Vascular endothelial growth factor-A (VEGF-A

Our preliminary studies revealed that oncogenic KRAS (KRAS/V12) dramatically suppressed the growth of immortalized airway epithelial cells (NHBE-T, with viral antigen-inactivated p53 and RB proteins). This process appeared to be a novel event, different from the so-called premature senescence that is induced by either p53 or RB, suggesting the existence of a novel tumor suppressor that functions downstream of oncogenic KRAS. After a comprehensive search for genes whose expression levels were modulated by KRAS/V12, we focused on DUSP6, a pivotal negative feedback regulator of the RAS-ERK pathway. A dominant-negative DUSP6 mutant, however, failed to rescue KRAS/V12-induced growth suppression, but conferred a stronger anchorage-independent growth activity to the surviving subpopulation of cells generated from KRAS/V12-transduced NHBE-T. DUSP6 expression levels were found to be weaker in most lung cancer cell lines than in NHBE-T, and DUSP6 restoration suppressed cellular growth. In primary lung cancers, DUSP6 expression levels decreased as both growth activity and histological grade of the tumor increased. Loss of heterozygosity of the DUSP6 locus was found in 17.7% of cases and was associated with reduced expression levels. These results suggest that DUSP6 is a growth suppressor whose inactivation could promote the progression of lung cancer. We have here identified an important factor involved in carcinogenesis through a comprehensive search for downstream targets of oncogenic KRAS.

BRN2 is a developmental neural cell-specific POU domain transcription factor and is crucial for cell lineage determination. We investigated the importance of BRN2 in the expression of the lineage-specific transcription factors (achaete-scute homolog-like 1 (ASCL1) and NeuroD1 (ND1)) and neural/neuroendocrine marker molecules (neural cell adhesion molecule 1 (NCAM1), synaptophysin (SYP) and chromogranin A (CHGA)) in small cell lung cancer (SCLC) using cultured lung cancer cells. All examined SCLC cell lines expressed BRN2, as well as ASCL1, ND1, NCAM1, SYP and CHGA. The expression levels of ASCL1, ND1, NCAM1, SYP and CHGA considerably decreased when BRN2 was knocked down in SCLC cells, and the addition of a BRN2 transgene into non-SCLC (NSCLC) cells induced the expression of ASCL1, ND1, NCAM1, SYP and CHGA. However, the BRN2 gene was not activated by the forced expression of ASCL1 or ND1 in NSCLC cells. The knockdown of BRN2 caused significant growth retardation with decrease of S to G2 phase population and mitotic cell rates and unaltered Ki-67-labeled or apoptotic cell rates in SCLC cells, indicating increase of G1 phase population. These findings suggest that BRN2 is a higher level regulator than ASCL1 and ND1 and BRN2 might be involved in aggressiveness of SCLC.

Small cell lung cancer (SCLC) exhibits insulin-like growth factor-dependent growth. SCLC is the most aggressive among known in vivo lung cancers, whereas in vitro growth of SCLC is paradoxically slow as compared with that of non-SCLC (NSCLC). In this study, we demonstrate that SCLC cells overexpress insulin-like growth factor binding protein (IGFBP)-2 via NeuroD, a neuroendocrine cell-specific transcription factor. Chromatin immunoprecipitation, electrophoretic mobility shift, and IGFBP-2 promoter assays all revealed that NeuroD binds to the E-box in the 5'-untranslated region of IGFBP-2. A NeuroD transgene in both airway epithelial and NSCLC cells up-regulated the transcription of IGFBP-2 and retarded cell growth. Recombinant IGFBP-2 repressed the growth of both airway epithelial and NSCLC cells in a dose-dependent manner. A NeuroD-specific small interfering RNA repressed IGFBP-2 expression in SCLC, and neutralization of IGFBP-2 and an IGFBP-2-specific small interfering RNA increased SCLC cell growth. Pathological samples of SCLC also expressed IGFBP-2 abundantly, as compared with NSCLC, and showed only rare (8%) IGFBP-2 promoter methylation, whereas the IGFBP-2 promoter was methylated in 71% of adenocarcinomas and 29% of squamous cell carcinomas. These findings suggest that 1) SCLC has an IGFBP-2 overexpression mechanism distinct from NSCLC, 2) secreted IGFBP-2 contributes to the slow growth of SCLC in vitro, and 3) the epigenetic alterations in the IGFBP-2 promoter contribute to the striking differences in IGFBP-2 expression between SCLC and NSCLC in vivo.