Specificity in cancer immunotherapy

Abstract

From the earliest days in the field of tumor immunology three questions have been asked: do cancer cells express tumor-specific antigens, does the immune system recognize these antigens and if so, what is their biochemical nature? We now know that truly tumor-specific antigens exist, that they are caused by somatic mutations, and that these antigens can induce both humoral and cell-mediated immune responses. Because tumor-specific antigens are exclusively expressed by the cancer cell and are often crucial for tumorigenicity, they are ideal targets for anti-cancer immunotherapy. Nevertheless, the antigens that are targeted today by anti-tumor immunotherapy are not tumor-specific antigens, but antigens that are normal molecules also expressed by normal tissues (so-called "tumor-associated" antigens). If tumor-specific antigens exist and are ideal targets for immunotherapy, why are they not being targeted? In this review, we summarize current knowledge of tumor-specific antigens: their identification, immunological relevance and clinical use. We discuss novel tumor-specific epitopes and propose new approaches that could improve the success of cancer immunotherapy, especially for the treatment of solid tumors.

Figure 1

Biosynthesis of N- and O-glycans. N-glycosylation occurs in three major steps in the ER and the Golgi resulting in long, branched sugar chains. O-glycosylation occurs by the stepwise addition of monosaccharides and begins with the linkage of N-acetylgalactosamine (GalNAc) to Ser or Thr. Further elongation leads to the formation of so-called Core structures (Core 1–8). Adapted from A. Varki et al., Essentials of Glycobiology, 1999.

Figure 2

Tumor-associated carbohydrate antigens (TACA). Shown are the most common carbohydrate structures known to be overexpressed by malignanies: Tn, T (TF, Core1), Sialyl-Tn, Lewis^a, Sialyl-Lewis^a, Lewis^x, Sialyl-Lewis^x (right panel). Left panel shows legend of carbohydrate symbols.

Figure 3

A somatic mutation in the chaperone Cosmc leads to the creation of a tumor-specific, glycopeptidic neo-epitope on a wild-type protein. Mutation of the chaperone gene, Cosmc, abolished the activity of the Core 1 β1,3 galactosyltransferase needed for the linkage between N-actelylgalactosamine (GalNAc) and Galactose (Gal) (= Core 1). This disrupted O-glycan Core 1 synthesis, leaving only GalNAc attached to the Thr residue. The novel combination of the monosaccharide GalNAc with a 10 amino acid wild-type protein sequence RGTKPPLEEL from the transmembrane protein OTS8 creates a tumor-specific antigen.

Figure 4

Pathways of tumor-specific antigen creation. On the left is shown the previous paradigm for tumor-specific antigen creation. On the right is the new pathway; in which a mutation in an upstream gene causes aberrant post-translational modification and converts a wild-type protein into a tumor-specific antigen.