Cancer Immunology, Immunotherapy

, Volume 62, Issue 2, pp 371–382

The role of gamma interferon in DNA vaccine-induced tumor immunity targeting simian virus 40 large tumor antigen

Authors

  • Joel F. Aldrich
    • Department of Immunology and Molecular MicrobiologyTexas Tech University Health Sciences Center
  • Michael H. Shearer
    • Department of Immunology and Molecular MicrobiologyTexas Tech University Health Sciences Center
  • Devin B. Lowe
    • Department of Immunology and Molecular MicrobiologyTexas Tech University Health Sciences Center
    • Department of DermatologyUniversity of Pittsburgh School of Medicine
  • Richard E. Winn
    • Department of Immunology and Molecular MicrobiologyTexas Tech University Health Sciences Center
    • Department of Internal MedicineTexas Tech University Health Sciences Center
  • Cynthia A. Jumper
    • Department of Immunology and Molecular MicrobiologyTexas Tech University Health Sciences Center
    • Department of Internal MedicineTexas Tech University Health Sciences Center
  • Ronald C. Kennedy
    • Department of Immunology and Molecular MicrobiologyTexas Tech University Health Sciences Center
    • Department of Internal MedicineTexas Tech University Health Sciences Center
    • Department of Immunology and Molecular MicrobiologyTexas Tech University Health Sciences Center
Original Article

DOI: 10.1007/s00262-012-1338-x

Cite this article as:
Aldrich, J.F., Shearer, M.H., Lowe, D.B. et al. Cancer Immunol Immunother (2013) 62: 371. doi:10.1007/s00262-012-1338-x
  • 186 Views

Abstract

The central role of CD4+ T lymphocytes in mediating DNA vaccine-induced tumor immunity against the viral oncoprotein simian virus 40 (SV40) large tumor antigen (Tag) has previously been described by our laboratory. In the present study, we extend our previous findings by examining the roles of IFN-γ and Th1-associated effector cells within the context of DNA immunization in a murine model of pulmonary metastasis. Immunization of BALB/c mice with plasmid DNA encoding SV40 Tag (pCMV-Tag) generated IFN-γ-secreting T lymphocytes that produced this cytokine upon in vitro stimulation with mKSA tumor cells. The role of IFN-γ as a mediator of protection against mKSA tumor development was assessed via in vivo IFN-γ neutralization, and these experiments demonstrated a requirement for this cytokine in the induction immune phase. Neutralization of IFN-γ was associated with a reduction in Th1 cytokine-producing CD4+ and CD8+ splenocytes, as assessed by flow cytometry analysis, and provided further evidence for the role of CD4+ T lymphocytes as drivers of the cellular immune response. Depletion of NK cells and CD8+ T lymphocytes demonstrated the expendability of these cell types individually, but showed a requirement for a resident cytotoxic cell population within the immune effector phase. Our findings demonstrate the importance of IFN-γ in the induction of protective immunity stimulated by pCMV-Tag DNA-based vaccine and help to clarify the general mechanisms by which DNA vaccines trigger immunity to tumor cells.

Keywords

DNA vaccineTumor immunologyIFN-γSV40 large tumor antigen

Introduction

Active immunization strategies have contributed significantly to the development of improved methods for cancer treatment and prevention [1, 2]. The successes of prophylactic hepatitis B virus and human papilloma virus vaccines in controlling the incidence of virally associated hepatocellular carcinoma and cervical carcinoma, respectively, have demonstrated the utility of targeting viral antigens in the prevention of cancer [35]. Additionally, the recent approval of Sipuleucel-T immunotherapy has expanded the role of cancer vaccines to include late-stage treatment options for metastatic, castrate-resistant prostate cancer [6, 7]. Given these achievements in treating and preventing malignancies, several next-generation cancer vaccines, including various nucleic acid-based vaccines, are currently under investigation in both laboratory and clinical settings [812]. Although DNA immunization approaches have characteristically suffered from low immunogenicity in human studies, the recent licensure of a DNA vaccine encoding xenogenic (human) tyrosinase for the treatment of canine oral melanoma has placed considerable promise on the further expansion of this vaccine modality [13] and has inspired correlative clinical trials in human patients [14, 15]. As DNA vaccines continue to be explored within the context of tumor immunity, efforts to maximize vaccine efficacy will likely require a thorough understanding of the antitumor immune mechanisms underlying DNA vaccine function.

The use of simian virus 40 large tumor antigen (SV40 Tag) as a representative tumor-associated antigen (TAA) for studies of vaccine-induced tumor immunity in a murine model of pulmonary metastasis has previously been described by our laboratory [16, 17]. Studies with recombinant SV40 Tag (rTag) vaccination have demonstrated the importance of CD4+ T lymphocytes in the induction immune phase and the required activities of multiple immune components (e.g., CD8+ T lymphocytes, natural killer (NK) cells, and antibody) in the effector immune phase of tumor immunity against the SV40 Tag-expressing tumor cell line, mKSA [1820]. Interestingly, a parallel vaccination scheme with plasmid DNA encoding SV40 Tag (pCMV-Tag) revealed a central role for CD4+ T lymphocytes in both the induction and effector immune phases, and nonessential roles for CD8+ T lymphocytes and anti-Tag antibodies in these processes [21]. Upon characterization of the humoral immune response, vaccination with pCMV-Tag was shown to generate higher yields of the IgG2a antibody subtype than previous vaccination schemes with rTag [18], possibly indicating a more prevalent role for gamma interferon (IFN-γ) and/or Th1 responses in DNA vaccine-induced tumor immunity. Our collective findings thus suggest that the dominant antitumor immune mechanisms may vary with vaccine modality, and provide an impetus for the further evaluation of IFN-γ and Th1 immune responses associated with DNA vaccination.

IFN-γ is a pleiotropic cytokine that plays a critical role in immunity to many microbes and tumors, but may also contribute to the pathology associated with both protective responses and autoimmune disorders [22]. Signaling through the IFN-γ receptor (IFNγR) engages the JAK/STAT pathway of transcription, initiating diverse biological processes that may include upregulation of MHC class I and class II molecules, increased antigen processing and presentation, differentiation of the Th1 subset of T lymphocytes, ‘classical’ activation of macrophages, and isotype switching to specific IgG subtypes in B cells. Within the context of tumor immunity, IFN-γ is generally thought to play a protective role, as indicated by early studies examining the development of spontaneous and carcinogen-induced tumors in immunologically impaired knockout mice [23, 24]. Alternatively, IFN-γ may enhance immune evasion in some scenarios of tumor immunity and has been reported to promote the expansion of aggressive phenotypes in certain human tumor cells [25]. Although IFN-γ is frequently linked to the efficacy of tumor immunotherapy and vaccination procedures, relatively few studies have directly investigated the role of this cytokine as a mediator of in vivo protection [26, 27].

In this study, we explore the contributions of IFN-γ to DNA vaccine-induced tumor immunity and extend our previous findings by further identifying the critical components of vaccine function. IFN-γ neutralization experiments were conducted in vivo and revealed a requirement for this cytokine in the induction phase of tumor immunity to mKSA tumor cells. Corresponding in vitro analyses demonstrated that T lymphocytes from pCMV-Tag-immunized mice secrete IFNg upon stimulation with mKSA and that the percentage of Th1 cytokine-producing CD4+ and CD8+ lymphocytes is reduced upon IFN-γ neutralization. Although free IFN-γ was not required in the effector immune phase, targeted depletions of CD8+ T lymphocytes and NK cells indicated a necessary role for a Th1-associated cytotoxic cell component in the elicitation of complete tumor immunity. Collectively, our data demonstrate the critical roles that IFN-γ and Th1-associated effector cells play in maximizing the efficacy of an SV40 Tag DNA vaccine, and help to further shape a putative model of tumor immunity within our system. Such findings have direct implications for the development of immunotherapeutic and vaccination protocols against virally associated cancers and provide insight into the general mechanisms of tumor immunity targeting TAAs.

Materials and methods

Cells and media

Experimental tumor cell challenges were performed in vivo using the SV40-transformed BALB/c mouse kidney fibroblast cell line mKSA. Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with l-glutamine (HyClone, Logan, UT, USA) and supplemented with 0.1 mM nonessential amino acids (Sigma, St. Loius, MO, USA), 100 U/ml penicillin (Sigma), 500 μg/ml streptomycin (Sigma), and 10 % heat-inactivated fetal bovine serum (HyClone). Cells were detached from flasks with 1 mM EDTA–phosphate-buffered saline (PBS), washed, and adjusted to 1 × 105 cells in a total volume of 50 μl sterile PBS prior to intravenous administration into mice.

In vivo neutralization of IFN-γ and depletion of CD8+ T lymphocytes and NK cells

Neutralization of murine IFN-γ was performed in vivo using the rat IgG1 monoclonal antibody (MAb) R4-6A2 [28, 29]. Purification of this reagent was performed via mouse anti-rat antibody affinity chromatography, and the total protein content was estimated by measuring the absorbance at 280 nm. Purified MAb was adjusted to 100 μg in 0.1 ml sterile PBS and administered to mice a total of 3–6 times via intraperitoneal injection. Figure 1 provides a schematic of the injection sequence for induction phase and effector phase neutralization studies.
https://static-content.springer.com/image/art%3A10.1007%2Fs00262-012-1338-x/MediaObjects/262_2012_1338_Fig1_HTML.gif
Fig. 1

Induction phase (a) and effector phase (b) treatment schemes for DNA immunization, IFN-γ neutralization/NK cell depletion, and tumor cell challenge in BALB/c mice. Mice were immunized with pcDNA3.1 or pCMV-Tag on days 0 and 14 and left untreated or injected with anti-IFN-γ MAb (R4-6A2) on days −2, −1, 1, 12, 13, and 15 for induction phase studies (a). In corresponding effector phase studies, mice were immunized with pcDNA3.1 or pCMV-Tag on days 0, 14, 28, and 42 and left untreated or injected with anti-IFN-γ MAb on days 54, 55, and 57 (b). In a separate effector phase study, mice were immunized with pcDNA3.1 or pCMV-Tag on days 0, 14, 28, and 42 with and injected with depleting antibodies to NK cells ± CD8+ T cells (anti-asialo GM1 ± 2.43) or a control rabbit IgG preparation on days 54, 44, and 57 (b). All mice were challenged with mKSA tumor cells on day 28 for induction phase studies and day 56 for effector phase studies. Challenged mice were then euthanized on day 46 for induction phase studies and day 74 for effector phase studies

In a separate study, murine CD8+ T lymphocytes and NK cells were depleted using the rat IgG2b MAb 2.43 and a polyclonal rabbit anti-mouse asialo GM1 preparation (Cedarlane Laboratories, Burlington, NC, USA), respectively. Flow cytometry analyses have previously shown that these depletion reagents remove >95 % of the targeted cell population [19]. Depleting antibodies were purified via affinity chromatography and prepared for in vivo administration as described above. Depletion schemes for this effector phase study are similarly detailed in Fig. 1.

Mice, immunization, and tumor cell challenge

Six- to eight-week-old female BALB/c mice were purchased from Charles River Laboratories (Wilmington, MA, USA). Institutional guidelines and the Animal Welfare Assurance Act were observed in regard to the treatment and maintenance of animals.

Groups of mice were immunized intramuscularly with 100 μg pCMV-Tag or an empty control vector, designated pcDNA3.1, a total of 2–4 times at 2-week intervals (Fig. 1). In some experiments, sera were collected from animals 2 weeks post-immunization and assessed for anti-SV40 Tag antibody titers via ELISA. Mice were challenged intravenously with 1 × 105 mKSA tumor cells 2 weeks following the final immunization and euthanized 18 days post-tumor cell challenge. In order to assess lung tumor burden, lungs were harvested from euthanized mice and prepared for tumor focus visualization via intratracheal injection of 10 % India ink and destaining in Fekete’s solution. Assessment of tumor burden was performed via macroscopic examination of the lung surface. The accuracy of macroscopic tumor examination has previously been verified by our laboratory via H&E staining [19].

IFN-γ ELISA

Detection of murine IFN-γ in lymphocyte culture supernatants was performed with a mouse interferon gamma Quantikine Immunoassay kit (R&D Systems, Minneapolis, MN, USA). Groups of mice were immunized twice with pcDNA3.1 or pCMV-Tag at 2-week intervals and challenged with 1 × 105 mKSA tumor cells 2 weeks following the final immunization. Spleens from 3 pcDNA3.1-immunized mice and 3 pCMV-Tag-immunized mice were harvested 18 days following tumor cell challenge, homogenized in PBS, and centrifuged over a Histopaque-1077 gradient solution (Sigma) according to the manufacturer’s instructions. The lymphocyte layer was collected, washed in PBS, and cells were frozen at −80 °C in fetal bovine serum (FBS) with 10 % DMSO. Splenocytes were subsequently pooled and incubated at 37 °C in the 96-well round-bottom plates at a 2:1 ratio with 1 × 105 anti-CD3/anti-CD28 antibody-coated magnetic beads (Invitrogen Dynal AS, Oslo, Norway)/well in RPMI media supplemented with 0.1 % β-mercaptoethanol and 10 ng/ml IL-2 (PeproTech, Rocky Hill, NJ, USA). After 3 days of culture, 50 % of the media was replaced with fresh RPMI media supplemented with 0.1 % β-mercaptoethanol, 10 ng/ml IL-2, and 10 ng/ml IL-7 (PeproTech), and cells were incubated at 37 °C for an additional 3 days. Cells were then split and incubated in fresh RPMI media supplemented with 0.1 % β-mercaptoethanol, 10 ng/ml IL-2, and 10 ng/ml IL-7 overnight at 37 °C. Beads were removed via magnetic field exposure, and cells were pooled, washed, and resuspended in DMEM supplemented with 5 ng/ml IL-2. Expanded T lymphocytes were then incubated at 37 °C in 24-well plates at a cell density of 1 × 106/well with mKSA target cells at a 1:1 target-to-effector ratio. After 24 h, culture supernatants were harvested via centrifugation and assayed for the concentration of IFN-γ, in pg/ml, according to the ELISA manufacturer’s instructions. Data represent the mean of triplicate determinations for each group.

Intracellular flow cytometry

Groups of mice were immunized with pCMV-Tag, neutralized of IFN-γ, and challenged with mKSA tumor cells as depicted in Fig. 1a. Spleens were harvested from 5 pCMV-Tag-immunized mice and 5 pCMV-Tag-immunized + anti-IFN-γ-treated mice 18 days following tumor cell challenge, homogenized in PBS, and loaded onto a Histopaque-1077 gradient solution (Sigma) according to the manufacturer’s instructions. Lymphocytes were harvested via centrifugation, washed in PBS, and frozen at −80 °C in FBS with 10 % DMSO. Splenocyte samples were subsequently resuspended in 5 ml RPMI medium supplemented with phorbol myristate acetate (PMA; 50 ng/ml), ionomycin (1 μM), and brefeldin A (1 μg/ml) and stimulated for 5 h at 37 °C.

Fluorescence-activated cell sorting (FACS) reagents were purchased from BD Biosciences (Franklin Lakes, NJ). For each splenocyte sample, 1 × 106 cells were pelleted, resuspended in 100 μl blocking buffer (1 % bovine serum albumin−0.5 % sodium azide in PBS), and incubated in the dark for 20 min at room temperature. Cell surface markers were stained by incubation with 100 ng fluorescein isothiocyanate (FITC)-conjugated rat anti-mouse CD4 (RM4-5) or 100 ng FITC-conjugated rat anti-mouse CD8α (53–6.7) for 20 min at room temperature. Fixed cells were pelleted, permeabilized by resuspension in 100 μl BD Perm/Wash, and stained for intracellular cytokines by incubation with 100 ng phycoerythrin (PE)-Cy5-conjugated rat anti-mouse IFN-γ (XMG1.2) or 100 ng PE-Cy5-conjugated rat anti-mouse IL-2 (JE56-5H4) for 20 min at room temperature.

Stained cells were pelleted, prepared for FACS analysis by resuspension in 1 ml FACS flow sheath fluid, and passed through a Becton–Dickinson FACSVantage SE instrument. Analysis was performed with FlowJo software by first gating on the mononuclear cell population and then examining the percentage of cells positive for the surface markers, CD4 or CD8, and the intracellular cytokines, IFN-γ or IL-2. Data are presented as the total percentage of double-positive splenocytes, determined by averaging the values for 5 mice per group.

Statistical analysis

Comparisons between groups were performed via Student’s unpaired two-sample t test, with significance indicated at P values of <0.05. Individual concentration values were reciprocally transformed in the IFN-γ ELISA prior to statistical analysis. All statistical analyses were performed using GraphPad InStat and Prism software.

Results

T lymphocytes from pCMV-Tag-immunized mice secrete IFN-γ in response to in vitro stimulation with mKSA

Previous studies performed by our laboratory have demonstrated that immunization with recombinant SV40 Tag (rTag) generates mKSA-reactive splenocytes that secrete IFN-γ upon tumor cell stimulation [19], despite the heavily Th2-skewed immune response associated with this vaccine modality. In addition, our previous investigations with pCMV-Tag immunization have implied a prominent role for IFN-γ in the development of DNA vaccine-induced immunity, as indicated by the generation of mixed Th1/Th2 immune response in vivo [21]. In order to further assess the ability of pCMV-Tag vaccination to stimulate the synthesis of IFN-γ, we activated T lymphocytes from pCMV-Tag-immunized mice in vitro and assayed culture supernatants for the presence of IFN-γ via ELISA.

Mice were immunized twice with pcDNA3.1 or pCMV-Tag at 2-week intervals and challenged with 1 × 105 mKSA tumor cells 2 weeks following the final immunization. Splenocyte samples were subsequently harvested and T cells were globally expanded in vitro via CD3/CD28 engagement. Expanded T cells were then incubated overnight with mKSA tumor cell targets, and culture supernatants were harvested and assayed for the presence of IFN-γ via ELISA. As shown in Fig. 2, T cells from pCMV-Tag-immunized mice secrete greater amounts of IFN-γ in response to mKSA stimulation than T cells from pcDNA3.1-immunized mice. Culture supernatants from pCMV-Tag-associated T cells contained 111.92 ± 21.07 pg/ml IFN-γ, compared to 68.37 ± 1.90 pg/ml IFN-γ from pcDNA3.1-associated T cells. Since only pCMV-Tag-associated T cells were primed for SV40 Tag at the time of tumor challenge, the secretion of IFN-γ by pcDNA3.1-associated T cells likely represents an artifact of antitumor immune stimulation following mKSA tumor cell challenge. Although it is interesting to speculate that T cell-derived TGF-β might also exhibit differential production in pCMV-Tag-immunized mice and pcDNA3.1-immunized mice, this cytokine is secreted in significant quantities by our irradiated mKSA tumor cell targets (unpublished observations) and is thus difficult to investigate in T-cell culture supernatants.
https://static-content.springer.com/image/art%3A10.1007%2Fs00262-012-1338-x/MediaObjects/262_2012_1338_Fig2_HTML.gif
Fig. 2

Secretion of IFN-γ by mKSA-stimulated splenocytes from DNA-immunized mice. Mice were immunized twice with pcDNA3.1 (control) or pCMV-Tag and challenged with mKSA tumor cells 2 weeks following the final immunization. Splenocyte samples were harvested 18 days following tumor cell challenge and pooled for 3 mice per group. T cells were expanded from pooled splenocytes via 7-day culture with anti-CD3/anti-CD28-coated magnetic beads and subsequently harvested and incubated in triplicate for 24 h at a 1:1 ratio with mKSA tumor cell targets. Supernatants from 24 h cultures were collected and assayed for the presence of IFN-γ via ELISA. Data are presented as the average concentration of IFN-γ, in pg/ml, from triplicate determinations per group. Standard error bars represent SEM

These results demonstrate that immunization with pCMV-Tag generates mKSA-reactive T cells that produce significantly more IFN-γ than T cells from control-immunized mice. Such findings were expected, given our previous observations with IFN-γ secretion from rTag-immunized mice and the apparent induction of Th1 immune mechanisms associated with pCMV-Tag vaccination. Although other pro-inflammatory T-cell cytokines, such as TNF-α and IL-17, might also be expected to contribute to tumor immunity within our model, Utermohlen and colleagues have previously ruled out a role for TNF-α in the SV40 Tag system [30], and we are unaware of any reports indicating a role for IL-17 in immunity to polyomaviruses or viral TAAs.

IFN-γ is required in the induction phase, but not the effector phase, of tumor immunity to mKSA tumor cells

Provided that immunization with pCMV-Tag enhances the contribution of Th1 immunity in vivo and results in the production of mKSA-reactive T lymphocytes that secrete IFN-γ in vitro, we thought it is pertinent to directly investigate the role of IFN-γ within the context of induction phase and effector phase tumor immunity. IFN-γ neutralization experiments were performed by injecting mice with anti-IFN-γ MAb during the course of immunization or at the time of tumor cell challenge for induction phase and effector phase studies, respectively. Mice were subsequently challenged with mKSA tumor cells and analyzed for the presence of lung tumor foci 18 days post-tumor cell challenge. Detailed schematics of the immunization, neutralization, tumor cell challenge, and euthanization regime for these studies are provided in Fig. 1.

Mice treated with IFN-γ-neutralizing MAb in the induction phase of tumor immunity exhibit increased susceptibility to mKSA tumor formation in the lungs, as shown in Fig. 3, panel I. Sixty percent (3/5) of these IFN-γ-neutralized mice were positive for lung tumor foci, whereas none (0/5) of the mice immunized with pCMV-Tag and left nonneutralized demonstrated signs of tumor formation (Table 1). Although immunity was reduced substantially upon anti-IFN-γ MAb treatment, these mice demonstrated some resistance to mKSA tumor formation that was not evident in mice treated with the empty control vector pcDNA3.1, which were uniformly positive for lung tumor nodules (3/3).
https://static-content.springer.com/image/art%3A10.1007%2Fs00262-012-1338-x/MediaObjects/262_2012_1338_Fig3_HTML.gif
Fig. 3

Representative images of lung tumor foci in pCMV-Tag-immunized mice ± anti-IFN-γ treatment in the induction immune phase (a) or effector immune phase (b). Mice treated with anti-IFN-γ MAb in the induction immune phase exhibit detectable mKSA tumor foci, whereas all other groups of mice show a lack of tumor foci

Table 1

Proportion of mice with visible lung tumor foci following immunization with plasmid DNA and IFN-γ neutralization in the induction or effector immune phase

Immune Phase

Treatment

% of mice with tumors (no. of unprotected/no. of challenged)

Induction Phase

pcDNA3.1

100 % (3/3)

pCMV-Tag

0 % (0/5)

pCMV-Tag + anti-IFN-γ

60 % (3/5)

Effector Phase

pcDNA3.1

100 % (3/3)

pCMV-Tag

0 % (0/5)

pCMV-Tag + anti-IFN-γ

0 % (0/5)

Prior to tumor cell challenge, mice were immunized twice with pcDNA3.1 or pCMV-Tag for induction phase studies and four times with pcDNA3.1 or pCMV-Tag for effector phase studies. Some groups of mice were injected with an anti-IFN-γ MAb during the course of immunization for induction phase neutralization studies, or at the time of tumor cell challenge for effector phase neutralization studies (Fig. 1). All mice were euthanized 18 days following tumor cell challenge, and the percentage (fraction) of mice bearing any visible lung tumor foci was recorded for each group

In contrast to the required activities of IFN-γ in the induction immune phase, free cytokine was not necessary for the elicitation of complete tumor immunity in the effector immune phase. All mice immunized with pCMV-Tag were completely protected from the development of lung tumor foci (0/5), regardless of any additional treatment with anti-IFN-γ MAb (Fig. 3, panel II, and Table 1). As demonstrated previously, all mice immunized with pcDNA3.1 were uniformly susceptible to the formation of tumor foci (3/3), indicating that the protection observed in pCMV-Tag + anti-IFN-γ-treated mice was not due to a failure of the tumor cell challenge.

Collectively, these in vivo neutralization experiments illustrate an important role for IFN-γ in the induction immune phase and the dispensability of this cytokine in the effector immune phase of tumor immunity driven by pCMV-Tag. Such findings suggest a role for IFN-γ in the development of antitumor effector cells, which may subsequently mediate tumor immunity through IFN-γ-independent mechanisms.

Neutralization of IFN-γ in the induction immune phase reduces the proportion of Th1 cytokine-producing CD4+ and CD8+ splenocytes

In order to further characterize the effect of induction phase IFN-γ neutralization on the development of tumor-reactive immune cell populations, we analyzed the percentage of Th1 cytokine-producing CD4+ and CD8+ splenocytes in pCMV-Tag-immunized mice ± anti-IFN-γ treatment in the immune induction phase. Groups of mice were immunized, neutralized of IFN-γ, challenged with mKSA, and euthanized as depicted in Fig. 1a. Splenocytes were subsequently harvested, activated in vitro with PMA-ionomycin, and analyzed for the presence of select surface markers and intracellular cytokines via flow cytometry.

Given the tendency for IFN-γ to establish Th1-skewed immune responses via positive feedback mechanisms, we first assessed the proportion of IFN-γ-producing CD4+ and CD8+ cells among total splenocytes. As shown in Fig. 4, IFN-γ+ CD4+ cells and IFN-γ+ CD8+ cells both constituted a much smaller percentage of total splenocytes in IFN-γ-neutralized mice than in nonneutralized mice. Specifically, IFN-γ+ CD4+ cells represented 1.85 ± 0.29 % of the total splenocyte population in pCMV-Tag + anti-IFN-γ mice, compared to 7.39 ± 1.01 % in pCMV-Tag mice. Consistent with these findings, IFN-γ+ CD8+ cells represented 0.74 ± 0.10 % of the total splenocyte population in pCMV-Tag + anti-IFN-γ mice, compared to 3.54 ± 0.87 % in pCMV-Tag mice. In all immunized mice, CD4+ splenocytes constituted a higher percentage of the IFN-γ+ cells than did CD8+ splenocytes, underscoring a potential role for CD4+ cells as drivers of the in vivo Th1 immune response. These results indicate that in vivo neutralization of free IFN-γ disrupts the production of this cytokine by murine splenocytes and that the CD4+ cell population contributes more to the total production of IFN-γ than the CD8+ cell population.
https://static-content.springer.com/image/art%3A10.1007%2Fs00262-012-1338-x/MediaObjects/262_2012_1338_Fig4_HTML.gif
Fig. 4

Intracellular production of IFN-γ in splenocytes obtained from pCMV-Tag-immunized mice ± anti-IFN-γ treatment in the induction immune phase. Mice were immunized with pCMV-Tag, treated with anti-IFN-γ MAb or left untreated during the course of immunization, and challenged with mKSA. Splenocyte samples were harvested from euthanized mice, stimulated in vitro with PMA-ionomycin and analyzed for the presence of CD4, CD8, and intracellular IFN-γ via flow cytometry. a Quantitative analyses of the CD4+ IFN-γ+ and CD8+ IFN-γ+ splenocyte populations from pCMV-Tag-immunized mice ± anti-IFN-γ-treatment. The total percentage of double-positive cells was determined by averaging data for 5 mice per group. Standard error bars represent SEM. b Representative flow cytometry dot plots of IFN-γ-producing CD4+ and CD8+ splenocytes from pCMV-Tag-immunized mice ± anti-IFN-γ treatment

In addition to IFN-γ, we assessed the percentage of CD4+ and CD8+ splenocytes producing IL-2, which is a critical T lymphocyte growth factor classically associated with Th1 responses [31, 32]. IL-2+ CD4+ cells and IL-2+ CD8+ cells represented smaller percentages of total splenocytes in IFN-γ-neutralized mice than in nonneutralized mice (Fig. 5), although this reduction was not as severe as that previously observed for IFN-γ. IL-2+ CD4+ cells constituted 1.88 ± 0.35 % of the total splenocyte population in pCMV-Tag + anti-IFN-γ mice, compared to 3.69 ± 0.56 % in pCMV-Tag mice. IL-2+ CD8+ cells constituted 0.73 ± 0.11 % of the total splenocyte population in pCMV-Tag + anti-IFN-γ mice, compared to 1.78 ± 0.37 % in pCMV-Tag mice. As previously observed for IFN-γ, CD4+ splenocytes contributed more to the total production of IL-2 than did CD8+ splenocytes. Collectively, our experiments demonstrate that IFN-γ neutralization in the immune induction phase broadly inhibits the production of Th1-associated cytokines in the spleen, and suggest a dominant role for CD4+ lymphocytes in the synthesis of these immunomodulatory proteins.
https://static-content.springer.com/image/art%3A10.1007%2Fs00262-012-1338-x/MediaObjects/262_2012_1338_Fig5_HTML.gif
Fig. 5

Intracellular production of IL-2 in splenocytes obtained from pCMV-Tag-immunized mice ± anti-IFN-γ treatment in the induction immune phase. Mice were immunized with pCMV-Tag, treated with anti-IFN-γ MAb or left untreated during the course of immunization, and challenged with mKSA. Splenocyte samples were harvested from euthanized mice, stimulated in vitro with PMA-ionomycin, and analyzed for the presence of CD4, CD8, and intracellular IL-2 via flow cytometry. a Quantitative analyses of the CD4+ IL-2+ and CD8+ IL-2+ splenocyte populations from pCMV-Tag-immunized mice ± anti-IFN-γ-treatment. The total percentage of double-positive cells was determined by averaging data for 5 mice per group. Standard error bars represent SEM. b Representative flow cytometry dot plots of IL-2-producing CD4+ and CD8+ splenocytes from pCMV-Tag-immunized mice ± anti-IFN-γ treatment

Th1-associated cytotoxic effector cells are required for the elicitation of complete tumor immunity against mKSA

Despite our previous observation that free IFN-γ is not required in the immune effector phase, we reasoned that certain Th1-associated effector cells might be required to mediate tumor cell killing at this stage of the immune response. Our previous studies with targeted T lymphocyte depletions have demonstrated that CD4+ T lymphocytes play a necessary role in the effector phase of the immune response, although the precise function of these cells as antitumor ‘helpers’ versus ‘direct effectors’ has not been elucidated. In contrast, singular depletion of CD8+ T lymphocytes did not impair protection from mKSA tumor cells, indicating that these adaptive immune effectors are not necessary for the elicitation of complete tumor immunity within the context of DNA immunization. In order to further examine the contribution of traditional Th1-associated cytotoxic cell populations in DNA vaccine-induced tumor immunity, we depleted pCMV-Tag-immunized mice of NK cells alone and in combination with CD8+ T lymphocytes in the effector immune phase and performed an assessment of in vivo protection.

Mice were immunized with pCMV-Tag, depleted of NK cells ± CD8+ T lymphocytes, challenged with mKSA tumor cells, and euthanized as shown in Fig. 1b. Akin to previous observations with CD8+ T lymphocyte depletion, all mice immunized with pCMV-Tag and treated with an anti-NK cell reagent alone or a control rat IgG preparation failed to develop detectable lung tumor foci (Table 2). In spite of this, 50 % of the mice treated with a combination of anti-NK cell and anti-CD8 T lymphocyte reagents developed lung tumor foci, providing evidence that these two cell populations perform a vital overlapping function within the context of DNA vaccine-induced tumor immunity. Although this precise overlapping characteristic has yet to be elucidated, both NK cells and CD8+ T lymphocytes can deliver direct lethal hits to tumor cells via the release of cytotoxic granules, and both cell types can perform their functions via adaptive immune mechanisms. As previously indicated by our laboratory [21], all mice immunized with pCMV-Tag were positive for IgG antibody to SV40 Tag at the time of immune cell depletion and tumor challenge (data not shown), validating the possibility of antibody-dependent cell-mediated cytotoxicity (ADCC) as a mechanism of tumor cell destruction via employment of NK cells.
Table 2

Proportion of mice with visible lung tumor foci following immunization with plasmid DNA and depletion of NK cells ± CD8+ T cells in the immune effector phase

Treatment

% of mice with tumors (no. of unprotected/no. of challenged)

pcDNA3.1

100 % (5/5)

pCMV-Tag + Rabbit IgG

0 % (0/5)

pCMV-Tag + anti-NK

0 % (0/5)

pCMV-Tag + anti-NK + anti-CD8

50 % (2/4)

Mice were immunized with pcDNA3.1 or pCMV-Tag a total of four times and depleted of NK cells ± CD8+ T lymphocytes at the time of tumor cell challenge (Fig. 1b). All mice were euthanized 18 days following tumor cell challenge, and the percentage (fraction) of mice bearing any visible lung tumor foci was recorded for each group

Taken together, our previous and current findings suggest a central role for CD4+ T lymphocytes in the effector phase of tumor immunity to mKSA, and overlapping effector functions for CD8+ T lymphocytes and NK cells in this process. Interestingly, both CD8+ T lymphocytes and NK cells were dispensable when depleted singularly; however, a double-depletion scheme demonstrated the importance of a resident cytotoxic effector cell component consisting of CD8+ T lymphocytes and/or NK cells within the effector phase of tumor immunity to mKSA.

Discussion

Our laboratory has endeavored to identify the critical antitumor immune mechanisms stimulated by both SV40 Tag recombinant protein and plasmid DNA immunization schemes within a model of pulmonary metastasis in BALB/c mice. Interestingly, we have elucidated distinct roles for various immune cell populations during the course of immunization and initial immune cell priming (i.e., the induction immune phase) versus the mobilization of antitumor immune effectors upon subsequent tumor cell challenge (i.e., the effector immune phase). Our previous work with recombinant SV40 Tag protein (rTag) immunization has identified CD4+ T lymphocytes as important mediators of the induction phase immune response [18], and anti-Tag antibodies, CD8+ T lymphocytes, and NK cells as critical immune effectors to mKSA tumor cells [19, 20]. Given that SV40 Tag is expressed both intracellularly and on the surface of SV40-transformed cells [33], the involvement of both cellular and humoral immunity is logical. Within this scenario of rTag-induced tumor immunity, we proposed that Th2 CD4+ T lymphocytes stimulated the secretion of anti-SV40 Tag antibodies from B cells, which in turn engaged NK cells to destroy tumor cells via ADCC [34]. CD8+ T lymphocytes were also required in the immune effector phase, presumably after exposure to neo-antigens via ADCC release and subsequent dendritic cell cross-priming. In a corresponding scenario of pCMV-Tag DNA vaccine-induced tumor immunity, we identified CD4+ T lymphocytes as drivers of a mixed Th1/Th2 antitumor immune response and demonstrated the nonessential roles of anti-Tag antibodies and CD8+ T lymphocytes in the elicitation of complete tumor immunity to mKSA [21]. We hypothesized that the striking differences in antitumor effector mechanisms associated with rTag and pCMV-Tag vaccination might correlate with the induction of an enhanced Th1 component in the pCMV-Tag vaccination scenario.

In this study, we explored the contribution of IFN-γ and Th1-associated effector cells to tumor immunity stimulated by vaccination with pCMV-Tag. Initial in vitro analysis demonstrated that T lymphocytes from pCMV-Tag-immunized mice secrete IFN-γ in response to mKSA stimulation, and subsequent in vivo IFN-γ neutralization experiments illustrated the important role of this cytokine in the induction phase of tumor immunity to mKSA tumor cells. Flow cytometry analysis showed a reduction in the percentage of IFN-γ-producing splenocytes and IL-2-producing CD4+ and CD8+ splenocytes upon IFN-γ neutralization and provided evidence for the role of CD4+ cells as the primary source of these Th1-associated cytokines. Although IFN-γ played a much less consequential role in the effector immune phase, a critical overlapping function for CD8+ T lymphocytes and NK cells was identified at this stage of the immune response, demonstrating that certain Th1-associated effector cells are required to mediate complete tumor immunity to mKSA tumor cells. Collectively, these findings aid in the construction of a putative model for pCMV-Tag-induced tumor immunity distinct from our previous model with rTag vaccination.

Given the requirement for CD4+ T lymphocytes in both the induction and effector phases of tumor immunity against SV40 Tag-expressing tumor cells, we believe that these cells serve to establish and maintain the mixed Th1/Th2 immune response stimulated by pCMV-Tag vaccination. Immunization with plasmid DNA likely stimulates a strong innate immune response via engagement of TLR-9 on antigen-presenting cells, which may in turn support the differentiation of Th1 CD4+ T lymphocytes via release of IL-12 and other immunostimulatory cytokines [35]. Moreover, the ability of CD4+ T lymphocytes to ignite Th1 immune responses to tumor antigens has been widely reported [3638], and our current flow cytometry data suggest that these cells produce the majority of the IFN-γ and IL-2 in our system. Our previous findings with rTag vaccination have also indicated the importance of CD4+ T lymphocytes in stimulating protective Th2 responses to mKSA tumor cells [18], implicating a role for these cells in the dual activation of protective Th1- and Th2-immune pathways in response to pCMV-Tag vaccination. The immune effectors mobilized by this mixed Th1/Th2 response may function within humoral immunity, such as anti-Tag antibodies and NK cells, as well as within cell-mediated immunity (CMI), such as CD8+ T lymphocytes and NK cells. NKT cells and other innate-like lymphocytes may also contribute to the protective antitumor response upon tumor cell challenge; however, these cells generally react with glycolipids or common microbial antigens and thus are not expected a play a major role in the induction of protective immune responses upon pCMV-Tag vaccination. Although it would be interesting to further evaluate mechanisms of antitumoral immune cell infiltration and antibody deposition via immunohistochemistry, the complete protection afforded by our vaccine excludes the possible acquisition of tumor masses for this purpose. The ability of DNA vaccines to simultaneously engage both the humoral and cell-mediated arms of the immune system is a hallmark of this vaccine modality [3941] and may explain the relative versatility of tumor immunity stimulated by pCMV-Tag compared to rTag.

Unlike our previous observations with rTag vaccination, singular depletion of CD8+ T lymphocytes or NK cells did not abrogate tumor immunity to mKSA tumor cells, suggesting the presence of alternative immune pathways lacking in recombinant protein vaccination schemes. Interestingly, induction phase depletion of CD8+ T lymphocytes enhanced the early antibody response to SV40 Tag, indicating that impairments in CMI may skew the immune response toward compensatory humoral immune mechanisms [21]. Likewise, complete tumor immunity to mKSA tumor cells could be achieved prior to the induction of a detectable antibody response, presumably via activation of Tag-specific CD8+ T lymphocytes. Previous studies with the DNA vaccine pSV3-neo demonstrated the importance of CD8+ T lymphocytes as primary mediators of immunity to mKSA tumor cells [42]; however, the native SV40 promoter in this plasmid failed to generate the threshold level of antigen required to provide complete immunity within a model of pulmonary metastasis [43]. In our present studies with the more potent pCMV-Tag, abrogation of vaccine function required the impairment of multiple effector pathways (i.e., CMI and ADCC via double depletion of CD8+ T lymphocytes and NK cells) or the impairment of a central induction component, such as CD4+ T lymphocytes or the pleiotropic cytokine IFN-γ. The central role of IFN-γ-secreting CD4+ T lymphocytes in tumor immunity has recently been described for a DNA vaccine encoding large tumor antigen from Merkel cell polyomavirus [44], perhaps indicating a conserved role for this cell subset in mobilizing tumor immunity against transforming polyomavirus oncoproteins upon vaccination with plasmid DNA.

The antitumor potential of host-derived IFN-γ has been described in multiple animal and human studies and remains an important focus of many current approaches to elicit immunity to tumors. Although most of these studies have examined the production of IFN-γ in vitro, a select few have directly investigated the role of this cytokine within the context of in vivo protection. An early report by Tuttle and colleagues indicated a significant reduction in the antitumor efficacy of adoptively transferred tumor-draining lymph node cells if administered in conjunction with an anti-IFN-γ antibody [45]. Furthermore, the role of IFN-γ in whole cell vaccination and adoptive cell transfer has been explored by Winter and colleagues in studies with IFN-γ knockout mice [27], and Wolchok and colleagues have used a similar IFN-γ knockout mouse model to describe a role for IFN-γ in the induction phase of tumor immunity stimulated by certain melanosomal antigen-encoding DNA vaccines [26]. In our present study, we directly identify the requirement for biologically active host-derived IFN-γ and Th1 network integrity in the stimulation of protective SV40 Tag DNA vaccine-induced tumor immunity in a model of pulmonary metastasis in immunocompetent BALB/c mice. In our assessment, the use of neutralization/depletion approaches in place of immunologically impaired knockout animals circumvents the issues of confounding immune skewing phenomena and provides an exceptionally accurate glimpse of the natural underlying antitumor immune mechanisms associated with antitumor vaccination. Additionally, to our knowledge, the direct in vivo role of IFN-γ in vaccine-induced tumor immunity to a viral TAA has not previously been described.

As novel vaccines continue to be explored for the treatment and prevention of cancer, an understanding of the antitumor mechanisms associated with these immunological agents will provide valuable insight into the construction of more effective immunization platforms. Our previous and current findings illustrate the critical roles that CD4+ T lymphocytes and IFN-γ play in driving DNA vaccine-induced tumor immunity to a viral TAA, and reconcile some of the key differences between recombinant protein and DNA vaccination strategies. Such findings have direct implications for the development of vaccines against certain SV40 Tag-expressing tumors, including malignant pleural mesothelioma and non-Hodgkin’s lymphoma, as well as other malignancies bearing distinct TAAs of viral or self-origin.

Acknowledgments

This work was supported, in part, by National Institutes of Health grant no. RR-12317 and the Office of Research at Texas Tech University Health Sciences Center. J. F. Aldrich is a recipient of the TTUHSC Dean’s Scholars Award and the Mary Lou Clements-Mann Endowed Scholarship. This manuscript is dedicated to the memory of Dr. Ronald C. Kennedy, whose insights and guidance continue to inspire the work of the authors.

Conflict of interest

The authors declare that they have no conflict of interest.

Copyright information

© Springer-Verlag 2012