Antibody-CD1d bifunctional fusion protein for targeting innate immunity to cancer cells
Invariant natural killer T lymphocytes (iNKT) are a unique subpopulation of T lymphocytes recognizing glycolipid antigens in the context of the MHC class I-like monomorphic molecule CD1d. Upon activation, iNKT cells are known to induce in turn a potent activation of cells of the innate and adaptive immune response, such as NK cells, dendritic cells (DCs), as well as T and B lymphocytes. In this context, iNKT cells have been shown to efficiently mediate antitumor activity. Thus, recent research has focused on the manipulation of these cells for antitumor therapies. Alpha-galactosylceramide (αGalCer) is a synthetic high affinity CD1d ligand used for activation of iNKT cells, and has been tested for therapy of cancer patients in clinical trials. However, a major drawback of αGalCer, as a free drug, is that a single injection leads to a short-lived iNKT cell activation followed by a long-term anergy, limiting its therapeutic use.
In contrast, we have recently (Stirnemann et al. 2008) shown that, when αGalCer was loaded on a recombinant soluble CD1d molecule (αGalCer/sCD1d), repeated injections led to a sustained iNKT and NK cell activation associated with interferon γ secretion, as well as with DC maturation. Most importantly, when the αGalCer/sCD1d was fused to an anti-HER2 scFv antibody fragment, potent inhibition of experimental lung metastasis and established subcutaneous tumors was obtained, when systemic treatment was started 7 days after tumor graft, whereas at this time, free αGalCer had no effect. The anti-tumor activity of the CD1d-anti-HER2 fusion protein was associated with its HER2-specific tumor localization and with the accumulation of iNKT, NK and T cells at the tumor site.
These results suggest the enormous potential of targeting these recombinant CD1d proteins to tumors for effective immunotherapy approaches. The aim of this project is to continue the preclinical development of this tool with a view to its future clinical development in the context of current immunotherapy cancer programs underway in the Lausanne environment.
Potential of the sustained iNKT and NK cell activation in promoting active antitumor immunization. The characterization of tumor antigens has promoted the development of cancer vaccines. However, despite good antigen-specific T cell responses, clinical results have remained limited, largely due to T cell tolerance at the tumor site. We have now started collaboration with Prof. Pedro Romero at the LICR to combine sCD1d-antitumor therapy with active immunization against melanoma antigens, such as TRP2 or GP100.
Understand the mechanism of the sustained iNKT cell activation that is obtained by repeated injections of recombinant αGalCer-loaded sCD1d molecules, in contrast to the short-lived activation induced by treatment with αGalCer alone, which is followed by long-term unresponsiveness. Interestingly, it was recently shown that the co-inhibitory receptor Program death-1 (PD-1) was mainly responsible for iNKT anergy. In contrast, our recent results indicate that iNKT cells activation by recombinant αGalCer-loaded CD1d proteins prevent PD-1 upregulation resulting in the prolonged reactivity of these cells.
Immunosuppressive mechanisms at the tumor site. Multiple roles have been ascribed to T regulatory cells (Tregs) and to Myeloid Derived Suppressor Cells (MDSCs) in repressing antitumor immunity and promoting tumor growth, and these two cell populations are often enriched in tumor tissue. Two mouse models, that have been developed to selectively deplete these cell populations, will be tested to evaluate their respective roles in limiting CD1d-mediated therapy:
- DEREG mice: this transgenic mouse line expresses the diphteria toxin receptor under the control of the FoxP3 promoter. Administration of diphteria toxin allows the transient depletion of Tregs cells. This model will be tested in collaboration with Pr. A. Knuth and Dr. Maries van den Broek, Oncology Department, University Hospital Zurich. It will allow evaluating the limiting role of Tregs on the efficacy of the CD1d treatment.
- Arginase 1 knock out mice within the myeloid and endothelial lineages have recently been developed by specific deletion under the Tie2 Cre promoter (Maastricht University). Arginase 1 activity is one of the major mechanisms of MDSCs to induce T cell immunosuppression. This mouse model will allow evaluating the inhibitory role of MDSCs on the iNKT/NK-mediated CD1d immunotherapy.
Combine sCD1d-antitumor therapy with low dose chemotherapy. Recent studies have shown that low dose chemotherapy had immunostimulatory properties, and some agents were shown to act against T regs and MDSCs. Our preliminary data demonstrate a potent synergy between iNKT-mediated antitumor therapy and low dose cyclophosphamide (CY).
Newly developed sCD1d-antitumor fusion proteins. So far, two sCD1d-antitumor scFv fusion proteins were developed and shown to have antitumor activity, respectively sCD1d-anti-HER2 (Stirnemann et al. 2008) and sCD1d-anti-CEA fusion proteins (manuscript in prep). Alternatively to the targeting of CD1d to tumor antigens, antigens expressed by the tumor stroma or by tumor neo-vessels are considered. In this context, a fusion protein of b2m-sCD1d-anti-VEGFR3 was developed and shown to bind efficiently on VEGFR-3-expressing cells. This approach may allow redirecting the iNKT-mediated immune response inside various tumors, whose neo-vessels are known to express VEGFR-3, independently of their own phenotype.
Tumor models. The different studies are performed in various tumor models:
Grafted tumors: subcutaneous or i.v. tumor grafts (melanoma, colon carcinoma)
Spontaneous tumors: transgenic model of Tyr::N-Ras autochtonous melanomas