Human cancers arising from different cell lineages display genotypes with characteristic features, but the lineage-specific factors that influence the differences in genetic profiles have not been identified. Such differences could be explained by models in which a cellular phenotype limits a cells vulnerability to certain mutations or by models in which observed mutations determine the ultimate cell phenotype. To evaluate the relative merits of these proposals, we are studying the initiation of cancers by genetically modifying cells at discrete stages of differentiation after derivation from human embryonic stem cells (hESCs). We have focused initially on small cell lung cancer (SCLC), the most aggressive type of human lung cancer, characterized by a poor prognosis, the rapid development of resistance to treatment, and nearly universal loss of function of multiple tumor suppressor genes, especially TP53 and RB. Earlier studies with mouse models indicate that the normal cell precursors to SCLCs are likely to be pulmonary neuro-endocrine cells (PNECs). Building on existing methods for sequentially differentiating hESCs in culture into several types of lung cells, we have now produced PNECs in significant numbers for the first time by blocking signaling through NOTCH receptors (using inhibitors of gamma-secretase) and by interfering with expression of the RB gene (using inhibitory RNAs) in lung progenitor cells. In contrast, expression of mutated EGFR or KRAS genes, common in human lung adenocarcinomas that arise from the alveolar cell lineage, has no evident effects on the formation of PNECs. Although PNECs induced by blocking NOTCH and RB signaling do not form xenografts in immune-deficient mice, the cells form subcutaneous tumors resembling early stage human SCLC and bearing neuroendocrine markers if expression of the TP53 gene is also impaired. These findings imply that PNECs or their committed precursors are vulnerable to oncogenic mutations found specifically in human SCLC, resulting in a significant increase in number of PNECs and conversion to neoplasia. This experimental system provides opportunities to study tumor progression and cancer drug susceptibility and resistance, using human lung cells grown in culture.