Elsevier

Molecular Immunology

Volume 42, Issue 11, July 2005, Pages 1311-1319
Molecular Immunology

Anti-neuroblastoma effect of ch14.18 antibody produced in CHO cells is mediated by NK-cells in mice

https://doi.org/10.1016/j.molimm.2004.12.018Get rights and content

Abstract

Successful treatment of stage 4 neuroblastoma remains a major challenge in pediatric oncology. In order to improve the outcome, passive immunotherapy using human–mouse chimeric monoclonal anti-disialoganglioside GD2 antibody ch14.18 has been evaluated in early phase clinical trials with promising results in progressing stage 4 neuroblastoma patients. In preparation of European phase III clinical trial (HR-NBL-1/ESIOP), the cell line used for production of ch14.18 was changed. Specifically, the plasmid encoding for ch14.18 antibody was recloned into CHO cells. Here, we report the in vitro and in vivo anti-neuroblastoma activity of antibody ch14.18 produced in CHO cells (ch14.18/CHO) compared to that of ch14.18 manufactured from SP2/0 (ch14.18/SP2/0) and NS0 cells (ch14.18/NS0).

First, we demonstrate identical binding of ch14.18/CHO to the nominal antigen disialoganglioside GD2 in vitro compared to ch14.18/SP2/0 and ch14.18/NS0. Binding was GD2-specific, since all precursor- and metabolite-gangliosides of GD2 tested were not recognized by ch14.18/CHO. Second, the functional properties of ch14.18/CHO were determined in complement-dependent cytotoxicity (CDC) and antibody-dependent cellular cytotoxicity (ADCC) reactions against GD2 positive neuroectodermal tumor cell lines in vitro. There was no difference in CDC mediated specific tumor cell lysis among the three different ch14.18 antibody preparations. Interestingly, ch14.18/CHO showed superior ADCC activity at low antibody concentrations. Third, the efficacy of ch14.18/CHO was evaluated in the NXS2 neuroblastoma model in vivo. Importantly, the ch14.18/CHO preparation was effective in suppression of experimental liver metastasis in this model. In vivo depletion of NK-cells completely abrogated this effect, suggesting that the mechanism involved in the ch14.18/CHO induced anti-neuroblastoma effect is mediated by NK-dependent ADCC.

Introduction

One of the major challenges in pediatric oncology is the effective treatment of stage 4 neuroblastoma, one of the most common solid tumors in children (Niethammer and Handgretinger, 1995). Over 50% of the patients present with nonresectable primary tumors and disseminated metastasis to distant organ sites at diagnosis. The overall survival rate of such patients has not been significantly improved during the last 20 years despite the introduction of radiotherapy and/or high dose chemotherapy followed by allogeneic or autologous hematopoietic stem cell transplantation and differentiation therapy with 13-cis retinoic acid (Matthay et al., 1999).

One concept in the treatment of neuroblastoma is targeting of tumor cells with monoclonal antibodies directed against the glycolipid antigen ganglioside GD2, which is expressed at high density in most neuroblastoma cases. In contrast, limited expression has been detected in normal tissues such as the peripheral nervous system and the cerebellum, which is protected from immunoglobulins by the blood brain barrier. A variety of murine and human–mouse chimeric anti-GD2 antibodies have been generated (Schulz et al., 1984, Saito et al., 1985, Mujoo et al., 1989, Mueller et al., 1990) two of which are under evaluation in the clinical setting: ch14.18 (Handgretinger et al., 1995, Uttenreuther-Fischer et al., 1995) and 3F8 (Kushner et al., 2001). Ch14.18 is a human–mouse chimeric construct consisting of variable regions derived from the murine anti-GD2 antibody 14G2a and constant regions of heavy and light chains from a human IgG1 molecule. 3F8 is a completely murine antibody. A major advantage of a human–mouse chimeric construct is the reduction of the immunogenicity in patients.

The rationale for passive immunotherapy in neuroblastoma is provided by the effector functions of anti-GD2 mAbs. In vitro, anti-GD2 antibodies facilitate antibody-dependent cellular cytotoxicity (ADCC), a process mediated primarily by Fc receptor bearing effector cells like monocytes, macrophages and natural killer (NK-cells) (Honsik et al., 1986, Barker et al., 1991). The second effector function of anti-GD2 mAbs against neuroblastoma cells is activation of the complement cascade through the classical pathway, leading to complement-dependent cytotoxicity (CDC) (Mujoo et al., 1989, Mueller et al., 1990). Reports suggest that a third mechanism contributing to the anti-tumor effect upon passive immunotherapy with tumor specific monoclonal antibodies involves the anti-idiotypic network (Cheung et al., 2000, Riethmuller et al., 1998). Specifically, the humoral response of the host against the therapeutic mAb (Ab1) leads to the generation of an anti-idiotype-antibody (Ab2) representing an internal image of Ab1. In a subsequent humoral immune response the host may generate an anti-anti-idiotypic antibody (Ab3), which is capable to specifically recognize the nominal antigen expressed by the tumor cell. The role of the anti-idiotypic network in the activity of ch14.18 is unknown so far.

First clinical results using ch14.18 anti-GD2 mAb antibody in children with neuroblastoma were obtained in two phase I trials using ch14.18 mAb produced in non-secreting murine myeloma cells Sp2/0 (ch14.18/Sp2/0). First, 10 patients were treated with increasing doses of ch14.18 mAb. Side effects consisted primarily of pain, tachycardia, hypertension, fever, and urticaria. Objective responses were observed in five patients: one partial (PR), four mixed responses (MRs) and one stable disease (SD) (Yu et al., 1998). In a second phase I trial nine patients with stage 4 neuroblastoma were treated with increasing doses of 30 up to 50 mg/m2/day for 5 days. Side effects were again pain and anaphylactic reactions such as pruritus and urticaria. Two complete responses (CR), 2 PRs, 1 MR, 1 SD and 3 PDs were observed (Handgretinger et al., 1995). Promising clinical responses were also observed in a phase II trial using murine anti-GD2 mAb 3F8 (Cheung et al., 1998a, Cheung et al., 1998b). Based on these results, the European HR-NBL-1/ESIOP study group decided to test the hypothesis that adjuvant immunotherapy with anti-GD2 mAb ch14.18 provides a survival benefit for stage 4 neuroblastoma patients.

For this purpose, a large quantity of ch14.18 antibody was produced following GMP production standards. The antibody producing cell lines used to generate recombinant ch14.18 so far are Sp2/0 and NS0 cells, which are both non-secreting murine myeloma cells commonly used for antibody production. However, these murine cell lines carry murine xenotropic retrovirus in contrast to the cells of hamster origin (Shepherd et al., 2003). Therefore, we decided to switch the production cell line from SP2/0 to chinese hamster ovary (CHO) cells, in order to avoid extensive viral clearance studies of murine xenotropic retrovirus during the downstream purification process. Recloning of the plasmid driving the expression of ch14.18 antibody was successfully accomplished in CHO cells. Ch14.18 protein (ch14.18/CHO) was purified from stably producing cell lines and a master working cell bank was generated for subsequent large-scale production.

Here we report for the first time the characterisation and anti-neuroblastoma activity of the new recloned ch14.18/CHO mAb produced for the HR-NBL-1/ESIOP trial in vitro and in vivo.

Section snippets

Animals

Female A/J mice (6–8 weeks of age) were obtained from Harlan Sprague–Dawley, Sulzfeld, Germany. All animals were housed in university-approved facilities and animal experiments were performed according to the German guide for the care and use of laboratory animals, i.e. “Tierschutzgesetz”.

Cell lines and murine–tumor model

The human NB cell lines SK-N-AS and Kelly and the human melanoma cell line M21 were propagated in RPMI (10% FCS) under standard tissue culture conditions (37 °C, 5% CO2).

The murine NXS2 NB cell line expressing

Protein analysis of ch14.18/CHO

The antibody concentrations were determined using two independent methods in order to assure for molar equivalence in all in vitro and in vivo experiments. The absence of protein aggregates in the antibody preparations was demonstrated by size exclusion chromatography (data not shown). Heavy and light chains of the three different ch14.18 preparations were analyzed by SDS-PAGE gel electrophoresis under reducing conditions (Fig. 1). The separation of heavy and light chains revealed identical

Discussion

Passive immunotherapy with antibodies directed against disialoganglioside GD2 provides an important strategy for the treatment of neuroblastoma. Results from early phase clinical trials suggest that this antibody may impact on the course of the disease based on the reported response rates (Handgretinger et al., 1995, Uttenreuther-Fischer et al., 1995, Yu et al., 1998).

In a recent study, the effect of treatment with chimeric anti-GD2-antibody ch14.18 on the consolidation of children older than 1

Acknowledgements

This work was supported by EU Grant SIOPEN-R-NET no. QLRI-CT-2002-01768 to RL, DFG Grant Lo 632 2-3 and Fördergesellschaft Kinderkrebs-Neuroblastom-Forschung e.V. to HNL, and grants from the Charité Forschungsförderung to SF and HNL. The authors thank Dr. S.D. Gillies for outstanding technical support and the supply of the ch14.18 plasmid, and Kerstin Hilt and Anne Strandsby for technical assistance. We also acknowledge the financial support of the following European institutions, which are

References (25)

  • C.J. Honsik et al.

    Lymphokine-activated killer cells targeted by monoclonal antibodies to the disialogangliosides GD2 and GD3 specifically lyse human tumor cells of neuroectodermal origin

    Proc. Natl. Acad. Sci. U.S.A.

    (1986)
  • B.H. Kushner et al.

    Phase II trial of the anti-G (D2) monoclonal antibody 3F8 and granulocyte-macrophage colony-stimulating factor for neuroblastoma

    J. Clin. Oncol.

    (2001)
  • Cited by (96)

    View all citing articles on Scopus
    1

    RL and HNL share authorships.

    View full text