TNF Plays an Essential Role in Tumor Regression after Adoptive Transfer of Perforin/IFN-γ Double Knockout Effector T Cells

Christian H. Poehlein, Hong-Ming Hu, Jane Yamada, Ilka Assmann, W. Gregory Alvord, Walter J. Urba and Bernard A. Fox
J Immunol February 15, 2003, 170 (4) 2004-2013; DOI: https://doi.org/10.4049/jimmunol.170.4.2004

Abstract

We have recently shown that effector T cells (TE) lacking either perforin or IFN-γ are highly effective mediators of tumor regression. To rule out compensation by either mechanism, TE deficient in both perforin and IFN-γ (perforin knockout (PKO)/IFN-γ knockout (GKO)) were generated. The adoptive transfer of PKO/GKO TE mediated complete tumor regression and cured wild-type animals with established pulmonary metastases of the B16BL6-D5 (D5) melanoma cell line. PKO/GKO TE also mediated tumor regression in D5 tumor-bearing PKO, GKO, or PKO/GKO recipients, although in PKO/GKO recipients efficacy was reduced. PKO/GKO TE exhibited tumor-specific TNF-α production and cytotoxicity in a 24-h assay, which was blocked by the soluble TNFRII-human IgG fusion protein (TNFRII:Fc). Blocking TNF in vivo by administering soluble TNFR II fusion protein (TNFRII:Fc) significantly reduced the therapeutic efficacy of PKO/GKO, but not wild-type TE. This study identifies perforin, IFN-γ, and TNF as a critical triad of effector molecules that characterize therapeutic antitumor T cells. These insights could be used to monitor and potentially tune the immune response to cancer vaccines.

Tumor-specific T cells can mediate regression of established tumor, prolong survival, and cure animals, and in some cases, patients with cancer; however, the therapeutic results in clinical studies have been discouraging and the mechanisms used by effector T cells (TE)3 to attack and eradicate solid tumors are poorly understood (1, 2, 3, 4, 5). Early studies investigating the mechanism of T cell-mediated tumor regression correlated tumor-specific cytolytic function with therapeutic activity (6, 7, 8, 9, 10). This concept was further buttressed by the finding that cytolytic CD8+ T cell bulk cultures or clones could mediate tumor regression (11, 12, 13, 14, 15, 16). The report by Barth et al. (17) was the first to document that T cells lacking tumor-specific cytolytic activity, measured by standard 4-h 51Cr-release assays, could effectively eliminate metastatic tumor deposits and subsequent studies confirmed this finding (5, 18). However, these experiments could not exclude the possible acquisition of cytolytic function by T cells in vivo following adoptive transfer. Matsumura et al. (19) confirmed this possibility, documenting that nonlytic CD8+ T cells could become cytolytic following interaction with tumor in situ. The development of gene knockout (KO) mice that lacked a defined lytic pathway provided the first opportunity to test directly whether cytolytic function was required for T cell-mediated tumor regression. Studies using perforin KO (PKO) and Fas ligand (FasL)-deficient (gld) mice clearly ruled out a requirement for these molecules in T cell-mediated regression of pulmonary metastases (20). Subsequent investigations using PKO mice confirmed the report of Winter et al. (20), and extended the findings to intracranial, but not s.c., tumors (21, 22).

In the absence of cytolytic effector function, IFN-γ might be responsible for the observed antitumor effect (23, 24). IFN-γ can inhibit proliferation of some tumors and has potent antiangiogenic activity (25, 26, 27). Importantly, IFN-γ can also orchestrate the recruitment and activation of the innate immune response (28, 29, 30), augment Ag processing, and up-regulate expression of MHC class I and II molecules on tumor cells and APC (30, 31, 32, 33). A role for IFN-γ in T cell-mediated tumor regression was first documented by Barth et al. (17), who showed that neutralizing IFN-γ reduced the therapeutic efficacy of adoptively transferred T cells that exhibited tumor-specific secretion of IFN-γ. Subsequent studies documented a strong correlation between tumor-specific IFN-γ secretion and therapeutic efficacy (34). However, studies using IFN-γ KO (GKO) mice provided the first direct evidence that IFN-γ was not required for T cell-mediated tumor regression (21, 35). Because both perforin and IFN-γ are known to have potent antitumor effects and yet the absence of either one resulted in tumor regression, we considered the possibility that in the absence of either perforin or IFN-γ the remaining effector mechanism might “compensate” for the lost function. We addressed this concept by transferring tumor-specific PKO TE into tumor-bearing IFN-γ receptor KO (GRKO) mice or tumor-bearing PKO recipients that were treated with neutralizing anti-IFN-γ mAb. Although both strategies failed to block therapeutic activity of transferred T cells, they could not eliminate the direct effect of IFN-γ on the tumor or the contribution of perforin-competent cells in the GRKO recipient in the first strategy, or formally exclude a role for IFN-γ in the second strategy (35).

To test directly whether either perforin or IFN-γ were required for T cell-mediated tumor regression, mice deficient in both genes were developed. These mice were unable to produce IFN-γ and failed to generate T cells or lymphokine-activated killer cells that could lyse appropriate targets in 6-h 51Cr-release assays. In this report, we examined whether tumor-specific T cells that lacked both IFN-γ and perforin could exhibit significant antitumor activity in vitro and in vivo.

Materials & Methods

Animals

PKO (C57BL/6-PFPtm1Sdz), GKO (C57BL/6-IFN-γtm1Ts), and wild-type (wt) mice on a C57BL/6 background were purchased from The Jackson Laboratory (Bar Harbor, ME). PKO and GKO mice were used to generate double KO mice. Because both perforin and IFN-γ are on the same chromosome, multiple backcrosses were required to generate a stable colony of PKO/GKO mice. Mice were screened for deficiency of both genes according to the protocols provided by The Jackson Laboratory (http://www.jax.org). These mice were submitted to The Jackson Laboratory and given the designation C57BL/6-Pfp<tm1Sdz>Ifng<tm1TS>. For all studies, mice were treated according to the principles of the Guide for the Care and Use of Laboratory Animals (National Research Council, 1996). All animal research protocols were approved by the Earle A. Chiles Research Institute animal care and use committee.

Tumor cell lines

B16BL6-D5 (D5) is a poorly immunogenic subclone of the spontaneously arising B16BL6 melanoma derived from C57BL/6 mice (provided by Dr. S. Shu, Cleveland Clinic Foundation, Cleveland, OH) (34, 36). D5-G6 is a stable clone of D5 that was transduced with a murine GM-CSF retroviral MFG vector (provided by Dr. M. Arca, University of Michigan, Ann Arbor, MI) (37). D5-G6 cells secrete ∼200 ng/ml/106cells/24 h GM-CSF. MCA-310 is a methyl-cholanthrene-induced, weakly immunogenic, fibrosarcoma of C57BL/6 mice (20).

Culture conditions

Lymphocytes and tumor cells were cultured in complete medium (CM), which consisted of RPMI 1640 supplemented with 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, 2 mM l-glutamine, 50 μg/ml gentamicin sulfate (all from BioWhittaker, Walkersville, MD), 50 μM 2-ME (Sigma-Aldrich, St. Louis, MO), and 10% FBS (Life Technologies, Grand Island, NY). Tumor cells were harvested two to three times per week by brief trypsinization (Trypsin; BioWhittaker) and maintained in T-75 or T-150 culture flasks.

Vaccination and challenge experiments

PKO/GKO or wt mice were injected s.c. with 10 × 106 irradiated D5-G6 tumor cells (10,000 rad) and challenged 14 days later with 2 × 104 D5 s.c. (10 × TD100 for D5 tumor). The TD100 is the dose at which 100% of the injected animals will develop tumor. In all experiments naive wt and PKO/GKO mice were included as controls and were challenged with the same tumor dose.

Generation of TE and adoptive immunotherapy

One million D5-G6 or MCA-310 tumor cells were injected s.c. in all four flanks of wt and PKO/GKO mice. Eight days following tumor inoculation, the tumor vaccine draining lymph nodes (TVDLN) were harvested, washed, resuspended at 2 × 106 cells per ml in CM and cultured in 24-well plates with anti-CD3 as described previously (20). After 2 days of activation, the T cells were harvested and expanded in CM containing 60 IU rhIL-2/ml (kind gift from Chiron, Emeryville, CA) for 3 additional days. T cells were then harvested, washed twice in HBSS, counted and used in cytotoxicity, intracellular cytokine staining, cytokine release assays, and adoptive transfer studies. For adoptive transfer experiments, wt, PKO, GKO, and PKO/GKO mice bearing 3-day established pulmonary metastases were generated by i.v. inoculation with 0.2 × 106 D5 or 0.3 × 106 MCA-310 tumor cells. T cells were administered i.v. and starting with the day of T cell transfer, animals received 90,000 IU IL-2 i.p. q.d. for 4 days. Where specified, 250 μg of TNFRII:Fc (38) (Amgen, Seattle, WA) or control human IgG (Sigma-Aldrich) were administered i.p at the time of adoptive transfer and again 2 days later. Mice were sacrificed 14 days following tumor inoculation, lungs were resected (in the case of MCA-310 metastases, lungs would be tracheal-infused with India ink for contrast), fixed in Fekete’s solution, and the number of pulmonary metastases were enumerated. In all experiments, TE used for in vivo studies were analyzed in parallel for their cytolytic capability and/or cytokine profile.

ELISA

PKO/GKO and wt TE were washed, resuspended in CM, and cultured at 4 × 106 cells/well in a 24-well plate. The cells were either left unstimulated (negative control) or stimulated with 2 × 105 tumor cells (D5, MCA-310) or anti-CD3 (positive control). Supernatants were harvested after 20–24 h and assayed for IFN-γ, TNF-α, IL-4, IL-5, and IL-10 using commercially available reagents (BD PharMingen, San Diego, CA).

In vitro cytotoxicity–6 and 24 h assay

For 6-h 51Cr-release assays, D5 or MCA-310 were incubated with 100 μCi Na51CrO4 for 90 min, washed and plated into 96-well round bottom plates with 104 tumor target cells/well. Target cells were incubated with TE in triplicate at the indicated E:T ratios in 200 μl of CM at 37°C in a CO2 incubator. The supernatant was harvested, counted, and the mean percent lysis was determined as previously described (20). For microscopic analysis, long-term (24 h) tumor-specific cytotoxicity assays were performed by culturing 0.2 × 106 D5 or MCA-310 cells in CM for 24 h with 4 × 106 D5-specific TE generated from wt or PKO/GKO mice in a 24-well flat-bottom plate. Tumor cells alone and tumor cells cultured with 4 × 106 TE generated from nonvaccinated naive wt and PKO/GKO lymph nodes were included as controls. After 24 h, TE and nonadherent cells were washed out and the remaining adherent tumor cell monolayer was fixed and stained with 0.05% crystal violet. The remaining fixed cells were documented using digital microscopic imaging.

For quantitative analysis of long-term cytotoxicity, 104 D5 melanoma cells were cultured in 200 μl of CM in a 96-well flat-bottom plate alone or with 106 TE in triplicates. After 24 h at 37°C in a CO2 incubator, plates were gently washed, adherent tumor cells were fixed (3:1 methanol-glacial acid for 5–10 min), rinsed with tap water, and stained with 0.05% crystal violet in 10% ethanol for 15 min. The plate was then washed thoroughly with tap water and stained cells were dissolved in 10% acetic acid/20% ethanol and the OD was determined using a 592-nm filter on a microplate reader (Bio-Tek Instruments, Winooksi, VT). Absorbance of wells with tumor cells alone was used as the reference for no cytotoxicity and buffer alone as reference for 100% lysis.

Intracellular TNF-α staining and flow cytometric analysis

PKO/GKO and wt TE were stimulated for 12 h in the presence of 5 μg/ml brefeldin A (Sigma-Aldrich) in CM only (no stimulation), with specific tumor (D5), unrelated syngeneic tumor (MCA-310), or immobilized anti-CD3. TE were harvested and stained with anti-CD8FITC and anti-CD3CyChrome mAbs, fixed and permeabilized in Cytofix/Cytoperm and stained intracellularly with anti-TNF-αPE or anti-IFN-γPE mAb (all from BD PharMingen). Analysis was performed on 50,000 gated CD8+/CD3+ TE with a FACSCalibur and CellQuest software (BD Biosciences, San Diego, CA). Data is presented as the percentage of CD8+/CD3+/TNF-α+ TE, or CD8+/CD3+/IFN-γ+ TE, over the total number of CD8+/CD3+ TE.

Statistical analysis

Statistical analysis of the number of pulmonary metastases was performed by Wilcoxon rank sum test and survival analysis was performed using Kaplan-Meier plots and log rank sum tests using S-plus 2000 software (Data Analysis Product Division; Mathsoft, Seattle, WA).

Results

Vaccination of PKO/GKO mice does not induce protective antitumor immunity

After documenting that PKO/GKO TE lacked functional expression of perforin and IFN-γ (data not shown), we investigated whether vaccination with irradiated D5-G6 cells would induce a protective immune response in these double KO mice. Previously we had observed that D5-G6 vaccination in wt or PKO mice was highly effective at protecting from a subsequent tumor challenge, but in GKO mice, vaccination failed to protect from a subsequent tumor challenge (35). To determine whether vaccination with D5-G6 would induce protective immunity in mice lacking both genes, we repeated this experiment in PKO/GKO mice. All vaccinated PKO/GKO mice eventually developed tumors, while 80% of vaccinated wt mice were immune and resisted tumor challenge (Table I⇓). Thus, in the absence of both perforin and IFN-γ, vaccination with D5-G6 is unable to protect mice from a subsequent tumor challenge.

View this table:
Table I.

Vaccination of PKO/GKO mice does not induce protective antitumor immunity

PKO/GKO TE are therapeutic in wt recipients

We next investigated whether PKO/GKO TE cells could mediate tumor regression. The adoptive transfer of 35 or 70 × 106 T cells from either wt or PKO/GKO mice significantly (p < 0.05) reduced the number of pulmonary metastases in wt mice. However, PKO/GKO TE were clearly less effective than wt TE, as evidenced by the resulting increased number of metastases in mice receiving 35 × 106 in all three experiments (Fig. 1⇓). Although, PKO/GKO TE were not as therapeutic as wt TE on a per cell basis, they were capable of mediating complete tumor regression in wt animals with systemic tumor when 70 × 106 TE were transferred. Furthermore, this finding demonstrates that neither perforin nor IFN-γ was required to prime TE in response to a GM-CSF-secreting tumor vaccine. Next, we examined whether the antitumor activity of PKO/GKO TE could improve survival of wt-recipient mice with established systemic tumor. Although all IL-2-treated control animals were dead by 18–20 days, animals treated with the adoptive transfer of 70 × 106 PKO/GKO TE and IL-2 treatment survived significantly longer (∗, p < 0.0001) and 50% of treated animals were cured (Fig. 2⇓).

FIGURE 1.
FIGURE 1.

PKO/GKO TE cells mediate significant regression of 3-day established D5 pulmonary metastases in wt-recipient mice (∗, p < 0.01). TE cells from PKO/GKO or wt mice were adoptively transferred into wt mice with 3-day established pulmonary metastases and 90,000 IU IL-2 i.p. was administered for four consecutive days. Mice receiving IL-2 alone served as controls. The number of T cells transferred (35 or 70 × 106) is presented for each group (n = 5) of animals for three consecutive and independent experiments. Mice were sacrificed 14 days after tumor inoculation, lungs were resected and fixed in Fekete’s solution, and the number of pulmonary metastases were counted. Each triangle represents an individual mouse.

FIGURE 2.
FIGURE 2.

PKO/GKO TE can cure wt recipients. Seventy million PKO/GKO (▴) and wt (▪) TE were adoptively transferred into wt mice bearing 3-day established D5 pulmonary metastases. Control mice did not receive TE and all mice were treated with 90,000 IU IL-2 i.p. for four consecutive days, starting at the time of adoptive transfer. Survival of animals receiving TE was significantly (∗, p < 0.0001) improved compared with mice treated with IL-2 alone (Embedded Image).

PKO/GKO TE antitumor capacity is not limited to melanoma

The ability of PKO/GKO TE to mediate tumor regression was not unique to the D5 melanoma system. The adoptive transfer of TE from MCA-310-vaccinated PKO/GKO mice were able to mediate regression of established 3-day MCA-310 pulmonary metastases (Fig. 3⇓). The results in this model resembled those seen with D5. The adoptive transfer of PKO/GKO TE resulted in significant (∗, p < 0.01), dose-dependent regression of pulmonary metastases from this weakly immunogenic fibrosarcoma, but on a per-cell basis PKO/GKO TE were not as efficient as wt TE.

FIGURE 3.
FIGURE 3.

PKO/GKO TE also mediate tumor regression in a fibrosarcoma model. PKO/GKO and wt TE generated from MCA-310 TVDLN mediated regression of 3-day established MCA-310 pulmonary metastases (∗, p < 0.01) compared with the IL-2 control. All groups received 90,000 IU IL-2 i.p. for four consecutive days beginning at the time of adoptive transfer. The number of transferred TE is presented in the figure. Mice were sacrificed 14 days following tumor inoculation and the number of pulmonary metastases. Data are representative of two experiments.

PKO/GKO TE mediate tumor regression in PKO- and GKO-recipient animals

The adoptive transfer studies noted above used wt mice as the tumor-bearing recipient animals. To determine whether the production of perforin or IFN-γ by the tumor-bearing recipient animal was responsible for the antitumor effects mediated by the PKO/GKO TE, 70 × 106 PKO/GKO TE were adoptively transferred into wt, PKO, and GKO mice bearing 3-day established D5 pulmonary metastases. The lack of either perforin or IFN-γ in the recipient animal did not compromise the efficacy of this high dose (70 × 106) of PKO/GKO TE; the antitumor effects were comparable regardless of the recipient animals phenotype (Table II⇓). Thus, PKO/GKO TE can mediate effective tumor regression even when the recipient animal cannot produce either perforin or IFN-γ.

View this table:
Table II.

Efficacy of PKO/GKO TE is not impaired after adoptive transfer into PKO- and GKO-recipient mice

PKO/GKO TE mediate tumor regression and cure PKO/GKO recipients

To rule out the possibility that therapeutic activity of PKO/GKO TE depends on the expression of either IFN-γ or perforin by the tumor-bearing recipients, we examined their ability to mediate tumor regression in PKO/GKO recipients. Although PKO/GKO TE were less effective than wt TE (†, p < 0.05), they still mediated significant (*p < 0.01) tumor regression (Fig. 4⇓), and provided a survival advantage (*p < 0.0001) compared with treatment with IL-2 alone (Fig. 5⇓) in PKO/GKO-recipient animals. However, only 12% of animals that received 70 × 106 PKO/GKO TE survived >120 days. The survival benefit was improved significantly to 40% (§, p < 0.01) by doubling the number of PKO/GKO TE transferred (Fig. 5⇓). This shows that TE can use mechanism(s), other than perforin or IFN-γ, to destroy tumor cells, extend survival, and cure animals with systemic disease. At the same time it is also clear that the therapeutic efficacy of PKO/GKO TE is significantly (†, p < 0.01) reduced compared with wt TE (Fig. 5⇓). Because the transfer of PKO/GKO TE into wt, PKO, and GKO recipients bearing established tumor (Table II⇑) resulted in similar antitumor efficacy, but the transfer of PKO/GKO TE into PKO/GKO recipients did not, it suggests that in the complete absence of both perforin and IFN-γ the immune system is less able to “compensate” effectively to mediate regression of pulmonary metastases. This suggests that perforin and IFN-γ compensate or are redundant for one another, directly or indirectly, in facilitating tumor regression in this model.

FIGURE 4.
FIGURE 4.

PKO/GKO TE mediate tumor regression in PKO/GKO recipients. TE from PKO/GKO or wt mice (35 or 70 × 106) were transferred into PKO/GKO-recipient mice with 3-day established pulmonary D5 metastases. All animals received 90,000 IU IL-2 i.p. for four consecutive days beginning at the time of adoptive transfer. Animals were sacrificed on day 14 after tumor inoculation and the number of pulmonary metastases was determined. Although PKO/GKO and wt TE mediated significant (∗, p < 0.01) tumor regression, the transfer of wt TE cells was significantly more therapeutic (†, p < 0.05) than the transfer of PKO/GKO TE. Three consecutive experiments are presented.

FIGURE 5.
FIGURE 5.

Adoptive transfer of PKO/GKO TE can cure PKO/GKO mice bearing systemic tumor. PKO/GKO mice bearing 3-day established D5 pulmonary metastases were treated with 70 × 106 wt TE (▪, n = 25), 70 × 106 PKO/GKO TE (▴, n = 26), or 140 × 106 PKO/GKO TE (○, n = 10). All groups received IL-2 as described before, the control group received IL-2 i.p. alone ( Embedded Image, n = 26). All groups receiving TE showed significantly increased (∗, p < 0.0001) survival compared with the IL-2 alone control group. Survival was significantly improved (†, p < 0.01) by transferring 140 × 106 PKO/GKO TE as compared with 70 × 106 PKO/GKO TE. But survival mediated by adoptive transfer of both PKO/GKO TE doses was still significantly decreased (§, p < 0.01) as compared with the survival after adoptive transfer of wt TE. Statistical differences were assessed via the log rank sum test.

PKO/GKO TE produce tumor-specific TNF-α

Because PKO/GKO TE have significant antitumor activity, we sought to characterize the effector mechanism(s) in these cells. Because IFN-γ is a type 1 cytokine and plays a role in attenuating type 2 cytokine responses, PKO/GKO TE might exhibit a tumor-specific type 2 cytokine profile. Therefore, simultaneous with the adoptive transfer studies, PKO/GKO TE were also examined for their capacity to release cytokines in response to tumor cells. Although PKO/GKO TE released IL-4, IL-5, and IL-10 when stimulated with anti-CD3 in 24-h assays, there was no release of these type 2 cytokines when they were stimulated with D5 or control tumor cells (data not shown). Thus, the priming of an immune response in the absence of perforin and IFN-γ did not lead to the development of a tumor-specific type 2 cytokine response. Therefore, it is unlikely that tumor regression was mediated by these cytokines.

Because Barth et al. (17) reported that both IFN-γ and TNF-α had a key role in T cell-mediated tumor regression, we examined whether PKO/GKO TE produced TNF-α in response to stimulation with specific tumor. When PKO/GKO TE, generated from D5-G6 TVDLN were stimulated with D5 tumor, no TNF-α was detected in the culture supernatant in our standard 24-h ELISA (data not shown). Thinking that we might have missed the period when TNF-α was secreted, a time course was performed and PKO/GKO TE assayed for TNF-α secretion 6, 12, and 24 h following stimulation. Again, although TNF-α was secreted in response to stimulation with anti-CD3, no secretion was detected following stimulation with D5 or MCA-310 (data not shown). However, interrogation by RT-PCR of the same PKO/GKO TE revealed tumor-specific up-regulation of message for TNF-α (data not shown). We considered that TNF-α produced in response to stimulation with specific tumor might be membrane-bound, consumed or bound by tumor stimulators and not be available for detection in the culture supernatant. To test this possibility we used intracellular cytokine analysis to assay PKO/GKO TE. Results of these studies identified a population of CD8+ T cells that up-regulated expression of TNF-α in response to stimulation with D5 (Fig. 6⇓A). In four consecutive experiments, a similar pattern of tumor-specific TNF-α expression was seen for both PKO/GKO and wt TE (Fig. 6⇓B). Although PKO/GKO TE appeared to have a lower frequency of tumor-specific TNF-α-expressing T cells than wt TE, the difference was not statistically significant. Simultaneously with the monitoring of TNF-α production, the frequency of tumor-specific IFN-γ-producing cells was also assessed. Fig. 6⇓B shows that the frequency of tumor-specific TNF-α- and IFN-γ-producing wt TE was similar. As expected, there was no detection of IFN-γ by PKO/GKO TE. These studies suggest that tumor-specific expression of TNF-α provides a potential effector molecule that may be responsible for the antitumor effects mediated by PKO/GKO TE.

FIGURE 6.
FIGURE 6.

A and B, PKO/GKO TE produce tumor-specific TNF-α and lack IFN-γ production. A, Both PKO/GKO and wt TE cells generated from D5-G6 TVDLN exhibit D5 tumor-specific TNF-α expression, while only wt TE exhibit complimentary tumor-specific IFN-γ. PKO/GKO or wt TE were stimulated for 12 h in CM alone (None), with D5, MCA-310, or with immobilized anti-CD3 mAb. The percentage of CD8+/CD3+/TNF-α+ TE, or CD8+/CD3+/IFN-γ+ TE, respectively, over the total number of CD8+/CD3+ TE (gated 50,000 events) is presented in the upper right quadrant. Data are representative of four independent experiments. B, The percentage of tumor-specific CD8+/CD3+/TNF-α+ TE, or CD8+/CD3+/IFN-γ+ TE, respectively, over the total number of CD8+/CD3+ TE in PKO/GKO TE (▪) and wt TE (□) is presented as the mean ± SEM for four independent experiments. Stimulation with anti-CD3 mAb served as a positive control for intracellular staining and is presented using a different scale.

PKO/GKO TE-mediated tumor-specific cytotoxicity is blocked by TNFRII:Fc

Although PKO/GKO TE are not cytolytic in 6-h 51Cr-release assays, long-term 24-h cytotoxicity assays were developed to investigate whether PKO/GKO TE might mediate TNF-dependent D5-specific killing. Microscopic evaluation of tumor cell-TE cultures revealed tumor-specific killing in 24-h assays (Fig. 7⇓A). As an additional specificity control, naive wt and PKO/GKO lymph nodes were harvested, activated with anti-CD3 mAb, expanded in IL-2, and used in parallel with the D5-specific TE generated from D5-G6 TVDLN. Anti-CD3/IL-2-activated “effector” T cells from naive lymph nodes (not tumor-primed) did not exhibit cytotoxicity for D5 or MCA-310, further supporting the specificity of TE generated from TVDLN of PKO/GKO mice. Although it is possible that several TNF family members might be responsible for the cytotoxicity seen in these assays, previous studies with the D5 melanoma document its resistance to killing by FasL, anti-Fas, or TRAIL (20), (H. Winter, unpublished observation). To directly address whether TNF was responsible for the tumor lysis seen in this assay, TNF was neutralized by adding soluble TNFRII:Fc to the cytotoxicity assays. PKO/GKO TE appeared less cytotoxic than wt TE, but both specifically killed D5 tumor cells (∗, p < 0.05) and not the MCA-310 control tumor cells (Fig. 7⇓B). There was in fact no significant difference in the level of D5-specific cytotoxicity mediated by PKO/GKO and wt TE. The addition of soluble TNFRII:Fc to these assays diminished significantly (∗, p < 0.05) the killing of D5 by PKO/GKO TE, but failed to inhibit the tumor-specific killing of wt TE (Fig. 7⇓B). Because soluble TNFRII:Fc can bind both TNF-α and lymphotoxin (LT)-α (TNF-β) we were unable to confirm which molecule was responsible. Nonetheless, these results document that tumor-specific TE generated from PKO/GKO mice maintain an in vitro cytolytic function that is detectable at 24 h and is mediated by TNF.

FIGURE 7.
FIGURE 7.

A and B, PKO/GKO TE exhibit tumor-specific killing in 24 h in vitro assays. A, Microscopic analysis of fixed and stained tumor cells cultured alone or with TE generated from D5-G6 TVDLN or naive lymph nodes are presented. Pictures shown are representative of triplicate cultures in two independent experiments. Tumor-specific cytotoxicity of PKO/GKO TE cells can be blocked by soluble TNFRII: Fc. B, PKO/GKO TE or wt TE were incubated for 24 h with D5 or MCA-310 tumor cells and the percentage of killed tumor cells was evaluated. PKO/GKO and wt TE mediate significant killing of D5 (∗, p < 0.05) compared with that observed for MCA-310. There is no significant difference between PKO/GKO- and wt TE-mediated D5 killing. The addition of 1 μg/ml TNFRII:Fc at the beginning of the experiment did not block cytotoxicity of wt TE, but eliminated completely PKO/GKO TE-mediated D5 killing. Data are presented as the mean ± SEM of triplicate determinations and is representative of two independent experiments.

PKO/GKO TE-mediated tumor regression is blocked by TNFRII:Fc administration

To determine whether TNF played a critical role in tumor regression mediated by PKO/GKO TE, soluble TNFRII:Fc (14) was administered i.p. to mice following the adoptive transfer of wt or PKO/GKO TE. In three of three experiments, treatment with soluble TNFRII:Fc significantly (∗, p < 0.01) reduced the therapeutic efficacy of PKO/GKO TE, while administration of human IgG (control) had no effect on TE function (two of two experiments). In contrast, treatment with TNFRII:Fc did not reduce the efficacy of wt TE (Fig. 8⇓). These results imply that in the absence of perforin and IFN-γ, TNF, either α and/or β, is the major mechanism for T cell-mediated tumor regression. The undisturbed efficacy of wt TE in animals treated with TNFRII:Fc also demonstrates the importance of perforin and/or IFN-γ in wt animals and further illustrates a redundant, but limited, repertoire of effector mechanisms available for T cells to eliminate established pulmonary metastases.

FIGURE 8.
FIGURE 8.

Tumor regression mediated by PKO/GKO TE cells can be blocked by soluble TNFRII:Fc. In vivo administration of soluble TNFRII:Fc significantly blocks (∗, p < 0.01) the therapeutic efficacy of 70 × 106 adoptively transferred PKO/GKO TE in mice bearing 3-day established systemic D5 tumor, identifying TNF as a critical in vivo effector mechanism. In contrast, the therapeutic efficacy of wt TE cells is not diminished by the fusion protein treatment. All animals received 90,000 IU IL-2 i.p. q.d. times four. On days 1 and 3 after adoptive transfer, specified groups received i.p. 250 μg of soluble TNFRII:Fc or human IgG (control). Animals receiving IL-2 alone served as controls; all animals were sacrificed on day 14 after tumor inoculation and the number of D5 pulmonary metastases was determined.

Discussion

The adoptive transfer of tumor-specific CD8+ T cells has been shown to mediate significant antitumor effects and cure animals of established metastases in numerous murine models. Considerable effort has been directed at dissecting the mechanisms of T cell-mediated tumor regression in these models in the hope that the information will lead to improved treatment for patients with cancer. Recent studies have documented that established pulmonary and intracranial metastases can be treated effectively by T cells that lack either perforin, FasL, or IFN-γ (20, 21, 22, 35). Furthermore, most studies reported efficacy in host animals that were deficient in the same molecule, thereby eliminating a requirement for that effector mechanism in T cell-mediated destruction of pulmonary or intracranial tumor. Because these studies demonstrated that there was not an absolute requirement for either perforin, FasL, or IFN-γ, three dominant effector systems of CD8+ T cells, it suggests there is redundancy in the effector mechanisms that can be mobilized by CD8+ T cells to destroy tumor cells in vivo.

This study also identified the importance of perforin or IFN-γ expression by cells of the tumor-bearing recipient in the antitumor effect mediated by adoptively transferred PKO/GKO TE. Although 70 × 106 D5-specific PKO/GKO TE were highly effective at eliminating pulmonary metastases in tumor-bearing wt, PKO, or GKO mice (Table I⇑), the transfer of the same number of PKO/GKO TE into tumor-bearing PKO/GKO mice rarely provided complete clearing of the lungs (1 of 15 mice, Fig. 4⇑, as compared with 11 of 15 mice, Fig. 1⇑). Similarly, in survival studies, the transfer of 140 × 106 D5-specific PKO/GKO TE was required to cure 40% of day 3 tumor-bearing PKO/GKO recipients, compared with 84% survival when 70 × 106 wt TE were used to treat the same recipients. In contrast, 70 × 106 PKO/GKO TE were about twice as effective in tumor-bearing wt mice, mediating long-term survival and curing 50% of treated animals (Fig. 2⇑). These results document that in addition to the effector mechanism(s) supplied by the transferred TE, the recipient mice still use either perforin or IFN-γ-based mechanisms to amplify the effects of the tumor-specific T cells. The observation that IFN-γ plays a role when perforin is absent (PKO/GKO TE in PKO tumor-bearing recipients, Table II⇑) is not unexpected given its direct antitumor effects, its function in amplifying the innate immune response and in augmenting Ag processing (30, 39). However, the significant role for perforin, when IFN-γ is absent (PKO/GKO TE in GKO tumor-bearing recipients), is somewhat surprising given that we have not observed substantial infiltrates of NK cells into the lungs of treated animals (35) (C. H. Poehlein, unpublished observation in PKO/GKO mice). This result hints that the endogenous host CD8+ T cell response is playing a bigger role than expected when the transferred TE are deficient in perforin and IFN-γ. Others have reported a requirement for host lymphocytes in T cell-mediated tumor regression following adoptive immunotherapy (40, 41). However, it is more generally reported that the adoptive transfer of tumor-specific CD8+ TE and IL-2 can mediate regression in animals depleted of endogenous lymphocytes (1, 42, 43, 44, 45).

These seemingly conflicting results may be explained by the redundancy of the immune system, which has a broad repertoire of effector mechanisms available to the transferred TE. TE with a high frequency of T cells exhibiting multiple destructive mechanisms would be expected to mediate tumor regression in the absence of a host component, whereas TE populations with a low frequency of tumor-reactive cells and/or those lacking robust effector mechanisms would require the “support” of host cells (46). This reasoning may also explain some of the differences reported by investigators studying antitumor properties of T cells in gene KO mice. For instance, Prevost-Blondel et al. (47) showed that IFN-γ was essential in their model, as TCR-transgenic T cells from IFN-γ single KO mice are unable to mediate regression in IFN-γ KO mice. Although we observed that adoptive transfer of in vitro-activated and -expanded GKO TE cured GKO mice bearing established pulmonary metastases and subsequently protected them from a subsequent s.c. tumor challenge (35).

One explanation for this difference may be the vaccination and activation process used to generate TE in our studies. We have previously shown that vaccination with a GM-CSF-secreting tumor vaccine effectively primes therapeutic CD8+ TE in CD4-depleted and MHC class II KO animals (48). Thus, it seems possible that such a strong vaccine induces a cytokine/costimulatory milieu that drives T cells more effectively and/or with a broader spectrum of effector mechanisms. Hence, the failure of tumor-specific GKO T cells to mediate tumor regression may have more to do with the frequency of primed T cells, their environment during activation and expansion, and/or their available repertoire of effector mechanisms than with an inherent deficiency associated with the loss of a single effector mechanism.

In contrast to our results, which demonstrate that PKO/GKO mice generate a reduced therapeutic antitumor response to vaccination with D5-G6, Badovinac and Harty (49) showed that the immune response to vaccination with Listeria was stronger and more therapeutic in PKO/GKO mice (49, 50). In their study, the significantly heightened protection against a Listeria challenge correlated with the frequency of Listeria monocytogenes Ag-specific T cells, which was up to 4.5-fold higher in PKO/GKO than in wt animals. Consistent with our observation of a reduced therapeutic response, D5-G6-vaccinated PKO/GKO mice had a lower frequency of tumor-specific CD8+ TNF-α+ T cells than vaccinated wt mice, but this difference was not significant (Fig. 6⇑B). Given that both attenuated Listeria and GM-CSF-secreting tumor vaccines are highly effective at priming protective immune responses in their respective models, it is unclear why there is such a striking difference in the frequency of responding T cells in these two models. Particularly, why does the TNF-α response of PKO/GKO cells increase for Listeria and not for D5-G6? Is this an inherent advantage of using a Listeria-based vector to prime an immune response? The difference may be due to L. monocytogenes being a foreign bacterial Ag or contributing some adjuvant property that augments the TNF-α response. Then again, the inherent difference between PKO/GKO mice on a C57BL/6 vs a BALB/c background may explain the observed differences. Alternatively, the difference between these two models may be related to the in vitro activation step that is required to mature fresh TVDLN to functional TE. In this case, the activation process may lead to an increased activation-induced cell death and loss of some of the tumor-specific T cells. Although we did not observe decreased recovery of PKO/GKO TE following activation with anti-CD3 and IL-2 it is still possible that there was a selective loss of Ag-specific TE during the activation step.

Our in vitro experiments imply that TNF-α played an important role in the tumor-specific cytotoxicity of PKO/GKO TE. This observation is consistent with reports of others showing that effector cells that lack perforin and/or FasL mediate potent Ag-specific function in vitro via TNF-α. Lee et al. (51) reported that lymphokine-activated killer cells generated from PKO/gld double KO mice mediated cytotoxic activity that could be blocked by anti-TNF Ab. Similarly, Prevost-Blondel et al. (47) recently showed a crucial role for TNF-α in a gp33-transfected 3LL-A9 lung cancer model. In their studies, perforin-deficient TCR transgenic T cells that recognize a lymphocytic choriomeningitis virus peptide Ag stably expressed on a transfected Lewis lung carcinoma cell line (3LL), mediate tumor cell lysis and tumor regression, but loose efficacy in vivo in the absence of TNF-α. Additionally, Seki et al. (22) demonstrated efficient perforin-independent antitumor efficacy of adoptively transferred CTL. In their model, blocking TNF-α reduced CTL-mediated cytotoxicity in vitro.

A critical role for TNF in animal models of autoimmunity and infectious disease has also been identified: T cell-induced acute diabetes occurs independently of either perforin or functional FasL, while animals deficient in TNFRII expression do not develop diabetes (52). Furthermore, in studies using perforin/IFN-γ double KO BALB/c mice, Harty and colleague (53) identified an important role for TNF in T cell-mediated immune responses to L. monocytogenes. They also showed that the immune response to vaccination with Listeria was stronger and more therapeutic in perforin/IFN-γ double KO compared with wt mice (49, 50). This is in contrast to our results, where the absence of perforin and IFN-γ resulted in T cells with reduced capacity to mediate tumor regression in both active specific and adoptive immunotherapy studies. Additional studies identified that perforin-independent immunity to L. monocytogenes requires CD8+ T cell-derived TNF-α (53) and is independent of Fas/FasL (53, 54). Although we cannot directly exclude Fas/FasL (CD95/CD95L) interactions in our PKO/GKO model, we have previously demonstrated that D5 tumor cells are resistant to apoptosis induced by anti-Fas mAb in vitro and that TE with a functional deficiency of FasL can still effectively cure animals with established tumors (20).

Our hypothesis that TNF plays a critical role in the tumor regression mediated by PKO/GKO TE is consistent with our data and the in vitro findings of others. However, a number of questions remain: although TNF is essential in this model, is it TNF-α and/or TNF-β (LT-α)? Because both molecules are bound by the soluble TNFRII:Fc, it is not possible to determine the answer from these experiments. Additionally, while TE-mediated in vitro cytotoxicity is effectively blocked by soluble TNFRII:Fc, the D5 tumor is resistant to cytotoxicity by rTNF-α (H. Winter, unpublished observation). It is unknown whether this difference is related to the relative effectiveness of membrane bound vs soluble TNF-α or whether LT-α (55) is the mediator. Second, is it the TNF produced by the transferred TE and/or TNF from the host that is essential for T cell-mediated regression of pulmonary metastases? The answer to these questions requires the generation of mice deficient in perforin, IFN-γ, and TNF-α or LT-α. Current efforts are directed toward addressing these questions.

Our current study clearly identifies TNF as one of three primary effector mechanisms, together with perforin and IFN-γ, for T cell-mediated tumor regression. Although these results do not exclude participation of additional effector molecules, in the absence of perforin, IFN-γ, and TNF, they are insufficient to mediate effective tumor regression in our model. The identification of three central effector mechanisms provides an important characterization of T cells with therapeutic capacity and has repercussions for immunological monitoring and development of surrogate endpoints for patients on cancer vaccine trials.

Acknowledgments

We thank Patricia Ruane for the excellent care of animals used in this study. We thank Drs. Andrew Weinberg, Suyu Shu, and Edwin Walker for their comments and assistance during the process of this study and Daniel P. Healey for excellent support in developing the intracellular flow cytometric analysis.

Footnotes

  • ↵1 This work was supported by National Institutes of Health Grants 1RO1CA80964-03 (to B.A.F.), 1RO1CA92254-01 (to B.A.F.), the M. J. Murdock Charitable Trust, and the Chiles Foundation. C.H.P. was a Chiles Foundation visiting fellow.

  • ↵2 Address correspondence and reprint requests to Dr. Bernard A. Fox, Laboratory of Molecular and Tumor Immunology, Earle A. Chiles Research Institute, 4805 NE Glisan Street, Portland, OR 97213. E-mail address: foxb{at}foxlab.org

  • ↵3 Abbreviations used in this paper: TE, effector T cell; KO, knockout; PKO, perforin KO; FasL, Fas ligand; GKO, IFN-γ KO; GRKO, IFN-γ receptor KO; wt, wild type; CM, complete medium; TVDLN, tumor vaccine draining lymph node; LT, lymphotoxin.

  • Received September 12, 2002.
  • Accepted December 6, 2002.

References

  1. ↵
    Cheever, M. A., P. D. Greenberg, A. Fefer. 1981. Adoptive therapy of established syngeneic leukemia by cells primarily sensitized in vitro. Cancer Res. 41: 2658
    OpenUrlAbstract/FREE Full Text
  2. ↵
    Rosenberg, S. A., M. T. Lotze, L. M. Muul, S. Leitman, A. E. Chang, J. T. Vetto, C. A. Seipp, C. Simpson. 1986. A new approach to the therapy of cancer based on the systemic administration of autologous lymphokine-activated killer cells and recombinant interleukin-2. Surgery 100: 262
    OpenUrlPubMed
  3. ↵
    Rosenberg, S. A., B. S. Packard, P. M. Aebersold, D. Solomon, S. L. Topalian, S. T. Toy, P. Simon, M. T. Lotze, J. C. Yang, C. A. Seipp, et al 1988. Use of tumor-infiltrating lymphocytes and interleukin-2 in the immunotherapy of patients with metastatic melanoma: a preliminary report. N. Engl. J. Med. 319: 1676
    OpenUrlCrossRefPubMed
  4. ↵
    Chang, A. E., A. Aruga, M. J. Cameron, V. K. Sondak, D. P. Normolle, B. A. Fox, S. Shu. 1997. Adoptive immunotherapy with vaccine-primed lymph node cells secondarily activated with anti-CD3 and IL-2. J. Clin. Oncol. 15: 796
    OpenUrlAbstract/FREE Full Text
  5. ↵
    Yoshizawa, H., A. E. Chang, S. Shu. 1991. Specific adoptive immunotherapy mediated by tumor-draining lymph node cells sequentially activated with anti-CD3 and IL-2. J. Immunol. 147: 729
    OpenUrlAbstract
  6. ↵
    Rosenau, W., H. D. Moon. 1966. Studies on the mechanism of the cytolytic effect of sensitized lymphocytes. J. Immunol. 96: 80
    OpenUrlAbstract/FREE Full Text
  7. ↵
    Kedar, E., M. Schwartzbach, Z. Raanan, S. Hefetz. 1977. In vitro induction of cell-mediated immunity to murine leukemia cells. II. Cytotoxic activity in vitro and tumor-neutralizing capacity in vivo of anti-leukemia cytotoxic lymphocytes generated in macrocultures. J. Immunol. Methods 16: 39
    OpenUrlCrossRefPubMed
  8. ↵
    Chauvenet, P. H., C. P. McArthur, R. T. Smith. 1979. Demonstration in vitro of cytotoxic T cells with apparent specificity toward tumor-specific transplantation antigens on chemically induced tumors. J. Immunol. 123: 2575
    OpenUrlAbstract/FREE Full Text
  9. ↵
    Fernandez-Cruz, E., B. Halliburton, J. D. Feldman. 1979. In vivo elimination by specific effector cells of an established syngeneic rat Moloney virus-induced sarcoma. J. Immunol. 123: 1772
    OpenUrlAbstract/FREE Full Text
  10. ↵
    Barth, R. J., Jr., S. N. Bock, J. J. Mule, S. A. Rosenberg. 1990. Unique murine tumor-associated antigens identified by tumor infiltrating lymphocytes. J. Immunol. 144: 1531
    OpenUrlAbstract
  11. ↵
    Dailey, M. O., E. Pillemer, I. L. Weissman. 1982. Protection against syngeneic lymphoma by a long-lived cytotoxic T-cell clone. Proc. Natl. Acad. Sci. USA 79: 5384
    OpenUrlAbstract/FREE Full Text
  12. ↵
    Rosenstein, M., S. A. Rosenberg. 1984. Generation of lytic and proliferative lymphoid clones to syngeneic tumor: in vitro and in vivo studies. J. Natl. Cancer Inst. 72: 1161
    OpenUrlPubMed
  13. ↵
    Greenberg, P. D.. 1986. Therapy of murine leukemia with cyclophosphamide and immune Lyt2+ cells: cytolytic T cells can mediate eradication of disseminated leukemia. J. Immunol. 136: 1917
    OpenUrlAbstract
  14. ↵
    Yoshizawa, H., A. E. Chang, S. Y. Shu. 1992. Cellular interactions in effector cell generation and tumor regression mediated by anti-CD3/interleukin 2-activated tumor-draining lymph node cells. Cancer Res. 52: 1129
    OpenUrlAbstract/FREE Full Text
  15. ↵
    Yamasaki, T., H. Handa, J. Yamashita, Y. Watanabe, Y. Namba, M. Hanaoka. 1984. Specific adoptive immunotherapy with tumor-specific cytotoxic T-lymphocyte clone for murine malignant gliomas. Cancer Res. 44: 1776
    OpenUrlAbstract/FREE Full Text
  16. ↵
    Spiess, P. J., J. C. Yang, S. A. Rosenberg. 1987. In vivo antitumor activity of tumor-infiltrating lymphocytes expanded in recombinant interleukin-2. J. Natl. Cancer Inst. 79: 1067
    OpenUrlPubMed
  17. ↵
    Barth, R. J., Jr., J. J. Mule, P. J. Spiess, S. A. Rosenberg. 1991. Interferon γ and tumor necrosis factor have a role in tumor regressions mediated by murine CD8+ tumor-infiltrating lymphocytes. J. Exp. Med. 173: 647
    OpenUrlAbstract/FREE Full Text
  18. ↵
    Shu, S., R. A. Krinock, T. Matsumura, J. J. Sussman, B. A. Fox, A. E. Chang, D. S. Terman. 1994. Stimulation of tumor-draining lymph node cells with superantigenic staphylococcal toxins leads to the generation of tumor-specific effector T cells. J. Immunol. 152: 1277
    OpenUrlAbstract
  19. ↵
    Matsumura, T., R. A. Krinock, A. E. Chang, S. Shu. 1993. Cross-reactivity of anti-CD3/IL-2 activated effector cells derived from lymph nodes draining heterologous clones of a murine tumor. Cancer Res. 53: 4315
    OpenUrlAbstract/FREE Full Text
  20. ↵
    Winter, H., H. M. Hu, W. J. Urba, B. A. Fox. 1999. Tumor regression after adoptive transfer of effector T cells is independent of perforin or Fas ligand (APO-1L/CD95L). J. Immunol. 163: 4462
    OpenUrlAbstract/FREE Full Text
  21. ↵
    Peng, L., J. C. Krauss, G. E. Plautz, S. Mukai, S. Shu, P. A. Cohen. 2000. T cell-mediated tumor rejection displays diverse dependence upon perforin and IFN-γ mechanisms that cannot be predicted from in vitro T cell characteristics. J. Immunol. 165: 7116
    OpenUrlAbstract/FREE Full Text
  22. ↵
    Seki, N., A. D. Brooks, C. R. Carter, T. C. Back, E. M. Parsoneault, M. J. Smyth, R. H. Wiltrout, T. J. Sayers. 2002. Tumor-specific CTL kill murine renal cancer cells using both perforin and Fas ligand-mediated lysis in vitro, but cause tumor regression in vivo in the absence of perforin. J. Immunol. 168: 3484
    OpenUrlAbstract/FREE Full Text
  23. ↵
    Hung, K., R. Hayashi, A. Lafond-Walker, C. Lowenstein, D. Pardoll, H. Levitsky. 1998. The central role of CD4+ T cells in the antitumor immune response. J. Exp. Med. 188: 2357
    OpenUrlAbstract/FREE Full Text
  24. ↵
    Fallarino, F., T. F. Gajewski. 1999. Cutting edge: differentiation of antitumor CTL in vivo requires host expression of Stat1. J. Immunol. 163: 4109
    OpenUrlAbstract/FREE Full Text
  25. ↵
    Sgadari, C., J. M. Farber, A. L. Angiolillo, F. Liao, J. Teruya-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
  26. ↵
    Sgadari, C., A. L. Angiolillo, B. W. Cherney, S. E. Pike, J. M. Farber, L. G. Koniaris, P. Vanguri, P. R. Burd, N. Sheikh, G. Gupta, et al 1996. Interferon-inducible protein-10 identified as a mediator of tumor necrosis in vivo. Proc. Natl. Acad. Sci. USA 93: 13791
    OpenUrlAbstract/FREE Full Text
  27. ↵
    Qin, Z., T. Blankenstein. 2000. CD4+ T cell-mediated tumor rejection involves inhibition of angiogenesis that is dependent on IFN γ receptor expression by nonhematopoietic cells. Immunity 12: 677
    OpenUrlCrossRefPubMed
  28. ↵
    Krasagakis, K., C. Garbe, C. C. Zouboulis, C. E. Orfanos. 1995. Growth control of melanoma cells and melanocytes by cytokines. Recent Res. Cancer Res. 139: 169
    OpenUrlCrossRefPubMed
  29. ↵
    Bohm, W., S. Thoma, F. Leithauser, P. Moller, R. Schirmbeck, J. Reimann. 1998. T cell-mediated, IFN-γ-facilitated rejection of murine B16 melanomas. J. Immunol. 161: 897
    OpenUrlAbstract/FREE Full Text
  30. ↵
    Boehm, U., T. Klamp, M. Groot, J. C. Howard. 1997. Cellular responses to interferon-γ. Annu. Rev. Immunol. 15: 749
    OpenUrlCrossRefPubMed
  31. ↵
    Bach, E. A., M. Aguet, R. D. Schreiber. 1997. The IFN γ receptor: a paradigm for cytokine receptor signaling. Annu. Rev. Immunol. 15: 563
    OpenUrlCrossRefPubMed
  32. ↵
    Rosenberg, S. A., J. R. Yannelli, J. C. Yang, S. L. Topalian, D. J. Schwartzentruber, J. S. Weber, D. R. Parkinson, C. A. Seipp, J. H. Einhorn, D. E. White. 1994. Treatment of patients with metastatic melanoma with autologous tumor- infiltrating lymphocytes and interleukin 2. J. Natl. Cancer Inst. 86: 1159
    OpenUrlAbstract/FREE Full Text
  33. ↵
    Restifo, N. P., F. Esquivel, A. L. Asher, H. Stotter, R. J. Barth, J. R. Bennink, J. J. Mule, J. W. Yewdell, S. A. Rosenberg. 1991. Defective presentation of endogenous antigens by a murine sarcoma: implications for the failure of an anti-tumor immune response. J. Immunol. 147: 1453
    OpenUrlAbstract/FREE Full Text
  34. ↵
    Hu, H. M., W. J. Urba, B. A. Fox. 1998. Gene-modified tumor vaccine with therapeutic potential shifts tumor-specific T cell response from a type 2 to a type 1 cytokine profile. J. Immunol. 161: 3033
    OpenUrlAbstract/FREE Full Text
  35. ↵
    Winter, H., H. M. Hu, K. McClain, W. J. Urba, B. A. Fox. 2001. Immunotherapy of melanoma: a dichotomy in the requirement for IFN-γ in vaccine-induced antitumor immunity versus adoptive immunotherapy. J. Immunol. 166: 7370
    OpenUrlAbstract/FREE Full Text
  36. ↵
    Blanckaert, V. D., M. E. Schelling, C. A. Elstad, G. G. Meadows. 1993. Differential growth factor production, secretion, and response by high and low metastatic variants of B16BL6 melanoma. Cancer Res. 53: 4075
    OpenUrlAbstract/FREE Full Text
  37. ↵
    Arca, M. J., J. C. Krauss, A. Aruga, M. J. Cameron, S. Shu, A. E. Chang. 1996. Therapeutic efficacy of T cells derived from lymph nodes draining a poorly immunogenic tumor transduced to secrete granulocyte-macrophage colony-stimulating factor. Cancer Gene Ther. 3: 39
    OpenUrlPubMed
  38. ↵
    Spencer-Green, G.. 2000. Etanercept (Enbrel): update on therapeutic use. Ann. Rheum. Dis. 59: (Suppl. 1):I46
    OpenUrlCrossRefPubMed
  39. ↵
    Kaplan, D. H., V. Shankaran, A. S. Dighe, E. Stockert, M. Aguet, L. J. Old, R. D. Schreiber. 1998. Demonstration of an interferon γ-dependent tumor surveillance system in immunocompetent mice. Proc. Natl. Acad. Sci. USA 95: 7556
    OpenUrlAbstract/FREE Full Text
  40. ↵
    Nagoshi, M., P. S. Goedegebuure, U. L. Burger, N. Sadanaga, M. P. Chang, T. J. Eberlein. 1998. Successful adoptive cellular immunotherapy is dependent on induction of a host immune response triggered by cytokine (IFN-γ and granulocyte/macrophage colony-stimulating factor) producing donor tumor-infiltrating lymphocytes. J. Immunol. 160: 334
    OpenUrlAbstract/FREE Full Text
  41. ↵
    Nagoshi, M., N. Sadanaga, H. G. Joo, P. S. Goedegebuure, T. J. Eberlein. 1999. Tumor-specific cytokine release by donor T cells induces an effective host anti-tumor response through recruitment of host naive antigen presenting cells. Int. J. Cancer 80: 308
    OpenUrlCrossRefPubMed
  42. ↵
    Rosenberg, S. A., P. Spiess, R. Lafreniere. 1986. A new approach to the adoptive immunotherapy of cancer with tumor-infiltrating lymphocytes. Science 233: 1318
    OpenUrlAbstract/FREE Full Text
  43. ↵
    Fox, B. A., P. J. Spiess, A. Kasid, R. Puri, J. J. Mule, J. S. Weber, S. A. Rosenberg. 1990. In vitro and in vivo antitumor properties of a T-cell clone generated from murine tumor-infiltrating lymphocytes. J. Biol. Response Modif. 9: 499
    OpenUrlPubMed
  44. ↵
    Chang, A. E., S. Y. Shu, T. Chou, R. Lafreniere, S. A. Rosenberg. 1986. Differences in the effects of host suppression on the adoptive immunotherapy of subcutaneous and visceral tumors. Cancer Res. 46: 3426
    OpenUrlAbstract/FREE Full Text
  45. ↵
    Alexander, R. B., S. A. Rosenberg. 1990. Long-term survival of adoptively transferred tumor-infiltrating lymphocytes in mice. J. Immunol. 145: 1615
    OpenUrlAbstract
  46. ↵
    Dobrzanski, M. J., J. B. Reome, R. W. Dutton. 2001. Role of effector cell-derived IL-4, IL-5, and perforin in early and late stages of type 2 CD8 effector cell-mediated tumor rejection. J. Immunol. 167: 424
    OpenUrlAbstract/FREE Full Text
  47. ↵
    Prevost-Blondel, A., E. Roth, F. M. Rosenthal, H. Pircher. 2000. Crucial role of TNF-α in CD8 T cell-mediated elimination of 3LL-A9 Lewis lung carcinoma cells in vivo. J. Immunol. 164: 3645
    OpenUrlAbstract/FREE Full Text
  48. ↵
    Hu, H. M., H. Winter, W. J. Urba, B. A. Fox. 2000. Divergent roles for CD4+ T cells in the priming and effector/memory phases of adoptive immunotherapy. J. Immunol. 165: 4246
    OpenUrlAbstract/FREE Full Text
  49. ↵
    Badovinac, V. P., J. T. Harty. 2000. Adaptive immunity and enhanced CD8+ T cell response to Listeria monocytogenes in the absence of perforin and IFN-γ. J. Immunol. 164: 6444
    OpenUrlAbstract/FREE Full Text
  50. ↵
    Badovinac, V. P., A. R. Tvinnereim, J. T. Harty. 2000. Regulation of antigen-specific CD8+ T cell homeostasis by perforin and interferon-γ. Science 290: 1354
    OpenUrlAbstract/FREE Full Text
  51. ↵
    Lee, R. K., J. Spielman, D. Y. Zhao, K. J. Olsen, E. R. Podack. 1996. Perforin, Fas ligand, and tumor necrosis factor are the major cytotoxic molecules used by lymphokine-activated killer cells. J. Immunol. 157: 1919
    OpenUrlAbstract
  52. ↵
    Herrera, P. L., D. M. Harlan, P. Vassalli. 2000. A mouse CD8 T cell-mediated acute autoimmune diabetes independent of the perforin and Fas cytotoxic pathways: possible role of membrane TNF. Proc. Natl. Acad. Sci. USA 97: 279
    OpenUrlAbstract/FREE Full Text
  53. ↵
    White, D. W., J. T. Harty. 1998. Perforin-deficient CD8+ T cells provide immunity to Listeria monocytogenes by a mechanism that is independent of CD95 and IFN-γ but requires TNF-α. J. Immunol. 160: 898
    OpenUrlAbstract/FREE Full Text
  54. ↵
    Ware, C. F., P. D. Crowe, T. L. Vanarsdale, J. L. Andrews, M. H. Grayson, R. Jerzy, C. A. Smith, R. G. Goodwin. 1991. Tumor necrosis factor (TNF) receptor expression in T lymphocytes: differential regulation of the type I TNF receptor during activation of resting and effector T cells. J. Immunol. 147: 4229
    OpenUrlAbstract
  55. ↵
    Gray, P. W., B. B. Aggarwal, C. V. Benton, T. S. Bringman, W. J. Henzel, J. A. Jarrett, D. W. Leung, B. Moffat, P. Ng, L. P. Svedersky, et al 1984. Cloning and expression of cDNA for human lymphotoxin, a lymphokine with tumour necrosis activity. Nature 312: 721
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section