The focus of the CMBS is pediatric neuroectodermal tumors including neuroblastoma, brain tumors and Ewings sarcoma. Retinoic Acid(RA) induced differentiation of neuroblastoma tumor cell lines continues to be our model for studies that define the signal transduction paths that mediate growth control, differentiation and cell death. Using a series of receptor selective retinoids ligands we found that RAR and RXR nuclear receptors are required to mediate maximal growth inhibiting and differentiating effects of retinoids. The most active combinations were ligands that activated RARbeta;RXR heterodimers followed by ligands activating RARalpha:RXR and RARgamma:RXR. We identified one agonist that was more potent than RA & 9cisRA in inhibiting growth and inducing differentiation, whether this compound has clinical utility remains to be evaluated. Differential display was utilized to identify genes that may be selectively activated by these agonists. One candidate differentially displayed gene is retSDR1, a retinal alcohol dehydrogenase which is involved in retinoid metabolism, and localizes to chromosome 1. We have found that one allele is deleted in Neuroblastoma cells our current working hypothesis is that the loss of this enzyme may render cells insensitive to physiologic concentrations of retinoids while cells are sensitive to pharmacologic concentrations of retinoids. We have determined that the molecular mechanisms by which retionids causes a G1 arrest of cell growth in NB cells is via a complete inhibition of G1 cyclin-dependent kinases. The decrease in kinase activity is caused by an increase in p27kip and its binding to G1 cyclin-dependent kinases. As the increase in p27 is not transcriptionally regulated, it may be due to the RA induced decrease in N-myc which leads to a desequestration of p27. The expression of G1 Cyclins D and E and cdk2 and cdk4 and other inhibitors are unchanged during RA treatment. However the levels of cdk6 are dramatically decreased during RA induced differentiation. The decease in cyclinD;cdk6;p27 complexes may contribute to the increase in p27. We believe the levels of Nmyc also affect the p27 levels. We have determined that Nmyc over-expression increases cyclinE and cyclin E dependent kinases and decreases p27 levels. The decrease in p27 levels is due to its phosphorylation by cyclinE dependent kinases which targets it to be degraded by the proteosome. Thus in R A treated NB cells as Nmyc decreases there is a decrease in phospohorylation of p27 and a decrease in its ability to be degraded by the proteosome. In NB tumors, Trks serve as tumor markers; TrkA is expressed in good prognosis tumors and most poor prognosis tumor express TrkB. We have found that the differential activation of these signal transduction pathways in NB may alter their growth, invasiveness, chemosensitivity and cell survival. Using tet-regulated expression vectors we have made a series of NB cells which express varying levels of Trk and are testing the effects of activation of these paths on cell growth. We find that NGF treatment of cells expressing high levels of TrkA causes a marked decrease in cell number by delaying but not arresting progression through the cell cycle. This is associated with decreases in N-myc and increases in p27 binding to cdk-kinases. Thus using RA or NGF we find decreases in N-myc and increases in p27 associated with growth inhibition. We also find that IGFII levels decrease in cell lines derived from adrenal primary NB tumors and this is consistent with the non-overlapping expression of TrkA and IGFII in adrenal primary NB tumors. Conversely TrkA levels and IGFII are co-expressed in extraadrenal neuroblastomas of infants. Transfection of TrkA into a thoracic NB cell line and activation of the TrkA signal transduction path leads to increased expression of IGFII. This indicates a tight link between activation of Trk A and regulation of IGFII expression in NB cells. A number of pediatric tumor have alterations in transcription factors. These mutant transcription factors may stimulate or repress genes important in controling cell growth and differentiation. We have studied an inhibitor of histone deacetylation, MS-27-275 and found it is active in a number of pediatric tumor cell lines in vitro. Ewing's sarcoma cells are characterized by a t(11;22) translocation that creates a chimeric transcription factor EWS/Fli. We have found that the EWS/Fli represses transcription of TGFB RII so that the TGFB signaling path is not active in these cells. We find that MS-27-275 increases TGFBRII and this may be due to the transcriptional relief of the EWS/Fli repression of the TGFBRII promoter. Thus MS-27-275 may target molecular alterations in Ewing's sarcoma.We have studied the activity of this drug in animal models and found that it is active in inhibiting the growth of sarcomas and neuroblastomas. In studying the biology of NB and Ewings' sarcoma tumors we have noted that the survival of selected tumor cell lines is dependent on an autocrine Stem Cell Factor/c-kt autocrine loop. Using the tyrosine kinase inhibitor, STI571 we have found that the growth of cell lines derived from these tumors is significantly decreased. In vitro, STI571 inhibits the SCF stimulated autophosphorylation of NB and EWS cells. As part of a project to identify genes that marked differentiated NB cells that may have clinical utility as tumor markers, the CMBS has isolated a retinoid regulated gene 37G1 that has homology to unc-33 a C. elegans gene important in neuronal pathfinding. We published the cDNA sequence, mRNA and protein analysis and regulation in neuroblastoma. Antibodies to hUlip mark differentiated ganglionic cells in neuroblastoma tumors and brain tumors. HUlip also marks gemistocytic cells in astrocytic tumors and glioblastoma as well as all the cellular components of the subependymal giant cell astrocytoma (SEGA). Both the human and mouse genes have been isolated; the human gene maps to 5q31 and contains 14 exons spanning over 50kb. The murine and human promoters of Ulip have been isolated and characterized and contain a highly conserved set of transcription regulatory sites that are being tested for tissue specific expression. Using the murine gene, constructs for knocking out Ulip have been made and Ulip +/- mice have been made. We currently have 2 Ulip colonies of mice in the SJ/129 strain and into 6 generations on a C57Bl background. In both strains there have been no detectable alterations in expected breeding patterns. Animals do not have gross morphologic alterations. An analysis of Ulip-/- mice will lead to better understanding of the effects of loss of Ulip during development. Since the KO construct contains a B-galactosidase expression vector we have been able to study the developmental expression of Ulip during embryogenesis. We have initiated studies to determine what signalling mechanisms induce Ulip expression during Retinoid treatment of NB cells.
