T Cell-Mediated Delay of Spontaneous Mammary Tumor Onset: Increased Efficacy with In Vivo versus In Vitro Activation
http://www.100md.com
免疫学杂志 2005年第8期
Abstract
Peripheral tolerance to shared Ags expressed on both tumors and normal self-tissues presents a major barrier to T cell-based immunotherapy as a treatment for cancer. To assess the activity of tumor-specific T cells against spontaneously arising carcinomas in the context of shared Ag expression, we developed a model system whereby an identified tumor Ag, tumor ERK (tERK), is expressed transgenically on both normal mammary tissue and spontaneous mammary carcinomas. Transfer of in vitro-activated, tERK-specific DUC18 T cells delayed spontaneous tumor development in tERK-expressing mice when T cells were given before the development of palpable carcinomas. However, antitumor activity mediated by in vitro-activated DUC18 T cells, as measured by responsiveness against a transplanted tERK-expressing fibrosarcoma challenge, was lost within days of transfer. This loss was due to expression of tERK as a self-Ag on normal tissues and was independent of the presence of mammary tumors. In contrast, transferred naive DUC18 T cells maintained a long-term protective function in tERK-expressing mice. Ten-fold fewer naive T cells activated in vivo were able to replicate the delay in spontaneous tumor development achieved by in vitro-activated T cells. These results are in contrast to our earlier studies using transplanted tumors alone, in which in vitro-activated DUC18 T cells were more efficacious than naive DUC18 T cells and highlight the need to perform tumor studies in the presence of tumor Ag expression on normal self-tissue.
Introduction
Tumors in mice and humans can be recognized and destroyed by CD8+ T cells. This capability has lead to attempts to use T cells to specifically target tumors by adoptive immunotherapy, i.e., the transfer of tumor-specific T cells into patients with established tumors, in the hopes of causing tumor regression or rejection. Because it is not currently practical to identify unique tumor Ags for every patient, most adoptive immunotherapies target shared Ags expressed by both the tumor and normal self-tissue. These shared Ags are often tissue-restricted Ags expressed by the tumor and the normal tissue from which the tumor originated. Examples of immunotherapy trials that target-shared Ags are the melanoma trials targeting melanocyte Ags, including Melan-A, Mart-1, and gp100 (1, 2, 3, 4, 5).
Although targeting shared Ags allows more universally applicable therapies, transferred T cells are subject to peripheral tolerance mechanisms that prevent self-specific immune responses. Clinical studies have revealed dramatic reductions of in vitro-expanded T cells within 2 wk of adoptive transfer into cancer patients (1, 5, 6). Although nondeletional phenomena such as T cell exhaustion may contribute to this T cell loss, deletion of self-Ag-specific T cells is likely a critical factor. However, it is difficult to differentiate between these possibilities in clinical trials, and the examination of such mechanistic questions is better suited to murine tumor models.
To date, the vast majority of work examining interactions between tumors and T cells in mice has been done using transplantable tumors. These are often tumor cell lines that are transfected with model Ags to allow recognition by transgenic T cells. Therefore, the Ag in question is confined to the tumor itself. Although much has been learned from these systems, they do not allow the study of T cell activity in the presence of Ags expressed on normal self tissue.
Experiments designed to study the impact of peripheral tolerance on tumor-specific T cell responses by transplanting tumors into mice that transgenically express Ags on normal tissue or by using mice that spontaneously develop Ag-expressing tumors have shown a variety of results. Experiments using transplantable tumors in the presence of transgenic Ag expression in the liver have shown that naive T cells are anergized and cannot be activated unless restimulated in vitro in the presence of IL-2, which rescues in vivo function (7). Alternatively, other systems specific for melanocyte Ags have shown that vaccination with recombinant vaccinia virus and administration of IL-2 can restore naive T cell function in vivo (8). Experiments using transgenic mice that develop spontaneous pancreatic carcinomas that share Ag expression with normal pancreas show that naive T cells are ignorant of the tumor Ag (9, 10). Only one of these studies has used in vitro-activated T cells, and none of them has observed the in vivo deletion seen in patients.
Several years ago, our laboratory generated a TCR-transgenic system (DUC18) in which adoptively transferred CD8+ T cells are able to efficiently eliminate a transplanted murine fibrosarcoma, CMS5. DUC18 T cells recognize an epitope, designated tumor ERK (tERK) 4, that is generated by a point mutation in the ERK2 kinase expressed by CMS5. Adoptively transferred naive DUC18 T cells are able to completely reject small CMS5 tumors that are transplanted 3 days before T cell transfer (11). When DUC18 T cells are activated in vitro with a tERK peptide before transfer into tumor-bearing recipients, the kinetics of tumor rejection are more rapid, and the DUC18 T cells are able to eliminate much larger tumors that have been allowed to grow for 8 days before T cell transfer (12). Because DUC18 T cells proved to be so proficient in rejecting transplantable CMS5, we wanted to examine the efficacy of naive vs in vitro-activated DUC18 T cells in the more physiologically relevant context of spontaneously arising mammary carcinomas.
To do this, we generated transgenic mice (Dt) expressing the rejection Ag (tERK) of the CMS5 fibrosarcoma driven by the mouse mammary tumor virus (MMTV) promoter. Therefore, Dt mice express tERK on normal mammary tissues. These Dt mice were bred to NeuT mice, which express the cNeu oncogene under the control of the same promoter (13). Double transgenic (Dt/NeuT) mice spontaneously develop mammary carcinomas that express the tERK Ag. Adoptive transfer of in vitro-activated, tERK-specific DUC18 T cells is able to delay the onset of mammary tumor development in Dt/NeuT mice by several weeks if T cells are given before the development of palpable tumors. However, antitumor activity of in vitro-activated DUC18 T cells is rapidly lost in Dt/NeuT mice, as evidenced by their failure to clear transplanted CMS5 challenges. This loss occurs even in Dt mice and is therefore due to the expression of tERK as a self-Ag on normal tissues and is completely independent of the presence of mammary tumors. In contrast, when naive DUC18 T cells are activated in vivo in tERK-expressing Dt mice by CMS5 challenge, they not only mount a productive primary response but also generate a protective memory population that is able to reject a secondary CMS5 challenge at least 2 months later. Surprisingly, the adoptive transfer of 10-fold fewer naive DUC18 T cells, activated in vivo by CMS5 challenge, is able to effect the same delay in spontaneous mammary tumor formation as is seen with larger numbers of in vitro-activated DUC18 T cells.
Our findings highlight the potential differences between spontaneous and transplantable model systems and the importance of peripheral tolerance as a barrier to adoptive immunotherapy. In contrast to our earlier findings made using the transplantable CMS5 system, naive DUC18 T cells were more effective and had longer activity than in vitro-activated DUC18 T cells in this spontaneous mammary tumor system. Because naive T cells activated in vivo are able to survive, our results indicate that it should be possible to manipulate in vitro culture conditions to generate activated, tumor Ag- specific T cells that persist and therefore have enhanced efficacy in the presence of shared Ags in vivo.
Materials and Methods
Mice and cell lines
All mice used in this study are on the BALB/c background. DUC18 mice, originally generated in our laboratory (11), were bred and housed in the animal facility at the Washington University School of Medicine. Thy 1.1+ BALB/c mice were obtained from Dr. H. Levitsky (Johns Hopkins University, Baltimore, MD) and bred to DUC18 to generate Thy 1.1+ DUC18 mice. NeuT mice, developed by Forni and colleagues (13), were kindly provided by Dr. J. Berzofsky (National Cancer Institute, National Institutes of Health, Bethesda, MD).
The construct used to generate transgenic mice expressing tERK in mammary tissue, designated Dt, was assembled by blunt ligation of tERK cDNA originally cloned from CMS5 into a MMTV vector containing the human growth hormone polyadenylation sequence (gift of R. Schreiber (Washington University School of Medicine, St. Louis, MO)). Purified DNA was injected into BALB/c pronuclei. Founder mice were screened by Southern blot using probes spanning the junction between the vector and tERK insert. Probes were generated by digesting the MMTVtERK plasmid with EcoRI and labeled using the Rediprime kit (Amersham Biosciences). A PCR screen was developed using primers that spanned the junction between the 3'-end of the tERK insert and the 5'-end of the human growth hormone poly-A sequence (forward: TATGACCCAAGTAATGAGCCCATT; reverse: CTGAGATTGGCCAAATACTGG).
Dt mice, designated Dt/NeuT, were generated by breeding male mice positive for both the cNeu and Dt transgenes to BALB/c females. Offspring were screened by PCR for the presence of the Dt and NeuT transgenes. Female mice were used for all studies.
CMS5 (14) was freshly thawed from frozen stock for every experiment and expanded through no more than three passages in RPMI 1640 medium supplemented with 10% FCS (HyClone), 1 mM glutamax (Invitrogen Life Technologies), and 50 μg/ml gentamicin (Invitrogen Life Technologies). All CMS5 challenges in this study were s.c. injections of 3 x 106 CMS5 cells in HBSS.
DUC18 T cell activation in vitro
In vitro-activated DUC18 T cells were generated as previously described by culturing DUC18 splenocytes at 5 x 106 cells/ml in RPMI 1640 medium supplemented with 10% FCS (HyClone), 1 mM glutamax (Invitrogen Life Technologies), 5 x 10–5 M 2-ME (Sigma-Aldrich), and 50 μg/ml gentamicin (Invitrogen Life Technologies) in the presence of 0.5 μM tERK peptide (QYIHSANVL) (12). Cells were harvested after 4 days in culture and purified over Ficoll-Paque (Amersham Biosciences). The resulting cell population was generally 70–90% DUC18 T cells, defined as being V8.3+and CD8+. Cells were analyzed by flow cytometry for the percentage of live lymphocytes that were V8.3+CD8+, and cell counts were compensated accordingly. T cell adoptive transfer was given i.v. in HBSS.
Detection of tERK message
The presence of tERK message in various tissues was assayed as previously described (12). Briefly, RNA was extracted from the tissue or cell line of interest using TRIzol (Invitrogen Life Technologies) and used to make cDNA. Primers spanning the tERK mutation (Integrated DNA Technologies) were used to amplify the cDNA. This PCR product was purified and digested overnight with SfcI (New England Biolabs). Digested PCR products were then resolved on a 2% agarose gel. The presence of a 150-bp band indicates the presence of tERK message.
Monitoring of tumor development
NeuT and Dt/NeuT mice were monitored twice weekly for the development of palpable mammary nodules. The development of persistent, progressive, palpable (3 mm in diameter) nodules marked tumor onset. In NeuT mice, mammary hyperplasia begins before 4 wk of age. By 10 wk of age, mice progress to carcinoma in situ. The development of palpable tumors typically occurs between 13 and 17 wk of age. Mammary tumor growth continued to be monitored until any single nodule exceeded 15 mm in diameter, or the mouse began showing visible signs of distress. These kinetics are similar to those previously published for this model (13).
CMS5 challenge consisted of 3 x 106 cells suspended in HBSS injected s.c. in the right hind flank. Mice were subsequently monitored for tumor development at the site of injection beginning 2 or 3 days after challenge. Tumor nodules were measured every other day for the duration of the experiment. Tumor area is calculated by multiplying two perpendicular diameters (11, 12).
Flow cytometry
For peripheral blood samples, RBC were lysed using buffered ammonium chloride before Ab staining. Peripheral blood samples were stained using directly conjugated anti-V8.3-FITC, anti-Thy1.1-PE, and anti-CD8-CyChrome. All Abs used in this study were purchased from BD Pharmingen. Data was collected either on a FACScan or a FACSCalibur and analyzed using CellQuest software (BD Pharmingen) or FlowJo (Tree Star).
Statistical analysis
Statistical significance of tumor incidence was analyzed by the Mann-Whitney U test using InStat software (GraphPad Software).
Results
Generation of tERK-transgenic mice
To assess the efficacy of transgenic T cells in treating spontaneous tumors expressing a shared tumor Ag, transgenic mice, referred to as Dt, were generated by expressing the tERK Ag under the control of the MMTV promoter. These Ag-expressing mice were bred to NeuT mice expressing the cNeu oncogene under the control of the MMTV promoter. The resulting double transgenic Dt/NeuT mice express tERK as an Ag shared by normal tissue and spontaneous mammary carcinomas. The effects of shared expression of tERK Ag on tERK-specific DUC18 T cells were then assessed in the context of spontaneous mammary carcinomas.
tERK message can be detected based on its susceptibility to digest by SfcI (Fig. 1A). Analysis of a variety of tissues from a 6-wk-old Dt virgin female mouse showed tERK expression in the mammary glands, salivary glands, and lungs, in agreement with previous characterization of the MMTV promoter (15, 16). There was also expression in the spleen (Fig. 1B). Dt mice were then crossed to NeuT mice, which have been described previously as spontaneously developing lobular carcinoma throughout their mammary tissue with complete penetrance. Tumor development in these mice is continuous, beginning when the mice are <4 wk old (13, 17). No changes in the kinetics of tumor development (Fig. 1C) or the pathology (data not shown) of mammary tumors that developed in double transgenic Dt/NeuT mice relative to NeuT mice were observed. The pathology and kinetics of the Dt/NeuT mice correlated with previously published characterization of NeuT mice (17). Analysis of mRNA taken from spontaneous tumors showed that tumors developing in Dt/NeuT double transgenic mice express tERK message (Fig. 1D). Therefore, the Dt transgene functions as an antigenic tag and does not alter the characteristics of tumors that develop in double transgenic Dt/NeuT mice. Therefore, tumors that develop in double transgenic mice are suitable targets for DUC18 T cell-mediated therapy.
FIGURE 1. Characterization of the Dt transgenic mouse. A, The tERK mutation generates a novel SfcI restriction site that can be used to assay cDNA for tERK message. Normal ERK (nERK) does not contain this restriction site. PCR products spanning the mutation are gene-cleaned and digested. tERK, but not nERK, message is susceptible to SfcI digestion. B, Survey of tERK expression in various organs from a 6-wk-old virgin Dt female. The lower band indicates the presence of tERK message, and the upper band is intact nERK message. C, Kinetics of tumor development in NeuT vs Dt/NeuT mice. Mice were monitored twice weekly for the development of palpable (>3 mm in diameter) tumors. The age at which mice first develop persistent, palpable tumors is the age of tumor onset. D, Assay for tERK message in mammary tumors taken from unmanipulated Dt/NeuT mice. The lower band indicates the tERK message, and the upper band indicates the nERK message (N = NeuT, Dt/N = Dt/NeuT).
Transfer of activated DUC18 T cells delays tumor development in Dt/NeuT mice
Dt/NeuT mice were bred directly to DUC18-transgenic mice to assess the ability of DUC18 T cells to delay mammary tumor formation. Approximately 80% of the CD8 T cells in the circulation of DUC18-transgenic mice are V8.3+CD8+. When Dt/NeuT/DUC18 triple transgenic mice were analyzed by flow cytometry, no such enrichment for DUC18 T cells was observed (data not shown). This indicated that DUC18 T cells were being deleted in these mice, likely due to low levels of tERK transgene expression in the thymus (Fig. 1B). Therefore, we used adoptive transfer to address whether DUC18 T cells were capable of delaying the formation of spontaneous mammary tumors in Dt/NeuT mice.
Previous studies using DUC18 T cells to treat transplantable tumors had indicated that T cells activated in vitro before transfer into the host were highly effective in rejecting large tumors (12). Therefore, a single dose of activated T cells (30 x 106) was administered to 4- or 9-wk-old Dt/NeuT mice to assess whether any delay in mammary tumor onset could be attained. In agreement with previous characterizations of NeuT mice, we find that at 4–5 wk of age, NeuT mice show mild mammary hyperplasia but no overt carcinoma (13). By 9–10 wk of age, there is histological evidence of carcinoma in situ but not palpable tumors (data not shown). There was a statistically significant delay in the onset of palpable mammary tumors in 4-wk-old Dt/NeuT mice that received activated T cells compared with untreated mice (Fig. 2A). A similar delay could be seen in mice that received T cells at 9–10 wk of age. The delay in tumor development seen in T cell-treated Dt/NeuT mice was also significant compared with T cell-treated NeuT mice (Fig. 2B). Thus, the delay in tumor formation depends both on the T cell and on the Ag expression of the host.
FIGURE 2. Treatment of Dt/NeuT mice with a single dose of in vitro-activated DUC18 T cells delays onset of mammary tumors. A, Kinetics of mammary tumor development in Dt/NeuT mice left untreated () or treated with 30 x 106 in vitro-activated DUC18 T cells at 4–5 wk of age () or 9–10 wk of age (). Statistical significance as determined by the Mann-Whitney U test for untreated vs treated at 4–5 wk of age, p = 0.0009, and at 9–10 wk of age, p 0.0001. B, Kinetics of tumor development in Dt/NeuT mice () or NeuT mice () treated at 4–5 wk of age with 30 x 106 in vitro-activated DUC18 T cells. Statistical significance as determined by the Mann-Whitney U test, p = 0.013. Mice shown in these experiments were collected over several months and are progeny from several different litters from several different breeding pairs.
These delays were noteworthy, given that the NeuT transgene induces continuous tumor formation. Thus, any T cell treatment can only delay tumor development as long as the T cells maintain activity in vivo. Mice > 16 wk of age exhibiting palpable mammary tumors showed no objective effect when treated with activated DUC18 T cells (data not shown). Therefore, treatment of Ag expressing Dt/NeuT mice with monospecific T cell immunotherapy could effect a delay in the development of spontaneous mammary carcinoma, provided that the T cells were administered before the onset of palpable tumors.
Loss of DUC18 T cell activity in the presence of Ag-expressing mammary tumors
Although the delay in tumor onset attained by the activated DUC18 T cells was encouraging, we sought to elucidate the reasons for the eventual development of mammary tumors in T cell-treated, Ag-expressing Dt/NeuT mice. To test whether tumors that developed in DUC18 T cell-treated mice were Ag loss variants, tumors were assayed for the presence of the tERK message. tERK message was present in multiple tumors taken from various Dt/NeuT mice that had been treated with DUC18 T cells at 4–5 wk of age (data not shown). Thus, Ag loss did not appear to be a mechanism of immune evasion in this model.
Because tumors develop continuously in NeuT mice, we wanted to determine whether tumor outgrowth was the result of a loss of DUC18 T cell activity in the T cell-treated mice. The transplantable CMS5 fibrosarcoma is susceptible to DUC18 T cell rejection and can therefore be used to assess the presence of systemic DUC18 T cell activity. Mice that received 30 x 106 activated DUC18 T cells at 4–5 or 9–10 wk of age were challenged with CMS5 shortly after the development of their first palpable mammary tumors to assess DUC18 T cell activity. At the time of mammary tumor development (17–20 wk of age), there was little to no DUC18 T cell activity present in the Dt/NeuT mice as indicated by the uninhibited growth of CMS5 in these mice (data not shown). In contrast, all of the non-Ag-expressing NeuT mice exhibited unimpaired DUC18 T cell function (data not shown).
We wanted to assess T cell activity over a shorter period of time than that described in the above experiments, which took place over a period of many months. To maximize the effects of the tumor environment, mice bearing palpable mammary tumors were used to assess the impact of shared Ag expression on short-term T cell function. Therefore, the dose of T cells given to the mice was lowered to 3 x 106, and the CMS5 challenge was administered 3 days after the T cell transfer. Even in this short time frame, there was a marked decrease in the activity of DUC18 T cells in Ag-expressing Dt/NeuT mice exhibited by a failure to control CMS5 outgrowth (Fig. 3). These results indicate a rapid loss of T cell function in the presence of palpable, Ag-expressing tumors and shared Ags expressed on normal tissues.
FIGURE 3. T cell activity is rapidly lost in mice bearing spontaneous, Ag-expressing mammary tumors. T cell activity was measured after T cell transfer into mammary tumor bearing Dt/NeuT or NeuT mice. Mice with palpable mammary tumors received 3 x 106 in vitro-activated DUC18 T cells. Three days following T cell transfer, mice received a s.c. challenge of CMS5 to test DUC18 T cell activity. Shown is CMS5 tumor area over time in individual mice from a total of three independent experiments. A total of 10 NeuT and 10 Dt/NeuT mice was tested.
In vitro-activated T cells are eliminated rapidly in Dt/NeuT mice
The previous experiments raised the question of whether DUC18 T cells were present but nonfunctional in the presence of tERK as a self-Ag or were eliminated altogether. To differentiate between these possibilities, we used flow cytometry to monitor the persistence of DUC18 T cells in the peripheral blood after their transfer into 4- to 5-wk-old Dt/NeuT or NeuT mice. For these experiments, 30 x 106 activated Thy 1.1+ DUC18 T cells were transferred into Thy 1.2+ Dt/NeuT or NeuT mice, and recipients were analyzed periodically for the presence of circulating Thy1.1+ T cells. Throughout the period of observation, levels of Thy1.1+ T cells were 5- to 10-fold higher in NeuT mice than in Dt/NeuT mice, beginning 3 days after T cell transfer (Fig. 4). Experiments using 9-to 10-wk-old recipients showed a similar loss of circulating Thy 1.1+ DUC18 T cells in Ag-expressing Dt/NeuT mice (data not shown). We surveyed organs for T cells but were unable to detect accumulated DUC18 T cells in the spleen, lymph nodes, or mammary tissues of Ag-expressing mice (data not shown). Therefore, it does not appear that the T cells are being trapped or sequestered in tERK- expressing tissues or in secondary lymphoid tissues. Rather, it appears that activated DUC18 T cells were being eliminated rapidly following transfer in Ag-expressing hosts.
FIGURE 4. In vitro-activated DUC18 T cells are rapidly lost in Dt/NeuT hosts. A total of 30 x 106 Thy 1.1+ in vitro-activated DUC18 T cells was transferred into 4- to 5-wk-old Thy 1.2+ Dt/NeuT () or NeuT () hosts. Peripheral blood was tested periodically for the percentage of peripheral blood lymphocytes that were Thy 1.1+,V8.3+ as an indication of transferred DUC18 T cells. Shown is the percentage of peripheral blood lymphocytes in individual NeuT () or Dt/NeuT mice () that are Thy1.1,V8.3+ over the course of the experiment. Results are representative of four independent experiments.
Loss of T cell activity is Ag, but not oncogene, dependent
The loss of DUC18 T cells in Dt/NeuT mice could be due to the expression of cognate Ag on self-tissue or to immunosuppressive effects of the tumors that develop. To determine whether transgenic tERK expression alone was sufficient to eliminate transferred DUC18 T cells in the periphery, 3 x 106 in vitro-activated T cells were transferred into Dt or BALB/c mice. Three days later, mice were challenged with CMS5. Dt transgenic mice showed a marked defect in the ability to reject this challenge (Fig. 5). Tracking experiments assaying DUC18 T cell persistence in the circulation of Dt mice showed a similar loss of T cells as that seen in Dt/NeuT mice (data not shown). Because Dt mice do not express the cNeu oncogene, they do not develop mammary tumors. Therefore, the loss of in vitro-activated DUC18 T cells seen in the Dt mice is due solely to the expression of tERK on normal peripheral tissues and not to any tumor-induced immune dysfunction.
FIGURE 5. Inactivation of DUC18 T cells is due to expression of tERK as a self-Ag. Groups of Dt or BALB/c mice were treated with 3 x 106 in vitro-activated DUC18 T cells and received a s.c. CMS5 challenge 3 days later. Subsequent CMS5 growth was measured to indicate in vivo DUC18 T cell activity. Shown are the areas of CMS5 tumors over time in individual Dt () or BALB/c () mice. Results are representative of three independent experiments.
Naive DUC18 T cells become activated in the presence of peripherally expressed tERK
Previous studies examining the ability of naive T cells to become activated in the presence of self-Ag have shown that naive T cells can become anergized and can only be rescued by restimulation in vitro in the presence of IL-2 (7). Because tERK expression in Dt transgenic mice appeared sufficient to delete in vitro-activated DUC18 T cells, we sought to assess the impact of the Dt transgene on the function of naive DUC18 T cells. Naive DUC18 T cells were transferred into Dt or BALB/c mice 24 h before CMS5 challenge to determine the effect of transgenic tERK expression on the activation and persistence of DUC18 T cells. Although 10- to 100-fold more DUC18 T cells were required to reject CMS5 challenge relative to BALB/c mice, naive DUC18 T cells were capable of rejecting CMS5 challenge in Dt mice when present in sufficient numbers (Fig. 6). In four independent experiments, 80% of Dt mice that received 3.5 x 105 naive DUC18 T cells were able to reject subsequent CMS5 challenge. Mice that received 3.5 x 104 naive DUC18 T cells rejected 40% of CMS5 challenges in three independent experiments (Fig. 6 and data not shown). These data indicate that, in this system, naive DUC18 T cells can acquire effector function and reject transplantable tumors, even in the presence of tERK expressed as a self-Ag.
FIGURE 6. Naive DUC18 T cells can demonstrate effector function in the presence of self-Ag. Naive DUC18 T cells were transferred into Dt or BALB/c mice in the doses indicated. Twenty-four hours later, mice received a s.c. CMS5 challenge. Shown is the tumor area over time for individual mice in one experiment representative of three independent experiments. The number in the upper left corner of each graph represents the dose of naive T cells transferred. BALB/c controls are shown on the left, and Ag-expressing Dt mice are in the right column.
T cells resulting from in vivo activation are capable of prolonged persistence in the presence of tERK as a self-Ag
Because the DUC18 T cells were able to mount an immune response to CMS5 even in the presence of tERK expressed on self-tissue, we were able to assess the impact of shared Ag expression on the development of a tumor-specific memory response. Mice that had rejected a previous challenge with CMS5 were rechallenged 2 mo later, and the growth of this second challenge was monitored to assess T cell activity. In every experiment, there was evidence of DUC18 T cell activity in Dt mice that had received naive cells and then were rechallenged with CMS5 (Fig. 7). When mice that were able to reject primary CMS5 challenges with either 3.5 x 105 or 3.0 x 106 DUC18 T cells were rechallenged 2 mo later, 70% of these secondary challenges were rejected in four independent experiments. These results contrast with similar experiments done using in vitro-activated T cells, which cannot sustain T cell activity in Dt mice for more than days or weeks at most (Figs. 3 and 4 and data not shown).
FIGURE 7. DUC18 T cells activated in vivo can maintain effector function for prolonged periods in the presence of the Dt transgene. Dt or BALB/c mice received transfers of naive DUC18 T cells and were challenged with CMS5 the next day. Mice that rejected this primary challenge were rechallenged 2 mo later with a second dose of CMS5. Shown are the growth kinetics of the secondary CMS5 challenge in individual Dt () and BALB/c () mice, as well as CMS5 growth in untreated BALB/c mice (). Results are representative of four individual experiments. In two experiments, an initial dose of 3.5 x 105 DUC18 T cells was used, and in two other experiments, a dose of 3 x 106 naive T cells was used. Similar outcomes were observed with both doses.
In vivo-activated T cells significantly delay spontaneous tumor development
Because in vivo-activated DUC18 T cells can maintain function for long periods of time in the presence of self-Ag, we evaluated whether they might be more efficient than in vitro-activated DUC18 T cells in delaying the development of spontaneous tumors in Dt/NeuT mice. Therefore, 4- to 5-wk-old NeuT or Dt/NeuT mice were treated with 3 x 106 naive T cells and then challenged with CMS5 to activate these cells in vivo. Mice that rejected this CMS5 challenge were then monitored for mammary tumor development. Surprisingly, this relatively low number of naive T cells achieved a delay in spontaneous tumor development (Fig. 8A). In fact, the kinetics of tumor development in mice treated with naive T cells activated in vivo was identical with that created by a 10-fold higher dose of in vitro-activated cells (Fig. 8B). Although these results are promising, the fact that the treated mice eventually developed mammary tumors indicates that there may be other factors dictating the efficacy of DUC18 T cells against spontaneous mammary tumors. It is also possible that the number of effector cells generated from transfer of 3 x 106 naive DUC18 T cells is insufficient to delay tumor development beyond that achieved by 30 x 106 in vitro-activated cells. These experiments indicate that, although longer lived T cells are much more efficient in preventing spontaneous tumor onset, there are still limitations to their efficacy that must be further investigated.
FIGURE 8. Low doses of naive T cells can delay spontaneous tumor development. A total of 3 x 106 naive DUC18 T cells was transferred into 4-wk-old Dt/NeuT or NeuT hosts. Twenty-four hours later, mice received a s.c. CMS5 challenge. Mice that rejected the CMS5 challenge were monitored for the development of spontaneous mammary tumors. A, Age of initial mammary tumor development of NeuT/Dt vs NeuT mice treated with in vivo-activated T cells. Statistical significance as determined by the Mann-Whitney U test, p = 0.02. B, Comparison of the kinetics of mammary tumor development in Ag-expressing mice treated with 3 x 106 naive T cells or with 30 x 106 in vitro-activated T cells at 4–5 or 9–10 wk of age. Data for in vitro-activated T cell treatment are taken from the experiments shown in Fig. 2.
Discussion
In the present study, we describe a system in which in vitro-activated T cells are able to effect a delay in tumor development in a spontaneous tumor model. We observed that in vitro-activated DUC18 T cells are eliminated rapidly in Ag-expressing Dt/NeuT mice, similar to what has been described in patients treated by adoptive transfer of in vitro-expanded tumor-infiltrating lymphocytes or PBL (1, 5, 6). This T cell loss is due to expression of tERK as a self-Ag because it also occurs in Dt mice lacking mammary tumors. Thus, in our model system, T cell deletion does not require the presence of tumors. Surprisingly, while naive DUC18 T cells are slightly less efficacious in the presence of the tERK transgene than in wild-type BALB/c mice, they are still able to become activated and reject transplanted CMS5 challenges. Naive DUC18 T cells activated in vivo in the presence of the tERK transgene are able to persist for months after activation, whereas in vitro-activated cells are eliminated within days. A transfer of only 3 x 106 naive DUC18 T cells is able to effect an identical delay in spontaneous tumor formation to that seen following transfer of 30 x 106 in vitro-activated DUC18 T cells. Overall, our findings suggest that in vitro-activated T cells are highly susceptible to peripheral deletion after an encounter with self-Ag in vivo, and this dramatically limits their antitumor efficacy in vivo.
These results are in sharp contrast to the efficacy of activated DUC18 T cells in the transplantable CMS5 system. Previous studies have shown that, when treating the CMS5 fibrosarcoma in non-tERK-expressing hosts, in vitro-activated DUC18 T cells were faster and more effective at treating rejectable tumors and could reject large tumors that naive DUC18 T cells were incapable of destroying (11, 12). From these results, we predicted that in vitro-activated DUC18 T cells also would be more effective at treating Dt/NeuT mice, in which spontaneous tERK-expressing tumors were present. However, this proved not to be the case because the antitumor response mediated by in vitro-activated DUC18 T cells is extremely short lived. The CMS5 system is among a small group of transplantable systems in which the T cell is able to control established tumors. In other systems, the ability of the tumor-specific T cell to control tumor growth is limited by tolerance of the T cells to the tumor (18, 19). We cannot eliminate the possibility that the differences in DUC18 T cell activity against Dt/NeuT tumors and transplantable CMS5 are a result of the unusual susceptibility of CMS5 to DUC18-mediated killing, rather than differences in T cell activity against spontaneous vs transplantable tumors. Although CMS5 may be unusual in the ability of DUC18 T cells to reject the tumor, the stark contrast between the efficacy of naive and activated T cells in the spontaneous vs transplantable systems does highlight the immunological differences between these systems and demonstrates that mechanisms and therapies discovered using transplantable systems should be confirmed in spontaneous tumor models.
In other model systems, there is a link between the levels of Ag expressed on peripheral tissues and the fate of T cells transferred into Ag-expressing mice. High levels of Ag expressed on normal tissue have been shown to delete Ag-specific T cells in the periphery. Indeed, this peripheral deletion was dependent on the dose of Ags expressed (20, 21). In these systems, naive cells transferred into Ag-expressing mice rapidly proliferated and were eliminated. In our model system, we believe that in vitro-activated DUC18 T cells are deleted rapidly following adoptive transfer into Dt or Dt/NeuT mice because we are unable to detect them in the peripheral blood, the spleen, the lymph nodes, or mammary tissues (Fig. 4 and data not shown). Because we detected high levels of tERK mRNA in the spleens of Dt and Dt/NeuT mice, it is possible that a high dose of Ags in this organ was mediating deletion of transferred DUC18 T cells. However, we found the same loss of in vitro-activated DUC18 T cells upon their transfer into splenectomized Dt mice (data not shown); thus, a potentially high dose of tERK Ag in this site does not appear to mediate T cell deletion in our model. Interestingly, in another model system, the presence of large Ag-expressing spontaneous pancreatic tumors was shown to actually facilitate Ag drainage, cross-presentation, and activation of naive T cells (22). Whether the dose of Ag in our system is insufficient to activate naive cells or the type or activation state of the APC presenting the Ag is not conducive to the activation of naive T cells is unclear. However, if DUC18 T cells are activated in vitro before transfer into the Ag expressing hosts, there is sufficient Ag expressed to delete these cells. This may suggest that different thresholds of antigenic stimulation exist for activation of naive T cells vs deletion of previously activated effector T cells.
NeuT and Dt/NeuT mice were treated with naive DUC18 T cells activated in vivo by CMS5 challenge in an attempt to address whether in vivo survival was the major factor limiting the efficacy of the in vitro-activated DUC18 T cells in delaying mammary tumor development. A 10-fold lower dose of naive T cells was able to induce a delay in tumor development comparable to that sustained by in vitro-activated T cells. The factors limiting the efficacy of naive T cells are unknown, but the delay achieved by this relatively low dose of T cells indicates that even a small population of persistent T cells can cause a marked delay in spontaneous tumor onset. Whether the final limitation is simply the number of specific precursors available or a more fundamental mechanistic barrier to tumor rejection remains to be seen.
The delay in tumor development achieved by the naive DUC18 T cells and their persistence in the Dt mice is even more noteworthy in light of studies showing the peripheral tolerance of naive and effector/memory T cells specific for self-Ag. T cells specific for self-Ags have been shown to lose function or be deleted altogether (7, 23, 24, 25, 26). The persistence of the in vivo-activated DUC18 T cells is even more surprising given the deletion of self-specific T cells at all stages of differentiation. Elucidating the reasons for the persistence of these T cells in the presence of self-Ag expression could lead to the development of in vitro regimens to produce effector T cells capable of persisting for long periods of time in the presence of shared tumor Ag.
Although all T cell-treated mice eventually developed tumors in this study, one important consideration is the fact that the NeuT transgene leads to continuous development of multiple, rapidly growing mammary carcinomas. Tumors develop throughout the murine mammary tissue, which is quite extensive. This means that even a short delay in tumor formation in this system could translate into a much more significant effect in a clinical setting where tumor formation is a singular event. Additionally, the continuous generation of new tumors means that the loss of T cell activity in vivo leads to immediate outgrowth of subsequent tumors, which would not happen in the case of singular tumor development.
The fact that naive DUC18 T cells activated by CMS5 challenge in vivo are able to persist and establish protective memory T cell populations indicates that it should be possible to prevent the loss of in vitro-activated T cells by altering their culture conditions. Unfortunately, our attempts achieve this were unsuccessful. Addition of IL-2 or IL-15 to the in vitro culture, provision of supplemental CD4 help, or stimulation with low concentrations of peptide failed to generate in vitro-activated T cells capable of persisting in Ag-expressing mice (data not shown). Another possibility is that in our in vitro culture, where bulk splenocytes are being treated with tERK peptide, the type of APC present is not providing optimal T cell stimulation. In a bulk splenocyte preparation, B cells and semimature dendritic cells (DC) are likely responsible for the majority of tERK presentation to naive DUC18 T cells. During in vivo stimulation of naive DUC18 T cells following CMS5 challenge, DC in tumor draining lymph nodes may have acquired a more mature phenotype and therefore may be more suited to differentiating DUC18 T cells into competent effectors. Altering the type and frequency of DC during in vitro stimulation could provide the signals required to generate long-lasting DUC18 effectors.
The induction of a productive, tumor-specific T cell response in the face of peripheral tolerance has long presented a problem for tumor immunologists. In animal models, tolerance of endogenous tumor-specific T cells has been observed against spontaneous tumors (27, 28), but isolation and transfer of immunotherapeutic clones have not been attempted in most of these systems. Polyclonal therapies, such as DNA vaccination or IL-12 treatment, have likewise been effective in achieving impressive delays in tumor development in NeuT mice. Indeed, allogeneic vaccination combined with IL-12 treatment in NeuT mice can prevent tumor formation for up to a year and limits the number of tumors that do develop in individual mice (13, 29). These effects are the result of Ab and IFN- produced in the NeuT mice (30). Unfortunately, these systems depend on the stimulation of a polyclonal response and cannot then address what might be happening to specific T cells in the course of the response. Systems in which identified T cell clones are used to treat spontaneous tumors show T cell ignorance of the tumors (9, 10).
In the present study, we introduce a system in which a characterized, identifiable, tumor-specific T cell can delay the development of spontaneous mammary carcinomas. Interestingly, in our system, although adoptively transferred in vitro-activated DUC18 T cells are short lived, naive DUC18 T cells activated in vivo are resistant to elimination. Elucidation of the factors responsible for this in vivo persistence would represent a significant advancement to the field of tumor immunotherapy and would then allow the additional characterization of the conditions that dictate survival or deletion of T cells specific for shared Ags in vivo. The fact that the therapeutic efficacy of these T cells in the presence of a shared Ag is a sharp contrast to the activity of this T cell in the presence of a unique tumor Ag highlights the need for shared Ag models to study T cell behavior in these more physiologic situations. The persistence of in vivo-activated T cells in the presence of persistent Ag indicates that there is a way to permit long-term survival of shared Ag-specific T cells in vivo. The recapitulation of this T cell survival by manipulation of in vitro expansion techniques would be a great benefit to the field of tumor immunology.
Disclosures
The authors have no financial conflict of interest.
Acknowledgments
We thank Dave Donermeyer and Dr. Ken Matsui for technical assistance in creating the Dt transgenic construct, Donna Thompson and Jerri Smith for clerical assistance, Kathy Frederick for technical assistance in performing splenectomies, Drs. Guido Forni and Jay Berzofsky for guidance in obtaining and maintaining the NeuT mice, and Silvia Kang, Gavin Dunn, and Drs. Laura Mandik-Nayak and Fei Shih for critical reading of this manuscript.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by grants from the National Institutes of Health. L.A.O. was supported by a Howard Hughes Predoctoral Training Grant. L.A.N. is supported by a postdoctoral fellowship from the Cancer Research Institute.
2 Current address: Emory University School of Medicine, Department of Microbiology and Immunology, Atlanta, GA 30322.
3 Address correspondence and reprint requests to Paul M. Allen, Washington University School of Medicine, Department of Pathology and Immunology, 660 South Euclid Avenue, Campus Box 8118, St. Louis, MO 63110. E-mail address: pallen{at}wustl.edu
4 Abbreviations used in this paper: tERK, tumor ERK; MMTV, mouse mammary tumor virus; DC, dendritic cell.
Received for publication November 23, 2004. Accepted for publication January 29, 2005.
References
Yee, C., J. A. Thompson, D. Byrd, S. R. Riddell, P. Roche, E. Celis, P. D. Greenberg. 2002. Adoptive T cell therapy using antigen-specific CD8+ T cell clones for the treatment of patients with metastatic melanoma: in vivo persistence, migration, and antitumor effect of transferred T cells. Proc. Natl. Acad. Sci. USA 99:16168.
Dudley, M. E., L. T. Ngo, J. Westwood, J. R. Wunderlich, S. A. Rosenberg. 2000. T-cell clones from melanoma patients immunized against an anchor-modified gp100 peptide display discordant effector phenotypes. Cancer J. 6:69.
Dudley, M. E., J. R. Wunderlich, P. F. Robbins, J. C. Yang, P. Hwu, D. J. Schwartzentruber, S. L. Topalian, R. Sherry, N. P. Restifo, A. M. Hubicki, et al 2002. Cancer regression and autoimmunity in patients after clonal repopulation with antitumor lymphocytes. Science 298:850.
Dudley, M. E., J. R. Wunderlich, J. C. Yang, P. Hwu, D. J. Schwartzentruber, S. L. Topalian, R. M. Sherry, F. M. Marincola, S. F. Leitman, C. A. Seipp, et al 2002. A phase I study of nonmyeloablative chemotherapy and adoptive transfer of autologous tumor antigen-specific T lymphocytes in patients with metastatic melanoma. J. Immunother. 25:243.
Dudley, M. E., J. Wunderlich, M. I. Nishimura, D. Yu, J. C. Yang, S. L. Topalian, D. J. Schwartzentruber, P. Hwu, F. M. Marincola, R. Sherry, S. F. Leitman, S. A. Rosenberg. 2001. Adoptive transfer of cloned melanoma-reactive T lymphocytes for the treatment of patients with metastatic melanoma. J. Immunother. 24:363.
Yee, C., J. A. Thompson, P. Roche, D. R. Byrd, P. P. Lee, M. Piepkorn, K. Kenyon, M. M. Davis, S. R. Riddell, P. D. Greenberg. 2000. Melanocyte destruction after antigen-specific immunotherapy of melanoma: direct evidence of T cell-mediated vitiligo. J. Exp. Med. 192:1637.
?hlen, C., M. Kalos, D. J. Hong, A. C. Shur, P. D. Greenberg. 2001. Expression of a tolerizing tumor antigen in peripheral tissue does not preclude recovery of high-affinity CD8+ T cells or CTL immunotherapy of tumors expressing the antigen. J. Immunol. 166:2863.
Overwijk, W. W., M. R. Theoret, S. E. Finkelstein, D. R. Surman, L. A. de Jong, F. A. Vyth-Dreese, T. A. Dellemijn, P. A. Antony, P. J. Spiess, D. C. Palmer, et al 2003. Tumor regression and autoimmunity after reversal of a functionally tolerant state of self-reactive CD8+ T cells. J. Exp. Med. 198:569.
Speiser, D. E., R. Miranda, A. Zakarian, M. F. Bachmann, K. McKall-Faienza, B. Odermatt, D. Hanahan, R. M. Zinkernagel, P. S. Ohashi. 1997. Self antigens expressed by solid tumors do not efficiently stimulate naive or activated T cells: implications for immunotherapy. J. Exp. Med. 186:645.
Lyman, M. A., S. Aung, J. A. Biggs, L. A. Sherman. 2004. A spontaneously arising pancreatic tumor does not promote the differentiation of naive CD8+ T lymphocytes into effector CTL. J. Immunol. 172:6558.
Hanson, H. L., D. L. Donermeyer, H. Ikeda, J. M. White, V. Shankaran, L. J. Old, H. Shiku, R. D. Schreiber, P. M. Allen. 2000. Eradication of established tumors by CD8+ T cell adoptive immunotherapy. Immunity 13:265.
Matsui, K., L. A. O‘Mara, P. M. Allen. 2003. Successful elimination of large established tumors and avoidance of antigen-loss variants by aggressive adoptive T cell immunotherapy. Int. Immunol. 15:797.
Boggio, K., G. Nicoletti, E. Di Carlo, F. Cavallo, L. Landuzzi, C. Melani, M. Giovarelli, I. Rossi, P. Nanni, C. De Giovanni, et al 1998. Interleukin 12-mediated prevention of spontaneous mammary adenocarcinomas in two lines of Her-2/neu transgenic mice. J. Exp. Med. 188:589.
DeLeo, A. B., H. Shiku, T. Takahashi, M. John, L. J. Old. 1977. Cell surface antigens of chemically induced sarcomas of the mouse. I. Murine leukemia virus-related antigens and alloantigens on cultured fibroblasts and sarcoma cells: description of a unique antigen on BALB/c Meth A sarcoma. J. Exp. Med. 146:720.
Muller, W. J., E. Sinn, P. K. Pattengale, R. Wallace, P. Leder. 1988. Single-step induction of mammary adenocarcinoma in transgenic mice bearing the activated c-neu oncogene. Cell 54:105.
Lucchini, F., M. G. Sacco, N. Hu, A. Villa, J. Brown, L. Cesano, L. Mangiarini, G. Rindi, S. Kindl, F. Sessa, P. Vezzoni, L. Clerici. 1992. Early and multifocal tumors in breast, salivary, harderian and epididymal tissues developed in MMTY-Neu transgenic mice. Cancer Lett. 64:203.
Di Carlo, E., M. G. Diodoro, K. Boggio, A. Modesti, M. Modesti, P. Nanni, G. Forni, P. Musiani. 1999. Analysis of mammary carcinoma onset and progression in HER-2/neu oncogene transgenic mice reveals a lobular origin. Lab. Invest. 79:1261
Sotomayor, E. M., I. Borrello, E. Tubb, F.-M. Rattis, H. Bien, Z. Lu, S. Fein, S. Schoenberger, H. I. Levitsky. 1999. Conversion of tumor-specific CD4+ T-cell tolerance to T-cell priming through in vivo ligation of CD40. Nat. Med. 5:780.
Staveley-O’Carroll, K., E. Sotomayor, J. Montgomery, I. Borrello, L. Hwang, S. Fein, D. Pardoll, H. Levitsky. 1998. Induction of antigen-specific T cell anergy: an early event in the course of tumor progression. Proc. Natl. Acad. Sci. USA 95:1178.
Morgan, D. J., H. T. Kreuwel, L. A. Sherman. 1999. Antigen concentration and precursor frequency determine the rate of CD8+ T cell tolerance to peripherally expressed antigens. J. Immunol. 163:723.
Kurts, C., R. M. Sutherland, G. Davey, M. Li, A. M. Lew, E. Blanas, F. R. Carbone, J. F. P. Miller, W. R. Heath. 1999. CD8 T cell ignorance or tolerance to islet antigens depends on antigen dose. Proc. Natl. Acad. Sci. USA 96:12703.
Kurt, R. A., R. Whitaker, A. Baher, S. Seung, W. J. Urba. 2000. Spontaneous mammary carcinomas fail to induce an immune response in syngeneic FVBN202 neu transgenic mice. Int. J. Cancer 87:688.
Nanni, P., G. Nicoletti, C. De Giovanni, L. Landuzzi, E. Di Carlo, F. Cavallo, S. M. Pupa, I. Rossi, M. P. Colombo, C. Ricci, et al 2001. Combined allogeneic tumor cell vaccination and systemic interleukin 12 prevents mammary carcinogenesis in HER-2/neu transgenic mice. J. Exp. Med. 194:1195.
Nanni, P., L. Landuzzi, G. Nicoletti, C. De Giovanni, I. Rossi, S. Croci, A. Astolfi, M. Iezzi, E. Di Carlo, P. Musiani, G. Forni, P.-L. Lollini. 2004. Immunoprevention of mammary carcinoma in HER-2/neu transgenic mice is IFN- and B cell dependent. J. Immunol. 173:2288.(Leigh A. O’Mara, Lyse A. )
Peripheral tolerance to shared Ags expressed on both tumors and normal self-tissues presents a major barrier to T cell-based immunotherapy as a treatment for cancer. To assess the activity of tumor-specific T cells against spontaneously arising carcinomas in the context of shared Ag expression, we developed a model system whereby an identified tumor Ag, tumor ERK (tERK), is expressed transgenically on both normal mammary tissue and spontaneous mammary carcinomas. Transfer of in vitro-activated, tERK-specific DUC18 T cells delayed spontaneous tumor development in tERK-expressing mice when T cells were given before the development of palpable carcinomas. However, antitumor activity mediated by in vitro-activated DUC18 T cells, as measured by responsiveness against a transplanted tERK-expressing fibrosarcoma challenge, was lost within days of transfer. This loss was due to expression of tERK as a self-Ag on normal tissues and was independent of the presence of mammary tumors. In contrast, transferred naive DUC18 T cells maintained a long-term protective function in tERK-expressing mice. Ten-fold fewer naive T cells activated in vivo were able to replicate the delay in spontaneous tumor development achieved by in vitro-activated T cells. These results are in contrast to our earlier studies using transplanted tumors alone, in which in vitro-activated DUC18 T cells were more efficacious than naive DUC18 T cells and highlight the need to perform tumor studies in the presence of tumor Ag expression on normal self-tissue.
Introduction
Tumors in mice and humans can be recognized and destroyed by CD8+ T cells. This capability has lead to attempts to use T cells to specifically target tumors by adoptive immunotherapy, i.e., the transfer of tumor-specific T cells into patients with established tumors, in the hopes of causing tumor regression or rejection. Because it is not currently practical to identify unique tumor Ags for every patient, most adoptive immunotherapies target shared Ags expressed by both the tumor and normal self-tissue. These shared Ags are often tissue-restricted Ags expressed by the tumor and the normal tissue from which the tumor originated. Examples of immunotherapy trials that target-shared Ags are the melanoma trials targeting melanocyte Ags, including Melan-A, Mart-1, and gp100 (1, 2, 3, 4, 5).
Although targeting shared Ags allows more universally applicable therapies, transferred T cells are subject to peripheral tolerance mechanisms that prevent self-specific immune responses. Clinical studies have revealed dramatic reductions of in vitro-expanded T cells within 2 wk of adoptive transfer into cancer patients (1, 5, 6). Although nondeletional phenomena such as T cell exhaustion may contribute to this T cell loss, deletion of self-Ag-specific T cells is likely a critical factor. However, it is difficult to differentiate between these possibilities in clinical trials, and the examination of such mechanistic questions is better suited to murine tumor models.
To date, the vast majority of work examining interactions between tumors and T cells in mice has been done using transplantable tumors. These are often tumor cell lines that are transfected with model Ags to allow recognition by transgenic T cells. Therefore, the Ag in question is confined to the tumor itself. Although much has been learned from these systems, they do not allow the study of T cell activity in the presence of Ags expressed on normal self tissue.
Experiments designed to study the impact of peripheral tolerance on tumor-specific T cell responses by transplanting tumors into mice that transgenically express Ags on normal tissue or by using mice that spontaneously develop Ag-expressing tumors have shown a variety of results. Experiments using transplantable tumors in the presence of transgenic Ag expression in the liver have shown that naive T cells are anergized and cannot be activated unless restimulated in vitro in the presence of IL-2, which rescues in vivo function (7). Alternatively, other systems specific for melanocyte Ags have shown that vaccination with recombinant vaccinia virus and administration of IL-2 can restore naive T cell function in vivo (8). Experiments using transgenic mice that develop spontaneous pancreatic carcinomas that share Ag expression with normal pancreas show that naive T cells are ignorant of the tumor Ag (9, 10). Only one of these studies has used in vitro-activated T cells, and none of them has observed the in vivo deletion seen in patients.
Several years ago, our laboratory generated a TCR-transgenic system (DUC18) in which adoptively transferred CD8+ T cells are able to efficiently eliminate a transplanted murine fibrosarcoma, CMS5. DUC18 T cells recognize an epitope, designated tumor ERK (tERK) 4, that is generated by a point mutation in the ERK2 kinase expressed by CMS5. Adoptively transferred naive DUC18 T cells are able to completely reject small CMS5 tumors that are transplanted 3 days before T cell transfer (11). When DUC18 T cells are activated in vitro with a tERK peptide before transfer into tumor-bearing recipients, the kinetics of tumor rejection are more rapid, and the DUC18 T cells are able to eliminate much larger tumors that have been allowed to grow for 8 days before T cell transfer (12). Because DUC18 T cells proved to be so proficient in rejecting transplantable CMS5, we wanted to examine the efficacy of naive vs in vitro-activated DUC18 T cells in the more physiologically relevant context of spontaneously arising mammary carcinomas.
To do this, we generated transgenic mice (Dt) expressing the rejection Ag (tERK) of the CMS5 fibrosarcoma driven by the mouse mammary tumor virus (MMTV) promoter. Therefore, Dt mice express tERK on normal mammary tissues. These Dt mice were bred to NeuT mice, which express the cNeu oncogene under the control of the same promoter (13). Double transgenic (Dt/NeuT) mice spontaneously develop mammary carcinomas that express the tERK Ag. Adoptive transfer of in vitro-activated, tERK-specific DUC18 T cells is able to delay the onset of mammary tumor development in Dt/NeuT mice by several weeks if T cells are given before the development of palpable tumors. However, antitumor activity of in vitro-activated DUC18 T cells is rapidly lost in Dt/NeuT mice, as evidenced by their failure to clear transplanted CMS5 challenges. This loss occurs even in Dt mice and is therefore due to the expression of tERK as a self-Ag on normal tissues and is completely independent of the presence of mammary tumors. In contrast, when naive DUC18 T cells are activated in vivo in tERK-expressing Dt mice by CMS5 challenge, they not only mount a productive primary response but also generate a protective memory population that is able to reject a secondary CMS5 challenge at least 2 months later. Surprisingly, the adoptive transfer of 10-fold fewer naive DUC18 T cells, activated in vivo by CMS5 challenge, is able to effect the same delay in spontaneous mammary tumor formation as is seen with larger numbers of in vitro-activated DUC18 T cells.
Our findings highlight the potential differences between spontaneous and transplantable model systems and the importance of peripheral tolerance as a barrier to adoptive immunotherapy. In contrast to our earlier findings made using the transplantable CMS5 system, naive DUC18 T cells were more effective and had longer activity than in vitro-activated DUC18 T cells in this spontaneous mammary tumor system. Because naive T cells activated in vivo are able to survive, our results indicate that it should be possible to manipulate in vitro culture conditions to generate activated, tumor Ag- specific T cells that persist and therefore have enhanced efficacy in the presence of shared Ags in vivo.
Materials and Methods
Mice and cell lines
All mice used in this study are on the BALB/c background. DUC18 mice, originally generated in our laboratory (11), were bred and housed in the animal facility at the Washington University School of Medicine. Thy 1.1+ BALB/c mice were obtained from Dr. H. Levitsky (Johns Hopkins University, Baltimore, MD) and bred to DUC18 to generate Thy 1.1+ DUC18 mice. NeuT mice, developed by Forni and colleagues (13), were kindly provided by Dr. J. Berzofsky (National Cancer Institute, National Institutes of Health, Bethesda, MD).
The construct used to generate transgenic mice expressing tERK in mammary tissue, designated Dt, was assembled by blunt ligation of tERK cDNA originally cloned from CMS5 into a MMTV vector containing the human growth hormone polyadenylation sequence (gift of R. Schreiber (Washington University School of Medicine, St. Louis, MO)). Purified DNA was injected into BALB/c pronuclei. Founder mice were screened by Southern blot using probes spanning the junction between the vector and tERK insert. Probes were generated by digesting the MMTVtERK plasmid with EcoRI and labeled using the Rediprime kit (Amersham Biosciences). A PCR screen was developed using primers that spanned the junction between the 3'-end of the tERK insert and the 5'-end of the human growth hormone poly-A sequence (forward: TATGACCCAAGTAATGAGCCCATT; reverse: CTGAGATTGGCCAAATACTGG).
Dt mice, designated Dt/NeuT, were generated by breeding male mice positive for both the cNeu and Dt transgenes to BALB/c females. Offspring were screened by PCR for the presence of the Dt and NeuT transgenes. Female mice were used for all studies.
CMS5 (14) was freshly thawed from frozen stock for every experiment and expanded through no more than three passages in RPMI 1640 medium supplemented with 10% FCS (HyClone), 1 mM glutamax (Invitrogen Life Technologies), and 50 μg/ml gentamicin (Invitrogen Life Technologies). All CMS5 challenges in this study were s.c. injections of 3 x 106 CMS5 cells in HBSS.
DUC18 T cell activation in vitro
In vitro-activated DUC18 T cells were generated as previously described by culturing DUC18 splenocytes at 5 x 106 cells/ml in RPMI 1640 medium supplemented with 10% FCS (HyClone), 1 mM glutamax (Invitrogen Life Technologies), 5 x 10–5 M 2-ME (Sigma-Aldrich), and 50 μg/ml gentamicin (Invitrogen Life Technologies) in the presence of 0.5 μM tERK peptide (QYIHSANVL) (12). Cells were harvested after 4 days in culture and purified over Ficoll-Paque (Amersham Biosciences). The resulting cell population was generally 70–90% DUC18 T cells, defined as being V8.3+and CD8+. Cells were analyzed by flow cytometry for the percentage of live lymphocytes that were V8.3+CD8+, and cell counts were compensated accordingly. T cell adoptive transfer was given i.v. in HBSS.
Detection of tERK message
The presence of tERK message in various tissues was assayed as previously described (12). Briefly, RNA was extracted from the tissue or cell line of interest using TRIzol (Invitrogen Life Technologies) and used to make cDNA. Primers spanning the tERK mutation (Integrated DNA Technologies) were used to amplify the cDNA. This PCR product was purified and digested overnight with SfcI (New England Biolabs). Digested PCR products were then resolved on a 2% agarose gel. The presence of a 150-bp band indicates the presence of tERK message.
Monitoring of tumor development
NeuT and Dt/NeuT mice were monitored twice weekly for the development of palpable mammary nodules. The development of persistent, progressive, palpable (3 mm in diameter) nodules marked tumor onset. In NeuT mice, mammary hyperplasia begins before 4 wk of age. By 10 wk of age, mice progress to carcinoma in situ. The development of palpable tumors typically occurs between 13 and 17 wk of age. Mammary tumor growth continued to be monitored until any single nodule exceeded 15 mm in diameter, or the mouse began showing visible signs of distress. These kinetics are similar to those previously published for this model (13).
CMS5 challenge consisted of 3 x 106 cells suspended in HBSS injected s.c. in the right hind flank. Mice were subsequently monitored for tumor development at the site of injection beginning 2 or 3 days after challenge. Tumor nodules were measured every other day for the duration of the experiment. Tumor area is calculated by multiplying two perpendicular diameters (11, 12).
Flow cytometry
For peripheral blood samples, RBC were lysed using buffered ammonium chloride before Ab staining. Peripheral blood samples were stained using directly conjugated anti-V8.3-FITC, anti-Thy1.1-PE, and anti-CD8-CyChrome. All Abs used in this study were purchased from BD Pharmingen. Data was collected either on a FACScan or a FACSCalibur and analyzed using CellQuest software (BD Pharmingen) or FlowJo (Tree Star).
Statistical analysis
Statistical significance of tumor incidence was analyzed by the Mann-Whitney U test using InStat software (GraphPad Software).
Results
Generation of tERK-transgenic mice
To assess the efficacy of transgenic T cells in treating spontaneous tumors expressing a shared tumor Ag, transgenic mice, referred to as Dt, were generated by expressing the tERK Ag under the control of the MMTV promoter. These Ag-expressing mice were bred to NeuT mice expressing the cNeu oncogene under the control of the MMTV promoter. The resulting double transgenic Dt/NeuT mice express tERK as an Ag shared by normal tissue and spontaneous mammary carcinomas. The effects of shared expression of tERK Ag on tERK-specific DUC18 T cells were then assessed in the context of spontaneous mammary carcinomas.
tERK message can be detected based on its susceptibility to digest by SfcI (Fig. 1A). Analysis of a variety of tissues from a 6-wk-old Dt virgin female mouse showed tERK expression in the mammary glands, salivary glands, and lungs, in agreement with previous characterization of the MMTV promoter (15, 16). There was also expression in the spleen (Fig. 1B). Dt mice were then crossed to NeuT mice, which have been described previously as spontaneously developing lobular carcinoma throughout their mammary tissue with complete penetrance. Tumor development in these mice is continuous, beginning when the mice are <4 wk old (13, 17). No changes in the kinetics of tumor development (Fig. 1C) or the pathology (data not shown) of mammary tumors that developed in double transgenic Dt/NeuT mice relative to NeuT mice were observed. The pathology and kinetics of the Dt/NeuT mice correlated with previously published characterization of NeuT mice (17). Analysis of mRNA taken from spontaneous tumors showed that tumors developing in Dt/NeuT double transgenic mice express tERK message (Fig. 1D). Therefore, the Dt transgene functions as an antigenic tag and does not alter the characteristics of tumors that develop in double transgenic Dt/NeuT mice. Therefore, tumors that develop in double transgenic mice are suitable targets for DUC18 T cell-mediated therapy.
FIGURE 1. Characterization of the Dt transgenic mouse. A, The tERK mutation generates a novel SfcI restriction site that can be used to assay cDNA for tERK message. Normal ERK (nERK) does not contain this restriction site. PCR products spanning the mutation are gene-cleaned and digested. tERK, but not nERK, message is susceptible to SfcI digestion. B, Survey of tERK expression in various organs from a 6-wk-old virgin Dt female. The lower band indicates the presence of tERK message, and the upper band is intact nERK message. C, Kinetics of tumor development in NeuT vs Dt/NeuT mice. Mice were monitored twice weekly for the development of palpable (>3 mm in diameter) tumors. The age at which mice first develop persistent, palpable tumors is the age of tumor onset. D, Assay for tERK message in mammary tumors taken from unmanipulated Dt/NeuT mice. The lower band indicates the tERK message, and the upper band indicates the nERK message (N = NeuT, Dt/N = Dt/NeuT).
Transfer of activated DUC18 T cells delays tumor development in Dt/NeuT mice
Dt/NeuT mice were bred directly to DUC18-transgenic mice to assess the ability of DUC18 T cells to delay mammary tumor formation. Approximately 80% of the CD8 T cells in the circulation of DUC18-transgenic mice are V8.3+CD8+. When Dt/NeuT/DUC18 triple transgenic mice were analyzed by flow cytometry, no such enrichment for DUC18 T cells was observed (data not shown). This indicated that DUC18 T cells were being deleted in these mice, likely due to low levels of tERK transgene expression in the thymus (Fig. 1B). Therefore, we used adoptive transfer to address whether DUC18 T cells were capable of delaying the formation of spontaneous mammary tumors in Dt/NeuT mice.
Previous studies using DUC18 T cells to treat transplantable tumors had indicated that T cells activated in vitro before transfer into the host were highly effective in rejecting large tumors (12). Therefore, a single dose of activated T cells (30 x 106) was administered to 4- or 9-wk-old Dt/NeuT mice to assess whether any delay in mammary tumor onset could be attained. In agreement with previous characterizations of NeuT mice, we find that at 4–5 wk of age, NeuT mice show mild mammary hyperplasia but no overt carcinoma (13). By 9–10 wk of age, there is histological evidence of carcinoma in situ but not palpable tumors (data not shown). There was a statistically significant delay in the onset of palpable mammary tumors in 4-wk-old Dt/NeuT mice that received activated T cells compared with untreated mice (Fig. 2A). A similar delay could be seen in mice that received T cells at 9–10 wk of age. The delay in tumor development seen in T cell-treated Dt/NeuT mice was also significant compared with T cell-treated NeuT mice (Fig. 2B). Thus, the delay in tumor formation depends both on the T cell and on the Ag expression of the host.
FIGURE 2. Treatment of Dt/NeuT mice with a single dose of in vitro-activated DUC18 T cells delays onset of mammary tumors. A, Kinetics of mammary tumor development in Dt/NeuT mice left untreated () or treated with 30 x 106 in vitro-activated DUC18 T cells at 4–5 wk of age () or 9–10 wk of age (). Statistical significance as determined by the Mann-Whitney U test for untreated vs treated at 4–5 wk of age, p = 0.0009, and at 9–10 wk of age, p 0.0001. B, Kinetics of tumor development in Dt/NeuT mice () or NeuT mice () treated at 4–5 wk of age with 30 x 106 in vitro-activated DUC18 T cells. Statistical significance as determined by the Mann-Whitney U test, p = 0.013. Mice shown in these experiments were collected over several months and are progeny from several different litters from several different breeding pairs.
These delays were noteworthy, given that the NeuT transgene induces continuous tumor formation. Thus, any T cell treatment can only delay tumor development as long as the T cells maintain activity in vivo. Mice > 16 wk of age exhibiting palpable mammary tumors showed no objective effect when treated with activated DUC18 T cells (data not shown). Therefore, treatment of Ag expressing Dt/NeuT mice with monospecific T cell immunotherapy could effect a delay in the development of spontaneous mammary carcinoma, provided that the T cells were administered before the onset of palpable tumors.
Loss of DUC18 T cell activity in the presence of Ag-expressing mammary tumors
Although the delay in tumor onset attained by the activated DUC18 T cells was encouraging, we sought to elucidate the reasons for the eventual development of mammary tumors in T cell-treated, Ag-expressing Dt/NeuT mice. To test whether tumors that developed in DUC18 T cell-treated mice were Ag loss variants, tumors were assayed for the presence of the tERK message. tERK message was present in multiple tumors taken from various Dt/NeuT mice that had been treated with DUC18 T cells at 4–5 wk of age (data not shown). Thus, Ag loss did not appear to be a mechanism of immune evasion in this model.
Because tumors develop continuously in NeuT mice, we wanted to determine whether tumor outgrowth was the result of a loss of DUC18 T cell activity in the T cell-treated mice. The transplantable CMS5 fibrosarcoma is susceptible to DUC18 T cell rejection and can therefore be used to assess the presence of systemic DUC18 T cell activity. Mice that received 30 x 106 activated DUC18 T cells at 4–5 or 9–10 wk of age were challenged with CMS5 shortly after the development of their first palpable mammary tumors to assess DUC18 T cell activity. At the time of mammary tumor development (17–20 wk of age), there was little to no DUC18 T cell activity present in the Dt/NeuT mice as indicated by the uninhibited growth of CMS5 in these mice (data not shown). In contrast, all of the non-Ag-expressing NeuT mice exhibited unimpaired DUC18 T cell function (data not shown).
We wanted to assess T cell activity over a shorter period of time than that described in the above experiments, which took place over a period of many months. To maximize the effects of the tumor environment, mice bearing palpable mammary tumors were used to assess the impact of shared Ag expression on short-term T cell function. Therefore, the dose of T cells given to the mice was lowered to 3 x 106, and the CMS5 challenge was administered 3 days after the T cell transfer. Even in this short time frame, there was a marked decrease in the activity of DUC18 T cells in Ag-expressing Dt/NeuT mice exhibited by a failure to control CMS5 outgrowth (Fig. 3). These results indicate a rapid loss of T cell function in the presence of palpable, Ag-expressing tumors and shared Ags expressed on normal tissues.
FIGURE 3. T cell activity is rapidly lost in mice bearing spontaneous, Ag-expressing mammary tumors. T cell activity was measured after T cell transfer into mammary tumor bearing Dt/NeuT or NeuT mice. Mice with palpable mammary tumors received 3 x 106 in vitro-activated DUC18 T cells. Three days following T cell transfer, mice received a s.c. challenge of CMS5 to test DUC18 T cell activity. Shown is CMS5 tumor area over time in individual mice from a total of three independent experiments. A total of 10 NeuT and 10 Dt/NeuT mice was tested.
In vitro-activated T cells are eliminated rapidly in Dt/NeuT mice
The previous experiments raised the question of whether DUC18 T cells were present but nonfunctional in the presence of tERK as a self-Ag or were eliminated altogether. To differentiate between these possibilities, we used flow cytometry to monitor the persistence of DUC18 T cells in the peripheral blood after their transfer into 4- to 5-wk-old Dt/NeuT or NeuT mice. For these experiments, 30 x 106 activated Thy 1.1+ DUC18 T cells were transferred into Thy 1.2+ Dt/NeuT or NeuT mice, and recipients were analyzed periodically for the presence of circulating Thy1.1+ T cells. Throughout the period of observation, levels of Thy1.1+ T cells were 5- to 10-fold higher in NeuT mice than in Dt/NeuT mice, beginning 3 days after T cell transfer (Fig. 4). Experiments using 9-to 10-wk-old recipients showed a similar loss of circulating Thy 1.1+ DUC18 T cells in Ag-expressing Dt/NeuT mice (data not shown). We surveyed organs for T cells but were unable to detect accumulated DUC18 T cells in the spleen, lymph nodes, or mammary tissues of Ag-expressing mice (data not shown). Therefore, it does not appear that the T cells are being trapped or sequestered in tERK- expressing tissues or in secondary lymphoid tissues. Rather, it appears that activated DUC18 T cells were being eliminated rapidly following transfer in Ag-expressing hosts.
FIGURE 4. In vitro-activated DUC18 T cells are rapidly lost in Dt/NeuT hosts. A total of 30 x 106 Thy 1.1+ in vitro-activated DUC18 T cells was transferred into 4- to 5-wk-old Thy 1.2+ Dt/NeuT () or NeuT () hosts. Peripheral blood was tested periodically for the percentage of peripheral blood lymphocytes that were Thy 1.1+,V8.3+ as an indication of transferred DUC18 T cells. Shown is the percentage of peripheral blood lymphocytes in individual NeuT () or Dt/NeuT mice () that are Thy1.1,V8.3+ over the course of the experiment. Results are representative of four independent experiments.
Loss of T cell activity is Ag, but not oncogene, dependent
The loss of DUC18 T cells in Dt/NeuT mice could be due to the expression of cognate Ag on self-tissue or to immunosuppressive effects of the tumors that develop. To determine whether transgenic tERK expression alone was sufficient to eliminate transferred DUC18 T cells in the periphery, 3 x 106 in vitro-activated T cells were transferred into Dt or BALB/c mice. Three days later, mice were challenged with CMS5. Dt transgenic mice showed a marked defect in the ability to reject this challenge (Fig. 5). Tracking experiments assaying DUC18 T cell persistence in the circulation of Dt mice showed a similar loss of T cells as that seen in Dt/NeuT mice (data not shown). Because Dt mice do not express the cNeu oncogene, they do not develop mammary tumors. Therefore, the loss of in vitro-activated DUC18 T cells seen in the Dt mice is due solely to the expression of tERK on normal peripheral tissues and not to any tumor-induced immune dysfunction.
FIGURE 5. Inactivation of DUC18 T cells is due to expression of tERK as a self-Ag. Groups of Dt or BALB/c mice were treated with 3 x 106 in vitro-activated DUC18 T cells and received a s.c. CMS5 challenge 3 days later. Subsequent CMS5 growth was measured to indicate in vivo DUC18 T cell activity. Shown are the areas of CMS5 tumors over time in individual Dt () or BALB/c () mice. Results are representative of three independent experiments.
Naive DUC18 T cells become activated in the presence of peripherally expressed tERK
Previous studies examining the ability of naive T cells to become activated in the presence of self-Ag have shown that naive T cells can become anergized and can only be rescued by restimulation in vitro in the presence of IL-2 (7). Because tERK expression in Dt transgenic mice appeared sufficient to delete in vitro-activated DUC18 T cells, we sought to assess the impact of the Dt transgene on the function of naive DUC18 T cells. Naive DUC18 T cells were transferred into Dt or BALB/c mice 24 h before CMS5 challenge to determine the effect of transgenic tERK expression on the activation and persistence of DUC18 T cells. Although 10- to 100-fold more DUC18 T cells were required to reject CMS5 challenge relative to BALB/c mice, naive DUC18 T cells were capable of rejecting CMS5 challenge in Dt mice when present in sufficient numbers (Fig. 6). In four independent experiments, 80% of Dt mice that received 3.5 x 105 naive DUC18 T cells were able to reject subsequent CMS5 challenge. Mice that received 3.5 x 104 naive DUC18 T cells rejected 40% of CMS5 challenges in three independent experiments (Fig. 6 and data not shown). These data indicate that, in this system, naive DUC18 T cells can acquire effector function and reject transplantable tumors, even in the presence of tERK expressed as a self-Ag.
FIGURE 6. Naive DUC18 T cells can demonstrate effector function in the presence of self-Ag. Naive DUC18 T cells were transferred into Dt or BALB/c mice in the doses indicated. Twenty-four hours later, mice received a s.c. CMS5 challenge. Shown is the tumor area over time for individual mice in one experiment representative of three independent experiments. The number in the upper left corner of each graph represents the dose of naive T cells transferred. BALB/c controls are shown on the left, and Ag-expressing Dt mice are in the right column.
T cells resulting from in vivo activation are capable of prolonged persistence in the presence of tERK as a self-Ag
Because the DUC18 T cells were able to mount an immune response to CMS5 even in the presence of tERK expressed on self-tissue, we were able to assess the impact of shared Ag expression on the development of a tumor-specific memory response. Mice that had rejected a previous challenge with CMS5 were rechallenged 2 mo later, and the growth of this second challenge was monitored to assess T cell activity. In every experiment, there was evidence of DUC18 T cell activity in Dt mice that had received naive cells and then were rechallenged with CMS5 (Fig. 7). When mice that were able to reject primary CMS5 challenges with either 3.5 x 105 or 3.0 x 106 DUC18 T cells were rechallenged 2 mo later, 70% of these secondary challenges were rejected in four independent experiments. These results contrast with similar experiments done using in vitro-activated T cells, which cannot sustain T cell activity in Dt mice for more than days or weeks at most (Figs. 3 and 4 and data not shown).
FIGURE 7. DUC18 T cells activated in vivo can maintain effector function for prolonged periods in the presence of the Dt transgene. Dt or BALB/c mice received transfers of naive DUC18 T cells and were challenged with CMS5 the next day. Mice that rejected this primary challenge were rechallenged 2 mo later with a second dose of CMS5. Shown are the growth kinetics of the secondary CMS5 challenge in individual Dt () and BALB/c () mice, as well as CMS5 growth in untreated BALB/c mice (). Results are representative of four individual experiments. In two experiments, an initial dose of 3.5 x 105 DUC18 T cells was used, and in two other experiments, a dose of 3 x 106 naive T cells was used. Similar outcomes were observed with both doses.
In vivo-activated T cells significantly delay spontaneous tumor development
Because in vivo-activated DUC18 T cells can maintain function for long periods of time in the presence of self-Ag, we evaluated whether they might be more efficient than in vitro-activated DUC18 T cells in delaying the development of spontaneous tumors in Dt/NeuT mice. Therefore, 4- to 5-wk-old NeuT or Dt/NeuT mice were treated with 3 x 106 naive T cells and then challenged with CMS5 to activate these cells in vivo. Mice that rejected this CMS5 challenge were then monitored for mammary tumor development. Surprisingly, this relatively low number of naive T cells achieved a delay in spontaneous tumor development (Fig. 8A). In fact, the kinetics of tumor development in mice treated with naive T cells activated in vivo was identical with that created by a 10-fold higher dose of in vitro-activated cells (Fig. 8B). Although these results are promising, the fact that the treated mice eventually developed mammary tumors indicates that there may be other factors dictating the efficacy of DUC18 T cells against spontaneous mammary tumors. It is also possible that the number of effector cells generated from transfer of 3 x 106 naive DUC18 T cells is insufficient to delay tumor development beyond that achieved by 30 x 106 in vitro-activated cells. These experiments indicate that, although longer lived T cells are much more efficient in preventing spontaneous tumor onset, there are still limitations to their efficacy that must be further investigated.
FIGURE 8. Low doses of naive T cells can delay spontaneous tumor development. A total of 3 x 106 naive DUC18 T cells was transferred into 4-wk-old Dt/NeuT or NeuT hosts. Twenty-four hours later, mice received a s.c. CMS5 challenge. Mice that rejected the CMS5 challenge were monitored for the development of spontaneous mammary tumors. A, Age of initial mammary tumor development of NeuT/Dt vs NeuT mice treated with in vivo-activated T cells. Statistical significance as determined by the Mann-Whitney U test, p = 0.02. B, Comparison of the kinetics of mammary tumor development in Ag-expressing mice treated with 3 x 106 naive T cells or with 30 x 106 in vitro-activated T cells at 4–5 or 9–10 wk of age. Data for in vitro-activated T cell treatment are taken from the experiments shown in Fig. 2.
Discussion
In the present study, we describe a system in which in vitro-activated T cells are able to effect a delay in tumor development in a spontaneous tumor model. We observed that in vitro-activated DUC18 T cells are eliminated rapidly in Ag-expressing Dt/NeuT mice, similar to what has been described in patients treated by adoptive transfer of in vitro-expanded tumor-infiltrating lymphocytes or PBL (1, 5, 6). This T cell loss is due to expression of tERK as a self-Ag because it also occurs in Dt mice lacking mammary tumors. Thus, in our model system, T cell deletion does not require the presence of tumors. Surprisingly, while naive DUC18 T cells are slightly less efficacious in the presence of the tERK transgene than in wild-type BALB/c mice, they are still able to become activated and reject transplanted CMS5 challenges. Naive DUC18 T cells activated in vivo in the presence of the tERK transgene are able to persist for months after activation, whereas in vitro-activated cells are eliminated within days. A transfer of only 3 x 106 naive DUC18 T cells is able to effect an identical delay in spontaneous tumor formation to that seen following transfer of 30 x 106 in vitro-activated DUC18 T cells. Overall, our findings suggest that in vitro-activated T cells are highly susceptible to peripheral deletion after an encounter with self-Ag in vivo, and this dramatically limits their antitumor efficacy in vivo.
These results are in sharp contrast to the efficacy of activated DUC18 T cells in the transplantable CMS5 system. Previous studies have shown that, when treating the CMS5 fibrosarcoma in non-tERK-expressing hosts, in vitro-activated DUC18 T cells were faster and more effective at treating rejectable tumors and could reject large tumors that naive DUC18 T cells were incapable of destroying (11, 12). From these results, we predicted that in vitro-activated DUC18 T cells also would be more effective at treating Dt/NeuT mice, in which spontaneous tERK-expressing tumors were present. However, this proved not to be the case because the antitumor response mediated by in vitro-activated DUC18 T cells is extremely short lived. The CMS5 system is among a small group of transplantable systems in which the T cell is able to control established tumors. In other systems, the ability of the tumor-specific T cell to control tumor growth is limited by tolerance of the T cells to the tumor (18, 19). We cannot eliminate the possibility that the differences in DUC18 T cell activity against Dt/NeuT tumors and transplantable CMS5 are a result of the unusual susceptibility of CMS5 to DUC18-mediated killing, rather than differences in T cell activity against spontaneous vs transplantable tumors. Although CMS5 may be unusual in the ability of DUC18 T cells to reject the tumor, the stark contrast between the efficacy of naive and activated T cells in the spontaneous vs transplantable systems does highlight the immunological differences between these systems and demonstrates that mechanisms and therapies discovered using transplantable systems should be confirmed in spontaneous tumor models.
In other model systems, there is a link between the levels of Ag expressed on peripheral tissues and the fate of T cells transferred into Ag-expressing mice. High levels of Ag expressed on normal tissue have been shown to delete Ag-specific T cells in the periphery. Indeed, this peripheral deletion was dependent on the dose of Ags expressed (20, 21). In these systems, naive cells transferred into Ag-expressing mice rapidly proliferated and were eliminated. In our model system, we believe that in vitro-activated DUC18 T cells are deleted rapidly following adoptive transfer into Dt or Dt/NeuT mice because we are unable to detect them in the peripheral blood, the spleen, the lymph nodes, or mammary tissues (Fig. 4 and data not shown). Because we detected high levels of tERK mRNA in the spleens of Dt and Dt/NeuT mice, it is possible that a high dose of Ags in this organ was mediating deletion of transferred DUC18 T cells. However, we found the same loss of in vitro-activated DUC18 T cells upon their transfer into splenectomized Dt mice (data not shown); thus, a potentially high dose of tERK Ag in this site does not appear to mediate T cell deletion in our model. Interestingly, in another model system, the presence of large Ag-expressing spontaneous pancreatic tumors was shown to actually facilitate Ag drainage, cross-presentation, and activation of naive T cells (22). Whether the dose of Ag in our system is insufficient to activate naive cells or the type or activation state of the APC presenting the Ag is not conducive to the activation of naive T cells is unclear. However, if DUC18 T cells are activated in vitro before transfer into the Ag expressing hosts, there is sufficient Ag expressed to delete these cells. This may suggest that different thresholds of antigenic stimulation exist for activation of naive T cells vs deletion of previously activated effector T cells.
NeuT and Dt/NeuT mice were treated with naive DUC18 T cells activated in vivo by CMS5 challenge in an attempt to address whether in vivo survival was the major factor limiting the efficacy of the in vitro-activated DUC18 T cells in delaying mammary tumor development. A 10-fold lower dose of naive T cells was able to induce a delay in tumor development comparable to that sustained by in vitro-activated T cells. The factors limiting the efficacy of naive T cells are unknown, but the delay achieved by this relatively low dose of T cells indicates that even a small population of persistent T cells can cause a marked delay in spontaneous tumor onset. Whether the final limitation is simply the number of specific precursors available or a more fundamental mechanistic barrier to tumor rejection remains to be seen.
The delay in tumor development achieved by the naive DUC18 T cells and their persistence in the Dt mice is even more noteworthy in light of studies showing the peripheral tolerance of naive and effector/memory T cells specific for self-Ag. T cells specific for self-Ags have been shown to lose function or be deleted altogether (7, 23, 24, 25, 26). The persistence of the in vivo-activated DUC18 T cells is even more surprising given the deletion of self-specific T cells at all stages of differentiation. Elucidating the reasons for the persistence of these T cells in the presence of self-Ag expression could lead to the development of in vitro regimens to produce effector T cells capable of persisting for long periods of time in the presence of shared tumor Ag.
Although all T cell-treated mice eventually developed tumors in this study, one important consideration is the fact that the NeuT transgene leads to continuous development of multiple, rapidly growing mammary carcinomas. Tumors develop throughout the murine mammary tissue, which is quite extensive. This means that even a short delay in tumor formation in this system could translate into a much more significant effect in a clinical setting where tumor formation is a singular event. Additionally, the continuous generation of new tumors means that the loss of T cell activity in vivo leads to immediate outgrowth of subsequent tumors, which would not happen in the case of singular tumor development.
The fact that naive DUC18 T cells activated by CMS5 challenge in vivo are able to persist and establish protective memory T cell populations indicates that it should be possible to prevent the loss of in vitro-activated T cells by altering their culture conditions. Unfortunately, our attempts achieve this were unsuccessful. Addition of IL-2 or IL-15 to the in vitro culture, provision of supplemental CD4 help, or stimulation with low concentrations of peptide failed to generate in vitro-activated T cells capable of persisting in Ag-expressing mice (data not shown). Another possibility is that in our in vitro culture, where bulk splenocytes are being treated with tERK peptide, the type of APC present is not providing optimal T cell stimulation. In a bulk splenocyte preparation, B cells and semimature dendritic cells (DC) are likely responsible for the majority of tERK presentation to naive DUC18 T cells. During in vivo stimulation of naive DUC18 T cells following CMS5 challenge, DC in tumor draining lymph nodes may have acquired a more mature phenotype and therefore may be more suited to differentiating DUC18 T cells into competent effectors. Altering the type and frequency of DC during in vitro stimulation could provide the signals required to generate long-lasting DUC18 effectors.
The induction of a productive, tumor-specific T cell response in the face of peripheral tolerance has long presented a problem for tumor immunologists. In animal models, tolerance of endogenous tumor-specific T cells has been observed against spontaneous tumors (27, 28), but isolation and transfer of immunotherapeutic clones have not been attempted in most of these systems. Polyclonal therapies, such as DNA vaccination or IL-12 treatment, have likewise been effective in achieving impressive delays in tumor development in NeuT mice. Indeed, allogeneic vaccination combined with IL-12 treatment in NeuT mice can prevent tumor formation for up to a year and limits the number of tumors that do develop in individual mice (13, 29). These effects are the result of Ab and IFN- produced in the NeuT mice (30). Unfortunately, these systems depend on the stimulation of a polyclonal response and cannot then address what might be happening to specific T cells in the course of the response. Systems in which identified T cell clones are used to treat spontaneous tumors show T cell ignorance of the tumors (9, 10).
In the present study, we introduce a system in which a characterized, identifiable, tumor-specific T cell can delay the development of spontaneous mammary carcinomas. Interestingly, in our system, although adoptively transferred in vitro-activated DUC18 T cells are short lived, naive DUC18 T cells activated in vivo are resistant to elimination. Elucidation of the factors responsible for this in vivo persistence would represent a significant advancement to the field of tumor immunotherapy and would then allow the additional characterization of the conditions that dictate survival or deletion of T cells specific for shared Ags in vivo. The fact that the therapeutic efficacy of these T cells in the presence of a shared Ag is a sharp contrast to the activity of this T cell in the presence of a unique tumor Ag highlights the need for shared Ag models to study T cell behavior in these more physiologic situations. The persistence of in vivo-activated T cells in the presence of persistent Ag indicates that there is a way to permit long-term survival of shared Ag-specific T cells in vivo. The recapitulation of this T cell survival by manipulation of in vitro expansion techniques would be a great benefit to the field of tumor immunology.
Disclosures
The authors have no financial conflict of interest.
Acknowledgments
We thank Dave Donermeyer and Dr. Ken Matsui for technical assistance in creating the Dt transgenic construct, Donna Thompson and Jerri Smith for clerical assistance, Kathy Frederick for technical assistance in performing splenectomies, Drs. Guido Forni and Jay Berzofsky for guidance in obtaining and maintaining the NeuT mice, and Silvia Kang, Gavin Dunn, and Drs. Laura Mandik-Nayak and Fei Shih for critical reading of this manuscript.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by grants from the National Institutes of Health. L.A.O. was supported by a Howard Hughes Predoctoral Training Grant. L.A.N. is supported by a postdoctoral fellowship from the Cancer Research Institute.
2 Current address: Emory University School of Medicine, Department of Microbiology and Immunology, Atlanta, GA 30322.
3 Address correspondence and reprint requests to Paul M. Allen, Washington University School of Medicine, Department of Pathology and Immunology, 660 South Euclid Avenue, Campus Box 8118, St. Louis, MO 63110. E-mail address: pallen{at}wustl.edu
4 Abbreviations used in this paper: tERK, tumor ERK; MMTV, mouse mammary tumor virus; DC, dendritic cell.
Received for publication November 23, 2004. Accepted for publication January 29, 2005.
References
Yee, C., J. A. Thompson, D. Byrd, S. R. Riddell, P. Roche, E. Celis, P. D. Greenberg. 2002. Adoptive T cell therapy using antigen-specific CD8+ T cell clones for the treatment of patients with metastatic melanoma: in vivo persistence, migration, and antitumor effect of transferred T cells. Proc. Natl. Acad. Sci. USA 99:16168.
Dudley, M. E., L. T. Ngo, J. Westwood, J. R. Wunderlich, S. A. Rosenberg. 2000. T-cell clones from melanoma patients immunized against an anchor-modified gp100 peptide display discordant effector phenotypes. Cancer J. 6:69.
Dudley, M. E., J. R. Wunderlich, P. F. Robbins, J. C. Yang, P. Hwu, D. J. Schwartzentruber, S. L. Topalian, R. Sherry, N. P. Restifo, A. M. Hubicki, et al 2002. Cancer regression and autoimmunity in patients after clonal repopulation with antitumor lymphocytes. Science 298:850.
Dudley, M. E., J. R. Wunderlich, J. C. Yang, P. Hwu, D. J. Schwartzentruber, S. L. Topalian, R. M. Sherry, F. M. Marincola, S. F. Leitman, C. A. Seipp, et al 2002. A phase I study of nonmyeloablative chemotherapy and adoptive transfer of autologous tumor antigen-specific T lymphocytes in patients with metastatic melanoma. J. Immunother. 25:243.
Dudley, M. E., J. Wunderlich, M. I. Nishimura, D. Yu, J. C. Yang, S. L. Topalian, D. J. Schwartzentruber, P. Hwu, F. M. Marincola, R. Sherry, S. F. Leitman, S. A. Rosenberg. 2001. Adoptive transfer of cloned melanoma-reactive T lymphocytes for the treatment of patients with metastatic melanoma. J. Immunother. 24:363.
Yee, C., J. A. Thompson, P. Roche, D. R. Byrd, P. P. Lee, M. Piepkorn, K. Kenyon, M. M. Davis, S. R. Riddell, P. D. Greenberg. 2000. Melanocyte destruction after antigen-specific immunotherapy of melanoma: direct evidence of T cell-mediated vitiligo. J. Exp. Med. 192:1637.
?hlen, C., M. Kalos, D. J. Hong, A. C. Shur, P. D. Greenberg. 2001. Expression of a tolerizing tumor antigen in peripheral tissue does not preclude recovery of high-affinity CD8+ T cells or CTL immunotherapy of tumors expressing the antigen. J. Immunol. 166:2863.
Overwijk, W. W., M. R. Theoret, S. E. Finkelstein, D. R. Surman, L. A. de Jong, F. A. Vyth-Dreese, T. A. Dellemijn, P. A. Antony, P. J. Spiess, D. C. Palmer, et al 2003. Tumor regression and autoimmunity after reversal of a functionally tolerant state of self-reactive CD8+ T cells. J. Exp. Med. 198:569.
Speiser, D. E., R. Miranda, A. Zakarian, M. F. Bachmann, K. McKall-Faienza, B. Odermatt, D. Hanahan, R. M. Zinkernagel, P. S. Ohashi. 1997. Self antigens expressed by solid tumors do not efficiently stimulate naive or activated T cells: implications for immunotherapy. J. Exp. Med. 186:645.
Lyman, M. A., S. Aung, J. A. Biggs, L. A. Sherman. 2004. A spontaneously arising pancreatic tumor does not promote the differentiation of naive CD8+ T lymphocytes into effector CTL. J. Immunol. 172:6558.
Hanson, H. L., D. L. Donermeyer, H. Ikeda, J. M. White, V. Shankaran, L. J. Old, H. Shiku, R. D. Schreiber, P. M. Allen. 2000. Eradication of established tumors by CD8+ T cell adoptive immunotherapy. Immunity 13:265.
Matsui, K., L. A. O‘Mara, P. M. Allen. 2003. Successful elimination of large established tumors and avoidance of antigen-loss variants by aggressive adoptive T cell immunotherapy. Int. Immunol. 15:797.
Boggio, K., G. Nicoletti, E. Di Carlo, F. Cavallo, L. Landuzzi, C. Melani, M. Giovarelli, I. Rossi, P. Nanni, C. De Giovanni, et al 1998. Interleukin 12-mediated prevention of spontaneous mammary adenocarcinomas in two lines of Her-2/neu transgenic mice. J. Exp. Med. 188:589.
DeLeo, A. B., H. Shiku, T. Takahashi, M. John, L. J. Old. 1977. Cell surface antigens of chemically induced sarcomas of the mouse. I. Murine leukemia virus-related antigens and alloantigens on cultured fibroblasts and sarcoma cells: description of a unique antigen on BALB/c Meth A sarcoma. J. Exp. Med. 146:720.
Muller, W. J., E. Sinn, P. K. Pattengale, R. Wallace, P. Leder. 1988. Single-step induction of mammary adenocarcinoma in transgenic mice bearing the activated c-neu oncogene. Cell 54:105.
Lucchini, F., M. G. Sacco, N. Hu, A. Villa, J. Brown, L. Cesano, L. Mangiarini, G. Rindi, S. Kindl, F. Sessa, P. Vezzoni, L. Clerici. 1992. Early and multifocal tumors in breast, salivary, harderian and epididymal tissues developed in MMTY-Neu transgenic mice. Cancer Lett. 64:203.
Di Carlo, E., M. G. Diodoro, K. Boggio, A. Modesti, M. Modesti, P. Nanni, G. Forni, P. Musiani. 1999. Analysis of mammary carcinoma onset and progression in HER-2/neu oncogene transgenic mice reveals a lobular origin. Lab. Invest. 79:1261
Sotomayor, E. M., I. Borrello, E. Tubb, F.-M. Rattis, H. Bien, Z. Lu, S. Fein, S. Schoenberger, H. I. Levitsky. 1999. Conversion of tumor-specific CD4+ T-cell tolerance to T-cell priming through in vivo ligation of CD40. Nat. Med. 5:780.
Staveley-O’Carroll, K., E. Sotomayor, J. Montgomery, I. Borrello, L. Hwang, S. Fein, D. Pardoll, H. Levitsky. 1998. Induction of antigen-specific T cell anergy: an early event in the course of tumor progression. Proc. Natl. Acad. Sci. USA 95:1178.
Morgan, D. J., H. T. Kreuwel, L. A. Sherman. 1999. Antigen concentration and precursor frequency determine the rate of CD8+ T cell tolerance to peripherally expressed antigens. J. Immunol. 163:723.
Kurts, C., R. M. Sutherland, G. Davey, M. Li, A. M. Lew, E. Blanas, F. R. Carbone, J. F. P. Miller, W. R. Heath. 1999. CD8 T cell ignorance or tolerance to islet antigens depends on antigen dose. Proc. Natl. Acad. Sci. USA 96:12703.
Kurt, R. A., R. Whitaker, A. Baher, S. Seung, W. J. Urba. 2000. Spontaneous mammary carcinomas fail to induce an immune response in syngeneic FVBN202 neu transgenic mice. Int. J. Cancer 87:688.
Nanni, P., G. Nicoletti, C. De Giovanni, L. Landuzzi, E. Di Carlo, F. Cavallo, S. M. Pupa, I. Rossi, M. P. Colombo, C. Ricci, et al 2001. Combined allogeneic tumor cell vaccination and systemic interleukin 12 prevents mammary carcinogenesis in HER-2/neu transgenic mice. J. Exp. Med. 194:1195.
Nanni, P., L. Landuzzi, G. Nicoletti, C. De Giovanni, I. Rossi, S. Croci, A. Astolfi, M. Iezzi, E. Di Carlo, P. Musiani, G. Forni, P.-L. Lollini. 2004. Immunoprevention of mammary carcinoma in HER-2/neu transgenic mice is IFN- and B cell dependent. J. Immunol. 173:2288.(Leigh A. O’Mara, Lyse A. )