New Tools for CXMS from Dr. Meng-Qiu Dong’s Laboratory

Dr. Meng-Qiu Dong’s lab and collaborators have recently published papers in Nature Communications to report technological advancement on mass spectrometry-based protein structural analysis. The titles of the two articles are "Improving mass spectrometry analysis of protein structures with arginine-selective chemical cross-linkers" and "A high-speed search engine pLink 2 with systematic evaluation for proteome-scale identification of cross-linked peptides.". The former, published on the 2^nd of September, 2019, reported the development and application of ArGO_1-3, a series of arginine-arginine (R-R) cross-linkers, and KArGO, an arginine-lysine (K-R) cross-linker. The latter, published on the 30^th of July, 2019, introduced a fast and accurate software pLink 2 for identification of cross-linked peptides and a set of methods for systematic evaluation of cross-link identification software.

Dr. Meng-Qiu Dong’s lab has been perfecting a technology called cross-linking of proteins coupled with mass spectrometry (CXMS or XL-MS). CXMS uses chemical cross-linker to covalently link two amino acids in the same protein or in different proteins that are spatially close to each other. After digestion of cross-linked protein samples into peptides, LC-MS/MS is used to identify cross-linked peptides and locate the exact cross-linking sites to obtain low-resolution structural information, which can help infer the folding of a protein and the approximate protein-protein binding interface. Compared to crystal and cryo-electron microscopy, CXMS does not requires a large quantity of highly purified samples. Besides, CXMS is fast and easy, and able to capture the dynamics of proteins in solution. As a result, CXMS has been widely used over the past decade, frequently lending a hand to resolving the atomic structures of large protein complexes and to finding regions of direct interactions between proteins. As a pioneer and promoter, Dr. Meng-Qiu Dong’s lab has worked closely with research teams in the fields of computational science, organic chemistry and protein science to advance the CXMS technology, including the development of software (Yang B, Nat Meth 2012; Lu S, Nat Meth 2015), crosslinkers (Tan D, eLife 2016; Zhang X, Anal Chem 2018), new workflows and applications (Lu S, Nat Meth 2015; Gong Z, Biophys Rep 2015; Ding YH, Anal Chem 2016; Ding YH, J Biol Chem 2017).

A significant limitation of current cross-linking mass spectrometry techniques is the overdependence of amino-specific NHS ester crosslinkers. Such crosslinkers primarily target lysine, resulting in limited structural information, especially for proteins or protein segments that lack lysine. Therefore, the development of new crosslinkers for other types of amino acids has become an urgent problem to be solved.

In order to obtain more structural information, Dr. Meng-Qiu Dong’s laboratory collaborated with Dr. Xiaoguang Lei’s laboratory of Peking University to develop crosslinkers that target arginine. This led to a series of crosslinkers called ArGO_1-3 each containing a polyethylene glycol chain with an aromatic glyoxal moiety attached to both ends. ArGO can covalently link the guanidine groups of two arginine residues with good selectivity. ArGO makes up for the deficiency of K-K crosslinkers, but the performance on different protein samples is uneven. To improve performance and increase the range of use, the authors replaced one end of the ArGO crosslinker with o-phtalaldehyde, which targets lysine, and named it KArGO. KArGO targets more amino acid residues on the surface of a protein, increases structural coverage, and performs stably well on multiple standard proteins. The authors further applied KArGO to multi-protein complexes such as CNGP and UtpA. The results showed that KArGO compensated for the deficiency of existing lysine crosslinkers and provided abundant complementary structure information.

Fig 1. The cross-linking reaction of KArGO

Another key aspect of CXMS is mass spectrometry data analysis. There are dozens of existing software tools for identification of cross-linking peptides. The analysis results of the same data using different software may differ significantly from one another. Therefore, there is an urgent need in the field for thorough and objective evaluation of data analysis software. So, the authors designed a comprehensive set of evaluation methods using simulation datasets, synthetic peptide datasets, ¹⁵N labeled datasets, and entrapment databases, and used them to gauge ten mainstream software tools, including pLink 1 (Yang B, Nat Meth 2012) and its upgraded version pLink 2, which integrates all the features of pLink 1 and those of the protein disulfide identification software pLink-SS (Lu S, Nat Meth 2015). pLink 2 and pLink 1 ranked at the top in sensitivity and accuracy, far exceeding the other software tools. Taking the simulation data set (the spectrum is almost perfect) test results as an example, at 5% false discovery rate, the sensitivity and accuracy of pLink 1 and pLink 2 reached and exceeded 99.8%, respectively. pLink 2 also has a very obvious speed advantage; in a total of 20 tests with varying data type, data size, and the size of the protein sequence database, pLink 2 is on average 40 times faster than pLink 1 and 3 times faster than Kojak, which ranked the third in sensitivity and accuracy. Since its free public release on the New Year's Day of 2018, pLink 2 attracted more than 1,000 users from all over the world.

Yong Cao (2015 PTN program) in Dr. Meng-Qiu Dong’s Lab at NIBS, Beijing is the co-first author in both articles. The two other co-first authors of the ArGO/KArGO paper are: Dr. Alexander X. Jones, a British postdoctoral fellow, and Yuliang Tang, a doctoral student, both from Dr. Xiaoguang Lei’s lab in Peking University. The corresponding authors are Dr. Xiaoguang Lei and Dr. Meng-Qiu Dong. Other authors include Jian-Hua Wang, Yue-He Ding, Hui Tan, Zhen-Lin Chen, Run-Qian Fang, Jili Yin, Rong-Chang Chen, Xing Zhu, Yang She, Niu Huang, Feng Shao, Keqiong Ye, and Rui-Xiang Sun. The two other co-first authors of the pLink 2 paper are: Zhenlin Chen, a doctoral student, and Jiaming Meng, a master’s student, both from the pFind team of the Institute of Computing Technology, CAS. The corresponding authors are Dr. Meng-Qiu Dong and Dr. Si-Min He. Other authors include Ji-Li Yin, Run-Qian Fang, Sheng-Bo Fan, Chao Liu, Wen-Feng Zeng, Yue-He Ding, Dan Tan, Long Wu, Wen-Jing Zhou, Hao Chi, and Rui-Xiang Sun. Both studies were funded by the Ministry of Science and Technology of the People's Republic of China, The National Natural Science Foundation of China, the Chinese Academy of Sciences, and the Beijing Municipal Government.