Tsr model coordinates courtesy of Sung-Hou Kim, UC Berkeley.


Research Topics

Transmembrane signaling is one of many fundamental processes mediated by membrane proteins that is also medically important, since the majority of current drug targets are membrane receptors. Prokaryotic membrane proteins are the typical targets of detailed structure/function studies in this field; our target is the bacterial chemotaxis receptor. Although extensive studies of these receptors have yielded crystal structures of soluble fragments and evidence for subtle ligand-induced conformational changes, important questions remain regarding the mechanism of signal transmission across the membrane. Our approach is to integrate multiple methods to address key mechanistic questions in bacterial chemoreceptors.


Transmembrane signaling by chemotaxis receptors: conformational changes

Solid-state NMR is an important tool for characterizing structure and function in membrane proteins. We pioneered site-directed solid-state NMR as a powerful approach for high resolution measurements of local structure at sites throughout large membrane proteins. Application to the serine chemotaxis receptor provided the first distance measurements in the intact, membrane-bound receptor with sufficient resolution to measure the subtle changes thought to transmit the signal in this receptor. Results of interhelical distance measurements in the periplasmic and transmembrane domains of the intact receptor are consistent with the proposed ligand-induced piston model for the signaling mechanism (Murphy et al, 2001a; Isaac et al, 2002). As described below, NMR experiments in progress aim to measure multiple simultaneous distances to map conformational changes between receptor signaling states within native-like receptor complexes.

Assembly of chemoreceptor signaling arrays for structure/function studies

In bacteria, chemoreceptors form extended arrays with two partner proteins, a kinase CheA and a scaffolding protein CheW. We have two methods for preparing native-like functional arrays of receptor cytoplasmic fragments for biophysical studies to determine how the receptor structure and dynamics change between signaling states. (A) We use vesicle template assembly, developed by our late colleague Bob Weis, to bind His-tagged receptor fragments to lipid vesicles and assemble functional complexes with CheA and CheW. (B) We discovered that molecular crowding agents are sufficient to mediate assembly of complexes with comparable kinase activity. In collaboration with Grant Jensen's lab (Caltech) we have used electron cryotomography to show that both of these are the best current in vitro preparations for biophysical studies, producing homogeneous arrays with native properties such as lateral dimensions ≈100 nm and 12 nm hexagonal spacing as observed for intact receptor arrays (left cartoons) in cells (Briegel et al, 2014a). Molecular crowding conditions stabilize formation of "sandwich" arrays (right cartoon) for both cytoplasmic fragments in vitro (tomogram) and naturally occuring cytoplasmic receptors in vivo (Briegel et al, 2014b).

We assemble these native-like chemoreceptor signaling arrays for cytoREDOR measurements of their activity, structure, and dynamics using functional assays, NMR, and mass spectrometry. For example, we have used NMR to measure a distance that tests current models and provides the first high-resolution structural constraint for the organization of functional receptor signaling arrays (Fowler et al, 2010a). In collaboration with Jochem Struppe (Bruker Instruments) and Ann McDermott's lab (Columbia University) we have obtained 2-dimensional 13C/13C and 15N/13C correlation spectra of receptors in these native-like arrays (Harris et al, 2016). 2DnmrArrayThe sensitivity and 1 ppm linewidths in these spectra set the stage for NMR experiments that will measure multiple distance constraints within a targeted region of the receptor to investigate conformational changes between signaling states, for example to test the ligand-induced expansion that we have proposed (Sferdean et al, 2012). For this and many systems for which complete structures of individual components or of one state of the complex are known, our approaches enable solid-state NMR to provide key information on multiple mechanistically important states of multi-protein complexes to understand how they operate in the cell.


We have also reconstituted the intact receptor with CheA and CheW to investigate a long-standing controversy over whether changes in receptor clustering are involved in signaling. Titration of ligand inhibition of kinase activity for receptors reconstituted at low and high lipid:protein ratios shows no change in apparent ligand affinity. This result indicates that receptor oligomerization is not involved in the mechanism of kinase control, and led us to propose a ligand-induced expansion signaling mechanism (Sferdean et al, 2012).


Transmembrane signaling by chemotaxis receptors: dynamics

It is widely accepted that the transmembrane signal begins with a ligand-induced 2 Å piston motion of an alpha helix that extends through the periplasmic and transmembrane domains, but how that signal is propagated through the cytoplasmic domain is unclear. Results from our lab (Seeley et al, 1996) and others have shown that the cytoplasmic domain has a dynamic tertiary structure. We have developed a novel hydrogen deuterium exchange mass spectrometry (HDX-MS) strategy (Koshy et al, 2013) for measuring the backbone dynamics of a protein within membrane-bound, functional complexes. Using this approach we demonstrated that the receptor cytoplasmic fragment is significantly stabilized within functional complexes. Such a stabilized structure is more consistent with its role in propagating a 2 Å piston to control the kinase activity of CheA bound at the membrane-distal tip 200 Å away.

We have also shown that a core of stable structure in the receptor cytoplasmic fragment is increased dramatically by changes in only a few residues, suggesting that this flexible domain is poised for stabilization as part of the signaling mechanism (Murphy et al, 2001b). More recently, the Parkinson and Falke labs have proposed that the signaling mechanism involves stabilization and destabilization of different subdomains in the cytoplasmic domain of the receptor. Using our HDX-MS approach, we have shown that differences in hydrogen exchange between signaling states localize to regions of the receptor that are critical for excitation and adaptation (Koshy et al, 2014). Thus it appears that changes in dynamics play a role in propagation of the signal through the cytoplasmic domain, and our HDX-MS approach is an ideal way to quantify these changes. This is a promising approach for measurements of functionally important dynamics within multi-protein complexes comparable to native complexes found in the cell.


These studies are helping to reveal the structural basis of the mechanism of transmembrane signaling.

Methods: solid-state NMR and membrane protein studies

Another area of interest in the lab is development of techniques for membrane protein studies. For example, our site-directed strategy combines molecular biology with solid-state NMR distance measurements to move beyond synthetic peptides or active sites to measure distances at any site in a large protein (Murphy et al, 2001a; Isaac et al, 2002). This review article (Kovacs et al, 2007) summarizes practical strategies for rotational resonance and REDOR distance measurements in large proteins. We developed a novel spicalixn diffusion method for measuring depths within membranes to characterize structure in membrane peptides and proteins (Kumashiro et al, 1998; Gallagher et al, 2004). We demonstrated the utility of a calixarene:fluorotoluene inclusion complex for calibration and optimization of carbon-fluorine REDOR for long distance measurements (Fowler et al, 2010b). We implemented a chemical shift thermometer with a sacrificial sample to demonstrate approaches to optimize NMR experiments while avoiding heat damage to retain functionality in complex NMR samples (Fowler et al, 2012). These reviews (Thompson 2011; Harris & Thompson, 2014) of magic-angle spinning studies of proteins illustrate how a number of laboratories are using solid-state NMR methods to address questions of fundamental interest and of medical importance, with emphasis on the strategies needed to expand solid-state NMR to large multi-protein complexes. Creative combinations of labeling strategies with emerging NMR methodologies will make solid-state NMR a key player in current efforts to understand how proteins work together in the multi-protein complexes that carry out the fundamental processes of life in the cell.










Home Department of Chemistry Research Publications MCB Program CBI Program