The first two authors contributed equally to this paper.
MIDGE/hNIS vaccination generates antigen-associated CD8+IFN-γ+ T cells and enhances protective antitumor immunity
Article first published online: 31 JAN 2007
Copyright © 2007 Wiley-Liss, Inc.
International Journal of Cancer
Volume 120, Issue 9, pages 1942–1950, 1 May 2007
How to Cite
Choi, Y., Jeon, Y.-H., Kang, J.-H., Chung, J.-K., Schmidt, M. and Kim, a. C.-W. (2007), MIDGE/hNIS vaccination generates antigen-associated CD8+IFN-γ+ T cells and enhances protective antitumor immunity. Int. J. Cancer, 120: 1942–1950. doi: 10.1002/ijc.22567
- Issue published online: 28 FEB 2007
- Article first published online: 31 JAN 2007
- Manuscript Accepted: 4 DEC 2006
- Manuscript Received: 28 AUG 2006
- Korean Science & Engineering Foundation (KOSEF)
- antitumor vaccination;
- MIDGE vectors;
- monitoring technique;
- CD8+ T cells;
- CT26 tumor model
Human sodium iodide symporter (hNIS) is a transmembrane protein that actively transports iodide ions into thyroid cells. hNIS is over-expressed in some cases of the thyroid cancers compared with the surrounding normal tissues and has been considered to be an attractive target for immunotherapy. The aim of this study is to determine the feasibility of utilizing the hNIS antigenic protein in enhanced-antigen-associated immunotherapy using image analysis with a gamma counter. To accomplish this, minimalistic immunogenically defined gene expression (MIDGE), either plain or coupled to a nuclear localization signal (NLS) peptide, was used as a vector system. Vaccination with MIDGE/hNIS, MIDGE/hNIS-NLS and pcDNA3.1/hNIS produced a significant increase in the number of hNIS-associated IFN-γ-secreting CD8+ T cells, with MIDGE/hNIS having the strongest effect. In addition, immunization with the hNIS encoding vectors induced antigen-mediated antitumor activity against NIS-expressing CT26 tumors in vivo, with the highest tumor free rate (100%) and lowest tumor growth being observed up to 40 days after the CT26/NIS tumor challenge with MIDGE/hNIS than those resulting from other immunization groups. Tumor progression could be followed noninvasively and repetitively by monitoring levels of hNIS gene expression in the tumors using scintigraphic image analysis. Overall, hNIS has a potential use as an antigen for immunization approaches, and vaccination with MIDGE/hNIS vectors is an effective means of generating hNIS-associated immune responses in mice. © 2007 Wiley-Liss, Inc.
DNA vaccines can induce both cellular and humoral immune responses and have become an attractive immunization strategy for protecting the host in various experimental models against diseases, such as infections, cancer and autoimmunity.1, 2, 3 It is believed that DNA vaccines are potentially safer than traditional vaccines, having the advantages of stability and cost effectiveness in terms of manufacturing and storage. In addition, multiple antigens could be combined into a single plasmid to target multiple pathogens or multiple components of a single pathogen.4 However, thus far, DNA vaccines have shown low immunogenicity when tested on large animals and humans.5
There is a need for strategies that can enhance the immunogenicity elicited by DNA vaccines expressing the antigens of interest. The potency of a DNA vaccine in terms of eliciting an effective immune response is associated with the levels the encoded proteins in eukaryotic cells.6 Several vector systems have been used in clinical gene transfer trials to increase the efficiency of DNA delivery with retroviral and adenoviral vectors being most common.7, 8 Although the viral vectors can effectively and stably transduce the target cells, they are also associated with risks, such as recombination with wild-type viruses and the activation of protooncogenes.9 A plasmid vector, which is comprised of double-stranded circular DNA, can avoid these risks. However, the efficiency of DNA delivery is usually lower than those of viral vectors.7 The expression plasmids used for DNA based vaccination usually contain a transcription unit as well as bacterial sequences, and the presence of constitutive genes in a plasmid can have a strong influence on the outcome of the immunization.10 Furthermore, the application of plasmid DNA as a vaccine can disseminate antibiotic resistance genes. The minimalistic immunogenically defined gene expression (MIDGE) vector system can address both issues by reducing the sequence content of the transfected DNA to the expression cassette encoding the antigen and eliminating the antibiotic resistance gene as well as other nonessential sequences.10 The MIDGE vector is linear plasmid vector containing only the expression cassette (CMV enhancer/promoter region, antigen encoding DNA and a chimeric intron and polyadenylation sequences) and lacks selection marker genes. Nuclear localization signal (NLS) peptides can be conjugated to DNA-based vectors in order to increase their transfection efficiency.11, 12, 13, 14 Previously Moreno et al.15 and Schirmbeck et al.16 reported the expression level of encoded antigens along with the immune responses obtained using the MIDGE vector administered through various injection sites. It was shown that MIDGE-NLS vectors are as efficient as plasmids in transferring genes into the various cell lines or through an in vivo injection. Moreover, immunization with MIDGE-NLS vectors has been reported to generate more antigen specific CD8+ T cells than immunization with the plasmid or MIDGE.15, 16
The human sodium-iodide symporter (hNIS) belongs to the sodium/glucose cotransporter family that can transport iodide into thyroid cells.17, 18 Most methods used for imaging gene expression from transplanted ‘reporter’ genes in rodents, primates and humans require the tissue to be obtained postmortem. However, new methods using hNIS as a reporter gene labeled with 124I, 125I, 131I or Tc-99m have allowed the visualization of various biochemical processes in the tissues of living subjects. As with other reporter genes, the hNIS gene has also been examined in the context of the long-term imaging of gene therapy experiments.19, 20, 21, 22
If a proper immune response can be induced through in vivo vaccination using the hNIS gene, this vaccination strategy would be helpful for 2 reasons, i.e., targeting immunotherapy against NIS expressing cancer and as an innovative and objective tool for evaluating the efficacy of a vaccine using scintigraphic imaging.
Therefore, this study examined whether MIDGE/hNIS or MIDGE/hNIS-NLS can induce more hNIS gene expression than pcDNA3.1/hNIS, and whether or not an immune response can be induced with hNIS as an antigenic peptide. The results showed that the plasmid vector, pcDNA3.1/hNIS, enhanced the expansion and activation of the hNIS-associated CD8+ T cells and generated significant CD8+ T cell dependent tumor preventive effects against NIS expressing CT26 tumors in vaccinated mice compared with the empty vector (pcDNA3.1) treated mice. On the other hand, both MIDGE/hNIS and MIDGE/hNIS-NLS generated much stronger hNIS-associated CD8+ T cell responses and a greater antitumor effect against NIS expressing CT26 tumors in vaccinated mice than pcDNA3.1/hNIS.
Material and methods
Construction of plasmids, MIDGE and modified MIDGE
The human sodium iodide symporter (hNIS) expressing vector, FL-hNIS/pcDNA3.1 (pcDNA3.1/hNIS), was a kind gift from Dr. S. Jhiang (Ohio State University, Columbus, OH) and controlled using a CMV promoter, which also contained a neomycin resistance gene under the control of the SV40 promoter. The (minimalistic immunogenically defined gene expression) MIDGE/hNIS and modified MIDGE/hNIS-NLS were generated using the method described elsewhere.10, 15, 16 Briefly, the MIDGE/hNIS construct was derived from the hNIS plasmid after amplification using PCR to generate a recognition site for the restriction enzyme, KpnI, at the 5′-end, and a SacI recognition site at the 3′-end of the PCR product. The plasmid, pcDNA3.1-hNIS, was used as a template. After digestion with KpnI and SacI (MBI Fermentas, Vilnius, Lithuania), the PCR product was ligated into the pMCV1.4-vector, which was verified by sequencing. The ends of the Eco31I (MBI Fermentas, Vilnius, Lithuania) digested plasmid were used to ligate the hairpin oligodeoxyribonucleotides (ODNs) with T4 DNA ligase. The MIDGE/hNIS construct was purified using anionic exchange column chromatography (Merck EMD-DMAE). The MIDGE/hNIS-NLS constructs were produced using the peptide (PKKKRKVEDPYC) coupled ODNs instead.16 The resulting NLS-coupled ODN was purified by HPLC.
Specific pathogen-free 6-week-old female Balb/C mice were obtained from SLC (Japan). All the animals were housed under specific pathogen-free conditions and were handled in accordance with the guidelines of the Seoul National University Animal Research Committee.
Transfection and antigen expression in vitro
The CT26 tumor cells (1 × 105) were transfected transiently with equimolar amounts of the different DNA vectors (pcDNA3.1 and pcDNA3.1/hNIS: 0.4 μg; MIDGE/hNIS and MIDGE/hNIS-NLS: 0.22 μg) using lipofectamine (Gibco BRL, UK). Two days later, hNIS expression was detected by the radioactive iodide uptake. Briefly, after transfection, the cells were seeded in 24-cell plates. The iodide uptake was examined by incubating the cells with 0.5 ml Hanks balanced salt solution (HBSS) containing 0.5% bovine serum albumin and 10 mM of 2-[4-(2-hydroxyethyl)-1-piperazinyl] ethanesulfonic acid-NaOH, 3.7 kBq carrier-free 125I and 10 μM NaI, at pH 7.4 to yield a specific activity of 740 MBq/mmol at 37°C for 30 min. The cells were washed twice with 2 ml cold iodide-free HBSS and lysed with 0.5 ml 0.2% SDS. The concentrated iodide was measured using a gamma counter (Canbera, Meriden, CT). The experiments were performed in triplicate and the iodide uptake is expressed as CPM.
DNA inoculation and antigen expression in vivo: Autoradiography of mouse muscle
The Balb/C mice were inoculated intramuscularly (i.m.) with 100 μg of pcDNA3.1 (hind-left leg) or pcDNA3.1/hNIS (hind-right leg). The groups inoculated with MIDGE/hNIS (hind-right leg) or MIDGE/hNIS-NLS (hind-right leg) received 54.8 μg per mouse, which is an equimolar concentration of the plasmid. Autoradiography was used to determine the distribution of Tc-99m in the mouse muscle. Forty-eight hours after DNA inoculation, Tc-99m (600 μCi, 0.1 ml) was injected intravenously into the female Balb/C mice weighing 20–25 g. Within 20 min of the Tc-99m injection, the animals were sacrificed by decapitation. The bilateral hind-leg muscles were removed and frozen immediately at −20°C. The radioactivity and muscle weights were measured using a gamma counter and an electric balance, respectively. The frozen muscles were sectioned at 10 μm and mounted on slides. Imaging plates (Fujifilm, Japan) were exposed to the slides overnight, and the exposed imaging plates were analyzed using a BAS-2000 (Fujifilm FLA-2000, Fujifilm, Japan).
The Balb/C mice were immunized i.m. with 100 μg of pcDNA3.1 or pcDNA3.1/hNIS in a 100 μl volume. The groups primed with MIDGE/hNIS or MIDGE/hNIS-NLS received 54.8 μg per mouse. After 1 week, the mice were boosted with an i.m. injection of the same amount of DNA.
Intracellular cytokine flowcytometry analysis
The Balb/C mice were immunized with the different DNA vectors twice i.m. at 1-week intervals. Ten days after the final immunization, the draining lymphoid cells (5 × 106) were extracted from the lymph nodes of the mice. The cells were then stimulated with the irradiated (50 Gy) CT26/NIS tumor cells (1 × 106). For intracellular cytokine staining, the 42 hr-stimulated draining lymphoid cells were treated with golgistop (PharMingen, San Diego, CA) and incubated for a further 6 hr. The cells were harvested and stained with FITC-conjugated anti-CD4 or CD8 antibody and PE-conjugated anti-IFN-γ antibody. The analyses were performed using a B-D FACScan equipped with CELLQuest software (Becton-Dickinson Immunocytometry System, Mountain View, CA).
Murine tumor cell line
The NIS transfected CT26 cells were kindly donated by Shin.22
In vivo tumor protection experiments
For the tumor protection studies, the mice (4 per group) were vaccinated i.m. with 100 μg per mouse of pcDNA3.1 or pcDNA3.1/hNIS in a 100 μl volume. The groups primed with MIDGE/hNIS or MIDGE/hNIS-NLS received 54.8 μg/100 μl per mouse, which is an equimolar concentration of the plasmids. One week later, the mice were boosted with the same amount of DNA. One week after the final vaccination, the mice were challenged subcutaneously (s.c.) with CT26/NIS tumor cells at 5 × 104 cells/mouse in the hind-right legs. The tumor volumes were measured with calipers every 5 or 7 days from 18 days after the tumor challenge. The animals were euthanized if the tumor became necrotic or exceeded 550 mm2 in size. The tumor volumes were defined as the length (mm) × width (mm).
Scintigraphy using Tc-99m
CT26/NIS tumor cells (5 × 104) were inoculated s.c. into the mice that had previously been immunized twice at 1-week intervals either with the empty plasmid (pcDNA3.1), pcDNA3.1/hNIS, MIDGE/hNIS or MIDGE/hNIS-NLS. The induction of the immune response by the plasmid vector (pcDNA3.1/hNIS) or MIDGE vector (MIDGE/hNIS, MIDGE/hNIS-NLS) was monitored by imaging the mice at 30, 35 and 40 days after the tumor challenge using a gamma camera. Briefly, 30 min after the intraperitoneal (i.p.) injection of 600 μCi of Tc-99m, the tumor-bearing mice were placed in a spread-prone position, and scanned with the gamma camera (ON 410; Ohio Nuclear, Solon, OH) equipped with a pinhole collimator. The regions of interest (ROI) were drawn over the tumor region and the total counts/total pixel numbers were determined.
The statistical significance was determined using an unpaired Student's t test. Kaplan-Meier curves were generated for the time to survival of the mice. The curves for 2 groups were compared using the log-rank test for the equality of survivor functions. A value of p < 0.05 was considered significant.
Human NIS encoding MIDGE vectors
The human NIS encoding sequence was inserted into the pcDNA3.1 plasmid vector to generate pcDNA3.1/hNIS (Fig. 1a). pcDNA3.1/hNIS was under the control of the human CMV promoter and contained a neomycin resistance gene under the control of a SV40 promoter. The MIDGE/hNIS construct was generated from the pcDNA3.1/hNIS plasmid, as described in Materials and Methods (Fig. 1b). MIDGE/hNIS is a linear construct with covalently closed ends. This expression vector contained only the promoter, intron, hNIS encoding and polyadenylation sequences (Fig. 1b). A nuclear localization sequence signal (NLS) was linked to the MIDGE/hNIS vector to increase the transfection efficiency and expression level of the antigen (Fig. 1c).
In vitro expression of pcDNA3.1/hNIS and MIDGE vectors
The in vitro transient expressions of hNIS from pcDNA3.1/hNIS, and the corresponding MIDGE/hNIS and MIDGE/hNIS-NLS vectors were tested in the transfected CT26 tumor cells, which were analyzed by the radioactive 125I uptake. The observed uptakes of 125I were 904 ± 60, 2,536 ± 130, 2,481 ± 78 and 2,301 ± 58 cpm/1 × 105 cells in the pcDNA3.1, pcDNA3.1/hNIS (p < 0.001 vs. pcDNA3.1), MIDGE/hNIS (p < 0.0001 vs. pcDNA3.1 or p = 0.037 vs. MIDGE/hNIS-NLS) and MIDGE/hNIS-NLS (p < 0.0001 vs. pcDNA3.1) transfected cells, respectively. The radioactive 125I uptake assays demonstrated that the hNIS activity in the pcDNA3.1/hNIS transfected CT26 tumor cells was 2.8 times higher than in their empty vector transfected counterparts.
The MIDGE/hNIS transfected cells expressed similar levels of the hNIS genes as pcDNA3.1/hNIS and MIDGE/hNIS-NLS transfected cells. Surprisingly, MIDGE/hNIS-NLS transfection did not result in increased gene expression compared with pcDNA3.1/hNIS or MIDGE/hNIS transfection (Fig. 2).
In vivo expression of pcDNA3.1/hNIS and MIDGE vectors
The in vivo expression of hNIS was determined by autoradiography in excised muscle samples from the mice inoculated with pcDNA3.1/hNIS, MIDGE/hNIS and MIDGE/hNIS-NLS. Autoradiography demonstrated that the hNIS activity was 40% higher in the pcDNA3.1/hNIS injected muscle than in the pcDNA3.1 injected muscle (Fig. 3).
Surprisingly, pcDNA3.1/hNIS expressed a slightly higher level of the hNIS protein than either MIDGE/hNIS or MIDGE/hNIS-NLS, but this difference was not significant. Attaching a nuclear localization signal (NLS) to the end of MIDGE/hNIS did not cause an increase in antigen (hNIS) expression in vivo (Figs. 3b and 3c).
Vaccination with MIDGE/hNIS significantly induced hNIS-associated CD8+T cell immune response
The ability of the MIDGE vector expressing hNIS to enhance the NIS associated CD8+T cell immune responses was examined using intracellular cytokine staining and flow cytometric analysis, which were used to count the number of hNIS-associated IFN-γ-secreting CD8+T cells generated in the vaccinated mice.
As shown in Figure 4a, there was a higher number of hNIS-associated IFN-γ-secreting CD8+T cells in the mice vaccinated with MIDGE/hNIS than in the other vaccination groups (pcDNA3.1, pcDNA3.1/hNIS and MIDGE/hNIS-NLS). MIDGE/hNIS generated a significant increase in the number of hNIS-associated IFN-γ-secreting CD8+T cells (286 ± 53/3 × 104 lymph node cells) compared with pcDNA3.1 (24 ± 5/3 × 104 cells, p = 0.013). This translates into an 11.9-fold increase in the number of hNIS-associated IFN-γ-secreting CD8+T cells (Figs. 4a and 4b). An injection of the MIDGE/hNIS-NLS (p = 0.009 vs. pcDNA3.1 or p = 0.033 vs. MIDGE/hNIS) and the pcDNA3.1/hNIS (p = 0.019 vs. pcDNA3.1 or p = 0.016 vs. MIDGE/hNIS) still enhanced the number of hNIS-associated IFN-γ-secreting CD8+T cells but to a lesser extent (152 ± 24/3 × 104 cells and 106 ± 22/3 × 104 cells, respectively). These results show that MIDGE/hNIS vaccination significantly enhanced the NIS associated CD8+ T cell immune responses in the vaccinated mice.
Vaccination with MIDGE/hNIS-NLS significantly enhanced the hNIS-associated CD4+T cell immune response
The ability of the hNIS-expressing MIDGE vectors to enhance the hNIS-associated CD4+T cell immune responses was assessed by analyzing the lymph node cells obtained from the vaccinated mice using flow cytometry via double staining of the CD4 surface marker and intracellular IFN-γ.
As shown in Figures 4c and 4d, the mice vaccinated with MIDGE/hNIS-NLS (129 ± 9/3 × 104 cells) showed a significantly higher number of hNIS-associated IFN-γ-secreting CD4+T cells than the pcDNA3.1 (52 ± 4/3 × 104 cells, p = 0.002), pcDNA3.1/hNIS (62 ± 8/3 × 104 cells, p = 0.001) or MIDGE/hNIS (57 ± 3/3 × 104 cells, p = 0.002). The mice vaccinated with pcDNA3.1/hNIS or MIDGE/hNIS showed only a slight increase in the number of hNIS-associated CD4+IFN-γ+ T cells compared with the mice vaccinated with pcDNA3.1. These results suggest that MIDGE/hNIS-NLS enhances the hNIS-associated CD4+ T cell immune responses in vaccinated mice.
Improved immunotherapy achieved through vaccination with MIDGE/hNIS
An in vivo tumor protection experiment was carried out to determine the ability of MIDGE/hNIS to protect vaccinated mice against CT26/NIS tumors. The mice were vaccinated with pcDNA3.1, pcDNA3.1/hNIS, MIDGE/hNIS or MIDGE/hNIS-NLS, and then challenged s.c. with 5 × 104 CT26/NIS cells in the right leg (Fig. 5a, blue arrow) on Day 7 after the final vaccination.
As shown in Figures 5b and 5c, 100% of the mice that received the MIDGE/hNIS vaccine remained tumor-free for 40 days after the CT26/NIS tumor challenge. This indicates that the MIDGE/hNIS vaccination provides protective antitumor immunity.
MIDGE/hNIS-NLS and pcDNA3.1/hNIS induced protection but was lower than that of MIDGE-hNIS (50% and 25% tumor-free, respectively). The MIDGE/hNIS vaccines give significantly better results than the pcDNA3.1 vaccines (p < 0.05 on Day 25, p < 0.01 on Days 30–40). MIDGE/hNIS-NLS (p < 0.05 on day 25, p < 0.01 on Days 30–40) and pcDNA3.1/hNIS (p < 0.05 on Days 25–40) still had inhibitory effects on tumor growth compared with the mice injected with pcDNA3.1, which produced an inferior result to that of MIDGE/hNIS (Fig. 5b). There was a significantly longer survival time of the mice treated with MIDGE/hNIS (p < 0.01) or MIDGE/hNIS-NLS (p < 0.05) compared with the mice vaccinated with pcDNA3.1, as shown in the Kaplan-Meier plots in Figure 5d. In contrast, vaccination with pcDNA3.1/hNIS did little produce survival benefit compared with pcDNA3.1 (MST = 52.3 ± 3.8 vs. 45.3 ± 2.5 days, respectively).
Vaccination with MIDGE/hNIS (p < 0.01 vs. pcDNA3.1; p < 0.01 vs. pcDNA3.1/hNIS) and MIDGE/hNIS-NLS (p < 0.01 vs. pcDNA3.1; p < 0.01 vs. pcDNA3.1/hNIS) increased the MST further (MST = 74.7 ± 2.5 and 67 ± 4.4 days, respectively) (Table I).
|pcDNA3.1||45.3 ± 2.5|
|pcDNA3.1/hNIS||52.3 ± 3.8|
|MIDGE/hNIS||74.7 ± 2.5*|
|MIDGE/hNIS-NLS||67 ± 4.4*|
Tc-99m scintigraphy using a gamma camera
The induction of the immune response by pcDNA3.1/hNIS or MIDGE vectors was monitored by imaging 30, 35 and 40 days after the tumor challenge using a gamma camera. As shown in Figure 6a, all mice vaccinated with MIDGE/hNIS remained tumor free for 40 days after the CT26/NIS tumor challenge. Naÿve mice injected with PBS were used as the control. The gamma count of the naÿve mouse was similar to that of MIDGE/hNIS (data not shown). In contrast, all the mice that received pcDNA3.1 developed tumors within 14 days of the tumor challenge. The mice receiving pcDNA3.1/hNIS or MIDGE/hNIS-NLS developed tumors within 25 days of the tumor challenge. The CT26/NIS tumors of all groups except for the MIDGE/hNIS group were clearly visible, with intensities comparable to those of the thyroid and bladder, which indicates that the tumor mass was viable and not necrotic (Fig. 6a). As shown in Figure 6b, the mice vaccinated with MIDGE/hNIS showed a significant decrease in the number of total counts/total pixel compared with the mice vaccinated with pcDNA3.1 (p < 0.05 on Days 30–35 and p < 0.01 on Day 40), pcDNA3.1/hNIS (p < 0.01 on Day 30 and p < 0.05 on Days 35–40) or MIDGE/hNIS-NLS (p < 0.05 on Days 35–40). The injection of the MIDGE/hNIS-NLS (vs. pcDNA3.1: p < 0.05 on Days 30–40; vs. pcDNA3.1/hNIS: p < 0.05 on Day 30) vaccine still produced a decrease in the total counts/total pixel.
These results confirmed that among the vaccine groups examined, the MIDGE/hNIS vaccine induced the most potent antitumor response.
DNA vaccines can induce both humoral and cellular immune responses, and are considered to be an effective basis for the development of vaccines against a wide variety of bacterial infections, viral diseases and cancers. However, various in vivo studies have been reported that DNA vaccines tend to produce relatively poor immunogenic responses, which means that they are often required in large amounts to be effective.23, 24, 25
Many strategies have been used in an attempt to enhance the immunogenicity of DNA vaccines.26, 27, 28, 29, 30, 31 It is evident that the potencies of these vaccines depends on amount of the antigen expressed in transfected cells.6 Therefore, a successful DNA vaccine strategy requires both the efficient delivery of DNA into somatic cells and the appropriate expression of the gene products in the targeted cells.12
The human sodium iodide symporter (hNIS) is a transmembrane glycoprotein that is responsible for iodide uptake by cells. Scintigraphic imaging using hNIS provide a useful means of noninvasively and repetitively monitoring the in situ status of tumor growth at earlier stages.20, 21, 22 To take this advantage, we used hNIS as an antigenic protein and we established the stable tumor cell line expressing hNIS (CT26/hNIS). Iodide is a major component for thyroid hormone synthesis and is transported by NIS protein. Therefore, NIS is an abundant protein in thyroid organ and plays an important role for thyroid function.18, 32 To minimize possible NIS DNA vaccination effect on thyroid function in mice, we employed human NIS rather than mouse NIS. Thyroid function appeared to be comparable in pcDNA3.1- and pcDNA3.1-hNIS-injected mice as monitored using scintigraphic imaging (data not shown).
The presence of hNIS on the basolateral membrane of thyroid follicular cells has been exploited for many years in the treatment of thyroid disease with radiolabeled iodide. Moreover, this noninvasive therapy has been proven to be a safe and effective means of monitoring and treating thyroid cancer.33
Usually NIS expression is decreased in tissue of thyroid cancer.17, 34, 35 However, the presence of NIS is unique and specific in thyroid cancer compared to other tissues.36, 37 In addition, several investigators have reported the presence of NIS in tissue of breast cancer.38 Therefore, hNIS can be used as a tumor antigen for immunotherapy against the hNIS over-expressing cancers. To date, there is no report showing that hNIS was used as a tumor associated antigen target for cancer DNA vaccines.
In this study, vaccination with the hNIS encoding plasmid (pcDNA3.1/hNIS) generated a higher number of hNIS-associated IFN-γ secreting CD8+ T cells than vaccination with pcDNA3.1 (Figs. 4a and 4b). Moreover, the induction of anti-hNIS immunity was strong enough to reduce the level of tumor growth (Figs. 5 and 6).
Some studies have reported the monitoring of tumor growth and the therapeutic response using gamma cameras.39, 40 Using sodium iodide symporter, targeted radionuclide gene therapy has been performed to promote treatment of various types of cancer. Some investigators found that effective gene expression is induced in living subjects by infection of NIS expressing nonviral or viral vector into cancer cell and it provides a good therapeutic index in preclinical model.41, 42 Other reports show that NIS gene transfer using tissue-specific promoter allows a mean of targeting NIS gene expression to cancer cells selectively, thereby maximizing tissue-specific cytotoxicity and minimizing toxic side-effects in normal cells.43, 44, 45 By NIS gene transfer into cancer cell, therapeutic radionuclide (I-131, Re-188) is concentrated in NIS expressing cancer. The emission characteristics and physical properties of therapeutic radionuclide could directly kill cancer cells by apoptosis. Kang et al. reported that NIS expressing human hepatocellular carcinoma could be selectively killed by the induced I-131 and Re-188 accumulation through NIS gene expression.46 In this study, the scintigraphic images showed a viable tumor mass without necrosis (Fig. 6) and may provide advantages over caliper-based tumor measurements in vivo.
A minimalistic, immunogenically defined gene expression (MIDGE) vector was used as an alternative to plasmids for vaccination. Previous studies have shown that the levels of IFN-γ producing cells are similar in mice immunized with a plasmid or MIDGE but were consistently higher in the mice immunized with MIDGE-NLS when using MIDGE encoding HBsAg.16 However, in this study, the results show that the MIDGE/hNIS vector increased the number of CD8+ T cells significantly in the draining lymph nodes of the vaccinated mice compared with the MIDGE/hNIS-NLS vector or pcDNA3.1/hNIS (Figs. 4a and 4b). Furthermore, vaccination with MIDGE/hNIS had a more inhibitory effect on tumor growth than vaccination with the other DNA vaccines (MIDGE/hNIS > MIDGE/hNIS-NLS > pcDNA3.1/hNIS > pcDNA3.1). All the mice vaccinated with MIDGE/hNIS remained tumor free for 40 days after the CT26/NIS challenge (Figs. 5b and 5c). Even at similar levels of gene expression, the MIDGE vectors may enhance the antigen presentation efficiency of hNIS in hNIS uptake cells through the MHC class I pathways to induce greater NIS associated CD8+ T cell responses in vivo. The hNIS proteins secreted from transfected myocytes could be taken up and processed by other APCs through the MHC class I-restricted pathway, which would increase T cell activation.47, 48
Another interesting result was that MIDGE/hNIS-NLS vaccination induced more CD4+ T cell immune responses than MIDGE/hNIS vaccination but without comparable inhibition of tumor growth. Further study will be needed to determine if these elevated CD4+ cells act as regulatory T cells or the Th2 type of helper T cells. Although there was no significant up-regulation of hNIS-associated-CD4+ T cells in the MIDGE/hNIS vaccinated mice, it is possible that some of the hNIS-associated-CD4+ T cells help to generate and expand the CD8+ T cells in these mice.49
In conclusion, a novel MIDGE vector encoding hNIS was found to generate a much stronger hNIS-associated CD8+ T cell response and enhanced antitumor effect against NIS-expressing CT26 tumors in vaccinated mice than the MIDGE/hNIS-NLS and plasmid (pcDNA3.1/hNIS) vectors. A hNIS vaccination strategy would be helpful for 2 reasons; a tool for targeting further immunotherapy trials against hNIS expressing cancers, and as a tool for evaluating the in vivo vaccine efficacy through scintigraphic imaging. hNIS vaccination might also provide a synergistic therapeutic model, e.g. 1 involving an radioimmunotherapeutic approach based on a combinatorial 131I and hNIS DNA vaccine therapy for NIS expressing cancer.
We thank Dr. Detlef Oswald for helping with cloning procedures and Dr. Y.H. Kim for the critical review the manuscript. Mr. Y. Choi, Dr. J.H. Kang and Mr. Y.H. Jeon were supported by the BK21 Project for Medicine, Dentistry and Pharmacy in Korea (2004).