Cancer Immunotherapy: Whence and Whither

Abstract

The current concepts and practice of cancer immunotherapy evolved from classical experiments that distinguished "self" from "non-self" and the finding that humoral immunity is complemented by cellular immunity. Elucidation of the biology underlying immune checkpoints and interactions between ligands and ligand receptors that govern the immune system's ability to recognize tumor cells as foreign has led to the emergence of new strategies that mobilize the immune system to reverse this apparent tolerance. Some of these approaches have led to new therapies such as the use of mAbs to interfere with the immune checkpoint. Others have exploited molecular technologies to reengineer a subset of T cells to directly engage and kill tumor cells, particularly those of B-cell malignancies. However, before immunotherapy can become a more effective method of cancer care, there are many challenges that remain to be addressed and hurdles to overcome. Included are manipulation of tumor microenvironment (TME) to enhance T effector cell infiltration and access to the tumor, augmentation of tumor MHC expression for adequate presentation of tumor associated antigens, regulation of cytokines and their potential adverse effects, and reduced risk of secondary malignancies as a consequence of mutations generated by the various forms of genetic engineering of immune cells. Despite these challenges, the future of immunotherapy as a standard anticancer therapy is encouraging. Mol Cancer Res; 15(6); 635-50. ©2017 AACR.

Figure 1. Complexities of cell-cell interactions and microenvironment in T cell activation and inhibition

Four cell types are depicted: T cell, NK cell, APC or a tumor cell transduced with a construct expressing CD80. Several other cell types, including regulatory T cells (T_regs), myeloid derived suppressor cells (MDSCs) tumor associated fibroblasts (TAFs) and tumor-associated macrophages (TAMs) that would normally appear in a tumor microenvironment are not shown. When a tumor cell is transduced with a CD80 construct (upper cell) the ectopically expressed CD80, in the context of MHC/antigen complex engagement of the T cell receptor (TCR), can engage CD28 on a T_eff cell to activate the T cell and cause it to become cytolytic. TCRs have an immunoglobulin-like heterodimeric structure with α and β chains containing variable (V) and constant (C) regions, but with an anchoring transmembrane domain. Associated with the TCR is the CD3 signaling molecule comprised of CD3γ/CD3ε and CD3δ/CD3ε dimers and a dimeric CD3ζ chain. Close to the carboxyl terminus of each CD3 ε, γ and δ subunit is an immunoreceptor tyrosine-based activation motif (ITAM) marked by a short black bar. The CD3ζ subunit has three such ITAMs. In addition to T cell activation as a consequence of direct interaction between the TCR and antigen-associated MHC and the CD80/CD86 and CD28 interaction, cytokines produced by NK cells, APCs, dendritic cells and T cells can act on T and NK cells in a paracrine or autocrine fashion.

Figure 2. Interactions between tumor cells and T cells that activate or inhibit T cells

The upper panel shows interactions between T cell surface markers PD1 and tumor cell ligands, PD-L1 and PD-L2 that inhibit T_eff cell activation. There is potential interaction with PD-L2 and an unknown receptor that requires validation. CD80 and CD86 can both engage with CD28 with different affinities and with subtly different T cell activating outcomes. They both can also interact with CTLA-4 in an inhibitory capacity. The lower panel shows that antibodies that interrupt the engagement of these surface molecules can reverse their activating or inhibitory functions.

Figure 3. The evolution of CARs

A soluble immunoglobulin molecule is depicted in the upper left of the figure with light chain and heavy chain variable regions highlighted. Three generations of CARs are illustrated showing a single chain fragment comprised of a heavy chain variable region linked by a flexible linker to a light chain variable region. Both variable antibody fragments are derived from a monoclonal antibody and are coupled to a hinge region bound to a transmembrane fragment. Alternatively a ligand or ligand fragment for a receptor that is overexpressed on tumor cells (e.g. EGFR family in epithelial malignancies) can substitute for the targeting antibody fragment in the CAR construct. In the first generation CARs, antigen recognition (antibody or ligand) activates intracellular signals generated by CD3 ζ component of the CAR derived from the TCR. To increase efficacy of activation and persistence of the activated state, the second generation CARs incorporate a co-stimulatory signal such as CD28. The third generation CARs incorporate two co-stimulatory signals (e.g. CD28 and 4-1BB) for even greater stimulation of the engineered T cells encountering their targeted antigen.