Introduction
Interleukin-13 Receptor α2 (IL-13Rα2) is a high affinity receptor for the Th2 derived cytokine IL-13 and a known cancer testis antigen [
1,
2]. IL-13Rα2 is over expressed in a variety of human cancers including malignant glioma, head and neck cancer, Kaposi's sarcoma, renal cell carcinoma, and ovarian carcinoma [
3‐
7]. We have demonstrated previously that IL-13Rα2 can be effectively targeted by a recombinant immunotoxin, consisting of IL-13 and truncated
pseudomonas exotoxin (IL-13-PE) [
8‐
11]. IL-13-PE is highly cytotoxic to tumor cells
in vitro and
in vivo that express high levels of IL-13Rα2 [
12]. Several phase I and II clinical trials, and one phase III clinical trial, evaluating the safety, tolerability, and efficacy of this agent have been completed in patients with recurrent glioblastoma multiforme [
13,
14]. Most recently, we have demonstrated expression of IL-13Rα2 in human pancreatic ductal adenocarcinoma [
15]. Seventy-one percent of pancreatic tumors overexpressed IL-13Rα2 chain. Pancreatic tumors were also successfully targeted by IL-13-PE in an animal model of human cancer [
15,
16]. Thus, IL-13Rα2 is currently being assessed as a cancer therapy in a variety of preclinical and clinical trials [
4,
17,
18]
The significance of IL-13Rα2 expression in cancer is not known and the mechanism of its upregulation is still not clear. Epigenetic mechanisms such as DNA methylation and histone modification are known to be involved in many disease pathogenesis including cancer [
19]. DNA methylation occurs on cytosines that are followed by guanines (CpG dinucleotides) and is usually associated with gene silencing [
20]. Histones are modified at several different amino acid residues and with many different modifications including methylation, acetylation, phosphorylation and ubiquitination. Some lysine residues can either be methylated or acetylated, and there are three different possibilities for each methylated site [
21]. Histone modification can be transiently altered by the cell environment [
22]. Mainly, gene expression is activated by histone acetylation and decreased by methylation. Histone acetylation induced by histone acetyltransferase (HAT) is associated with gene transcription, while histone hypoacetylation induced by histone deacetylase (HDAC) is associated with gene silencing [
23].
HDAC inhibition results in increased acetylation in histones and causes over expression of some genes. HDAC inhibitors are grouped into various classes based on their structures [
24]. Trichostatin A (TSA), suberoylanilide hydroxamic acid (SAHA), and sodium butyrate (NaB) are commonly studied HDAC inhibitors. These inhibitors induce cell growth arrest and apoptosis in a broad spectrum of transformed cells [
25]. Because of these characteristics, HDAC inhibitors are being tested in the clinic for cancer therapy. Two HDAC inhibitors, SAHA and Romidepsin, are licensed by FDA for the treatment of cutaneous T-cell lymphoma [
26].
In the present study, we have examined the epigenetic regulation of the IL-13Rα2 gene in pancreatic cancer cell lines and investigated whether the IL-13Rα2 gene can be modulated by epigenetic mechanisms. We have also examined the effect of HDAC inhibitors on IL-13Rα2 expression. We demonstrate for the first time that three different HDAC inhibitors dramatically upregulate IL-13Rα2 in pancreatic cancer cell lines expressing no or low levels of IL-13Rα2. These inhibitors also modestly upregulated IL-13Rα2 in cells expressing higher levels of IL-13Rα2. More importantly, HDAC inhibitors sensitized pancreatic tumor cells to IL-13-PE and mediated enhanced sensitivity even though these cells did not naturally express IL-13Rα2. A combination therapy of HDAC inhibitors and IL-13-PE demonstrated a pronounced anti-tumor effect in human tumor bearing immunodeficient mice indicating a synergistic impact on tumor response. Thus, a novel combination of HDAC inhibitors and IL-13-PE may have a prominent role in pancreatic cancer or other cancer therapies in the clinic.
Materials and methods
Cell culture and reagents
Pancreatic cancer cell lines and human umbilical vein endothelial cell line (HUVEC) were obtained from the American Type Culture Collection (Manassas, VA). Human normal gingival fibroblasts (HGF) was obtained from Sciencell (San Diego, CA) and human pancreatic ductal epithelial cells (HPE) from Cell Systems (Kirkland, WA). Renal cell carcinoma (PM-RCC) cell line was developed in our laboratory [
4]. Recombinant IL-13-PE was produced and purified in our laboratory [
9,
11,
27]. Trichostatin A (TSA), sodium butyrate (NaB) and SP600125 were purchased from Sigma-Aldrich (St. Louis, MO). SR11302 was purchased from Tocris Bioscience (Ellisville, MO). Suberoylanilide Hydroxamic Acid (SAHA) was purchased from Selleck (Houston, TX).
Reverse transcription-PCR
Quantitative reverse transcription-PCR (qRT-PCR) and RT-PCR were performed as described previously [
28,
29] using a SYBR 1 reagent kit (Bio-Rad, Hercules, CA). Mouse IL-13Rα2 and β-actin primers were purchased from QIAGEN (Valencia, CA). Gene expression was normalized to β-actin before the fold change in gene expression was determined.
Chromatin immunoprecipitation (ChIP) assays
ChIP assays were performed using a ChIP assay kit (Millipore, Billerica, MA). To cross-link DNA with chromatin, 1 × 106 cells were incubated for 5 min in 1% formaldehyde at 37°C. The cells were harvested, washed with phosphate buffered saline (PBS), resuspended in lysis buffer and 200-1000 bp fragments of DNA from chromatin were prepared as recommended by the manufacturer. One hundredth of the resultant solution was used as an internal control. The remainder was immunoprecipitated for 16 hours at 4°C using anti-acetylated histone H3 and anti-acetylated histone H4 antibodies (Millipore, Billerica, MA). The precipitated immune complexes were recovered using protein A-agarose, and then purified using QIAamp DNA mini kit (QIAGEN). Samples were analyzed by qPCR to determine a ratio of histone acetylation at the IL-13Rα2 promoter site using propriety primers Hs04516601_cn for IL-13Rα2 gene and RNase P/TERT reference copy number primers after following the manufacturer's instructions (Applied Biosystems, Foster City, CA).
Bisulfite-PCR and sequencing
Bisulfite sequencing was performed using CpGenome Fast DNA Modification Kit (Millipore, Billerica, MA). Briefly, 1 μg of genome DNA was incubated for 16 hours at 50°C with sodium bisulfite solution. The modified DNA was purified by DNA binding column. The promoter region of IL-13Rα2 gene was amplified by PCR using specific primer pairs, FW: 5'-TTGGGGAGAAAGAGAGATTTG-3', and BW: 5'-CAAACTTACCCCACCCAAAA-3'. The PCR products were cloned into pCR2.1 vector using a TOPO-cloning KIT (Invitrogen, Carlsbad, CA) and sequenced using an ABI377 automated sequencer. At least 10 clones were sequenced for each cell line.
AP-1 activation assay
Nuclear extracts from cell lines were collected using the Transfactor Extract Kit (Active Motif, Carlsbad, CA) and tested for DNA binding activity using the AP-1 family TransAM Kit (Active Motif) according to the manufacturer's instructions [
28].
Immunohistochemistry (IHC) and Immunocytochemistry (ICC)
Expression of human and mouse IL-13Rα2 protein in pancreatic cancer cell lines and mouse organs was observed by indirect immunofluorescence-immunostaining as described previously [
28,
30] using anti-mouse monoclonal and anti-human IL-13Rα2 polyclonal antibodies (R&D, Minneapolis, MN). Tissue samples were fixed in 10% formalin solution for IHC and human cells were fixed by 4% paraformaldehyde (PFA) for ICC. The nucleus was counterstained by DAPI.
IL-13Rα2 gene knockdown by RNA interference
Retrovirus-mediated RNA interference was performed using the pSuper RNAi system (Oligoengine, Seattle, WA) following the manufacturer's instructions as described previously [
16,
28].
Protein synthesis inhibition assay
In vitro cytotoxic activity of IL-13 cytotoxin (IL-13-PE) was measured by the inhibition of protein synthesis as described earlier [
11]. All assays were performed in quadruplicate and data are shown as mean ± SD.
Tumor xenograft studies
Panc-1 and ASPC-1 cells (2 × 106) were injected s.c. in the left flank of female athymic nude mice. From day 4 after tumor implantation, 5 mg/kg TSA was subcutaneously (s.c.) injected every alternative days or 25 mg/kg SAHA were intraperitoneally (i.p.) injected daily for 14 days. From day 5, 50 or 100 μg/kg IL-13-PE or PBS/0.2% human serum albumin (vehicle) were intratumorally (i.t.) injected daily for 14 days. Mice body weight and tumor size was measured every 4-7 days from day 4. Measurement was continued until more than one tumor reached 20 mm in diameter in each group. Their appearances were observed through out the entire experiment for detecting toxic side effects from the treatment. Animal studies were conducted under an approved protocol in accordance with the principles and procedures outlined in the NIH Guide for the Care and Use of Laboratory Animals.
Statistical analysis
The data were analyzed for statistical significance using Student's t test for comparison between two groups and ANOVA among more than two groups. All experiments including the animal model were repeated at least twice.
Discussion
We demonstrate for the first time that IL-13Rα2, a tumor antigen, is highly susceptible to epigenetic modulation in pancreatic cancer cell lines. Interestingly, DNA methylation and histone acetylation were differentially regulated in cells overexpressing or not overexpressing IL-13Rα2. Histones (H3 and H4) were highly acetylated at the promoter region of IL-13Rα2 in IL-13Rα2-positive pancreatic cancer cell lines, but not in IL-13Rα2-negative cell lines. In contrast, histones in IL-13Rα2-negative pancreatic cell lines and normal cell lines were highly methylated, but not in IL-13Rα2 positive cell lines. The reason for the differential histone acetylation and methylation is not known but appears to correlate with IL-13Rα2 expression and may be responsible for variability of IL-13Rα2 expression in cancer cells.
The role of histone acetylation was explored further using histone deacetylase (HDAC) inhibitors. Interestingly, in the presence of HDAC inhibitors (TSA and NaB), IL-13Rα2 expression was significantly induced in IL-13Rα2-negative cell lines whose histones were not acetylated compared to IL-13Rα2-positive cell lines in which histones were acetylated. The mechanism of differential IL-13Rα2 regulation was examined. IL-13 signals through IL-13Rα2 via the AP-1 pathway and inactivation of this pathway by JNK and AP-1 inhibition suppressed IL-13Rα2 expression in IL-13Rα2-positive cell lines. Additionally, inactivation of the AP-1 pathway also suppressed induction of IL-13Rα2 by HDAC inhibitors in IL-13Rα2-negative cell lines. In accordance, Wu et al. have reported the importance of c-jun, which is a member of AP-1 transcription factor, in IL-13Rα2 expression [
32]. These observations indicate a strong correlation between transcription factor and histone acetylation in the IL-13Rα2 at the promoter region.
The significance of IL-13Rα2 upregulation by HDAC inhibitors was examined. As expected, IL-13 induced STAT6 phosphorylation in IL-13Rα2-negative pancreatic cancer cell lines (Supplementary Figure
5). Interestingly, TSA increased IL-13Rα2 expression, but suppressed STAT6 phosphorylation induced by IL-13 treatment. The suppression of STAT6 phosphorylation by TSA was inhibited by IL-13Rα2 RNAi indicating that IL-13Rα2 is directly involved in this counter-regulation (data not shown). Similarly, as expected, IL-13 did not induce MMPs expression in IL-13Rα2-negative pancreatic cancer cell lines [
28]. However, when cells were treated with TSA, IL-13 could increase MMP-9, 12 and 14 mRNA as IL-13Rα2 expression was upregulated. In contrast, MMPs were not induced by TSA when IL-13Rα2 was knocked-down by RNAi or IL-13 signaling was inhibited by JNK inhibitor.
We took advantage of upregulation of IL-13Rα2 in pancreatic cancer cell lines and hypothesized that HDAC inhibitors may enhance the sensitivity of IL-13 receptor-targeted immunotoxin, IL-13-PE, in pancreatic cancers. We have previously demonstrated that IL-13-PE is a powerful anti-cancer agent, causing regression of IL-13Rα2-positive human tumors derived from variety of human cancers including pancreatic cancer [
15,
16]. However, for efficacy, these tumors must express high levels of IL-13Rα2. Since cancer is a heterogeneous disease, drug-induced upregulation of IL-13Rα2 could be used in cancers expressing even low levels of IL-13 α2 to enhance the intensity of the immunotoxin anti-cancer response. Indeed, we demonstrate that pre-treatment of tumor cell lines
in vitro with TSA enhanced their sensitivity to IL-13-PE and made IL-13Rα2-negative cell lines extremely sensitive to IL-13-PE. In contrast, TSA treatment did not sensitize normal epithelial cell lines, thus providing a therapeutic advantage of targeting tumors but not normal tissues. Consequently, the use of HDAC inhibitors may open a new avenue of treating pancreatic cancer when combined with IL-13-PE. It is possible that HDAC inhibitors may also sensitize tumors to other immunotoxins targeting different antigens or cell surface receptors.
The reason why normal epithelial cells are not sensitized to IL-13-PE by TSA is not clear. Epithelial cells exhibit a similar histone modification pattern to IL-13Rα2-negative pancreatic cancer cell lines but, IL-13Rα2 is not upregulated in normal epithelial cells by HDAC inhibitors. This may be because normal cell lines show no c-jun activity, while IL-13Rα2-negative pancreatic cancer cell lines show a 2-6 fold increase in c-jun activity indicating that TSA induction of high levels of IL-13Rα2 is dependent on the AP-1/c-jun pathway.
We also demonstrate that HDAC inhibitors when combined with IL-13-PE cause more dramatic tumor responses than those caused by either agent alone in two pancreatic cancer models. Pancreatic cancers in situ were not sensitive to IL-13-PE as they do not naturally express IL-13Rα2 and TSA or SAHA alone showed only modest to moderate anti-tumor effects. However, when TSA or SAHA were combined with IL13-PE a dramatic inhibition of tumor growth was observed. In agreement with our observations, HDAC inhibition has been reported in combination therapies for other types of cancer. Combination therapy of SAHA and retinoic acid has been examined for resistant acute promyelocytic leukemia in which SAHA enhanced the anti-cancer effect of retinoic acid [
33]. Another HDAC inhibitor, LAQ824, is reported to be effective in combination with adoptive T-cell transfer therapy against mouse model of melanoma [
34]. These authors hypothesized that LAQ824 increases the tumor-associated antigen expression enhancing the anti-tumor effectiveness of T cell therapy.
It is important to note that while HDAC inhibition enhanced the remarkable anti-cancer effects of IL-13-PE in pancreatic cancer models in vivo by upregulating IL-13Rα2 in the tumors, no significant upregulation of IL-13Rα2 expression was observed in any vital organs. In addition, no detectable histological changes were observed in any vital organs. Although IL-13-PE was injected locally, our findings confirm that this novel combination therapeutic approach is safe. Future studies will examine systemic administration of IL-13-PE in combination with HDAC inhibitors in syngenic animal tumor models. Taken together, our results provide support for testing this novel combination in the clinic for the therapy of human cancer including pancreatic cancer for which no therapeutic options are currently available.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
Conceived and designed the experiments: TF, BHJ, RKP. Performed the experiments: TF. Analyzed the data: TF. Wrote the paper: TF, BHJ, RKP.
All authors have read and approved the final manuscript.