Conserved among eukaryotes and prokaryotes, potassium channels modulate neuronal signaling in the brain, regulate cell volume and the flow of salt across epithelia, and control heart rate, vascular tone, and the release of hormones such as insulin. They further protect neurons and muscles under metabolic stress.
Calcium-activated chloride channels (CaCCs) also serve a broad range of physiological functions, including the control of secretion from airway epithelia and exocrine glands. In the nervous system, these chloride channels may regulate neuronal excitability. In green algae and plants that lack voltage-gated sodium channels, electric signal generation may depend on calcium-activated chloride channels.
To understand how these ion channels allow only the physiologically appropriate ions to go through, how they alter channel activity in response to electrical and chemical signals, and how they contribute to neuronal signaling and synaptic plasticity, it is important to determine their molecular identity so that these channels can be studied one at a time—a challenge both for potassium channels and for CaCCs, owing to the heterogeneity, low abundance, and ubiquity of these channels.
Starting with positional cloning of the Shaker voltage-gated potassium (Kv) channel gene in the fruit fly in the 1980s, expression cloning of a mammalian inwardly rectifying potassium (Kir) channel in the 1990s, and the recent expression cloning of the calcium-activated chloride channel (CaCC) from frog oocytes, we have been studying potassium channels of two large and distantly related families, as well as calcium-activated chloride channels belonging to a novel ion channel family—the TMEM16 family of transmembrane proteins with unknown function. All three ion channel families are highly conserved in evolution.
Potassium channel mutations cause diseases of the brain (epilepsy, episodic ataxia), ear (deafness), heart (arrhythmia), muscle (myokymia, periodic paralysis), kidney (hypertension), and pancreas (hyperinsulinemic hypoglycemia, neonatal diabetes) and developmental abnormalities of neural crest–derived tissues (Andersen's syndrome). Moreover, potassium channel openers and blockers have been developed for pharmaceutical purposes.
Calcium-activated chloride channels and the CFTR (cystic fibrosis transmembrane regulator) chloride channels act jointly to maintain the luminal airway surface liquid at the right level for airway epithelial cilia to remove irritants and pathogens. In the smooth muscle, calcium release from internal store due to G protein-coupled receptor activation will activate CaCCs, leading to membrane potential changes that open calcium channels to sustain smooth muscle contraction. Therefore, CaCC modulators may be considered for the treatment of hypertension, cystic fibrosis, and other pulmonary diseases, such as asthma and chronic bronchitis. Furthermore, like the KCNK9 potassium channel gene, the TMEM16A calcium-activated chloride channel gene is amplified or up-regulated in several types of human carcinomas. This raises the question whether channel modulators could be considered for treatments.
One unique advantage in channel studies is the possibility of examining one channel at a time, with submillisecond resolution, for many seconds, in experimentally determined intracellular and extracellular environments. In addition to conducting biophysical, biochemical, and cell biological studies of channel assembly, trafficking, regulation, and function, we need to learn how these channels are targeted to specific subcellular compartments of neurons and how they respond dynamically to neuronal activity and in turn modulate neuronal signaling.
Potassium Channel Functions Probed by Yeast Screens of Randomly Mutagenized Mammalian Kir Channels and Plant Kv Channels
To complement structure-function studies based on site-directed mutagenesis, we took advantage of the ability of animal and plant potassium channels to rescue potassium-transport-deficient yeast for growth in low-potassium medium. We screened many hundreds of thousands of randomly mutagenized channels and found these unbiased mutant screens instructive.
For the Kir channels with two transmembrane segments per α subunit, we deduced a helix-packing model for the Kir2.1 channel distinct from that based on the KcsA structure. We verified our model by isolating second-site suppressors predicted to be near the conditional lethal mutations they suppress. In sequence minimization experiments, we showed that 40 M1 residues predicted to face lipid can all be substituted with the same hydrophobic amino acid, regardless of its size, without losing channel function. Moreover, simultaneous substitution of 16 M2 residues predicted to line the pore with the negatively charged aspartate residues also results in functional channels. Remarkably, these 40 lipid-facing and 16 pore-lining residues are situated, as predicted by our model, in the KirBac1.1 channel structure subsequently solved by Declan Doyle's laboratory (University of Oxford), and—as to be expected—all positions intolerant of substitution are buried within the channel protein.
For the Kv channels with six transmembrane segments per α subunit, we employed a similar strategy to deduce a helix-packing model for the plant channel KAT1 at the "down" state adopted by the channel when the electrical potential on the intracellular side of the membrane is much more negative than that outside the cell (i.e., when the membrane potential is hyperpolarized). We have experimentally confirmed predicted interactions between the voltage sensor and the pore domain, thus verifying our model for the channel at the down state.
The unbiased yeast screens have also provided insights on features that enable potassium channels to allow potassium but not the slightly smaller sodium ions to go through (known as potassium selectivity). Starting with random mutagenesis of a mutant Kir3.2 channel that is constitutively active but allows sodium as well as potassium to go through its pore, thereby compromising yeast growth, we have isolated channels that have "evolved" to become selective for potassium permeation and hence can support yeast growth. The surprising finding that potassium selectivity can be restored by electrostatic stabilization of ions, in a region of the pore (the channel cavity) that is much wider than the selectivity filter, can be accounted for in a kinetic model for the long pore of a potassium channel that harbors multiple ions.
The ability of constitutively active Kir3.2 channels to influence yeast growth according to ion selectivity has enabled yeast mutant screens leading to the identification of evolutionarily conserved proteins important for channel traffic and function.
Axonal Targeting and Local Translation of Kv1 Channels in Hippocampal Neurons
Axonal Kv1 channels in the Shaker family enable action potentials to invade a physiologically appropriate number of axonal branches without bouncing back from the nerve terminals; hyperexcitability caused by altered Kv1 channel activity accounts for the symptoms of patients with episodic ataxia type 1 and the shaking phenotype of Shaker mutant flies. To understand the regulation of axonal Kv1 channels, we identified the axonal targeting machinery, and found that the microtubule plus end–binding protein EB1 and the KIF3 kinesin motor are required for Kv1 channel axonal targeting.
Kv1 channels also reside in somatodendritic regions of central neurons. We have found Kv1.1 mRNA in the dendrites, and uncovered activity regulation of dendritic Kv1.1 local translation—a positive feedback in which increased excitatory synaptic inputs causing activation of the NMDA glutamate receptors leads to suppression of dendritic Kv1.1 channel expression and enhanced excitability.
