Mouse models in which tumors arise spontaneously due to the transgenic expression of an oncoprotein provide an opportunity to test approaches that target the immune-mediated control of tumor progression. In this report we investigated the role of SV40 Tag-specific CD8+ T cells in the control of advanced choroid plexus tumor progression using large tumor Ag (Tag) transgenic mice. Since mice of the SV11 line are tolerant to the immunodominant SV40 Tag-derived CTL epitopes, mice with advanced stage tumors were reconstituted with naive C57BL/6 spleen cells following a low dose of γ-irradiation. This led to the priming of CTLs specific for the H2-Kb-restricted epitope IV by the endogenous Tag and a significant increase in the life span of Tag transgenic mice. Epitope IV-specific CD8+ T cells accumulated and persisted in the brains and tumors of SV11 mice, as determined by analysis with epitope-specific MHC class I tetramers. Brain-infiltrating epitope IV-specific T cells were capable of producing IFN-γ as well as lysing syngeneic Tag-transformed cells in vitro. In addition, the adoptive transfer of spleen cells from Tag-immune C57BL/6 mice resulted in a dramatic increase in the control of tumor progression in SV11 mice and was associated with the accumulation of CD8+ T cells specific for multiple Tag epitopes in the brain. These results indicate that the control of advanced stage spontaneous choroid plexus tumors is associated with the induction of a strong and persistent CD8+ T cell response to Tag.
One of the primary goals of tumor immunotherapy is to induce an effective immune response against pre-existing tumors. CD8+ CTLs are particularly well suited for the destruction of tumor cells due to their ability to recognize specific tumor Ags. This interaction is mediated through the presentation of tumor-derived peptides by MHC class I molecules at the surface of the tumor cell. The ability of tumor Ag-specific CTLs to control tumor progression in murine models is well established (1). Although adaptation of these strategies toward the immunotherapy of human tumors previously had limited success, more recent results are encouraging (2, 3). The majority of evidence supporting the role of CTLs in the control of tumor growth has been derived from experiments using transplantable murine tumor models. A number of transgenic mouse models have been developed in which tumors arise spontaneously at distinct anatomic sites (4, 5, 6, 7, 8). These models more closely mimic the course of tumor development in humans, providing opportunities to develop successful approaches toward the immunotherapy of spontaneous cancer.
The SV40 large tumor Ag (Tag)3 is a well-characterized oncoprotein capable of inducing tumor progression in mice when expressed as a transgene. Tag is also the target of a strong cell-mediated immune response (9). In C57BL/6 mice, immunization with full-length Tag results in a hierarchical response to multiple CTL epitopes (10, 11). Four CTL epitopes have been defined, including epitope I, Tag residues 206–215; epitope II/III, Tag residues 223–231; epitope IV, Tag residues 404–411; and epitope V, Tag residues 489–497 (12, 13, 14, 15, 16). Epitopes I, II/III, and V are H2-Db restricted, and epitope IV is H2-Kb restricted. Epitopes I, II/III, and IV are immunodominant, with the highest number of CD8+ T cells directed against epitope IV following immunization with full-length Tag (11). In contrast, epitope V-specific CD8+ T cells are detected only following immunization with Tag variants in which epitopes I, II/III, and IV have been inactivated (16) or following immunization with a recombinant vaccinia virus expressing epitope V as a minigene (17).
We previously studied the CD8+ T cell response to Tag in mice of the SV11 lineage (18), which express full-length Tag from the SV40 enhancer/promoter (19). Tag is expressed in the choroid plexus and thymus, leading to the development of choroid plexus papillomas. Neoplastic clusters of cells are detected by 40 days of age in Tag transgene-positive (SV11+) mice, with progressive tumor growth leading to death at a mean age of 104 days (20). Due to expression of Tag in the thymus, SV11+ mice are tolerant to the immunodominant Tag CTL epitopes I, II/III, and IV (18). Reconstitution of SV11+ mice with naive C57BL/6 spleen cells before tumor development resulted in priming of donor CTL precursors specific for epitope IV by the endogenous Tag. This CTL response correlated with the control of choroid plexus tumor progression.
These results indicated that the activation of T lymphocytes specific for a dominant Tag CTL epitope could inhibit the progression of tumors in SV11+ mice if adoptive transfers were given before tumor appearance at 45 days of age. The question of whether this approach will be feasible in controlling advanced tumors in SV11+ transgenic mice was not addressed. To approach this question, 80-day-old SV11+ mice carrying large, spontaneously arising tumors in the ventricles of the brain were reconstituted with naive spleen cells from C57BL/6 mice after low dose irradiation. These mice were monitored for both tumor growth and the presence of Tag epitope-specific CTL in brain and tumor sites.
The results of this study indicate that effective control of advanced choroid plexus tumors in SV11+ mice is associated with the recruitment of significant numbers of functional epitope IV-specific CD8+ T cells into brain and tumor. In addition, the transfer of Tag-immune spleen cells from C57BL/6 mice into irradiated SV11+ mice with advanced tumors resulted in a dramatic enhancement of tumor immunity with concomitant expansion and localization of CTL specific for T Ag epitopes to the brain.
Materials and Methods
C57BL/6 (H2b) mice were purchased from The Jackson Laboratory (Bar Harbor, ME) and maintained at the animal facility of the Milton S. Hershey Medical Center. Mice were routinely used between the ages of 6–12 wk. SV11 mice were generated by Palmiter et al. (19), and the kinetics of tumor progression have been described previously (20). SV11 (H2b) mice, expressing the SV40 Tag transgene under its own enhancer/promotor, were originally obtained from Dr. T. Van Dyke (University of North Carolina, Chapel Hill, NC) via Dr. E. Roy (University of Illinois, Urbana-Champaign, IL). The SV11 line has been maintained in the animal facility of the Milton S. Hershey Medical center by backcrossing Tag transgene-positive males with C57BL/6 females. SV11 transgene-positive mice (SV11+) were identified by PCR amplification of the transgene as described previously (18).
Cell lines and media
B6/WT-19 is an SV40-transformed C57BL/6 mouse embryo fibroblast line that expresses wild-type Tag (21). B6/T122B1 cells express a Tag derivative in which all four H2b-restricted CTL epitopes in Tag (I, II/III, IV, and V) have been inactivated by alanine substitutions of critical MHC class I anchor residues (N210A, N227A, F408A, and N493A) (11). All Tag-transformed cell lines were maintained in DMEM supplemented with 100 U penicillin/ml, 100 μg streptomycin/ml, 100 μg kanamycin/ml, 2 mM l-glutamine, 10 mM HEPES buffer, 0.075% (w/v) NaHCO3, and 5–10% FBS. RMA (H2b) cells (22) were maintained in suspension using RPMI 1640 medium supplemented with 10% FBS, 100 U penicillin/ml, 100 μg streptomycin/ml, 2 mM l-glutamine, and 50 μM 2-ME.
All peptides used were synthesized at the Macromolecular Core Facility of the Milton S. Hershey Medical Center by F-moc chemistry using an automated peptide synthesizer (model 9050 PepSynthesizer; MilliGen, Bedford, MA). Peptides were solubilized in DMSO and diluted to the appropriate concentration with RPMI 1640. Peptides used in these experiments correspond to SV40 Tag epitopes I (SAINNYAQKL), a variant of epitope II/III containing a Ser substitution for Cys at the N terminus (SKGVNKEYL) that leads to enhanced recognition by epitope II/III-specific CTL (data not shown), a variant of epitope IV containing a Leu substitution for Cys at the C terminus (VVYDFLKL) that leads to enhanced binding to H2-Kb (data not shown), and V (QGINNLDNL). In addition, the H2-Kb-restricted peptide 498–505 (SSIEFARL) from HSV glycoprotein B was used as a control peptide.
Irradiation and adoptive transfer
Spleen cells used for adoptive transfer were obtained from the appropriate sex of normal C57BL/6 mice. Eighty-day-old SV11 mice were reconstituted by i.v. injection of 5 × 107 RBC-depleted naive C57BL/6 spleen cells suspended in 0.2 ml HBSS into the tail vein. Mice received 400 rad gamma irradiation from a 60Co source (GammaCell 220; MDS Nordion, Ottawa, Canada) 1 day before reconstitution with spleen cells. In some experiments irradiated mice were reconstituted with spleen cells derived from C57BL/6 mice that had been immunized i.p. with 3 × 107 B6/WT-19 cells 10 days before transfer.
Isolation of lymphocytes from brain
Lymphocytes were isolated from the brain as previously described (23). Briefly, brains were harvested from mice following exsanguination under a lethal dose of anesthesia. Brains were minced in complete RPMI 1640 using a razor blade, and single cells were dissociated from larger tissue fragments by repeated pipetting. After allowing the debris to settle, the supernatant containing cells was collected, and the remaining clumps were disrupted by pipetting in fresh RPMI 1640. The supernatants containing cells were combined, and the cells were pelleted by centrifugation. Cells were resuspended in 3 ml 70% Percoll (Sigma, St. Louis, MO) in 15-ml conical tubes. 5 ml 35% Percoll was layered on top of the cells, and the gradients were centrifuged at 500 × g for 20 min at 4°C. Cells were harvested from the gradient interphase and washed once with complete RPMI 1640 before use. In some experiments choroid plexus tumors were dissected from brains, and the brain and tumor tissues were processed separately.
Preparation of MHC class I tetramers, staining of epitope-specific T cells, and flow cytometry
Production and characterization of MHC tetramers was achieved as described previously (11). For staining of lymphocyte populations, RBC-depleted lymphocytes were incubated with rat anti-mouse CD16/CD32 (BD PharMingen, San Diego, CA) and 50 μg/ml streptavidin (Molecular Probes, Eugene, OR) for 30 min on ice to block Fc receptors and nonspecific binding of streptavidin-conjugated tetramers, respectively. After a single wash, spleen cells were incubated with PE-labeled tetramers and FITC-labeled rat anti-mouse CD8a (53-6.7, BD PharMingen) for 1 h on ice. Cells were fixed with 2% paraformaldehyde and analyzed using a FACScan (BD Biosciences, San Jose, CA) flow cytometer, and the data were analyzed and prepared using CellQuest software (BD Biosciences). The percentage of CD8+ cells that stained with epitope-specific MHC tetramers was determined by subtracting the percentage of CD8+ cells that stained with an irrelevant MHC tetramer in the same cell population. Determination of TCR β-chain variable region (Vβ) usage by CTL lines was accomplished by staining with a panel of Vβ-specific Abs (BD PharMingen).
Intracellular cytokine assay
For staining of intracellular IFN-γ, RBC-depleted lymphocyte suspensions were prepared as described above, and 2.5 × 106 spleen cells or 1 × 105 brain-derived lymphocytes were incubated with 1 μM of the indicated synthetic peptides, representing Tag or control epitopes, and 1 μg/ml brefeldin A in complete RPMI 1640 containing 10% FBS for 6 h at 37°C in 5% CO2. Cells were stained for intracellular IFN-γ using the Cytofix/Cytoperm Kit (BD PharMingen) according to the manufacturer’s specifications. Briefly, stimulated cells were washed twice and then FcR were blocked by incubation with rat anti-mouse CD16/CD32 for 20 min, followed by staining with PE-labeled rat anti-mouse CD8a for 30 min. After fixation and permeabilization for 20 min, cells were stained with FITC-labeled rat anti-mouse IFN-γ (BD PharMingen) or an isotype control Ab for 30 min and then analyzed by flow cytometry as described above. The percentage of CD8+ cells that produced intracellular IFN-γ in response to a specific peptide was calculated by subtracting those cells that stained positively for IFN-γ following incubation with an unrelated peptide.
In vitro stimulation of bulk CTL and CTL lines
Lymphocytes isolated from spleen were restimulated in vitro with gamma-irradiated B6/WT-19 cells as previously described (18). Briefly, 1 × 107 spleen cells were mixed with 5 × 105 gamma-irradiated (10,000 rad) B6/WT-19 cells in 4 ml complete RPMI 1640 medium supplemented with 10% FBS/well of a 12-well tissue culture plate. To establish CTL lines from brains and choroid plexus tumors, 5 × 105 to 1 × 106 lymphocytes isolated from the brain or tumor of SV11+ mice that had received irradiation and adoptive transfer with naive C57BL/6 spleen cells were stimulated in vitro with 5 × 104 gamma-irradiated B6/WT-19 cells in 200 μl complete RPMI 1640 in a 96-well flat-bottom plate. After 7 days, cells from wells containing positive growth were expanded to 12-well plates with 5 × 105 gamma-irradiated B6/WT-19 cells and 5 U/ml rIL-2 (provided by Amgen, Thousand Oaks, CA). CTL were expanded every 7 days with fresh stimulator cells and rIL-2.
Histology and immunohistochemistry
Histological analysis of formalin-fixed and paraffin-embedded tissues was performed as described previously (18). Tumor burden was determined by completely sectioning the brains of individual mice and identifying the largest tumor area that could be detected for each mouse. The average tumor size detected in 100-day-old SV11+ mice that had been irradiated at 80 days of age was 7 ± 3 mm2 (range, 1.6–14 mm2). Mice that had tumors with an area <1 mm2 were considered to have a significantly reduced tumor burden compared with 100-day-old SV11+ mice that received only irradiation.
Kaplan-Meier survival plots were constructed with DeltaGraph software (Deltapoint, Monterey, CA), and statistical analysis was performed by a single-factor ANOVA and was validated using Fisher’s protected least significant difference test. p < 0.05 was considered significant.
Control of advanced tumor progression in SV1l+ mice following priming of CTL against the endogenous Tag
To study the role of the immune response in the control of advanced tumors, SV11+ mice were used at approximately 80 days of age, when the majority of mice had accumulated significant tumor mass in the choroid plexus (Fig. 1⇓A). Mice were irradiated and reconstituted with naive C57BL/6 spleen cells and sacrificed at various time points to assess the level of tumor burden by histologic analysis. Mice given only irradiation without subsequent adoptive transfer developed massive end-stage tumors by 100 days of age, which typically filled and substantially distorted the ventricles (Fig. 1⇓B). In contrast, mice given adoptive transfer with naive B6 spleen cells following irradiation had comparatively small tumors that rarely filled the ventricles by 100 days of age (Fig. 1⇓C). In total, three of six mice had significant reductions in tumor burden compared with control animals that received only irradiation. Suppression of tumor growth continued to be effective at 110 and even 150 days of age (Fig. 1⇓C), as four of five mice and two of three mice examined, respectively, had reduced tumor burden compared with 100-day-old control mice that received only irradiation. The majority of tumors detected in mice at 110 and 150 days of age were smaller than those typically found in unmanipulated mice at 80 days of age, suggesting that some tumors had regressed. These tumors retained expression of nuclear Tag as determined by immunohistochemistry (Fig. 1⇓C).
SV11+ mice that received both irradiation and adoptive transfer with naive C57BL/6 spleen cells had strong epitope IV-specific CTL responses at 100, 110, and 150 days of age (Fig. 2⇓A). Occasionally, CTLs specific for one of the other immunodominant epitopes were detected, but only at the later time points. Thus, epitope spreading to the less dominant CTL epitopes I and II/III may occur once the response to epitope IV has been established. This result indicates that control of tumor progression in SV11+ mice is associated with priming of epitope IV-specific CTLs against the endogenous Tag.
In accordance with the reduced tumor burden observed in SV11+ mice, transfer of naive C57BL/6 spleen cells into irradiated 80-day-old SV11+ mice resulted in a profound increase in the mean life span of SV11+ mice from 107 to 170 days of age (Fig. 2⇑C). SV11+ mice given adoptive transfers without prior irradiation had no significant increase in their life span, while SV11+ mice given only irradiation showed a modest increase in life span to 125 days of age. Although this increase was not statistically significant, irradiation might have a limited effect on the progression of advanced choroid plexus tumors. Together these results indicate that efficient priming of Tag-specific CTLs against the endogenous Tag in SV11+ mice is associated with the control of advanced choroid plexus tumors.
Epitope IV-specific CD8+ T cells are localized to the tumor site in SV11+ mice
The ability to expand epitope IV-specific CTLs after in vitro culture of spleen-derived lymphocytes with Tag-transformed cells indicates that a systemic response to Tag was induced in SV11+ mice following irradiation and adoptive transfer with naive C57BL/6 spleen cells. To determine the magnitude of this response, splenocytes from treated mice were stained directly using H2-Kb/epitope IV tetramers to elucidate epitope IV-specific CD8+ T cells. Epitope IV-specific T cells represented between 5 and 17% of the splenic CD8+ T cells in SV11+ mice at 10 days postadoptive transfer, but were not detected in Tag-negative littermates (Fig. 3⇓A). To determine whether epitope IV-specific CD8+ T cells were associated with the tumor site, lymphocytes were isolated from the brains of irradiated and reconstituted SV11+ mice and stained with epitope IV tetramer. Infiltrating CD8+ T cells were detected, of which a significant proportion were specific for Tag epitope IV (Fig. 3⇓A). In contrast, only a limited number of CD8+ T cells were isolated from the brains of Tag-negative littermates given the identical treatment, none of which was specific for epitope IV.
Epitope IV-specific CD8+ T cells isolated from spleen or brain were analyzed for the expression of cell surface markers considered to be indicative of the level of T cell activation. This analysis revealed that the spleen-derived CD8+, Kb/IV tetramer+ cells were CD44high, CD62 ligand (CD62L)low, and CD69low, typical of memory T cells, while CD8+, Kb/IV tetramer+ cells isolated from the brain were CD44high, CD62Llow, but expressed increased levels of CD69, typical of recently activated T cells (24). Gating on the CD8+, Kb/IV tetramer-negative cells in the spleen revealed a mixed population of naive and memory cells. The CD8+, Kb/IV tetramer-negative cells in the brain, however, uniformly displayed an activated phenotype. The specificity and functionality of these Kb/IV tetramer-negative CD8+ T cells remain unknown. The activation phenotype of CD8+ T cells isolated from the brains of SV11 transgene-negative littermates could not be determined due to the low number of total cells. Thus, priming of naive C57BL/6 spleen cells against the endogenous Tag in SV11+ mice led to the recruitment of epitope IV-specific CD8+ T cells that migrated to the site of Tag production and tumor formation early during the immune response.
Epitope IV-specific cells continued to represent a significant proportion of the brain-derived CD8+ T cells in SV11+ mice even at 40 days postadoptive transfer of C57BL/6 spleen cells (Fig. 4⇓A), indicating that epitope IV-specific CD8+ T cells persisted in the brains of SV11+ mice. CD8+ T cells specific for the H2-Db-restricted epitopes I and II/III were not detected in these mice using MHC tetramers constructed using epitopes I and II/III (Fig. 4⇓A). To more conclusively demonstrate that epitope IV-specific CD8+ T cells are localized to the tumor site, several treated SV11+ mice were identified that had detectable choroid plexus tumors at the time of sacrifice. Tumors were dissected from the brains, and lymphocytes were isolated from the tumors and stained with MHC tetramers. Epitope IV-specific CD8+ T cells were readily detected among the tumor-infiltrating lymphocytes (Fig. 4⇓B), indicating that epitope IV-specific T cells are associated with tumors in SV11+ mice.
Epitope IV-specific CD8+ T cells retain effector function in SV11+ mice
Progression of tumors in vivo has been associated with the induction of anergy among tumor Ag-specific T cells (25, 26). To determine whether the epitope IV-specific CD8+ T cells from SV11+ mice could mount a functional response to Ag, the ability of lymphocytes to produce IFN-γ following a short incubation with epitope IV peptide was determined. Significant numbers of IFN-γ-producing CD8+ T lymphocytes were detected in the spleen and brain-derived lymphocytes of 100-day-old SV11+ mice that had received both irradiation and adoptive transfer with naive C57BL/6 spleen cells at 80 days of age (Fig. 4⇑C). This result indicates that a significant population of epitope IV-specific CD8+ T cells retains their effector function after exposure to endogenous Tag in SV11+ mice. Only a limited number of epitope IV-specific CD8+ T cells were detected in SV11+ mice not given irradiation before adoptive transfer, and epitope IV-specific CD8+ T cells were not detected in transgene negative littermates (Fig. 4⇑C).
The ability of brain-infiltrating lymphocytes to produce IFN-γ in response to stimulation with peptides corresponding to epitopes I and II/III was tested in a separate experiment. Although no significant response to epitope I was detected, a low percentage (0.1–0.5%) of epitope II/III-specific CD8+ T cells could be detected in the brain-derived lymphocytes from 100-day-old SV11+ mice that had been irradiated and given adoptive transfer with naive C57BL/6 spleen cells at 80 days of age (data not shown). The ability to detect low amounts of epitope II/III-specific CD8+ T cells among brain-derived lymphocytes using peptide-induced IFN-γ production, but not with Db/II/III tetramer (Fig. 4⇑A) might be due to increased sensitivity of the IFN-γ assay. We have shown previously that the reactivity of the Db/II/III tetramer with epitope-specific CD8+ T cells is relatively weak compared with the reactivity of the Db/I and Kb/IV tetramers (11). These results suggest that some SV11+ mice develop a weak epitope II/III-specific CD8+ T cell response that can be detected by staining for peptide-induced IFN-γ production.
As a second measure of T cell function, the lytic capacity of brain- and tumor-infiltrating CD8+ T cells was assessed. Lymphocytes were isolated from a 120-day-old SV11+ mouse that had received irradiation and adoptive transfer with naive C57BL/6 spleen cells at 80 days of age. These lymphocytes were cultured in vitro with irradiated Tag-transformed cells to establish CTL lines. The responding cells lysed B6/WT-19 cells, which express wild-type Tag, but not B6/T122B1 cells, which express a Tag variant in which epitopes I, II/III, IV, and V have been inactivated (Fig. 5⇓A). CTL lines derived from both brain- and tumor-associated lymphocytes lysed RMA cells pulsed with peptide corresponding to epitope IV, but not epitopes I, II/III, and V. The brain- and tumor-derived CTL lines represent distinct isolates, as shown by their differential TCR Vβ region usage in which the brain-derived line expresses Vβ13, and the tumor-derived line expresses Vβ8 (Fig. 5B⇓). These results indicated that CD8+ T cells derived from the brain and tumor of an SV11+ mouse with active immunity to Tag were cytolytic. We have confirmed this result in a second animal (data not shown).
Adoptive transfer with spleen cells from Tag immune C57BL/6 mice leads to increased control of tumor progression
The data presented here indicate that priming against the endogenous Tag in SV11+ mice preferentially leads to the induction of an epitope IV-specific CD8+ T cell response and that this response correlates with the control of tumor progression in SV11+ mice with advanced choroid plexus tumors. If control of choroid plexus tumors is mediated by Tag-specific immunity, then the transfer of spleen cells from Tag-immune C57BL/6 mice should result in increased control of tumors. Thus, 80-day-old SV11+ mice were irradiated and reconstituted with spleen cells from C57BL/6 mice that had been immunized with the Tag-transformed cell line B6/WT-19. The donor cells derived from Tag-immune C57BL/6 mice contained epitope I, II/III, and IV-specific CD8+ CTLs as determined by 51Cr release assay and analysis with MHC tetramers (data not shown). Some SV11+ mice were examined at 110 and 150 days of age to assess the level of tumor burden. This analysis revealed that five of five and two of two mice examined, respectively, had reduced tumor burden, compared with SV11+ mice that received only irradiation at 80 days of age (Fig. 6⇓A, compare with Fig. 1⇑A). In addition, these mice showed a dramatic increase in life span, with a mean survival of 277 days (Fig. 6⇓B). Thus, adoptive transfer with Tag-immune spleen cells results in enhanced tumor immunity compared with the transfer of naive C57BL/6 spleen cells.
The CTL response of SV11+ mice reconstituted with Tag-immune spleen cells was assessed at 100, 110, and 150 days of age. At each time point, CTL specific for epitopes I, II/III, and IV were detected (Fig. 6⇑C), indicating that systemic immunity to the immunodominant epitopes was maintained for an extended period in vivo. To obtain a more quantitative estimate of the CD8+ T cell response, we analyzed lymphocytes from SV11+ mice for staining with MHC tetramers and peptide-induced IFN-γ production. Analysis of lymphocytes from the spleen revealed that epitope I-specific T cells represented 6–7% of CD8+ T cells, while 8–10% of CD8+ T cells were specific for epitope IV (Table I⇓). Analysis of brain-derived CD8+ T cells revealed an increased ratio of epitope IV- vs epitope I-specific cells, suggesting that epitope IV-specific T cells have an advantage in accumulation in the brain. Analysis of epitope II/III-specific CD8+ T cells was not determined using MHC tetramers.
Peptide-induced IFN-γ staining revealed similar numbers of epitope I- and IV-specific CD8+ T cells compared with analysis with MHC tetramers. In addition, significant numbers of epitope II/III-specific CD8+ T cells were detected by IFN-γ production in the spleens and brains of SV11+ mice that received both irradiation and adoptive transfer with Tag-immune C57BL/6 spleen cells. Only low percentages of Tag-specific CD8+ T cells were detected in the spleen and brain of SV11+ mice that received adoptive transfer of Tag-immune spleen cells without prior irradiation, indicating that irradiation leads to enhanced expansion of Tag-specific CD8+ T cells in SV11+ mice. Transfer of Tag-immune spleen cells into irradiated Tag-negative littermates resulted in the expansion of comparatively low numbers of Tag epitope-specific CD8+ T cells compared with the results in SV11+ mice. Thus, efficient expansion of Tag-epitope specific CD8+ T cells required both prior irradiation and expression of endogenous Tag. Importantly, CD8+ T cells specific for all three immunodominant epitopes were localized to the brains of SV11+ mice following transfer of Tag-immune spleen cells.
Immunotherapy of established tumors remains a major challenge in both experimental models and clinical settings. Based on our current understanding of CD8+ T lymphocytes, several factors contribute to the effective induction of tumor-specific immunity. First, a tumor-specific epitope must be generated that can be recognized in complex with the host MHC molecules by CD8+ T cells. This epitope might represent a novel epitope produced due to specific mutations that have occurred during tumorigenesis (1). Alternatively, the epitope might be derived from a normal self protein that is overexpressed in the tumor cells or from a foreign protein that has been introduced into the cell, such as those derived from an oncogenic virus. Second, CD8+ T cell precursors must be present in the host that can respond to this particular epitope. The T cell repertoire specific for self Ags is limited by negative selection in the thymus (27, 28), such that the remaining T cells that can respond to peripheral self Ags have a relatively low avidity (29, 30, 31, 32). In addition, tolerogenic mechanisms active in the periphery might limit the number of T cells that can respond to a particular tumor Ag (29, 33, 34). Presentation of tumor-derived Ags by professional APCs may be critical for efficient activation of naive T cells in the tumor-bearing host, allowing either the direct priming of CD8+ T cells or inducing the proper CD4+ T cell help required for activation of CD8+ T cells (35). Once active, the responding T cells must expand to an effective level, migrate to the tumor site, and persist long enough to eradicate the tumor.
The major finding of this study is that advanced stage choroid plexus tumors arising spontaneously in SV11 Tag transgenic mice remained susceptible to immune-mediated inhibition following the induction of immunity against the transforming protein, Tag. The events that consistently accompanied the successful control of tumor progression were the induction and maintenance of immunodominant epitope IV-specific CD8+ T cells capable of migrating to the tumor site that displayed both cytolytic activity and the ability to produce IFN-γ. Importantly, epitope IV-specific CD8+ T cells were primed efficiently by the endogenous Tag following the transfer of naive C57BL/6 spleen cells into irradiated SV11+ mice. Thus, the accumulation of high levels of epitope IV-specific CD8+ T cells in the brain might partially explain the effectiveness of this approach in the control of aggressively growing tumors at this advanced stage.
Other recent studies have found that T cell tolerance, rather than immunity, is induced following the transfer of naive T cells into mice expressing model self Ags as a transgene. Adoptive transfer of naive TCR transgenic CD8+ T cells into mice that express either OVA or influenza hemagglutinin as a transgene from the insulin promoter resulted in the initial proliferation of CD8+ T cells followed by their deletion (36, 37, 38). Although these studies did not use self Ags expressed on endogenous tumors, a similar result has been found after transfer of CD4+ or CD8+ TCR transgenic T cells into mice containing influenza hemagglutinin-expressing transplantable tumors (25, 39). The basis for CD8+ T cell survival and expansion in our model might be the induction of Tag-specific CD4+ T cell help in SV11+ mice following adoptive transfer with naive C57BL/6 spleen cells. This reasoning is supported by studies demonstrating that specific CD4+ T cell help prevented the deletion of TCR transgenic CD8+ T cells in mice expressing endogenous OVA or containing hemagglutinin-expressing transplantable tumors. (39, 40). We have made a similar observation using line 501 Tag transgenic mice that express Tag from the α-amylase promoter and develop osteosarcomas (41). These mice developed progressive tolerance to the Tag epitopes, including epitope I. The onset of tolerance to epitope I was accelerated in 501/CD4 knockout mice vs 501 mice, indicating that CD4+ T cell help is important for the survival of Tag-specific CD8+ T cells exposed to the endogenous Tag.
Although we did not test specifically for Tag-specific CD4+ T cells in the experiments shown here, we have found previously that Tag-specific Ab can be detected in SV11+ mice given adoptive transfers with C57BL/6 spleen cells (data not shown), indicative of T cell help. Although specific T cell help could potentially come from the endogenous CD4+ T cells of SV11+ mice, we have shown previously that SV40 Tag-specific Abs are detected in only 40% of SV11+ mice following immunization with syngeneic SV40 Tag-transformed cells (42). Thus, the CD4+ T cell compartment of SV11+ mice might be partially tolerant to Tag epitopes. Future studies will address the role of CD4+ T cells in the development of epitope IV-specific CTL responses in SV11+ mice and the control of tumor progression.
Our results demonstrate that irradiation of tumor-bearing SV11+ mice before adoptive transfer of naive C57BL/6 spleen cells led to dramatically enhanced priming of epitope IV-specific CD8+ T cells. Recent studies investigating the role of APCs in the regulation of T cell responses may provide some clues to this immune-enhancing effect of irradiation. Experiments designed to determine the role of the APC in direct stimulation of naive CD8+ T cells revealed that conditioning of APCs via engagement of CD40 led to the induction of CD8+ T cell responses in the absence of CD4+ T cell help (43, 44, 45). This suggested that the APCs require a signal that results in their maturation to induce immunity among naive CD8+ T cells. Dendritic cells (DCs) have been heavily investigated for their role in directing the immune response (35). Under steady state conditions, a subset of DCs has been shown to endocytose material from normal tissues and then migrate to the T cell areas of the regional lymph nodes, but does not induce immunity (46). These results and others have led to the hypothesis that a T cell encounter with resting or immature APCs in the lymphoid tissues, particularly with DCs, might lead to tolerance rather than the induction of immunity (47).
Based on these recent findings it is tempting to speculate that irradiation of SV11+ mice triggers events that result in the activation of DCs for the priming of epitope IV-specific T cells. The finding that irradiation of 80-day-old SV11+ mice alone led to a small increase in the life span of SV11+ mice suggests that irradiation induces some death of tumor cells. Immature DCs that encounter necrotic cells have been shown to take up cell-associated Ags and subsequently mature (48, 49), suggesting that necrotic cells release factors that lead to the maturation of DCs (50). These mature DCs are potent inducers of CD8+ T cell responses. In addition, immature DCs that have captured Ag via apoptotic cells can be induced to mature by proinflammatory stimuli (51). Basu et al. (52) recently demonstrated that maturation of DC is induced by heat shock proteins derived from necrotic cells. In addition, the heat shock protein gp96 has been found to induce the accumulation of mature DC in draining lymph nodes (53). Since heat shock proteins also are capable of carrying tumor Ag-derived peptides (54), these molecules can potentially deliver both tumor Ag and maturation signals to the APC, resulting in T cell activation. In the case of SV11+ mice, irradiation might induce some tumor cell death, which results in Tag uptake and subsequent activation of DCs. Alternatively, irradiation might directly signal the maturation of DCs that have already endocytosed Tag from apoptotic tumor cells under normal conditions (55). Although we did not directly address the site of CD8+ T cell activation, recent evidence suggests that naive donor C57BL/6 spleen cells might need to encounter mature Tag epitope-bearing APCs in the regional lymph nodes to achieve T cell priming (56). It is interesting to note that CD8+ T cells specific for epitopes I, II/III, and IV expanded to much higher levels in irradiated SV11+ mice vs unirradiated SV11+ mice after transfer of Tag-immune spleen cells at 80 days of age, indicative of increased inflammatory conditions.
Epitope IV-specific CTL clearly dominated the CD8+ T cell response in SV11+ mice. The explanation for this dominance over epitopes I and II/III in this setting is unclear. This result could be a reflection of the initial precursor frequency in the donor lymphocytes, the efficiency of epitope presentation, or the susceptibility of CTL precursors to tolerance induction. We have shown previously that the CD8+ T cell response to immunization with Tag in C57BL/6 mice is predominated by epitope IV-specific T cells (11). This predominance is exaggerated depending upon the vehicle used for immunization. Whether the frequency of effector T cells is directly related to the naive precursor frequency has not been determined. Thus, too few epitope I- or epitope II/III-specific CTL precursors may graft into SV11+ mice to be recruited. Another possible explanation is that epitope I- and epitope II/III-specific CTL precursors are tolerized shortly following exposure to endogenous Tag in SV11+ mice, while epitope IV-specific CTL are resistant to tolerance induction and are primed. We recently demonstrated that CTL tolerance to epitopes I and II/III precedes the development of tolerance to epitope IV using C57BL/6-derived mice that express Tag as a transgene from the α-amylase promoter and develop osteosarcomas (41). In addition, epitope I-specific CTL were not induced following adoptive transfer of naive C57BL/6 spleen cells into SV11+ mice and subsequent immunization with a recombinant vaccinia virus expressing full-length Tag or epitope I as a minigene (18). Tolerance to the H2-Kb-restricted epitope IV, however, was not observed in the current study, as evidenced by high levels of epitope IV-specific CD8+ T cell persistence in the brains and spleens of SV11+ mice. Thus, the ability to maintain an active immune response to Tag epitope IV at the tumor site for an extended period of time is associated with the control of tumor progression.
Derivation of CTL from SV11+ mice that had survived for an extended period of time, e.g., > 120 days, occasionally resulted in low levels of activity against epitopes I and II/III. Thus, in a few mice epitope I- and II/III-specific CD8+ T cells might have remained ignorant of the endogenous Tag until the later stages of the immune response. This is reminiscent of the epitope spreading that occurs in the later stages of autoimmunity (57), which recently has been shown to occur in the CD8+ T cell response to tumors (58, 59, 60). Several studies using transgenic mouse models have revealed that precursor T lymphocytes can remain ignorant of endogenous Ags unless inflammatory conditions are induced (61, 62, 63, 64, 65). Our results differ from these studies in that the donor lymphocyte population was derived from normal mice, which contained CTL precursors specific for multiple epitopes.
A significant proportion of the CD8+ T lymphocytes isolated from the brains of SV11+ mice that were irradiated and reconstituted with naive C57BL/6 spleen cells were specific for epitope IV and displayed an activated phenotype. In addition, a population of CD8+ T cells was detected in the brain that did not stain with the epitope IV tetramer. While these cells also displayed an activated phenotype, they were not specific for Tag epitopes I or II/III. In addition, CTL specific for the immunorecessive epitope V were not detected in any SV11+ mice that had been irradiated and reconstituted with naive C57BL/6 spleen cells. The possibility that these CD8+ T cells might be specific for a previously uncharacterized Tag epitope seems unlikely, since CTL lysis of a cell line expressing a Tag variant in which epitopes I, II/III, IV, and V had been inactivated were not detected (data not shown). The brain-infiltrating CD8+ T cells might represent bystander cells that have been drawn into the immune milieu of an active response to the tumor. Regardless, their specificity remains unknown.
Multiple studies have assessed the ability of preimmune lymphocytes to control the progression of tumors in vivo. Our results demonstrate that the adoptive transfer of irradiated SV11+ mice with Tag-immune lymphocytes results in a dramatic increase in the life span of SV11+ mice compared with transfer of lymphocytes from naive C57BL/6 mice. The basis for this increased response might be attributed to the presence of CTL specific for multiple Tag epitopes compared with only epitope IV-specific CTL. Significant numbers of CD8+ T cells specific for epitopes I, II/III, and IV were detected in the brains of irradiated SV11+ mice that received Tag immune spleen cells. Alternatively, the transfer of previously activated T cells might reduce the lag time between adoptive transfer and the recruitment of an effective number of T cells to the tumor site, such that tumors are targeted more rapidly than in mice that received unprimed lymphocytes.
The results of this and previous studies investigating immune control of spontaneously arising tumors demonstrate that although significant control of tumor progression was achieved, all mice eventually succumbed to tumors. Multiple mechanisms have been suggested to explain the escape of tumors from immune control including down-regulation of MHC molecules or components of the Ag-processing machinery (66), the outgrowth of tumors expressing variants of the original T cell epitope, the production of immunosuppressive factors by the tumor or tumor microenvironment, and the direct killing of responding T cells by the tumor (67). In a previous study we demonstrated that tumor cell lines derived from choroid plexus tumors of SV11+ mice that developed after the induction of Tag-specific CTLs express high levels of MHC class I and Tag and are lysed efficiently by CTL clones specific for the wild-type Tag epitopes. Roy et al. (68) have recently demonstrated that advanced stage choroid plexus tumors as well as the surrounding tumor vascular in SV11+ mice are capable of producing the immunosuppressive factor TGF-β2, providing a basis for down-regulation of an active immune response.
The ability of some tumors to remain sequestered from the immune system also has been observed (56, 69). Ganns et al. (69) demonstrated that the microenvironment of Tag-induced pancreatic tumors is not susceptible to infiltration by Tag-specific CD4+ TCR transgenic T cells unless the T cells are activated ex vivo and the tumor-bearing animals are irradiated before reconstitution. Preactivated T cells did not infiltrate tumors from mice that received adoptive transfers without prior irradiation. The authors suggest that the conditioning of tumors using ionizing radiation allowed lymphocyte infiltration due to an increase in the formation of high endothelial venules. Thus, in addition to the suggestion that irradiation might result in tumor cell death in SV11+ mice and subsequent activation of naive T lymphocytes, irradiation also might condition the tumors for access by immune cells.
Our results indicate that the control of choroid plexus tumor progression is associated with the recruitment of high levels of Tag epitope-specific CD8+ T cells. One common finding among a number of studies investigating the role of epitope-specific CTLs in the control of spontaneously arising tumors in Tag transgenic mice has been the transient nature of the inhibitory effect (63, 70, 71). This control of tumor progression subsided even though tumor Ag-specific memory T cells remained that could be reactivated. Together these results suggest that the CD8+ T cell response to tumor Ag is down-regulated after the initial response to the tumor. Without reinitiation of the conditions that led to the first response, the tumor continues to progress without activating the CTL memory response. Thus, approaches that limit down-regulation of the immune response, such as administration of anti-CTLA-4 mAbs (72), or repetitive boosting of the immune response (63) might further prolong the response against Tag-induced tumors.
We thank Melanie Epler and Andrew Gaydos for excellent technical assistance, and Dr. Sandra Hutchinson for helpful comments and discussion.
↵1 This work was supported by Research Grant CA25000 from the National Cancer Institute, National Institutes of Health (to S.S.T.), and a grant from the Four Diamonds Fund (to T.D.S.).
↵2 Address correspondence and reprint requests to Dr. Todd D. Schell, Department of Microbiology and Immunology, H107, Pennsylvania State University College of Medicine, 500 University Drive, Hershey, PA 17033. E-mail address:
↵3 Abbreviations used in this paper: Tag, large tumor Ag; CD62L, CD62 ligand; DC, dendritic cells; SV11+, SV11 transgene positive.
- Received August 27, 2001.
- Accepted October 18, 2001.
- Copyright © 2001 by The American Association of Immunologists