WO1995029698A1

WO1995029698A1 - Tumor-associated antigen peptides and their use in tumor diagnosis and therapy

Info

Publication number: WO1995029698A1
Application number: PCT/US1995/004912
Authority: WO
Inventors: Lea Eisenbach; Gideon Berke; Michael Feldman; Matityahu Fridkin; Ofer Mandelboim
Original assignee: Yeda Research And Development Co., Ltd. At The Weizmann Institute Of Science; Rycus, Avigail
Priority date: 1994-05-03
Filing date: 1995-05-03
Publication date: 1995-11-09
Also published as: IL109540A0

Abstract

Synthetic peptides of the sequence Phe-Glu-Gln-Asn-Thr-Ala-Gln-X1-X2 wherein X1 is Ala or Pro, and X2 is Gly or OH, and biologically active analogs thereof, are capable of: (i) binding to H-2Kb molecules on RMA-S cells, (ii) sensitizing RMA-S cells to lysis by specific anti-tumor CTL, (iii) inducing specific CTL in vivo, and (iv) protecting mice from metastasis of 3LL lung carcinoma. These tumor-associated peptides are suitable for tumor diagnosis and therapy, and particularly for treatment of tumor metastasis.

Description

TUMOR-ASSOCIATED ANTIGEN PEPTIDES AND THEIR USE IN TUMOR DIAGNOSIS AND THERAPY

FIELD OF THE INVENTION

The present invention is generally in the field of tumor immunology, and more specifically, concerns synthetic tumor-associated peptides useful for tumor diagnosis and vaccination and for the preparation of specific anti-tumor immunotherapeutic agents e.g. specific anti-tumor CTL or anti-tumor antibodies.

BACKGROUND OF THE INVENTION AND PRIOR ART

Many mouse and human tumors express major histocompatibility complex (MHC) class I-associated antigens that constitute targets for syngeneic cytotoxic T-lymphocytes (CTL). Genes encoding such antigens were isolated from a mouse mastocytoma and from human melanomas by genetic methods (Van Den Eynde et al, 1991; Van Der Bruggen et al., 1991). Isolation and characterization of MHC class I-associated peptides has enabled the recent identification of specific anchor residues typical of peptides that bind to distinct class I molecules (Falk et al., 1991). Moreover, CTL specific to particular MHC-peptide combinations have been used to identify naturally occurring antigenic peptides in cell extracts and enabled their direct sequencing (Rotzschke et al, 1990; Van Bleek and Nathenson, 1990; Udaka et al., 1992). Most known MHC ligands are of viral origin or self peptides, derived from normal proteins such as minor histocompatibility antigens, enzymes, histones or heat shock proteins (reviewed in Rammensee et al., 1993).

CTL directed against peptides presented by MHC class I molecules (reviewed in Urban and Schreiber, 1992), constitute powerful effectors of the immune system against animal tumors. The in-vivo efficacy of CTL was demonstrated in experimental systems by active vaccination with various MHC class I and cytokine-transduced tumor cells and by adoptive transfer studies (Dranoff et al., 1993; Melief, 1992). The notion that CTL also have anti-tumor activity in human has received support from the therapeutic activity of expanded tumor infiltrating lymphocytes (TEL) in melanoma patients (Rosenberg, 1988). CD8+ CTL from melanoma patients were shown to be directed against recently defined melanoma- associated antigens like MAGE 1, tyrosinase, gplOO, and MART 1 (Boon et al., 1994). Peptides derived from melanoma tumor-associated antigens (TAAs) or from a murine mastocytoma P815 induced CTL responses in-vitro and in-vivo, yet thus far there is no clear evidence, that such peptides may have direct therapeutic effects in advanced tumors.

The high metastatic clone D122 of the Lewis lung (3LL) carcinoma is characterized by low cell surface expression of H-2K molecules, and poor immunogenicity. Transfection of H-2K converts D 122 to a non-metastatic and highly immunogenic phenotype (Plaksin et al., 1988). The response against the H-2K transfectants is mediated by CD8+ T-cells as deduced from their malignancy in Nude/Nude and CD8-depleted mice (Mandelboim et al., 1992). H-2D molecules, expressed on D 122 cells, do not serve as a restriction element for anti-tumor CTL (Plaksin et al., 1992).

Regression of highly malignant residual tumor is the major goal in cancer immuno- therapy. It has been a long-felt need to isolate specific tumor-associated antigens or active peptides derived therefrom, which can serve as targets for CTLs, and thereby facilitate a way for combatting such tumors. Moreover, isolation of such tumor-associated peptides is also important for various diagnostic methods for the early detection of tumors. Accordingly, it is an ongoing need to provide more and more such tumor-associated antigens or peptides to expand the range of tumors that may be detected with the above diagnostic method, or any other existing diagnostic methods which employ such antigens or peptides. Moreover, isolation and characterization of such antigens or peptides is also important for the development of specific anti-tumor antibodies which may be used in the various existing immunotherapeutic methods to combat tumor growth.

In contrast to the above-mentioned art concerning various other tumor-associated antigens, the present invention concerns the isolation, by total acid extractions and repeated HPLC fractionations, and sequencing of a heretofore not described mammalian tumor- associated peptide and analogs thereof, which peptide is derived from the above-noted mouse Lewis lung carcinoma (3LL).

It is accordingly an object of the present invention to provide such a peptide and its analogs. It is also an object of the present invention to provide uses of this peptide and its analogs in vaccines for tumor therapy and in methods for induction of tumor-specific CTLs and methods for tumor diagnosis.

SUMMARY OF THE INVENTION

According to the present invention, it has been found that a mammalian tumor- specific peptide and the synthetic peptides based on this peptide's sequence, are capable of binding so-called "empty" H-2K molecules expressed on RMA-S cells, are also able to sensitize RMA-S cells to lysis by specific anti-3LL CTLs, can induce the production of peptide/analog-specific CTLs in vivo, and can protect mice from metastases of 3LL lung carcinoma.

Accordingly the present invention provides a synthetic peptide of the sequence I: Phe-Glu-Gln-Asn-Thr-Ala-Gln-X] -X2 (I) wherein X] is Ala or Pro, and X2 is Gly or OH, and biologically active analogs thereof, said peptide of formula I and the biologically active analogs thereof being capable of: (i) binding to "empty" H-2K molecules on RMA-S cells, (ii) sensitizing RMA-S cells to lysis by specific anti-tumor CTL, (iii) inducing specific anti-tumor CTL in vivo, and (iv) protecting mice from metastases of 3LL lung carcinoma.

The biological activities of the peptides of sequence (I) and their analogs and derivatives are determined by the assay procedures described herein in the specification.

In preferred embodiments the synthetic peptides according to the invention are the peptides herein designated MUT 1 and MUT 2 and that have amino acid sequences: MUT 1 : Phe-Glu-Gln-Asn-Thr-Ala-Gln-Pro

MUT 2: Phe-Glu-Gln-Asn-Thr- Ala-Gin- Ala The peptides of the invention are for use in tumor diagnosis and therapy and for the preparation of anti-tumor antibodies or of specific anti-tumor CTLs.

Since human connexin 37 is 86% homologous to mouse and rat connexin 37 (Reed KE, Westphale EM, Larson DM, Wang HZ, Veenstra RD, Beyer EC, (1993), Molecular cloning and functional expression of human connexin 37, an endothelial cell gap junction protein. J. Clin Invest., 91: 997-1004) and 100% homologous in the domain 52-60 to which MUT 1 and MUT 2 belong (and in similar areas around functional cysteins) there is ground to believe that similar mutations may be relevant to human tumors. Thus the peptides of formula I can be used, for example, in tumor diagnosis - by measuring the response of peripheral blood lymphocytes (PBL) from patients to synthetic peptides (free or loaded on appropriate presenting cells) by increased proliferation or in CTL assays. Such result would indicate whether mutations relevant to connexin exist in the patients' tumor and whether such patients may be suitable candidates for peptide related therapy. The peptides may further be used in tumor therapy: peptide vaccines with appropriate adjuvants may be used to vaccinate tumor-bearing patients pre- or post-surgery. Such cancer vaccines induce specific CTL and possibly antibodies that recognize mutated connexin 37 peptides in the primary tumor or metastases.

DESCRIPTION OF THE DRAWINGS

Figs, la-d depict, graphically, the H-2K stabilization and immunogenicity of total acid-extracted peptides, wherein: Fig. la shows the fluorescent anti-K antibody staining of RMA-S cells without peptides (al) and of RMA-S cells loaded with total peptide extract (a2); Fig. lb shows the in vitro lytic activity of CTL induced by peptide-bound RMA-S cells, the various target cells tested being indicated by differently-patterned bars, the key to which is included in the figure; and Figs, lc, Id show the mean tumor diameters (in mm) and lung weights (in mg) of D122-bearing mice vaccinated with peptide-loaded RMA-S cells.

P<0.0001 between the RMA-S+ peptide vaccinated group and all other groups in (c) and P vvaalluueess aarree 00..002288 aanndd 00..00000044 ffoorr ggrroouuppss vvaacccciinnaatteedd wwiitthh KK 3399..55 and peptide-loaded RMA-S cells, respectively, as compared to the non-vaccinated group (d). Figs. 2a-c depict, graphically, the reversed-phase HPLC separation of total acid extracted peptides and the detection of naturally-occurring CTL epitopes, the upper panels depicting the HPLC separation profile (solid lines) and the acetonitrile gradients used in the elutions (broken lines), and the lower panels depicting the CTL activity of the fractions eluted from the HPLC column. The arrows indicate the CTL-active peaks of the HPLC- separated fractions.

Figs. 3a-e depict, graphically, the stabilization of K expression and lysis of RMA-S cells as mediated by the peptides herein designated MUT 1 and MUT 2, wherein: Fig. 3a shows the H-2K and H-2D cell surface expression of RMA-S cells incubated with the various indicated peptides, the upper panels showing cells fluorescently labelled with monoclonal anti-K antibody 20-8-4, the lower panels showing cells fluorescently labelled with monoclonal anti-D^b antibody 28-14-8 (Ozato K, Sachs DH, (1981) Monoclonal antibodies to MHC antigens. Hybridoma antibodies reactive to antigens of H-2K haplotypes reveal genetic control of isotype expression. J. Immunol, 126: 317-321), the left panels showing cells stained with FITC-conjugated goat anti-mouse immunoglobulin (background) or K /D^b and second antibody, and the other panels showing peptide-loaded RMA-S cells stained with anti-K /D compared to anti-K /D -stained RMA-S cells; Figs. 3b, 3c show the in vitro lytic activity of CTL induced by K 39.5 cells, tested on the various target RMA-S cells loaded with the various indicated peptides or the unloaded controls, each graph having a symbol, the meaning of which is shown in the key next to Fig. 3c; Fig. 3d shows the dose response of the various target S-methionine-labelled RMA-S cells (see key to each graph next to Fig. 3c) loaded with varying amounts of the different peptides and subjected to CTL lysis; and Fig. 3e shows the in vitro lytic activity of CTL induced by MUT 1 and MUT 2 RMA-S-loaded cells, the meanings of each bar being in the key next to the figure. Experimental details were as in Fig. 1.

Figs. 4a-c show PCR amplification and sequences of connexin 37 cDNA in 3LL clones and in normal lungs. Fig. 4a is a schematic presentation of connexin 37 showing the Cys to Gin mutation site and the primers used for amplification and sequencing. Fig. 4b shows PCR products amplified from lung, K 39.5 and D122 cDNA using primers 1+3 (marked as primer 1), or primers 2+3 (marked as primer 2). Fig. 4c shows sequencing of PCR products amplified by primers 1+3 from lung, K 39.5 and D122 (panel 1) and by primers 2+3 from K 39.5 and D122 (panel 2). Only one example of each of the sequences is shown.

Fig. 5 shows that MUT 1 and MUT 2 peptides protect mice from D122 metastasis. Groups of C57BL mice (10 mice/group) were immunized i.p. three times at seven day intervals with irradiated (5000 Rad) RMA-S, D122 and K^b39.5 tumor cells, with peptides loaded on RMA-S, with peptides in PBS (20 μg/mouse), or with the same peptides in EFA (20 μg/mouse). Ten days after the last immunization, the mice were inoculated i.f.p. with 2X10^ D122 tumor cells. When tumors reached diameters of 8mm, the tumor-bearing legs were amputated and 23 days later, according to the death of the control group, mice were sacrificed and metastatic loads were determined. Unpaired student T-test showed that relative to the control group (non-immunized mice), p-values were 0.5923, 0.0059, 0.5421 for D122 K^b39.5, RMA-S tumor cells, respectively, 0.3441, 0.7474, 0.8858 for MUT 1, MUT 2, Con 37 in PBS, respectively, 0.0058, 0.0012 and 0.8317 for MUT 1, MUT 2 and Con 37 in IFA. and 0.0393, 0.0174, or 0.8451 for MUT 1, MUT 2, or Con 37 on RMA-S cells. Immunization with IFA alone had no effect, the p value was 0.3988 relative to the non-immunized control group, p value for D122 cells injected in anesthetized mice was 0.8017 compared to control. Normal lung weight (200mg) is marked with a line.

Fig. 6 shows that post-surgical immunotherapy is mediated in mice by MUT 1 and MUT 2 peptides. D122 tumor cells (2X10⁵) were injected i.f.p. in C57BL mice (10 mice/ group). Thirty days later, when tumors reached a diameter of 6-6.5mm, tumor bearing legs were amputated. Two days after the amputations, immunizations started using irradiated (5000 Rad) D122, K^b39.5, RMA-S cells and MUT 1, MUT 2 and Con 37 peptides. Immunizations were done s.c. with peptides in PBS, in IFA or loaded on RMA-S cells as described in Example 5. Each group of mice was immunized four times at seven-day intervals. Thirty-33 days after amputation, according to the death of the control groups, mice were sacrificed and metastatic loads were determined. Lungs of one mouse of each group were stained using 0.2ml India ink injected into the trachea (d,e). Lung tissues are shown in black, metastases are in white. Unpaired student T-test showed that relative to the control group p values were 0.1734, 0.0206 and 0.2907 for D122, K^b39.5 and RMA-S immunized mice, respectively, 0.8130, 0.9490, 0.8576 for mice immunized with MUT 1, MUT 2 and Con 37 peptides in PBS, respectively, 0.0034, 0.0027 and 0.9521 for mice immunized with the same peptides in IFA, respectively and 0.0055, 0.0073 and 0.7688 for mice immunized with peptides loaded on RMA-S cells. P values for the control groups in relation to non- immunized mice were 0.7562 and 0.4250 for mice immunized with adjuvant alone, or mice injected with PBS, respectively.

Fig. 7 shows survival of MUT 1 and MUT 2 peptide vaccinated mice in post-surgical immunotherapy experiments. D122 tumor cells (2X10^) were injected i.f.p. in C57BL mice. Thirty days later, when tumors reached diameters of 6-6.5mm, tumor-bearing legs were amputated. Two days after the amputations, immunizations started using irradiated D122, K^b39.5, RMA-S cells and peptides loaded RMA-S cells. Survival was followed: ( blank squares): non-immunized mice; (blank triangles): mice immunized with irradiated RMA-S cells; (filled tri.angles) mice immunized with Con 37 loaded RMA-S cells; (blank circles): mice immunized with irradiated K^b39.5 cells; (filled circles): mice immunized with MUT 2 loaded RMA-S cells; (filled squares) mice immunized with MUT 1 loaded RMA-S cells.

Fig. 8 shows the involvement of CD4 and CD8 T-cells in the immunotherapeutic effect mediated by MUT 1 and MUT 2 peptides. Groups of C57BL mice (30 mice/group) were injected i.f.p. with 2x10^ D122 tumor cells. When tumors reached diameter of 5.5mm, groups of mice were divided into three. One third of the mice were depleted of CD4+ T-cell by i.v. injection of 100 μl GK1.5 ascites fluid (anti-CD4 mAb diluted 1 :5 in PBS). One week and 4 weeks later, mice were i.p. boosted with the same antibody. One-third of mice were depleted of CD8+ T-cells by i.v. injection of 100 μl YTS- 169.4 ascites fluid (anti-CD8 mAb). Boosts were given weekly, i.p. The third group was not depleted. Thirty days after the injection of D122 tumor cells, when tumors reached diameter of 6-6.5mm, tumor-bearing legs were amputated. Two days after the amputations, mice were immunized with irradiated tumor cells and peptides as described in Fig 6. For undepleted mice, P values relative to the control group were>0.05 in groups receiving D122, RMA-S, and Con 37 by all administration routes and for all peptides given in PBS. For the groups immunized by MUT 1 or MUT 2 peptides loaded on RMA-S or given in IFA, p values were 0.0134, 0.0328, 0.0131, 0.0142, respectively, p value for the K 39.5 immunized group was 0.0188. In the groups of mice that were depleted of CD4 T-cells, p values for D122, RMA-S, or Con 37 were>0.05. p Values for MUT 1, MUT 2 on RMA-S, in IFA, or in PBS were 0.0019, 0.0008, and 0.0006 for MUT 1, respectively, and 0.0008, 0.0006 and 0.0002 for MUT 2 peptide, respectively, p value for K^b39.5 cells was 0.0016. In all groups of mice that were depleted from CD8 T-cell, p values were>0.05. P values for the control groups: anesthetized mice and mice injected with PBS were >0.05.

Fig 9 shows the immune response in mice as a function of the site of vaccination with peptides MUT 1 and MUT 2. Groups of C57BL mice (10-20 mice/group) were injected i.f.p. with 2X10^ D122 tumor cells. When tumors reached diameter of 5.5mm groups of mice were divided into three. One third of the mice were depleted of CD4+ T-cell by i.v. injection of 100 μl of GK1.5 ascites fluid (anti-CD4 mAb diluted 1:5 in PBS). One week and 4 weeks later, mice were i.p. boosted with the same antibody. One-third of mice were depleted from CD8+ T-cells by i.v. injection of 100 μl YTS 169.4 of ascites fluid (anti-CD8 mAb). Boosts were given weekly, i.p. The third group was not depleted. Thirty days after the injection of D122 tumors, when tumors reached diameter of 6-6.5mm, tumor-bearing legs were amputated. Two days after the amputations, mice were immunized with peptides solubilized in PBS. Immunizations were done four times at seven-day intervals and peptides were given i.p. (groups designated MUT 1, MUT 2, Con 37 i.p.), or the first injection was given intradermally (i.d.) in anesthetised mice and the following injections were given s.c. (groups designated MUT 1, MUT 2, Con 37 s.c). For undepleted mice, P values relative to the control group were>0.05 in groups receiving D122. For the groups of mice depleted of CD4+ T cells p values relative to the control (D122 bearing mice, CD4+ depleted) were>0.05 for all peptides injected i.p. and 0.0002, 0.0001, 0.1147 for MUT 1 s.c, MUT 2 s.c. and Con 37 s.c. respectively. For mice depleted of CD8+ T cells, P values relative to the control group (D122 bearing mice, CD8+ depleted) were>0.05. P values for the control groups: anesthetized mice and mice injected with PBS were >0.05. DETAILED DESCRIPTION OF THE INVENTION

The new mammalian tumor-associated peptides which have been isolated, sequenced and characterized according to the present invention, and later synthesized, are derived from a Lewis lung carcinoma (3LL) and have the following amino acid sequence in the first seven positions: Phe-Glu-Gln-Asn-Thr-Ala-Gln, whilst in the eighth position there is Ala or Pro, and an optional ninth residue may be Gly.

Synthetic octamers based on these sequences were prepared, these synthetic peptides having the sequences Phe-Glu-Gln-Asn-Thr-Ala-Gln-Pro (herein designated peptide MUT 1) and Phe-Glu-Gln-Asn-Thr-Ala-Gln-Ala (herein designated peptide MUT 2), and were shown to bind to so-called "empty" H-2K molec lies expressed on the surface of RMA-S cells as well as to sensitize RMA-S cells to lysis by specific anti-3LL CTL, to induce specific CTL in vivo, and to protect mice from 3LL lung carcinoma metastases. Both synthetic peptides MUT 1 and MUT 2 were shown to have essentially the biological activity as the originally isolated peptide(s).

Another synthetic peptide was also prepared, being that of the closely related sequence from position 52-59 of the mouse gap-junction protein connexin 37 i.e. the synthetic connexin 37 peptide of sequence: Phe-Glu-Cys-Asn-Thr-Ala-Gln-Pro. Thus, the main difference between the MUT 1 and MUT 2 peptides and the connexin 37 peptide is in position 3, i.e. both MUT 1 and MUT 2 have a Gin residue whilst the connexin 37 peptide has a Cys residue. When tested for biological activity, the connexin 37 peptide showed none of the above noted biological activities of the MUT 1, MUT 2 or originally isolated peptides, indicating that these activities are directly dependent on the presence of a Gin or Gin-like residue in position 3 of the peptide, the activities being lost when a Cys residue is in this position.

It is known that connexin 37 is abundant in the lungs and belongs to the family of connexin proteins which are transmembrane proteins which form intercellular hydrophilic channels called gap-junctions. Formation of such gap-junctions is dependent on a small number of cysteine residues in the extracellular domains of connexins, amongst which is Cys 54 (which is Cys in position 3 of the above connexin 37 peptide) that is located on the first extracellular domain of connexin 37, and is replaced by Gin in the lung carcinoma peptides. Other connexins, for example, human connexins 26 and 43 have been suggested to act as tumor suppressor genes in breast and are down-regulated in human breast carcinomas. The observation that gap-junctional intercellular communications contributes to normal growth regulation is supported by the absence of such communications in many solid tumors and by the correlation between transformation and the loss of gap junctional intracellular communications in experimental systems.

Thus, the present invention provides a way for isolating and characterizing tumor- associated peptides, a way for inducing CTLs specific for such peptides which can be used for combatting tumors by established CTL-based immunotherapy procedures, and provides the synthetic peptides in HPLC-purified form. The purified peptides may be used in vaccines for tumor metastasis therapy and to produce specific CTLs or specific antibodies which are useful as anti-tumor therapeutic agents.

The peptides of the present invention have a hydrophilic nature and accordingly, the preparation of peptide analogs by way of substitution of one or more amino acids should ensure that this hydrophilic nature is conserved. For example, as noted above, the Gin in position 3 must not be replaced by Cys, but it is feasible that this Gin be replaced by Asn or a synthetic residue of similar nature to Gin. Further, as is readily apparent to all of skill in the art, the residues in positions 1-8 of the peptides MUT 1 and MUT 2 may each be substituted by residues (natural or synthetic) of similar nature. The feasibility of such substitutions may readily be tested by testing the biological activity of such analogs in accordance with the h assay procedures set forth herein i.e. capability of binding "empty" H-2K molecules on RMA-S cell; capability of sensitizing RMA-S cells to lysis by specific anti-3LL CTL; and capability of inducing specific CTL in vivo.

The present invention will now be described in greater detail in the following non- limiting examples and their accompanying figures. EXAMPLES EXAMPLE 1

The H-2K stabilization and immunogenicity of total acid-extracted peptides from tumor cells.

In preliminary experiments, the H-2K -D122 transfectant K^b39.5 was used for isolating K -bound peptides, according to schemes outlined by Falk et al, 1991(a) and Van Bleek and Nathenson, 1990. Sensitization of H-2 "empty" RMA-S cells, by the HPLC fractionated peptides, to lysis by anti-K 39.5 CTL repeatedly revealed the presence of prominent active peptide fractions (not shown). However, fractions from 2x10 tissue cultured cells were insufficient in quantity for sequence analysis by mass-spectroscopy.

Successful identification of a tumor-associated peptide necessitates both large quantities of tumor cells as well as a sensitive bio-assay and therefore large groups of nude mice were injected with K 39.5 tumor cells. Total acid extraction rather than isolation of K - bound peptide complexes was used to increase the amount of extractable peptides, assuming that only a part of the total amount of cellular peptides are associated with K molecules. The total acid peptide fraction (Mr<5,000) was tested for presence of K -binding peptides and anti -tumor properties.

Total acid extracted peptides were prepared from K 39.5 cells grown as subcutaneous tumors in 150 CD1 nude mice. Non-necrotic (1-2 cm) tumors were homogenized in phosphate-buffered saline (PBS) containing 0.5% NP40; lOμg/ml soybean trypsin inhibitor; 5μg/ml leupeptin; 8μg ml aprotinin and 0.5mM PMSF (phenylmethylsulfonyl fluoride, a protease inhibitor), were then stirred for 30 min at 4°C, titrated with 10% trifluoroacetic acid (TFA) to a final concentration of 0.1% TFA and further processed as described in Falk et al, 1991 (b). One third of the peptidic fraction (Mr<5,000) was dissolved in Iscove's Medium (IMDM) containing antibiotics and 5x10 M β-mercaptoethanol (IMDM medium), and the remaining 2/3 of the material was dissolved in 0.1 % TFA for HPLC separation.

For peptide loading, RMA-S cells were precultured, 36-48 h at 26°C in RPMI medium supplemented with 10% fetal calf serum (FCS); ImM glutamine; 1 mM non- essential amino acids; 1 mM sodium pyruvate; combined antibiotics; lOmM HEPES pH 7.4; 12 and 2x10^"5 M β-mercaptoethanol (RPMI-HEPES medium). Cells were irradiated (5,000

35 Rads) for immunization in vivo, labeled with S- methionine for target preparation, or left untreated and then incubated in a small volume (50-100 μl, containing 5x10 -8x10 cells) with an equal volume of 250 μg/ml total extracted peptides, for 30 min at 26°C and 3,5 hr at

37°C. For flow cytometry, washed cells were stained with monoclonal anti-K antibody 20-

8-4. For CTL assays, effector lymphocytes were derived from C57BL/6 mice that had been immunized with one of the following types of cells: 2x10 irradiated (5,000 Rads) D122,

K 39.5, RMA-S, and peptide-loaded RMA-S cells, by procedures as described in

Mandelboim et al., 1992. Splenocytes were restimulated on irradiated (5,000 Rad) and mitomycin-C treated (80 μg/ml per 10 cells) K 39.5 cells for four days, and reacted with 5000 35 S-methionine-labeled target cells, at an effector to target (E:T) ratio of 50: 1, for 5h at

37°C as described in Mandelboim et al., 1992. Spontaneous release did not exceed 30% of maximal release. The standard error was less than 5% of the means of the wells.

For in vivo vaccination assays, C57BL/6 mice were inoculated into the footpad with 2x10 D122 cells per mouse. Eighteen days later, groups of mice were immunized intraperitoneally (i.p) five times at 7-day intervals with one of the following: inactivated (5,000 Rad) parental D122 cells, K^b39.5 cells, RMA-S cells, and peptide-loaded RMA-S cells. Controls were non-immunized mice. Tumor growth and metastasis were determined as described in Mandelboim et al., 1992.

The results of the tests performed on the above-noted total acid-extracted peptide fraction (Mr<5000) with regards to the presence of K -binding peptides in this fraction and anti-tumor properties thereof, are set forth in Figs, la-d: Fig. la shows the control experiment of fluorescent anti-K antibody staining of RMA-S cells without peptides (method according to Townsend et al., 1989) and Fig. 2a shows the fluorescent anti-K antibody staining of RMA-S cells loaded with the total peptide extract. From the comparison of (al) and (a2) it is apparent that stabilization of K molecules on RMA-S, H-2 negative cells, occurred following incubation with the total peptide extract. It should be noted that D expression on the peptide-loaded RMA-S cells was also induced (results not shown). Fig. lb shows the in vitro lytic activity of CTL induced by peptide-bound RMA-S cells, the various effector cells (D122; K^b39,5; RMA-S; and RMA-S-peptide-loaded) being tested on the various target cells. From Fig. lb it is therefore apparent that the immunization of syngeneic C57BL/6 mice with peptide-loaded RMA-S cells produced a tumor-specific CTL response that was comparable to that obtained after immunization with K 39.5 tumor cells.

Fig lc and Fig. Id show, respectively, the mean tumor diameters (in mm) and lung weights (in mg) of D 122 tumor-bearing mice vaccinated with peptide-loaded RMA-S cells. It is apparent that vaccination of C57BL/6 mice bearing 18-day-old D122 tumors with peptide- loaded RMA-S cells retarded tumor-growth and metastasis formation. Accordingly, the crude peptide fraction contains peptides presentable by syngeneic RMA-S cells that conferred anti- tumor activity.

EXAMPLE 2

Reversed-phase HPLC separation of total acid extracted peptides, detection of naturally-occurring CTL-specific peptide epitopes and preparation of synthetic analogs of the peptides.

Crude peptide fractions (see Example 1) were separated by reversed-phase HPLC. The separations were performed using a Merck Hibar column (250x4 mm) (Merck, Darmstadt, Germany) repacked with Lichrosorb RP-8 (5μm) (Merck). Eluents: solution A, 0.1% TFA in water (v/v); solution B, 0.1% TFA in 70% acetonitrile in water (v/v). The gradients used for each separation are described in the upper panels of Figs. 2a-c: in Fig. 2a the gradient was 0-70% acetonitrile and in Figs. 2b and 2c the gradient was 0-44% acetonitrile.

Individual fractions from two of eight separations were freed of acetonitrile and TFA by repeated drying in a vacuum centrifuge. Neutral fractions were resuspended in 0.6 ml RPMI-HEPES medium for use in the CTL assay. RMA-S cells were incubated at 26°C for 48 o r 3 h, then labeled with S-methionine for 8 hr, and 5x10 cells in 50 μl RPMI medium were incubated with 50 μl of peptides from each tested fraction, in duplicates, at 26°C for 30 min. Restimulated lymphocytes from K 39.5 immunized mice were added at a 100: 1 ratio. After 5 h at 37°C, plates were centrifuged and processed (see Mandelboim et al., 1992). Percent specific lysis was calculated as set forth in Mandelbiom et al., 1992. RMA-S cells without peptides were used as a negative control and K 39.5 cells were used as a positive control. The results shown in Figs. 2(a)-(c) represent the lysis of RMA-S cells loaded with individual fractions minus the spontaneous lysis of RMA-S cells (3-5%).

Fig. 2a shows the HPLC profile (upper panel) and CTL activity (lower panel) of the first separation (one of eight). Fractions of 1 ml, at a flow rate of 1 ml/min were collected. The individual fractions (1-60) were incubated with RMA-S cells and tested (six independent assays) for lysis by anti-K 39.5 CTL. Five active fractions were detected, namely, nos. 6, 19, 25, 31 and 32. The active fractions were pooled, in particular, fractions 31 and 32 (marked with an arrow), from all of the separations. The pooled active fractions were concentrated to 1 ml and rechromatographed, the results of which are shown in Fig. 2b. The individual fractions 1-54 of Fig. 2b (upper panel) were collected and tested as before for CTL activity (lower panel), from which it was observed that fractions 41 and 42 were the active fractions (marked with arrow). These active fractions were pooled and further processed by a third HPLC separation, the results of which are shown in Fig. 2c

Following the third HPLC separation (Fig. 2c), the active fraction no. 30 (arrow, 100 pmol) was collected and sequenced by standard Edman degradation on an Applied Biosystem protein sequencer model 475 A with an on-line 120A PTH analyzer. A major peptide Phe- Glu-Gln-Asn-Thr-Ala-Gln was unambiguously determined for the first seven residues. The eighth residue was questionably Ala or Pro, and a possible ninth residue, Gly, could also be determined. Position ten showed no significant signal for any residue. An NBRF data bank search revealed homology with a peptide from the mouse gap-junction protein connexin 37 (Phe-Glu-Cys-Asn-Thr-Ala-Gln-Pro-Gly, positions 52-60). Because most known K - restricted peptides are eight amino-acid long (see Rammensee et al., 1993), octameric synthetic peptides were prepared representing connexin 37 positions 52-59, and two variants of the deduced sequences with Pro (MUT 1) or Ala (MUT 2) at position eight. These three synthetic peptides were prepared by standard Merrifield solid-phase procedure using t-Boc chemistry on an Applied Biosystem 430A peptide/protein synthesizer, followed by purification of the synthesized peptides by standard HPLC procedures. The sequences of the so-prepared synthetic peptides are:

MUT 1: H-Phe-Glu-Gln-Asn-Thr-Ala-Gln-Pro-OH

MUT 2: H-Phe-Glu-Gln-Asn-Thr-Ala-Gln-Ala-OH

Connexin-37: H-Phe-Glu-Cys-Asn-Thr-Ala-Gln-Pro-OH

EXAMPLE 3

Stabilization of K expression and lysis of RMA-S cells mediated by the MUT 1 and

MUT 2 peptides.

Following procedures similar to those set forth in Example 1, the capacity of the synthetic peptides MUT 1 and MUT 2 to mediate stabilization of K expression and lysis of peptide-loaded RMA-S cells was tested. The results of these tests are set forth in Figs. 3a-e. Fig. 3a shows the H-2K^D and H-2D cell surface expression of RMA-S cells incubated with 5μM synthetic peptides MUT 2, MUT 1, connexin 37 (a.a. 52-59), or VSV (a.a. 52-59). RMA-S (controls) and RMA-S-peptide loaded cells were stained with the monoclonal anti- K antibody 20-8-4 (upper panels) and the monoclonal anti-D^b antibody 28-14-8 (lower panels). Left panels (upper and lower): RMA-S control cells stained with FITC-conjugated goat anti-mouse antibodies, or anti-K /D antibodies and a second antibody. Other panels (upper and lower): Peptide-loaded RMA-S cells stained with anti-K /D antibodies compared to anti-K /D antibody-stained RMA-S cells. Backgrounds of peptide loaded RMA-S cells overlap with the background of RMA-S cells in the left panels and are not shown in other panels. Thus, it is apparent from Fig. 3a that MUT 1 and MUT 2, as well as the prototype K -restricted peptide VSV 52-59, all stabilized K expression on RMA-S cells, while connexin-37 had no effect (upper panels). Similar testing with anti-D -antibodies revealed no binding (lower panels).

Figs. 3b and 3c show the in vitro lytic activity of CTL induced by K 39.5 cells, tested on RMA-S cells loaded with 5μM of the various synthetic peptides and inhibition of lysis by anti-K antibody. RMA-S target cells used in Fig. 3b were labeled as described in Example 1. For inhibition of lysis by antibodies, peptide-loaded and labeled RMA-S cells were incubated with 1 :40, 1:80 and 1 :160 dilutions of monoclonal antibodies anti-K 20-8-4, or anti-D 28-14-8, for 1 hr at room temperature. Effector cells were added at an E:T ratio of 50:1 for 5 hr as described in Example 1. The data shown (see single point not on any of the curves in Fig. 3b) are from a single 20-8-4 antibody concentration (1 :40), with MUT 1- RMA-S loaded cells as targets and K 39.5-induced CTL as effectors. Similar inhibition was observed with CTL induced by MUT 1 -RMA-S loaded cells incubated with MUT 1 -RMA-S or K 39.5 cells as targets. Anti-D antibody 28-14-8 did not inhibit lysis (data not shown). Thus, in Fig. 3b there is shown, amongst others, the inhibition of lysis of MUT 1 -loaded,

³⁵S-labelled RMA-S cells by a 1:40 dilution of the anti-K 20-8-4 antibody.

In Fig. 3c the RMA-S target cells were labelled with 51 Cr (instead of 35 S as in Fig.

3b). The ⁵¹Cr labelling of RMA-S cells was by incubation of lOxlO⁶ RMA-S cells in lOOμl IMDM medium supplemented with lOOμci of Cr-Na2Crθ4 for 90 min at 31°C. Following labelling, the cells were washed and used as targets in 5 h cytotoxicity assays as described in Example 1. In both Fig. 3b and 3c the data shown represent an average of four independent assays.

In addition to the above mentioned concerning the anti-K antibody inhibition of lysis, the results presented in Figs. 3b and 3c clearly show that both MUT 1 and MUT 2, but not connexin 37 or VSV peptides, were active in cytotoxic assays with anti-K 39.5 CTL, i.e. ³⁵S-labelled- or ⁵¹Cr-labelled-RMA-S cells loaded with either MUT 1 or MUT 2 peptides were effectively lysed by the anti-K 39.5 CTL, whilst those loaded with connexin 37 or VSV peptides were not lysed by the CTL (their lysis being the same as the non-loaded RMA- S control target cells). In Fig. 3b and 3c each graph represents a different kind of target cell and is indicated by a different symbol, the key to which is provided next to Fig. 3c, wherein it should be noted that the target "RMA-S" means non-loaded, labelled RMA-S cells, whilst the remaining targets indicate with which peptide the labelled RMA-S cells were loaded. Further, it should also be noted (as described above) that the antibody inhibition of lysis result is presented only in Fig. 3b as a single point, the target cells being "MUT 1 preincubated with 20-8-4", and this because in this instance only a 50: 1 effector: target (E:T) cell ratio was carried out. In all other experiments a range of effectortarget ratios was examined (E:T of 12: 1 to 100: 1) as indicated in Figs. 3b and 3c.

In Fig. 3d there is shown how the amount of peptide loaded onto the target cell affects the percentage of lysis of the target cell by the specific CTL, i.e. a dose response. 5-500nM of the various peptides were incubated with 35 S-methionine-labelled RMA-S cells, and reacted with CTL for 5 hr as described in Example 1. The effectortarget cell ratio in the lysis reaction was 50:1. Each curve in Fig. 3d is represented by a different symbol representing the different peptide with which the target cells were loaded, the key to which is provided next to

Fig. 3c and which is explained above in respect of Figs. 3b and 3c. From the results presented in Fig. 3d, it is apparent that sensitization for lysis by synthetic peptides showed a maximal effect at 100 nM for both MUT 1 and MUT 2, but lysis was detected even after sensitization with peptides at 5nM. The other peptides, connexin 37 and VSV, showed no effect, i.e. the same effect as the non-loaded RMA-S control target cells, even in amounts of up to 500 nM.

In Fig. 3e there is shown the in vivo lytic activity of CTL induced by MUT 1 and

MUT 2 RMA-S-loaded cells. Mice were immunized with RMA-S cells loaded with the various peptides (MUT 1, MUT 2 and connexin 37) or with unloaded cells (RMA-S, controls) and their splenocytes were subsequently obtained and used as effector cells by the procedures described in Example 1. The other effector cells examined for comparative purposes were D122 and K 39.5 (39.5) cells prepared as described in Example 1. For the lytic assays an effectoπtarget cell ratio of 25:1 was used, and the procedure was as described in Example 1. In Fig. 3e the differently patterned bars represent the different target cells subjected to lysis by each kind of effector cell, the key to which is provided next to the figure. From the results presented in Fig. 3e it is apparent that the splenocytes from mice immunized with MUT 1 and MUT 2 RMA-S-loaded cells exhibited similar specificities as the anti-K 39.5 CTL (positive control), whilst the splenocytes from mice immunized with connexin 37 RMA-S-loaded cells exhibited a lack of specific CTL activity, as observed with the splenocytes from mice immunized with non-loaded RMA-S cells (negative controls). EXAMPLE 4

PCR Amplification and sequences of connexin 37 cDNA in 3LL clones and in normal lungs.

To examine whether MUT 1 and/or MUT 2 are actually derived from a mutated connexin 37 gene, reverse transcription-polym erase chain reactions (RT-PCR) were done on RNAs isolated from D122, K 39.5, or normal C57BL/6 lungs using a 3' primer derived from the connexin 37 sequence (501-518) and variant 5' primers overlapping the putative mutation sites in their 3' end (Fig. 4a) as described by Huang et al., 1991. Priming with the TGT (Cys)-containing primer, amplified products of -300 bp from all complementary DNA (Fig. 4b, primer 1). In contrast, the mutated CAG primer, which represents a Gin sequence, amplified similar products only in tumor cells (Fig. 4b, primer 2). Direct sequencing confirmed that connexin 37 sequences were amplified, indicating that Gin-mutated, as well as normal connexin 37 are expressed in the tumour (Fig. 4c). The 8th amino acid of the peptide was found to be Pro (CCG) and not Ala (GCN) (Fig. 4c). Whether mutations in connexin 37 contribute to the malignancy of lung carcinomas or represent sporadic events during tumor progression is not clear. But transfection of connexin 43 caused growth retardation of glioma cells in vitro and transfection of connexin 32 caused growth retardation of hepatoma cells in vivo, indicating negative correlation between intercellular communication and tumorigenicity (Zhu et al., 1992; Eghabali et al., 1991).

EXAMPLE 5

In vitro CTL activity using MUT 1 and MUT 2 peptides in Incomplete Freund

Adjuvant (IFA) (A) or loaded on RMA-S cells (B)

In this example, we examined the therapeutic efficacy and the immunological consequences of synthetic peptide vaccines. To test possible routes of immunizations that will efficiently induce specific anti-tumor CTL, we immunized mice with peptides presented on antigen-presenting cells (RMA-S), or with peptides injected in IFA.

Groups of C57BL mice were immunized 3 times at 7-day intervals either with inactivated tumor cells RMA-S, D122 and K 39.5 (irradiated 5000 Rad) or with peptides in TFA. Mice received each 20 μg peptide (synthetic MUT 1 or MUT 2, or the homologous normal connexin 37 (Con 37) peptide) in 50 μl of PBS mixed with 50 μl of IFA. The first immunization was done intradermally in anesthetized mice (i.p. injection of Nembutal (Sanofi), 10 mg/kg mouse), whereas the additional immunizations were given subcutaneously (s.c). Ten days after the last immunization, spleens were removed. Splenocytes were restimulated on irradiated (5000 Rad) and mytomicin-treated (80 μg/ml/10 cells for 1 hr) K 39.5 cells for 4 days, and reacted with 5000 S-methionine labeled target cells, at effector to target (E:T) ratio of 50: 1, for 5 hr at 37°C. Anesthetized mice or mice injected with IFA alone did not induce any CTL. The results are shown in Section A of Table 1.

The in-vitro lytic activity of CTL induced by MUT 1 and MUT 2 RMA-S loaded cells was examined. Effector lymphocytes were derived from C57BL mice that had been immunized with 2xl0⁶ irradiated (5000 Rad) D122, K^b39.5, RMA-S and peptide-loaded RMA-S cells. Splenocytes were restimulated as above and reacted with S-methionine labeled target cells. Only the E:T ratio of 25: 1 is shown in Section B of Table 1. Spontaneous release did not exceed 30% of maximal release. Table 1 shows one (out of three) CTL experiment performed. Standard error was under 5% of the means of wells.

As shown in Table 1, immunization with MUT 1 and MUT 2 and K 39.5 cells induced CTL that moderately killed D122 tumor cells or RMA-S cells loaded with MUT 1 and MUT 2 synthetic peptides and efficiently killed K^b39.5 tumor cells. The control groups immunized with RMA-S cells or RMA-S cells loaded with the Con 37 peptide were hardly killed by these CTL. Immunization with the Con 37 peptide in IFA did not induce almost any anti-tumor CTL. CTL that were induced against MUT 1 and MUT 2 peptides in IFA were as efficient as CTL that were induced against the same peptides loaded on RMA-S cells (Table IB). Immunization with peptides solubilized in serum-free medium or PBS were performed either with peptides injected intraperitoneally (i.p). or with peptides injected intradermally (i.d.), followed by subcutaneous (s.c.) injections. None of these immunizations induced anti- tumor CTL activity. TABLE 1

Target cells (%specific lvsis)

Effectors RMA-S D122 K^b39. RMA-S+ RMA-S+ RMA-S+ 25:1 MUT 1 MUT 2 con 37

(100 nM) (100 nM)) (100 nM)

RMA-S 3 0 0 5 6 6 D122 0 0 13 3 7 0

K^b39.5 0 13 42 23 28 0 MUT 1 0 18 50 30 26 0 MUT 2 0 20 39 26 32 0 Con 37 0 0 0 5 6 0

B

RMA-S 0 0 0 0 0 0 D122 0 0 16 12 11 0

K^b39.5 0 20 52 21 24 0 MUT 1 0 15 35 21 25 0 MUT 2 0 18 25 21 21 0 Con 37 0 0 0 0 0 0

EXAMPLE 6

MUT 1 and MUT 2 peptides protect mice from D122 metastasis

In this experiment, we tested whether the ability of the MUT 1 and MUT 2 peptides to induce CTL correlates with their ability to protect mice from D122 metastasis. Groups of mice were immunized three times with peptides in PBS, IFA, or loaded on RMA-S cells, as described in Fig. 5. D122, RMA-S and K^b39.5 cells served as controls for the effectiveness of peptide vaccination. Groups of mice that were not immunized, mice that were anesthetized only and mice that were immunized with PBS in adjuvant, served as negative controls. Ten days after the last immunization, mice were inoculated intrafootpad (i.f.p). with 2x10^ D122 tumor cells. The growth rates of the local tumors were not affected by any of the immunizations. Peptides that were injected i.p. or i.d. in PBS did not protect mice from D122 metastasis. MUT 1 and MUT 2 peptides that were given either on RMA-S or in IFA reduced significantly metastatic spread (p values are presented in the legend to Fig 5), as did immunization with K 39.5 cells. All other groups, including mice that were immunized with the Con 37 peptide (in any of the tested forms), were highly metastatic. In an additional protection experiment, mice were immunized four times with the same peptides, in the same three forms. Using the four immunization protocol, MUT 1 and MUT 2 peptides loaded on RMA-S reduced significantly the growth rates of D 122 local tumors, while immunization in IFA did not affect local tumor growth (not shown).

EXAMPLE 7

Post-surgical immunotherapy is mediated in mice by MUT 1 and M T 2 peptides.

In order to imitate more closely a clinical situation in which established residual tumor or metastases have to be treated, we examined the kinetics of lung metastasis establishment in D122-bearing mice. C57BL mice were injected i.f.p. with 2x10^ D122 tumor cells. At various time points three mice of each group were sacrificed and histological analysis of the lungs was carried out by Haematoxylin/Eosin/Light-Green staining of embedded tissues. Evidence for existence of metastasis was seen from day 14 on: multiple metastatic foci were seen on day 22 post i.f.p. injection which is equivalent to primary tumor size of about 5mm (not shown)

Post-surgical immunotherapy experiments were employed to test whether in these tumor-bearing mice with established micrometastases, therapeutic effects of peptide vaccination may be achieved. Fig. 6 shows one of the post-surgical experiments performed (one out of three). Mice were injected i.f.p. with 2 10^ D122 tumor cells and when local tumors reached diameters of 6-6.5mm, the tumor-bearing feet were amputated and two days later immunizations begun. Mice were immunized with the same vaccination protocols as in Fig. 5, except the group that received peptides in PBS injected i.d./s.c. Thirty days later, according to the death of the control groups, mice were sacrificed and metastatic loads were determined. Mice that were immunized with MUT 1 and MUT 2 peptides either on RMA-S or in IFA showed significantly reduced metastatic loads (p values, Fig. 6). The same effect was seen in mice immunized with K 39.5 cells. In contrast, peptides that were given in PBS i.d./s.c. did not immunize at all. The negative controls, non-immunized mice, anesthetized mice, mice immunized with irradiated D122 or RMA-S cells, or mice immunized with the Con 37 peptide, were all highly metastatic (Fig. 6abc). In each group, the lungs of two mice were stained using India ink for visual presentation. Fig. 6(d) shows the anti-metastatic effect mediated by MUT 1 and MUT 2 peptides loaded on RMA-S (metastatic nodules are seen in white), and Fig. 6(e) shows the anti-metastatic effect mediated by MUT 1 and MUT 2 peptides in IFA.

In one of the post-surgical immunotherapy experiments, mice were not killed thirty days after the amputation, instead they were left for monitoring of survival. The results show that groups of mice that were immunized with MUT 1 and MUT 2 peptides in IFA died from metastasis 45-50 days after the amputation (not shown). Mice immunized with K^b39.5 died on days 36-52 (Fig. 7). Control groups died from metastases 34-36 days after amputation. Seven out of nine mice in the group of mice immunized with RMA-S+MUT 1 peptide were still alive 400 days after amputations, whereas only 2 out of 9 mice were alive in mice immunized with RMA-S+MUT 2 peptide (Fig. 7). Thus MUT 1 is a slightly more immunogenic peptide than MUT 2 and vaccination by peptide presented on RMA-S cells seems to be more long lasting than vaccination in IFA.

EXAMPLE 8

Involvement of CD4 and CD8 T-cells in the immunotherapeutic effect mediated by

MUT 1 and MUT 2 peptides

To determine the effector T-cell populations mediating the anti-metastatic effects of peptide vaccines, we depleted C57BL mice from CD4+ T-cells or CD8+ T-cells, using the GK1.5 monoclonal antibody (anti-CD4) or YTS-169.4 monoclonal antibody (anti-CD8). Depletions were performed during the post-surgical immunotherapy experiments. Fig. 8 shows one of these experiments (one out of two). Groups of mice were injected with D122 cells and immunized with the same vaccines as in Example 6. Mice that were not depleted from CD4 or CD8 cells (Fig. 8, upper panel) showed metastatic profiles similar to the previous immunotherapy experiments (Fig. 6). Mice that were depleted from CD8+ T-cells were all highly metastatic (Fig. 8, lower panel). It seems therefore that the effector cells acting in immunotherapy are indeed CD8 T-cells. Surprisingly, depletion of CD4 T-cells had no effect on metastatic profiles of vaccinated mice, immunization with MUT 1 and MUT 2 peptides on RMA-S and in TFA, or immunization with K 39.5 tumor cells significantly reduced metastatic loads as in non-depleted mice (Fig. 8, middle). Moreover MUT 1 or MUT 2 peptides that were injected i.d./s.c. in PBS also showed anti-metastatic effects in CD4 depleted mice. It seems therefore that depletion of CD4+ T-cells increases the immunity mediated by MUT 1 and MUT 2 peptides.

EXAMPLE 9

The immune response in mice as a function of the site of peptide vaccination.

To test whether the route of injection can influence the immunogenic capacity of soluble peptides, post-surgical immunotherapy experiments were performed as before. Mice were depleted either from CD4+, CD8+ or not depleted, and immunized with MUT 1, MUT 2 or Con 37 peptides in PBS, given i.p. or i.d./s.c. Mice that were not depleted (Fig. 9, upper panel), or mice that were depleted from CD8+ T-cells (Fig 9, lower panel), were all highly metastatic (p values in legend to Fig 9). In CD4+ depleted mice, groups that were injected i.p. with MUT 1, MUT 2 or Con 37 peptides, were all highly metastatic. In contrast, mice that were injected i.d./s.c. with MUT 1 or MUT 2 peptides were almost non-metastatic (Fig. 9, middle panel). Mice immunized with Con 37 either i.p. or i.d./s.c, were as metastatic as the control group. Thus, peptides solubilized in PBS have a distinct therapeutic effect in CD4+ depleted mice when injected i.d./s.c, but not when injected i.p.

The above results demonstrate that established micrometastases of a highly malignant, nonimmunogenic lung carcinoma can be rejected by treatment with TAA peptides. MUT 1 and MUT 2, but not Con 37, induced similar levels of CTL, mediated similarly protection against metastasis, and mediated rejection of metastases in diseased mice. However, in a survival experiment, MUT 1 was shown to be more effective than MUT 2 in complete erradication of post-operative metastases (Fig. 7).

MUT 1, which constitutes the natural D122 TAA peptide, contains a proline residue in position 8, while MUT 2 contains an alanine at the same position. Although the amino acid at position 8 of the peptides are contained within pocket F of the K groove and do not seem to contribute to CTL binding, fine changes in binding to MHC may influence also the affinity to the T-cell receptor. Vaccination with peptides in IFA or loaded on RMA-S seem to have generally a similar effect in-vivo, yet the mechanisms involved in the two modes of vaccination are probably different. Intradermal injection of peptides in IFA probably recruits professional antigen presenting cells (APC) like dentritic cells or subsets of macrophages. Anti-tumor immunity is CD8+ T-cell dependent as judged from depletion experiments (Fig. 8). In contrast, CD4+ T-cell depletion has little effect on therapy by peptides in TFA or loaded on RMA-S cells (Fig. 8). The ability of TAA peptides in PBS to vaccinate against tumors when delivered i.d. but not i.p. might be a function of local availability of APC in the skin, like Langerhans cells, rather than site-associated suppression.

Potentially, peptide vaccination is an attractive mode of anti-tumor therapy. Studies by Feltkamp et al., 1993, have shown that protection against an HPV-16 induced C57BL tumor can be achieved by preimmunization with E7-derived synthetic peptides. According to the present invention, it is for the first demonstrated that established metastases can be cured by a CTL epitope which represent a mutated cellular gene and that most cured mice at a given protocol achieve long term survival.

REFERENCES

1. Boon, T. et al., Annu. Rev. Immunol. 12, 337 (1994).

2. Dranoff , G. et al ., Proc. Natl. Acad. Sci. USA. 90, 3539 (1993).

3. Eghabali, B., Kessler, J.A., Reid, L.M. Rou, C. & Spray, D.C. Proc natn. Acad. Sci. U.S.A. 88, 10701-10705 (1991).

4. Falk, K., Rotzschke, O., Stevanovic, S., Jung, G, & Rammensee, H.G., Nature, 351, 290-296 (1991(a)) and Falk, K., et al., J. Exp. Med. 174, 425-434 (1991(b)). 5. Feltkamp et al. Eur. J. Immunol. 23, 2242 (1993).

6. Huang, M.M. et al., Nucleic Acids Res. 20, 4567-4573 (1991).

7. Mandelboim, O., Feldman, M., & Eisenbach, L., J. Immun. 148, 3666-73 (1992).

8. Melief, C.J.M., Adv. in Cancer. Res. 58, (1992).

9. Plaksin, D., Gelber, C, & Eisenbach, L., Int. J. Cancer, 52, 771-777 (1992).

10. Plaksin, D., Gelber, C, Vadai, E., Feldman, M., & Eisenbach, L., Proc. Natl. Acad. Sci. USA, 85, 4463-4470 (1988).

11. Rammensee, H.G., Falk, K., & Rotzschke, O., Rev. Immun. 11, 213-244 (1993).

12. Rosenberg, S. A., New Engl. J. Exp. Med. 319, 1676 (1988).

13. Rotzschke, 0. et al., Nature, 348, 252-254 (1990).

14. Townsend, A. et al., Nature, 340, 443-448 (1989).

15. Udaka, K., Tsomides, T.J. & Eisen, N.H., Cell, 69, 989-998 (1992).

16. Urban, J. L. and H. Schreiber, Annu. Rev. Immunol. 10, 617 (1992).

17. Van Bleek, GM. & Nathenson, S.G., Nature, 348, 213-216 (1990).

18. Van Den Eynde, B., Lethe, B., Van Pel, A., De Plaen E., & Boon, T. J., Exp. Med., 173, 1373-1384 (1991).

19. Van Der Bruggen, P. et al., Science, 254, 1643-1647 (1991). 20. Zhu, D., Kidder, G.M., Caveney, S. & Naus. C.C. Proc. Natl. Acad. Sci. U.S.A. 89, 10218.

Claims

1. A synthetic peptide of the sequence I:

Phe-Glu-Gln-Asn-Thr-Ala-Gln-Xi -X₂ (I) wherein X] is Ala or Pro, and X2 is Gly or OH, and biologically active analogs thereof, said peptides of sequence I and their biologically active analogs being capable of: (i) binding to H-2K molecules on RMA-S cells, (ii) sensitizing RMA-S cells to lysis by specific anti- tumor CTL, and (iii) inducing specific CTL in vivo, and (iv) protecting mice from metastasis of 3LL lung carcinoma.

2. A synthetic peptide according to claim 1, designated MUT 1, of the sequence:

Phe-Glu-Gln-Asn-Thr-Ala-Gln-Pro

3. A synthetic peptide according to claim 1, designated MUT 2, of the sequence:

Phe-Glu-Gln-Asn-Thr-Ala-Gln-Ala

4. Use of a peptide according to any one of claims 1 to 3 in a method for the diagnosis of tumors.

5. Use of a peptide according to any one of claims 1 to3 for the preparation of anti-tumor antibodies.

6. Use of a peptide according to any one of claims 1 to 3 for the preparation of specific anti-tumor cytotoxic T lymphocytes.

7. A vaccine for the treatment of tumor metastasis comprising an effective amount of a peptide according to any one of claims 1 to 3 in a suitable adjuvant.

8. Use of a peptide according to any one of claims 1 to 3 for the preparation of a composition suitable for regression of carcinoma metastasis.

9. A method of treatment to cause regression of carcinoma metastasis in a patient, said method comprising administering to a patient in need thereof effective amounts of a peptide according to claim 1.