Structural mass spectrometric analysis of protein assemblies

Abstract

Structural analyses of protein complexes reveal active sites, subunit interfaces and allosteric regulatory sites. Mass Spectrometry (MS) complements and enhances these approaches and can provide biologically meaningful insights in cases where high resolution structures are not available due to difficulties associated with in vitro reconstitution, instability or the transient nature of the interactions. Additionally, removing a complex from its cellular environment may affect changes in molecular structure; thus, the ability to characterize protein-protein interactions in vivo is invaluable. In this thesis, I focused on the biophysical characterization of assemblies that modulate the actin cytoskeleton. The work is divided into two major areas: (1) The structural dynamics of the de novo actin filament nucleator, Arp2/3; and (2) The development of new technologies to examine protein interactions in vivo, using the actin severing protein cofilin as a model. First, I describe efforts to define the activation mechanism of Arp2/3, a stable protein complex consisting of seven subunits that promotes branching of new actin filaments when ATP and a small activator peptide are bound. The apo-form of this complex has been crystallographically characterized, but attempts to obtain crystals of the peptide-activated form have not been successful, leaving unanswered questions about conformational changes and specific binding interfaces. To address this problem, amide hydrogen/deuterium exchange (HDX) coupled to MS was applied to the Arp2/3 complex in the absence and presence of ATP and activating peptide, providing novel insights into the structural and dynamic alterations that accompany activation. I also describe initial efforts to develop an in vivo protein footprinting approach, using cofilin as a model. The structure of cofilin is known and a number of in vitro structural models exist for the cofilin-actin complex; however, questions remain about the detailed aspects of its in vivo regulation. Although these experiments proved inconclusive, they represent initial steps towards the realization of a method that could support the in vivo structural analysis of protein surfaces, which does not require recovery of the intact complex or its in vitro reconstitution, potentially providing models that are unobtainable with current methods.