| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Microbiology and Immunology, Pennsylvania State University College of Medicine, Hershey, PA 17033; and
The Jackson Laboratory, Bar Harbor, ME 04609
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
TCD8 recognize peptides bound to MHC class I molecules on the tumor cell surface. These peptides are derived from the processing of intracellular proteins, with each MHC class I allele binding different combinations of the available peptides for presentation at the cell surface (23). Thus, a wide variety of peptide/MHC complexes is presented at the tumor cell surface, and the tumor-specific TCD8 response can be targeted against multiple epitopes simultaneously, although with varying efficiencies (24). This phenomenon, known as immunodominance, is controlled by multiple factors affecting Ag processing and presentation as well as the T cell repertoire of the host (25). Because multiple tumor epitopes could potentially be targeted, assessing the nature of the immune response toward epitopes that lead to the most effective control of tumor progression in experimental models will provide clues toward better approaches to the immunotherapy of cancer.
Our previous studies using two different transgenic mouse lines expressing SV40 large tumor Ag (T Ag) indicate that immunodominance among T Ag epitopes contributes toward the efficiency of TCD8-mediated control of progressively growing tumors due to differential inactivation of T Ag-specific TCD8 (3, 6). SV40 T Ag encodes four H-2b-restricted epitopes (26). Epitopes I (residues 206–215), II/III (residues 223–231), and V (residues 489–497) are H-2Db (Db)-restricted, whereas epitope IV (residues 404–411) is H-2Kb (Kb)-restricted. These four T Ag-derived, TCD8-recognized epitopes fall on a hierarchical scale, with epitope IV being the most immunodominant followed by epitopes I and II/III (27, 28). Epitope V is immunorecessive and not immunogenic unless the three dominant epitopes are inactivated in T Ag (29, 30). The ability of TCD8-specific for T Ag to control tumor progression was established by showing that the transfer of TCD8 specific for the immunodominant T Ag epitopes into SV11 T Ag transgenic mice induced regression of advanced stage spontaneous tumors (31).
To better understand the role of the endogenous TCD8 repertoire in the control of spontaneous tumor progression, we took advantage of the well-characterized tumor model of rat insulin promoter (RIP) 1-Tag4 transgenic mice on the C57BL/6 (B6) background (32). In RIP1-Tag4 mice, T Ag is expressed as a transgene from the rat insulin II promoter in pancreatic 
cells, which leads to highly predictable formation of cell tumors. The transformation of cells by T Ag is a gradual process, initially inducing development of islet hyperplasia by three months of age and followed by the appearance of macroscopic insulinomas between 5 and 6 mo of age (18, 33). The average life span of RIP1-Tag4 mice is 
200 days, with 100% of mice developing tumors. Transformed cells secrete high amounts of insulin leading to the development of hypoglycemic shock and death of tumor-bearing mice. Previously, Knowles and colleagues (18) demonstrated that immunization of RIP1-Tag4 mice with SV40 before or during T Ag-induced tumorigenesis resulted in the development of TCD8 capable of lysing syngeneic T Ag-transformed cells in vitro. However, only immunization before the expression of T Ag in the cells (at 3 wk of age) led to long-term protection against tumor development. Whether the composition of T Ag-specific TCD8 response was altered by expression of the endogenous T Ag was not addressed. Although this previous study demonstrated that the presence of T Ag-specific TCD8 is associated with long-term control of tumor progression in RIP1-Tag4 mice, it remained to be determined whether peripheral tolerance limits TCD8 responding to individual T Ag epitopes over the course of tumor progression and whether long-term protection from tumor progression is associated with a particular T Ag epitope.
Our results show that activation of endogenous TCD8 specific for the dominant SV40 T Ag epitope IV before T Ag expression in the pancreatic cells leads to lifelong surveillance against T Ag-induced spontaneous tumors. Importantly, functional epitope IV-specific but not epitope I-specific TCD8 persist in the host without inducing tissue destruction or diabetes in RIP1-Tag4 mice.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
B6 (H-2b) mice were purchased from The Jackson Laboratory, and B6.SJL mice (CD45.1+) were procured from Taconic Farms. Line RIP1-Tag4 mice are on the B6 background and express the SV40 early region as a transgene under control of the rat insulin promoter and were described previously (32). TCR-I mice express epitope I-specific TCR
and TCR-chains as a transgene (15) and are available from The Jackson Laboratory as line B6.Cg-Tg(Tcra TcrB)416Tev/J. Approximately 95% of TCD8 express the TCR transgenes. All mice were maintained at the animal facility of the Milton S. Hershey Medical Center (Hershey, PA). Animal studies were performed in accordance with guidelines established by the Pennsylvania State University College of Medicine Institutional Animal Care and Use Committee (Hershey, PA) under an approved protocol.
Reagents and immunizations
The B6/WT-19 cell line was derived previously by transformation of B6 mouse embryo fibroblasts with wild-type SV40 strain VA45-54 (34). B6/15Bb cells express a T Ag variant in which epitopes I, II/III, and V have been inactivated by site-directed mutagenesis but which encodes wild-type epitope IV (27). Cell lines were maintained in DMEM supplemented with 100 U/ml penicillin, 100 µg/ml streptomycin, 100 µg/ml kanamycin, 2 µM L-glutamine, 10 mM HEPES, 0.075% (w/v) NaHCO3, and 5–10% FBS. Peptides were synthesized at the Macromolecular Core Facility of the Milton S. Hershey Medical Center by Fmoc chemistry using an automated peptide synthesizer (9050 MilliGen PepSynthesizer; Millipore). Peptides used in these experiments correspond to the SV40 T Ag epitope I (SAINNYAQKL; 206–215), epitope IV containing a Leu replacement of Cys at residue 411 (VVYDFLKL; 404–411), and an epitope from HSV glycoprotein B (gB) (SSIEFARL; 498–505). Immunization of mice with T Ag-expressing cell lines was conducted by i.p. injection of 3–5 x 107 cells in 0.5 ml of HBSS at the indicated ages.
Flow cytometric analysis and in vitro stimulation of bulk CTL cultures
Ex vivo staining of TCD8 with MHC tetramers and primary conjugated Abs was performed on single-cell suspensions prepared from spleens as described previously (27). For dissociation of tumors and pancreas, tissue was minced with a razor blade and incubated in HBSS containing 1% FBS, 1 mg/ml collagenase D (Roche), and 50 U/ml DNase I (Roche) for 1 h at 37°C. Any remaining clumps were dissociated by gentle pipetting, and cells were washed before tetramer staining. In some cases, spleen cells were cultured in vitro for 6 days in the presence of 
-irradiated B6/WT-19 stimulator cells as described previously (3). After staining, cells were fixed with 2% paraformaldehyde in PBS and analyzed using a FACScan or FACSCalibur flow cytometer (BD Biosciences). Data were analyzed using FlowJo software (Tree Star). Production and characterization of the Db/T Ag epitope I (Db/I), Kb/T Ag epitope IV (Kb/IV), Db/influenza virus (Flu) nucleoprotein (NP) epitope 366–374 (Db/Flu), and Kb/HSV gB epitope 498–505 (Kb/gB) tetramers were described previously (27). The following mAbs were purchased from BD Pharmingen: PE- and CyChrome-labeled rat anti-CD8 (clone 53-6-7), FITC-labeled rat anti-mouse IFN- (clone XMG1.2), and PE-labeled anti-mouse CD45.1 (clone A20). The percentage of CD8+ cells that stained positive for the T Ag-specific tetramer was determined by subtracting the percentage of cells that stained positive with the control Db/Flu or Kb/gB tetramers within the same population.
Adoptive transfer and proliferation assay
Adoptive transfers were performed using sex-matched, RBC-depleted lymphocytes derived from spleens and lymph nodes of TCR-I transgenic mice. Recipient mice received i.v. injections of lymphocytes derived from spleen and lymph nodes of TCR-I mice that contained 1–2 x 106 clonotypic CD8+ T cells as determined by prior staining with MHC tetramers. Mice were immunized the following day. For proliferation assays, TCR-I-derived lymphocytes were labeled with 5 µM CFSE (Invitrogen Life Sciences) for 10 min at 37° C in PBS and 0.1% BSA followed by three washes before i.v. injection into recipient mice. Proliferating cells were detected after three days by measuring the mean fluorescent intensity of CFSE staining retained in CD8+ Db/I tetramer-positive cells recovered from the spleens of recipient mice.
Intracellular cytokine assays
For staining of intracellular IFN-, RBC-depleted lymphocyte suspensions were prepared as described above and incubated at 5 x 106/ml with the indicated synthetic peptides (1 µM) representing T Ag or control epitopes and 1 µg/ml brefeldin A in complete RPMI 1640 containing 10% FBS for 4 h at 37°C in 5% CO2. The Cytofix/Cytoperm kit (BD Pharmingen) was used to stain for IFN- production according to the manufacturer’s instructions and as described previously (27). For peptide titration assays, splenocytes from B6/WT-19-immunized mice were restimulated in vitro with -irradiated B6/WT-19 cells for 5 days at 37° C with 5% CO2. Responder cells were plated at 5 x 105 cells/well in the presence of titrated amounts of the epitope IV or control Kb-binding peptide 498–505 from HSV gB (35) and 1 µg/ml brefeldin A. After a 6-h incubation at 37°C with 5% CO2, cultures were stained as described above.
In vivo cytotoxicity assay
Splenocytes from B6.SJL mice (CD45.1+) were incubated in the presence of the indicated peptides (1 µM) in RPMI 1640 with10% FBS at 37°C for 90 min and washed three times to remove excess peptide. Peptide-coated targets were then labeled with varying concentrations (5 µM CFSEhigh, 0.5 µM CFSEmedium, and 0.025 µM CFSElow) of CFSE (Invitrogen Life Sciences) for 10 min at 37°C in PBS and 0.1% BSA, washed 2 times, and mixed together at 1:1:1 ratio, and then 6 x 106 cells were injected i.v. into the tail vein of sex-matched mice in 0.2 ml HBSS. The elimination of CFSE-labeled targets was assessed after 16 h by analyzing spleen cells for the presence of CD45.1+CFSE+ cells. The following formula was used: percentage of specific killing = [1 – (ratio of control mice/ratio of immunized mice) x 100], where ratio equals the percentage of CFSElow/CFSEhigh or medium. CFSElow targets were pulsed with control peptide gB and used as an internal control for nonspecific lysis.
Tumor histology, immunohistochemistry, and measurement of islet size
For T Ag immunohistochemistry, pancreata were fixed with 10% buffered formalin before being embedded in paraffin blocks. Eight-micrometer-thick sections were cut on a microtome and collected onto positively charged slides. Sections were deparaffinized in xylene and rehydrated in ethanol. After two washes in PBS, slides were treated with Ag-unmasking solution (Vector Laboratories) for 30 min in a steamer. Cooled slides were washed in PBS, incubated with 10% normal goat serum in PBS containing 0.1% Tween 20 for 30 min, and washed and incubated with primary anti-T Ag Ab for 1 h at room temperature. Primary Ab consisted of an equal mixture of mAb Pab419 (36) and Pab901 (37), culture supernatants that bind to epitopes in the amino-terminal and carboxyl-terminal portions of SV40 T Ag, respectively. Sections were then incubated for 1 h with biotinylated anti-mouse Ab and visualized with peroxidase/diaminobenzidine detection (ChemMate detection kit; Ventana) for 2–10 min, as needed. Parallel sections were stained with H&E. Sections were examined using a Nikon Microphot-FXA microscope, and representative images were captured using a Sony DKC-ST5 digital camera. To measure the size of individual islets, H&E-stained pancreatic sections were examined under the microscope using a micrometer. The radius of all islets per section was measured, and their individual sizes were plotted. Ten sections were examined for each mouse.
Life span analysis and measurement of blood glucose levels
Mice in life span analyses were monitored for symptoms of hypoglycemia, which are indicative of advanced stage tumors, and sacrificed when blood glucose levels dropped below 50 mg/dl. Blood glucose levels were measured using an Elite glucometer (Bayer). Kaplan-Meier survival plots were constructed using Prism software (GraphPad).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
To determine whether TCD8 specific for the immunodominant T Ag epitopes remain sensitive to immunization in aged RIP1-Tag4 mice, groups of three RIP1-Tag4 mice aged 35, 90, and 180 days as well as 90-day-old B6 mice were immunized i.p. with T Ag-transformed B6/WT-19 cells that express wild-type T Ag. Nine days later, splenic TCD8 lymphocytes were stained with anti-CD8 Ab and MHC class I tetramers capable of detecting T cells specific for epitope I (Db/I) or epitope IV (Kb/IV). A tetramer composed of the influenza NP epitope NP366–374 (Db/Flu) was used as a control tetramer. The data in Fig. 1A show that epitope IV-specific T cells expanded in 35-day-old RIP1-Tag4 mice to levels similar to those for the control B6 mice. The number of epitope IV-specific TCD8 recruited by immunization gradually declined in older RIP1-Tag4 mice but remained significant even at 180 days of age. In contrast, the TCD8 response against epitope I already was decreased in 35-day old RIP1-Tag4 mice compared with B6 mice and was undetectable by tetramer staining in 180-day-old RIP1-Tag4 mice.
|
in response to short-term incubation with epitope IV peptide as determined by intracellular cytokine staining (data not shown). Staining of tumor-derived TCD8 with the Db/I tetramer did not show positive cells above the background staining obtained with control Db/Flu tetramer. These results show that a proportion of epitope IV-specific but not epitope I-specific TCD8 survive peripheral tolerance in tumor-bearing RIP1-Tag4 mice and remain responsive to immunization, resulting in the infiltration of pancreatic tumors. Epitope I-specific TCR transgenic T cells are rapidly tolerized in RIP1-Tag4 mice
Following T Ag immunization of 5-wk-old naive RIP1-Tag4 mice, the response to epitope IV is 15-fold greater than that directed toward epitope I (Fig. 1). Whether this difference can be traced to fewer precursors in the naive T cell repertoire targeting epitope I vs epitope IV is unknown. If such a discrepancy exists, however, the time required to induce tolerance among epitope I-specific TCD8 vs epitope IV-specific TCD8 may be significantly shorter. To determine whether an increased frequency of epitope I-specific TCD8 would lead to a more enduring response in RIP1-Tag4 mice, we used adoptive transfer of naive TCD8 from epitope I-specific TCR transgenic mice in which >95% of the TCD8 are specific for epitope I (15). In a previous study, TCR-I transgenic T cells were shown to be triggered by the endogenous T Ag after transfer into mice expressing T Ag as a transgene from the -amylase promoter (15). To monitor the consequences of TCR-I T cell transfer into RIP1-Tag4 mice, CFSE-labeled TCR-I-derived lymphocytes were transferred into RIP1-Tag4 mice at various ages representing the various stages of tumor progression. The ability of the transferred TCD8 to proliferate in response to the endogenous T Ag was monitored after 3 days. The frequency of TCR-I cells detected in the spleens of RIP1-Tag4 mice increased slightly over that detected in B6 mice, with the highest number of cells detected in 6-mo-old RIP1-Tag4 mice (Fig. 2A, upper panels). This increase corresponded with the presence of proliferating cells in the RIP1-Tag4 mice (Fig. 2A, bottom panels). Although TCR-I T cells initially proliferated in RIP1-Tag4 mice, these cells were rapidly eliminated by day 10 as evidenced by only minimal detection in the spleen (Fig. 2B) and lack of detection in the pancreas (data not shown) of 6-mo-old mice. Similar results were obtained when TCR-I cells were transferred into younger mice (data not shown). These results indicate that epitope I-specific TCR transgenic TCD8 are rapidly activated and eliminated from RIP1-Tag4 hosts following recognition of the endogenous T Ag, consistent with loss of the endogenous population of epitope I-specific TCD8.
|
1% of splenic TCD8 following immunization of RIP1-Tag4 mice (Fig. 1A). This level was increased to 9.5% of TCD8 following adoptive transfer with TCR-I cells and subsequent immunization (Fig. 2C), a level comparable to that of the epitope IV-specific TCD8 that accumulate from the endogenous T cell repertoire (Fig. 1A; 3 mo). We noted that 2.5-fold more epitope I-specific TCD8 accumulated in the spleens of B6 mice compared with RIP1-Tag4 mice following adoptive transfer of TCR-I cells and immunization (Fig. 2C). By 40 days postimmunization, epitope I-specific TCD8 had decreased below the limit of detection by tetramer analysis in RIP1-Tag4 mice but represented 5% of CD8+ cells in B6 mice, a 5-fold decrease from day 10. To determine whether a residual population of TCR-I T cells was present in RIP1-Tag4 mice at this later time point, some mice were given a booster immunization with B6/WT-19 cells at day 34, 6 days before analysis. The results demonstrate that epitope I-specific TCD8 remained undetectable in RIP1-Tag4 mice despite extensive expansion in B6 mice. Thus, immunization successfully induces the expansion of TCR-I T cells in RIP1-Tag4 mice leading to initial accumulation at levels similar to that obtained with the endogenous epitope IV-specific TCD8, but these TCD8 fail to persist long term despite this increased frequency. Early immunization of RIP1-Tag4 mice against epitope IV prevents the development of tumors without inducing diabetes
Because we observed that TCD8 specific for epitope IV but not epitope I could be triggered even in tumor-bearing RIP1-Tag4 mice (Fig. 1A), we determined whether specific immunization against the immunodominant epitope IV alone could mediate control of tumor progression in RIP1-Tag4 mice. A group of 35-day-old RIP1-Tag4 mice were immunized i.p. with B6/15Bb cells expressing a T Ag variant (epitope IV-only) containing epitope IV but lacking the other H-2b-restricted epitopes I, II/III, and V (27). Immunization with B6/15Bb cells results in a quantitatively similar TCD8 response to epitope IV as observed following immunization with B6/WT-19 cells expressing the wild-type T Ag (data not shown). Another group of RIP1-Tag4 mice was immunized at 180 days of age and survival was monitored. The data in Fig. 3A demonstrate that early immunization at 35 days of age with epitope IV-only T Ag-expressing cells prevented the progression of tumors, because all mice in the group (n = 8) lived for two years, at which time these mice were euthanized for further analysis (see below). In contrast, immunization of RIP1-Tag4 mice with epitope IV-only T Ag-expressing cells at 180 days of age was not effective at controlling tumor progression (n = 8, median 204 days), and their life span was similar to untreated RIP1-Tag4 mice (n = 10, median 194 days). Thus, immunization of RIP1-Tag4 mice at 35 days of age against epitope IV results in a normal disease-free life span, whereas immunization of tumor-bearing mice had a minimal effect on survival.
|
Immunization against the T Ag might be expected to induce an autoimmune phenotype and subsequent diabetes if cells are specifically destroyed by responding immune cells as has been observed in similar transgenic models (39, 40). However, we found that mice immunized with B6/15Bb cells at 5 wk of age that survived long term maintained normal levels of blood glucose (Fig. 3B), indicative of continued cell function. In contrast, naive RIP1-Tag4 mice became hypoglycemic by 27 wk of age, consistent with the progression of cell tumors producing high levels of insulin (32). Thus, immunization against T Ag epitope IV results in continuous and long-term protection from tumor progression without disrupting pancreatic function.
Immunized RIP1-Tag4 mice survive long term despite continued expression of T Ag
Long-term survival of RIP1-Tag4 mice immunized with T Ag-transformed cells at 35 days of age could be explained by the loss of cells expressing the T Ag transgene. To assess this possibility, we determined whether long-term surviving RIP1-Tag4 mice that were immunized at 35 days of age maintain T Ag expression in the pancreatic islets as detected by immunohistochemistry (Fig. 4). Islet cells from unimmunized B6 mice lacked T Ag expression (Fig. 4C), whereas the majority of cells in the islets of 6-mo-old naive RIP1-Tag4 mice were positive for T Ag expression (Fig. 4D). We then analyzed pancreatic tissue obtained from long-term surviving RIP1-Tag4 mice that were immunized at 35 days of age. The histological analysis of pancreatic sections revealed that T Ag-positive cells were detected in the islets of 1-year-old (Fig. 4, G and H) and 2-year-old (Fig. 4, K and L) RIP1-Tag4 mice. H&E staining of pancreatic tissue from these T Ag-immune mice showed that most islets were smaller than the islets found in naive 6-mo-old RIP1-Tag4 mice, although some hyperplastic islets were slightly larger (Fig. 4F). It should be noted that the number of islet cells positive for T Ag from immunized mice was generally reduced (Fig. 4, G, K, and L) compared with the number of T Ag-positive cells detected in tumors from unimmunized RIP1-Tag4 mice (Fig. 4D) unless a particular islet appeared hyperplastic (Fig. 4H). Hyperplastic islets were rare, however, in immunized mice (Fig. 3D). Thus, despite continuous expression of T Ag in the islets, large tumors did not develop in immunized mice. These results demonstrate that immunization at 35 days of age with T Ag epitope IV-only expressing cells prevented the development of tumors in RIP1-Tag4 mice. These data also suggest that T Ag-expressing islet cells may be susceptible to TCD8-mediated elimination but that enough cells persist to produce insulin over the lifespan of the mice.
|
We next determined whether epitope IV-specific TCD8 induced by immunization with Tag IV-only cells at 35 days of age persist and remain functional in 2-year old RIP1-Tag4 survivor mice. Representative survivors were sacrificed to quantitate the epitope IV-specific TCD8 present in the spleens and pancreas by tetramer staining and to examine T cell function by in vivo cytotoxicity assay (Fig. 5A). This analysis revealed that epitope IV-specific TCD8 were present at detectable levels in the spleens (8.1% of total TCD8) of long-term survivors and expanded dramatically (53% of total TCD8) following in vitro culture (Fig. 5A; group 5). Analysis of epitope IV-specific in vivo cytotoxicity in the same animal showed that epitope IV-pulsed targets were completely eliminated. To determine whether these persisting epitope IV-specific TCD8 could expand following antigenic challenge in vivo, representative mice were immunized with B6/WT-19 cells. After nine days, epitope IV-specific TCD8 had expanded to 15% of TCD8 in the spleen (Fig. 5A; group 6), indicating that these persisting TCD8 remained sensitive to antigenic challenge. Epitope IV-pulsed target cells were also completely eliminated in these mice although no epitope I-specific lysis was detected, indicating that these RIP1-Tag4 mice remained fully tolerant to epitope I despite the lack of tumor progression. Thus, in contrast to the fate of epitope I-specific TCD8, epitope IV-specific TCD8 primed from the endogenous repertoire by specific immunization persist long term in RIP1-Tag4 mice and remain functional even in aged mice.
|
Epitope IV-specific TCD8 activated in tumor-bearing mice are not functionally compromised
One possible explanation for the inability of immunization against T Ag to induce control of advanced-stage tumors in RIP1-Tag4 mice is that the responding epitope IV-specific TCD8 in older mice are rendered anergic or that higher avidity TCD8 are deleted following recognition of self Ag (7, 15, 41, 42, 43). Thus, we determined whether RIP1-Tag4 mice immunized at 6 mo of age could eliminate target cells pulsed with epitope IV peptide. Epitope IV-specific TCD8 were recruited to 2.2% of TCD8 following immunization of 171-day-old RIP1-Tag4 mice with wild-type T Ag-expressing B6/WT-19 cells (Fig. 5A; group 4). In addition, epitope IV-pulsed targets were completely eliminated in the in vivo killing assay, whereas no significant lysis of epitope I-pulsed target cells was observed. In contrast, both epitope I-specific and IV-specific killing was detected in control B6 mice immunized with B6/WT-19 cells (Fig. 5A; group 2). We considered that because epitope IV-specific TCD8 persisted in 6-mo-old naive RIP1-Tag4 mice, some spontaneous priming against the T Ag in tumor-bearing RIP1-Tag4 mice might occur. Spleen cells from unimmunized 6-mo-old RIP1-Tag4 mice were analyzed directly by tetramer staining. The results in Fig. 5A (group 3) show that although epitope IV-specific TCD8 were not detected directly ex vivo by MHC tetramer analysis, a modest level of epitope IV-specific killing was detected (39%). Moreover, epitope IV-specific TCD8 were expanded after in vitro culture (6.2% of TCD8), indicating that RIP1-Tag4 mice spontaneously develop low numbers of activated epitope IV-specific TCD8, a finding consistent with those reported by Ye et al. (18) in which SV40-specific CTL activity was detected ex vivo. No epitope IV-specific TCD8 were detected following in vitro culture of splenocytes from naive B6 mice (Fig. 5A; group 1). Thus, the TCD8 triggered by immunization in tumor-bearing RIP1-Tag4 mice remain functional in vivo.
To determine whether the avidity of epitope IV-specific TCD8 was altered in tumor-bearing RIP1-Tag4 mice, we immunized groups of 2-mo-old and tumor-bearing 8.5-mo-old RIP1-Tag4 mice as well as nontransgenic B6 mice with B6/WT-19 cells expressing wild-type T Ag. After 10 days, spleen cells were harvested and cultured for 5 days in vitro with irradiated B6/WT-19 cells to expand the low number of cells that are detected in the older RIP1-Tag4 mice (Fig. 1). All mice in the 8.5-mo-old group had extensive tumor burden, whereas those in the 2-mo-old group lacked detectable tumors (data not shown). Cells in the responding lymphocyte cultures were tested for their ability to produce IFN- in response to titrated amounts of epitope IV peptide in an intracellular cytokine assay. The results demonstrate that TCD8 derived from tumor-bearing RIP1-Tag4 mice recognized epitope IV peptide as efficiently as those derived from 2-mo-old RIP1-Tag4 mice (Fig. 6). In addition, the avidity of epitope IV-specific TCD8 derived from RIP1-Tag4 mice did not vary from B6-derived TCD8. These results indicate that the avidity and function of epitope IV-specific TCD8 does not change significantly as tumors progress in RIP1-Tag4 mice, although the frequency of TCD8 that can be recruited decreases with age (Fig. 1).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
The overall goal of this study was to determine the extent to which peripheral tolerance affects the ability to recruit TCD8 specific for two immunodominant T Ag epitopes derived from a single oncogenic protein and whether control of spontaneous tumor growth correlates with the presence of one or both of these TCD8 populations. Our results demonstrate that tumor progression in RIP1-Tag4 mice can be controlled for the natural lifespan by targeting TCD8 specific for the single immunodominant Kb-restricted epitope IV only if mice are immunized at the time of initial oncogene expression. Control of tumor development is associated with the persistence of functional epitope IV-specific TCD8 in the peripheral lymphoid organs as well as in the pancreas. Despite the persistence of epitope IV-specific TCD8 in these mice, TCD8 to a second T Ag epitope, the Db-restricted epitope I, are deleted or anergized during the early stages of tumor development as a result of their interaction with the endogenous T Ag, even when the precursor frequency is dramatically increased by the addition of epitope I-specific TCR transgenic T cells. Thus, the kinetics of tolerance is epitope-dependent, even though both epitopes are derived from the same protein.
The finding that complete TCD8 tolerance toward epitope I occurs despite only partial loss of responsiveness to epitope IV is consistent with a previous study using a different line of SV40 T Ag transgenic mice in which epitope I-specific TCD8 were shown to be more rapidly tolerized by the endogenous T Ag than epitope IV-specific TCD8. Line 501 mice express T Ag from the -amylase promoter, resulting in the formation of osteosarcomas around 1 year of age (44). Similar to RIP1-Tag4 mice, TCD8 specific for both T Ag epitopes are initially detected following immunization of young 501 mice (6), but these mice become tolerant to epitope I by 6 mo of age when high levels of T Ag expression are detected. Epitope IV-specific TCD8 were efficiently recruited in 501 mice at 6 mo of age, but unlike the present study these cells were eventually tolerized following the appearance of tumors. One possible mechanism to explain the early loss of epitope I-specific TCD8 in these two different T Ag transgenic mouse models is that a difference in the TCD8 precursor frequency specific for epitopes I and IV may lead to more rapid loss of epitope I responsiveness. This suggestion is based on the finding that epitope IV-specific TCD8 accumulate at higher levels than epitope I-specific TCD8 in both B6 and RIP1-Tag4 mice following immunization with wild-type T Ag. Whether the magnitude of the response can be attributed to differences in the naive precursor frequency or to other mechanisms that control immunodominance remains to be determined. Our experiments here tested this possibility by seeding in large numbers of TCR-I T cells to increase the naive precursor frequency as well as the frequency of responder cells following immunization. In both cases, the epitope I-specific TCD8 were lost rapidly from the RIP1-Tag4 hosts, suggesting that the differential tolerance observed is not explained simply by the presence of fewer epitope I-specific TCD8. Rather, our data support a model in which TCD8 specific for epitope I may be more susceptible to peripheral tolerance than TCD8 specific for epitope IV. Some factors that have been shown to contribute to peripheral T cell tolerance include the level of Ag expression (45, 46), lack of CD4+ T cell help (47), and the avidity of self-reactive T cells (7). Because both epitopes are derived from the same protein, it is unlikely that protein levels contribute toward this difference. Epitope I, however, is known to form highly stable complexes with Db (48), perhaps leading to prolonged high level presentation in vivo that results in more rapid deletion of the responding T cells. Whether endogenous T cells specific for epitopes I and IV from RIP1-Tag4 mice have comparable avidities is unknown.
A unique finding in this study is that a proportion of epitope IV-specific TCD8 survive the effects of peripheral tolerance in RIP1-Tag4 mice even after the appearance of tumors. This contrasts with other models of T Ag-induced spontaneous cancer in which epitope IV-specific TCD8 are eventually tolerized in mice with progressively growing tumors (6, 21). Epitope IV-specific TCD8 are apparently deleted from the repertoire of line 501 mice developing osteosarcomas, as they are undetectable following immunization of tumor-bearing animals but respond to immunization in age-matched littermates without tumors (6). This difference might be due to contrasting properties of osteosarcomas developing in 501 mice and insulinomas in RIP1-Tag4 mice. Tumors in 501 mice appear at a later age, grow invasively into surrounding tissues, and are metastatic (44). In contrast, tumors in RIP1-Tag4 mice do not infiltrate surrounding tissues and do not form metastases. In addition, T Ag is expressed in multiple tissues in 501 mice, whereas expression is localized to the pancreatic cells in RIP1-Tag4 mice.
A different scenario is observed when a silent T Ag transgene, expressed from the -actin/-globin promoter, is activated and leads to sporadic tumor formation in various tissues (21). In this case, epitope IV-specific TCD8 appear to be rendered noncytolytic by the tumor environment, perhaps due to the presence of immunosuppressive cytokines such as TGF-. This finding contrasts with our observation in which epitope IV-specific TCD8 remained cytotoxic even in mice with advanced-stage tumors. These differences are a reminder that unique tumor environments can differentially affect TCD8 responding to the same tumor-specific epitope and that the results cannot be generally applied to all tumor models.
Tumor progression in the presence of competent epitope IV-specific TCD8
We found no evidence that tumor progression in naive RIP1-Tag4 mice altered the quality of epitope IV-specific TCD8, although the number of responding cells is reduced 3-fold in tumor-bearing mice vs 5-wk old mice. Indeed, epitope IV-specific TCD8 generated in tumor-bearing mice acquired in vivo killing activity and the ability to produce IFN- and had a similar avidity for epitope IV peptide as those from normal B6 mice. Yet these TCD8 failed to significantly control tumor growth. A different result was obtained in mice coexpressing the lymphocytic choriomeningitis virus (LCMV) glycoprotein and T Ag from RIP in which immunization of tumor-bearing mice with LCMV was shown to provide a delay in tumor progression that waned following elimination of the virus (49). Similar to our findings, tumor progression continued despite the presence of persisting functional TCD8 targeting the immunodominant LCMV epitope. Our findings also contrast with those from some other tumor systems (50) in which tumor-infiltrating TCD8 were impaired in their ability to effectively destroy progressing tumors due to production of immunosuppressive cytokines (51, 52), inhibition by regulatory T cells (53, 54, 55), induction of apoptosis (56), or failure to infiltrate existing tumors (22). Whether TCD8 within the tumor stroma of RIP1-Tag4 mice are inhibited in situ remains to be determined.
Because epitope IV-specific TCD8 remain responsive and functional in RIP1-Tag4 mice with progressing tumors, escape from immune-mediated control of tumor progression may be related to changes in the tumor cells themselves. One such mechanism is the down-regulation of MHC class I expression, enabling tumor escape from immunosurveillance (57, 58, 59). Extensive analysis of MHC class I expression on islet cells in RIP1-Tag4 mice by Ye et al. (60) showed that normal islets in B6 and RIP1-Tag4 mice express MHC class I, whereas insulinomas from unimmunized RIP1-Tag4 mice contained a heterogeneous population of cells that were either negative or weakly positive for MHC class I expression. This observation suggests that heterogeneity of MHC class I expression on tumor cells could limit the efficiency of immunosurveillance in 6-mo-old RIP1-Tag4 mice by preventing the recognition of T Ag-expressing tumor cells.
The basis for complete protection in RIP1-Tag4 mice immunized before but not after tumor formation may lie in the steps that lead to T Ag transformation and tumorigenesis. T Ag-expressing cells undergo immortalization with the capacity to proliferate followed by acquisition of a tumorigenic phenotype (61). A brief interaction with activated functional T Ag-specific TCD8 can interfere with this process. Our previous studies (62) have shown that T Ag-transfected primary mouse embryo fibroblasts abort the transformation process upon a brief interaction with a T Ag-specific CTL clone. In this study we show that a portion of pancreatic islets of RIP1-Tag4 mice that were immunized at 35 days of age still contain cells positive for T Ag expression at 2 years of age, and most do not appear to be hyperplastic. The mechanism by which epitope IV-specific TCD8 suppress tumor progression in vivo could be mediated by the production of cytokines such as IFN- or TNF- (63, 64). Alternatively, those cells expressing the highest levels of T Ag might be eliminated before transitioning to tumor cells. This mechanism requires sufficient levels of MHC class I expression to present T Ag epitopes to functional TCD8, whereas escape variants arising because of low MHC class I expression would be expected to quickly advance to progressively growing tumors.
Lifelong tumor control without development of diabetes
Our results provide support for an earlier observation (18) that RIP1-Tag4 mice immunized with SV40 before endogenous T Ag expression led to long-term control of tumor progression (average lifespan of 622 days in the previous study). However, immunization with SV40 was ineffective after T Ag was expressed from the host’s own cells despite the persistence of SV40-specific CTL. Although the vehicle used for immunization (SV40 vs T Ag transformed cells) differs from that in the previous study, our findings here suggest that control of tumor progression following SV40 infection of young RIP1-Tag4 mice in the previous study was likely mediated by epitope IV-specific TCD8. A similar result was obtained recently in the related RIP1-Tag5 line of mice following immunization with soluble T Ag in the presence of CpG-containing oligonucleotides (22), although no quantitative data on the TCD8 response was presented.
We found that long-term control of tumor development by epitope IV-specific TCD8 does not result in diabetes in RIP1-Tag4 mice, because sufficient numbers of T Ag-expressing cells survive. This is despite the persistence of functionally competent epitope IV-specific TCD8. Thus, an equilibrium is reached in which tumor progression is controlled but organ function is maintained, an ideal scenario for tumor immunotherapy. Our results contrast with those observed in mice expressing either the glycoprotein or NP from LCMV in cells. In this case, virus infection resulted in T cell infiltration followed by diabetes due to destruction of the cells (39, 40). Indeed, we found that T Ag-expressing cells persisted in the islets of long-term survivors. Whether continued -islet cell survival in these mice is dependent on the level of T Ag expression is unknown. Defining the basis for the establishment of this equilibrium may yield new concepts for targeting immunotherapy against self-Ags.
A similar scenario has been observed in some transplantable tumor models in which immunization against a self-tumor Ag induces tumor regression without leading to autoimmunity (5, 65, 66). Thus, the number of insulin-secreting T Ag-positive cells that remain produce enough insulin to prevent the development of diabetes. It should be pointed out that all cells in RIP1-Tag4 transgenic mice are expected to express T Ag, although the expression levels may vary (32). In the RIP-hemagglutinin (HA) transgenic model, immunization with Flu also induces the onset of diabetes due to the activation of endogenous self-reactive TCD8 (42). In this model, however, RIP-HA mice must be immunized before 1 wk of age to induce diabetes, because high avidity HA-specific TCD8 are eliminated by peripheral tolerance at later time points (7). Preliminary experiments suggest that epitope IV-specific TCD8 recovered from long-term surviving RIP1-Tag4 mice have avidities similar to those derived from B6 mice. The possibility remains that the LCMV and Flu HA proteins are expressed at higher levels in the pancreas than T Ag, such that cells in these mice are more readily destroyed.
A recently published study argued against a role for immunosurveillance of spontaneous cancer (21). That suggestion may not be generalized to all immunogenic spontaneous models of cancer (67). Although the role of spontaneously primed TCD8 to control autochthonous tumors will continue to be debated, data presented here and by others (18, 21) argue that long-term immunosurveillance against spontaneous tumors can be effective if T cells are activated before the appearance of tumors or if the tumor-bearing host is preconditioned before immunotherapy (14, 31, 68). A key to effective T cell-mediated therapy is to target epitopes less likely to lead to peripheral tolerance following prolonged exposure to the endogenous Ag. In addition, understanding the mechanisms by which tumor-specific T cell responses can mediate control of tumor progression without disrupting organ function will be a major advance toward developing immunotherapeutic strategies.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This study was supported by National Cancer Institute/National Institutes of Health Grants CA-25000 (to S.S.T.) and PO3 CA-34196 (to B.B.K.) and American Cancer Society Research Scholar Grant 04-059-01-LIB (to T.D.S.). S.C.H. was supported by National Cancer Institute Training Grant T32 CA60395. 
2 Current address: Institute of Molecular Genetics, Academy of Sciences of the Czech Republic, Prague, Czech Republic.
3 Current address: Department of Biology, Sinclair Community College, Dayton, OH 45402.
4 Address correspondence and reprint requests to Dr. Satvir S. Tevethia, Department of Microbiology and Immunology, The Pennsylvania State University College of Medicine, Hershey, PA 17033. E-mail address: sst1{at}psu.edu
5 Abbreviations used in this paper: TCD8, CD8+ T lymphocyte; B6, C57BL/6; Db, H-2Db; Db/I, Db epitope I; Flu, influenza virus; gB, glycoprotein B; HA, hemagglutinin; Kb, H-2Kb; Kb/IV, Kb epitope IV; LCMV, lymphocytic choriomeningitis virus; NP, nucleoprotein; RIP, rat insulin promoter; T Ag, tumor Ag.
Received for publication January 12, 2006. Accepted for publication June 19, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
-cell tumours in transgenic mice expressing recombinant insulin/simian virus 40 oncogenes. Nature 315: 115-122. [Medline] cells. Nature 325: 223-228. [Medline] cells cultured from individual preneoplastic foci in a multistage tumorigenesis pathway: a potentially general technique for isolating physiologically representative cell lines. Mol. Cell. Biol. 13: 4223-4232. cells can break peripheral ignorance and lead to spontaneous autoimmune diabetes. Immunity 3: 727-738. [Medline]
