Dina Schuster

Flexible domains on membrane proteins and how to study them – combining cryo-EM, limited proteolysis-coupled and crosslinking mass spectrometry

Diane C.A. Barret1 *, Dina Schuster1, 2, 3 *, Matthew J. Rodrigues1, Alexander Leitner2, Paola Picotti2, Gebhard F.X. Schertler1, U. Benjamin Kaupp4, 5, Volodymyr M. Korkhov1, 3 and Jacopo Marino1

1 Laboratory of Biomolecular Research, Paul Scherrer Institute, Villigen, Switzerland.
2 Institute of Molecular Systems Biology, Department of Biology, ETH Zürich, Zürich, Switzerland.
3 Institute of Molecular Biology and Biophysics, Department of Biology, ETH Zürich, Zürich, Switzerland.
4 Life and Medical Sciences Institute LIMES, University of Bonn, Bonn, Germany.
5 Max Planck Institute for Multidisciplinary Sciences, Göttingen, Germany
* Contributed equally

Introduction: Membrane proteins, due to their low expression levels and high hydrophobicity, are difficult to study and underrepresented in common bottom-up proteomics experiments. They often contain large stretches of flexible, partially disordered domains, which are not amenable to conventional structural biology techniques, such as cryo-EM and X-ray crystallography. We use a combination of cryo-EM, limited proteolysis-coupled mass spectrometry (LiP-MS) and crosslinking mass spectrometry (XL-MS) to obtain a comprehensive understanding of how calmodulin (CaM) interacts with the cyclic nucleotide-gated ion channel (CNG) of retinal rods. We apply LiP-MS on native retinal membrane suspensions and XL-MS on purified protein to gain insight in structural features that cryo-EM could not resolve.

Methods: For LiP-MS experiments, suspensions of bovine retinal membranes were titrated with increasing concentrations of CaM ranging from 0-3 µg, followed by pulse proteolysis with proteinase K and a tryptic digest under denaturing conditions. Data was acquired in DIA mode on a Thermo Scientific Orbitrap Exploris 480 mass spectrometer and analyzed using Spectronaut (Biognosys). The purified CNG channel was crosslinked with CaM using primary amine crosslinking with disuccinimidyl suberate. Data was acquired on a Thermo Scientific Fusion Lumos Tribrid mass spectrometer and analyzed using xQuest.  For single-particle cryo-EM, proteins were solubilized from membrane preparations and purified by affinity-  and size-exclusion chromatography. Samples were blotted onto carbon grids, plunge-frozen, and data was collected on a Thermo Scientific Krios G4 electron microscope.

Preliminary data: Our data shows that LiP-MS and XL-MS provide complementary information about the interaction of the CNG channel with CaM. In particular, these methods give structural information on flexible domains that are invisible to cryo-EM. LiP-MS, a method based on the use of an unspecific protease (proteinase K), which cleaves accessible protein regions, shows changes in protease accessibility upon interaction. XL-MS provides proximity information through covalent linkage of primary amines.  LiP-MS enables structural and interaction studies in situ without the need for protein purification and due to the high sensitivity of mass spectrometry requires only minute amounts of the sample. The method is advantageous for structural studies of membrane proteins as these proteins can be studied directly in their native lipid environment, avoiding the use of detergent and laborious purification procedures. The additional use of XL-MS together with LiP-MS helps distinguish direct interaction sites from conformational changes, thereby overcoming a limitation of LiP-MS. We show that the CNG channel behaves similar in detergent and in its native lipid environment. We identify that, in the presence of saturating Ca2+ concentrations, CaM binds to two different sites on the CNG channel, with one of them being close to the coiled-coil region of the protein. Cryo-EM studies in presence and absence of CaM show that the channel adopts a more compact structure upon interaction. LiP-MS identifies the interaction sites of CaM and additional accessibility changes on flexible domains of the protein. XL-MS helps interpret both the cryo-EM and LiP-MS results by identifying the specific binding site of CaM visible in the cryo-EM structure but unresolved due to low resolution. Taken together, the combination of cryo-EM, LiP-MS, and XL-MS can help better understand the function of membrane proteins and the role of their flexible domains.

Novel aspect: LiP-MS and XL-MS help study flexible domains on membrane proteins and features that cryo-EM cannot resolve.

Event Timeslots (1)

ictadmin ictadmin