|Home | About | Journals | Submit | Contact Us | Français|
DC engineered in vitro by DNA encoding OVAhsp70 and IL-15 up-regulated their expressions of CD80, CD86, CCR7 and IL-15Rα and promoted their productions of IL-6, IL-12 and TNF-α. Transcriptional IL-15-directed in vivo DC targeting DNA vaccine encoding OVAhsp70 elicited long-lasting Th1 and CTL responses and anti-B16OVA activity. CD8 T cell-mediated primary tumor protection was abrogated by DC or CD4 T cell depletion during the induction phase of immune responses. However, CD4 T cell depletion during immunization did not impair CD8 T cell-dependent long-lasting tumor protection. Furthermore, in vivo DC-derived IL-15 exerted the enhancements of cellular and humoral immune responses and antitumor immunity elicited by OVAhsp70 DNA vaccine. Importantly, the potency of this novel DNA vaccine strategy was proven using a self/tumor Ag (TRP2) in a clinically relevant B16 melanoma model. These findings have implications for developing next generation DNA vaccines against cancers and infectious diseases in both healthy and CD4 deficient individuals.
DNA vaccines are safe, relatively inexpensive and effective in eliciting innate and adaptive immune responses and protective immunity (1-3). DNA vaccination therefore is a practical and promising strategy aimed at developing effective vaccines against tumors and infectious diseases (1-3). Weak immunogenicity of DNA vaccines is a major obstacle to current DNA vaccine development and may be responsible for the ineffective vaccination in humans (1-3). Many strategies have been designed and tested to overcome this problem including administration of DNA encoding cytokines, chemokines, costimulatory molecules, survival factor or helper Ag (4-8), injection of mAb (9), addition of chemical adjuvant (10), enhancement of Ag expression (11), improvement of DNA delivery (12) and modification of Ag (3, 13).
Dendritic cells (DC) are essential in initiating, programming and regulating Ag-specific T cell immunity and the key target of DNA vaccines (5, 14-20). DNA vaccine encoding single-chain antibodies (via receptor CD205) to target Ag to DC enhances its potency (21-22). Targeting Ag to DC alone without providing proper DC activation signaling can result in tolerance not immunity (23-24). Thus, DC targeting strategies should include means to activate DC. Although DNA self triggering and some receptor (such as FcR, TLR) triggering may provide some DC activation stimuli (25-28), these stimuli were usually insufficient for optimal DC activation. Strong Ag-specific CD4 and CD8 T cell responses are critical to reject tumors and chronic infectious diseases (1-3, 13, 22). To achieve this, DNA vaccines need to target DC for Ag processing and presentation via MHC class I and II pathways and optimal DC activation. In particular, DNA vaccines encoding Ag fused to hsp70 (Aghsp70) elicited better Ag-specific CD4 and CD8 T cell responses and antitumor immunity (29-31). The mechanisms behind the observations were linked to target Ag to DC (via CD40 and LOX-1) and activate DC (via CD40) (30-34).
IL-15 is a critical cytokine for the development and function of innate and adaptive immune cells (35). DNA-derived IL-15 has been coadministrated as an adjuvant to enhance the potency of DNA vaccines encoding Ag (36-37). Accumulated data have indicated that DC-derived IL-15 is necessary for the induction and maintenance of CD8 T cell responses and may be a signal for DC activation to promote immunity (35, 38-40). Whether transcriptional IL-15-directed in vivo DC targeting can enhance the efficacy of a DNA vaccine has not yet been proven.
In this study, we constructed a novel DNA vaccine with the ubiquitous expression and secretion of a fusion Ag (Aghsp70) and the DC-specific expression and secretion of an optimized human IL-15 (IL-15) (36). We tested the ability of DC-derived IL-15 to improve DC activation, cellular and humoral immune responses and tumor protection in B16 melanoma with a model Ag (OVA) or self/tumor Ag (TRP2). We also examined the mechanisms underlying this novel DNA vaccine strategy.
C57BL/B6 and CD11c-DTR (B6.FVB-Tg(Itgax-DTR/EGFP)57Lan/J) mice (female, 6-8 wks) were purchased from Taconic (Germantown, New York) and JAX (Bar Harbor, Maine), and housed in specific pathogen-free conditions in the University of Pittsburgh animal facility. All animal procedures were performed according to approved protocols and in accordance with recommendations for the proper use and care of laboratory animals. Murine melanoma tumor cell B16OVA (a gift from Drs. E. Lord and J. Frelinger at Univ. of Rochester, Rochester, New York) (41) or B16 (ATCC, Manassas, Virginia) was maintained in DMEM (IRVINE Scientific, Santa Ana, CA) supplemented with 10% fetal bovine serum (FBS) (Hyclone, Logan, Utah), 2mM glutamine (Invitrogen, Carlsbad, California) and 1× antibiotic antimycotic solution (Sigma, St. Louis, Missouri).
Restricted enzymes, T4 DNA ligase and antarctic phosphatase were purchased from NEB (Beverly, Massachusetts). AccuPrime pfx DNA polymerase and one shot Top10 chemically competent E. coli were purchased from Invitrogen. All primers were synthesized by IDTDNA (Coralville, Iowa). Detailed vector construction is described in supplemental materials. Briefly, pcDNA3.1 (+) (Invitrogen) was digested with EcoRV and Bst1107I and self-ligated with T4 DNA Ligase (resultant vector 1). PCR amplified NheI and XhoI-digested BGHpA was cloned into NheI and XhoI-digested vector 1 (resultant vector 2). NotI and XhoI-digested CD11c-Eαd fragment (CD11c-pDOI-5 vector is a gift from Dr. T. Brocker at Ludwig-Maximilians-Universitat Munchen, Germany) was cloned into NotI and XhoI-digested vector 2 (resultant vector 3: CMV-AscI-CD11c-EcoRI). OVA, hsp70 or OVAhsp70 was cloned into AscI-digested vector 3 (resultants CMV-OVA-CD11c, CMV-hsp70-CD11c and CMV-OVAhsp70-CD11c). PCR amplified EcoRI-digested optimized human IL-15 (36) (Human IL-15 cDNA is a gift from Dr. T. A. Waldmann at NIH, Bethesda, Maryland) was cloned into EcoRI-digested CMV-OVA-CD11c, CMV-hsp70-CD11c or CMV-OVAhsp70-CD11c (resultants CMV-OVA-CD11c-IL-15, CMV-hsp70-CD11c-IL-15 or CMV-OVAhsp70-CD11c-IL-15). PCR amplified blunt end fragment of epidermal keratin K14-EcoRI-BGHpA (K14) (42) was cloned into pCR-Blunt II-TOPO (Invitrogen) (resultant vector 4). NotI and XhoI-digested K14 was cloned into NotI and XhoI-digested CMV-OVAhsp70-CD11c vector (resultant CMV-OVAhsp70-K14). EcoRI-digested optimized human IL-15 was cloned into EcoR I-digested CMV-OVAhsp70-K14 (resultant CMV-OVAhsp70-K14-IL-15). PCR amplified AscI-digested TRP2 or TRP2hsp70 (7) was cloned into AscI-digested CMV-OVA-CD11c or CMV-OVA-CD11c-IL-15 (resultants CMV-TRP2-CD11c-IL-15, CMV-TRP2hsp70-CD11c or CMV-TRP2hsp70-CD11c-IL-15). Inserted genes were confirmed by both enzyme digestion and DNA sequencing. DNA was purified using EndoFree plasmid kits (QIAGEN, Valencia, California).
Bone marrow-derived DC were generated as described previously (43). Briefly, bone marrow cells (1×106/ml) taken from naïve C57BL/B6 mice were cultured in RPMI1640 (IRVINE Scientific) supplemented with 10% FBS (Hyclone), 2mM glutamine (Invitrogen), 1× antibiotic antimycotic solution (Sigma), recombinant mouse GM-CSF (1000 unites/ml) and IL-4 (1000 unites/ml) (R&D Systems, Minnneapolis, Minnesota). On d 5-6 of DC culture, CD11c+ DC were purified using anti-CD11c micro-beads (Miltenyi Biotec., Auburn, California). CD11c+ DC (1×106) in 40μl medium in a 12-well plate were shot with a bullet (3μg DNA/bullet) using a Gene Gun (Bio-Rad, Hercules, California) at 260 psi pressure as described in ref. (44). After transfection, cells were continually cultured in 1ml BM-DC culture medium for 2 or 3 d. D 2 post transfection, the pooled DC culture supernatant (3ml) was harvested and concentrated to 300μl using a Centricon-10 (Amicon Millipore, Billerica, Massachusetts) following the vendor's instruction. The concentration of human IL-15 in concentrated supernatants was determined by human IL-15 ELISA kit (BD Biosciences, San Jose, California). 293T cells (ATCC) were transfected with DNA as described previously (31). OVAhsp70 in DC lysates or 293T cell supernatants was detected by Western blotting (WB) using mouse anti-human hsp70 mAb (Assay Designs Stressgene, Ann Arbor, Michigan) as described previously (31). D 3 post transfection, culture supernatants and DC were harvested. The concentration of IL-6, IL-12p40 or TNF-α in DC culture supernatants was determined by ELISA kits (eBioscience and Biolegend, San Diego, California). To examine DC phenotypes, harvested DC were incubated with anti-mouse CD16/32 (eBioscience), and then stained with APC-anti-mouse CD11c (HL3), PE-anti-mouse CD86 (GL1), PE-anti-mouse CD80 (16-10A1), PE-anti-mouse CCR7 (4B12) or FITC-anti-mouse IL-15Rα (isotype control of each antibody was used in control staining) (eBioscience; R&D Systems; BD Biosciences), and analyzed by flow cytometry on a BD LSRII with CellQuest software (BD Biosciences). The flow cytometric data were analyzed using Flowjo software (Tree star, Ashland, Oregon).
C57BL/B6 mice (3mice/group) were immunized by Gene Gun with CMV-OVAhsp70-CD11c-IL-15 DNA and related control vectors once (3μg DNA/bullet, 4 bullets at 4 sites, 400 psi pressure) as described in ref. (14). D 14 or 40 post immunization, CD4 T cells were isolated from splenocytes using anti-mouse CD4 micro-beads (Miltenyi Biotec.). Isolated CD4 T cells (6×106) were cultured with OVA-specific MHC class II peptides (OVA323-339, purity>95%, synthesized and HPLC purified in the core facility of University of Pittsburgh) (200ng/ml) in the presence of splenocytes (as APCs) (3×105) in 1ml RPMI 1640 10%FBS at 37°C, 5% CO2 for 3 d. The concentration of IFN-γ, IL-4 or IL-5 in the culture supernatants was determined by ELISA kit (BD Biosciences).
C57BL/B6 mice (3mice/group) were immunized by Gene Gun with CMV-OVAhsp70-CD11c-IL-15 DNA and related control vectors once as described above. D 14 post immunization, naive syngeneic splenocytes were labeled with either 0.5 μM or 5 μM carboxyfluorescein succinimidyl ester (CFSE, Molecular Probes, Eugene, Oregon) for 10 minutes at 37°C, washed, and then pulsed for 1 h with OVA-specific MHC class I peptides (OVA257-264, 200ng/ml, purity>95%, synthesized and HPLC purified in the core facility of University of Pittsburgh). Labeled and pulsed cells were subsequently mixed at a 1:1 ratio and approximately 107 cells were injected intravenously into immunized mice. 12 h later, mice were killed and splenocytes were analyzed by flow cytometry as described above. The percentage of OVA-specific lysis was calculated as follows: % specific lysis = (1 − [ratio of CFSElo/CFSEhi in naive mice ÷ ratio of CFSElo/CFSEhi in immunized mice]) × 100.
C57BL/B6 mice (3mice/group) were immunized by Gene Gun with CMV-OVAhsp70-CD11c-IL-15 DNA and related control vectors once as described above. D 40 post immunization, splenocytes (6 × 106/ml) were cultured with OVA257-264 peptides (200ng/ml) at 37°C, 5% CO2 for 48 h. To quantify the target-cell killing activity mediated by OVA-specific T lymphocytes, the flow cytometry-based CTL assay (Cytoxilux, OncoImmunin, Inc., Gaithersburg, Maryland) was used to detect the specific cleaved caspase in the target cells according to vendor's protocol. Briefly, 4 × 106 B16OVA (target cells) in 2ml culture medium were cultured with OVA257-264 peptides (200ng/ml) at 37°C, 5% CO2. After 45 min, TFL4 (2μl) was added and B16OVA cells were continually cultured for 15 min. The target cells were then washed with serum-free RPMI and resuspended in wash buffer at 2 × 106/ml. Target cells (T) (2 × 105/well) were mixed with titrated lymphocyte effectors (E) at ratios of 50:1, 25:1 or 12.5:1 (E:T) in 96-well plates in 200μl wash buffer. The cell mixture was centrifuged at 1,400rpm for 5 min at room temperature. Cell pellets were resuspended in 75μl caspase substrate solution and incubated at 37°C, 5% CO2 for 2 h. The caspase activity in target cells was analyzed by flow cytometry.
C57BL/B6 mice (2mice/group) were immunized by Gene Gun with CMV-OVAhsp70-CD11c-IL-15 or CMV-OVAhsp70-K14-IL-15 DNA once as described above on d 1. Sera were collected on d 21, 28, 35 and 42. Anti-OVA Ig was monitored by standard ELISA. Briefly, 96-well plate was coated with OVA protein (Grade V, Sigma) in carbonate buffer (PH9.5) (10μg/ml) at 4°C overnight. Samples of sera were diluted 10 times in 10% FCS and added into OVA-coated well for incubation for 2 h at room temperature. After extensive washing with wash buffer, HRP-goat anti-mouse Ig (Millipore) was added for incubation for 1 h at room temperature. After extensive washing with wash buffer, substrates of HRP were added and plates were placed in the dark. After 15 min, the plate was read by an ELISA reader SpectraMax (Molecular Devices, Sunnyvale, California) at OD450.
C57BL/B6 mice (3-5mice/group) were immunized by Gene Gun with CMV-OVAhsp70-CD11c-IL-15 DNA and related control vectors once as described above on d 1. D 14 or 40 post immunization, mice were challenged s.c. with B16OVA (1×105). In some experiments, mice were immunized by Gene Gun with CMV-TRP2hsp70-CD11c-IL-15 DNA and related control vectors as described above on d 1, 7 and 14. On d 21, mice were challenged s.c. with B16 (5×104). Tumors were measured using a digital slide calipers (Fisher Scientific, Pittsburgh, Pennsylvania) in the two perpendicular diameters every 3 d. Mice were sacrificed when tumor reached 10mm in mean diameter. To transiently deplete endogenous CD11c+ DC, CD11c-DTR transgenic mice were injected i.p. once with diphtheria toxin (DT) (Sigma) at 5 ng/g mouse body weight (45). After 16-18 h, these mice were immunized by Gene Gun with CMV-OVAhsp70-CD11c-IL-15 DNA as described above. To deplete CD4 T cells during immunization, mice immunized by Gene Gun with CMV-OVAhsp70-CD11c-IL-15 DNA on d 1 were injected i.p. with anti-mouse CD4 mAb (GK1.5) (200μg/injection) on d -3, 2 and 5. To deplete CD8 T cells during tumor challenge, mice immunized by Gene Gun with CMV-OVAhsp70-CD11c-IL-15 DNA on d 1 were injected i.p. with anti-mouse CD8 mAb (53-6.7) (200μg/injection) on d 13, 15, 17, 20 and 23 (tumor challenge at d 14) or d 38, 41, 44, 48, 55 (tumor challenge at d 40). Depletion was confirmed by flow cytometry and resulted in greater than 95% reduction of relevant cell types.
Data were statistically analyzed using Student's t test (Graph Pad Prism version 5, San Diego, California). Data from animal survival experiments were statistically analyzed using Log Rank test (Graph Pad Prism version 5). P < 0.05 is considered to be statistically significant.
We designed a novel DNA vaccine with CMV promoter-driven ubiquitous expression of Aghsp70 and CD11c promoter-driven DC-specific expression of optimized IL-15 (Fig.1A). Bone marrow-derived CD11c+ DC or 293T cells were transfected with CMV-OVAhsp70-CD11c-IL-15, CMV-OVAhsp70-CD11c or CMV-CD11c DNA and continually cultured for 48 h. OVAhsp70 was detected in the lysates of CD11c+ DC or culture supernatants of 293T cells transfected with CMV-OVAhsp70-CD11c-IL-15 and CMV-OVAhsp70-CD11c but not CMV-CD11c DNA (Fig.1B) (Due to the interference by FBS, secreted OVAhsp70 was detected in FBS-free culture supernatants of transfected 293T cells but not in FBS-containing culture supernatants of transfected DC.). Furthermore, CD11c+ DC transfected by CMV-OVAhsp70-CD11c-IL-15 DNA secreted significant IL-15 (Fig. 1C). The data suggest that OVAhsp70 and IL-15 are expressed well by CD11c+ DC after DNA transfection.
The function of Aghsp70 was linked to activate DC (34). To determine whether transcriptional IL-15-directed DC targeting can further improve Aghsp70-engineered DC activation, cultured CD11c+ DC were transfected with CMV-OVAhsp70-CD11c-IL-15 DNA or related control vectors. D 3 post transfection, DC and culture supernatants were harvested and analyzed by flow cytomerty and ELISA, respectively, to monitor DC phenotypes and cytokines. CMV-OVAhsp70-CD11c-IL-15-engineered DC not only markedly improved expressions of CD80, CD86, CCR7 and IL-15Rα (Fig.2A), but also dramatically increased the productions of IL-6, IL-12 and TNF-α (Fig.2B). The data suggest that transcriptional IL-15-directed DC targeting improves Aghsp70-engineered DC activation in vitro.
DNA vaccines encoding Aghsp70 have been shown to enhance Ag-specific T cell responses and antitumor immunity (29-31). To assess whether transcriptional IL-15-directed in vivo DC targeting can further enhance OVA-specific long-lasting T cell responses induced by OVAhsp70 DNA vaccine, mice were immunized once by Gene Gun with CMV-OVAhsp70-CD11c-IL-15 DNA or related control vectors. To determine CD4 T cell responses, d 14 or 40 post immunization, CD4 T cells were isolated from immunized or non-treatment mice and restimulated with OVA323-339 peptides in the presence of APCs in vitro. To determine CTL responses, d 14 post immunization, naive syngeneic splenocytes were labeled with CFSE and then pulsed with OVA257-264 peptides in vitro. The target-cell (splenocytes-OVA) killing activity mediated by OVA-specific T lymphocytes was measured by the in vivo CTL killing assay. D 40 post immunization, splenocytes from immunized or non-treatment mice were restimulated with OVA257-264 peptides in vitro. The target-cell (B16OVA) killing activity mediated by OVA-specific T lymphocytes was measured by the flow cytometry-based CTL assay. CMV-OVAhsp70-CD11c-IL-15 DNA vaccine was superior in eliciting potent long-lasting OVA-specific IFN-γ-producing Th1 (Fig. 3A, C) and CTL responses (Fig. 3B, D).
To evaluate protective tumor immunity induced by this DNA vaccine strategy, mice were immunized once by Gene Gun with CMV-OVAhsp70-CD11c-IL-15 DNA or related control vectors. D 14 later, these mice were challenged with B16OVA. CMV-OVAhsp70-CD11c DNA vaccine generated tumor protection when compared with CMV-OVA-CD11c-IL-15 or CMV-hsp70-CD11c-IL-15 DNA vaccine (Fig. 4A). However, CMV-OVAhsp70-CD11c-IL-15 DNA vaccine elicited better tumor protection than CMV-OVAhsp70-CD11c DNA vaccine (p<0.05) (Fig. 4A). Moreover, CMV-OVAhsp70-CD11c-IL-15 DNA vaccine also generated long-lasting tumor protection (tumor challenge at d 40 post DNA immunization) (Fig. 4C). The data show that transcriptional IL-15-directed in vivo DC targeting enhances OVAhsp70 DNA vaccine-elicited long-lasting Th1 and CTL responses and tumor protection.
To dissect the cellular mechanisms underlying this DNA vaccine strategy, endogenous DC or CD4 T cells were depleted during immunization and CD8 T cells were depleted during tumor challenge (d 14 after DNA immunization). DC, CD4 or CD8 T cell depletion abrogated tumor protection elicited by CMV-OVAhsp70-CD11c-IL-15 DNA vaccine (Fig. 4B), suggesting that CD8 T cell-mediated primary tumor protection elicited by this DNA vaccine strategy is DC- or CD4 T cell-dependent during the induction phase of immune responses.
To further examine the role of T cells in long-lasting tumor protection elicited by this DNA vaccine strategy, CD4 T cells were depleted during immunization and CD8 T cells were depleted during tumor challenge (d 40 post immunization). CD4 T cell depletion did not diminish, but slightly increased long-lasting tumor protection elicited by CMV-OVAhsp70-CD11c-IL-15 DNA vaccine (Fig. 4C). However, CD8 T cells were absolutely required to clear tumor cells (Fig. 4C). This data indicates that CD8 T cell-dependent long-lasting tumor protection can be elicited by this DNA vaccine strategy in both normal and CD4 T cell deficient mice.
DC and epidermal cells are transfected by Gene Gun immunization. To investigate whether the resource of IL-15 is critical in enhancing the efficacy of the DNA vaccine, we compared cellular and humoral immune responses and tumor protection elicited by CMV-OVAhsp70-CD11c-IL-15 (DC-derived IL-15) and CMV-OVAhsp70-K14 (a well-defined epidermal keratin K14 promoter, 42)-IL-15 (epidermal cell-derived IL-15) to define the resource of IL-15 in enhancing the efficacy of Aghsp70 DNA vaccine. When compared with transcriptional IL-15-directed in vivo epidermal cell targeting, transcriptional IL-15-directed in vivo DC targeting increased CTL, Th1 and antibody responses (Fig. 5A-C). Importantly, as shown in Fig.5D, DC-derived, not epidermal cell-derived, IL-15 was able to promote tumor protection. The data suggest that in vivo DC-derived IL-15 is important to enhance immune responses and tumor protection.
To test whether this novel DNA vaccine strategy can be effective in a murine B16 melanoma using the self/tumor Ag (TRP2), mice were immunized by Gene Gun with CMV-TRP2-CD11c-IL-15, CMV-TRP2hsp70-CD11c or CMV-TRP2hsp70-CD11c-IL-15 DNA on d 1, 7 and 14. On d 21, these mice were challenged with B16. In consistent with previous report (7), mice immunized with CMV-TRP2hsp70-CD11c DNA demonstrated longer survival when compared with CMV-TRP2-CD11c-IL-15 DNA (p<0.05). However, CMV-TRP2hsp70-CD11c-IL-15 DNA vaccine was much more effective in eliciting tumor protection than CMV-TRP2hsp70-CD11c DNA vaccine (p<0.005) (Fig.6). This data suggests that transcriptional IL-15-directed in vivo DC targeting DNA vaccine is effective in a self/tumor Ag (TRP2) B16 melanoma model, implying that this novel strategy might be translatable into a clinically relevant setting.
There is currently much interest in exploring in vivo DC targeting DNA vaccine strategies which may provide a practical way to develop bulk quantities of effective vaccines for cancer and infectious diseases (18-19, 22). IL-15, which was ranked number one in NCI Immunotherapy Agent Workshop Proceedings, has been tested to enhance the effects of tumor vaccine, chemotherapy or adoptive cell transfer and HIV-1 vaccine (35-37). However, IL-15 has a short half-life and high doses are needed to achieve biological responses in vivo. DNA-delivered IL-15 has been shown to increase the magnitude of the response to DNA vaccines (36-37). Also, DC-derived IL-15 may serve as a signal for DC activation and maturation to promote immunity (35, 38-40). It is unknown whether transcriptional IL-15-directed in vivo DC targeting can greatly enhance the efficacy of DNA vaccines.
DC-specific CD11c promoter-driven Ag expression is insufficient to elicit optimal T cell immunity (46), therefore, the ubiquitous expression and secretion of Aghsp70 by mammalian cells was designed. Although IL-15 mRNA was found in various cell populations, IL-15 protein is secreted at very low levels and is not easily detectable under physiological conditions (47). Furthermore, DC-derived IL-15 may have some unique abilities to its biological functions (39, 48-49). Therefore, the restricted expression and secretion of IL-15 by DC was designed. This study was aimed at demonstrating that transcriptional IL-15-directed DC targeting can improve DC activation ex vivo, determining if transcriptional IL-15-directed in vivo DC targeting can augment Aghsp70 DNA vaccine-elicited Ag-specific Th1 and CTL responses and protective immunity in both normal and CD4 deficient mice, and proving if in vivo DC-derived IL-15 is crucial for enhancing immune responses and antitumor immunity.
IL-15 transfection stimulates DC function and protects them from tumor-induced apoptosis in vitro (50). Also, DC-derived IL-15 is involved in the resistance to DC maturation-accompanied apoptosis in vitro (51). In this study, transcriptional IL-15-directed DC targeting increased the levels of CD80 and CD86 on DC (Fig.2A), suggesting an enhanced DC ability to stimulate T cells. DC present IL-15 in trans to CD8 T cells and NK cells through IL-15Rα binding IL-15 (52-53). Transcriptional IL-15-directed DC targeting increased the expression of IL-15Rα on DC (Fig.2A), suggesting an enhanced DC ability to promote T and/or NK cell immunity. Furthermore, transcriptional IL-15-directed DC targeting increased the productions of IL-6, IL-12 and TNF-α by DC (Fig.2B). This further suggests that transcriptional IL-15-directed DC targeting might modulate DC activation. In this study, BM-derived DC were generated in the presence of mouse GM-CSF and IL-4 and purified immature CD11c+ DC were transfected by DNA encoding Aghsp70 and IL-15. However, some previous reports have indicated that DC can be produced by the direct action of IL-15 on precursor cells (38, 54-59). Whether DC precursors transfected by DNA encoding Aghsp70 and IL-15 in the presence or absence of mouse GM-CSF and/or IL-4 can influence DC activation needs to be further investigated in future studies.
After immunization, DNA-engineered DC or other cells produce, secrete and process Aghsp70. Soluble unprocessed Aghsp70 drains to lymph nodes where resident DC process and present it to naïve T cells. A second wave of OVAhsp70 reaches the lymph nodes later, by an influx of transfected Aghsp70-bearing DC from the skin, which is required for the effective T cell priming. Transcriptional IL-15-directed DC targeting elevated the level of CCR7 on DC (Fig.2A). This indicates that DC engineered by Aghsp70 and IL-15 may be effective in migration to lymph nodes for potent T cell priming (60). However, whether IL-15-targeted Aghsp70-bearing DC are primed with targeted Aghsp70 in skin or in draining lymph nodes is unclear. It should be established in future studies.
One major goal of vaccine strategies is to elicit Ag-specific CD4 T cells, which are important to generate persisting Ag-specific memory CTL (61-63). Transcriptional IL-15-directed in vivo DC targeting DNA vaccine encoding OVAhsp70 elicited strong long-lasting OVA-specific IFN-γ-producing CD4 T cell responses (these cells did not produce IL-4 or IL-15 upon in vitro restimulation, data not shown) (Fig.3A, C). Therefore, it is not surprising that this DNA vaccine strategy also elicited long-lasting CTL responses (Fig.3B, D), and more importantly, effective tumor protection (Fig.4A). The induction of strong protective Th1 and CTL responses may be associated with the in vivo activation of a minute subset of DC. The precise DC subpopulations that are targeted by IL-15 in this strategy need to be further delineated.
It has been shown that IL-15 codelivered with vaccines overcomes CD4 T cell deficiency for promoting CD8 T cells and DC-derived IL-15 during priming is also necessary for optimal induction of long-lasting CD8 T cells even in the presence of CD4 T cell helpers (49). In this DNA vaccine strategy, even though CD4 T cell depletion during immunization abrogated CD8 T cell-mediated primary tumor protection (Fig.4B), an effective CD8 T cell-dependent long-lasting tumor protection was generated in both normal and CD4 T cell deficient mice (Fig.4C). This suggests that transcriptional IL-15-directed in vivo DC targeting may be important to immunize CD4 T cell deficient individual for an effective long-lasting CTL-based protection.
DC and other skin cells such as epidermal cells transfected by DNA encoding Aghsp70 and IL-15 produce and secrete these molecules. Overexpression of IL-15 in the epidermis promotes cutaneous immunity (64). It was assumed that epidermal cell-derived IL-15 may improve Aghsp70-bearing DC function via a paracrine manner, thereby eliciting the enhanced Ag-specific immune responses and antitumor immunity. However, DC-derived, not epidermal cell-derived, IL-15 improved cellular and humoral immune responses and tumor protection in this DNA vaccine (Fig.5). This suggests that DC-derived IL-15 may modulate DC via an autocrine manner to elicit the optimal immune responses and protective immunity. However, we could not rule out the possibility that the enhanced immune responses and antitumor immunity might be due to the difference in the amount of IL-15 produced at the vaccination site. Regardless of the precise mechanism(s), this observation emphasizes the importance of DC as the resource of IL-15 in DNA vaccinations. We were not able to examine whether IL-15 produced by in vivo DNA vaccine-engineered DC is sufficient to enhance the potency of DNA vaccine using IL-15-/- mice because of the concern that IL15-/- mice lack NK cells and have marked reductions in memory CD8 T cells, NK1.1 T cells, Thy1-CD8a intraepithelial lymphocytes, and abnormal development of DC from early ontogeny (58, 65).
Obviously, the use of OVA model Ag in C57BL/B6 mice can not properly predict the possibility of this DNA vaccine strategy in a clinically relevant setting (19). To create an effective DNA vaccine with the possible potential clinical applicability, we tested the efficacy of protective tumor immunity elicited by this DNA vaccine strategy in a clinically relevant B16 melanoma using the self/tumor Ag (TRP2). Transcriptional IL-15-directed in vivo DC targeting greatly enhanced the TRP2hsp70 DNA vaccine-elicited tumor protection in a clinically relevant B16 melanoma model (p<0.005) (Fig.6). Likely, this novel DNA vaccine strategy can be translated into a clinically relevant setting.
The data suggest that transcriptional IL-15-directed in vivo DC targeting enhanced the efficacy of DNA vaccine encoding Aghsp70 but not Ag alone (Figs.2--4,4, ,6).6). Persistent Ag delivery and stabilization of Ag presentation may be an important consequence of DC activation in vivo (66). It is therefore possible that the expression, secretion and processing and presentation of Aghsp70 by DNA-engineered DC may be important to collaborate with DC-derived IL-15 for optimal improvement of DC activation and inducing T cell immunity. The precise mechanisms need to be investigated in future studies. Furthermore, because enhancements in immunity to tumors are mediated by breaking tolerance to self Ag, the immunity to the coexpressed IL-15, which is a ‘self Ag’, and the potential consequences for clinical applications are worthwhile to be addressed in future studies.
In summary, we demonstrated the proof-of-concept of a novel transcriptional IL-15-directed in vivo DC targeting DNA vaccine that improves DC activation, elicits long-lasting Ag-specific Th1 and CTL responses and tumor protection, generates CD8 T cell-dependent long-lasting protective immunity in both normal and CD4 deficient mice and is effective in a clinically relevant tumor model. These findings will be important for developing next generation vaccines against cancers and viral infections in both healthy and CD4 deficient individuals.
We are indebted to T. A. Waldmann (NIH), T. Brocker (Ludwig-Maximilians-Universitat Munchen), G. Erdos (University of Pittsburgh) and E. Lord and J. Frelinger (University of Rochester) for provide reagents. We also are indebted to R. M. Steinman (The Rockefeller University) for his constructive help. This work was supported by NIH grant R01CA108813 (to Z.Y.), R01 AI076060, CA106662, and P01 CA73743 (to L.D.F.).
Disclosures: The authors have no financial conflict of interest.