Generation and characterization of a single-gene encoded single-chain immunoglobulin-interleukin-2 fusion protein

Immunotcchnology I (1995) 231-241

Generation .and characterization of a single-gene encoded single-chain immunoglobulin-interleukin-2 fusion protein Liming Shu’, Chen-F. Qi2, Patricia Horan Hand, Jeffrey Schlom*, S.V. S. Kashmiri Laboratory of Tumor Immunologyand Biology, National Cancer Institute, National Institutes of Health, Building 10. Room 8BO7, 10 Center Drive MSC 1750, Bethesda. MD H)892-1750, USA Received 31 July 1995; revision received 13 October 1995; accepted 19 October 1995

Buckground: Interleukin-2 (IL-2), a potent inducer of cellular immune responses, has been used for biological therapy of human cancer; however, the high doses of IL-2 required to mediate patients’ immune responses can cause considerable systemic toxicity. The murine monoclonal antibody (MAb) CC49, which reacts with tumor-associated glycoprotein (TAG)-72, expressed on a variety of human carcinomas, has shown excellent tumor localization in recent clinical trials. Objectives: Development and characterization of a singlechain immunoglobulin-IL-2 (SCIg-IL-2) fusion protein which, by delivering IL-2 selectively to the tumor site, can serve as an effective reagent for CC49/IL-2 combination therapy. Study design: A single-gene encoding the SCIg-IL-2 fusion protein derived from the chimeric (c) CC49 was designed, generated and inserted in an expression vector. The monomeric single-chain protein consisted of the CC49 heavy and light chain variable domains covalently joined through a (GGGGS)3 linker peptide. The carboxyl end of the variable domain of the light chain was linked to the amino terminus of the human yl Fc through the hinge region, and the carboxyl end of the CH3 domain was linked to the amino terminus of the human IL-2 through a GGGSGGG linker peptide. The SCIg-IL-2, expressed from the murine myeloma cells transfected with the expression construct, was characterixed for its antigen-binding specifity, antibody effector functions and IL-2 biological activity. Reds and conclusion: Transfection of murine myeloma cells with the single-gene expression construct SCIg-IL-2 expressed a single-chain protein of approximately 70 kD, which was secreted into tissue culture tluid as a homodimer of approximately 140 kD. SCIg-IL-2 competed completely with CCC49 for binding to the TAG-72 antigen, but approximately three- to four-fold more of the SCIg-IL-2 was required to achieve levels of competition similar to those observed with the murine or chimeric CC49. With human effeetor cells, the fusion protein mediated lysis of TAG-72positive human carcinoma cells. Prior treatment of human effector cells with 100 U/ml of human IL-2 enhanced the fusion protein-mediated cytolysis from 32 to 65%. At doses of L 1 ng/ml, the stimulatory effect of SCIg-IL-2 on IL-2 dependent murine HT-2 cell proliferation was comparable to that of the recombinant human IL-2. The single-gene construct may also facilitate inoculation of the gene in animal tissue for in vivo expression of the fusion protein. Keywords: Single&sin Carcinoma

itumunoglobulin;

Single gene; Fusion

Abbreviations: ADCC, antibody-dependent cellular cytotoxicity; BSM, bovine submaxillary mucin; cCC49, mousehuman chin&c Cc49; IgG, immunoglobuhn G; H-2, intcrleukin-l; MAb, monoclonal antibody; nCC49, native MAb Cc49; rh-IL2, recombinant human interleukin-2; SDS, sodium dodecyl sulfate; TAG, tumor-associated glycoprotein. Elsevier Science B.V. SSDI 1380-2933(95)00025-P

protein;

Monoclonal

antibody

CC49; Interleukin-2;

??Corresponding author, Tel.: +l 301 4964343; Fax: +I 301 4%2756. ’Present Address: IGEN, Inc., 1602 Industrial Drive, Gaithersburg, MD 20877, USA. ‘Present Address: Bldg. 7, Room 302, NIAID, NIH, Bethesda, MD 20892, USA.

232

L. Shu et al. /Immunotechnology I (1995) 231-241

1. Introduction Interleukin-2 (IL-2) is a potent inducer of cellular immune responses [ 11. This characteristic has prompted numerous investigations involving the use of IL-2 for biological therapy of human cancers (for review, see [2]). IL-2 has been administered alone [3] and in conjunction with lymphokine-activated killer (LAK) cells [4], which are generated by IL-2 stimulation of immune cells [5]. Results from these clinical studies suggest that high doses of IL-2 may be required to effect a therapeutic response in patients. These studies also make it clear that rapid clearance of IL-2 from the circulation [6] and systemic toxicity [7] are two major impediments of its optimal use as a therapeutic agent. To circumvent these difficulties, researchers are attempting to devise ways of delivering high concentrations of IL-2 exclusively to the tumor site. Adoptive cellular therapy is one such approach being tested in preclinical studies. Tumor cells, genetically manipulated to secrete IL2, have been adoptively transferred into animals to secrete high concentrations of IL-2 in a local tumor environment [8-121. Also, tumor-infiltrating lymphocytes (TILs), which preferentially localize to tumor sites, have been engineered to secrete IL-2 [ 131. An alternative approach is the use of antibodies to tumor associated antigens as vehicles for delivering IL-2 to the tumor site. With this intention, Fell et al. [14] developed a fusion protein consisting of human IL-2 attached to a chimeric F(ab’) derived from murine monoclonal antibody (MAb) L6, which is selectively reactive to several types of human carcinomas. The fusion protein retained antigen-binding specificity of MAb L6 and the biological activity of the IL-2. However, it could not carry out antibody effector functions, such as antibody-dependent cellular cytotoxicity (AD CC), because it lacked the Fc region. LeBerthon et al. [IS] chemically linked IL-2 to the carboxy1 end of the heavy chain of murine MAb B72.3 to increase tumor targeting of the antibody, presumably by increasing local vascular permeability. However, the chemical link rendered the IL-2 moiety of the fusion protein biologically inactive [ 151. Fusion proteins, in which human IL-2

was genetically linked to the CH3 domain of the human Fc, were later derived from the anti ganglioside GD2 antibody (ch14.18) [16] and the anti-human epidermal growth factor antibody ch225 [I 71. These antibody-IL-2 fusion proteins retained the antigen-binding specificity and biological activity of IL-2. In addition, the human effector cells’ ability to kill the specific target cells was augmented when the target cells were coated with the fusion protein. The therapeutic potential of the tumoricidal antibody-cytokine conjugates might be more efticiently harnessed if it were feasible to express, in vivo, these molecules or carry out ex vivo transfection of cells for their delivery to the tumor site by adoptive transfer. This requires, however, the introduction and expression of two genes (one encoding antibody light chain and the other encoding heavy chain-IL-2 fusion protein) in a single cell to generate a functional antibody-IL-2 conjugate. To facilitate ex vivo transfection and in vivo gene expression, we earlier developed a single-gene construct SGAC,_C,.t1, which encodes a single-chain immunoglobulin (SCIg) molecule derived from MAb CC49 [ 181. Native MAb CC49 (nCC49) is a murine IgGl reactive with tumor-associated glycoprotein (TAG)-72, which is expressed on a variety of carcinomas, including those of gastrointestinal, ovarian, endometrial, prostatic, pancreatic, non-small cell lung, and mammary origin [19,20]. In recent clinical trials, ‘3’I-labeled nCC49 has shown excellent tumor localization to colorectal, ovarian, and prostate carcinomas [21-231. Several genetic variants, including a mouse-human chimeric CC49 (cCC49), a ‘humanized’ CDR-grafted CC49 [24], a CC49 sFv (251, and the single-gene-encoded single-chain immunoglobulin molecule designated SCAACLCHl (SCIgACHl) [18], have recently been developed. SCIgACHl , which is derived from cCC49, consists of CC49 sFv covalently joined to the human yl Fc through the hinge region. SP210 murine myeloma cells transfected with the singlegene construct expressed a single-chain protein that is secreted as a dimeric molecule of 108 kD. This molecule retains the antigen binding properties, cytolytic activity and tumor targeting characteristics of the parental cCC49 molecule

L. Shu et al. /Immunotechnology

[ 181.In an effort to develop a molecule which may efficiently target IL-2 to the tumor site and serve as a reagent for antibody/IL-2 combination therapy of human carcinomas, we have generated a single-gene-encoded single chain fusion protein SCIg-IL-2, which has a human IL-2 molecule attached to the carboxyl end of the Fc region of SCIgACH 1. We report here the characterization of the antigen-binding specificity, antibody effector functions, and IL-2 biological activity of this molecule in comparison with those of SCIgACul and ccc49.

I (1995)

231-241

233

2.2. Templates for DNA amplification Sequences encoding Cu2 and Cu3 domains of the human y 1 were amplified by using the expression construct pLGCKSN49 as a template. pLGCKSN49, which carries genes encoding chimeric heavy and light chains of MAb CC49 in pLNCX, has been previously described [ 181. A human IL-2 cDNA of 1.2 kb, cloned in pBR322 [26], provided by Dr. Howard Young (National Cancer Institute-Frederick Cancer Research and Development Center, Frederick, MD), was used as a template to amplify human IL-2 sequences.

2. Materials and methods 2.1. Polymerase chain reaction (PCR)

and the

synthetic oligonucleotide primers

For DNA amplification, PCR was carried out essentially as previously described [ 181. The sequences of the oligonucleotides used as primers were oligonucleotide 1 (5’ CH2, coding), 5’GTGCATAATGCCAAGACA-3 ’; oligonucleotide 2 (3’ CH3, noncoding), 5’GCCACCTCC TGATCCCCTCCTTTACCCGGAGACAGGGAGAG-3’; oligonucleotide 3 (5’ IL-2, coding), 5’AAGGAGGCGGATCAGGAGGTGGCGC ACCTACTTCAAGTTCTACA-3 ’; and oligonucleotide 4 (3’ IL-2, noncoding), 5’-GACATATGaagcttCAAG’ITAGTG’ITGAGATGATG-3 ‘. The overlapping complementary sequences are underlined, and the Hind111 recognition site is in lower case letters. Oligonucleotide 1 anneals to the sequence, 3 bp upstream from the unique SstII site, of the noncoding strand of the gene that codes for the C,2 region of human yl. Oligonucieotide 2 is complementary to the coding sequence of the Cu3 domain of human yl, starting at 33 bp downstream from the unique NsiI site and stretching to the 3 ’end. Oligonucleotide 3 has complementarity to the 5’ end of the noncoding strand of the human IL-2 gene, whereas parts of the overlapping sequences of oligonucleotides 2 and 3 bear informational content for the peptide linker GGGSGGG. Oligonucleotide 4 anneals to the 3 ’ end of the coding strand of the human IL-2 gene. Additionally, it carries a termination codon and a site for Hind111 at the flank.

2.3. Eukaryotic expression vector and transfection The expression construct pLNCS23-IL-2 is derived from the vector pLNCX, which has been described previously 1271.Briefly, pLNCX carries the human cytomegalovirus promoter/enhancer complex to drive expression of the target gene, which can be inserted in a multiple cloning site located downstream. The vector also carries the neomycin resistance gene, which is driven by the 5 ’ long terminal repeat of the Moloney murine sarcoma virus. The expression construct was delivered into SPUO cells by electroporation, which was carried out by the Cell-Porator system (Gibco/BRL, Gaithersburg, MD), as described previously [28]. Transfectants were selected in complete Dulbecco’s Modified Eagle’s Medium containing G418 (Gibco/BRL) at 1.5 mg/ml (50% effective).

2.4. Cell lines and h4Abs The SP2/0-Ag14 cell line was obtained from Dr. Sherie Morrison (University of California School of Medicine, Los Angeles, CA). The KLE-B human endometrial carcinoma cell line [29] was provided by Dr. George Richardson (Massachusetts General Hospital, Boston, MA). KLE-B cells have been characterized by flow cytometry in our laboratory and constitutively express the TAG-72 antigen in vitro in 35%60% of cells. The IL-Zdependent murine cell line HT-2 has been previously described [ 131. The A375 human melanoma cell line was received from Dr. Stuart

234

L. Shu et al. /Immunotechnology 1 (1995) 231-241

Aaronson (National Cancer Institute, National Institutes of Health, Bethesda, MD). The development and properties of MAb CC49 have been previously described [30]. CCC49 y 1 and SCIgACnl were purified as reported earlier [ 181. PttriEed human IgG was supplied by Jackson Immune Research Laboratory, Inc. (West Grove, PA). 2.5. Purification and physical characterization of the fusion protein The fusion protein, SCIg-IL-2, was purified from tissue culture supematant of the producer clone grown in protein-free hybridoma medium (PFHM-II, Gibco/BRL). Protein G chromatography was carried out by using recombinant protein G-agarose (GibcoBRL) according to the supplier’s recommendation. A centricon 30 microconcentrator (Amicon, Beverly, MA) was used to concentrate the material eluted from the protein G column. The eluted material was electrophoreticaIly analyzed on precast sodium dodecyl sulfate (SDS) 8%16% polyacrylamide Tris glycine gels (Novex System, San Diego, CA) with and without 2-mercaptoethanol. The proteins were visualized by staining with Coomassie Blue R250. 2.6. Enzyme-linked immunoabsorbent assay and competition immunoassays Transfectoma culture supernatants and eluted material from protein G column chromatography were assayed for antigen binding by solid-phase enzyme-linked immunoabsorbent assay. Individual wells of the 96well polyvinyl microtiter plates were coated with 1 pg of the TAG-7Zpositive bovine submaxillary mucin (BSM) type I-S (Sigma Chemical Co., St. Louis, MO) or 20 rg of protein extract of TAG-7Znegative A375 human melanoma xenograft. The remainder of the assay was performed as previously described [31]. Competition assays were performed to determine the relative binding of the immunoglobulins to TAG-72 by a procedure described earlier [ 181. After the final step, the absorbance was read at a 490-nm

wavelength. Percent binding is the ratio of observed absorbance to the total absorbance multiplied by 100. An assay without the competitor was included to obtain a total absorbance value. 2.7. Antibody dependent cellular cytolytic activity ADCC activity of the immunoglobulins was measured by a 24-h ’’‘In-release assay using KLE-B human endometrial carcinoma cells as targets. To check whether IL-2 augments ADCC activity of the fusion protein, human peripheral blood mononuclear cells cultured for 24 h with and without recombinant human IL-2 (rh-IL-2) (100 U/ml) were used as effector cells in paraBe assays. The remainder of the assay and cakmlation of percent lysis have previously been described 1321. 2.8. Bioassays of IL-2 activity The biological activity of the IL-2 moiety of SCIg-IL-2 was determined by an assay previously described [33]. For this assay, which is based on proliferation of the IL-2-dependent cell line HT-2, the cells cultured in IL-2-containing medium were washed free of IL-2 and then seeded in complete RPM1 medium without IL-2 at a density of 1 x lo4 cells/well. The cell cultures were supplemented with different amounts of the purified SCIg-IL-2 or rh-IL-2 (Cetus Corporation, Emeryville, CA). After 24 h of stimulation, the cell lines were pulsed with ‘H-thymidine (1 &i/well) for 18-24 h and the incorporated radioactivity was measured by scintillation spectroscopy. For conversion of the molarity of SCIg-IL-2 to units of IL-Uml, the molar weight of the IL-2 moiety (M, 15) in the fusion protein (M, 139) was taken into consideration. IL-2 units were calculated on the basis of 4 U/ng of IL-2 protein. 3. Results 3.1. Design of the fusion protein The dimeric form of SCIg-IL-2 is schematically presented in Fig. 1. The design of the fusion pro-

L. Shu et al. /Immunotechnology1 (1995) 231-241

235

SCIgACu 1 that was described earlier [ 181, cys*” was replaced by a proline residue. 3.2. Generation of a single-gene encoding fusion proVH

s-s S-S

CH2

CH3

IL-2

tein and its expression in myeloma cells

Generation of the single-gene SGACr_Cu1IL-2 and its expression construct pLNCS23-IL-2 required the insertion of the human IL-2 gene, in construct pLNCS23 [18], 3’ to the sequence encoding the human yl chain. A 60 bp NsiIlHindIII fragment encoding the carboxyl end of the Cu3 domain was replaced by a PCR-generated sequence encoding the carboxyl end of Ct.,3 attached to the human IL-2 through the GGGSGGG peptide linker. The procedure is outlined as a flow sheet of DNA manipulations in Fig. 2. Briefly, a DNA sequence encoding the Cu3 and the carboxy1 half of the Ct.,2 domain of the human yl was amplified by using the construct pLGCKSN49

Fig. 1. Schematic diagram of the dimeric SCIg-IL-2 fusion protein. VH and VL, variable domains of the heavy and light chains; Cn2 and Cn3, heavy chain constant region domains of human IgGl; IL-2, human interleukin-2.

tein monomer, which is a modification of the single-chain immunoglobulin SCIgACul developed earlier [18], consists of the following: the carboxy terminus of the Vu and the amino end of the Vr domains of the MAb CC49 are covalently joined through a (GGGGS)s peptide linker, and the carboxyl end of the VL domain is linked to the amino terminus of the Cu2 region of the human yl Fc through the hinge region. The carboxy terminus of the Cu3 domain of the Fc, in turn, is covalently attached to the amino terminus of the human IL-2 through a short peptide linker, GGGSGGG. The formation of a dimer is ensured by retaining CYSTIC and CYSTIC of the functional hinge, which provide disulfide links between two heavy chains. Also, this design requires that cys**s, which is located in the genetic hinge and is involved in making a disulfide bond with cys214of the K light chain, be replaced with some other amino acid residue. In the single-chain immunoglobulin fusion protein SCIg-IL-2, like in

Fig. 2. Schematic Ilowsheet for the generation of the singiegene SGACLC~I-IL2 and its expression construct pLNCS23H-2. A flowsheet of DNA manipulations starting from the template DNAs, pLGCRSN49, and human IL2 cDNA clone and leading to the generation of the final gene SGACtCnIIL-2 and the expression construct pLNCS23-IL-2 are shown schematically. Horizontal thin arrows, primers used for DNA amplifmtion; CMV, human cytomegalovirus promoter; H-2, human interleukin-2 sequence; LTR, long terminal repeat; Neo, neomycin-resistance gene; Vt, and VR, sequences encoding variable domains of CC49 heavy and light chains; CR 1, Cu2, and Cn3, sequences encoding constant region domains of human y 1 chain.

236

L. Shu et al. /Immnotechnology

[ 181as a template and oligonucleotides 1 and 2 as the 5 ’ and 3 ’ primers, respectively. Similarly, a DNA sequence carrying informational content of the human IL-2 was amplified by using the human IL-2 cDNA clone [26] as a template and oligonucleotides 3 and 4 as the 5’ and 3 ’ primers, respectively. Primers 2 and 3 had a 25 bp complementing overlap, and part of the overlap encoded the linker peptide GGGSGGG. Thus, an overlap was generated between the 3 ’ end of the amplified DNA-encoding immunoglobulin domains and the 5 ’ end of the DNA amplified from the human IL-2 template. By using the two amplified DNAs as templates and oligonucleotides 1 and 4 as the end primers, a recombinant PCR was carried out to generate a single DNA molecule encompassing part of the human Fc, the linker peptide, and the human IL-2 sequences. This DNA was doubly digested with NsiIIHindIII, thereby generating a 470 bp DNA fragment that encoded the carboxyl end of the human yl Cn3 domain attached, through a short peptide linker, to the amino terminus of the human IL-2. The 470 bp NsiIIHindIII fragment was joined, in a three way ligation, to 2.3 kb BamHl/NsiI and 5.8 kb BamHl/HindIII DNA fragments that were generated from the construct pLNCS23. The resulting expression construct, pLNCS23-IL-2, carrying an approximately 1.9 kb gene (SGAC,_Cnl-IL-2) downstream from the human cytomegalovirus promoter, is shown in Fig. 2. The expression construct pLNCS23-IL-2 was electroporated into the SP2/0 mouse myeloma cells. Two weeks after the onset of selection in G418 medium, tissue culture supematants were assayed for reactivity to the TAG-R-positive BSM. Five of the twenty four culture supematants assayed were positive for TAG-72 reactivity. None of the supematants showed reactivity to the cell extract from the TAG-7Znegative A375 human melanoma. The cell clone with the highest reactivity to TAG-72 was chosen for further characterization. Cells were adapted to grow in serum-free and protein-free medium. The secreted fusion protein, designated SCIg-IL-2, was purified from the culture medium by protein G column chromatography. Approximately 200 ng of immunoreactive

I (1995) 231-241

SCIg-IL-2 was obtained per ml of the tissue culture supematant. SDS-polyacrylamide gel analysis of the fusion protein

electrophoresis

3.3.

The SCIg-IL-2 fusion protein, eluted from the protein G column, was analyzed by SDS-polyacrylamide gel electrophoreais to assess its purity and to determine its size. The results shown in Fig. 3 include purified SC1g~Cnl [ 181 for comparison. Under nonreducing conditions, SCIgACnl showed a single band with a molecular weight of approximately 140-150 kD (lane 1). Migration of SCIgACnl was slower than what would be expected of its 108-kD molecular mass. This phenomenon has been reported previously [18], and was attributed to the presence of intact sulfide bonds in the protein. SCIg-IL-2 showed one major band of approximately 180- 190 kD and a minor band of 65-70 kD under nonreducing conditions (lane 2). The migration pattern of the minor band suggests that the molecular mass of this protein is approximately one-half that of the protein that appeared as the major band. Under reducing conditions, the SCIgACnl appeared as a single band of 55-60 kD (lane 3),

1

2

3

4

M

kD

-

205

- 116 - 97 - 66 -

45

-

29

Fig. 3. SDS-PAGE analysis of the fusion protein SCIg-IL-Z. Protein G purified SCIgACHl (lanes 1 and 3) and SCIg-IL-2 (lanes 2 and 4) loaded with (lanes 3 and 4) and without (lanes 1 and 2) 2mercaptoethanol were analyzed on precast SDS/I% 16%polyacrylamide Tris glycine gels. The sizes of the molecular weight markers (M) are given on the right in kD.

L. Shu et al. /Immunotechnology 1 (1995) 231-241

while SCIg-IL-2 appeared as a single band of approximately 65-70 kD (lane 4). Thus, the molecular mass of SCIg-IL-2 is approximately 140 kD, which conforms with the value deduced from the nucleotide sequence of the synthetic gene. The size of the minor band observed for SCIg-IL-2 under nonreducing condition (lane 2) and the appearance of only one band of the reduced SCIgIL-2 (lane 4) strongly suggest that the minor band in lane 2 represents half-molecules generated because of partial reduction of the protein as artifacts of the purification process.

231

observed with SC&AC&l, cCC49, and nCC49. For example, 50% inhibition of the binding of biotinylated cCC49 with TAG-72 required approximately 7 nM of SCIg-IL-2 vs. approximately 2 nM of nCC49, cCC49, or SCIgACul. 3.5. Biological activity of the IL-2 moiety of sczg-IL-2

Experiments were undertaken to determine if the human IL-2 molecule, covalently attached through its amino terminus to the human Fc yl, was biologically functional. To this end, we compared stimulatory effect of SCIg-IL-2 and rh-IL-2 on the proliferation of IL-Zdependent HT-2 cells (Fig. 5). At lower doses (14 U/ml), rh-IL-2 showed a greater stimulatory effect on HT-2 cell proliferation. At doses of r5 U/ml, the stimulatory effect of SCIg-IL-2 on HT-2 cell proliferation was comparable to that of rh-IL-2. The reason for the relatively poor stimulatory potential of SCIg-IL-2 at lower concentrations remains unclear.

3.4. Antigen binding A solid-phase competition immunoassay was developed to compare the immune reactivity and relative binding affinity of SCIg-IL-2 with SCIgACul, cCC49, and nCC49. Unlabeled antibodies were used to compete with biotinylated cCC49 for binding to the TAG-R-positive BSM. The competition profile presented in Fig. 4 shows that all species of MAb CC49 competed completely and with similar slopes. However, approximately three-fold more of the SCIg-IL-2 was required to achieve similar levels of competition as those

100

80

60

40

0

20

12

3

4

5

6

7

8

9

10

Uhl

0 .01

10

100

1000

1c

nM

Fig. 4. Competition assay for binding of SCIg-IL-2 to TAG-72 antigen. SCIg-IL2 (a), SCIgAC”I (0), nCC49 (n), and CCC49@) were usfxl in increasing concentrations to compete for the binding of biotinylated CCC49to the TAG-72positive BSM.

0

Fig. 5. IL-2 activity of SCIg-IL-Z. IL-Z-dependent HT-2 cells suspended in complete RPM1 medium without IL-2 were supplementedwith different amounts of purified SCIg-IL-2 (0) or rh-IL-2 (0). Concentrations of the two proteins were normalized to U/ml of IL-2 by taking the molarities of the two proteins into consideration. After 24 h of stimulation, the cells were pulsed with [3H]-thymidine (1 &i/well). IL-2 activity was measured by 13H]incorporation into the HT-2 cells.

L. Shu et al. / Immunotechnology1 (1995) 231-241

238

3.6. Cytolytic activity

MAb cCC49 yl and the single-chain antibody SCIgACul mediate ADCC with human effector cells against cultured KLE-B human endometrial carcinoma cells that constitutively express TAG-72 on their cell surface [ 181. The antibodymediated lysis is augmented by prior treatment of the human effector cells with IL-2. Because the Fc region in SC&-IL-2 was modified by covalently joining its carboxyl end to a human IL-2 molecule, it was imperative to check whether it affected the cytolytic function of the molecule. Using KLE-B as target cells in a 24-h “‘In-release assay, we compared ADCC activity of SCIgACnl and SCIg-IL-2. Untreated and IL-Ztreated human peripheral blood mononuclear cells were used as effecters. The results presented in Fig. 6 show that at each effecter/target cell ratio the extent of cytolysis mediated by SCIgACul and SCIg-IL-2 was comparable. At an effecter/target ratio of 50: 1, both antibodies mediated approximately 32% of cell lysis, whereas an irrelevant human IgGl showed very low cytolytic activity (Panel A). Prior treatment of the human effector cells with 100 U/ml of human IL-2 substantially increased the level of cytolysis. At the highest effectoritarget cell ratio, SCIgACul- and SCIg-IL-Zmediated cytolysis reached 60% and 65%, respectively (Panel B).

A

Fig. 6. Antibody-dependent cell-mediated cytotoxicity of SCIg-IL-2. A 24-h “‘In-release assay was performed by using different ratios of the human peripheral blood mononuclear cells as effector cells (E) and lllIn-laheled KLE-B human endomttrial carcinoma cells as target cells (T). Five cg of SCIgAC”I (O), SCIg-IL-2 (0), or human IgGl (A) were used with effector cells cultured without (panel A) or with (panel B) 100 U/ml of rh-IL-2.

The increase in percent lysis using the control antibody is most likely attributed to increased LAK activity of effector cells due to IL-2 treatment. 4. Discussion The ability to genetically engineer antibodies has provided impetus to develop multifunctional reagents for cancer immunotherapy [ 14,16,17,3436]. Among these reagents, the IL-2 conjugates of anti-tumor antibodies are especially attractive, because such molecules offer several advantages, including (1) IL-2 per se is a potential anti-cancer reagent by virtue of its ability to activate endogenous immune mechanisms [ 11;(2) the selective reactivity of anti-tumor antibodies would facilitate targeting of IL-2 to the tumor site [37]; (3) the conjugated IL-2 moiety augments cytolytic activity of the antibody by stimulating effector cells [ 14,16,17]; and (4) the IL-2 moiety increases tumor vascularity, thus resulting in potential enhanced tumor uptake of the antibody [15]. MAb CC49, has shown a high degree of tumor localization in recent clinical trials [21-231. Based on the rationale that therapeutic efficacy of MAb CC49 might be enhanced by delivering human IL-2 to the tumor site with the MAb, a cCC49-IL-2 fusion protein has been developed. Unlike the antibody-IL-2 fusion proteins developed by others [16,17], this molecule, SCIgIL-2, is a single-chain protein composed of the CC49 heavy chain and light chain variable regions, the human yl Fc domain, and the human IL-2 molecule covalently linked to the carboxyl end of the Ct.,3 domain. The SCIg-IL-2 fusion protein, like nCC49 and SCIgACul, competed completely with cCC49 for binding to the TAG-72 antigen. However, SCIgIL-2 showed two- to three-fold lower binding affinity than the other variants of MAb CC49. This is in accord with our earlier report that the insect cell derived SCIg-IL-2 has approximately 3-fold lower binding affinity to TAG-72 than the insect cell derived SCIgACul, while the latter has the same binding affinity as the hybridoma derived CCC49 and SCIgACul [38]. The presence of the halfmolecules in the hybridoma derived SCIg-IL-2 preparation (Fig. 3, lane 2) could account for the

L. Shu et al. / Immunotechnology 1 (1995) 231-241

apparently lower binding affinity of SCIg-IL-2. The monomeric variants of CC49 have been shown to have approximately eight-fold lower relative binding affinity than their dimeric counterparts [39]. The lower antigen-binding affinity of SCIg-IL-2 could alternatively be attributed to an adverse effect of the low pH affinity purification. However, this explanation is not plausible, because SCIgACul, cCC49 and nCC49 that were used to compare the relative binding affinity of the SCIg-IL-2 were also purified by the same low pH affinity scheme. The cytolytic activity of the SCIg-IL-2 fusion protein for the TAG-72 positive KLEB cells was comparable to that of SCIgACul. When the human effector cells were pretreated with IL-2, SCIg-IL-Zmediated lysis reached a level of 60%-65%. This is in contrast to the report that ch14.18-IL-2 fusion protein potentiated the lysis of SK-N-AS neuroblastoma cells by human LAK cells more effectively than that by either ch14.18 or an admixture of ch14.18 and rh-IL-2 [37]. A 4-h “0-release assay that was used to measure the cytolytic activity of the ch14.18-IL-2 fusion protein could, perhaps, have demonstrated a more pronounced cytolytic activity of the SCIg-IL-2 compared to that of the SCIgACn 1. Nevertheless, our results suggest that administration of SCIgIL-2 in patients may stimulate natural killer cells to generate LAK activity, as was shown in neuroblastoma patients on administration of rh-IL-2 [40,41]. The IL-2 moiety of SCIg-IL-2 was shown to be biologically active in stimulating proliferation of IL-Zdependent HT-2 cells. At doses of r 1 rig/ml, the IL-2 moiety of SCIg-IL-2 was as active as equimolar concentrations of rh-IL-2. However, at doses of < 1 @ml, SCIg-IL-2 had substantially lower stimulating activity on I-IT-2 cell proliferation than rh-IL-2 did. It is also plausible that the nature of the linkage affects the cytokine activity of the fusion protein. ch225-IL-2 [17] and ch14.18-IL-2 [ 161are biologically as active as molar equivalents of rh-IL-2. In these fusion proteins, IL-2 is directly fused to the Cu3 domains of the respective immunoglobulin molecules. In contrast, the L6-IL-2 fusion protein, which was generated by joining IL-2 to the Fab of MAb L6 through a lfamino-acid residue peptide linker, was found to

239

be 200-fold less active than rh-IL-2 [14]. In SCIgIL-2, a seven-amino-acid GS linker peptide was used to join the IL-2 molecule to the Cu3 domain of the immunoglobulin. This linker may be responsible for the somewhat reduced biological activity of the SCIg-IL-2. A comparison of the SCIg-IL-2 molecules and the cCC49-IL-2 fusion protein molecules, made with and without GGGSGGG linker, could help determine whether the linker has any deleterious effect on the biological activity of the IL-2 moiety. Nevertheless, it has recently been shown that following particle mediated delivery of the pLNCS23-IL-2 construct in Balb/c mice, the serum level of the in vivo expressed SCIg-IL-2 reached a concentration of 2.6 ng/ml within 24 h [42]. At this concentration, the IL-2 moiety of SCIg-IL-2 would be as active as the equimolar concentration of rh-IL-2. Taken together, the results of antigen reactivity, cytolytic activity, and the IL-2 activity assay suggest that, in vivo, SCIg-IL-2 may target IL-2 to the tumor site, activate endogenous effector cells, and efficiently carry out lysis of the targeted tumor cells. It has been shown that ch14.18-IL-2 fusion protein suppresses growth of human neuroblastoma in a hepatic metastases model in SCID mice [37]. Similarly, SCIg-IL-2 may be a useful therapeutic reagent in treatment protocols for TAG-7Zpositive human carcinomas. It remains to be seen, whether this reagent will live up to its potential in preclinical testing and clinical trials. These studies will require large-scale production and purification of SCIg-IL-2 in various tissue culture systems. In searching for a cost effective system for large scale production and purification, the SCIg-IL-2 has recently been expressed in insect cells, using the baculovirus expression system [38]. A more attractive and cost-effective method for preclinical and clinical testing of this reagent may utilize the gene delivery approach. It may be pointed out that, unlike other immunoglobulin-IL-2 fusion proteins developed previously, SCIg-IL-2 is a single chain protein encoded by a single-gene construct. This characteristic may facilitate inoculation of the gene into animal tissue for in vivo expression of the fusion protein and ex vivo transfection of patients’ cells for adoptive therapy.

240

L. Shu et al. /Immunotechnology I (1995) 231-241

Acknowledgmeots

w1 Thor, A., Ohuchi, N., Szpak, C.A., Johnston, W.W. and

We thank Neema Soares and Diane J. Poole for their excellent technical assistance.

[211 Scot, A. M., Divgi, C. R., Welt, S., Kemeny, N., Finn,

Schlom, J (1986) Cancer Res. 46, 3118-3124.

References [II Smith, K.A. (1993) Blood 81, 1414-1423. 121Marincola, F.M. and Rosenberg, S.A. (1995) In Biologic Therapy of Cancer (DeVita, V.T., Hellman, S. and Rosenberg, S.A., eds.), J.B. Lippincott Co., Philadelphia, PA, pp. 250-262. [31 Lotze, M.T., Chang, A.E., Seipp, C.A., Simpson, C., Vetto, J.T. and Rosenberg, S.A. (1986) JAMA 256, 3117-3124. [41 Rosenberg, S.A., Lotze, M.T., Yang, J.C., Topalian, S.L., Chang, A.E., Schwartzentruber, D.J., Abersold, P., Lehman, S., Linehan, W.W., Seip, C.A., White, D.E. and Steinberg, SM. (1993) J. Natl. Cancer Inst. 85622-632. 151 Grimm, E.A., Robb, R.J., Roth, J.A., Neckers, L.M., Lachman, L.B., Wilson, D.J. and Rosenberg, S.A. (1983) J. Exp. Med. 1356-1361. 161 Komad, M.W., Hemstreet, G., Hersh, E.M., Manse], P.W.A., Metelsmann, R., Kolitz, J.E. and Bradley, E.C. (1990) Cancer Res. 50, 2009-2017. 171 Whittington, R. and Faulds, D. (1993) Drugs 46, 446-514. VI Feoron, E.R., Pardoll, D.M., Itaya, T., Golumbek, P. and Levitsky, HI. (1990) Cell 60, 397-403. PI Gansbacher, B., Zier, K., Daniels, B., Cronin. K., Banerji, R. and Gilboa, E. (1990) J. Exp. Med. 172, 1217-1227. DOI Ley, V., Langlade-Demoyen, P., Kourilsky, P. and Larson-Sciard, E. (1991) Eur. J. Immunol. 21, 851-854. [Ill Karp, S.E., Farber, A., Sale, J.C., Hwu, P., Jaffe, G., Asher, A.L., Shiloni, E., Restifo, N.P., Mule, J.J. and Rosenberg, S.A. (1993) J. Immunol. 150, 896-908. WI Primus, F.J., Finch, M.D., Masci, A.M., Schlom, J. and Kashmiri, S.V.S. (1993) Cancer Res. 53, 3355-3361. [I31 Yamaue, H., Kashmiri, S.V.S., Filippi, R., Nieroda, C., Yannelli, J.R., Tsang, K.W. and Schlom, J. (1994) J. Immunother. 16, 262-274. 1141 Fell, H.P., Gayle, M.A., Grosmaire, L. and Ledbetter, J.A. (1991) J. Immunol. 146, 2446-2452. u51 LeBerthon, B., Khawli, L.A., Alauddin, M., Miller, G.K., Charak, B.S., Mazumdur, A. and Epstein, A.L. (1991) Cancer Res. 51, 2694-2698. 1161 Gillies, S.D., Reilly, E.B., Lo, K-M. and Reisfeld, R.A. (1992) Proc. Natl. Acad. Sci. USA 89, 1428-1432. [I71 Naramura, M., Gillies, S.D., Mendelsohn, J., Reisfeld, R.A. and Mueller, B.M. (1994) Immunol. Lett. 39, 91-99. WI Shu, L., Qi, C-F., SchIom, J. and Kashmiri, S.V.S. (1993) Proc. Nat]. Acad. Sci. USA 90,7995-7999. WI Colcher, D., Hand, P.H., Nuti, M. and Schlom, J. (1981) Proc. Natl. Acad. Sci. USA 78, 3199-3203.

R. D., Pentlow, K., Old, L.J., Schlom, J. and Larson, SM. (1992) Eur. J. Nut. Med. 19, 569-570. w1 Gallinger, S., Reilly, R.M., Kirsh, J.C., Odze, R.D., Schmocker, B.J., Hay, K., Polihronis, J., Damani, M.T., Shpitz, B. and Stern, H.S. (1993) Cancer Res. 53, 271-278. 1231 Murray, J. L., Macey, D. J., Rosenblum, M. G., Kasi, L. P., Grant, E. J., Zhang, H-Z., Katz, R. L., Rieger, P. T., LeBherz, D., Bhadkamkar, V. P., South, M. S., Booser, D. J., Dhingra, K., Greiner, J. W., Schlom, J. and Podoloff, D. A. (1994) Ninth Intl. Conf. on Monoclonal Antibody Immunoconjugates for Cancer (Abstract) 1241 Kashmiri, S.V.S., Shu, L., Padian, E.A., Milenic, D.E., Schlom, J. and Hand, P.H. (1995) Hybridoma 14, 461-473. v51 Sawyer, J.R., Schlom, J. and Kashmiri, S.V.S. (1994) Protein Eng. 7, 1401-1406. WI Holbrook, N.J., Smith, K.A., Fomace, A.J., Comeau, CM., Wiscocil, R.L. and Crabtree, G.R. (1984) Proc. Nat]. Acad. Sci. USA 81, 1634-1638. 1271 Tsang, K.Y., Kashmiri, S.V.S., DeFillipi, R., Qi, C-F., Calve, B., Shu, L., Nieroda, C.A., Greiner, J.W. and Schlom, J. (1993) J. Immunother. 13, 143-152. P81 Slavin-Chiorini, D.C., Hand, P.H., Kashmiri, S.V.S., Calvo, B., Zaremba, S. and Schlom, J. (1993) Int. J Cancer 53, 97-103. WI Richardson, G.S., Dickerson, G.R., Atkins, L., MacLaughlin, D.T., Roam, S., Merck, L.P. and Bradley, F.M. (1984) Gynecol. Oncol. 17, 213-230 I301 Muraro, R., Koroki, M., Wunderlich, D., Poole, D.J., Colcher, D., Thor, A., Greiner, J.W., Simpson, J.F., Molinolo, A., Noguchi, P. and Schlom J. (1988) Cancer Res. 48, 4588-4596. [311 Hand, P.H., Calvo, B., Milenic, D., Yokota, T., Finch, M., Snoy, P., Garmestani, K., Gansow, O., Schlom, J. and Kashmiri, S.V.S. (1992) Cancer Immunol. Immunother. 35, 165-174. 1321 Hutzell, P., Kashmiri, S., Colcher, D., Primus, F.J., Hand, P.H., RoseIIi, M., Finch, M., Yarranton, G., Bodmer, M., Whittle, N., King, D., Loullis, C.C., McCoy, D.W., Callahan, R. and Schlom, J (1991) Cancer Res. 51, 181-189. [331 Abrams, S.I., Wang, R., Munger, W.E. and Russel, J.H. (1993) Cell. Immunol. 147, 388-202. 1341 Gillies, S.D., Young, D., Lo, K-M. and Roberts, S. (1993) Bioconjugate Chem. 4, 230-235. [351 Gillies, SD., Young, D., Lo, K-M., Foley, S.F. and Reisfeld, R.A. (1991) Hybridoma 10, 347-356. [361 Shin, S.U. and Morrison, S.L. (1990) Proc. Nat]. Acad. Sci. USA 87, 5322-5326. 1371 Sabzevari, H., Gillies, S.D., Mueller, B.M., Pancook, J.D. and Reisfeld, R.A. (1994) Proc. Natl. Acad. Sci. USA 91, 9626-9630. [381 Bei, R., Scblom, J. and Kashmiri, S.V.S. (1995) J. Immunol. Methods 186, 245-255.

L. Shu et al. /Immunotechnology 1 (1995) 231-241 [39] Milenic, D.E., Yokota, T., Filpula, D.R., Finkelman, M.A.J., Dodd, SW., Wood, J.F., Whitlow, M., boy, P. and Schlom, J. (1991) Cancer Res. 51, 6363-6371. [40] Hank, J.A., Weil-Hillman, G., Surfus, J.E., Sosman, J.A. and Sondel, P.M. (1990) Cancer Bnmunol. Immunother. 31, 53-59.

241

[41] Truit, R.L., Piaskowski, V., Kirchner, P., McOlash, L., Camitta, B.M. and Caspter, J.T. (1993) Immunother. 11, 274-285. [42] Nicolet, C.M., Burkholder, J.K., Gan, J., Gulp, J., Kashmiri, S.V.S., Schlom, J., Yang, N-S. and Sondel, P.M. (1995) Cancer Gene Ther. 2, 161-170.