Manipulation of avidity to improve effectiveness of adoptively transferred CD8(+) T cells for melanoma immunotherapy in human MHC class I-transgenic mice.

The adoptive transfer of tumor-reactive CD8(+) T cells into tumor-bearing hosts provides an attractive alternative to vaccination-based active immunotherapy of melanoma. The development of techniques that result in the preferential expansion of tumor-reactive T cells is therefore of great importance. In this study, we report the generation of HLA-A*0201-restricted CD8(+) T cell populations that recognize either tyrosinase(369-376) or gp100(209-217) from tolerant human class I MHC-transgenic mice by using single amino acid-substituted variant peptides. Low peptide concentration or restimulation with the parent peptide was used to enhance the functional avidity, defined by stimulation of IFN-gamma accumulation, and cross-reactivity of the resulting T cell populations. We found a direct correlation between the ability of a T cell population to respond in vitro to low concentrations of the precise peptide expressed on the tumor and its ability to delay the outgrowth of B16 melanoma after adoptive transfer. Surprisingly, we found that some T cells that exhibited high functional avidity and were effective in controlling tumor outgrowth exhibited low structural avidity, as judged by MHC-tetramer staining. Our results establish strategies for the development and selection of CD8(+) T cell populations that persist despite peripheral tolerance, and that can control melanoma outgrowth. Furthermore, they support the use of human MHC class I-transgenic mice as a preclinical model for developing effective immunotherapies that can be rapidly extended into therapeutic settings.

Immune responses to the HLA-A*0201-restricted epitopes of tyrosinase and glycoprotein 100 enable control of melanoma outgrowth in HLA-A*0201-transgenic mice.

Many of the Ags recognized by human melanoma-reactive CTL are derived from proteins that are also expressed in melanocytes. The possibility of self-tolerance to these epitopes has led to questions about their utility for antitumor immunotherapy. To investigate the issue, we established a preclinical model based on transgenic mice expressing a recombinant HLA-A*0201 molecule and B16 melanoma transfected to express this molecule. HLA-A*0201-restricted epitopes from the melanocyte differentiation proteins (MDP) tyrosinase and gp100 are expressed in both tumor cells and melanocytes, and the former is associated with self-tolerance. However, adoptive transfer of tyrosinase or gp100-reactive CTL developed from tolerant mice delayed tumor outgrowth, as did immunization with MDP peptide-pulsed dendritic cells. Protection was enhanced by the use of peptide ligands containing conservative substitutions that were cross-reactive with the original Ags. These data establish that CTL populations reactive against MDP-derived self-Ags can be activated to mount effective antitumor immunity and strongly support their continued development for tumor immunotherapy in humans.

Direct identification of human tumor-associated peptide antigens and a preclinical model to evaluate their use.

Although the arsenal of a healthy immune system includes both circulating antibodies and cellular components such as T cells, the latter seem to be particularly important in tumor immunology. Under normal conditions, the immune system does not react to the body's cells, which may be described as expressing "self" antigens on the cell surface. When a cell becomes cancerous, however, novel antigens are expressed on the cell surface. These novel "tumor" antigens are recognized as foreign by the body's immune system, and the cells that express them are destroyed or incapacitated. Whereas antibodies may react directly with protein antigens, T cells instead recognize peptide antigens presented by class I and class II molecules of the major histocompatibility complex (MHC). All cells normally break down proteins that they have made. The class I antigen-processing pathway has evolved to display peptides produced by this breakdown process as a way to provide information to cytotoxic T cells about what the cell is making. The display of new peptides as a result of infection or transformation can stimulate cytotoxic T cells to kill the cell. In addition, antigen-processing cells such as dendritic cells engulf dead or dying cells and degradeproteins into peptide fragments. These peptides are then displayed by the MHC class II molecules and presented to T helper cells, which augment the activity of the cytotoxic T cells. Cytotoxic T lymphocytes have recently been isolated from human tumors (especially melanoma) and are critical to the development of promising immunotherapeutic agents. As we shall discuss, these cells can recognize antigens that are common to tumors from different patients. We shall also explore how advances in instrumentation and the use of transgenic mice have increased our understanding of tumor-associated peptides to the point where we can begin to strive for a peptide-based therapeutic vaccine. The caveats for such therapy will also be addressed.

Self-tolerance to the murine homologue of a tyrosinase-derived melanoma antigen: implications for tumor immunotherapy.

The human tyrosinase-derived peptide YMDGTMSQV is presented on the surface of human histocompatibility leukocyte antigen (HLA)-A*0201(+) melanomas and has been suggested to be a tumor antigen despite the fact that tyrosinase is also expressed in melanocytes. To gain information about immunoreactivity and self-tolerance to this antigen, we established a model using the murine tyrosinase-derived homologue of this peptide FMDGTMSQV, together with transgenic mice expressing the HLA-A*0201 recombinant molecule AAD. The murine peptide was processed and presented by AAD similarly to its human counterpart. After immunization with recombinant vaccinia virus encoding murine tyrosinase, we detected a robust AAD-restricted cytotoxic T lymphocyte (CTL) response to FMDGTMSQV in AAD transgenic mice in which the entire tyrosinase gene had been deleted by a radiation-induced mutation. A residual response was observed in the AAD(+)tyrosinase(+) mice after activation under certain conditions. At least some of these residual CTLs in AAD(+)tyrosinase(+) mice were of high avidity and induced vitiligo upon adoptive transfer into AAD(+)tyrosinase(+) hosts. Collectively, these data suggest that FMDGTMSQV is naturally processed and presented in vivo, and that this presentation leads to substantial but incomplete self-tolerance. The relevance of this model to an understanding of the human immune response to tyrosinase is discussed.

The density of peptides displayed by dendritic cells affects immune responses to human tyrosinase and gp100 in HLA-A2 transgenic mice.

Several HLA-A*0201-restricted peptide epitopes that can be used as targets for active immunotherapy have been identified within melanocyte differentiation proteins. However, uncertainty exists as to the most effective way to elicit CD8+ T cells with these epitopes in vivo. We report the use of transgenic mice expressing a derivative of HLA-A*0201, and dendritic cells, to enhance the activation of CD8+ T cells that recognize peptide epitopes derived from human tyrosinase and glycoprotein 100. We find that by altering the cell surface density of the immunizing peptide on the dendritic cells, either by pulsing with higher concentrations of peptide, or by changing the MHC-peptide-binding affinity by generating variants of the parent peptides, the size of the activated CD8+ T cell populations can be modulated in vivo. Significantly, the density of peptide that produced the largest response was less than the maximum density achievable through short-term peptide pulsing. We have also found, however, that while some variant peptides are effective at eliciting both primary and recall CD8+ T cell responses that can recognize the parental epitope, other variant epitopes lead to the outgrowth of CD8+ T cells that only recognize the variant. HLA-A*0201 transgenic mice provide an important model to define which peptide variants are most likely to stimulate CD8+ T cell populations that recognize the parental, melanoma-specific peptide.

Dinitrophenyl-modified autologous melanoma vaccine induces a T cell response to hapten-modified, melanoma peptides.

Active specific immunotherapy with dinitrophenyl (DNP)-modified autologous melanoma vaccine elicits inflammatory responses in metastatic tumor sites. Postsurgical adjuvant immunotherapy with this vaccine prolongs survival in stage III melanoma patients. We have reported that, after administration of DNP-modified melanoma vaccine, T cell responses to DNP-modified autologous tumor cells are demonstrable in vivo and in vitro. These responses are hapten specific and MHC restricted. To elucidate this phenomenon, we investigated the immune response to DNP-modified peptides eluted from autologous cells. Short peptides were extracted from DNP-modified and unmodified autologous melanoma cells by an acid elution technique and HPLC fractionation. Peptides were also extracted from DNP-modified and unmodified, EB virus-transformed, autologous B lymphoblasts. These various peptide fractions were loaded onto autologous B lymphoblasts and tested for ability to elicit a response by a DNP-specific T cell line as measured by IFN-gamma production. Unexpectedly, stimulatory activity of peptides from DNP-modified melanoma cells was confined to a single HPLC fraction. Spectrometric analysis of this fraction confirmed modification of peptides with DNP. A weaker T cell response was observed to a single HPLC fraction of DNP-modified peptides from the patient's B lymphoblasts. No T cell response was elicited by corresponding fractions of peptides eluted from unmodified melanoma cells or B lymphoblasts. These findings demonstrate the human T cell response to DNP-modified autologous melanoma cells is mediated by hapten-modified, MHC-associated peptides. Further investigation of these peptides could lead to a new strategy for peptide-based cancer immunotherapy.

Initiation codon scanthrough versus termination codon readthrough demonstrates strong potential for major histocompatibility complex class I-restricted cryptic epitope expression.

Accumulating evidence shows that the repertoire of major histocompatibility complex class I-restricted epitopes extends beyond conventional translation reading frames. Previously, we reported that scanthrough translation, where the initiating AUG of a primary open reading frame is bypassed, is most likely to account for the presentation of cryptic epitopes from alternative reading frames within the influenza A PR/8/34 nucleoprotein gene. Here, we confirm and extend these findings using an epitope cassette construct that features two well-defined CD8(+) T cell (TCD8+) epitopes in alternative reading frames, each preceded by a single start codon. Expression of one epitope depends on scanning of the ribosome over the first AUG with translation initiation occurring at the second AUG. We find that scanthrough translation has great potency in our system, with its impact being modulated, as predicted, by the base composition surrounding the first initiation codon, the number of start codons preceding the point of alternate reading frame initiation, and the efficiency with which the epitope itself is generated. Additionally, we investigated the efficiency of eukaryotic translation termination codons, to assess codon readthrough as a mechanism for cryptic epitope expression from 3' untranslated regions. In contrast with initiation codons, eukaryotic stop codons appear to be highly efficient at preventing expression of epitopes encoded in 3' untranslated regions, suggesting that 3' untranslated regions are not a common source of cryptic epitope substrate. We conclude that scanthrough is a powerful mechanism for the expression of epitopes encoded in upstream alternative open reading frames that may contribute significantly to TCD8+ responses and to tolerance induction.

Ribosomal scanning past the primary initiation codon as a mechanism for expression of CTL epitopes encoded in alternative reading frames.

An increasing amount of evidence has shown that epitopes restricted to MHC class I molecules and recognized by CTL need not be encoded in a primary open reading frame (ORF). Such epitopes have been demonstrated after stop codons, in alternative reading frames (RF) and within introns. We have used a series of frameshifts (FS) introduced into the Influenza A/PR/8 /34 nucleoprotein (NP) gene to confirm the previous in vitro observations of cryptic epitope expression, and show that they are sufficiently expressed to prime immune responses in vivo. This presentation is not due to sub-dominant epitopes, transcription from cryptic promoters beyond the point of the FS, or internal initiation of translation. By introducing additional mutations to the construct exhibiting the most potent presentation, we have identified initiation codon readthrough (termed scanthrough here, where the scanning ribosome bypasses the conventional initiation codon, initiating translation further downstream) as the likely mechanism of epitope production. Further mutational analysis demonstrated that, while it should operate during the expression of wild-type (WT) protein, scanthrough does not provide a major source of processing substrate in our system. These findings suggest (i) that the full array of self- and pathogen-derived epitopes available during thymic selection and infection has not been fully appreciated and (ii) that cryptic epitope expression should be considered when the specificity of a CTL response cannot be identified or in therapeutic situations when conventional CTL targets are limited, as may be the case with latent viral infections and transformed cells. Finally, initiation codon readthrough provides a plausible explanation for the presentation of exocytic proteins by MHC class I molecules.