DNA vaccination against tissue-restricted antigens is a strategy for cancer therapy. Immune tolerance and ignorance of self antigens has been a hurdle for this approach. We have shown that immunization with xenogeneic DNA orthologues elicits tumor immunity. One model that we have developed entails immunization of mice against tyrosinase-related protein-2 (Tyrp2) using cDNA encoding homologous human Tyrp2. A subset of mice immunized with human Tyrp2 developed antibody responses to Tyrp1. Unexpectedly, this was not simply due to cross-reactivity, as mice with anti-Tyrp1 antibodies were not usually the same animals with anti-Tyrp2 antibodies. Although autoimmune vitiligo was frequently observed in mice that had been immunized with Tyrp2, its occurrence was not correlated with the development of antibodies to Tyrp1. This implies that the appearance of anti-Tyrp1 antibodies was not simply a consequence of the destruction of melanocytes by T-cells recognizing Tyrp2. This represents an example of intermolecular determinant recognition, but is not simply due to epitope spreading since antibodies against the antigen targeted by DNA vaccination are not typically detected.
This article was published in Cancer Immunity, a Cancer Research Institute journal that ceased publication in 2013 and is now provided online in association with Cancer Immunology Research.
The melanosomal membrane glycoproteins Tyrp1 and Tyrp2 (tyrosinase related protein-1 and -2) are recognized by antibodies and CD8+ T-cells, respectively, in patients with melanoma (1, 2, 3, 4). As such, these and other components of the melanosome have been investigated as potential antigens for melanoma vaccines. We and others have evaluated recombinant proteins, recombinant poxvirus, peptides and DNA vaccines in an effort to develop the most potent means for inducing antibodies and T-cell responses against these antigens. Using xenogeneic (human) DNA vaccines in a C57BL/6 mouse model system, we have previously shown that mice produce antibodies and T-cells recognizing gene products of both the mouse and human orthologues (5, 6). Specifically, we have observed immunity against mouse Tyrp1 or Tyrp2 when mice were immunized with human TYRP1 or DCT (DOPAchrome tautomerase, hTyrp2) respectively, while no response occurs when plasmids encoding the syngeneic (mouse) antigens are used. The immune responses following immunization with xenogeneic genes resulted in tumor rejection of syngeneic B16 melanoma and autoimmune coat hypopigmentation.
One of the advantages of vaccination with full-length cDNA is the presentation of multiple potential epitopes contained within the DNA sequence. An additional possibility is that immune recognition can spread to determinants on related proteins encoded by different genes (paralogues). In this regard, TYRP1 and DCT are paralogues that presumably diverged from an ancestral tyrosinase gene possibly more than 500 million years ago. The Tyrp1 and Tyrp2 proteins have 52% amino acid identity and 67% amino acid homology, with most of the similarity located in the melanosome lumenal domain, including the metal binding residues. We were therefore interested in investigating whether immunization with DCT might also result in the production of cross-reactive antibodies to Tyrp1.
Immunization with human Tyrp2 cDNA induces autoantibodies
In previous studies, we showed that mice immunized with xenogeneic (human) Tyrp1 cDNA generated antibodies that recognized syngeneic Tyrp1 from B16 mouse melanoma (5). These antibodies were found to be of the Th2 subtype (IgG1, IgG2b). In this report, we show that, as expected, mice immunized with xenogeneic hTyrp2 cDNA developed antibodies that recognized syngeneic mouse Tyrp2 protein from B16 melanoma (Figure 1). These antibodies were found to be of the Th2 isotype (IgG1 and IgG2b, data not shown). In these experiments the immunoprecipitation-Western blot assays were carried out with antibodies TA99 and anti-pep8, which precipitate only the specified antigen (Tyrp1 or Tyrp2, respectively). Anti-pep8 was shown to recognize only Tyrp2, and not Tyrp1, in the original report describing the production of these anti-melanosomal antigen peptide sera (7). To demonstrate that TA99 precipitates Tyrp1 but not Tyrp2, we performed an immunoprecipitation-Western blot experiment in which B16 melanoma lysate was first immunoprecipitated with TA99 and the filter probed with either anti-pep8 or anti-pep1 (recognizing Tyrp1). No bands were detected by anti-pep8 while the anti-pep1 recognized the expected Tyrp1 band (Figure 2).
Sera from a subset of mice immunized with hTyrp2 contain antibodies to Tyrp1
Immunoprecipitation-Western blot assays were carried out with sera from mice immunized with human Tyrp2 DNA to assess their ability to recognize Tyrp1. A subset of mice immunized with hTyrp2 DNA developed antibodies that recognized mouse Tyrp1 (Figure 3). These antibodies showed a similar recognition of human Tyrp1 in SK-Mel-188 lysate (data not shown). We analyzed 61 sera from mice immunized with hTyrp2 DNA and found that 21% (13/61) of these animals had antibodies recognizing Tyrp1. As a comparison, 30% (18/61) of the same animals developed autoantibodies to the intended antigen, Tyrp2. By analyzing the antibody profile of the 61 individual mice, it was found that only four of the animals had antibodies to both Tyrp1 and Tyrp2 (Figure 4). Nine of the thirteen mice had antibodies that only recognized Tyrp1 and not Tyrp2, eliminating cross-reactivity as the explanation for the anti-Tyrp1 antibodies. Analysis of sera collected at different time points after immunization showed that anti-Tyrp1 antibodies were not evident until the completion of five weekly immunizations, consistent with the kinetics of antibody production observed when mice are immunized directly with hTyrp1 cDNA (Figure 5).
Autoantibodies from mice immunized with hTyrp2 recognize syngeneic mouse Tyrp1 and not a related protein
Even though the immunoprecipitation-Western blot assay for autoantibodies against Tyrp1 was performed using a monoclonal antibody specific for Tyrp1, we wanted to ensure that the antibodies that were detected in the sera of Tyrp2-vaccinated mice were actually recognizing Tyrp1 and not another related protein. To answer this question, we repeated the immunoprecipitation-Western blot assay using recombinant mouse Tyrp1 produced in a baculovirus expression system. The results (Figure 6) show that the sera that were previously found to contain anti-Tyrp1 antibodies in the immunoprecipitation-Western blot assays using melanoma cell lysates also recognized baculovirus-derived recombinant mouse Tyrp1.
Production of anti-Tyrp1 antibodies does not correlate with hypopigmentation
We and others have previously observed autoimmune hypopigmentation of mouse coat resulting from immunization with hTyrp2 or hTyrp1 DNA. This begins within three weeks of the start of immunization and can progress from scattered patches of white hair on the immunized abdomen to involve hypopigmentation of most of the coat. One possible mechanism for the induction of anti-Tyrp1 antibodies could involve the destruction of melanocytes following immunization with hTyrp2, leading to increased accessibility of Tyrp1; consequently, Tyrp1 released from dying melanocytes could induce an autoantibody response. Because we had compiled hypopigmentation scores for a proportion of the immunized animals, the correlation between anti-Tyrp1 antibody production and hypopigmentation was determined. The results of 12 mice for which such data are available are shown in Table 1. Clearly, the appearance of hypopigmentation is not sufficient for the induction of anti-Tyrp1 antibodies.
DNA vaccination is a rapid and efficient way to induce both antibody and T cell responses against otherwise poorly immunogenic self antigens normally found on tumor cells. One of the advantages of immunization with full-length DNA is the ability to present numerous potential epitopes within the sequence. We have found that immunization of mice with human Tyrp2 DNA not only results in the induction of T-cells and antibodies recognizing Tyrp2, but also of antibodies specifically binding to Tyrp1, a related melanosomal differentiation antigen.
Intermolecular epitope spreading is a potential mechanism for the anti-Tyrp1 antibody response. Epitope spreading is a process in which the target of an immune response extends from the intended antigen to other epitopes present on the cell involved in the primary response. The additional epitopes may be within the same molecule (intramolecular) or on distinct molecules (intermolecular). Epitope spreading has classically been associated with experimental autoimmune encephalitis and diabetes mellitus in mice; however, several recent publications describe epitope spreading in human and mouse systems for peptide-based tumor vaccines (8, 9, 10, 11, 12, 13). The basic mechanism of intermolecular epitope spreading involves the destruction of tissue based on the recognition of the intended antigenic target, followed by the release of other potential antigens from this tissue and the presentation of these exogenous antigens by MHC class II molecules or via cross-priming onto MHC class I molecules.
Following immunization against Tyrp2, normal melanocytes are destroyed by CD8+ T-cells recognizing Tyrp2 epitopes, evidenced by the hypopigmentation that occurs within three weeks of initiating vaccination (6). We hypothesized that since Tyrp1 is the most abundant glycoprotein found in melanocytic cells (1, 2), the inflammatory milieu induced by the T-cell response to Tyrp2 may have allowed for a secondary response to Tyrp1. However, there was no correlation between the amount of hypopigmentation of an individual mouse and the probability of developing anti-Tyrp1 antibodies. Presentation of Tyrp1 epitopes could also occur as a result of the skin injury caused by gene gun immunization in the presence of immunostimulatory CpG motifs in plasmid DNA (14, 15). This is not likely given the absence of Tyrp1 antibodies in the serum of mice immunized with either mouse Tyrp1 DNA or empty vector (5).
An alternative explanation lies in the extensive homology between members of the tyrosinase family of melanosomal membrane glycoproteins. Tyrp2 and Tyrp1 share 52% identity at the amino acid level, allowing for the possibility that a common determinant may be presented in the subset of animals that have anti-Tyrp1 antibodies. These antibodies are evident following the fifth weekly immunization, the same time course found in mice immunized directly with human Tyrp1 DNA (5). This supports the idea that the response to Tyrp1 is induced directly by the Tyrp2 DNA vaccine and not by the Tyrp1 protein released from normal melanocytes. Selective recognition of a particular shared Tyrp2 and Tyrp1 determinant in some animals may occur because of higher affinity of antibodies for the determinant in the context of the Tyrp1 protein or more efficient presentation of such epitopes when contained in Tyrp2 versus Tyrp1.
Another possibility is that T cell help is preferentially provided by Tyrp1 epitopes over Tyrp2 epitopes. Tyrp1 is the most abundant glycoprotein in the endosome compartment of melanocytes and could out-compete Tyrp2 peptides for MHC presentation. However as class II MHC epitopes from mouse Tyrp1 and Tyrp2 are not defined at this time, it is difficult to address this possibility experimentally.
Anti-Tyrp1 antibodies produced in response to immunization with Tyrp2 DNA are predominantly of the IgG1 and IgG2b isotypes, indicating a Th2 response (data not shown). This contrasts with the Th1 cytotoxic T-cell response that accounts for much of the tumor immunity and autoimmunity induced by Tyrp2 DNA vaccination. Observation of both Th1 and Th2 responses to the same vaccine has significant implications for the design of strategies to enhance responses using cytokines. With the knowledge that Tyrp2 DNA vaccines elicit both Th1 and Th2-based immunity, one would be hesitant to use polarizing cytokines, such as IFN-gamma or IL-4, to augment the response. A more reasonable approach in this situation would be to include a 'Th-neutral' cytokine, such as GM-CSF, as a molecular adjuvant.
Materials and methods
C57BL/6 mice (6 to 8-wk-old females) were acquired through the National Cancer Institute breeding program. These mice were kept in a pathogen-free Memorial Sloan-Kettering Cancer Center vivarium according to National Institutes of Health Animal Care guidelines. All mice entered the study between 7 and 10 weeks of age.
Cell lines and tissue culture
B16F10/LM3, a pigmented mouse melanoma cell line of C57BL/6 origin and derived from the B16F10 line, was provided by Dr. Isaiah Fidler (M.D. Anderson Cancer Center, Houston, TX) (16).
The human DCT/TRP-2 (hTyrp2) expression vector (supplied by Drs. S. A. Rosenberg and N. Restifo, National Cancer Institute, Bethesda, MD) has been described (4).
The method of DNA immunization has been reported (17). In brief, plasmid DNA encoding hTyrp2 was coated onto 1.0 µm gold bullets. Animals were immunized by delivering gold-DNA complexes using a helium-driven gun (Accell®; PowderJect Vaccines, Madison, WI) into each abdominal quadrant (1 µg plasmid DNA/quadrant) for a total of four injections. Mice were immunized weekly for a total of 5 weeks.
Hypopigmentation experiments were performed as described (5, 6). In brief, after the final immunization, mice were shaved and depilated over the posterior flank and observed for 8 wks. Scoring of hypopigmentation was performed by dividing the abdomen into four equal quadrants. Quadrants were recorded as positive when they had an estimated >50% depigmented hairs. Hypopigmentation was scored 0-4+ according to the number of quadrants that were depigmented in each mouse (e.g., 3+ if three of four quadrants are depigmented >50%).
Antibody responses to Tyrp1 or Tyrp2
Immunoprecipitation-Western blot assays were used to detect antibody responses to Tyrp1 and Tyrp2, as previously described (5, 6). For Tyrp1, B16 melanoma cells were lysed and the lysate precipitated with monoclonal antibody TA99, which recognizes both human and mouse Tyrp1 (18). For Tyrp2, anti-pep8, a rabbit polyclonal antiserum raised against a carboxyl-terminal peptide (kindly provided by Vincent Hearing, National Cancer Institute) was used as the precipitating antibody. Immunoprecipitates were washed and separated on 8% SDS-PAGE under reducing conditions and transferred to Immobilon membrane (Millipore, Bedford, MA). The membranes were blocked overnight at 4°C in 5% non-fat dry milk, washed, and probed with mouse sera or preimmune mouse sera diluted 1:60 in 0.2% Tween-20, phosphate-buffered saline. Blots were then washed and incubated with goat anti-mouse IgG-horseradish peroxidase and detected using an ECL kit (Amersham, Arlington Heights, IL) and Kodak BioMax MR-2 film (Kodak, Rochester, NY). For the determination of specificity, rabbit anti-pep1 antibody against Tyrp-1, which is not cross-reactive with Tyrp2, was used (7).
In some experiments, baculovirus-derived recombinant mTyrp1 was used as a target for immunoprecipitation in place of B16 melanoma lysate. Sf9 insect cells were infected with baculovirus with or without mTyrp1 (a generous gift from ImClone Systems, New York, NY) for one hour, after which protein expression was allowed to proceed for 48 hours. The cells were then harvested and lysates used for immunoprecipitation-Western blot assays as described above.
This work was supported by NIH grants K12CA01712-08 and R0158621, the Quentin J. Kennedy Foundation, the Rubin Family Foundation and Swim Across America.
- Received June 21, 2002.
- Accepted June 21, 2002.
- Copyright © 2002 by Jedd D. Wolchok