Neuroblastoma (NB) is the most common solid tumor of early infancy. Despite a significant improvement in the general survival rate for children with cancer, the prognosis of high-risk NB remains low, at about 30%, despite the use of intensive chemo-radiotherapy followed by differentiation therapy with retinoic acid (RA). Relapses in this category of NB are often due to the emergence of multi-drug and RA-resistant minimal residual cancer cells. The use of natural 13-cis RA, as a single chemo-preventive agent, has improved the survival rate to 50% for high-risk NB patients. However, the prevalence of RA-resistance is high in high-risk NB, and in solid cancers, in general. RA-resistance in cancer cells is mediated by a number of factors. Loss of RA-induced expression of the putative tumor suppressor gene, retinoic acid receptor-beta (RARβ), is one of the most common factors that have been reported in RAresistant phenotypes of a wide range of cancer cells. The transcriptional regulation of RAR(β) gene and other retinoid responsive-genes is believed to be regulated by the ligand-dependent transactivation of the homo- or heterodimer complexes of the retinoic acid receptor (RAR) and retinoid X receptor (RXR) subtypes, namely alpha (α), beta (β) and gamma (γ). It is believed that the anti-cancer activities of natural all-trans RA and 13-cis RA are mediated through activation of RAR-complexes. The loss of RA-induced RAR β expression can be caused by aberrant recruitment of chromatin structure modifying enzymes, histone deacetylases (HDACs), which have major roles in the global regulation of gene transcription. However, the mechanism of RA-resistance in NB cells is unclear. This thesis set out to identify the molecular mechanism of RA-resistance and to develop a new therapeutic approach to overcome RA-resistance in NB cells. The data in this thesis demonstrated that deregulation or over-expression of proto-oncogene MYCN caused a total RA-resistance in NB cells in vitro and in vivo, despite the strong induction of RARI3 expression. The data also indicated that the activation of RAR-dependent pathways by aRA or 13RA alone is not sufficient to overcome MYCN-mediated RA-resistance in NB cells. In the light of this observation, this thesis went on to examine whether combined targeting activation of RAR and RXR subtypes with receptor specific ligands could enhance the therapeutic efficacy of the retinoid signaling pathway. NB cells were treated with a panel of receptor-specific retinoids, namely aRA, l3RA, 9RA (RAR-specific), CD 417, CD 2314 (RARβ-specific), CD 666 (RARγ-specific), CD 336 (RARα-specific), CD 3640, CD 2872 (RXR-specific), as a single agent or in combination at a low concentration of 0.1 ??M. The results showed that combined targeting activation of RARα and RXR was not only the most effective combination, but also overcame MYCN-mediated RA-resistance in NB cells in vitro.Collectively, these data demonstrated the combined targeting activation of RAR and RXRs as a new approach to enhance the efficacy of retinoid therapy and overcome RA-resistance in the treatment of high-risk NB, and other cancers. The emerging therapeutic potential of HDAC inhibitors (HDACi) as front line anti-cancer agents, or adjuvants to other agents such as RA, has suggested a new approach in the treatment of cancer. However, the molecular mechanism of the remarkably specific anticancer actions of HDACi is still largely speculation. The data presented in this study was the first to demonstrate a novel sequential order and the dosage-dependent roles of basal p21Wafl expression and G2/M arrest as protective mechanisms against HDACi-induced apoptosis. In addition, this thesis also examined and compared the therapeutic efficacy of HDACi as a single agent and in combination with other anti-cancer agents such as RA, IFNα and chemotherapeutic agents. Evaluation of the therapeutic effects of combinations of aRA, IFN and HDACi showed that combination of HDACi and IFNα exhibited the strongest synergy against NB cells in vitro. Treatment of MYCN transgenic mice, which consistently develop abdominal NB tumors that closely mirror the human disease in both physiological and biological aspects, with hydroxamic acid-based HDACi, trichostatin A (TSA), alone reduced tumor growth by nearly 50%, compared to the solvent control and IFNα alone, which had no effect on NB tumor growth. The most exciting finding was that the combination of HDACi and IFNα synergistically reduced tumor mass and angiogenesis by over 80% without any apparent systemic side-effects. The therapeutic effect of treatment with HDACi correlated with the induction of acetylation of histone 4 protein (H4) in both tumor and organ tissues, indicating a wide therapeutic index for HDACi in vivo. Collectively, the data in this study have demonstrated basal p21 Wafl expression as a potential marker of sensitivity to HDACi-based therapy, and the therapeutic efficacy of a novel combination of HDACi with IFNα in vivo. These preclinical data have provided an evidence-based rationale for a clinical trial of the combination of HDACi and IFNα in the treatment of high risk NB.