Modulation of Ion Channels by Protein Phosphorylation: How the Brain Works

Irwin B. Levitan    Biochemistry Department and Volen Center for Complex Systems, Brandeis University, Waltham, Massachusetts 02454

INTRODUCTION

The subtitle of this chapter is not entirely facetious. The electrical activity of individual nerve cells, and communication among them, are determined by the movement of ions across the plasma membrane through the highly specialized intrinsic membrane proteins known as ion channels. Because the essence of brain function is information transfer, mediated by ion channels, within and between nerve cells, it is safe to say that detailed knowledge about how ion channels work and how they are regulated is necessary for even the most rudimentary understanding of how the brain controls behavior. Of course, ion channels are not restricted to nerve cells, or even to excitable cells, and many of the chapters in this volume will describe the fundamental role of ion channels and their modulation in the functioning of cells of many kinds.

It is instructive to recall that the term “ion channel modulation” was not long ago regarded as something of an oxymoron. The classic studies of Alan Hodgkin and Andrew Huxley defined the roles of voltage-dependent sodium and potassium currents in the generation and propagation of squid axon action potentials as “all-or-none.” Although Hodgkin and Huxley themselves never explicitly excluded the possibility that these ionic currents might be subject to cellular regulation, the very brilliance of their work and the completeness of their description of action potential properties led others to conclude that the properties of the ion channels responsible for these currents might be immutable. Similarly, the pioneering studies of Bernard Katz and his colleagues on synaptic transmission at the frog neuromuscular junction led to the belief that all synapses behave like the neuromuscular junction, and engendered considerable skepticism about early claims for a role for intracellular second messengers in some forms of synaptic transmission.

It is now widely accepted, however, that patterns of neuronal electrical activity and intercellular communication vary greatly. Even closely neighboring neurons can exhibit distinct electrical properties, reflecting their distinct complement of ion channels that are functional under a given set of conditions. At least as important for brain function, and more important from the perspective of this volume, is the fact that these patterns of electrical activity are not fixed, but can be modulated by the actions of neurotransmitters, hormones, or sensory inputs. The medley of modulatory changes observed in excitable cells matches the diversity of endogenous electrical properties themselves, but some patterns can be discerned: modulation generally results in a change in action potential amplitude or duration, in endogenous action potential firing, or in the efficacy of synaptic transmission (Fig. 1). These may not be mutually exclusive—for example, a change in action potential duration will influence neurotransmitter release, hence will modulate synaptic transmission indirectly. As this volume illustrates, cells make use of an enormous assortment of molecular mechanisms to produce these changes. I will focus my discussion on the ubiquitous mechanism that has been most thoroughly studied, modulation by protein phosphorylation.

IONIC CURRENTS ARE MODULATED BY PROTEIN PHOSPHORYLATION AND DEPHOSPHORYLATION

One of the earliest and best understood examples of modulation by protein phosphorylation comes from studies of the prolongation of the cardiac action potential by β-adrenergic agonists. In principle, such an increase in action potential duration could come about either by an increase in the calcium current that underlies the depolarizing phase of the cardiac action potential or by a decrease in the potassium current(s) responsible for repolarization (or both). By the early 1970s, work from the laboratories of Harald Reuter, Wolfgang Trautwein, Richard Tsien, and others had established that voltage-dependent calcium currents in the heart can be enhanced by stimulation of β-adrenergic receptors, and that this modulation is probably mediated by that archetypal intracellular second messenger, cyclic AMP (1). This was followed by the demonstration of a role for cyclic AMP in the modulation of the electrical properties of several different molluscan neurons (2–4). It soon became evident that these actions of cyclic AMP are mediated by protein phosphorylation, by the cyclic-AMP-dependent protein kinase (PKA). For example, injection of the active catalytic subunit of PKA into both cardiac muscle cells and molluscan neurons can mimic the effects of neurotransmitters and cyclic AMP on ionic currents (5–8), and intracellular injection of a specific protein inhibitor of PKA can block them (9,10); some of these early studies are reviewed by Kennedy (11). Subsequent experiments demonstrated that other kinds of protein kinase, as well as phosphoprotein phosphatases, are involved in the modulation of ionic currents in a variety of cell types (12–15).

MOLECULAR MECHANISMS OF MODULATION OF IONIC CURRENTS

Studies such as those just described made it evident that protein phosphorylation is both necessary and sufficient for certain actions of hormones and neurotransmitters on cellular electrical properties, but they did not speak to the question of which proteins are phosphorylated by the kinases. The ion channel proteins, responsible for the ionic currents that are subject to modulation, were obvious candidates, but only with the advent of single channel recording and molecular cloning approaches did it become possible to address this issue directly.

In this chapter I will summarize biophysical, biochemical, and molecular evidence that potassium channels are phosphorylated directly by protein kinases and that this phosphorylation results in channel modulation. It is clear that ion channels other than potassium channels also are subject to this kind of regulation, and I will refer here to several notable examples that help to illustrate some of the points I wish to convey. However, work with voltage-dependent sodium and calcium channels and ligand-gated ion channels is described in more detail in Chapters 2 and 3, and my detailed discussion is restricted to the modulation of several different classes of potassium channels by phosphorylation. Because this literature is becoming rather extensive, I will be selective rather than comprehensive in this summary and will focus on the historical development of the field. I then will review the emerging evidence that some modulatable ion channels exist as part of a regulatory protein complex, which includes the protein kinase and phosphoprotein phosphatase activities that participate in channel modulation. Finally, I will return from this reductionist excursion to a brief consideration of the cellular physiological consequences of channel phosphorylation, and of the intimate association of channels with regulatory enzymes.

POTASSIUM CHANNEL PROTEINS ARE SUBSTRATES FOR PROTEIN KINASES AND PHOSPHATASES

Biochemical Measurements of Potassium Channel Phosphorylation

Because there are tissue sources containing high concentrations of nicotinic acetylcholine receptor/channels and voltage-gated sodium and calcium channels, and specific affinity reagents that can be used for their purification and assay, it has been known for a long time from biochemical measurements that these channels are substrates for a variety of protein kinases. For example, the nicotinic acetylcholine receptor/channel purified from Torpedo is a substrate for both PKA (16) and an endogenous tyrosine kinase (17), and both sodium channels (18) and calcium channels (19) from various tissues can be phosphorylated by PKA. However, no rich tissue sources of potassium channels exist, and thus in only a few cases have comparable protein biochemistry experiments been done on potassium channels. One of the best studied examples is the Kv1.3 potassium channel, which has been cloned from several mammalian tissues and is the major voltage-gated potassium channel expressed in T lymphocytes (where it is known as the type n potassium channel) (20,21). Cai and Douglass (22) prepared antibodies against fusion proteins encoding Kv1.3 sequences and demonstrated using the human Jurkat T lymphocyte cell line that the type n potassium channel is phosphorylated on serine and threonine residues by both PKA and protein kinase C (PKC). The same antibodies were used to demonstrate that cloned Kv1.3, expressed heterologously in a human embryonic kidney (HEK) 293 cell line, can be phosphorylated on tyrosine residues by either endogenous or cotransfected exogenous tyrosine kinases (23).

Potassium Channel Phosphorylation Inferred from Modulation of Function

Because of the difficulty of such biochemical experiments with low abundance membrane proteins, a key approach to determining whether ion...