Activation of cytomegalovirus-specific CD8+ T-cell response by antibody-mediated peptide-major histocompatibility class I complexes.

Imposing antigenicity on tumor cells is a key step toward successful cancer-immunotherapy. A cytomegalovirus-derived peptide recombinantly fused to a major histocompatibility class I complex and a monoclonal antibody can be targeted to tumor cells by antibody-mediated delivery and activate a strong and specific CD8+ T cell response.

Introduction

Cancer-immunotherapy holds promise for becoming a breakthrough treatment of advanced tumors. T lymphocytes are the key mediators of tumor rejection but can be subjugated by an immunosuppressive environment. Furthermore, lack or loss of antigenicity often prevents elimination of cancer by specific T lymphocytes. To address these problems, bispecific T-cell engagers (BiTEs), adoptive T-cell therapy and immunomodulators were recently developed for clinical use.

BiTEs, which indiscriminately activate T cells through binding to the CD3 component of the T-cell receptor (TCR) complex and a tumor-specific antigen, have been approved for EpCAM (catumaxomab) and CD19 (blinatumomab). Treatment with BiTEs causes considerable toxicity such as neutropenia, anemia, neurologic effects and in some cases cytokine release syndrome, necessitating premedication with glucocorticoids along with a stepwise dose increase or even interruption of treatment in a proportion of patients with neurologic side effects. Due to the progress in antibody engineering, the field is moving from non-Fc-based T-cell bispecifics such as BiTE or Tandab formats to more complex, Fc-based, IgG-derived T-cell bispecifics with improved antigen binding and increased half-life. However, to date all T-cell engagers are directed against the CD3ε TCR subunit and therefore trigger polyclonal activation of T cells irrespective of their subtype and specificity.

Adoptive T–cell transfer therapy has made considerable progress in the past ten years both for naturally occurring and genetically engineered lymphocytes. It has been tested only in small clinical trials, yet, but appears to have the potential of becoming a curative treatment for some advanced-stage cancers. The therapy regimen is complex as it requires ex vivo expansion and reinfusion of the tumor-specific T cells and may also involve their genetic modification as well as non-myeloablative chemotherapy as preconditioning (lymphodepletion). In addition, it may be accompanied by severe adverse effects such as cytokine-release syndrome, neurological dysfunction requiring hospitalization and even intensive care support which may limit a broader clinical application.

Immunomodulators such as antibodies recognizing checkpoint inhibitory molecules like cytotoxic T-lymphocyte antigen-4 (CTLA-4) and programmed cell death 1 (PD-1) or its ligand PD-L1 have demonstrated clinical successes with favorable safety profiles in treatment of some solid malignancies such as melanoma, non-small cell lung carcinoma and renal cell carcinoma. However, the observed response rates remain below 30% for anti-PD-1 therapy in unselected cancer patients. Interestingly, cancers with higher somatic mutation rates appear to respond best to immune checkpoint blockage. Most likely, a high mutational load produces neoantigens thus increasing the antigenicity of tumor cells which otherwise express only a limited number of poorly immunogenic self-antigens. Some of these neoepitopes obviously share homology with viral and bacterial antigens which may indicate that not only the number but also the nature of the mutations in a tumor triggers their recognition by T cells.

Imposing CMV-antigenicity to Tumor Cells

We believe that in order to increase the success rate of cancer immunotherapy novel approaches allowing controlled manipulation of tumor cell antigenicity need to be developed. Pursuing this goal, we have recently introduced a new technology for selective delivery of a cytomegalovirus (CMV)-derived peptide-major histocompatibility class I complex (pMHCI) to tumor cells. Chronic CMV infection affects the vast majority of humans and results in generation of constantly renewing, antigen-specific and differentiated cytotoxic effector T lymphocytes persisting both in the blood and various organs at high frequencies. CMV-specific CD8+ T-cell responses are mainly focused on a few immunodominant peptides and a single recombinant pMHCI-IgG fusion is sufficient to redirect a large proportion of CMV-specific T lymphocytes against CMV-negative tumor cells expressing the chosen cell surface target (). Following exposure to pMHCI of relevant specificity, tumor antigen–expressing cancer cells are decorated with fusion proteins composed of a complete tumor antigen–specific antibody connected to a single MHC class I:peptide complex bearing a covalently linked CMV-derived peptide (pMHCI–IgG). The tumor cells can be specifically eliminated in vitro through engagement of antigen-specific CD8+ T cells from peripheral blood mononuclear cell preparations of CMV-infected humans independently of the level of endogenous MHC class I expression on the target. Thus, the paradigm of immune-mediated tumor eradication can be extended even to tumor variants characterized by total loss of MHC expression, which is frequently observed in a sizable proportion of different tumors. Activation of CMV-specific T cells requires surprisingly low pMHCI–IgG concentrations without additional expansion, pre-activation, or provision of T-cell co-stimulatory signals. Our favored molecular format possesses a number of advantageous features related to protein production, stability, IgG-like pharmacokinetics and antigen-binding properties. Due to a single pMHCI complex per molecule and low pMHC-I:TCR binding affinity, target-independent activation of T cells and peripheral sink should not interfere with efficient in vivo tumor targeting. In contrast to pan-T-cell recruiters, application of pMHCI-IgGs is HLA-allotype restricted that limits the patient cohort to 30–40% of the population in the case of HLA A*0201. However, it remains to be seen how pMHCI-IgGs compare to conventional T-cell engagers when it comes to safety and the type of activation/death programs induced in T cells in vivo. In a side by side comparison with BiTEs, we found that pMHCI-IgGs induce reduced secretion of cytokines despite comparable tumor cell killing in vitro. We believe that dressing up tumor cells with CMV-peptide MHCI complexes and subsequent engagement of virus-specific CD8+ T cell subpopulation will be advantageous in clinical settings.

Figure 1.

Mechanism of action for tumor-targeted CMV-peptide: MHC class I-antibody fusion proteins (pMHCI-IgG). CMV-pMHCI-IgG selectively recruit CMV-specific CD8+ T cells. Upon crosslinking of the antibody on antigen-specific tumor cells virus-specific T cells mediate cell lysis.