byimmunofluorescence to preferentially accumulate in all threetypes of tumors (Fig. 2B).A quantitative biodistribution analysis was performed using

radioiodinated protein preparations in immunocompetentmice bearing F9 tumors. Although the parental F8 antibodyin small immune protein (SIP) format exhibited selectivetumor uptake in line with previously published results (8),F8–IFN-g did not accumulate in the tumorwhen used at a doseof 15 mg (Fig. 2C). However, pre-administration of 20 mgunlabeled F8–IFN-g , followed 15 minutes later by the injectionof radioiodinated F8–IFN-g , resulted in a biodistribution pro-file similar to that of the parental antibody (Fig. 2C), indicatingthat a trapping interaction could be blocked and saturated atthe doses used. It is unlikely that this interaction took place atthe level of circulating leukocytes, because an ex vivo incuba-tion of radiolabeled F8–IFN-g (at 1 mg/mL and at 100 mg/mL)with mouse blood followed by centrifugation led to the recov-ery of the immunocytokine in the supernatant (Fig. 2D).A quantitative biodistribution analysis performed 1 hour

after intravenous injection of radioiodinated protein prepara-tions revealed that, in contrast with the parental F8 antibody inSIP format, the immunocytokine F8–IFN-g mainly localized tothe liver and the spleen (Fig. 2E). The pre-administration ofunlabeled F8–IFN-g reduced the liver and spleen uptake,whereas F8-B7.2, a highly glycosylated F8-based fusion protein(25) that did not contain the IFN-g moiety, showed no inhib-itory effect. These data further support the hypothesis that invivo trapping of F8–IFN-g at low doses is associated with thepresence of the IFN-g moiety.In a therapy experiment, the dose of F8–IFN-g could be

escalated from 7.8 (3-mg equivalents of IFN-g) to 78 mg (30-mgequivalents of IFN-g), with a clear dependence of tumor growthinhibition on the administered dose (Fig. 3A). None of theregimens used was toxic to mice (Fig. 3B).Biodistribution experiments were repeated in theWEHI-164

and LLCmodels, using pre-administration of unlabeled immu-nocytokine and comparing the tumor-targeting activity of F8-IFN-g with the corresponding negative control protein, KSF–IFN-g . In all cases, a preferential tumor accumulation wasobserved for F8–IFN-g . The tumor uptake (measured as%ID/g)was greater in F9 tumors, compared with the other two tumormodels (Fig. 4A, C, and E). In all three syngeneic tumor modelstested, the tumor growth inhibition of the targeted F8–IFN-gimmunocytokine was significantly greater (P < 0.0001; forstatistical analysis of therapy data, see Supplementary DataS2), compared with KSF–IFN-g (Fig. 4B, D, and F). At thechosen dose of 200 mg of F8–IFN-g (corresponding to 78-mgequivalents of IFN-g), no signs of toxicity were observed(Supplementary Data S3). A histologic analysis of tumor sec-tions following immunocytokine treatment, staining for vas-

cular structures (CD31), CD45þ leukocytes, CD4þ and CD8þ

lymphocytes, FoxP3 (a marker expressed in regulatory T cells),CD25þ lymphocytes, CD45Rþ cells (mainly B cells), AsialoGM1þ cells (mainly NK cells), and F4/80þ cells (mainly macro-phages) revealed an increased infiltration of leukocytes in thetargeted IFN-g treatment groups, while the infiltration ofFoxP3þ cells was decreased (Fig. 5).

Combination studies with immunocytokinesC57BL/6 mice bearing subcutaneous LLC tumors were used

to test whether the therapeutic activity of F8–IFN-g could beimproved by combining it with other immunocytokines. Whentumors were clearly palpable, mice were grouped and therapystarted by the injection of either PBS (buffer vehicle), F8–IFN-g(200 mg), F8–TNF (2 mg), or a mixture of F8–IFN-g and F8–TNF(administered together at the same dose used as single agents).Although F8–IFN-g had a more potent tumor growth retar-dation effect than F8–TNF and the single-agent treatment wasnot toxic (Supplementary Data S4), all mice in the combinationtreatment group died after the second injection (Fig. 6A andB). Figure 6C shows the therapy results obtained with 3injections of F8–IFN-g (200 mg), F8–IL-4 (90 mg), or the com-bination of both. Substantial tumor growth retardation wasobserved for the combination regimen compared with thesingle agents (P < 0.0001; for statistical analysis of therapydata, see Supplementary Data S2), but the treatment had no

Figure3. Dose-titration of F8–IFN-g in subcutaneousF9 teratocarcinoma-bearing mice. A, 4 days after tumor implantation, when tumors wereclearly palpable, mice were given 3 injections of either F8–IFN-g orPBS (as negative control) at different concentrations of the fusion protein(7.8 mg corresponding to 3 mg IFN-g ; 26 mg corresponding to 10 mg IFN-g ;78 mg corresponding to 30 mg IFN-g ) in different time intervals (every48 or 72 hours). B, assessment of toxicity was done by observation ofchanges in weight. Results are expressed as percentage weight changecompared with starting weight over time.






curative results (Fig. 6D). No signs of toxicity were observed(Supplementary Data S4).

DiscussionWhile studying the tumor-targeting properties of immuno-

cytokines, we observed that some fusion proteins based oncertain cytokines or growth factors (e.g., IL-2, IL-12, IL-15, TNF,GM-CSF, and VEGF-120) retain the tumor-targeting propertiesof the parental antibody, while others (e.g., those based on TATpeptides, calmodulin, or dual cytokine fusions) completelyabrogate tumor targeting in vivo, even though they were fullyimmunoreactive in vitro (1). We have reported that antibody–IFN-g fusions do not selectively localize to tumors in wild-typemice, but retain the tumor-targeting ability when injected

into mice defective for the IFN-g receptor (29). In this study,we show that injecting unlabeled IFN-g–based immunocyto-kine before injecting labeled fusion protein could overcome thetrapping by the IFN-g receptor, reduce the uptake of radiola-beled F8-IFN-g in the liver and the spleen, and allow theimmunocytokine to selectively accumulate at the tumor site.The clear dependence of the tumor growth inhibition profile onthe dose of the immunocytokine (Fig. 3) indicates that the IFN-g receptor has to be saturated before the development of theantitumor effect. Our data suggest that the receptor trapping ofF8–IFN-g can be overcome by the administration of sufficientamounts of the immunocytokine without any additional tox-icity at the doses used.

We have described therapy results obtained with the com-bination of IL-2– and IL-12–based immunocytokines (20), IL-2–

Figure 4. Targeting properties andanticancer activity of F8–IFN-g inthree different syngeneic mousemodels of cancer. Forbiodistribution experiments, micewere given 20 mg of F8–IFN-gfollowed by 15 mg of radioiodinatedprotein. Mice were sacrificed after24 hours. Organswere excised andradioactivity counted, expressingresults as%ID/g� SE. For therapyexperiments, tumor-bearing micewere injected with either PBS(negative control, buffer vehicle),200 mg KSF–IFN-g (untargetedIFN-g ; 78-mg equivalents of IFN-g ),or 200 mg F8–IFN-g (78-mgequivalents of IFN-g ) to a total of600 mg/mouse in 3 intravenousinjections, administered every 72hours. Results are expressed astumor volume � SEM (n ¼ 5). F9teratocarcinoma (A and B), WEHI-164 sarcoma (C and D), and LLC(E and F).

Hemmerle and Neri





and TNF-based immunocytokines (35), IL-4– and IL-2–basedimmunocytokines, and IL-4– and IL-12–based immunocyto-kines (30). The additive effects of F8–IL-4 plus F8–IFN-g aresurprising in terms of the opposite T-cell polarization proper-ties of the two cytokines. However, a possible synergy betweenTh1 and Th2 responses has been predicted previously (28, 36).We have described recently that receptor-trapping mechan-

isms of certain immunocytokines can be conveniently tested invitro by incubation of radiolabeled protein preparations withfresh blood, followed by centrifugation and counting. In thecase of F8–IFN-g , however, this method was not predictive forin vivo activity (Fig. 2D), possibly because of receptor expres-sion in the endothelium (37). Indeed, a quantitative biodis-tribution analysis performed at an early time point revealed anunusually high uptake of the immunocytokine in the liver andthe spleen, a process dependent on the presence of the IFN-gmoiety in the fusion protein.The doses of F8–IFN-g used in themouse to achieve efficient

tumor targeting and tumor growth inhibition were in the 10-mg/kg dose range. It is interesting to note that human IFN-g(Actimmune in the United States, Vidara Therapeutics; Imu-kin, outside the United States, Canada, and Japan, Boehringer

Ingelheim) is administered to patients at a recommended doseof 1.5 mg/kg. This discrepancy may reflect different activities ofthe cytokines among species, making clinical developmentmore difficult. Indeed, while we have demonstrated a strongsingle-agent tumor growth inhibition with F8–IFN-g in thethree syngeneic immunocompetent mouse models of cancertested, it remains to be considered whether the fully humanF8–IFN-g immunocytokine or the F8–IL-12 immunocytokinewould be the better candidate for clinical development (20, 38).Indeed, F8–IL-12 also targets tumors efficiently at low doses(�0.5 mg/kg) and is able to potently induce IFN-g overexpres-sion at the tumor site (18, 20). On the other hand, clinicaldevelopment of F8–IFN-g as a "me better" (biosuperior)product may be facilitated by the fact that recombinantIFN-g is an approved biopharmaceutical, while the industrialdevelopment of recombinant IL-12 has been stopped at thelevel of several phase II clinical trials. Irrespective of futureindustrial plans, the lessons learned in this article shed lighton the nature of immunocytokine receptor-trapping mechan-isms in vivo and suggest possible solutions, such as the use ofhigh doses or the use of mutants with decreased affinity to thereceptor.

Figure 5. Immunofluorescence analysis of tumor-infiltrating immune cells after therapy. Tumor-bearing mice were injected with 3 doses of PBS,KSF–IFN-g , or F8–IFN-g (every 72 hours, 200 mg). 10-mm sections of tumors were stained for CD31 (vascular control staining, Alexa Fluor 594; red) andfor different surface markers (Alexa Fluor 488; green). Magnification, �20; scale bar, 100 mm.






Disclosure of Potential Conflicts of InterestT. Hemmerle is a consultant/advisory board member for Philochem AG. D.

Neri has an ownership interest (including patents) and is a consultant/advisoryboard member for Philogen AG.

Authors' ContributionsConception and design: T. Hemmerle, D. NeriDevelopment of methodology: D. NeriAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): T. Hemmerle, D. NeriAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): T. Hemmerle, D. NeriWriting, review, and/or revision of the manuscript: T. Hemmerle, D. NeriAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): D. NeriStudy supervision: D. Neri

AcknowledgmentsThe authors thank Phillipp Probst for his help with the cloning procedure and

protein preparation.

Grant SupportThis study was financially supported by the Swiss National Science Foun-

dation, the ETH Zurich, the Commission for Technology and Innovation (CTI)Switzerland, the Swiss Cancer League, and the European Union (FP7 ProjectPRIAT).

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

Received October 16, 2013; revised February 27, 2014; accepted February 28,2014; published OnlineFirst March 24, 2014.

References1. Pasche N, Neri D. Immunocytokines: a novel class of potent armed

antibodies. Drug Discov Today 2012;17:583–90.2. Kontermann RE. Antibody–cytokine fusion proteins. Arch Biochem

Biophys 2012;526:194–205.3. Schrama D, Reisfeld RA, Becker JC. Antibody targeted drugs as

cancer therapeutics. Nat Rev Drug Discov 2006;5:147–59.4. Dela Cruz JS, Huang TH, Penichet ML, Morrison SL. Antibody–cyto-

kine fusion proteins: innovative weapons in the war against cancer.Clin Exp Med 2004;4:57–64.

5. Brack SS, Silacci M, Birchler M, Neri D. Tumor-targeting properties ofnovel antibodies specific to the large isoform of tenascin-C. ClinCancer Res 2006;12:3200–8.

6. Kaspar M, Zardi L, Neri D. Fibronectin as target for tumor therapy. IntJ Cancer 2006;118:1331–9.

7. Neri D, Bicknell R. Tumour vascular targeting. Nat Rev Cancer 2005;5:436–46.

8. Villa A, Trachsel E, Kaspar M, Schliemann C, Sommavilla R, Rybak JN,et al. A high-affinity human monoclonal antibody specific to the

Figure 6. Therapeutic activity of F8–IFN-g in combinationwith F8–TNF or F8–IL-4 against subcutaneous LLC. A, when tumorswere clearly palpable,micewererandomly grouped and injected intravenously either with PBS, 200 mg F8–IFN-g , 2 mg F8–TNF, or the combination of the single agents (200 mg F8–IFN-g with 2mg F8–TNF). Results are expressed as tumor volume � SEM. B, Kaplan–Meier plot of the combination therapy with F8–IFN-g and F8–TNF. C, combinationtherapy of F8–IFN-g (200 mg) with F8–IL-4 (90 mg). D, survival plot of the combination therapy of F8–IFN-g with F8–IL-4.


Hemmerle and Neri




alternatively spliced EDA domain of fibronectin efficiently targetstumor neo-vasculature in vivo. Int J Cancer 2008;122:2405–13.

9. Schwager K, Bootz F, Imesch P, Kaspar M, Trachsel E, Neri D. Theantibody-mediated targeted delivery of interleukin-10 inhibits endo-metriosis in a syngeneicmousemodel.HumReprod2011;26:2344–52.

10. Rybak JN, Trachsel E, Scheuermann J, Neri D. Ligand-based vasculartargeting of disease. ChemMedChem 2007;2:22–40.

11. Li J, Li X, Zhang Y, Zhou XK, YangHS, Chen XC, et al. Gene therapy forpsoriasis in the K14-VEGF transgenic mouse model by topical trans-dermal delivery of interleukin-4 using ultradeformable cationic lipo-some. J Gene Med 2010;12:481–90.

12. Pedretti M, Soltermann A, Arni S, Weder W, Neri D, Hillinger S.Comparative immunohistochemistry of L19 and F16 in non–small celllung cancer and mesothelioma: two human antibodies investigated inclinical trials in patients with cancer. Lung Cancer 2009;64:28–33.

13. Xia YP, Li B, Hylton D, Detmar M, Yancopoulos GD, Rudge JS.Transgenic delivery of VEGF to mouse skin leads to an inflammatorycondition resembling human psoriasis. Blood 2003;102:161–8.

14. Borsi L, Balza E, Carnemolla B, Sassi F, Castellani P, Berndt A, et al.Selective targeted delivery of TNFalpha to tumor blood vessels. Blood2003;102:4384–92.

15. Carnemolla B, Borsi L, Balza E, Castellani P, Meazza R, Berndt A, et al.Enhancement of the antitumor properties of interleukin-2 by its tar-geted delivery to the tumor blood vessel extracellular matrix. Blood2002;99:1659–65.

16. Frey K, Schliemann C, Schwager K, Giavazzi R, Johannsen M, Neri D.The immunocytokine F8-IL2 improves the therapeutic performance ofsunitinib in a mouse model of renal cell carcinoma. J Urol 2010;184:2540–8.

17. Gafner V, Trachsel E, Neri D. An engineered antibody–interleukin-12fusion protein with enhanced tumor vascular targeting properties. Int JCancer 2006;119:2205–12.

18. Halin C, Gafner V, Villani ME, Borsi L, Berndt A, Kosmehl H, et al.Synergistic therapeutic effects of a tumor targeting antibody fragment,fused to interleukin 12 and to tumor necrosis factor alpha. Cancer Res2003;63:3202–10.

19. Halin C, Niesner U, Villani ME, Zardi L, Neri D. Tumor-targetingproperties of antibody-vascular endothelial growth factor fusion pro-teins. Int J Cancer 2002;102:109–16.

20. Pasche N, Wulhfard S, Pretto F, Carugati E, Neri D. The antibody-based delivery of interleukin-12 to the tumor neovasculature eradi-cates murine models of cancer in combination with Paclitaxel. ClinCancer Res 2012;18:4092–103.

21. KasparM, Trachsel E, Neri D. The antibody-mediated targeted deliveryof interleukin-15 and GM-CSF to the tumor neovasculature inhibitstumor growth and metastasis. Cancer Res 2007;67:4940–8.

22. Pasche N, Frey K, Neri D. The targeted delivery of IL17 to the mousetumor neo-vasculature enhances angiogenesis but does not reducetumor growth rate. Angiogenesis 2012;15:165–9.

23. PascheN,WoytschakJ,WulhfardS, Villa A, FreyK,Neri D.Cloning andcharacterization of novel tumor-targeting immunocytokines based onmurine IL7. J Biotechnol 2011;154:84–92.

24. Schwager K, Kaspar M, Bootz F, Marcolongo R, Paresce E, Neri D,et al. Preclinical characterization of DEKAVIL (F8-IL10), a novel clinical-stage immunocytokine which inhibits the progression of collagen-induced arthritis. Arthritis Res Ther 2009;11:R142.

25. Hemmerle T, Wulhfard S, Neri D. A critical evaluation of the tumor-targeting properties of bispecific antibodies based on quantitativebiodistribution data. Protein Eng Des Sel 2012;25:851–4.

26. Curnis F, Gasparri A, Sacchi A, Cattaneo A, Magni F, Corti A. Targeteddelivery of IFNgamma to tumor vessels uncouples antitumor fromcounterregulatory mechanisms. Cancer Res 2005;65:2906–13.

27. Mizokami MM, Hu P, Khawli LA, Li J, Epstein AL. Chimeric TNT-3antibody/murine interferon-gamma fusion protein for the immunother-apy of solid malignancies. Hybrid Hybridomics 2003;22:197–207.

28. Quezada SA, Peggs KS, Simpson TR, Allison JP. Shifting the equi-librium in cancer immunoediting: from tumor tolerance to eradication.Immunol Rev 2011;241:104–18.

29. Ebbinghaus C, Ronca R, Kaspar M, Grabulovski D, Berndt A, KosmehlH, et al. Engineered vascular-targeting antibody–interferon-gammafusion protein for cancer therapy. Int J Cancer 2005;116:304–13.

30. Hemmerle T, Neri D. The antibody-based targeted delivery of inter-leukin-4 and 12 to the tumor neovasculature eradicates tumors in threemouse models of cancer. Int J Cancer 2014;134:467–77.

31. Hemmerle T, Probst P, Giovannoni L, Green AJ, Meyer T, Neri D. Theantibody-based targeted delivery of TNF in combination with doxo-rubicin eradicates sarcomas in mice and confers protective immunity.Br J Cancer 2013;109:1206–13.

32. Hoogenboom HR, Winter G. By-passing immunisation. Human anti-bodies from synthetic repertoires of germline VH gene segmentsrearranged in vitro. J Mol Biol 1992;227:381–8.

33. Borsi L, Castellani P, AllemanniG, Neri D, Zardi L. Preparation of phageantibodies to the ED-A domain of human fibronectin. Exp Cell Res1998;240:244–51.

34. Farrar MA, Schreiber RD. The molecular cell biology of interferon-gamma and its receptor. Annu Rev Immunol 1993;11:571–611.

35. Schwager K, Hemmerle T, Aebischer D, Neri D. The immunocytokineL19-IL2 eradicates cancer when used in combination with CTLA-4blockade or with L19-TNF. J Invest Dermatol 2013;133:751–8.

36. Quezada SA, Peggs KS. Tumor-reactive CD4þ T cells: plasticitybeyond helper and regulatory activities. Immunotherapy 2011;3:915–7.

37. JanewayC. Immunobiology: the immune system inhealth anddisease.6th ed. New York: Garland Science; 2005. xxiii, 823.

38. Sommavilla R, Pasche N, Trachsel E, Giovannoni L, Roesli C, Villa A,et al. Expression, engineering and characterization of the tumor-targeting heterodimeric immunocytokine F8-IL12. Protein Eng DesSel 2010;23:653–61.






Published OnlineFirst March 24, 2014.Cancer Immunol Res Teresa Hemmerle and Dario Neri

In VivoMechanism Fusion Proteins Reveals an Unexpected Receptor-Trapping

γIFN-−The Dose-Dependent Tumor Targeting of Antibody

Updated version

10.1158/2326-6066.CIR-13-0182doi:

Access the most recent version of this article at:

Material

Supplementary

.DC1

http://cancerimmunolres.aacrjournals.org/content/suppl/2014/03/24/2326-6066.CIR-13-0182Access the most recent supplemental material at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://cancerimmunolres.aacrjournals.org/content/early/2014/05/02/2326-6066.CIR-13-0182To request permission to re-use all or part of this article, use this link



http://cancerimmunolres.aacrjournals.org/lookup/doi/10.1158/2326-6066.CIR-13-0182

http://cancerimmunolres.aacrjournals.org/content/suppl/2014/03/24/2326-6066.CIR-13-0182.DC1

http://cancerimmunolres.aacrjournals.org/content/suppl/2014/03/24/2326-6066.CIR-13-0182.DC1

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

mailto:[email protected]

http://cancerimmunolres.aacrjournals.org/content/early/2014/05/02/2326-6066.CIR-13-0182


the dose-dependent tumor targeting of antibody ifn-g fusion … · 2014. 5. 2. · tumor therapy...

Documents