The zebrafish model organism has been used extensively for studies of genetic pathways in development, indicating its potential applicability to cancer. Here we show that targeted expression of MYCN in cells of the pancreatic islet induces neuroendocrine carcinoma. Four transgenic fish developed abdominal tumors between 4 and 6 months of age, and histologic analysis revealed lobulated arrangements of neoplastic cells with expression of the MYCN transgene. The tumors also expressed insulin mRNA, and pancreatic exocrine cells and ducts were identified within the neoplasms, indicating a pancreatic origin for the tumor. Transmission electron microscopy revealed cytoplasmic, endocrine-dense core granules, analogous to those found in human neuroendocrine tumors. Our studies establish a zebrafish transgenic model of pancreatic neuroendocrine carcinoma, setting the stage to evaluate molecular pathways downstream of MYCN in this vertebrate forward genetic model system.

In order to clone candidate tumor suppressor genes whose loss contributes to the pathogenesis of neuroblastoma (NB), we performed polymerase chain reaction (PCR) screening using a high-density sequence tagged site-content map within a commonly deleted region (chromosome band 1p36) in 24 NB cell lines. We found a *480 kb homozygously deleted region at chromosome band 1p36.2 in one of the 24 NB cell lines, NB-1, and cloned the human homologue (KIF1B-b) of the mouseKif1B-b gene in this region. The KIF1B-b gene had at least 47 exons, all of which had a classic exon ± intron boundary structure. Mouse Kif1B is a microtubule-based putative anterograde motor protein for the transport of mitochondria in neural cells. We performed mutational analysis of the KIF1B-b gene in 23 cell lines using 46 sets of primers and also an allelic imbalance (AI) analysis of KIF1B-b in 50 fresh NB samples. A missense mutation at codon 1554, GTG (Gly) to ATG (Met), silent mutations at codon 409 (ACG to ACA) and codon 1721 (ACC to ACT), and polymorphisms at codon 170, GAT (Asp) to GAA (Glu), and at codon 1087, TAT (Tyr), to TGT (Cys), were all identi®ed, although their functional signi®cances remain to be determined. The AI for KIF1B-b was slightly higher (38%) than those for the other two markers (D1S244, D1S1350) (35 and 32%) within the commonly deleted region (1p36). Reverse transcriptase-PCR analysis of the KIF1B-b gene revealed obvious expression in all NB cell lines except NB-1, although decreased expression of the KIF1B-b gene was found in a subset of early-and advanced-stage NBs. These results suggest that the KIF1B-b gene may not be a candidate for tumor suppressor gene of NB. Oncogene (2001) 20, 5075 ± 5083.

No abstract

Recently, loss of heterozygosity (LOH) studies suggest that more than two tumor suppressor genes lie on the short arm of chromosome 1 (1p) in neuroblastoma (NB). To identify candidate tumor suppressor genes in NB, we searched for homozygous deletions in 20 NB cell lines using a high-density STS map spanning chromosome 1p36, a common LOH region in NB. We found that the 45-kDa subunit of the DNA fragmentation factor (DFF45) gene was homozygously deleted in an NB cell line, NB-1. DFF45 is the chaperon of DFF40, and both molecules are necessary for caspase 3 to induce apoptosis. DFF35, a splicing variant of DFF45, is an inhibitor of DFF40. We examined 20 NB cell lines for expression and mutation of DFF45 gene by reverse transcription (RT)-polymerase chain reaction (PCR) and RT-PCR-single-strand conformation polymorphism. Some novel variant transcripts of the DFF45 gene were found in NB cell lines, but not in normal adrenal gland and peripheral blood. These variants may not serve as chaperons of DFF40, but as inhibitors like DFF35, thus disrupting the balance between DFF45 and DFF40. No mutations of the DFF45 gene were found in any NB cell line, suggesting that the DFF45 is not a tumor suppressor gene for NB. However, homozygous deletion of the DFF45 gene in the NB-1 cell line may imply the presence of unknown tumor suppressor genes in this region.