Transmembrane protein translocation is a fundamental and ubiquitous process that operates under several conserved principles. Protein conducting machines, known as translocase channels, catalyze the unfolding and subsequent transport of substrate proteins through narrow pores in an energy-dependent manner. Investigating how translocase machines rectify external energy inputs is key to understanding their force transduction mechanisms and overall function. Molecular ratchets are believed to be an integral mechanism for providing directional polypeptide transport through translocase channels, and several types of ratchets have been identified in these machines.

In this work, anthrax toxin, a protein translocase that possesses a hydrophobic gasket known as the φ clamp, is used as a model system to investigate the role of hydrophobic/steric ratcheting in

protein translocation. Planar lipid bilayer electrophysiology was used to establish that the PA φ clamp binds hydrophobic/aromatic residues in peptide substrates, and to determine the consequences this binding interaction has on transport of these peptides. Under conditions where protein unfolding is shown to be rate-limiting, sequences identified as φ-clamp interacting increase translocation of folded downstream domains.

Single channel electrophysiology was used to record substrate and driving force dependent changes in the conductance of a single PA channel. Open, bound, and two intermediate conductance states are identified, which we hypothesize correspond to different conformations of the catalytic φ-clamp loop. Crystallographic evidence, which supports this theory, is presented. A kinetic analysis of single channel peptide binding and translocation events shows that the translocation pathway for hydrophobic/aromatic sequences has committed steps. A hydrophobic ratchet model is proposed to explain φ-clamp catalysis and how the favorable interaction between the clamp and hydrophobic/aromatic sequences may favor unfolding and increase directional movement of the translocating chain.

The motivations for this study stem from an uncertainty about whether on-demand ridehailing services such as Uber, Lyft and others, will exacerbate existing transportation issues, or help alleviate them. To that end, the goals of the project were to learn about the perspectives of stakeholders from a variety of sectors, on their reactions to policies and other actions that might enable on-demand services to help alleviate existing transportation issues including congestion, emissions and inequality of access and mobility.

This study aims to address the following three questions: How well do stakeholders in different sectors and regions, agree about the potential outcomes related to on-demand ridehailing and sustainable transportation goals? What are stakeholder perspectives on the policies and strategies that might facilitate emerging on-demand transportation services to most effectively enhance sustainability and mobility outcomes? What decision making venues and approaches are supported by different stakeholders in the process, and how can these approaches be pursued in order to realize policy goals related to sustainability of on-demand ridehailing? I.e., what venues, and at what level can most effective policies be introduced to facilitate sustainability improvements in transportation by embracing on-demand ridehailing services.

To answer these questions, a series of interviews were completed with stakeholders from California Metropolitan Planning Organizations (MPOs) and Regional Transportation Planning Agencies (RTPAs), from state agencies, the ridehailing industry, and local planning agencies or transportation divisions of cities. The results of this study indicate that policy makers must consider the varied systems and contexts throughout the state; and likely throughout the US. Further, there is an existing dialogue on these topics among transportation professionals, public interest groups, academics and policy makers. In this study, the researchers took a systematic approach to documenting this dialogue and identifying meaningful messages and policy guidance that is not possible without a rigorous scientific approach.

View the NCST Project Webpage

Molecular machines face a number of challenges in transporting a protein either across a membrane or into a proteolytic complex. In many cases, a substrate protein must first be unfolded before being transported through a narrow channel. Despite its importance and relevance to a variety of different processes in the cell, translocation-coupled protein unfolding is still not well understood. In effort to determine the general biophysical mechanisms of this process, anthrax toxin is used as a model system.

In order to understand how a protein unfolds on a translocase channel, planar lipid bilayer electrophysiology, site-directed mutagenesis and thermodynamic stability studies were used to first identify the barriers in the translocation pathway and determine which barrier corresponds to substrate unfolding. Working under conditions where substrate unfolding is rate-limiting, we were then able to map how LF_N actually unfolds on the surface of the PA channel.

Next, the role of the channel in substrate unfolding and translocation is discussed. In particular, a novel substrate binding site on the surface of PA was identified from the crystal structure of a PA octamer bound to four LF_N substrates. This structure, which was solved by my colleague, Geoffrey Feld, reveals that the first α helix and β strand of each LF_N unfold and dock into a deep amphipathic cleft, termed the α clamp. Through extensive mutatgenesis studies on both PA and LF_N, Geoff and I determined that this site can bind a broad array of polypeptide substrates. The role of the α clamp in substrate unfolding, channel oligomerization and translocation is investigated and discussed.

Finally, in effort to further probe the α clamp's role in translocation, binding to the site is disrupted and the effects on translocation are investigated. Preliminary hypotheses and future directions are discussed.

Because proteins carry out their cellular functions in specific locations, protein transport represents a crucial step in the central dogma of biochemistry. Transport across a membrane barrier requires energy; therefore, nature has provided a class of molecular machines known as translocases that catalyze protein translocation. Often, these machines contain channels that are too narrow to transport a folded protein; thus, substrate unfolding adds another layer of complexity to the poorly understood transport mechanism. The anthrax toxin system represents a useful model for elucidating the mechanisms of translocation-coupled protein unfolding, both in terms of structure and function.

In order to understand the mechanism by which an unfolding machine interacts with its substrate, the X-ray structure of an anthrax protective antigen (PA) oligomer prechannel was solved in complex with the amino-terminal PA-binding domain of the substrate lethal factor (LF_N). The structure revealed how PA interacts with unfolded polypeptide via a hydrophobic cleft we call the &alpha clamp. Working with my colleague, Katie Thoren, we show how the &alpha clamp is critical for toxin assembly, substrate binding, unfolding, and translocation through a nonspecific binding mechanism.

The recent discovery that PA can form two oligomeric states exemplifies the complexity of multimeric protein nanomachine assembly. In order to better understand the molecular mechanism of heterogenous assembly, a series of PA constructs with varying length crosslinks was produced, their structures were solved to high resolution, and their assembly products were analyzed by electron microscopy. The flexibility of PA's receptor-binding domain (D4) relative to the main body of the protein provides a mechanism for controlling oligomeric stoichiometry, whereby D4 can adopt a pro-heptamer or pro-octamer conformation.

Finally, an overall model for protein translocation is presented, based on data discussed herein and other recent studies involving anthrax toxin transport.

Abstract

Structure-Function Studies of Anthrax Protective Antigen Octamer

by

Alexander Frederick Kintzer

Doctor of Philosophy in Chemistry

University of California, Berkeley

Professor Bryan Krantz, Chair

The assembly of bacterial toxins and virulence factors is critical to their function, but the regulation of holotoxin assembly during infection has not been studied. The protective antigen (PA) component of the toxin assembles into ring-shaped homooligomers that bind the two other enzyme components of the toxin, lethal factor (LF) and edema factor (EF), to form toxic complexes. To disrupt the host, these toxic complexes are endocytosed, such that the PA oligomer forms a membrane-spanning channel that LF and EF translocate through to enter the cytosol.

We found using single-channel electrophysiology that PA channels contain two populations of conductance states, which correspond with two different PA pre-channel oligomers observed by electron microscopy--the well-described heptamer and a novel octamer. Octameric complexes are likely integral to the PA assembly mechanism, as dimeric and tetrameric intermediates are the intial populated species during assembly. While complexes are functional translocases, a 3.2Å crystal structure suggests that the PA octamer prechannel may be more thermodynamically stable than the heptamer. Indeed, the octamer comprises 20-30% of the oligomers on cells, but predominates (70-80%) during assembly in physiological plasma.

To probe the mechanism of octamer stability, we measured the pH-dependence of channel formation using circular dichroism, mass spectrometry, and electron microscopy. We found that octamers form channels at lower pH than heptamers. This results in heptamer inactivation under physiological conditions, while LT complexes containing octameric PA maintain maximal cytotoxic activity. Thus the PA octamer is a physiological, stable, and active assembly state capable of forming lethal toxins that may withstand the hostile conditions encountered in the bloodstream. This assembly mechanism demonstrates a novel means to control cytotoxicity during anthrax infection.

Finally, we studied the mechanism of channel formation. We found from structural, thermodynamic, and kinetic studies that two barriers define the prechannel to channel transition--a pH-dependent and independent process. Further analysis revealed that the protonation of the membrane insertion loop, dissociation of domains 2 and 4, and interactions with phenylalanine-clamp comprise the rate-limiting, pH-dependent barrier.

Sensory systems rely on neuromodulators, such as serotonin, to provide flexibility for information processing in the face of a highly variable stimulus space. Serotonergic neurons broadly innervate the optic ganglia of Drosophila melanogaster, a widely used model for studying vision. The role for serotonergic signaling in the Drosophila optic lobe and the mechanisms by which serotonin regulates visual neurons remain unclear. Here we map the expression patterns of serotonin receptors in the visual system, focusing on a subset of cells with processes in the first optic ganglion, the lamina, and show that serotonin can modulate visual responses. Serotonin receptors are expressed in several types of columnar cells in the lamina including 5 HT2B in lamina monopolar cell L2, required for the initial steps of visual processing, and both 5 HT1A and 5 HT1B in T1 cells, whose function is unknown. Subcellular mapping with GFP-tagged 5 HT2B and 5 HT1A constructs indicates that these receptors localize to layer M2 of the medulla, proximal to serotonergic boutons, suggesting that the medulla is the primary site of serotonergic regulation for these neurons. Serotonin increases intracellular calcium in L2 terminals in layer M2 and alters the kinetics of visually induced calcium transients in L2 neurons following dark flashes. These effects were not observed in flies without functional 5-HT2B, which displayed severe differences in the amplitude and kinetics of their calcium response to both dark and light flashes. While we did not detect serotonin receptor expression in L1 neurons, they also undergo serotonin-induced calcium changes, presumably via cell non-autonomous signaling pathways. We provide the first functional data showing a role for serotonergic neuromodulation of neurons required for initiating visual processing in Drosophila. These findings demonstrate that tracing the molecular mechanisms of serotonergic modulation require a combination of receptor mapping to individual cells and subcellular compartments with complementary functional assays. We have identified a testable model system for these studies and developed new tools for studying the serotonin transporter.

A fundamental problem in biochemistry is how molecular machines convert chemical potential energy into mechanical work. Here, this problem is addressed in terms of how a proton gradient (ΔpH) drives anthrax toxin transmembrane protein translocation with a given directionality. One model to explain this phenomenon is a charge-state Brownian ratchet, wherein movement is biased by asymmetries across the membrane. Acidic residues in the substrate are protonated on the lower pH starting side, thus allowing them to pass through the cation-selective protective antigen (PA) channel. Movement can occur in either direction according to Brownian motion, but acidic residues that reach the higher pH of the destination are deprotonated, preventing retrotranslocation and resolving movement in one direction.

This model is probed through the use of planar lipid bilayer electrophysiology to analyze the charge requirements of the model substrate LF_N, the binding domain of the natural substrate lethal factor. Acidic residues are necessary and sufficient for ΔpH-driven translocation provided the starting side pH is low enough to sufficiently protonate the residues. Their position in the sequence, just before LF_N's folded domain, plays a key role in substrate unfolding. Basic charges are important for initiation into the channel and to chaperone deprotonated acidic residues during translocation.

Further work by Sarah Wynia-Smith confirms the importance of acidic residues just before LF_N's folded domain and points to the existence of electrostatic translocation barriers in the PA channel. I identify key residues in the upper portion of the channel's β barrel that contribute to its cation selectivity. These residues also prove to play a significant role in substrate initiation and translocation, supporting the role for a charge gate in the channel that prevents retrotranslocation.

Finally, secondary structure of the translocating substrate is analyzed. LF_N translocates in a compact helical state nucleated by the channel's α clamp, a nonspecific helix-binding site recently identified by Geoffrey Feld. α clamp-induced helix formation lowers a previously identified but uncharacterized translocation barrier, and substrates that cannot form helices cannot translocate. This suggests a mechanism wherein helix melting, coupled to the electrostatic ratchet, drives translocation-coupled substrate unfolding.

Parkinson's disease (PD) is a neurodegenerative disorder that affects millions of people worldwide and is characterized by a loss of dopaminergic neurons along with other deficits. While rare inheritable forms of PD exist, the vast majority of cases are sporadic and their etiology remains poorly understood. Epidemiological studies have increasingly linked exposures to environmental toxins, including a variety of pesticides, to an increased risk of contracting PD. However, it remains unclear as to the mechanism by which these pesticides cause neurotoxicity and the extent to which individual genetic variability plays a role. Alterations in the vesicular monoamine transporter (VMAT) gene, which packages dopamine and other amines into synaptic vesicles, have been shown to affect PD susceptibility in animal models and epidemiological studies alike. Using live-imaging techniques at the Drosophila melanogaster neuromuscular junction, I investigated the functional consequences of altered VMAT trafficking at an intact synapse. I show that mutations in putative VMAT trafficking domains result in abnormal protein localization during synaptic vesicle exo- and endocytosis, in a manner that is dependent on nerve terminal identity. Using the same approach, I show that the PD-linked pesticide ziram differentially disrupts neuronal activity and synaptic vesicle cycling at functionally distinct nerve terminals. In particular, I demonstrate that endocytosis at aminergic nerve terminals is particularly sensitive to both disruption of VMAT trafficking motifs and exposure to ziram, and that ziram appears to selectivity induce the spontaneous depolarization of these terminals. Additionally, using traditional toxicological approaches, I developed a novel model of paraquat and maneb induced PD in Drosophila melanogaster and demonstrate the general utility of this model for interrogating specific mechanisms implicated in PD etiology. This dissertation discusses my research using Drosophila melanogaster to address basic biological questions fundamental to understanding gene-environment interactions implicated in PD.