Generation Of Tumor Antigen Epitopes

Multiple transcriptional, translational, and post-translational mechanisms have been shown to be involved with generating the epitopes recognized by tumor reactive T cells. Intronic sequences that have been retained in a small percentage of the processed mRNA transcripts derived from certain genes appear to encode T cell epitopes. These include a mutated sequence derived from the intronic region of a gene that was termed MUM-1 (2), as well as non-mutated intronic sequences derived from the gp100 (81) and TRP-2 (46). Normal human melanocytes did not appear to express the intronic TRP-2 transcript at detectable levels in vivo, indicating that this actually represented a tumor specific antigen.

Several tumor antigens have been shown to represent the products of alternative open reading frames (ORFs), which in the majority of cases result from the initiation of translation at a methionine codon located downstream from the normal initiation codon. These include an HLA-A31 restricted epitope of TRP-2 (3), HLA-A2 restricted (4) and HLA-DR* 11 and 12 restricted epitopes of LAGE-1 (51), and an HLA-A2 restricted epitope of the BING-4 molecule (82). In addition, nucleotide deletions resulting from mutations within the CDKN2A locus, which encodes 2 products that share a single exon 2 but that are translated in 2 open reading frames termed p14ARF and p16INK4a, have recently also been shown to result in the generation of a T cell epitope (60). A mutated p14ARF transcript containing a single nucleotide deletion in exon 2 gave rise to a product that was translated in the third alternative open reading frame (AORF), and a T cell epitope that was recognized by an HLA-A11 restricted, tumor reactive TIL was present within this AORF. A melanoma that expressed a p16INK4a transcript containing a 2 base pair exon 2 deletion that resulted in the translation of the this AORF was also recognized by the HLA-A11 restricted TIL. Frame-shifted CDKN2A products are expressed at high frequencies in certain tumor types (83, 84), and thus this represents a highly shared mutated epitope that could be utilized as a target for immunotherapy in patients bearing certain cancers.

The proteasome plays an important role in protein processing steps that leads to the generation of MHC class I-binding peptides (85). The results of detailed analysis of processing suggest that many of the peptides that are produced are rapidly degraded by the proteasome (86), while at the same time, the proteasome plays an important role in generating the proper carboxy terminus of processed HLA class I restricted T cell epitopes. Thus, only a small percentage of candidate peptides from a given protein that are capable of binding exogenously to a particular HLA allele appear to be naturally processed endogenously (87, 88).

The importance of proteasomal processing mechanisms in generating the epitopes recognized by tumor reactive T cells has been demonstrated by several studies. The product of a short ORF of 11 amino acids encoded by a pseudogene that was similar to a portion of the 3' untranslated region of the homeoprotein HPX42B was recognized by HLA-B*1302 restricted, tumor reactive T cells (89). Recognition of the pseudogene was critically dependent on the presence of a stop codon in the pseudogene following the codon that encoded the carboxy terminal residue in the peptide epitope, which would generate a partially processed epitope. Similar mechanisms may also enhance the recognition of a subset of the tumor antigens encoded by short alternative ORFs (82, 90).

As mentioned above, some studies have failed to confirm initial reports indicating that T cells stimulated with certain candidate epitopes were capable of recognizing tumor cell targets. Initial reports indicated that tumor reactive T cells were generated following in vitro stimulation with an HLA-A2 restricted candidate epitope identified from MAGE-3, MAGE-3:271-279 (40). Subsequent studies indicated that T cells generated with this peptide failed to recognize tumor cell targets, and that incubation of long peptides that encompassed this epitope with purified proteasomes failed to generate the proper peptide carboxy terminus (41). Cultures of peptide reactive T cells recognized tumor cells that had been treated with the proteasomal inhibitor lactacystin, however, indicating that peptide bonds within this peptide sequence are normally cleaved by the proteasome, resulting in the inefficient recognition of un-manipulated cells. Initial results indicated that T

cells that were generated by stimulating PBMC in vitro with an HLA-A2 binding peptide from the human telomerase catalytic subunit hTERT, hTERT:540-548, recognized a variety of tumor cell targets (91). Subsequent studies failed to confirm the recognition of tumor targets and indicated that the peptide could not be generated by incubation of long peptide precursors with purified proteasomes (52).

Several recent studies have focused on identifying candidate T cell epitopes that can be efficiently processed by the proteasome. In one study, long peptides that encompassed candidate HLA-A2 restricted T cell epitopes derived from PRAME were incubated with purified proteasomes (92). The analysis of the peptide digests revealed that the appropriate carboxy terminal residue was generated from only 4 out of the 19 high affinity peptides that were subjected to degradation using this procedure. The T cells that were generated following in vitro sensitization with peptides that were identified using this approach appeared to efficiently recognize tumor cells that expressed PRAME and HLA-A2, indicating that this process might facilitate the identification of relevant peptide vaccine target epitopes. An epitope of the SSX-2 antigen that was initially identified by SEREX analysis was identified by incubation of long peptides with purified proteasomes (16).

Expression of the immunoproteasome, which has a distinct peptide cleavage specificity from the proteasome and which can be induced in a variety of cells by stimulation with interferon gamma (IFN-y), may also affect the processing of certain T cell epitopes. The treatment of tumor cells with IFN-y nearly eliminated the ability of those cells to stimulate T cells that recognize multiple HLA class I restricted T cell epitopes, including the gp100:209-217 and MART-1:26-35 peptides (93), while at the same time, the treated tumor cells did not demonstrate a diminished capacity to stimulate T cells that recognized additional epitopes. These findings are consistent with the altered cleavage specificity of the immunoproteasome. In contrast, IFN-y treated, but not un-treated melanoma cells were capable of stimulating T cells that recognized a class I HLA-B40 restricted MAGE-3 epitope (94). Un-treated tumor cells that were transfected with the catalytic LMP7 subunit, which is only expressed in the immunoproteasome, were also recognized by the MAGE-3 reactive T cells, providing further evidence that an alteration in antigen processing was responsible for these findings.

Intracellular protein targeting mechanisms may influence the ability of T cells to recognize certain epitopes. A point mutation in the CDC27 transcript that was not directly involved with generating the HLA class II restricted T cell epitope resulted in re-localization of this nuclear protein to the cytoplasm, thereby presumably resulting in enhanced processing of a distal T cell epitope (7). A similar mechanism may be involved with the processing of a mutated HLA class II restricted epitope derived from neo-poly(A) polymerase (95).

Post-translational processing has been shown to alter certain epitopes recognized by tumor reactive T cells. The HLA-A2 restricted tyrosinase epitope tyr:369-377 contains an asparagine residue at position 371 that represents an N-linked glycosylation site (31). HLA-A2 positive target cells that were pulsed with a peptide containing a substitution of an aspartic acid residue for asparagine residue, however, were recognized at significantly lower concentrations than target cells pulsed with the peptide that was encoded by the normal tyrosinase gene. A mammalian enzyme that removes oligosaccharide side chains from glycoproteins, PNGase (96), may be responsible for conversion of the amino acid sequence of the tyrosinase peptide. Subsequent studies indicated that tyrosinase molecules that are initially generated in the endoplasmic reticulum are subsequently transported into the cytoplasm where processing resulting in the generation of peptides that can then be transported back into the endoplasmic reticulum where they associate with HLA class I molecules (97).

Recent observations suggest that a process that has previously not been described in mammalian cells, protein splicing, is also responsible for generating T cell epitopes recognized by tumor reactive T cells. An epitope of the FGF-5 protein that was recognized by the HLA-A3 restricted T cells reactive with renal carcinomas resulted from the joining of sequences of 5 and 4 amino acids that were separated by 40 amino acids in the primary FGF-5 protein (54). Mechanisms such as alternative mRNA splicing or ribosomal skipping did not appear to be involved with generating this epitope, indicating that this peptide was derived by splicing of the FGF-5 protein. Two 3 and 6 amino acid peptides that were separated by a 4 amino acids in the gp100 protein appeared to be spliced to yield an epitope that was recognized in the context of HLA-A32 (6). Incubation of a 13 amino acid precursor with purified proteasomes resulted in the generation of the HLA-A32 restricted T cell epitope, indicating that peptide processing within this cellular compartment was capable of generating a spliced T cell epitope. Protein splicing also appeared to be involved with generation of a tyrosinase epitope that was recognized in the context of HLA-A24 (Robbins et al., manuscript in preparation). The tyrosinase epitope resulted from the linkage of a 5 amino acid peptide corresponding to residues 368 to 372 to a 4 amino acid peptide corresponding to residues 335 to 340. This T cell epitope appears to have result from splicing as well as re-ordering of the primary tyrosinase amino acid sequence. The HLA-A24 restricted tyrosinase epitope also contained 2 asparagines corresponding to residues 337 and 371 that represent N-linked glycosylation sites that appear to be converted to aspartic acid residues during the processing of the T cell epitope, as described for the partially overlapping HLA-A2 restricted tyrosinase epitope (31).

How To Prevent Skin Cancer

How To Prevent Skin Cancer

Complete Guide to Preventing Skin Cancer. We all know enough to fear the name, just as we do the words tumor and malignant. But apart from that, most of us know very little at all about cancer, especially skin cancer in itself. If I were to ask you to tell me about skin cancer right now, what would you say? Apart from the fact that its a cancer on the skin, that is.

Get My Free Ebook


Post a comment