4'-α/β-碳氨鏈尿苷類似物分子庫之建立:醯胺尿苷之合成

Synthesis of 4'-α/β-aminomethyl uridine analogs for construction of libraries via amide-bond formation

王仁聰(Jen-Tsung Wang)

指導教授 ： 俞鐘山

國立清華大學/原子科學院/生醫工程與環境科學系/碩士(2006年)

摘要

許多核苷(Nucleosides)的修飾物在抗病毒、抑制腫瘤細胞生長與核醫藥物上扮演相當重要的角色；目前已開發的藥物對於抗病毒有不錯之成效，但病毒的種類眾多與各式各樣新穎的病毒株出現，又加上面對環境的變遷，突變出更能適應的病毒株，使得原有藥性失效或降低。在本論文的研究裡以尿苷(Uridine)為起始物，主要合成出4’-α/β-aminomethyl uridine之相關核苷修飾化合物，來與市售各式各樣的酸基耦合形成醯胺鍵以建立起大量化分子庫。
合成路徑中，尿苷2',3',5'氫氧基先以TBDMSCl保護，接著HOAc將5'去掉保護，以EDC來將5'氫氧基氧化成醛基；置於NaOH鹼性條件下與甲醛反應，經NaBH4還原可製備出延伸出之4'碳氫氧支鏈尿苷。
後續將4'碳氫氧支鏈尿苷引入離去基；使用tosyl (TsO-)離去基無法成功的以N3-來置換，既使升高溫度(110℃)亦沒有反應；改以Triflate (TfO-)強離去基時，當接上一個當量的離去基，以LiN3行置換反應得到的化合物是經過環化路徑的2',4'-BNA (N-type)產物。由先前研究過文獻NMR光譜來比較，在C1',C2',C3'上個別的氫皆呈現單峰(偶合常數為零)。
接上兩個當量的TfO-離去基時，以LiN3行置換反應中，並無發生預期的環化現象，而N3-可將兩個離去基置換出；推測這可能是立障的關係，當存在兩個TfO-離去基時，結構上會造成立體障礙排擠使6'的離去基與2'矽保護基無法有效發生環化。之後以TBAF將TBDMS矽保護基移除，通入氫氣以Pd/C催化，使azido還原得到兩個氨基(amino)的尿苷類似物。
合成出的4',4'-bis-aminomethyl uridine，未來計劃與各式各樣的酸耦合，獲得大量化的醯氨尿苷類似物，以提供後續細胞活性試驗，期許高效率地篩選出具有潛力的新穎化合物。

關鍵字

尿苷 ； 分子庫

並列摘要

Numerous nucleoside analogs play an important role in acting as anti-viral, anti-tumor and nuclear imaging agents. However, there are a variety of new viruses and various mutants appear, which may results in resistance or inefficiency of anti-viral activity. In this study, as the starting material, uridine was derivatized to afford 4'-α/β-aminomethyl uridine analogs followed by coupling with various acids for construction of libraries via amide-bond formation.
2',3',5'-hydroxyl groups of uridine were first protected with TBDMSCl, followed by restoring 5'-OH group with HOAc. After oxidation with EDC, the aldehyde could be produced. Treatment with
aqueous HCHO and NaOH and subsequent reduction with NaBH4, the 4'α-C-branched chain could be generated.
Attempt to introduce azido group by nucleophilic substitution of the tosyl group was unsuccessful even under 110 oC. Whereas triflate proved to be an excellent leaving group, only cyclized product 2',4'-BNA was obtained when one free hydroxy moieties were reacted with lithium azide. This was confirmed by comparing with the reported NMR spectra, which showed only singlet for all the three protons (coupling constant was zero) on C1', C2' and C3'.
Ditriflates could be prepared when two equivalents of triflic anhydride was used. Repeating the treatment with lithium azide, the desired azido groups could be introduced and no cyclized product was observed. The steric interference raised by the second trifluoromethan
-sulfonyl group was proposed to account for the unfavorable nucleophilic attack of 6'-OTf on 2'-TBDMS. Removal of TBDMS group was easily achieved using TBAF, which was followed by transformation of the azido groups to the diamino groups under reduction conditions of hydrogen gas and Pd/C.
As the target compound 4',4'-bis-aminomethyl uridine analog was in hand, further coupling with various carboxylic acids to provide the compound libraries through amide linkage are performable. Relevant biological assay studies to find out the potential compound are in progress.

參考文獻

[1] Robert J. Jones, Norbert Bischofberger Antiviral Res. 1995, 27,

[3] Peter Wutzler, Rudolf Thust Antiviral Res. 2001, 49, 55-74.

[4] Rakesh Kumar Curr. Med. Chem. 2004, 11, 2749-2766.

Masayuki ujii, and Naoki Sugimoto J. Am. Chem. Soc. 2003, 125,

[8] Matt A. Peterson, BrDLEY l. Nilsson, Sanchita Sarker, Bogdan