Mapping the membrane topology of the diphtheria toxin channel: A study in molecular cartography

Abstract

Diphtheria Toxin (DT) is a 65 kD, two chain, protein. Its N-terminal A fragment kills cells by enzymatically inactivating their protein synthetic machinery, while its C-terminal B-fragment (40 kD) is required for binding of toxin to the cell surface and for translocation of the A fragment from the extracellular space into the cytosol. This B fragment forms ion-conducting channels in cell membranes as well as in artificial lipid bilayers. Channels formed by DT in lipid bilayers are highly pH dependent in specific activity as well as in single-channel properties such as conductance and selectivity.;In this work I explore the pH dependence of single-channel conductance and selectivity in DT channels. In particular, I have tried to identify those residues responsible for channel pH dependence by using a combination of site-directed mutagenesis and single-channel electrophysiology in lipid bilayer membranes.;After an introduction to diphtheria toxin in Chapter 1, I present results demonstrating that a 61 amino acid region within the B fragment is sufficient to form channels with single-channel conductances and selectivities virtually identical to those of the wild-type channel (Chapter 2).;In Chapters 3 and 4, I discuss the properties of channels formed by toxins with point mutations that alter the charge of each potentially titratable residue in the channel-forming region. Of nine such residues, changes at four affect single channel conductance; these are also apparently responsible for the pH dependence of single channel conductance. One, Aspartate 352, is entirely responsible for the effect of trans (i.e. the side opposite toxin addition) pH on conductance; it also affects selectivity in a manner consistent with an entirely electrostatic interaction with permeant ions. I further show that this group is on the trans side of the membrane, and its pK{dollar}\sb{lcub}\rm a{rcub}{dollar} is roughly 5.5 (Chapter 3).;The channel's dependence on cis pH is more complicated (Chapter 4). At least three groups are involved here, one of which is predicted by secondary structure models to be within the pore itself. This group, Glu 362, has proven harder to locate functionally (than Asp 352). It is highly sensitive to the pH in the cis compartment, but seems also slightly accessible to the pH of the trans solution. Based on these observations I predict that it is near, but not necessarily on, the cis side of the membrane. The remaining two groups, Glutamates 326 and 327, seem to work in concert and may well be on the cis side of the membrane.;Based on all of these observations, I propose a model for the membrane topology of the channel-forming region, and speculate on the possible geometry of the pore itself.