- Email: [email protected]
Smad4 alterations in pancreatic cancer cell lines than in primary pancreatic adenocarcinomas
Cancer Letters 139 (1999) 43±49
Higher frequency of DPC4/Smad4 alterations in pancreatic cancer cell lines than in primary pancreatic adenocarcinomas Detlef Bartsch a,*, Peter Barth b, Daniel Bastian a, Annette Ramaswamy b, Berthold Gerdes a, Brunhilde Chaloupka a, Yvonne Deiss a, Babette Simon c, Andreas Schudy a a
Department of Surgery, Philipps-University Marburg, Baldingerstraûe, 35043Marburg, Germany Department of Pathology, Philipps-University Marburg, Baldingerstraûe, 35043Marburg, Germany c Department of Internal Medicine, Philipps-University Marburg, Baldingerstraûe, 35043Marburg, Germany b
Received 1 September 1998; accepted 2 December 1998
Abstract The tumor suppressor gene DPC4/Smad4 at 18q21.1 is inactive in about 50% of pancreatic carcinoma xenografts and cell lines. However, the role of DPC4 in the multistep carcinogenesis of primary pancreatic adenocarcinomas remains uncertain. Therefore, we examined 45 primary human pancreatic adenocarcinomas and 12 pancreatic cancer cell lines for DPC4 alterations by single-strand conformational variant (SSCV) analysis and a PCR-based deletion assay. DPC4 was inactivated by either homozygous deletion or point mutation in 6 of 12 cell lines (50%). None of the primary pancreatic carcinomas carried a DPC4 mutation, although 66% revealed LOH of 18q21 sequences. These ®ndings suggest that inactivation of DPC4 occurs more frequently in tumor-derived cell lines than in primary pancreatic adenocarcinomas. In addition, another, yet unidenti®ed, tumor suppressor gene(s) may be linked with the frequent LOH of 18q21 in primary pancreatic adenocarcinomas. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Pancreatic carcinoma; DPC4; Smad4; Tumor suppressor genes; Mutations
1. Introduction The putative tumor suppressor gene DPC4 (deleted in pancreatic carcinoma locus 4; Smad4) was recently cloned at human chromosome 18q21.1 [1]. DPC4 is a member of the highly conserved family of Smad proteins which have been linked to the TGF-b superfamily of cytokines. Smad proteins are involved in the regulation of cell differentiation as well as the inhibition of cellular proliferation, and their alterations * Corresponding author. Tel.: 1 49-6421-286443; fax: 1 496421-288995. E-mail address: [email protected] (D. Bartsch)
could confer resistance to TGF-b and thereby contribute to tumorigenesis [2]. DPC4 was shown to be inactivated, either by homozygous deletion (30%) or by intragenic mutation (20%), in half of pancreatic cancer xenografts and cell lines [1]. DPC4 inactivation was also detected in 17% of primary colorectal cancers [3] and 16% of primary biliary tract carcinomas [4], but only in a minority ( , 7%) of cancers from other anatomic sites such as esophageal, gastric, ovarian, head and neck, lung and prostate areas [5± 14]. So far, the mutations were con®rmed for only two primary pancreatic carcinomas used for establishing xenografts [1]. Thus, the frequency of DPC4 altera-
0304-3835/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0304-383 5(98)00380-2
44
D. Bartsch et al. / Cancer Letters 139 (1999) 43±49
tions in primary pancreatic adenocarcinomas remains uncertain, especially since it is well known that cell lines can acquire genetic abnormalities during tissue culture and passage through nude mice [15,16]. We examined the entire coding region of DPC4 for mutations in 45 primer, pancreatic adenocarcinomas and 12 pancreatic cancer cell lines to assess the role of DPC4 in the tumorigenesis of this aggressive cancer. Primary tumors with suf®cient neoplastic cellularity were also evaluated for loss of heterozygosity (LOH) in the DPC4 region. 2. Materials and methods 2.1. Tumor specimens and cell lines Pancreatic cancer and normal tissue specimens used in this study included both frozen and paraf®nembedded specimens from patients undergoing surgery for pancreatic adenocarcinoma of all tumor stages from January 1991 to September 1997. Neoplastic cellularity was assessed on hematoxylin and eosin stained slides by two quali®ed pathologists (P.B. and A.R.). Thirty-eight fresh-frozen and seven paraf®n-embedded primary ductal pancreatic adenocarcinomas had a neoplastic cellularity greater than 40% after microdissection of the non-neoplastic tissue. The following cell lines were obtained from the American Type Culture Collection: ASPC1, BxPC3, Miapaca2, Capan1, CFPAC1, HPAF, Panc1, PC44, PC3 and Hs776T. The cell lines 8902 and 8988T were generously provided by Dr. P. ElsaÈsser, Department of Anatomy, Philipps-University Marburg, Germany. Genomic DNA of pancreatic xenografts PX23, PX74, PX28, PX86, PX101 and PX102 were kindly provided by Dr. S. Kern, Department of Oncology and Pathology, The Johns Hopkins Medical Institutions, Baltimore, MD. Genomic DNA from fresh-frozen tissue samples and cell lines was isolated using the QIAamp Tissue Kit according to the manufacturer's recommendations (Qiagen GmbH, Hilden, Germany). Paraf®n-embedded samples were extracted by the salting-out method as described previously [17]. 2.2. Mutation and LOH-analysis PCR±SSCV analysis was performed using 15
primer sets with intronic sequences designed to amplify DPC4 exons 1±11. Primer pairs for exons 1 to 7 were as follows: EX1/1S 5 0 TCAGAAATTGGAGACATATTTG3 0 and EX1/1AS 5 0 TCTTTTTTCTCCTTCAGCTTC3 0 ; EX1/2S 5 0 AGTGAAACATTTGCAAAAAGAG3 0 and EX1/2AS 5 0 GTTTTTAAATCTGCCACCATAG3 0 ; EX2S 5 0 AAAGTGTCTTGCATAATGTGAC3 0 and EX2AS 5 0 TTCTTAGGATGAAACAAACTAC3 0 ; EX3S 5 0 CTTTCATTGTAATGATTAATGTTTC3 0 and EX3AS 5 0 AAGAGAAAGTAGTAAGAAACAG3 0 ; EX4/1S 5 0 GATTTTAGGTGTTATTATATTACTTG3 0 and EX4/1AS 5 0 CTCTGTCGATGCACGATTAC3 0 ; EX4/2S 5 0 TTGTCCACTGAAGGACATTC3 0 and EX4/2AS 5 0 GCTGACTACA-TCTGATTCTAG3 0 ; EX5S 5 0 CATAAGATGACA-TCTATGAATG3 0 and EX5AS 5 0 GCTTTTAT-AAAGGCTGCCTAC3 0 ; EX6S 5 0 TAAAAGCAAATTAACCCATGTG3 0 and EX6AS 5 0 CCCTTACAACAAAAACAAGAG3 0 ; EX7S 5 0 TGAAAGTTTTAGCATTAGACAAC3 0 and EX7AS 5 0 CGTTTCAATCACCACTAAATC3 0 . Primers for exons 8 to 11 were published previously [4]. The PCR products were electrophoresed through 0.5% MDE gels (Mutation Detection Enhancement, Serogel, Biozym) with and without 5% glycerol at room temperature for 12 h at 6 W as reported [18]. Each PCR reaction was repeated at least once to con®rm the initial result. Pancreatic cancer xenografts, PX23, PX74, PX28, PX86, PX101 and PX102, known to harbor DPC4 mutations [1,5], were used as controls for the sensitivity of mutation detection under these SSCV conditions. In addition, titration experiments with the pancreatic cancer cell line ASPC1, carrying a homozygous mutation in exon 2 (Q100T) [5], and the ®broblast cell line W138 with wild-type DPC4 sequences were performed. The SSCV system used has been shown to be capable of detecting a DPC4 mutation, even if the neoplastic cells comprise only 10% of the tissue under study. Aberrant and wild type SSCV bands were excised, reampli®ed, and manually sequenced by cycle sequencing (SequiTherm, Epicentre) and visualized by autoradiography. All sequencing results were con®rmed by repeat PCR ampli®cations and sequencing on both strands. All cell lines were evaluated for homozygous deletion of DPC4 by multiplex PCR including co-ampli®cation of DPC4 exon 8 and K-ras exon 1. The
D. Bartsch et al. / Cancer Letters 139 (1999) 43±49
45
Fig. 1. (A) SSC variant of cell line Capan1 in exon 8 (arrowhead). (B) DNA sequencing identi®ed the nonsense mutation S343X (TCA to TGA) as indicated by the arrow. WT, wildtype.
forward primer sequence for K-ras exon 1 was 5 0 GGCCTGCTGAAAATGACTGA3 0 , the reverse 5 0 TTGTTGGATCATATTCGTCC3 0 . Multiplex PCR was performed in a ®nal volume of 10 ml containing 10±20 ng of genomic DNA, 1.25 mM MgCl2, 5% dimethyl sulfoxide, 200 mM of each dNTP, 0.5 units of Taq polymerase, and 1 mM of each primer. The PCR products were resolved on 10% non-denaturing polyacrylamide gels, stained with ethidium bromide and visualized by UV transillumination. Microsatellite markers D18S46, D18S474 and D18S363 from the region 18q21.1 were used to determine LOH in the DPC4 region. PCR ampli®cations
and PCR product analyses were carried out as described previously [4]. LOH of the tumor was de®ned as a reduction of the intensity with a minimal value of 50% by visual inspection in either of the two alleles as compared to those in the constitutional DNA. 3. Results Forty-®ve primary pancreatic ductal adenocarcinomas and 12 pancreatic carcinoma cell lines were examined for DPC4 mutations by SSCV and direct sequencing of the variants. All eleven exons of the DPC4 gene were successfully ampli®ed by PCR in
Fig. 2. (A) SSC variant of tumor 5584/95T in exon 11 (arrowhead). (B) DNA sequencing revealed a silent 1527I mutation (ATT to ATC) as indicated by the arrow. WT, wildtype.
46
D. Bartsch et al. / Cancer Letters 139 (1999) 43±49
Fig. 3. Examples of LOH at 18q21.1 in primary pancreatic adenocarcinoma. Upper or lower alleles that are clearly reduced or absent in the tumor DNA compared to the corresponding normal DNA are indicated by arrows. T, tumor tissue; N, corresponding normal constitutional DNA.
the 45 primary ductal adenocarcinomas, the six control xenografts and the eight pancreatic cancer cell lines, respectively. The cell lines CFPAC1, BxPC3, Hs776T and 8988T did not amplify for any SSCV amplicon or in the deletion assay indicating homozygous deletion of the DPC4 gene. SSCV analyses showed reproducible abnormal banding patterns in the six control xenografts and in the cell lines Capan1 and ASPC1, respectively (Fig. 1A). Sequence analysis of the variant bands con®rmed the described mutations in all six xenografts [1]. Table 1 LOH analysis of the DPC4 region in primary pancreatic adenocarcinoma a Tumor
NPC(%)
D18S474
D18S46
D18S363
58aT 80aT 18558/91T 7839/95T 12356/95T 832/96T 14951/96T 7732/96T 7aT 10aT 10826/95T 14089/95T
90 70 80 80 80 100 70 70 80 80 90 80
NI LOH LOH / LOH LOH LOH NI / / / NI
LOH / LOH NI LOH LOH NI LOH / / / /
NI / NI LOH NI LOH / LOH / NI NI /
a NPC (%), percentage of neoplastic cellularity; LOH, loss of heterozygosity; NI, not informative; /, retention of both alleles.
The cell line ASPC1 showed a Q100T mutation (AGG to ACG) in exon 2 and the cell line Capan1 a S343X mutation (TCA to TGA) in exon 8 (Fig. 1B), which is in agreement with a previous report [5]. In contrast, only one case (5584T) of the 45 primary tumor samples showed an aberrant band pattern (Fig. 2A). Upon direct sequencing of the variant band a base substitution at codon 527 in exon 11 (ATT to ATC) was detected, causing no change in the predicted amino acid (Fig. 2B). Since this base change is silent and is also found in the constitutional DNA, it is likely to be a polymorphism. Twelve of the 45 primary pancreatic carcinomas had a neoplastic cellularity greater than 70% after microdissection and were, therefore, eligible for LOH analysis. Eight of these 12 tumors (66%) showed LOH of at least one informative marker in the DPC4 region (Fig. 3, Table 1). 4. Discussion In the present study, the overall frequency of DPC4 alterations (deletions and mutations) observed in pancreatic carcinoma cell lines was 50%, whereas only one of the primary pancreatic adenocarcinomas revealed a point mutation, likely to be a polymorphism. Given the mutation rate of 20% in pancreatic cancer xenografts reported by Hahn et al. [1], nine of our 45 primary pancreatic carcinomas
D. Bartsch et al. / Cancer Letters 139 (1999) 43±49
would be expected to have a DPC4 mutation. Considering the eight primary tumors revealing LOH of 18q21 in the current study, about 20% or 1.6 tumors would be expected to harbor a DPC4 mutation, which is statistically not distinguishable from 0 mutated tumors observed. However, our observations raise the question as to the true rate of DPC4 mutations and whether the differences in observed rates in pancreatic carcinoma cell lines, xenografts and primary tumors might be due to something other than chance variation. Differences in the methodological approaches used may account for the different ®ndings obtained in Hahn's [1] and our studies. The xenograft approach used by Hahn et al. [1] limits contamination from non-neoplastic cells by enriching the tumor cells [19]. Thus, analyzing primary tumors after microdissection might underestimate the frequency of DPC4 mutations in primary pancreatic adenocarcinomas. However, this possibility seems less likely in light of the sensitivity of our test and the DNA quality, since we identi®ed p53 mutations in 45% of the samples and K-ras mutations in 74% of the same samples (unpublished data) in concordance with the reported rates by others [20,21]. On the other hand, the xenograft approach could overestimate the mutation frequency as a result of additional genetic changes conferring an in vivo selective growth advantage analogous to those described for the pl6 INK4a and p53 tumor suppressor genes in pancreatic cancer cell lines and orthotopic pancreatic cancer xenografts [15,22,23]. The possibility, however, that DPC4 mutations were not detected by SSCV or lie outside the regions studied (e.g. the 5 0 and 3 0 untranslated regions) must also be considered. Furthermore, PCR-based detection of homozygous deletions, which is an important mechanism of gene inactivation for DPC4 [1,5,6] in pancreatic cancer cell lines and xenografts, was precluded in our microdissected primary pancreatic carcinomas by contaminating non-neoplastic cells (up to 60%). Finally, national biases, based on population differences and environmental exposures, might be another reason for the different mutation frequencies in the studies. The presented results are in agreement with those of previous studies revealing a higher frequency of DPC4 alterations in human tumor-derived cell lines
47
than their corresponding primary tumors as demonstrated for human lung, head and neck, gastric and ovarian cancers [5,7±9]. Mutational inactivation or homozygous loss of DPC4 in tumor cell lines may represent in vitro acquired genetic alterations as has been proposed for loss or inactivation of the tumor suppressor gene p16 INK4a in pancreatic carcinomas and in a variety of other tumor types [22,23]. It is of note that DPC4 was only mutated in 20% of pancreatic cancer xenografts showing an LOH frequency of 90% at this locus [1]. Although we could not identify any DPC4 mutation in primary pancreatic adenocarcinomas, 66% of primary tumors with suf®cient neoplastic cellularity revealed also LOH of 18q21. These data and the recent ®nding of additional 18q11 deletion breakpoints in pancreatic carcinomas [24] make it reasonable to suggest the presence of an additional tumor suppressor gene on 18q important in pancreatic carcinogenesis. In favor of yet another tumor suppressor gene(s) on 18q are recent molecular genetic ®ndings in other tumors such as esophageal, gastric, colon, lung, breast and prostate carcinomas, which also show high frequencies of LOH at 18q21 and rarely alterations of the DPC4 gene [5,6,8,9,11,12,25]. Smad2 and DCC, both putative tumor suppressor genes located in this region, are probably not the candidate genes, since no inactivating mutations in these genes have been reported in pancreatic carcinomas [26,27]. Alternatively, DPC4 itself might be inactivated by other molecular mechanisms such as aberrant hypermethylation leading to transcription silencing. Tumor-speci®c aberrant hypermethylation at 18q21 has been reported in lung cancers [28], a putative mechanism of DPC4 gene inactivation yet to be investigated in pancreatic tumorigenesis. In conclusion, our studies of DPC4 alterations in primary ductal pancreatic carcinomas and cell lines suggest that, although this tumor suppressor plays a role in the carcinogenesis of pancreatic adenocarcinomas, abnormalities in this gene occur more frequently in tumor-derived cell lines than in primary tumors. DPC4 warrants further investigation in pancreatic carcinomas, including functional approaches such as gene replacement, to gain more insight into the pathogenesis of this fatal disease. In addition, future studies may lead to the identi®cation of another, yet unidenti®ed, tumor suppressor gene(s)
48
D. Bartsch et al. / Cancer Letters 139 (1999) 43±49
linked with frequent 18q21 deletions in pancreatic adenocarcinomas. Acknowledgements This work was supported by a grant of the Deutsche Forschungsgemeinschaft (Ba 1467/2-1) to D Bartsch References [1] S.A. Hahn, M. Schutte, A.T.M. Hoque, C.A. Moskaluk, L.T. da Costa, E. Rozenblum, C.L. Weinstein, A. Fischer, C.J. Yeo, R.H. Hruban, S.E. Kern, DPC-4, a candidate tumor suppressor gene at human chromosome 18q21.1, Science 271 (1996) 350±353. [2] Y. Zhang, X.-H. Feng, R.-Y. Wu, R. Derynck, Receptor-associated Mad homologues synergize as effectors of the TGF-b response, Nature 383 (1996) 168±172. [3] Y. Takagi, H. Kohmura, M. Futamura, H. Kida, H. Tanemura, K. Shimokawa, S. Shigetoyo, Somatic alterations of the DPC4 gene in human colorectal cancers in vivo, Gastroenterology 111 (1996) 1369±1372. [4] S.A. Hahn, D. Bartsch, A. Schroers, H. Galehdari, M. Becker, A. Ramaswamy, I. Schwarte-Waldhoff, H. Maschek, W. Schmiegel, Mutations of the DPC4/Smad4 gene in biliary tract carcinoma, Cancer Res. 58 (1998) 1124±1126. [5] M. Schutte, R.H. Hruban, L. Hedrick, K.R. Cho, G.M. Nadasdy, C.L. Weinstein, G.S. Boom, W.B. Isaacs, P. Cairns, H. Nawroz, D. Sidransky, R.A. Casero Jr., P.S. Meltzner, S.A. Hahn, S.E. Kern, DPC4 gene in various tumor types, Cancer Res. 56 (1996) 2527±2530. [6] S. Thiagalingam, C. Lengauer, F.S. Leach, M. Schutte, S.A. Hahn, J. Overhauser, J.K.V. Willson, S. Markowitz, S.R. Hamilton, S.E. Kern, K.W. Kinzler, B. Vogelstein, Evaluation of candidate tumor suppressor genes on chromosome 18 in colorectal cancers, Nat. Genet. 13 (1996) 342±346. [7] S.K. Kim, Y. Fan, V. Papadimitrakopoulou, G. Clayman, W.N. Hittelman, W.K. Hong, R. Lotan, L. Mao, DPC4, a candidate tumor suppressor gene, is altered infrequently in head and neck squamous cell carcinoma, Cancer Res. 56 (1996) 2519±2521. [8] M. Nagatake, Y. Takagi, H. Osada, K. Uchida, T. Mitsudomi, S. Saji, K. Shimokata, T. Takahashi, T. Takahashi, Somatic in vivo alterations of the DPC4 gene at 18q21 in human lung cancers, Cancer Res. 56 (1996) 2718±2720. [9] S. Nishizuka, G. Tamura, C. Maesawa, K. Sakata, Y. Suzuki, T. Iwaya, M. Terashima, K. Saito, R. Satodate, Analysis of the DPC4 gene in gastric carcinoma, Jpn. J Cancer Res. 88 (1997) 335±339. [10] A.T.M. Hoque, S.A. Hahn, S.E. Kern, DPC4 gene mutation in colitis associated neoplasia, Gut 40 (1997) 120±122. [11] M.T. Barrett, M. Schutte, S.E. Kern, B.J. Reid, Allelic loss and mutational analysis of the DPC4 gene in esophageal adenocarcinoma, Cancer Res. 56 (1996) 4351±4353.
[12] J. Lei, T.T. Zou, Y.-Q. Shi, X. Zhou, K.N. Smolinski, J. Yin, R.F. Souza, R. Appel, S. Wang, K. Cymes, O. Chan, J.M. Abraham, N. Harpaz, S.J. Meltzer, Infrequent DPC4 gene mutation in esophageal cancer, gastric cancer and ulcerative colitis-associated neoplasms, Oncogene 13 (1996) 2459±2462. [13] X.-T. Kong, S.H. Choi, A. Inoue, F. Xu, T. Chen, J. Takita, J. Yokota, F. Bessho, M. Yanagisawa, R. Hanada, K. Yamamoto, Y. Hayashi, Expression and mutational analysis of the DCC, DPC4, and MADR2/JV18-1 genes in neuroblastoma, Cancer Res. 57 (1997) 3772±3778. [14] D. McGrogan, M. Pegram, D. Slamon, R. Bookstein, Comparative mutational analysis of DPC4 (Smad) in prostatic and colorectal carcinomas, Oncogene 15 (1997) 1111±1114. [15] G. Reyes, A. Villanueva, C. Garcia, F.J. Sancho, J. Piulats, F. Luis, G. Capella, Orthotopic xenografts of human pancreatic carcinomas acquire genetic aberrations during dissemination in nude mice, Cancer Res. 56 (1996) 5713±5719. [16] C. Lengauer, K.W. Kinzler, B. Vogelstein, Genetic instability in colorectal cancers, Nature 386 (1996) 623±625. [17] D.K. Wright, M.M. Manos, Sample preparation from paraf®nembedded tissues, in: M.A. Innis, D.H. Gelfand, J.J. Sninsky, T.J. White (Eds.), PCR Protocols: a Guide to Methods and Applications, Academic Press, San Diego, CA, 1990, pp. 153±158. [18] D. Bartsch, D.W. Shevlin, W.S. Tung, O. Kisker, S.A. Wells Jr., P.J. Goodfellow, Frequent mutations of CDKN2 in primary pancreatic adenocarcinomas, Gene Chromosome Cancer 14 (1995) 189±195. [19] C. Caldas, S.A. Hahn, L.T. da Costa, M.S. Redston, M. Schutte, A.B. Seymour, C.L. Weinstein, R.H. Hruban, C.J. Yeo, S.E. Kern, Frequent somatic mutations and homozygous deletions of the p16 (MTS1) gene in pancreatic adenocarcinoma, Nat. Genet. 8 (1994) 27±32. [20] M.S. Redston, C. Caldas, A.B. Seymour, R.H. Hruban, L. da Costa, C.J. Yeo, S.E. Kern, p53 mutations in pancreatic carcinoma and evidence of common involvement of homocopolymer tracts in DNA microdeletions, Cancer Res. 54 (1994) 3025±3033. [21] C. Almoguera, D. Shibata, K. Forrester, J. Martin, N. Arnheim, M. Perucho, Most human carcinomas of the exocrine pancreas contain mutant c-K-ras genes, Cell 53 (1988) 549±554. [22] P. Cairns, L. Mao, A. Merlo, D.J. Lee, D. Schwab, Y. Eby, K. Tokino, P. van der Riet, J.E. Blaugrund, D. Sidransky, Rates of p16 (MTS1) mutations in primary tumor with 9p loss, Science 265 (1994) 415±417. [23] L. Huang, T.L. Goodrow, S.-H. Zhang, A.J. Klein-Szanto, H. Chang, B. Ruggeri, Deletion and mutation analysis of the p16/ MTS1 tumor suppressor gene in human ductal pancreatic cancer reveals a higher frequency of abnormalities in tumorderived cell lines than in primary ductal adenocarcinomas, Cancer Res. 56 (1996) 1137±1141. [24] M. HoÈglund, L. Gorunova, T. Jonson, S. Dawiskiba, A. Andren-Sandberg, G. Stenman, B. Johansson, Cytogenetic and FISH analyses of pancreatic carcinoma reveal breaks in 18q11 with consistent loss of 18q12-qter and frequent gain of 18p, Br. J. Cancer 77 (1998) 1893±1899.
D. Bartsch et al. / Cancer Letters 139 (1999) 43±49 [25] K. Uchida, M. Nagatake, H. Osadya, Y. Yatabe, M. Kondo, T. Mitsudomi, A. Masuda, T. Takahashi, T. Takahashi, Somatic in vivo alterations of the JV18-1 gene al 18q21 in human lung cancers, Cancer Res. 56 (1996) 5583±5585. [26] G.J. Riggins, K.W. Kinzler, B. Vogelstein, S. Thiagalingam, Frequency of Smad gene mutations in human cancers, Cancer Res. 57 (1997) 2578±2580. [27] B. Simon, R. Weinel, M. HoÈhne, J. Watz, J. Schmidt, G.
49
Kortner, R. Arnold, Frequent alteration of the tumor suppressor genes p53 and DCC in human pancreatic carcinoma, Gastroenterology 106 (1994) 1645±1651. [28] M. Nagatake, H. Osada, M. Kondo, K. Uchida, M. Nishio, K. Shimokata, T. Takahashi, Aberrant hypermethylation at the blc-2 locus at 18q21 in human lung cancers, Cancer Res. 56 (1996) 1886±1891.
Higher frequency of DPC4/Smad4 alterations in pancreatic cancer cell lines than in primary pancreatic adenocarcinomas Detlef Bartsch a,*, Peter Barth b, Daniel Bastian a, Annette Ramaswamy b, Berthold Gerdes a, Brunhilde Chaloupka a, Yvonne Deiss a, Babette Simon c, Andreas Schudy a a
Department of Surgery, Philipps-University Marburg, Baldingerstraûe, 35043Marburg, Germany Department of Pathology, Philipps-University Marburg, Baldingerstraûe, 35043Marburg, Germany c Department of Internal Medicine, Philipps-University Marburg, Baldingerstraûe, 35043Marburg, Germany b
Received 1 September 1998; accepted 2 December 1998
Abstract The tumor suppressor gene DPC4/Smad4 at 18q21.1 is inactive in about 50% of pancreatic carcinoma xenografts and cell lines. However, the role of DPC4 in the multistep carcinogenesis of primary pancreatic adenocarcinomas remains uncertain. Therefore, we examined 45 primary human pancreatic adenocarcinomas and 12 pancreatic cancer cell lines for DPC4 alterations by single-strand conformational variant (SSCV) analysis and a PCR-based deletion assay. DPC4 was inactivated by either homozygous deletion or point mutation in 6 of 12 cell lines (50%). None of the primary pancreatic carcinomas carried a DPC4 mutation, although 66% revealed LOH of 18q21 sequences. These ®ndings suggest that inactivation of DPC4 occurs more frequently in tumor-derived cell lines than in primary pancreatic adenocarcinomas. In addition, another, yet unidenti®ed, tumor suppressor gene(s) may be linked with the frequent LOH of 18q21 in primary pancreatic adenocarcinomas. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Pancreatic carcinoma; DPC4; Smad4; Tumor suppressor genes; Mutations
1. Introduction The putative tumor suppressor gene DPC4 (deleted in pancreatic carcinoma locus 4; Smad4) was recently cloned at human chromosome 18q21.1 [1]. DPC4 is a member of the highly conserved family of Smad proteins which have been linked to the TGF-b superfamily of cytokines. Smad proteins are involved in the regulation of cell differentiation as well as the inhibition of cellular proliferation, and their alterations * Corresponding author. Tel.: 1 49-6421-286443; fax: 1 496421-288995. E-mail address: [email protected] (D. Bartsch)
could confer resistance to TGF-b and thereby contribute to tumorigenesis [2]. DPC4 was shown to be inactivated, either by homozygous deletion (30%) or by intragenic mutation (20%), in half of pancreatic cancer xenografts and cell lines [1]. DPC4 inactivation was also detected in 17% of primary colorectal cancers [3] and 16% of primary biliary tract carcinomas [4], but only in a minority ( , 7%) of cancers from other anatomic sites such as esophageal, gastric, ovarian, head and neck, lung and prostate areas [5± 14]. So far, the mutations were con®rmed for only two primary pancreatic carcinomas used for establishing xenografts [1]. Thus, the frequency of DPC4 altera-
0304-3835/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0304-383 5(98)00380-2
44
D. Bartsch et al. / Cancer Letters 139 (1999) 43±49
tions in primary pancreatic adenocarcinomas remains uncertain, especially since it is well known that cell lines can acquire genetic abnormalities during tissue culture and passage through nude mice [15,16]. We examined the entire coding region of DPC4 for mutations in 45 primer, pancreatic adenocarcinomas and 12 pancreatic cancer cell lines to assess the role of DPC4 in the tumorigenesis of this aggressive cancer. Primary tumors with suf®cient neoplastic cellularity were also evaluated for loss of heterozygosity (LOH) in the DPC4 region. 2. Materials and methods 2.1. Tumor specimens and cell lines Pancreatic cancer and normal tissue specimens used in this study included both frozen and paraf®nembedded specimens from patients undergoing surgery for pancreatic adenocarcinoma of all tumor stages from January 1991 to September 1997. Neoplastic cellularity was assessed on hematoxylin and eosin stained slides by two quali®ed pathologists (P.B. and A.R.). Thirty-eight fresh-frozen and seven paraf®n-embedded primary ductal pancreatic adenocarcinomas had a neoplastic cellularity greater than 40% after microdissection of the non-neoplastic tissue. The following cell lines were obtained from the American Type Culture Collection: ASPC1, BxPC3, Miapaca2, Capan1, CFPAC1, HPAF, Panc1, PC44, PC3 and Hs776T. The cell lines 8902 and 8988T were generously provided by Dr. P. ElsaÈsser, Department of Anatomy, Philipps-University Marburg, Germany. Genomic DNA of pancreatic xenografts PX23, PX74, PX28, PX86, PX101 and PX102 were kindly provided by Dr. S. Kern, Department of Oncology and Pathology, The Johns Hopkins Medical Institutions, Baltimore, MD. Genomic DNA from fresh-frozen tissue samples and cell lines was isolated using the QIAamp Tissue Kit according to the manufacturer's recommendations (Qiagen GmbH, Hilden, Germany). Paraf®n-embedded samples were extracted by the salting-out method as described previously [17]. 2.2. Mutation and LOH-analysis PCR±SSCV analysis was performed using 15
primer sets with intronic sequences designed to amplify DPC4 exons 1±11. Primer pairs for exons 1 to 7 were as follows: EX1/1S 5 0 TCAGAAATTGGAGACATATTTG3 0 and EX1/1AS 5 0 TCTTTTTTCTCCTTCAGCTTC3 0 ; EX1/2S 5 0 AGTGAAACATTTGCAAAAAGAG3 0 and EX1/2AS 5 0 GTTTTTAAATCTGCCACCATAG3 0 ; EX2S 5 0 AAAGTGTCTTGCATAATGTGAC3 0 and EX2AS 5 0 TTCTTAGGATGAAACAAACTAC3 0 ; EX3S 5 0 CTTTCATTGTAATGATTAATGTTTC3 0 and EX3AS 5 0 AAGAGAAAGTAGTAAGAAACAG3 0 ; EX4/1S 5 0 GATTTTAGGTGTTATTATATTACTTG3 0 and EX4/1AS 5 0 CTCTGTCGATGCACGATTAC3 0 ; EX4/2S 5 0 TTGTCCACTGAAGGACATTC3 0 and EX4/2AS 5 0 GCTGACTACA-TCTGATTCTAG3 0 ; EX5S 5 0 CATAAGATGACA-TCTATGAATG3 0 and EX5AS 5 0 GCTTTTAT-AAAGGCTGCCTAC3 0 ; EX6S 5 0 TAAAAGCAAATTAACCCATGTG3 0 and EX6AS 5 0 CCCTTACAACAAAAACAAGAG3 0 ; EX7S 5 0 TGAAAGTTTTAGCATTAGACAAC3 0 and EX7AS 5 0 CGTTTCAATCACCACTAAATC3 0 . Primers for exons 8 to 11 were published previously [4]. The PCR products were electrophoresed through 0.5% MDE gels (Mutation Detection Enhancement, Serogel, Biozym) with and without 5% glycerol at room temperature for 12 h at 6 W as reported [18]. Each PCR reaction was repeated at least once to con®rm the initial result. Pancreatic cancer xenografts, PX23, PX74, PX28, PX86, PX101 and PX102, known to harbor DPC4 mutations [1,5], were used as controls for the sensitivity of mutation detection under these SSCV conditions. In addition, titration experiments with the pancreatic cancer cell line ASPC1, carrying a homozygous mutation in exon 2 (Q100T) [5], and the ®broblast cell line W138 with wild-type DPC4 sequences were performed. The SSCV system used has been shown to be capable of detecting a DPC4 mutation, even if the neoplastic cells comprise only 10% of the tissue under study. Aberrant and wild type SSCV bands were excised, reampli®ed, and manually sequenced by cycle sequencing (SequiTherm, Epicentre) and visualized by autoradiography. All sequencing results were con®rmed by repeat PCR ampli®cations and sequencing on both strands. All cell lines were evaluated for homozygous deletion of DPC4 by multiplex PCR including co-ampli®cation of DPC4 exon 8 and K-ras exon 1. The
D. Bartsch et al. / Cancer Letters 139 (1999) 43±49
45
Fig. 1. (A) SSC variant of cell line Capan1 in exon 8 (arrowhead). (B) DNA sequencing identi®ed the nonsense mutation S343X (TCA to TGA) as indicated by the arrow. WT, wildtype.
forward primer sequence for K-ras exon 1 was 5 0 GGCCTGCTGAAAATGACTGA3 0 , the reverse 5 0 TTGTTGGATCATATTCGTCC3 0 . Multiplex PCR was performed in a ®nal volume of 10 ml containing 10±20 ng of genomic DNA, 1.25 mM MgCl2, 5% dimethyl sulfoxide, 200 mM of each dNTP, 0.5 units of Taq polymerase, and 1 mM of each primer. The PCR products were resolved on 10% non-denaturing polyacrylamide gels, stained with ethidium bromide and visualized by UV transillumination. Microsatellite markers D18S46, D18S474 and D18S363 from the region 18q21.1 were used to determine LOH in the DPC4 region. PCR ampli®cations
and PCR product analyses were carried out as described previously [4]. LOH of the tumor was de®ned as a reduction of the intensity with a minimal value of 50% by visual inspection in either of the two alleles as compared to those in the constitutional DNA. 3. Results Forty-®ve primary pancreatic ductal adenocarcinomas and 12 pancreatic carcinoma cell lines were examined for DPC4 mutations by SSCV and direct sequencing of the variants. All eleven exons of the DPC4 gene were successfully ampli®ed by PCR in
Fig. 2. (A) SSC variant of tumor 5584/95T in exon 11 (arrowhead). (B) DNA sequencing revealed a silent 1527I mutation (ATT to ATC) as indicated by the arrow. WT, wildtype.
46
D. Bartsch et al. / Cancer Letters 139 (1999) 43±49
Fig. 3. Examples of LOH at 18q21.1 in primary pancreatic adenocarcinoma. Upper or lower alleles that are clearly reduced or absent in the tumor DNA compared to the corresponding normal DNA are indicated by arrows. T, tumor tissue; N, corresponding normal constitutional DNA.
the 45 primary ductal adenocarcinomas, the six control xenografts and the eight pancreatic cancer cell lines, respectively. The cell lines CFPAC1, BxPC3, Hs776T and 8988T did not amplify for any SSCV amplicon or in the deletion assay indicating homozygous deletion of the DPC4 gene. SSCV analyses showed reproducible abnormal banding patterns in the six control xenografts and in the cell lines Capan1 and ASPC1, respectively (Fig. 1A). Sequence analysis of the variant bands con®rmed the described mutations in all six xenografts [1]. Table 1 LOH analysis of the DPC4 region in primary pancreatic adenocarcinoma a Tumor
NPC(%)
D18S474
D18S46
D18S363
58aT 80aT 18558/91T 7839/95T 12356/95T 832/96T 14951/96T 7732/96T 7aT 10aT 10826/95T 14089/95T
90 70 80 80 80 100 70 70 80 80 90 80
NI LOH LOH / LOH LOH LOH NI / / / NI
LOH / LOH NI LOH LOH NI LOH / / / /
NI / NI LOH NI LOH / LOH / NI NI /
a NPC (%), percentage of neoplastic cellularity; LOH, loss of heterozygosity; NI, not informative; /, retention of both alleles.
The cell line ASPC1 showed a Q100T mutation (AGG to ACG) in exon 2 and the cell line Capan1 a S343X mutation (TCA to TGA) in exon 8 (Fig. 1B), which is in agreement with a previous report [5]. In contrast, only one case (5584T) of the 45 primary tumor samples showed an aberrant band pattern (Fig. 2A). Upon direct sequencing of the variant band a base substitution at codon 527 in exon 11 (ATT to ATC) was detected, causing no change in the predicted amino acid (Fig. 2B). Since this base change is silent and is also found in the constitutional DNA, it is likely to be a polymorphism. Twelve of the 45 primary pancreatic carcinomas had a neoplastic cellularity greater than 70% after microdissection and were, therefore, eligible for LOH analysis. Eight of these 12 tumors (66%) showed LOH of at least one informative marker in the DPC4 region (Fig. 3, Table 1). 4. Discussion In the present study, the overall frequency of DPC4 alterations (deletions and mutations) observed in pancreatic carcinoma cell lines was 50%, whereas only one of the primary pancreatic adenocarcinomas revealed a point mutation, likely to be a polymorphism. Given the mutation rate of 20% in pancreatic cancer xenografts reported by Hahn et al. [1], nine of our 45 primary pancreatic carcinomas
D. Bartsch et al. / Cancer Letters 139 (1999) 43±49
would be expected to have a DPC4 mutation. Considering the eight primary tumors revealing LOH of 18q21 in the current study, about 20% or 1.6 tumors would be expected to harbor a DPC4 mutation, which is statistically not distinguishable from 0 mutated tumors observed. However, our observations raise the question as to the true rate of DPC4 mutations and whether the differences in observed rates in pancreatic carcinoma cell lines, xenografts and primary tumors might be due to something other than chance variation. Differences in the methodological approaches used may account for the different ®ndings obtained in Hahn's [1] and our studies. The xenograft approach used by Hahn et al. [1] limits contamination from non-neoplastic cells by enriching the tumor cells [19]. Thus, analyzing primary tumors after microdissection might underestimate the frequency of DPC4 mutations in primary pancreatic adenocarcinomas. However, this possibility seems less likely in light of the sensitivity of our test and the DNA quality, since we identi®ed p53 mutations in 45% of the samples and K-ras mutations in 74% of the same samples (unpublished data) in concordance with the reported rates by others [20,21]. On the other hand, the xenograft approach could overestimate the mutation frequency as a result of additional genetic changes conferring an in vivo selective growth advantage analogous to those described for the pl6 INK4a and p53 tumor suppressor genes in pancreatic cancer cell lines and orthotopic pancreatic cancer xenografts [15,22,23]. The possibility, however, that DPC4 mutations were not detected by SSCV or lie outside the regions studied (e.g. the 5 0 and 3 0 untranslated regions) must also be considered. Furthermore, PCR-based detection of homozygous deletions, which is an important mechanism of gene inactivation for DPC4 [1,5,6] in pancreatic cancer cell lines and xenografts, was precluded in our microdissected primary pancreatic carcinomas by contaminating non-neoplastic cells (up to 60%). Finally, national biases, based on population differences and environmental exposures, might be another reason for the different mutation frequencies in the studies. The presented results are in agreement with those of previous studies revealing a higher frequency of DPC4 alterations in human tumor-derived cell lines
47
than their corresponding primary tumors as demonstrated for human lung, head and neck, gastric and ovarian cancers [5,7±9]. Mutational inactivation or homozygous loss of DPC4 in tumor cell lines may represent in vitro acquired genetic alterations as has been proposed for loss or inactivation of the tumor suppressor gene p16 INK4a in pancreatic carcinomas and in a variety of other tumor types [22,23]. It is of note that DPC4 was only mutated in 20% of pancreatic cancer xenografts showing an LOH frequency of 90% at this locus [1]. Although we could not identify any DPC4 mutation in primary pancreatic adenocarcinomas, 66% of primary tumors with suf®cient neoplastic cellularity revealed also LOH of 18q21. These data and the recent ®nding of additional 18q11 deletion breakpoints in pancreatic carcinomas [24] make it reasonable to suggest the presence of an additional tumor suppressor gene on 18q important in pancreatic carcinogenesis. In favor of yet another tumor suppressor gene(s) on 18q are recent molecular genetic ®ndings in other tumors such as esophageal, gastric, colon, lung, breast and prostate carcinomas, which also show high frequencies of LOH at 18q21 and rarely alterations of the DPC4 gene [5,6,8,9,11,12,25]. Smad2 and DCC, both putative tumor suppressor genes located in this region, are probably not the candidate genes, since no inactivating mutations in these genes have been reported in pancreatic carcinomas [26,27]. Alternatively, DPC4 itself might be inactivated by other molecular mechanisms such as aberrant hypermethylation leading to transcription silencing. Tumor-speci®c aberrant hypermethylation at 18q21 has been reported in lung cancers [28], a putative mechanism of DPC4 gene inactivation yet to be investigated in pancreatic tumorigenesis. In conclusion, our studies of DPC4 alterations in primary ductal pancreatic carcinomas and cell lines suggest that, although this tumor suppressor plays a role in the carcinogenesis of pancreatic adenocarcinomas, abnormalities in this gene occur more frequently in tumor-derived cell lines than in primary tumors. DPC4 warrants further investigation in pancreatic carcinomas, including functional approaches such as gene replacement, to gain more insight into the pathogenesis of this fatal disease. In addition, future studies may lead to the identi®cation of another, yet unidenti®ed, tumor suppressor gene(s)
48
D. Bartsch et al. / Cancer Letters 139 (1999) 43±49
linked with frequent 18q21 deletions in pancreatic adenocarcinomas. Acknowledgements This work was supported by a grant of the Deutsche Forschungsgemeinschaft (Ba 1467/2-1) to D Bartsch References [1] S.A. Hahn, M. Schutte, A.T.M. Hoque, C.A. Moskaluk, L.T. da Costa, E. Rozenblum, C.L. Weinstein, A. Fischer, C.J. Yeo, R.H. Hruban, S.E. Kern, DPC-4, a candidate tumor suppressor gene at human chromosome 18q21.1, Science 271 (1996) 350±353. [2] Y. Zhang, X.-H. Feng, R.-Y. Wu, R. Derynck, Receptor-associated Mad homologues synergize as effectors of the TGF-b response, Nature 383 (1996) 168±172. [3] Y. Takagi, H. Kohmura, M. Futamura, H. Kida, H. Tanemura, K. Shimokawa, S. Shigetoyo, Somatic alterations of the DPC4 gene in human colorectal cancers in vivo, Gastroenterology 111 (1996) 1369±1372. [4] S.A. Hahn, D. Bartsch, A. Schroers, H. Galehdari, M. Becker, A. Ramaswamy, I. Schwarte-Waldhoff, H. Maschek, W. Schmiegel, Mutations of the DPC4/Smad4 gene in biliary tract carcinoma, Cancer Res. 58 (1998) 1124±1126. [5] M. Schutte, R.H. Hruban, L. Hedrick, K.R. Cho, G.M. Nadasdy, C.L. Weinstein, G.S. Boom, W.B. Isaacs, P. Cairns, H. Nawroz, D. Sidransky, R.A. Casero Jr., P.S. Meltzner, S.A. Hahn, S.E. Kern, DPC4 gene in various tumor types, Cancer Res. 56 (1996) 2527±2530. [6] S. Thiagalingam, C. Lengauer, F.S. Leach, M. Schutte, S.A. Hahn, J. Overhauser, J.K.V. Willson, S. Markowitz, S.R. Hamilton, S.E. Kern, K.W. Kinzler, B. Vogelstein, Evaluation of candidate tumor suppressor genes on chromosome 18 in colorectal cancers, Nat. Genet. 13 (1996) 342±346. [7] S.K. Kim, Y. Fan, V. Papadimitrakopoulou, G. Clayman, W.N. Hittelman, W.K. Hong, R. Lotan, L. Mao, DPC4, a candidate tumor suppressor gene, is altered infrequently in head and neck squamous cell carcinoma, Cancer Res. 56 (1996) 2519±2521. [8] M. Nagatake, Y. Takagi, H. Osada, K. Uchida, T. Mitsudomi, S. Saji, K. Shimokata, T. Takahashi, T. Takahashi, Somatic in vivo alterations of the DPC4 gene at 18q21 in human lung cancers, Cancer Res. 56 (1996) 2718±2720. [9] S. Nishizuka, G. Tamura, C. Maesawa, K. Sakata, Y. Suzuki, T. Iwaya, M. Terashima, K. Saito, R. Satodate, Analysis of the DPC4 gene in gastric carcinoma, Jpn. J Cancer Res. 88 (1997) 335±339. [10] A.T.M. Hoque, S.A. Hahn, S.E. Kern, DPC4 gene mutation in colitis associated neoplasia, Gut 40 (1997) 120±122. [11] M.T. Barrett, M. Schutte, S.E. Kern, B.J. Reid, Allelic loss and mutational analysis of the DPC4 gene in esophageal adenocarcinoma, Cancer Res. 56 (1996) 4351±4353.
[12] J. Lei, T.T. Zou, Y.-Q. Shi, X. Zhou, K.N. Smolinski, J. Yin, R.F. Souza, R. Appel, S. Wang, K. Cymes, O. Chan, J.M. Abraham, N. Harpaz, S.J. Meltzer, Infrequent DPC4 gene mutation in esophageal cancer, gastric cancer and ulcerative colitis-associated neoplasms, Oncogene 13 (1996) 2459±2462. [13] X.-T. Kong, S.H. Choi, A. Inoue, F. Xu, T. Chen, J. Takita, J. Yokota, F. Bessho, M. Yanagisawa, R. Hanada, K. Yamamoto, Y. Hayashi, Expression and mutational analysis of the DCC, DPC4, and MADR2/JV18-1 genes in neuroblastoma, Cancer Res. 57 (1997) 3772±3778. [14] D. McGrogan, M. Pegram, D. Slamon, R. Bookstein, Comparative mutational analysis of DPC4 (Smad) in prostatic and colorectal carcinomas, Oncogene 15 (1997) 1111±1114. [15] G. Reyes, A. Villanueva, C. Garcia, F.J. Sancho, J. Piulats, F. Luis, G. Capella, Orthotopic xenografts of human pancreatic carcinomas acquire genetic aberrations during dissemination in nude mice, Cancer Res. 56 (1996) 5713±5719. [16] C. Lengauer, K.W. Kinzler, B. Vogelstein, Genetic instability in colorectal cancers, Nature 386 (1996) 623±625. [17] D.K. Wright, M.M. Manos, Sample preparation from paraf®nembedded tissues, in: M.A. Innis, D.H. Gelfand, J.J. Sninsky, T.J. White (Eds.), PCR Protocols: a Guide to Methods and Applications, Academic Press, San Diego, CA, 1990, pp. 153±158. [18] D. Bartsch, D.W. Shevlin, W.S. Tung, O. Kisker, S.A. Wells Jr., P.J. Goodfellow, Frequent mutations of CDKN2 in primary pancreatic adenocarcinomas, Gene Chromosome Cancer 14 (1995) 189±195. [19] C. Caldas, S.A. Hahn, L.T. da Costa, M.S. Redston, M. Schutte, A.B. Seymour, C.L. Weinstein, R.H. Hruban, C.J. Yeo, S.E. Kern, Frequent somatic mutations and homozygous deletions of the p16 (MTS1) gene in pancreatic adenocarcinoma, Nat. Genet. 8 (1994) 27±32. [20] M.S. Redston, C. Caldas, A.B. Seymour, R.H. Hruban, L. da Costa, C.J. Yeo, S.E. Kern, p53 mutations in pancreatic carcinoma and evidence of common involvement of homocopolymer tracts in DNA microdeletions, Cancer Res. 54 (1994) 3025±3033. [21] C. Almoguera, D. Shibata, K. Forrester, J. Martin, N. Arnheim, M. Perucho, Most human carcinomas of the exocrine pancreas contain mutant c-K-ras genes, Cell 53 (1988) 549±554. [22] P. Cairns, L. Mao, A. Merlo, D.J. Lee, D. Schwab, Y. Eby, K. Tokino, P. van der Riet, J.E. Blaugrund, D. Sidransky, Rates of p16 (MTS1) mutations in primary tumor with 9p loss, Science 265 (1994) 415±417. [23] L. Huang, T.L. Goodrow, S.-H. Zhang, A.J. Klein-Szanto, H. Chang, B. Ruggeri, Deletion and mutation analysis of the p16/ MTS1 tumor suppressor gene in human ductal pancreatic cancer reveals a higher frequency of abnormalities in tumorderived cell lines than in primary ductal adenocarcinomas, Cancer Res. 56 (1996) 1137±1141. [24] M. HoÈglund, L. Gorunova, T. Jonson, S. Dawiskiba, A. Andren-Sandberg, G. Stenman, B. Johansson, Cytogenetic and FISH analyses of pancreatic carcinoma reveal breaks in 18q11 with consistent loss of 18q12-qter and frequent gain of 18p, Br. J. Cancer 77 (1998) 1893±1899.
D. Bartsch et al. / Cancer Letters 139 (1999) 43±49 [25] K. Uchida, M. Nagatake, H. Osadya, Y. Yatabe, M. Kondo, T. Mitsudomi, A. Masuda, T. Takahashi, T. Takahashi, Somatic in vivo alterations of the JV18-1 gene al 18q21 in human lung cancers, Cancer Res. 56 (1996) 5583±5585. [26] G.J. Riggins, K.W. Kinzler, B. Vogelstein, S. Thiagalingam, Frequency of Smad gene mutations in human cancers, Cancer Res. 57 (1997) 2578±2580. [27] B. Simon, R. Weinel, M. HoÈhne, J. Watz, J. Schmidt, G.
49
Kortner, R. Arnold, Frequent alteration of the tumor suppressor genes p53 and DCC in human pancreatic carcinoma, Gastroenterology 106 (1994) 1645±1651. [28] M. Nagatake, H. Osada, M. Kondo, K. Uchida, M. Nishio, K. Shimokata, T. Takahashi, Aberrant hypermethylation at the blc-2 locus at 18q21 in human lung cancers, Cancer Res. 56 (1996) 1886±1891.