Candidate Genes Associated with Tumor Regression Mediated by Intratumoral Il-12 Electroporation Gene Therapy

ARTICLE

doi:10.1016/j.ymthe.2003.11.022

Candidate Genes Associated with Tumor Regression Mediated by Intratumoral Il-12 Electroporation Gene Therapy Shulin Li,1,* Xueqing Xia,2 Francesca M. Mellieon,1 Jianguo Liu,1 and Stacy Steele1 1

Department of Comparative Biomedical Sciences and 2 Department of Pathological Biomedical Sciences, Louisiana State University, Skip Bertman Drive, Baton Rouge, LA 70803, USA

*To whom correspondence and reprint request should be addressed. Fax: +1-225-578-9895. E-mail: [email protected].

Interleukin-12 (IL-12) is one of the most effective cytokines for treating malignancy. Intratumoral delivery of the murine Il-12 gene, using electroporation, has been found effective in inducing regression of established tumors in mice, and more effective than intramuscular injection of this gene by electroporation, but what is not known is the molecular mechanism by which IL-12 exerts an antitumor effect. To define these candidate genes, the gene expression profiles of tumors treated with and without intratumoral Il-12 electroporation gene therapy were analyzed by cDNA array. Mig (Cxcl9), Stat1, and IRF7 are the three genes that are the most altered at the level of expression after administration of Il-12 via intratumoral electroporation, when subjected to further characterization by Northern blot, Western blot, and immunostaining. The results from Northern blot and immunostaining analyses indicate that intratumoral delivery of the murine Il-12 gene via electroporation induces accumulation of IRF7 in the nuclei of tumor cells and upregulates Mig and Stat1 expression by 15- and 5-fold, respectively, compared to intratumoral electroporation of control plasmid DNA. Intramuscular injection of the same Il-12 gene using electroporation upregulates Mig and Stat1 by only 6- and 2.9-fold, respectively, but does not induce any IRF7 accumulation in the nuclei. Further functional analyses of Mig indicate that expression in tumors can induce CD4+ but not CD8+ T cell infiltration. Further functional analysis of Stat1 indicates that a lack of Stat1 expression inhibits the Il-12-mediated induction of IP10, a known antiangiogenic gene. These data suggest that these three genes may positively correlate with the antitumor efficacy of intratumoral Il-12 electroporation gene therapy. Key Words: Il-12, gene therapy, electroporation, Stat1, IRF7, cDNA array

INTRODUCTION Interleukin-12 (IL-12) is a 70-kDa heterodimer protein composed of 40- and 35-kDa subunits [1]. It is one of the most effective cytokines for treating malignancy among many other cytokines, including IL-2, IL-4, IL7, tumor necrosis factor (TNF)-a, and granulocyte-macrophage colony-stimulating factor [2 – 8]. A significant antitumor effect has been found for IL-12 in multiple tumor models, using either recombinant protein or various gene therapy approaches, including cell-based IL-12 gene therapy, viral vector-based IL-12 gene therapy, and nonviral plasmid DNA-based IL-12 gene therapy [2,4,9]. Antitumor efficacy induced by IL-12 is due to

MOLECULAR THERAPY Vol. 9, No. 3, March 2004 Copyright B The American Society of Gene Therapy 1525-0016/$30.00

pleiotropic effects including antiangiogenesis, stimulation of interferon-g (IFN-g) production by natural killer (NK) cells, enhancement of the cytolytic activity of NK cells and cytotoxic T lymphocytes, and stimulation of the differentiation of Th1 cells [2 – 4,10,11]. Some reports suggest that antiangiogenesis may play the most important role in the antitumor effect [12], while others suggest that T cells are the most significant contributor to the IL-12-mediated antitumor effect [13 – 15]. It is not clear which downstream genes regulated by IL-12 lead to these therapeutic effects. In lymphoid cells, IL-12 induces multiple gene expression [16], multiple cytokine secretion [17], and

347

ARTICLE

doi:10.1016/j.ymthe.2003.11.022

STAT1 activation. Activation of STAT4 drives the differentiation of Th0 to Th1 cells, the production of IFNg from NK cells [18], and the activation of STAT1 [19]. There are no reports, however, on how intratumoral IL12 treatment affects gene expression in the tumor during tumor regression, which is important for understanding the underlying molecular mechanism of IL12’s in vivo antitumor effect. Intratumoral IL-12 treatment has shown greater efficacy and less toxicity than intramuscular IL-12 delivery [20]. Using the SCCVII tumor model, we found that intratumoral Il-12 electroporation gene therapy (EGT) not only eradicates 40% of established tumors, but also leads to the generation of antitumor immuno-memory cells for at least a year [21]. However, systemically high levels of Il-12 expression, mediated by intramuscular injection of plasmid DNA encoding Il-12 via electroporation, does not lead to significant tumor regression, only to inhibition of established tumor growth [22]. Similar results were found in a melanoma model with the same delivery approach [23]. It is not known, however, which downstream genes mediate the significant antitumor effect by intratumoral Il-12 EGT. To reveal the molecular candidate genes that induce significant tumor regression by intratumoral Il-12 EGT, we used a cDNA array to determine gene expression profiles in tumors treated with and without intratumoral Il-12 EGT. We also compared the differences in the levels of gene expression in tumors, using intramuscular and intratumoral Il-12 EGT. The gene expression analysis between different routes of administration in combination with gene interference confirmed the array analysis, that Stat1 and Mig (Cxcl9) are two important candidate

genes, which are associated with Il-12-mediated tumor eradication.

RESULTS Determination of the Candidate Genes Induced by Intratumoral Il-12 EGT Intratumoral Il-12 EGT refers to delivery of plasmid DNA encoding murine Il-12 via electroporation, directly into the tumor [21]. Intramuscular Il-12 EGT refers to delivery of Il-12 plasmid DNA with electroporation into muscle tissue that is distant from the tumor site [22]. Intratumoral Il-12 EGT resulted in 40% regression of established tumors and generation of long-term antitumor immune memory in an SCCVII tumor model [21]. To determine systematically the key candidate genes responsible for such a potent antitumor efficacy, we analyzed the expression of 1176 known murine genes in tumors treated with Il-12 or control plasmid DNA with the Atlas Mouse 1.2 Array (BD Biosciences, CA, USA). A ratio of z3 and an absolute difference of z100 in the levels of gene expression between treated and control tumors were chosen as criteria for selecting key candidate genes mediated by Il12 for further analysis. The rationale for choosing such stringent criteria is to increase the probability of discovering the genes most altered by Il-12. Using these criteria, we found 14 genes that were either upregulated or downregulated in tumors by intratumoral Il-12 EGT (Table 1). Among these 14 genes, 5 (36%) are transcriptional factors (Stat1, IRF7, IRF1, and IRF2) or regulator (Ski1) and 6 (43%) are either regulators or end products of the Jak-Stat (Janus tyrosine kinase-signal transducers and activator of transcription) signal transduction pathway [16,24,25], which include Stat1, IRF1, IRF2, IRF7, Pim1, and Mig. Four of

TABLE 1: Change in gene expression in the tumors treated with or without intratumoral IL-12 gene therapy Gene

Gamma interferon-induced monokine precursor (MIG) Signal transducer and activator of transcription 1 (STAT1) WSB2 protein Interferon Regulatory Factor 7 (IRF 7) Semaphorin B Interferon inducible protein 1(IFi1) Interferon regulatory factor 1 (IRF1) Glutathione S-transferace 5 (GST5-5) Pim1 proto-oncogene Interferon regulatory factor 2 (IRF 2) Glutathione reductase Ski proto-oncogene Granzyme A Non-histone chromosomal protein in HGM-14

Accession #

M34815 U06924 AF033188 U73037 X85991 U19119 M21065 J04696 M13945 J03168 X76341 U14173 M13226 U14173

Signal intensity pCtrl

pIL12

9 122 21 27 45 135 230 54 74 49 130 59 20 636

862 1484 220 264 241 647 863 200 269 170 446 201 132 158

Intensity difference 853 885 199 237 196 512 633 146 195 121 316 142 112 478

Ratio pI12/ pCtrl >20 12.2 10.5 9.7 5.4 4.8 3.8 3.7 3.6 3.5 3.4 3.4 up .24

Ratio, level of gene expression in tumors with IL-12 gene therapy vesus control plasmid DNA; pCtrl, control plasmid DNA; pIL12, plasmid DNA containing IL-12 transgene. All the data are derived from cDNA array analysis.

348

MOLECULAR THERAPY Vol. 9, No. 3, March 2004 Copyright B The American Society of Gene Therapy

ARTICLE

doi:10.1016/j.ymthe.2003.11.022

these 14 genes were induced 10-fold or more, compared to controls, based on array analysis (Table 1). Three of these 4 genes are known molecules in the interferon-signal transduction pathway and possibly play significant roles in the regulation of antitumor immune-regulatory gene expression [9,24,26], inhibition of angiogenesis, and enhancement of lymphoid cell infiltration [27 – 29]. The fourth gene, Wsb2, is less known. These results indicate that Mig, Stat1, Wsb2, and IRF7 are important candidate genes related to intratumoral Il-12 EGT-mediated tumor eradication [21]. The Candidate Genes Selected by cDNA Array Were Confirmed Independently by Northern Blot Analysis To confirm that the four candidate genes selected by cDNA array are truly altered by intratumoral Il-12 EGT, we analyzed the levels of these genes’ expression from independent experiments further by Northern blot. Analysis of Wsb2 expression was omitted because the DNA template for this gene could not be generated. Northern blot data confirm the cDNA array result, i.e., intratumoral Il-12 gene therapy induces the expression of Mig, Stat1, and IRF7 (Figs. 1A and 1B), which is in the same order as is found in the cDNA array analysis (Tables 1 and 2). Although the absolute levels of alteration between cDNA array and Northern blot results are different (Tables 1 and 2), the magnitude of Mig mRNA alteration in both assays is the most significant and the level of IRF7 mRNA is the least significant (Table 2).

FIG. 1. Intratumoral Il-12 EGT induces high levels of Mig, Stat1, and IRF7 mRNA in tumors. For intratumoral EGT, Il-12 or control plasmid DNA (10 Ag) was injected into the tumor followed by two electric pulses. The electric parameters were 450 V/cm and two pulses for a duration of 20 ms. Ten micrograms of RNA for each tumor sample was used to determine the level of gene expression by Northern blot. ‘‘pIL12’’ and ‘‘pCtrl’’ represent tumors injected with Il-12 plasmid DNA and empty plasmid DNA, respectively. Each lane represents a tumor sample isolated from a separate animal. Three mice were treated with control plasmid DNA and another three mice with plasmid DNA encoding Il-12. (A) Induction of Mig by intratumoral Il-12 EGT. (B) Induction of Stat1 and IRF7 by intratumoral Il-12 EGT.

MOLECULAR THERAPY Vol. 9, No. 3, March 2004 Copyright B The American Society of Gene Therapy

TABLE 2: Comparison of the magnitute change of gene expression in tumors after intratumoral or intramuscular (systemic) injection of IL-12 DNA plasmid via electroporation Genes

Intratumoral IL-12 Intramuscular IL-12 pCtrl: pIL12

P value

pCtrl: pIL12

P value

Gamma interferon-induced 1:15.1 monokine precursor (MIG) Signal transducer and 1:5.0 activator of transcription 1 (STAT1) Interferon Regulatory 1:4.0 Factor 7 (IRF 7)

0.0016

1:6.0

0.039

0.0009

1:2.9

0.061

0.0027

1:2.8

0.006

Data are derived from Northen blot analysis. pCtrl, control DNA plasmid; pIL12, plasmid DNA containing transgene IL-12.

Different Levels of Induction of Mig and Stat1 mRNA Correlate to Differences in Antitumor Efficacy between Intramuscular and Intratumoral Il-12 EGT To confirm further that increased expression of these genes in tumors may correlate to significant antitumor effect, we subject the expression levels of these genes in the tumors from tumor-bearing mice treated with intramuscular Il-12 EGT to Northern blot analysis. The hypothesis for this gene expression study is that the expression of these genes in the tumors treated with intramuscular Il-12 EGT will be lower than in those treated with intratumoral Il-12 EGT because intramuscular Il-12 EGT leads to a weaker antitumor effect than intratumoral Il-12 EGT. Indeed, Northern blot results do

FIG. 2. Intramuscular Il-12 EGT also increases the levels of Mig, Stat1, and IRF7 expression in tumors. For intramuscular EGT, plasmid DNA (10 Ag) with and without the Il-12-expressing cassette was injected into muscle followed by electric pulses, as described under Materials and Methods. The electric parameters used for intramuscular delivery were 350 V/cm and two pulses for a duration of 20 ms. Northern blot analysis and data labeling were the same as described for Fig. 1. A total of eight animals (n = 4) was used, four for control and four for Il-12 treatment.

349

ARTICLE

doi:10.1016/j.ymthe.2003.11.022

FIG. 3. Intratumoral Il-12 EGT induces a higher level of Stat1 protein in tumors than intramuscular treatment. Stat1 protein levels in tumors were compared in intratumoral and intramuscular Il-12 EGT by Western blot analysis, using a polyclonal anti-mouse Stat1 antibody. (A) Stat1 proteins are primarily induced in the tumors treated with intratumoral Il-12 EGT. (B) Stat1 is detected in tumors receiving intramuscular Il-12 EGT only when the membrane was overexposed on a film. The data labeling is the same as in Fig. 1. A nonspecific binding protein was also detected in all samples from intramuscular and intratumoral Il-12 EGT. The detection of this nonspecific binding protein indicates that the lack of Stat1 in some of the control mice (pCtrl) is not due to missed or reduced sample loading.

demonstrate that intramuscular Il-12 EGT induces the expression of Stat1, Mig, and IRF7 (Figs. 2A and 2B), but at a lower magnitude than intratumoral Il-12 EGT (Table 2). Compared with expression levels induced by control plasmid DNA, levels of Mig and Stat1 induction by intramuscular Il-12 EGT are, on average, 2.9- and 6.0fold higher, respectively. This is much lower than those mediated by intratumoral Il-12 EGT, which are 5- and 15-fold higher, respectively. Because intramuscular Il-12 EGT is less effective than intratumoral Il-12 EGT [21 – 23] in the inhibition of tumor growth and induction of Mig and Stat1, a plausible explanation is that the higher magnitude of induction of Mig and Stat1 mRNA may correlate with the potent therapeutic effect mediated by intratumoral Il-12 EGT. Significant Induction of Stat1 Protein and IRF7 in the Tumor Is Associated with the Potent Antitumor Efficacy Mediated by Intratumoral Il-12 EGT To demonstrate if the protein levels of these candidate genes are induced similar to the induction of mRNA by intratumoral and intramuscular Il-12 EGT, we performed immunostaining of tumor sections and Western blot analysis of tumor protein extract, depending on the feasibility of antibody detection. Because anti-

mouse Mig antibody is not available, Mig protein determination is not included in this study. Intratumoral Il-12 EGT induces a higher level of Stat1 protein in the tumor than intramuscular Il-12 EGT, as demonstrated in Figs. 3A and 3B. The Stat1 expression shown on Fig. 3A was captured by an image camera directly, and Stat1 shown on Fig. 3B was overexposed on the film, demonstrating that the induced Stat1 expression is not due to overloading. Importantly, expression and accumulation of IRF7 protein are primarily detected in tumors treated with intratumoral Il-12 EGT, but are not found in any tumors treated with intramuscular Il-12 EGT (Fig. 4), suggesting that accumulation of IRF7 in the nuclei of tumor cells is associated with the potent antitumor efficacy mediated by intratumoral Il-12 EGT [21]. Intratumoral Il-12 EGT Increases the Level of Il-12 Expression in the Tumor To determine whether the difference in the level of Il-12 expression after intratumoral or intramuscular gene delivery via electroporation accounts for the different antitumor efficacy [21,22] and the different levels of Mig, Stat1, and IRF7 (Table 2), we determined the level of Il-12 expression using enzyme-linked immunosorbent assay

FIG. 4. Analysis of IRF7 localization in the tumor cells after intratumoral or intramuscular Il-12 EGT. Tumor samples from the tumor-bearing mice treated with intratumoral and intramuscular Il-12 EGT were obtained on day 10. Frozen tumor sections were stained with anti-mouse IRF7 polyclonal antibody and the representative image was taken under a microscope at 40 magnification for (A) to (D). (E) A 10-fold reduction of the image. Arrows point to the positively stained anti-IRF7 antibody in the nuclei.

350

MOLECULAR THERAPY Vol. 9, No. 3, March 2004 Copyright B The American Society of Gene Therapy

doi:10.1016/j.ymthe.2003.11.022

ARTICLE

in the tumor but intramuscular Il-12 EGT increases the level of distribution in the serum (Fig. 5). This opposite distribution pattern is associated with a difference in the expression levels of the candidate genes and antitumor efficacy between different routes of administration.

FIG. 5. Analysis of the distribution of Il-12 protein after different routes of administration of Il-12 plasmid DNA using electroporation. The purpose of this experiment is to compare the levels of Il-12 expression and the difference in distribution of Il-12 between tumor and serum. Tumor and serum samples (n = 5) were obtained on day 3 after administration of Il-12 DNA plasmid. The procedures for intratumoral and intramuscular EGT are the same as described in the legends to Figs. 1 and 2, respectively. The level of gene expression was determined using ELISA as detailed under Materials and Methods.

(ELISA) after delivery of the Il-12 gene. Interestingly, the maximum levels of Il-12 expression by different administration routes are similar, but the distributions of this protein between tumor and serum are opposite. Intratumoral Il-12 EGT increases the level of Il-12 distribution

Interference of Stat1 and Mig Expression Inhibits the Expression of Antiangiogenic Gene IP10 and Infiltration of CD4+ T Cells, Respectively To determine the antitumor function of Mig, one of the primary candidate genes associated with intratumoral Il-12 EGT-mediated antitumor effects, we cloned a mouse Mig gene into a gene expression vector and injected it into tumors using electroporation, as described under Materials and Methods. This cloned Mig cDNA has been confirmed by sequence and restriction analysis, as well as expression analysis, using Northern blot (data not shown). We determined infiltration of CD4+ and CD8+ T cells into the tumors. We detected significantly increased infiltration of CD4+ T cells but not CD8+ T cells in the tumors treated with Mig EGT, indicating that Mig may be a chemoattractant for CD4+ T cells (Figs. 6B and 6C). Stat1 is a primary transcription factor that mediates induction of gene expression in the interferon signal transduction pathway [26]. IFN-g-inducible protein 10 (IP-10) is a known chemoattractant and induces antitumor effects by inducing antiangiogenic effects and immune cell infiltration [27,30]. To determine whether expression of IP10 is mediated by Stat1, Stat1 knockout FIG. 6. The effects of interference of Il-12-mediated candidate genes on gene expression and immune cell infiltration. (A) Interference of Stat1 inhibits the expression of antiangiogenic gene IP10. Northern blot analysis of the expression of IP10 was performed in Stat1 / and wild-type mice treated with and without Il-12 gene in the muscles. Ten micrograms of plasmid DNA of pCtrl or pIL12 was injected intramuscularly via electroporation, as described in the Fig. 2 legend. (B and C) Injection of Mig (pMig) into the tumor via electroporation increases infiltration of CD4+ T cells into the tumor. Thirty micrograms of plasmid DNA with and without Mig was injected into tumors (n = 5) once a week for a total of two administrations. Three days after the second administration, tumor-bearing mice were sacrificed and infiltration of CD4+ and CD8+ T cells was analyzed from tumor sections using immunostaining, as described under Materials and Methods.

MOLECULAR THERAPY Vol. 9, No. 3, March 2004 Copyright B The American Society of Gene Therapy

351

ARTICLE mice (Stat1 / ) and wild-type mice were injected with Il-12 via intramuscular electroporation. IP10 expression was not detected in the Stat1 / mice but was greatly increased in the wild-type mice (Fig. 6A), confirming that Stat1 plays an important role in the intratumoral Il-12 EGT-mediated antitumor effect.

DISCUSSION Despite the rapid progress of micro- and macroarray technology, no effort has been made to use this technique to reveal the IL-12-induced alteration of the gene expression profile of tumors. Using this powerful approach, we have found that 13 genes are induced and 1 is inhibited in the tumor by intratumoral Il-12 EGT (Table 1), in which Il-12 plasmid DNA is injected into the tumor directly, followed by electric pulses. A 10-fold or higher induction, compared to injection of control plasmid DNA via the same electroporation delivery, is found for the genes Mig, Stat1, IRF7, and Wsb2 (Table 1). Detailed analysis and comparison of the levels and cellular localization of these candidate genes between intratumoral and intramuscular Il-12 EGT reveal one novel discovery: accumulation of IRF7 in the nuclei of tumor cells is associated with potent tumor eradication mediated by intratumoral Il-12 EGT [21]. As far as we know, this is the first demonstration of IRF7 induction in the tumor mediated by intratumoral Il-12 EGT. This finding may serve as an important end point to predict tumor eradication or significant regression because intratumoral Il-12 EGT-induced accumulation of IRF7 in the tumor tissues (Figs. 4 and 5) is associated with 40% tumor regression mediated by intratumoral Il-12 EGT [21], and the lack of IRF7 accumulation in the nuclei after intramuscular Il-12 EGT is associated with the failed regression of established tumors by intramuscular Il-12 EGT [22]. The association of IRF7 expression with antitumor activity was also reported by others [31,32]. It was reported that induction of IRF7 inhibits tumor cell growth and activation of IRF7 parallels the activation of p53 [31,32]. To confirm the antitumor significance of the three candidate genes mediated by intratumoral Il-12 EGT and selected by cDNA array analysis, the ideal approach would be to test Il-12-mediated therapeutic effect in both transgenic and knockout tumor-bearing mice that either overexpress or lack expression of these genes. Obtaining these tumor models and knockout mice is costly and time-consuming. In addition, it is possible that multiple genes interact together to affect the cellular and immune system, generating the Il-12-mediated antitumor effect. To elucidate the possible roles of these genes without those transgenic or knockout mice, an indirect approach is adopted here, based on the hypothesis that the candidate genes, induced by intratumoral but not by intramuscular Il-12 EGT, will be the key genes that mediate

352

doi:10.1016/j.ymthe.2003.11.022

tumor eradication. This hypothesis is based on our therapeutic efficacy results in which intramuscular Il-12 EGT failed to eradicate established tumors as effectively as intratumoral Il-12 EGT [21,22]. To test this hypothesis, the expression of Mig, Stat1, and IRF7 in the tumor was compared between tumors receiving either intratumoral or intramuscular Il-12 EGT. We found that a high level of Mig and Stat1 induction in tumors may positively correlate with intratumoral Il-12 EGT-mediated tumor eradication, as reported previously [21]. There is a 15-fold induction of Mig, compared with injection of control plasmid DNA via electroporation, when using intratumoral Il-12 EGT. However, there is only a 6-fold induction when using intramuscular Il-12 EGT. The difference in the level of Stat1 protein induction between intratumoral and intramuscular Il-12 EGT clearly demonstrates that the high level of Stat1 induction in the tumor may also be important for Il-12-mediated antitumor effect (Fig. 3). To strengthen this conclusion, Il-12mediated induction of a known antiangiogenic gene, IP10 [27], was tested in the Stat1 / and wild-type mice. IP10 is induced in the wild-type mice but abolished in the Stat1 / mice (Fig. 6A). This result is in agreement with the significant antiangiogenic effect mediated by Il-12 EGT [21]. To strengthen this conclusion further, the antitumor function of Mig gene therapy was explored. It was found that injection of Mig by electroporation induces infiltration of CD4+ T cells (Figs. 6B and 6C). This is in agreement the observations of others in a murine mammary tumor model, in which transduced-Mig tumor cells induced infiltration of CD4+ T cells but not CD8+ T cells [33]. These observations conflict with earlier reports, which suggested that Mig plays a role in CD8+ T cell infiltration [34]. The discovery in this study that Stat1 may play a major role in the Il-12-mediated inhibition of tumor growth is also supported by in vitro studies, in which Stat1 enhances expression of immunomodulatory genes [35] but inhibits the expression of the cell cycle genes [9], and in vivo studies, in which the knockout of Stat1 in the host mice will compromise Il-12-mediated antitumor effects [36]. Taken together, these observations indicate that Stat1 is required in both the host and the tumor cells for Il-12-mediated tumor regression. Induction of Stat1 in the tumor and accumulation of IRF7 in the nuclei of tumor cells mediated by tumor-local Il-12 EGT may be beneficial to tumor eradication, but induction of other genes in the tumor may be harmful. Induction of Pim1 by Il-12 EGT (Table 1) may suppress Jak-Stat signaling and T cell development [25,37]. The detailed analysis of Stat1, Mig, and IRF7 in this study is not meant to downplay the potential significance of the other genes listed in Table 1. For example, c-Ski, which was initially known as a proto-oncogene, may play a significant role in suppression of tumor growth in at least some cell types because Ski-deficient heterozygous

MOLECULAR THERAPY Vol. 9, No. 3, March 2004 Copyright B The American Society of Gene Therapy

doi:10.1016/j.ymthe.2003.11.022

mice are susceptible to tumorigenesis, and introduction of Ki-ras into Ski-deficient mouse embryonic fibroblasts results in neoplastic transformation [38]. Altered expression of the chromosomal protein Hmg-14, the only gene downregulated by Il-12 in our cDNA array analysis (Table 1), is also downregulated by a well-known antimicrotubule agent, TZT-1027, in combination with transfection of TNF-a [39]. TNF-a, also a secondary cytokine of Il-12, may induce another independent pathway to inhibit tumor growth. Thus, this study opens up a new approach to examine Il-12-mediated molecular events. Despite the fact that molecular change has been highly emphasized in this report, other alterations such as induction of immunogenicity of tumor cells will also greatly contribute to the potent therapeutic effect mediated by intratumoral Il-12 EGT. However, such an alteration should be detected by cDNA array analysis, if this is the case. Studying the molecular alteration with and without Il12 treatment is a valuable approach in discovering novel Il-12-mediated genes that are responsible for tumor growth and eradication. It is interesting to demonstrate, at the molecular level, that injection of the same Il-12 gene using a different route leads to different levels of specific gene expression (Table 2) and even to different profiles of gene expression. The difference in the induction of Mig, Stat1, and IRF7 expression when intratumoral and intramuscular Il-12 EGT are compared is most likely caused by the tumor-local concentration of Il-12 rather than a systemic level of Il-12. This is because intramuscular injection of Il-12 using electroporation yields 170 pg of Il-12/ml of serum, but intratumoral injection yields 69 pg of Il-12/ml of serum. On the contrary, intramuscular injection yields 65 pg of Il-12/mg of total protein in the tumor extract, but intratumoral injection yields 143 pg of Il-12/mg of total protein (Fig. 5). It is unlikely that the difference in electroporation voltages used for intramuscular and intratumoral plasmid DNA injection caused the different antitumor effects, because others have reported a similar difference between the effects of intratumoral and intramuscular EGT using an entirely different set of electroporation fields [23]. Thus, the tumor-local concentration of Il-12 is the key to what causes the different levels of gene expression and antitumor efficacy.

MATERIALS AND METHODS Generation of SCCVII tumors, measurement of tumor growth, injection of plasmid DNA, and determination of cytokine expression. We purchased 6- to 8-week-old female C3H/HeJ mice, weighing 18 – 20 g, from Jackson Laboratories (Bar Harbor, ME, USA) and maintained them under National Institutes of Health guidelines that were approved by the Institutional Animal Care and Use Committee of the University of Arkansas for Medical Sciences, where the work began. SCCVII is a spontaneously arising murine squamous cell carcinoma. We obtained this cell line from Dr. Candace Johnson’s laboratory at the

MOLECULAR THERAPY Vol. 9, No. 3, March 2004 Copyright B The American Society of Gene Therapy

ARTICLE

University of Pittsburgh (PA, USA) and maintained the cells in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum (Life Technologies, Rockville, MD, USA). Tumors were generated by inoculating mice subcutaneously with 1  105 SCCVII cells in a 30-Al volume. Tumor growth was measured with a caliper and tumor volume was calculated with the formula V = k/8(a  b2), where V is the tumor volume, a is the maximum tumor diameter, and b is the diameter at 90j to a [36,40]. Intratumoral and intramuscular injection of Il-12 or control plasmid DNA followed by electroporation, as well as determination of tumor regression and survival, was performed according to protocols described previously [21,22]. In each experiment, five animals for each treatment or control group were used to study tumor regression or survival. Both of the optimal electroporation parameters for intramuscular injection and intratumoral injection, optimized previously, were used [41,42] for gene injection into muscles and tumors, respectively. The parameters used for intratumoral electroporation were 450 V/cm and 20-ms pulse duration for two pulses. The parameters for muscle administration were 350 V/cm and 20-ms pulse duration for two pulses. Plasmid DNA of the Il-12 construct was obtained from Valentis, Inc. (Burlingame, CA, USA); the backbone of this construct was described in a previous publication [43]. The control plasmid DNA consisted of a deletion of the Il-12 gene from the Il-12 construct. Mig and IP10 genes were amplified from total RNA isolated from mouse spleen by RT-PCR as described in a previous publication [42]. Mig was cloned into a TA2.1 vector (Invitrogen, Inc., San Diego, CA, USA) and subcloned into an expression vector containing a h-actin promoter and h-globulin polyadenylation signal, depicted previously [44]. To determine the expression of Il-12 and IFN-g, we used the same protocols as were previously used [21,22] to prepare samples and perform ELISA in this study. For gene expression analysis, three to five mice from each treatment or control group were used in each experiment. Gene expression profile and expression analysis of Stat1, Mig, and IRF7. To determine systematically the candidate genes responsible for the antitumor effect mediated by intratumoral Il-12 gene therapy, the expression of 1176 known murine genes in tumors treated with Il-12 or control plasmid DNA (10 Ag) were analyzed with the Atlas Mouse 1.2 Array (BD Biosciences). To obtain tumor samples for gene expression profile analysis with the cDNA array, tumors in the tumor-bearing mice were injected with Il-12 or control plasmid DNA, and electroporation followed, as described previously [21]. All tumor samples subjected to either cDNA array or Northern blot analysis were obtained 2 days after administration of 10 Ag of Il-12 or control plasmid DNA by intratumoral and intramuscular delivery, respectively. Detailed cDNA array and Northern blot analysis methods were the same as were described previously [42,45]. Immunostaining analysis of gene expression. The procedure for frozenblock preparation, tissue sectioning, and immunostaining were the same as described previously [42]. In brief, a 10-Am section was cut and fixed with cold acetone ( 20jC). The sections were rinsed with phosphatebuffered saline and blocked with 2% dry milk. The primary antibodies applied to the sections were anti-IRF7, -CD4, and -CD8 antibodies (1:200; Santa Cruz Biotechnology, Santa Cruz, CA, USA). After the slides were rinsed three times, the biotinylated secondary antibody (1:400 dilution) was incubated for 30 min. Streptavidin/HRP was then applied and incubated for another 30 min. After the slides were rinsed, DAB was applied for 10 to 15 min. All the antibody and immunostaining reagents were purchased from Santa Cruz Biotechnology. Western blot analysis of gene expression. Protein extraction was conducted as described previously [42]. In brief, frozen tumor tissues were homogenized using a mini bead-beater (Biospec Products, Bartlesville, OK, USA) with silica beads for 1 min in 1 ml of enzyme extraction buffer (Promega, Madison, WI, USA). An equal amount of protein extract from each tumor (100 Ag/sample) was subjected to Western blot analysis as described previously [46]. ECL reagents (Amersham, UK) and Image

353

ARTICLE

Station 440 (Perkin-Elmer Life Sciences, Boston, MA, USA) were used to detect and capture the image. Statistical analysis. Expression levels of Stat1, Mig, and IRF7 were the primary outcomes measured. We used the two-sided Student’s t test to compare the means of individual treatments when the primary outcome was statistically significant. Survival analysis was performed with the m2 test. P values less than 0.05 were considered statistically significant.

doi:10.1016/j.ymthe.2003.11.022

20.

21.

22. 23.

ACKNOWLEDGMENTS The authors thank Valentis, Inc., and Gentronics, Inc., for allowing us to use their gene constructs and electroporator, respectively, for this study. This work is supported by grants from the NCI/NIH (RO1CA98928) and NIDCR/NIH (R21 DE14682).

24.

25. 26.

RECEIVED FOR PUBLICATION JUNE 9, 2003; ACCEPTED NOVEMBER 21, 2003.

REFERENCES 1. Kobayashi, M., et al. (1989). Identification and purification of natural killer cell stimulatory factor (NKSF), a cytokine with multiple biologic effects on human lymphocytes. J. Exp. Med. 170: 827 – 845. 2. Golab, J., and Zagozdzon, R. (1999). Antitumor effects of interleukin-12 in pre-clinical and early clinical studies. Int. J. Mol. Med. 3: 537 – 544. 3. Trinchieri, G. (1994). Interleukin-12: a cytokine produced by antigen-presenting cells with immunoregulatory functions in the generation of T-helper cells type 1 and cytotoxic lymphocytes. Blood 84: 4008 – 4027. 4. Burke, F. (1999). Cytokines (IFNs, TNF-alpha, IL-2 and IL-12) and animal models of cancer. Cytokines Cell Mol. Ther. 5: 51 – 61. 5. Rakhmilevich, A. L., Janssen, K., Turner, J., Culp, J., and Yang, N. S. (1997). Cytokine gene therapy of cancer using gene gun technology: superior antitumor activity of interleukin-12. Hum. Gene Ther. 8: 1303 – 1311. 6. Watanabe, Y., et al. (1989). Exogenous expression of mouse interferon gamma cDNA in mouse neuroblastoma C1300 cells results in reduced tumorigenicity by augmented anti-tumor immunity. Proc. Natl. Acad. Sci. USA 86: 9456 – 9460. 7. Colombo, M. P., Modesti, A., Parmiani, G., and Forni, G. (1992). Local cytokine availability elicits tumor rejection and systemic immunity through granulocyte – T-lymphocyte cross-talk. Cancer Res. 52: 4853 – 4857. 8. Forni, G., et al. (1993). Cytokine-induced tumor immunogenicity: from exogenous cytokines to gene therapy. J. Immunother. 14: 253 – 257. 9. Bramson, J. L., et al. (1996). Direct intratumoral injection of an adenovirus expressing interleukin-12 induces regression and long-lasting immunity that is associated with highly localized expression of interleukin-12. Hum. Gene Ther. 7: 1995 – 2002. 10. Soiffer, R. J., Robertson, M. J., Murray, C., Cochran, K., and Ritz, J. (1993). Interleukin12 augments cytolytic activity of peripheral blood lymphocytes from patients with hematologic and solid malignancies. Blood 82: 2790 – 2796. 11. Yoshida, A., Koide, Y., Uchijima, M., and Yoshida, T. O. (1994). IFN-gamma induces IL-12 mRNA expression by a murine macrophage cell line, J774. Biochem. Biophys. Res. Commun. 198: 857 – 861. 12. Andrews, K. J., et al. (2000). Adenovirus – interleukin-12-mediated tumor regression in a murine hepatocellular carcinoma model is not dependent on CD1-restricted natural killer T cells. Cancer Res. 60: 6457 – 6464. 13. Moller, P., et al. (2000). Increased non-major histocompatibility complex-restricted lytic activity in melanoma patients vaccinated with cytokine gene-transfected autologous tumor cells. Cancer Gene Ther. 7: 976 – 984. 14. Brunda, M. J., et al. (1993). Antitumor and antimetastatic activity of interleukin 12 against murine tumors. J. Exp. Med. 178: 1223 – 1230. 15. Nastala, C. L., et al. (1994). Recombinant IL-12 administration induces tumor regression in association with IFN-gamma production. J. Immunol. 153: 1697 – 1706. 16. Rogge, L., et al. (2000). Transcript imaging of the development of human T helper cells using oligonucleotide arrays. Nat. Genet. 25: 96 – 101. 17. Rodolfo, M., and Colombo, M. P. (1999). Interleukin-12 as an adjuvant for cancer immunotherapy. Methods 19: 114 – 120. 18. Murphy, K. M., et al. (1999). T helper differentiation proceeds through Stat1-dependent, Stat4-dependent and Stat4-independent phases. Curr. Top. Microbiol. Immunol. 238: 13 – 26. 19. Gollob, J. A., et al. (2000). Impairment of STAT activation by IL-12 in a patient

354

27. 28.

29. 30. 31.

32. 33.

34. 35. 36. 37.

38. 39.

40.

41.

42. 43.

44.

45.

46.

with atypical mycobacterial and staphylococcal infections. J. Immunol. 165: 4120 – 4126. Chen, L., et al. (1997). Eradication of murine bladder carcinoma by intratumor injection of a bicistronic adenoviral vector carrying cDNAs for the IL-12 heterodimer and its inhibition by the IL-12 p40 subunit homodimer. J. Immunol. 159: 351 – 359. Li, S., Zhang, X., and Xia, X. (2002). Regression of tumor growth and induction of longterm antitumor memory by interleukin 12 electro-gene therapy. J. Natl. Cancer. Inst. 94: 762 – 768. Hanna, E., et al. (2001). Intramuscular electroporation delivery of IL-12 gene for treatment of squamous cell carcinoma located at a distant site. Cancer Gene Ther. 8: 1 – 7. Lucas, M. L., Heller, L., Coppola, D., and Heller, R. (2002). IL-12 plasmid delivery by in vivo electroporation for the successful treatment of established subcutaneous B16, F10 melanoma. Mol. Ther. 5: 668 – 675. Chatterjee-Kishore, M., Wright, K. L., Ting, J. P., and Stark, G. R. (2000). How Stat1 mediates constitutive gene expression: a complex of unphosphorylated Stat1 and IRF1 supports transcription of the LMP2 gene. EMBO J. 19: 4111 – 4122. Chen, X. P., et al. (2002). Pim serine/threonine kinases regulate the stability of Socs-1 protein. Proc. Natl. Acad. Sci. USA 99: 2175 – 2180. Meraz, M. A., et al. (1996). Targeted disruption of the Stat1 gene in mice reveals unexpected physiologic specificity in the JAK-STAT signaling pathway. Cell 84: 431 – 442. Angiolillo, A. L., et al. (1995). Human interferon-inducible protein 10 is a potent inhibitor of angiogenesis in vivo. J. Exp. Med. 182: 155 – 162. Rizzuto, G., et al. (1999). Efficient and regulated erythropoietin production by naked DNA injection and muscle electroporation. Proc. Natl. Acad. Sci. USA 96: 6417 – 6422. Sgadari, C., et al. (1996). Interferon-inducible protein-10 identified as a mediator of tumor necrosis in vivo. Proc. Natl. Acad. Sci. USA 93: 13791 – 13796. Taub, D. D. (1996). Chemokine – leukocyte interactions. The voodoo that they do so well. Cytokine Growth Factor Rev. 7: 355 – 376. Kim, T. K., Kim, T., Kim, T. Y., Lee, W. G., and Yim, J. (2000). Chemotherapeutic DNA-damaging drugs activate interferon regulatory factor-7 by the mitogen-activated protein kinase kinase-4 – cJun NH2-terminal kinase pathway. Cancer Res. 60: 1153 – 1156. Ko, L. J., and Prives, C. (1996). p53: puzzle and paradigm. Genes Dev. 10: 1054 – 1072. Dorsey, R., et al. (2002). Immunotherapy with interleukin-10 depends on the CXC chemokines inducible protein-10 and monokine induced by IFN-gamma. Cancer Res. 62: 2606 – 2610. Tannenbaum, C. S., et al. (1998). The CXC chemokines IP-10 and Mig are necessary for IL-12-mediated regression of the mouse RENCA tumor. J. Immunol. 161: 927 – 932. Boehm, U., Klamp, T., Groot, M., and Howard, J. C. (1997). Cellular responses to interferon-gamma. Annu. Rev. Immunol. 15: 749 – 795. Fallarino, F., and Gajewski, T. F. (1999). Cutting edge: differentiation of antitumor CTL in vivo requires host expression of Stat1. J. Immunol. 163: 4109 – 4113. Xu, Y. X., Gao, X., Janakiraman, N., Chapman, R. A., and Gautam, S. C. (2001). IL-12 gene therapy of leukemia with hematopoietic progenitor cells without the toxicity of systemic IL-12 treatment. Clin. Immunol. 98: 180 – 189. Shinagawa, T., et al. (2001). Increased susceptibility to tumorigenesis of ski-deficient heterozygous mice. Oncogene 20: 8100 – 8108. Natsume, T., et al. (2002). Enhanced antitumor activities of TZT-1027 against TNFalpha or IL-6 secreting Lewis lung carcinoma in vivo. Cancer Chemother. Pharmacol. 49: 35 – 47. Puisieux, I., et al. (1998). Canarypox virus-mediated interleukin 12 gene transfer into murine mammary adenocarcinoma induces tumor suppression and long-term antitumoral immunity. Hum. Gene Ther. 9: 2481 – 2492. Li, S., Xia, X., Zhang, X., and Suen, J. (2002). Regression of tumors by IFN-alpha electroporation gene therapy and analysis of the responsible genes by cDNA array. Gene Ther. 9: 390 – 397. Li, S., et al. (2001). Intramuscular electroporation delivery of IFN-alpha gene therapy for inhibition of tumor growth located at a distant site. Gene Ther. 8: 400 – 407. Coleman, M., et al. (1998). Nonviral interferon alpha gene therapy inhibits growth of established tumors by eliciting a systemic immune response. Hum. Gene Ther. 9: 2223 – 2230. Li, S., Leonard, D., and Wilkinson, M. F. (1997). T cell receptor (TCR) mini-gene mRNA expression regulated by nonsense codons: a nuclear-associated translation-like mechanism. J. Exp. Med. 185: 985 – 992. Hanna, E., et al. (2001). A novel alternative approach for prediction of radiation response of squamous cell carcinoma of head and neck. Cancer Res. 61: 2376 – 2380. Li, S., and Budde, R. (1995). Recombinant CSK monoclonal antibody production and characterization. Hybridoma 14: 341 – 346.

MOLECULAR THERAPY Vol. 9, No. 3, March 2004 Copyright B The American Society of Gene Therapy