Native State of Complement Protein C3d Analyzed via Hydrogen Exchange and Conformational Sampling

Background: Hydrogen/deuterium exchange detected by mass spectrometry (HDX-MS) is an experimental technique that provides valuable information about a protein’s structure and dynamics. Data produced by HDX-MS experiments is often interpreted using a crystal structure of the protein. However, it has been suggested that more accurate interpretations can be derived from conformational ensembles produced by molecular dynamics simulations than from crystal structures. This assumes that the experimental data can be first computationally replicated from such conformation(s), using a prediction model to derive HDX-MS data from a protein’s structure. Results: In this paper, we analyze the complement protein C3d through HDX-MS data, and we evaluate several methodologies to interpret this data, using an existing HDX-MS prediction model. Although crystal structures of C3d are available, little is known about the variability of its native state. To bridge this gap, since HDX-MS data is known to reflect a protein’s inherent flexibility, we perform an HDX-MS experiment on C3d. To interpret and refine the obtained HDX-MS data, we then need to find a conformation (or a conformational ensemble) of C3d that allows computationally replicating this data. First, we confirm that a crystal structure may not be a good choice for that. Second, we suggest that, even though they are indeed a better choice, conformational ensembles produced by molecular dynamics simulations might not always allow replicating experimental data. Third, we show that coarse-grained conformational sampling of C3d can produce a conformation from which C3d’s HDX-MS data can be computationally replicated and refined from the peptide to the residue level. Conclusions: In the case of the model protein C3d, the similarity between the conformation generated by coarse-grained conformational sampling and available crystal structures establishes that C3d’s native state presents little conformational variability. Yet the ability to obtain a structural model of C3d in solution that provides a good correspondence to experimental HDX-MS data and allows refining this data may prove highly valuable. This could impact many HDX-based applications, from structural analyses to ligand-interaction studies.