Specific cytotoxicity of a long-term cultured T-cell clone on human autologous mammary cancer cells

We established an autologous specific T-cell killer clone, TcHMC-1, that has been cultured and has retained its function for over 1 year. TcHMC-1 and target cells (HMC-1-8) were derived from the metastatic pleural effusion of a patient with mammary carcinoma. At culture initiation, pleural exudative lymphocytes (PLEL) already demonstrated a high cytotoxic activity against uncloned HMC-1 breast tumor cell targets but not against autologous fibroblasts and K562 targets, and phenotypically these cells showed 100 and 90% reactivity with OKT3 and OKT8 monoclonal antibodies, respectively. However, at the early phase of cultivation under interleukin 2, PLEL had a relatively high cytotoxicity against some allogeneic tumor cells. Furthermore, the longer these PLEL were cultured with interleukin 2 and stimulated with MMC-treated HMC-1, the less cytotoxic activity of PLEL against HMC-1 targets became. We then cloned PLEL as well as HMC-1 tumor cells, and an autologous pair of TcHMC-1 and a target cell clone, HMC-1-8, was successfully obtained. TcHMC-1 showed more than 60% specific cytotoxicity against HMC-1-8, and it was confirmed, using cold target inhibition assays, that TcHMC-1 did not demonstrate nonspecific cytotoxicity against allogeneic targets as well as the natural killer cell activity. Moreover, we examined the in vivo action of TcHMC-1 against HMC-1-8 cells by the Winn assay using nude mice. The data showed that s.c. injections with a mixture of TcHMC-1 and HMC-1-8 clearly resulted in a failure of tumor development in the nude mice even 12 weeks after injections, whereas mice given injections of HMC-1-8 and allogeneic T-lymphocytes cultured with interleukin 2 developed tumors. The autologous pair of a killer T-cell clone and tumor line could be very useful for future investigations of the specific destruction of autologous tumor cells by cytotoxic T-lymphocytes, including analysis for tumor-specific antigens possibly of rejection type and clonotypic T-cell antigen receptors.     

Identification of a human T cell clone with the cytotoxic T lymphocyte and natural killer-like cytotoxic function against autologous mammary carcinoma and K562 line

Pleural exudative lymphocytes (PLEL) from a 60-year-old female patient showed high cytotoxicity against the autologous mammary tumor line, HMC-2, and NK-susceptible K562 cells, although PLEL demonstrated only weak cytotoxic potentials against several allogeneic tumor lines. We successfully obtained seven cytotoxic T cell clones from PLEL bulk populations, and assessed the possibility that these lymphocytes are simply natural killer (NK)-like cells or have the dual cytotoxic activity of cytotoxic T lymphocytes (CTL) and NK-like cells. These clones, designated as TcHMC-2, showed strong cytotoxicity against both HMC-2 and K562 cells. In contrast, allogeneic human peripheral blood-derived NK cells could not kill HMC-2 targets. Furthermore, a blocking study of TcHMC-2 cytotoxicity using monoclonal antibodies against CD3, CD8 and human MHC class I products showed that all of these antigen molecules were involved in the cytotoxicity of TcHMC-2 clone against autologous HMC-2 cells, indicating MHC class I recognitive cytotoxicity. These data indicate that the TcHMC-2 clone may have dual cytotoxicity with CTL- and NK-like activity against autologous HMC-2 mammary tumor and K562 cells, respectively.     

Characterization of n-butyl alcohol solubilized, breast tumor specific antigens recognized by a human autologous cytotoxic T-cell clone

We demonstrated previously the establishment of a human cytotoxic T-cell clone, TcHMC-1, under culturing with recombinant interleukin that showed the specific cytotoxicity against an autologous breast tumor cell line, HMC-1-8. In the present study, the autologous tumor specific antigens that could be involved in this cytotoxicity were extracted by using n-butyl alcohol and were analyzed for their biochemical profiles. The cytotoxicity of TcHMC-1 against HMC-1-8 was inhibited by adding OKT3 and OKT8 monoclonal antibodies into the cultures, or by pre-sensitizing HMC-1-8 target cells by anti-major histocompatibility complex class I monoclonal antibodies. This suggests that T-cell antigen receptor molecule complexes Ti/T3 on TcHMC-1 and corresponding specific tumor antigens on HMC-1-8 are involved in the cytotoxicity under the restriction of major histocompatibility complex class I products. Precultures of TcHMC-1 with crude n-butyl alcohol extracts from HMC-1-8 cells enhanced the cytotoxic potentials of this clone as seen as mixed lymphocyte tumor cell cultures. This enhancement was dependent on dosage of crude n-butyl alcohol extracts and these TcHMC-1 cells were still cytotoxic specifically for HMC-1-8 targets, but not for other allogenic tumor lines including K562. However, HMC-1-8 crude n-butyl alcohol extracts could not enhance DNA synthesis of TcHMC-1 as assessed by incorporation of [3H]thymidine in the cells. Biochemical purification studies demonstrated that the HMC-1-8 tumor specific antigens were eluted into fractions containing molecules with molecular weights of approximately 200,000 on Sephadex G-200 column chromatography. The antigens were further separated into the fraction that was eluted with 0.4-0.5 M NaCl in an ionic strength on Mono Q fast protein liquid chromatography. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of this fraction demonstrated three molecules with molecular weights of 26,000, 30,000, and 32,000 under reduced molecular conditions. The data suggest that these molecules could be tumor specific antigens that are involved in the cytotoxicity of cytotoxic T-cells against a human autologous tumor.     

Identification of a Human T Cell Clone with the Cytotoxic T Lymphocyte and Natural Killer-like Cytotoxic Function against Autologous Mammary Carcinoma and K562 Line

Pleural exudative lymphocytes (PLED from a 60-year-old female patient showed high cytotoxicity against the autologous mammary tumor line, HMC-2, and NK-susceptible K562 cells, although PLEL demonstrated only weak cytotoxic potentials against several allogeneic tumor lines. We successfully obtained seven cytotoxic T cell clones from PLEL bulk populations, and assessed the possibility that these lymphocytes are simply natural killer (NK)-like cells or have the dual cytotoxic activity of cytotoxic T lymphocytes (CTL) and NK-like cells. These clones, designated as T_cHMC-2, showed strong cytotoxicity against both HMC-2 and K562 cells. In contrast, allogeneic human peripheral blood-derived NK cells could not kill HMC-2 targets. Furthermore, a blocking study of T_cHMC-2 cytotoxicity using monoclonal antibodies against CD3, CD8 and human MHC class I products showed that all of these antigen molecules were involved in the cytotoxicity of T_cHMC.2 clone against autologous HMC-2 cells, indicating MHC class I recognitive cytotoxicity. These data indicate that the T_cHMC-2 clone may have dual cytotoxicity with CTL- and NK-like activity against autologous HMC-2 mammary tumor and K562 cells, respectively.     

Biological behavior of human breast carcinoma-associated antigens expressed during cellular proliferation

The cell surface-binding properties of two murine monoclonal antibodies reactive with human mammary tumor cells are described. Fluorescence-activated cell sorter analyses demonstrate that both monoclonal antibodies, B6.2 and B38.1, are reactive with the surface of the majority of human breast tumor cell lines tested but are unreactive with a variety of normal human cell lines, melanomas, sarcomas, and lymphoid tumors. Antibody B6.2 was also reactive with selective carcinomas, while antibody B38.1 showed even broader reactivity. The two monoclonal antibodies were unreactive with the surface of a variety of normal human tissues obtained at biopsy, including lymph nodes, bone marrow, and spleen, but were reactive with mammary tumor cells obtained from four of six pleural effusions. Surface binding to mammary tumor cells by both monoclonal antibodies was shown to decrease during density-dependent arrest; further cell cycle analysis demonstrated differential antibody surface binding during S phase. Prolonged exposure of mammary tumor cells to antibody showed no evidence of antigen capping or internalization. Both monoclonal antibodies were shown to lyse mammary tumor cells in antibody-dependent cell-mediated cytotoxicity.     

Characterization of human autotumor-reactive T-cell clones obtained from tumor-infiltrating lymphocytes in liver metastasis of gastric carcinoma.

Three autotumor-reactive T-cell clones have been established from tumor-infiltrating lymphocytes isolated from a metastatic lesion of human gastric carcinoma in the liver. The clones all were shown to be CD3+, CD8+, CD4-, CD16-, T-cell receptor alpha/beta +, and T-cell receptor gamma/delta-, and they have retained both their autotumor reactivity and the same phenotype for over a year in culture. Each clone had a different rearrangement of the T-cell receptor gamma chain genes as indicated by Southern blot analysis. Tested against a panel of 18 tumor cell targets, the clones preferentially lysed autologous tumor (AuTu) cells, but each clone also showed weak cytotoxicity against one allogeneic cholangiocarcinoma cell line. At the same time, each clone showed appreciable cytotoxicity against K562 targets. In blocking experiments, anti-CD3, anti-WT31, anti-CD8, or anti-HLA class I monoclonal antibodies blocked AuTu cytotoxicity but not cytotoxicity against K562. In contrast, allocytotoxicity against the cholangiocarcinoma was blocked only by anti-CD3 monoclonal antibody. All 10 subclones of one T-cell clone had high levels of AuTu cytotoxicity but variable levels of anti-K562 cytotoxic activity. Proliferation of the T-cell clones was significantly stimulated by the addition of irradiated autologous but not allogeneic tumor cells. Preincubation of cultured AuTu cells with tumor necrosis factor alpha or gamma-interferon (IFN-gamma), but not with IFN-alpha, increased their susceptibility to lysis by the T-cell clones; however, it increased resistance of AuTu to lysis by interleukin 2-activated natural killer cells. The expression of an adhesion molecule, intercellular adhesion molecule 1, on the surface of AuTu was also up-regulated by tumor necrosis factor alpha or IFN-gamma, but not by IFN-alpha. All three cytokines up-regulated HLA-class-I antigens on AuTu. Pretreatment of K562 targets or allogeneic cholangiocarcinoma cells with the same cytokines increased their resistance to lysis by the T-cell clones. Overall, the results indicate that these T-cell clones show specificity for AuTu but also independently recognize a limited number of allogeneic tumor targets and lyse K562 targets. The mechanisms involved in the recognition by the T-cell clones of autologous, allogeneic, and K562 tumor targets appeared to be distinct.     

Targeting of interleukin-2 to human MK-1-expressing carcinoma by fusion with a single-chain Fv of anti-MK-1 antibody

BACKGROUND: Targeting of cytokines into the tumor sites using antibody-cytokine fission proteins represents a novel approach in cancer immunotherapy. We previously reported a novel monoclonal antibody, FU-MK-1, which recognizes a glycoprotein antigen (termed MK-1 antigen) that is overexpressed on the surface of a majority of carcinomas. MATERIALS AND METHODS: To target IL-2 and cytotoxicity of effector cells to MK-1-expressing tumor cells, we genetically fused recombinant human interleukin-2 (rhIL-2) to a single chain variable fragment (scFv) antibody derived from FU-MK-1. The resulting fission protein, designated FUscFv/IL-2 was expressed in Pichia pastoris, purified by Ni-affinity chromatography, and characterized for the MK-1-binding specificity and the IL-2 biological activity. RESULTS: The FUscFv/IL-2 fusion protein effectively introduced a specific cytotoxicity of lymphokine-activated killer cells to the tumor cells and consequently suppressed the tumor growth in a SCID mouse xenograft model. CONCLUSION: This approach may be used for in vivo administration to localize IL-2 to tumor tissues, enhancing the immune response to human MK-1-expressing tumors while reducing systemic side-effects.     

Identification of a tumor-associated target antigen, ATM-1, for a human T-cell clone with activated killer activity and its existence in sera of cancer patients

Three human T-cell clones with activated killer activity (5B5, 5C1, and 7B5) which could lyse various tumor cell lines were established. The cytotoxic activity of these clones was decreased by incubation with anti-CD3 monoclonal antibody, suggesting that they recognized tumor cells by T-cell antigen receptor. A monoclonal antibody which blocked the cytotoxic activity of clone 5B5 was obtained. This antibody (N1977) blocked the binding and cytotoxic activity of clone 5B5 at the target cell level, suggesting that the antigen defined by N1977 antibody, designated as ATM-1, was a target molecule recognized by 5B5 cells. ATM-1 in the conditioned medium of a cancer cell line (NBT-2) and serum from a patient with lung cancer was characterized by following its immunoreactivity. On gel filtration, both the conditioned medium and the serum gave three peaks of ATM-1 immunoreactivity, corresponding to approximate molecular weights of 1,200,000, 700,000, and 120,000, respectively. They were chromatofocused at pH 4.0, 4.8, and 6.5, respectively. The high molecular weight forms were shown to be molecules with the disulfide-linked elementary glycoprotein with ATM-1 immunoreactivity and approximate molecular weight of 120,000. Most of the molecules with ATM-1 immunoreactivity bound to both concanavalin A and wheat germ agglutinin, and their binding activity to the antibodies was lost by treatment at 60 degrees C for 30 min. An assay of ATM-1 level in sera was performed by a sandwich enzyme immunoassay. The following positive percentages were obtained from preliminary clinical studies: breast cancer, 67% (8 of 12 cases); hepatocellular carcinoma, 83% (10 of 12 cases); gastric cancer, 58% (7 of 12 cases); lung cancer, 41% (5 of 12 cases); hematological malignancies, 0% (0 of 9 cases); systemic lupus erythematosus, 0% (0 of 8 cases); rheumatoid arthritis, 0% (0 of 8 cases).     

Brefeldin A Blocks the Cytotoxicity of T Cell Receptor α/β and γ/δ Cytotoxic T Lymphocyte Clones Reacting against Human Autologous Cancer Cells

We studied the effector mechanism of T cell receptor (TCR) α/β - and γ/δ -type cytotoxic T lymphocyte (CTL) clones that react with human autologous tumor cells. Treatment of tumor cells with a fungal antibacterial reagent, brefeldin A (BFA), resulted in the inhibition of cytotoxicity of an autologous tumor (HST-2)-specific CD8⁺ TCR α/β -type CTL, TcHST-2. Other anti-metabolites such as chloroquine, cycloheximide and colchicine did not affect the cytotoxicity. The cell-surface antigen expression, including MHC class I molecules, was not influenced by BFA treatment. Furthermore, BFA did not influence the cytotoxicity of lymphokine-activated killer cells and natural killer cells. Since BFA blocks the transport of peptides from endoplasmic reticulum to the Golgi apparatus, the above data suggest that BFA could affect washing out of the peptide fragments from the MHC class I groove. Consequently, target tumor cells were protected from killing by CTL, Moreover, we obtained a CD4⁻, 8⁻, TCR γ/δ -type (Vδ1⁺) CTL clone, TcHOT, that reacts against an autologous ovarial carcinoma, HOT. BFA could also inhibit this cytotoxicity, and it is likely that different presenting molecules other than MHC class I proteins participate in the cytotoxicity of this TCR gm/δ - type CTL. These studies suggest that both TCR α/β - and γ/δ -type CTL may require antigenic peptides that are most likely derived from the BFA-sensitive, intracellular endogenous target proteins.     

X-irradiation to human malignant glioma cells enhances the cytotoxicity of autologous killer lymphocytes under specific conditions

PURPOSE: To clarify the effect of combining x-irradiation and human killer lymphocytes against autologous malignant glioma cells, we analyzed not only the alteration of surface antigen expression in irradiated tumor cells, but also the cytotoxic effects of human killer lymphocytes on the autologous tumor cells with and without x-irradiation. METHODS AND MATERIALS: Six malignant glioma cell-lines (MG 1-6) established from each patient with a malignant glioma in our institute, and U87MG, were used as materials. They were irradiated by 0-50 Gy of X-rays, and the alternations of their human histocompatability leukocyte antigen (HLA), HLA-ABC, HLA-DR, -DP, -DQ, and FAS expressions were examined. Then, three sets of autologous natural killer (NK) cells, and autologous tumor-specific T lymphocytes (ATTL) were induced from the peripheral blood mononuclear cells (PBMCs) of three patients, and in vitro cytotoxic effects of these killer cells on the irradiated autologous tumor cells were analyzed. RESULTS: Irradiation-enhanced HLA-DR, -DP, -DQ, and FAS expression in glioma cell lines with low p53 expression. However, there was no correlation between HLA-ABC expression and X-ray dose. After irradiation of the tumor cells, cytotoxicity was enhanced in four of six effectors; in particular, it was significantly elevated in two killer lymphocytes. It was speculated that the enhancing effect was influenced not only by the p53 status of the tumor, but also by the types of killer lymphocytes; the alteration of cytotoxicity in NK cells on irradiated tumor cells may be compensatory for alteration in ATTLs. CONCLUSION: It was indicated that irradiation of malignant tumor cells enhanced killer cell-mediated cytotoxicity in autologous models under specific conditions. These basic data should contribute to clinical trials using local radiotherapy and systemic adoptive immunotherapy with killer lymphocytes.     

Cytolysis of mammary tumor targets by resting, interleukin-2 stimulated and in vitro cultured peripheral blood lymphocytes from breast cancer patients

The cytotoxic capacity of resting, interleukin-2 (IL-2)-stimulated and in vitro cultured (3-5 days in 10 U/ml IL-2 containing media) peripheral blood lymphocytes (PBLs) from breast cancer patients to a panel of established mammary tumor cell lines was ascertained. Significant cytolysis (ranging from 7.8 to 12.4%, at an effector: target ratio of 20:1) of all mammary tumor targets (MCF-7, 734B, ZR-75-1, ZR-75-30, BT-20 and Hs578T) by PBLs was demonstrable in 18 h chromium release assays. Natural killer (NK) cytotoxicity was distinct from IL-2 stimulated (5 U/ml) and in vitro cultured PBL cytotoxicity in that resting PBLs were not cytolytic to RAJI cells, normal breast epithelia (Hs578Bst) and fibroblasts. Basal NK activity against mammary tumor targets was significantly reduced in patients receiving chemotherapy when compared to both untreated patients and normal controls. In criss-cross cold target inhibition studies, ZR-75-1 and K562 targets were not mutually competitive in NK cell assays (using resting PBLs) but were mutually competitive in lymphokine-activated killer (LAK) assays (using in vitro cultured PBLs). In eleven independent experiments, basal NK activity of ZR-75-1 cells was increased by a cold target excess of K562 (8.2 +/- 2.4% vs 30.5 +/- 5.2%, mean +/- SE, p greater than 0.01, cold:hot target ratio = 10:1). Interestingly, no such parallel increase of cytolysis of 734B targets by K562 cells was observed. Basal cytotoxicity against ZR-75-1 and K562 targets was serologically depleted using antibodies to natural killer cells HNK-1 and Leu 11b. Thus mammary tumor cell lines parallel autologous tumor cells, yet show features that are distinct from NK-resistant and sensitive lymphoid cell lines in their susceptibility to natural resistant cytolytic mechanisms.     

A monoclonal antibody (B72.3) defines patterns of distribution of a novel tumor-associated antigen in human mammary carcinoma cell populations

We report here both the range and patterns of reactivity of an IgG1 monoclonal antibody, B72.3, prepared against human, metastatic mammary carcinoma cells. When the avidin-biotin complex (ABC) immunoperoxidase technique was used on tissue sections, monoclonal B72.3 reacted with 19 of 41 (46%) primary mammary carcinomas and 13 of 21 (62%) metastatic lesions, either in axillary lymph nodes or at distal sites. Variable concentrations of antigen, recognized by B72.3, were observed among mammary tumors, as well as among different cell populations of a given tumor mass. Several patterns of antigen distribution were observed: membrane, diffuse cytoplasmic, focal and marginal. No reactivity was observed to normal mammary epithelium, stroma, or lymphocytes of the breast, nor to any cell types in a variety of other normal human tissues, melanomas, and sarcomas. Reactivity with all of four colon carcinomas was also observed. Assay of serial sections of mammary carcinomas with B72.3 and a monoclonal antibody directed against carcinoembryonic antigen demonstrated that these antigens were both distinct and non-coordinately expressed.     

Characterization of a membrane-associated glycoprotein (gp 43) on human hepatocellular carcinoma by a monoclonal antibody.

A monoclonal antibody, Hepama-1, produced by immunizing mice with cells of a human hepatocellular carcinoma cell line, has been used to identify and characterize a previously unreported antigen present on the surface of human hepatocellular carcinoma cells. The antigen occurred on the membranes of human hepatoma cell lines and tumor biopsies but was not detectable in tumors of other origin or normal tissues. Binding was determined by enzyme-linked immunoabsorbent assay and immunofluorescence on cell lines and by immunoperoxidase staining of tissue sections. In immunofluorescence studies, Hepama-1 antibodies stained five out of six human hepatoma cell lines, showed only slight binding to breast tumor cell lines, but failed to stain colon tumor or normal cell lines. The antihepatoma antibody exhibited positive immunoperoxidase staining of human liver tumor sections but did not stain tumors of other origin. Hepama-1 bound specifically to a membrane glycoprotein with an approximate molecular weight of 43,000. Western blot and solid phase enzyme-linked immunoabsorbent assay analysis showed that the 43-kD antigen occurred on five of six human hepatoma cell lines and was expressed by every human hepatocellular carcinoma biopsy tested. This cell surface molecule represents a potentially useful target for immunotherapy and localization of human hepatocellular carcinomas.     

Cytotoxicity of Histocompatibility Leukocyte Antigen–DR8–restricted CD4+ Killer T Cells against Human Autologous Squamous Cell Carcinoma

Although CD8⁺ killer T cells reacting against human autologous tumor cells have recently been studied in detail, little is known about the cytotoxic mechanism of CD4⁺ T cells against such tumor cells. In order to investigate this, we have established CD4⁺ cytotoxic T lymphocyte TcOSC–20 lines. TcOSC–20 showed selective cytotoxic activity against autologous OSC–20 cells, derived from a cancer of the tongue, in an HLA–DR–restricted fashion. HLA–DR8 (DRB1* 08032) is the only DR molecule expressed on OSC–20 cells, and anti–DRS monoclonal antibody could inhibit the Cytotoxicity, suggesting that HLA–DRB1^★ 08032 is the tumor rejection antigen–presenting moleculeto TcOSC–20. The Fas ligand was expressed on TcOSC–20 lines, and its expression was induced upon mixed lymphocyte–tumor cell culture of autologous peripheral blood lymphocytes. Furthermore, the Cytotoxicity of TcOSC–20 was inhibited by anti–Fas ligand antibody.These data imply that TcOSC–20 lines recognize the tumor antigenic peptide presented by HLA–DR8, and exert Cytotoxicity against autologous tumor cells via a Fas–mediated cytotoxic pathway.     

Immune responses of adenovirus type-9 infected (tumor-free) male rats against adenovirus type-9 induced mammary fibroadenomas and a chemically induced mammary carcinoma

Lymphocytes from five non-tumor-bearing male W/Fu rats infected with human adenovirus type-9 as newborns or as adults share a common reactivity against rat mammary fibroadenoma target cells and rat mammary carcinoma target cells, as demonstrated by microcytotoxicity tests. Mammary fibroadenomas were induced by adenovirus type-9 infection of newborn female W/Fu rats, littermates of three of the male lymphocyte donors. The mammary carcinoma target cells were derived from a single rat mammary carcinoma induced by 3,2′-dimethyl-4-amino-biphenyl (DMABP). Lymphocytes from the male adenovirus type-9 infected rats were not cytotoxic to normal rat breast epithelial cells, nor to target cells explanted from a polyoma-virus-induced sarcoma. Sera from each of five mammary fibroadenoma-bearing female rats inhibited the cytotoxic effect of lymphocytes from the adenovirus type-9 infected males against both mammary fibroadenoma and carcinoma target cells. Sera from the female W/Fu rat bearing the DMABP induced mammary carcinoma also inhibited the male lymphocyte cytotoxicity against both fibroadenoma and carcinoma target cells. Serum from rats bearing a polyoma-virus-induced sarcoma, though actively blocking in its own system, had no effect on lymphocytes from the adenovirus type-9 infected males. Anti-viral antibody response was demonstrated 8 days following adenovirus type-9 infection of adult males. No anti-tumor humoral response was demonstrated in either newborn-infected or adult adenovirus type-9 infected W/Fu male rats. Sera were checked for cytotoxicity against fibroadenoma and mammary carcinoma target cells in the presence of active homologous complement. No complement-dependent cytotoxicity was evident. Sera from infected males were admixed to known blocking sera from mammary fibroadenoma- and carcinoma-bearing rats, with no significant abrogation of the blocking effect. These results indicate that, despite the lack of tumor development, infection of male W/Fu rats by adenovirus type-9 has produced an antigen which is at least partially common to the antigen (s) shared by adenovirus type-9 induced mammary fibroadenomas and a DMABP induced mammary carcinoma.     

Visualization of effective tumor targeting by CD8+ natural killer T cells redirected with bispecific antibody F(ab')(2)HER2xCD3

HER2 is an attractive immunotherapeutic target for neoplastic disease because this cell surface molecule is overexpressed on a large fraction of malignant tumor cells. To directly assess therapeutic responses to targeted therapy by noninvasive in vivo imaging in small animals, human HER2-expressing ovarian carcinoma cells were genetically modified with a firefly luciferase gene, and light emission was used for visualization of tumor growth and response to therapy. This imaging approach was able to demonstrate in real-time tumor regression in a HER2 xenograft mouse model by adoptive transfer of in vitro induced and expanded cytotoxic CD8+ natural killer T (NKT) cells retargeted with a humanized bispecific antibody F(ab')(2)HER2xCD3. Immunotherapy with effector cells alone or a humanized monoclonal antibody anti-p185(HER2) (4D5-8) resulted in significant but slower reduction in tumor burden. Long-term survival of tumor xenografts correlated inversely with visible residual tumor burden. In vitro, F(ab')(2)HER2xCD3 substantially augmented cytotoxic activity of CD8+ NKT cells. By flow-sorting, CD8+ NKT cells coexpressing CD56 were found to have the highest redirected killing ability. Treatment with concanamycin A or EGTA abrogated CD8+ NKT cytotoxicity indicating that perforin is a major pathway of tumor cell lysis. In contrast, when CD8+ NKT cell were cross-linked with F(ab')(2)HER2xCD3 neither the immunosuppressants cyclosporine A and FK506, nor the increase of intracellular cyclic AMP by dibutyryl cyclic AMP were able to inhibit cytotoxicity demonstrating that signaling via the CD3 antigen changes the biological activity of non-MHC-restricted effector cells. These studies have demonstrated that CD8+ NKT cells can be successfully redirected to tumor cells using bispecific antibodies and offer a promising strategy for adoptive immunotherapy of neoplastic diseases.     

Induction of human monocyte-mediated tumor cell killing by a plant lectin, wheat germ agglutinin

Human peripheral blood monocytes from healthy donors were separated by discontinuous gradient centrifugation and adherence to yield highly purified adherent cell populations (greater than 99% monocytes). Five different plant lectins were tested for ability to induce lectin-dependent monocyte-mediated cytotoxicity (LDMC). Only one lectin, wheat germ agglutinin (WGA), induced significant and reproducible LDMC activity. All the tumor target cells tested were sensitive to variable extents to cytotoxicity mediated by WGA-treated monocytes. Pretreatment of monocytes with WGA did not result in development of LDMC. N-Acetylglucosamine, which specifically binds WGA, inhibited WGA-dependent monocyte-mediated cytotoxicity. Treatment of adherent monocyte-rich monolayers with monoclonal anti-natural killer cell antibody (anti-Leu-11b) and complement did not affect the LDMC activity induced by WGA. These results indicate that the plant lectin WGA, which binds specifically to both human monocytes and tumor cells, renders human blood monocytes cytotoxic to human tumor cells.     

Autologous mixed lymphocyte-tumor reaction and autologous mixed lymphocyte reaction. II. Generation of specific and non-specific killer T cells capable of lysing autologous tumor

The specific and non-specific nature of autotumor cytotoxicity induced in autologous mixed lymphocyte-tumor culture (AMLTC) and autologous mixed lymphocyte culture (AMLC) was studied in patients with carcinomatous pleural effusions. Small- and medium-sized blood lymphocytes that were isolated by centrifugation on discontinuous Percoll gradients did not lyse autologous, freshly isolated effusion tumor cells. In vitro activation of the small lymphocytes, but not of the medium lymphocytes, with autologous tumor cells generated cytotoxic potential restricted to autologous tumor. When stimulated with autologous non-malignant non-T cells, the medium lymphocytes, but not small lymphocytes, were triggered to cytotoxicity that acted not only on autologous tumor cells but also on allogeneic tumor cells, T blasts, and tumor cell lines. Experiments using monoclonal antibodies (MAb) and complement (C') showed that both types of killer cells were CD2+ CD3+ CD16- T cells. Autotumor cytotoxicity developed in AMLTC was mediated by the CD4- CD8+ T cell subset in 6 of 9 cases and the CD4+ CD8- subset in the other 3 cases. In contrast, cytotoxicity induced in AMLC was exerted exclusively by the CD8+ subset. The enrichment of blasts from cultured T cells on discontinuous density gradients enhanced autotumor killing activity, with no reactivity recorded for blast-depleted, resting T cells. Addition of mitomycin-C-treated large granular lymphocytes (LGL) to AMLTC abolished the induction of autotumor killer cells, whereas non-specific killer cells were generated in AMLC irrespective of the presence of LGL. These results indicate that stimulation of autoreactive T cells in AMLTC and in AMLC could induce 2 distinct types of autotumor killer cells.     

A novel monoclonal antibody (7B10) with differential reactivity between human mammary carcinoma and normal breast

A monoclonal antibody (7B10) which displays differential reactivity with breast carcinomas compared to benign lesions or normal breast tissue was selected by fusion of spleen cells from BALB/c mice immunized with the T47D human mammary carcinoma cell line. The antigen, recognized by 7B10 on T47D cells, appeared to be both surface and cytoplasm localized, as demonstrated by indirect immunofluorescence, immunoperoxidase, and electron microscopy studies. This antibody (IgG1) bound with four human breast cancer cell lines (T47D, MCF7, ZR-75-1, and HSL53) which express estrogen receptors. No binding was observed with cancer cell lines of other origin or with normal cells. In vivo, by immunoperoxidase staining of frozen sections of normal breast, the antigen recognized by 7B10 appeared to be located on epithelial cell membranes, whereas in benign and malignant mammary disorders, staining also involved the cytoplasm, as confirmed by electron microscopy on fresh cancer tissue. On formalin-fixed, paraffin-embedded sections, cytoplasmic staining was detected in breast cancer, but no immunostaining was observed with benign lesions or normal breast. In paraffin sections, most normal tissues investigated did not react with 7B10 antibody. However, ducts in the parotid gland, tubules in the kidney and some biliary ducts, and apocrine glands in the skin showed irregular, diffuse weak staining. 7B10 was unreactive with adenocarcinomas of origin other than breast, except for some cells in ovarian clear cell carcinoma. No reactivity was observed with squamous carcinomas, lymphomas, or melanomas. The antigen recognized by 7B10 appeared to be a Mr 32,000 protein, as identified by immunoprecipitation from extracts of T47D after labeling with [35S]methionine. Since the antigen was present only on the membrane of differentiated normal mammary epithelial cells, and was also expressed in the cytoplasm of tumor cells, it may be of interest in immunological studies of mammary epithelial cell differentiation. Moreover, since in formalin-fixed tissues immunostaining is virtually confined to mammary carcinomas, monoclonal antibody 7B10 may have diagnostic applications in breast cancer.     

Lymphocytes infiltrating human ovarian tumors: synergy between tumor necrosis factor alpha and interleukin 2 in the generation of CD8+ effectors from tumor-infiltrating lymphocytes

Tumor-infiltrating lymphocytes (TIL) were isolated by enzymatic digestion and gradient centrifugation from 18 human ovarian carcinomas. These cells were cultured in a complete medium supplemented with recombinant interleukin 2 (IL2) alone or recombinant IL2 plus recombinant tumor necrosis factor alpha (TNF-alpha), and their growth and antitumor cytotoxicity were determined. TIL cultured in the presence of IL2 plus TNF-alpha (1000 units/ml each) for 6 days showed significantly higher cytotoxicity against fresh autologous tumor targets than did TIL cultured with IL2 alone (e.g., mean lytic units/10(7) cells for 8 TIL preparations were 290 versus 74; P less than 0.05). No differences in [3H]thymidine uptake or natural killer cell activity were observed among these TIL cultures. In titration experiments, optimal synergistic concentrations of IL2 and TNF-alpha were determined as 10(2) and 10(3) units/ml, respectively. Using these concentrations for culturing the TIL, effector cells developed which preferentially lysed autologous tumor and displayed a CD8+ phenotype (up to 75% positive). However, the autologous tumor cytotoxicity mediated by these cultured TIL on day 6 was short lived. By day 12, it was replaced by non-major histocompatibility complex-restricted, lymphokine-activated killer cell-like activity mediated by CD3-CD56+ effector cells. Simultaneously, the production of gamma-interferon and interleukin 1 decreased in these cultures. In contrast to TNF-alpha, anti-CD3 antibody synergized with IL2 to increase 2-3-fold TIL proliferation but not their cytotoxic activity against autologous tumor cell targets. These data suggest that TNF-alpha and IL2 synergize early in culture to induce tumor-reactive CD8+ effectors, some of which may be specific for autologous ovarian tumor cells. However, the conditions needed to sustain the specific autologous tumor responses in long-term cultures of human TIL remain to be determined.