A transgenic mouse model of metastatic carcinoma involving transdifferentiation of a gastric epithelial lineage progenitor to a neuroendocrine

Andrew J. Syder*†, Sherif M. Karam†‡§, Jason C. Mills*¶, Joseph E. Ippolito*, Habib R. Ansari§, Vidya Farook§, and Jeffrey I. Gordon*ʈ

Departments of *Molecular Biology and Pharmacology and ¶Pathology and Immunology, Washington University School of Medicine, St. Louis, MO 63110; ‡Department of Anatomy, Faculty of Medicine and Health Sciences, United Arab Emirates University, Al-Ain 17666, United Arab Emirates; and §Department of Anatomy, Faculty of Medicine, Kuwait University, Safat 13110, Kuwait

Edited by Bert Vogelstein, The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Baltimore, MD, and approved January 28, 2004 (received for review December 2, 2003) neuroendocrine cancers (NECs) arise in various endoderm- cinoids’’ rarely metastasize, whereas small cell carcinomas are derived epithelia, have diverse morphologic features, exhibit a almost universally fatal and have commonly metastasized by the wide range of growth , and generally have obscure time of initial diagnosis (5–7). cellular origins and ill-defined molecular mediators of initiation Because NECs are typically discovered long after the initiating and progression. We describe a transgenic mouse model of met- event, the features of their cell of origin, and the molecular astatic gastric cancer initiated by expressing simian virus 40 large pathways they follow during their progression, are poorly un- tumor antigen (SV40 TAg), under control of regulatory elements derstood. Thus, it would be advantageous if NE carcinogenesis from the mouse Atp4b , in the progenitors of acid-producing could be modeled in mice. Unfortunately, there are only a few parietal cells. Parietal cells normally do not express endocrine or reported mouse models where there is reproducible develop- neural features, and Atp4b-Cre bitransgenic mice with a Cre re- ment of NECs from a known cell of origin. For example, NE cells porter confirmed that the Atp4b regulatory elements are not active represent one of the three lineages that constitute the normal in gastric enteroendocrine cells. GeneChip analyses were per- mouse prostate epithelium. Forced expression of simian virus 40 CELL BIOLOGY formed on laser capture microdissected SV40 TAg-expressing cells large tumor antigen (SV40 TAg) in a subset of these cells in preinvasive foci and invasive tumors. that distinguish produces a metastatic NEC with a rapid and stereotypic pattern invasive from preinvasive cells were then hierarchically clustered of evolution in multiple pedigrees of transgenic mice (8). Intra- with DNA microarray datasets obtained from human lung and tracheal infection of mice, harboring conditional null alleles of gastric cancers. The results, combined with immunohistochemical Rb1 and p53, with an engineered adeno-Cre virus leads to and electron microscopy studies of Apt4b-SV40 TAg stomachs, somatic inactivation of both genes and development of a met- revealed that progression to invasion was associated with trans- astatic small cell-type NEC, although the cell of origin in this differentiation of parietal cell progenitors to a neuroendocrine model has not been reported (9). phenotype, and that invasive cells shared molecular features with In the present report, we describe development of a metastatic NECs arising in the human pulmonary epithelium, including tran- NEC from a committed gastric epithelial progenitor that nor- scription factors that normally regulate differentiation of various mally gives rise to a lineage that expresses no NE features: the endocrine lineages and maintain neural progenitors in an undif- acid-producing parietal cell (PC) (10). This transgenic mouse ferentiated state. The 399 mouse genes identified as regulated model is informative because it illustrates how NECs can arise during acquisition of an invasive phenotype and concomitant by transdifferentiation of a progenitor normally dedicated to a neuroendocrine transdifferentiation, plus their human orthologs non-NE fate. associated with lung NECs, provide a foundation for molecular Gastric cancers, the second leading cause of cancer deaths classification of NECs arising in other tissues and for genetic tests worldwide (11, 12), are predominantly adenocarcinomas with of the molecular mechanisms underlying NEC pathogenesis. two general, occasionally overlapping, phenotypes: those with an intestine-like morphology (‘‘intestinal-type’’) and those that neuroendocrine carcinogenesis ͉ metastatic gastric cancer ͉ functional infiltrate diffusely (‘‘diffuse-type’’) (13). NE tumors have tradi- Ϸ genomic studies ͉ mouse͞human epithelial cancers tionally been considered uncommon in the stomach ( 1% of all gastric cancers; e.g., ref. 14). However, during the last three decades, the prevalence of NE tumors of the carcinoid-type has umors with neuroendocrine (NE) features arise in tissues increased dramatically, either because of an increased incidence Tlocated throughout the human body. They exhibit diverse of hyperplasia of enteroendocrine cells in response to wide- morphologic patterns and range from indolent to highly aggres- spread use of long-term acid-suppressive therapy and͞or be- sive (1). This diversity has impeded development of a consensus cause more pathologists are recognizing this entity (15). As many system for their classification (2, 3). Neuroendocrine cancers as 10% of the tumors diagnosed as intestinal- or diffuse-type may (NECs) that have a neuroectodermal origin (‘‘neural-type,’’ such have NE features (16, 17). as those that arise in the adrenal medulla) usually express The mouse is a good model for studying human gastric neurofilaments, whereas those that have an epithelial, most tumorigenesis because a considerable amount of information is commonly endodermal, origin express cytokeratins. Epithelial- type NECs occurring in organs such as the lung, prostate, or gastrointestinal tract are thought not to have a neuroectodermal This paper was submitted directly (Track II) to the PNAS office. origin (1, 4). Abbreviations: NE, neuroendocrine; NEC, NE cancer; SV40 TAg, simian virus 40 large tumor In the lung, NECs account for up to a quarter of epithelial antigen; PC, parietal cell; pPC, pre-PC; iGC, invasive gastric cancer cell; mGC, metastatic tumors. Travis and coworkers (5) have developed a system to gastric cancer cell; EM, electron microscopy; CgA, . classify these NECs on the basis of their morphological pheno- †A.J.S. and S.M.K. contributed equally to this work. type. Such distinctions have important therapeutic and prognos- ʈTo whom correspondence should be addressed. E-mail: [email protected]. tic implications: relatively indolent NE tumors known as ‘‘car- © 2004 by The National Academy of Sciences of the USA

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0307983101 PNAS ͉ March 30, 2004 ͉ vol. 101 ͉ no. 13 ͉ 4471–4476 Downloaded by guest on September 24, 2021 known about the morphologic features of its epithelial progen- itors (18–23). As in , the mucosa contains innumerable tubular invaginations known as gastric units (24). The epithelium lining these units is continuously renewed. Renewal is driven by multipotent stem cells functionally anchored in a niche posi- tioned in the middle of each unit (the ‘‘isthmus’’; see Fig. 4A, which is published as supporting information on the PNAS web site). In the corpus region of the stomach (Fig. 4) this stem cell gives rise to five descendant lineages. Three lineages constitute the majority of the Ϸ200 cells that populate the gastric unit. Mucus-producing pit cells differentiate as they undergo a rapid upward migration from the isthmus through the so-called pit region to the surface epithelium, where they are removed by exfoliation and͞or apoptosis (turnover time, 3 days; ref. 19). Zymogenic cells differentiate during a slower downward migra- tion from the isthmus to the base of the unit (turnover time Ϸ190 days; refs. 20 and 25). Acid-producing PCs are unique in that they differentiate within the isthmus. They arise from a com- mitted preparietal cell (pPC) progenitor with well described morphologic features and then undergo a bipolar migration to the pit and base (10, 22). The enteroendocrine and caveolated lineages represent the remaining, albeit much less prevalent, cell types (21). Transcriptional regulatory elements from the mouse gene encoding the noncatalytic ␤-subunit of Hϩ,Kϩ-ATPase (Atp4b) have been used to selectively deliver a variety of gene products to pPCs (e.g., refs. 26–28). Our earlier studies of 4- to 11-week- old mice containing a transgene consisting of these regulatory elements linked to SV40 TAg revealed that expression of the viral oncoprotein produces a 50- to 70-fold expansion of pPCs and an accompanying block in their differentiation to mature PCs (27). We now show that members of two pedigrees of these Fig. 1. Invasive gastric cancer in Atp4b-SV40 TAg mice. (A) Hematoxylin and transgenic mice develop metastatic gastric cancer associated eosin (H&E)-stained section from a 44-week-old transgenic mouse showing with the transdifferentiation of pPCs to a NE phenotype. This invasion of tumor into the muscle layers of the stomach and into lymph͞ evolution was defined by GeneChip profiling of laser-capture vascular spaces (arrows highlight nests and cords of tumor cells expanding vessels). (B) H&E-stained section of a liver metastasis containing a solid mass of microdissected cell populations and was confirmed by multilabel small cells with exceedingly scant cytoplasm. (C–F) Serial sections from an immunohistochemistry, transmission electron microscopy (EM), 11-month-old Atp4b-SV40 TAg stomach. (C) Low-magnification image of an as well as by hierarchical cluster analyses of our GeneChip H&E-stained section demonstrating a tumor spanning the entire mucosa. datasets with publicly available DNA microarray datasets gen- Tumor cells (lower left, asterisk; compare with tubuloglandular profiles in the erated from a variety of human NECs. upper right) form a nearly solid mass and have scant cytoplasm. (D) Higher- magnification view of the area boxed in C. Note that tumor cells do not form Materials and Methods glands: instead, they are organized into cords, loose ribbons, and nests Details of the methods used to (i) generate and maintain (insulae). (E) Adjacent serial section demonstrating that hyperplastic pPCs in pedigrees of Atp4b-SV40 TAg and Atp4b-Cre transgenic mice, glandular profiles exhibit more intense staining with antibodies to SV40 TAg (purple) compared with invasive gastric cancer cells (iGCs). (F) Adjacent serial (ii) conduct multilabel immunohistochemical and transmission section demonstrating Atp4b labeling (purple) of cells in glands and no detect- EM studies, (iii) perform GeneChip analysis of navigated laser- able staining in iGCs. (Bars: 25 ␮minA,50␮minB and D–F, 120 ␮minC.) capture microdissection cell populations harvested from Atp4p- SV40 TAg animals and their normal littermates, and (iv) un- dertake hierarchical clustering studies of DNA microarray published as supporting information on the PNAS web site). datasets are all provided in Supporting Materials and Methods, There was focal dysplasia in some regions containing the ex- which is published as supporting information on the PNAS web panded pPC population. Dysplastic cells exhibited nuclear het- site. erogeneity and abnormal polarity and were located in gland Results and Discussion profiles featuring cellular stratification and occasional ‘‘cribri- form’’ characteristics. At 38 weeks, focally invasive cancer was Development of Metastatic Gastric Cancer in Mice That Express SV40 identified in Ϸ10% of animals. These lesions were typically TAg in Their pPCs. Given that expression of SV40 TAg under the located at the apex of the gastric units (i.e., near the lumen) and control of nucleotides Ϫ1035 to ϩ24 of the mouse Apt4b gene showed loss of glandular architecture. Invasive cells were orga- produces a 50- to 70-fold increase in the number of normally rare pPCs by 12 weeks of age (Fig. 4 B and C), we postulated that nized in loose trabeculae or ribbons, were smaller than hyper- continued expression of the viral oncoprotein would predispose plastic pPCs, and had increased nuclear-to-cytoplasmic ratios to development of carcinoma. Therefore, we examined trans- with granular (Fig. 4F). Carcinoma encompassing the genic and normal littermates from two pedigrees as they aged full thickness of the stomach occurred in 60% of mice by 48 ϭ from 12 to 52 weeks. Transgenic animals from both pedigrees weeks (n 16). Sheets of small relatively monotonous tumor showed the same phenotype, indicating it was not dependent on cells infiltrated the submucosa, often accompanied by histologic the site of chromosomal insertion of their transgene. evidence of lymphatic–vascular invasion (Fig. 1 A, C, and D). By Between 12 and 32 weeks of age, pPC hyperplasia produced 1 year of age, all animals studied (n ϭ 53) had invasive gastric progressive epithelial thickening as well as cyst formation in all cancer. Grossly visible nodular masses projected from the serosal animals studied (n ϭ 32) (Fig. 4 D and E and Table 1, which is surface of the stomachs of 60% of these mice, and there were

4472 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0307983101 Syder et al. Downloaded by guest on September 24, 2021 associated local lymph node metastases. Macroscopic (Ն2-mm) sion across different species and different GeneChip platforms, hepatic metastases were present in 20% of animals and were we developed a two-step process that involved initial extraction composed of solid sheets of small cells with high nuclear-to- of a molecular signature of genes that clearly differentiate the cytoplasmic ratios (Fig. 1B). None of the age- and gender- iGC and mGC populations (invasive cancers) from SV40 TAg- matched normal FVB͞N mice in our colony developed gastric expressing pPCs and gastric epithelium from normal mice. We lesions (n ϭ 100). then used the human orthologs of these mouse genes to cluster a previously published group of 203 GeneChip profiles of lung Functional Genomics Analysis Reveals That Development of Invasive carcinomas with diverse phenotypes, as well as normal lung Cancer Is Associated with Expression of a NE Phenotype. Multilabel tissue (30). The first step was performed by clustering the pPC, immunohistochemical studies of 48-week-old Atp4b-SV40 TAg iGC, and mGC GeneChip datasets described above with three mice revealed that whereas hyperplastic pPCs expressed SV40 GeneChip datasets obtained from normal adult total gastric unit TAg and the normal PC-specific marker ␤-subunit of Hϩ,Kϩ- epithelium (harvested by laser-capture microdissection from the ATPase (Atp4b), invasive tumor cells in the stomach and liver corpus region of the stomach; ref. 31). A cluster of 399 unique were weakly positive for SV40 TAg and had no detectable Atp4b probe sets was identified representing transcripts that differen- (Fig. 1 E and F). Intrigued by these differences, as well as by the tiated iGCs and mGCs from their presumptive pPC precursors morphologic differences observed between hyperplastic pPCs and members of all other normal gastric epithelial lineages (see and invasive cells, we conducted a broad survey of their molec- boxed area with an asterisk in Fig. 5A and Table 5, which are ular properties. published as supporting information on the PNAS web site). For A navigated form of laser-capture microdissection (see Sup- the second step, we performed a hierarchical cluster analysis that porting Materials and Methods) was used to harvest (i) hyper- was limited to the human orthologs of these mouse invasive plastic SV40 TAg- and Atp4b-positive cells with clear pPC gastric cancer-associated genes to determine whether this subset morphology, (ii) weakly SV40 TAg-positive, Atp4b-negative, of orthologs could also help differentiate human lung tumors locally invasive gastric cancer cells (iGCs), and (iii) the small from normal pulmonary epithelium, and whether they are monotonous metastatic gastric cancer cells (mGCs) in the liver preferentially expressed in particular types of human lung car- (10,000 cells per mouse per population; n ϭ two 48-week-old cinoma. mice). Each RNA sample from each mouse was used to generate In this cross-platform analysis, we first matched the 399 mouse complementary RNA targets that were subsequently hybridized U74Av2 GeneChip probe set identifiers to human U95Av2 to Affymetrix MG-U74Av2 GeneChips (all Affymetrix CEL files GeneChip probe set identifiers by using UniGene Gene Symbols CELL BIOLOGY are available at http:͞͞gordonlab.wustl.edu). After designating corresponding to their respective probe sets. Of the 399 genes, the pPC GeneChips as a baseline, we used Affymetrix MAS 199 were found to be common to both the human and mouse version 5.0 software to select for transcripts that had received a arrays (Table 6, which is published as supporting information on ‘‘Present’’ call and an ‘‘Increase’’ or ‘‘Decrease’’ call in duplicate the PNAS web site). This 199-member list was then used as a comparisons, and had an average signal log ratio Ն1 (average filter for hierarchical clustering of U95Av2 GeneChips repre- fold change Ն2) across duplicate arrays. A total of 629 probe sets senting NE tumors of the carcinoid type (n ϭ 20), NE small cell (592 unique UniGene entries) satisfied these selection criteria lung carcinoma (n ϭ 6), squamous cell carcinoma (n ϭ 21), when comparing iGCs to pPCs (Table 2, which is published as adenocarcinomas (n ϭ 127), and normal lung control specimens supporting information on the PNAS web site). Our comparison (n ϭ 17). All comparisons were performed using the hierarchical of mGCs and pPC yielded 794 probe sets representing 752 clustering algorithm incorporated in DCHIP software (32). UniGene entries (Table 3, which is published as supporting Among the transcripts detected by these 199 probe sets, 48 information on the PNAS web site), whereas a comparison of the (24%) were enriched in carcinoids (green dendrogram labeled iGC versus mGC populations revealed only 73 differences LCa-1 in Fig. 5B). Eight of these 48 transcripts were detected (Table 4, which is published as supporting information on the almost exclusively in carcinoids, where they had uniformly high PNAS web site). levels of expression (red in Fig. 5B), and low to undetectable The majority of tumors arising in the human stomach have expression (green-black) in the 171 other tumors or normal lung clear features of adenocarcinoma. In the intestinal-type, mucus- controls. They included mRNAs encoding AMPA2 inotropic secreting cells are organized in a glandular fashion (tubes or glutamate (GRIA2), dopa decarboxylase (DDC), trans- papillae). In the diffuse type, glandular architecture is less thyretin (TTR), proprotein convertase 1 (PCSK1), secretogra- prominent but tumor cells are still mucus filled and thought to nin II (SCG2), chromogranin A (CGA), chromogranin B be derived from glandular elements that have down-regulated (CGB), and secretory granule neuroendocrine 1 expression of genes that mediate cell–cell adhesion (13, 29). We (SGNE1). had expected that forced expression of SV40 TAg in a normal The 199-gene filter identified five other tumor clusters: (i)12 gastric epithelial progenitor would lead to expansion of glandu- pulmonary adenocarcinomas that had NE characteristics [den- lar elements, and thus we were not surprised that pPC hyper- drogram labeled LCa-2 in Fig. 5B; genes responsible for defining plasia initially produced tubuloglandular structures. However, this cluster included calcitonin (CALCA), as well as those whose the invasive tumors (iGC and mGC populations) that arose in expression was enriched in carcinoids (e.g., DDC, PCSK1)]; (ii) areas of glandular hyperplasia (Fig. 1 C and D) did not have the a group of pure squamous cell carcinomas (LCa-3); (iii) a group typical mucus-glandular features of adenocarcinomas that arise of pure small cell lung carcinomas (SCLC; LCa-4); (iv) a group in either the corpus or antral regions of the human stomach. of squamous cell cancers and adenocarcinomas (LCa-5); and (v) Therefore, we first compared the profiles of the a group of 17 adenocarcinomas and one squamous cell carci- invasive mouse gastric cell populations (iGCs plus mGCs) with noma (LCa-6) that had high levels of expression of 9 genes that profiles documented in cancers arising in another endoderm- function predominantly in the synthesis, maintenance, or remod- derived tissue where a more diverse mixture of epithelial tumors eling of extracellular matrix (ECM) components [e.g., lysyl commonly form: the human lung. oxidase (LOX), fibronectin (FN1), osteonectin (SPARC), and multiple collagens]. (Additional information concerning clus- Hierarchical Clustering with DNA Microarray Datasets Obtained from tered lung samples is provided in Fig. 5C.) Human Pulmonary Neoplasms. A large number of global gene Bhattacharjee and colleagues (30) provided only survival data expression profiles for human lung carcinomas are available in for adenocarcinomas in their study. Using this information, the public domain. Because we needed to compare gene expres- Kaplan–Meier analysis (WINSTAT EXCEL plug-in software from

Syder et al. PNAS ͉ March 30, 2004 ͉ vol. 101 ͉ no. 13 ͉ 4473 Downloaded by guest on September 24, 2021 Fig. 2. The transition from hyperplasia to neoplasia is characterized by increased expression of NE markers and loss of a PC marker. (A) Multilabel immunohistochemical study of serial stomach sections from an 11-month-old Atp4b-SV40 TAg mouse. Chromogranin A (CgA, green cytoplasm), a marker for enteroendocrine cells, is expressed in cells that have weak SV40 TAg staining (red, nuclear; e.g., arrowheads). Arrows point to larger, more in- tensely staining enteroendocrine cells that are SV40 TAg-negative. (B) Adja- cent section demonstrating cells with prominent dopa decarboxylase (Ddc, green) staining, and weak SV40 TAg expression. Note the sharp demarcation between the clusters of strongly and weakly staining SV40 TAg-expressing cells. (C) The ␤-subunit of Hϩ,Kϩ-ATPase (Atp4b; green) is limited to the intensely SV40 TAg-positive pPC population. (Bar: 25 ␮m.)

Fig. 3. Invasive gastric carcinoma associated with transdifferentiation of A-Prompt, Lehigh Valley, PA) disclosed that, compared with all pPCs to a NE cell phenotype. (A) Multilabel study showing that SV40 TAg (red) other tumor samples in the dataset, only the LCa-2 cluster colocalizes with Dolichos biflorus agglutinin (DBA) binding (blue; a PC lineage (adenocarcinomas with NE features) had a significantly less marker) but does not localize with the enteroendocrine cell marker CgA favorable outcome (P ϭ 0.002). (green) in the stomach of a 9-week-old Atp4b-SV40 TAg transgenic mouse. (B Together, these results demonstrate that the great majority of and C) Section from the stomach of a 7-week-old Atp4b-Cre, R26R bitrans- genes that differentiate the invasive gastric carcinoma cell genic mouse. B shows a combined bright-field and immunofluorescence image demonstrating that none of the ␤-galactosidase (␤-gal)-expressing cells populations (iGCs, mGCs) in Atp4b-SV40 TAg mice from pPCs (blue, due to Cre-mediated recombination of R26R) belong to the CgA- and other epithelial lineages in the normal mouse stomach also positive (green) enteroendocrine lineage. C shows the section presented in B help segregate human lung carcinomas that have NE features. viewed with immunofluorescence. pPCs that appear blue in B appear red after staining with DBA. CgA-positive enteroendocrine cells remain green. Nuclei Immunohistochemical and Transmission EM Evidence That pPCs Un- are stained dark blue with bisbenzimide. (D and E) Transmission EM of a region dergo Transdifferentiation to iGCs with NE Phenotypes. To confirm of focal neoplasia in the stomach of a 14-month-old Atp4b-SV40 TAg mouse. that invasive gastric cancer cells acquire NE gene expression, D presents a high-power view of four cells with a range of morphological follow-up immunohistochemical studies were performed using features. Cell 1 is representative of a normal pPC on the basis of the presence of multiple large mitochondria (M), tubulovesicular structures (tv), and long stomachs from 48-week-old Atp4b-SV40 TAg mice and antibod- microvilli (L). Cell 4 is representative of a normal enteroendocrine cell on the ies directed against three well established NE biomarkers: dopa basis of the presence of many electron-dense secretory granules (eg). Cells 2 decarboxylase (Ddc; increased 50-fold in the GeneChip com- and 3 have a mixture of large mitochondria, tubular vesicular structures, and parison of iGC vs. pPC populations and highly expressed in lung electron-dense secretory granules, indicating transdifferentiation from a pPC carcinoids), chromogranin A (CgA; increased 20-fold and also toward a NE phenotype. E shows a lower-magnification view at the border of expressed at high levels in carcinoids), and tryptophan hydrox- a focus of neoplasia. Cells with the pPC phenotype located in the upper left ylase (Tph1; increased 830-fold). All three were prom- lack electron-dense secretory granules; enteroendocrine cells at the bottom right have abundant granules. The remaining cells have a variable number of inently expressed in areas of focal invasive neoplasia but not in ␮ adjacent normal-appearing epithelium or in hyperplastic epithe- secretory granules, indicating that they are acquiring NE features. (Bars: 20 m in A–C; original magnifications: ϫ15,600 in D, ϫ3,200 in E.) lium containing SV40 TAg and Atp4b-positive pPCs (e.g., Fig. 2). These results establish that NE biomarkers are expressed in we generated two pedigrees of FVB͞N transgenic mice that neoplastic SV40 TAg-positive cells. At least two scenarios could contained these regulatory elements linked to Cre recombinase. account for this observation: initiated pPCs undergo transdif- Recombination was scored in bitransgenic mice possessing the ferentiation to a NE phenotype, or SV40 TAg expression is Cre-containing transgene and the R26R Cre reporter [R26R initiated, or becomes more pronounced, in members of the contains a floxed ‘‘stop sequence’’ located directly in front of an gastric stem cell’s descendant enteroendocrine lineage as ani- ␤ mals age and invasive cancer appears. The latter scenario implies Escherichia coli lacZ ( -galactosidase) ORF deposited in the that nucleotides Ϫ1035 to ϩ24 of Atp4b are active in enteroen- otherwise ubiquitously expressed Rosa26 (33). Expression docrine cells. One piece of evidence arguing against such a of Cre induces lacZ expression through excision of the stop scenario is that our immunohistochemical studies did not detect sequence]. coexpression of SV40 TAg and CgA in serially sectioned stom- Analysis of multiple tissues harvested from Atp4b-Cre, R26R ␤ achs harvested from young (9-week-old) transgenic animals, bitransgenic mice confirmed that E. coli -galactosidase expres- whereas there was invariable coexpression of SV40 TAg and sion [detected with 5-bromo-4-chloro-3-indolyl ␤-D-galactoside PC-specific glycans recognized by Dolichos biflorus agglutinin (X-Gal)] was limited to PCs, and not detectable in any CgA- (Fig. 3A). positive or Ddc-positive gastric enteroendocrine cells (e.g., Fig. To further rule out the possibility that nucleotides Ϫ1035 to 3 B and C). (X-Gal staining of tissue sections also demonstrated ϩ24 of Atp4b may be active (or become active) in non-PC types, that Cre was not expressed in adult kidney, small intestinal, and

4474 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0307983101 Syder et al. Downloaded by guest on September 24, 2021 colonic epithelial lineages, or in liver, spleen, and lung; data not of NECs (i.e., poorly differentiated with high nuclear-to- shown.) cytoplasmic ratios, clustering together to form focal (or multi- These results support the scenario that SV40 TAg-expressing focal) solid (nonglandular) lesions (e.g., ref. 49), the results of pPCs undergo transdifferentiation to a NE phenotype as they global gene expression profiling of SV40 TAg-expressing tumors become invasive. Subsequent transmission EM studies docu- originating in cells that are not neurons or members of a NE mented this transdifferentiation (Fig. 3 D and E). Normal pPCs lineage have not been deposited in publicly accessible databases. are characterized by multiple large mitochondria and rudimen- If these datasets are forthcoming, the 399-element filter de- tary canaliculi within their cytoplasm, whereas enteroendocrine scribed above may be useful for cluster analyses to determine cells have distinct, electron-dense cytoplasmic secretory gran- whether SV40 TAg commonly produces carcinomas with the ules. Most cells in areas of invasive neoplasia (iGCs) contained molecular properties of NECs. both of these features. The relative proportion of expressed pPC versus endocrine features varied between cells: i.e., within an NE Features in Human Gastric Cancers. If acquisition of NE markers area of focal neoplasia, a gradient of mixed-feature cells was is more common than generally thought in gastric epithelial evident. This spectrum in aggregate captures the transdifferen- carcinomas, it could have important implications for diagnosis tiation events that pPCs undergo to a NE-like cell. and͞or treatment. Currently, the incidence, classification, and prognosis of NECs are far better described in the human lung Mechanistic Considerations. We sought to identify factors that may than in the stomach. Older studies have concluded that gastric be potential regulators͞mediators of this transdifferentiation. As carcinoids represent only 1% of stomach cancers, and 3% of all noted above, transcripts recognized by 399 genes on the mouse carcinoids, whereas a more recent analysis concluded that gastric U74Av2 GeneChips were enriched in the iGC͞mGC populations carcinoids represent up to 10–30% of all carcinoids (15). The compared with hyperplastic pPCs and all normal gastric epithe- paucity of information about the incidence of NE features in lial cell types. (GO) terms were used to extract gastric (or intestinal) adenocarcinomas may reflect the absence the 32 genes that encode transcription factors and DNA-binding of a robust set of molecular biomarkers for systematic analysis. proteins from the dataset (Table 7, which is published as With this in mind, the signature of 399 transcripts enriched in supporting information on the PNAS web site). mouse mGC and iGC populations was used to probe human Some of the extracted transcription factors are expressed in gastric cancer cDNA microarray datasets (http:͞͞genome- pancreatic endocrine cells. For example, neurogenic differenti- www5.stanford.edu) originally published by Leung et al. (53). ation 1 (NeuroD1), paired box gene 6 (Pax6), and Nkx2.2 are key These datasets were generated from 90 primary gastric adeno- CELL BIOLOGY regulators of insulin gene expression (34). Studies of mice carcinomas with intestinal, diffuse, mixed, and indeterminate deficient in NeuroD1 and Pax6 have established that these phenotypes, plus 14 lymph node metastases from 14 of the 90 factors are necessary for development of alpha, beta, and delta primary cancers, and were accompanied by datasets from 22 cells in pancreatic islets (35, 36). NeuroD1, which interacts with nonneoplastic gastric mucosal samples. The 399 transcripts were the promoter region of Nkx2.2 (37), is also expressed in differ- matched to 341 cDNAs on the microarray (Table 8, which is entiated small intestinal enteroendocrine cells (38). published as supporting information on the PNAS web site). In addition to transcription factors known to effect endocrine Hierarchical clustering with DCHIP and this 341-member panel cell differentiation, other factors were identified that normally disclosed distinctive molecular signatures in a subset of gastric maintain neuronal precursors in a relatively undifferentiated carcinomas (Fig. 6, which is published as supporting information state. SRY-box 2 (Sox2), a member of the SoxB1 family, on the PNAS web site). The GCa-1 cluster was enriched for maintains neural progenitor characteristics in the CNS: inhibi- genes associated with neural͞endocrine cells: e.g., SOX2, DDC tion of Sox2 expression in these progenitors causes them to exit (dopa decarboxylase), GAD1 (glutamic acid decarboxylase), the cell cycle and differentiate (39). Another transcription factor SNAP25 (synaptosomal associated protein), SYT1 (synaptotag- in the list, Hes-related with YRPW motif 1 (Hey1), also main- min 1), ENO2 (neuronal enolase), and RTN1 (reticulon 1). tains the neural precursor phenotype in brain: it negatively Tissues in the GCa-2 category were enriched in genes involved regulates basic helix–loop–helix (bHLH) transcription factors, in extracellular matrix remodeling [e.g., LOX (lysyl oxidase) and such as mouse achaete scute homolog 1 (mAsh1) and mouse multiple collagens], paralleling the molecular signature of the atonal homolog 3 (Math3) that induce neuronal differentiation extracellular matrix-enriched lung cancers shown in Fig. 5B). (40). (The bHLH transcription factor mAsh1 is essential for Using clinical data provided by Leung and coworkers, (http:͞͞ differentiation of NE cell lineages; mAsh1Ϫ/Ϫ mice die at birth genome-www.stanford.edu͞Gastric࿝Cancer), we performed and lack NE cells in their lungs and thyroid (41–43)). Thus, the three individual Kaplan–Meier analyses comparing each cluster induction of Sox2 and Hey1 as pPCs transdifferentiate to NE-like (GCa-1, GCa-2, or GCa-3) to all other gastric carcinomas in the iGCs may be important in maintaining their relatively undiffer- dataset. No statistically significant differences (i.e., P Ͻ 0.05) entiated͞invasive growth properties. were observed in patient survival. In addition, our hierarchical Both p53 and Rb are inactivated in up to 90% of small cell lung clustering of the microarray datasets without the 341-gene filter cancers (SCLCs; ref. 44). As noted in the Introduction, condi- failed to segregate gastric cancers on the basis of their histopa- tional inactivation of pRB and p53 by adenovirus-mediated thology (diffuse, intestinal, mixed, or indeterminate; data not delivery of Cre to the pulmonary epithelium of mice results in shown). Intriguingly, while no correlation was observed between metastatic small cell type lung NECs. This is not an effect of tumor stage and the clustering of a sample into the GCa-1 or adenoviral infection per se: when this approach was used to GCa-2 categories, all samples in the GCa-3 category were engineer over-expression of K-ras, an oncogene associated with categorized in the American Joint Committee on Cancer lung cancer, the resulting tumors had the histopathologic fea- (AJCC) classification scheme as IIIA or IIIB (invasive gastric tures of adenomas (papillary subtype) and adenocarcinomas, cancers locally metastatic to regional lymph nodes). This cluster rather than SCLCs (45). Because SV40 TAg is known to represented 16% (7͞43) of the stage IIIA and IIIB samples. inactivate p53 and pRB, an obvious question is whether NE Genes in this cluster encode membrane proteins involved in features are commonly expressed in other epithelial neoplasms cytoskeletal reorganization, cell motility, adhesion, and antiapo- produced in other mouse models by expression of this viral ptotic functions [e.g., ena͞VASP-like protein (EVL; ref. 54), oncoprotein [e.g., stomach (46), prostate (8, 47, 48), thyroid (49), Wiskott–Aldrich syndrome protein interacting protein mammary gland (50), lung (51), and retina (52)]. While tumor (WASPIP or WIP; ref. 55), cytoplasmic FMR1 interacting cells in some of these mouse models have morphologic features protein 2 (CYFIP2; ref. 56), and the tetraspanin CD53 (57)].

Syder et al. PNAS ͉ March 30, 2004 ͉ vol. 101 ͉ no. 13 ͉ 4475 Downloaded by guest on September 24, 2021 These functions may support an aggressive, invasive phenotype ing hypothesis-based gain-of-function or loss-of-function genetic in these tumors. tests of their contributions to this process. The Atp4b transcrip- In summary, our cluster analysis has established that the NE tional regulatory elements identified as active in pPCs, as well as molecular signature obtained from a comparison of iGCs and the Atp4b-Cre mice described in this report, should facilitate mGCs versus pPCs and normal epithelium is directly applicable such studies. to human lung NECs, and it may prove useful for further subcategorization of human gastric cancers. We are grateful to Sabrina Wagoner, Maria Karlsson, and David O’Donnell for assistance with mouse husbandry, Qiutang Li and Janaki Prospectus. Our finding that an epithelial lineage progenitor, Guruge for supplying materials during the early phases of this study, which normally gives rise to a nonneural nonendocrine cell type, Doug Leip for his help with the GeneSymbol matching algorithm, Bill can transdifferentiate to metastatic cells with a NE phenotype Coleman and Jamie Dant for help with histochemical analyses, and has implications for the molecular pathogenesis of neuroendo- Saeed Tariq and Armstrong Gbewonyo for help with some EM studies. crine cancers in various endoderm-derived epithelia. The list of This work was supported, in part, by grants from the National Institutes genes whose expression changed during transdifferentiation of Health (DK58529 and DK63483 to J.I.G.) and from the Terry Fox (e.g., Sox2, Hey1, NeuroD1) provides a starting point for design- Fund for Cancer Research (to S.M.K.).

1. DeLellis, R. A. (2001) Am. J. Clin. Pathol. 115, Suppl, S5–S16. 33. Soriano, P. (1999) Nat. Genet. 21, 70–71. 2. Wick, M. R. (2000) Am. J. Clin. Pathol. 113, 331–335. 34. Cissell, M. A., Zhao, L., Sussel, L., Henderson, E. & Stein, R. (2003) J. Biol. 3. Solcia, E. K. G. & Sobin, L. H. (2000) Histological Typing of Endocrine Tumours Chem. 278, 751–756. (Springer, Berlin), 2nd Ed. 35. Naya, F. J., Huang, H. P., Qiu, Y., Mutoh, H., DeMayo, F. J., Leiter, A. B. & 4. Komminoth, P., Seelentag, W. K., Saremaslani, P., Heitz, P. U. & Roth, J. Tsai, M. J. (1997) Genes Dev. 11, 2323–2334. (1996) Histochem. Cell Biol. 106, 551–562. 36. St-Onge, L., Sosa-Pineda, B., Chowdhury, K., Mansouri, A. & Gruss, P. (1997) 5. Travis, W. D., Linnoila, R. I., Tsokos, M. G., Hitchcock, C. L., Cutler, G. B., Nature 387, 406–409. Jr., Nieman, L., Chrousos, G., Pass, H. & Doppman, J. (1991) Am. J. Surg. 37. Watada, H., Scheel, D. W., Leung, J. & German, M. S. (2003) J. Biol. Chem. Pathol. 15, 529–553. 278, 17130–17140. 6. Travis, W. D., Rush, W., Flieder, D. B., Falk, R., Fleming, M. V., Gal, A. A. 38. Rindi, G., Ratineau, C., Ronco, A., Candusso, M. E., Tsai, M. & Leiter, A. B. & Koss, M. N. (1998) Am. J. Surg. Pathol. 22, 934–944. (1999) Development (Cambridge, U.K.) 126, 4149–4156. 7. Simon, G. R. & Wagner, H. (2003) Chest 123, 259S–271S. 39. Graham, V., Khudyakov, J., Ellis, P. & Pevny, L. (2003) Neuron 39, 749–765. 8. Hu, Y., Ippolito, J. E., Garabedian, E. M., Humphrey, P. A. & Gordon, J. I. 40. Sakamoto, M., Hirata, H., Ohtsuka, T., Bessho, Y. & Kageyama, R. (2003) (2002) J. Biol. Chem. 277, 44462–44474. J. Biol. Chem. 278, 44808–44815. 9. Meuwissen, R., Linn, S. C., Linnoila, R. I., Zevenhoven, J., Mooi, W. J. & 41. Borges, M., Linnoila, R. I., van de Velde, H. J., Chen, H., Nelkin, B. D., Mabry, Berns, A. (2003) Cancer Cell 4, 181–189. M., Baylin, S. B. & Ball, D. W. (1997) Nature 386, 852–855. 10. Mills, J. C., Syder, A. J., Hong, C. V., Guruge, J. L., Raaii, F. & Gordon, J. I. 42. Huber, K., Bruhl, B., Guillemot, F., Olson, E. N., Ernsberger, U. & Unsicker, (2001) Proc. Natl. Acad. Sci. USA 98, 13687–13692. K. (2002) Development (Cambridge, U.K.) 129, 4729–4738. 11. Parkin D. M., Pisani, P. & Ferlay, J. (1999) Int. J. Cancer 80, 827–841. 43. Lanigan, T. M., DeRaad, S. K. & Russo, A. F. (1998) J. Neurobiol. 34, 126–134. 12. Alberts, S. R., Cervantes, A. & van de Velde, C. J. (2003) Ann. Oncol. 14, 44. Gouyer, V., Gazzeri, S., Bolon, I., Drevet, C., Brambilla, C. & Brambilla, E. ii31–ii36. (1998) Am. J. Respir. Cell Mol. Biol. 18, 188–196. 13. Lauren, P. (1965) Acta. Pathol. Microbiol. Scand. 64, 31–49. 45. Jackson, E. L., Willis, N., Mercer, K., Bronson, R. T., Crowley, D., Montoya, 14. Fenoglio-Preiser, C. M. (2001) Am. J. Clin. Pathol. 115, Suppl, S79–S93. R., Jacks, T. & Tuveson, D. A. (2001) Genes Dev. 15, 3243–3248. 15. Gilligan, C. J., Lawton, G. P., Tang, L. H., West, A. B. & Modlin, I. M. (1995) 46. Thompson, J., Epting, T., Schwarzkopf, G., Singhofen, A., Eades-Perner, Am. J. Gastroenterol. 90, 338–352. A.-M., Van Der Putten, H. & Zimmerman, W. (2000) Int. J. Cancer 86, 16. Rogers, L. W. & Murphy, R. C. (1979) Am. J. Surg. Pathol. 3, 195–202. 863–869. 17. Waldum, H. L., Aase, S., Kvetnoi, I., Brenna, E., Sandvik, A. K., Syversen, U., 47. Perez-Stable, C., Altman, N. H., Mehta, P. P., Deftos, L. J. & Roos, B. A. (1997) Johnsen, G., Vatten, L. & Polak, J. M. (1998) Cancer 83, 435–444. Cancer Res. 57, 900–906. 18. Karam, S. M. & Leblond, C. P. (1993) Anat. Rec. 236, 259–279. 48. Masumori, N., Thomas, T. Z., Chaurand, P., Case, T., Paul, M., Kasper, S., 19. Karam, S. M. & Leblond, C. P. (1993) Anat. Rec. 236, 280–296. Caprioli, R. M., Tsukamoto, T., Shappell, S. B. & Matusik, R. J. (2001) Cancer 20. Karam, S. M. & Leblond, C. P. (1993) Anat. Rec. 236, 297–313. Res. 61, 2239–2249. 21. Karam, S. M. & Leblond, C. P. (1993) Anat. Rec. 236, 333–340. 49. Bosse´, P., Bernex, F., De Sepulveda, P., Salau¨n, P. & Panthier, J.-J. (1997) 22. Karam, S. M. (1993) Anat. Rec. 236, 314–332. Oncogene 14, 2661–2670. 23. Karam, S. M., Li, Q. & Gordon, J. I. (1997) Am. J. Physiol. 272, G1209–G1220. 50. Green, J. E., Shibata, M. A., Yoshidome, K., Liu, M. L., Jorcyk, C., Anver, 24. Karam, S. M. & Leblond, C. P. (1992) Anat. Rec. 232, 231–246. M. R., Wigginton, J., Wiltrout, R., Shibata, E., Kaczmarczyk, S., et al. (2000) 25. Mills, J. C., Andersson, N., Stappenbeck, T. S., Chen, C. C. M. & Gordon, J. I. Oncogene 19, 1020–1027. (2003) J. Biol. Chem. 278, 46138–46145. 51. Wikenheiser, K. A., Clark, J. C., Linnoila, R. I., Stahlman, M. T. & Whitsett, 26. Lorenz, R. G. & Gordon, J. I. (1993) J. Biol. Chem. 268, 26559–26570. J. A. (1992) Cancer Res. 52, 5342–5352. 27. Li, Q., Karam, S. M. & Gordon, J. I. (1995) J. Biol. Chem. 270, 15777–15788. 52. Penna, D., Schmidt, A. & Beermann, F. (1998) Oncogene 17, 2601–2607. 28. Syder, A. J., Guruge, J. L., Li, Q., Hu, Y., Oleksiewicz, C. M., Lorenz, R. G., 53. Leung, S. Y., Chen, X., Chu, K. M., Yuen, S. T., Mathy, J., Ji, J., Chan, A. S., Karam, S. M., Falk, P. G. & Gordon, J. I. (1999) Mol. Cell 3, 263–274. Li, R., Law, S., Troyanskaya, O. G., et al. (2002) Proc. Natl. Acad. Sci. USA 99, 29. Carpenter, P. M., Al-Kuran, R. A. & Theuer, C. P. (2002) Am. J. Clin. Pathol. 16203–16208. 118, 887–894. 54. Krause, M., Dent, E. W., Bear, J. E., Loureiro, J. J. & Gertler, F. B. (2003) 30. Bhattacharjee, A., Richards, W. G., Staunton, J., Li, C., Monti, S., Vasa, P., Annu. Rev. Cell Dev. Biol. 19, 541–564. Ladd, C., Beheshti, J., Bueno, R., Gillette, M., et al. (2001) Proc. Natl. Acad. 55. Anton, I. M., Saville, S. P., Byrne, M. J., Curcio, C., Ramesh, N., Hartwig, J. H. Sci. USA 98, 13790–13795. & Geha, R. S. (2003) J. Cell Sci. 116, 2443–2451. 31. Syder, A. J., Oh, J. D., Guruge, J. L., O’Donnell, D., Karlsson, M., Mills, J. C., 56. Schenck, A., Bardoni, B., Langmann, C., Harden, N., Mandel, J. L. & Bjorkholm, B. M. & Gordon, J. I. (2003) Proc. Natl. Acad. Sci. USA 100, Giangrande, A. (2003) Neuron 38, 887–898. 3467–3472. 57. Okochi, H., Mine, T., Nashiro, K., Suzuki, J., Fujita, T. & Furue, M. (1999) 32. Li, C. & Wong, W. H. (2001) Proc. Natl. Acad. Sci. USA 98, 31–36. J. Clin. Gastroenterol. 29, 63–67.

4476 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0307983101 Syder et al. Downloaded by guest on September 24, 2021