METHYLATION OF DNA MAY BE USEFUL AS A COMPUTATIONAL TOOL: EXPERIMENTAL EVIDENCE
METHYLATION OF DNA MAY BE USEFUL AS A COMPUTATIONAL TOOL: EXPERIMENTAL EVIDENCE
复制标题
DNA 甲基化可能可用作计算工具：实验证据
DOI：
10.1142/9789812706799_0001
复制
发表时间：
2007
期刊：
影响因子：
--
通讯作者：
T. Head
中科院分区：
文献类型：
--
作者：
S. Gal;N. Monteith;Sara Shkalim;Hu Huang;T. Head
之前，我们已经解释了我们称之为“水计算”的抽象概念，并用具体的湿实验室结果来说明它。在这里，我们探讨了使用甲基化酶在双链DNA分子的位点上“写”，当且仅当这些位点以前没有被甲基化时，限制性内切酶会切割这些位点。如果位点已被甲基化，则该位点表示位0（假，F），如果未被甲基化，则该位点表示位1（真，T）。“阅读”是通过在每个部位尝试切割来完成的。我们发现了8种市售的甲基化酶和8种相应的限制性内切酶，它们在其中一种甲基化酶作用后不会切割。我们能够证实，这8种酶中的每一种的甲基化仅单独阻断与该位点相关的限制性内切酶的切割，而不阻断任何其他酶的切割。然后，我们使用这些酶来接近一个3变量，4子句的可满足性（SAT）的问题，使用质粒DNA（pBluescript）或PCR产物从该区域含有质粒上的限制性内切酶位点。定义成对的甲基化酶来代表算子p、q和r的每一种状态，一个甲基化酶用于p，另一个用于p '，等等。我们在两个位点平行地甲基化DNA，所以要么p位点被甲基化（使p为假），要么p'位点被甲基化（使p'为假）。我们对其他两个变量也这样做，以创建一组逻辑上一致的DNA片段。然后，我们使用限制性内切酶应用这4个子句来切割不满足它们的DNA片段。最后，我们发现了完整DNA的证据，表明答案满足所有条款。为了确认每个布尔运算符的状态，我们使用了适当的限制性内切酶的切割。我们发现，在质粒和PCR产物的计算中，一个位点对在两个位点中都显示错误;例如，q和q '。这是不可能的。我们怀疑这些限制性内切酶中的一种，特别是BssHII，在从句中进行了不完全切割。总之，我们成功地证明了DNA甲基化在数学计算中的有用性。因此，我们已经为我们的武器库添加了以水的方式执行DNA计算的潜在方法。
Previously we have explained the abstract concept we call ‘aqueous computing’ and illustrated it with concrete wet lab results. Here, we explore the use of methylase enzymes to ‘write’ on double-stranded DNA molecules at sites where restriction enzymes will cut if, and only if, the sites have not previously been methylated. A site represents the bit zero (False, F) if the site has been methylated and the bit one (True, T) if it has not been methylated. ‘Reading’ is done by attempting a cut at each of the sites. We found 8 commercially available methylases and 8 corresponding restriction enzymes that would not cut after the action of one of the methylases. We were able to confirm that methylation by each of these 8 enzymes individually blocked cleavage only by the restriction enzyme associated with that site and not any other enzyme. We then used these enzymes to approach a 3-variable, 4-clause satisfiability (SAT) problem using either plasmid DNA (pBluescript) or PCR product made from the region containing the restriction enzyme sites on the plasmid. Pairs of methylases were defined to represent each of the states of the operators p, q and r, one methylase for p and another for p’, etc. We methylated the DNA in parallel at the two sites so either the p site was methylated (making p false) or the p’ site was methylated (making p’ false). We did that for the other two variables as well to create a set of logically consistent DNA fragments. Then we applied the 4 clauses using restriction enzymes to cut DNA fragments that did not satisfy them. At the end, we found evidence for intact DNA indicating an answer satisfying all of the clauses. To confirm the state of each of the Boolean operators, we used cleavage by the appropriate restriction enzyme. We found in the computation with both the plasmid and the PCR product, one site pair to show false in both sites; q and q’, for instance. This should not be possible. We suspected incomplete cutting during the clauses by one of these restriction enzymes, specifically BssHII. In summary, we did successfully show the usefulness of DNA methylation in a scheme to do a mathematical computation. Thus, we have added to our arsenal of potential methods of performing DNA computing in the aqueous style.