以質體為基礎之活體內DNA修復試驗研究

Abstract

為了研究 DNA 的修復機制，常使用具有錯誤或受損傷的 DNA 序列作為反應物，使用這樣的 DNA受質可在試管中(in vitro)進行分析，或轉形至細胞以生物體(in vivo)的方式分析修復情形。常用的DNA 受質的形式可為質體/噬菌體為基礎的 DNA 或寡核苷酸，本實驗室的過去噬菌體DNA 受質進行試管內分析的方式，雖然有利於酵素功能分析及作用機制的研究，但由於反應的條件常會被質疑無法反映真實生理狀況，為了能充分印證在試管中(in vitro)實驗所推論出的結果，設計出一套 in vivo 的實驗系統有其重要性。 本實驗室曾製備噬菌體 DNA 為基礎的受質，以感染大腸桿菌進行 in vivo 分析，但以噬菌體DNA 受質，會有以下問題:第一，噬菌體的生活史當中，有以其中一股為模板進行滾筒複製的過程，有這樣可能會遺失另一股鹼基序列的資訊(strand loss )；第二，以噬菌體感染大腸桿菌的溶菌斑，在(LacZ)藍白篩選的分析中所顯現的顏色，較不如單一菌落容易判別，需要進一步的改善。因此本研究設計質體 DNA 為基礎的受質以其建立 in vivo DNA 修復分析系統，並評估其在核酸修復研究之應用。 本研究選用帶有選擇標誌為 ampr 及 lacZα 之質體pUC18，插入設計的雙股寡核苷酸插入子(insert)並送定序確認序列。設計好的 pUC18 vector 以 DNA 切口內切酶(nicking endonuclease)在其中一股的兩段特定序列進行水解，並移除一小段序列後分離純化 gap DNA。再將 gap DNA 與含有亞黃嘌呤核酸(inosine)的單股寡核苷酸以接合酶接合，製成 TI 配對 DNA 受質。設計好的 TI 配對 DNA 受質分別進行 in vitro 及 in vivo 分析。 實驗以野生株與核酸內切酶第五型突變株大腸桿菌萃取物與 TI 配對 DNA 受質in vitro反應 分析。結果野生株的修復效率約為 65%；核酸內切酶第五型(EndoV)突變株的修復效率約為 10%，少於野生株的結果的六分之一。將 TI 配對 DNA受質分別轉形至各種突變株及其對應野生株大腸桿菌後，塗布培養在含有 ampicillin 的瓊脂平板，挑出單一菌落並培養增幅。將增幅的大腸桿菌抽取質體，將質體轉形至 DH5α，塗布培養在含有ampicillin，X-gal 及 IPTG 的瓊脂平板，觀察菌落顏色，以分析 in vivo 修復情形結果，野生株 BW25113為 90%，EndoV 突變株 JW5547 為 33%，可顯示野生株與突變株在修復效率的差異，然而 polA+菌株 KA796 為 88%與 DNA 聚合酶I(DNA polymerase I) 3’ 端外切酶 KA796mutant (polA exo- 突變株 )為 84% ，變異株修復效率並無顯著降低，推測可能為複製過程其中含有 inosine 的一股，因 3’外切酶活性突變的 DNA polymerase I 阻礙了該股的複製的進行，導致該股修復訊息的損失，而無法以此分析方法釐清 DNA polymerase I 的 3 端外切酶活性在此修復路徑的重要性。

To study DNA repair mechanism, mispaired or damaged DNA sequences were commonly used as substrates. These DNA substrates could be analyzed by either in vitro analysis or in vivo analysis. The common DNA substrate types for example, are plasmid/phage-based DNA or oligonucleotides. In our previous study, phage DNA substrate was used for in vitro assay. Although in vitro study is beneficial to analyze the mechanism and function of enzyme kinetics, it has been questioned if it could represent the real physiological environment. Therefore in vivo model will be useful to validate the results from in vitro study. Our lab has used M13,f1 phage-based DNA substrates to transfect Escherichia coli (E. coli) in vivo study. However, phage-based DNA substrates may lead to the following disadvantages. First, during the life cycle of M13 phages, one strand of DNA is utilized as template for rolling circle replication, which may leads to stand loss of the other DNA stand. Second, the plaques of E. coli infected by phages could not be easily differentiated in Lac Z blue/white screening. Therefore, in this study, plasma-based DNA substrates were used to develop in vivo DNA repair analysis system and its application in the research of nucleic acid repair was evaluated. In this study, plasmids with selecting marker ampr and lacZα were used such as pUC18. These plasmids were inserted with the designed double-strand oligonucleotide and its sequence was confirmed by sequencing. The designed pUC18 vector was digested by nicking endonuclease in two specific site of one DNA strand. After a small sequence was removed, gap DNA was separated and purified. Subsequently, gap DNA was connected to the single-strand oligonucleotide using ligase to form TI-mismatch DNA substrates for in vitro and in vivo analysis. We used cell extract of E. coli wild type or EndoV mutant strain to repair the TI substrate. The repair efficiency of wild type extract was about 65% and the efficiency of EndoV mutant was around 10%, less than one sixth of the wild type. After the TI-mismatched DNA substrates transformed into different strain of mutants and their isogenic wild type with selection ampicillin; individual ampicillin-resistant colony was picked for amplification. The plasmid DNA of the amplified E. coli was extracted, transform into DH5α and spread on an agar plate containing ampicillin, X-gal and IPTG to observe the color of colonies in order to evaluate the in vivo repair efficacy. The obtained results show that the repair efficacy of wild type BW25113 was and EndoV mutant were 90% and it was 33% respectively . This result is consistent with in vitro results mentioned above. However, the repair efficiency of polA+ strain KA796 was 88% and the DNA polymerase I 3’→5’ exonuclease mutant strain was 84%. No statistic difference was observed in the repair efficacy between the mutant and wild type. It is speculated that the phenomenon may result from the DNA polymerase I with the mutant 3’ exonuclease inhibited the ligation of the inosine-containing DNA strand during replication, strand containing inosine collapsed. Therefore, the repair information of the strand was lost and thus this analytic method could not be used to clarify the importance of the 3’ exonuclease of DNA polymerase I in this repair pathway.