Satellite RNA Increases DNA Damage and Accelerates Tumor Formation

Published OnlineFirst May 10, 2018; DOI: 10.1158/1541-7786.MCR-18-0139

DNA Damage and Repair Molecular Cancer Research Satellite RNA Increases DNA Damage and Accelerates Tumor Formation in Mouse Models of Pancreatic Cancer Takahiro Kishikawa, Motoyuki Otsuka, Tatsunori Suzuki, Takahiro Seimiya, Kazuma Sekiba, Rei Ishibashi, Eri Tanaka, Motoko Ohno, Mari Yamagami, and Kazuhiko Koike

Abstract

Highly repetitive tandem arrays such as satellite sequences in which may be due to impaired nuclear localization of Y-Box the centromeric and pericentromeric regions of chromosomes, Binding Protein 1 (YBX1), a component of the DNA damage which were previously considered to be silent, are actively repair machinery. In addition, when crossed with pancreas- transcribed in various biological processes, including cancers. speciﬁc Kras-mutant mice, MajSAT RNA expression resulted in In the pancreas, this aberrant expression occurs even in Kras- an earlier increase in PanIN formation. These results suggest mutated pancreatic intraepithelial neoplasia (PanIN) tissues, that aberrant MajSAT RNA expression accelerates oncogenesis which are precancerous lesions. To determine the biological by increasing the probability of a second driver mutation, thus role of satellite RNAs in carcinogenesis in vivo,weconstructed accelerating cells to exit from the breakthrough phase to the mouse major satellite (MajSAT) RNA-expressing transgenic expansion phase. mice. However, these transgenic mice did not show spontaneous malignant tumor formation under normal breeding. Implications: Aberrant expression of satellite RNAs accelerates Importantly, however, DNA damage was increased in pancre- oncogenesis through a mechanism involving increased DNA atic tissues induced by caerulein treatment or high-fat diet, damage. Mol Cancer Res; 1–8. 2018 AACR.

Introduction somes (13), were believed to be transcriptionally silenced through continuous heterochromatin formation in the normal state. Pancreatic ductal adenocarcinoma (PDAC) is one of the most However, recent studies have demonstrated that these sequences intractable diseases and is ranked the third leading cause of can be transcribed to yield satellite noncoding RNAs with impor- cancer-related death in the United States (1). The genetic muta- tant roles in the organization and regulation of genomes (14). tion proﬁle of PDAC is relatively simple. So-called "driver gene Also, aberrant expression and high levels of heterogenous tran- mutations" occur mostly in four major genes: KRAS, TP53, scripts from satellite regions were recently found in various CDKN2A, and SMAD4 (2–5). The mutations occur sequentially epithelial cancers, especially in pancreatic cancer tissues (15). during pancreatic carcinogenesis (6, 7). However, a constitutively This deregulated expression begins during PanIN formation in active KRAS mutation is likely required for initiation of carcino- Kras-mutated mice through the development of invasive carci- genesis, because it is detected at an extremely high probability noma (15, 16). Moreover, satellite RNA is released into the (over 95%) in PDAC and in 36%–87% of pancreatic intraepithe- bloodstream and detected in the serum in both pancreatic cancer lial neoplasia (PanIN) tissues, which are precancerous lesions for patients and patients with intraductal papillary mucinous neo- PDAC (5, 8, 9). In addition, genetically engineered mouse models plasm, another type of precancerous lesion in the pancreas (17). that harbor mutated Kras genes in the pancreas develop PanIN We initially hypothesized that the expression of these satellite tumors, whereas mice with other mutations do not (10–12). RNAs at such an early stage of carcinogenesis may play a role Satellite sequences, highly repetitive noncoding arrays mostly in the subsequent oncogenic processes. In our previous study, in the centromeric and pericentromeric regions of the chromo- we showed that overexpression of mouse major satellite RNA (MajSAT RNA) increases the number of genomic and mitochondrial DNA mutations in mouse primary cell lines derived from Department of Gastroenterology, Graduate School of Medicine, The University Kras-mutated PanIN tumors, and that this leads to the cellular of Tokyo, Tokyo, Japan. transformation of precancerous cells (16). We also found that Note: Supplementary data for this article are available at Molecular Cancer MajSAT RNA speciﬁcally binds to YBX1, one of which functions is Research Online (http://mcr.aacrjournals.org/). an enhancer of several DNA damage repair pathways (18). In Corresponding Author: Motoyuki Otsuka, Department of Gastroenterology, normal cells, YBX1 translocates into the nucleus from the cyto- Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, plasm under various genotoxic stresses, including oxidative stress, Tokyo 113-8655, Japan. Phone: 813-3815-5411, ext. 37966; Fax: 813-3814-0021; to repair the induced DNA damage. However, in MajSAT RNA– E-mail: [email protected] expressing cells, YBX1 is trapped in the cytoplasm by MajSAT doi: 10.1158/1541-7786.MCR-18-0139 RNA, resulting in delayed recovery from DNA damage and an 2018 American Association for Cancer Research. increase in the random gene mutation rate (16).

www.aacrjournals.org OF1

Kishikawa et al.

Although we previously showed a contribution of MajSAT RNA Immunoreaction Enhancer Solution (Toyobo Life Science). Sig- expression in carcinogenesis in vitro, additional investigation was nals were enhanced by Vectastain ABC kit (Vector Laboratories) needed to determine the role of MajSAT RNA in DNA damage according to the manufacturer's protocol and visualized with 3,30- repair and tumorigenesis in vivo. Therefore, in this study, we diaminobenzidine in buffered substrate (Nichirei Bioscience). examined whether the oncogenic functions of MajSAT RNA For the primary antibodies derived from mice, Histofine Mouse detected in vitro were similarly observed in vivo using a novel Stain Kit (Nichirei Bioscience) was used to block endogenous MajSAT RNA–expressing transgenic mouse model. immunoglobulin in the tissue. The following antibodies were used for the assay: anti-8-OHdG (#MOG-202P, JaICA), anti-YB1 Methods (#ab12148; Abcam), anti-phospho-Histone H2A.X (#2577; Cell Signaling Technology). The number of positive cells was deter- Generation of MajSAT RNA–expressing transgenic mice mined by counting 100 nuclei in every eighth field of view from To generate MajSAT RNA–expressing transgenic mice, plasmids three mice in each group. expressing MajSAT RNA were modified as follows: the pLVSIN- EF1a-MajSAT vector was constructed as described previously (16). The linearized transgene was excised from this plasmid by Mitochondrial DNA copy number measurement digestion at the AccI and NotI sites, which were microinjected into Mitochondrial DNA copy number was measured as described C57BL/6J mouse embryos (Center for Disease Biology and Inter- previously (16). Briefly, 100 ng isolated total DNA were subjected active Medicine of the University of Tokyo, Tokyo, Japan). Geno- to quantitative PCR using SYBR green (StepOnePlus Realtime typing was performed by PCR using DNA isolated from tail snips. PCR system; Thermo Fisher Scientific), and the relative copy cre/þ C Two different mouse lines were maintained. Pdx1 (19), number was calculated using the DD t method. H19 gene levels, G12D/þ cre/þ and LSL-Kras (20) were intercrossed to generate Pdx1 ; as nuclear DNA copy numbers, were used for normalization (23). G12D/þ LSL-Kras (PK) mice on a >95% C57BL/6 background (19). Each quantification was performed in triplicate. The primers used were as follows: mt-Co1, Fw: 50-CCC AAT CTC TAC CAG CAT C-30 and Rv: 50-GGC TCA TAG TAT AGC TGG AG-30, nuclear-H19, Fw: Detection of MajSAT RNA expression 50-GTA CCC ACC TGT CGT CC-30 and Rv: 50-GTC CAC GAG ACC To confirm the expression of MajSAT RNA in MajSAT RNA– AAT GAC TG-30. expressing transgenic mice, mouse tissues were immediately frozen in liquid nitrogen after resection and stored 80C. Frozen Statistical analysis tissues were crushed without thawing using SK mill (Tokken) and Statistically significant differences between two groups immediately immersed in ice-cold Isogen II reagent (Nippon were identified using Student t test when the variances were gene). The expression of MajSAT RNA was quantified using the equal and Welch t test when variances were unequal. P values TRAP-ddPCR method as described previously (17). Briefly, total less than 0.05 were considered to indicate statistical signifi- RNA was hybridized with biotin-labeled RNA oligonucleotide cance. A survival curve was created using Kaplan–Meier meth- probes, which were complementary to the core MajSAT RNA 0 od. All analyses were performed using JMP Pro software sequence. The sequence of the probe was 5 -CCUUCAGUGUG- 0 (SAS Institute Inc). CAUUUCUCAUU-3 . Consequently, nonhybridized RNAs were digested by RNase A/T1, which selectively targets single-stranded Study approval RNA. The protected double-stranded core sequence was reverse All of the experimental protocols were approved by the internal transcribed using the TaqMan MicroRNA RT Kit (Thermo Fisher ethics committee for animal experimentation (approval number: Scientific) and quantified using the QX200 Droplet Digital PCR #H17-042) and conducted in accordance with the Guidelines for system (Bio-Rad Laboratories). the Care and Use of Laboratory Animals of the Graduate School of Medicine, the University of Tokyo (Tokyo, Japan). Caerulein treatment and high-fat diet feeding To induce acute pancreatitis, 50 mg/kg/body weight caerulein (Sigma-Aldrich) was injected intraperitoneally every 8 hours for 2 Results consecutive days (21). Mice were 7–11 weeks old and weighed Inflammation in MajSAT RNA–expressing transgenic mice 20–30 g. The final day of caerulein injection was considered day 0. On the basis of our recent results showing that MajSAT RNA has For high-fat diet feeding, special forage with 60 kcal% fat an oncogenic role in vitro (16), we constructed MajSAT RNA– (D12492, Research Diets) was fed to each mouse for 12 weeks. expressing mice to determine the role of these RNAs in vivo. All mice were 9–13 weeks old and weighed 22–30 g. Endogenous MajSAT RNAs are expressed as heterogenous transcripts ranging from 200 to 8,000 bases in length with numerous IHC sequence variations, including a number of tandem repeats IHC was performed as described previously (22). Antigen (15, 16). However, it is technically difficult to reproduce such retrieval was performed by incubating the slides in a microwave diversity via external overexpression. Therefore, we adopted a oven in 10 mmol/L sodium citrate buffer (pH 6.0) for 15 minutes transgene approach that included approximately three tandem following deparaffinization. Endogenous peroxidase activity was repeats of basic MajSAT repetitive sequences, which are 855 bp in blocked by incubation in 3% hydrogen peroxide buffer for 13 length, to cover junctional parts at least twice, similar to our in vitro minutes other than when dying 8-OHdG. To minimize nonspe- analyses (Fig. 1A) (16). cific background staining, slides were blocked in 5% normal goat MajSAT RNA expression in transgenic mice was confirmed serum (Dako Corporation, Carpinteria) for 15 minutes at room using the TRAP method followed by droplet digital PCR (ddPCR), temperature. Tissues were incubated overnight at 4C with pri- which enables precise quantification of heterogenous MajSAT mary antibodies diluted with Can Get Signal Immunostain transcripts. We originally developed this method to measure

OF2 Mol Cancer Res; 2018 Molecular Cancer Research

MST RNA Increases DNA Damage and Tumor Formation

A B * pLVSIN-EF1a-MajSAT 120

SV40 * EF1a Maj Maj Maj polyA 80 855 bp

40 Absolute expression (copies/µL)

0 Wild MajSAT Wild MajSAT Pancreas Liver

C Wild MajSAT

Pancreas

Liver

Kidney

Figure 1. Establishment of MajSAT RNA–expressing mice. A, A schematic of the transgene construct. Approximately three tandem repeats of MajSAT RNA core sequences were inserted between the EF1a promotor and SV40 polyA signal sequences. B, Confirmation of MajSAT RNA expression by the TRAP-ddPCR method. MajSAT RNA levels were measured from 100 ng total RNA extracted from the pancreatic tissues and livers of wild-type and MajSAT RNA–expressing mice. The absolute copy number of MajSAT RNA represents the mean SE of four mice each. , P < 0.05. C, Pathologic images of MajSAT RNA–expressing mice. Representative H&E images are shown. The left six panels show the images of the pancreas, liver, and kidneys from wild-type mice, and the right panels show images from MajSAT RNA–expressing mice. All mice were 2 years old (n ¼ 5 for wild and n ¼ 8 for MajSAT). Scale bar, 100 mm. serum human satellite RNAs with ultrahigh sensitivity (17). Using of vascular walls were reproducibly detected in the pancreas, this method, only core sequences are concentrated and aligned in liver, and kidneys in 2-year-old MajSAT RNA–expressing mice a uniform length from heterogeneous MajSAT RNAs, and they can (Fig. 1C). This was not observed in the 1-year-old mice definitely be quantified using quantitative PCR. In pancreas and liver tissues (Supplementary Fig. S1A). Consistently, biochemical tests of from the transgenic mice, MajSAT RNA levels were approximately the sera of 2-year-old MajSAT RNA–expressing mice showed a double those of control mice, indicating that the transgene was significant increase in aspartate transaminase levels (Supplemen- indeed expressed (Fig. 1B). tary Fig. S1C). These results suggest that MajSAT RNA expression After 1 year of breeding, we did not observe any tumorigenic leads to spontaneous chronic inflammation during long-term changes in any organs in these transgenic mice (Supplementary breeding. Fig. S1A), except for spontaneous lymphomas in the liver, spleen, thymus, and thyroid in 5 cases out of 11 mice (Supplementary MajSAT RNA–expressing mice show increased DNA damage Fig. S1B), which were rarely observed in the wild-type mice. In Next, we determined the effects of inflammation on carcino- addition, greater inflammatory cell infiltration and thickening genic processes. As we previously reported that MajSAT RNA–

www.aacrjournals.org Mol Cancer Res; 2018 OF3

Kishikawa et al.

expressing pancreatic cells show impaired DNA damage repair which reflects repair of the DNA damage. In contrast, recovery and increased DNA mutation rates (16), acute pancreatitis was was significantly prolonged in MajSAT RNA–expressing mice induced by constitutive hourly intraperitoneal injections of caer- (Fig. 2B). ulein (21, 24). The degree of inflammation was not significantly As the copy number of mitochondrial DNA is correlated different between wild-type and MajSAT RNA–expressing mice inversely with the degree of mitochondrial DNA damage and (Supplementary Fig. S2A). In addition, serum amylase and lipase mutation (25), we measured the gene copy number of mitochon- levels, which reflect pancreatic damage, did not differ significantly drial cytochrome c oxidase I (mt-Co1), normalized to genomic between wild-type and MajSAT RNA–expressing mice (Supple- H19 gene levels. The mt-Co1 copy number decreased immediately mentary Fig. S2B), while alanine aminotransferase, alkaline phos- after caerulein treatment and gradually recovered over 7 days in phatase, triglyceride, and glucose levels were occasionally differ- wild-type mice, while restoration of the copy number was signif- ent. Therefore, we determined DNA damage levels in the pancre- icantly impaired in MajSAT RNA–expressing mice (Fig. 2C). This atic tissues of MajSAT RNA–expressing mice by evaluating the suggests greater accumulation of mitochondrial DNA damage in expression of 8-hydroxy-20-deoxyguanosine (8-OHdG), a repre- MajSAT RNA–expressing mice, which is consistent with our in vitro sentative oxidative DNA damage marker, using IHC. 8-OHdG results (16). stained was seen in the nucleus in MajSAT RNA–expressing pancreatic tissues, in a dot pattern that was similar to the position MajSAT RNA impairs YBX1 nuclear localization, leading to of chromatin 7 days after caerulein treatment (Fig. 2A). The accumulation of DNA damage percentage of 8-OHdG–positive cells was gradually decreased in We previously reported that MajSAT RNA binds to YBX1, wild-type mice as time passed after caerulein treatment (Fig. 2B), which functions as an enhancer of the DNA damage repair

ABC Wild (%) Wild * * 60 MajSAT 1.5 * Wild * MajSAT * 40 1.0 MajSAT *

20 0.5 Relative copy number Frequencies of 8-OHdG-positive cells 0 0 Control Day 0 Day 1 Day 3 Day 7 Control Day 0 Day 1 Day 3 Day 7

DE Wild * (%) * Wild * MajSAT 60

40 MajSAT

Frequencies of nuclear YBX1-positive cells Frequencies of nuclear 0 Control Day 0 Day 1 Day 3 Day 7

Figure 2. DNA damage was increased in MajSAT RNA–expressing pancreatic tissue after caerulein-induced pancreatitis. A, Representative IHC images of 8-OHdG (in brown) at day 7 after caerulein treatment (n ¼ 3 in each group). The right panels are magnified views of the rectangles in the left panels. Scale bar, 30 mm. Arrows show spots of strong indicating the colocalization of 8-OHdG with chromatin in the nucleus. B, Percentage of 8-OHdG–positive cells after caerulein treatment. 8-OHdG–positive cells were visually counted in every eighth field of view. Data represent the mean SE of three mice each. , P < 0.05. C, Mitochondria encoded gene (mt-Co1) levels in mouse pancreas were quantified by quantitative PCR. The decreased levels indicate an increase in mitochondrial DNA damage and mutations. Values from the nontreated control were set as 1.0. Data represent the mean SE of three mice each. , P < 0.05. D, Representative IHC images of YBX1 protein at day 0 just after caerulein treatment. The right panels are magnified views of the rectangles in the left panels (n ¼ 3 in each group). Scale bar, 30 mm. YBX1 is stained brown and the nucleus blue. Arrows show the colocalization of YBX1 with chromatin in the nucleus. E, Percentage of nuclear YBX1-positive cells after caerulein treatment. Cells in which YBX1 was detected in the nucleus were visually counted in every eighth field of view. Data represent the mean SE of three mice each. , P < 0.05.

OF4 Mol Cancer Res; 2018 Molecular Cancer Research

MST RNA Increases DNA Damage and Tumor Formation

AB 120

PK PKM PK * 80 * Foci/slide 40

PKM *

0 9w 14w 19w 24w 30w CD

1.0

PK 0.8 PK PKM

0.6 P = 0.028 0.4 Survival rate

0.2 PKM

0.0 10 20 30 40 50 60 70 80 90 week

Figure 3. MajSAT RNA–expressing mice show more, and earlier, PanIN foci in Kras-mutated pancreatic tissues. A, Representative pathologic images of 14-week-old PK and PKM mouse pancreatic tissues (n ¼ 3 in each group). The top panels show the pancreas of PK mice and the lower panels the pancreas of PKM mice. PanIN foci are circled in white. Right panels are magnified views of the rectangles in the left panels. Scale bar, 200 mm. B, The number of PanIN foci is higher in PKM mice at earlier ages. Data represent the mean SE of three independent mice each. , P < 0.05. C, Representative H&E-stained image of 1-year-old PK and PKM mice (n ¼ 3 in each group). Scale bar, 400 mm. D, Kaplan–Meier survival curve of PK (n ¼ 21) and PKM mice (n ¼ 21). The blue and red lines represent PK and PKM mice, respectively. The P value was calculated using a log-rank test. pathway. In addition, we found that MajSAT RNA inhibits (Supplementary Fig. S2C and D), suggesting that damaged DNA the nuclear translocation of YBX1 after oxidative stress, result- accumulation was probably induced by impairment of the rela- ing in suppression of the YBX1 DNA repair function (16). tively mild DNA damage repair system, which includes base Therefore, we evaluated YBX1 in tissues from wild-type excision repair (BER). As YBX1 interacts mainly with proteins and MajSAT RNA–expressing mice after caerulein treatment involved in BER to enhance its activities (18), the inhibition using IHC. Nuclear YBX1 was detected as strong spots in the of YBX1 by MajSAT RNA could cause an increase in 8-OHdG, nucleus, suggesting that YBX1 coexisted with chromatin where but not gH2AX. 8-OHdG was also strongly stained (Fig. 2D). Consistent with To confirm these results in other models, we examined the in vitro data, the percentage of cells with nuclear YBX1 DNA damage and YBX1 localization using a high-fat diet- staining was significantly increased in wild-type mice imme- induced inflammation model, which induces relatively low- diately after caerulein treatment. In contrast, MajSAT RNA– grade inflammation in the pancreas (26–28). Although there expressing mice had a relatively poor response (Fig. 2D and E), were no significant differences in increase of body weight suggesting that spatial depletion of YBX1 at DNA damage during the 12-week treatment (Supplementary Fig. S3A), Maj- sites by MajSAT RNA expression caused a delay in DNA SAT RNA–expressing mice showed more inflammatory changes damage repair. in the pancreas compared with wild-type mice (Supplementary We further examined the rate of gH2AX-positive cells after Fig. S3B). Consistent with the caerulein treatment protocol, caerulein treatment, to determine the degree of DNA double- more 8-OHdG–positive cells were detected, and the nuclear stranded breaks. Although we expected that these rates were also translocation of YBX1 was inhibited in MajSAT RNA–expres- higher in MajSAT RNA–expressing mice, there were no significant sing mice compared with wild-type mice (Supplementary differences between wild-type and MajSAT RNA–expressing cells Fig. S3C and S3D).

www.aacrjournals.org Mol Cancer Res; 2018 OF5

Kishikawa et al.

Kras Oxidative stress mutation

MajSAT RNA

DNA damage Figure 4. MajSAT RNA increases DNA damage and accelerates tumorigenesis by inhibiting DNA repair via YBX1. A YBX1 Random schematic of the proposed MajSAT mutation RNA function in vivo. Kras mutation accumlation causes oxidative stress and low levels of continuous DNA damage. YBX1, which is normally translocated to the nucleus where it is involved in DNA damage repair, is trapped by MajSAT Driver mutation RNA in the cytoplasm. Unrepaired oxidized DNAs cause accumulation of Normal acinar PanIN random point mutations and increase the probability of accidental driver gene mutations, which trigger cells to enter the expansion phase from the breakthrough phase during the oncogenesis.

Morphological change

Time cource Breakthrough phase Expansion phase

MajSAT RNA accelerates PanIN formation in pancreatic of PKM mice was significantly lower compared with PK mice Kras-mutant mice (Fig. 3D). PKM mice tended to have more severe chronic pancre- As spontaneous tumorigenesis was not observed in the pan- atitis with a greater depletion of acinar and islet cells than PK mice creatic tissues of MajSAT RNA–expressing mice, we crossed at 30–40 weeks of age (Supplementary Fig. S4A). Furthermore, PK the MajSAT RNA–expressing mice with mice that have a consti- mice developed various skin tumors in neck, eyes, whisker roots, cre/þ tutively active Kras mutation in their pancreatic cells (Pdx1 ; anal canal, and vulvo-vaginal skin, where Pdx1 is reported to be G12D/þ LSL-Kras mice). The Kras-mutant mice spontaneously expressed (31). Although these tumors were pathologically local- develop adenomatous tumors in the pancreas, which resemble ized papillomas and did not show carcinogenetic changes, their human PanIN tissues, but they rarely develop adenocarcinoma size and number were significantly larger in PKM than PK mice even after more than 1 year of breeding (29, 30). We compared the (Supplementary Fig. S4B). cre/þ G12D/þ pancreatic phenotypes between Pdx1 ; LSL-Kras mice cre/þ G12D/þ þ/ (PK) and Pdx1 ; LSL-Kras ; MajSAT mice (PKM). In the PK model, small PanIN foci began to develop sporadically at Discussion approximately 6 weeks of age, and almost all normal acinar In this report, we showed that MajSAT RNA expression delays regions were replaced by benign tumors and fibrous tissues at the repair of mild DNA damage by inhibiting nuclear transloca- 40 weeks of age. As shown in Fig. 3, the number of PanIN foci was tion of YBX1, thus increasing the development of PanIN in Kras- significantly increased in PKM mice compared with PK mice at an mutated mice (Fig. 4). These results were consistent with our early age (within 24 weeks; Fig. 3A and B). These foci increased previous in vitro results using Kras-mutated primary cell lines. and expanded, eventually merging, and became uncountable after Although MajSAT RNA expression without Kras mutation did 30 weeks of age. At this point, there were no significant differences not result in tumor development in the pancreas, the inflamma- in the number of PanIN foci between PK and PKM mice. No tion induced by caerulein or high-fat diet led to the accumulation morphologic changes were observed, including malignant trans- of DNA damage and the absence of YBX1 nuclear localization. formation such as vessel invasion and metastasis, even though the This suggests that DNA repair mechanisms that are executed mice were bred for over a year (Fig. 3C). However, the survival rate by YBX1, such as BER, were impaired after the induction of

OF6 Mol Cancer Res; 2018 Molecular Cancer Research

MST RNA Increases DNA Damage and Tumor Formation

genotoxic stress in MajSAT RNA–expressing mice. As we currently no adequate method to express all heterogeneous described previously, these phenomena led to an increase in MajSAT RNAs, we speculate this is the reason why the impact of random mutations in vitro (16).Wewerenotabletoclearly the artificial overexpression was lost after PanIN formation. demonstrate the colocalization of MajSAT RNA and YBX1 in While significant differences in malignant transformation vivo, because costaining of RNA and protein is technically between PK and PKM mice were not observed, the prognosis was difficult. This is particularly true in formalin-fixed, paraffin- worse for PKM than PK mice. Although the precise reason for the embedded tissue due to the incompatibility of in situ hybrid- differences in prognosis remains to be elucidated, PKM mice ization and IHC techniques. However, we speculate that Maj- showed worse chronic pancreatitis, which may have resulted in SAT RNA expression impaired the nuclear localization of YBX1 pancreatic endocrine and exocrine dysfunction. In addition, skin after DNA damage, similar to what we observed in vitro (16). In tumors were more severe in PKM mice, which may have affected fact, 8-OHdG expression was reduced in the chromatin of wild- their physical strength. type mice, where YBX1 was concentrated. Conversely, 8-OHdG In summary, this study confirmed that aberrant expression of expression was significantly observed in the chromatin, where MajSAT RNA is one of the oncogenic promotors through impaired localization of YBX1 was not detected, in MajSAT RNA–expres- DNA damage repair, at least partly through YBX1 mislocalization. sing mice. These results suggest that MajSAT RNA expression Further elucidating the molecular mechanisms of deregulated inhibits the nuclear translocation of YBX-1, which theoretically MajSAT RNA expression as well as the downstream events may localizes in DNA-damaged regions for their repair (18). result in additional preventative strategies against pancreatic We also showed that the emergence of PanIN foci in Kras- cancer. mutated pancreatic tissues occurred significantly earlier by ectopic MajSAT RNA expression compared with canonical PK mice. Disclosure of Potential Conflicts of Interest Generally, Kras mutations increase oxidative DNA damage spon- No potential conflicts of interest were disclosed. taneously through abnormal production of reactive oxygen spe- cies (32). This suggests that in PK and PKM mice, mutated Kras Authors' Contributions itself may function in a similar manner as the inflammation Conception and design: T. Kishikawa, M. Otsuka, K. Sekiba induced by caerulein or a high calorie diet, thus leading to Development of methodology: T. Kishikawa, K. Sekiba Acquisition of data (provided animals, acquired and managed patients, accumulation of genomic mutations and earlier PanIN formation provided facilities, etc.): T. Kishikawa, T. Suzuki, T. Seimiya, K. Sekiba, in PKM mice. R. Ishibashi, E. Tanaka, M. Ohno, M. Yamagami Carcinogenesis is divided into three phases: the breakthrough, Analysis and interpretation of data (e.g., statistical analysis, biostatistics, expansion, and invasion phase (8). In the breakthrough phase, a computational analysis): T. Kishikawa, M. Otsuka, K. Sekiba cell acquires a driver gene mutation and begins to proliferate Writing, review, and/or revision of the manuscript: T. Kishikawa, M. Otsuka, abnormally. After many years, these cells randomly obtain a K. Sekiba Administrative, technical, or material support (i.e., reporting or organizing second driver gene mutation that enables the cells to thrive in data, constructing databases): T. Kishikawa, K. Sekiba their local environment, thus entering the expansion phase. This Study supervision: K. Sekiba, K. Koike phase is followed by the acquisition of a greater number of driver gene mutations, which allows the cells to enter the invasive phase. Acknowledgments In human tissues, there is usually a long period between the This work was supported by Grants-in-Aid from the Ministry of Education, breakthrough and expansion phases. On the basis of our obser- Culture, Sports, Science and Technology, Japan (#16H05149, #16KT0109, vation that MajSAT RNA expression in mice results in significantly #26860492, #17K15923, and #15H04807; to M. Otsuka, T. Kishikawa, earlier PanIN formation, we speculate that aberrant MajSAT RNA and K. Koike), and by the Project for Cancer Research And Therapeutic Evolution (P-CREATE) from Japan Agency for Medical Research and Development expression may accelerate the expansion phase by increasing the (AMED; #JP17cm0106419, to T. Kishikawa; #JP19cm0106602, to M. Otsuka), probability of developing a second driver gene mutation (Fig. 4). and by Pancreatic disease research promotion award from Pancreas Research Although earlier PanIN formation was observed in PKM mice, Foundation of Japan (to T. Kishikawa). we did not observe any significant differences in tumor formation between PK and PKM mice after 1 year of breeding. This may be The costs of publication of this article were defrayed in part by the payment of because after PanIN formation in transgenic mice, endogenous page charges. This article must therefore be hereby marked advertisement in MajSAT RNA is expressed at much higher levels than those of the accordance with 18 U.S.C. Section 1734 solely to indicate this fact. transgene. In fact, we expressed MajSAT RNAs with three repeats of the consensus sequence. Endogenous MajSAT RNAs, however, are Received February 7, 2018; revised March 15, 2018; accepted April 24, 2018; more heterogeneous and have additional repeats. While there is published first May 10, 2018.

References 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2017. CA Cancer J Clin 4. Jones S, Zhang X, Parsons DW, Lin JC, Leary RJ, Angenendt P, et al. Core 2017;67:7–30. signaling pathways in human pancreatic cancers revealed by global geno- 2. Biankin AV, Waddell N, Kassahn KS, Gingras MC, Muthuswamy LB, Johns mic analyses. Science 2008;321:1801–6. AL, et al. Pancreatic cancer genomes reveal aberrations in axon guidance 5. Knudsen ES, O'Reilly EM, Brody JR, Witkiewicz AK. Genetic diversity of pathway genes. Nature 2012;491:399–405. pancreatic ductal adenocarcinoma and opportunities for precision med- 3. Waddell N, Pajic M, Patch AM, Chang DK, Kassahn KS, Bailey P, et al. icine. Gastroenterology 2016;150:48–63. Whole genomes redeﬁne the mutational landscape of pancreatic cancer. 6. Bardeesy N, DePinho RA. Pancreatic cancer biology and genetics. Nat Rev Nature 2015;518:495–501. Cancer 2002;2:897–909.

www.aacrjournals.org Mol Cancer Res; 2018 OF7

Kishikawa et al.

7. Hruban RH, Goggins M, Parsons J, Kern SE. Progression model for 21. Niederau C, Ferrell LD, Grendell JH. Caerulein-induced acute necrotizing pancreatic cancer. Clin Cancer Res 2000;6:2969–72.. pancreatitis in mice: protective effects of proglumide, benzotript, and 8. Vogelstein B, Kinzler KW. The path to cancer—three strikes and you're out. secretin. Gastroenterology 1985;88:1192–204.. N Engl J Med 2015;373:1895–8. 22. Yoshikawa T, Wu J, Otsuka M, Kishikawa T, Suzuki N, Takata A, et al. 9. Lohr M, Kloppel G, Maisonneuve P, Lowenfels AB, Luttges J. Frequency of Repression of MicroRNA function mediates inflammation-associated K-ras mutations in pancreatic intraductal neoplasias associated with pan- colon tumorigenesis. Gastroenterology 2017;152:631–43. creatic ductal adenocarcinoma and chronic pancreatitis: a meta-analysis. 23. Hayashi M, Imanaka-Yoshida K, Yoshida T, Wood M, Fearns C, Tatake RJ, Neoplasia 2005;7:17–23. et al. A crucial role of mitochondrial Hsp40 in preventing dilated cardio- 10. Hingorani SR, Wang L, Multani AS, Combs C, Deramaudt TB, Hruban RH, myopathy. Nat Med 2006;12:128–32. et al. Trp53R172H and KrasG12D cooperate to promote chromosomal 24. Saeki K, Kanai T, Nakano M, Nakamura Y, Miyata N, Sujino T, et al. CCL2- instability and widely metastatic pancreatic ductal adenocarcinoma in induced migration and SOCS3-mediated activation of macrophages are mice. Cancer Cell 2005;7:469–83. involved in cerulein-induced pancreatitis in mice. Gastroenterology 11. Aguirre AJ, Bardeesy N, Sinha M, Lopez L, Tuveson DA, Horner J, et al. 2012;142:1010–20 e9. Activated Kras and Ink4a/Arf deficiency cooperate to produce metastatic 25. Verschoor ML, Ungard R, Harbottle A, Jakupciak JP, Parr RL, Singh G. pancreatic ductal adenocarcinoma. Genes Dev 2003;17:3112–26. Mitochondria and cancer: past, present, and future. BioMed Res Int 12. Bardeesy N, Cheng KH, Berger JH, Chu GC, Pahler J, Olson P, et al. 2013;2013:1–10. Smad4 is dispensable for normal pancreas development yet critical in 26. Khasawneh J, Schulz MD, Walch A, Rozman J, Hrabe de Angelis M, progression and tumor biology of pancreas cancer. Genes Dev 2006; Klingenspor M, et al. Inflammation and mitochondrial fatty acid beta- 20:3130–46. oxidation link obesity to early tumor promotion. Proc Natl Acad Sci U S A 13. Plohl M, Mestrovic N, Mravinac B. Centromere identity from the DNA 2009;106:3354–9. point of view. Chromosoma 2014;123:313–25. 27. Philip B, Roland CL, Daniluk J, Liu Y, Chatterjee D, Gomez SB, et al. A high- 14. Ferreira D, Meles S, Escudeiro A, Mendes-da-Silva A, Adega F, Chaves R. fat diet activates oncogenic Kras and COX2 to induce development of Satellite non-coding RNAs: the emerging players in cells, cellular pathways pancreatic ductal adenocarcinoma in mice. Gastroenterology 2013;145: and cancer. Chromosome Res 2015;23:479–93. 1449–58. 15. Ting DT, Lipson D, Paul S, Brannigan BW, Akhavanfard S, Coffman EJ, et al. 28. Dawson DW, Hertzer K, Moro A, Donald G, Chang HH, Go VL, et al. High- Aberrant overexpression of satellite repeats in pancreatic and other epi- fat, high-calorie diet promotes early pancreatic neoplasia in the conditional thelial cancers. Science 2011;331:593–6. KrasG12D mouse model. Cancer Prev Res 2013;6:1064–73. 16. Kishikawa T, Otsuka M, Yoshikawa T, Ohno M, Ijichi H, Koike K. Satellite 29. Tuveson DA, Shaw AT, Willis NA, Silver DP, Jackson EL, Chang S, et al. RNAs promote pancreatic oncogenic processes via the dysfunction of Endogenous oncogenic K-ras(G12D) stimulates proliferation and wide- YBX1. Nat Commun 2016;7:13006. spread neoplastic and developmental defects. Cancer Cell 2004;5:375–87. 17. Kishikawa T, Otsuka M, Yoshikawa T, Ohno M, Yamamoto K, Yamamoto 30. Ijichi H, Chytil A, Gorska AE, Aakre ME, Fujitani Y, Fujitani S, et al. N, et al. Quantitation of circulating satellite RNAs in pancreatic cancer Aggressive pancreatic ductal adenocarcinoma in mice caused by pancre- patients. JCI Insight 2016;1:e86646. as-specific blockade of transforming growth factor-beta signaling in coop- 18. Eliseeva IA, Kim ER, Guryanov SG, Ovchinnikov LP, Lyabin DN. Y-box- eration with active Kras expression. Genes Dev 2006;20:3147–60. binding protein 1 (YB-1) and its functions. Biochemistry 2011;76: 31. Mazur PK, Gruner BM, Nakhai H, Sipos B, Zimber-Strobl U, Strobl LJ, et al. 1402–33. Identification of epidermal Pdx1 expression discloses different roles of 19. Hingorani SR, Petricoin EF, Maitra A, Rajapakse V, King C, Jacobetz MA, Notch1 and Notch2 in murine Kras(G12D)-induced skin carcinogenesis et al. Preinvasive and invasive ductal pancreatic cancer and its early in vivo. PLoS One 2010;5:e13578. detection in the mouse. Cancer Cell 2003;4:437–50.. 32. Liou GY, Doppler H, DelGiorno KE, Zhang L, Leitges M, Crawford HC, et al. 20. Jackson EL, Willis N, Mercer K, Bronson RT, Crowley D, Montoya R, et al. Mutant KRas-induced mitochondrial oxidative stress in acinar cells upre- Analysis of lung tumor initiation and progression using conditional gulates EGFR signaling to drive formation of pancreatic precancerous expression of oncogenic K-ras. Genes Dev 2001;15:3243–8. lesions. Cell Rep 2016;14:2325–36.

OF8 Mol Cancer Res; 2018 Molecular Cancer Research

Satellite RNA Increases DNA Damage and Accelerates Tumor Formation in Mouse Models of Pancreatic Cancer

Takahiro Kishikawa, Motoyuki Otsuka, Tatsunori Suzuki, et al.

Mol Cancer Res Published OnlineFirst May 10, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/1541-7786.MCR-18-0139

Supplementary Access the most recent supplemental material at: Material http://mcr.aacrjournals.org/content/suppl/2018/05/05/1541-7786.MCR-18-0139.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/early/2018/05/24/1541-7786.MCR-18-0139. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.