IFN-γ Can Promote Tumor Evasion of the Immune System In Vivo by Down-Regulating Cellular Levels of an Endogenous Tumor Antigen

Gregory L. Beatty and Yvonne Paterson
J Immunol November 15, 2000, 165 (10) 5502-5508; DOI: https://doi.org/10.4049/jimmunol.165.10.5502

Abstract

Although IFN-γ has been generally thought to enhance antitumor immune responses, we found that IFN-γ can promote tumor escape in the CT26 colon carcinoma by down-regulating the protein expression of an endogenous tumor Ag. gp70, the env product of the endogenous ecotropic murine leukemia virus, has been reported to be the immunodominant Ag of CT26. We show that IFN-γ down-regulates intracellular and surface levels of gp70 protein resulting in a reduced lysis by CTL, which is restored by pulsing IFN-γ-treated CT26 with the Ld-restricted immunodominant AH1 epitope derived from gp70. To investigate the role of CT26 sensitivity to IFN-γ in vivo, we constructed two variants of CT26, CT26.mugR and CT26.IFN, that are unresponsive to IFN-γ or express IFN-γ, respectively. We demonstrate using these variants that tumor responsiveness to IFN-γ promotes a reduction in tumor immunogenicity in vivo that is correlated with an increased tumor incidence in immune mice. Analysis of the tumors from mice challenged with CT26 or CT26.mugR revealed infiltration of CD8 T cells secreting IFN-γ. We conclude that IFN-γ secreted by tumor-infiltrating T cells promotes tumor escape through the down-regulation of the endogenous tumor Ag gp70. These findings have impact on the design of effective antitumor vaccine strategies.

Tumors have evolved many mechanisms to evade recognition by the immune system. Tumor evasion of an immune response may involve down-regulation of MHC class I molecules or components of the Ag presentation process (1, 2, 3), secretion of immunosuppressive molecules (e.g., IL-10, TGF-β, or PGE2) (4, 5), up-regulation of antiapoptotic molecules (e.g., Bcl-2 or Bcl-xL) (6, 7, 8), or expression of Fas ligand (9, 10). Thus, a challenge for immunotherapeutic approaches to cancer is the generation of an immune response effective in countering these strategies of immune evasion.

IFN-γ has been recognized as a critical cytokine involved in effective antitumor immune responses. Depletion of IFN-γ in several tumor models eliminates the effectiveness of a particular therapeutic strategy (11, 12). The effectiveness of IFN-γ in these therapies has been attributed to its ability to up-regulate MHC class I expression and other Ag presentation components (1), induce chemokines that inhibit angiogenesis (e.g., monokine induced by IFN-γ and IFN-inducible protein 10) (13, 14), and down-regulate the expression of immunosuppressive molecules secreted by tumors (15, 16). Recently, several investigators have identified the importance of tumor cell responsiveness to IFN-γ for the effectiveness of tumor-based vaccine strategies and IL-12 therapy. Tumor cell unresponsiveness to IFN-γ reduces the immunogenicity of such tumor cells in vivo (17). In addition, tumor cell responsiveness to IFN-γ is critical for the effectiveness of IL-12 to inhibit angiogenesis within the tumor (18).

We chose to analyze the role of tumor cell responsiveness to IFN-γ in the murine CT26 colon carcinoma model where the immunodominant tumor Ag is reported to be gp70, an envelope glycoprotein of an endogenous ecotropic murine leukemia virus (19). Surprisingly, we found that IFN-γ promotes tumorigenicity despite its ability to up-regulate MHC class I expression. We found that this increase in tumorigenicity correlated with a down-regulation of gp70 protein levels promoted by IFN-γ. We demonstrate here the first report where IFN-γ secreted by tumor-infiltrating T cells promotes tumor escape through the down-regulation of an immunodominant endogenous tumor Ag.

Materials and Methods

Mice and cell lines

Female BALB/c mice, 6–8 wk of age, were purchased from Charles River Laboratories (Wilmington, MA). CT26 is an N-nitroso-N-methylurethane-induced colon carcinoma (20) that is syngeneic to BALB/c. EL4 is a carcinogen-induced lymphoma syngeneic with C57BL/6. Both cell lines were cultured in RPMI 1640 containing 10% FCS and penicillin/streptomycin. In addition, 0.05 mM 2-ME was added to EL4 cultures.

Transfections and retroviral infections

The plasmids pEF2.mugR and pL(mIFNγ-KDEL)SN were contributed by William Lee (University of Pennsylvania, Philadelphia, PA). The expression plasmid pEF2.mugR contains a truncated murine IFN-γRα cDNA under the control of the eukaryotic translation elongation factor 1α promoter (18, 21). To create CT26 cells expressing this mutant IFN-γRα, parental cell lines were transfected with pEF2.mugR using Lipofectin reagent (Life Technologies, Grand Island, NY). G418-resistant cells were sorted by flow cytometry for cells overexpressing IFN-γRα and subsequently maintained in medium containing G418. The retroviral vector pL(mIFNγ-KDEL)SN contains a mutant mIFN-γ containing the carboxyl-terminal endoplasmic reticulum retention signal Lys-Asp-Glu-Leu (KDEL) inserted into the retroviral vector LXSN (22, 23). The pL(mIFNγ-KDEL)SN vector was transfected into the BOSC-23 packaging cell line by standard calcium phosphate coprecipitation, and the 24-h culture supernatant was used to infect CT26 cells for 4 h in the presence of polybrene (4 μg/ml). G418-resistant clones were then analyzed for reduced expression of cell surface gp70 by flow cytometry.

Detection of gp70 protein by Western blot

Lysates were generated by resuspension of cells in lysis buffer (50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.1% SDS, 1% Triton X-100, and 20 μg/ml PMSF). The total protein levels of cleared lysates were determined using the Bio-Rad Dc Protein Assay (Bio-Rad, Richmond, CA). Samples (25 μg of total protein) were boiled for 3 min at 95°C in reducing buffer (0.125 M Tris-HCl (pH 6.8), 4% SDS, 0.005% bromophenol blue, 20% glycerol, and 0.7 M 2-ME) before being analyzed by electrophoresis in 4–12% Tris-glycine precast gels (NOVEX, San Diego, CA). After electrophoresis, proteins were transferred to PolyScreen polyvinylidene difluoride transfer membranes (NEN Life Science Products, Boston, MA) using a Bio-Rad semidry transfer cell. Probing of blots was conducted using Vectastain ABC Rat IgG peroxidase kit (Vector Laboratories, Burlingame, CA). Probed blots were developed using enhanced chemiluminescence Western blotting detection reagents (Amersham Pharmacia Biotech, Arlington Heights, IL) and were analyzed by autoradiography.

Cell-mediated cytotoxicity assay

For the generation of CT26-specific CTLs, mice were immunized with 2 × 106 irradiated (15,000 rad) CT26 cells on day 0 and challenged with 5 × 105 CT26 on day 21. Single-cell suspensions of splenocytes from mice remaining tumor free for 60 days were prepared, and RBC were removed by NH4Cl treatment. Splenocytes were cultured at 4 × 106/ml with 2 μM AH1 peptide in RPMI 1640 containing 10% FCS, penicillin/streptomycin, and 2-ME (0.05 mM). After 6 days, responder cells were harvested, washed twice, and incubated with 51Cr (1 mCi/ml; NEN)-labeled CT26, CT26 treated with mIFN-γ for 3 days at 1000 U/ml, or MHC class I-mismatched EL4 cells at different E:T cell ratios. Where indicated, target cells were incubated for 30 min at 37°C with AH1 peptide and washed with medium before addition of responders. Following an incubation period of 4 h, supernatants were assayed for radioactivity using a Wallac 1450 MicroBeta liquid scintillation counter (Wallac, Gaithersburg, MD). The percentage of specific lysis was calculated as [(sample cpm − spontaneous cpm)/(maximal cpm − spontaneous cpm)] × 100%. All assays were performed in quadruplicate. Spontaneous release was <15%.

Ags, mAbs, and cytokines

The AH1 peptide (SPSYVYHQF) is the class I Ld immunodominant epitope recognized within gp70 (19) and was provided by Drew Pardoll (Johns Hopkins University, Baltimore, MD). The E7 peptide (RAHYNIVTF) is a class I Db epitope from the human papilloma virus E7 protein (24). Abs against CD62 ligand (CD62L;3 clone MEL-14), CD8β.2 (clone 53-5.8), CD45 (clone 30-F11), IFN-γRα (clone GR.20), and murine IFN-γ (clone XMG1.2) were purchased from PharMingen (San Diego, CA). Secondary goat anti-rat FITC IgG and goat anti-mouse IgG FITC Abs were purchased from Sigma (St. Louis, MO). The 28-14-8S hybridoma (anti-Ld) was obtained from American Type Culture Collection (Manassas, VA) (2). The 35/299 hybridoma (α-gp70) was a gift from Drew Pardoll (Johns Hopkins University) and expresses a rat IgG2a mAb specific for gp70 (25). 35/299 and 28-14-8S hybridoma supernatants were purified over a protein G column (Amersham Pharmacia Biotech). Recombinant murine IFN-γ was purchased from Genzyme (Cambridge, MA).

Flow cytometry

To determine surface expression of murine IFN-γRα, cells were stained with primary monoclonal rat anti-IFN-γRα Ab and secondary goat anti-rat IgG FITC-conjugated Ab. Surface expression of MHC class I Ld was determined using purified 28-18-8S (anti-Ld) Ab and secondary goat anti-mouse IgG FITC Ab.

Intracellular cytokine expression

For assessment of tumor immunogenicity, irradiated (15,000 rad) tumor cells (5 × 106) were s.c. inoculated into BALB/c mice. Ten days later, single-cell suspensions of splenocytes from immunized mice were prepared, and RBC were removed by NH4Cl treatment. Splenocytes were cultured at 4 × 106/ml with 2 μM AH1 peptide in RPMI 1640 containing 10% FCS, penicillin/streptomycin, and 2-ME (0.05 mM). After 6 days, responder cells were harvested, washed, and restimulated with 2 μM AH1 or 2 μM E7 control peptide for 5 h in the presence of brefeldin A. For detection of cytoplasmic cytokine expression, cells were stained with PE-anti-CD8 mAb and APC-anti-CD62L mAb, fixed and permeabilized with Cytofix/CytoPerm solution (PharMingen), and stained with FITC-conjugated anti-IFN-γ mAb for 30 min on ice. The percentage of cells expressing cytoplasmic IFN-γ was determined by flow cytometry (FACScalibur; Becton Dickinson, Mountain View, CA).

Vaccinations and tumor growth

Irradiated (15,000 rad) CT26 (2 × 106) cells were inoculated s.c. into the right flank of BALB/c mice. After 21 days, mice were challenged with 5 × 105 CT26 or CT26.mugR cells s.c. in the left flank. Alternatively, mice were immunized i.p. with a recombinant vaccinia construct containing the full-length gp70 gene that was a gift from Drew Pardoll (Johns Hopkins University). After 28 days immunized mice were challenged with 5 × 104 CT26 or CT26.mugR cells s.c. in the left flank. Mice were monitored three times per week for the development of tumor nodules. In each experiment, eight mice were used per group.

Processing of tumors for ex vivo analysis

Tumors of 10 mm in diameter were excised, processed in PBS containing 1 mg/ml DNase I (Life Technologies), 2 mg/ml collagenase P (Roche, Indianapolis, IN), and penicillin/streptomycin for 1 h at 37°C as previously described (26). For analysis of intracellular cytokine expression, 2 × 106 cells were cultured per well in a 96-well round-bottom plate for 7 h in the presence of brefeldin A. Tumor cells were analyzed for gp70 expression by gating on CD45-negative cells.

Results

IFN-γ reduces tumor lysis by CTL

We undertook this study to determine whether IFN-γ would enhance the susceptibility of CT26 to recognition and lysis by tumor-specific CD8+ CTL. The dominant tumor Ag recognized on CT26 is gp70, an envelope protein of the endogenous murine leukemia virus. Previous work has demonstrated that the immunodominant epitope recognized within gp70 is AH1, a peptide nonamer presented in the context of MHC class I Ld (19). As shown in Fig. 1⇓A, IFN-γ enhances the expression of the MHC class I molecule Ld. Nevertheless, treatment of CT26 cells with IFN-γ for 3 days markedly reduced the level of CT26 lysis by AH1-specific CTL (Fig. 1⇓B). Recognition and lysis of IFN-γ-treated CT26 were restored by addition of the AH1 peptide, indicating that loss of recognition was due to lack of AH1 presentation.

FIGURE 1.
FIGURE 1.

IFN-γ enhances MHC class I while reducing CTL recognition of CT26. A, CT26 cells were cultured for 3 days in the presence of medium alone or 1000 U/ml IFN-γ and assessed for MHC class I Ld expression. B, CT26 cells were treated with medium alone or 1000 U/ml IFN-γ for 3 days and used as target cells in a CTL assay with CT26-specific CTL. Target cells were incubated for 30 min with or without AH1 peptide before the assay. MHC class I-mismatched EL4 cells were used as control targets.

IFN-γ down-regulates gp70 expression

We investigated the loss of AH1 presentation on CT26 cells following treatment with IFN-γ by analyzing the ability of IFN-γ to down-regulate gp70 protein levels. The gp70 envelope protein is normally expressed at modest levels on the cell surface of CT26. As shown in Fig. 2⇓A, treatment of CT26 with IFN-γ results in the loss of gp70 expression from the cell surface. Analysis of cell supernatants by Western blot demonstrated that this loss in gp70 surface expression is not due to shedding of the protein into the surrounding medium (data not shown). Rather, loss of surface expression is a result of reduced total gp70 protein levels, as shown in Fig. 2⇓B. Analysis of gp70 protein levels by Western blot revealed two molecular species, with the more slowly migrating product representing the gp70 protein. The other molecular species that migrates faster in the gel has previously been observed (27). This product, which probably represents nontranslocated, cytoplasmic gp70 polypeptides, lacks N-linked oligosaccharides as well as inter- or intramolecular disulfide bonds and is rapidly degraded in vivo (27). Our results indicate that this molecular species is not processed for presentation to CTL (Fig. 1⇑B).

FIGURE 2.
FIGURE 2.

IFN-γ down-regulates gp70 surface and total protein levels. A, CT26 cells were cultured with or without 1000 U/ml IFN-γ and subsequently stained for surface expression of gp70. B, Lysates from CT26 cells treated with 1000 U/ml IFN-γ for 1–4 days were analyzed by Western blot for expression of gp70. The upper band corresponds to the gp70 protein, whereas the lower band represents a nontranslocated form previously described (27 ).

Tumor responsiveness to IFN-γ promotes tumor escape in vivo

Since CT26 responsiveness to IFN-γ reduces the level of recognition by AH1-specific CTL, we hypothesized that CT26 cells rendered unresponsive to IFN-γ may be rejected more efficiently than IFN-γ-responsive CT26 cells in immune mice. To test this hypothesis, we took advantage of a dominant negative mutant of IFN-γRα that has been described previously (17, 18). By overexpressing this mutant in CT26 we generated cells designated CT26.mugR that present high levels of a truncated form of the IFN-γR α-chain at the cell surface (Fig. 3⇓A). These cells are unresponsive to IFN-γ, as shown by the inability of IFN-γ to down-regulate gp70 surface expression (Fig. 3⇓B). We next compared the levels of tumorigenicity of CT26 and CT26.mugR in immune mice. BALB/c mice were immunized with 2 × 106 irradiated CT26 cells s.c. in the right flank and 3 wk later were challenged with 5 × 105 cells of either CT26 or CT26.mugR. As shown in Fig. 4⇓A, the onset of tumor growth for naive animals challenged with either CT26 or CT26.mugR was the same. However, CT26 immune animals were found to reject CT26.mugR more efficiently than CT26. We also observed this same finding in mice vaccinated with a recombinant vaccinia construct only expressing the gp70 gene (Fig. 4⇓B). These observations support our hypothesis that tumor responsiveness to IFN-γ favors tumor outgrowth in the CT26 model.

FIGURE 3.
FIGURE 3.

CT26.mugR overexpresses IFN-γRα and is unresponsive to IFN-γ. A, CT26 and CT26.mugR were stained for expression of IFN -γRα. B, CT26 and CT26.mugR were cultured with medium alone or 1000 U/ml IFN-γ for 3 days and subsequently analyzed for expression of gp70 by flow cytometry.

FIGURE 4.
FIGURE 4.

Tumor responsiveness to IFN-γ promotes tumor outgrowth in immune mice. A, BALB/c mice were immunized with either saline (• and ○), 2 × 106 irradiated CT26 s.c. (▪ and □), or 2 × 106 irradiated CT26.IFN s.c. (▴ and ▵) on day −21. On day 0, mice were challenged with 5 × 105 CT26 (○, □, and ▵) or CT26.mugR (•, ▪, and ▴). B, BALB/c mice were immunized i.p. on day −28 with saline (•) or a vaccinia recombinant expressing gp70 (□). On day 0, mice were challenged with 5 × 104 CT26 (left panel) or CT26.mugR (right panel).

Tumor responsiveness to IFN-γ reduces tumor immunogenicity in vivo

Previous studies using tumor cells unresponsive to IFN-γ demonstrated that these tumor cells display a decreased level of immunogenicity in vivo. However, our findings would suggest that unresponsiveness to IFN-γ might increase immunogenicity. To test the role of tumor responsiveness to IFN-γ on tumor immunogenicity in vivo, we constructed an additional variant of CT26 that overexpresses IFN-γ (CT26.IFN). This variant was constructed by transducing CT26 with a retroviral vector coding for a mutant IFN-γ containing the carboxyl-terminal endoplasmic reticulum retention signal Lys-Asp-Glu-Leu (KDEL). Inclusion of the KDEL signal sequence anchors IFN-γ in the endoplasmic reticulum (22, 23). Therefore, IFN-γ signaling occurs only in the tumor cell, and no IFN-γ is secreted from the cell. CT26.IFN displays a loss of gp70 expression at both the surface and total protein levels, while exhibiting enhanced expression of MHC class I Ld as shown in Fig. 5⇓. We compared the level of immunogenicity of CT26, CT26.mugR, and CT26.IFN in vivo by immunizing mice with 5 × 106 irradiated tumor cells. As shown in Fig. 6⇓A, CT26.mugR induced the highest level of AH1-specific CD8+ cells to secrete IFN-γ in a 5-h intracellular cytokine assay, while CT26.IFN induced the fewest number of AH1-specific CD8+ cells. These findings demonstrate that CT26 responsiveness to IFN-γ decreases the level of tumor immunogenicity in vivo. A similar trend was observed using live tumor cells injected s.c. (Fig. 6⇓B). Although live tumor cells are less immunogenic than irradiated tumor cells, their level of immunogenicity is still inversely dependent on responsiveness to IFN-γ. In agreement with this finding, we have also observed that immunization of mice with irradiated CT26.IFN is less effective in protecting against challenge with CT26 (Fig. 4⇑A). Therefore, tumor responsiveness to IFN-γ reduces immunogenicity in the CT26 model.

FIGURE 5.
FIGURE 5.

IFN-γ reduces gp70 protein expression in CT26.CT26 or CT26.IFN cells cultured in vitro were assessed by flow cytometry for gp70 and MHC class I Ld expression (A) or Western blot for expression of gp70 (B).

FIGURE 6.
FIGURE 6.

Tumor responsiveness to IFN-γ reduces the level of in vivo tumor immunogenicity. Irradiated (15,000 rad, A, 5 × 106) or live (B, 1 × 105) CT26, CT26.mugR, or CT26.IFN cells were transplanted s.c. into individual BALB/c mice. Spleens were harvested 10 days later and set up in culture with 2 μM AH1 peptide. After 6 days, cells were restimulated with 2 μM AH1 (top panel) or control E7 peptide (bottom panel) for 7 h in the presence of brefeldin A. The levels of intracellular IFN-γ expressed by CD8+ CD62L− cells were quantified by intracellular cytokine staining. The percentage of CD8+ CD62L− cells secreting IFN-γ is shown in the top left corner of each plot.

Secretion of IFN-γ by T lymphocytes infiltrating tumor

We hypothesized that if IFN-γ is the dominant factor regulating gp70 expression in vivo, then ex vivo analysis of CT26 and CT26.mugR tumors should reveal differential levels of gp70. To test this hypothesis, BALB/c mice were challenged with 5 × 105 tumor cells s.c. and harvested when they reached 10 mm in diameter. By flow cytometric analysis, gating on tumor cells revealed that CT26 tumors had lost gp70 surface expression, whereas gp70 surface expression was retained in CT26.mugR tumors (Fig. 7⇓A). Further, preparation of ex vivo tumor lysates revealed a marked down-regulation of total gp70 levels in CT26 compared with CT26.mugR tumors (Fig. 7⇓B). We sought to determine the source of the IFN-γ responsible for the loss of gp70 expression in CT26 by analyzing ex vivo tumor homogenates. We found by flow cytometry that mice challenged with CT26 and CT26.mugR display a vast tumor infiltration of CD45-positive cells containing CD8+, CD4+, and MAC-1+ cells. We show here in Fig. 8⇓ that tumor-infiltrating CD8+ T cells secrete IFN-γ in both CT26 and CT26.mugR. We have also observed secretion of IFN-γ by tumor-infiltrating CD4+ cells, but not MAC-1+ cells (data not shown). These findings indicate that CT26 has evolved to evade an immune response by down-regulating expression of the gp70 tumor Ag in response to tumor infiltrating T lymphocytes secreting IFN-γ.

FIGURE 7.
FIGURE 7.

CT26 loss of gp70 expression in vivo is IFN-γ dependent. CT26 or CT26.mugR cells (5 × 104) were transplanted s.c. into BALB/c mice. A, When tumors reached 10 mm in diameter, they were excised and analyzed by flow cytometry for tumor expression of gp70. B, Lysates prepared from in vitro cultured CT26 (lane 1) or ex vivo processed CT26 (lane 2) or CT26.mugR (lane 3) tumors were analyzed by Western blot for expression of gp70.

FIGURE 8.
FIGURE 8.

CD8 cells infiltrating CT26 and CT26.mugR secrete IFN-γ. CT26 or CT26.mugR tumors of 10 mm in diameter were excised from individual BALB/c mice. The intracellular IFN-γ levels of CD8+ CD62L− cells infiltrating CT26 (top panel) or CT26.mugR (bottom panel) tumors were measured by intracellular cytokine staining following a 7-h culture with brefeldin A. Cells were first stained for surface expression of CD8 and CD62L and then for intracellular expression of IFN-γ using either an anti-IFN-γ Ab (left panel) or control isotype-matched Ab (right panel). The percentage of CD8+CD62L− cells secreting IFN-γ is shown in the top right corner of each plot.

Discussion

Unexpectedly, we have found that CT26 tumor responsiveness to IFN-γ increases MHC class I expression at the expense of reduced tumor recognition and lysis by CTL. We demonstrate here that this loss of recognition and lysis is explained by the ability of IFN-γ to down-regulate the protein levels of the immunodominant tumor Ag gp70. We investigated the significance of this observation in vivo by using IFN-γ-insensitive CT26 tumors to evaluate the role of tumor responsiveness to IFN-γ for rejection in immune mice. Perhaps the most significant observation of this study is the demonstration that tumor responsiveness to IFN-γ reduces tumor immunogenicity in vivo. This loss of immunogenicity is shown to be the result of a reduction in gp70 expression in vivo in response to IFN-γ secretion by tumor-infiltrating lymphocytes. As a result, immune mice display an enhanced rejection of IFN-γ-insensitive compared with IFN-γ-responsive CT26 tumors.

The findings reported in this study are the first to demonstrate a role for tumor responsiveness to IFN-γ in promoting tumor escape in vivo. Other tumor models have demonstrated the requirement of tumor responsiveness to IFN-γ for therapeutic effectiveness. Dighe et al. (17) demonstrated that sensitivity to IFN-γ in the MethA tumor increases tumor immunogenicity in vivo. The authors explained that MethA tumor cells are not inherently very immunogenic, but acquire immunogenicity when exposed to host-derived IFN-γ in vivo. As a result, established IFN-γ-insensitive MethA tumors remain unresponsive to treatment with LPS, whereas IFN-γ-responsive tumors are rejected. Similarly, Coughlin et al. (18) have shown that tumor insensitivity to IFN-γ enhances tumorigenicity in vivo. In this study tumor responsiveness to IFN-γ was required for the effectiveness of IL-12 to slow tumor growth through inhibition of angiogenesis. However, tumor responsiveness to IFN-γ was recently shown not to be required for the ability of adoptively transferred CD4+ T cells to induce tumor rejection (28). Therefore, it is likely that both the form of immune therapy adopted and the characteristics of the tumor targeted will be important in determining the requirement for tumor responsiveness to IFN-γ.

The gp70 protein was identified as a tumor Ag on CT26 by a method involving the fractionation of tumor cell antigenic content and the assaying of each fraction for recognition by tumor-specific CTL. This method, using in vitro cultured CT26 cells as a source of Ag, led to the identification of the AH1 antigenic epitope of gp70 (19). However, we have found that gp70 is differentially expressed on freshly ex vivo isolated and in vitro cultured CT26 cells. Our results demonstrating loss of Ag expression in vivo urge the use of freshly isolated autologous tumor cells rather than cultured tumor cell lines for identification of potential tumor rejection Ags. Indeed, distinct CTL can often display differential activity against fresh and cultured autologous tumor cells (29). In addition, although previous work has described gp70 as the immunodominant tumor Ag of CT26, our finding that vaccinia virus-expressing gp70 is ineffective in providing protection against challenge with the parental tumor suggests that gp70 may not be a good target for tumor immunotherapy. These results further suggest the existence of other potential CT26-specific rejection Ags, as enhanced protection is provided with whole irradiated tumor cells. Nevertheless, the ability of vaccinia-expressing gp70 to protect against challenge with an IFN-γ-resistant variant of CT26 implies that gp70 can act as a tumor rejection Ag under these conditions. These results provide support for IFN-γ-dependent loss of endogenous tumor Ag expression as a mechanism of tumor escape that is associated with decreased tumor immunogenicity.

The ability of IFN-γ to regulate tumor immunogenicity in vivo has previously been investigated in the MethA tumor model. In this system, it was proposed that IFN-γ may induce a subset of proteasomes that contains latent membrane protein-2 and -7, called the immunoproteasome, which may be critical in enhancing immunogenicity (17). However, recent work by Morel et al. (30) has demonstrated that induction of the immunoproteasome by IFN-γ results in less efficient processing of some tumor Ags, allowing for evasion of recognition by CTL. In this study, in agreement with our results, IFN-γ was found to decrease tumor immunogenicity. However, in contrast to our findings, the majority of tumor Ags examined were not significantly regulated by IFN-γ (30). Although we cannot dismiss the possibility in the CT26 tumor model that induction of the immunoproteasome by IFN-γ reduces efficient processing of the AH1 epitope of gp70, we have shown that IFN-γ significantly down-regulates gp70 to levels that would result in reduced CTL recognition and lysis.

Loss of tumor Ag expression is recognized as a mechanism by which tumors may escape immune recognition. In murine models, elimination or down-regulation of entire genes encoding tumor Ags has been correlated with decreased immunogenicity and increased tumorigenicity of Ag loss variants (31, 32). In humans, several reports have documented the loss of immunogenic tumor Ags during tumor progression or following therapy. Mai et al. (33) reported two cases of advanced prostatic adenocarcinoma following hormonal therapy, showing complete loss of three tissue immunoreactive prostatic markers, including prostate-specific Ag. In melanoma, loss of the targeted Ags tyrosinase and MART-1/Melan-A is often observed in patients undergoing peptide-based immunization (3). These reports suggest that selective pressure facilitates the emergence of Ag-negative tumors. Indeed, the observation that tumor Ag expression can be down-modulated by IFN-γ may help explain the emergence of tumor Ag loss variants following immunotherapy. In fact, the ability of IFN-γ to decrease the levels of human papilloma virus E6 and E7 transcripts in several cervical cancer cell lines (34) may explain the ineffectiveness of E7 peptide-based vaccine therapies (35, 36) and the limited correlation between intratumoral expression of IFN-γ and clinical outcome (37). In addition, in melanoma IFN-γ down-regulation of the tissue-specific tumor Ag MART-1/Melan-A (30) as well as several other tumor-associated Ags has been reported (38). Certainly, IFN-γ down-regulation of MART-1/Melan-A expression may explain the frequently observed loss of MART-1/Melan-A in patients receiving peptide-based therapy. Therefore, although our study addresses the regulation of a retrovirally derived murine endogenous tumor Ag, it is likely that the findings presented here may be extended to human tumor viral Ags as well as to human tissue-specific tumor Ags.

If loss of tumor Ag expression following immunotherapy targeting the induction of type I cell-mediated immunity is a frequent event, development of strategies to counteract this phenomenon must be explored. Recent reports have demonstrated that type I and type II therapies are equally effective in inducing tumor regression (39, 40). However, our findings suggest that tumor Ag-based therapies exploiting type I immune responses may in some cases be ineffective due to the down-regulation of tumor Ags by type I cytokine-secreting tumor-infiltrating lymphocytes. In such cases, therapies promoting type II immune responses may be worthy of investigation and prove more successful.

Finally, our results emphasize the importance of understanding the regulation of tumor Ag expression in vivo for enhancing the effectiveness of tumor-Ag based immunotherapies. Although tumor exposure to IFN-γ has previously been recognized to enhance tumor immunogenicity, it is clear that down-regulation of tumor Ag expression or reduced Ag processing by immunoproteasomes can reduce tumor immunogenicity. These findings may provide an explanation for the lack of correlation between induction of antitumor CTL capable of secreting IFN-γ and tumor regression as well as account for some observations demonstrating increased tumor progression following systemic IFN-γ treatment (41, 42). In addition, our results support a need for investigating the effectiveness of type II compared with type I immunotherapy as an alternative therapeutic approach for targeting endogenous tumor Ags that may be otherwise lost during type I immune responses.

Acknowledgments

We thank Drs. William M. F. Lee, Drew Pardoll, and Denise La Temple for helpful discussions and providing reagents.

Footnotes

  • ↵1 This work was supported by National Institutes of Health Grant CA69632 (to Y.P.) and National Institutes of Health Training Grant T32CA09140 (to G.L.B.).

  • ↵2 Address correspondence and reprint requests to Dr. Yvonne Paterson, University of Pennsylvania, 323 Johnson Pavilion, 3610 Hamilton Walk, Philadelphia, PA 19104-6076. E-mail address: yvonne{at}mail.med.upenn.edu

  • ↵3 Abbreviation used in this paper: CD62L, CD62 ligand.

  • Received May 18, 2000.
  • Accepted August 21, 2000.

References

  1. ↵
    Webber, J. S., S. A. Rosenberg. 1998. Modulation of murine tumor major histocompatibility antigens by cytokines in vivo and in vitro. Cancer Res. 48: 5818
    OpenUrlAbstract/FREE Full Text
  2. ↵
    Ferrone, S., F. M. Marincola. 1995. Loss of HLA class I antigens by melanoma cells: molecular mechanisms, functional significance and clinical relevance. Immunol. Today 16: 487
    OpenUrlCrossRefPubMed
  3. ↵
    Maeurer, M. J., S. M. Gollin, D. Martin, W. Swaney, J. Bryant, C. Castelli, P. Robbins, G. Parmiani, W. J. Storkus, M. T. Lotze. 1996. Tumor escape from immune recognition: lethal recurrent melanoma in a patient associated with downregulation of the peptide transporter protein TAP-1 and loss of expression of the immunodominant MART-1/Melan-A antigen. J. Clin. Invest. 98: 1633
    OpenUrlCrossRefPubMed
  4. ↵
    Bodmer, S., K. Strommer, K. Frei, C. Siepl, N. de Tribolet, I. Heid, A. Fontana. 1989. Immunosuppression and transforming growth factor-β in glioblastoma: preferential production of transforming growth factor-β2. J. Immunol. 143: 3222
    OpenUrlAbstract/FREE Full Text
  5. ↵
    Hirte, H., D. A. Clark. 1991. Generation of lymphokine-activated killer cells in human ovarian carcinoma ascitic fluid: identification of transforming growth factor-β as a suppressive factor. Cancer Immunol. Immunother. 32: 296
    OpenUrlCrossRefPubMed
  6. ↵
    Campana, D., E. Coustan-Smith, A. Manabe, M. Buschle, S. C. Raimondi, F. G. Behm, R. Ashmun, M. Arico, A. Biondi, C. H. Pui. 1993. Prolonged survival of B-lineage acute lymphoblastic leukemia cells is accompanied by overexpression of bcl-2 protein. Blood 81: 1025
    OpenUrlAbstract/FREE Full Text
  7. ↵
    Miyake, H., I. Hara, K. Yamanaka, K. Gohji, S. Arakawa, S. Kamidono. 1999. Overexpression of Bcl-2 enhances metastatic potential of human bladder cancer cells. Br. J. Cancer. 79: 1651
    OpenUrlCrossRefPubMed
  8. ↵
    Liu, R., C. Page, D. R. Beidler, M. S. Wicha, G. Nunez. 1999. Overexpression of Bcl-x(L) promotes chemotherapy resistance of mammary tumors in a syngeneic mouse model. Am. J. Pathol. 155: 1861
    OpenUrlCrossRefPubMed
  9. ↵
    Buzyn, A., F. Petit, M. Ostankovitch, S. Figueiredo, B. Varet, J.-G. Guillet, J.-C. Ameisen, J. Estaquier. 1999. Membrane-bound Fas (Apo-1/CD95) ligand on leukemic cells: a mechanism of tumor immune escape in leukemia patients. Blood 94: 3135
    OpenUrlAbstract/FREE Full Text
  10. ↵
    O’Connell, J., G. C. O’Sullivan, J. K. Collins, F. Shanahan. 1996. The Fas counterattack: Fas-mediated T cell killing by colon cancer cells expressing Fas ligand. J. Exp. Med. 184: 1075
    OpenUrlAbstract/FREE Full Text
  11. ↵
    Ogawa, M., W. G. Yu, K. Umerhara, M. Iwasaki, R. Wijesuriya, T. Tsujimura, T. Kubo, H. Fujiwara, T. Hamaoka. 1998. Multiple roles of interferon-γ in the mediation of interleukin 12-induced tumor regression. Cancer Res. 58: 2426
    OpenUrlAbstract/FREE Full Text
  12. ↵
    Nastala, C. L., H. D. Edington, T.G. McKinney, H. Tahara, M. A. Nalesnik, M. J. Brunda, M. K. Gately, S. F. Wolf, R. D. Schreiber, W. J. Storkus, et al 1994. Recombinant IL-12 administration induces tumor regression in association with IFN-γ production. J. Immunol. 153: 1697
    OpenUrlAbstract/FREE Full Text
  13. ↵
    Sgadari, C., A. L. Angiolillo, G. Tosato. 1996. Inhibition of angiogenesis by interleukin-12 is mediated by the interferon-inducible protein-10. Blood 87: 3877
    OpenUrlAbstract/FREE Full Text
  14. ↵
    Sgadari, C., J. M. Farber, A. L. Angiolillo, F. Liao, J. Teryua-Feldstein, P. R. Burd, L. Yao, G. Gupta, C. Kanegane, G. Tosato. 1997. Mig, the monokine induced by interferon-γ, promotes tumor necrosis in vivo. Blood 89: 2635
    OpenUrlAbstract/FREE Full Text
  15. ↵
    Naganuma, H., A. Sasaki, E. Satoh, M. Nagasaka, S. Nakano, S. Isoe, H. Nukui. 1998. Down-regulation of transforming growth factor-β and interleukin-10 secretion from malignant glioma cells by cytokines and anticancer drugs. J. Neurooncol. 39: 227
    OpenUrlPubMed
  16. ↵
    Boraschi, D., S. Censini, A. Tagliabue. 1984. Interferon-γ reduces macrophage-suppressive activity by inhibiting prostaglandin E2 release and inducing interleukin 1 production. J. Immunol. 133: 764
    OpenUrlAbstract
  17. ↵
    Dighe, A. S., E. Richards, L. J. Old, R. D. Schreiber. 1994. Enhanced in vivo growth and resistance to rejection of tumor cells expressing dominant negative IFNγ receptors. Immunity 1: 447
    OpenUrlCrossRefPubMed
  18. ↵
    Coughlin, C. M., K. E. Salhany, M. S. Gee, D.C. LaTemple, S. Kotenko, X. Ma, G. Gri, M. Wysocka, J. E. Kim, L. Liu, et al 1998. Tumor cell responses to IFN-γ affect tumorigenicity and response to IL-12 therapy and antiangiogenesis. Immunity 9: 25
    OpenUrlCrossRefPubMed
  19. ↵
    Huang, A. Y. C., P. H. Gulden, A. S. Woods, M. C. Thomas, C. D. Tong, W. Wang, V. H. Engelhard, G. Pasternack, R. Cotter, D. Hunt, et al 1996. The immunodominant major histocompatibility complex class I-restricted antigen of a murine colon tumor derives from an endogenous retroviral gene product. Proc. Natl. Acad. Sci. USA 93: 9730
    OpenUrlAbstract/FREE Full Text
  20. ↵
    Corbett, T. H., D. J. Griswold, B. J. Roberts, J. C. Peckham, F. J. Schabel. 1975. Tumor induction relationships in development of transplantable cancers of the colon in mice for chemotherapy assays, with note on carcinogen structure. Cancer Res. 35: 2434
    OpenUrlAbstract/FREE Full Text
  21. ↵
    Goldman, L. A., E. C. Cutrone, S. V. Kotenko, C. D. Krause, J. A. Langer. 1996. Modification of vectors pEF-BOS, pcDNA-1, pcDNA-3 result in improved convenience and expression. BioTechniques 21: 1013
    OpenUrlPubMed
  22. ↵
    Miller, A. D., G. J. Rosman. 1989. Improved retroviral vectors for gene transfer and expression. BioTechniques 8: 980
    OpenUrl
  23. ↵
    Will, A., U. Hemmann, F. Horn, M. Rollinghoff, A. Gessner. 1996. Intracellular murine IFN-γ mediates virus resistance, expression of oligoadenylate synthetase, and activation of STAT transcription factors. J. Immunol. 157: 4576
    OpenUrlAbstract
  24. ↵
    Feltkamp, M. C. W., H. L. Smits, M. P. M. Vierboom, R. P. Minnaar, B. M. de Jongh, J. W. Drijfhout, J. ter Schegget, C. J. M. Melief, W. M. Kast. 1993. Vaccination with cytotoxic T lymphocyte epitope-containing peptide protects against a tumor induced by human papillomavirus type 16-transformed cells. Eur. J. Immunol. 23: 2242
    OpenUrlCrossRefPubMed
  25. ↵
    Pinter, A., W. J. Honnen, J. S. Tung, P.V. O’Donnell, U. Hammerling. 1982. Structural domains of endogenous murine leukemia gp70s containing specific antigenic determinants by monoclonal antibodies. Virology 116: 499
    OpenUrlCrossRefPubMed
  26. ↵
    Kedar, E., B. L. Ikejiri, G. D. Bonnard, R. B. Herberman. 1982. A rapid technique for isolation of viable tumor cells from solid tumors: use of the tumor cells for induction and measurement of cell-mediated cytotoxic responses. Eur. J. Cancer Clin. Oncol. 18: 991
    OpenUrlCrossRefPubMed
  27. ↵
    Opstelten, D. E., M. Wallin, H. Garoff. 1998. Moloney murine leukemia virus envelope protein subunits, gp70 and pr15e, form a stable disulfide-linked complex. J. Virol. 72: 6537
    OpenUrlAbstract/FREE Full Text
  28. ↵
    Mumberg, D., P. A. Monach, S. Wanderling, M. Philip, A. Y. Toledano, R. D. Schreiber, H. Schreiber. 1999. CD4+ T cells eliminate MHC class II-negative cancer cells in vivo by indirect effects of IFN-γ. Proc. Natl. Acad. Sci. USA 96: 8633
    OpenUrlAbstract/FREE Full Text
  29. ↵
    Faure, F., J. Even, P. Kourilsky. 1998. Tumor-specific immune response: current in vitro analyses may not reflect the in vivo immune status. Crit. Rev. Immunol. 18: 77
    OpenUrlCrossRefPubMed
  30. ↵
    Morel, S., F. Levy, O. Burelet-Schlitz, F. Brasseur, M. Probst-Kepper, A. L. Peitrequin, B. Monsarrat, R. V. Velthoven, J. C. Cerottini, T. Boon, et al 2000. Processing of some antigens by the standard proteasome but not by the immunoproteasome results in poor presentation by dendritic cells. Immunity 12: 107
    OpenUrlCrossRefPubMed
  31. ↵
    De Plaen, E., C. Lurquin, A. Van Pel, B. Mariame, J. Szikora, T. Wolfel, C. Sibille, P. Chomez, T. Boon. 1988. Immunogenic (tum-) variants of mouse tumor P815: cloning of the gene of tum-antigen P19A and identification of the tum-mutation. Proc. Natl. Acad. Sci. USA 85: 2274
    OpenUrlAbstract/FREE Full Text
  32. ↵
    Van den Eynde, B., B. Lethe, A. Van Pel, E. De Plaen, T. Boon. 1991. The gene coding for a major tumor rejection antigen to tumor P815 is identical to the normal gene of syngeneic DBA/2 mice. J. Exp. Med. 173: 1373
    OpenUrlAbstract/FREE Full Text
  33. ↵
    Mai, K. T., A. S. Commons, D. G. Perkins, H. M. Yazdi, J. P. Collins. 1996. Absence of serum prostate-specific antigen and loss of tissue immunoreactive prostatic markers in advanced prostatic adenocarcinoma after hormonal therapy: a report of two cases. Hum. Pathol. 27: 1377
    OpenUrlCrossRefPubMed
  34. ↵
    Kim, K. Y., L. Blatt, M. W. Taylor. 2000. The effects of interferon on the expression of human papillomavirus oncogenes. J. Gen. Virol. 81: 695
    OpenUrlAbstract/FREE Full Text
  35. ↵
    Steller, M. A., K. J. Gurski, M. Murakami, R. W. Daniel, K. V. Shah, E. Celis, A. Sette, E. L. Trimble, R. C. Park, F. M. Marincola. 1998. Cell-mediated immunological responses in cervical and vaginal cancer patients immunized with a lipidated epitope of human papillomavirus type 16 E7. Clin. Cancer Res. 4: 2103
    OpenUrlAbstract
  36. ↵
    Ressing, M. E., W. J. van Driel, R. M. Brandt, G. G. Kenter, J. H. de Jong, T. Bauknecht, G. J. Fleuren, P. Hoogerhout, R. Offringa, A. Sette, et al 2000. Detection of T helper responses, but not of human papillomavirus-specific cytotoxic T lymphocyte responses, after peptide vaccination of patients with cervical carcinoma. J. Immunother. 23: 255
    OpenUrlCrossRef
  37. ↵
    Tartour, E., A. Gey, X. Sastre-Garau, I. L. Surin, V. Mosseri, W. H. Fridman. 1998. Prognostic value of intratumoral interferon γ messenger RNA expression in invasive cervical carcinomas. J. Natl. Cancer Inst. 90: 287
    OpenUrlAbstract/FREE Full Text
  38. ↵
    Mortarini, R., F. Belli, G. Parmiani, A. Anichini. 1990. Cytokine-mediated modulation of HLA-class II, ICAM-1, LFA-3 and tumor-associated antigen profile of melanoma cells: comparison with anti-proliferative activity by rIL1-β, rTNF-α, rIFN-γ, rIL4 and their combinations. Int. J. Cancer 45: 334
    OpenUrlCrossRefPubMed
  39. ↵
    Nishimura, T., K. Iwakabe, M. Sekimoto, Y. Ohmi, T. Yahata, M. Nakui, T. Sato, S. Habu, H. Tashiro, M. Sato, et al 1999. Distinct role of antigen-specific T helper type (Th1) and Th2 cells in tumor eradication in vivo. J. Exp. Med. 190: 617
    OpenUrlAbstract/FREE Full Text
  40. ↵
    Dobrzanski, M. J., J. B. Reome, R. W. Dutton. 2000. Type 1 and type 2 CD8+ effector T cell subpopulations promote long-term tumor immunity and protection to progressively growing tumor. J. Immunol. 164: 916
    OpenUrlAbstract/FREE Full Text
  41. ↵
    Meyskens, F. L., K. Kopecky, M. Samson, E. Hersh, J. Macdonald, H. Jaffe, J. Crowley, C. Coltman. 1990. Recombinant human interferon γ: adverse effects in high-risk stage I and II cutaneous malignant melanoma. J. Natl. Cancer Inst. 82: 1071
    OpenUrlFREE Full Text
  42. ↵
    Kowalzick, L., U. Weyer, P. Lange, E. W. Breitbart. 1990. Systemic therapy of advanced metastatic malignant melanoma with a combination of fibroblast interferon-β and recombinant interferon-γ. Dermatologica 181: 298
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section