9月9日出版的Cell杂志封面是一双错配的袜子，喻意是关于DNA错配和修复的。文章的通讯作者为华人科学家——李国民（音译），在武汉大学生物系获得理学学士学位，在美国Wayne State University (韦恩州立大学)取得理学博士学位，目前是美国肯塔基大学医学中心（University of Kentucky Medical Center）的副教授。
李国民教授多年来从事DNA复制错配与修复的分子机制研究。DNA除了外界环境影响会造成损伤外，在其本身的复制过程中也会出现错配。尽管这种出错的几率十分微小，但是生命这台精密的仪器不允许任何差错。因为就算是一个碱基的错配也可能造成严重的疾病。如果错配修复基因存在缺陷的话就可能造成癌症，比如：遗传性非多发性息肉结肠癌（hereditary nonpolyposis colorectal cancer）。
对于复制过程中出现的错误，比如碱基错配、少量插入、缺失，机体内本身就存在有校正机制。这种校正机制通常可以由DNA聚合酶3’→5’的校对功能立即进行纠正。如果DNA聚合酶没有发现错配，就需要错配修复（mismatch repair,
MMR）系统来完成。
在大肠杆菌中的错配修复系统已经研究得十分透彻了。这个系统主要包括酶错配矫正酶（mismatch correction enzyme）、DNA聚合酶Ⅲ和DNA连接酶等11种蛋白参与。可以分为步骤：启始、切除和修复。启始：根据序列上GATC中腺嘌呤（A）的甲基化程度不同，酶错配矫正酶中的MutS部分识别错配区域，而MutH和MutL识别没有甲基化的GATC序列。之后，将新合成的单链上切出一个缺口（nick）。切除：由核酸外切酶将距离GATC到错配区域间的DNA链切除。为了防止剩下的暴露的单链遭到降解，需要有单链结合蛋白（Ssb）将其保护起来。修复：由DNA聚合酶Ⅲ全酶进行合成，连接酶对缺口进行连接。
错配修复系统是一种高度保守的途经。理论上，它在真核系统中的修复过程应该与原核系统中的相类似。原核系统是依靠甲基化程度来判断哪条是新合成链，而在真核系统中是如何识别错配单链的机制尚未明确。但是，人们在体外可以利用含有MutSα或者MutSβ、MutLα、RPA、PCNA、EXO1、HMGB1、RFC和DNA聚合酶δ的系统将含有缺口的双链DNA进行修复。
李国民教授这篇文章Reconstitution of 5′-Directed Human Mismatch Repair in a Purified System，文章利用纯化后的人源蛋白在体外对5’方向存在缺口的错配DNA进行了修复。这一系统包括了MutSα或者MutSβ、MutLα、RPA、PCNA、EXO1、HMGB1、RFC和DNA聚合酶δ蛋白。在这个系统中，MutSβ对于碱基错配的修复作用十分有限，它对于插入/缺失错配的修复率比MutSα高；MutSα对于两种类型的错配修复率是一样的。MutLα可以减低EXO1的酶切活性，并且在错配区域被切除后MutLα可以终止EXO1的催化切除效果。如果缺乏MutLα蛋白EXO1就会无法停止作用而切除过多的DNA。
RPA和HMGB1在MutSα激活的EXO1催化切除的错配修复启始、切除过程中起到相似但是互补的作用，但是RPA在协助MutLα调节切除过错配区域后的终止过程中起到完全不同的作用。一个碱基有效的错配修复需要MutSα-MutLα多分子复合物的参与。这一研究描述了一种由错配引起的DNA切除的启始和终止模式。
在体外重新构建的错配修复系统能执行的只是体内错配修复系统的一部分功能。它必须以出现缺口的双链DNA为底物进行反应，而在体内这种缺口产生的机制仍不清楚，特别是先导链（leading strand）缺口的产生。在这项研究中将真核系统错配修复系统中MutLα的作用机理进行了进一步的阐明，是MutLα使EXO1失去酶切的活性，从而使切除反应终止。这对于错配修复系统是否能正常工作是一个关键的步骤。
李国民教授联系方式：
Graduate Center for Toxicology and Markey Cancer Center, University of Kentucky Medical Center, Lexington, Kentucky 40536
Phone: (859) 257-7053
E-mail: gmli@uky.edu
研究兴趣
Research Interest: Molecular mechanism of DNA mismatch repair.
DNA mismatch repair maintains genomic stability by correcting mismatches generated during normal DNA metabolism and by mediating DNA damage-induced apoptosis. Defects in mismatch repair genes are the genetic basis of certain types of human cancer, including hereditary nonpolyposis colorectal cancer. To fully understand the molecular mechanisms of the mismatch repair system in human cells, our work is focused on three areas: 1) enzymology of mismatch repair; 2) identification and characterization of proteins involved in DNA damage-induced, mismatch repair-dependent apoptosis; and 3) genetic alterations of mismatch repair genes in human cancer and other diseases.
小常识：DNA错配修复的机制
Mismatch repair
To repair mismatched bases, the system has to know which base is the correct one. In E. coli, this is achieved by a special methylase called the "Dam methylase", which can methylate all adenines that occur within (5')GATC sequences. Immediately after DNA replication, the template strand has been methylated, but the newly synthesized strand is not methylated yet. Thus, the template strand and the new strand can be distinguished.
. Mismatch repair.
The repairing process begins with the protein MutS which binds to mismatched base pairs. Then, MutL is recruited to the complex and activates MutH which binds to GATC sequences. Activation of MutH cleaves the unmethylated strand at the GATC site. Subsequently, the segment from the cleavage site to the mismatch is removed by exonuclease (with assistance from helicase II and SSB proteins). If the cleavage occurs on the 3' side of the mismatch, this step is carried out by exonuclease I (which degrades a single strand only in the 3' to 5' direction). If the cleavage occurs on the 5' side of the mismatch, exonuclease VII or RecJ is used to degrade the single stranded DNA. The gap is filled by DNA polymerase III and DNA ligase.
The distance between the GATC site and the mismatch could be as long as 1,000 base pairs. Therefore, mismatch repair is very expensive and inefficient.
Mismatch repair in eukaryotes may be similar to that in E. coli. Homologs of MutS and MutL have been identified in yeast, mammals, and other eukaryotes. MSH1 to MSH5 are homologous to MutS; MLH1, PMS1 and PMS2 are homologous to MutL. Mutations of MSH2, PMS1 and PMS2 are related to colon cancer.
In eukaryotes, the mechanism to distinguish the template strand from the new strand is still unclear.
