ORIGINAL ARTICLE
Biphasic Alterations in Expression and Subcellular Localization of MUC1 in Pancreatic Ductal Carcinogenesis in Syrian Hamsters Tsukasa Kitahashi, PhD, Mami Takahashi, PhD, and Toshio Imai, DVM, PhD
Objectives: The aim of the present study was to characterize molecular targets for the prevention/diagnosis of pancreatic cancer using a chemically induced hamster pancreatic carcinogenesis model, in which background injuries to the parenchyma, for example, chronic pancreatitis or acinar atrophy, are limited. Methods: Gene expression profiles in atypical hyperplasias were first investigated using a microarray technique. Immunohistochemical analyses of early lesions and invasive ductal carcinoma (IDC) were then conducted for MUC1, of which mRNA levels were prominent among the up-regulated genes, in contrast with the coexpression of epithelial-mesenchymal transition (EMT)-related proteins. Results: Immunohistochemistry for MUC1 cytoplasmic domain (MUC1CD), which was not detected in normal-like pancreatic ducts, was positive in the apical surfaces of the epithelia of hyperplasias with and without atypia and IDC areas with distinct tubular patterns. In contrast, cytoplasmic/ nuclear positivity for MUC1-CD was observed in the invasive front of IDCs. The coexpression of EMT-related proteins, such as slug and vimentin, with cytoplasmic/nuclear MUC1-CD was also detected. Conclusions: Alterations in the expression and subcellular localization of MUC1 represent a biphasic phenomenon, and the latter may be associated with EMT in pancreatic carcinogenesis in hamsters, which indicates that changes in MUC1 are important targets for pancreatic cancer prevention and chemotherapy. Key Words: pancreatic ductal adenocarcinoma, MUC1, cytoplasmic domain, epithelial-mesenchymal transition (Pancreas 2015;44: 76Y86)
P
ancreatic cancer is the fourth to sixth cause of cancer death, with a high mortality ranking in developed countries.1Y3 Invasive ductal carcinomas (IDCs) are the most common histological type in human pancreatic cancer cases. The IDC gene/protein expression profiles have been investigated with the aim of improving diagnostic and prognostic techniques and developing therapies.4Y8 Genetic studies have also identified the signature molecular profile of IDCs, consisting of mutations in the KRAS and TP53 genes and loss/inactivation of CDKN2A/p16 and SMAD4. Such genetic alterations seem to occur in a temporal sequence in putative progressive From the Central Animal Division, National Cancer Center Research Institute, Tokyo, Japan. Received for publication September 6, 2013; accepted April 9, 2014. Reprints: Toshio Imai, DVM, PhD, Central Animal Division, National Cancer Center Research Institute, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan (evA3 was confirmed to identify AHs.24
Total RNA Extraction and Gene Expression Array Total RNA was extracted from each isolated pancreatic duct of 3 BOP-treated hamsters and 2 control hamsters using ISOGEN (Nippon Gene, Toyama, Japan). After RNA purification, a cocktail of total RNA at 239 and 400 ng from 3 AH lesions and 2 normal ducts, respectively, were applied to the Amino Allyl MessageAmp II aRNA Amplification Kit (Ambion) for cDNA synthesis. The synthesized cDNAs were labeled with Cy3 fluorescein dye according to NimbleGen’s labeling method (Roche NimbleGen, Madison, WI).
FIGURE 1. Images of pancreatic ducts around the head portion of the pancreas. Pancreatic ducts were stained with RNA later containing Reactive Black 5 solution (A), an isolated duct (B), scale bar of 200 Km. Microscopic features of a normal pancreatic duct from a control hamster (C) and atypically hyperplastic duct from a BOP-treated hamster (D) stained with hematoxylin and eosin, scale bar of 100 Km. * 2014 Lippincott Williams & Wilkins
www.pancreasjournal.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
77
Pancreas
Kitahashi et al
A total of 2 Kg of Cy3 labeled cDNAs were subjected to the following gene expression analysis on the NimbleGen expression array platform. The NimbleGen hamster gene expression 385 K Microarray (Roche NimbleGen) representing 1699 genes (70 of the 1699 genes were duplicated) in the form of optimized 60-mer oligonucleotides were spotted onto the microarray slide. Array data were normalized by the robust multichip average algorithm 32 and analyzed using Arraystar Software version 11.1.0 build 49 (DNASTAR, Madison, WI). Differential expression was defined as an at least 3-fold change versus the respective controls, removing expressed sequence tags, and the Student t test was applied.
qRT-PCR Analysis Aliquots of total RNA were extracted from the normal pancreatic tissue of control animals and normal-like pancreatic tissue, and the histopathologically confirmed IDCs of BOP-treated animals were subjected to reverse transcription reactions. Each mRNA transcript was measured using a previously described qRT-PCR method.33 The sequences of the primers used were as follows: MUC1 (forward, 5’GACACCTACCTATGAG-3’; reverse, 5’-CTACAAGTTGGCAG AAGTGG-3’), slug (forward, 5’-ACTGTGACAAGGAATATGTG-3’; reverse, 5’-GACATTCTGGAGAAGGTTTTG-3’), snail1 (forward, 5’-CATCCGAAGCCACACGCTG-3’; reverse, 5’-GAAGGGCTTC TCGCCAGTG-3’), twist (forward, 5’-GATGGCAAGCTGCAGCT ATG-3’; reverse, 5’-CAGCTCCAGAGTCTCTAGAC-3’), E-cadherin (forward, 5’-CTGCAGGTCTCATCATGGC-3’; reverse, 5’-ACC TGTAGACCTCGGCACTG-3’), cdh2 (N-cadherin; forward, 5’-CT GATATATGCCCAAGACAAAG-3’; reverse, 5’-GTCTCTCTTCT GCCTTTGTAG-3’), vimentin (forward, 5’-GATTCAGGAACAGC ATGTCC-3’; reverse, 5’-CATCCACTTCACAGGTGAG-3’), transforming growth factor (TGF)-A1 (forward, 5’-TTCCTGCTTCTCATG GCCACCC-3’; reverse, 5’-TGCCGCACGCAGCAGTTCTT-3’), and 40S ribosomal protein S7 (RPS7; forward, 5’-CCAGAAAATCCA AGTCCGGC-3’; reverse, 5’-AGTCCTCAAGGATGGCATCG-3’). Data were calculated as ratios to RPS7 mRNA.
Histopathology and Immunohistochemistry Pancreatic ductal proliferative lesions were classified as hyperplasias (Hs), AHs as premalignant lesions, and adenocarcinomas (ACs)-early with minimal invasive growth and ACs-advanced according to the criteria described earlier with minor modifications.24,31 Antigen retrieval for immunohistochemistry was performed in an autoclave for 10 minutes at 121-C in 10-mM citrate buffer (pH 6.0) for MUC1-CD, vimentin, and E-cadherin. The anticytoplasmic tail of human MUC1 and antihuman slug rabbit polyclonal antibodies were purchased from Abcam (Cambridge, UK), and the antihuman vimentin mouse monoclonal (clone V9) and antihuman E-cadherin mouse monoclonal (clone 36) antibodies were from DAKO Cytomation (Glostrup, Denmark) and BD Biosciences (San Jose, CA), respectively. The streptavidin-biotinperoxidase complex method (StreptABComplex/HRP, DAKO) was used to determine the expression and distribution/localization of each antigen. Sections were lightly counterstained with hematoxylin for microscopic examinations. Negative controls without primary antibody reactions were set for each antigen using serial sections. Slug-positive ratios were counted per approximately 100 to 1000 epithelial cell nuclei in normal (like) pancreatic ducts and each lesion.
Histopathological Evaluation of Pancreatic Tissue and Isolated Ducts The incidence and multiplicity of pancreatic ductal lesions including Hs, AHs, ACs-early, and ACs-advanced in the pancreases
78
www.pancreasjournal.com
Volume 44, Number 1, January 2015
of 29 BOP-treated hamsters 8 to 20 weeks after the BOP treatment were summarized in Table 1. The Hs were flat or partly papillary epithelial lesions composed of columnar cells without distinct atypia (Fig. 2B) relative to normal ducts composed of small epithelia in control animals (Fig. 2A). The AHs had more papillary structures than Hs and may have partly had a cribriform pattern. These lesions had some nuclear abnormalities, for example, enlargement, clouding, and some loss of polarity; however, invasion through the basement membrane was absent (Fig. 2C). Typical ACs had a distinct tubular, cribriform, or anaplastic pattern with severely atypical columnar or cuboidal epithelia (Fig. 3A). The incidence of Hs and particularly ACs-advanced increased over time from weeks 8 to 20 after the BOP treatment, whereas AHs and ACs-early were observed at lower incidences at all time points (Table 1). All ductal lesions, except for 2 AC-advanced cases, seemed to originate in the intralobular/interlobular pancreatic ducts/ductules. The 2 cases found in the main pancreatic ducts were harboring mucinous characteristics and may have been similar to the intraductal papillary mucinous neoplasm-originated IDCs in humans.34 Therefore, these cases were excluded from counting in Table 1 and further evaluations. The macroscopically enlarged pancreatic lymph nodes found in 2 of the 8 animals at week 20 were histopathologically evaluated, and the metastasis of ACs was confirmed (Figs. 3B, C). Regarding histopathological changes other than ductal/ductular lesions, diffuse chronic pancreatitis with acinar atrophy and fatty infiltration were observed in 1 of the 9 hamsters at week 16 and 3 of the 8 hamsters at week 20. No histopathological changes were found in the pancreases of 3 control animals at week 12. Isolated pancreatic ducts included epithelial cells and surrounding connective/muscular tissue. Those in the control animals were lined with epithelial cells with scant cytoplasm and condensed small round nuclei (Fig. 1C). On the other hand, some parts of the isolated pancreatic ducts in 3 of the 5 BOP-treated hamsters had cuboidal to columnar epithelial cells, which were still monolayered or partly multilayered and were characterized by enlarged nuclei (Fig. 1D), and these were identified as AH changes.
Gene Expression Profiles of Isolated Pancreatic Ducts The expression profiles of pancreatic AHs in BOP-treated hamsters were compared with those of normal pancreatic ducts in control animals, and a total of 112 genes, including 70 upregulated and 42 down-regulated ones, were identified as being greater than 3-fold differentially expressed in the AHs of BOPtreated hamsters (Tables 2, 3). Complete hybridization data TABLE 1. Incidence and Multiplicity of Hs, AHs, and ACs in the Pancreases of Hamsters With and Without the BOP Treatment Week Lesions Hyperplasia AH AC-early
RESULTS
&
AC-advanced
No. animals
8
12
16
20
6 6 9 8 1* (17) 3 (50) 5 (56) 6 (75) 0.2 (0.4)† 0.5 (0.6) 0.6 (0.5) 1.0 (0.8) 0 0 2 (22) 1 (13) 0.2 (0.4) 0.1 (0.4) 1 (17) 3 (50) 0 0 0.2 (0.4) 0.7 (0.8) 1 (17) 2 (33) 5 (56) 7 (88) 0.3 (0.8) 0.3 (0.5) 0.6 (0.5) 2.0 (1.3)
*
Number of animals with lesions (%). Mean number and SD of each lesion per animal.
†
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pancreas
&
Volume 44, Number 1, January 2015
Biphasic Alterations in MUC1 in Hamster Pancreas
FIGURE 2. Histopathology and immunohistochemistry of normal pancreatic ducts in a control hamster (A) and hyperplastic (B) and atypically hyperplastic (C) ducts in BOP-treated hamsters; hematoxylin and eosin staining, scale bars of 50 Km. Immunohistochemistry for MUC1 cytoplasmic domain (DYF), slug (GYI), vimentin (JYL), and E-cadherin (MYO) of each serial section of (AYC). Note the clear staining for the MUC1-CD in the apical surfaces of hyperplastic and atypically hyperplastic ducts (E, F). Nuclear/cytoplasmic reactions to antislug antibodies were sometimes observed not only in hyperplastic and atypically hyperplastic ducts (H, I) but also in normal pancreatic ducts and islet cells (G). * 2014 Lippincott Williams & Wilkins
www.pancreasjournal.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
79
Pancreas
Kitahashi et al
&
Volume 44, Number 1, January 2015
FIGURE 3. Histopathology and immunohistochemistry of AC in a BOP-treated hamster (A) and pancreatic lymph node with metastatic lesions in another BOP-treated hamster (B), scale bars of 200 Km. C, A higher magnification of the box in B, scale bar of 50 Km. Immunohistochemistry for the MUC1-CD (DYF), slug (GYI), vimentin (JYL), and E-cadherin (MYO) of each serial section of A or C, scale bars of 50 Km. E, Clear cytoplasmic expression of the MUC1-CD in the invasive front of AC. F, Nuclear accumulation of the MUC1-CD in metastatic AC cells (arrows). K, Note the cytoplasmic expression of vimentin in the invasion front of AC (arrows).
80
www.pancreasjournal.com
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pancreas
&
Volume 44, Number 1, January 2015
Biphasic Alterations in MUC1 in Hamster Pancreas
TABLE 2. Up-Regulated Genes Identified as Being Greater Than 3-Fold Differently Expressed in AHs Relative to Normal Pancreatic Ducts Fold Change
Accession
Description
P
21.58 16.79 12.60 12.36 11.30 10.71 10.30 10.28 9.95 9.71 9.14
DQ369045.1 AY535011.1 AF221841.1 AF060820.1 AJ431372.2 M13933.1 DQ367730.1 DQ207608.1 U36918.1 D26473.1 AY649405.1
1.77E-09 9.93E-11 1.89E-08 8.93E-09 7.03E-08 9.99E-08 2.17E-08 4.11E-11 3.45E-07 1.31E-08 3.44E-09
8.99
DQ403102.1
8.71 8.52 8.10 7.86 7.48 7.29
L40382.1 DQ402953.1 D15058.1 U41499.1 AY508121.1 DQ403002.1
7.12 6.79 6.77 6.61 6.58 6.52 6.34 6.30 6.18 5.99 5.91
U41737.1 AJ312092.1 AF487552.1 X53139.1 AF004814.1 AF364317.1 DQ092501.1 AY550277.1 AF260255.1 J00059.1 DQ202704.1
5.90 5.70 5.66 5.62 5.50 5.29 5.18 4.88 4.88 4.70 4.58 4.49 4.33 4.23 4.14 4.10 4.07 3.95 3.84 3.83 3.79 3.78
AY057063.1 M26640.1 AY100688.1 AF532110.1 M88136.1 X04850.1 U58672.1 M10230.1 X87227.1 AF294427.1 M34665.1 AY553637.1 AJ582075.1 U08342.1 M17169.1 DQ473436.1 DQ202702.1 L40381.1 U73375.1 AB194396.1 X13175.1 DQ403027.1
Mucin 1 mRNA, partial CDS Putative mitochondrial ATP synthase mRNA, partial CDS Thioredoxin peroxidase II mRNA, complete CDS Aldehyde reductase mRNA, partial CDS mRNA for CAP1 protein Ribosomal protein RPS17 mRNA, complete CDS Complement protein C1qBP mRNA, partial CDS Beta-actin mRNA, partial CDS Mucin (MUC1) mRNA, complete CDS RPS4 mRNA for ribosomal protein S4, complete CDS Mitochondrial NADH-ubiquinone oxidoreductase ESSS subunit precursor, mRNA, complete CDS Mitochondrial ATP synthase, H + transporting F1 complex beta subunit (ATP5B) mRNA, partial CDS GST pi enzyme mRNA, complete CDS Mitochondrial malate dehydrogenase 2, NAD mRNA, partial CDS mRNA for DAD-1, complete CDS Arsenite-resistance protein mRNA, complete CDS Clone 1(10) adrenal gland 15-kDa selenoprotein-like mRNA sequence Succinate dehydrogenase complex subunit B mRNA, partial CDS Pancreatic beta cell growth factor (INGAP) mRNA, complete CDS mRNA for beta actin (actb gene) Nucleotide-binding protein G(s) alpha subunit (GNAS) mRNA, complete CDS mRNA for nucleotide-binding coupling protein alpha subunit Ubiquitin conjugating enzyme mRNA, complete CDS Protein disulfide-isomerase mRNA, complete CDS MHC class II antigen alpha chain mRNA, partial CDS Adrenonovelin mRNA, complete CDS Tissue inhibitor of matrix metalloproteinase-2 mRNA, partial CDS Preproglucagon mRNA, complete CDS Guanine nucleotide binding protein, alpha stimulating (GNAS) mRNA, complete CDS ERP57 protein mRNA, complete CDS Sulfated glycoprotein 2 mRNA, 3’ end Calreticulin mRNA, complete CDS Peroxiredoxin 2 mRNA, complete CDS Seryl-tRNA synthetase mRNA, partial CDS mRNA fragment for glucose regulated protein (grp 94) Ornithine decarboxylase antizyme mRNA, complete CDS >-tubulin I mRNA, 3’ flank mRNA for vimentin Proliferating cell nuclear antigen mRNA, complete CDS T-complex protein 1 mRNA, complete CDS Inhibitor of DNA binding 2 mRNA, 3’ UTR Partial mRNA for Bcl-2 associated protein (bax gene) Beta-tubulin isotype I mRNA, complete CDS Glucose-regulated protein GRP78 mRNA, complete CDS Immunoglobulin heavy chain mRNA, partial CDS Guanine nucleotide binding protein alpha inhibiting 2 mRNA, complete CDS GST pi enzyme mRNA, complete CDS Pancreas cancer-associated protein 4 mRNA, complete CDS mRNA for ovca1, partial CDS KG4 mRNA related to cellular proliferation Ribosomal protein L18 mRNA, partial CDS
1.14E-10 8.1E-10 3.61E-09 2.22E-08 2.81E-06 3.15E-07 5.75E-07 1.26E-06 6.12E-08 2.44E-08 1.16E-05 5.68E-08 3.64E-08 1.79E-07 1.89E-08 8.13E-08 1.32E-07 1.51E-05 1.41E-10 1.17E-08 1.85E-06 1.07E-06 6.46E-08 8.01E-08 3.13E-08 1.60E-06 2.36E-07 3.78E-07 2.17E-08 2.42E-08 5.72E-05 1.01E-07 1.68E-05 1.24E-06 9.12E-06 6.84E-08 2.03E-08 5.97E-05 5.97E-10 6.94E-08
(continued on next page) * 2014 Lippincott Williams & Wilkins
www.pancreasjournal.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
81
Pancreas
Kitahashi et al
&
Volume 44, Number 1, January 2015
TABLE 2. (Continued) Fold Change
Accession
3.68 3.66 3.59 3.54 3.41 3.40 3.35 3.33 3.31 3.29 3.24 3.23 3.20 3.19 3.15 3.15 3.14 3.12 3.09
AF046214.1 M11241.1 J02983.1 U50395.1 L18986.1 M99692.1 AB014876.2 X57112.1 X62678.1 X77631.1 DQ017755.1 AF320819.1 D31814.1 EF210569.1 AF281149.1 AY688356.1 AF020767.1 X57029.1 AY004869.1
are available as part of the supplementary information (SDC 1, at http://links.lww.com/MPA/A312). Some up-regulated genes related to cell differentiation, proliferation, or motility were identified based on the potential molecular function of the proteins encoded by the differentially expressed genes, for example, mucin 1 (MUC1, 21.58- and 9.95-fold), tissue inhibitor of matrix metalloproteinase-2 (6.18-fold), vimentin (4.88-fold), and proliferating cell nuclear antigen (4.70-fold). Genes encoding redox proteins, such as thioredoxin peroxidase II (12.60-fold) and peroxiredoxin 2 (5.62-fold), and metabolizing enzymes, such as aldehyde reductase (12.36-fold) and mitochondrial malate dehydrogenase 2 (8.52-fold), were also up-regulated in AHs. On the other hand, genes encoding cell adhesion proteins such as connexin 45 (0.17-fold), gap junction membrane channel protein alpha 1 (0.27-fold), and connexin 47 (0.31-fold) were found to be down-regulated. However, normal expression levels and the significance of the down-regulation of these genes in the pancreatic duct remain unclear.
Gene Expression and Immunohistochemical Analysis of MUC1 in Normal and Precancerous and Cancerous Pancreatic Tissue Among the up-regulated and down-regulated genes identified in the microarray analysis, expression of the MUC1 gene in AHs was markedly different from that in normal pancreatic ducts. In addition, MUC1 was previously shown to be expressed in PanIN lesions together with IDCs in humans using immunohistochemistry.34 To validate the microarray results obtained, qRT-PCR and immunohistochemical analysis on the MUC1 gene and protein expression, respectively, were performed using the pancreatic tissues of control and BOP-treated hamsters and AC tissues from BOP-treated hamsters. Gene expression levels in the normal pancreases of control animals and normal-like pancreases of BOP-treated animals were similar, whereas those in BOP-induced AC tissues were slightly higher (Fig. 4A). Immunohistochemical analysis revealed that the acinar cell apices were positive for MUC1-CD, whereas the intercalated ductal epithelia were not in both control and BOP-treated
82
www.pancreasjournal.com
P
Description TGF-A mRNA, partial CDS Chinese hamster ovary ribosomal protein S14 mRNA, complete CDS Insulinoma rig mRNA, encoding a putative DNA-binding protein Wild-type tumor suppressor P53 (p53) mRNA, complete CDS Lysosomal membrane glycoprotein A mRNA, complete CDS Iron-binding protein (ferritin) mRNA mRNA for ribosomal protein L13, complete CDS mRNA for beta 2-microglobulin mRNA for P5 protein IFAP300 mRNA Placental cathepsin L mRNA, partial cds Basigin precursor, mRNA, partial cds mRNA for polyprotein precursor, complete CDS Regenerating protein 2 mRNA, complete CDS Spermidine/spermine N1-acetyltransferase mRNA, complete CDS Cu/Zu superoxide dismutase 1-like mRNA, partial sequence Skeletal muscle neutral protease mekratin mRNA, complete CDS mRNA for >-globin chain Cyclophilin mRNA, partial CDS
6.09E-06 4.77E-07 1.80E-07 0.000150 3.91E-06 4.31E-07 4.90E-07 1.26E-05 1.12E-05 3.06E-06 0.000123 6.98E-05 4.95E-06 0.000796 0.000345 8.87E-08 1.71E-08 3.23E-05 6.53E-07
hamsters (Fig. 2D). The main and branch (intralobular and interlobular) pancreatic ducts were negative for MUC1-CD but were occasionally slightly positive at their apical surface (Fig. 2D). In contrast, the apical surfaces of the epithelia of all Hs, AHs, ACs-early, and AC-advanced areas with distinct tubular patterns were positive for MUC1-CD, being observed in all BOP-treated hamster tissues (Figs. 2E, F; 3D), and no significant change was observed in staining intensities between the different grades of lesions. In contrast, surface staining for MUC1-CD was weaker in the invasive front of ACs-advanced, and cytoplasmic staining was characteristic (Fig. 3E). In addition, the nuclear accumulation of MUC1-CD was clearly observed in 1 lymph node metastasis case (Fig. 3F).
Gene Expression and Immunohistochemical Analysis of EMT-Related Proteins in Normal and Precancerous and Cancerous Pancreatic Tissue MUC1-CD was previously shown to play a direct role in MUC1 by initiating EMT.30 Therefore, the coexpression of EMTrelated proteins was analyzed to clarify the role of MUC1 in the progression processes of pancreatic carcinogenesis in vivo. The gene expression levels of slug in the normal pancreases of control animals were negligible, whereas those in the normal-like pancreases of BOP-treated animals were slightly higher and were significantly increased in the AC tissues of BOP-treated animals (Fig. 4B). Immunohistochemical analysis revealed that nuclear slug-positive ratios were slightly higher in Hs, AHs, ACsearly, and AC-advanced areas with distinct tubular patterns and higher in the invasive front of ACs-advanced than those in the normal(-like) pancreatic ducts of control and BOP-treated animals (Fig. 5). The increased nuclear staining of slug seemed to indicate its coexpression with cytoplasmic MUC1-CD (Fig. 3H), but not with apical MUC1-CD (Figs. 2GYI, 3G). Mesenchymal tissues/ cells, including fibroblasts, vascular endothelia, and immunocytes, were positive for vimentin, whereas the epithelia of normal pancreatic ducts, Hs, AHs, or AC-advanced areas with distinct tubular patterns were not (Figs. 2JYL, 3J). Carcinoma cells in the invasion front were positive for vimentin in 22 of the 24 AC-advanced * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pancreas
&
Volume 44, Number 1, January 2015
Biphasic Alterations in MUC1 in Hamster Pancreas
TABLE 3. Down-Regulated Genes Identified as Being Greater Than 3-Fold Differently Expressed in AHs Relative to Normal Pancreatic Ducts Fold Change 0.03 0.06 0.07 0.09 0.11 0.11 0.11 0.12 0.12 0.14 0.15 0.17 0.17 0.18 0.18 0.19 0.19 0.20 0.20 0.21 0.21 0.21 0.21 0.22 0.22 0.24 0.24 0.24 0.25 0.25 0.25 0.25 0.27 0.29 0.29 0.29 0.30 0.31 0.31 0.32 0.33 0.33
Accession
Description
P
AF052847.1 AJ746220.1 M27242.1 AF052838.1 L38710.1 AF052848.1 AY221595.1 AF052842.1 AF052841.1 Z66540.1 AF052846.1 AF052850.1 AF508038.1 AF052853.1 M27243.1 AF052844.1 AF052845.1 AF052849.1 AF052851.1 AY329083.1 AF052852.1 AB028235.1 AF055671.1 U71279.1 DQ837221.1 X64179.1 D50578.1 D28566.1 AY788096.1 DQ237898.1 DQ237889.1 DQ015678.1 AY049064.1 U65889.1 AF492475.1 X63022.1 AY162470.1 AF508037.1 AF052843.1 DQ237897.1 CF355260.1 D89286.1
Clone e23 retroviral-like pol protein (pol) mRNA, partial CDS Partial mRNA for eIF4GI protein Serum amyloid A (SAA1) mRNA, complete CDS Clone e26 retroviral-like pol protein (pol) mRNA, partial CDS 3-beta-hydroxysteroid dehydrogenase/5-4-eneisomerase (HSD3B) mRNA, complete CDS Clone e22 retroviral-like pol protein (pol) mRNA, partial CDS Baculoviral IAP repeat-containing protein 6 mRNA, partial CDS Clone ut16 retroviral-like pol protein (pol) mRNA, partial CDS Clone e8 retroviral-like pol protein (pol) mRNA, partial CDS mRNA (440 bp) related to sexual dimorphism Clone s25 retroviral-like pol protein (pol) mRNA, partial CDS Clone s1 retroviral-like pol protein (pol) mRNA, partial CDS Connexin 45 mRNA, partial CDS Clone ut10 retroviral-like pol protein (pol) mRNA, partial CDS Serum amyloid (SAA2) mRNA, complete CDS Clone s12 retroviral-like pol protein (pol) mRNA, partial CDS Clone t12 retroviral-like pol protein (pol) mRNA, partial CDS Clone e18 retroviral-like pol protein (pol) mRNA, partial CDS Clone s19 retroviral-like pol protein (pol) mRNA, partial CDS Per1-interacting protein mRNA, partial CDS Clone t3 retroviral-like pol protein (pol) mRNA, partial CDS mRNA for IL1 alpha, partial CDS Collagenase 2 (MMP-8) mRNA, partial CDS Adrenocorticotropin receptor (MC2) mRNA, complete CDS Secretoglobin precursor, mRNA, complete CDS mRNA for betalike y-globin gene mRNA for carboxylesterase, complete CDS mRNA for carboxylesterase, complete CDS Mitochondrial-associated cysteine-rich protein mRNA, complete CDS Progesterone receptor mRNA, partial CDS BMPR-IB mRNA, partial CDS 5-hydroxytryptamine receptor 2A mRNA, complete CDS Gap junction membrane channel protein alpha 1 mRNA, partial CDS Bone sialoprotein mRNA, complete CDS LIF receptor mRNA, partial CDS mRNA for cytochrome P450IIC Parotid secretory protein mRNA, complete CDS Connexin 47 mRNA, partial CDS Clone t2 retroviral-like pol protein (pol) mRNA, partial CDS N-cadherin mRNA, partial CDS cDNA similar to the androgen receptor, mRNA sequence mRNA for interalpha-trypsin inhibitor heavy chain 2, complete CDS
2.12E-07 2.39E-11 1.70E-09 7.63E-10 1.16E-07 3.82E-11 1.47E-08 6.71E-09 3.70E-08 1.18E-08 6.21E-08 2.43E-11 3.01E-07 1.59E-11 8.76E-06 1.10E-06 1.91E-06 1.74E-12 1.44E-12 1.05E-08 5.82E-12 1.51E-08 2.38E-08 4.66E-06 4.64E-09 5.85E-06 2.57E-07 7.52E-07 2.11E-06 4.60E-06 1.33E-05 0.00043 4.64E-05 1.03E-06 4.47E-06 1.09E-05 1.99E-06 3.68E-06 0.00292 0.00205 1.91E-05 5.76E-08
cases; however, the staining intensities and expression ratios in each lesion were weak and low, respectively (Fig. 3K). Epithelial tissues/cells, including acinar cell and main and branch pancreatic ducts, showed membranous positivities for E-cadherin (Fig. 2M), and its subcellular expression patterns were similar; however, the staining intensities seemed to be higher in Hs, AHs, or ACs (Figs. 2N, O; 3M) than in normal pancreatic ducts. The coexpression of cytoplasmic MUC1, nuclear slug, and cytoplasmic vimentin and weak staining intensities of E-cadherin (Figs. 3N, O) in carcinoma cells in the invasion front were generally confirmed, as were those of nuclear MUC1 and nuclear slug (Fig. 3I), whereas vimentin expression did not seem to be correlated with nuclear MUC1 * 2014 Lippincott Williams & Wilkins
expression (Fig. 3L) in the lymph node metastatic case. The gene expression levels of snail1, twist, E-cadherin, cdh2, vimentin, and TGF-A1 were similar between the normal pancreases of control animals and normal-like pancreases of BOP-treated animals, whereas those in the AC tissues of BOP-treated animals were increased, except for E-cadherin (Figs. 4CYH).
DISCUSSION Among the up-regulated genes in the AHs of pancreatic ducts analyzed using a microarray technique in the hamster carcinogenesis model, the expression of the MUC1 gene was markedly different from that in normal pancreatic ducts. To www.pancreasjournal.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
83
Pancreas
Kitahashi et al
&
Volume 44, Number 1, January 2015
FIGURE 4. Gene expression levels of (A) MUC1, (B) slug, (C) snail1, (D) twist, (E) E-cadherin, (F) cdh2, (G) vimentin, and (H) TGF-A1 in the normal pancreases of control animals and normal-like pancreases and AC tissues of BOP-treated animals (sample numbers were 4, 8, and 4, respectively). *P G 0.01 and **P G 0.05 vs normal control; Student t test.
84
www.pancreasjournal.com
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pancreas
&
Volume 44, Number 1, January 2015
FIGURE 5. Immunohistochemical slug-positive ratios in the nuclei of each lesion from control and BOP-treated animals. The numbers of animals counted were 3, 5, 15, 3, 5, 15, and 15 for normal pancreatic ducts in control animals, normal-like ducts in BOP(+) animals, Hs, AHs, ACs-early, AC areas with distinct tubular patterns, and the invasive front of ACs, respectively. **P G 0.01 vs all other lesions except for AC-early (P G 0.05); Tukey test.
validate the microarray results obtained, qRT-PCR for MUC1 mRNA levels was performed using the pancreatic tissues of control and BOP-treated hamsters and carcinoma tissues from BOP-treated hamsters. The results revealed similar expression levels between the normal pancreases of control animals and normal-like pancreases of BOP-treated animals. Immunohistochemistry for MUC1 revealed that its expression was clearly detected in the apices of acinar cells, which constitute a marked portion of normal(-like) pancreatic tissue, and BOP-induced alterations were absent in acinar cells, the major source of MUC1 expression. This was considered to be the reason for the absence of a significant difference in MUC1 mRNA levels in normal(-like) pancreatic tissue between control and BOPtreated animals. MUC1 has been detected in the cell apices of the centroacinar cells, intercalated ducts, and intralobular ducts of the normal pancreas in humans by every antibody for the different glycoforms of MUC1, for example, NCL-MUC-1CORE (clone Ma552, reacts with the core peptide of the tandem repeat domain),35 DF3 (reacts with the core peptide of the tandem repeat domain),36 MY.1E12 (reacts with sialylated MUC1 mucin), and HMFG-1 (reacts with fully glycosylated MUC1 mucin).34 The results of the present study showing acinar cell positivity by anti-MUC1-CD antibodies in the hamster pancreas were consistent with the abovementioned results in humans. The qRT-PCR revealed that MUC1 mRNA expression levels were slightly higher in hamster carcinoma tissues than in normal(-like) pancreatic tissues, which demonstrated that MUC1 gene expression levels in carcinoma cells were higher than those in acinar cells and the unique function of this molecule in pancreatic ductal carcinogenesis. The immunohistochemical expression of MUC1-CD was negative in normal(-like) interlobular/intralobular pancreatic ducts and positive in the apical surfaces of not only AHs but also Hs without atypia, and no significant change was observed in staining intensities between the different grades of lesions. Previous studies demonstrated that MUC1 was expressed in the cell apices of intralobular ducts and intercalated ducts in the normal human pancreas, and the increase in PanINs was correlated with their grades by every antibody for the different * 2014 Lippincott Williams & Wilkins
Biphasic Alterations in MUC1 in Hamster Pancreas
glycoforms of MUC1.34,37 The reason for the rapid induction of MUC1 in the hyperplastic ductal cell apices of the present hamster pancreatic carcinogenesis model remains unclear. In the histopathological evaluation, Hs without atypia and advanced ACs increased over time throughout the experiment period, whereas AHs and early carcinomas with minimal invasive growth were observed at lower incidences at all time points. This may indicate that a rapid progression process from Hs to carcinomas was present in the hamster carcinogenesis model used, and hyperplastic ductal epithelia may harbor some early malignant characteristics including the distinct up-regulation of MUC1. Another possibility may be related to the anti-MUC1CD antibody-specific immunoreactivities observed in the present study. Further immunohistochemical studies using anti-MUC1CD antibodies have yet to be conducted in human PanINs and IDC samples. One of the most interesting aspects of the present study was the cytoplasmic positivity of MUC1-CD characterized in the invasive front of ACs. In addition, the nuclear accumulation of MUC1-CD was clearly observed in 1 lymph node metastasis case. A similar phenomenon was previously reported in relation to the aggressive behavior of human pancreatic IDCs using MUC1/DF3 antibodies (reacts with the core peptide of the tandem repeat domain), which were expressed not only at the cell apex but also in the cytoplasm and lateral membrane.27 MUC1 is encoded by a single transcript and, after transcription, undergoes autocleavage into 2 subunits to form a stable transmembrane heterodimer.38 The MUC1 N-terminal subunit is a mucin component that has been extensively glycosylated, of which the core peptide contains tandem repeats. MUC1 Cterminal subunit is recognized as an oncogene, has 7 tyrosine residues, and was first reported to interact with A-catenin.38 MUC1 has recently been shown to regulate the nuclear localization and function of the epidermal growth factor receptor,39 or MUC1-CD is translocated to the nucleus, in which it promotes the transcription of metastatic genes associated with EMT in human pancreatic carcinoma cells.40 The present study is the first to demonstrate the in vivo cytoplasmic/nuclear positivity of MUC1-CD characterized in the invasive front or the metastasis of ACs in contrast to its apical positivity in carcinoma areas with distinct tubular patterns. In addition, the coexpression of EMT-related proteins, such as slug and vimentin, with cytoplasmic MUC1 was also shown. Despite the distinct coexpression of cytoplasmic/nuclear MUC1 and slug/vimentin in the invasive front or the metastasis of ACs, nuclear slug positivity was slightly more frequent in Hs, AHs, and AC areas with distinct tubular patterns, in which MUC1 overexpression was observed in their epithelial apices only. In addition, slug mRNA levels were slightly higher in the normal-like pancreases of BOP-treated hamsters than in the normal pancreases of controls. Therefore, further studies are needed to clarify the overexpression of MUC1 before its translocalization to the cytoplasm/ nuclei, which could be a preparatory stage for EMT in the epithelial cells of preneoplastic lesions, for example, Hs and AHs. In addition, the key signaling molecules contributing EMT via MUC1 intracellular translocalization should be investigated by a comparative gene expression analysis using DNA microarrays between AC areas with distinct tubular patterns and those in the invasive front, in combination with MUC1 gene knockdown or knockout studies in vitro and/or in vivo. In any case, the purpose of the present immunohistochemical analysis was to demonstrate a biphasic alteration in the expression level and subcellular localization of MUC1 in pancreatic ductal carcinogenesis using an animal model, and the results obtained indicate that both an elevation in MUC1 expression levels and the www.pancreasjournal.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
85
Pancreas
Kitahashi et al
cleavage/translocation of MUC1-CD to the cytoplasm or nuclei were important targets for pancreatic cancer prevention and chemotherapy. Further immunohistochemical studies using anti-MUC1-CD antibodies should be conducted in human IDC samples. ACKNOWLEDGMENT We thank Ms Satomi Kohno for her expert technical assistance and the National Cancer Center Research Core Facility for some of the analyses in this study. The Core Facility was supported by the National Cancer Center Research and Development Fund (23-A-7). REFERENCES 1. Matsuda A, Matsuda T, Shibata A, et al. Cancer incidence and incidence rates in Japan in 2007: a study of 21 population-based cancer registries for the Monitoring of Cancer Incidence in Japan (MCIJ) project. Jpn J Clin Oncol. 2013;43:328Y336. 2. Krejs GJ. Pancreatic cancer: epidemiology and risk factors. Dig Dis. 2010;28:355Y358. 3. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J Clin. 2013;63:11Y30. 4. Frampton AE, Krell J, Giovannetti E, et al. Defining a prognostic molecular profile for ductal adenocarcinoma of the pancreas highlights known key signaling pathways. Expert Rev Anticancer Ther. 2012;12:1275Y1278. 5. Cutts RJ, Gadaleta E, Hahn SA, et al. The Pancreatic Expression database: 2011 update. Nucleic Acids Res. 2011;39:D1023YD1028. 6. Kim HN, Choi DW, Lee KT, et al. Gene expression profiling in lymph node-positive and lymph node-negative pancreatic cancer. Pancreas. 2007;34:325Y334. 7. Crnogorac-Jurcevic T, Gangeswaran R, Bhakta V, et al. Proteomic analysis of chronic pancreatitis and pancreatic adenocarcinoma. Gastroenterology. 2005;129:1454Y1463. 8. Logsdon CD, Simeone DM, Binkley C, et al. Molecular profiling of pancreatic adenocarcinoma and chronic pancreatitis identifies multiple genes differentially regulated in pancreatic cancer. Cancer Res. 2003; 63:2649Y2657. 9. Feldmann G, Beaty R, Hruban RH, et al. Molecular genetics of pancreatic intraepithelial neoplasia. J Hepatobiliary Pancreat Surg. 2007;14:224Y232. 10. Hruban RH, Adsay NV, Albores-Saavedra J, et al. Pancreatic intraepithelial neoplasia: a new nomenclature and classification system for pancreatic duct lesions. Am J Surg Pathol. 2001;25:579Y586. 11. Bardeesy N, DePinho RA. Pancreatic cancer biology and genetics. Nat Rev Cancer. 2002;2:897Y909. 12. Buchholz M, Braun M, Heidenblut A, et al. Transcriptome analysis of microdissected pancreatic intraepithelial neoplastic lesions. Oncogene. 2005;24:6626Y6636. 13. Pan S, Chen R, Reimel BA, et al. Quantitative proteomics investigation of pancreatic intraepithelial neoplasia. Electrophoresis. 2009;30:1132Y1144. 14. Hingorani SR, Petricoin EF, Maitra A, et al. Preinvasive and invasive ductal pancreatic cancer and its early detection in the mouse. Cancer Cell. 2003;4:437Y450. 15. Hruban RH, Adsay NV, Albores-Saavedra J, et al. Pathology of genetically engineered mouse models of pancreatic exocrine cancer: consensus report and recommendations. Cancer Res. 2006;66:95Y106. 16. Morris JPT, Cano DA, Sekine S, et al. Beta-catenin blocks Kras-dependent reprogramming of acini into pancreatic cancer precursor lesions in mice. J Clin Invest. 2010;120:508Y520. 17. Corcoran RB, Contino G, Deshpande V, et al. STAT3 plays a critical role in KRAS-induced pancreatic tumorigenesis. Cancer Res. 2011;71:5020Y5029. 18. Fujii H, Egami H, Chaney W, et al. Pancreatic ductal adenocarcinomas induced in Syrian hamsters by N-nitrosobis(2-oxopropyl)amine contain a c-Ki-ras oncogene with a point-mutated codon 12. Mol Carcinog. 1990;3:296Y301.
86
www.pancreasjournal.com
&
Volume 44, Number 1, January 2015
19. Muscarella P, Knobloch TJ, Ulrich AB, et al. Identification and sequencing of the Syrian Golden hamster (Mesocricetus auratus) p16(INK4a) and p15(INK4b) cDNAs and their homozygous gene deletion in cheek pouch and pancreatic tumor cells. Gene. 2001;278:235Y243. 20. Li J, Qin D, Knobloch TJ, et al. Expression and characterization of Syrian golden hamster p16, a homologue of human tumor suppressor p16 INK4A. Biochem Biophys Res Commun. 2003;304:241Y247. 21. Li J, Weghorst CM, Tsutsumi M, et al. Frequent p16INK4A/CDKN2A alterations in chemically induced Syrian golden hamster pancreatic tumors. Carcinogenesis. 2004;25:263Y268. 22. Birt DF, Patil K, Pour PM. Comparative studies on the effects of semipurified and commercial diet on longevity and spontaneous and induced lesions in the Syrian golden hamster. Nutr Cancer. 1985;7:167Y177. 23. Takahashi M, Pour P. Spontaneous alterations in the pancreas of the aging Syrian golden hamster. J Natl Cancer Inst. 1978;60:355Y364. 24. Kitahashi T, Yoshimoto M, Imai T. Novel immunohistochemical marker, integrin alpha(V)beta(3), for BOP-induced early lesions in hamster pancreatic ductal carcinogenesis. Oncol Lett. 2011;2:229Y234. 25. Konishi Y, Tsutsumi M, Tsujiuchi T. Mechanistic analysis of pancreatic ductal carcinogenesis in hamsters. Pancreas. 1998;16:300Y306. 26. Terada T, Ohta T, Sasaki M, et al. Expression of MUC apomucins in normal pancreas and pancreatic tumours. J Pathol. 1996;180:160Y165. 27. Yonezawa S, Goto M, Yamada N, et al. Expression profiles of MUC1, MUC2, and MUC4 mucins in human neoplasms and their relationship with biological behavior. Proteomics. 2008;8:3329Y3341. 28. Tsutsumida H, Swanson BJ, Singh PK, et al. RNA interference suppression of MUC1 reduces the growth rate and metastatic phenotype of human pancreatic cancer cells. Clin Cancer Res. 2006;12:2976Y2987. 29. Yuan Z, Liu X, Wong S, et al. MUC1 knockdown with RNA interference inhibits pancreatic cancer growth. J Surg Res. 2009;157:e39Ye46. 30. Roy LD, Sahraei M, Subramani DB, et al. MUC1 enhances invasiveness of pancreatic cancer cells by inducing epithelial to mesenchymal transition. Oncogene. 2011;30:1449Y1459. 31. Konishi Y, Mizumoto K, Kitazawa S, et al. Early ductal lesions of pancreatic carcinogenesis in animals and humans. Int J Pancreatol. 1990;7:83Y89. 32. Bolstad BM, Irizarry RA, Astrand M, et al. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics. 2003;19:185Y193. 33. Hori M, Kitahashi T, Imai T, et al. Enhancement of carcinogenesis and fatty infiltration in the pancreas in N-nitrosobis(2-oxopropyl)amine-treated hamsters by high-fat diet. Pancreas. 2011;40:1234Y1240. 34. Nagata K, Horinouchi M, Saitou M, et al. Mucin expression profile in pancreatic cancer and the precursor lesions. J Hepatobiliary Pancreat Surg. 2007;14:243Y254. 35. Baeckstrom D, Nilsson O, Price MR, et al. Discrimination of MUC1 mucins from other sialyl-Le(a)-carrying glycoproteins produced by colon carcinoma cells using a novel monoclonal antibody. Cancer Res. 1993;53:755Y761. 36. Siddiqui J, Abe M, Hayes D, et al. Isolation and sequencing of a cDNA coding for the human DF3 breast carcinoma-associated antigen. Proc Natl Acad Sci U S A. 1988;85:2320Y2323. 37. Kigure S. Immunohistochemical study of the association between the progression of pancreatic ductal lesions and the expression of MUC1, MUC2, MUC5AC, and E-cadherin. Rinsho Byori. 2006;54:447Y452. 38. Kufe DW. Functional targeting of the MUC1 oncogene in human cancers. Cancer Biol Ther. 2009;8:1197Y1203. 39. Bitler BG, Goverdhan A, Schroeder JA. MUC1 regulates nuclear localization and function of the epidermal growth factor receptor. J Cell Sci. 2010;123:1716Y1723. 40. Lau SK, Shields DJ, Murphy EA, et al. EGFR-mediated carcinoma cell metastasis mediated by integrin alphavbeta5 depends on activation of c-Src and cleavage of MUC1. PLoS One. 2012;7:e36753.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Biphasic Alterations in Expression and Subcellular Localization of MUC1 in Pancreatic Ductal Carcinogenesis in Syrian Hamsters Tsukasa Kitahashi, PhD, Mami Takahashi, PhD, and Toshio Imai, DVM, PhD
Objectives: The aim of the present study was to characterize molecular targets for the prevention/diagnosis of pancreatic cancer using a chemically induced hamster pancreatic carcinogenesis model, in which background injuries to the parenchyma, for example, chronic pancreatitis or acinar atrophy, are limited. Methods: Gene expression profiles in atypical hyperplasias were first investigated using a microarray technique. Immunohistochemical analyses of early lesions and invasive ductal carcinoma (IDC) were then conducted for MUC1, of which mRNA levels were prominent among the up-regulated genes, in contrast with the coexpression of epithelial-mesenchymal transition (EMT)-related proteins. Results: Immunohistochemistry for MUC1 cytoplasmic domain (MUC1CD), which was not detected in normal-like pancreatic ducts, was positive in the apical surfaces of the epithelia of hyperplasias with and without atypia and IDC areas with distinct tubular patterns. In contrast, cytoplasmic/ nuclear positivity for MUC1-CD was observed in the invasive front of IDCs. The coexpression of EMT-related proteins, such as slug and vimentin, with cytoplasmic/nuclear MUC1-CD was also detected. Conclusions: Alterations in the expression and subcellular localization of MUC1 represent a biphasic phenomenon, and the latter may be associated with EMT in pancreatic carcinogenesis in hamsters, which indicates that changes in MUC1 are important targets for pancreatic cancer prevention and chemotherapy. Key Words: pancreatic ductal adenocarcinoma, MUC1, cytoplasmic domain, epithelial-mesenchymal transition (Pancreas 2015;44: 76Y86)
P
ancreatic cancer is the fourth to sixth cause of cancer death, with a high mortality ranking in developed countries.1Y3 Invasive ductal carcinomas (IDCs) are the most common histological type in human pancreatic cancer cases. The IDC gene/protein expression profiles have been investigated with the aim of improving diagnostic and prognostic techniques and developing therapies.4Y8 Genetic studies have also identified the signature molecular profile of IDCs, consisting of mutations in the KRAS and TP53 genes and loss/inactivation of CDKN2A/p16 and SMAD4. Such genetic alterations seem to occur in a temporal sequence in putative progressive From the Central Animal Division, National Cancer Center Research Institute, Tokyo, Japan. Received for publication September 6, 2013; accepted April 9, 2014. Reprints: Toshio Imai, DVM, PhD, Central Animal Division, National Cancer Center Research Institute, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan (evA3 was confirmed to identify AHs.24
Total RNA Extraction and Gene Expression Array Total RNA was extracted from each isolated pancreatic duct of 3 BOP-treated hamsters and 2 control hamsters using ISOGEN (Nippon Gene, Toyama, Japan). After RNA purification, a cocktail of total RNA at 239 and 400 ng from 3 AH lesions and 2 normal ducts, respectively, were applied to the Amino Allyl MessageAmp II aRNA Amplification Kit (Ambion) for cDNA synthesis. The synthesized cDNAs were labeled with Cy3 fluorescein dye according to NimbleGen’s labeling method (Roche NimbleGen, Madison, WI).
FIGURE 1. Images of pancreatic ducts around the head portion of the pancreas. Pancreatic ducts were stained with RNA later containing Reactive Black 5 solution (A), an isolated duct (B), scale bar of 200 Km. Microscopic features of a normal pancreatic duct from a control hamster (C) and atypically hyperplastic duct from a BOP-treated hamster (D) stained with hematoxylin and eosin, scale bar of 100 Km. * 2014 Lippincott Williams & Wilkins
www.pancreasjournal.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
77
Pancreas
Kitahashi et al
A total of 2 Kg of Cy3 labeled cDNAs were subjected to the following gene expression analysis on the NimbleGen expression array platform. The NimbleGen hamster gene expression 385 K Microarray (Roche NimbleGen) representing 1699 genes (70 of the 1699 genes were duplicated) in the form of optimized 60-mer oligonucleotides were spotted onto the microarray slide. Array data were normalized by the robust multichip average algorithm 32 and analyzed using Arraystar Software version 11.1.0 build 49 (DNASTAR, Madison, WI). Differential expression was defined as an at least 3-fold change versus the respective controls, removing expressed sequence tags, and the Student t test was applied.
qRT-PCR Analysis Aliquots of total RNA were extracted from the normal pancreatic tissue of control animals and normal-like pancreatic tissue, and the histopathologically confirmed IDCs of BOP-treated animals were subjected to reverse transcription reactions. Each mRNA transcript was measured using a previously described qRT-PCR method.33 The sequences of the primers used were as follows: MUC1 (forward, 5’GACACCTACCTATGAG-3’; reverse, 5’-CTACAAGTTGGCAG AAGTGG-3’), slug (forward, 5’-ACTGTGACAAGGAATATGTG-3’; reverse, 5’-GACATTCTGGAGAAGGTTTTG-3’), snail1 (forward, 5’-CATCCGAAGCCACACGCTG-3’; reverse, 5’-GAAGGGCTTC TCGCCAGTG-3’), twist (forward, 5’-GATGGCAAGCTGCAGCT ATG-3’; reverse, 5’-CAGCTCCAGAGTCTCTAGAC-3’), E-cadherin (forward, 5’-CTGCAGGTCTCATCATGGC-3’; reverse, 5’-ACC TGTAGACCTCGGCACTG-3’), cdh2 (N-cadherin; forward, 5’-CT GATATATGCCCAAGACAAAG-3’; reverse, 5’-GTCTCTCTTCT GCCTTTGTAG-3’), vimentin (forward, 5’-GATTCAGGAACAGC ATGTCC-3’; reverse, 5’-CATCCACTTCACAGGTGAG-3’), transforming growth factor (TGF)-A1 (forward, 5’-TTCCTGCTTCTCATG GCCACCC-3’; reverse, 5’-TGCCGCACGCAGCAGTTCTT-3’), and 40S ribosomal protein S7 (RPS7; forward, 5’-CCAGAAAATCCA AGTCCGGC-3’; reverse, 5’-AGTCCTCAAGGATGGCATCG-3’). Data were calculated as ratios to RPS7 mRNA.
Histopathology and Immunohistochemistry Pancreatic ductal proliferative lesions were classified as hyperplasias (Hs), AHs as premalignant lesions, and adenocarcinomas (ACs)-early with minimal invasive growth and ACs-advanced according to the criteria described earlier with minor modifications.24,31 Antigen retrieval for immunohistochemistry was performed in an autoclave for 10 minutes at 121-C in 10-mM citrate buffer (pH 6.0) for MUC1-CD, vimentin, and E-cadherin. The anticytoplasmic tail of human MUC1 and antihuman slug rabbit polyclonal antibodies were purchased from Abcam (Cambridge, UK), and the antihuman vimentin mouse monoclonal (clone V9) and antihuman E-cadherin mouse monoclonal (clone 36) antibodies were from DAKO Cytomation (Glostrup, Denmark) and BD Biosciences (San Jose, CA), respectively. The streptavidin-biotinperoxidase complex method (StreptABComplex/HRP, DAKO) was used to determine the expression and distribution/localization of each antigen. Sections were lightly counterstained with hematoxylin for microscopic examinations. Negative controls without primary antibody reactions were set for each antigen using serial sections. Slug-positive ratios were counted per approximately 100 to 1000 epithelial cell nuclei in normal (like) pancreatic ducts and each lesion.
Histopathological Evaluation of Pancreatic Tissue and Isolated Ducts The incidence and multiplicity of pancreatic ductal lesions including Hs, AHs, ACs-early, and ACs-advanced in the pancreases
78
www.pancreasjournal.com
Volume 44, Number 1, January 2015
of 29 BOP-treated hamsters 8 to 20 weeks after the BOP treatment were summarized in Table 1. The Hs were flat or partly papillary epithelial lesions composed of columnar cells without distinct atypia (Fig. 2B) relative to normal ducts composed of small epithelia in control animals (Fig. 2A). The AHs had more papillary structures than Hs and may have partly had a cribriform pattern. These lesions had some nuclear abnormalities, for example, enlargement, clouding, and some loss of polarity; however, invasion through the basement membrane was absent (Fig. 2C). Typical ACs had a distinct tubular, cribriform, or anaplastic pattern with severely atypical columnar or cuboidal epithelia (Fig. 3A). The incidence of Hs and particularly ACs-advanced increased over time from weeks 8 to 20 after the BOP treatment, whereas AHs and ACs-early were observed at lower incidences at all time points (Table 1). All ductal lesions, except for 2 AC-advanced cases, seemed to originate in the intralobular/interlobular pancreatic ducts/ductules. The 2 cases found in the main pancreatic ducts were harboring mucinous characteristics and may have been similar to the intraductal papillary mucinous neoplasm-originated IDCs in humans.34 Therefore, these cases were excluded from counting in Table 1 and further evaluations. The macroscopically enlarged pancreatic lymph nodes found in 2 of the 8 animals at week 20 were histopathologically evaluated, and the metastasis of ACs was confirmed (Figs. 3B, C). Regarding histopathological changes other than ductal/ductular lesions, diffuse chronic pancreatitis with acinar atrophy and fatty infiltration were observed in 1 of the 9 hamsters at week 16 and 3 of the 8 hamsters at week 20. No histopathological changes were found in the pancreases of 3 control animals at week 12. Isolated pancreatic ducts included epithelial cells and surrounding connective/muscular tissue. Those in the control animals were lined with epithelial cells with scant cytoplasm and condensed small round nuclei (Fig. 1C). On the other hand, some parts of the isolated pancreatic ducts in 3 of the 5 BOP-treated hamsters had cuboidal to columnar epithelial cells, which were still monolayered or partly multilayered and were characterized by enlarged nuclei (Fig. 1D), and these were identified as AH changes.
Gene Expression Profiles of Isolated Pancreatic Ducts The expression profiles of pancreatic AHs in BOP-treated hamsters were compared with those of normal pancreatic ducts in control animals, and a total of 112 genes, including 70 upregulated and 42 down-regulated ones, were identified as being greater than 3-fold differentially expressed in the AHs of BOPtreated hamsters (Tables 2, 3). Complete hybridization data TABLE 1. Incidence and Multiplicity of Hs, AHs, and ACs in the Pancreases of Hamsters With and Without the BOP Treatment Week Lesions Hyperplasia AH AC-early
RESULTS
&
AC-advanced
No. animals
8
12
16
20
6 6 9 8 1* (17) 3 (50) 5 (56) 6 (75) 0.2 (0.4)† 0.5 (0.6) 0.6 (0.5) 1.0 (0.8) 0 0 2 (22) 1 (13) 0.2 (0.4) 0.1 (0.4) 1 (17) 3 (50) 0 0 0.2 (0.4) 0.7 (0.8) 1 (17) 2 (33) 5 (56) 7 (88) 0.3 (0.8) 0.3 (0.5) 0.6 (0.5) 2.0 (1.3)
*
Number of animals with lesions (%). Mean number and SD of each lesion per animal.
†
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pancreas
&
Volume 44, Number 1, January 2015
Biphasic Alterations in MUC1 in Hamster Pancreas
FIGURE 2. Histopathology and immunohistochemistry of normal pancreatic ducts in a control hamster (A) and hyperplastic (B) and atypically hyperplastic (C) ducts in BOP-treated hamsters; hematoxylin and eosin staining, scale bars of 50 Km. Immunohistochemistry for MUC1 cytoplasmic domain (DYF), slug (GYI), vimentin (JYL), and E-cadherin (MYO) of each serial section of (AYC). Note the clear staining for the MUC1-CD in the apical surfaces of hyperplastic and atypically hyperplastic ducts (E, F). Nuclear/cytoplasmic reactions to antislug antibodies were sometimes observed not only in hyperplastic and atypically hyperplastic ducts (H, I) but also in normal pancreatic ducts and islet cells (G). * 2014 Lippincott Williams & Wilkins
www.pancreasjournal.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
79
Pancreas
Kitahashi et al
&
Volume 44, Number 1, January 2015
FIGURE 3. Histopathology and immunohistochemistry of AC in a BOP-treated hamster (A) and pancreatic lymph node with metastatic lesions in another BOP-treated hamster (B), scale bars of 200 Km. C, A higher magnification of the box in B, scale bar of 50 Km. Immunohistochemistry for the MUC1-CD (DYF), slug (GYI), vimentin (JYL), and E-cadherin (MYO) of each serial section of A or C, scale bars of 50 Km. E, Clear cytoplasmic expression of the MUC1-CD in the invasive front of AC. F, Nuclear accumulation of the MUC1-CD in metastatic AC cells (arrows). K, Note the cytoplasmic expression of vimentin in the invasion front of AC (arrows).
80
www.pancreasjournal.com
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pancreas
&
Volume 44, Number 1, January 2015
Biphasic Alterations in MUC1 in Hamster Pancreas
TABLE 2. Up-Regulated Genes Identified as Being Greater Than 3-Fold Differently Expressed in AHs Relative to Normal Pancreatic Ducts Fold Change
Accession
Description
P
21.58 16.79 12.60 12.36 11.30 10.71 10.30 10.28 9.95 9.71 9.14
DQ369045.1 AY535011.1 AF221841.1 AF060820.1 AJ431372.2 M13933.1 DQ367730.1 DQ207608.1 U36918.1 D26473.1 AY649405.1
1.77E-09 9.93E-11 1.89E-08 8.93E-09 7.03E-08 9.99E-08 2.17E-08 4.11E-11 3.45E-07 1.31E-08 3.44E-09
8.99
DQ403102.1
8.71 8.52 8.10 7.86 7.48 7.29
L40382.1 DQ402953.1 D15058.1 U41499.1 AY508121.1 DQ403002.1
7.12 6.79 6.77 6.61 6.58 6.52 6.34 6.30 6.18 5.99 5.91
U41737.1 AJ312092.1 AF487552.1 X53139.1 AF004814.1 AF364317.1 DQ092501.1 AY550277.1 AF260255.1 J00059.1 DQ202704.1
5.90 5.70 5.66 5.62 5.50 5.29 5.18 4.88 4.88 4.70 4.58 4.49 4.33 4.23 4.14 4.10 4.07 3.95 3.84 3.83 3.79 3.78
AY057063.1 M26640.1 AY100688.1 AF532110.1 M88136.1 X04850.1 U58672.1 M10230.1 X87227.1 AF294427.1 M34665.1 AY553637.1 AJ582075.1 U08342.1 M17169.1 DQ473436.1 DQ202702.1 L40381.1 U73375.1 AB194396.1 X13175.1 DQ403027.1
Mucin 1 mRNA, partial CDS Putative mitochondrial ATP synthase mRNA, partial CDS Thioredoxin peroxidase II mRNA, complete CDS Aldehyde reductase mRNA, partial CDS mRNA for CAP1 protein Ribosomal protein RPS17 mRNA, complete CDS Complement protein C1qBP mRNA, partial CDS Beta-actin mRNA, partial CDS Mucin (MUC1) mRNA, complete CDS RPS4 mRNA for ribosomal protein S4, complete CDS Mitochondrial NADH-ubiquinone oxidoreductase ESSS subunit precursor, mRNA, complete CDS Mitochondrial ATP synthase, H + transporting F1 complex beta subunit (ATP5B) mRNA, partial CDS GST pi enzyme mRNA, complete CDS Mitochondrial malate dehydrogenase 2, NAD mRNA, partial CDS mRNA for DAD-1, complete CDS Arsenite-resistance protein mRNA, complete CDS Clone 1(10) adrenal gland 15-kDa selenoprotein-like mRNA sequence Succinate dehydrogenase complex subunit B mRNA, partial CDS Pancreatic beta cell growth factor (INGAP) mRNA, complete CDS mRNA for beta actin (actb gene) Nucleotide-binding protein G(s) alpha subunit (GNAS) mRNA, complete CDS mRNA for nucleotide-binding coupling protein alpha subunit Ubiquitin conjugating enzyme mRNA, complete CDS Protein disulfide-isomerase mRNA, complete CDS MHC class II antigen alpha chain mRNA, partial CDS Adrenonovelin mRNA, complete CDS Tissue inhibitor of matrix metalloproteinase-2 mRNA, partial CDS Preproglucagon mRNA, complete CDS Guanine nucleotide binding protein, alpha stimulating (GNAS) mRNA, complete CDS ERP57 protein mRNA, complete CDS Sulfated glycoprotein 2 mRNA, 3’ end Calreticulin mRNA, complete CDS Peroxiredoxin 2 mRNA, complete CDS Seryl-tRNA synthetase mRNA, partial CDS mRNA fragment for glucose regulated protein (grp 94) Ornithine decarboxylase antizyme mRNA, complete CDS >-tubulin I mRNA, 3’ flank mRNA for vimentin Proliferating cell nuclear antigen mRNA, complete CDS T-complex protein 1 mRNA, complete CDS Inhibitor of DNA binding 2 mRNA, 3’ UTR Partial mRNA for Bcl-2 associated protein (bax gene) Beta-tubulin isotype I mRNA, complete CDS Glucose-regulated protein GRP78 mRNA, complete CDS Immunoglobulin heavy chain mRNA, partial CDS Guanine nucleotide binding protein alpha inhibiting 2 mRNA, complete CDS GST pi enzyme mRNA, complete CDS Pancreas cancer-associated protein 4 mRNA, complete CDS mRNA for ovca1, partial CDS KG4 mRNA related to cellular proliferation Ribosomal protein L18 mRNA, partial CDS
1.14E-10 8.1E-10 3.61E-09 2.22E-08 2.81E-06 3.15E-07 5.75E-07 1.26E-06 6.12E-08 2.44E-08 1.16E-05 5.68E-08 3.64E-08 1.79E-07 1.89E-08 8.13E-08 1.32E-07 1.51E-05 1.41E-10 1.17E-08 1.85E-06 1.07E-06 6.46E-08 8.01E-08 3.13E-08 1.60E-06 2.36E-07 3.78E-07 2.17E-08 2.42E-08 5.72E-05 1.01E-07 1.68E-05 1.24E-06 9.12E-06 6.84E-08 2.03E-08 5.97E-05 5.97E-10 6.94E-08
(continued on next page) * 2014 Lippincott Williams & Wilkins
www.pancreasjournal.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
81
Pancreas
Kitahashi et al
&
Volume 44, Number 1, January 2015
TABLE 2. (Continued) Fold Change
Accession
3.68 3.66 3.59 3.54 3.41 3.40 3.35 3.33 3.31 3.29 3.24 3.23 3.20 3.19 3.15 3.15 3.14 3.12 3.09
AF046214.1 M11241.1 J02983.1 U50395.1 L18986.1 M99692.1 AB014876.2 X57112.1 X62678.1 X77631.1 DQ017755.1 AF320819.1 D31814.1 EF210569.1 AF281149.1 AY688356.1 AF020767.1 X57029.1 AY004869.1
are available as part of the supplementary information (SDC 1, at http://links.lww.com/MPA/A312). Some up-regulated genes related to cell differentiation, proliferation, or motility were identified based on the potential molecular function of the proteins encoded by the differentially expressed genes, for example, mucin 1 (MUC1, 21.58- and 9.95-fold), tissue inhibitor of matrix metalloproteinase-2 (6.18-fold), vimentin (4.88-fold), and proliferating cell nuclear antigen (4.70-fold). Genes encoding redox proteins, such as thioredoxin peroxidase II (12.60-fold) and peroxiredoxin 2 (5.62-fold), and metabolizing enzymes, such as aldehyde reductase (12.36-fold) and mitochondrial malate dehydrogenase 2 (8.52-fold), were also up-regulated in AHs. On the other hand, genes encoding cell adhesion proteins such as connexin 45 (0.17-fold), gap junction membrane channel protein alpha 1 (0.27-fold), and connexin 47 (0.31-fold) were found to be down-regulated. However, normal expression levels and the significance of the down-regulation of these genes in the pancreatic duct remain unclear.
Gene Expression and Immunohistochemical Analysis of MUC1 in Normal and Precancerous and Cancerous Pancreatic Tissue Among the up-regulated and down-regulated genes identified in the microarray analysis, expression of the MUC1 gene in AHs was markedly different from that in normal pancreatic ducts. In addition, MUC1 was previously shown to be expressed in PanIN lesions together with IDCs in humans using immunohistochemistry.34 To validate the microarray results obtained, qRT-PCR and immunohistochemical analysis on the MUC1 gene and protein expression, respectively, were performed using the pancreatic tissues of control and BOP-treated hamsters and AC tissues from BOP-treated hamsters. Gene expression levels in the normal pancreases of control animals and normal-like pancreases of BOP-treated animals were similar, whereas those in BOP-induced AC tissues were slightly higher (Fig. 4A). Immunohistochemical analysis revealed that the acinar cell apices were positive for MUC1-CD, whereas the intercalated ductal epithelia were not in both control and BOP-treated
82
www.pancreasjournal.com
P
Description TGF-A mRNA, partial CDS Chinese hamster ovary ribosomal protein S14 mRNA, complete CDS Insulinoma rig mRNA, encoding a putative DNA-binding protein Wild-type tumor suppressor P53 (p53) mRNA, complete CDS Lysosomal membrane glycoprotein A mRNA, complete CDS Iron-binding protein (ferritin) mRNA mRNA for ribosomal protein L13, complete CDS mRNA for beta 2-microglobulin mRNA for P5 protein IFAP300 mRNA Placental cathepsin L mRNA, partial cds Basigin precursor, mRNA, partial cds mRNA for polyprotein precursor, complete CDS Regenerating protein 2 mRNA, complete CDS Spermidine/spermine N1-acetyltransferase mRNA, complete CDS Cu/Zu superoxide dismutase 1-like mRNA, partial sequence Skeletal muscle neutral protease mekratin mRNA, complete CDS mRNA for >-globin chain Cyclophilin mRNA, partial CDS
6.09E-06 4.77E-07 1.80E-07 0.000150 3.91E-06 4.31E-07 4.90E-07 1.26E-05 1.12E-05 3.06E-06 0.000123 6.98E-05 4.95E-06 0.000796 0.000345 8.87E-08 1.71E-08 3.23E-05 6.53E-07
hamsters (Fig. 2D). The main and branch (intralobular and interlobular) pancreatic ducts were negative for MUC1-CD but were occasionally slightly positive at their apical surface (Fig. 2D). In contrast, the apical surfaces of the epithelia of all Hs, AHs, ACs-early, and AC-advanced areas with distinct tubular patterns were positive for MUC1-CD, being observed in all BOP-treated hamster tissues (Figs. 2E, F; 3D), and no significant change was observed in staining intensities between the different grades of lesions. In contrast, surface staining for MUC1-CD was weaker in the invasive front of ACs-advanced, and cytoplasmic staining was characteristic (Fig. 3E). In addition, the nuclear accumulation of MUC1-CD was clearly observed in 1 lymph node metastasis case (Fig. 3F).
Gene Expression and Immunohistochemical Analysis of EMT-Related Proteins in Normal and Precancerous and Cancerous Pancreatic Tissue MUC1-CD was previously shown to play a direct role in MUC1 by initiating EMT.30 Therefore, the coexpression of EMTrelated proteins was analyzed to clarify the role of MUC1 in the progression processes of pancreatic carcinogenesis in vivo. The gene expression levels of slug in the normal pancreases of control animals were negligible, whereas those in the normal-like pancreases of BOP-treated animals were slightly higher and were significantly increased in the AC tissues of BOP-treated animals (Fig. 4B). Immunohistochemical analysis revealed that nuclear slug-positive ratios were slightly higher in Hs, AHs, ACsearly, and AC-advanced areas with distinct tubular patterns and higher in the invasive front of ACs-advanced than those in the normal(-like) pancreatic ducts of control and BOP-treated animals (Fig. 5). The increased nuclear staining of slug seemed to indicate its coexpression with cytoplasmic MUC1-CD (Fig. 3H), but not with apical MUC1-CD (Figs. 2GYI, 3G). Mesenchymal tissues/ cells, including fibroblasts, vascular endothelia, and immunocytes, were positive for vimentin, whereas the epithelia of normal pancreatic ducts, Hs, AHs, or AC-advanced areas with distinct tubular patterns were not (Figs. 2JYL, 3J). Carcinoma cells in the invasion front were positive for vimentin in 22 of the 24 AC-advanced * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pancreas
&
Volume 44, Number 1, January 2015
Biphasic Alterations in MUC1 in Hamster Pancreas
TABLE 3. Down-Regulated Genes Identified as Being Greater Than 3-Fold Differently Expressed in AHs Relative to Normal Pancreatic Ducts Fold Change 0.03 0.06 0.07 0.09 0.11 0.11 0.11 0.12 0.12 0.14 0.15 0.17 0.17 0.18 0.18 0.19 0.19 0.20 0.20 0.21 0.21 0.21 0.21 0.22 0.22 0.24 0.24 0.24 0.25 0.25 0.25 0.25 0.27 0.29 0.29 0.29 0.30 0.31 0.31 0.32 0.33 0.33
Accession
Description
P
AF052847.1 AJ746220.1 M27242.1 AF052838.1 L38710.1 AF052848.1 AY221595.1 AF052842.1 AF052841.1 Z66540.1 AF052846.1 AF052850.1 AF508038.1 AF052853.1 M27243.1 AF052844.1 AF052845.1 AF052849.1 AF052851.1 AY329083.1 AF052852.1 AB028235.1 AF055671.1 U71279.1 DQ837221.1 X64179.1 D50578.1 D28566.1 AY788096.1 DQ237898.1 DQ237889.1 DQ015678.1 AY049064.1 U65889.1 AF492475.1 X63022.1 AY162470.1 AF508037.1 AF052843.1 DQ237897.1 CF355260.1 D89286.1
Clone e23 retroviral-like pol protein (pol) mRNA, partial CDS Partial mRNA for eIF4GI protein Serum amyloid A (SAA1) mRNA, complete CDS Clone e26 retroviral-like pol protein (pol) mRNA, partial CDS 3-beta-hydroxysteroid dehydrogenase/5-4-eneisomerase (HSD3B) mRNA, complete CDS Clone e22 retroviral-like pol protein (pol) mRNA, partial CDS Baculoviral IAP repeat-containing protein 6 mRNA, partial CDS Clone ut16 retroviral-like pol protein (pol) mRNA, partial CDS Clone e8 retroviral-like pol protein (pol) mRNA, partial CDS mRNA (440 bp) related to sexual dimorphism Clone s25 retroviral-like pol protein (pol) mRNA, partial CDS Clone s1 retroviral-like pol protein (pol) mRNA, partial CDS Connexin 45 mRNA, partial CDS Clone ut10 retroviral-like pol protein (pol) mRNA, partial CDS Serum amyloid (SAA2) mRNA, complete CDS Clone s12 retroviral-like pol protein (pol) mRNA, partial CDS Clone t12 retroviral-like pol protein (pol) mRNA, partial CDS Clone e18 retroviral-like pol protein (pol) mRNA, partial CDS Clone s19 retroviral-like pol protein (pol) mRNA, partial CDS Per1-interacting protein mRNA, partial CDS Clone t3 retroviral-like pol protein (pol) mRNA, partial CDS mRNA for IL1 alpha, partial CDS Collagenase 2 (MMP-8) mRNA, partial CDS Adrenocorticotropin receptor (MC2) mRNA, complete CDS Secretoglobin precursor, mRNA, complete CDS mRNA for betalike y-globin gene mRNA for carboxylesterase, complete CDS mRNA for carboxylesterase, complete CDS Mitochondrial-associated cysteine-rich protein mRNA, complete CDS Progesterone receptor mRNA, partial CDS BMPR-IB mRNA, partial CDS 5-hydroxytryptamine receptor 2A mRNA, complete CDS Gap junction membrane channel protein alpha 1 mRNA, partial CDS Bone sialoprotein mRNA, complete CDS LIF receptor mRNA, partial CDS mRNA for cytochrome P450IIC Parotid secretory protein mRNA, complete CDS Connexin 47 mRNA, partial CDS Clone t2 retroviral-like pol protein (pol) mRNA, partial CDS N-cadherin mRNA, partial CDS cDNA similar to the androgen receptor, mRNA sequence mRNA for interalpha-trypsin inhibitor heavy chain 2, complete CDS
2.12E-07 2.39E-11 1.70E-09 7.63E-10 1.16E-07 3.82E-11 1.47E-08 6.71E-09 3.70E-08 1.18E-08 6.21E-08 2.43E-11 3.01E-07 1.59E-11 8.76E-06 1.10E-06 1.91E-06 1.74E-12 1.44E-12 1.05E-08 5.82E-12 1.51E-08 2.38E-08 4.66E-06 4.64E-09 5.85E-06 2.57E-07 7.52E-07 2.11E-06 4.60E-06 1.33E-05 0.00043 4.64E-05 1.03E-06 4.47E-06 1.09E-05 1.99E-06 3.68E-06 0.00292 0.00205 1.91E-05 5.76E-08
cases; however, the staining intensities and expression ratios in each lesion were weak and low, respectively (Fig. 3K). Epithelial tissues/cells, including acinar cell and main and branch pancreatic ducts, showed membranous positivities for E-cadherin (Fig. 2M), and its subcellular expression patterns were similar; however, the staining intensities seemed to be higher in Hs, AHs, or ACs (Figs. 2N, O; 3M) than in normal pancreatic ducts. The coexpression of cytoplasmic MUC1, nuclear slug, and cytoplasmic vimentin and weak staining intensities of E-cadherin (Figs. 3N, O) in carcinoma cells in the invasion front were generally confirmed, as were those of nuclear MUC1 and nuclear slug (Fig. 3I), whereas vimentin expression did not seem to be correlated with nuclear MUC1 * 2014 Lippincott Williams & Wilkins
expression (Fig. 3L) in the lymph node metastatic case. The gene expression levels of snail1, twist, E-cadherin, cdh2, vimentin, and TGF-A1 were similar between the normal pancreases of control animals and normal-like pancreases of BOP-treated animals, whereas those in the AC tissues of BOP-treated animals were increased, except for E-cadherin (Figs. 4CYH).
DISCUSSION Among the up-regulated genes in the AHs of pancreatic ducts analyzed using a microarray technique in the hamster carcinogenesis model, the expression of the MUC1 gene was markedly different from that in normal pancreatic ducts. To www.pancreasjournal.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
83
Pancreas
Kitahashi et al
&
Volume 44, Number 1, January 2015
FIGURE 4. Gene expression levels of (A) MUC1, (B) slug, (C) snail1, (D) twist, (E) E-cadherin, (F) cdh2, (G) vimentin, and (H) TGF-A1 in the normal pancreases of control animals and normal-like pancreases and AC tissues of BOP-treated animals (sample numbers were 4, 8, and 4, respectively). *P G 0.01 and **P G 0.05 vs normal control; Student t test.
84
www.pancreasjournal.com
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pancreas
&
Volume 44, Number 1, January 2015
FIGURE 5. Immunohistochemical slug-positive ratios in the nuclei of each lesion from control and BOP-treated animals. The numbers of animals counted were 3, 5, 15, 3, 5, 15, and 15 for normal pancreatic ducts in control animals, normal-like ducts in BOP(+) animals, Hs, AHs, ACs-early, AC areas with distinct tubular patterns, and the invasive front of ACs, respectively. **P G 0.01 vs all other lesions except for AC-early (P G 0.05); Tukey test.
validate the microarray results obtained, qRT-PCR for MUC1 mRNA levels was performed using the pancreatic tissues of control and BOP-treated hamsters and carcinoma tissues from BOP-treated hamsters. The results revealed similar expression levels between the normal pancreases of control animals and normal-like pancreases of BOP-treated animals. Immunohistochemistry for MUC1 revealed that its expression was clearly detected in the apices of acinar cells, which constitute a marked portion of normal(-like) pancreatic tissue, and BOP-induced alterations were absent in acinar cells, the major source of MUC1 expression. This was considered to be the reason for the absence of a significant difference in MUC1 mRNA levels in normal(-like) pancreatic tissue between control and BOPtreated animals. MUC1 has been detected in the cell apices of the centroacinar cells, intercalated ducts, and intralobular ducts of the normal pancreas in humans by every antibody for the different glycoforms of MUC1, for example, NCL-MUC-1CORE (clone Ma552, reacts with the core peptide of the tandem repeat domain),35 DF3 (reacts with the core peptide of the tandem repeat domain),36 MY.1E12 (reacts with sialylated MUC1 mucin), and HMFG-1 (reacts with fully glycosylated MUC1 mucin).34 The results of the present study showing acinar cell positivity by anti-MUC1-CD antibodies in the hamster pancreas were consistent with the abovementioned results in humans. The qRT-PCR revealed that MUC1 mRNA expression levels were slightly higher in hamster carcinoma tissues than in normal(-like) pancreatic tissues, which demonstrated that MUC1 gene expression levels in carcinoma cells were higher than those in acinar cells and the unique function of this molecule in pancreatic ductal carcinogenesis. The immunohistochemical expression of MUC1-CD was negative in normal(-like) interlobular/intralobular pancreatic ducts and positive in the apical surfaces of not only AHs but also Hs without atypia, and no significant change was observed in staining intensities between the different grades of lesions. Previous studies demonstrated that MUC1 was expressed in the cell apices of intralobular ducts and intercalated ducts in the normal human pancreas, and the increase in PanINs was correlated with their grades by every antibody for the different * 2014 Lippincott Williams & Wilkins
Biphasic Alterations in MUC1 in Hamster Pancreas
glycoforms of MUC1.34,37 The reason for the rapid induction of MUC1 in the hyperplastic ductal cell apices of the present hamster pancreatic carcinogenesis model remains unclear. In the histopathological evaluation, Hs without atypia and advanced ACs increased over time throughout the experiment period, whereas AHs and early carcinomas with minimal invasive growth were observed at lower incidences at all time points. This may indicate that a rapid progression process from Hs to carcinomas was present in the hamster carcinogenesis model used, and hyperplastic ductal epithelia may harbor some early malignant characteristics including the distinct up-regulation of MUC1. Another possibility may be related to the anti-MUC1CD antibody-specific immunoreactivities observed in the present study. Further immunohistochemical studies using anti-MUC1CD antibodies have yet to be conducted in human PanINs and IDC samples. One of the most interesting aspects of the present study was the cytoplasmic positivity of MUC1-CD characterized in the invasive front of ACs. In addition, the nuclear accumulation of MUC1-CD was clearly observed in 1 lymph node metastasis case. A similar phenomenon was previously reported in relation to the aggressive behavior of human pancreatic IDCs using MUC1/DF3 antibodies (reacts with the core peptide of the tandem repeat domain), which were expressed not only at the cell apex but also in the cytoplasm and lateral membrane.27 MUC1 is encoded by a single transcript and, after transcription, undergoes autocleavage into 2 subunits to form a stable transmembrane heterodimer.38 The MUC1 N-terminal subunit is a mucin component that has been extensively glycosylated, of which the core peptide contains tandem repeats. MUC1 Cterminal subunit is recognized as an oncogene, has 7 tyrosine residues, and was first reported to interact with A-catenin.38 MUC1 has recently been shown to regulate the nuclear localization and function of the epidermal growth factor receptor,39 or MUC1-CD is translocated to the nucleus, in which it promotes the transcription of metastatic genes associated with EMT in human pancreatic carcinoma cells.40 The present study is the first to demonstrate the in vivo cytoplasmic/nuclear positivity of MUC1-CD characterized in the invasive front or the metastasis of ACs in contrast to its apical positivity in carcinoma areas with distinct tubular patterns. In addition, the coexpression of EMT-related proteins, such as slug and vimentin, with cytoplasmic MUC1 was also shown. Despite the distinct coexpression of cytoplasmic/nuclear MUC1 and slug/vimentin in the invasive front or the metastasis of ACs, nuclear slug positivity was slightly more frequent in Hs, AHs, and AC areas with distinct tubular patterns, in which MUC1 overexpression was observed in their epithelial apices only. In addition, slug mRNA levels were slightly higher in the normal-like pancreases of BOP-treated hamsters than in the normal pancreases of controls. Therefore, further studies are needed to clarify the overexpression of MUC1 before its translocalization to the cytoplasm/ nuclei, which could be a preparatory stage for EMT in the epithelial cells of preneoplastic lesions, for example, Hs and AHs. In addition, the key signaling molecules contributing EMT via MUC1 intracellular translocalization should be investigated by a comparative gene expression analysis using DNA microarrays between AC areas with distinct tubular patterns and those in the invasive front, in combination with MUC1 gene knockdown or knockout studies in vitro and/or in vivo. In any case, the purpose of the present immunohistochemical analysis was to demonstrate a biphasic alteration in the expression level and subcellular localization of MUC1 in pancreatic ductal carcinogenesis using an animal model, and the results obtained indicate that both an elevation in MUC1 expression levels and the www.pancreasjournal.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
85
Pancreas
Kitahashi et al
cleavage/translocation of MUC1-CD to the cytoplasm or nuclei were important targets for pancreatic cancer prevention and chemotherapy. Further immunohistochemical studies using anti-MUC1-CD antibodies should be conducted in human IDC samples. ACKNOWLEDGMENT We thank Ms Satomi Kohno for her expert technical assistance and the National Cancer Center Research Core Facility for some of the analyses in this study. The Core Facility was supported by the National Cancer Center Research and Development Fund (23-A-7). REFERENCES 1. Matsuda A, Matsuda T, Shibata A, et al. Cancer incidence and incidence rates in Japan in 2007: a study of 21 population-based cancer registries for the Monitoring of Cancer Incidence in Japan (MCIJ) project. Jpn J Clin Oncol. 2013;43:328Y336. 2. Krejs GJ. Pancreatic cancer: epidemiology and risk factors. Dig Dis. 2010;28:355Y358. 3. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J Clin. 2013;63:11Y30. 4. Frampton AE, Krell J, Giovannetti E, et al. Defining a prognostic molecular profile for ductal adenocarcinoma of the pancreas highlights known key signaling pathways. Expert Rev Anticancer Ther. 2012;12:1275Y1278. 5. Cutts RJ, Gadaleta E, Hahn SA, et al. The Pancreatic Expression database: 2011 update. Nucleic Acids Res. 2011;39:D1023YD1028. 6. Kim HN, Choi DW, Lee KT, et al. Gene expression profiling in lymph node-positive and lymph node-negative pancreatic cancer. Pancreas. 2007;34:325Y334. 7. Crnogorac-Jurcevic T, Gangeswaran R, Bhakta V, et al. Proteomic analysis of chronic pancreatitis and pancreatic adenocarcinoma. Gastroenterology. 2005;129:1454Y1463. 8. Logsdon CD, Simeone DM, Binkley C, et al. Molecular profiling of pancreatic adenocarcinoma and chronic pancreatitis identifies multiple genes differentially regulated in pancreatic cancer. Cancer Res. 2003; 63:2649Y2657. 9. Feldmann G, Beaty R, Hruban RH, et al. Molecular genetics of pancreatic intraepithelial neoplasia. J Hepatobiliary Pancreat Surg. 2007;14:224Y232. 10. Hruban RH, Adsay NV, Albores-Saavedra J, et al. Pancreatic intraepithelial neoplasia: a new nomenclature and classification system for pancreatic duct lesions. Am J Surg Pathol. 2001;25:579Y586. 11. Bardeesy N, DePinho RA. Pancreatic cancer biology and genetics. Nat Rev Cancer. 2002;2:897Y909. 12. Buchholz M, Braun M, Heidenblut A, et al. Transcriptome analysis of microdissected pancreatic intraepithelial neoplastic lesions. Oncogene. 2005;24:6626Y6636. 13. Pan S, Chen R, Reimel BA, et al. Quantitative proteomics investigation of pancreatic intraepithelial neoplasia. Electrophoresis. 2009;30:1132Y1144. 14. Hingorani SR, Petricoin EF, Maitra A, et al. Preinvasive and invasive ductal pancreatic cancer and its early detection in the mouse. Cancer Cell. 2003;4:437Y450. 15. Hruban RH, Adsay NV, Albores-Saavedra J, et al. Pathology of genetically engineered mouse models of pancreatic exocrine cancer: consensus report and recommendations. Cancer Res. 2006;66:95Y106. 16. Morris JPT, Cano DA, Sekine S, et al. Beta-catenin blocks Kras-dependent reprogramming of acini into pancreatic cancer precursor lesions in mice. J Clin Invest. 2010;120:508Y520. 17. Corcoran RB, Contino G, Deshpande V, et al. STAT3 plays a critical role in KRAS-induced pancreatic tumorigenesis. Cancer Res. 2011;71:5020Y5029. 18. Fujii H, Egami H, Chaney W, et al. Pancreatic ductal adenocarcinomas induced in Syrian hamsters by N-nitrosobis(2-oxopropyl)amine contain a c-Ki-ras oncogene with a point-mutated codon 12. Mol Carcinog. 1990;3:296Y301.
86
www.pancreasjournal.com
&
Volume 44, Number 1, January 2015
19. Muscarella P, Knobloch TJ, Ulrich AB, et al. Identification and sequencing of the Syrian Golden hamster (Mesocricetus auratus) p16(INK4a) and p15(INK4b) cDNAs and their homozygous gene deletion in cheek pouch and pancreatic tumor cells. Gene. 2001;278:235Y243. 20. Li J, Qin D, Knobloch TJ, et al. Expression and characterization of Syrian golden hamster p16, a homologue of human tumor suppressor p16 INK4A. Biochem Biophys Res Commun. 2003;304:241Y247. 21. Li J, Weghorst CM, Tsutsumi M, et al. Frequent p16INK4A/CDKN2A alterations in chemically induced Syrian golden hamster pancreatic tumors. Carcinogenesis. 2004;25:263Y268. 22. Birt DF, Patil K, Pour PM. Comparative studies on the effects of semipurified and commercial diet on longevity and spontaneous and induced lesions in the Syrian golden hamster. Nutr Cancer. 1985;7:167Y177. 23. Takahashi M, Pour P. Spontaneous alterations in the pancreas of the aging Syrian golden hamster. J Natl Cancer Inst. 1978;60:355Y364. 24. Kitahashi T, Yoshimoto M, Imai T. Novel immunohistochemical marker, integrin alpha(V)beta(3), for BOP-induced early lesions in hamster pancreatic ductal carcinogenesis. Oncol Lett. 2011;2:229Y234. 25. Konishi Y, Tsutsumi M, Tsujiuchi T. Mechanistic analysis of pancreatic ductal carcinogenesis in hamsters. Pancreas. 1998;16:300Y306. 26. Terada T, Ohta T, Sasaki M, et al. Expression of MUC apomucins in normal pancreas and pancreatic tumours. J Pathol. 1996;180:160Y165. 27. Yonezawa S, Goto M, Yamada N, et al. Expression profiles of MUC1, MUC2, and MUC4 mucins in human neoplasms and their relationship with biological behavior. Proteomics. 2008;8:3329Y3341. 28. Tsutsumida H, Swanson BJ, Singh PK, et al. RNA interference suppression of MUC1 reduces the growth rate and metastatic phenotype of human pancreatic cancer cells. Clin Cancer Res. 2006;12:2976Y2987. 29. Yuan Z, Liu X, Wong S, et al. MUC1 knockdown with RNA interference inhibits pancreatic cancer growth. J Surg Res. 2009;157:e39Ye46. 30. Roy LD, Sahraei M, Subramani DB, et al. MUC1 enhances invasiveness of pancreatic cancer cells by inducing epithelial to mesenchymal transition. Oncogene. 2011;30:1449Y1459. 31. Konishi Y, Mizumoto K, Kitazawa S, et al. Early ductal lesions of pancreatic carcinogenesis in animals and humans. Int J Pancreatol. 1990;7:83Y89. 32. Bolstad BM, Irizarry RA, Astrand M, et al. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics. 2003;19:185Y193. 33. Hori M, Kitahashi T, Imai T, et al. Enhancement of carcinogenesis and fatty infiltration in the pancreas in N-nitrosobis(2-oxopropyl)amine-treated hamsters by high-fat diet. Pancreas. 2011;40:1234Y1240. 34. Nagata K, Horinouchi M, Saitou M, et al. Mucin expression profile in pancreatic cancer and the precursor lesions. J Hepatobiliary Pancreat Surg. 2007;14:243Y254. 35. Baeckstrom D, Nilsson O, Price MR, et al. Discrimination of MUC1 mucins from other sialyl-Le(a)-carrying glycoproteins produced by colon carcinoma cells using a novel monoclonal antibody. Cancer Res. 1993;53:755Y761. 36. Siddiqui J, Abe M, Hayes D, et al. Isolation and sequencing of a cDNA coding for the human DF3 breast carcinoma-associated antigen. Proc Natl Acad Sci U S A. 1988;85:2320Y2323. 37. Kigure S. Immunohistochemical study of the association between the progression of pancreatic ductal lesions and the expression of MUC1, MUC2, MUC5AC, and E-cadherin. Rinsho Byori. 2006;54:447Y452. 38. Kufe DW. Functional targeting of the MUC1 oncogene in human cancers. Cancer Biol Ther. 2009;8:1197Y1203. 39. Bitler BG, Goverdhan A, Schroeder JA. MUC1 regulates nuclear localization and function of the epidermal growth factor receptor. J Cell Sci. 2010;123:1716Y1723. 40. Lau SK, Shields DJ, Murphy EA, et al. EGFR-mediated carcinoma cell metastasis mediated by integrin alphavbeta5 depends on activation of c-Src and cleavage of MUC1. PLoS One. 2012;7:e36753.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
