Purpose: Cancer vaccines targeting tumor-associated antigens are being investigated for the therapy of tumors. Numerous strategies, including the direct intratumoral (i.t.) vaccination route, have been examined. For tumors expressing carcinoembryonic antigen (CEA) as a model tumor-associated antigen, we previously designed poxviral vectors that contain the transgenes for CEA and a triad of T-cell costimulatory molecules, B7-1, intercellular adhesion molecule-1, (ICAM-1), and leukocyte function associated antigen-3 (LFA-3) (CEA/TRICOM). Two types of poxvirus vectors were developed: replication-competent recombinant vaccinia and replication-defective recombinant fowlpox. We have shown previously that a vaccine regimen composed of priming mice s.c. with recombinant vaccinia-CEA/TRICOM and boosting i.t. with recombinant fowlpox-CEA/TRICOM was superior to priming and boosting vaccinations using the conventional s.c. route in inducing T-cell responses specific for CEA. These studies also showed that CEA was needed to be present both in the vaccine and in the tumor for therapeutic effects.

Experimental Design: To determine specific immune responses associated with vaccination-mediated tumor regression, CEA-transgenic mice bearing CEA+ tumors were vaccinated with the CEA/TRICOM s.c./i.t. regimen, and T-cell immune responses were assessed.

Results: In CEA+ tumor-bearing mice vaccinated with the CEA/TRICOM s.c./i.t. regimen, T-cell responses could be detected not only to CEA encoded in vaccine vectors but also to other antigens expressed on the tumor itself: wild-type p53 and an endogenous retroviral epitope of gp70. Moreover, the magnitude of CD8+ T-cell immune responses to gp70 was far greater than that induced to CEA or p53. Finally, the predominant T-cell population infiltrating the regressing CEA+ tumor after therapy was specific for gp70.

Conclusion: These studies show that the breadth and magnitude of antitumor immune cascades to multiple antigens could be critical in the therapy of established tumors.

Cancer vaccines are under investigation for the therapy of many types of cancers. Numerous strategies, including the route of vaccination, are being evaluated using a variety of vaccines (1–8).

We have previously described poxviral vectors that contain the transgenes for carcinoembryonic antigen (CEA) and a triad of T-cell costimulatory molecules, B7-1, intercellular adhesion molecule-1 (ICAM-1), and leukocyte function associated antigen-3 (LFA-3) (TRICOM) as well as vectors coexpressing the combination (CEA/TRICOM). Two types of poxvirus vectors were developed: replication-competent recombinant vaccinia (rV) and replication-defective recombinant fowlpox (rF; refs. 1, 9–11). It has been shown that a diversified systemic prime and boost vaccine regimen, in which rV-CEA/TRICOM was used as a prime and rF-CEA/TRICOM as a boost, was more efficacious than continuous vaccination with rF-CEA/TRICOM in the induction of CEA-specific T-cell responses and antitumor activity (9, 11).

We have shown recently the advantage of an immunization regimen involving a s.c. prime vaccination with rV-CEA/TRICOM followed by intratumoral (i.t.) boosting with rF-CEA/TRICOM. These studies were done in CEA-transgenic mice, where CEA is a self-antigen (8). This vaccine regimen was shown to be superior to s.c. priming followed by s.c. boosting. In addition, when wild-type viruses were used or when vaccines not encoding CEA (TRICOM) were used for either s.c. priming or i.t. boosting, tumor growth was not suppressed (8). It is also important to note that this regimen was not effective on CEA tumors. These studies showed the requirement of having the tumor-associated antigen (TAA) CEA both in the vaccine and in the tumor for the induction of the optimal antitumor activity as well as prime vaccination with rV virus and the addition of granulocyte macrophage colony-stimulating factor (GM-CSF). It was not known, however, what role, if any, other TAAs present in the target tumor played in the antitumor immune response.

We show here that in CEA+ tumor-bearing CEA-transgenic mice vaccinated with the CEA/TRICOM s.c./i.t. regimen, T-cell immune responses could be detected not only to CEA encoded in vaccine vectors but also to other antigens expressed in the tumor: wild-type p53 and an endogenous retroviral epitope of gp70. Moreover, the magnitude of CD8+ T-cell immune responses to gp70 was far greater than that induced to CEA or p53. A majority of the T-cell population infiltrating the CEA+ tumor after therapy were specific for gp70. These studies show the s.c./i.t. CEA/TRICOM vaccine regimen induces an antigen cascade, which in turn facilitates the regression of established tumors. These studies also underscore the need, in clinical vaccine studies, to examine host responses after vaccination not only to the antigen in the vaccine but also to other TAAs known to be associated with a given tumor type.

Animals and Tumor Cells. Female C57BL/6 mice transgenic for human CEA (CEA-transgenic) were obtained from a breeding pair provided by Dr. John Thompson (Institute of Immunobiology, University of Freiburg, Freiburg, Germany). The generation and characterization of the CEA-transgenic mouse has been described previously (12, 13). Mice were housed and maintained under pathogen-free conditions in microisolator cages until used for experiments at 6 to 8 weeks of age. The following tumor cell lines (H-2b) were used: lymphoma EL-4 cells, parental murine colon adenocarcinoma MC38 cells (14), MC38 cells expressing human CEA (MC38-CEA+; ref. 14), melanoma B16 cells (CEAgp70+), and mammary Mtag cells (CEAgp70; ref. 15); Mtag expressing human CEA (Mtag-CEA+) and Mtag expressing murine gp70 (Mtag-gp70+) were established from parental Mtag cells via retroviral transduction with PLNSX2 (Retro-X, BD Biosciences, Palo Alto, CA). These cells, except EL-4 cells, were trypsinized, and all cells were washed in PBS before use.

Recombinant Poxvirus Vaccines. The rV and rF viruses containing the murine B7-1, ICAM-1, and LFA-3 genes (rV-TRICOM and rF-TRICOM, respectively) or in combination with the human CEA gene (rV-CEA/TRICOM and rF-CEA/TRICOM, respectively) have been described (1, 2). The rF virus containing the gene for murine GM-CSF (rF-GM-CSF) has been described (16). Therion Biologics Corp. (Cambridge, MA) kindly provided all orthopox viruses.

Therapy of CEA+ Tumors with CEA/TRICOM Subcutaneous/Intratumoral Vaccine Regimen. CEA-transgenic mice were transplanted s.c. with MC38-CEA+ tumor cells (3 × 105 per mouse) into the right flank. In protocols for the early-phase therapy of tumors, mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and then i.t. with rF-CEA/TRICOM on days 11, 18, and 25. In protocols for the late-phase therapy of tumors, mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then i.t. with rF-CEA/TRICOM on days 15, 22, and 29. In indicated experiments, mice received s.c. vaccination in place of i.t. boost vaccination. Each virus was injected at 108 plaque-forming units per mouse admixed with 107 plaque-forming units per mouse of rF-GM-CSF. Tumor size was measured using calipers once to twice a week.

Tumor Rechallenge of CEA-Transgenic Mice Cured of CEA+ Tumors by the CEA/TRICOM Subcutaneous/Intratumoral Vaccine Regimen. CEA-transgenic mice cured by the CEA/TRICOM s.c./i.t. vaccine regimen were rechallenged with MC38-CEA+ tumors on day 60 after the first tumor transplantation. Mice rejecting the second MC38-CEA+ tumor challenge for 30 days were challenged with the parental CEA tumors (MC38). Mice rejecting the MC38 tumor challenge for 30 days were challenged with B16 melanoma cells. Tumor size was measured using calipers once to twice a week. In indicated experiments, cured mice were challenged with Mtag tumor cells 3 months after MC38-CEA+ tumor implantation.

Lymphocyte Proliferation Assay. To evaluate CD4+ T-cell responses specific for tumor antigens, splenic T cells were tested for proliferation in response to CEA protein as described previously (17).

Cytotoxicity Assay. To evaluate CD8+ T-cell responses specific for tumor antigens, spleens were pooled and dispersed into single-cell suspensions and then stimulated with the H-2Db-restricted peptide CEA526-533 (10 μg/mL, EAQNTTYL; refs. 13, 18), H-2Db-restricted peptide p53232-240 (2 μg/mL, KYMCNSSCM; refs. 19, 20), or H-2Kb-restricted peptide p15E604-611 (1 μg/mL, KSPWFTTL, the gp70 peptide; ref. 21). Six days later, bulk lymphocytes were separated by centrifugation through a Ficoll-Hypaque gradient.

Using these recovered lymphocytes, tumor-killing activity was tested as described previously (17). Briefly, the recovered lymphocytes, 51Cr-labeled target cells (EL-4, 5 × 103 per well), and each peptide were incubated for 5 hours (96-well U-bottomed plates), and radioactivity in supernatants was measured using a gamma counter (Corba Autogamma, Packard Instruments, Downers Grove, IL). As control peptides, VSV-N52-59 (RGYVYQGL; ref. 22) was used for H-2Db-restricted peptides, and ovalbumin257-264 (SIINFEKL; ref. 23) was used for H-2Kb-restricted peptides. In some experiments, MC38-CEA+ or Mtag cells were used as a target without peptide. The percentage of tumor lysis was calculated as follows: % tumor lysis = [(experimental counts per minute − spontaneous counts per minute) / (maximum counts per minute − spontaneous counts per minute)] × 100. Nonspecific 51Cr release in response to each control peptide was subtracted from that induced by the appropriate tumor antigen peptide. Data were averaged and graphed as Δ% ± SD. Where mentioned in text, CTL activity was converted to lytic units as described previously (24).

Cytokine Production Assay. The recovered lymphocytes separated as described above were also tested for cytokine production from CD8+ T cells in response to defined tumor antigen peptides. Recovered lymphocytes (5 × 105 per well) were restimulated with fresh irradiated naive splenocytes as antigen-presenting cells (5 × 106 per well) and with each peptide (10 μg/mL CEA peptide, 2 μg/mL p53 peptide, or 1 μg/mL gp70 peptide). Twenty-four hours later, the supernatant fluid was collected and analyzed for murine cytokines using the Cytometric Bead Array kit (BD PharMingen, San Diego, CA). Nonspecific cytokine production in response to each control peptide was subtracted from that induced by the appropriate tumor antigen peptide.

Flow Cytometric Analysis. To analyze T-cell populations in mice treated with the CEA/TRICOM s.c./i.t. vaccination regimen, tumors were harvested and mechanically dispersed into single-cell suspensions. Peripheral blood was collected into 4% citrated PBS and removed RBC. The following antibodies were purchased from BD PharMingen and used for analysis: FITC-conjugated anti-CD3e (hamster IgG1), CyChrome-conjugated anti-CD8 monoclonal antibody (mAb; rat IgG2a), and appropriate isotype control antibodies. To evaluate the generation of CEA-specific CTLs in mice after vaccination, cells were stained with phytoerythrin-conjugated CEA526-533/H-2Db-tetramer (CEA-tetramer, Beckman Coulter, Fullerton, CA), FITC-conjugated anti-CD3e, and CyChrome-conjugated anti-CD8 mAb. To evaluate the generation of gp70-specific CTLs in mice after vaccination, cells were stained with phytoerythrin-conjugated p15E604-611/H-2Kb-tetramer (the gp70-tetramer, NIH Tetramer Facility, National Institute of Allergy and Infectious Diseases, Bethesda, MD), FITC-conjugated anti-CD3e, and CyChrome-conjugated anti-CD8 mAb. Immunofluorescence staining was done after Fc receptor blocking with anti-CD16/CD32 mAb. The immunofluorescence was compared with the appropriate isotype-matched controls and analyzed with CellQuest software using a FACSCalibur cytometer (Becton Dickinson, Mountain View, CA).

Statistical Analysis. Significant differences were evaluated using ANOVA with repeated measures using Statview 4.1 (Abacus Concepts, Inc., Berkeley, CA). For graphical representation of data, y-axis error bars indicate the SD of the data for each point on the graph.

Therapy of CEA+ Tumors with Subcutaneous Priming Vaccination following Intratumoral Boosting Vaccinations. CEA-transgenic mice were transplanted with MC38-CEA+ tumors on day 0. All mice showed progressive tumor growth, and at day 4, the mean tumor size was 20.3 ± 5.5 mm3. As a control, mice were vaccinated with PBS vehicle s.c. on day 4 and then vaccinated with PBS i.t. on days 11, 18, and 25 (Fig. 1A). When mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and then boosted s.c. with rF-CEA/TRICOM on days 11, 18, and 25, tumor growth was suppressed somewhat significantly compared with that in the control group (Fig. 1B; P = 0.003), but no tumor regression was noted. In contrast, when mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and boosted i.t. with rF-CEA/TRICOM on days 11, 18, and 25, tumor growth was markedly and significantly inhibited (P = 0.0001 compared with mice primed with rV-CEA/TRICOM s.c. and boosted with rF-CEA/TRICOM s.c.; Fig. 1B). Finally, 8 of 10 mice in this group showed complete regression of tumors (Fig. 1C).

Fig. 1

Efficacy of rV-CEA/TRICOM s.c. priming followed by rF-CEA/TRICOM i.t. boosting. CEA-transgenic mice were transplanted s.c. with MC38-CEA+ tumors on day 0. Vaccine therapy was initiated on day 4 after tumor transplant. A, control mice were vaccinated with PBS vehicle s.c. on day 4 and i.t. on days 11, 18, and 25. B, mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and then boosted s.c. with rF-CEA/TRICOM on days 11, 18, and 25. C, mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and then boosted i.t. with rF-CEA/TRICOM on days 11, 18, and 25. D-F, therapy of CEA tumors. CEA-transgenic mice were transplanted s.c. with parental MC38 tumors on day 0. D, control mice were vaccinated with PBS vehicle s.c. on day 4 and i.t. on days 11, 18, and 25. E, mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and then boosted s.c. with rF-CEA/TRICOM on days 11, 18, and 25. F, mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and then boosted i.t. with rF-CEA/TRICOM on days 11, 18, and 25. G-I, therapy of CEA+ tumors. Vaccine therapy was withheld until day 8 after tumor transplant. G, control mice were vaccinated with PBS vehicle s.c. on day 8 and i.t. on days 15, 22, and 29. H, mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted s.c. with rF-CEA/TRICOM on days 15, 22, and 29. I, mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted i.t. with rF-CEA/TRICOM on days 15, 22, and 29. A-C and G-I, Ps on day 28 compared with the PBS control group. D-F, Ps on day 25 compared with the PBS control group. All recombinant viruses were admixed with rF-GM-CSF. Tumor volume was monitored twice a week. Compilation of four separate experiments. Each experiment was repeated three to five times.

Fig. 1

Efficacy of rV-CEA/TRICOM s.c. priming followed by rF-CEA/TRICOM i.t. boosting. CEA-transgenic mice were transplanted s.c. with MC38-CEA+ tumors on day 0. Vaccine therapy was initiated on day 4 after tumor transplant. A, control mice were vaccinated with PBS vehicle s.c. on day 4 and i.t. on days 11, 18, and 25. B, mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and then boosted s.c. with rF-CEA/TRICOM on days 11, 18, and 25. C, mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and then boosted i.t. with rF-CEA/TRICOM on days 11, 18, and 25. D-F, therapy of CEA tumors. CEA-transgenic mice were transplanted s.c. with parental MC38 tumors on day 0. D, control mice were vaccinated with PBS vehicle s.c. on day 4 and i.t. on days 11, 18, and 25. E, mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and then boosted s.c. with rF-CEA/TRICOM on days 11, 18, and 25. F, mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and then boosted i.t. with rF-CEA/TRICOM on days 11, 18, and 25. G-I, therapy of CEA+ tumors. Vaccine therapy was withheld until day 8 after tumor transplant. G, control mice were vaccinated with PBS vehicle s.c. on day 8 and i.t. on days 15, 22, and 29. H, mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted s.c. with rF-CEA/TRICOM on days 15, 22, and 29. I, mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted i.t. with rF-CEA/TRICOM on days 15, 22, and 29. A-C and G-I, Ps on day 28 compared with the PBS control group. D-F, Ps on day 25 compared with the PBS control group. All recombinant viruses were admixed with rF-GM-CSF. Tumor volume was monitored twice a week. Compilation of four separate experiments. Each experiment was repeated three to five times.

Close modal

To determine if there was any antigen specificity to the antitumor responses noted above, CEA-transgenic mice were transplanted with parental (CEA) MC38 tumors. Control mice were vaccinated with PBS vehicle s.c. on day 4 and i.t. on days 11, 18, and 25 (Fig. 1D). When mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and then boosted s.c. with rF-CEA/TRICOM on days 11, 18, and 25, tumor growth was not significantly suppressed (Fig. 1E). Similarly, when mice were vaccinated s.c. with rV-CEA/TRICOM on day 4 and then boosted i.t. with rF-CEA/TRICOM on days 11, 18, and 25, tumor growth was not significantly suppressed (Fig. 1F; P = 0.27 compared with mice primed s.c. with rV-CEA/TRICOM and boosted s.c. with rF-CEA/TRICOM; Fig. 1E). These results thus show that the CEA/TRICOM s.c. prime/i.t. boost regimen required CEA in the tumor for induction of antitumor activity.

Next, we attempted to treat larger tumors with the s.c. prime and i.t. boost vaccine regimen by withholding vaccine therapy until day 8 after tumor transplant. When therapy was initiated at this time, the mean tumor size was 51.2 ± 16.7 mm3 (Fig. 1G-I). Control mice were vaccinated with PBS s.c. on day 8 and i.t. on days 15, 22, and 29 (Fig. 1G). When mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted s.c. with rF-CEA/TRICOM on days 15, 22, and 29, no significant antitumor effects were noted compared with that of the control group (Fig. 1H; P = 0.08). When mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted i.t. with rF-CEA/TRICOM on days 15, 22, and 29, tumor growth was again significantly inhibited (Fig. 1I; P = 0.0001 compared with mice primed with rV-CEA/TRICOM s.c. and boosted with rF-CEA/TRICOM s.c.; Fig. 1H). Finally, 5 of 10 mice in this group showed complete regression of tumors (Fig. 1I). These results, taken together, show that s.c. priming followed by i.t. boosting is superior to s.c. priming followed by s.c. boosting.

CEA-Transgenic Mice Cured of CEA+ Tumors Reject Challenge with Diverse Tumor Types. To examine the extent of immunologic memory generated in mice cured by CEA/TRICOM s.c. prime/i.t. boost vaccine therapy, eight CEA-transgenic mice cured after the therapy started on day 4 (Fig. 1C) and five mice cured after the therapy started on day 8 (Fig. 1I) were rechallenged with MC38-CEA+ tumors (at day 60 after the first tumor transplantation). Nonvaccinated CEA-transgenic mice were transplanted with MC38-CEA+ tumors at the same time as a control, and all mice developed progressive tumors. As shown in Table 1, 100% of mice in the vaccinated group rejected the rechallenged MC38-CEA+ tumors. Next, to further examine the specificity of the antitumor immunity to the CEA+ tumors, these mice were challenged with parental (CEA) MC38 tumors 30 days after the rechallenge of MC38-CEA+ tumors. The MC38 tumors were rejected in 92% (12 of 13) of mice cured by CEA/TRICOM s.c. prime/i.t. boost vaccine therapy (Table 1). To determine if the antitumor immunity could translate to another tumor type, mice were challenged with syngeneic B16 melanoma cells 30 days after the challenge of parental MC38 tumors. None of these mice rejected the B16 tumors. However, tumor growth was significantly suppressed compared with that in the control mice (P = 0.0001; Table 1). The antitumor effects seen with the B16 melanoma and the CEA tumors led us to believe that immune responses to other antigens expressed on MC38 tumors and potentially other tumor types, such as B16 melanoma, were induced by the CEA/TRICOM s.c. prime/i.t. boost vaccine regimen. It has been reported previously that MC38 tumors overexpress wild-type p53 (19) and an endogenous retroviral env epitope gp70 (21). Reverse transcription-PCR analysis of MC38 parental cells, MC38-CEA+ cells, and B16 tumor cells showed that these cells expressed substantial quantities of mRNA encoding p53 and gp70 (data not shown).

Table 1

Long-term protection to diverse tumor challenges in mice cured by s.c./i.t. vaccination with CEA/TRICOM

GroupNo. cured mice% Tumor-free mice after challenge with
Second MC38-CEA+*Parental MC38B16 tumors
Vaccination 13/20§ 100 (13/13) 92 (12/13) 0 (0/12) [341 ± 244] 
Control 0/25 0 (0/5) 0 (0/5) 0 (0/5) [1707 ± 607] 
GroupNo. cured mice% Tumor-free mice after challenge with
Second MC38-CEA+*Parental MC38B16 tumors
Vaccination 13/20§ 100 (13/13) 92 (12/13) 0 (0/12) [341 ± 244] 
Control 0/25 0 (0/5) 0 (0/5) 0 (0/5) [1707 ± 607] 
*

Mice cured by vaccine regimen were rechallenged with MC38-CEA+ tumor cells on day 60 after first MC38-CEA+ tumor implantation.

Mice that rejected the secondary MC38-CEA+ tumors were challenged with parental (CEA) MC38 tumor cells on day 90 after first MC38-CEA+ tumor implantation.

Mice that rejected the MC38 tumors were challenged with syngeneic B16 melanoma cells on day 120 after the first MC38-CEA+ tumor implantation. Data indicate mean tumor volume ± SD on day 18 after challenge.

§

Ten CEA-transgenic mice were vaccinated with rV-CEA/TRICOM (s.c.) on day 4 and then boosted with rF-CEA/TRICOM (i.t.) on days 11, 18, and 25 after tumor implantation. Another 10 CEA-transgenic mice were vaccinated with rV-CEA/TRICOM (s.c.) on day 8 and then boosted with rF-CEA/TRICOM (i.t.) on days 15, 22, and 29 after tumor implantation. rF-GM-CSF was admixed with vaccines.

P = 0.0001 compared with tumor volume in the control.

Control mice were treated s.c. or i.t. with PBS according to the same schedule of vaccination.

Analysis of Immune Responses Induced by CEA/TRICOM Subcutaneous Prime/Intratumoral Boost Vaccination Therapy. To determine possible therapeutic mechanisms associated with the CEA/TRICOM s.c. prime/i.t. boost vaccine regimen, immunologic studies were conducted using MC38-CEA+ tumor-transplanted CEA-transgenic mice. This study consisted of four groups. As a control, group 1 was vaccinated with PBS vehicle s.c. on day 8 and i.t. on days 15 and 22 after tumor transplantation (Fig. 2A). Mice in group 2 were vaccinated with rV-CEA/TRICOM s.c. on day 8 and boosted s.c. with rF-CEA/TRICOM on days 15 and 22 (Fig. 2B). Again, no significant antitumor activity was observed in this group. When mice were vaccinated with rV-CEA/TRICOM s.c. on day 8 and boosted i.t. with rF-CEA/TRICOM on days 15 and 22 (Fig. 2C), tumor growth was significantly inhibited compared with that in the control (Fig. 2C; P = 0.0001); 50% (5 of 10) of the mice had regressing tumors, and the remaining mice (5 of 10) had growing tumors. To compare the immune responses between mice failing the CEA/TRICOM s.c./i.t. vaccine regimen and mice responding to the regimen, these mice were separated into two groups based on tumor volume (<200 versus >200 mm3 on day 28) and designated vaccine responder (Fig. 2E) and vaccine nonresponder (Fig. 2D), respectively. These mice were sacrificed on day 29 and cells were used for in vitro immune assays.

Fig. 2

Analysis of therapeutic mechanisms induced by CEA/TRICOM s.c. prime/i.t. boost vaccination therapy. CEA-transgenic mice were transplanted s.c. with MC38-CEA+ tumors on day 0 (n = 10). A, control mice were vaccinated with PBS vehicle s.c. on day 8 and i.t. on days 15 and 22. B, mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted s.c. with rF-CEA/TRICOM on days 15 and 22. C, mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted i.t. with rF-CEA/TRICOM on days 15 and 22. Each virus was admixed with rF-GM-CSF. Ps on day 28 compared with the PBS control group. Mice in C were separated into two groups based on the tumor volume at day 28 and were used for subsequent immunologic analyses (Figs. 3 and 4) 29 days after tumor transplantation. Representative of three independent experiments.

Fig. 2

Analysis of therapeutic mechanisms induced by CEA/TRICOM s.c. prime/i.t. boost vaccination therapy. CEA-transgenic mice were transplanted s.c. with MC38-CEA+ tumors on day 0 (n = 10). A, control mice were vaccinated with PBS vehicle s.c. on day 8 and i.t. on days 15 and 22. B, mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted s.c. with rF-CEA/TRICOM on days 15 and 22. C, mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted i.t. with rF-CEA/TRICOM on days 15 and 22. Each virus was admixed with rF-GM-CSF. Ps on day 28 compared with the PBS control group. Mice in C were separated into two groups based on the tumor volume at day 28 and were used for subsequent immunologic analyses (Figs. 3 and 4) 29 days after tumor transplantation. Representative of three independent experiments.

Close modal

To examine CEA-specific CD8+ T-cell responses, we tested splenic T cells from mice that had received the CEA/TRICOM s.c./i.t. vaccine regimen for both CTL activity and cytokine production. Figure 3A depicts CTL activity of CD8+ T cells in response to CEA peptide (tumor target = EL-4 cells). CD8+ T cells from PBS-vaccinated CEA-transgenic mice showed low CEA-specific CTL activity (○). This activity was significantly enhanced in mice showing antitumor activity following s.c./i.t. vaccination with CEA/TRICOM (vaccine responders, ▪; P = 0.013 at 0.63:1 E:T ratio compared with control mice) as well as in mice failing to show antitumor activity to this vaccine regimen (vaccine nonresponders, ▴; P = 0.012 compared with control mice). When the data were converted to lytic units: LU25 was <2.5 in the PBS control (○), 5.0 in vaccine nonresponders (▴), and 25.0 in vaccine responders (▪). Figure 3D depicts IFN-γ production from CD8+ T cells in response to CEA peptide. CD8+ T cells from vaccine responders produced the greatest levels of IFN-γ (>3-fold greater than that noted from vaccine nonresponders). As shown in Fig. 3G, tumor necrosis factor-α production was also enhanced in the vaccine responders compared with that of vaccine nonresponders. There were no detectable levels of interleukin-2, interleukin-4, and interleukin-5 from CD8+ T cells in both vaccine responders and nonresponders. These results show that CEA-specific T-cell immune responses can be induced in a tumor-bearing self-antigen system and that the CD8+ T-cell responses specific for CEA show increased activity from mice showing antitumor activity in response to vaccine as opposed to mice failing the vaccine regimen. Next, CD4+ T cells were examined for proliferation in response to CEA protein (50 μg/mL). CD4+ T cells from PBS-vaccinated control mice showed baseline proliferation in response to CEA protein. These responses were enhanced in mice showing antitumor activity following s.c./i.t. vaccination with CEA/TRICOM (P = 0.0062 compared with control mice). The increased CEA-specific proliferation was also noted from mice failing to show antitumor activity to this vaccine regimen (P = 0.0015 compared with control mice).

Fig. 3

Induction of CD8+ T-cell responses to CEA, p53, and gp70 after the CEA/TRICOM s.c. prime/i.t. boost vaccine regimen. Splenic lymphocytes from CEA-transgenic mice depicted in Fig. 2A , D, and E were used 29 days after tumor transplantation (n = 5). A, CEA-specific CTL activity. B, p53-specific CTL activity. C, gp70-specific CTL activity. Tumor lysis was measured by 51Cr release in supernatants. Control mice treated with PBS (○), nonresponders to CEA/TRICOM s.c./i.t. vaccine therapy (▴), and responders to CEA/TRICOM s.c./i.t. vaccine therapy (▪). D-F, antigen-specific IFN-γ production from CD8+ T cells. G-I, antigen-specific tumor necrosis factor-α production from CD8+ T cells. Representative of three independent experiments.

Fig. 3

Induction of CD8+ T-cell responses to CEA, p53, and gp70 after the CEA/TRICOM s.c. prime/i.t. boost vaccine regimen. Splenic lymphocytes from CEA-transgenic mice depicted in Fig. 2A , D, and E were used 29 days after tumor transplantation (n = 5). A, CEA-specific CTL activity. B, p53-specific CTL activity. C, gp70-specific CTL activity. Tumor lysis was measured by 51Cr release in supernatants. Control mice treated with PBS (○), nonresponders to CEA/TRICOM s.c./i.t. vaccine therapy (▴), and responders to CEA/TRICOM s.c./i.t. vaccine therapy (▪). D-F, antigen-specific IFN-γ production from CD8+ T cells. G-I, antigen-specific tumor necrosis factor-α production from CD8+ T cells. Representative of three independent experiments.

Close modal

Induction of an Antigenic Cascade in Mice after CEA/TRICOM Subcutaneous Prime/Intratumoral Boost Vaccine Therapy. It has been reported that MC38-CEA+ tumors overexpress wild-type p53 (19) and an endogenous retroviral env epitope gp70 (21). To determine whether immune responses specific for these additional tumor antigens were induced in vaccinated mice, MHC class I–restricted p53 and gp70 peptides were used to evaluate CD8+ T-cell responses. We first tested CD8+ T-cell responses by CTL activity and cytokine production. Figure 3B depicts CTL activity of specific CD8+ T cells in response to p53 peptide (tumor target = EL-4 cells). CTLs specific for p53 were not detected from PBS-vaccinated control mice (○). When mice were vaccinated with the CEA/TRICOM s.c./i.t. regimen, this CTL activity was enhanced albeit weakly [P = 0.0005 versus vaccine nonresponders at 0.63:1 E:T ratio (▴) and P = 0.0092 versus vaccine responders at 0.63:1 E:T ratio (▪)]. IFN-γ production from CD8+ T cells in response to p53 peptide is shown in Fig. 3E. CD8+ T cells from vaccine responders produced IFN-γ in response to p53 at 5-fold greater levels than that of vaccine nonresponders. As shown in Fig. 3H, tumor necrosis factor-α production was low in both vaccine responders and nonresponders.

In contrast to p53-specific responses, strong CD8+ T-cell responses specific for gp70 were seen in mice receiving the CEA/TRICOM s.c./i.t. vaccine regimen. Figure 3C depicts CTL activity of CD8+ T cells in response to gp70 peptide (tumor target = EL-4 cells). CD8+ T cells from the PBS control mice showed weak responses (○). This activity was significantly enhanced by vaccination with CEA/TRICOM vectors [P =0.002 versus vaccine nonresponders at 0.63:1 E:T ratio (▴) and P = 0.01 versus vaccine responders at 0.63:1 E:T ratio (▪)]. The gp70-specific CTL activity was most significantly increased in the vaccine responder group (▪) compared with that of vaccine nonresponders (P = 0.026 at 0.625:1 E:T ratio). On conversion of the data to lytic units, the LU25 was <2.5 in the control (○), 14.3 in vaccine nonresponders (▴), and 33.3 in vaccine responders (▪).

IFN-γ production from CD8+ T cells in response to gp70 peptide is shown in Fig. 3F. All groups receiving CEA/TRICOM s.c./i.t vaccine therapy showed increased IFN-γ production compared with that seen in the control, but the increment was highest in vaccine responders. The amount was 2.7-fold greater than that in vaccine nonresponders. Moreover, as shown in Fig. 3I, tumor necrosis factor-α production in response to gp70 peptide was 3.5-fold greater in the vaccine responder group compared with that of vaccine nonresponders.

These results show that the CEA/TRICOM s.c./i.t. vaccine regimen could induce T-cell immune responses not only to CEA encoded in vectors but also to other antigens expressed on tumors but not encoded in vaccines. In addition, the gp70-specific CD8+ T-cell responses were much greater than that seen in response to other tumor antigens such as CEA or p53 (Fig. 3C , F, and I). These results suggest that the success of the vaccine regimen could be influenced by induction of T-cell immune responses to other predominant antigens expressed on tumors, such as gp70.

Induction of gp70-Specific CTL Activity in Mice Vaccinated with the CEA/TRICOM Subcutaneous/Intratumoral Vaccine Regimen. CD8+ T-cell responses to gp70, not encoded in vaccines but expressed on tumors, were much greater than that seen in response to other tumor antigens (Fig. 3C,, F, and I). To further examine the potential role of gp70-specific CD8+ T cells in the antitumor response, studies were conducted to examine the CTL activity against MC38-CEA+ tumor cells, which were implanted in vivo. We first tested CEA-specific CD8+ T cells obtained from mice vaccinated with the CEA/TRICOM s.c./i.t. regimen. These cells were stimulated with CEA peptide for 6 days. CEA-specific CD8+ T cells from mice vaccinated with the CEA/TRICOM s.c./i.t. regimen showed significantly enhanced EL-4 cell killing in response to CEA peptide compared with those from the PBS-vaccinated mice (Fig. 4A; P = 0.003). There were no significant differences in EL-4 killing between nonresponders and responders to the s.c./i.t. vaccine regimen (P = 0.4349). In contrast, the CTL activity directed against MC38-CEA+ tumors (without addition of CEA peptide) was significantly enhanced in responders than that in nonresponders (P = 0.0285), although CEA-specific CD8+ T cells from nonresponders significantly killed MC38-CEA+ tumors compared with those from the control mice (Fig. 4A; P = 0.0237). In addition, when CEA and gp70 Mtag mammary tumor cells were used as a tumor targets, the CEA-specific CD8+ T cells did not show CTL activity (Fig. 4A).

Fig. 4

Induction of gp70-specific CTL responses in responders to the CEA/TRICOM s.c. prime/i.t. boost vaccine regimen. CEA-transgenic mice were vaccinated according to the same protocol described in Fig. 2, and splenic lymphocytes were used 29 days after tumor transplantation (n = 4). A, CTL activity of CEA peptide-stimulated CTLs (CEA-CTLs). Splenic lymphocytes stimulated with CEA peptide for 6 days were used for 51Cr release assay (E:T ratio = 40:1). B, CTL activity of gp70 peptide-stimulated CTLs (gp70-CTLs). Splenic lymphocytes stimulated with gp70 peptide for 6 days were used for 51Cr release assay (E:T ratio = 40:1). C, CTL activity of gp70 peptide-induced CTLs from cured mice. CEA-transgenic mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted s.c. with rF-CEA/TRICOM on days 15, 22, and 29. Cured mice after the therapy were used 88 days after tumor implantation. Splenic lymphocytes stimulated with gp70 peptide for 6 days were used for 51Cr release assay (E:T ratio = 40:1). D, tumor challenge into cured mice. CEA-transgenic mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted s.c. with rF-CEA/TRICOM on days 15, 22, and 29. Cured mice after the therapy were challenged with tumor cells (3 × 105) 88 days after tumor implantation (n = 5, bold lines). As a control, age/sex-matched CEA-transgenic mice were implanted with same tumors (n = 3, thin lines). Tumor volume was monitored twice a week.

Fig. 4

Induction of gp70-specific CTL responses in responders to the CEA/TRICOM s.c. prime/i.t. boost vaccine regimen. CEA-transgenic mice were vaccinated according to the same protocol described in Fig. 2, and splenic lymphocytes were used 29 days after tumor transplantation (n = 4). A, CTL activity of CEA peptide-stimulated CTLs (CEA-CTLs). Splenic lymphocytes stimulated with CEA peptide for 6 days were used for 51Cr release assay (E:T ratio = 40:1). B, CTL activity of gp70 peptide-stimulated CTLs (gp70-CTLs). Splenic lymphocytes stimulated with gp70 peptide for 6 days were used for 51Cr release assay (E:T ratio = 40:1). C, CTL activity of gp70 peptide-induced CTLs from cured mice. CEA-transgenic mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted s.c. with rF-CEA/TRICOM on days 15, 22, and 29. Cured mice after the therapy were used 88 days after tumor implantation. Splenic lymphocytes stimulated with gp70 peptide for 6 days were used for 51Cr release assay (E:T ratio = 40:1). D, tumor challenge into cured mice. CEA-transgenic mice were vaccinated s.c. with rV-CEA/TRICOM on day 8 and then boosted s.c. with rF-CEA/TRICOM on days 15, 22, and 29. Cured mice after the therapy were challenged with tumor cells (3 × 105) 88 days after tumor implantation (n = 5, bold lines). As a control, age/sex-matched CEA-transgenic mice were implanted with same tumors (n = 3, thin lines). Tumor volume was monitored twice a week.

Close modal

We next tested gp70-specific CD8+ T cells obtained from mice vaccinated with the CEA/TRICOM s.c./i.t. regimen. These cells were stimulated with gp70 peptide for 6 days. Gp70-specific CD8+ T cells from vaccine responders showed greatest CTL activity (Fig. 4B; P = 0.0002 versus the control mice and P = 0.0005 versus vaccine nonresponders). In addition, these gp70-specific CD8+ T cells from vaccine responders strongly killed MC38-CEA+ tumors (Fig. 4B; P = 0.0003 versus the control mice and P = 0.0004 versus vaccine nonresponders). The CTL activity of gp70-specific CD8+ T cells was much higher than that of CEA-specific CD8+ T cells in vaccine responders. When Mtag cells (gp70) were used as a tumor target, CTL activity of gp70-specific CD8+ T cells was not observed (Fig. 4B). To confirm that gp70-specific CTL from vaccinated mice could specifically recognize gp70+ tumor cells, either CEA+ or gp70+ Mtag cells were used as a tumor target. There, cured mice were used 3 months after the CEA/TRICOM s.c./i.t. vaccine regimen. When MC38-CEA+ cells (CEA+, gp70+), Mtag-gp70+ cells (CEA-, gp70+), or parental Mtag cells (CEA-, gp70-) admixed with gp70 peptide were used as a target of gp70 peptide-induced CTLs, CTL activity was noted (Fig. 4C). However, Mtag-CEA+ cells (CEA+, gp70-) or parental Mtag cells were not killed. This result shows that gp70-specific CTLs could specifically kill gp70+ tumors and that CTLs specific for additional tumor antigens could be induced in mice receiving the CEA/TRICOM s.c./i.t. vaccine regimen.

We then examined the relevancy of gp70-specific T cells induced by the CEA/TRICOM s.c./i.t. vaccine regimen in vivo. When cured mice (vaccine responders) were challenged with MC38-CEA+ tumors (CEA+, gp70+), tumor growth was strongly suppressed and all mice (5 of 5) were tumor free 16 days after tumor challenge (Fig. 4D). When mice were challenged with Mtag-CEA+ tumors (CEA+, gp70-), however, only 60% of mice rejected tumor. In contrast, when mice were challenged with Mtag-gp70+ tumors (CEA-, gp70+), 100% of mice rejected tumor. Mice challenged with parental Mtag cells (CEA-, gp70-) failed to reject these tumors (Fig. 4D). These results suggest that both CEA-specific and gp70-specific antitumor immune responses were induced by the CEA/TRICOM s.c./i.t. vaccine therapy and could mediate antitumor activity. These results, taken together, suggest that CTL activity not only to CEA encoded in vaccine vectors but also to other tumor antigens not encoded in vaccines are responsible for successful vaccine therapy of established tumors and subsequent immune memory.

Detection of CD8+/Tetramer+ T Cells Specific for Tumor Antigens in Mice Vaccinated with the CEA/TRICOM Subcutaneous/Intratumoral Vaccine Regimen. To further analyze the CTL induction specific for tumor antigens in mice receiving the CEA/TRICOM s.c./i.t. vaccine therapy, we examined the level of tetramer-binding CD8+ T cells in the local tumor sites and systemic circulation. CEA-transgenic mice transplanted with MC38-CEA+ tumors were vaccinated s.c. with rV-CEA/TRICOM on day 8 after tumor transplant and boosted i.t. with rF-CEA/TRICOM on days 15 and 22. On day 29, cells were harvested from individual s.c. transplanted tumors and peripheral blood from vaccine responders and nonresponders and analyzed for CEA-tetramer or gp70-tetramer binding using flow cytometry (n = 4 per group). As seen in Fig. 5G, CD8+ T cells specific for CEA or gp70 were markedly increased in tumors of mice receiving the s.c./i.t. vaccine therapy compared with that seen in PBS-vaccinated mice. The increment of both CEA-specific and gp70-specific T cells was significantly higher in vaccine responders than that in nonresponders (P = 0.0011 in CEA-specific CTLs and P = 0.0004 in gp70-specific CTLs). The level of gp70-specific CD8+ T-cell induction was higher than that of CEA-specific CD8+ T-cell induction in local tumor sites. Peripheral blood cells were analyzed for CEA-tetramer or gp70-tetramer binding (Fig. 4G). Both CEA-specific and gp70-specific CD8+ T cells were significantly increased in peripheral blood of vaccine responders compared with that seen in the control mice and vaccine nonresponders (P = 0.0014) as well as local tumor sites. These results suggest that s.c./i.t. vaccination with CEA/TRICOM vectors could induce CD8+ T cells specific for additional TAAs in mice and that mice showing antitumor activity (vaccine responders) had significantly greater recruitment of antigen-specific CTLs (CEA, gp70) in tumors.

Fig. 5

Induction of tumor antigen-specific CD8+ T cells in mice receiving CEA/TRICOM s.c./i.t. vaccine therapy. CEA-transgenic mice were vaccinated according to the same protocol described in Fig. 2 and then used 29 days after tumor transplantation (n = 4). Tumors were harvested and mechanically dispersed into single-cell suspensions. Peripheral blood was collected into 4% citrated PBS and removed RBC. Cells were individually immunostained with anti-CD3 mAb, anti-CD8 mAb, and CEA-tetramer or gp70-tetramer. A-C, CEA-specific CTLs in tumor-infiltrating cells. A, control mice treated with PBS; B, nonresponders to CEA/TRICOM s.c./i.t. vaccine therapy; C, responders to CEA/TRICOM s.c./i.t. vaccine therapy; D-F, gp70-specific CTLs in tumor-infiltrating cells: (D) control mice treated with PBS, (E) nonresponders to CEA/TRICOM s.c./i.t. vaccine therapy, and (F) responders to CEA/TRICOM s.c./i.t. vaccine therapy. G, % Tetramer-binding CD3+CD8+ cells in tumor-infiltrating cells and peripheral blood. Ps comparing the % tetramer-binding CD3+CD8+ cells between nonresponders and responders to CEA/TRICOM s.c./i.t. vaccine therapy. Representative of three independent experiments.

Fig. 5

Induction of tumor antigen-specific CD8+ T cells in mice receiving CEA/TRICOM s.c./i.t. vaccine therapy. CEA-transgenic mice were vaccinated according to the same protocol described in Fig. 2 and then used 29 days after tumor transplantation (n = 4). Tumors were harvested and mechanically dispersed into single-cell suspensions. Peripheral blood was collected into 4% citrated PBS and removed RBC. Cells were individually immunostained with anti-CD3 mAb, anti-CD8 mAb, and CEA-tetramer or gp70-tetramer. A-C, CEA-specific CTLs in tumor-infiltrating cells. A, control mice treated with PBS; B, nonresponders to CEA/TRICOM s.c./i.t. vaccine therapy; C, responders to CEA/TRICOM s.c./i.t. vaccine therapy; D-F, gp70-specific CTLs in tumor-infiltrating cells: (D) control mice treated with PBS, (E) nonresponders to CEA/TRICOM s.c./i.t. vaccine therapy, and (F) responders to CEA/TRICOM s.c./i.t. vaccine therapy. G, % Tetramer-binding CD3+CD8+ cells in tumor-infiltrating cells and peripheral blood. Ps comparing the % tetramer-binding CD3+CD8+ cells between nonresponders and responders to CEA/TRICOM s.c./i.t. vaccine therapy. Representative of three independent experiments.

Close modal

Here, we examined if direct overexpression of both signal-1 (CEA) and signal-2 (three costimulatory molecules; TRICOM) in tumor cells could potentiate antitumor activity. This was tested by using a vaccine regimen composed of first priming mice s.c. with rV-CEA/TRICOM and then boosting i.t. with rF-CEA/TRICOM, each admixed with rF-GM-CSF. Previously, we showed that both CEA and TRICOM were essential as transgenes in the prime vaccination vector (8, 9, 11, 25). Here, it is shown that a s.c. prime and i.t. boost vaccination regimen with CEA/TRICOM vectors was effective in inhibiting the growth of CEA+ tumors (Fig. 1), and cured mice subsequently rejected a secondary challenge of MC38-CEA+ tumors (Table 1), showing long-term antitumor immunity. The s.c. prime/i.t. boost vaccination therapy with CEA/TRICOM vectors, however, was not effective in eliminating the growth of CEA MC38 tumors (Fig. 1F). This result further shows the specificity of this therapy for CEA+ tumors and highlights the requirement of CEA in the vaccine and in the tumors for optimal vaccine therapy. It should be pointed out that these studies used CEA-transgenic mice in which CEA is a self-antigen; these mice express CEA in normal gastrointestinal tissue and fetal tissue in a manner similar to that expressed in humans (12, 13). These CEA-transgenic mice also contain CEA protein in sera at levels similar to that found in patients with advanced colorectal carcinoma (5-100 ng/mL).

When splenic T cells were tested for CEA-specific immune responses after the CEA/TRICOM s.c./i.t vaccine regimen, stronger CEA-specific CD8+ T-cell responses were noted in mice responding to the vaccine regimen than in mice failing the vaccine regimen (Figs. 3D and G and 4B). These observations agree with studies showing that antitumor effects induced by CEA/TRICOM vaccination in a murine peripancreatic metastasis model of MC38-CEA+ tumors were completely abrogated by depletion of the CD8+ cell population (25).

CEA-transgenic mice cured of tumors by the CEA/TRICOM s.c./i.t. vaccine regimen rejected not only MC38-CEA+ tumors but also CEA parental MC38 tumors. In addition, these mice significantly suppressed the growth of B16 melanoma tumors (Table 1). These results suggested that the tumor therapy could potentially be accompanied by the induction of immune responses to other antigens expressed on these tumors. Splenic T-cell immune responses to p53 and gp70, which were expressed on MC38-CEA+ tumors but not encoded in vaccines, were observed in mice vaccinated with the CEA/TRICOM s.c./i.t regimen, and the CD8+ T-cell responses were most strongly induced in responders to the therapy (Figs. 3 and 4). Moreover, CD8+ T-cell responses specific for gp70 were much greater than those specific for CEA or p53. Strikingly, the gp70-specific T-cell response level was correlative to the therapeutic efficacy, showing the greatest antitumor responses in responders to the s.c. prime/i.t. boost vaccination therapy.

Many CTLs specific for CEA and gp70 were seen to be infiltrating the tumors of mice that had received the CEA/TRICOM s.c./i.t. therapy (Fig. 5). Of note, these CTLs predominantly comprised CD8+ T-cell population in tumors and peripheral blood of vaccine responders. The increment of tumor antigen-specific CTLs was correlated to the therapeutic efficacy (i.e., vaccine nonresponders showed CTL increase in tumors and peripheral blood but to a lower degree than that observed in the vaccine responder group). Interestingly, the concentration of gp70-specific CTLs was greater than that of CEA-specific CTLs in local tumor sites of vaccine responders. These results indicate that the s.c. prime/i.t. boost vaccination with CEA/TRICOM vectors was essential for the initiation of antitumor effects, but the subsequent induction of antigen-specific immune responses by cross-priming led to efficient expansion and recruitment of gp70-specific CTLs into tumors, where these cells could play a crucial role in antitumor responses.

Meazza et al. reported that mice could reject parental TS/A tumors and CT-26 tumors challenged after rejection of MHC II–transfected TS/A tumors but could not reject gp70 tumors (26). Rosato et al. showed that gp70-specific CTLs were generated in mice that rejected TS/A tumors transfected with costimulatory molecules or cytokines (27). These studies suggested that gp70 is an influential CTL-immunodominant epitope that is commonly overexpressed on mouse tumors. Our findings confirm and extend these observations.

The results reported here of an “antigen cascade” support and extend previous results by several groups. Recently, Pilon et al. reported that HER-2/neu+ tumor models cured after HER-2-encoding DNA vaccine therapy rejected HER-2/neu tumors. It was hypothesized that this could have resulted from the broadening of epitope recognition (28). Another study also suggested the epitope spreading induced by cross-priming, showing that CTLs specific for another epitope were seen in cured mice after immunotherapy using a P815-derived tumor peptide (29). Such a phenomenon has also been reported in clinical studies. Cavacini et al. reported that after vaccination therapy of cancer patients with PSA+ prostate tumors with rV/rF-PSA, sera were cross-reactive to multiple prostate tumor cell lines, including some that did not express prostate-specific antigen (30). Butterfield et al. observed that spreading of immune reactivity was induced only in melanoma patients with a complete response after i.t. vaccination therapy with MART1-pulsed dendritic cells (31). These studies support our hypothesis that the breadth and magnitude of antitumor immune cascades to multiple antigens expressed on tumors could be critical in the therapy of established tumors.

Taken together, these studies show that the TAA (in this case, CEA) must be present in both the vaccine and the tumor to initiate the antitumor response. Subsequent cross-priming of an antigen cascade could then be initiated/potentiated by additional antigens (in this case, p53 and gp70) expressed on the tumors. The potentiation of the antitumor response that led to complete tumor regression seemed to be related to an immune response generated to a predominant antigen (gp70) not encoded in the vaccine. These studies thus have important implications in the analysis of a given patient's immune response in vaccine clinical trials. To this point, the vast majority of studies have followed only the generation of T-cell responses directed against the TAA in the vaccine. The studies reported here provide further evidence that the magnitude and/or breadth of an antigen cascade (distinct from that antigen in vaccines) may determine the effectiveness of the antitumor therapy.

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.

We thank Marion Taylor for excellent technical assistance and Debra Weingarten for her editorial assistance.

1
Hodge JW, Sabzevari H, Yafal AG, et al. A triad of costimulatory molecules synergize to amplify T-cell activation.
Cancer Res
1999
;
59
:
5800
–7.
2
Morse MA. Technology evaluation: CEA-TRICOM, Therion Biologics Corp.
Curr Opin Mol Ther
2001
;
3
:
407
–12.
3
Peter I, Nawrath M, Kamarashev J, et al. Immunotherapy for murine K1735 melanoma: combinatorial use of recombinant adenovirus expressing CD40L and other immunomodulators.
Cancer Gene Ther
2002
;
9
:
597
–605.
4
Kim SH, Carew JF, Kooby DA, et al. Combination gene therapy using multiple immunomodulatory genes transferred by a defective infectious single-cycle herpes virus in squamous cell cancer.
Cancer Gene Ther
2000
;
7
:
1279
–85.
5
Colmenero P, Chen M, Castanos-Velez E, Liljestrom P, Jondal M. Immunotherapy with recombinant SFV-replicons expressing the P815A tumor antigen or IL-12 induces tumor regression.
Int J Cancer
2002
;
98
:
554
–60.
6
Heinzerling L, Dummer R, Pavlovic J, et al. Tumor regression of human and murine melanoma after intratumoral injection of IL-12-encoding plasmid DNA in mice.
Exp Dermatol
2002
;
11
:
232
–40.
7
Jourdier TM, Moste C, Bonnet MC, et al. Local immunotherapy of spontaneous feline fibrosarcomas using recombinant poxviruses expressing interleukin 2 (IL2).
Gene Ther
2003
;
10
:
2126
–32.
8
Kudo-Saito C, Schlom J, Hodge JW. Intratumoral vaccination and diversified subcutaneous/intratumoral vaccination with recombinant poxviruses encoding a tumor antigen and multiple costimulatory molecules.
Clin Cancer Res
2004
;
10
:
1090
–9.
9
Grosenbach DW, Barrientos JC, Schlom J, Hodge JW. Synergy of vaccine strategies to amplify antigen-specific immune responses and antitumor effects.
Cancer Res
2001
;
61
:
4497
–505.
10
Greiner JW, Zeytin H, Anver MR, Schlom J. Vaccine-based therapy directed against carcinoembryonic antigen demonstrates antitumor activity on spontaneous intestinal tumors in the absence of autoimmunity.
Cancer Res
2002
;
62
:
6944
–51.
11
Aarts WM, Schlom J, Hodge JW. Vector-based vaccine/cytokine combination therapy to enhance induction of immune responses to a self-antigen and antitumor activity.
Cancer Res
2002
;
62
:
5770
–7.
12
Eades-Perner AM, van der Putten H, Hirth A, et al. Mice transgenic for the human carcinoembryonic antigen gene maintain its spatiotemporal expression pattern.
Cancer Res
1994
;
54
:
4169
–76.
13
Kass E, Schlom J, Thompson J, et al. Induction of protective host immunity to carcinoembryonic antigen (CEA), a self-antigen in CEA transgenic mice, by immunizing with a recombinant vaccinia-CEA virus.
Cancer Res
1999
;
59
:
676
–83.
14
Robbins PF, Kantor JA, Salgaller M, et al. Transduction and expression of the human carcinoembryonic antigen gene in a murine colon carcinoma cell line.
Cancer Res
1991
;
51
:
3657
–62.
15
Liu K, Abrams SI. Alterations in Fas expression are characteristic of, but not solely responsible for, enhanced metastatic competence.
J Immunol
2003
;
170
:
5973
–80.
16
Kass E, Panicali DL, Mazzara G, Schlom J, Greiner JW. Granulocyte/macrophage-colony stimulating factor produced by recombinant avian poxviruses enriches the regional lymph nodes with antigen-presenting cells and acts as an immunoadjuvant.
Cancer Res
2001
;
61
:
206
–14.
17
Kantor J, Irvine K, Abrams S, et al. Antitumor activity and immune responses induced by a recombinant carcinoembryonic antigen-vaccinia virus vaccine.
J Natl Cancer Inst
1992
;
84
:
1084
–91.
18
Kalus RM, Kantor JA, Gritz L, et al. The use of combination vaccinia vaccines and dual-gene vaccinia vaccines to enhance antigen-specific T-cell immunity via T-cell costimulation.
Vaccine
1999
;
17
:
893
–903.
19
Hilburger Ryan M, Abrams SI. Characterization of CD8+ cytotoxic T lymphocyte/tumor cell interactions reflecting recognition of an endogenously expressed murine wild-type p53 determinant.
Cancer Immunol Immunother
2001
;
49
:
603
–12.
20
Lacabanne V, Viguier M, Guillet JG, Choppin J. A wild-type p53 cytotoxic T cell epitope is presented by mouse hepatocarcinoma cells.
Eur J Immunol
1996
;
26
:
2635
–9.
21
Yang JC, Perry-Lalley D. The envelope protein of an endogenous murine retrovirus is a tumor-associated T-cell antigen for multiple murine tumors.
J Immunother
2000
;
23
:
177
–83.
22
Esquivel F, Yewdell J, Bennink J. RMA/S cells present endogenously synthesized cytosolic proteins to class I-restricted cytotoxic T lymphocytes.
J Exp Med
1992
;
175
:
163
–8.
23
el-Shami K, Tirosh B, Bar-Haim E, et al. MHC class I-restricted epitope spreading in the context of tumor rejection following vaccination with a single immunodominant CTL epitope.
Eur J Immunol
1999
;
29
:
3295
–301.
24
Hodge JW, Rad AN, Grosenbach DW, et al. Enhanced activation of T cells by dendritic cells engineered to hyperexpress a triad of costimulatory molecules.
J Natl Cancer Inst
2000
;
92
:
1228
–39.
25
Hodge JW, Grosenbach DW, Aarts WM, Poole DJ, Schlom J. Vaccine therapy of established tumors in the absence of autoimmunity.
Clin Cancer Res
2003
;
9
:
1837
–49.
26
Meazza R, Comes A, Orengo AM, Ferrini S, Accolla RS. Tumor rejection by gene transfer of the MHC class II transactivator in murine mammary adenocarcinoma cells.
Eur J Immunol
2003
;
33
:
1183
–92.
27
Rosato A, Santa SD, Zoso A, et al. The cytotoxic T-lymphocyte response against a poorly immunogenic mammary adenocarcinoma is focused on a single immunodominant class I epitope derived from the gp70 Env product of an endogenous retrovirus.
Cancer Res
2003
;
63
:
2158
–63.
28
Pilon SA, Kelly C, Wei WZ. Broadening of epitope recognition during immune rejection of ErbB-2-positive tumor prevents growth of ErbB-2-negative tumor.
J Immunol
2003
;
170
:
1202
–8.
29
Markiewicz MA, Fallarino F, Ashikari A, Gajewski TF. Epitope spreading upon P815 tumor rejection triggered by vaccination with the single class I MHC-restricted peptide P1A.
Int Immunol
2001
;
13
:
625
–32.
30
Cavacini LA, Duval M, Eder JP, Posner MR. Evidence of determinant spreading in the antibody responses to prostate cell surface antigens in patients immunized with prostate-specific antigen.
Clin Cancer Res
2002
;
8
:
368
–73.
31
Butterfield LH, Ribas A, Dissette VB, et al. Determinant spreading associated with clinical response in dendritic cell-based immunotherapy for malignant melanoma.
Clin Cancer Res
2003
;
9
:
998
–1008.