Murine monoclonal antibodies cytotoxic to human glioma cells in vitro.

Six monoclonal antibodies (mABs) against human glioma cells (T2) were produced. T2 cells grown as solid tumors in nude mice, were dissociated and used to immunize Balb/c mice. After fusion of splenocytes with myeloma cells, eight hybrids secreting mABs were selected according to their ability to react immunohistochemically with T2 cells, but not with normal adult human brain. Cytotoxicity of mABs was tested using (3H)-thymidine incorporation assays in vitro. Four mABs showed complement-mediated cytotoxicity for T2 cells, other human glioma cells (T1), and a human melanoma cell line. Incubation with one antibody, mAb2A1, lowered (3H)-thymidine incorporation in the T2 and T1 cells to ca. 10%, and in melanoma cells to ca. 35% of control levels. Another antibody, mAb3B2, displayed a similar cytotoxicity for T2 and T1 cells, but did not show measurable cytotoxicity for melanoma cells and rat primary astrocyte cultures. Moreover, this antibody did not crossreact with haematopoietic cells from patients bearing CNS tumors or normal subjects. MAb3B2, therefore, appears to recognize and epitope associated to human gliomas, will be a useful glioma tumor marker and may have some potential therapeutical value.     

Identification of a structure on human melanoma cells recognized by CTL exhibiting anomalous killer cell function

The structures involved in the recognition of melanoma cells by nonspecific cytotoxic T lymphocytes (CTL) activated in mixed lymphocyte culture were investigated with monoclonal antibodies (MAb) which blocked this anomalous killer (AK) function. Of over 2000 MAb raised against melanoma cells, only three inhibited killing; one of these, an IgMk termed Leo Me13, was investigated in detail. In antibody-binding studies using a large range of cultured tumor cells, it was shown that Leo Me13 was relatively specific for melanoma cells. Of more importance, Leo Me13 inhibited conjugate formation between AK cells and melanoma target cells by 60 to 80% and caused an eight- to 10-fold reduction in killing. The MAb did not immunoprecipitate protein from melanoma cells surface-labeled with 125I, and thin-layer chromatography followed by immunoblotting of the separated glycolipids from melanoma cells indicated that the epitope was on acidic glycolipids migrating between GM1 and GD1a; moreover, treatment of melanoma cells with neuraminidase resulted in complete loss of binding of Leo Me13 but not of other anti-melanoma antibodies which did not inhibit AK cell-mediated lysis. Other melanoma-reactive MAb of the same isotype as Leo Me13 did not block killing of melanoma cells, but one documented antibody, R24, an IgG3 with specificity for the ganglioside GD3, was found to inhibit this function. These data suggest that the AK cells recognize and bind to melanoma cells by a secondary "lectin-type" receptor for a carbohydrate moiety.     

Monoclonal antibodies NORM-1 and NORM-2 induce more normal behavior of tumor cells in vitro and reduce tumor growth in vivo

In this study, large numbers of hybridomas (produced by syngeneic immunization with B16 mouse melanoma and fusion with NS-1 myeloma cells) were screened for the production of antibodies that affected morphology and growth of animal and human tumor cells in vitro. Two such antibodies, NORM-1 and NORM-2 (both IgG2a), inhibited the growth of B16 melanoma cells in soft agar and increased the serum requirements of tumor cells in tissue culture. Antibody NORM-2 also inhibited the growth of SV40-transformed 3T3 cells in agar and caused them to deposit more fibronectin into extracellular matrix. These antibodies thus seem to induce a more normal behavior of tumor cells in vitro. In vivo both antibodies reduced the number of growing lung tumors of B16 melanoma in C57BL/6 mice by 70%-90% when injected 3 days after the tumor cells. By immunoprecipitation of 35S-methionine-labeled cell extracts, NORM-2 antibody recognized a 59 kd protein in B16 mouse and in A375 human melanoma cells but not in 3T3 fibroblasts.     

Endogenous Tumor necrosis factor alpha unmasks the binding sites for N-acetylglucosaminyl-(beta1-4)-N-acetylmuramyl-alanyl-D-isoglutamine on the surface of melanoma cells

The surface of the melanoma BRO cells was shown to contain binding sites for N-acetylglucosaminyl-(beta 1-4)-N-acetylmuramyl-alanyl-D-isoglutamine (GMDP). Their number (1500 +/- 200 per cell) and affinity (Kd = 10 +/- +/- 1.2 nM) were determined. The occurrence of these sites was found to correlate with the ability of the melanoma cells to react in vitro with GMDP by increasing the expression of melanoma-associated antigens (MAA). An increased number of the GMDP binding sites (5200 +/- 500 per cell) was observed upon treating the melanoma BRO cells with tumor necrosis factor alpha (TNF-alpha). The mechanism of the TNF-alpha action most likely involves the unmasking of GMDP binding sites, initially expressed on the cell surface, by activating the endogenous protease that hydrolyzes surface proteins, in particular, highly glycosylated LAMP-2 protein exposed on the melanoma cell surface.     

WGA binding to the surface of two autologous human melanoma cell lines: different expression of sialyl and N-acetylglucosaminyl residues

Two autologous human melanoma cell lines were studied to determine their capacities to bind wheat germ agglutinin (WGA). Both cell lines were derived from the same patient, the first, IGR 39, originated from the primary tumor, the second, IGR 37, was established from a metastatic lymph node. WGA binding sites on the surface of these cell lines were compared before and after sialidase and/or tunicamycin treatments. IGR 39 cells exhibited two classes of WGA binding sites with high and low affinities, whereas IGR 37 cells had only one class of high affinity binding sites. After tunicamycin treatment, the capacity of IGR 39 cells to bind WGA was markedly altered, since only one class of WGA binding sites with high affinity was observed under these conditions, whereas tunicamycin did not induce significant changes in the lectin binding of IGR 37 cells. The low affinity WGA binding sites, which were only found on IGR 39 cells, corresponded to sialyl residues present in N-linked glycoproteins. The high affinity binding sites present on both cell lines probably involved sialyl and N-acetyl-glucosaminyl residues associated with O-linked glycoproteins and/or glycolipids. No direct correlation could be drawn between the number of WGA binding sites and the overall sialic acid levels exposed to sialidase treatment. The 3-fold increase in the amount of cell surface glycopeptides obtained after pronase digestion and specifically binding to WGA-Sepharose was in good agreement with the overall higher number of WGA binding sites on IGR 39 compared to IGR 37 cells. Thus, subtle carbohydrate changes of cell surface glycoconjugates might account for the differences between the biological properties of human melanoma cell lines of low and high tumorigenicity.     

Interactions between growth factor receptors and corresponding monoclonal antibodies in human tumors

Monoclonal antibodies (MAbs) to the human epidermal growth factor (EGF) receptor, the type I insulin-like growth factor (IGF) receptor, and the nerve growth factor (NGF) receptor were used to study the growth regulation of malignant cells. Anti-EGF receptor MAb 425 inhibited the growth of A 431 squamous carcinoma cells which express high numbers of EGF receptors on their surfaces. Growth inhibition induced by MAb 425 was accompanied by alterations of the cell-cycle distribution of these cells, indicating the ability of a monoclonal antibody to act as a biologically active ligand. Growth stimulation of melanoma cells by EGF was unrelated to EGF receptor expression on the cell surface. Insulin- and IGF-I-induced growth stimulation of melanoma cells was inhibited by MAb alpha IR-3 which reacts with the type I IGF receptor. This result indicates that the type I IGF receptor mediated growth stimulation not only by IGF-I but also by insulin. Normal melanocytes and cells of all stages of tumor progression expressed in tissue culture the receptor for NGF, but no effect on the growth of these cells has been observed.     

Localization of a tumor cell adhesion domain of laminin by a monoclonal antibody

Monoclonal antibodies were prepared to localize the domain(s) of laminin to which tumor cells adhere. Rat Y3-Ag 1.2.3 myeloma cells were fused with spleen cells from a rat immunized with a purified 440-kDa fragment of chymotrypsin-digested laminin. Three monoclonal antibodies (AL-1 to AL-3) that bound to intact laminin in a solid-phase radioimmunoassay were chosen for further analysis. The epitopes recognized by these antibodies were characterized by radioimmunoassays, immunoblotting, radioimmunoprecipitation, and immunoaffinity chromatography. In cell adhesion assays, monoclonal antibody AL-2 inhibited the binding of the highly metastatic melanoma cell line, K-1735-M4, to both intact laminin and the 440-kDa fragment of laminin. Electron microscopic examination of laminin-monoclonal antibody interactions showed that monoclonal antibody AL-2 reacted with the long arm of laminin directly below the cross-region. Two monoclonal antibodies that failed to inhibit tumor cell adhesion to laminin reacted with epitopes on the lateral short arms or cross-region of laminin as seen by electron microscopy. These results suggest that a new tumor cell binding domain of laminin may be located close to the cross-region on the long arm of laminin.     

Construction,expression and characterization of the engineered antibody against tumor surface antigen,p185^c—erbB—2

The c-erbB-2 proto-oncogene encodes a 185kD protein p185,which belongs to epidermal growth factor receptor family.Amplification of this gene has been shown to correlate with poor clinical prognosis for certain cancer patients.The monoclonal antibody A21 which directed against p185 specifically inhibits proliferation of tumor cells overexpressing p185,hence allows it to be a candidate for targeted therapy.In order to overcome several drawbacks of murine MAb,we cloned its VH and VL genes and constructed the single-chain FV(scFv)through a peptide linker.The recombinant scFv A21 was expressed in Escherichia coli and purified by the affinity column.Subsequently it was characterized by ELISA,Western blot,cell immunohistochemistry and FACS.All these assays showed the binding activity to extracellular domain(ECD)of p185.Based on those properties of scFvA21,we further constructed the scFv-Fc fusion molecule with a homodimer form and the recombinant product was expressed in mammalian cells.In a series of subsequent analysis this fusion protein showed identical antigen binding site and activity with the parent antibody.These anti-p185 engineered antibodies have promised to be further modified as a tumor targeting drugs,with a view of application in the diagnosis and treatment of human breast cancer.     

Development and characterization of human monoclonal antibodies that neutralize multiple TGFβ isoforms

Transforming growth factor (TGF)β levels are elevated in, and drive the progression of, numerous disease states such as advanced metastatic cancer and systemic and ocular fibrosis. There are 3 main isoforms, TGFβ1, 2, and 3. As multiple TGFβ isoforms are involved in disease processes, maximal therapeutic efficacy may require neutralization of 2 or more of the TGFβ isoforms. Fully human antibody phage display libraries were used to discover a number of antibodies that bind and neutralize various combinations of TGFβ1, 2 or 3. The primary panning did not yield any uniformly potent pan-isoform neutralizing antibodies; therefore, an antibody that displayed potent TGFβ 1, 2 inhibition, but more modest affinity versus TGFβ3, was affinity matured by shuffling with a light chain sub-library and further screening. This process yielded a high affinity pan-isoform neutralizing clone. Antibodies were analyzed and compared by binding affinity, as well as receptor and epitope competition by surface plasmon resonance methods. The antibodies were also shown to neutralize TGFβ effects in vitro in 3 assays: 1) interleukin (IL)-4 induced HT-2 cell proliferation; 2) TGFβ-mediated IL-11 release by A549 cells; and 3) decreasing SMAD2 phosphorylation in Detroit 562 cells. The antibodies’ potency in these in vitro assays correlated well with their isoform-specific affinities. Furthermore, the ability of the affinity-matured clone to decrease tumor burden in a Detroit 562 xenograft study was superior to that of the parent clone. This affinity-matured antibody acts as a very potent inhibitor of all 3 main isoforms of TGFβ and may have utility for therapeutic intervention in human disease.     

Isolation of a tumor cell laminin receptor

BL6 murine melanoma cells contain approximately 110,000 cell surface binding sites for the basement membrane glycoprotein laminin. Treatment of isolated melanoma cell plasma membranes with detergent yields a single class of laminin receptor. The receptor was purified 900 fold by laminin affinity chromatography. The isolated receptor has a Mr of 67,000 and binds laminin with high affinity: kd = 2 nm. The binding affinity of the isolated receptor was similar to that of the plasma membranes or the whole cells. Such a laminin receptor, isolated here for the first time, could facilitate the interaction of metastasizing tumor cells with the basement membrane.     

Generation of cell surface-bound plasmin by cell-associated urokinase-type or secreted tissue-type plasminogen activator: a key event in melanoma cell invasiveness in vitro.

Recently, we have shown that plasminogen activators (PAs) of both types, urokinase-type (uPA) as well as tissue-type (tPA), are involved in the in vitro invasiveness of human melanoma cells. The present study is focused on the generation and importance of cell surface-bound plasmin in this process. The human melanoma cell lines MelJuso and MeWo expressed plasminogen binding sites on the cell surface. Plasminogen binding was saturable and not species-specific, since human and bovine plasminogen bound to the cells with comparable efficiency. The activation of the proenzyme plasminogen bound on MelJuso cells, which expressed surface-associated uPA activity, occurred almost synchronously with binding to the cell surface. Removal of cell-associated uPA considerably reduced plasmin generation on these cells. In contrast, plasminogen activation on MeWo cells, which secreted tPA into the culture supernatant and which were devoid of surface-associated PA activity, was by far less effective. The efficiency of the activation process could be increased by addition of exogenous tPA. With both cell lines, plasmin generation on the cell surface was suppressed by inhibitory monoclonal antibodies specific for the respective PA type. Selective inhibition of cell surface-associated plasmin by preincubating the cells with an inhibitory monoclonal antibody or with aprotinin, as well as removal of plasmin from the cell surface, led to a significant decrease in cellular invasiveness of both cell lines into various biological substrates such as fibrin gel, the basement membrane extract Matrigel, or intact extracellular matrix. Both cell lines were able to penetrate an intact cell layer of the human keratinocyte line HaCaT, a process, which also proved to be dependent on cell-associated plasmin. In conclusion, these data provide evidence that plasminogen activation associated with the surface of human melanoma cells is catalyzed much more efficiently by cell-associated uPA (MelJuso) than by secreted tPA (MeWo). Cell-associated plasmin, which is protected from inactivation by serum inhibitors, represents the essential component of the proteolytic cascade of plasminogen activation during in vitro invasiveness of human melanoma cells.     

Characterization of an immunodominant antigenic site on GB virus C glycoprotein E2 that is involved in cell binding

GB virus type C (GBV-C) is a human flavivirus that may cause persistent infection, although most infected individuals clear viremia and develop antibodies to the envelope glycoprotein E2. To study GBV-C E2 antigenicity and cell binding, murine anti-E2 monoclonal antibodies (MAbs) were evaluated to topologically map immunogenic sites on GBV-C E2 and for the ability to detect or block recombinant E2 binding to various cell lines. Five competition groups of MAbs were identified. Groups I and II did not compete with each other. Group III competed with both groups I and II. Group IV did not compete with group I, II, or III. One MAb competed with all of the other MAbs, suggesting that the epitopes bound by these MAbs are intimately related. Individually, none of the MAbs competed extensively with polyclonal human convalescent antibody (PcAb); however, combinations of all five MAb groups completely blocked PcAb binding to E2, suggesting that the epitopes bound by these MAbs form a single, immunodominant antigenic site. Only group I and III MAbs detected purified recombinant E2 bound to cells in binding assays. In contrast, group II MAbs neutralized the binding of E2 to cells. Both PcAb and MAbs were conformation dependent, with the exception of one group II MAb (M6). M6 bound to a five-amino-acid sequence on E2 if the peptide included four C-terminal or eight N-terminal residues, suggesting that the GBV-C E2 protein contains a single immunodominant antigenic site which includes a complex epitope that is involved in specific cellular binding.     

Tumor cell killing by in vitro affinity-matured recombinant human monoclonal antibodies

We have developed a method that allows the rapid improvement of the affinity of phage-displayed antibody fragments by selection on intact eukaryotic cells. A single chain Fv fragment, specific for the tumor-associated Ep-Cam molecule, was mutagenized by shuffling of the immunoglobulin light chain variable region and DNA shuffling of both heavy and light chain variable regions. Higher-affinity mutants were selected from small phage display libraries by cell panning under stringent conditions. When converted to an intact fully human antibody, the mutagenized anti-tumor monoclonal antibody displayed an affinity of 0.4 nM, a 15-fold improvement over the affinity of the original antibody. Compared to previously reported affinity maturation schemes, panning on intact cells does not require purified targets for selection and may be particularly useful when the target molecule can not be expressed as a recombinant molecule or easily purified without disrupting its native configuration. In vitro tumor cell killing assays demonstrated an improved performance of the higher-affinity antibody in complement-mediated tumor cell killing. In contrast, the lower-affinity antibody performed somewhat better in antibody-dependent cellular cytotoxicity assays and penetrated better in multicell spheroids of tumor cells, an in vitro model for the tumor penetration capacity of antibodies. Received: 26 February 2000 / Accepted: 26 January 2001     

Recognition of human activated CD44 by tumor vasculature-targeted antibody

TES-23 monoclonal antibody (MAb), which targets rat CD44H on tumor vascular endothelial cells (TEC), dominantly reacted to human activated CD44 rather than human inactive CD44. TES-23 MAb reacted to HT-1080 fibrosarcoma cells almost comparably to anti-human CD44 MAb and moderately to HUVEC; however, it hardly reacted to PBMC. The binding of soluble hyaluronate to HT-1080 cells and HUVEC was clearly noted, but not to PBMC. In addition, stimulation with phorbol 12-myristate 13-acetate induced soluble hyaluronate binding of MOLT-4 human T lymphoma cells and relatively increased the reactivity of TES-23 MAb. Our results suggest that TES-23 MAb can potentially recognize human activated CD44 and hence might be potentially useful for the treatment of human solid tumors containing TEC that express activated CD44.     

A 3-dimensional tumor growth inhibition assay for testing monoclonal antibody cytotoxicity

Summary A 3-dimensional tumor growth inhibition assay [18] has been adapted to test the cytotoxic activity of a panel of monoclonal antibodies directed to various antigenic determinants on the surface of mouse mammary tumor cells. Target cells can be prepared from either cultured cells or from pieces of fresh tumor. Antibody and complement are added when cells are growing actively and cell growth can be measured, non destructively, over a 7–10-day period. Effective diffusion of antibody through collagen gel and binding to target cells embedded in the gel is demonstrated by indirect immunofluorescent staining. The specificity of monoclonal antibody AMT 101 cytotoxicity for mouse mammary tumor cells is the same in trypan blue exclusion assays of single-cell suspensions as in collagen gel assays, with complete killing seen in the collagen gel assay only. The collagen gel assay allows the testing of repeated treatments in vitro, as well as combined treatment with multiple antibodies. It also allows cell-cell interaction and preserves all cell components in the tumor. The collagen gel assay has potential as a method of predicting the outcome of monoclonal antibody treatment of solid tumors.     

Selective macrophage activation by muramyldipeptide bound to monoclonal antibodies specific for mouse tumor cells

Summary IgM monoclonal antibodies directed against tumor cells which do not mediate antibody-dependent macrophage cytotoxicity (ADMC) even when they are cytotoxic in the presence of complement, have been shown to render macrophages tumoricidal when they carry an immunomodulating agent, i.e., muramyldipeptide (MDP).This statement is based on experiments using two IgM monoclonal antibodies selected for their ability to bind L1210 leukemia cells (F₂-10-23-IgM) and 3LL Lewis lung carcinoma cells (6B6-IgM) specifically, as shown by flow cytofluorometry analysis.The MDP-IgM conjugates, containing 45 MDP molecules per IgM molecule, were prepared by allowing MDP-hydroxy-succinimide ester to react with IgM monoclonal antibodies.The MDP-IgM conjugates are shown to bind to relevant tumor cells and to induce the activation of thioglycolate-elicited peritoneal mouse macrophages leading to 80% growth inhibition of target cells at optimum concentrations of bound MDP. These concentrations of bound MDP were 10 times lower than the concentration of free MDP, giving a maximum activation that is limited to 20% growth inhibition.No macrophage activation was evidenced when tumor cells were incubated in the presence of irrelevant MDP-IgM conjugates and macrophages or when macrophages were preincubated in the presence of MDP-IgM conjugates and then incubated in the presence of relevant or irrelevant tumor cells but in the absence of the MDP-IgM conjugates.The reported results are discussed with reference to the mechanism of activation of macrophage by muramyldipeptide and to the usefulness of such MDP-IgM conjugates as potential antitumor agents in cancer therapy.Abbreviations ADMC antibody dependent macrophage mediated cytotoxicity - F-GAM fluoresceinylthiocarbamyl goat anti mouse antibody - -Man-BSA -mannopyranosyl-phenylthiocarbamyl bovine serum albumin containing some 25 mannose residues (neoglycoprotein) - MDP muramyldipeptide, 2-acetamido-3(2-0-d-lactyl-l-alanyl-d-glutamyl amine) glucopyranose - MDP-F₂-10-23-IgM Murine monoclonal antibody specific of L1210 leukemia cells and substituted with 45 MDP molecules - MDP—6B6-IgM Murine monoclonal antibody specific of 3LL Lewis lung carcinoma cells and substituted with 45 MDP molecules - MEM minimal essential medium - TDM Trehalose-dimycolate     

Topographical separation of the catalytic sites of asparagine synthetase explored with monoclonal antibodies

Monoclonal antibodies which inhibited the enzymatic activity of bovine pancreatic asparagine synthetase were mapped to two topographically separate regions of the enzyme surface using competitive binding assays. Three antibodies which all inhibited glutamine- and NH3-dependent synthesis of asparagine bound to a common antigenic region. A fourth monoclonal antibody which interfered with glutamine binding or cleavage but not with NH3-dependent synthesis of asparagine was mapped to a separate region of the enzyme surface. These findings suggest a topographical separation between the aspartyl-AMP and glutamine-binding sites of bovine pancreatic asparagine synthetase. Three noninhibitory antibodies exhibited conformation-dependent binding and were mapped to a third region of the enzyme. Binding assays were used to demonstrate extensive cross-reaction of these antibodies with asparagine synthetases isolated from bovine liver and sheep pancreas. Substantial cross-reactions were also seen with the enzyme isolated from rat liver or pancreas, a human tumor cell line, and a mouse tumor cell line. Of the four antibodies that inhibited glutamine- and NH3-dependent synthesis of asparagine from ruminant species, only one bound to and inhibited the enzyme from rat liver or mouse cells, which suggests significant structural differences between the ruminant and rodent enzymes.     

Growth inhibition of tumor cells in vitro by using monoclonal antibodies against gonadotropin-releasing hormone receptor

As the continuation of a previous study, synthetic peptides corresponding to the extracellular domains of human gonadotropin-releasing hormone (GnRH) receptor were used to generate additional monoclonal antibodies which were further characterized biochemically and immunologically. Among those identified to recognize GnRH receptor, monoclonal antibodies designated as GHR-103, GHR-106 and GHR-114 were found to exhibit high affinity (Kd ≤ 1 × 10⁻⁸ M) and specificity to GnRH receptor as judged by the whole cell binding immunoassay and Western blot assay. Both anti-GnRH receptor monoclonal antibodies and GnRH were shown to compete for the same binding site of GnRH receptor on the surface of cultured cancer cells. Growth inhibitions of cancer cells cultured in vitro were demonstrated by cellular apoptosis experiments (TUNEL and MTT assays) under different conditions of treatment with GHR-106 monoclonal antibody or GnRH analogs. It was generally observed that both GnRH I and GHR-106 effectively induce the apoptosis of cultured cancer cells as determined by TUNEL and MTT assays. Consistently, suppressions of gene expressions at mRNA levels were demonstrated with several ribosomal proteins (P0, P1, P2 and L37), when cancer cells were incubated with GnRH or GHR-106. The widespread expressions of GnRH receptor in almost all of the studied human cancer cell lines were also demonstrated by RT-PCR and Western blot assay, as well as indirect immunofluorescence assay with either of these monoclonal antibodies as the primary antibody. In view of the longer half life of antibodies as compared to that of GnRH or its analogs, anti-GnRH receptor monoclonal antibodies in humanized forms could function as GnRH analogs and serve as an ideal candidate of anti-cancer drugs for therapeutic treatments of various cancers in humans as well as for fertility regulations.     

Monoclonal antibodies to endogenous galactose-specific tumor cell lectins.

A monoclonal antibody, 5D7, was obtained after immunization of syngeneic mice with B16 melanoma cell extracts enriched for endogenous lectin activity and screening for inhibition of lectin-mediated hemagglutination. Binding of this antibody to affinity-purified B16 melanoma galactoside-specific lectin was revealed by solid-phase radioimmunoassay and binding to the surface of viable B16 cells was demonstrated by indirect immunofluorescence. Inhibition of lectin activity and cell surface labeling by 5D7 antibody were also found with several types of cultured human and murine cells including melanoma, sarcoma and carcinoma. This monoclonal antibody should be useful for evaluating the role of tumor cell surface lectins in intercellular interactions and metastasis.     

Phage-selected primate antibodies fused to superantigens for immunotherapy of malignant melanoma

 The high-molecular-weight melanoma-associated antigen, HMW-MAA, has been demonstrated to be of potential interest for diagnosis and treatment of malignant melanoma. Murine monoclonal antibodies (mAb) generated in response to different epitopes of this cell-surface molecule efficiently localise to metastatic lesions in patients with disseminated disease. In this work, phage-display-driven selection for melanoma-reactive antibodies generated HMW-MAA specificities capable of targeting bacterial superantigens (SAg) and cytotoxic T cells to melanoma cells. Cynomolgus monkeys were immunised with a crude suspension of metastatic melanoma. A strong serological response towards HMW-MAA demonstrated its role as an immunodominant molecule in the primate. Several clones producing monoclonal scFv antibody fragments that react with HMW-MAA were identified using melanoma cells and tissue sections for phage selection of a recombinant antibody phage library generated from lymph node mRNA. One of these scFv fragments, K305, was transferred and expressed as a Fab-SAg fusion protein and evaluated as the tumour-targeting moiety for superantigen-based immunotherapy. It binds with high affinity to a unique human-specific epitope on the HMW-MAA, and demonstrates more restricted crossreactivity with normal smooth-muscle cells than previously described murine mAb. The K305 Fab was fused to the superantigen staphylococcal enterotoxin A (D227A) [SEA(D227A)], which had been mutated to reduce its intrinsic MHC class II binding affinity, and the fusion protein was used to demonstrate redirection of T cell cytotoxicity to melanoma cells in vitro. In mice with severe combined immunodeficiency, carrying human melanoma tumours, engraftment of human lymphoid cells followed by treatment with the K305Fab-SEA(D227A) fusion protein, induced HMW-MAA-specific tumour growth reduction. The phage-selected K305 antibody demonstrated high-affinity binding and selectivity, supporting its use for tumour therapy in conjunction with T-cell-activating superantigens. Received: 9 September 1999 / Accepted: 21 October 1999