關遺傳學是在不改變DNA序列下，進一步調控基因的表現。包含DNA的甲基化和組蛋白後轉錄修飾。其中一種組蛋白修飾為組蛋白甲基和去甲基化，被甲基轉移酵素和組蛋白去甲基酵素調控。在許多種癌症中發現，甲基轉移酵素和組蛋白去甲基酵素的作用已經失調。主要的組蛋白去甲基酵素都包含一段Jumonji-C domain，同時利用2-oxoglutarate和二價鐵離子進行去甲基的功能。KDM4B是其中一種的去甲基酵素，作用於H3K9me3/me2以及H3K36me3/me2，另一方面KDM4B在許多種癌症中的表現量都偏高，例如:胃癌，腎臟癌，直腸癌，乳癌和前列腺癌。其中在前列腺癌上，KDM4B和AR receptor共同作用促進細胞增長因此我們假設KDM4B可以作為一個治療癌症的目標蛋白。在這篇論文中，我們希望可以探討可以抑制KDM4B活性的天然抑制劑。在蛋白質和分子結構的研究下，我們發現一個天然物M2可以抑制KDM4B(IC50:12.2 μM，Ki:2.72μM)，同時會和反應物有一個競爭的關係。另一方面，M2也可以抑制其他KDM4的去甲基酵素。在細胞的實驗中，M2會在CWR22Rv1，LNCaP，LNCaP C4-2B和PC-3的細胞上，產生毒性，同時也發現KDM4B的主要反應物H3K9me3在細胞的表現量也提升。最後總結我們的實驗，認為M2可以被視為可以發展為抗癌藥物的方向，未來我們將開發M2的衍生物，進一步提升抑制性和靈敏度。 Epigenetics is the study of transcriptional and physiological trait variations that do not involve changes in the DNA sequence. DNA methylation and post-translational modification in the histone tail are major mechanisms that produce such changes. One of histone modifications, methylation/demethylation, is controlled by histone methyltransferase and histone demethylase has been recently studied. Dysregulation of histone methylation/demethylation is found to be associated with various types of cancer. The major type of lysine demethylase (KDM) consists of Jumonji-C (JmjC) domain (KDM2?KDM8) and requires 2-oxoglutarate and Fe(II) as cofactors to remove the methyl moiety from the histone tail. Of those, KDM4B, a member of the KDM4 family can specifically remove the methyl group from H3K9me3/me2 and H3K36me3/me2. KDM4B are overexpressed in many types of cancer including gastric cancer, renal cancer, colon cancer, breast cancer and prostate cancer. In prostate cancer, KDM4B is a co-activator of androgen receptor (AR) to stimulate cell growth. Given the important role of AR in prostate carcinogenesis, we propose KDM4B as a useful target protein. In this study, we aim to discover natural compounds that inhibit KDM4B based on a structure-guided approach. We have successfully identified a compound M2 with IC50 = 12.2 ± 1.6 μM, Ki= 2.72 μM. M2 exhibits a competitive inhibition mode toward KDM4B. M2 also shows inhibition against other members of the KDM4 family but not against PHF8. In cell-based experiment, M2 had cytoxicity in AR-dependent (LNCaP) as well as AR-independent prostate cancer cells (CWR22Rv1, LNCaP C4-2B and PC-3). Addition of M2 in LNCaP and CWR22Rv1 cells led to the increase of H3K9m3 signal. Together, M2 is a specific KDM4 inhibitor and shows anti-prostate cancer effect. Future work is to derive M2-based compounds with higher specificity and potency.