T cells stand at the apex of the immune surveillance system. T cells become activated after the T cell receptor (TCR) binds its cognate peptide in the context of cell surface proteins encoded by the Major Histocompatibility Complex (MHC) family of genes on an antigen presenting cell (APC). The requirement of recognizing both MHC and peptide makes TCR binding a more complex interaction than the relatively simple binding of B cell antibody to antigen. Once a TCR binds to its cognate peptide-MHC the T cell requires a second signal via costimulatory receptors in order to achieve full activation. Activated T cells then trigger cytokine release and a cascade of proliferation and differentiation begins. T cell responsiveness is thus limited in a way that B cell responsiveness is not: T cells only respond to peptides that have been processed and are presented on native MHCs, and require a second signal for full activation. Extracellular proteins, viruses or bacteria presented without processing or in the context of a foreign MHC complex do not elicit an immune response. Absence of a costimulatory signal results in T cell anergy.
The ability of T cells to detect non-self antigens and initiate an immune response led researchers to investigate whether T cells were able to detect and eliminate tumor cells. Experiments with mice lacking different components of the immune system illustrated that immune surveillance does play a role in detecting and eliminating cancer (reviewed in Swann, 2007); however cancers frequently evade immune detection. Evasion is achieved via several
different mechanisms. First, cancers may fail to elicit an immune response because there are no “non-self” peptides or proteins presented to T and B cells. However, cancer cells do upregulate the expression of some stress-related genes that induce a T cell response. If such a response fails to completely eliminate the malignant cells, surviving cells may then evade immune detection by downregulating the expression of MHC proteins on the surface of the cell or slowing the
proteolytic process that generates peptides for display in MHC proteins. Mouse models involving cancer cells that have been engineered to present a foreign peptide or protein on the cell surface have demonstrated that the immune system is efficient at eradicating cancer cells if it can identify them as non-self (Swann, 2007). With this noted, researchers began to examine alternate ways to trigger an immune response to cancer in humans.
CARs are chimeric antigen receptors that target specific (generally native) antigens. The first generation of CAR-T cells were composed of a single chain variable fragment (scFv) from an antibody targeting the antigen of interest fused with the activating domains of CD3ζ or Fc receptor γ. These first generation CARs were able to induce a cytolytic response, but were unable to produce cytokines or undergo expansion since they lacked a secondary activation signal (reviewed in Sadelain, 2013). Subsequent generations of CAR design incorporated co-stimulatory molecules along with the activating domains to produce the full range of T cell responses. These second generation CARs are comprised of the CD3ζ activation domain fused to the cytoplasmic domain of a co-stimulatory molecule such as CD28, 4-1BB, DAP10, OX40, or ICOS. Second generation CAR-T cells are capable of cytokine secretion and continued expansion in the face of repeated exposure to antigens (Sadelain, 1998). Currently researchers are investigating the efficacy of “triple decker” CAR-T cells that couple two costimulatory receptor cytoplasmic domains with the activating domain of CD3ζ (reviewed in Sadelain, 2013)