A potent and specific immunotoxin for tumor cells expressing disialoganglioside GD2

GD2-targeting therapy: a comparative analysis of approaches and promising directions

Disialoganglioside GD2 is a promising target for immunotherapy with expression primarily restricted to neuroectodermal and epithelial tumor cells. Although its role in the maintenance and repair of neural tissue is well-established, its functions during normal organism development remain understudied. Meanwhile, studies have shown that GD2 plays an important role in tumorigenesis. Its functions include proliferation, invasion, motility, and metastasis, and its high expression and ability to transform the tumor microenvironment may be associated with a malignant phenotype. Structurally, GD2 is a glycosphingolipid that is stably expressed on the surface of tumor cells, making it a suitable candidate for targeting by antibodies or chimeric antigen receptors. Based on mouse monoclonal antibodies, chimeric and humanized antibodies and their combinations with cytokines, toxins, drugs, radionuclides, nanoparticles as well as chimeric antigen receptor have been developed. Furthermore, vaccines and photoimmunotherapy are being used to treat GD2-positive tumors, and GD2 aptamers can be used for targeting. In the field of cell therapy, allogeneic immunocompetent cells are also being utilized to enhance GD2 therapy. Efforts are currently being made to optimize the chimeric antigen receptor by modifying its design or by transducing not only αβ T cells, but also γδ T cells, NK cells, NKT cells, and macrophages. In addition, immunotherapy can combine both diagnostic and therapeutic methods, allowing for early detection of disease and minimal residual disease. This review discusses each immunotherapy method and strategy, its advantages and disadvantages, and highlights future directions for GD2 therapy.

Biology of GD2 ganglioside: implications for cancer immunotherapy

Part of the broader glycosphingolipid family, gangliosides are composed of a ceramide bound to a sialic acid-containing glycan chain, and locate at the plasma membrane. Gangliosides are produced through sequential steps of glycosylation and sialylation. This diversity of composition is reflected in differences in expression patterns and functions of the various gangliosides. Ganglioside GD2 designates different subspecies following a basic structure containing three carbohydrate residues and two sialic acids. GD2 expression, usually restrained to limited tissues, is frequently altered in various neuroectoderm-derived cancers. While GD2 is of evident interest, its glycolipid nature has rendered research challenging. Physiological GD2 expression has been linked to developmental processes. Passing this stage, varying levels of GD2, physiologically expressed mainly in the central nervous system, affect composition and formation of membrane microdomains involved in surface receptor signaling. Overexpressed in cancer, GD2 has been shown to enhance cell survival and invasion. Furthermore, binding of antibodies leads to immune-independent cell death mechanisms. In addition, GD2 contributes to T-cell dysfunction, and functions as an immune checkpoint. Given the cancer-associated functions, GD2 has been a source of interest for immunotherapy. As a potential biomarker, methods are being developed to quantify GD2 from patients' samples. In addition, various therapeutic strategies are tested. Based on initial success with antibodies, derivates such as bispecific antibodies and immunocytokines have been developed, engaging patient immune system. Cytotoxic effectors or payloads may be redirected based on anti-GD2 antibodies. Finally, vaccines can be used to mount an immune response in patients. We review here the pertinent biological information on GD2 which may be of use for optimizing current immunotherapeutic strategies.

Targeted Delivery of Endosomal Escape Peptides to Enhance Immunotoxin Potency and Anti-cancer Efficacy

This work describes use of anti-carcinoembryonic antigen antibodies (10H6, T84.66) for targeted delivery of an endosomal escape peptide (H6CM18) and gelonin, a type I ribosome inactivating protein. The viability of colorectal cancer cells (LS174T, LoVo) was assessed following treatment with gelonin or gelonin immunotoxins, with or without co-treatment with T84.66-H6CM18. Fluorescent microscopy was used to visualize the escape of immunoconjugates from endosomes of treated cells, and efficacy and toxicity were assessed in vivo in xenograft tumor-bearing mice following single- and multiple-dose regimens. Application of 25 pM T84.66-H6CM18 combined with T84.66-gelonin increased gelonin potency by ~ 1,000-fold and by ~ 6,000-fold in LS174T and LoVo cells. Intravenous 10H6-gelonin at 1.0 mg/kg was well tolerated by LS174T tumor-bearing mice, while 10 and 25 mg/kg doses led to signs of toxicity. Single-dose administration of PBS, gelonin conjugated to T84.66 or 10H6, T84.66-H6CM18, or gelonin immunotoxins co-administered with T84.66-H6CM18 were evaluated. The combinations of T84.66-gelonin + 1.0 mg/kg T84.66-H6CM18 and 10H6-gelonin + 0.1 mg/kg T84.66-H6CM18 led to significant delays in LS174T growth. Use of a multiple-dose regimen allowed further anti-tumor effects, significantly extending median survival time by 33% and by 69%, for mice receiving 1 mg/kg 10H6-gelonin + 0.1 mg/kg T84.66-H6CM18 (p = 0.0072) and 1 mg/kg 10H6-gelonin + 1 mg/kg T84.66-H6CM18 (p = 0.0017). Combined administration of gelonin immunoconjugates with antibody-targeted endosomal escape peptides increased the delivery of gelonin to the cytoplasm of targeted cells, increased gelonin cell killing in vitro by 1,000-6,000 fold, and significantly increased in vivo efficacy.

Neuroblastoma Origin and Therapeutic Targets for Immunotherapy

Neuroblastoma is a pediatric solid cancer of heterogeneous clinical behavior. The unique features of this type of cancer frequently hamper the process of determining clinical presentation and predicting therapy effectiveness. The tumor can spontaneously regress without treatment or actively develop and give rise to metastases despite aggressive multimodal therapy. In recent years, immunotherapy has become one of the most promising approaches to the treatment of neuroblastoma. Still, only one drug for targeted immunotherapy of neuroblastoma, chimeric monoclonal GD2-specific antibodies, is used in the clinic today, and its application has significant limitations. In this regard, the development of effective and safe GD2-targeted immunotherapies and analysis of other potential molecular targets for the treatment of neuroblastoma represents an important and topical task. The review summarizes biological characteristics of the origin and development of neuroblastoma and outlines molecular markers of neuroblastoma and modern immunotherapy approaches directed towards these markers.

Botulinum neurotoxin type C protease induces apoptosis in differentiated human neuroblastoma cells

Neuroblastomas constitute a major cause of cancer-related deaths in young children. In recent years, a number of translation-inhibiting enzymes have been evaluated for killing neuroblastoma cells. Here we investigated the potential vulnerability of human neuroblastoma cells to protease activity derived from botulinum neurotoxin type C. We show that following retinoic acid treatment, human neuroblastoma cells, SiMa and SH-SY5Y, acquire a neuronal phenotype evidenced by axonal growth and expression of neuronal markers. Botulinum neurotoxin type C which cleaves neuron-specific SNAP25 and syntaxin1 caused apoptotic death only in differentiated neuroblastoma cells. Direct comparison of translation-inhibiting enzymes and the type C botulinum protease revealed one order higher cytotoxic potency of the latter suggesting a novel neuroblastoma-targeting pathway. Our mechanistic insights revealed that loss of ubiquitous SNAP23 due to differentiation coupled to SNAP25 cleavage due to botulinum activity may underlie the apoptotic death of human neuroblastoma cells.

Anti-GD2 mAbs and next-generation mAb-based agents for cancer therapy

Tumor-specific monoclonal antibodies (mAbs) have demonstrated efficacy in the clinic, becoming an important approach for cancer immunotherapy. Due to its limited expression on normal tissue, the GD2 disialogangloside expressed on neuroblastoma cells is an excellent candidate for mAb therapy. In 2015, dinutuximab (an anti-GD2 mAb) was approved by the US FDA and is currently used in a combination immunotherapeutic regimen for the treatment of children with high-risk neuroblastoma. Here, we review the extensive preclinical and clinical development of anti-GD2 mAbs and the different mechanisms by which they mediate tumor cell killing. In addition, we discuss different mAb-based strategies that capitalize on the targeting ability of anti-GD2 mAbs to potentially deliver, as monotherapy, or in combination with other treatments, improved antitumor efficacy.

Augmenting the Efficacy of Immunotoxins and Other Targeted Protein Toxins by Endosomal Escape Enhancers

The toxic moiety of almost all protein-based targeted toxins must enter the cytosol of the target cell to mediate its fatal effect. Although more than 500 targeted toxins have been investigated in the past decades, no antibody-targeted protein toxin has been approved for tumor therapeutic applications by the authorities to date. Missing efficacy can be attributed in many cases to insufficient endosomal escape and therefore subsequent lysosomal degradation of the endocytosed toxins. To overcome this drawback, many strategies have been described to weaken the membrane integrity of endosomes. This comprises the use of lysosomotropic amines, carboxylic ionophores, calcium channel antagonists, various cell-penetrating peptides of viral, bacterial, plant, animal, human and synthetic origin, other organic molecules and light-induced techniques. Although the efficacy of the targeted toxins was typically augmented in cell culture hundred or thousand fold, in exceptional cases more than million fold, the combination of several substances harbors new problems including additional side effects, loss of target specificity, difficulties to determine the therapeutic window and cell type-dependent variations. This review critically scrutinizes the chances and challenges of endosomal escape enhancers and their potential role in future developments.

Engineering anti-GD2 monoclonal antibodies for cancer immunotherapy

Ganglioside GD2 is highly expressed on neuroectoderm-derived tumors and sarcomas, including neuroblastoma, retinoblastoma, melanoma, small cell lung cancer, brain tumors, osteosarcoma, rhabdomyosarcoma, Ewing's sarcoma in children and adolescents, as well as liposarcoma, fibrosarcoma, leiomyosarcoma and other soft tissue sarcomas in adults. Since GD2 expression in normal tissues is restricted to the brain, which is inaccessible to circulating antibodies, and in selected peripheral nerves and melanocytes, it was deemed a suitable target for systemic tumor immunotherapy. Anti-GD2 antibodies have been actively tested in clinical trials for neuroblastoma for over the past two decades, with proven safety and efficacy. The main limitations have been acute pain toxicity associated with GD2 expression on peripheral nerve fibers and the inability of antibodies to treat bulky tumor. Several strategies have been developed to reduce pain toxicity, including bypassing complement activation, using blocking antibodies, or targeting of O-acetyl-GD2 derivative that is not expressed on peripheral nerves. To enhance anti-tumor efficacy, anti-GD2 monoclonal antibodies and fragments have been engineered into immunocytokines, immunotoxins, antibody drug conjugates, radiolabeled antibodies, targeted nanoparticles, T-cell engaging bispecific antibodies, and chimeric antigen receptors. The challenges of these approaches will be reviewed to build a perspective for next generation anti-GD2 therapeutics in cancer therapy.

In silico driven redesign of a clinically relevant antibody for the treatment of GD2 positive tumors

Ganglioside GD2 is a cell surface glycolipid that is highly expressed on cancer cells of neuroectodermal origin, including neuroblastoma, retinoblastoma, melanoma, sarcomas, brain tumors and small cell lung cancer. Monoclonal antibodies (MoAb) that target GD2 have shown clinical efficacy in the treatment of GD2 expressing tumors, and are expected to be the new standard of care for the treatment of pediatric neuroblastoma. In this study, the crystal structure of anti-GD2 murine MoAb 3F8 was solved to 1.65 Å resolution and used as a template for molecular docking simulations of its antigen, the penta-saccharide head group of GD2. Molecular docking revealed a binding motif composed of 12 key interacting amino acid side-chains, involving an extensive network of interactions involving main-chain and side-chain hydrogen bonding, two Pi-CH interactions, and an important charged interaction between Arg95 of the H3 loop with the penultimate sialic acid residue of GD2. Based on in silico scanning mutagenesis of the 12 interacting amino acids from the docked 3F8:GD2 model, a single point mutation (Heavy Chain: Gly54Ile) was engineered into a humanized 3F8 (hu3F8) MoAb and found to have a 6-9 fold enhancement in antibody-dependent cell-mediated cytotoxicity of neuroblastoma and melanoma cell lines. With enhanced tumor-killing properties, the re-engineered hu3F8 has the potential be a more effective antibody for the treatment of GD2-positive tumors.

Cell-targeting fusion constructs containing recombinant gelonin

Therapeutic agents capable of targeting tumor cells present as established tumors and micrometastases have already demonstrated their potential in clinical trials. Immunotoxins targeting hematological malignancies and solid tumors have additionally demonstrated excellent clinical activity. This review focuses on our design and characterization studies of constructs composed of recombinant gelonin toxin fused to either growth factors or single-chain antibodies targeting solid tumor cells, tumor vasculature or hematological malignancies. These agents demonstrate cytotoxicity at nanomolar or sub-nanomolar levels. All of these constructs display impressive selectivity and specificity for antigen-bearing target cells in vitro and in vivo and are excellent clinical trial candidates.

Destination 'lysosome': a target organelle for tumour cell killing?

Lysosomes and lysosome-related organelles constitute a system of acid compartments that interconnect the inside of the cell with the extracellular environment via endocytosis, phagocytosis and exocytosis. In recent decades it has been recognized that lysosomes are not just wastebaskets for disposal of unused cellular constituents, but that they are involved in several cellular processes such as post-translational maturation of proteins, degradation of receptors and extracellular release of active enzymes. By complementing the autophagic process, lysosomes actively contribute to the maintenance of cellular homeostasis. Proteolysis by lysosomal cathepsins has been shown to mediate the death signal of cytotoxic drugs and cytokines, as well as the activation of pro-survival factors. Secreted lysosomal cathepsins have been shown to degrade protein components of the extracellular matrix, thus contributing actively to its re-modelling in physiological and pathological processes. The malfunction of lysosomes can, therefore, impact on cell behaviour and fate. Here we review the role of lysosomal hydrolases in several aspects of the malignant phenotype including loss of cell growth control, altered regulation of cell death, acquisition of chemoresistance and of metastatic potential. Based on these observations, the lysosome is proposed as a potential target organelle for the chemotherapy of tumours. We will also present some recent data concerning the technologies for delivering chemotherapeutic drugs to the endosomal-lysosomal compartment and the strategies to improve their efficacy.

Immunotoxins for targeted cancer therapy

Immunotoxins constitute a new modality for the treatment of cancer, since they target cells displaying specific surface-receptors or antigens. Immunotoxins contain a ligand such as a growth factor, monoclonal antibody, or fragment of an antibody which is connected to a protein toxin. After the ligand subunit binds to the surface of the target cell, the molecule internalizes and the toxin kills the cell. Bacterial toxins which have been targeted to cancer cells include Pseudomonas exotoxin and diphtheria toxin, which are well suited to forming recombinant single-chain or double-chain fusion toxins. Plant toxins include ricin, abrin, pokeweed antiviral protein, saporin and gelonin, and have generally been connected to ligands by disulfide-bond chemistry. Immunotoxins have been produced to target hematologic malignancies and solid tumors via a wide variety of growth factor receptors and antigens. Challenges facing the clinical application of immunotoxins are discussed.

Superantigen-staphylococcal-enterotoxin-A-dependent and antibody-targeted lysis of GD2-positive neuroblastoma cells

Superantigens such as the staphylococcal enterotoxin A (SEA) are among the most potent T cell activators known. They bind to major histocompatibility complex (MHC) class II molecules and interact with T cells depending on their T cell receptor (TCR) V beta expression. Superantigens also induce a variety of cytokines and trigger a direct cytotoxic effect against MHC-class-II-positive target cells. In order to extend superantigen-dependent cell-mediated cytotoxicity (SDCC) to MHC-class-II-negative neuroblastoma cells, SEA was linked to the anti-ganglioside GD2 human/mouse chimeric monoclonal antibody (mAb) ch14.18. Ganglioside GD2 is expressed on most tumours of neuroectodermal origin but is expressed to a lesser extent on normal tissues. The linkage of ch14.18 to SEA was achieved either with a protein-A-SEA fusion protein or by chemical coupling. Both constructs induced T-cell-mediated cytotoxicity towards GD2-positive neuroblastoma cells in an effector-to-target(E:T)-ratio- and dose-dependent manner in vitro. To reduce the MHC class II affinity of SEA, a point mutation was introduced in the SEA gene (SEAm9) that resulted in 1000-fold less T cell killing of MHC-class-II-expressing cells as compared to native SEA. However, a protein-A-SEAm9 fusion protein mediated cytotoxicity similar to that of protein-A-SEA on ch14.18-coated, MHC-class-II-negative neuroblastoma cells. Taken together, these findings suggest that superantigen-dependent and monoclonal-antibody-targeted lysis may be a potent novel approach for neuroblastoma therapy.

An antimelanoma immunotoxin containing recombinant human tumor necrosis factor: tissue disposition, pharmacokinetic, and therapeutic studies in xenograft models

The ability of monoclonal antibody conjugates to re-direct plant or bacterial toxins, chemotherapeutic agents and radionuclides to selected target cells has been well-documented. Recombinant human tumor necrosis factor (TNF) is a macrophage-derived, non-glycosylated (17 kDa) peptide with a broad range of biological and immunological effects including antiviral activity, cytotoxic and cytostatic effects. A conjugate of the antimelanoma antibody ZME-018 and TNF in previous studies has shown melanoma-selective cytotoxic effects in vitro. Pharmacokinetic studies of the ZME-TNF immunotoxin showed that the agent cleared from plasma biphasically with alpha- and beta-phase half-lives similar to that of ZME itself (72 min and 36 h compared to 84 min and 41 h respectively). In contrast, TNF itself was cleared rapidly from plasma with a terminal-phase half-life of only 2.7 h. The clearance rate of ZME-TNF from plasma (Clp) was almost tenfold more rapid than for ZME (1.1 versus 0.16 ml/kg x min) but was threefold slower than the clearance for TNF itself (3.4 ml/kg x min). Tissue distribution studies in nude mice bearing human melanoma xenografts showed similar tumor localization of the immunotoxin compared to the free antibody and slightly higher concentrations in liver and kidney compared to ZME itself. Treatment of nude mice bearing well-developed A375 tumors with the immunotoxin resulted in a statistically significant (P < 0.002) suppression in tumor growth rate (fivefold increase) compared to saline-treated controls, which increased 20-fold over the same period. These studies demonstrate the feasibility of this approach and suggest that TNF may represent a non-antigenic alternative to immunotoxins containing plant and bacterial toxins.

In vitro and in vivo properties of an anti-CD5-momordin immunotoxin on normal and neoplastic T lymphocytes

An anti-CD5 monoclonal antibody (mAb) was linked to the plant toxin momordin, a type-1 ribosome-inactivating protein purified from Momordica charantia. The in vitro cytotoxicity of the immunotoxin was evaluated as the inhibition of protein and/or DNA synthesis on isolated peripheral blood mononuclear cells (PBMC) and on human T cell leukemia Jurkat. The potency of the immunotoxin on PBMC was very high (IC50 = 1 - 10 pM) and was not affected by blood components. The conjugate was also very efficient in the inhibition of the proliferative response in a mixed lymphocyte reaction (IC50 = 10 pM). Moreover, the in vitro performance of the immunotoxin compared favorably with those reported for other anti-CD5-based immunoconjugates containing ricin A chain. The in vivo activity of the immunotoxin was assessed in the model of nu/nu mice bearing Jurkat leukemia. A significant inhibition of the tumour development (80%, P < 0.01) in the animals treated with immunotoxin was observed. Taken together, the in vitro and in vivo results suggest that the anti-CD5-momordin conjugate may be useful for graft-versus-host disease therapy and potentially in the treatment of CD5-positive leukemias and lymphomas.

Antitumor effects of a novel monoclonal antibody with high binding affinity to ganglioside GD3

Ganglioside GD3, which is one of the major gangliosides expressed on the cell surface human tumors of neuroectodermal origin, has been studied as a target molecule for passive immunotherapy. We established ten kinds of anti-GD3 monoclonal antibodies (mAb) of the mouse IgG3 subclass by immunization with purified GD3 and melanoma cells. One of the established mAb, KM641, showed major reactivity with GD3 and minor reactivity with GQ1b out of 11 common gangliosides in an enzyme-linked immunosorbent assay. Immunostaining of gangliosides, separated on thin-layer chromatography plates, using KM641 revealed that most of the melanoma cell lines contained immunoreactive GD3 and GD3-lactone at a high level, but only the adrenal gland and the urinary bladder out of 21 human normal tissues had immunoreactive GD3. In immunofluorescence, KM641 bound to a variety of living tumor cell lines especially melanoma cells, including some cell lines to which another anti-GD3 mAb R24, established previously, failed to bind. High-affinity binding of KM641 to a tumor cell line was quantified by Scatchard analysis (Kd = 1.9 x 10(-8) M). KM641 exerted tumor-killing activity in the presence of effector cells or complement against melanoma cells expressing GD3 at a high level. Not only natural killer cells but also polymorphonuclear cells were effective as the effector cells in antibody-dependent cellular cytotoxicity. Intravenous injection of KM641 markedly suppressed the tumor growth of a slightly positive cell line, C24.22 (7.2 x 10(5) binding sites/cell), as well as a very GD3-positive cell line, G361 (1.9 x 10(7) binding sites/cell), inoculated intradermally in nude mice. KM641, characterized by a high binding affinity for GD3, has the potential to be a useful agent for passive immunotherapy of human cancer.