Ion channels are basic elements of molecular hardware in the nervous system. These membrane-spanning proteins directly mediate the transmembrane ionic fluxes giving rise to electrical signals in neurons and other electrically active cells. All proteins of this type have a common structure: that of a water-filled pore spanning the cell membrane. As a result, channels act as "leak" pathways for ions down their transmembrane gradients. The ion channels involved in neuronal function are highly intelligent leaks: they select strongly among the different species of inorganic ions present in the aqueous solutions bathing the cell membrane, and they have the ability to open and close their conduction pores in response to external signals, such as binding of neurotransmitters (ligand-gated channels) or changes in transmembrane electric field (voltage-dependent channels).
My research is aimed at questions of fundamental molecular mechanisms of ion channel operation and the underlying protein structures involved. Until recently, when the first direct structure determinations of ion channels began to emerge, it was necessary to draw structural inferences from close examination of ion channel function. This can be done because ion channels can be studied at the single-molecule level, both in the cellular environment (using patch-recording techniques) or after reconstitution into biochemically defined artificial membranes. These high-resolution recording methods can now be combined with direct structure determination to illuminate the operations of these membrane proteins. My lab is focusing on several ion channels of known structure that provide opportunities to address mechanistically important questions about channel structure and function. We are also in the midst of determining the structure of a polyamine transporter.
Cl– Channels and Transporters in the Androgynous CLC Family
A class of Cl–-conducting ion channels has recently been recognized as centrally involved in many cellular electrical processes. This large molecular family of CLC-type Cl– channels is expressed in nearly all cells. For example, CLC channels set the electrical excitability of mammalian skeletal muscle, permit acidification of endosomes, and regulate blood volume (and hence pressure) via renal epithelial transport. Certain inherited human myotonias and several human renal and bone diseases result from disruption of CLC channel genes. Several years ago, we identified CLC homologs in bacterial and archaeal genomes and showed that one of these conducts Cl– when reconstituted in liposomes. The high-resolution structure of this bacterial homolog was determined in Roderick MacKinnon's lab (HHMI, Rockefeller University). We recently stumbled upon a startling conclusion: the bacterial homolog is not an ion channel at all. Instead, it moves Cl– ion across the membrane by an entirely different mechanism: stoichiometric exchange for protons. Thus, the CLC molecular family is androgynous, containing isoforms of two mechanistically different types, Cl– channels and Cl–/H+ exchange-transporters. This situation is now recognized to be a general feature of eukaryotic CLCs as well. The human genome contains nine CLCs: four channels and five transporters. We are currently trying to understand the mechanistic implications of this unusual finding.
We are examining the mechanism by which the bacterial CLC protein couples Cl– to H+ movement, by relating the electrophysiological and kinetic behavior of functionally altered mutants to their x-ray crystal structures. In particular, HHMI associate Hariharan Jayaram is working on structures of mutants of the transporter that have been "broken" so as to form true Cl–-selective channels. These channels are much slower in ion-throughput rate than CLC channels designed by evolution, but they are much faster than any known transporter protein. The bacterial homolog studied here is so far the only prokaryotic CLC protein that has been functionally reconstituted. Jayaram is working to fill out our picture of this unusual Cl– channel–transporter family by expressing, functionally analyzing, and crystallizing a panel of bacterial CLC homologs of the channel subtype. In parallel, HHMI associate Hyun-Ho Lim is seeking to discover the specific pathway by which protons move through the transporter protein in exchange with Cl–.
The Ion Pathway of a K+ Channel
K+ channels are arguably the most deeply understood membrane proteins, since several of them permit both examination of high-resolution structures and close-up analysis of single-molecule function. Graduate student Kene Piasta is exploiting this combination of capabilities to understand the energetics of ion binding to the conduction pore of KcsA, a prokaryotic K+ channel of known structure. Recent work from Eduardo Perozo's laboratory (University of Chicago) has shown how to produce mutants of KcsA that never close. The binding of K+ and its ionic analogs to these constitutively open pores can be examined rigorously via a method developed in our lab many years ago. This method uses Ba2+, a hard divalent cation with a crystal radius precisely matching that of K+. This ion, a strong and sticky blocker of many K+ channels, including KcsA, may be used to probe the ion pathway.
Ion Transport in Bacterial Extreme Acid Resistance
The extreme acid resistance response (XAR) by which enteric bacteria such as Escherichia coli protect themselves against the low pH of the stomach involves several amino acid decarboxylase genes working in concert with several membrane transport proteins (including the CLC Cl– transporter). Among these is an arginine transporter of the "APC" superfamily that functions physiologically as a "virtual proton pump" to prevent acid overload of the cytoplasm during acid shock. HHMI associates Yiling Fang and Hariharan Jayaram have recently determined the structure of this transporter as a first step toward understanding the mechanism by which the protein selects finely for substrates—an essential requirement for its proton-extrusion function.
