(2004) 18, 636–644 & 2004 Nature Publishing Group All rights reserved 0887-6924/04 $25.00 www.nature.com/leu A novel recombinant bispecific single-chain , bscWue-1 Â CD3, induces T-cell-mediated cytotoxicity towards human cells

DHo¨nemann1,3,4, P Kufer2,4, MM Rimpler1,3,4, M Chatterjee1,3,4, S Friedl3, F Riecher1, K Bommert1,3,BDo¨rken1,3 and RC Bargou1,3

1Department of , and Tumor-Immunology, Helios Clinics, Robert-Ro¨ssle Cancer Center, University Medical Center Charite´, Berlin, Germany; 2Institute of Immunology, University of Munich, Munich, Germany; and 3Max- Delbru¨ck Center for Molecular Medicine, Berlin, Germany

The development of antibody-based strategies for the treatment the treatment of minimal residual disease after high-dose of multiple myeloma (MM) has been hampered so far by the fact .5,6 Another immunotherapeutic strategy is the that suitable -specific surface have been missing. However, recently a novel , application of monoclonal that have been shown to designated Wue-1, has been generated that specifically recog- be efficacious in various including malignant 7–9 nizes normal and malignant human plasma cells. Therefore, . However, in contrast to other B-cell non- Wue-1 is an interesting and promising candidate to develop Hodgkin lymphomas, antibody-based therapies in MM are not novel immunotherapeutic strategies for the treatment of MM. established. One main reason for this is the fact that, so far, One variant for an antibody-based strategy is the bispecific suitable plasma cell-specific surface antigens have been antibody approach. Recombinant bispecific single-chain (bsc) 1,10–12 antibodies are especially interesting candidates because they missing. Although normal and malignant plasma cells show exceptional biological properties. We have generated a express a number of well characterized surface markers, none novel MM-directed recombinant bsc antibody, bscWue-1 Â CD3, have turned out to be plasma cell specific: for example, CD38 is and analyzed the biological properties of this antibody using an activation rather than a differentiation-associated the MM cell line NCI-H929 and primary cells from the bone lacking lineage restriction;13 CD56 is an N-CAM splice variant marrow of patients with MM. We were able to show that also expressed on NK-cells;14 CD138 (syndecan-1) is also bscWue-1 Â CD3 induces efficient and selective T-cell-mediated 15 cell death of NCI-H929 cells and primary myeloma cells in nine expressed on epithelia, and a recently described plasma cell- out of 11 cases. The bscWue-1 Â CD3 Ab is efficacious even at associated antigen HM1.24 has turned out to be also expressed 16 low E:T ratios, and with or without additional T-cell pre- or by other cell types such as stromal cells. The costimulation. Target cell lyses were specific for Wue-1 expression of Muc-1 core was reported to be expressed antigen-positive cells and could be blocked by the Wue-1 only on subsets of MM cells.11 Moreover, typical B-cell lineage monoclonal antibody. markers such as CD19 or CD20 are not, or only poorly, Leukemia (2004) 18, 636–644. doi:10.1038/sj.leu.2403264 1,4 Published online 22 January 2004 expressed by MM cells. Keywords: multiple myeloma; immunotherapy; singlechain Recently, a novel monoclonal antibody, designated Wue-1, bispecific antibody; bsp Wue-1 Â CD3; Wue-1 was generated that specifically binds to normal and malignant human plasma cells. When analyzing various tissues by immunohistochemistry, Wue-1 mAb was not reactive with all normal tissues tested. Strong staining could be detected Introduction exclusively in normal plasma cells and in all cases of MM analyzed.12 Expression profile and biochemical characteristics Multiple myeloma (MM) is a malignant disorder of the discriminate the Wue-1 antigen from other plasma cell- terminally differentiated (for instance, plasma cell) mainly associated antigens described so far,12 although the molecular localized in the bone marrow. The disorder is frequently identity is still unclear and therefore under current investigation. accompanied by monoclonal (M) protein production and either As of the highly selective expression pattern, Wue-1 could be a diffuse osteoporosis or lytic bone lesions. Common complica- promising candidate to develop novel immunotherapeutic tions of overt MM include recurrent bacterial infections, , strategies for the treatment of MM. 1 bone destruction, hypercalcemia, and renal failure. An interesting variant of an antibody-based strategy are Despite some progress with high-dose chemotherapy and bispecific antibodies. By joining two different antigen-binding autologous /bone marrow transplantation, MM is still sites in one antibody molecule, redirection of effector cells to a an incurable disease with a median survival rate of about 3–5 predefined tumor target can be achieved. Although bispecific 1–4 years. This bleak situation has stimulated the search for antibodies are efficient in recruiting cytotoxic effector cells alternative therapeutic strategies among which immunothera- against tumor cells in vitro and in vivo,17–29 large randomized peutic strategies play an increasing role. Tumor vaccination clinical trials as a final test of therapeutic efficacy are still strategies have been developed and shown to activate idiotype- lacking. The main reason for this are difficulties in producing specific T-cell clones in MM patients in different settings. sufficient amounts of clinical grade material. The production The therapeutic significance of this observation is currently and purification processes are inefficient, with low yields being investigated in clinical trials, particularly as an option for regardless of whether a hybrid–hybridoma approach, chemical linkage, or renaturation from bacterial inclusion bodies of Correspondence: RC Bargou, Department of Hematology, Oncology recombinant Fab or Fv fragments was followed. In addition, and Tumor-Immunology, Helios Clinics, Robert-Ro¨ssle Cancer Center, nearly all of these products are plagued by contamination with ´ University Medical Center Charite, Lindenberger Weg 80, 13125 ill-defined by-products.26,27,30–33 As we have previously shown Berlin, Germany; Fax: þ 49 30 9417 1398 4The first four authors contributed equally to this work with two different bispecific single-chain (bsc) antibodies Received 10 September 2003; accepted 17 November 2003; (EpCAM Â CD3 and CD19 Â CD3), these handicaps can be Published online 22 January 2004 overcome with a molecular format linking the variable regions bscWue-1 Â CD3 in the treatment of MM DHo¨nemann et al 637 of two antibodies in the configuration V light-chain (VL)1–V anti-CD3 scFv fragment. Subsequently, the DNA fragment 34–36 heavy-chain (VH)1–VH2–VL2 on one peptide chain. The encoding the bsc antibody Wue-1 Â CD3 with the domain resulting recombinant molecule produced by mammalian cells arrangement VLWue-1–VHWue-1–VHCD3–VLCD3 was sub- is secreted at high yields in a fully active form that requires no cloned into the EcoRI/SalI sites of the expression vector pEF- further renaturation. Another advantage of these recombinant DHFR.34 The resulting plasmid DNA was transfected into bsc antibodies compared to other antibody formats is their DHFR-deficient Chinese hamster ovary cells by electroporation. exceptionally strong biological activity.37,38 Moreover, the small Selection and gene amplification were performed as described size is expected to improve the penetration into malignant previously.34 tissues, thus further increasing the potential therapeutic effi- 39 cacy. Owing to the promising biological properties of bsc 0 0 antibodies, the very same molecular format was chosen to List of primers (5 –3 , IUPAC code) produce a recombinant antibody construct directed against the 0 Wue-1 and the CD3 antigen (bscWue-1 Â CD3) in pursuit of a 5 VLWue-1-L1: GAAGCACGCGTAGATATCKTGMTSACC- 0 future therapeutic trial in patients with MM. To our knowledge, CAAWCTCCA; 3 VLWue-1-K: GAAGATGGATCCAGCGGCCG- 0 the bscWue-1 Â CD3 molecule presented here is the first MM CAGCATCAGC; 5 VHWue-1-H4: CCGGCCATGGCGCAGGTG 0 cell-directed bispecific antibody described so far. CAGCTGCAGCAGTCTGG; 3 VHWue-1-G: ACCAGGGGCCAG 0 TGGATAGACAAGCTTGGGTGTCGTTTT; 5 VLB5RRV: AGGTG 0 TACACTCCATATCCAGCTGACCCAGTCTCCA; 3 VLGS15: GG Materials and methods AGCCGCCGCCGCCAGAACCACCACCTTTGATCTCGAGCTTG 0 GTCCC; 5 VHGS15: GGCGGCGGCGGCTCCGGTGGTGGT 0 Cell lines and chemicals GGTTCTCAGGTSMARCTGCAGSAGTCWGG; 3 VHBspEI: AAT CCGGAGGAGACGGTGACCGTGGTCCCTTGGCCCCAG. NCI-H929 and NALM-6 cells were obtained from the ‘Deutsche Sammlung von Mikroorganismen und Zellkulturen’ (DSMZ, Protein purification Braunschweig, Germany). Cell lines were cultured in complete RPMI 1640 (Biochrom, Berlin, Germany) with 10% fetal calf The bsc antibody was purified from the cell-culture supernatant serum (FCS) (GIBCO, Karlsruhe, Germany). All chemicals were in a three-step process including ion exchange chromatography, of research grade and obtained from Sigma, Deisenhofen, immobilized metal affinity chromatography via a C-terminal Germany or Merck, Darmstadt, Germany. 6 Â Histidin-Tag and gel filtration. Protein concentrations were determined using the Bio-Rad protein assay with IgG (Bio-Rad, Mu¨nchen, Germany) as Cloning of variable immunoglobulin domains standard protein. The Aekta FPLC System (Pharmacia, Stockholm, Sweden) and 12 The VL and VH domains from the Wue-1 hybridoma were Unicorn Software were used for chromatography. All chemicals cloned according to standard PCR methods as described were of research grade and purchased from Sigma (Deisenho- before.40 cDNA synthesis was carried out with oligo-dT primers fen, Germany) or Merck (Darmstadt, Germany). and reverse transcriptase. For the amplification of the V domains by PCR, we used degenerate primers based on the description of SDS-PAGE analysis Du¨bel et al.41 The resulting fragments were cloned into the pGEM-T plasmid (Promega, Mannheim, Germany) and se- SDS-PAGE under reducing conditions was performed with quenced. The DNA of the anti-CD3 scFv fragment was provided precast 4–12% Bis/Tris gels (NOVEX). The preparation and 42 by Traunecker. application of samples was according to the manufacturer’s protocol. The gel was stained with colloidal Coomassie. The molecular weight was determined with MultiMark protein Construction of bsc fragments standard (NOVEX/Invitrogen, Karlsruhe, Germany).

To obtain the anti-Wue-1 scFv fragment, the corresponding VL Immunocytochemistry and VH regions were cloned into separate plasmid vectors as templates for a VL- and VH-specific PCR using the oligonucleo- 0 0 0 tide primer pairs 5 VLB5RRV/3 VLGS15 and 5 VHGS15/ (1) Patients and specimen: Normal and tumor biopsy tissues 0 1 3 VHBspEI, respectively. Overlapping complementary se- were kept at À70 C as snap-frozen blocks until sections quences were introduced into the PCR products that combined were prepared at the time of the experiments. Tumor to the coding sequence of a 15-amino-acid glycine–serine samples were selected from the German lymph node (G4S1)3 linker during the subsequent fusion PCR. This amplifica- registry of the Institute of Pathology in Wu¨rzburg, and 0 43 tion step was performed with the primer pair 5 VLB5RRV/ classified according to WHO classification. For diagnostic 0 3 VHBspEI, and the resulting product (encoding the anti-Wue-1 purposes, the morphological and immunophenotypical scFv fragment) was directly cloned into the TA-cloning vector analyses of paraffin-embedded and fresh-frozen sections pGEM-T (Promega, Mannheim, Germany). The anti-Wue-1 were performed by standard methods.12 fragment was excised from the pGEM-T vector by partial (2) Immunostaining: Immunostaining was performed on 4-mm digestion with EcoRV and complete digestion with BspEI and sections or on cytospins. The immunoperoxidase method cloned EcoRV/BspEI into the bluescript KS vector (Stratagene, La was applied by a three-step incubation procedure with Jolla, USA) containing the (EcoRI/SalI-cloned) coding sequence diluted affinity-purified anti-mouse antibodies (Dako, Ham- of the bsc antibody EpCAM Â CD3.34 Thereby, the anti-EpCAM burg, Germany) as described in detail elsewhere.12,44 specificity was replaced by the anti-Wue-1 fragment, preserving BscWue-1 Â CD3 Ab was detected by an anti-His-Tag the 5-amino-acid (G4S1) linker that connects to the C-terminal antibody (Dianova, Hamburg, Germany).

Leukemia bscWue-1 Â CD3 in the treatment of MM DHo¨nemann et al 638 FACS analysis of Wue-1 and CD19 expression cell-membrane-dye PKH-26 (Sigma, Deisenhofen, Germany), according to the manufacturer’s protocol. Cells were incubated A total of 5 Â 105 cells were washed with phosphate-buffered- in 96- or 24-well flat-bottom culture plates (Nunc, Wiesbaden, saline (PBS) and resuspended in 100 ml PBS with 10% Germany) in RPMI/10% FCS at indicated effector:target ratios. Venimmun (Aventis Behring, Mu¨nchen, Germany) and 0.1% The experiments with enriched primary myeloma and primary NaN3. For the analysis of Wue-1 antigen, cells were incubated CLL cells were performed in the presence of PHA (10 ng/ml) and with Wue-1-mouse for 30 min. Mouse IgG1 was used as IL-2 (60 U/ml). No additional stimulation of T cells with PHA or isotype control. After washing, cells were incubated with a IL-2 was performed in the experiments with cell lines. After 48 h phycoerythrin (PE)-labeled goat-anti-mouse F(ab2) fragment (cell lines) or 4–6 days (primary MM and CLL samples), cells (Dianova, Hamburg, Germany) for 30 min. For the analysis of were washed twice with PBS and stained with propidium iodide CD19 surface expression, cells were incubated with a PE- (PI) (2.5 mg/ml). Analysis was performed on a ‘FACScalibur’ flow labeled anti-CD19 mAb (Biosource, Solingen, Germany) for cytometer using ‘Cell-Quest’ software (Becton Dickinson, 30 min. Mouse IgG1-PE was used as isotype control. Flow Heidelberg, Germany) by counting 5–10 Â 104 cells. Specific cytometry analysis was performed in a ‘FACScalibur’ flow cytotoxicity was calculated by comparing the relative number of cytometer, using ‘Cell-Quest’ software (Becton Dickinson, PI-negative/PKH-positive cells (viable target cell fraction) in the Heidelberg, Germany). bscWue-1 Â CD3 samples with the viable target cell fraction in the bscEpCAM Â CD3 Ab control samples using the following formula: 100 Â (1À(number of viable target cell fraction in the Enrichment of myeloma cells from bone marrow bscWue-1 Â CD3 sample)/(number of viable target fraction in aspirates of MM patients the bscEpCAM Â CD3 control)). The same formula was used to calculate specific cytotoxicity against the CLL cells induced by Mononuclear cells from bone marrow aspirates of MM patients the bscCD19 Â CD3 antibody. To rule out that the decrease of were separated by Ficoll density-gradient centrifugation, in- the fraction of target cells was due to the proliferation of T cells cubated for 15 min with anti-CD138-coated paramagnetic during the assay time, a correlation to absolute cell counts was beads at 41C, and separated in a magnet (Miltenyi Biotec, carried out by two methods: counting viable cells in the Bergisch Gladbach, Germany) according to the manufacturer’s Neubauer chamber under the microscope at the beginning and protocol. To estimate the efficiency and specificity of the at the end of the incubation period or by FACS analysis using isolation procedure, cytospins of the enriched cells were CaliBRITE beads (Becton Dickinson, San Jose, USA). With both analyzed by morphology after Pappenheim staining. The purity methods no increase in absolute cell number could be detected, of the myeloma cell preparations was 490% in all experiments. indicating that T-cell proliferation is not responsible for the Patient cells were taken from routine diagnostic specimens after decrease of the fraction of target cells (error o715%). Since no informed consent of the patients and approval by the local ethics cell aggregates could be detected morphologically or in the committee. FACS analysis, it is unlikely that the bscWue-1 Â CD3 interferes with the ability to count single cells. Enrichment of primary CLL cells from the peripheral blood of CLL patients Cytotoxicity assays with the mononuclear cell fraction from bone marrow aspirates of MM patients: Mono- Primary CLL cells were enriched by Ficoll density-gradient nuclear cells from bone marrow aspirates of MM patients centrifugation of peripheral blood from two different CLL (including T cells and MM cells from the same patient) were patients with leukocyte counts 4150 000/ml containing over separated by Ficoll density-gradient centrifugation. The T-cell to 95% CD19 þ . No further processing was per- myeloma cell ratio was determined by flow cytometry using formed. CD3 as the T-cell marker and CD138 as the myeloma cell marker. Cells were incubated in 24-well flat-bottom culture plates (Nunc, Wiesbaden, Germany) in RPMI/10% FCS. No Enrichment of PBMC and T cells from buffy coats additional stimulation of T cells with PHA or IL-2 was performed. After 4 to 6 days, cells were washed twice with Human PBMC were isolated from fresh buffy coats of random PBS and stained with PI (2.5 mg/ml) and FITC-labeled anti- donors by density-gradient centrifugation. Monocytes and CD138-mAb (Biosource, Solingen, Germany). FACS analysis macrophages were depleted by adhesion to cell-culture flasks was performed as described above. Specific cytotoxicity was for 2 h. CD19 þ B cells were depleted from the remaining cell calculated by comparing the number of viable target cells in the suspension using anti-CD19-coated paramagnetic beads (Dynal, respective samples with the bscEpCAM Â CD3 Ab control Hamburg, Germany), according to the manufacturer’s protocol. sample using the same formula as described above. This procedure resulted in a T-cell population of 490% purity, as determined by flow-cytometry after staining CD3 (data not shown). In the experiments with primary myeloma cells, the Results depletion of CD19 þ cells was omitted. bscWue-1 Â CD3 Ab binds to Wue-1 antigen-positive myeloma cells Cytotoxicity assays The specific expression of the Wue-1 antigen on normal and Cytotoxicity assays using myeloma cell lines, enriched malignant plasma cells has been demonstrated in a previous primary myeloma cells from bone marrow aspirates study.12 Within malignant disorders of the lymphohematopoie- and CLL cells from the peripheral blood as target cells tic system, the Wue-1 antigen is exclusively expressed by MM (PKH assay): Cytotoxicity assays were performed as de- and immunocytoma cells, reflecting its association with terminal scribed elsewhere.36 Briefly, target cells were labeled with the B-cell differentiation.12 Here, the expression of the Wue-1

Leukemia bscWue-1 Â CD3 in the treatment of MM DHo¨nemann et al 639 antigen was studied in a larger number of MM cases to confirm bscWue-1 Â CD3 Ab, and performed an experiment with the previous findings. Bone marrow biopsies from 51 MM enriched effector T cells of healthy human donors and a mixed patients were analyzed by immunohistochemistry. The Wue-1 target cell population consisting of NCI-H929 and NALM-6 antigen was found to be strongly expressed in 50/51 MM cases cells. This mixed target cell population was used in order to rule analyzed (see, for example, Greiner et al ).12 out any ‘bystander’ cytotoxicity induced by the bscWue- This selective expression pattern prompted us to develop a 1 Â CD3 Ab by demonstrating that even in the presence of new bsc antibody directed against the Wue-1 antigen. activated T cells cytotoxicity is strictly dependent on antibody- The bscWue-1 Â CD3 Ab was cloned, expressed, and purified mediated interaction of effector and target cells. Therefore, we as described before for similar constructs.34,37 SDS-PAGE labeled either NCI-H929 or NALM-6 cells with the membrane- analysis revealed a purity of 495% and an apparent molecular dye PKH-26. The 1:1 mixture of labeled and unlabeled target weight of approximately 54 kDa for monomeric bscWue- cells was incubated with unlabeled effector T cells (E:T ratio 1 Â CD3 (Figure 1c). By immunocytochemistry, we could 10:1) and with the bscWue-1 Â CD3 Ab, bscEpCAM Â CD3 Ab demonstrate the binding of the bscWue-1 Â CD3 Ab to Wue- or bscCD19 Â CD3 Ab. All three antibodies share the same 1-positive NCI-H929 MM cells (Figure 1a), whereas the molecular format but recognize the EpCAM antigen (mainly bscCD19 Â CD3 antibody showed no binding to NCI-H929 found on epithelial tissues)34, the CD19 antigen (specifically cells (Figure 1b). expressed on normal B cells and B- cells37), or the Wue-1 antigen (specifically expressed on plasma and MM cells), while sharing an identical anti-CD3-binding site. Upon incuba- bscWue-1 Â CD3 Ab induces specific lysis of the Wue-1 tion with the bscWue-1 Â CD3 Ab, we observed specific lysis of antigen-positive myeloma cell line NCI-H929 but not of NCI-H929 cells only, whereas no lysis of Wue-1-negative Wue-1-negative NALM-6 lymphoma cells NALM-6 cells was seen. On the other hand, the bscCD19 Â CD3 Ab induced lysis of NALM-6 cells, but had After we showed that the bscWue-1 Â CD3 Ab binds to MM no cytotoxic effect towards NCI-H929 cells. Incubation with the cells, we next tested whether this novel recombinant antibody bscEpCAM Â CD3 mAb did not induce lysis of NCI-H929 or was able to specifically induce T-cell-mediated lysis of MM NALM-6 cells (Figure 3). Neither cell line expressed the EpCAM cells. antigen (data not shown). In the absence of effector T cells, no The MM cell line NCI-H929 expresses the Wue-1 antigen but cytotoxic effects of bscWue-1 Â CD3 could be observed since lacks the expression of CD19. On the other hand, the B- cytotoxicity was reduced by 87% after the removal of T cells. lymphoma cell line NALM-6 is negative for the Wue-1 antigen To further characterize the cytotoxic effect of bscWue- but positive for CD19 (Figure 2). Therefore, we used these two 1 Â CD3 Ab, NCI-H929 cells were incubated with effector lines to analyze target cell and antigen specificity of the cells derived from random healthy donors, and increasing amounts of bscWue-1 Â CD3 Ab. Here, we could demonstrate dose-dependent lysis of NCI-H929 cells (Figure 4a). Cytotoxicity assays with NCI-H929 cells were performed without any additional T-cell stimulation. As shown in Figure 4b, the bscWue-1 Â CD3 Ab was effective even at E:T ratios of 5:1. Since the Wue-1 antigen has not been cloned so far, we further aimed at proving that the bscWue-1 Â CD3 Ab-mediated lysis of NCI-H929 cells is antigen specific. Therefore competi- tion experiments with the monoclonal Wue-1 Ab were performed. As demonstrated in Figure 5, incubation with an eight-fold molar excess of monoclonal Wue-1 Ab resulted in the strong inhibition of the cytotoxic activity of the bscWue- 1 Â CD3 Ab. In contrast, the incubation with an appropiate isotype control antibody did not affect the cytotoxic activity.

bscWue-1 Â CD3 Ab induces specific lysis of primary human MM cells

After we demonstrated the specific cytotoxicity of the bscWue- 1 Â CD3 Ab towards the MM cell line NCI-H929, we next tested whether this bsc antibody was also able to induce lysis of primary human MM cells. Tumor samples derived from 11 different MM patients were analyzed. The characteristics of these patients are summarized in Table 1. The mean specific lysis of all patient samples (#1–11) analyzed with bscWue- 1 Â CD3 in this study was 45.5% with a standard deviation (s.d.) of 723.8%. Based on this calculation, we defined 20% as cutoff Figure 1 Binding specificity and purification of the bscWue- level and scored lysis of primary MM cells below 20% as 1 Â CD3 Ab. (a) Immunocytochemical analysis using the bscWue- negative. Â 1 CD3 Ab and the MM cell line NCI-H929 (immunoperoxidase The cytotoxicity experiments were carried out in two different staining of cytospins). (b) The bispecific antibody CD19 Â CD3 served as negative control. (c) Coomassie-stained gel of the purified bscWue- experimental settings. In the first setting, MM cells from bone 1 Â CD3 Ab. M denotes the molecular weight marker. The molecular marrow aspirates of MM patients were isolated and enriched by mass (kDa) is indicated. magnetic separation using the CD138 marker. The purified MM

Leukemia bscWue-1 Â CD3 in the treatment of MM DHo¨nemann et al 640

Figure 2 Wue-1 antigen and CD19 antigen expression on NCI-H929 MM cells and NALM-6 lymphoma cells. (a) FACS analysis of Wue-1 antigen expression was performed with monoclonal Wue-1 Ab and a secondary PE-labeled goat-anti-mouse F(ab2) fragment. Mouse IgG1 (filled lines) was used as isotype control. (b) FACS analysis of CD19 expression was performed with FITC-labeled anti-CD19 mAb. FITC-labeled mouse IgG1 mAb was used as an isotype control.

Figure 4 Dose-dependent and E:T-ratio-dependent cytotoxicity of bscWue-1  CD3 Ab towards NCI-H929 cells. (a) Cytotoxicity assay (48 h) with increasing amounts of bscWue-1  CD3 Ab at an E:T ratio of 10:1. (b) Cytotoxicity assay (48 h) with varying E:T ratios in the Figure 3 Target cell-specific cytotoxicity of the bscWue-1  CD3 presence of bscWue-1  CD3 Ab (2 mg/ml). Specific lysis was Ab. Cytotoxicity assays with effector T cells from healthy human determined as described in Materials and methods. Experiments were donors and a mixed target cell population consisting of NCI-H929 MM performed with effector T cells isolated from three different random cells and NALM-6 lymphoma cells were performed. Either NCI-H929 donors (T-cell samples #1–3). No stimulation of T cells was performed. cells or NALM-6 cells were stained with the membrane-dye PKH-26. Experiments were performed in triplicate, the s.d. is indicated. The 1:1 mixture of labeled and unlabeled target cells was incubated with unlabeled effector T cells (E:T ratio ¼ 10:1) for 48 h in the  m  m presence of bscWue-1 CD3 Ab (2 g/ml), bscCD19 CD3 Ab (1 g/ cells were then incubated with either allogeneic T cells from ml), or bscEpCAM  CD3 Ab (2 mg/ml). Cytotoxicity was determined by PI staining and FACS analysis. No stimulation of T cells was healthy donors or with autologous T cells, which had been performed. Experiments were performed in triplicate, the s.d. is isolated from the peripheral blood of the respective patient. The indicated. assays were performed in the presence of 10 ng/ml PHA and

Leukemia bscWue-1 Â CD3 in the treatment of MM DHo¨nemann et al 641 To approach the in vivo situation more closely a second experimental setting was established. Here, the unmanipulated mononuclear cell fraction containing both myeloma cells and effector T cells from the same bone marrow aspirate was used for cytotoxicity experiments. The E:T ratio in these samples reflects more likely the physiological conditions of the bone marrow because no further enrichment of myeloma cells or T cells was performed. The E:T ratio was determined by FACS analysis. The mononuclear cell fractions were incubated with the bscWue-1 Â CD3 Ab or the bscEpCAM Â CD3 control Ab without additional pre- or costimulation of the effector T cells.

Table 2 Cytotoxicity of the bscWue-1 Â CD3 Ab towards enriched primary human MM cells from bone marrow aspirates

Pat. PBMC E:T ratio Assay time Spec. lysis Figure 5 Antigen specificity of cytotoxicity induced by bscWue- # (days) (%) 1 Â CD3 towards NCI-H929 cells. Cytotoxicity assay (48 h) with an eight-fold molar excess of monoclonal Wue-1 Ab or unspecific control 1 Autol. 10:1 6 52 Ig in the presence of bscWue-1 Â CD3 Ab (2 mg/ml) and an E:T-ratio of 2 Allog. 10:1 5 52 10:1. Specific lysis was determined as described in Materials and 3 Autol. 10:1 6 55/52 methods. Experiments were performed with effector T cells isolated 4* Autol. 10:1 6 28/13/6 from three different random donors (T-cell samples #1–3). No 5 Allog. 10:1 7 52/34 stimulation of T cells was performed. Experiments were performed 6 Allog. 10:1 6 61/50 in triplicate, the s.d. is indicated. 7* Autol. 10:1 6 0 8 Allog. 10:1 6 65/65/74 Table 1 MM patient characteristics Primary human MM cells were isolated and enriched from bone marrow aspirates of eight different patients by immunomagnetic Pat # Age Sex Isotype Stage (Salmon Pretreatment separation using the CD138 marker. PBMCs from healthy donors (years) and Durie) (allog.) or from the respective patients (autol.) were added resulting in an E:T ratio of 10:1. Cytotoxicity experiments were performed with 1 69 M IgG III A MP bscWue-1 Â CD3 Ab at a concentration of 1 mg/ml. As unspecific T-cell 276MLCl III A No stimuli PHA (10 ng/ml) and IL-2 (60 U/ml) were added at day 0. After 5– 3 47 M IgG III A No 7 days, specific lysis was determined as described in Materials and 4 60 M IgA III A Mult. (HDCT) methods. Depending on the amount of multiple myeloma cells 5 69 F IgG II A No obtained, one to three independent experiments were performed, 667MLCk IA No and the specific lysis of each experiment is indicated in the last 7 74 F IgG III A MP column. The bscEpCAM Â CD3 Ab was used as a negative control. 8 56 F IgG III A No Samples marked with * were scored as not having achieved specific 9 65 M IgG III A Mult. (no HDCT) cytotoxicity. 10 66 M IgG III A Rad. 11 64 F IgG III A No Primary MM cells derived from diagnostic bone marrow aspirates of 11 different patients were used for cytotoxicity experiments with the bscWue-1 Â CD3 Ab in this study (see, for results, Tables 2 and 3). The characteristics of these patients are summarized in this table. LC: light chain; MP: melphalan+prednisone; Mult. (HDCT): multiple treatments including high-dose chemotherapy; Mult. (no HDCT): multiple treatments without high-dose chemotherapy; Rad.: radio- therapy.

60 U/ml IL-2. As shown in Table 2, specific lysis of primary MM cells after the addition of 1 mg/ml bscWue-1 Â CD3 Ab could be observed in 6/8 cases analyzed, and ranged from 34 to 74%. Thereby, a specific lysis above 50% could be observed in most experiments. No significant difference was observed between medium control samples and samples with the bscEp- CAM Â CD3 control antibody (data not shown). From patient 3 enough MM cells were purified to perform a dose–response curve, suggesting that the maximal cytotoxic activity was not Figure 6 Dose–response curve of bscWue-1 Â CD3 with primary reached with 1 mg/ml of bscWue-1 Â CD3 in this specific MM cells derived from patient 3. Cytotoxicity assay with primary MM experimental setting (Figure 6). Cytotoxicity assays performed cells from patient 3 (Table 1) with increasing amounts of bscWue- with CLL cells from two different patients as negative controls 1 Â CD3 Ab at an E:T ratio of 10:1. As unspecific T-cell stimuli PHA (10 ng/ml) and IL-2 (60 U/ml) were added at day 0. Specific lysis was showed no or only minor lysis induced by the bscWue-1 Â CD3 determined as described in Materials and methods. Experiments were Ab, but displayed high susceptibility to bscCD19 Â CD3- performed with autologous T cells isolated from the peripheral blood mediated cytotoxicity (Figure 7). of the same patient. Experiments were performed in duplicate.

Leukemia bscWue-1 Â CD3 in the treatment of MM DHo¨nemann et al 642 As shown in Table 3, incubation with 2.5 mg/ml bscWue- binding of bscWue-1 Â CD3 to Wue-1-expressing cells was 1 Â CD3 Ab for 4 or 6 days resulted in specific lysis of primary demonstrated by immunocytochemistry. MM cells in the range of 36–88%. We analyzed the biological activity of the recombinant The mean specific lysis achieved with allogeneic T cells (# 2, bscWue-1 Â CD3 antibody using the human MM cell line 5, 6, 8) was 54.5% with an s.d. of 7 5.2%. The mean specific NCI-H929 and primary myeloma cells freshly isolated from lysis achieved with autologous T cells (# 1, 3, 4, 7, 9, 10, 11) diagnostic bone marrow aspirates of 11 MM patients. We could was 40.5% with an s.d. of 710.7%. This difference is not demonstrate that bscWue-1 Â CD3 induces T-cell-mediated statistically significant as determined by the unpaired t-test cytotoxicity against both the MM cell line and primary MM (P ¼ 0.37). Thus, no significant difference in specific cytotoxicity cells in nine out of 11 cases. No significant differences were between experiments with either autologous or allogeneic found between samples containing autologous or allogeneic T effector T cells could be observed. cells. Since an allogeneic response is MHC restricted, this observation underscores the ability of bscWue-1 Â CD3 to elicit MHC-independent cytotoxicity. This could be of great impor- Discussion tance for a potential clinical application of this antibody. The bscWue-1 Â CD3 Ab-induced cytotoxic effects were We have generated a novel recombinant MM cell-directed bsc target cell and antigen specific and were mediated by T cells antibody, designated bscWue-1 Â CD3. The antibody consists of as demonstrated by the following results: two different single-chain Fv fragments joined by a glycine– serine linker. One specificity is directed against the CD3 antigen (1) control bsc antibodies of different target specificity but of human T cells and the other antigen-binding site engages the identical molecular structure, bscEpCAM Â CD3 and recently described plasma cell-specific Wue-1 antigen.12 The bscCD19 Â CD3, failed to lyse myeloma cells, bscWue-1 Â CD3 Ab was cloned, expressed, and purified as (2) cytotoxicity could be blocked by the Wue-1 monoclonal previously described for similar bsc constructs.34,37 Specific antibody, (3) the bscWue-1 Â CD3 Ab failed to lyse the B-lymphoma cell line NALM-6 that does not express the Wue-1 antigen, and (4) in the absence of T cells no cytotoxic effects of the bscWue- 1 Â CD3 Ab could be observed. In addition, we have shown the capability of the bscWue- 1 Â CD3 Ab to induce MM cell-directed cytotoxicity with autologous T cells derived from the bone marrow of patients at an E:T ratio of 1:2 and without any additional T-cell pre- or costimulation. The fact that the bscWue-1 Â CD3 Ab is able to mediate specific cytotoxicity under these conditions is a promising property of the antibody with regard to its potential clinical application. This finding is also of interest in the context of previous reports about the presence of myeloma- idiotype-specific T cells located in the bone marrow that seem to be dormant in MM patients with active disease.45,46 One could speculate that the bscWue-1 Â CD3 Ab might have the potential to activate these dormant myeloma-specific T cells, and thus might trigger a sustained myeloma-directed immune response. Figure 7 Lack of activity of bscWue-1 Â CD3 on CLL cells in The observation that bscWue-1 Â CD3 Ab-induced cytotoxi- contrast to bscCD19 Â CD3. Cytotoxicity assay with CLL cells from city is mediated by T cells at low E:T ratios without any need of two different donors in the presence of bscWue-1 Â CD3 Ab (1 mg/ml) additional pre- or costimulation is in accordance with previous or bscCD19 Â CD3 (100 pg/ml). As unspecific T-cell stimuli PHA studies using similar bsc antibodies of the same molecular (10 ng/ml) and IL-2 (60 U/ml) were added at day 0. Specific lysis was format, for instance, bscEpCAM Â CD3 Ab and bscCD19 Â CD3 determined as described in Materials and methods. Experiments were 35,36 performed with effector T cells isolated from two different random Ab. The findings with these single-chain antibodies are in donors. Experiments were performed in triplicate, the s.d. is indicated. sharp contrast to findings with conventional bispecific anti- bodies including bispecific antibody constructs that require Table 3 Cytotoxicity of the bscWue-1 Â CD3 Ab towards non- much higher E:T ratios and extensive prestimulation (eg IL-2) or enriched primary human MM cells costimulation (for example, anti-CD28) of the effector T cells to be efficacious.47–53 Thus, it seems that the biological character- Pat. # E:T ratio Assay time (days) Spec. lysis (%) istics described above are unique properties of this particular bsc format that discriminates these molecules from other T-cell 9 1:2 4 36 recruiting bispecific antibodies generated until now. The 10 5:1 6 88 mechanism underlying these biological properties is still unclear 11 6:1 6 38 and under current investigation. bscWue-1 Â CD3 Ab (2.5 mg/ml) was added to the mononuclear cell Previously, it has been reported that ligation of Wue-1 by the fractions derived from bone marrow aspirates from three different MM + + monoclonal antibody was able to enhance the proliferation of patients. The proportion of T cells (CD3 ) and MM cells (CD138 ) in the NCI-H929 cells.12 Thus, one might speculate that a potential respective aspirates was determined by FACS analysis at day 0. The resulting calculated E:T ratios (CD3+:CD138+) are indicated. No therapeutic effect of Wue-1 antibody constructs could be additional effector T cells from the peripheral blood were added and antagonized by the induction of myeloma cell proliferation. no additional T-cell stimulation was performed. After 4 or 6 days, However, in our cytotoxicity experiments with the bscWue- specific lysis was determined as described in Materials and methods. 1 Â CD3 antibody, a decrease (rather than increase) of the viable

Leukemia bscWue-1 Â CD3 in the treatment of MM DHo¨nemann et al 643 myeloma cell fraction was observed. Furthermore, in the References presence of the bscWue-1 Â CD3 Ab alone no enhanced proliferation of NCI-H929 cells could be observed (data not 1 Bataille R, Harousseau JL. Multiple myeloma. N Engl J Med 1997; shown). Therefore, we assume that an undesirable proliferative 336: 1657–1664. stimulus due to treatment with bscWue-1 Â CD3 is unlikely. 2 Peest D, Deicher H, Coldewey R, Leo R, Bartl R, Bartels H et al. A Several investigators have reported on antibody-based im- comparison of polychemotherapy and melphalan/prednisone for primary remission induction and interferon-alpha for main- munotherapeutic approaches in MM using anti-CD38, anti- tenance treatment in multiple myeloma A prospective trial of HM1.24, or anti-CD138 antibody constructs. However, since the German Myeloma Treatment Group. Eur J Cancer 1995; 31a: the targeted antigens are also expressed on various normal 146–151. tissues including normal hematopoietic cells, these approaches 3 Attal M, Harousseau JL, Stoppa AM, Sotto JJ, Fuzibet JG, Rossi JF might induce side effects.13–16,54 In contrast, the expression of et al. A prospective, randomized trial of autologous bone marrow the Wue-1 antigen is restricted to normal and malignant plasma transplantation and chemotherapy in multiple myeloma. Inter- groupe Francais du Myelome. N Engl J Med 1996; 335: 91–97. cells and could not be detected on other normal tissues or other 12 4 Hallek M, Leif Bergsagel P, Anderson KC. Multiple myeloma: normal cell types. increasing evidence for a multistep transformation process. Blood Among the anti-MM antibodies described previously, chi- 1998; 91: 3–21. meric and humanized forms of anti-CD38 and anti-HM1.24 5 Massaia M, Borrione P, Battaglio S, Mariani S, Beggiato E, Napoli P mAbs have been reported to show cytotoxic effects against et al. Idiotype vaccination in human myeloma: generation of myeloma cells in the presence of effector cells. The recently tumor-specific immune responses after high-dose chemotherapy. Blood 1999; 94: 673–683. described humanized anti-HM1.24 mAb has been shown to 6 Reichardt VL, Okada CY, Liso A, Benike CJ, Stockerl Goldstein KE, mediate antibody-dependent cellular cytotoxicity against both Engleman EG et al. Idiotype vaccination using dendritic cells after myeloma cell lines and myeloma cells from patients in the autologous peripheral blood stem cell transplantation for multiple presence of human peripheral blood mononuclear cells myeloma–a feasibility study. Blood 1999; 93: 2411–2419. (PBMCs). To achieve efficient specific target cell lysis (30– 7 McLaughlin P, Grillo Lopez AJ, Link BK, Levy R, Czuczman MS, 40%), an E:T ratio of 50:1 to 100:1 was required in these Williams ME et al. chimeric anti-CD20 monoclonal 55 antibody therapy for relapsed : half of patients experiments. Similar results were obtained with a humanized respond to a four-dose treatment program. J Clin Oncol 1998; 16: anti-CD38 antibody, which was tested with the Wien-133 cell 2825–2833. line and allogeneic T cells.56 The cytotoxic potential of the anti- 8 Maloney DG, Liles TM, Czerwinski DK, Waldichuk C, Rosenberg CD138 antibody B-B4 was tested primarily without effector cells J, Grillo LA, Levy R. Phase I clinical trial using escalating single- as an immunotoxin conjugate in preclinical studies.15,57 In dose infusion of chimeric anti-CD20 monoclonal antibody (IDEC- addition, sperm protein 17 (Sp17) was described as a suitable C2B8) in patients with recurrent B-cell lymphoma. Blood 1994; 84: 2457–2466. target for the immunotherapy of MM with antigen-specific 9 Coiffier B, Lepage E, Briere J, Herbrecht R, Tilly H, Bouabdallah R cytotoxic T cells, but was not yet evaluated for an antibody- et al. CHOP chemotherapy plus rituximab compared with CHOP based immunotherapeutic approach.58 alone in elderly patients with diffuse large-B-cell lymphoma. N In contrast, in our experiments with the bscWue-1 Â CD3 Ab Engl J Med 2002; 346: 235–242. using primary myeloma cells and autologous T cells, an E:T ratio 10 Maloney DG, Donovan K, Hamblin TJ. Antibody therapy for of 1:2 was sufficient for myeloma cell lysis. Another possible treatment of multiple myeloma. Semin Hematol 1999; 36: 30–33. 11 Treon SP, Shima Y, Grossbard ML, Preffer FI, Belch AR, Pilarski advantage of the single-chain antibody construct presented LM, Anderson KC. Treatment of multiple myeloma by antibody compared to other plasma-cell-associated antibodies described, mediated immunotherapy and induction of myeloma selective so far, may be its low molecular weight that might improve its antigens. Ann Oncol 2000; 11 (Suppl 1): 107–111. penetration into malignant tissue, as has been shown for Fab or 12 Greiner A, Neumann M, Stingl S, Wassink S, Marx A, Riechert F Fv antibody fragments.39 With regard to immunogenicity of the et al. Characterization of Wue-1, a novel monoclonal antibody covalently linked 4 V-regions devoid of all constant immuno- that stimulates the growth of cells. Virchows Arch 2000; 437: 372–379. globulin domains, we anticipate a reduction in the HAMA 13 Alessio M, Roggero S, Funaro A, De Monte LB, Peruzzi L, Geuna (human anti-mouse antibody) response that has been observed M et al. CD38 molecule: structural and biochemical analysis on with other conventional constructs. human T lymphocytes, thymocytes, and plasma cells. J Immunol In summary, we have shown that a recombinant bsc antibody, 1990; 145: 878–884. bscWue-1 Â CD3, specifically induces cytotoxicity against a 14 Rees RC. MHC restricted and non-restricted killer lymphocytes. human MM cell line and primary MM cells derived from bone Blood Rev 1990; 4: 204–210. 15 Vooijs WC, Post J, Wijdenes J, Schuurman HJ, Bolognesi A, Polito marrow aspirates of patients. BscWue-1 Â CD3 is efficacious at L et al. Efficacy and toxicity of plasma-cell-reactive monoclonal low E:T ratios and without additional T-cell stimulation and, antibodies B-B2 and B-B4 and their immunotoxins. Cancer therefore, shares the exceptional biological properties with other Immunol Immunother 1996; 42: 319–328. members of this particular bsc format. To our knowledge, the 16 Ohtomo T, Sugamata Y, Ozaki Y, Ono K, Yoshimura Y, Kawai S bscWue-1 Â CD3 Ab is the first MM-directed bispecific antibody et al. Molecular cloning and characterization of a surface antigen described so far, and the first immunotherapeutic approach preferentially overexpressed on multiple myeloma cells. Biochem Biophys Res Commun 1999; 258: 583–591. directed towards the Wue-1 antigen, which has thus turned out 17 Demanet C, Brissinck J, De Jonge J, Thielemans K. Bispecific to be a promising target for an antibody-based treatment of MM. antibody-mediated immunotherapy of the BCL1 lymphoma: increased efficacy with multiple injections and CD28-induced costimulation. Blood 1996; 87: 4390–4398. 18 Haagen IA, Geerars AJ, de Lau WB, Bast BJ, de Gast BC. The efficacy of CD3 Â CD19 bispecific monoclonal antibody (BsAb) Acknowledgements in a clonogenic assay: the effect of repeated addition of BsAb and interleukin-2. Blood 1995; 85: 3208–3212. We thank A Greiner, University of Wu¨rzburg (Germany), for 19 Honeychurch J, Tutt AL, Valerius T, Heijnen IA, Van De Winkel providing the immunocytochemistry and immunohistochemistry JG, Glennie MJ. Therapeutic efficacy of FcgammaRI/CD64- data. This work was supported in part by the SFB 506 and SFB 456 directed bispecific antibodies in B-cell lymphoma. Blood 2000; grants from the Deutsche Forschungsgemeinschaft. 96: 3544–3552.

Leukemia bscWue-1 Â CD3 in the treatment of MM DHo¨nemann et al 644 20 Daniel PT, Kroidl A, Kopp J, Sturm I, Moldenhauer G, Do¨rken B 40 Orlandi R, Gussow DH, Jones PT, Winter G. Cloning immunoglo- et al. Immunotherapy of B-cell lymphoma with CD3 Â 19 bulin variable domains for expression by the polymerase chain bispecific antibodies: costimulation via CD28 prevents ‘veto’ reaction. Proc Natl Acad Sci USA 1989; 86: 3833–3837. of antibody-targeted cytotoxic T cells. Blood 1998; 92: 41 Du¨bel S, Breitling F, Fuchs P, Zewe M, Gotter S, Welschof M et al. 4750–4757. Isolation of IgG antibody Fv-DNA from various mouse and rat 21 Nitta T, Sato K, Yagita H, Okumura K, Ishii S. Preliminary trial of hybridoma cell lines using the polymerase chain reaction with a specific targeting therapy against malignant glioma. Lancet 1990; simple set of primers. J Immunol Methods 1994; 175: 89–95. 335: 368–371. 42 Traunecker A, Lanzavecchia A, Karjalainen K. Bispecific single 22 Canevari S, Stoter G, Arienti F, Bolis G, Colnaghi MI, Di Re EM chain molecules (Janusins) target cytotoxic lymphocytes on HIV et al. Regression of advanced ovarian carcinoma by intraperitoneal infected cells. EMBO J 1991; 10: 3655–3659. treatment with autologous T lymphocytes retargeted by a bispecific 43 Harris NL, Jaffe ES, Diebold J, Flandrin G, Muller-Hermelink HK, monoclonal antibody. J Natl Cancer Inst 1995; 87: 1463–1469. Vardiman J et al. World Health Organization classification of 23 De Gast GC, Van Houten AA, Haagen IA, Klein S, De Weger RA, neoplastic diseases of the hematopoietic and lymphoid tissues: Van Dijk A et al. Clinical experience with CD3 Â CD19 bispecific report of the Clinical Advisory Committee Meeting – Airlie House, antibodies in patients with B cell malignancies. J Hematother Virginia, November 1997. J Clin Oncol 1999; 17: 3835–3849. 1995; 4: 433–437. 44 Marx A, Kirchner T, Greiner A, Muller-Hermelink HK, Schalke B, 24 Hartmann F, Renner C, Jung W, Deisting C, Juwana M, Eichentopf Osborn M. Neurofilament epitopes in thymoma and antiaxonal B et al. Treatment of refractory Hodgkin’s disease with an anti- in myasthenia gravis. Lancet 1992; 339: 707–708. CD16/CD30 bispecific antibody. Blood 1997; 89: 2042–2047. 45 Fagerberg J, Yi Q, Gigliotti D, Harmenberg U, Ruden U et al. 25 Borchmann P, Schnell R, Fuss I, Manzke O, Davis T, Lewis LD T-cell-epitope mapping of the idiotypic monoclonal IgG heavy et al. Phase 1 trial of the novel bispecific molecule H22 Â Ki-4 in and light chains in multiple myeloma. Int J Cancer 1999; 80: patients with refractory . Blood 2002; 100: 671–680. 3101–3107. 46 Kay N, Leong T, Kyle RA, Greipp P, Van Ness B, Bone N et al. 26 Staerz UD, Bevan MJ. Hybrid hybridoma producing a bispecific Altered T cell repertoire usage in CD4 and CD8 subsets of multiple monoclonal antibody that can focus effector T-cell activity. Proc myeloma patients, a Study of the Eastern Cooperative Oncology Natl Acad Sci USA 1986; 83: 1453–1457. Group (E9487). Leukemia Lymphoma 1999; 33: 127–133. 27 Lanzavecchia A, Scheidegger D. The use of hybrid hybridomas to 47 Bohlen H, Manzke O, Patel B, Moldenhauer G, Do¨rken B, von target human cytotoxic T lymphocytes. Eur J Immunol 1987; 17: Fliedner V et al. Cytolysis of leukemic B-cells by T-cells activated 105–111. via two bispecific antibodies. Cancer Res 1993; 53: 4310–4314. 28 Kroesen BJ, ter HA, Spakman H, Willemse P, Sleijfer DT, de VE 48 Manzke O, Berthold F, Huebel K, Tesch H, Diehl V, Bohlen H. et al. Local antitumour treatment in carcinoma patients with CD3 Â CD19 bispecific antibodies and CD28 bivalent antibodies bispecific-monoclonal-antibody-redirected T cells. Cancer Immu- enhance T-cell reactivity against autologous leukemic cells in nol Immunother 1993; 37: 400–407. pediatric B-ALL bone marrow. Int J Cancer 1999; 80: 715–722. 29 Zhu Z, Lewis GD, Carter P. Engineering high affinity humanized anti- 49 Haagen IA, van de Griend R, Clark M, Geerars A, Bast B, de Gast p185HER2/anti-CD3 bispecific F(ab0)2 for efficient lysis of p185HER2 B. Killing of human leukaemia/lymphoma B cells by activated overexpressing tumor cells. Int J Cancer 1995; 62: 319–324. cytotoxic T lymphocytes in the presence of a bispecific mono- 30 Mallender WD, Voss EJ. Construction, expression, and activity of a clonal antibody (alpha CD3/alpha CD19). Clin Exp Immunol 1992; bivalent bispecific single-chain antibody. J Biol Chem 1994; 269: 90: 368–375. 199–206. 50 Haagen IA, Geerars AJ, de Lau WB, Clark MR, van de Griend RJ, 31 Mallender WD, Ferreira ST, Voss EJ, Coelho ST. Inter-active-site Bast BJ et al. Killing of autologous B-lineage using distance and solution dynamics of a bivalent-bispecific single- CD3 Â CD19 bispecific monoclonal antibody in end stage chain antibody molecule. Biochemistry 1994; 33: 10100–10108. leukemia and lymphoma. Blood 1994; 84: 556–563. 32 Kostelny SA, Cole MS, Tso JY. Formation of a bispecific antibody 51 Cochlovius B, Kipriyanov SM, Stassar MJ, Christ O, Schuhmacher by the use of zippers. J Immunol 1992; 148: 1547–1553. J, Strauss G et al. Treatment of human B cell lymphoma xenografts 33 Gruber M, Schodin BA, Wilson ER, Kranz DM. Efficient tumor cell with a CD3 Â CD19 diabody and T cells. J Immunol 2000; 165: lysis mediated by a bispecific single chain antibody expressed in 888–895. Escherichia coli. J Immunol 1994; 152: 5368–5374. 52 Bohlen H, Hopff T, Manzke O et al. Lysis of malignant B cells from 34 Mack M, Riethmuller G, Kufer P. A small bispecific antibody patients with B-chronic lymphocytic leukemia by autologous construct expressed as a functional single-chain molecule with T cells activated with CD3 Â CD19 bispecific antibodies in high tumor cell cytotoxicity. Proc Natl Acad Sci USA 1995; 92: combination with bivalent CD28 antibodies. Blood 1993; 82: 7021–7025. 1803–1812. 35 Mack M, Gruber R, Schmidt S, Riethmuller G, Kufer P. Biologic 53 Bauer S, Renner C, Juwana JP, Held G, Ohnesorge S, Gerlach K properties of a bispecific single-chain antibody directed against et al. Immunotherapy of human tumors with T-cell-activating 17-1A (EpCAM) and CD3: tumor cell-dependent T cell stimulation bispecific antibodies: stimulation of cytotoxic pathways in vivo. and cytotoxic activity. J Immunol 1997; 158: 3965–3970. Cancer Res 1999; 59: 1961–1965. 36 Dreier T, Lorenczewski G, Brandl C, Hoffmann P, Syring U, 54 Treon SP, Shima Y, Preffer FI, Doss DS, Ellman L, Schlossman RL Hanakam F et al. Extremely potent, rapid and costimulation- et al. Treatment of by antibody-mediated independent cytotoxic T-cell response against lymphoma cells immunotherapy. Semin Oncol 1999; 26: 97–106. catalyzed by a single-chain bispecific antibody. Int J Cancer 2002; 55 Ozaki S, Kosaka M, Wakahara Y, Ozaki Y, Tsuchiya M, Koishihara 100: 690–697. Y et al. Humanized anti-HM1.24 antibody mediates myeloma cell 37 Lo¨ffler A, Kufer P, Lutterbu¨se R, Zettl F, Daniel PT, Schwenkenbe- cytotoxicity that is enhanced by cytokine stimulation of effector cher JM et al. A recombinant bispecific single-chain antibody, cells. Blood 1999; 93: 3922–3930. CD19 Â CD3, induces rapid and high lymphoma-directed cytotoxi- 56 Ellis JH, Barber KA, Tutt A, Hale C, Lewis AP, Glennie MJ et al. city by unstimulated T lymphocytes. Blood 2000; 95: 2098–2103. Engineered anti-CD38 monoclonal antibodies for immunotherapy 38 Lo¨ffler A, Gruen M, Wuchter C, Schriever F, Kufer P, Dreier T et al. of multiple myeloma. J Immunol 1995; 155: 925–937. Efficient elimination of chronic lymphocytic leukaemia B cells by 57 Post J, Vooijs WC, Bast BJ, De Gast GC. Efficacy of an anti-CD138 autologous T cells with a bispecific anti-CD19/anti-CD3 single- immunotoxin and doxorubicin on drug-resistant and drug-sensitive chain antibody construct. Leukemia 2003; 17: 900–909. myeloma cells. Int J Cancer 1999; 83: 571–576. 39 Yokota T, Milenic DE, Whitlow M, Schlom J. Rapid tumor 58 Chiriva-Internati M, Wang Z, Salati E, Bumm K, Barlogie B, Lim penetration of a single-chain Fv and comparison with other SH. Sperm protein 17 (Sp17) is a suitable target for immunotherapy immunoglobulin forms. Cancer Res 1992; 52: 3402–3408. of multiple myeloma. Blood 2002; 100: 961–965.

Leukemia