Collision cross sections in both helium and nitrogen gases were measured directly using a drift cell with RF ion confinement inserted within a quadrupole/ion mobility/time-of-flight hybrid mass spectrometer (Waters Synapt HDMS, Manchester, U.K.). Collision cross sections for a large set of denatured peptide, denatured protein, native-like protein, and native-like protein complex ions are reported here, forming a database of collision cross sections that spans over 2 orders of magnitude. The average effective density of the native-like ions is 0.6 g cm(-3), which is significantly lower than that for the solvent-excluded regions of proteins and suggests that these ions can retain significant memory of their solution-phase structures rather than collapse to globular structures. Because the measurements are acquired using an instrument that mimics the geometry of the commercial Synapt HDMS instrument, this database enables the determination of highly accurate collision cross sections from traveling-wave ion mobility data through the use of calibration standards with similar masses and mobilities. Errors in traveling-wave collision cross sections determined for native-like protein complexes calibrated using other native-like protein complexes are significantly less than those calibrated using denatured proteins. This database indicates that collision cross sections in both helium and nitrogen gases can be well-correlated for larger biomolecular ions, but non-correlated differences for smaller ions can be more significant. These results enable the generation of more accurate three-dimensional models of protein and other biomolecular complexes using gas-phase structural biology techniques.

Collapse to compact states in the gas phase, with smaller collision cross sections than calculated for their native-like structure, has been reported previously for some protein complexes although not rationalized. Here we combine experimental and theoretical studies to investigate the gas-phase structures of four multimeric protein complexes during collisional activation. Importantly, using ion mobility-mass spectrometry (IM-MS), we find that all four macromolecular complexes retain their native-like topologies at low energy. Upon increasing the collision energy, two of the four complexes adopt a more compact state. This collapse was most noticeable for pentameric serum amyloid P (SAP) which contains a large central cavity. The extent of collapse was found to be highly correlated with charge state, with the surprising observation that the lowest charge states were those which experience the greatest degree of compaction. We compared these experimental results with in vacuo molecular dynamics (MD) simulations of SAP, during which the temperature was increased. Simulations showed that low charge states of SAP exhibited compact states, corresponding to collapse of the ring, while intermediate and high charge states unfolded to more extended structures, maintaining their ring-like topology, as observed experimentally. To simulate the collision-induced dissociation (CID) of different charge states of SAP, we used MS to measure the charge state of the ejected monomer and assigned this charge to one subunit, distributing the residual charges evenly among the remaining four subunits. Under these conditions, MD simulations captured the unfolding and ejection of a single subunit for intermediate charge states of SAP. The highest charge states recapitulated the ejection of compact monomers and dimers, which we observed in CID experiments of high charge states of SAP, accessed by supercharging. This strong correlation between theory and experiment has implications for further studies as well as for understanding the process of CID and for applications to gas-phase structural biology more generally.

SummaryIs the order in which proteins assemble into complexes important for biological function? Here, we seek to address this by searching for evidence of evolutionary selection for ordered protein complex assembly. First, we experimentally characterize the assembly pathways of several heteromeric complexes and show that they can be simply predicted from their three-dimensional structures. Then, by mapping gene fusion events identified from fully sequenced genomes onto protein complex assembly pathways, we demonstrate evolutionary selection for conservation of assembly order. Furthermore, using structural and high-throughput interaction data, we show that fusion tends to optimize assembly by simplifying protein complex topologies. Finally, we observe protein structural constraints on the gene order of fusion that impact the potential for fusion to affect assembly. Together, these results reveal the intimate relationships among protein assembly, quaternary structure, and evolution and demonstrate on a genome-wide scale the biological importance of ordered assembly pathways.