5. Conclusions and Perspectives
Efforts to map a proteome-wide PPI network have made significant contributions to our basic understanding of biological systems (Rual et al., 2005). Scientists increasingly have deeper and comprehensive insights into aberrant PPIs occurring in various pathological conditions to strategically come up with effective therapeutic avenues (Ryan and Matthews 2005). To this end, rational design of a potent modulator that best counteracts abnormal PPIs would require detailed knowledge on a binding interface of PPI with high fidelity comprising not only atomic resolution but also “reality” existing in native environments. CLMS and HDMS discussed herein are not competent enough for producing atomic resolution of an interface comparable to that by X-ray crystallography and EM. However, in terms of reality, they are especially useful in extracting structural information in biological, native environment including contact residues at interfaces, inter-residue distances, peptide-level solvent accessibilities, and temporal dynamics of interfacial areas, from which refined and near-physiological interfacial landscape can be remodeled preferably using a preexisting structure as a template. Exciting computational advances in data processing and structural biology are greatly improving the fidelity of integration of data obtained from both convention methods and CLMS/HDMS. High-fidelity information of an interface thus acquired would serve as an excellent cornerstone, increasing the likelihood of successful rational drug discovery. Designer agonists or antagonists that recognize hot residues in an interface responsible for a targeted PPI in a specific pathological condition are expected to outperform conventional binders with regard to both potency and efficacy.
Limitations of CLMS and HDMS remain to be overcome in future advances. Specifically, CLMS users often suffer from excessive crosslinking condition optimization and non-specific reactions that would compromise precise measurement of spatial distances. Back exchange of deuterium in HDMS hampers the reliable interpretation of solvent accessibility dataset, necessitating effective sample treatment protocols by tricky trial and error approach. Developments of new crosslinkers with residue-specificity, bifunctionality, and/or reduced background mass signals are expected to address the current challenges of CLMS (Ding et al., 2016; Leitner et al., 2014; Lössl et al., 2014; Schneider et al., 2018). Developments of improved workflows in HDMS should prove to be innovative in the coming era (Hamuro and Zhang 2019; Lau et al., 2019; Oganesyan et al., 2018). Together with novel MS instrumentation and computational data processing techniques, all the efforts will continue to advance CLMS and HDMS as excellent polishing tools for high-fidelity structural determination of protein complexes and binding interfaces thereof, thereby fueling the growth of structural proteomics and biology, and, further, the advent of structure-based drug design regime with exceptional reliability and probability (Figure 4).