Folate targeting of haptens to cancer cell surfaces ... - Springer Link

8 downloads 30 Views 260KB Size Report
therapies by activating autologous immune effector cells. [e.g. cytotoxic T cells, natural ... these modulators enhance immune cell killing of anti- body-coated cells ...
Cancer Immunol Immunother (2002) 51: 153–162 DOI 10.1007/s00262-002-0266-6

O R I GI N A L A R T IC L E

Yingjuan Lu Æ Philip S. Low

Folate targeting of haptens to cancer cell surfaces mediates immunotherapy of syngeneic murine tumors

Received: 25 September 2001 / Accepted: 3 January 2002 / Published online: 19 March 2002  Springer-Verlag 2002

Abstract A variety of human cancers overexpress a cell surface receptor with high affinity for the vitamin, folic acid (Kd 10–10 M). Covalent attachment of therapeutic agents to folic acid has been shown to allow efficient targeting of the folate–drug conjugates to folate receptor-expressing cancer cells, with little or no uptake by normal tissues except the kidneys. We report here the use of folate’s ability to deliver attached molecules specifically to cancer cells to convert poorly immunogenic tumors into highly immunogenic tissue targets. By linking folic acid to a model hapten, we have been able to decorate folate receptor-expressing cancer cell surfaces with >106 haptens/cell in vivo. Following marking of such cells with haptens, the cells are observed to become opsonized with autologous anti-hapten antibodies, which is presumed to mediate cell removal via antibodydependent cellular cytotoxicity (ADCC). Supplemental administration of low levels of ADCC-activating cytokines [e.g. interleukin-2 (IL-2) and interferon-alpha (IFN-a)] has been shown to synergize with the folatetargeted immunotherapy. Thus, using M109 syngeneic lung cancer cells injected intraperitoneally into Balb/c mice that were previously immunized against fluorescein, a significant extension of life span is observed following treatment with folate–fluorescein conjugates, and complete cures are observed upon supplementation with moderate levels of IL-2 and IFN-a. Because control tumor-bearing mice treated with the same cytokines but with non-targeted fluorescein show no extension of life span, we conclude that tumor-specific opsonization is an

Y. Lu Æ P.S. Low (&) Department of Chemistry, 1393 Brown Building, Purdue University, West Lafayette, IN 47907, USA E-mail: [email protected] Tel.: +765-494-5273 Fax: +765-494-0239 Present address: Y. Lu Endocyte Inc., 1205 Kent Avenue, West Lafayette, IN 47906, USA

essential step in this immunotherapy. Finally, because the anti-fluorescein antibodies are unable to access the folate receptors on the apical membranes of the kidney proximal tubules, no kidney or other normal tissue cytotoxicity is observed. These data suggest that retargeting of haptens to folate receptor-expressing cancers might constitute a method for mobilizing the immune system specifically against poorly immunogenic tumors. Keywords Cancer immunotherapy Æ Cytokine Æ Folate receptor Æ Folate-targeted hapten Æ Humoral immune response Abbreviations ADCC antibody-dependent cellular cytotoxicity Æ BSA bovine serum albumin Æ ELISA enzyme-linked immunosorbent assay Æ FITC fluorescein isothiocyanate Æ FR folate receptor Æ IFN interferon Æ IL interleukin Æ mAb monoclonal antibody

Introduction Eradication of tumor cells by stimulation of the host’s immune system has long been considered a promising strategy for the treatment of cancer [12]. With the intention of recruiting greater immune system involvement, stimulation of a cancer patient with cytokines [4], vaccination with tumor-associated antigens [8], treatment with monoclonal antibodies (mAbs) [41], supplementation with tumor-specific T cells [22], infusion of antigenprimed dendritic cells [34], and administration of immune-stimulating innate antigens [33] have all been explored as means of promoting systemic elimination of diseased tissue. Ideally, patients should respond to such therapies by activating autologous immune effector cells [e.g. cytotoxic T cells, natural killer (NK) cells, and macrophages] that destroy all tumor cells and induce sustained immunity to further tumor challenge. In practice, while several of the strategies indeed show considerable promise, others have failed in clinical trials due to tumorinduced immune dysfunction arising primarily from the

154

ability of the tumor to: (a) down-regulate its HLA molecules; (b) suppress expression of immune cell co-receptors; (c) inactivate the TAP-dependent peptide processing pathway; (d) secrete immune inhibitory factors; and/or (e) induce tolerance in host immune cells [3, 11, 31]. Apparently, the natural immunogenicity of many tumors is insufficient to enable complete tumor eradication by many of the proposed methodologies. In an effort to improve the immunogenicity of cancer cells that would normally escape immune surveillance, we have envisioned a two-step method that would force the thorough scrutiny of cancer cells by the immune system. In the first step, the tumor cells are converted from a poorly immunogenic state to a highly immunogenic state by the targeted enrichment of their cell surfaces with a novel hapten against which a strong immune response has already been elicited. While in humans the hapten could conceivably involve an antigenic determinant from any pathogen against which the patient has been previously immunized (e.g. tetanus, typhoid, measles, etc.), in our mouse tumor models we have employed the hapten, fluorescein. Further, because many human and murine tumor cells overexpress a high affinity receptor for the vitamin folic acid [5, 18, 32, 36, 39], this hapten has been specifically targeted to tumor cells by covalent ligation to folic acid [38]. In step two of the immunotherapy, the immune system, which can now recognize the tumor cells because of their extensive opsonization with anti-hapten antibodies, is stimulated with non-toxic levels of immunostimulatory cytokines to assure that Fc-expressing immune cells mediate removal of all marked tumor cells. For this study, we have selected interleukin 2 (IL-2) and interferon-alpha (IFN-a) as the stimulatory cytokines, since these modulators enhance immune cell killing of antibody-coated cells (ADCC) [15, 16, 20, 37]. We report here the use of folate–fluorescein conjugates to decorate the surfaces of M109 syngeneic murine lung cancer cells with large numbers of haptens against which a strong antibody titer has been induced. We demonstrate that while a minor extension of life span is observed upon treatment with folate–fluorescein conjugate alone, significant improvement of life span or complete cures can be achieved by supplementation with low to moderate concentrations of IL-2 and IFN-a.

monoclonal anti-FITC antibody IgG1j was purchased from Alexis (San Diego, Calif.). Human recombinant IL-2 (‡1·107 IU/mg) was obtained from Chemicon Int. (Temecula, Calif.) as a lyophilized powder with no additives. Human recombinant IFN-a A/D (1.17·108 IU/mg) was purchased from Research Diagnostics (Flanders, N.J.). For in vivo use, IL-2 or IL-2 plus IFN-a were prepared in sterile phosphate-buffered saline (PBS; 136.9 mM NaCl, 2.68 mM KCl, 8.1 mM Na2HPO4, 1.47 mM KH2PO4, pH 7.4) containing 1% syngeneic mouse serum and stored at –80C in small aliquots. Cell culture Two spontaneous lung carcinoma cell lines of Balb/c origin were used in these studies. M109, a high folate receptor (FR)-expressing, Madison lung carcinoma subclone was a generous gift from A. Gabizon (Sharet Institute of Oncology, Hadassah-Hebrew University Medical Center and Medical School, Jerusalem, Israel) [17]. Line 1, a low FR-expressing, Balb/c alveolar lung carcinoma, was a kind gift from Aventis Gencell (Hayward, Calif.). Both cell lines were cultured in folate-deficient RPMI 1640 medium (Gibco BRL) supplemented with 10% v/v heat-inactivated fetal calf serum (FCS), 100 IU/ml penicillin, and 100 lg/ml streptomycin. Murine tumor models All animal experiments were carried out in accordance with procedures approved by the Purdue Animal Care and Use Committee. Female Balb/c mice were purchased from Harlan Sprague–Dawley (Indianapolis, Ind.) and used when they reached 6 to 8 weeks of age. M109 tumors were maintained in Balb/c mice by subcutaneous passage at 3-week intervals. Tumors were harvested when desired, and regenerated according to an established procedure [17]. For intraperitoneal (i.p.) tumor implants, 5·105 viable tumor cells at an early passage (P0 or P1) were suspended in 400 ll of folate-deficient RPMI 1640 medium supplemented with 1% syngeneic mouse serum (antibiotic-free) and injected into the peritoneal cavity. For subcutaneous (s.c.) tumor implants, 1·106 cells per 100 ll volume were injected into the shoulder region. Line 1 i.p. tumors were similarly established by injection of 2.5·105 cultured tumor cells per animal. Folate–hapten (FITC) conjugate preparation The structure of folate–FITC is shown in Fig. 1. Folate–FITC (Mr 872.2) was synthesized by coupling FITC with folate–(c)-ethyleneamine in anhydrous dimethylsulfoxide. Ether-precipitated crude product was purified by preparative high-performance liquid chromatography (HPLC) and analyzed by mass spectrometry. All folate–FITC preparations were at least 95% pure, as demonstrated by analytical HPLC. The affinity of folate–FITC for cell surface FR was examined by fluorescence measurements of its binding to cancer cells in the presence and absence of excess folic acid [38]. These analyses confirmed that the folate–FITC conjugate exhibits a similar affinity for FR to free folic acid (data not shown).

Materials and methods Reagents Bovine serum albumin (BSA), keyhole limpet hemocyanin, goat anti-mouse IgG–horseradish peroxidase (HRP) and IgG–phycoerythrin conjugates were purchased from Sigma Chemical Co. (St Louis, Mo.). Biotin-conjugated goat anti-mouse IgG1, IgG2a, IgG2b, IgG3, IgM, IgA, and streptavidin-conjugated HRP were obtained from Caltag Laboratories. Folate-(c)-ethyleneamine was supplied by Endocyte (West Lafayette, Ind.). Fluorescein isothiocyanate (FITC; isomer I) and (5)-aminofluorescein (single isomer) were purchased from Molecular Probes (Eugene, Ore.). TiterMax Gold adjuvant was obtained from CytRx (Norcross, Ga.). Mouse

Fig. 1 Structure of the folate–hapten construct, folate–ethylenediamine–fluorescein (folate–FITC). Dashed lines indicate the boundaries between the three parts of the molecule

155 Immunization For induction of anti-FITC antibodies, BSA was used as the carrier and was coupled to FITC at a FITC-to-protein molar ratio of 13, according to published procedures [9]. Mice were immunized at adjacent sites at the base of the tail with 50 lg of BSA–FITC in 100 ll TiterMax Gold adjuvant (emulsions prepared according to the manufacturer’s instructions). Four weeks later, mice were boosted with 80% of the primary antigen dose injected on the back of the neck. Ten days after this secondary immunization, blood samples were collected (under anesthesia) through retro-orbital sinus puncture and the serum was stored at –20C until enzymelinked immunosorbent assay (ELISA) analysis. ELISA analysis Mouse serum samples collected from immunized animals were assayed for anti-FITC antibody production according to a published method, with minor modifications [24]. Briefly, 96-well ELISA plates coated with keyhole limpet hemocyanin–FITC (synthesized as described for BSA–FITC; 2 lg/well, overnight) were saturated with 0.2% gelatin in PBS containing 0.05% Tween-20 for 1 h at 37C. After washing twice with buffer consisting of 0.05% Tween20 in PBS, serial dilutions of pooled FITC-antiserum or preimmune serum were added to the plates (50 ll/well) and incubated for 1 h at 37C. The plates were washed four times and then incubated with a goat anti-mouse IgG–HRP conjugate (Sigma). Following another 1 h incubation at 37C and a subsequent wash, the presence of mouse anti-FITC antibodies was revealed by adding 200 ll/well of o-phenylenediamine dihydrochloride in substrate buffer (Sigma Fast o-Phenylenediamine Dihydrochloride Tablet Sets). The enzymatic reaction was stopped after 30 min in the dark at room temperature by the addition of 50 ll 3 N HCl, and optical density was evaluated at 490 nm using a 96-plate reader. Frequently, serum samples from individual mice were analyzed at a fixed serum dilution factor to evaluate the consistency of immunization within each set of animals. For quantitation of anti-FITC IgG1, IgG2a, IgG2b, IgG3, IgM and IgA antibodies, ELISA was carried out using biotin-conjugated goat anti-mouse IgG1, IgG2a, IgG2b, IgG3, IgM and IgA (Caltag) as the primary antibodies and streptavidin-conjugated HRP (Caltag) as the secondary antibody. Flow cytometric analyses Two-colored flow cytometric analyses were performed on M109 cells for different purposes. To examine whether FITC-antiserum could bind to folate-FITC decorated tumor cells, monolayers of cells were gently dissociated from culture flasks using a non-enzymatic cell dissociation solution (Sigma). Cells were centrifuged at 800 rpm for 7 min and suspended in culture medium at 1·107 cells/ ml. The cell suspension was then incubated with 0.1 ml of medium containing 0.2 lM folate–FITC for 30 min on ice, washed three times in PBS containing 1% BSA and 0.1% sodium azide, pelleted as above, and resuspended in the same buffer. The folate–FITC bound cells were then treated for 1 h on ice with the following: (a) PBS; (b) 10 lg/ml of anti-FITC mAb IgG1j; or (c) a 1:55 dilution of pooled FITC-antiserum or preimmune serum. Thereafter, the cells were again washed and stained at room temperature for 30 min with a 1:40 dilution of goat anti-mouse IgG–phycoerythrin conjugate (as the secondary antibody). The cells were then subjected to a final wash and suspended in wash buffer for flow cytometric analysis. To examine the binding of preformed folate–FITC–IgG immune complexes to FR-bearing tumor cells, pooled FITC-antiserum or preimmune serum was incubated at room temperature for 1 h with folate–FITC at a concentration of 11 nmol folate–FITC/ ml serum. Unbound folate–FITC was removed by gel filtration on a PD-10 column (Pharmacia) eluted with PBS. Protein fractions collected in small elution volumes were combined and the volume was adjusted to achieve a final serum dilution of 1:10. Equal

volumes of pooled FITC-antiserum or preimmune serum (without folate–FITC incubation) were also passed through a PD-10 column under identical conditions and used as controls. M109 cells at a density of 1·107/ml were suspended in culture medium and incubated for 2 h at 4C with 1:20 serum dilutions of pooled FITCantiserum complexes or preimmune serum pretreated with or without folate–FITC. Thereafter, the cells were washed and stained with a 1:40 dilution of goat anti-mouse IgG-phycoerythrin conjugate. After 30 min incubation at room temperature, the cells were again washed and suspended in wash buffer for analysis.

Immunofluorescence evaluation To investigate the co-localization of anti-FITC antibodies with folate–FITC conjugates in M109 tumors during immunotherapy, immunized mice were implanted s.c. on the shoulders with 1·106 tumor cells and tumors were grown to 50 mm3 in a week (designated as day 0). The mice were then i.p. injected at 48-h intervals (days 0, 2, 4, 6, 8, and 10) with PBS or folate–FITC (1,500 nmol/kg) in addition to five daily treatments with IL-2 (5,000 IU/dose/day) on days 5 to 9. On days when both folate– FITC and IL-2 were administered, an 8-h interval between the separate injections was maintained. One day after the last treatment of folate–FITC, the tumors were excised, washed briefly with PBS, and blotted dry. After embedding in OCT resin (Miles Inc., Kankakee, Il.), the tumor tissues were frozen by dipping in liquid N2 and stored sealed at –80C. Thereafter, 5-lm cryostat sections of tissue were mounted on polylysine-coated Superfrost microscopy slides (Fisher Scientific) and used for immunofluorescence analysis. To prepare slides for confocal microscopy, frozen tissue sections were allowed to equilibrate at room temperature and air-dried in the dark for at least 1 h. Tissue sections were then fixed in cold acetone (4C) for 10 min, allowed to dry, and rehydrated in PBS for 10 min. After removal of excess PBS and OCT resin, nonspecific binding was blocked by incubating the tissue sections first with PBS containing 1% BSA and 0.1% sodium azide for 5 min followed by 2·2 min washing in PBS. Tissue sections were then incubated with 300 ll/slide of phycoerythrin-conjugated antimouse IgG secondary antibody at 0 or 1:27 dilution in PBS containing 1% BSA and 0.1% sodium azide. Thereafter, tissue sections were subjected to a final round of washing in PBS (3·2 min) and dried before examination under a confocal microscope. Tissue-associated FITC and phycoerythrin fluorescence were imaged with a BioRad MRC-1024 UV/Vis Confocal Laser Scanning Microscopy system equipped with a Nikon Diaphot 300 inverted microscope, a 488-nm argon laser, a 10· Fluor Phase objective, a 522/35 PMT filter for FITC and a 580/32 PMT filter for phycoerythrin. To avoid autofluorescence, tissue sections from the PBS-treated mice were processed identically and used as controls to guide instrument settings. Tissue transmission images were collected on the identical image fields.

Survival studies Unless otherwise specified, all FITC-immunized animals were i.p. implanted with M109 tumor cells 18 to 19 days after the secondary immunization against fluorescein. To reduce mouse serum folate levels to values characteristic of humans (20 nM), mice were maintained on a folate-deficient diet from the day of secondary immunization until the end of the treatment period, which was normally not more than 4 weeks. The desired immunotherapy regimens were administered (all i.p.) on the days indicated in the figure legends. After treatment, mice were randomized to avoid cage effects and evaluated daily for survival. Although the median survival of untreated tumor-bearing mice varied from 19 to 23 days over the course of the studies, all mice in any one experiment were inoculated on the same day with the same batch of cells, thereby assuring that the data in any one experiment were comparable.

156 Treatment regimen I In this immunotherapy regimen, folate–FITC was administered prior to the initiation of cytokine treatment. Thus, on days 4 and 7 after tumor implantation, BSA–FITC immunized mice were injected with either 0.3 ml PBS or 1,500 nmol/kg folate–FITC in PBS. Treatment with IL-2 (5,000 IU/dose in 0.2 ml PBS containing 1% syngeneic serum) was then given on days 8 to 12. Treatment regimen II In this immunotherapy regimen, folate–FITC and immunomodulatory cytokines were administered simultaneously. Thus, BSA– FITC immunized mice bearing i.p. M109 tumors were treated with folate–FITC (1,500 nmol/kg) in combination with IL-2 (5,000 or 250,000 IU/dose) and IFN-a (25,000 IU/dose) on days 7, 8, 9, 11, and 14 post-tumor cell implantation. (5)-aminofluorescein (1,500 nmol/kg), used as the control for a non-targeted hapten, was examined under the same protocol as described in the figure legends. When desired, long-term survivors (>60 days) that appeared healthy and tumor-free were challenged i.p. with fresh M109 tumor cells on both day 62 (5·105 cells) and day 96 (1.5·106 cells) after the initial tumor cell injection. An observation period of >30 days was allowed after each new challenge to monitor animal tumor development/survival. No additional folate–FITC and/or cytokine treatments were given to any of these rechallenged mice. Finally, rechallenged mice that remained healthy 127 days after the initial tumor implantation were again challenged by an i.p. injection of 2.5·105 line 1 syngeneic lung cancer cells and their daily survival was monitored.

Statistical analysis Statistical analysis of survival data was performed using the logrank test in the computer program GraphPad Prism (GraphPad Software Inc., San Diego, Calif.). Differences in survival were considered significant when P £ 0.05.

Results Induction of anti-hapten antibodies in mice To generate anti-FITC antibodies, mice were immunized twice with a total of 90 lg BSA–FITC in TiterMax Gold, a non-ionic block co-polymer surfactant known to stimulate a strong and prolonged antigenspecific IgG response in experimental animals [1]. ELISA analysis of individual or pooled mouse serum collected 10 days after the second immunization demonstrated a high titer of anti-FITC IgG in all immunized animals, with 50% of maximum binding at 1:3,220 serum dilution. Analysis of anti-FITC isotypes further indicated a strong IgG1 response, a moderate IgG2a response, and a relatively weak induction of FITCspecific IgG2b and IgG3 (titer ratios of 1: 0.28: 0.11: 0.11). FITC-specific IgA and IgM were not measurable. The consistency of immunization within an immunotherapy protocol was examined by assaying a fixed dilution of individual serum samples collected from all immunized animals. In general, inter-individual variations of anti-FITC antibody responses were less than 16%, and any occasional low responders were excluded from the study.

Targeting of specific antibodies to FR-positive tumor cells in vitro The ability of folate–FITC to promote opsonization of FR-positive tumor cell surfaces with anti-FITC antibodies was evaluated under two different conditions (Fig. 2). First, to determine whether cancer cell-associated folate–FITC is capable of promoting anti-FITC antibody binding, M109 cells pre-incubated with folate– FITC were treated with pooled FITC-antiserum, commercial anti-FITC mAb (positive control), or preimmune serum followed by phycoerythrin-conjugated secondary antibody. As shown in Fig. 2B and C, flow cytometric analyses detected both FITC and phycoerythrin fluorescence on cells treated with either pooled FITC-antiserum or commercial anti-FITC mAb. In contrast, the phycoerythrin fluorescence associated with preimmune serum exposure was negligible (Fig. 2A). These results suggest that anti-FITC antibodies are capable of recognizing the hapten on folate– FITC molecules already bound to FR on cancer cell surfaces. Since i.v. administered folate–hapten conjugates could also conceivably form complexes with circulating anti-hapten antibodies prior to extravasation into the tumor, it was also important to determine whether preformed folate–FITC/anti-FITC antibody complexes could still recognize cell surface FR. Therefore, folate– FITC/anti-FITC antibody complexes were preformed by incubating folate–FITC with pooled FITC-antiserum, anti-FITC mAb, or preimmune serum, and the purified immune complexes were tested for binding to M109 tumor cells. Cell surface-bound immune complexes were then detected with the same anti-mouse IgG–phycoerythrin conjugate. As seen in Fig. 2E and F, immune complex binding was again observed on the M109 cancer cell surfaces, suggesting that the sequence of formation of folate–FITC/anti-FITC immune complexes does not affect the opsonization of the tumor cells. Therefore, as long as both cell surface FR and circulating/interstitial anti-FITC IgG are not simultaneously saturated with folate–FITC, the vitamin–hapten conjugate should be capable of bridging anti-hapten antibodies to the cancer cell surface. Folate–FITC mediated binding of antibodies to tumor tissues in vivo To provide evidence for co-localization of folate–FITC and anti-FITC antibodies in the tumor during folatetargeted immunotherapy, BSA–FITC immunized mice bearing 50 mm3 s.c. M109 tumors were treated with folate–FITC (1,500 nmol/kg) or PBS in conjunction with low levels of IL-2 (5,000 IU/dose). One day following termination of folate–FITC treatment, all mice were killed, residual tumors were resected, and tumor sections were stained with goat anti-mouse IgG–phycoerythrin in preparation for viewing by

157

Fig. 2 Flow cytometric analyses of the ability of folate–FITC to mediate anti-fluorescein IgG binding to FR-expressing cancer cells. A–C M109 lung cancer cells were incubated with folate–FITC to saturate cell surface FR and washed to remove any unbound folate–hapten conjugates. The hapten-decorated cells were then incubated with A) preimmune serum; B) pooled FITC-antiserum at 1:55 serum dilution; or C) anti-FITC mAb at 10 lg/ml, and cell surface IgG was quantitated by labeling with phycoerythrinconjugated goat anti-mouse IgG antibody prior to analysis by flow cytometry. D–F Recognition of M109 tumor cells by preformed folate–FITC/anti-FITC antibody complexes was determined by incubating D) preimmune serum, E) pooled FITC-antiserum, or F) anti-FITC mAb with folate–FITC and antibody–hapten complexes were separated from free folate–FITC by gel filtration chromatography on a PD-10 column. The preformed non-immune or immune complexes were allowed to bind M109 tumor cells, and cell surface IgG were detected with the goat anti-mouse IgG–phycoerythrin conjugate, as described above. Two-color histograms show logarithmically increasing fluorescence intensity of FITC plotted on the X-axis and phycoerythrin (shown as PE) plotted on the Y-axis. The percentages of FITC-positive, phycoerythrin-positive, or dual positive cells are indicated in the corresponding quadrants

immunofluorescence microscopy. As noted in Fig. 3, strong FITC and phycoerythrin fluorescence was observed in tumor specimens from folate–FITC treated mice (Fig. 3H and I), but not in specimens from PBStreated mice (Fig. 3B and C). The phycoerythrin fluorescence associated with non-stained adjacent tumor specimens from folate–FITC treated mice was negligible (Fig. 3F). These data thus confirm that folate– FITC can aggressively mark tumor tissues in vivo and that autologous antibody binding follows hapten localization, as observed in vitro. Importantly, folate– FITC binding was not observed in any normal tissues except the kidneys, and antibody opsonization was not

observed in the kidneys. We presume that the absence of kidney opsonization derives from the fact that the kidney FR resides on the apical membrane of the proximal tubules [2, 10, 40], i.e. a site that is accessible to solutes excreted into the urine, but inaccessible to serum proteins retained in the blood. These data, therefore, suggest that intravenous folate–FITC administration provides a means of inducing the selective opsonization of tumor tissues in vivo. Synergistic effect of folate–FITC and IL-2 on the survival of mice implanted intraperitoneally with M109 tumors Because the FR is especially over-expressed in metastatic tumors [36], we decided to evaluate folate-targeted immunotherapy using a metastatic i.p. tumor model expressing high levels of FR. For this purpose, M109 lung cancer cells were injected i.p. into syngeneic Balb/c mice, and the tumor-bearing animals were treated on days 4 and 7 with 1,500 nmol/kg folate– FITC followed by five daily doses of IL-2 (5,000 IU/ dose), as outlined in regimen I. As shown in Fig. 4, treatment with either folate–FITC or IL-2 alone did not result in significant antitumor activity, as measured by prolongation of median survival time. However, a combination of the two drugs dramatically increased median survival from 18 to 42 days (P106 FR /cell are distributed over the cancer cell surface. Where examined, FR expression appears to be even further enhanced in cancers resistant to standard chemotherapy [28] and in higher grade and later stage neoplasms [36, 42]. Thus, cancers that are difficult to treat by classical methods may be most easily attacked with folate-targeted therapies. Because folate conjugates can bind to cell surface FR with affinities similar to free folate [30], a number of new cancer treatments have been explored based on folate targeting of: (a) protein toxins; (b) chemotherapeutic agents; (c) genes; (d) oligonucleotides; (e) ribozymes; (f) liposomes with entrapped drugs; (g) radiotherapeutic agents; and (h) enzyme constructs for prodrug therapy [30, 35]. We have presented here the first application of folate targeting to the field of immunotherapy. Our strategy has been to first mark the cancer cells with foreign haptens and their accompanying high affinity antibodies, and then to boost the immune system to attack marked cells by stimulation with non-toxic levels of cytokines. Marking the tumor cells with haptens and anti-hapten antibodies was shown to be important, since: (a) substitution of a non-targeted hapten, (b) failure to immunize against the targeted hapten, or (c) reduction in the quantity of folate-hapten administered all abrogated the therapy. Similarly, boosting the immune system was

161

found to be required, since failure to treat with IL-2 or IFN-a also rendered the therapy ineffective. In contrast, when both folate–FITC and IL-2/IFN-a were employed either sequentially or simultaneously, strong anti-tumor activity was observed. Clearly, synergy between both arms of the bimodal therapy constitutes an important element of the hapten-targeted response. One of the attractive features of the therapy lies in the apparent absence of toxicity to normal tissues. While cytokine toxicity can be readily controlled by limiting cytokine doses to non-toxic levels, folate–FITC toxicity must naturally correspond to the tissue distribution of the opsonizing anti-hapten antibodies. Because FR is expressed in substantial quantities not only in cancer tissue, but also in the kidneys and choroid plexus [39, 40], one would logically anticipate that the immune system would target these tissues also. However, our data reveal no such opsonization of kidneys or choroid plexus (M. Kennedy and P. S. Low, unpublished data), and toxicology studies detect no damage to the tissues either. The explanation for this outcome probably lies in FR accessibility. Thus, brain FR have been found to reside on the brain side of the blood–brain barrier [29], and kidney FR are known to be located on the luminal/ apical membrane of the proximal tubule epithelium [2, 10, 40]. In neither case would circulating antibodies be expected to gain access to the hapten-enriched membranes, unless the normal epithelial barrier became pathologically disrupted. For most cancers, this should not be the case. Although no mechanistic studies were conducted here, it is still possible to speculate on the processes involved in hapten-mediated tumor cell destruction. Following opsonization, tumor cells should be primed for antibody-dependent cellular cytotoxicity (ADCC). In this pathway, Fc receptor-expressing immune cells, such as NK cells and macrophages, recognize opsonized cell surfaces and initiate killing/phagocytosis of the marked cells. Complement, if present, can also enhance the killing/removal process. Long-term protective immunity might then arise through subsequent presentation of tumor cell components to T cells that somehow recognize a tumor antigen among the presented material. Expansion of tumor-specific CD8+ T cell clones could then enable the observed rejection of subsequent tumor cell inocula in the absence of further treatment. Obviously, further studies will be required to verify this possible pathway. Because the approach described above was not fully optimized, there are obviously a number of methodological variables that could still be refined. Starting from the immunization step, the selection of a more appropriate hapten, carrier and adjuvant could significantly affect antibody titer, affinity, and isotype distribution. The chemistry of the folate–hapten conjugate might also be explored to assure the highest affinity for FR binding, and the dose of folate–hapten conjugate could be varied to promote the optimal antibody density on cancer cell surfaces. In order to

induce the most specific immune response, a more thorough evaluation of cytokines should be conducted, and improvements in the dosing schedule for the tumor-targeted hapten and immunomodulatory cytokine(s) present additional opportunities for refinement. While most cancer patients may respond normally to immunization against a foreign hapten, alternative methods for the induction of a high and sustained antihapten antibody titer must also be considered for immunocompromised patients. It may also be beneficial to explore the use of folate conjugates of common haptens against which most patients are already immunized (i.e. dominant antigenic determinants from tetanus toxoid, tuberculosis, measles, etc.). Clearly, further optimization could lead to improvements in both the specificity and potency of folate-targeted immunotherapy. Acknowledgment The authors would like to thank Dr. Chris Leamon for valuable discussions and Chakri Abburi for technical assistance. This work was supported by grants from Endocyte Pharmaceuticals, Inc., West Lafayette, Indiana, USA.

References 1. Bennett B, Check IJ, Olsen MR, Hunter RL (1992) A comparison of commercially available adjuvants for use in research. J Immunol Methods 153:31 2. Birn H, Nielsen S, Christensen EI (1997) Internalization and apical-to-basolateral transport of folate in rat kidney proximal tubule. Am J Physiol 272:F70 3. Bodey B, Bodey B, Jr., Siegel SE, Kaiser HE (2000) Failure of cancer vaccines: the significant limitations of this approach to immunotherapy. Anticancer Res 20:2665 4. Borden EC, Sondel PM (1990) Lymphokines and cytokines as cancer treatment. Immunotherapy realized. Cancer 65:800 5. Brigle KE, Spinella MJ, Westin EH, Goldman ID (1994) Increased expression and characterization of two distinct folate binding proteins in murine erythroleukemia cells. Biochem Pharmacol 47:337 6. Bukowski RM (2000) Cytokine combinations: therapeutic use in patients with advanced renal cell carcinoma. Semin Oncol 27:204 7. Burke F (1999) Cytokines (IFNs, TNF-alpha, IL-2 and IL-12) and animal models of cancer. Cytokines Cell Mol Ther 5:51 8. Chamberlain RS (1999) Prospects for the therapeutic use of anticancer vaccines. Drugs 57:309 9. Cherukuri A, Durack G, Voss EW Jr (1997) Evidence for hapten recognition in receptor-mediated intracellular uptake of a hapten–protein conjugate by murine macrophage. Mol Immunol 34:21 10. Christensen EI, Birn H, Verroust P, Moestrup SK (1998) Membrane receptors for endocytosis in the renal proximal tubule. Int Rev Cytol 180:237 11. Costello RT, Gastaut JA, Olive D (1999) Tumor escape from immune surveillance. Arch Immunol Ther Exp 47:83 12. Davis ID (2000) An overview of cancer immunotherapy. Immunol Cell Biol 78:179 13. Eisenthal A, Cameron RB, Rosenberg SA (1990) Induction of antibody-dependent cellular cytotoxicity in vivo by IFN-alpha and its antitumor efficacy against established B16 melanoma liver metastases when combined with specific anti-B16 monoclonal antibody. J Immunol 144:4463 14. Eisenthal A, Rosenberg SA (1989) The effect of various cytokines on the in vitro induction of antibody-dependent cellular cytotoxicity in murine cells. Enhancement of IL-2-induced

162

15.

16.

17.

18.

19.

20.

21.

22. 23.

24.

25.

26. 27.

antibody-dependent cellular cytotoxicity activity by IL-1 and tumor necrosis factor-alpha. J Immunol 142:2307 Flieger D, Spengler U, Beier I, Kleinschmidt R, Hoff A, Varvenne M, Sauerbruch T, Schmidt-Wolf I (1999) Enhancement of antibody dependent cellular cytotoxicity (ADCC) by combination of cytokines. Hybridoma 18:63 Flieger D, Spengler U, Beier I, Sauerbruch T, Schmidt-Wolf I (2000) Combinations of the cytokines IL-12, IL-2 and IFNalpha significantly augment whereas the cytokine IL-4 suppresses the cytokine-induced antibody-dependent cellular cytotoxicity of monoclonal antibodies 17–1A and BR55–2. Cytokine 12:756 Gabizon A, Horowitz AT, Goren D, Tzemach D, Mandelbaum-Shavit F, Qazen MM, Zalipsky S (1999) Targeting folate receptor with folate linked to extremities of poly(ethylene glycol)-grafted liposomes: in vitro studies. Bioconjug Chem 10:289 Gates SB, Mendelsohn LG, Shackelford KA, Habeck LL, Kursar JD, Gossett LS, Worzalla JF, Shih C, Grindey GB (1996) Characterization of folate receptor from normal and neoplastic murine tissue: influence of dietary folate on folate receptor expression. Clin Cancer Res 2:1135 Gazit Z, Weiss DW, Shouval D, Yechezkeli M, Schirrmacher V, Notter M, Walter J, Kedar E (1992) Chemo-adoptive immunotherapy of nude mice implanted with human colorectal carcinoma and melanoma cell lines. Cancer Immunol Immunother 35:135 Hank JA, Robinson RR, Surfus J, Mueller BM, Reisfeld RA, Cheung NK, Sondel PM (1990) Augmentation of antibody dependent cell mediated cytotoxicity following in vivo therapy with recombinant interleukin 2. Cancer Res 50:5234 Kedar E, Rutkowski Y, Leshem B (1992) Chemo-immunotherapy of murine solid tumors: enhanced therapeutic effects by interleukin-2 combined with interferon alpha and the role of specific T cells. Cancer Immunol Immunother 35:63 Li Q, Chang AE (1999) Adoptive T-cell immunotherapy of cancer. Cytokines Cell Mol Ther 5:105 Longenecker BM, Reddish M, Koganty R, MacLean GD (1994) Specificity of the IgG response in mice and human breast cancer patients following immunization against synthetic sialyl–Tn, an epitope with possible functional significance in metastasis. Adv Exp Med Biol 353:105 Lussow AR, Fanget L, Gao L, Block M, Buelow R, Pouletty P (1996) Targeting of antihapten antibodies to activated T cells via an IL-2-hapten conjugate prolongs cardiac graft survival. Transplantation 62:1703 MacLean GD, Reddish MA, Koganty RR, Longenecker BM (1996) Antibodies against mucin-associated sialyl–Tn epitopes correlate with survival of metastatic adenocarcinoma patients undergoing active specific immunotherapy with synthetic STn vaccine. J Immunother Emphasis Tumor Immunol 19:59 Morse MA (2000) Technology evaluation: Theratope, Biomira Inc. Curr Opin Mol Ther 2:453 Oldham RK, Blumenschein G, Schwartzberg L, Birch R, Arnold J (1992) Combination biotherapy utilizing interleukin-2

28.

29. 30. 31. 32.

33.

34. 35. 36. 37.

38.

39.

40.

41. 42.

and alpha interferon in patients with advanced cancer: a National Biotherapy Study Group Trial. Mol Biother 4:4 Ottone F, Miotti S, Bottini C, Bagnoli M, Perego P, Colnaghi MI, Menard S (1997) Relationship between folate-binding protein expression and cisplatin sensitivity in ovarian carcinoma cell lines. Br J Cancer 76:77 Patrick TA, Kranz DM, van Dyke TA, Roy EJ (1997) Folate receptors as potential therapeutic targets in choroid plexus tumors of SV40 transgenic mice. J Neurooncol 32:111 Reddy JA, Low PS (1998) Folate-mediated targeting of therapeutic and imaging agents to cancers. Crit Rev Ther Drug Carrier Syst 15:587 Restifo NP, Esquivel F, Kawakami Y, Yewdell JW, Mule JJ, Rosenberg SA, Bennink JR (1993) Identification of human cancers deficient in antigen processing. J Exp Med 177:265 Ross JF, Chaudhuri PK, Ratnam M (1994) Differential regulation of folate receptor isoforms in normal and malignant tissues in vivo and in established cell lines. Physiologic and clinical implications. Cancer 73:2432 Seya T, Begum NA, Nomura M, Tsuji S, Matsumoto M, Hayashi A, Azuma I, Toyoshima K (2000) Innate immune therapy for cancer. Screen for molecules capable of activating the innate immune system. Adv Exp Med Biol 465:229 Sinkovics JG, Horvath JC (2000) Vaccination against human cancers (review). Int J Oncol 16:81 Sudimack J, Lee RJ (2000) Targeted drug delivery via the folate receptor. Adv Drug Deliv Rev 41:147 Toffoli G, Cernigoi C, Russo A, Gallo A, Bagnoli M, Boiocchi M (1997) Overexpression of folate binding protein in ovarian cancers. Int J Cancer 74:193 Vuist WM, Visseren MJ, Otsen M, Bos K, Vyth-Dreese FA, Figdor CG, Melief CJ, Hekman A (1993) Enhancement of the antibody-dependent cellular cytotoxicity of human peripheral blood lymphocytes with interleukin-2 and interferon alpha. Cancer Immunol Immunother 36:163 Wang S, Luo J, Lantrip DA, Waters DJ, Mathias CJ, Green MA, Fuchs PL, Low PS (1997) Design and synthesis of [111In]DTPA–folate for use as a tumor-targeted radiopharmaceutical. Bioconjug Chem 8:673 Weitman SD, Lark RH, Coney LR, Fort DW, Frasca V, Zurawski VR, Jr., Kamen BA (1992) Distribution of the folate receptor GP38 in normal and malignant cell lines and tissues. Cancer Res 52:3396 Weitman SD, Weinberg AG, Coney LR, Zurawski VR, Jennings DS, Kamen BA (1992) Cellular localization of the folate receptor: potential role in drug toxicity and folate homeostasis. Cancer Res 52:6708 White CA, Weaver RL, Grillo-Lopez AJ (2001) Antibodytargeted immunotherapy for treatment of malignancy. Annu Rev Med 52:125 Wu M, Gunning W, Ratnam M (1999) Expression of folate receptor type alpha in relation to cell type, malignancy, and differentiation in ovary, uterus, and cervix. Cancer Epidemiol Biomarkers Prev 8:775