targeting of interleukin 2 to human ovarian carcinoma by fusion … · targeting of interleukin 2...

[CANCER RESEARCH 58, 4146-4154, September 15, 19981

Targeting of Interleukin 2 to Human Ovarian Carcinoma by Fusion with a Single-Chain Fv of Antifolate Receptor Antibody1

Cecilia Molimi,2 Mariangela Figini, Daniela Nicosia, Elena LuisÃ³n, Venkatesh Ramakrishna,3 Giulia Casorati,

Giorgio Parmiani, Zelig Eshhar, Silvana Canevari, and Mario P. ColomboDivision of Experimental Oncology "D" 1C. M., D. N., V. R., G. P., M. P. C.j and "E" [M. F.. E. L, S. C.], Istillilo Nazionale Tumori, 20133 Milan, Italy; Imnmnochemistry Unit,

DIBIT. H. S. Raffaele. 20132 Milan, Italy Â¡G.C.Â¡;and Department of Immunology, The Weizmann Institute of Science, 76100 Rehovot. Israel [Z. E.Â¡

ABSTRACT

To provide a new tool for the immunotherapy of human ovariancarcinoma, we constructed a fusion protein between interleukin-2 (IL-2)and the single-chain Fv (scFv) of MOV19, a monoclonal antibody directedagainst Â»-folate receptor (a-FR), known to be overexpressed on human

nonmucinous ovarian carcinoma. This was accomplished by fusing thecoding sequences in a single open reading frame and expressing theIL-2/MOV19 scFv chimera under the control of the murine immunoglob-

ulin K promoter in J558L plasmacytoma cells. The design allowed theconstruction of a small molecule combining the specificity of MOV19 withthe immunostimulatory activity of IL-2. This might improve the tissue

penetration and distribution of the fusion protein within the tumor,reduce its immunogenicity, and avoid the toxicity related to the systemicadministration of IL-2. The IL-2/MOV19 fusion protein was stable on

purification from the cell supernatant and was biologically active. Importantly, this construct was able to target IL-2 onto the surface of a-FR-overexpressing tumor cells and stimulated the proliferation of the IL-2-dependent CTLL-2 cell line as well as that of human resting peripheralblood lymphocytes. In a syngeneic mouse model, IL-2/MOV19 scFv specifically targeted Â»-FR gene-transduced metastatic tumor cells without

accumulating in normal tissues, due to its fast clearance from the body.Prolonged release of IL-2/MOV19 scFv by in vivo transplanted JS58-

EF6.1 producer cells protected 60% of mice from the development of lungmÃ©tastasescaused by an i.v. injection of a-FR gene-transduced tumorcells. Moreover, treatment with IL-2/MOV19 scFv, but not with recombinant IL-2, significantly reduced the volume of s.c. tumors. The phar-macokinetics and biological characteristics of IL-2/MOV19 scFv might

allow us to combine the systemic administration of this molecule with theadoptive transfer of in vitro retargeted T lymphocytes for the treatment ofovarian cancer, thereby providing local delivery of IL-2 without toxicity.

INTRODUCTION

IL-24 has been used for the immunotherapy of a variety of human

malignancies; however, its rapid clearance, severe toxicity whengiven at high doses, and suboptimal concentration at the tumor sitehave limited its efficacy in vivo. Genetic manipulation of tumor cellsor fibroblasts allowed IL-2 to be secreted in situ. In vivo, suchIL-2-secreting cells induce an inflammatory reaction that may result

in the elimination of tumor cells, the generation of a systemic immuneresponse, and immunological memory (1-3). However, data fromanimal models indicate that immunotherapy with IL-2 gene-trans

duced cells is only effective for limited tumor loads, suggesting its

Received 5/1/98: accepted 7/17/98.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1Supported by the Associazione Italiana Ricerca sul Cancro and the Italian Ministry

of Health.2 To whom request for reprints should be addressed, at Division of Experimental

Oncology "D," Istituto Nazionale Tumori, via Venezian 1, 20133 Milan, Italy. E-mail:

melani @istitutotumori.mi.it.3 Present address: Argonex Pharmaceuticals, Inc., 706 Forrest Street, Charlottesville,

VA 22903.4 The abbreviations used are: IL, interleukin; scFv. single-chain Fv; mAb, monoclonal

antibody: a-FR, a-folate receptor; TAA, tumor-associated antigen; Ab. antibody; HPLC,high-performance liquid chromatography; PBL, peripheral blood lymphocyte; bsAb,

bispecific AB; %ID/g, the percentage of injected dose per gram of tissue.

optimal use in the treatment of minimal residual disease. Furthermore,on gene transduction, IL-2 secretion may vary according to the vector

used, the strength of the promoter driving the transgene expression,and the efficiency of protein synthesis in the targeted cell.

The selective delivery of IL-2 at the tumor site would avoid the

toxicity associated with systemic administration, overcome the concerns due to ex vivo genetic manipulation of autologous or allogeneiccells, and solve the difficulties of transducing tumor cells in vivo.Taking advantage of the targeting specificity of mAbs directed toTAAs, fusion proteins have been constructed to allow IL-2 targeting

at the tumor microenvironments and the activation of antitumor immune responses in murine models (4-8). Targeting of IL-2 to tumor

sites is expected to sustain in situ the antitumor cytotoxicity of both Tlymphocytes activated by direct immunization and lymphocytes infused in the context of adoptive immunotherapy.

Among human tumors, ovarian carcinoma has insidious dissemination patterns and is often diagnosed at advanced stages. Althoughplatinum combination chemotherapy improved the response rate ofthese patients, satisfactory long-term survival has not been achieved

(9). Immunotherapeutic approaches can be envisioned for these ovarian carcinoma patients, because their tumors express a panel of knownTAAs, thereby providing a target for antitumor vaccination (10, 11).Moreover, mAbs have been raised against ovarian carcinoma antigensthat can be used for delivery drugs and radionuclides for eitherdiagnosis or therapy (12-14). Finally, the natural course of the diseaseincludes a long-term phase of i.p. growth, providing the rationale for

locoregional immunotherapy. After the chemotherapy of refractoryovarian cancer, immunotherapy treatment with mAbs or tumor necrosis factor can lead to the reduction of malignant ascites (15). Partialresponses were obtained when IL-2 was given i.p. in combinationwith lymphokine-activated killer cells or tumor-infiltrating lympho

cytes. In fact, signs of immune activation such as an increase ofCD8+, CD16+, and CD56+ cells and the secretion of IFN-y, granu-

locyte macrophage colony-stimulating factor, and IL-IO in the peri

toneal cavity were observed in these patients as well as in patientstreated with i.p. IL-2 only (16-18).

Recently, i.p. tumor regression was achieved after adoptive immunotherapy with T lymphocytes retargeted in vitro by the bsAb OC/TR(19) directed to CD3 and to the a-FR, a MT,38,000 glycoprotein TAA

overexpressed in more than 90% of human nonmucinous ovariancarcinomas (12). The need for IL-2-dependent preactivation of Tlymphocytes in vitro provides the rationale for targeting IL-2 to tumor

sites to sustain the activation and function of retargeted effector cellsin a combinatorial immunotherapy. To this end, we constructed afusion protein between human IL-2 and the scFv of mAb MOV 19 thatspecifically recognizes the a-FR. MOV 19 is not internalized onbinding to the a-FR (20) and therefore allows the IL-2 portion of the

fusion protein to remain on the cell surface.Here we show that the IL-2/MOV19 scFv fusion protein was

expressed at high levels in murine myeloma cells, and that oncepurified, it exhibited the functional activity of its two components, i.e.specific binding to a-FR-expressing tumors and the stimulation oflymphocyte proliferation. To assess the activity of IL-2/MOV19 scFv

4146

Research. on October 19, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

ANTIFOLATE RECEPTOR scFv/IL-2 FUSION PROTEIN

in vivo, a murine tumor model was prepared by the retroviral trans-duction of the gene coding for the human a-FR into murine coloncarcinoma C26 (C26/a-FR). The pharmacokinetics and biodistribu-tion of IL-2/MOV19 scFv showed a rapid clearance after i.p. administration and a specific retention within C26/a-FR but not in C26 or innormal tissues. Moreover, IL-2/MOV19 scFv impaired C26/a-FR

outgrowth, thus suggesting that a therapeutic application of this reagent is warranted.

MATERIALS AND METHODS

Construction and Expression of IL-2/MOV19 scFv. The full-length human IL-2 cDNA sequence was amplified by PCR from the Lh2SN vector (21)with primers containing the suitable restriction sites ///ndIII-/faÂ«nHI and

cloned into the pCDNA3 vector (Invitrogen, Leek, the Netherlands). Thesequences encoding the variable chains of mouse niAb MOV 19 were reverse-

transcribed from the producing hybridoma, amplified, and assembled to encode the scFv fragment using PCR essentially as described previously (22. 23).The assembled MOV 19 scFv was cloned into pHENl to be screened by phagepanning on cells overexpressing the a-FR (23), and then the full-length DNA

coding sequence was amplified with primers containing the EcoR\ and Xholsites and cloned 3' to the IL-2 sequence into the corresponding sites of vector

pCDNA3. The synthetic linker GGGTCGACTTCCGGTAGCGGCAAATC-CTCTGAAGGCAAAGGT was then ligated in-frame into the BÂ«mHI-Â£roRIsites between IL-2 and MOV 19 scFv to encode a 13-amino acid spacer that hasbeen shown to facilitate the correct folding of the fusion protein.5 The entire

open reading frame was sequenced and transfected into J558L myeloma cells(24) by protoplast fusion. Clones of G4l8-resistant J558L cells were screenedfor the secretion of IL-2/MOV19 scFv by ELISA for human IL-2 (Duoset kit:Genzyme. Cambridge. MA) according to the manufacturer's instructions. To

increase the production of the fusion protein, the expression cassette wasfurther modified and cloned into the pH PCRy III vector (25). This expressionvector encodes the constant portion of murine K light chain (CÂ«)under theimmunoglobulin K promoter and contains the murine immunoglobulin H andimmunoglobulin K enhancers. Using an appropriate PCR primer, the stopcodon was removed from the 3' end of the MOV 19 Vk and substituted with a

splicing consensus sequence to ensure the fusion between IL-2/MOV19 scFvand CK. The modified cassette was sequenced. blunt-ended, and cloned into the

pH PCRylll vector. Vector pHIL2/MOVSD was transfected into J558L cellsby protoplast fusion, and cells were selected by G4I8. The supernatants ofresistant cells were screened by ELISA for human IL-2. and the most positivecolonies were subcloned by limiting dilution to obtain two high-expressingclones. J558-EF6.1 and J558-EF6.2.

The molecular si/,e and immunoreactivity of the fusion protein in thesupernatant of the expressing clones were tested by Western blotting. Totalsupernatant proteins (50 jug) were subjected to 10% SDS-PAGE and electro-

blotted onto Immobilen membranes (Millipore Corp.. Bedford. MA). Filterswere hybridized to biotinylated goat antimurine CK Ab (Southern Biotechnology Associates. Birmingham. AL; [0.1 /Â¿g/ml])or rabbit antihuman IL-2 Ab

(Gen/.yme; 10 fig/ml) in 5% BSA/PBS/0.1 % Tween for 1 h, washed severaltimes in PBS/0.1% Tween 20. and incubated for l h with either Avidin BiotinComplex (Vector Laboratories. Burlingame. CA) or horseradish peroxidase-

conjugated goat antirabbit Ab (0.05 /Ag/ml) in 5% BSA/PBS/0.1 % Tween 20.After extensive washing, the filters were treated with enhanced chemilumi-

nescence detection solutions (Amity: Amersham Life Science. Buckinghamshire. United Kingdom) and exposed to Hyperfilm (Amity: Amersham LifeScience) from 10 s to 5 min.

Functional Assays for IL-2/MOV19 scFv. The IL-2 biological activity ofIL-2/MOV19 scFv was tested in the supernatant of transduced J558L cells and

as a purified fusion protein by the ability to sustain the proliferation ofIL-2-dependent CTLL-2 cells. A serial dilution of supernatant from the J558-EF6.1 and J558-EF6.2 clones or purified fusion protein and recombinanthuman IL-2 (Proleukin; Chiron Corp., Emeryville) were incubated for 48 hwith 5 x 10' CTLL-2 cells that had been starved of IL-2. One /j.Ci of['Hlthymidine (Amity; Amersham Life Science) was added to 200 Â¡Ãºof cell

s C. Melani. unpublished observations.

culture for the last 18 h. and cell proliferation was measured by |'H]thymidine

incorporation.The MOV 19 binding reactivity of the IL-2/MOVI9 scFv fusion protein was

assayed by immunofluorescence on tumor cells expressing or not expressingthe a-FR; 5 x 10s cells were incubated for 1 h on ice with serial dilutions of

supernatant and with 5 /Â¿g/mlpurified IL-2/MOV19 scFv or MOV 19 mAb and

diluted in 2% BSA/PBS/0.05% sodium azide. Cells were then washed fourtimes with the same buffer and incubated with either biotinylaled untimouscCK (Southern Biotechnology Associates) or biotinylated antihuman IL-2 (Gen-

zyme) at 2.5 /ng/ml for 1 h on ice. After repeated washing, the cells wereincubated for 40 min with streptavidin-phycoeritrin conjugate (5 /ig/ml:

Pharmingen. San Diego. CA) in ice. washed, and analyzed using a FACScan(Becton Dickinson). For competition assays. 100 /Â¿g/mlchimeric MOV 19 (26)were added to the IL-2/MOV19 scFv in the first incubation, and hiotinylatcd

antimouse CK was used as the secondary Ab.Purification of IL-2/MOV19 scFv Fusion Protein. The rat antimouse CK

mAb 187.1 (ATCC number HB58) was purified from hybridoma supernatantby affinity chromatography on Hi Trap Protein G (Pharmacia Biotech. Uppsala. Sweden) and coupled to CNBr-activated Sepharose 4B (Pharmacia Biotech) according to the manufacturer's instructions. The EF6.1 cell supernatant

was dialyzed against 0.1 M Tris-HCl (pH 8.0). filtered, and loaded on the

column. The fusion protein was eluted by acid buffer, equilibrated withTris-HCl (pH 9.5), and dialyzed extensively in 0.9 g/liter NaCI or in 3.4 mMPBS-EDTA. After purification, the fusion protein was analyzed on SDS-PAGE

and by HPLC gel filtration on a SuperÃ³se 12HR 10/30 column (PharmaciaBiotech), and its activity was assayed as described above for the cell supernatant.

Murine Tumor Model and Cell Lines. To analyze the in vivo antitumoreffect of the fusion protein in a syngeneic context, the C26 udenocarcinomu ofthe BALB/c strain was transduced with the human a-FR gene by means of aretroviral vector to obtain the C26/a-FR clone. A 910-bp Â£cÂ»RIfragmentencoding the human a-FR was excised from pBluescript KS+ (27) and cloned

into the corresponding site of retroviral vector LXSN to obtain the Lgp38SNvector. Recombinant retroviral particles were obtained by the transtection ofthe amphotropic packaging cell line gp+AM12 and the infection of theecotropic packaging cell line gp+E86 as described previously (21). Subcon-

fluent plates of the murine adenocarcinoma cell line C26 were infected withviral supernatant in the presence of polybrene as described previously (21 ). andcolonies of transduced cells were isolated by G418 selection and subcloned bylimiting dilution to obtain clones displaying homogeneous expression of thea-FR. as measured by immunofluorescence using mAb MOV 19. Clone 5A6

that expressed the transgene al a high level was chosen for additional experiments and was designated C26/a-FR.

The cell lines INT.Ov-1. INT.Ov-2, INT.Ov-3. and INT.Ov-7 were established in our laboratory as described previously (II) from primanÂ»'and meta-static ovarian carcinomas that were either serous or mucinous. IGROV-I cells

were kindly provided by Dr J. BÃ©nard(Institute Gustave Roussy, Villejuife.France: Ref. 28); SK-CO-1 and HT-29 colon carcinoma cell lines as well as theSK-BR3 breast carcinoma cell line were from the American Type CultureCollection. a-FR expression was assayed by immunofluorescence using the

MOV 19 mAb.PBLs were collected from normal donors and stored frozen until used. All

human tumor cell lines were maintained in RPMI 1640 supplemented with10% PCS or 10% pooled human AB serum, whereas murine cell lines weregrown in DMEM supplemented with 5% FCS. and transduced cells weregrown in DMEM supplemented with 0.8 mg/ml G418.

Tumor Targeting and IL-2 Bioactivity in Vitro. Irradiated (20.IXK) rads)tumor cells with or without a-FR expression were incubated with purifiedIL-2/MOV19 scFv (5 /Â¿g/ml)for 30 min on ice. washed repeatedly with freshmedium, plated in triplicate in serial dilutions into 96-well plates, and incubated with or without 5 X 10' CTLL-2 cells for 48 h, with the addition of 1/xCi of |'H|thymidine for the last 18 h. CTLL-2 proliferation was measured as:

cpm incorporated by CTLL-2 - cpm incorporated by irradiated tumor cells.

Lymphocyte Proliferation Assay. Irradiated (20.000 rads) INT.Ov-2 andINT.Ov-3 ovarian carcinoma cells were incubated or not incubated withIL-2/MOVI9 scFv as described above. After repeated washing with freshmedium. 10J treated and untreated tumor cells were plated in triplicate into

round-bottomed 96-well plates, and Iflr human lymphocyte antigen-matchedPBLs were added to each well (E:T. 10:1); 100IU/ml recombinant human IL-2

4147




or IL-2/MOV19 scFv or 1 /xg/ml of phytohemagglutinin were added to controlwells. Cell proliferation was measured with an 18-h [3H]thymidine incorpo

ration assay after 5 days of culture.Pharmacokinetics and Biodistribution. The purified IL-2/MOV19 scFv

was labeled with 12Ã•Ias described previously (29) at a specific activity of 4

/*Ci//j.g. Mice received Lugol solution (0.02% I,) and KC1O4 (0.6 mg/ml) intheir drinking water starting 3 days before radiolabeled fusion protein administration and throughout the experiments to block free iodine uptake by thethyroid gland and stomach mucosae. Groups of three BALB/c mice wereinjected i.p. with a single dose of I25l-labeled IL-2/MOV19 scFv diluted in 150/al of 0.9% (w/v) saline (dose. 2.08 X 10~" mol; activity, 4.9 /Â¿Ci).Blood

samples were drawn from the retroorbital sinus and allowed to clot for 30 minat 37Â°C, and the sera were collected by cemrifugation. Pharmacokinetic

analysis was done with serum samples as described previously (30). Thepharmacokinetics of purified IL-2/MOV19 scFv were also assayed by injecting

10 /ig of fusion protein i.p. into BALB/c mice and collecting blood at differenttime intervals. The amount of IL-2/MOV19 scFv was measured in mouseserum with an ELISA for human IL-2. Pharmacokinetic parameters were

calculated by using PCNONLIN software (SCI Software, Lexington KY).The biodistribution of IL-2/MOV19 scFv was similarly tested in BALB/c

mice bearing 7-day lung mÃ©tastasesof C26/a-FR or C26 obtained by i.v.injection of 5 X IO5tumor cells. At 2, 6, 18. 24, and 30 h after the i.p. injection

of radiolabeled IL-2/MOV19 scFv, groups of three mice were sacrificed by

cervical dislocation, and their organs were weighed and counted for radioactivity. The %ID/g was calculated, and the fusion protein distribution wasexpressed as the ratio between the %ID/g of tissue and blood.

In Vivo Treatment with IL-2/MOV19 scFv. To obtain a continuous

release of the fusion protein, groups of 10 BALB/c mice injected s.c. with5 x IO4 or i.v. with IO4 C26 or C26/a-FR cells received IO6 live J558-EF6.1

producer cells s.c. in the opposite flank once or three times on days 1, 7, and14. The mice were monitored three times a week for tumor growth andsurvival, as described below. Differences in survival were analyzed by thelog-rank test and were considered significant when P < 0.05.

Treatment with purified IL-2/MOV19 scFv was performed in groups of fiveBALB/c mice each injected s.c. with 5 X IO4 C26 or C26/a-FR cells in the left

flank. Starting after 6 h, 10 ^g of IL-2/MOV19 scFv in PBS were injected i.p.

daily. 5 days/week for 3 weeks, whereas control mice received the same

volume (200 fil) of PBS or equivalent International units of recombinant IL-2.

The mice were monitored three times a week for tumor growth by measuringthe two major tumor diameters with a caliper. Tumor volume was expressed as:(minor diameter)2 X major diameter. The results of the two separate experi

ments were pooled, and the statistical significance was determined by Student's t test; differences in tumor volumes were considered statistically sig

nificant when two-tailed P were S0.05.

RESULTS

Expression and Purification of IL-2/MOV19 scFv. Our firstattempt to express the fusion protein IL-2/MOV19 scFv in J558L cells

by the pCDNA3 vector under the control of the cytomegaloviruspromoter yielded a low but biologically active (data not shown)amount of secreted fusion protein (range, 0.1-0.12 fig/ml, as measured by an ELISA for human IL-2).

To improve the yield, the pH PCR7 III vector that contains theimmunoglobulin K promoter, both K and H enhancers, and the murineCK sequence was modified to give pHIL2/MOV19SD (Fig. IA). TheCK tail allows easy purification of the fusion protein by an immuno-affinity column charged with anti-CÂ«antibody. Two clones of trans-fected J558L cells secreted a high amount of IL-2/MOV19 scFv (upto 3.5 ju.g/ml/48 h) as assayed by the IL-2 ELISA. Western blotanalysis performed with anti-IL-2 and anti-CK Ab revealed a mono-

meric protein of Mr 52,500 that partially dimerized under nonreducingconditions (Fig. 1, B and C). The dimerization of the fusion proteinmay occur due to free SH present on CÂ«and also on IL-2 (Cys-125).The secreted fusion protein was shown to bind the a-FR antigenexpressed on IGROV-1 cells in the immunofluorescence assay anddisplayed IL-2 activity in a CTLL-2 proliferation assay (data not

shown).IL-2/MOV19 scFv was purified from the supernatants of the J558-

EF6.1 producer clone with an immunoaffinity column using ananti-CK antibody, and the yield ranged from 2-3.5 mg of protein/liter

IL-E MOV19 scFv SD Â«A stopspacer I

"9 VHiVkV ^ XX1w- irN

k enhancer H enhancer

anti IL-2 anti CkkD64 -

50 -

36 -

98 -64 -

50 -

B 15-

11.25

1234 1 3 4

22.50

Time (min)33.75 45.00

Fig. 1. A. schematic representation of the expression vector pHIL2/MOVSD. The arrow indicates the transcriptional start from the immunoglobulin x promoter. â€¢represents thehuman IL-2 leader sequence, and SD and SA are the splicing donor and acceptor sites, respectively. B and C Western blot analysis of the cell supernatants under reducing (B) andnonreducing (C) conditions revealed by antihuman IL-2 {left panels) and antimurine CK mAbs (right panels). Lane I, transfected J558EF6 cells: Lane 2, clone J558EF6.1 : Lane 3. cloneJ558EF6.2; Lane 4, nontransfected parental J558L cells. The molecular size marker is indicated. D. HPLC profile of purified IL-2/MOVI9 scFv. The fusion protein eluted from theaffinity chromatography column and stored in PBS-EDTA (3.4 mM) was analyzed on a SuperÃ³se 12HR 10/30 column. The first peak corresponds to the fusion protein dimers (11%),and the second peak corresponds to the monomeric form of IL-2/MOV19 scFv (89%). The calibration of the column was done with the gel filtration standard (molecular weight markers

ranging from M, 1.350 lo M, 670.000) from BioRad.

4148




mO

Eo.o

125-1

100-

75-

50-

25-

Concentration

C26/a-FR

(pmoles/ml)

IGROV-1

0>O

10Â° 10' 10s 103 10'VA

10Â° 10' 102 1D3 10*

10Â° 10' 102 10' IO- 10' 102 10a 10'

Log Fluorescence Intensity

Fig. 2. A, biological activity of the IL-2 portion of the fusion protein measured byCTLL-2 cell proliferation. â€¢.purified IL-2/MOV19 scFv. O, recombinant human IL-2(Proleukin). B. antigen recognition by purified IL-2/MOV19 scFv fusion protein, a-FR-transduced murine tumor cells (C26/a-FR) and antigen-positive human ovarian carcinomacells (IGROV-1) were equally recognized by the MOV19 mAb (boli! line) and the fusionprotein (Â¡hinline), as revealed by using an antimurine CK Ab as a secondary Ab. Dolledline, negative control. C, the binding of IL-2/MOV19 scFv to antigen-positive tumor cellsdetected by an antihuman IL-2 as a secondary Ab. Dotted Â¡ine,negative control.

A (Fig. 2A). Purified IL-2/MOV19 scFv and mAb MOV19 were compared for binding to human carcinoma cell lines expressing variableamounts of the a-FR antigen. Immunofluorescence analysis showedno difference between the two reagents in the percentage of positivityand the mean intensity of fluorescence (Table 1; Fig. IB). The bindingof IL-2/MOV19 scFv to the a-FR antigen could be displaced by anexcess (20x) of chimeric mAb MOVI9, whose human constantportion is undetectable by the antimurine CK used as the secondaryAb. Competition for antigen binding was less evident in C26/a-FR;here, the percentage of positivity remained unchanged, although themean intensity of fluorescence was reduced by 1 log, because thetransgene expression was extremely high in these cells.

Moreover, the binding of IL-2/MOV19 scFv but not that of MOV 19to the cell surface was detectable by an antihuman IL-2 Ab used as asecondary Ab (Fig. 2C).

Lymphocyte Proliferation Assay. The IL-2-dependentCTLL-2cell line was cocultured with antigen-positive or -negative cells thatwere previously incubated with purified IL-2/MOV19 scFv. On antigen binding and repeated washing, IL-2/MOV19 scFv remainedbound to the surface of a-FR-expressing cells and efficiently stimulated CTLL-2 cell proliferation (Fig. 3/4), proving that fusion proteinbound to the surface of antigen-positive cells could stimulate the

D high-affinity IL-2 receptor on CTLL-2 cells.We also tested whether fusion protein bound to tumor cells could

stimulate human PBLs through the binding of its IL-2 portion. Lymphocyte proliferation was induced by the antigen-positive INT.Ov-2cells coated with IL-2/MOV19 scFv, but neither untreated INT.Ov-2nor antigen-negative INT.Ov-3 preincubated with the fusion proteinwas able to sustain lymphocyte proliferation (Fig. 35). Positive controls of lymphocyte proliferation were the fusion protein in solution,recombinant IL-2, or phytohemaglutinin (data not shown).

f\ Pharmacokinetics and Biodistribution. I25l-labeled IL-2/

MOV 19 scFv was used for pharmacokinetics studies. The fusionprotein was injected into mice, and the i.p. route was chosen in viewof the prospective treatment of ovarian carcinoma patients by i.p.administration. The iodinated IL-2/MOV19 scFv reached the bloodstream rapidly (5 min), peaked at 1h, and was completely secreted by24 h. The clearance curve of radiolabeled IL-2/MOV19 scFv wasbiphasic, fitting a two-compartment model [a very rapid absorptionphase (r1/2a = 5 min) and a slower elimination phase (fl/2|3 = 105.45min)]. Similar results were obtained by measuring the amount ofhuman IL-2 in the serum after i.p. injection of 10 /Â¿gof purifiedIL-2/MOV19 scFv (data not shown).

In tumor-bearing animals, the fusion protein did not accumulate in

of supernatant. When purified IL-2/MOV19 scFv was eluted in salineor PBS, the HPLC gel filtration revealed the presence of variableamounts of dimers and high molecular weight aggregates that reached50% of the total proteins and increased with time, especially onstorage at 4Â°C(data not shown). However, the addition of 3.4 mM

EDTA to PBS during dialysis and storage resulted in the best yield ofmonomeric fusion protein, as shown by the HPLC profile in Fig. ID(range, 82-86%).

Characterization and Function of IL-2/MOV19 scFv. The IL-2bioactivity of the purified fusion protein was tested several times in aCTLL-2 proliferation assay and compared with that of recombinanthuman IL-2. We found that 1 pg of purified IL-2/MOV19 scFvcorresponded to 7,500-12,500 IU of recombinant human IL-2 (Proleukin), depending on the different batches of purified fusion protein

Table 1 Comparison between 1L-2/MOV19 scFv and MOV19 mAb for antigenrecognition

a-FR expression was analyzed in a panel of tumor cell lines using either purifiedIL-2/MOV19 scFv fusion protein or MOV 19 mAb at 5 /ig/ml and an antimurine CK as a

secondary Ab. Competition for antigen binding was tested by adding an excess (20X) ofchimeric MOV 19 mAb to the fusion protein and assessing the binding with an antimurineCK secondary Ab that is unable to bind to the chimeric MOV 19.

CelllineIGROV-I

INT.Ov-1INT.Ov-2INT.Ov-3INT.Ov-7HT-29SK-CO-1C26C26/a-FRMOV

19mAb99.14"(112. 1/'

99.72(195.1)85.9 (258.7)0.96(21.3)2.9(71.6)

67.2 (45.2)63.6 (49.4)

0.45 (25.8)99 (437.6)IL-2/MOV19

scFv99.5(112.6)

100(225.5)87.2 (280.66)

0.6(41.4)2.6(71.4)

67.4 (47.6)64.3 (50.7)

1.2(14.1)99.9 (352.7)IL-2/MOV19

scFv + chimericMOV1921.85(16.3)

13.17(18.06)6.7 (29.6)

NDr

2.4 (70.4)8.4 (36.8)

10.2 (30.6)ND

93.8 (36.04)" Percentage of cell positivity.h Mean intensity of fluorescence.' ND, not done.

4149




Fig. 3. Tumor binding activity and lymphocytestimulation by IL-2/MOV19 scFv. A, IL-2-dependentCTLL-2 cell proliferation stimulated by the fusion protein bound on the cell surface of antigen-positive tumor

cells (see also Table 1), but not by untreated or bytreated but antigen-negative tumor cells. ^, pretreatment with fusion protein (5 /Â¿g/ml);â€¢,no treatment. B,human ovarian carcinoma cells targeted with IL-2/MOV 19 scFv induce or do not induce the proliferationof human resting PBLs according to their or-FR expression (inset). Tumor cell lines pretreated or not pre-treated with the fusion protein and freshly isolatedhuman lymphocyte antigen-matched PBLs were mixedat a 1:10 ratio and cocultured for 5 days, la, pretreatment with fusion protein (5 fig/ml); â€¢no treatment.

Q.

i2o.

Ã¼

150000-1

100000-

50000-

cell line

cc-FR

INT.Ov-2 INT.Ov-3 SKBR-3 IGROV-1 C26

Stimulating tumor cells

C26/a-FR

+

B

7500-I

Eo.

tC

|OQ.

_lCDO.

5000-

2500-

INT-Ow2

ieÂ»""iÃ¯Â«""I4J""Ã®4Â»li"

INT-Ov3

INT.Ov-2 INT.Ov-3

Stimulating tumor cells

normal organs or in lungs containing micrometastases of parental non-transduced C26, whereas a specific accumulation of IL-2/MOV19 scFvwas detectable in lungs bearing C26/a-FR micrometastases (Fig. 4).

In Vivo Treatment with IL-2/MOV19 scFv. The C26/a-FR clonewas characterized in both immunocompetent and immunocompro-

mised hosts and was found to behave like the parental cell line interms of growth rate, tumor take, and metastatic potential.

Using the producer J558EF6.1 cells as a living pump to obtain aprolonged release of the IL-2/MOV19 scFv in vivo, we analyzedwhether targeting IL-2 to C26/a-FR would affect tumor growth.

Indeed, producer J558EF6.1 cells released detectable levels of humanIL-2 in mouse serum for up to 4 days without forming tumors when

injected into syngeneic BALB/c or nu/nu mice (Fig. 5A), thus behav

ing like J558L transduced with the IL-2 gene (31). BALB/c mice wereinjected s.c. with either C26 or C26/a-FR and injected withJ558EF6.1 cells in the opposite flank. Growth of C26/a-FR was

completely inhibited in 22% of mice, whereas C26 tumor growth wasunaffected by IL-2/MOV19 scFv (Fig. 5ÃŸ).The same treatmentprotected 60% of mice injected i.v. with C26/a-FR but did not protectmice injected with C26 cells (P = 0.011; Fig. 5Q.

Because data from the pharmacokinetic experiments showed thatthe fusion protein had a fast clearance from the blood, we tested theeffect of repeated i.p. injection of purified IL-2/MOV19 scFv in miceinoculated s.c. with C26/a-FR. In different experiments, treatment

with the fusion protein significantly reduced the tumor volume(P = 0.03-0.003) in comparison with those of both control and

4150




LIVER MUSCLEMETASTATIC

LUNG2

O

I,s

m 1 -

0.5-

10 20 10 20 10 20 30

TIME POST-INJECTION (h)

Fig. 4. Biodislribution of l:5I-Iabeled IL-2/MOV19 scFv in BALB/c mice bearing

experimental lung mÃ©tastasesof C26 (filled symbols) or C26/a-FR (open symbols). Theratio between tissue and blood %ID/g was measured in three animals at each time pointterror bars, SE).

IL-2-treated mice (Fig. 6, A and B), whereas tumor onset was notsignificantly affected by treatment with either recombinant IL-2 orIL-2/MOV19 scFv.

DISCUSSION

Here we report the construction of a fusion protein between IL-2and the scFv of anti-a-FR Ab MOV 19 to provide a new tool forcombinatorial immunotherapy of human ovarian carcinoma. IL-2/

MOV 19 scFv was designed to be used first in a preclinical murinemodel, in which it was shown to specifically target antigen-positive

tumor cells and to cooperate with the antitumor immune response. Toour knowledge, this is the first demonstration of biologically activefusion proteins with IL-2 and scFv directed against a tumor antigen

whose therapeutic effect has been tested in murine syngeneic models(6, 32).

In designing the IL-2/MOV19 scFv fusion protein, we considered

several requirements: (a) the Ab must recognize an antigenic epitopedifferent from that of the MOV 18 mAb, which is part of the bsAbOC/TR that is used to retarget T lymphocytes for adoptive immunotherapy of ovarian carcinoma patients ( 18), thus allowing the combination of fusion protein and OC/TR treatments; (b) the moleculeshould remain on the tumor cell surface on antigen binding and shouldnot be internalized to allow IL-2 to bind its receptor on effector

lymphocytes; and (c) the fusion protein should have a small size toeasily penetrate into the tumor masses and should have a fast bloodclearance, thereby reducing the possibility of systemic toxicity. Therefore, we chose the MOV 19 mAb, which does not compete withMOV 18 for binding sites and is not internalized on binding (20), tofuse its scFv sequence with IL-2. In fact, the scFv fragment displaysmany of the required characteristics including a better tumor micro-

distribution in comparison with the intact IgG, which accumulates intothe perivascular regions of the tumor (33).

Different cloning strategies have been used to express fusion proteins between cytokines and antibodies or their fragments (5, 34-36).

Bacterial expression has been obtained for some fusion proteins, butalthough it is very convenient, it can result in low yields, especiallywhen the protein is not released into the periplasmic space butaccumulates in the inclusion bodies; moreover, bacterial expressioncan generate endotoxins that may be difficult to remove (32, 34).Attempts to express IL-2/MOV19 scFv in bacteria failed, because no

fusion protein could be rescued from the cell supernatant or periplasmic extract,5 suggesting that the protein was confined to inclusion

bodies. Refolding of the antibodies from inclusion bodies could beobtained, but this is a cumbersome procedure whose yield can varygreatly for distinct Fv regions (32, 34, 35). Therefore, we turned toeukaryotic expression, which may offer advantages, especially interms of posttranslational modification; moreover, the use of myelomacells for protein production allows for a continuous manufacturing ofsoluble protein with a high yield (35). We found that cloning the IL-2

1.5-1

ITn

24 48 72

Time (h)

96

100

uEo>O>

o3

50-

B

10 20 30 40

Days

100'

50-

CO

20 60 80 100

Days

Fig. 5. Treatment in vivo with J558EF6.1 cells secreting IL-2/MOVI9 scFv. A. normalmice injected s.c. with IO7 (D) or 5 x IO6 (D) producer cells were bled at the indicated

time points, and human IL-2 was measured by ELISA in serum samples. Values are themean of three individual mice (error bars, SE), ÃŸ.tumor onset in mice injected s.c. withC26 (circle) or C26/0-FR (st/uare) and treated with a s.c. injection of IO6 J558EF6.1 cells

(filled symbols) or PBS (open symbols) on day 1 in the opposite flank. C survival of micebearing lung mÃ©tastasesof C26 (circle) or C26/or-FR (square) treated with three s.c.injections of lO6J558EF6.1 cells (filled symbols) or PBS (open symbols) on days 1.7. and14. n = 10 animals/group. * = /> fi 0.01.

4151




2000-1

1500-1

1000-

a 500-

0-

10 20 30

Days

40

2000-1

1500-

2 500-

I 0-â€”Iâ€”

10

â€”Iâ€”

20

â€”Iâ€”

30 40

Days

Fig. 6. Treatment of mice hearing s.c. C26/a-FR with purified IL-2/MOVI9 scFv. A.mean tumor volume in mice treated i.p. with 10 jig/day purified fusion protein (â€¢)or PBS(D). A pool of two different experiments is shown, groups of five mice for eachexperiment. * - P r~ 0.03: *Â»= />Â£ 0.003. B. mean tumor volume in mice treated i.p.with 10 Â¿ig/daypurified fusion protein (â€¢)or an equivalent amount of recomhinant IL-2(75.000 lU/day: (O), n = 5 animals/group. * = P -^ 0.03: ** = />Â£ 0.003.

open reading frame with its leader sequence at the 5' end and intro

ducing the 13-amino acid spacer GSTSGSGKSSEGK to separate IL-2

from scFv represented the best configuration for a functional fusionprotein, as far as antigen recognition and IL-2 bioactivity are con

cerned. However, the expression of such a construct driven by thecytomegalovirus promoter Â¡nJ558L cells resulted in an active fusionprotein but a low yield. To optimize fusion protein production, wetook advantage of the pH PCRy III vector containing both immuno-

globulin H and K enhancers and the immunoglobulin Â«promoter.Moreover, by linking the fusion protein to the CK chain, we couldeasily purify it by affinity chromatography without introducing exogenous, potentially immunogenic sequences.

Aggregation often occurs in scFv due to the coupling of VH and VLbelonging to different molecules, and this can affect the pharmacoki-

netics and biodistribution of the molecule, reducing its bioavailabilityand function. We were able to drastically reduce aggregation by theaddition of 3.4 mM EDTA to both the dialyzing and storage buffers,thereby conserving the functional properties of the fusion protein.

When compared with fusion proteins with a similar structure, suchas L6 F(ab')/IL-2 described by Fell et al. (5), IL-2/MOV19 scFv

displayed a higher IL-2 bioactivity that was equivalent on a molarbasis with that of recombinant human IL-2. This feature is presumablyrelated to the characteristic structure of IL-2/MOV19 scFv, with theIL-2 cDNA at the 5' end of the coding sequence, and the scFv spacedby a "rigid" linker that allows a better interaction between the cyto-

kine at the amino terminus of the fusion protein and its receptor (5).

Purified IL-2/MOV19 scFv retained the specificity of the MOV 19

A mAb, because it competed with MOV 19 mAb for antigen binding andtargeted tumor cells expressing a-FR. Whereas other fusion proteinsbetween IL-2 and mAb or their fragments were shown to bind to theIL-2 high-affinity receptor (5, 36), thus requiring the preactivation oflymphocytes, IL-2/MOV19 scFv, once bound to the cell surface oftumors such as INT-Ov.2, stimulated the proliferation of resting

lymphocytes from human donor PBLs.The biodistribution of l25I-IL-2/MOV19 scFv injected i.p. into

naive BALB/c mice or BALB/c mice bearing experimental lungmicrometastases of C26 showed no specific accumulation of thefusion protein in the lungs, spleen, liver, kidneys, stomach, large andsmall bowel, blood, and muscle. Conversely, mice bearing lung mÃ©tastases of C26/a-FR tumor cells revealed an accumulation of IL-2/

MOV 19 in the lungs, as evaluated by the progressive increase of the%ID/g ratio between lung and blood, thus underlining the binding

D specificity of IL-2/MOV19 scFv. Because only part of the lung tissue

is affected by the metastatic process, the ratio of labeled fusion proteinlocalizing to the tumor can be expected to be even higher than thatindicated by the experimental results.

The very rapid pharmacokinetics of IL-2/MOV19 scFv resemble

that of other scFv (33). but the decay seems to be more accelerated,probably due to the IL-2 moiety, which, thanks to its short half-life

and fast clearance, reduces the r,/2 of the fusion protein, a phenomenon that has also been described for IL-2 fused with an entire immu

noglobulin molecule (37, 38).We designed an initial set of experiments in which IL-2/MOV19

scFv was directly secreted in vivo by J558EF6.1 cells injected s.c.The treatment protected 22% of mice from the development ofC26/a-FR tumors inoculated in the opposite flank and, as ex

pected, did not affect the growth of C26 tumor cells. The sametreatment was shown to significantly prolong the survival of micebearing experimental lung mÃ©tastasesof C26/a-FR but not that of

mice bearing parental C26 (Fig. 5C). These results prompted us touse purified IL-2/MOV19 scFv for repeated i.p. injections. Al

though the tumor volume at different time points was significantlyreduced in treated versus untreated mice (Fig. 6A) or versusIL-2-treated mice (Fig. 6ÃŸ),neither of the treatments significantlyaffected tumor onset and progression, suggesting that IL-2 targeting by the fusion protein, but not systemic IL-2, impairs the growthof C26/Ã–-FR without being able to cure such an aggressive tumor

as a single therapeutic agent.Reisfeld et al. (4), Xiang et al. (7), Becker et al. (39, 40), and Lode

et al. (8) have extensively analyzed the therapeutic activity of fusionproteins between IL-2 and the chimeric anti-GD2 ganglioside Ab

eh 14.18 that recognizes melanoma and neuroblastoma tumor cells orthe huKSl/4 Ab that recognizes the colon carcinoma-associated antigen Ep-CAM: using syngeneic murine models, they demonstratedthat these fusion proteins were able to eradicate established TAA-

transduced tumors and their experimental mÃ©tastasesthrough theactivation of CD8+ T lymphocytes (4, 7, 39, 40) or natural killer cells

(8). These investigators also demonstrated that the chl4.18-IL-2 fusion protein induced a CD8+ T-cell-mediated, long-lived, transferable

tumor-protective immunity in mice after the eradication of establishedmelanoma mÃ©tastases(40). In comparison with the chl4.18-IL-2 orhuKSl/4-IL-2 fusion proteins, the therapeutic efficacy of IL-2/

MOV 19 scFv alone seems to be limited, because our reagent was notable to cure mice bearing the C26/a-FR tumor. The differences between

the therapeutic efficacy of the two types of fusion proteins can beattributed mainly to the molecular structure, because chl4.18-IL-2 andhuKSl/4-IL-2 are immunoglobulin fusion proteins, and IL-2/MOV19

4152




scFv uses the scFv, thus resulting in a smaller molecule with differentpharmacokinetics and biological activity. Moreover, the tumor modelsare different, and the different release of various immunomodulatoryfactors by the neoplastic cells may affect the type of antitumor effectorcells activated by the IL-2/Ab fusion protein (8, 39, 40).

The reduced size of the IL-2/MOV19 scFv molecule is expected to

increase its tissue penetration and reduce its immunogenicity; moreover, the lack of the Fc tail avoids the activation of nonspecific effectorcells through antibody-dependent cell-mediated cytotoxicity, thus favoring the IL-2-mediated activation of T lymphocytes, a feature that could be

helpful when the fusion protein is administered in combination withadoptive transfer of T cell redirected by the bsAb OC/TR.

Although successful immunotherapy in preclinical models has often been achieved with single cytokines either given systemically orreleased by transduced cells, in humans, treatment with IL-2 that was

either systemic, secreted by ex vivo engineered cells, or secreted by invivo gene transduction resulted in limited therapeutic effects (41-44).

Combination of different immunotherapeutic approaches shouldtherefore be considered for the treatment of established tumors.

The pharmacokinetic characteristics of the murine IL-2/MOV19

scFv fusion protein call for different approaches to overcome its shortserum half-life and to prolong its therapeutic efficacy. In this regard,

minipumps implanted into animals may be helpful, because the constant release of the fusion protein would mimic the continuous i.v.administration achievable in patients. Finally, to be successful in ahuman setting, the molecule will have to be modified to reduce itsimmunogenicity by using human MOV 19 F(ab), whose variable regions have recently been isolated by guided selection from an antibody phage library (23).

REFERENCES

12.

13.

14.

15.

16.

17.

IX.

19.

ACKNOWLEDGMENTS

We thank Prof. Ralph A. Reisfeld (Department of Immunology. The Scripps 2Â°-

Research Institute. La Jolla. CA) for critically reviewing the manuscript. Wealso thank Ivano Arioli and Mariella Parenza for technical assistance and Dr.Donatella Negri for skillful help in the pharmacokinetics experiments. 21.

22.1. Fearon, E. R.. Pardoll. D. M. llaya. T.. Golumbcck. P.. Leviisky, H. !.. Simons. J. W..

Karasuyama, H.. Vogelstein, B., and Frost. P. Interleukin-2 production by tumor cells 23.

bypasses T helper function in the generation of an antitumor response. Cell, 60:397-403, 1990.

2. Gansbacher. B.. Zier, K.. Daniels. B.. Cronin, K.. Bannerji. R., and Gilboa. E.Interleukin 2 gene transfer into tumor cells abrogates tumorigenicity and induces 24.protective immunity. J. Exp. Med.. 172: 1217-1224. 1990.

3. Maass, G., Schmidt, W., Berger, M.. Schilcher. F.. Koszik, F.. Schneeberger. A., 25Stingi. G.. Birnstiel. M. L., and Schweighoffer. T. Priming of tumor-specific T cellsin the draining lymph nodes after immunization with interleukin 2-secreting tumorcells: three consecutive stages may be required for successful tumor vaccination. 26,Proc. Nati. Acad. Sci. USA, 92: 5540-5544. 1995.

4. Reisfeld. R. A.. Becker. J. C. and Gillies, S. D. Immunocytokines: a new approachto immunotherapy of melanoma. Melanoma Res.. 7 (Suppl. 2): S99-S106. 1997.

5. Fell, H. P.. Gayle, M. A.. Grosmaire. L., and Ledbetter, J. A. Genetic construction and -,-Â¡characterization of a fusion protein consisting of chimeric F(ab') with specificity for

carcinomas and human IL-2. J. Immunol.. 146: 2446-2452. 1991.6. Xiang. J., Liu. E.. Moyana, T.. and Qi, Y. Single-chain antibody variable region-

targeted interleukin-2 stimulates T cell killing of human colorectal carcinoma cells. .-Immunol. Cell Biol., 72: 275-285. 1994.

7. Xiang, R., Lode, H. N.. Dolman. C. S., Dreier, T., Varki, N. M., Qian, X., Lo, K., Lan,Y.. Super. M.. Gillies. S. D., and Reisfeld, R. A. Elimination of established murinecolon carcinoma mÃ©tastasesby antibody-interleukin 2 fusion protein therapy. CancerRes., 57: 4948-4955, 1997.

8. Lode. H. N., Xiang. R.. Dreier, T., Varki. N. M.. Gillies. S. D., and Reisfeld. R. A. 30Natural killer cell-mediated eradication of neuroblastoma mÃ©tastasesto bone marrowby targeted interleukin-2 therapy. Blood, 91: 1706-1715. 1998.

9. Taylor. A. E., Wiltshaw. E., Gore, M. E., Fryatt, !.. and Fisher. C. Long-term 3'follow-up of the first randomized study of cysplatin versus carboplatin for advancedepithelial ovarian cancer. J. Clin. Oncol., 12: 2066-2070. 1994.

10. Kuiper. M.. Peakman, M.. and Farzaneh. F. Ovarian tumor antigens as potentialtargets for immune gene therapy. Gene Then, 2: 7-15. 1995. 32.

4153

29.

Ramakrishna. V.. Negri. D.. Bmsic. V., Fomanelli. R.. Canevari. S.. Bolis. G., Castell,C.. and Parmiani. G. Generation and phenotypic characterization of new humanovarian cancer cell lines with the identification of antigens potentially recognizable byHLA-reslricted cytotoxic T cells. Int. J. Cancer, 73: 143-150, 1997.

Miotti. S., Canevari, S., Menard. S., Mezzanzanica. D., Porro, G., Pupa, S. M.,Regazzoni. M.. Tagliabue. E.. and Colnaghi, M. I. Characterization of human ovariancarcinoma associated antigens defined by novel monoclonal antibodies with tumorrestricted specificity. Int. J. Cancer, 39: 297-303. 1987.

Epenetos. A. A.. Munro. A. J.. Stewart. S.. Rampling. R.. Lambert. H. E.. McKenzie.C. G., Soutier. P.. Rahemtulla. A.. Hooker, G.. Sivolapenko. G. B.. Snook, D..Courtcnay-Luck. N., Dhokia, B.. Krausz. T.. Taylor-Papadimitriou. J., Durhin. H..and Bodmer. W. Antibody-guided irradiation of advanced ovarian cancer with intra-

periloneally administered radiolabelled monoclonal antibodies. J. Clin. Oncol., 5:1890-1899. 1987.

Crippa. F.. Bolis, G.. Seregni, E.. Gavoni. N.. Bombardieri. E., Scarfone. G.. Ferrari.C., and Buraggi. G. L. Single dose intraperiloneal radioimmunolherapy with themurine monoclonal antibody L"l-MOvl8: clinical results in patients with minimal

residual disease of ovarian cancer. Eur. J. Cancer. 31A: 686-690. 1995.

Rath, U., Kaufmann. M.. Schmid. H.. Hofmann, J.. Wiedemann. B., Kist. A..Kempeni. J., Schlick, E.. Basten. G.. Kommerell, B., and Mannel. D. Effect ofintraperitoneal recombinant human tumor necrosis factor a on malignant ascites. Eur.J. Cancer. 27: 121-125. 1991.

Sleis. R. G., Urba, W. J., Van der Molen. L. A.. Bookman, M. A.. Smith, J. W.. Clark,J. W., Miller. L. R.. Cnim. E. D.. Beckner, S. K.. McKnight, J. E., Ozols, R. F.,Stevenson, H. C., Young, R. C.. and Longo. D. L. Intraperitoneal lymphokine-activated killer-cell and interleukin-2 therapy for malignancies limited to the peritoneal cavity. J. Clin. Oncol., 8: 1618-1629, 1990.

Freedman. R. S.. Edwards, C. L.. Kavanagh. J. J., Kudelka, A. P.. Katz, R. L.,Carrasco, C. H.. Atkinson. E. N., Scott, W.. Tomasovich, B., Templin, S., andPlatsoucas. C. D. Intraperitoneal adoptive immunotherapy of ovarian carcinoma withtumor-infiltrating lymphocytes and low-dose recombinant interleukin-2: a pilot trial.J. Immunother., 16: 198-210. 1994.

Canevari, S.. Sloter. G.. Arienti. F., Bolis. G., Colnaghi, M. I.. Di Re, E. M.,Eggermont. A. M. M.. Goey. S. H., Gratama, J. W., Lamers, C. H. J.. Nooy, M. A.,Parmiani. G.. Raspagliesi. F.. Ravagnani. F.. Scarfone. G.. Trimbos, J. B., Wamaar,S. O., and Bolhuis. R. L. H. Regression of advanced ovarian carcinoma by intrap-

eritoneal treatment with autologous T lymphocytes retargeted by a bispecific monoclonal antibody. J. Nati. Cancer Inst.. 87: 1463-1469. 1995.

Freedman, R. S.. Gibbons. J. A.. Giedlin. M.. Kudelka. A. P.. Kavanagh. J. J..Edwards. C. L.. Carrasco. C. H., Nash, M. A., and Platsoucas. C. D. Immunophar-macology and cytokine production of a low-dose schedule of intraperitoneally administered human recomhinant interleukin-2 in patients with advanced epithelialovarian carcinoma. J. Immunother, 19: 443-451. 1997.

Casalini, P.. Mezzanzanica. D.. Canevari. S.. Delia Torr. G.. Miotti. S., Colnaghi,M. !.. and Matzku. S. Use of a combination of monoclonal antibodies directed againstthree distinct epitopes of a tumor-associated antigen. Analysis of cell binding anditernalizalion. Int. J. Cancer, 48: 284-290. 1991.Melani, C., Chiodoni. C.. Arienti. F.. Maccalli. C.. SulÃ©-Suso,J., Anichini. A.,

Colombo. M. P.. and Parmiani. G. Cytokine gene transduction in tumor cells:interleukin (ILl-2 or IL-4 gene transfer in human melanoma cells. Nat. Immun., 13:76-84. 1994.

Clackson. T.. Hoogenboom, H. R.. Griffith. A. D., and Winter. G. Making antibodyfragments using phage display libraries. Nature (Lond.). 352: 624-628, 1991.

Figini. M.. Obici. L.. Mezzanzanica. D., Griffith. A., Colnaghi. M. !.. Winter. G.. andCanevari, S. Panning phage antibody libraries on cells: isolation of human Fabfragments against ovarian carcinoma using guided selection. Cancer Res., 58: 991-

996. 1998.Oi, V. T., Morrison, S. L.. Herzenberg. L. A., and Berg, P. Immunoglobulin expression in transformed lymphoid cells. Proc. Nail. Acad. Sci. USA. 80: 825-829. 1983.

Traunecker. A.. Schneider. J.. Kiefer. H., and Karjalainen. K. Highly efficientneutralization of HIV recombinant CD4-immunoglobulin molecules. Nature (Lond.).339: 68-70, 1989.

Coney. L. R., Mezzanzanica, D., Sanborn. D.. Casalini, P.. Colnaghi. M. I., andZurawsky. V. R.. Jr. Chimeric murine-human antibodies directed against folate

binding receptor are efficient mediators of ovarian carcinoma cell killing. CancerRes., 54: 2448-2455, 1994.

Bollero. F., Tomasselli, A., Canevari, S-, Miotti. S.. MÃ©nard,S., and Colnaghi. M. I.Gene transfcction and expression of the ovarian carcinoma marker lolate-binding

protein on NIH/3T3 cells increases cell growth in vitro and in vivo. Cancer Res., 53:5791-5796. 1993.

BÃ©nard.J., De Silva. J.. De Blois, M. C., Boyer. P.. Duvilland. P.. Chiric. Eâ€žandRiou. G. Characterization of a human ovarian adenocarcinoma line. IGROVI, intissue culture and in nude mice. Cancer Res.. 45: 4970-4979. 1985.

Marchalonis. J. J. An enzymatic method for the trace iodination of immunoglobulinand other proteins. Biochem. J.. 113: 299-305. 1969.

Gadina. M.. Canevari, S., Ripamonti. M., Mariani. M., and Colnaghi. M.I. Preclinicalpharmacokinetics and localization studies of the radioiodinated anti-ovarian carcinoma MOvlS. NucÃ.Med. Biol., 18: 403-408, 1991.

Hock. H., Dorsch, M., Kunzendorf, U.. Qin. Z.. Diamanstein. T.. and Blankenstein.T. Mechanisms of rejection induced by tumor cell-targeted gene transfer of interleukin-2. inlerleukin-4, interleukin-7. tumor necrosis factor, or interferon y. Proc. Nati.Acad. Sci. USA, 90: 2774-2778, 1993.

Boleti. E.. Deonarain. M. P.. Spooner. R. A.. Smith. A. J.. Epenetos, A. A., and




George, A. J. T. Construction, expression and characterization of a single chain antibody-interleukin 2 fusion protein overcomes tumor heterogeneity hy induction ofanti-tumour antibody (scFv)-IL-2 fusion protein. Ann. Oneol., 6: 945-947, 1995. a cellular immune response. Proc. Nail. Acad. Sci. USA, 93: 7826-7831, 1996.

33. Yokota. T., Milenic, D. E., Whitlow, M., and Schlom, J. Rapid lumor penetration of 39. Becker, J. C, Pancook, J. D., Gillies. S. D., Furukawa, K., and Reisfeld, R. A. Ta single-chain Fv and comparison with other immunoglobulin forms. Cancer Res.. 52: cell-medialed eradication of murine metastatic melanoma induced by targeted inter-3402-3408, 1992. leukin 2 therapy. J. Exp. Med., 183: 2361-2366, 1996.

34. Savage. P., So, A.. Spooner, R. A., and Epenetos. A. A. A recombinanl single chain 40. Becker. J. C.. Varki. N.. Gillies. S. D.. Furukawa. K.. and Reisfeld. R. A. Long-livedantibody interleukin-2 fusion protein. Br. J. Cancer. 67: 304-310, 1993. and transferable tumor immunity in mice after targeted interleukin-2 therapy. J. Clin.

35. Dorai, H., McCartney. J. E.. Hudziak. R. M., Tai. M. S., Lamine!. A. A.. Houston. Investig., 98: 2801-2804. 1996.L. L.. Huslon. J. S., and Oppermann, H. Mammalian cell expression of single-chain 41. Rosenberg. S. A.. Lolze, M. T.. Yang, J. C.. Aebersold. P. M.. Linehan. W. M.. Seipp.Fv (sFv) antibody proteins and their C-terminal fusion with interleukin-2 and other C., and White. D. E. Experience with the use of high-dose interleukin-2 in theeffector domains. Bio-Technology, 12: 890-897, 1994. treatment of 652 cancer patients. Ann. Surg., 210: 474-485. 1989.

36. Gillies. S. D., Reilly. E. B., Lo. K. M.. and Reisfeld, R. A. Antibody-targeted 42. Parmiani, G. An explanation of the variable clinical response to interleukin 2 andinterleukin 2 stimulates T-cell killing of autologous tumor cells. Proc. Nati. Acad. Sci. LAK cells. Imrnunol. Today. //: 113-115. 1990.USA. 89: 1428-1432. 1992. 43. Parmiani. G.. Colombo, M. P., Melani. C.. and Arienti, F. Cytokine gene transduction

37. Gillies, S. D., Young, D., Lo, K. M.. and Roberts, S. Biological activity and in vivo in the immunotherapy of cancer. Adv. Pharmacol.. 40: 259-307, 1997.clearance of antitumor antibody/cytokine fusion proteins. Biocon. Chem.. 4: 230- 44. Shawler. D. L., Fakhrai, H., Van Beveren, C.. MercÃ³la. D., Gold, D. P., Bartholomew,

235, 1993. R. M.. Royston, I., and Soboi, R. E. Gene therapy approaches to enhance antitumor38. Becker. J. C.. Varki, N., Gillies, S. D.. Furukawa. K., and Reisfeld. R. A. An immunity. Adv. Pharmacol., 40: 309-337, 1997.

4154



1998;58:4146-4154. Cancer Res Cecilia Melani, Mariangela Figini, Daniela Nicosia, et al. Fusion with a Single-Chain Fv of Antifolate Receptor AntibodyTargeting of Interleukin 2 to Human Ovarian Carcinoma by

Updated version

http://cancerres.aacrjournals.org/content/58/18/4146

Access the most recent version of this article 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. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/58/18/4146To request permission to re-use all or part of this article, use this link



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

mailto:[email protected]



targeting of interleukin 2 to human ovarian carcinoma by fusion … · targeting of interleukin 2...

Documents