Epigenetic reprogramming using 5-azacytidine promotes an anti-cancer response in pancreatic adenocarcinoma cells

Epigenetic reprogramming using 5-azacytidine promotes an anti-cancer response in pancreatic... Curative management of pancreatic adenocarcinoma is limited because this malignancy remains resistant to most chemotherapeutic drugs. Strategies that reverse epigenetic alterations offer a unique opportunity for cancer cell reprogramming, which is valuable for development of new treatments. The aim of this work was to reprogram pancreatic ductal adenocarcinoma (PDAC) cells toward a less aggressive and drug-responsive phenotype. The process applied is called “epigenetic reprogramming”. To evaluate the efficiency of PDAC epigenetic reprogramming, we assessed tumor growth and drug response in PANC-1 cells after exposure to non-cytotoxic doses of the demethylating agent 5-azacytidine (5-AZA). Here, we showed that an epigenetic regimen using 5-AZA promoted an anti-cancer response by inhibiting PDAC tumor growth in vivo after the engraftment of treated cells. Remarkably, the subsequent addition of gemcitabine (GEM) to the 5-AZA-mediated reprogramming resulted in a marked growth inhibition effect in GEM-resistant pancreatic cancer cells. We observed that various characteristic peptides expressed in the pancreas, which included the antiproliferative hormone somatostatin (SST) and the SST receptor 2 (SSTR2), were significantly upregulated in the epigenetically reprogrammed PDAC cells. The inhibitory effect of octreotide (OCT), an SST analog, was tested on PDAC cells and found to be improved after cell reprogramming. Furthermore, we found that SST gene expression restoration following 5-AZA treatment or following knockdown of the DNA methyltransferase (DNMT)1 enzyme was associated with the reversion of SST epigenetic silencing through regional CpG demethylation. Lastly, we confirmed the efficacy of 5-AZA-based epigenetic reprogramming in vivo using a PDAC tumor growth model. In conclusion, this study demonstrates that epigenetic reprogramming using the demethylating compound 5-AZA shows anti-cancer effects in PANC-1 cells and is potentially attractive for the treatment of solid tumors. Introduction Diagnosis is frequently late because of the absence of disease- Pancreatic cancer is one of the most aggressive and resis- specific symptoms and new patients usually present with tant forms of malignancy . Mainly represented by pancreatic advanced or metastatic diseases. The deoxycytidine analog ductal adenocarcinoma (PDAC), it represents the fifth lead- gemcitabine (GEM) and GEM-based combination therapies ing cause of cancer-related death in industrialized countries . have been considered as standard treatments for limiting 3,4 pancreatic cancer progression . However, tumor ablation remains the only potentially curative option for pancreatic Correspondence: Luc Gailhouste (lgailhou@ncc.go.jp) or Takahiro Ochiya cancer. Given that only 15–20% of PDAC patients are con- (tochiya@ncc.go.jp) Division of Molecular and Cellular Medicine, National Cancer Center Research sidered to be appropriate candidates for surgical resection Institute, Tokyo, Japan and rapidly develop local recurrence , new therapeutic Graduate School of Medicine, The University of Tokyo, Tokyo, Japan alternatives are urgently required. Full list of author information is available at the end of the article Edited by P. Bouillet © The Author(s) 2018 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to theCreativeCommons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Official journal of the Cell Death Differentiation Association 1234567890():,; 1234567890():,; Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 2 of 12 Epigenetic regulations are crucial for orchestrating key first evaluated the effect of the demethylating agent biological events in eukaryotic organisms including 5-AZA on the four human PDAC cell lines PANC-1, embryonic development, cell differentiation, and mod- Capan-2, PL45, and SU.86-86. Cell viability assays were ulation of tissue-specific gene expression . Epigenetic performed and showed a clear dose–response effect, marks, such as DNA cytosine methylation and histone resulting in a gradual decrease in cell growth and sig- modifications, help to ensure the integrity of the genome nificant toxicity after 3 days for the PDAC cells that had been treated with high doses of 5-AZA (Fig. 1). Calcula- and maintain methylation states over the course of 7,8 repeated cell divisions . The significance of DNA tion of 5-AZA IC50 showed that Capan-2 and PANC-1 methylation has been extensively described in cancer cells, cells were the most resistant to the demethylating drug, in which oncogenes and tumor-suppressor genes acquire with an IC50 of 71.3 and 45.6 μM, respectively, after 48 h 9,10 cancer-specific methylation patterns . Unlike oncogenic of exposure. Conversely, SU.86.86 cells appeared more mutations, which are permanent changes in the cancer responsive to 5-AZA (IC50 = 19.2 μM, 48 h). genome, epigenetic alterations are potentially reversible, Next, we assessed the in vivo tumorigenic ability of offering a unique therapeutic opportunity . The cytidine PDAC cells after epigenetic reprogramming. Experiments analogs 5-azacytidine (5-AZA, azacytidine) and its deoxy were carried out using the PANC-1 cell line because these derivative 5-aza-2′-deoxycytidine (5-AZA-dC, decitabine) cells exhibited one of the most aggressive phenotypes have shown efficacy for the treatment of myelodysplastic among the four PDAC cell lines previously characterized. syndromes . Regarding the treatment of solid tumors, Based on our experience, the PANC-1 cells also showed development of epigenetic therapies has started to regain tumorigenic ability in vivo. The concentration of 5-AZA attention despite the variable efficacies reported so used for the epigenetic reprogramming of PANC-1 cells 13,14 far . was determined based on the MTT assays and IC50 values The development of relevant strategies erasing “cancer to minimize the cytotoxic effect of the compound. imprinting” and aberrantly hypermethylated marks Accordingly, PANC-1 cells were implanted into mice after represents a valuable asset for the therapeutic manage- a 2-week reprogramming regimen using 3 μM 5-AZA ment of pancreatic adenocarcinoma. The aim of this work with daily replacement, and tumor size was monitored for was to investigate the feasibility of reversing the malignant 12 weeks (Fig. 2a). As a result, a significant and persistent phenotype of pancreatic cancer cells by epigenetic inhibition of tumor growth was observed with the epi- reprogramming using the human PDAC cell line PANC- genetically reprogrammed cells compared with the 1. We first evaluated PANC-1 cell growth in response to PANC-1 cells that were not treated prior to inoculation 5-AZA treatment in vitro to determinate the optimal (p < 0.001, t-test) (Fig. 2b). Remarkably, the repro- concentration for cell reprogramming. Next, PDAC grammed PANC-1 cells nearly lost their ability to form tumor growth was analyzed in vivo after the engraftment tumors in vivo as tumor nodules were barely measurable of epigenetically reprogrammed PANC-1 cells into mice up to 9 weeks after cell engraftment. to validate the efficiency of the procedure. Importantly, we investigated whether 5-AZA-based epigenetic repro- Epigenetic reprogramming with non-cytotoxic doses of gramming could potentiate the cytotoxic effect of the 5-AZA sensitizes PANC-1 cells to gemcitabine chemotherapeutic agent GEM on resistant PDAC cells. In We evaluated whether 5-AZA-based epigenetic repro- addition, we explored the molecular mechanism under- gramming could potentiate the cytotoxic effect of the lying the reversion of the epigenetic silencing of char- chemotherapeutic agent GEM in resistant pancreatic acteristic markers expressed the pancreas, in particular for cancer cells. First, PANC-1 cells were reprogrammed by the antiproliferative hormone somatostatin (SST), which 5-AZA treatment (3 μM) for 2 weeks (Fig. 3a). Following was seen in reprogrammed pancreatic cancer cells. To this epigenetic reprogramming regimen, the cells were this end, the correlations between the expression and reseeded without 5-AZA and treated with increasing methylation profiles of the SST gene were analyzed after concentrations of GEM. Importantly, we confirmed that a 5-AZA-mediated epigenetic reprogramming and DNA 2-week reprogramming of PANC-1 cells using 3 μM methyltransferase (DNMT) 1 knockdown. Lastly, we 5-AZA did not significantly affect cell viability (Supple- assessed the potential anti-cancer action of an epigenetic mentary Figure 1). The growth inhibitory effect of GEM regimen on PDAC tumors in vivo. on PANC-1 cells was assessed by an MTT assay after 48 h of treatment, and the IC50 was calculated. Measurement Results of cell growth confirmed that non-reprogrammed PANC- Exposure to the epigenetic drug 5-AZA inhibits PDAC 1 cells were resistant to GEM with an IC50 greater than tumor growth 1000 μM (Fig. 3b). Conversely, growth of epigenetically To investigate the prospective therapeutic use of epi- reprogrammed PDAC cells was significantly inhibited by genetic reprogramming in pancreatic adenocarcinoma, we GEM in a concentration-dependent manner, with an IC50 Official journal of the Cell Death Differentiation Association Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 3 of 12 100,000 35,000 PANC-1 Capan-2 75,000 2.5 25,000 50,000 7.5 15,000 25,000 0 5,000 0 1234 5 01 2345 Time (day) Time (day) 5-AZA (µM) 55,000 55,000 PL45 SU.86.86 45,000 45,000 35,000 35,000 25,000 25,000 15,000 15,000 5,000 5,000 0 1234 5 01 2345 Time (day) Time (day) Fig. 1 Inhibition of PDAC cell growth in response to 5-AZA treatment. Time and dose-dependent cytotoxicity of 5-AZA as evaluated in the four human PDAC cell lines PANC-1, Capan-2, PL45, and SU.86.86. Twenty-four hours after seeding, cells were treated with 5-AZA at the indicated concentrations (day-0). 5-AZA medium was prepared and replaced daily. Cell number was estimated at the indicated times using a cell viability assay (MTT). The data depicted show the mean ± standard deviation (SD) and are representative of three distinct experiments equivalent to 111.6 μM after 48 h of treatment (p < 0.001 AZA-treated cells, with an induction ratio greater than compared with the control cells, t-test). While GEM had a 55-fold (p < 0.001, t-test). Using the publicly available data limited effect on PANC-1 cells, the sequential combina- sets from the Human Protein Atlas Program, we con- tion of 5-AZA-based epigenetic reprogramming and GEM firmed that SST protein is expressed in normal pancreatic increased sensitivity of PDAC cells toward GEM in vitro. tissues, but strongly repressed in pancreatic tumors (Supplementary Figure 2). Next, we quantified the mRNA 5-AZA-based epigenetic reprogramming enhances SST levels of the five human somatostatin receptors (SSTRs) expression in PANC-1 cells and restores SST analog before and after epigenetic reprogramming. As shown in response Fig. 4b, SSTR2, SSTR3, SSTR4, and SSTR5 but not SSTR1 To assess the molecular phenotype of PANC-1 cells in were expressed in PANC-1 cells. Among the four detected response to the 5-AZA-mediated epigenetic reprogram- receptors, SSTR2 exhibited the highest expression, ming, the expression level of several endocrine markers whereas SSTR3, SSTR4, and SSTR5 were expressed at was analyzed by RT-qPCR. Significant differences were similar low levels. The epigenetically reprogrammed cells obtained with some of the most characteristic peptides showed a significant 3.1-, 2.2-, and 2.0-fold induction of produced by the pancreas, such as insulin (INS), glucagon SSTR2, SSTR4, and SSTR5 mRNA, respectively, compared (GCG), amylin (islet amyloid polypeptide, IAPP), and SST, with the control cells (p < 0.001, t-test). which were all consistently upregulated in response to SST analogs have been used for treating gastro- 5-AZA treatment (Fig. 4a). In this study, SST was con- enteropancreatic neuroendocrine tumors, a rare form of sidered for further investigation because of the potential malignancy . However, assessment of the therapeutic tumor-suppressor activity of this antiproliferative hor- index of SST analogs in the management of PDAC tumors 15,16 18 mone . We observed that the mRNAs of SST were at are still needed . Here, we evaluated the effect of the SST extremely low levels in non-reprogrammed PDAC cells, analog octreotide (OCT) on PDAC cell growth. As pre- whereas SST expression was remarkably increased in 5- viously described, PANC-1 cells were reprogrammed by a Official journal of the Cell Death Differentiation Association Cell number Cell number Cell number Cell number Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 4 of 12 PANC-1 Xenograft seeding (8 mice/condition) Epigenetic reprogramming (in vitro) 3 days Tumor growth (in vivo) 02 0 12 (week) +/- 5-AZA (3 µM); daily replacement Tumor size measurement 1,500 PANC-1 Reprog_PANC-1 1,000 (5-AZA) p<0.001 0123 456789 10 11 12 Time (week) Fig. 2 Suppression of PANC-1 cell tumorigenicity after exposure to non-cytotoxic doses of 5-AZA. a Experimental design for the assessment of in vivo PDAC cell growth after epigenetic reprogramming. PANC-1 cells were pretreated with 3 µM 5-AZA for 2 weeks (in vitro). 5-AZA medium was prepared and replaced daily. After 3 days without 5-AZA, the reprogrammed and control cells were subcutaneously implanted in athymic nude mice. An equal number of viable cells was injected for each inoculum. The tumor nodules were monitored twice a week for 12 weeks (in vivo). b Xenograft tumorigenicity assay following the epigenetic reprogramming of PANC-1 cells. The data depicted show the mean ± standard error of the mean (SEM). The p value was calculated with a t-test to statistically evaluate the difference in tumor growth between the control (N = 8) and the reprogrammed PANC-1 cell group (N = 8) exhibited significant demethylation of the analyzed CpG 5-AZA regimen (3 μM for 2 weeks) before the beginning of the treatment with different dosages of OCT. We sites (p < 0.001, t-test), which was consistent with the observed that PANC-1 cells were resistant to OCT and reexpression of the SST gene observed after 5-AZA did not show any modification of their growth. Con- treatment. Interestingly, the COBRA data also identified a versely, epigenetic reprogramming was able to improve correlation between the restoration of INS expression in OCT-induced PANC-1 cell response (Fig. 4c). Although 5-AZA-treated cells (Fig. 4a) and demethylation of the the growth inhibitory effect of the SST analog observed in CpG sites located in INS promoter (Supplementary Fig- the reprogrammed PDAC cells was moderate, statistical ure 3a and 3b). significance was reached, with an inhibition of 10.5 ± 5.2 CpG methylation is primarily controlled by three major and 14.4 ± 4.2% using 10 and 100 μM OCT, respectively DNMT enzymes, of which DNMT1 plays a critical role in (p < 0.05 and p < 0.01 compared with the control repro- the maintenance of methylation patterns during cellular 8,19 grammed cells not exposed to OCT, t-test). replication . The expression and methylation profile of SST was evaluated after siRNA-based silencing to sub- 5-AZA treatment and DNMT1 knockdown reverts the stantiate the impact of DNMT1 repression in the epige- epigenetic silencing of SST netic reprogramming process. As a result, the specific To address whether 5-AZA acts by directly influencing knockdown of DNMT1 was correlated with a significant the methylation state of the SST gene, the correlation increase in SST expression levels (Fig. 5c). In addition, the between DNA methylation and SST expression levels was COBRA data revealed that SST reexpression was accom- evaluated in reprogrammed and control PDAC cells. In panied by a substantial demethylation of the CpG sites silico genomic analysis revealed that SST contained a located in its promoter region (Fig. 5d). A similar result CpG-rich region in its promoter (Fig. 5a). We performed was observed regarding the expression and methylation combined bisulfite restriction analysis (COBRA) to level of the INS gene after DNMT1 experimental silencing examine the methylation status of the identified CpG sites (Supplementary Figure 3c and 3d), demonstrating the and found prominent hypermethylation of the SST pro- efficiency of the DNMT1 siRNA in mimicking the effects moter in PANC-1 cells, with a methylation rate of nearly of 5-AZA treatment. Taken together, these data impli- 100% (Fig. 5b). By contrast, reprogrammed PANC-1 cells cated DNMT1 in the maintenance of SST epigenetic Official journal of the Cell Death Differentiation Association Tumor volume (mm ) Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 5 of 12 PANC-1 Seeding in seeding 96-well plate MTT assay Epigenetic reprogramming IC50 12 14 0 2 (day) +/- 5-AZA (3 µM); daily replacement GEM treatment 1.25 PANC-1 1.00 Reprog_PANC-1 IC50>1,000 µM 0.75 *** 0.50 IC50=111.6 µM 0.25 0.0 001 0 0.0 .01 1 0 0.1 .1 1 1 1 10 0 1 10 00 0 1,000 10 1000 00 (Log) GEM (µM) Fig. 3 Increased response to GEM treatment in epigenetically reprogrammed PANC-1 cells. a Experimental design for the measurement of GEM cytotoxicity on PANC-1 cells after epigenetic reprogramming. The PANC-1 cells were pretreated with 3 µM 5-AZA for 2 weeks with daily replacement. After 48 h without 5-AZA, the reprogrammed and control cells were seeded in 96-well plates. Twenty-four hours after seeding, GEM was added to the medium without 5-AZA and cell viability was measured using an MTT assay. b IC50 of GEM in the reprogrammed and control PANC-1 cells. Cell viability was measured at the indicated concentrations 48 h after the beginning of GEM treatment to determine the IC50 of the compound. Control PANC-1 cells showed low response to GEM (IC50 > 1000 µM). GEM IC50 was significantly reduced in epigenetically reprogrammed PANC-1 cells compared with the control cells (IC50 = 111.6 µM; p < 0.001 using the t-test). The statistical significance of the differences between the reprogrammed and control PANC-1 cells was ***p < 0.001 (t-test). All data shown in the figure are mean ± SD silencing in pancreatic adenocarcinoma and supported smaller in response to injections of the demethylating the contribution of 5-AZA-mediated DNMT1 inactiva- agent (Fig. 6b). Thus, the average size of the resected tion, which could be responsible for the demethylation of tumors was 253.9 ± 124.0 and 131.7 ± 75.2 mm for saline- the SST promoter and SST reexpression in epigenetically and 5-AZA-treated mice, respectively. As presented in reprogramed PANC-1 cells. Fig. 6c, nodule size monitoring showed a significant inhibition of tumor progression from 2.5 weeks of treat- Epigenetic reprogramming regimen shows efficacy in vivo ment (p < 0.001, t-test, N = 8 mice per group). and suppresses PANC-1 tumor growth To confirm that 5-AZA-mediated epigenetic repro- To evaluate the relevance and consistency of the epi- gramming was effective in vivo, SST and INS expression genetic reprogramming regimen, we tested whether levels were analyzed by RT-qPCR in the PDAC tumor 5-AZA treatment could modify the malignant phenotype tissues after treatment. After measurement, SST and INS of tumors generated from PDAC cells in vivo. A pilot mRNA levels were found to be relatively low in the dose–response assay with daily injections determined that untreated mice, which confirmed the silencing of these 5-AZA concentrations up to 3 mg/kg did not significantly genes in the tumors generated from the PANC-1 cells affect animal survival (Supplementary Figure 4a) and body (Fig. 6d). Conversely, 5-AZA treatment was associated weight (Supplementary Figure 4b). PANC-1 cells were with a remarkable reexpression of these two endocrine used and implanted into nude mice as xenograft models. peptides (p < 0.001, t-test). The subsequent observations 5-AZA treatment started when tumors reached a palpable demonstrated that the increased expression of SST and size (≥100 mm ). The epigenetic therapeutic procedure INS was consistent with the demethylation of their was based on an intraperitoneal (IP) injection of 3 mg/kg respective promoters, as shown by the COBRA data. 5-AZA, 6 times/week for 4 weeks, at which point the mice Spearman’s rank correlation analyses revealed that the were euthanized and the tumors collected (Fig. 6a). At the CpG methylation ratios were inversely correlated with the end of treatment, the tumors appeared to be markedly expression levels of SST (ρ = −0.770, p = 0.0029) and INS Official journal of the Cell Death Differentiation Association Cell growth (relative to control) Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 6 of 12 Fig. 4 Effect of 5-AZA-mediated epigenetic reprogramming on SST and SSTR gene expression, and SST analog response. a Relative expression levels of four major endocrine lineage markers, insulin (INS), glucagon (GCG), amylin (islet amyloid polypeptide, IAPP), and somatostatin (SST), and the insulin promoter factor (pancreatic and duodenal homeobox 1, PDX1) in epigenetically reprogrammed PANC-1 cells. The total RNAs were extracted from the PANC-1 cells after reprogramming with 3 µM 5-AZA for 14 and 16 days (T1 and T2, respectively), and the relative mRNA expression levels were determined by RT-qPCR. The non-reprogrammed PANC-1 cells were used as controls. Statistically significant differences in the gene expression levels were achieved at *p < 0.05, **p < 0.01, and ***p < 0.001 (t-test). b Expression of somatostatin receptors (SSTRs) in reprogrammed and control PANC-1 cells. Relative mRNA levels are expressed with regard to the expression level of SSTR2 measured in the non- reprogrammed PANC-1. Statistical significance: **p < 0.01 (t-test). ND not detected. c Evaluation of SST analog effect on reprogrammed PDAC cell growth. Cell viability was measured 120 h after starting treatment with the SST analog OCT at the indicated concentrations. Statistical significance: *p < 0.05 and **p < 0.01 (t-test). All data shown in the figure are mean ± SD (ρ = -0.564, p = 0.0290) in PDAC tumors. Consequently, epigenetic silencing of the antiproliferative hormone SST these results indicated that the correlation between the was reverted in epigenetically reprogrammed PDAC cells. reversion of epigenetic silencing and 5-AZA treatment In humans, ~70% of annotated gene promoters contain was effective in vivo. CpG-rich regions, which might be potentially affected by “cancer imprinting” . Our current knowledge indicates Discussion that abnormal DNA methylation is usually observed in So far, progress in treating pancreatic cancer has been pancreatic neoplasms and is correlated with the progres- 21–23 limited because of the poor response of PDAC cells to sion of the disease . However, the epigenetic silencing chemotherapies. In the present work, we reported that an of SST and its reversion by epigenetic reprogramming has epigenetic therapy regimen using the demethylating agent never been described in pancreatic cancer cells. In addi- 5-AZA significantly inhibited pancreatic tumor growth tion, this work is the first to report the therapeutic and sensitized the PDAC cell line PANC-1 to GEM and potential of a demethylating compound using an in vivo SST analog treatment. Furthermore, we showed that model of pancreatic ductal adenocarcinoma. Official journal of the Cell Death Differentiation Association Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 7 of 12 Fig. 5 Expression levels and methylation profiles of SST after 5-AZA treatment and DNMT1-knockdown. a In silico analysis of the human SST gene. The figures show the GC percentages, CpG sites, and the COBRA-amplified genomic region. The CpG sites in SST promoter were identified using the UCSC Genome Bioinformatics tool. The SST gene exhibits a CpG island of 262 bp (green horizontal bar), which contains 25 CpG sites. b COBRA was performed to evaluate CpG methylation (%) in the promoter of the SST gene in the control and epigenetically reprogrammed PANC-1 cells. Reprogrammed cells were treated with 5-AZA (3 µM) for 14 and 16 days (T1 and T2, respectively) before genomic DNA extraction. Representative data of three COBRA are shown. c Relative expression of DNMT1 and SST following DNMT1 silencing in PANC-1 cells. Two distinct siRNAs were used to target DNMT1 (siDNMT1_A and B), and two scrambled siRNAs were used as negative controls (siCtrl_A and B). The histograms show the mean ± SD of SST and DNMT1 expression levels, measured 6 and 8 days after transfection. d SST promoter methylation percentage after DNMT1 knockdown, as determined by COBRA. The data are representative of three COBRA. Genomic DNA was extracted from PANC-1 cells 6 and 8 days after transfection. Statistical significance: *p < 0.05, **p < 0.01, and ***p < 0.001 (t-test). U unmethylated, M methylated SST acts as an endogenous inhibitory regulator of var- neuroendocrine malignancies , their efficacy for PDAC ious cellular functions including hormone secretion, cell tumor management would require further investigations 15,16 18 motility, and cell proliferation . Although SST analogs and considering SSTR status . Consistent observations have been used for treating particular pancreatic regarding the epigenetic silencing of the antiproliferative Official journal of the Cell Death Differentiation Association Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 8 of 12 PANC-1 Tumor size Tumor 5-AZA (3 mg/Kg) xenograft 100mm resection or saline solution Tumor growth (in vivo) Epigenetic reprogramming (in vivo) x 0 12 3 4 (week) Tumor size <100 mm +/- 5-AZA by IP (6 times/week) 8 mice/condition Tumor size measurement b Saline (mock) c in vivo reprogramming Saline 5-AZA 5-AZA (3 mg/Kg) p<0.001 0 0.5 1 1.5 2 2.5 3 3.5 4 Time (week) SST INS *** *** Saline 5-AZA 30 4.5 rho=-0.770 rho=-0.564 p=0.0029 3.5 p=0.0290 2.5 10 1.5 0.5 20 40 60 80 100 10 20 30 40 50 60 Methylation (%) Methylation (%) Fig. 6 Efficacy of epigenetic reprogramming regimen in repressing PANC-1 tumor progression in vivo. a Schematic outline that illustrates the epigenetic reprogramming procedure. The PANC-1 cells were used to generate tumors in nude mice. In vivo cell reprogramming was performed by treating the mice with 5-AZA for 4 weeks (6 times/week) at a dose of 3 mg/kg by intraperitoneal injection (IP). Saline solution was injected for the control group. See Supplementary Figure 4 for survival curves and animal weight in response to 5-AZA exposure. b Representative size of the tumors 3 3 at the end of the experimental protocol. Saline-treated mice: 253.9 ± 124.0 mm . 5-AZA-treated mice: 131.7 ± 75.2 mm . c PDAC tumor growth in response to 5-AZA treatment in vivo. The size of the tumor nodules in the reprogrammed (N = 8) and control group (N = 8) were monitored twice a week for 4 weeks. The data show the mean ± SEM. Statistical significance was evaluated with the t-test. d Relative expression levels and methylation profiles of SST and INS after epigenetic reprogramming. The total RNAs and genomic DNAs were extracted for analysis from the resected tumors (4 weeks of treatment). The horizontal bars show the averages and statistical significance was ***p < 0.001 (t-test). The scatter plots show the correlation between the expression of SST and INS and CpG methylation levels of their respective promoters (Spearman’s rank coefficient). The red and blue dots show the 5-AZA-treated mice (N = 8) and saline-treated mice (N = 8), respectively Official journal of the Cell Death Differentiation Association Relative expression Expression Tumor volume (mm ) Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 9 of 12 peptide SST have been seen in various types of malig- study evaluated DNMT1 gene expression in 88 PDAC nancy. For example, SST promoter hypermethylation was tumors and 10 normal pancreatic tissues and showed that reported as a common event in human esophageal car- DNMT1 was expressed in 46.6% of PDAC tissues but not cinoma and was related to early neoplastic progression in in the normal pancreatic tissues analyzed . We pre- Barrett’s esophagus . Several studies have demonstrated viously reported the inverse correlation between DNMT1 the decrease in cell proliferation induced by SST and SST expression and the tumor-suppressor microRNA-148a in 16,25 34 analogs in SSTR2-positive pancreatic cancer cells . liver cancer . Similarly, Robert and collaborators Using the PANC-1 cell line, the group of Li and colleagues demonstrated that DNMT1 was required to maintain revealed a synergistic inhibitory effect on cell growth aberrant CpG methylation in colon cancer cells . The mediated by SST analog treatment after SSTR1 and authors showed that DNMT1 knockdown significantly SSTR2 reexpression . In our study, epigenetic repro- promoted the ability of 5-AZA-dC to reactivate the gramming of PANC-1 cells was associated with an tumor-suppressor genes silenced by hypermethylation. increase in SSTR2 expression and a significant inhibition Therapeutic strategies using DNA demethylating agents of cell growth in response to the SST analog OCT. Pre- represent attractive alternatives for the treatment of solid vious data indicated the frequent downregulation of tumors . Here, we show that 5-AZA exerted its antitumor SSTRs in pancreatic adenocarcinoma tissues and derived effect by reducing the tumorigenic potential of PANC-1 27,28 cell lines . Torrisani and coworkers found an upstream cells while also increasing GEM treatment response. promoter of SSTR2 that was controlled by CpG methy- Importantly, our data demonstrated that a pretreatment lation and was hypermethylated in various pancreatic cell model rather than the use of a combination of two drugs at lines, including PANC-1 cells . SSTR1 gene inactivation once was able to reprogram PDAC cells to leave them through a similar epigenetic mechanism was also identi- primed for killing by another anti-cancer agent, such as fied in Epstein-Barr virus-positive gastric cancer .We GEM. Recent data support the idea that reactivation of performed in silico analyses and confirmed the presence specific genes by hypomethylation drugs holds the key to of large CpG islands in the coding sequences of the five therapeutic benefits in non-small cell lung cancer , mel- 38 39 40 41 SSTR genes (Supplementary Figure 5). anoma , glioma , hepatoma , and epithelial tumors . Surprisingly, we found that other endocrine lineage Previous studies in pancreatic cancer have also reported genes were significantly increased in 5-AZA-treated the use of the deoxy derivative of 5-AZA, 5-AZA-dC. PDAC cells, which included INS, GCG, and IAPP.As However, the effect of 5-AZA itself on PDAC tumor evidenced for SST, the reexpression of INS was the con- growth and GEM-drug resistance has been poorly inves- sequence of demethylation of the CpG sites located in the tigated. Missiaglia and colleagues showed that 5-AZA-dC promoter of this gene. In line with this finding, Lefebvre treatment resulted in global DNA demethylation and and colleagues previously showed that 5-AZA-dC was apoptosis of pancreatic cancer cell lines . More recently, able to induce Ngn3, a major marker for islet progenitors, 5-AZA-dC was tested in an aggressive stroma-rich mouse 31 43 and endocrine differentiation in PANC-1 cells . How- model of pancreatic adenocarcinoma . Wang and cow- ever, we were not able to detect INS protein in repro- orkers also showed that the MEK inhibitor PD98059 grammed PANC-1 cells using the ELISA method (data potentiated the capability of 5-AZA-dC to mediate growth not shown). It is reasonable to hypothesize that SST arrest in pancreatic cancer cells . Currently, a phase II augmentation observed in our cells might participate in trial is recruiting participants to determine the effect of 5- the alteration of INS production at the protein level AZA (oral azacytidine) on progression-free survival and because of its inhibitory effect. Nevertheless, our data are outcomes in patients with resected pancreatic adeno- encouraging and support the importance of further carcinoma at high risk of recurrence . developing epigenetic methods to induce the differentia- The establishment of epigenetic therapies for solid tion of PDAC cell lines toward an endocrine-like lineage tumor treatment will remain challenging. First, given that in order to generate relevant insulin-producing human our study was performed using one PDAC cell line, cell models. additional experiments using other cancer cell types and In our study, the specific depletion of the DNMT1 other DNA demethylating drugs will be required to enzyme was able to mimic 5-AZA-mediated reprogram- evaluate the potential clinical application of epigenetic ming and increased expression of SST and INS. DNMT1 reprogramming-based therapies. Next, it will be critical to expression is known to be frequently increased in PDAC accurately determine the optimal dosage of the deme- tumors and associated with poor prognosis. For example, thylating agents to maximize the possibility of long-term Wang and coworkers examined the expression of DNMT1 treatment and ensure patient response and tolerance. In by immunohistochemistry staining in PDAC and benign addition, further investigations will be required to address pancreatic tissues and found that DNMT1 protein levels the specificity of cancer cell reprogramming with regard increased from precursor to advanced lesions . Another to undesirable gene reexpression and possible side effects Official journal of the Cell Death Differentiation Association Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 10 of 12 on non-neoplastic cells. Even though epigenetic drugs SU.86.86 cells were seeded at 10,000 cells/well (6 wells/ seem to preferentially reactivate genes that have been condition). The next day, the medium was changed and abnormally silenced in cancer cells , the reason why cells were treated with the indicated concentrations of these cancer imprinted genes are more susceptible to 5-AZA for one to five days (daily replacement). Cell via- reactivation by demethylation treatment remains unclear. bility was measured at the indicated times using the Cell In summary, our study emphasizes an effective Counting Kit-8 (Dojindo), according to the manu- facturer’s instructions (MTT assay). The absorbance at method for the epigenetic-based reprogramming of pan- creatic adenocarcinoma cells. We demonstrate that epi- 450 nm was measured using the Synergy H4 Microplate genetically reprogrammed PANC-1 cells using 5-AZA Reader system (BioTek). For the evaluation of exhibit a less aggressive phenotype with impaired tumor concentration-dependent cytotoxicity of GEM (IC50), growth and improved GEM response. The results from reprogrammed and control PANC-1 cells were seeded in ongoing investigations will be essential to determine the 96-well plates (10,000 cells/well; 6 wells/condition). The therapeutic value of epigenetic compounds for potential next day, the medium was changed, and reprogrammed applications in the treatment of solid tumors. Never- and control PANC-1 cells were cultured in medium theless, it is appealing to consider that such a repro- containing different concentrations of GEM for 48 h. gramming strategy may pave the way for further Treatment with 5-AZA was discontinued two days before controlling aggressive cancers and promote development seeding in 96-well plates, and cells were maintained of alternative therapies for inoperable or drug-resistant without 5-AZA until the end of the experiments. Cell tumors. viability was measured as mentioned above. A similar protocol was used for the assessment of the effect of the Materials and methods SST analog on PANC-1 cell growth before and after Cells and reagents epigenetic reprogramming. The human pancreatic adenocarcinoma cell lines, Capan-2, PL45, and SU.86.86, were purchased from the Xenograft establishment and tumorigenicity assay American Type Culture Collection. The PANC-1 cells Female athymic nude mice were purchased at 4–5weeks were obtained from the Public Health England Culture old and housed in isolator units under controlled humidity Collection. Cultured PANC-1 and PL45 cells were main- and temperature, with a 12-h light–dark cycle. The animals tained in DMEM (Gibco) supplemented with penicillin (50 received standard sterilized food and water ad libitum. The IU/mL; Gibco), streptomycin (50 μg/mL; Gibco), and 10% epigenetically reprogrammed cells (in vitro reprogram- fetal bovine serum (FBS; Thermo Scientific). Capan-2 and ming) and control PANC-1 cells were subcutaneously SU.86.86 cells were cultured in McCoy’s 5 A (Gibco) and implanted into the right flanks of the mice at a density of RPMI 1640 medium (Gibco), respectively, and both were 8× 10 cells by inoculation in DMEM without serum (100 supplemented with penicillin (50 IU/mL; Gibco), strepto- μL/mouse). The tumor nodules were monitored twice a mycin (50 μg/mL; Gibco), and 10% FBS. The demethylating week by palpation using a digital caliper. The tumor size agent 5-azacytidine (5-AZA; PubChem CID: 9444) was was determined using the formula (length × width²)/2 from Sigma (#A2385). The drug was dissolved in (mm ). The experiments continued until tumors reached phosphate-buffered saline as a 10 mM stock, filtered (0.22 the maximum allowable size dictated by the animal care μM), and stored at −20 °C in aliquots that were thawed guidelines of our institute. Animal experimental protocols immediately prior to use. The in vitro epigenetic repro- were approved by the National Cancer Center Institutional gramming procedure was performed by addition of 5-AZA Animal Care and Use Committee. to the PANC-1 cells at a concentration of 3 μM. Given the short half-life of the compound in culture media, 5-AZA Cell transfection medium was prepared and replaced daily. Gemcitabine The PANC-1 cells were seeded at a density of 40,000 (GEM; PubChem CID: 60750) and the SST analog cells/cm² in 35-mm-diameter culture dishes and trans- octreotide (OCT; PubChem CID: 448601) were purchased fected the next day using the TransFectin lipid reagent from Sigma (#G6423 and O1014, respectively). The com- (Bio-Rad Laboratories). The cells were incubated with the pounds were dissolved in H O as a 10 mM stock solution transfection mix containing 100 nM of siRNA and 5 μL of for GEM and 1 mM stock solution for the SST analog, TransFectin in a 1.2 mL total volume of serum- and filtered (0.22 μM), and stored at −20 °C. antibiotic-free OptiMEM (Invitrogen) for 5 h. The two human DNMT1 siRNAs were purchased from Ambion Cell growth assay and IC50 (ID #s4215 and #s4217; siDNMT1_A and B, respectively). For the evaluation of the time and dose-dependent The two control siRNAs, AllStars Negative Control (ID cytotoxicity of 5-AZA, PANC-1 cells were seeded at 7,500 #1027281; siCtrl_A) and Silencer Select Negative Control cells per well in 96-well plates and Capan-2, PL45, and Official journal of the Cell Death Differentiation Association Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 11 of 12 siRNA (ID #4390843; siCtrl_B), were purchased from amplification, 3 μL of amplified products were digested Qiagen and Life Technologies, respectively. with three units of restriction enzyme. Finally, the restric- tion products were separated by 10% PAGE and visualized Total RNA and genomic DNA isolation by ethidium bromide staining. The bands were densito- The mRNAs were purified using the miRNeasy Mini Kit metrically analyzed using the software ImageJ (v1.38×, (Qiagen), according to the manufacturer’s protocol. The National Institutes of Health, USA; http://rsb.info.-nih.gov/ ij) to quantify the unmethylated (U) and methylated (M) total RNAs were quantified using a NanoDrop 1000 spectrophotometer (Thermo Scientific), and the restriction fragments. The methylation levels were calcu- integrity of the RNA was evaluated with an Agilent 2100 lated for each locus using the formula (M × 100)/(M+ U) Bioanalyzer (Agilent Technologies). The genomic DNA and were expressed as a methylation percentage. was extracted using the GenElute Mammalian Genomic DNA Miniprep Kit (Sigma) and was quantified on a In vivo epigenetic reprogramming NanoDrop 1000 spectrophotometer. To determine the optimal dose of the demethylating drug for in vivo administration, the mice received a daily Real-time reverse transcription quantitative polymerase IP injection of 5-AZA diluted in sterile saline solution at chain reaction (RT-qPCR) concentrations ranging from 1.5 to 15 mg/kg (N = 5 for To evaluate the gene expression levels, the total RNAs each concentration) 6 times/week. The animals’ condi- were first treated with DNase using the TURBO DNA-free tions and weights were monitored twice a week. For kit (Ambion). Then, cDNAs were synthesized from 1 μg of in vivo epigenetic reprogramming, the PDAC cells were purified mRNA using SuperScript III Reverse Transcriptase first implanted in athymic nude mice as described above. (Invitrogen), according to the manufacturer’s recommen- After the tumors reached a palpable size (≥100 mm ), the dations. SYBR Green RT-qPCR was performed to evaluate animals were added to the study and randomly separated the mRNA levels in each sample (Platinum SYBR Green into two groups. Mice that did not develop tumors were qPCR SuperMix-UDG, Invitrogen) using the Step One Plus excluded from the study. Next, the mice received an IP Real-time PCR system (Applied Biosystems). After an injection of 3 mg/kg 5-AZA (N = 8) or saline solution initial denaturation at 95 °C for 2 min, the thermal cycles (N = 8) 6 times/week for 4 weeks. The tumor size was were repeated 40 times as follows: 95 °C for 15 s and 60 °C monitored twice a week, as described above. The animals for 30 s. The housekeeping genes glyceraldehyde 3- were euthanized at the study endpoint dictated by the phosphatase dehydrogenase (GAPDH) and ribosomal animal care guidelines of our institute. The tumors were protein S18 (RPS18) were used to normalize the cDNA immediately removed and snap frozen in liquid nitrogen levels. The sequences of the human primers used for gene for storage until RNA and DNA extraction. amplification are shown in Supplementary Table 1. Statistical analysis DNA methylation analysis The experimental data are presented as the means ± SD, COBRA was used to assess the methylation status of except for the in vivo tumorigenicity assay, in which error the specific CpG sites located in the promoter regions of bars show the SEM. Student’s t-test was performed to somatostatin (SST) and insulin (INS). An in silico analysis estimate the statistical significance of the data. The using the UCSC Genome Bioinformatics Site (http:// equality of the variances was tested using an F-test, and genome.ucsc.edu) was performed to identify the CpG correction was performed in the case of unequal variances sites associated with the proximal promoter for each gene. (Welch’s t-test). All p-values were two-tailed. The corre- AproximalCpG was defined as a CpG located within 500 lations between the gene expression (RT-qPCR) and DNA bp (±) of the transcription start site. MethPrimer (http:// methylation levels (COBRA) were assessed by calculating www.urogene.org/methprimer) was used to design the the Spearman’s rank coefficient. All statistical analyses COBRA primers required to amplify the genomic regions were performed using the MedCalc software. The containing the CpG of interest (Supplementary Table 2). experimental data are representative of at least three Briefly, 1 μg of genomic DNA was subjected to bisulfite independent experiments and were considered statisti- modification treatment using the EpiTect Plus kit (QIA- cally significant at a p < 0.05. GEN). Then, COBRA PCR was performed as follows: after an initial denaturation step at 94 °C for 3 min, the following Acknowledgements thermal cycles were repeated 40 times: 94 °C for 10 s, 55 °C We thank A. Inoue for her excellent technical support. This study was for 50 s, and 72 °C for 1 min. Each COBRA PCR was per- supported by grants from the Ministry of Health, Labour, and Welfare of Japan (H24-Bsou-Kanen-Ippan-009, -011, -013) and the Research Program on formed in a total volume of 10 μL, which contained 0.5 Hepatitis from Japan Agency for Medical Research and Development, and the units of Hot Start Taq polymerase (Takara), 10 pmol of joint Research Programs of the Institute for Molecular and Cellular Regulation, primers, and 1 μL of bisulfite-treated DNA. After PCR Gunma University (Maebashi, Japan). Official journal of the Cell Death Differentiation Association Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 12 of 12 Author details 22. van Kampen, J. G. et al. Epigenetic targeting in pancreatic cancer. Cancer Treat. Division of Molecular and Cellular Medicine, National Cancer Center Research Rev. 40,656–664 (2014). Institute, Tokyo, Japan. Graduate School of Medicine, The University of Tokyo, 23. Nones, K. et al. Genome-wide DNA methylationpatterns inpancreatic ductal Tokyo, Japan. Laboratory of Genome Science, Biosignal Genome Resource adenocarcinoma reveal epigenetic deregulation of SLIT-ROBO, ITGA2 and MET Center, Institute for Molecular and Cellular Regulation, Gunma University, signaling. Int. J. Cancer 135, 1110–1118 (2014). Maebashi, Japan. National Cancer Center, Tokyo, Japan 24. Jin, Z. et al. Hypermethylation of the somatostatin promoter is a common, early event in human esophageal carcinogenesis. Cancer 112,43–49 (2008). 25. Guillermet, J. et al. Somatostatin receptor subtype 2 sensitizes human pan- Conflict of interest creatic cancer cells to death ligand-induced apoptosis. Proc. Natl. Acad. Sci. The authors declare that they have no conflict of interest. USA 100,155–160 (2003). 26. Li, M. et al. Transfection of SSTR-1 and SSTR-2 inhibits Panc-1 cell proliferation and renders Panc-1 cells responsive to somatostatin analogue. J. Am. Coll. Surg. Publisher's note 201,571–578 (2005). Springer Nature remains neutral with regard to jurisdictional claims in 27. Li, M. et al. Characterization of somatostatinreceptorexpressioninhuman published maps and institutional affiliations. pancreatic cancer usingreal-time RT-PCR. J. Surg. Res. 119,130–137 (2004). 28. Buscail, L. et al. Loss of sst2 somatostatin receptor gene expression in human Supplementary Information accompanies this paper at (https://doi.org/ pancreatic and colorectal cancer. Cancer Res. 56,1823–1827 (1996). 10.1038/s41419-018-0487-z). 29. Torrisani,J.et al. Identification of an upstream promoter of the human somatostatin receptor, hSSTR2, which is controlled by epigenetic modifica- Received: 6 February 2018 Revised: 9 March 2018 Accepted: 14 March 2018 tions. Endocrinology 149,3137–3147 (2008). 30. Zhao, J. et al. Somatostatin receptor 1, a novel EBV-associated CpG hyper- methylated gene, contributes to the pathogenesis of EBV-associated gastric cancer. Br.J.Cancer 108, 2557–2564 (2013). 31. Lefebvre, B. et al. 5’-AZA induces Ngn3 expression and endocrine differ- References entiation in the PANC-1 human ductal cell line. Biochem. Biophys. Res. Com- 1. Ryan,D.P., Hong,T.S. & Bardeesy, N. Pancreatic adenocarcinoma. N. Engl. J. mun. 391, 305–309 (2010). Med. 371,2139–2149 (2014). 32. Wang, W., Gao, J., Man, X. H., Li, Z. S. & Gong, Y. F. Significance of DNA 2. Torre, L. A. et al. Global cancer statistics, 2012. Cancer J. Clin. 65,87–108 (2015). methyltransferase-1 and histone deacetylase-1 in pancreatic cancer. Oncol. 3. Li,D., Xie, K.,Wolff,R. & Abbruzzese,J.L.Pancreaticcancer. Lancet 363, Rep. 21, 1439–1447 (2009). 1049–1057 (2004). 33. Gao, J. et al. Aberrant DNA methyltransferase expression in pancreatic ductal 4. Von Hoff, D. D. et al. Increased survival in pancreatic cancer with nab-paclitaxel adenocarcinoma development and progression. J. Exp. Clin. Cancer Res. 32,86 plus gemcitabine. N. Engl. J. Med. 369, 1691–1703 (2013). (2013). 5. Konstantinidis, I. T. et al. Pancreatic ductal adenocarcinoma: is there a survival 34. Gailhouste, L. et al. miR-148a plays a pivotal role in the liver by promoting the difference for R1 resections versus locally advanced unresectable tumors? hepatospecific phenotype and suppressing the invasiveness of transformed What is a “true” R0 resection? Ann. Surg. 257,731–736 (2013). cells. Hepatology 58,1153–1165 (2013). 6. Li, E. Chromatin modification and epigenetic reprogramming in mammalian 35. Robert, M. F. et al. DNMT1 is required to maintain CpG methylation and development. Nat. Rev. Genet. 3,662–673 (2002). aberrant gene silencing in human cancer cells. Nat. Genet. 33,61–65 (2003). 7. Deaton,A.M.&Bird,A.CpG islandsand theregulationoftranscription. Genes 36. Christman,J.K.5-Azacytidineand 5-aza-2’-deoxycytidine as inhibitors of DNA Dev. 25, 1010–1022 (2011). methylation: mechanistic studies and their implications for cancer therapy. 8. Jaenisch,R. & Bird,A.Epigeneticregulationofgeneexpression: howthe Oncogene 21,5483–5495 (2002). genome integrates intrinsic and environmental signals. Nat. Genet. 33, 37. Juergens, R. A. et al. Combination epigenetic therapy has efficacy in patients 245–254 (2003). with refractory advanced non-small cell lung cancer. Cancer Discov. 1, 598–607 9. Esteller, M. Epigenetics in cancer. N. Engl. J. Med. 358,1148–1159 (2008). (2011). 10. Jones,P.A.&Baylin, S.B.The epigenomicsofcancer. Cell 128,683–692 (2007). 38. Alcazar, O. et al. Epigenetic regulation by decitabine of melanoma differ- 11. Dawson, M. A. & Kouzarides, T. Cancer epigenetics: from mechanism to entiation in vitro and in vivo. Int. J. Cancer 131,18–29 (2012). therapy. Cell 150,12–27 (2012). 39. Turcan, S. et al. Efficient induction of differentiation and growth inhibition in 12. Fenaux, P. et al. Efficacy of azacitidine compared with that of conventional IDH1 mutant glioma cells by the DNMT inhibitor decitabine. Oncotarget 4, care regimens in the treatment of higher-risk myelodysplastic syndromes: a 1729–1736 (2013). randomised, open-label, phase III study. Lancet Oncol. 10,223–232 (2009). 40. Gailhouste, L. et al. Epigenetic reprogramming of human hepatoma cells: a 13. Pfister, S. X. & Ashworth, A. Marked for death: targeting epigenetic changes in low-cost option for drug metabolism assessment. Cell. Mol. Gastroenterol. cancer. Nat. Rev. Drug Discov. 16,241–263 (2017). Hepatol. 5,454–457 (2018). 14. Azad, N.,Zahnow, C. A., Rudin,C.M.& Baylin, S.B.The future of epigenetic 41. Tsai, H. C. et al. Transient low doses of DNA-demethylating agents exert therapy in solid tumours--lessons from the past. Nat. Rev. Clin. Oncol. 10, durable antitumor effects on hematological and epithelial tumor cells. Cancer 256–266 (2013). Cell 21,430–446 (2012). 15. Pyronnet, S. et al. Antitumor effects of somatostatin. Mol. Cell. Endocrinol. 286, 42. Missiaglia, E. et al. Growth delay of human pancreatic cancer cells by 230–237 (2008). methylase inhibitor 5-aza-2’-deoxycytidine treatment is associated with acti- 16. Li, M. et al. Somatostatin, somatostatin receptors, and pancreatic cancer. World vation of the interferon signalling pathway. Oncogene 24,199–211 (2005). J. Surg. 29, 293–296 (2005). 43. Shakya, R. et al. Hypomethylating therapy in an aggressive stroma-rich model 17. Oberg, K. E., Reubi, J. C., Kwekkeboom, D. J. & Krenning, E. P. Role of soma- of pancreatic carcinoma. Cancer Res. 73,885–896 (2012). tostatins in gastroenteropancreatic neuroendocrine tumor development and 44. Wang, X. et al. Effect of inhibition of MEK pathway on 5-aza-deoxycytidine- therapy. Gastroenterology 139,742–753 (2010). suppressed pancreatic cancer cell proliferation. Genet. Mol. Res. 12,5560–5573 18. Hejna, M., Schmidinger, M. & Raderer, M. The clinical role of somatostatin (2013). analogues as antineoplastic agents: much ado about nothing? Ann. Oncol. 13, 45. ClinicalTrials.gov. NCT01845805. 653–668 (2002). 46. Liang,G., Gonzales,F.A., Jones, P. A.,Orntoft,T.F. & Thykjaer,T. Analysisof 19. Bird, A. DNA methylation patterns and epigenetic memory. Genes Dev. 16, gene induction in human fibroblasts and bladder cancer cells exposed to the 6–21 (2002). methylation inhibitor 5-aza-2’-deoxycytidine. Cancer Res. 62,961–966 (2002). 20. Saxonov, S., Berg, P. & Brutlag, D. L. A genome-wide analysis of CpG dinu- 47. Xiong, Z. & Laird, P. W. COBRA: a sensitive and quantitative DNA methylation cleotides in the human genome distinguishes two distinct classes of pro- assay. Nucleic Acids Res. 25,2532–2534 (1997). moters. Proc.Natl. Acad.Sci.USA 103, 1412–1417 (2006). 21. Silverman,B.R.&Shi, J. Alterations ofepigeneticregulators in pancreatic cancer and their clinical implications. Int. J. Mol. Sci. 17, 2138 (2016). Official journal of the Cell Death Differentiation Association http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Cell Death & Disease Springer Journals

Epigenetic reprogramming using 5-azacytidine promotes an anti-cancer response in pancreatic adenocarcinoma cells

Free
12 pages
Loading next page...
 
/lp/springer_journal/epigenetic-reprogramming-using-5-azacytidine-promotes-an-anti-cancer-Ac68NMxhDL
Publisher
Nature Publishing Group UK
Copyright
Copyright © 2018 by The Author(s)
Subject
Life Sciences; Life Sciences, general; Biochemistry, general; Cell Biology; Immunology; Cell Culture; Antibodies
eISSN
2041-4889
D.O.I.
10.1038/s41419-018-0487-z
Publisher site
See Article on Publisher Site

Abstract

Curative management of pancreatic adenocarcinoma is limited because this malignancy remains resistant to most chemotherapeutic drugs. Strategies that reverse epigenetic alterations offer a unique opportunity for cancer cell reprogramming, which is valuable for development of new treatments. The aim of this work was to reprogram pancreatic ductal adenocarcinoma (PDAC) cells toward a less aggressive and drug-responsive phenotype. The process applied is called “epigenetic reprogramming”. To evaluate the efficiency of PDAC epigenetic reprogramming, we assessed tumor growth and drug response in PANC-1 cells after exposure to non-cytotoxic doses of the demethylating agent 5-azacytidine (5-AZA). Here, we showed that an epigenetic regimen using 5-AZA promoted an anti-cancer response by inhibiting PDAC tumor growth in vivo after the engraftment of treated cells. Remarkably, the subsequent addition of gemcitabine (GEM) to the 5-AZA-mediated reprogramming resulted in a marked growth inhibition effect in GEM-resistant pancreatic cancer cells. We observed that various characteristic peptides expressed in the pancreas, which included the antiproliferative hormone somatostatin (SST) and the SST receptor 2 (SSTR2), were significantly upregulated in the epigenetically reprogrammed PDAC cells. The inhibitory effect of octreotide (OCT), an SST analog, was tested on PDAC cells and found to be improved after cell reprogramming. Furthermore, we found that SST gene expression restoration following 5-AZA treatment or following knockdown of the DNA methyltransferase (DNMT)1 enzyme was associated with the reversion of SST epigenetic silencing through regional CpG demethylation. Lastly, we confirmed the efficacy of 5-AZA-based epigenetic reprogramming in vivo using a PDAC tumor growth model. In conclusion, this study demonstrates that epigenetic reprogramming using the demethylating compound 5-AZA shows anti-cancer effects in PANC-1 cells and is potentially attractive for the treatment of solid tumors. Introduction Diagnosis is frequently late because of the absence of disease- Pancreatic cancer is one of the most aggressive and resis- specific symptoms and new patients usually present with tant forms of malignancy . Mainly represented by pancreatic advanced or metastatic diseases. The deoxycytidine analog ductal adenocarcinoma (PDAC), it represents the fifth lead- gemcitabine (GEM) and GEM-based combination therapies ing cause of cancer-related death in industrialized countries . have been considered as standard treatments for limiting 3,4 pancreatic cancer progression . However, tumor ablation remains the only potentially curative option for pancreatic Correspondence: Luc Gailhouste (lgailhou@ncc.go.jp) or Takahiro Ochiya cancer. Given that only 15–20% of PDAC patients are con- (tochiya@ncc.go.jp) Division of Molecular and Cellular Medicine, National Cancer Center Research sidered to be appropriate candidates for surgical resection Institute, Tokyo, Japan and rapidly develop local recurrence , new therapeutic Graduate School of Medicine, The University of Tokyo, Tokyo, Japan alternatives are urgently required. Full list of author information is available at the end of the article Edited by P. Bouillet © The Author(s) 2018 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to theCreativeCommons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Official journal of the Cell Death Differentiation Association 1234567890():,; 1234567890():,; Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 2 of 12 Epigenetic regulations are crucial for orchestrating key first evaluated the effect of the demethylating agent biological events in eukaryotic organisms including 5-AZA on the four human PDAC cell lines PANC-1, embryonic development, cell differentiation, and mod- Capan-2, PL45, and SU.86-86. Cell viability assays were ulation of tissue-specific gene expression . Epigenetic performed and showed a clear dose–response effect, marks, such as DNA cytosine methylation and histone resulting in a gradual decrease in cell growth and sig- modifications, help to ensure the integrity of the genome nificant toxicity after 3 days for the PDAC cells that had been treated with high doses of 5-AZA (Fig. 1). Calcula- and maintain methylation states over the course of 7,8 repeated cell divisions . The significance of DNA tion of 5-AZA IC50 showed that Capan-2 and PANC-1 methylation has been extensively described in cancer cells, cells were the most resistant to the demethylating drug, in which oncogenes and tumor-suppressor genes acquire with an IC50 of 71.3 and 45.6 μM, respectively, after 48 h 9,10 cancer-specific methylation patterns . Unlike oncogenic of exposure. Conversely, SU.86.86 cells appeared more mutations, which are permanent changes in the cancer responsive to 5-AZA (IC50 = 19.2 μM, 48 h). genome, epigenetic alterations are potentially reversible, Next, we assessed the in vivo tumorigenic ability of offering a unique therapeutic opportunity . The cytidine PDAC cells after epigenetic reprogramming. Experiments analogs 5-azacytidine (5-AZA, azacytidine) and its deoxy were carried out using the PANC-1 cell line because these derivative 5-aza-2′-deoxycytidine (5-AZA-dC, decitabine) cells exhibited one of the most aggressive phenotypes have shown efficacy for the treatment of myelodysplastic among the four PDAC cell lines previously characterized. syndromes . Regarding the treatment of solid tumors, Based on our experience, the PANC-1 cells also showed development of epigenetic therapies has started to regain tumorigenic ability in vivo. The concentration of 5-AZA attention despite the variable efficacies reported so used for the epigenetic reprogramming of PANC-1 cells 13,14 far . was determined based on the MTT assays and IC50 values The development of relevant strategies erasing “cancer to minimize the cytotoxic effect of the compound. imprinting” and aberrantly hypermethylated marks Accordingly, PANC-1 cells were implanted into mice after represents a valuable asset for the therapeutic manage- a 2-week reprogramming regimen using 3 μM 5-AZA ment of pancreatic adenocarcinoma. The aim of this work with daily replacement, and tumor size was monitored for was to investigate the feasibility of reversing the malignant 12 weeks (Fig. 2a). As a result, a significant and persistent phenotype of pancreatic cancer cells by epigenetic inhibition of tumor growth was observed with the epi- reprogramming using the human PDAC cell line PANC- genetically reprogrammed cells compared with the 1. We first evaluated PANC-1 cell growth in response to PANC-1 cells that were not treated prior to inoculation 5-AZA treatment in vitro to determinate the optimal (p < 0.001, t-test) (Fig. 2b). Remarkably, the repro- concentration for cell reprogramming. Next, PDAC grammed PANC-1 cells nearly lost their ability to form tumor growth was analyzed in vivo after the engraftment tumors in vivo as tumor nodules were barely measurable of epigenetically reprogrammed PANC-1 cells into mice up to 9 weeks after cell engraftment. to validate the efficiency of the procedure. Importantly, we investigated whether 5-AZA-based epigenetic repro- Epigenetic reprogramming with non-cytotoxic doses of gramming could potentiate the cytotoxic effect of the 5-AZA sensitizes PANC-1 cells to gemcitabine chemotherapeutic agent GEM on resistant PDAC cells. In We evaluated whether 5-AZA-based epigenetic repro- addition, we explored the molecular mechanism under- gramming could potentiate the cytotoxic effect of the lying the reversion of the epigenetic silencing of char- chemotherapeutic agent GEM in resistant pancreatic acteristic markers expressed the pancreas, in particular for cancer cells. First, PANC-1 cells were reprogrammed by the antiproliferative hormone somatostatin (SST), which 5-AZA treatment (3 μM) for 2 weeks (Fig. 3a). Following was seen in reprogrammed pancreatic cancer cells. To this epigenetic reprogramming regimen, the cells were this end, the correlations between the expression and reseeded without 5-AZA and treated with increasing methylation profiles of the SST gene were analyzed after concentrations of GEM. Importantly, we confirmed that a 5-AZA-mediated epigenetic reprogramming and DNA 2-week reprogramming of PANC-1 cells using 3 μM methyltransferase (DNMT) 1 knockdown. Lastly, we 5-AZA did not significantly affect cell viability (Supple- assessed the potential anti-cancer action of an epigenetic mentary Figure 1). The growth inhibitory effect of GEM regimen on PDAC tumors in vivo. on PANC-1 cells was assessed by an MTT assay after 48 h of treatment, and the IC50 was calculated. Measurement Results of cell growth confirmed that non-reprogrammed PANC- Exposure to the epigenetic drug 5-AZA inhibits PDAC 1 cells were resistant to GEM with an IC50 greater than tumor growth 1000 μM (Fig. 3b). Conversely, growth of epigenetically To investigate the prospective therapeutic use of epi- reprogrammed PDAC cells was significantly inhibited by genetic reprogramming in pancreatic adenocarcinoma, we GEM in a concentration-dependent manner, with an IC50 Official journal of the Cell Death Differentiation Association Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 3 of 12 100,000 35,000 PANC-1 Capan-2 75,000 2.5 25,000 50,000 7.5 15,000 25,000 0 5,000 0 1234 5 01 2345 Time (day) Time (day) 5-AZA (µM) 55,000 55,000 PL45 SU.86.86 45,000 45,000 35,000 35,000 25,000 25,000 15,000 15,000 5,000 5,000 0 1234 5 01 2345 Time (day) Time (day) Fig. 1 Inhibition of PDAC cell growth in response to 5-AZA treatment. Time and dose-dependent cytotoxicity of 5-AZA as evaluated in the four human PDAC cell lines PANC-1, Capan-2, PL45, and SU.86.86. Twenty-four hours after seeding, cells were treated with 5-AZA at the indicated concentrations (day-0). 5-AZA medium was prepared and replaced daily. Cell number was estimated at the indicated times using a cell viability assay (MTT). The data depicted show the mean ± standard deviation (SD) and are representative of three distinct experiments equivalent to 111.6 μM after 48 h of treatment (p < 0.001 AZA-treated cells, with an induction ratio greater than compared with the control cells, t-test). While GEM had a 55-fold (p < 0.001, t-test). Using the publicly available data limited effect on PANC-1 cells, the sequential combina- sets from the Human Protein Atlas Program, we con- tion of 5-AZA-based epigenetic reprogramming and GEM firmed that SST protein is expressed in normal pancreatic increased sensitivity of PDAC cells toward GEM in vitro. tissues, but strongly repressed in pancreatic tumors (Supplementary Figure 2). Next, we quantified the mRNA 5-AZA-based epigenetic reprogramming enhances SST levels of the five human somatostatin receptors (SSTRs) expression in PANC-1 cells and restores SST analog before and after epigenetic reprogramming. As shown in response Fig. 4b, SSTR2, SSTR3, SSTR4, and SSTR5 but not SSTR1 To assess the molecular phenotype of PANC-1 cells in were expressed in PANC-1 cells. Among the four detected response to the 5-AZA-mediated epigenetic reprogram- receptors, SSTR2 exhibited the highest expression, ming, the expression level of several endocrine markers whereas SSTR3, SSTR4, and SSTR5 were expressed at was analyzed by RT-qPCR. Significant differences were similar low levels. The epigenetically reprogrammed cells obtained with some of the most characteristic peptides showed a significant 3.1-, 2.2-, and 2.0-fold induction of produced by the pancreas, such as insulin (INS), glucagon SSTR2, SSTR4, and SSTR5 mRNA, respectively, compared (GCG), amylin (islet amyloid polypeptide, IAPP), and SST, with the control cells (p < 0.001, t-test). which were all consistently upregulated in response to SST analogs have been used for treating gastro- 5-AZA treatment (Fig. 4a). In this study, SST was con- enteropancreatic neuroendocrine tumors, a rare form of sidered for further investigation because of the potential malignancy . However, assessment of the therapeutic tumor-suppressor activity of this antiproliferative hor- index of SST analogs in the management of PDAC tumors 15,16 18 mone . We observed that the mRNAs of SST were at are still needed . Here, we evaluated the effect of the SST extremely low levels in non-reprogrammed PDAC cells, analog octreotide (OCT) on PDAC cell growth. As pre- whereas SST expression was remarkably increased in 5- viously described, PANC-1 cells were reprogrammed by a Official journal of the Cell Death Differentiation Association Cell number Cell number Cell number Cell number Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 4 of 12 PANC-1 Xenograft seeding (8 mice/condition) Epigenetic reprogramming (in vitro) 3 days Tumor growth (in vivo) 02 0 12 (week) +/- 5-AZA (3 µM); daily replacement Tumor size measurement 1,500 PANC-1 Reprog_PANC-1 1,000 (5-AZA) p<0.001 0123 456789 10 11 12 Time (week) Fig. 2 Suppression of PANC-1 cell tumorigenicity after exposure to non-cytotoxic doses of 5-AZA. a Experimental design for the assessment of in vivo PDAC cell growth after epigenetic reprogramming. PANC-1 cells were pretreated with 3 µM 5-AZA for 2 weeks (in vitro). 5-AZA medium was prepared and replaced daily. After 3 days without 5-AZA, the reprogrammed and control cells were subcutaneously implanted in athymic nude mice. An equal number of viable cells was injected for each inoculum. The tumor nodules were monitored twice a week for 12 weeks (in vivo). b Xenograft tumorigenicity assay following the epigenetic reprogramming of PANC-1 cells. The data depicted show the mean ± standard error of the mean (SEM). The p value was calculated with a t-test to statistically evaluate the difference in tumor growth between the control (N = 8) and the reprogrammed PANC-1 cell group (N = 8) exhibited significant demethylation of the analyzed CpG 5-AZA regimen (3 μM for 2 weeks) before the beginning of the treatment with different dosages of OCT. We sites (p < 0.001, t-test), which was consistent with the observed that PANC-1 cells were resistant to OCT and reexpression of the SST gene observed after 5-AZA did not show any modification of their growth. Con- treatment. Interestingly, the COBRA data also identified a versely, epigenetic reprogramming was able to improve correlation between the restoration of INS expression in OCT-induced PANC-1 cell response (Fig. 4c). Although 5-AZA-treated cells (Fig. 4a) and demethylation of the the growth inhibitory effect of the SST analog observed in CpG sites located in INS promoter (Supplementary Fig- the reprogrammed PDAC cells was moderate, statistical ure 3a and 3b). significance was reached, with an inhibition of 10.5 ± 5.2 CpG methylation is primarily controlled by three major and 14.4 ± 4.2% using 10 and 100 μM OCT, respectively DNMT enzymes, of which DNMT1 plays a critical role in (p < 0.05 and p < 0.01 compared with the control repro- the maintenance of methylation patterns during cellular 8,19 grammed cells not exposed to OCT, t-test). replication . The expression and methylation profile of SST was evaluated after siRNA-based silencing to sub- 5-AZA treatment and DNMT1 knockdown reverts the stantiate the impact of DNMT1 repression in the epige- epigenetic silencing of SST netic reprogramming process. As a result, the specific To address whether 5-AZA acts by directly influencing knockdown of DNMT1 was correlated with a significant the methylation state of the SST gene, the correlation increase in SST expression levels (Fig. 5c). In addition, the between DNA methylation and SST expression levels was COBRA data revealed that SST reexpression was accom- evaluated in reprogrammed and control PDAC cells. In panied by a substantial demethylation of the CpG sites silico genomic analysis revealed that SST contained a located in its promoter region (Fig. 5d). A similar result CpG-rich region in its promoter (Fig. 5a). We performed was observed regarding the expression and methylation combined bisulfite restriction analysis (COBRA) to level of the INS gene after DNMT1 experimental silencing examine the methylation status of the identified CpG sites (Supplementary Figure 3c and 3d), demonstrating the and found prominent hypermethylation of the SST pro- efficiency of the DNMT1 siRNA in mimicking the effects moter in PANC-1 cells, with a methylation rate of nearly of 5-AZA treatment. Taken together, these data impli- 100% (Fig. 5b). By contrast, reprogrammed PANC-1 cells cated DNMT1 in the maintenance of SST epigenetic Official journal of the Cell Death Differentiation Association Tumor volume (mm ) Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 5 of 12 PANC-1 Seeding in seeding 96-well plate MTT assay Epigenetic reprogramming IC50 12 14 0 2 (day) +/- 5-AZA (3 µM); daily replacement GEM treatment 1.25 PANC-1 1.00 Reprog_PANC-1 IC50>1,000 µM 0.75 *** 0.50 IC50=111.6 µM 0.25 0.0 001 0 0.0 .01 1 0 0.1 .1 1 1 1 10 0 1 10 00 0 1,000 10 1000 00 (Log) GEM (µM) Fig. 3 Increased response to GEM treatment in epigenetically reprogrammed PANC-1 cells. a Experimental design for the measurement of GEM cytotoxicity on PANC-1 cells after epigenetic reprogramming. The PANC-1 cells were pretreated with 3 µM 5-AZA for 2 weeks with daily replacement. After 48 h without 5-AZA, the reprogrammed and control cells were seeded in 96-well plates. Twenty-four hours after seeding, GEM was added to the medium without 5-AZA and cell viability was measured using an MTT assay. b IC50 of GEM in the reprogrammed and control PANC-1 cells. Cell viability was measured at the indicated concentrations 48 h after the beginning of GEM treatment to determine the IC50 of the compound. Control PANC-1 cells showed low response to GEM (IC50 > 1000 µM). GEM IC50 was significantly reduced in epigenetically reprogrammed PANC-1 cells compared with the control cells (IC50 = 111.6 µM; p < 0.001 using the t-test). The statistical significance of the differences between the reprogrammed and control PANC-1 cells was ***p < 0.001 (t-test). All data shown in the figure are mean ± SD silencing in pancreatic adenocarcinoma and supported smaller in response to injections of the demethylating the contribution of 5-AZA-mediated DNMT1 inactiva- agent (Fig. 6b). Thus, the average size of the resected tion, which could be responsible for the demethylation of tumors was 253.9 ± 124.0 and 131.7 ± 75.2 mm for saline- the SST promoter and SST reexpression in epigenetically and 5-AZA-treated mice, respectively. As presented in reprogramed PANC-1 cells. Fig. 6c, nodule size monitoring showed a significant inhibition of tumor progression from 2.5 weeks of treat- Epigenetic reprogramming regimen shows efficacy in vivo ment (p < 0.001, t-test, N = 8 mice per group). and suppresses PANC-1 tumor growth To confirm that 5-AZA-mediated epigenetic repro- To evaluate the relevance and consistency of the epi- gramming was effective in vivo, SST and INS expression genetic reprogramming regimen, we tested whether levels were analyzed by RT-qPCR in the PDAC tumor 5-AZA treatment could modify the malignant phenotype tissues after treatment. After measurement, SST and INS of tumors generated from PDAC cells in vivo. A pilot mRNA levels were found to be relatively low in the dose–response assay with daily injections determined that untreated mice, which confirmed the silencing of these 5-AZA concentrations up to 3 mg/kg did not significantly genes in the tumors generated from the PANC-1 cells affect animal survival (Supplementary Figure 4a) and body (Fig. 6d). Conversely, 5-AZA treatment was associated weight (Supplementary Figure 4b). PANC-1 cells were with a remarkable reexpression of these two endocrine used and implanted into nude mice as xenograft models. peptides (p < 0.001, t-test). The subsequent observations 5-AZA treatment started when tumors reached a palpable demonstrated that the increased expression of SST and size (≥100 mm ). The epigenetic therapeutic procedure INS was consistent with the demethylation of their was based on an intraperitoneal (IP) injection of 3 mg/kg respective promoters, as shown by the COBRA data. 5-AZA, 6 times/week for 4 weeks, at which point the mice Spearman’s rank correlation analyses revealed that the were euthanized and the tumors collected (Fig. 6a). At the CpG methylation ratios were inversely correlated with the end of treatment, the tumors appeared to be markedly expression levels of SST (ρ = −0.770, p = 0.0029) and INS Official journal of the Cell Death Differentiation Association Cell growth (relative to control) Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 6 of 12 Fig. 4 Effect of 5-AZA-mediated epigenetic reprogramming on SST and SSTR gene expression, and SST analog response. a Relative expression levels of four major endocrine lineage markers, insulin (INS), glucagon (GCG), amylin (islet amyloid polypeptide, IAPP), and somatostatin (SST), and the insulin promoter factor (pancreatic and duodenal homeobox 1, PDX1) in epigenetically reprogrammed PANC-1 cells. The total RNAs were extracted from the PANC-1 cells after reprogramming with 3 µM 5-AZA for 14 and 16 days (T1 and T2, respectively), and the relative mRNA expression levels were determined by RT-qPCR. The non-reprogrammed PANC-1 cells were used as controls. Statistically significant differences in the gene expression levels were achieved at *p < 0.05, **p < 0.01, and ***p < 0.001 (t-test). b Expression of somatostatin receptors (SSTRs) in reprogrammed and control PANC-1 cells. Relative mRNA levels are expressed with regard to the expression level of SSTR2 measured in the non- reprogrammed PANC-1. Statistical significance: **p < 0.01 (t-test). ND not detected. c Evaluation of SST analog effect on reprogrammed PDAC cell growth. Cell viability was measured 120 h after starting treatment with the SST analog OCT at the indicated concentrations. Statistical significance: *p < 0.05 and **p < 0.01 (t-test). All data shown in the figure are mean ± SD (ρ = -0.564, p = 0.0290) in PDAC tumors. Consequently, epigenetic silencing of the antiproliferative hormone SST these results indicated that the correlation between the was reverted in epigenetically reprogrammed PDAC cells. reversion of epigenetic silencing and 5-AZA treatment In humans, ~70% of annotated gene promoters contain was effective in vivo. CpG-rich regions, which might be potentially affected by “cancer imprinting” . Our current knowledge indicates Discussion that abnormal DNA methylation is usually observed in So far, progress in treating pancreatic cancer has been pancreatic neoplasms and is correlated with the progres- 21–23 limited because of the poor response of PDAC cells to sion of the disease . However, the epigenetic silencing chemotherapies. In the present work, we reported that an of SST and its reversion by epigenetic reprogramming has epigenetic therapy regimen using the demethylating agent never been described in pancreatic cancer cells. In addi- 5-AZA significantly inhibited pancreatic tumor growth tion, this work is the first to report the therapeutic and sensitized the PDAC cell line PANC-1 to GEM and potential of a demethylating compound using an in vivo SST analog treatment. Furthermore, we showed that model of pancreatic ductal adenocarcinoma. Official journal of the Cell Death Differentiation Association Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 7 of 12 Fig. 5 Expression levels and methylation profiles of SST after 5-AZA treatment and DNMT1-knockdown. a In silico analysis of the human SST gene. The figures show the GC percentages, CpG sites, and the COBRA-amplified genomic region. The CpG sites in SST promoter were identified using the UCSC Genome Bioinformatics tool. The SST gene exhibits a CpG island of 262 bp (green horizontal bar), which contains 25 CpG sites. b COBRA was performed to evaluate CpG methylation (%) in the promoter of the SST gene in the control and epigenetically reprogrammed PANC-1 cells. Reprogrammed cells were treated with 5-AZA (3 µM) for 14 and 16 days (T1 and T2, respectively) before genomic DNA extraction. Representative data of three COBRA are shown. c Relative expression of DNMT1 and SST following DNMT1 silencing in PANC-1 cells. Two distinct siRNAs were used to target DNMT1 (siDNMT1_A and B), and two scrambled siRNAs were used as negative controls (siCtrl_A and B). The histograms show the mean ± SD of SST and DNMT1 expression levels, measured 6 and 8 days after transfection. d SST promoter methylation percentage after DNMT1 knockdown, as determined by COBRA. The data are representative of three COBRA. Genomic DNA was extracted from PANC-1 cells 6 and 8 days after transfection. Statistical significance: *p < 0.05, **p < 0.01, and ***p < 0.001 (t-test). U unmethylated, M methylated SST acts as an endogenous inhibitory regulator of var- neuroendocrine malignancies , their efficacy for PDAC ious cellular functions including hormone secretion, cell tumor management would require further investigations 15,16 18 motility, and cell proliferation . Although SST analogs and considering SSTR status . Consistent observations have been used for treating particular pancreatic regarding the epigenetic silencing of the antiproliferative Official journal of the Cell Death Differentiation Association Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 8 of 12 PANC-1 Tumor size Tumor 5-AZA (3 mg/Kg) xenograft 100mm resection or saline solution Tumor growth (in vivo) Epigenetic reprogramming (in vivo) x 0 12 3 4 (week) Tumor size <100 mm +/- 5-AZA by IP (6 times/week) 8 mice/condition Tumor size measurement b Saline (mock) c in vivo reprogramming Saline 5-AZA 5-AZA (3 mg/Kg) p<0.001 0 0.5 1 1.5 2 2.5 3 3.5 4 Time (week) SST INS *** *** Saline 5-AZA 30 4.5 rho=-0.770 rho=-0.564 p=0.0029 3.5 p=0.0290 2.5 10 1.5 0.5 20 40 60 80 100 10 20 30 40 50 60 Methylation (%) Methylation (%) Fig. 6 Efficacy of epigenetic reprogramming regimen in repressing PANC-1 tumor progression in vivo. a Schematic outline that illustrates the epigenetic reprogramming procedure. The PANC-1 cells were used to generate tumors in nude mice. In vivo cell reprogramming was performed by treating the mice with 5-AZA for 4 weeks (6 times/week) at a dose of 3 mg/kg by intraperitoneal injection (IP). Saline solution was injected for the control group. See Supplementary Figure 4 for survival curves and animal weight in response to 5-AZA exposure. b Representative size of the tumors 3 3 at the end of the experimental protocol. Saline-treated mice: 253.9 ± 124.0 mm . 5-AZA-treated mice: 131.7 ± 75.2 mm . c PDAC tumor growth in response to 5-AZA treatment in vivo. The size of the tumor nodules in the reprogrammed (N = 8) and control group (N = 8) were monitored twice a week for 4 weeks. The data show the mean ± SEM. Statistical significance was evaluated with the t-test. d Relative expression levels and methylation profiles of SST and INS after epigenetic reprogramming. The total RNAs and genomic DNAs were extracted for analysis from the resected tumors (4 weeks of treatment). The horizontal bars show the averages and statistical significance was ***p < 0.001 (t-test). The scatter plots show the correlation between the expression of SST and INS and CpG methylation levels of their respective promoters (Spearman’s rank coefficient). The red and blue dots show the 5-AZA-treated mice (N = 8) and saline-treated mice (N = 8), respectively Official journal of the Cell Death Differentiation Association Relative expression Expression Tumor volume (mm ) Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 9 of 12 peptide SST have been seen in various types of malig- study evaluated DNMT1 gene expression in 88 PDAC nancy. For example, SST promoter hypermethylation was tumors and 10 normal pancreatic tissues and showed that reported as a common event in human esophageal car- DNMT1 was expressed in 46.6% of PDAC tissues but not cinoma and was related to early neoplastic progression in in the normal pancreatic tissues analyzed . We pre- Barrett’s esophagus . Several studies have demonstrated viously reported the inverse correlation between DNMT1 the decrease in cell proliferation induced by SST and SST expression and the tumor-suppressor microRNA-148a in 16,25 34 analogs in SSTR2-positive pancreatic cancer cells . liver cancer . Similarly, Robert and collaborators Using the PANC-1 cell line, the group of Li and colleagues demonstrated that DNMT1 was required to maintain revealed a synergistic inhibitory effect on cell growth aberrant CpG methylation in colon cancer cells . The mediated by SST analog treatment after SSTR1 and authors showed that DNMT1 knockdown significantly SSTR2 reexpression . In our study, epigenetic repro- promoted the ability of 5-AZA-dC to reactivate the gramming of PANC-1 cells was associated with an tumor-suppressor genes silenced by hypermethylation. increase in SSTR2 expression and a significant inhibition Therapeutic strategies using DNA demethylating agents of cell growth in response to the SST analog OCT. Pre- represent attractive alternatives for the treatment of solid vious data indicated the frequent downregulation of tumors . Here, we show that 5-AZA exerted its antitumor SSTRs in pancreatic adenocarcinoma tissues and derived effect by reducing the tumorigenic potential of PANC-1 27,28 cell lines . Torrisani and coworkers found an upstream cells while also increasing GEM treatment response. promoter of SSTR2 that was controlled by CpG methy- Importantly, our data demonstrated that a pretreatment lation and was hypermethylated in various pancreatic cell model rather than the use of a combination of two drugs at lines, including PANC-1 cells . SSTR1 gene inactivation once was able to reprogram PDAC cells to leave them through a similar epigenetic mechanism was also identi- primed for killing by another anti-cancer agent, such as fied in Epstein-Barr virus-positive gastric cancer .We GEM. Recent data support the idea that reactivation of performed in silico analyses and confirmed the presence specific genes by hypomethylation drugs holds the key to of large CpG islands in the coding sequences of the five therapeutic benefits in non-small cell lung cancer , mel- 38 39 40 41 SSTR genes (Supplementary Figure 5). anoma , glioma , hepatoma , and epithelial tumors . Surprisingly, we found that other endocrine lineage Previous studies in pancreatic cancer have also reported genes were significantly increased in 5-AZA-treated the use of the deoxy derivative of 5-AZA, 5-AZA-dC. PDAC cells, which included INS, GCG, and IAPP.As However, the effect of 5-AZA itself on PDAC tumor evidenced for SST, the reexpression of INS was the con- growth and GEM-drug resistance has been poorly inves- sequence of demethylation of the CpG sites located in the tigated. Missiaglia and colleagues showed that 5-AZA-dC promoter of this gene. In line with this finding, Lefebvre treatment resulted in global DNA demethylation and and colleagues previously showed that 5-AZA-dC was apoptosis of pancreatic cancer cell lines . More recently, able to induce Ngn3, a major marker for islet progenitors, 5-AZA-dC was tested in an aggressive stroma-rich mouse 31 43 and endocrine differentiation in PANC-1 cells . How- model of pancreatic adenocarcinoma . Wang and cow- ever, we were not able to detect INS protein in repro- orkers also showed that the MEK inhibitor PD98059 grammed PANC-1 cells using the ELISA method (data potentiated the capability of 5-AZA-dC to mediate growth not shown). It is reasonable to hypothesize that SST arrest in pancreatic cancer cells . Currently, a phase II augmentation observed in our cells might participate in trial is recruiting participants to determine the effect of 5- the alteration of INS production at the protein level AZA (oral azacytidine) on progression-free survival and because of its inhibitory effect. Nevertheless, our data are outcomes in patients with resected pancreatic adeno- encouraging and support the importance of further carcinoma at high risk of recurrence . developing epigenetic methods to induce the differentia- The establishment of epigenetic therapies for solid tion of PDAC cell lines toward an endocrine-like lineage tumor treatment will remain challenging. First, given that in order to generate relevant insulin-producing human our study was performed using one PDAC cell line, cell models. additional experiments using other cancer cell types and In our study, the specific depletion of the DNMT1 other DNA demethylating drugs will be required to enzyme was able to mimic 5-AZA-mediated reprogram- evaluate the potential clinical application of epigenetic ming and increased expression of SST and INS. DNMT1 reprogramming-based therapies. Next, it will be critical to expression is known to be frequently increased in PDAC accurately determine the optimal dosage of the deme- tumors and associated with poor prognosis. For example, thylating agents to maximize the possibility of long-term Wang and coworkers examined the expression of DNMT1 treatment and ensure patient response and tolerance. In by immunohistochemistry staining in PDAC and benign addition, further investigations will be required to address pancreatic tissues and found that DNMT1 protein levels the specificity of cancer cell reprogramming with regard increased from precursor to advanced lesions . Another to undesirable gene reexpression and possible side effects Official journal of the Cell Death Differentiation Association Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 10 of 12 on non-neoplastic cells. Even though epigenetic drugs SU.86.86 cells were seeded at 10,000 cells/well (6 wells/ seem to preferentially reactivate genes that have been condition). The next day, the medium was changed and abnormally silenced in cancer cells , the reason why cells were treated with the indicated concentrations of these cancer imprinted genes are more susceptible to 5-AZA for one to five days (daily replacement). Cell via- reactivation by demethylation treatment remains unclear. bility was measured at the indicated times using the Cell In summary, our study emphasizes an effective Counting Kit-8 (Dojindo), according to the manu- facturer’s instructions (MTT assay). The absorbance at method for the epigenetic-based reprogramming of pan- creatic adenocarcinoma cells. We demonstrate that epi- 450 nm was measured using the Synergy H4 Microplate genetically reprogrammed PANC-1 cells using 5-AZA Reader system (BioTek). For the evaluation of exhibit a less aggressive phenotype with impaired tumor concentration-dependent cytotoxicity of GEM (IC50), growth and improved GEM response. The results from reprogrammed and control PANC-1 cells were seeded in ongoing investigations will be essential to determine the 96-well plates (10,000 cells/well; 6 wells/condition). The therapeutic value of epigenetic compounds for potential next day, the medium was changed, and reprogrammed applications in the treatment of solid tumors. Never- and control PANC-1 cells were cultured in medium theless, it is appealing to consider that such a repro- containing different concentrations of GEM for 48 h. gramming strategy may pave the way for further Treatment with 5-AZA was discontinued two days before controlling aggressive cancers and promote development seeding in 96-well plates, and cells were maintained of alternative therapies for inoperable or drug-resistant without 5-AZA until the end of the experiments. Cell tumors. viability was measured as mentioned above. A similar protocol was used for the assessment of the effect of the Materials and methods SST analog on PANC-1 cell growth before and after Cells and reagents epigenetic reprogramming. The human pancreatic adenocarcinoma cell lines, Capan-2, PL45, and SU.86.86, were purchased from the Xenograft establishment and tumorigenicity assay American Type Culture Collection. The PANC-1 cells Female athymic nude mice were purchased at 4–5weeks were obtained from the Public Health England Culture old and housed in isolator units under controlled humidity Collection. Cultured PANC-1 and PL45 cells were main- and temperature, with a 12-h light–dark cycle. The animals tained in DMEM (Gibco) supplemented with penicillin (50 received standard sterilized food and water ad libitum. The IU/mL; Gibco), streptomycin (50 μg/mL; Gibco), and 10% epigenetically reprogrammed cells (in vitro reprogram- fetal bovine serum (FBS; Thermo Scientific). Capan-2 and ming) and control PANC-1 cells were subcutaneously SU.86.86 cells were cultured in McCoy’s 5 A (Gibco) and implanted into the right flanks of the mice at a density of RPMI 1640 medium (Gibco), respectively, and both were 8× 10 cells by inoculation in DMEM without serum (100 supplemented with penicillin (50 IU/mL; Gibco), strepto- μL/mouse). The tumor nodules were monitored twice a mycin (50 μg/mL; Gibco), and 10% FBS. The demethylating week by palpation using a digital caliper. The tumor size agent 5-azacytidine (5-AZA; PubChem CID: 9444) was was determined using the formula (length × width²)/2 from Sigma (#A2385). The drug was dissolved in (mm ). The experiments continued until tumors reached phosphate-buffered saline as a 10 mM stock, filtered (0.22 the maximum allowable size dictated by the animal care μM), and stored at −20 °C in aliquots that were thawed guidelines of our institute. Animal experimental protocols immediately prior to use. The in vitro epigenetic repro- were approved by the National Cancer Center Institutional gramming procedure was performed by addition of 5-AZA Animal Care and Use Committee. to the PANC-1 cells at a concentration of 3 μM. Given the short half-life of the compound in culture media, 5-AZA Cell transfection medium was prepared and replaced daily. Gemcitabine The PANC-1 cells were seeded at a density of 40,000 (GEM; PubChem CID: 60750) and the SST analog cells/cm² in 35-mm-diameter culture dishes and trans- octreotide (OCT; PubChem CID: 448601) were purchased fected the next day using the TransFectin lipid reagent from Sigma (#G6423 and O1014, respectively). The com- (Bio-Rad Laboratories). The cells were incubated with the pounds were dissolved in H O as a 10 mM stock solution transfection mix containing 100 nM of siRNA and 5 μL of for GEM and 1 mM stock solution for the SST analog, TransFectin in a 1.2 mL total volume of serum- and filtered (0.22 μM), and stored at −20 °C. antibiotic-free OptiMEM (Invitrogen) for 5 h. The two human DNMT1 siRNAs were purchased from Ambion Cell growth assay and IC50 (ID #s4215 and #s4217; siDNMT1_A and B, respectively). For the evaluation of the time and dose-dependent The two control siRNAs, AllStars Negative Control (ID cytotoxicity of 5-AZA, PANC-1 cells were seeded at 7,500 #1027281; siCtrl_A) and Silencer Select Negative Control cells per well in 96-well plates and Capan-2, PL45, and Official journal of the Cell Death Differentiation Association Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 11 of 12 siRNA (ID #4390843; siCtrl_B), were purchased from amplification, 3 μL of amplified products were digested Qiagen and Life Technologies, respectively. with three units of restriction enzyme. Finally, the restric- tion products were separated by 10% PAGE and visualized Total RNA and genomic DNA isolation by ethidium bromide staining. The bands were densito- The mRNAs were purified using the miRNeasy Mini Kit metrically analyzed using the software ImageJ (v1.38×, (Qiagen), according to the manufacturer’s protocol. The National Institutes of Health, USA; http://rsb.info.-nih.gov/ ij) to quantify the unmethylated (U) and methylated (M) total RNAs were quantified using a NanoDrop 1000 spectrophotometer (Thermo Scientific), and the restriction fragments. The methylation levels were calcu- integrity of the RNA was evaluated with an Agilent 2100 lated for each locus using the formula (M × 100)/(M+ U) Bioanalyzer (Agilent Technologies). The genomic DNA and were expressed as a methylation percentage. was extracted using the GenElute Mammalian Genomic DNA Miniprep Kit (Sigma) and was quantified on a In vivo epigenetic reprogramming NanoDrop 1000 spectrophotometer. To determine the optimal dose of the demethylating drug for in vivo administration, the mice received a daily Real-time reverse transcription quantitative polymerase IP injection of 5-AZA diluted in sterile saline solution at chain reaction (RT-qPCR) concentrations ranging from 1.5 to 15 mg/kg (N = 5 for To evaluate the gene expression levels, the total RNAs each concentration) 6 times/week. The animals’ condi- were first treated with DNase using the TURBO DNA-free tions and weights were monitored twice a week. For kit (Ambion). Then, cDNAs were synthesized from 1 μg of in vivo epigenetic reprogramming, the PDAC cells were purified mRNA using SuperScript III Reverse Transcriptase first implanted in athymic nude mice as described above. (Invitrogen), according to the manufacturer’s recommen- After the tumors reached a palpable size (≥100 mm ), the dations. SYBR Green RT-qPCR was performed to evaluate animals were added to the study and randomly separated the mRNA levels in each sample (Platinum SYBR Green into two groups. Mice that did not develop tumors were qPCR SuperMix-UDG, Invitrogen) using the Step One Plus excluded from the study. Next, the mice received an IP Real-time PCR system (Applied Biosystems). After an injection of 3 mg/kg 5-AZA (N = 8) or saline solution initial denaturation at 95 °C for 2 min, the thermal cycles (N = 8) 6 times/week for 4 weeks. The tumor size was were repeated 40 times as follows: 95 °C for 15 s and 60 °C monitored twice a week, as described above. The animals for 30 s. The housekeeping genes glyceraldehyde 3- were euthanized at the study endpoint dictated by the phosphatase dehydrogenase (GAPDH) and ribosomal animal care guidelines of our institute. The tumors were protein S18 (RPS18) were used to normalize the cDNA immediately removed and snap frozen in liquid nitrogen levels. The sequences of the human primers used for gene for storage until RNA and DNA extraction. amplification are shown in Supplementary Table 1. Statistical analysis DNA methylation analysis The experimental data are presented as the means ± SD, COBRA was used to assess the methylation status of except for the in vivo tumorigenicity assay, in which error the specific CpG sites located in the promoter regions of bars show the SEM. Student’s t-test was performed to somatostatin (SST) and insulin (INS). An in silico analysis estimate the statistical significance of the data. The using the UCSC Genome Bioinformatics Site (http:// equality of the variances was tested using an F-test, and genome.ucsc.edu) was performed to identify the CpG correction was performed in the case of unequal variances sites associated with the proximal promoter for each gene. (Welch’s t-test). All p-values were two-tailed. The corre- AproximalCpG was defined as a CpG located within 500 lations between the gene expression (RT-qPCR) and DNA bp (±) of the transcription start site. MethPrimer (http:// methylation levels (COBRA) were assessed by calculating www.urogene.org/methprimer) was used to design the the Spearman’s rank coefficient. All statistical analyses COBRA primers required to amplify the genomic regions were performed using the MedCalc software. The containing the CpG of interest (Supplementary Table 2). experimental data are representative of at least three Briefly, 1 μg of genomic DNA was subjected to bisulfite independent experiments and were considered statisti- modification treatment using the EpiTect Plus kit (QIA- cally significant at a p < 0.05. GEN). Then, COBRA PCR was performed as follows: after an initial denaturation step at 94 °C for 3 min, the following Acknowledgements thermal cycles were repeated 40 times: 94 °C for 10 s, 55 °C We thank A. Inoue for her excellent technical support. This study was for 50 s, and 72 °C for 1 min. Each COBRA PCR was per- supported by grants from the Ministry of Health, Labour, and Welfare of Japan (H24-Bsou-Kanen-Ippan-009, -011, -013) and the Research Program on formed in a total volume of 10 μL, which contained 0.5 Hepatitis from Japan Agency for Medical Research and Development, and the units of Hot Start Taq polymerase (Takara), 10 pmol of joint Research Programs of the Institute for Molecular and Cellular Regulation, primers, and 1 μL of bisulfite-treated DNA. After PCR Gunma University (Maebashi, Japan). Official journal of the Cell Death Differentiation Association Gailhouste et al. Cell Death and Disease (2018) 9:468 Page 12 of 12 Author details 22. van Kampen, J. G. et al. Epigenetic targeting in pancreatic cancer. Cancer Treat. Division of Molecular and Cellular Medicine, National Cancer Center Research Rev. 40,656–664 (2014). Institute, Tokyo, Japan. Graduate School of Medicine, The University of Tokyo, 23. Nones, K. et al. Genome-wide DNA methylationpatterns inpancreatic ductal Tokyo, Japan. Laboratory of Genome Science, Biosignal Genome Resource adenocarcinoma reveal epigenetic deregulation of SLIT-ROBO, ITGA2 and MET Center, Institute for Molecular and Cellular Regulation, Gunma University, signaling. Int. J. Cancer 135, 1110–1118 (2014). Maebashi, Japan. National Cancer Center, Tokyo, Japan 24. Jin, Z. et al. Hypermethylation of the somatostatin promoter is a common, early event in human esophageal carcinogenesis. Cancer 112,43–49 (2008). 25. Guillermet, J. et al. Somatostatin receptor subtype 2 sensitizes human pan- Conflict of interest creatic cancer cells to death ligand-induced apoptosis. Proc. Natl. Acad. Sci. The authors declare that they have no conflict of interest. USA 100,155–160 (2003). 26. Li, M. et al. Transfection of SSTR-1 and SSTR-2 inhibits Panc-1 cell proliferation and renders Panc-1 cells responsive to somatostatin analogue. J. Am. Coll. Surg. Publisher's note 201,571–578 (2005). Springer Nature remains neutral with regard to jurisdictional claims in 27. Li, M. et al. Characterization of somatostatinreceptorexpressioninhuman published maps and institutional affiliations. pancreatic cancer usingreal-time RT-PCR. J. Surg. Res. 119,130–137 (2004). 28. Buscail, L. et al. Loss of sst2 somatostatin receptor gene expression in human Supplementary Information accompanies this paper at (https://doi.org/ pancreatic and colorectal cancer. Cancer Res. 56,1823–1827 (1996). 10.1038/s41419-018-0487-z). 29. Torrisani,J.et al. Identification of an upstream promoter of the human somatostatin receptor, hSSTR2, which is controlled by epigenetic modifica- Received: 6 February 2018 Revised: 9 March 2018 Accepted: 14 March 2018 tions. Endocrinology 149,3137–3147 (2008). 30. Zhao, J. et al. Somatostatin receptor 1, a novel EBV-associated CpG hyper- methylated gene, contributes to the pathogenesis of EBV-associated gastric cancer. Br.J.Cancer 108, 2557–2564 (2013). 31. Lefebvre, B. et al. 5’-AZA induces Ngn3 expression and endocrine differ- References entiation in the PANC-1 human ductal cell line. Biochem. Biophys. Res. Com- 1. Ryan,D.P., Hong,T.S. & Bardeesy, N. Pancreatic adenocarcinoma. N. Engl. J. mun. 391, 305–309 (2010). Med. 371,2139–2149 (2014). 32. Wang, W., Gao, J., Man, X. H., Li, Z. S. & Gong, Y. F. Significance of DNA 2. Torre, L. A. et al. Global cancer statistics, 2012. Cancer J. Clin. 65,87–108 (2015). methyltransferase-1 and histone deacetylase-1 in pancreatic cancer. Oncol. 3. Li,D., Xie, K.,Wolff,R. & Abbruzzese,J.L.Pancreaticcancer. Lancet 363, Rep. 21, 1439–1447 (2009). 1049–1057 (2004). 33. Gao, J. et al. Aberrant DNA methyltransferase expression in pancreatic ductal 4. Von Hoff, D. D. et al. Increased survival in pancreatic cancer with nab-paclitaxel adenocarcinoma development and progression. J. Exp. Clin. Cancer Res. 32,86 plus gemcitabine. N. Engl. J. Med. 369, 1691–1703 (2013). (2013). 5. Konstantinidis, I. T. et al. Pancreatic ductal adenocarcinoma: is there a survival 34. Gailhouste, L. et al. miR-148a plays a pivotal role in the liver by promoting the difference for R1 resections versus locally advanced unresectable tumors? hepatospecific phenotype and suppressing the invasiveness of transformed What is a “true” R0 resection? Ann. Surg. 257,731–736 (2013). cells. Hepatology 58,1153–1165 (2013). 6. Li, E. Chromatin modification and epigenetic reprogramming in mammalian 35. Robert, M. F. et al. DNMT1 is required to maintain CpG methylation and development. Nat. Rev. Genet. 3,662–673 (2002). aberrant gene silencing in human cancer cells. Nat. Genet. 33,61–65 (2003). 7. Deaton,A.M.&Bird,A.CpG islandsand theregulationoftranscription. Genes 36. Christman,J.K.5-Azacytidineand 5-aza-2’-deoxycytidine as inhibitors of DNA Dev. 25, 1010–1022 (2011). methylation: mechanistic studies and their implications for cancer therapy. 8. Jaenisch,R. & Bird,A.Epigeneticregulationofgeneexpression: howthe Oncogene 21,5483–5495 (2002). genome integrates intrinsic and environmental signals. Nat. Genet. 33, 37. Juergens, R. A. et al. Combination epigenetic therapy has efficacy in patients 245–254 (2003). with refractory advanced non-small cell lung cancer. Cancer Discov. 1, 598–607 9. Esteller, M. Epigenetics in cancer. N. Engl. J. Med. 358,1148–1159 (2008). (2011). 10. Jones,P.A.&Baylin, S.B.The epigenomicsofcancer. Cell 128,683–692 (2007). 38. Alcazar, O. et al. Epigenetic regulation by decitabine of melanoma differ- 11. Dawson, M. A. & Kouzarides, T. Cancer epigenetics: from mechanism to entiation in vitro and in vivo. Int. J. Cancer 131,18–29 (2012). therapy. Cell 150,12–27 (2012). 39. Turcan, S. et al. Efficient induction of differentiation and growth inhibition in 12. Fenaux, P. et al. Efficacy of azacitidine compared with that of conventional IDH1 mutant glioma cells by the DNMT inhibitor decitabine. Oncotarget 4, care regimens in the treatment of higher-risk myelodysplastic syndromes: a 1729–1736 (2013). randomised, open-label, phase III study. Lancet Oncol. 10,223–232 (2009). 40. Gailhouste, L. et al. Epigenetic reprogramming of human hepatoma cells: a 13. Pfister, S. X. & Ashworth, A. Marked for death: targeting epigenetic changes in low-cost option for drug metabolism assessment. Cell. Mol. Gastroenterol. cancer. Nat. Rev. Drug Discov. 16,241–263 (2017). Hepatol. 5,454–457 (2018). 14. Azad, N.,Zahnow, C. A., Rudin,C.M.& Baylin, S.B.The future of epigenetic 41. Tsai, H. C. et al. Transient low doses of DNA-demethylating agents exert therapy in solid tumours--lessons from the past. Nat. Rev. Clin. Oncol. 10, durable antitumor effects on hematological and epithelial tumor cells. Cancer 256–266 (2013). Cell 21,430–446 (2012). 15. Pyronnet, S. et al. Antitumor effects of somatostatin. Mol. Cell. Endocrinol. 286, 42. Missiaglia, E. et al. Growth delay of human pancreatic cancer cells by 230–237 (2008). methylase inhibitor 5-aza-2’-deoxycytidine treatment is associated with acti- 16. Li, M. et al. Somatostatin, somatostatin receptors, and pancreatic cancer. World vation of the interferon signalling pathway. Oncogene 24,199–211 (2005). J. Surg. 29, 293–296 (2005). 43. Shakya, R. et al. Hypomethylating therapy in an aggressive stroma-rich model 17. Oberg, K. E., Reubi, J. C., Kwekkeboom, D. J. & Krenning, E. P. Role of soma- of pancreatic carcinoma. Cancer Res. 73,885–896 (2012). tostatins in gastroenteropancreatic neuroendocrine tumor development and 44. Wang, X. et al. Effect of inhibition of MEK pathway on 5-aza-deoxycytidine- therapy. Gastroenterology 139,742–753 (2010). suppressed pancreatic cancer cell proliferation. Genet. Mol. Res. 12,5560–5573 18. Hejna, M., Schmidinger, M. & Raderer, M. The clinical role of somatostatin (2013). analogues as antineoplastic agents: much ado about nothing? Ann. Oncol. 13, 45. ClinicalTrials.gov. NCT01845805. 653–668 (2002). 46. Liang,G., Gonzales,F.A., Jones, P. A.,Orntoft,T.F. & Thykjaer,T. Analysisof 19. Bird, A. DNA methylation patterns and epigenetic memory. Genes Dev. 16, gene induction in human fibroblasts and bladder cancer cells exposed to the 6–21 (2002). methylation inhibitor 5-aza-2’-deoxycytidine. Cancer Res. 62,961–966 (2002). 20. Saxonov, S., Berg, P. & Brutlag, D. L. A genome-wide analysis of CpG dinu- 47. Xiong, Z. & Laird, P. W. COBRA: a sensitive and quantitative DNA methylation cleotides in the human genome distinguishes two distinct classes of pro- assay. Nucleic Acids Res. 25,2532–2534 (1997). moters. Proc.Natl. Acad.Sci.USA 103, 1412–1417 (2006). 21. Silverman,B.R.&Shi, J. Alterations ofepigeneticregulators in pancreatic cancer and their clinical implications. Int. J. Mol. Sci. 17, 2138 (2016). Official journal of the Cell Death Differentiation Association

Journal

Cell Death & DiseaseSpringer Journals

Published: Apr 27, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off