Whole tumor antigen vaccination using dendritic cells: Comparison of RNA electroporation and pulsing with UV-irradiated tumor cells

Because of the lack of full characterization of tumor associated antigens for solid tumors, whole antigen use is a convenient approach to tumor vaccination. Tumor RNA and apoptotic tumor cells have been used as a source of whole tumor antigen to prepare dendritic cell (DC) based tumor vaccines, but their efficacy has not been directly compared. Here we compare directly RNA electroporation and pulsing of DCs with whole tumor cells killed by ultraviolet (UV) B radiation using a convenient tumor model expressing human papilloma virus (HPV) E6 and E7 oncogenes. Although both approaches led to DCs presenting tumor antigen, electroporation with tumor cell total RNA induced a significantly higher frequency of tumor-reactive IFN-gamma secreting T cells, and E7-specific CD8+ lymphocytes compared to pulsing with UV-irradiated tumor cells. DCs electroporated with tumor cell RNA induced a larger tumor infiltration by T cells and produced a significantly stronger delay in tumor growth compared to DCs pulsed with UV-irradiated tumor cells. We conclude that electroporation with whole tumor cell RNA and pulsing with UV-irradiated tumor cells are both effective in eliciting antitumor immune response, but RNA electroporation results in more potent tumor vaccination under the examined experimental conditions.


Introduction
Because tumor-associated antigens are not well characterized for the majority of human tumors, polyvalent vaccines prepared with whole tumor antigen are an attractive approach to induce tumor vaccination [1,2]. Recent advances in generation and manipulation of DCs provide opportunities to design powerful tumor vaccines. DCs are ideal vehicles for polyvalent tumor vaccination, as they readily process and present tumor antigen taken up from dying tumor cells.
DCs pulsed with apoptotic tumor cells have been used successfully to induce tumor vaccination [3][4][5][6][6][7][8][9][10][11][12]. Although controversy surrounds the ability of necrotic versus apoptotic tumor cells to serve as a source of multivalent antigen to pulse DCs [10,[13][14][15], UVB irradiation has been shown to result in a mixed population of apoptotic and necrotic tumor cells [16]. Tumor cells exposed to lethal ultraviolet-B (UVB) radiation have been shown to provide a suitable source of tumor antigen for DCs [16,17]. For example, UV-irradiated primary tumor cells provide sufficient tumor antigen to elicit expansion of tumor-reactive autologous T cells ex vivo in patients with advanced ovarian cancer [17], suggesting that this approach can be used clinically to induce therapeutic vaccination.
Several reports have described the use of tumor-extracted RNA as source of tumor antigen for the preparation of DCs and have indicated its potential use for antigen-specific or polyvalent tumor vaccination in the absence of identified tumor antigens [18][19][20][21][22][23]. Such approach may address important limitations in the procurement of tumor antigen, as primary tumor cell cultures are not feasible for a large number of patients. Although the feasibility and efficacy of electroporation of DCs with RNA for the preparation of polyvalent tumor vaccines has been convincingly demonstrated, a direct comparison of DC vaccines prepared with tumor RNA versus dying whole tumor cells has not been performed.
In this study, we compared tumor RNA to apoptotic tumor cells as a source of tumor antigen to generate a DCbased vaccine against tumors expressing the early gene products E6 and E7 of the human papilloma virus (HPV). We report that the use of tumor RNA as a source of tumor antigen is valuable alternative and superior to UV-irradiated tumor cells.
ID8, a cell line derived from spontaneous in vitro malignant transformation of C57BL/6 mouse ovarian surface epithelial cells, was a generous gift from Dr. Paul F. Terranova, University of Kansas [24]. ID8 cells were maintained in DMEM medium (Invitrogen) supplemented with 4% FBS, penicillin, streptomycin, insulin (5 µg/ml), transferrin (5 µg/ml), and sodium selenite (5 ng/ml, all Roche, Indianapolis, IN) in a 5% CO 2 atmosphere at 37°C. An ID8 cell line expressing the HPV16 E6 and E7 antigens was generated by transducing ID8 cells with the retroviral vector LXSN16E6E7 (American Type Culture Collection, Rockville, MD, donated by Dr. D. Galloway), which encodes the HPV16 E6 and E7 genes, as well as the neomycin phosphotransferase gene. The PA317 cell line was used to generate the retroviral vectors as previously described [25]. Selection of ID8 cells transduced with E6 and E7 (ID8-E6/7) or ID8 cells transduced with a control retroviral vector (LXSN) was achieved under neomycin pressure (1 mg/ml) [25]. The murine L929 (ATCC) immortalized cell line was grown in RPMI 1640 with 10% FBS and penicillin/streptomycin. All lines tested negative for Mycoplasma by PCR.

Animals and tumors
Six to eight week old female C57BL/6 (H-2K b ) and BALB/ c (H-2K d ) mice (Charles River Laboratories, Wilmington, MA) were used in protocols approved by the Institutional Animal Care and Use Committee and the University of Pennsylvania. TC-1 tumors were generated in C57BL/6 mice by s.c. inoculation of 2 × 10 4 TC-1 cells in 0.2 ml of PBS. Tumors were detectable ten days later and were measured weekly using a Vernier caliper. Tumor volumes were calculated by the formula V= 1/2 L × W 2 , where L is length (longest dimension) and W is width (shortest dimension). For some in vivo studies, CD8 + cells were depleted with rat anti-mouse CD8 (MCA1768XZ) and CD4 + cells with rat anti-mouse CD4 (MCA1767XZ) (both Serotec, Raleigh, NC). The antibodies were administered intravenously (100 µg/animal) on the day of tumor injection and a second dose one week later.

Generation of bone marrow-derived DCs
Murine dendritic cells were generated from bone marrow precursor cells with recombinant murine granulocytemacrophage colony-stimulating factor (GM-CSF; Peprotech, Rocky Hill, NJ; 20 ng/ml) as described previously [26]. Cells were counted using Trypan blue. Differentiation into immature DCs was documented through flow cytometry detection of CD80, CD86 and major histocompatibility complex class II (MHC-II). DC maturation was induced by culturing cells in RPMI media under standard conditions in the presence of 10 ng/ml murine GM-CSF supplemented with 0.1 µg/ml lipopolysaccharide (LPS, Sigma Chemical Co, Saint Louis, MO) and 20 ng/ml tumor necrosis factor-alpha (TNF-α, Peprotech).

DC electroporation with tumor RNA
Total cellular RNA was extracted from TC-1 cells using TRIzol Reagent (Invitrogen). Cells grown in 75 cm 2 flasks were resuspended and lysed using TRIzol reagent. Chloroform (0.2 ml per ml of TRIzol reagent) was added and incubated at room temperature for 2 min. The samples were centrifuged at 12,000 × g for 15 min at 4°C, and the aqueous phase was transferred to a new tube. Cold isopropanol was added at 0.5 ml per ml TRIzol reagent to precipitate RNA. Following 10 min incubation at room temperature the samples were centrifuged at 12,000 × g for 10 min at 4°C. The RNA pellet was washed once with 70% DEPC-ethanol and centrifuged at 7500 × g for 5 min. The pellet was briefly dried and dissolved in DEPC water. The quality and quantity of the total RNA was checked using RNA Nano LabChip ® (Agilent Technologies, Palo Alto, CA) according to the protocol provided. Two million DCs were resuspended with the appropriate amount of total TC-1 RNA in a 0.2-cm cuvette and electroporated using Gene Pulser II (BIO-RAD Laboratories, Hercules, CA) under different voltage and capacitance settings. DCs electroporated in the absence of TC-1 RNA (mock) were used as controls for some experiments.

DC pulsing with apoptotic tumor cells
Subconfluent cultures of TC-1 cells were rinsed twice in phosphate buffer saline (PBS) and exposed to ultraviolet-B (UVB) radiation at various doses up to 1500 µW/cm 2 for various times. Apoptosis at 24 hours was quantified by flow cytometry detection of annexin-V staining using the TACS™ Annexin-Biotin Apoptosis detection kit (R&D Systems, Minneapolis, MN) and confirmed with the ApopTag peroxidase in situ detection kit (Intergen, Purchase, NY) and the Apoptotic DNA-Ladder Kit (Roche), according to the manufacturers' instructions. Twenty-four hours after UVB radiation tumor cells were incubated with immature DCs at a 2:1 ratio (tumor cells, DCs). Twenty-four hours later, TNF-α (Peprotech; 20 ng/ml) and LPS (0.1 µg/ml) were added for additional 48 hours. DCs were harvested, rinsed and counted by trypan blue exclusion. In some experiments, radiated tumor cells were labeled with PKH26 fluorescent dye (Sigma; 5 µM) for 5 min at room temperature. RPMI supplemented with 10% FBS was added to stop the reaction and cells were rinsed three times prior to using them for pulsing DCs.

Animal vaccination
Animals received one intraperitoneally (i.p.) and then twice subcutaneously (s.c.) seven days apart, DCs (5 × 10 5 per dose) electroporated with TC-1 RNA or loaded with TC-1 cells killed by UVB and incubated with TNF-α (20 ng/ml) and LPS (0.1 µg/ml) were injected once. DCs. Control animals were injected with DCs mock electroporated and matured with TNF-α and LPS. Animals were challenged with tumor cells seven days after the last DC vaccination.

Flow cytometry
Cells were subjected to four-color flow cytometry on a FACSCalibur flow cytometer using CellQuest 3.2.

Chemotaxis assay
Migration of DCs towards murine macrophage inflammatory protein (MIP)-3α or MIP-3β (R&D Systems, Minneapolis, MN) was assessed in 96-well chemotaxis chambers using an 8 µm-pore nitrocellulose membrane (Neuroprobe, Gathersburg, MD) [29]. Pyrogen-free RPMI 1640 containing 1% BSA was used as chemotactic media. Results are presented as chemotactic index (CI), defined as fold increase in cell migration in the presence of chemotactic factors compared to chemotactic media alone. Each experiment was performed in triplicate.
For immunofluorescense of DCs following electroporation with TC-1 RNA, electroporated DCs were seeded on glass coverslips and cultured in RPMI in the presence of 10% FBS and 10 ng/ml GM-CSF for two days. Cells were fixed with acetone and consecutively stained with anti-HPV-16 E7 antibody (ED-17, Santa Cruz Biotechnologies) and anti-rabbit FITC (BD Pharmingen). Slides were counterstained with DAPI. Images were acquired through Cool SNAP Pro color digital camera (Media Cybernetics, Carlsbad, CA). Ten different fields for each sample at × 400 magnification were evaluated for cell counting.
The AbiPrism 7700 Sequence Detection System and SYBR green I PCR kits (both Applied Biosystems, Foster City, CA) were used for Real-Time PCR as described previously [30]. The following primers were used: E6: F 5'-GAC TTT GCT TTT CGG GAT TTA TGC -3', R 5'-TCA CAC AAC GGT TTG TTG TAT TGC-3'; E7: F 5'-CTG GAC AAG CAG AAC CGG ACA-3', R 5'-TGC TTT GTA CGC ACA CCG AA-3'. We normalized the cDNA load to mouse glyceraldehyde-3-phosphate dehydrogenase (GAPDH) with primers GAPDH F 5'-CCT GCA CCA CCA ACT GCT TA-3' and GAPDH R 5'-CAT GAG TCC TTC CAC GAT ACC A-3'. Data were expressed as relative units to GAPDH mRNA molecules. Molecules were considered to be present if more than five copies of mRNA were detected for every 10 6 copies of GAPDH mRNA.

Statistical analysis
A two-tailed Student's t-test was applied to determine differences between two groups. For multiple comparisons we performed ANOVA with post-analysis comparisons by the Tukey-Kramer multiple comparisons test. Non-parametric studies were performed by using the Mann-Whitney U test. A value of p < 0.05 was considered significant. Data are expressed as mean ± SD. Data was analyzed using Graph Pad Instat software (GraphPad Software, Inc., San Diego, CA).
To determine the best conditions for RNA electroporation in our system, bone marrow-derived DCs were electroporated with tumor cell RNA using different capacitance and voltage settings and 25 µg RNA per 10 6 DCs, which represents RNA from approximately one tumor cell per DC. The reaction was performed in a total volume of 200 µL. As shown in Figure 1C, the highest expression of E7 antigen in live electroporated DCs, was obtained at 300 V and 150 µF. These settings yielded 50% viability in electroporated DCs, as determined by flow cytometry PI exclusion analysis ( Figure 1D). Similar viability levels were obtained in DCs electroporated in the absence of RNA (mock, not shown).
To optimize DC electroporation, we electroporated DCs with different amounts of TC-1 RNA (5-50 µg/10 6 DCs) using the previously determined settings (300 V, 150 µF). E6-transcripts are longer than E7. As shown in Figure 1E, high E6 expression was observed only after electroporating 50 µg TC-1 RNA/10 6 DCs. Higher RNA amounts did not result in increased E6 expression (not shown). Elec-Expression of tumor-associated HPV E6 and E7 antigens by bone marrow-derived DCs after RNA electroporation troporation with 50 µg RNA/10 6 DCs also ensured high expression of E7-antigen in DCs ( Figure 1E). As shown in Figure 1F, high expression of E6 and E7 protein was detectable by flow cytometry for at least 4 days after RNA electroporation with 50 µg of TC-1 RNA/10 6 DCs. Immunofluorescence staining confirmed the presence of E7 protein in the cytoplasm of DCs 24 hours after electroporation with 50 µg TC-1 RNA/10 6 DCs ( Figure 1G).

Preparation of DCs pulsed with UV irradiated TC-1 cells (UV-DCs)
We tested various doses of ultraviolet-B (UVB) light and exposure times to identify UVB conditions that kill greater than 95% of TC-1 cells (not shown). Irradiation with 1500 µW/cm 2 UVB for 10 min induced apoptosis in TC-1 cells as assessed by DNA fragmentation detectable by TUNEL assay (Figure 2A) and DNA laddering ( Figure 2B), while by flow cytometry the majority of cells were annexin-V positive or propidium iodide and annexin-V double positive within 24 hours ( Figure 2C).
To verify the uptake of UV-irradiated cells by DCs, UVirradiated TC-1 cells were labeled with PKH26 fluorescent dye prior to pulsing of DCs. DCs and UV-irradiated tumor cells were cocultured for 18 hours to allow the uptake of tumor cell material by DCs. Cells were then stained with antibody against CD11c, and CD11c + DCs were analyzed for PKH26 expression with flow cytometry. More than 60% of DCs had taken up fluorescent tumor cells ( Figure  2D), compared with a background level of 1% in the control sample containing DCs and UV-irradiated cells admixed just before analysis (not shown).

Maturation of RNA-DCs and UV-DCs
We assessed whether DCs prepared by RNA electroporation or pulsing with UV-irradiated tumor cells respond differently to maturation stimuli such as TNF-α and LPS. Significant surface expression of CD86 and CD80 and MHC-II molecules was noted in DCs 48 hours post-electroporation with tumor cell RNA, as described above, in the presence of TNF-α and LPS ( Figure 3A). Similarly, immature DCs incubated with TC-1 cells exposed 24 hour earlier to lethal dose of UVB radiation upregulated CD80, CD86 and MHC-II 48 hours post-phagocytosis, in the presence of TNF-α and LPS ( Figure 3B).

MIP 3 alpha MIP 3 beta
A switch in chemokine receptors is a hallmark of DC maturation. Among others, this entails upregulation of CCR7 and downregulation of CCR6 [35]. RNA-DCs and UV-DCs matured with TNF-α and LPS exhibited similar migratory properties. As shown in Figure 3D, both DCs lost the ability to migrate towards macrophage inflammatory protein 3-alpha (MIP3-α), the ligand for CCR6, and acquired the ability to migrate towards MIP-3β, the ligand for CCR7.

RNA-DCs and UV-DCs are immunogenic in vivo
We compared the efficacy of RNA-DCs and UV-DCs to induce tumor vaccination. To minimize differences in the amount of tumor antigen between pulsing procedures, DCs were incubated with UV-irradiated cells at 1:2 ratio (DC, tumor cell) or electroporated with an amount of total RNA equivalent to two tumor cells per DC, as above. DCs were matured with TNF-α and LPS. Mock DCs were prepared by electroporation in the absence of tumor RNA followed by maturation with TNF-α and LPS. Healthy animals were vaccinated with RNA-DCs or UV-DCs, while control animals were vaccinated with mock DCs or left unvaccinated (naïve). Splenocytes isolated from the above animals were tested for reactivity against E6 and E7 antigens by incubating them with apoptotic TC-1, ID8-E6/7, ID8 or L-929 cells.
When incubated with TC-1 cells, splenocytes from animals vaccinated with RNA-DCs proliferated significantly more than splenocytes from animals vaccinated with UV-DCs ( Figure 5A). No proliferation was detected in lymphocytes from animals vaccinated with mock-electroporated DCs or naïve animals ( Figure 5A). Similar proliferation was detected when splenocytes from mice vaccinated with RNA-DCs were incubated with TC-1 or ID8-E6/7 cells ( Figure 5B), but no proliferation was detected against control ID8 cells lacking E6 or E7, or L-929 cells. This shows that the presence of HPV antigens was critical for T cell proliferation ( Figure 5B). Splenocytes from all four groups of animals showed similar proliferative response when stimulated with phytohemagglutinin (not shown), indicating no functional impairment.
We evaluated the frequency of tumor-reactive T cells among splenocytes in each group of animals by IFN-γ ELISPOT analysis. We used ID8-E6/E7 cells as target cells. A significantly higher frequency of tumor-reactive IFN-γ producing cells was found in spleens from animals vaccinated with RNA-DCs compared to animals vaccinated with UV-DCs. No response was observed in splenocytes from mock vaccinated or naïve animals ( Figure 5C). Similar response was seen in splenocytes from mice vaccinated with RNA-DCs incubated with TC-1 or ID8-E6/7 cells, while no response was observed against ID8 pr L-929 cells ( Figure 5D). Moreover, in the RNA vaccinated group, IFN-γ producing cells were mainly CD8 + cells, since immunodepletion of CD8 + cells in vivo decreased the number of IFN-γ producing cells among isolated splenocytes ( Figure 5E).
Higher levels of cytotoxic lymphocyte activity were detected in lymphocytes obtained from animals vaccinated with RNA-DCs relative to the UV-DCs group ( Figure  5F). Moreover, CTL activity of splenocytes was abrogated to control levels by immunodepletion of CD8 + cells in vivo ( Figure 5F). No lytic activity was observed when control ID8 cells expressing no E6 or E7, or L-929 cells were used as target cells, confirming the specificity of the reaction (not shown).
TC-1 tumors express HPV E6 and E7, offering the opportunity to quantify T cell responses against tumor-associated HPV epitopes through well-characterized tetramers [37]. We compared the ability of RNA-DCs and UV-DCs to generate T cell response against the H2-D b -restricted HPV E7 epitope RAHYNIVTF [34]. A four-fold higher frequency of tetramer-positive CD3 + CD8 + cells was detected in splenocytes from animals vaccinated with RNA-DCs compared to animals vaccinated with UV-DCs ( Figure  5G).

Vaccination with RNA-DCs or UV-DCs
To test the efficacy of RNA-DCs and UV-DCs in vivo, healthy mice were vaccinated with three injections of RNA-DCs or UV-DCs prepared as above and administered one week apart. Seven days after the last DC vaccination, animals were challenged with flank subcutaneous TC-1 tumors. Tumor growth was significantly delayed in animals vaccinated with RNA-DCs as well as animals vaccinated with UV-DCs ( Figure 6A). Although the difference in tumor growth between mice vaccinated with RNA-DCs and mice vaccinated with UV-DCs was not large, tumors were significantly smaller in mice vaccinated with RNA-DCs. Similar results were obtained in three independent experiments.
Tumors from mice vaccinated with RNA-DCs as well as from animals vaccinated with UV-DCs exhibited significantly higher frequency of CD3 + tumor-infiltrating cells relative to mice vaccinated with mock DCs. A higher frequency of CD3 + tumor-infiltrating cells was detected in animals vaccinated with RNA-DCs compared to animals vaccinated with UV-DCs ( Figure 6B, C). Finally, a higher proportion of CD3 + cells were CD8 + in animals vaccinated with RNA-DCs compared to animals vaccinated with UV-DCs or mock DCs ( Figure 6D).

Discussion
The present work addressed the relative efficacy of tumor vaccines prepared with DCs either electroporated with tumor RNA or with dead tumor cells. We used total RNA from tumor cells isolated with a common laboratory method and UVB irradiated tumor cells and optimized the conditions to minimize difference in amount of tumor cell material used to pulse DCs. We used the E7 MHC-I-restricted epitope RAHYNIVTF to quantify the intensity of CD8 + T cell response against tumor-associated antigens. The present results show that electroporation with whole tumor cell RNA and pulsing with UVB-irradiated tumor cells are both effective in eliciting antitumor immune response, but RNA electroporation results in more potent tumor vaccination. The efficacy of the vaccination by RNA-electroporated DCs was dependent on the presence of CD8 + cells, since in vivo depletion of these cells abrogated the reported effect. Importantly, vaccination with RNA-electroporated DCs expressing E6 and E7 significantly enhanced infiltration of CD3 + CD8 + into the tumors.
To our knowledge this is the first direct comparison between irradiated cells and whole RNA as a source of whole tumor antigen to prepare DC based tumor vaccines. Because the same number of cells was used to derive tumor antigen with both methodologies, the above findings indicate that RNA electroporation is an efficient methodology for loading DCs with tumor antigen. A previous report also found that apoptotic tumor cell pulsing is not an efficient approach to tumor vaccination [38]. Several reasons may account for our findings. First, antigen up-take may be less efficient with pulsing dying cells compared to RNA electroporation. Since UVB irradiation has been shown to result in a mixed population of apoptotic and necrotic tumor cells [16], it is possible that either process leads to degradation of important antigens [39], resulting in suboptimal antigen processing or presentation. Second, apoptotic DNA may bind to MHC class molecules and interfere with antigen presentation [40]. Third, although non-viral methods of DNA transfection of DCs are inefficient, the efficiency of gene transfer with RNA electroporation resembles that of transfection with recombinant viruses [41]. Finally, another advantage of RNA transfer of DC over pulsing DC with protein antigens is that endogenously synthesized antigens have better access to the class I MHC pathway [42].
It is noteworthy to comment that different maturation protocols can modify the capability of DCs to effectively present antigens upon RNA electroporation. In a series of elegant studies, Zobylawski et al., have shown that human DCs treated with a maturation cocktail formulated with TNF, IL-1β, IFN-γ, prostaglandin E2, and the Toll like receptor (TLR) 8 agonist R848 were able to generate an efficient immune response upon RNA electroporation; while addition of the TLR3 ligand poly(I:C) to the maturation cocktail rendered DCs unable to express proteins from electroporated RNA [43]. Indeed, the formulation of appropriate maturation cocktails is one of the challenges that faces the generation of effective DC-based vaccines for clinical use [44]. Thus, the use of different activation protocols might have produced different results in our studies. Further, since in our studies we used murine bone marrow-derived DCs, they may not directly translate to human monocyte-derived dendritic cells used for clinical studies. Additional studies with human monocytederived dendritic cells must be performed in order to determine the clinical relevance of our findings. Finally, it should be noted that our work compared RNA to dead tumor cells and our findings on superiority of RNA as a source of whole tumor antigen may not be relevant to alternate methods of preparing whole tumor cell protein, such as tumor lysates [45], which may preserve tumor antigens or do not interfere with antigen presentation.

Conclusion
Collectively, our data suggest that electroporation of whole tumor RNA represents a direct and effective way of delivering tumor antigen to DCs ex vivo. Coupled with the easier procurement of tumor RNA compared to the generation of tumor cell lines, these findings suggest that RNA electroporation should be a preferred method of loading DCs with whole tumor antigen in clinical trials.