Targeted Disruption of the 3P12 Gene, Dutt1/Robo1, Predisposes Mice to Lung Adenocarcinomas and Lymphomas with Methylation of the Gene Promoter

[CANCER RESEARCH 64, 6432–6437, September 15, 2004] Targeted Disruption of the 3p12 Gene, Dutt1/Robo1, Predisposes Mice to Lung Adenocarcinomas and Lymphomas with Methylation of the Gene Promoter

Jian Xian,1 Alan Aitchison,1 Linda Bobrow,2 Gerard Corbett,1 Richard Pannell,3 Terence Rabbitts,3 and Pamela Rabbitts1 1Department of Oncology, University of Cambridge, Medical Research Council Centre, Cambridge, United Kingdom; 2Department of Histopathology, Addenbrookes Hospital, Cambridge, United Kingdom; and 3Medical Research Council Laboratory of Molecular Biology, Cambridge, United Kingdom

ABSTRACT within a gene, DUTT1 (deleted in U-twenty-twenty; ref. 7). This intragenic deletion removes a segment encompassing exon 2 of the The DUTT1 gene is located on human chromosome 3, band p12, within gene, creating a mutant form of the protein lacking the first immu- a region of nested homozygous deletions in breast and lung tumors. It is noglobulin domain. The DUTT1 gene was independently isolated as therefore a candidate tumor suppressor gene in humans and is the homologue (ROBO1)oftheDrosophila axonal guidance receptor gene, Round- the human homologue (ROBO1)oftheDrosophila gene Roundabout about. We have shown previously that mice with a targeted homozygous (Robo; ref. 8). Dutt1/Robo1 is a trans-membrane receptor responding deletion within the Dutt1/Robo1 gene generally die at birth due to incom- to the Slit family of ligands and mediates signals that control cell plete lung development: survivors die within the first year of life with migration and cell position in a range of cell types from Drosophila to epithelial bronchial hyperplasia as a common feature. Because Dutt1/ mammals (9). It is not known whether it performs a similar role in Robo1 heterozygous mice develop normally, we have determined their epithelial cells, but abrogation of this function in tumors could explain tumor susceptibility. Mice with a targeted deletion within one Dutt1/Robo1 the contribution of DUTT1/ROBO1 to malignancy because inappro- allele spontaneously develop lymphomas and carcinomas in their second priate migration and incorrect cell position are such defining features year of life with a 3-fold increase in incidence compared with controls: of the malignant phenotype (10). However, although positional clon- invasive lung adenocarcinomas are by far the predominant carcinoma. In addition to the mutant allele, loss of heterozygosity analysis indicates that ing indicates that DUTT1/ROBO1 is a candidate tumor suppressor these tumors retain the structurally normal allele but with substantial gene, like other candidates on 3p, tumor-associated point mutations methylation of the gene’s promoter. Substantial reduction of Dutt1/Robo1 have not been detected (11). If the mechanism of DUTT1/ROBO1 protein expression in tumors is observed by Western blotting and immu- gene inactivation is epigenetic, a functional demonstration is required nohistochemistry. This suggests that Dutt1/Robo1 is a classic tumor sup- to conclusively confirm its involvement in tumor development. Tumor pressor gene requiring inactivation of both alleles to elicit tumorigenesis in formation in mice with targeted inactivation of the candidate gene is these mice. recognized to fulfill this requirement (12). To investigate the possibility that inactivation of this gene predisposes to tumor development, we generated a targeted intragenic INTRODUCTION deletion within the Dutt1/Robo1 gene in the mouse germ line, reca- The identification of genes associated with consistent tumor- pitulating the loss of exon 2 detected in a human small-cell lung specific chromosomal abnormalities has been an effective method of carcinoma (13). Frequent perinatal mortality is observed in mice isolating genes important in cancer development (1). Deletion map- homozygous for the disrupted Dutt1/Robo1 gene. The few surviving ping of chromosome 3 has indicated that somatic genetic loss to the homozygous mice die within 12 months, having developed extensive short arm is a frequent event in a variety of common carcinomas and bronchial hyperplasia. We now demonstrate that heterozygous mice, raised the expectation that tumor suppressor genes will be found with this deletion within only one allele, spontaneously develop located within the minimally deleted regions (2). Lung cancer has lymphomas and carcinomas in their second year of life: invasive received the most attention because many studies have indicated a carcinoma of the lung is by far the predominant epithelial tumor. very high incidence of loss on 3p (91% for small-cell lung carcinoma, Dutt1/Robo1 expression is substantially reduced in these tumors and 95% for squamous cell carcinoma, and 71% for adenocarcinoma) and is accompanied by substantial methylation of the gene’s promoter. We because 3p loss is detectable in low-grade preinvasive lesions, imply- conclude that Dutt1/Robo1 acts as a classic tumor suppressor gene ing an early, possibly initiating, role for 3p loss in lung tumorigenesis with inactivation of both alleles in the development of malignant (3). Several distinct regions along the chromosome have been delin- tumors in these mice. eated, and their boundaries have been reduced by the discovery of homozygous deletions nesting within the region of heterozygous loss (2). A number of genes within these deletions have been examined for MATERIALS AND METHODS frequent point mutations in tumors, but none has shown this classical Observation of Abnormal Mice Phenotypes. The Dutt1/Robo1 mutant feature of tumor suppressor genes (4, 5), although reduced expression mice (mixed C57BL/6 and 129/SV genetic background) were generated by due to promoter hypermethylation is common (6). gene targeting (13). Dutt1/Robo1 genotype was determined by standard South- Our search for a tumor suppressor gene has focused on proximal 3p ern blot hybridization analysis of tail DNA as described previously (13). The and identified three nested homozygous deletions, one of which is mice were housed in a pathogen-free environment, and their care was subject to United Kingdom Home Office regulations. They were observed frequently Received 7/19/04; accepted 8/3/04. and sacrificed on decline in their health or obvious tumor burden. Complete Grant support: Grant G9703123 from the Medical Research Council. T. Rabbitts and necropsy was carried out on all mice. Histopathological examination was R. Pannell are funded by a Medical Research Council grant to the Laboratory of Molecular performed on all tissues found with visible abnormalities in mutant mice and Biology. The costs of publication of this article were defrayed in part by the payment of page controls. Tissues were fixed in buffered 4% paraformaldehyde and then em- charges. This article must therefore be hereby marked advertisement in accordance with bedded in paraffin wax, and the wax blocks were sectioned at 5 to 8 ␮m and 18 U.S.C. Section 1734 solely to indicate this fact. stained with hematoxylin and eosin (H&E). Large tumors were also frozen for Requests for reprints: Pamela Rabbitts, Department of Oncology, University of extraction of DNA, RNA, and protein using standard procedures. Cambridge, Medical Research Council Centre, Hills Road, Cambridge CB2 2QH, United Kingdom. E-mail: [email protected]. Southern Blotting. Genotyping of tumors was performed as described ©2004 American Association for Cancer Research. previously for genotyping of tail DNA (13). 6432

Reverse Transcription-Polymerase Chain Reaction. Total RNA was prepared from mouse tumors, fibroblasts, and normal lung with Trizol reagent (GIBCO/BRL, Gaithersburg, MD). Reverse transcription-polymerase chain reaction (RT-PCR) was performed as described previously (13). Mouse Gapdh mRNA was amplified as a template control. For polymerase chain reaction (PCR) of Dutt1/Robo1, the primers used were as follows: (a) exon 1 forward (5Ј-AGGGATTGACAAGCCTCCGG-3Ј) and exon 2 reverse (5Ј-AGCTAC- CTCCAGCGATGCGT-3Ј); this PCR fragment includes exon 1 and exon 2 (527 bp); and (b) exon 3 forward (5Ј-ATACTACGGGATGACTTCAG-3Ј) and exon 5 reverse (5Ј-CTGGATTTGGGCAGCTCTCC-3Ј); this PCR fragment includes exon 3 to exon 5 (230 bp). For PCR of mouse Gapdh mRNA, the primers used were as follows: forward, 5Ј- TGAACGGATTTGGCCG- TATT-3Ј; and reverse, 5Ј-TGCCGTGAGTGGAGTCATAC-3Ј. The products were visualized by ethidium bromide staining after electrophoresis on agarose gels. Bisulfite Modification and Methylation Analysis. Bisulfite DNA sequencing was performed as described previously (14). One microgram of genomic DNA from normal or tumor tissues in a volume of 50 ␮L was denatured in 0.2 mol/L NaOH for 15 minutes at 37°C. Thirty microliters of 10 mmol/L hydroquinone (Sigma, St. Louis, MO) were added to each sample, mixed gently, and left at room temperature for 5 minutes; 520 ␮L of freshly prepared 3 mol/L sodium bisulfite (pH 5; Sigma) were added; and samples were incubated under mineral oil at 50°C for 16 hours. Modified DNA was purified using Wizard DNA purification resin according to the manufacturer’s instructions (Promega, Madison, WI) and eluted with 60 ␮L of water. Modi- fication was completed by incubation with NaOH to 0.25 mol/L for 15 minutes at 37°C, followed by ethanol precipitation. DNA was resuspended in Tris- EDTA and used immediately or stored at Ϫ20°C. Bisulfite-modified DNA (100 ng) was amplified using the mouse Dutt1/ Robo1 promoter region-specific primers designed by the MethPrimer program. The promoter region was amplified in two parts. CpG island region 1 from position Ϫ793 to Ϫ616 was amplified using forward primer 5Ј-GAAGTTT- GGGTTGTAGTTTG-3Ј and reverse primer 5Ј-AACCCTTATTCCCATCAC- 3Ј. CpG island region 2 from position Ϫ443 to the beginning of exon 1 (position Ϫ310) was amplified using forward primer 5Ј-GAGGAGTATTTTT- TGGGGTG-3Ј and reverse primer 5Ј-TTTCCTACCTCCACAATCC-3Ј. PCR conditions were 95°C for 12 minutes and 30 cycles of 95°C for 50 seconds, 55°C for 50 seconds, and 72°C for 50 seconds. PCR products were purified using QIAquick PCR purification kit (Qiagen). Methylated cytosine residues were identified by direct sequencing of PCR products with the same primers used for PCR. ϩ Ϫ Protein Detection by Western Blotting and Immunohistochemistry. Fig. 1. Tumor incidence and histopathology in Dutt1/Robo1 / mice. Immunohistochemical staining with a polyclonal antibody raised against the COOH-terminal peptide of DUTT1/ROBO1 (CYERGEDNNEELEETES) was The germ-line disruption in mice of other tumor suppressor genes performed as described previously (15). Western blotting, using the same associated with lung cancer development, such as p53, p16INK4a, and antibody, was as described previously (13). Rb, does not show increased incidence of lung cancer development (16). In contrast, 26.5% of the tumors that developed in the Dutt1/ ϩ/Ϫ RESULTS Robo1 mice were adenocarcinomas of the lung (Fig. 1B). Their development was spontaneous and not dependent on any guidance of Tumor Incidence in Dutt1/Robo1 Heterozygous Mice. Dutt1/ genetic damage to the bronchial epithelium. Ϫ Ϫ Robo1 / mice die at birth or within the first year of life, on average Phenotypic Characteristics of Lung Tumors Developing in /at 7 months, preventing evaluation of the effect of the homozygous Dutt1/Robo1؉/؊ Mice. The lung tumors that developed in the Dutt1 deletion within this gene, which results in a truncated protein, on Robo1ϩ/Ϫ mice were all adenocarcinomas, ranged in size from 2 to 10 tumor development (13). To assess whether tumors develop in het- mm2, and generally had an irregular, distorted appearance (Fig. 2A) ϩ Ϫ erozygotes, a cohort of 154 Dutt1/Robo1 / mice was aged together and showed clear invasion (Fig. 2B) with pleiomorphic nuclei (Fig. with 106 wild-type littermates. Eight control mice (7.5%) developed 2C). Histologic subtypes were typical, papillary, and glandular (data tumors (three mice developed lung tumors, and five mice developed not shown). In contrast, the three lung tumors arising in wild-type lymphomas), on average at 20.2 months. The incidence of lung tumors mice were very similar to each other and were small (Ͻ2mm2) with (2.8%) and lymphomas (5%) is within the range expected from the a regular pebble-like appearance (Fig. 2D), well circumscribed (Fig. genetic background of the mice used in these studies.4 In the Dutt1/ 2E), and with near normal nuclear morphology (Fig. 2F). Direct ϩ Ϫ Robo1 / group, 34 mice (22.0%) developed a total of 53 tumors comparison of tumor incidence (Fig. 1A) in wild-type and mutant (Fig. 1A). This is a significant increase in tumor predisposition com- mice masks these important differences in the degree of malignancy pared with wild-type mice; (P Ͻ 0.01, ␹2 test). Other than lympho- observed in the lung tumors. Although invasion was clearly detected mas, the tumors (both benign and malignant) were almost all epithelial in the lung adenocarcinomas that developed in the Dutt1/Robo1ϩ/Ϫ in origin (Fig. 1B). mice (Fig. 2B), secondary deposits were not observed macroscopi- cally. Examples of the other malignant tumors and preinvasive lesions ϩ Ϫ 4 http://tumor.informatics.jax.org/straingrid.html. occurring in the Dutt1/Robo1 / group are illustrated in Fig. 2GϪN. 6433

Fig. 2. Abnormal phenotypes in lung tumors from ϩ Ϫ Dutt1/Robo1 / mice. A, a pair of lungs from a Dutt1/ ϩ Ϫ Robo1 / mouse sacrificed at 18 months. The arrow- head indicates the lung distorted by tumor. B, H&E-stained ϩ Ϫ section of lung adenocarcinoma from a Dutt1/Robo1 / mouse showing stromal invasion (arrowheads). C, higher power view of B showing pleiomorphic nuclei. D, a pair of lungs from a wild-type mouse sacrificed at 18 months; tumor is indicated by the arrow. E, H&E staining of a lung tumor from a wild-type mouse; tumor is indicated by the arrow. F, higher power view of E showing virtually normal nuclei. G, H&E-stained section of lymphoma. HϪN are H&E-stained sections of other mainly epithelial malignant tumors and preinvasive lesions. H, hepatocel- lular carcinoma; I, sweat gland carcinoma; J, renal cell carcinoma; K, thecal cell tumor; L, hepatoctye severe dysplasia; M, ovarian epithelial atypia; N, gallbladder metaplasia (bars,50␮m).

Dutt1/Robo1 Expression in Tumors from Dutt1/Robo1؉/؊ allele (exons 3–5) and specifically from the wild-type allele (exons Mice. Because the mutant allele is an intragenic deletion, transcrip- 1–2; Fig. 3B). Although transcripts were detectable when mouse lung tion could occur from both mutant and wild-type alleles (Fig. 3A and from wild-type mice was the source of template RNA (Fig. 3C, Lane B). Primers for RT-PCR were used to detect transcription from either 1), products were barely detectable when RNA from the Dutt1/

Fig. 3. Partial structure and expression of wild-type and mutant Dutt1/ Robo1 gene. A, wild-type allele exons 1 to 5 and mutant allele with exon 2 replaced by a neo construct (13). B, the transcripts that arise from the mutant and wild-type genes and the priming sites and PCR products that distinguish the transcripts. C, examination of expression of Dutt1/Robo1 alleles in mutant mice and controls. RT-PCR analysis of RNA from lung tumors and lymphomas amplified with primers from exons 1 and 2 (only RNA from the wild-type Dutt1/Robo1 allele can act as a template) and from exons 3 and 5 (RNA from both wild-type and mutant Dutt1/Robo1 alleles can act as a template). Amplification with primers from Gapdh acts as a template control. Lane 1, RNA from normal lung from a wild-type littermate; Lane 2, RNA ϩ Ϫ from uninvolved lung of Dutt1/Robo1 / mice; Lanes 3–6, RNA from lung tumors; Lanes 7–10, RNA from lymphomas; Lane 11, RNA from fibroblasts from Dutt1/Robo1 homozygous mice lacking exon 2; Lane 12, no template control.

6434

Fig. 4. Protein expression in tumors from Dutt1/Robo1 mice. Immunohistochemistry of various tissues using anti-DUTT1/ ϩ Ϫ ROBO1 antibody. A, Dutt1/Robo1 / lung adenocarcinoma showing the tumor with no detectable staining (pink) surrounded by weakly positive (purple-gray) stroma. B, enlargement of the boxed area in A. C, normal mouse bronchus with occasional strongly stained cells in the bronchial epithelium. D, thecal cell tumor, a benign tumor of the ovary, with strongly positive cells. E, hepatocyte dysplasia. F, positively stained hyperplastic sweat glands. G, Western blotting with anti-DUTT1/ROBO1 antibody of the following tissues: Lane 1, newborn mouse brain; Lanes 2–5, ϩ Ϫ lymphomas from Dutt1/Robo1 / mice; Lane 6, NCI-H69 small- cell lung carcinoma cells; and Lanes 7–10, lung tumors from ϩ Ϫ Dutt1/Robo1 / mice.

Robo1ϩ/Ϫ tumors was used as template (Fig. 3C). In parallel with lymphomas, lung adenocarcinomas, and controls was treated with apparent low transcription levels of the Dutt1/Robo1 gene, using sodium bisulfite and amplified with primers flanking two potentially immunohistochemistry, very low levels of Dutt1/Robo1 protein were methylated regions (Fig. 5B), and the PCR products were sequenced. detected in 9 of 12 lung tumors examined that developed in Dutt1/ DNA from normal lung was unmethylated, as was that from a benign Robo1ϩ/Ϫ mice in contrast to surrounding normal tissue (Fig. 4A and tumor control. In contrast, DNA from the six lymphomas and six lung B) and to normal bronchial epithelium (Fig. 4C). Dutt1/Robo1 protein tumors assessed for methylation of the Dutt1/Robo1 promoter region was also virtually undetectable by immunohistochemistry in the three was all found to be methylated (Fig. 5C). Because the antibody to the other carcinomas listed in Fig. 1B (data not shown) and in 12 of 17 COOH-terminal peptide detects both mutant and wild-type protein, lymphomas tested (data not shown), but the benign neoplasms tested we conclude that the very low levels of any Dutt1/Robo1 protein in showed positive staining (Fig. 4DϪF), implying that inactivation was those malignant tumors tested imply that both alleles are methylated associated with the malignant phenotype. The low levels of Dutt1/ in the malignant tumors of Dutt1/Robo1ϩ/Ϫ mice. Robo1 protein in the malignant tumors (lymphomas and lung adenocarcinomas) were confirmed by Western blotting using mouse brain DISCUSSION and a human tumor cell line as positive controls (Fig. 4G). Nonmi- crodissected tumor samples contain variable amounts of normal cells Before the observations described here, only circumstantial evi- (e.g., blood vessels, stroma, lymphocytes, and so forth) that may dence linked the human DUTT1/ROBO1 gene to tumor formation (7, express Dutt1/Robo1 protein, presumably resulting in the trace of 11). In our earlier work, we used gene targeting in mice to recapitulate protein detectable by Western blotting (Fig. 4G, Lanes 4 and 8). a specific human intragenic deletion that occurs within the DUTT1/ Dutt1/Robo1 Gene Structure in Tumors from Dutt1/Robo1؉/؊ ROBO1 gene. The abnormal phenotypes that developed in mice Mice. The RNA and protein expression data suggested that Dutt1/ homozygous for the deletion provided a direct link between this gene Robo1 gene silencing had occurred in the etiology of the malignant and epithelial lung abnormalities in both developing and adult mice tumors. When tumors develop in mice with targeted disruption of a (13). In this study, we demonstrate that tumors arise in the heterozy- tumor suppressor gene, the remaining allele is often lost or inacti- gous mutant mice as a direct consequence of the specific genetic vated, mimicking the behavior of tumor suppressor genes in sponta- disruption of the Dutt1/Robo1 gene, providing the first unequivocal neously arising human tumors. However, Southern blotting of DNA demonstration of a functional association between this candidate from lung tumors and lymphomas showed that the wild-type allele tumor suppressor gene and tumor formation. No other candidate was retained in tumors developing in these Dutt1/Robo1 heterozygous tumor suppressor gene mapping within regions of homozygous dele- mice (Fig. 5A). tion on 3p has been directly linked in this way to spontaneous To determine whether this apparent gene silencing was linked to epithelial tumor development. hypermethylation of the promoter region of the Dutt1/Robo1 gene, we Further support for the Dutt1/Robo1 gene as a tumor suppressor identified the related CpG islands by referring to the location of those gene is provided by comparison with HIC1 (17). After confirmation of associated with the human DUTT1/ROBO1 gene (11). DNA from HIC1 as a tumor suppressor gene, not only by the observation of 6435

Fig. 5. Analysis of Dutt1/Robo1 gene structure ϩ Ϫ and epigenetic modification in Dutt1/Robo1 / mice. A, Southern blot analysis of the Dutt1/Robo1 gene in lung tumors and lymphomas indicates that both the wild-type allele (6.9 kb) and the mutated allele (7.4 kb) are detectable in all tumors examined. Lane 1, normal mouse DNA; Lane 2, mouse ϩ Ϫ tail DNA from a Dutt1/Robo1 / mouse; Lanes ϩ Ϫ 3–7, DNA from Dutt1/Robo1 / lung tumors; ϩ Ϫ Lanes 8–12, DNA from Dutt1/Robo1 / lymphomas. B, a CpG map of the promoter region of the mouse Dutt1/Robo1 gene (accession number NW_0000108). The blue-shaded areas indicate the CpG islands predicted using CpG plot program, ϩ1 indicates the translation start site. The primer pairs for CpG islands 1 and 2 are F1 and R1 and F2 and R2, respectively. C, the sequence within CpG region 1 (Ϫ653 to Ϫ708) before bisulfite treatment (unconverted) and after treatment if the CpG islands are methylated (converted methylated) or unmethylated (converted unmethylated). Chromato- graphs highlighting the different methylation patterns of normal lung (unmethylated), benign tumor (unmethylated), lung tumor (methylated), and lymphoma (methylated). The arrows indicate the locations of nucleotides that can be methylated.

tumor formation in mice heterozygous for a targeted deletion of Hic1, the observed DUTT1/ROBO1 protein expression in these experimen- but also by virtue of methylation of its promoter in these mouse tal mouse tumors and human tumors may be accounted for by reduced tumors, Chen et al. (18) have proposed that the detection of methyl- or absent expression at the early stage of lung tumor development and ation arising in the course of tumor development in genetically mod- increased expression at a later stage of malignancy, possibly perform- ified mice provides a “prime strategy” for confirming tumor suppres- ing some oncogenic function. Transforming growth factor ␤, which sor gene status on genes without somatic mutations as a defining suppresses the early stages of tumorigenesis but promotes tumor feature. progression and invasion, is an example of such a dual function (20). However, although we believe we have provided compelling evi- The expression pattern of E-cadherin in tumors also has similar dence that Dutt1/Robo1 behaves as a tumor suppressor gene in the parallels to DUTT1/ROBO1. Loss of E-CAD expression is associated development of these mouse tumors, we cannot at present extrapolate with promotion of invasion, but subsequent reexpression of E-CAD these observations directly to human lung carcinomas because North- may encourage survival of metastatic deposits. In a study of hetero- ern blotting has shown that the DUTT1/ROBO1 gene is expressed in geneity of expression of this gene in breast cancer, the fluctuation was a variety of human tumor cell lines (19). Furthermore, protein is easily shown to be modulated by variation in methylation of its promoter detectable by immunohistochemistry in tissue sections of primary (21). During the active selection processes that are part of tumor human tumors.5 The function of DUTT1/ROBO1 in normal epithelial development, cells with partial reactivation of tumor suppressor genes cells is not known, making it currently impossible to investigate may be selected because of the growth advantage conferred by reex- whether the DUTT1/ROBO1 protein detected in human tumors is pression of the gene. For genes with pleiomorphic and opposing performing some aberrant function. The apparent disparity between functions at different stages of tumor progression, inactivation by an epigenetic process rather than mutation provides a selective advantage 5 G. Corbett, J. Xian, M. Arends, J. Catto, and P. Rabbitts, unpublished observations. during tumor development because reactivation is possible. 6436

SLIT2, a likely ligand for DUTT1/ROBO1, also appears to have 6. Pfeifer GP, Yoon JH, Liu L, et al. Methylation of the RASSF1A gene in human both tumor-suppressing and -promoting roles. Dallol et al. (22, 23) cancers. Biol Chem 2002;383:907–14. 7. Sundaresan V, Chung G, Heppell-Parton A, et al. Homozygous deletions at 3p12 in consider the SLIT2 gene to be an excellent candidate tumor suppressor breast and lung cancer. Oncogene 1998;17:1723–9. due to its growth-suppressing properties and frequent methylation in 8. Kidd T, Brose K, Mitchell KJ, et al. Roundabout controls axon crossing of the CNS midline and defines a novel subfamily of evolutionarily conserved guidance recep- several human tumors including colorectal, lung, and breast carcino- tors. Cell 1998;92:205–15. mas. In contrast, Wang et al. (24) detected SLIT2 expression in a wide 9. Rao Y, Wong K, Ward M, Jurgensen C, Wu JY. Neuronal migration and molecular variety of tumor cell lines and primary tumors and showed in a series conservation with leukocyte chemotaxis. Genes Dev 2002;16:2973–84. 10. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100:57–70. of colon preinvasive lesions of increasing grade that the degree of 11. Dallol A, Forgacs E, Martinez A, et al. Tumour specific promoter region methylation SLIT2 expression correlated with the grade of malignancy, consistent of the human homologue of the Drosophila Roundabout gene DUTT1 (ROBO1) in with tumor-promoting properties. Clearly, the roles of DUTT1/ human cancers. Oncogene 2002;21:3020–8. 12. Baylin SB, Herman JG. Promoter hypermethylation: can this change alone ever ROBO1 and its ligand(s) in tumorigenesis require further detailed designate true tumor suppressor gene function? J Natl Cancer Inst (Bethesda) 2001; investigation, but these observations, together with this report, suggest 93:664–5. 13. Xian J, Clark KJ, Fordham R, et al. Inadequate lung development and bronchial that a new signaling pathway in tumorigenesis has been identified. hyperplasia in mice with a targeted deletion in the Dutt1/Robo1 gene. Proc Natl Acad Because signaling mediated by this pathway is likely to affect cell Sci USA 2001;98:15062–6. motility, an understanding of the abrogation of this pathway in tu- 14. Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB. Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci morigenesis may provide insight into the mechanisms of tumor inva- USA 1996;93:9821–6. sion and metastatic spread. 15. Clark K, Hammond E, Rabbitts P. Temporal and spatial expression of two isoforms of the Dutt1/Robo1 gene in mouse development. FEBS Lett 2002;523:12–6. 16. Tulverson DA, Jacks T. Modelling lung cancer in mice: similarities and shortcoming. ACKNOWLEDGMENTS Oncogene 1999;18:5318–24. 17. Wales MM, Biel MA, el Deiry W, et al. p53 activates expression of HIC-1, a new Thanks to Anindo Banerjee and Adeline Nicholas for help with preparation candidate tumour suppressor gene on 17p13.3. Nat Med 1995;1:570–7. 18. Chen WY, Zeng X, Carter MG, et al. Heterozygous disruption of Hic1 predisposes of the manuscript, to Mark Arends and Linda Ko Ferrigno for helpful com- mice to a gender-dependent spectrum of malignant tumours. Nat Genet 2003;33:197– ments on the manuscript, and to the Medical Research Council Laboratory of 202. Molecular Biology Biomedical Facilities staff for animal husbandry. 19. Sundaresan V, Roberts I, Bateman A, et al. The DUTT1 gene, a novel NCAM family member, is expressed in developing murine neural tissues and has an unusually broad pattern of expression. Mol Cell Neurosci 1998;11:29–35. REFERENCES 20. Derynck R, Akhurst RJ, Balmain A. TGF-␤ signalling in tumour suppression and cancer progression. Nat Genet 2001;29:117–29. 1. Albertson DG, Collins C, McCormick F, Gray JW. Chromosome aberrations in solid 21. Graff JR, Gabrielson E, Fujii H, Baylin SB, Herman JG. Methylation patterns of the tumours. Nat Genet 2003;34:369–76. E-cadherin 5Ј CpG island are unstable and reflect the dynamic, heterogeneous loss of 2. Zabarovsky ER, Lerman MI, Minna JD. Tumour suppressor genes on chromosome 3p E- cadherin expression during metastatic progression. J Biol Chem 2000;275:2727– involved in the pathogenesis of lung and other cancers. Oncogene 2002;21:6915–35. 32. 3. Wistuba II, Behrens C, Virmani AK, et al. High resolution chromosome 3p allelo- 22. Dallol A, Da Silva NF, Viacava P, et al. A human homologue of the Drosoplila Slit2 typing of human lung cancer and preneoplastic/preinvasive bronchial epithelium gene has tumor suppressor activity and is frequently inactivated in lung and breast reveals multiple, discontinuous sites of 3p allele loss and three regions of frequent cancers. Cancer Res 2002;62:5874–80. breakpoints. Cancer Res 2000;60:1949–60. 23. Dallol A, Morton D, Maher ER, Latif F. SLIT2 axon guidance molecule is frequently 4. Lerman MI, Minna JD. The 630kb lung cancer homozygous deletion region on human inactivated in colorectal cancer and suppresses growth of colorectal carcinoma cells. chromosome 3p21.3: identification and evaluation of the resident candidate tumor Cancer Res 2003;63:1054–8. suppressor genes. Cancer Res 2000;60:6116–33. 24. Wang B, Xiao Y, Ding B, et al. Induction of tumour angiogenesis by Slit-Robo 5. Huebner K. Tumor suppressors on 3p: a neoclassic quartet. Proc Natl Acad Sci USA signalling and inhibition of cancer growth by blocking Robo activity. Cancer Cell 2001;98:14763–5. 2003;4:19–29.

6437

Downloaded from cancerres.aacrjournals.org on October 3, 2021. © 2004 American Association for Cancer Research. Targeted Disruption of the 3p12 Gene, Dutt1/Robo1, Predisposes Mice to Lung Adenocarcinomas and Lymphomas with Methylation of the Gene Promoter

Jian Xian, Alan Aitchison, Linda Bobrow, et al.

Cancer Res 2004;64:6432-6437.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/64/18/6432

Cited articles This article cites 24 articles, 9 of which you can access for free at: http://cancerres.aacrjournals.org/content/64/18/6432.full#ref-list-1

Citing articles This article has been cited by 8 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/64/18/6432.full#related-urls

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

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

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/64/18/6432. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.