Membrane-spanning proteins must segregate to specific cellular locations to perform physiological roles, such as rapid signaling in neurons and heart, that distinguish vertebrates from other metazoans. Our laboratory has discovered the ankyrin family of adapters and their role in the organization of a surprisingly diverse set of proteins and membrane domains that are likely to play pervasive roles in vertebrate physiology. We have resolved a simple "code" for ankyrin binding by membrane proteins that has independently evolved in many protein families, including protein families with specialized roles in vertebrates. We have found that ankyrin proteins are required for localization of membrane partners and for formation of specialized membrane domains, including axon initial segments of neurons and lateral membrane domains of epithelial cells. We are exploring the molecular mechanisms underlying ankyrin function as well as the pathological consequences—such as cardiac arrhythmia and diabetes—when these mechanisms fail.
Ankyrin-G Is the Master Organizer of Axon Initial Segments
Axon initial segments are critical sites of integration in the vertebrate nervous system, where thousands of dendritic signals are summed, resulting in a single neuronal response in the form of action potentials. Action potentials are generated by voltage-gated sodium channels (Nav), which are highly concentrated at axon initial segments. Our laboratory became interested in Nav and axon initial segments through a serendipitous route. We initially discovered ankyrin and its role in connecting the spectrin-based membrane skeleton to the cytoplasmic domain of the anion exchanger in human erythrocytes. We then found that ankyrin-related proteins are expressed in most cells, and we began to search for their membrane protein partners. We discovered with collaborators that an uncharacterized ankyrin was associated with voltage-gated sodium channels isolated from brain tissue. We next determined that this ankyrin co-localized with sodium channels in excitable membranes in the mammalian nervous system. We cloned this sodium channel-associated ankyrin, which we termed ankyrin-G.
We selectively knocked out expression of ankyrin-G in the postnatal cerebellum in mice and found that ankyrin-G is required for cerebellar neurons to fire action potentials and for localization of Nav1.6 at their axon initial segments. Surprisingly, loss of ankyrin-G also results in coordinated loss of two other ankyrin-binding proteins, β4-spectrin and neurofascin. Neurofascin is a cell adhesion molecule that we had found co-localized with ankyrin-G at initial segments and nodes of Ranvier. Josh Huang (Cold Spring Harbor Laboratory) found that loss of neurofascin in ankyrin-G–null Purkinje neurons is associated with loss of interneuron synapses at their axon initial segments. Moreover, Edward Cooper (University of Pennsylvania) found that KCNQ2/3 channels bind to ankyrin-G and also are no longer clustered at axon initial segments of ankyrin-G–null Purkinje neurons. We have recently reported (in collaboration with Christian Schultz, Justus Liebig University Giessen) that ankyrin-G–null Purkinje neuron initial segments lose their axonal properties and become dendrites. These studies support a central organizing role for ankyrin-G at axon initial segments in clustering multiple membrane-spanning proteins in the plane of the membrane, as well as in promoting transcellular synaptic interactions with interneurons.
Cardiac Voltage-Gated Sodium Channel Nav1.5 Requires Ankyrin-G for Cellular Targeting
We discovered that ankyrin-G co-localizes with cardiac voltage-gated sodium channels (Nav1.5) at two strategic cellular domains in cardiomyocytes: intercalated discs, where the action potential is transmitted to adjacent cells, and in microdomains on the T tubule, close to voltage-gated calcium channels activated by the sodium-based action potential. Moreover, we also have found that E1053K mutation of Nav1.5 within its ankyrin-binding motif eliminates ankyrin-binding activity of Nav1.5 and causes a cardiac arrhythmia, termed Brugada syndrome. We found that the same E1053K mutation also abolishes targeting of Nav1.5 to the cell surface of cardiomyocytes. The finding that Nav1.5 requires ankyrin-binding activity for cellular targeting is important for several reasons. First, it further establishes the principle that mis-sorting of ion channels is functionally equivalent to loss of channel activity. It also suggests that ankyrin-Gbased localization of voltage-gated sodium channels is a conserved mechanism shared by both cardiomyocytes and neurons.
Diverse Ankyrin-G Partners in Early Embryos, Epithelial Cells, Skeletal Muscle, and Photoreceptors
We have recently expanded the repertoire of ankyrin-binding proteins and membrane domains to include E-cadherin at sites of cell-cell contact in early embryos and epithelial tissues, dystroglycan at costameres and neuromuscular junctions of skeletal muscle, and the cyclic nucleotidegated (CNG) channel β subunit in rod outer segments of photoreceptors. In each case, the initial observation was that ankyrin-G co-localized with these membrane proteins within specialized regions on the cell surface. We subsequently found that ankyrin-G is required for cellular localization of its membrane partners and associates with specific residues in their cytoplasmic domains. In skeletal muscle, we found that loss of ankyrin-G or ankyrin-B results in loss of dystrophin and inability of muscles to withstand exercise. We were able to show in photoreceptors that substitution of an unrelated ankyrin-binding sequence in the cytoplasmic domain of the CNG channel is sufficient to direct targeting from the cell body to rod outer segments. We also observed in cultured epithelial cells that ankyrin-G is required for biogenesis of the lateral membrane and, in photoreceptors, that reduction of ankyrin-G results in shorter rod outer segments. These observations, together with the finding that ankyrin-G is the master regulator of axon initial segments, suggest that ankyrin-G has a conserved role in formation as well as functional organization of diverse specialized membrane domains.
A Precise yet Easily Evolved Code for Ankyrin Recognition
Evolution of physiological domains required a mechanism to co-localize multiple unrelated membrane-spanning proteins. For example, the axon initial segment contains voltage-gated sodium channels, KCNQ2/3 channels that modulate sodium channels, and neurofascin (a cell adhesion molecule) that directs formation of inhibitory synapses that also modulate sodium channels. What is the code that directs these and other ankyrin-binding proteins to associate with ankyrin? Currently known ankyrin-binding sites in membrane proteins have independently evolved and do not exhibit obvious consensus in primary sequence. However, many ankyrin-binding sites are either predicted or demonstrated to be extended peptides lacking secondary structure. A possible binding site for unstructured peptides could be the ankyrin groove that runs the 240-Å length of the repeat stack. A groove of this length with variation in surface-exposed residues could potentially accommodate multiple types of partners.
Natively unstructured domains of proteins are widely utilized in protein recognition. One advantage of such a code is that unstructured proteins can multitask and also engage endocytic machinery and other adapter proteins, depending on cellular requirements. Another advantage is that the affinity for ankyrin can vary: the Kd for ankyrin is 10 nM for the anion exchanger, where a stable interaction is required to preserve erythrocyte membrane integrity, but 500 nM for E-cadherin, which can be rapidly internalized in response to signals. This variable affinity allows for flexibility in the tightness of ankyrin binding, depending on the physiological context. Finally, intrinsically unstructured proteins represent the most rapidly evolving part of the genome and have the potential to adjust readily to physiological demands in vertebrate evolution, such as the rapid acquisition of myelination.
We and our collaborators have discovered that a mutation in human ankyrin-B causes a dominantly inherited cardiac arrhythmia and is a risk factor for sudden cardiac death. We also demonstrated that elevated calcium transients combined with sympathetic stress provide an electrical mechanism for arrhythmia. A cellular rationale for elevated calcium is the observed coordinated reduction of three ankyrin-binding proteins classically involved in calcium homeostasis: Na/K ATPase (NKA), Na/Ca exchanger (NCX), and the inositol triphosphate (IP3) receptor. Together, these findings established the principle of a "channelopathy" due to abnormal localization of ion channels/transporters, and provided a novel mechanism for cardiac arrhythmia.
We next analyzed additional ankyrin-B mutations identified by Mark Keating (Harvard Medical School) and Sylvia Priori (University of Pavia) in screens of patients with cardiac arrhythmias. We demonstrated that these mutations resulted in loss of function of ankyrin-B in cardiomyocytes. The additional clinical data allowed a more precise definition of the ankyrin-B mutant phenotype as sick sinus syndrome with bradycardia (OMIM).
Recently we have discovered that ankyrin-B is required for cholinergic augmentation of insulin release and that R1788W mutation of ankyrin-B is a potential risk factor for type 2 diabetes expressed in about 1 percent of Caucasian and Hispanic adult diabetics. We found that ankyrin-B acts through stabilization of IP3 receptors in pancreatic β cells. Moreover, loss of ankyrin-B function results in decreased calcium transients in response to cholinergic stimulation, which is the opposite of the increased calcium transients in ankyrin-B–deficient cardiomyocytes. These findings place ankyrin-B as a risk factor in multiple human diseases associated with aberrant calcium signaling.
We have found that ankyrin-B may play a role in early senescence based on observations in ankyrin-Bdeficient mice, which have a reduced life span and accelerated age-related degenerative changes in multiple tissues. In future work we will address the mechanism to determine if ankyrin-B has specific roles in tissue repair and maintenance. We are excited by the possibility that understanding this mechanism could lead to a personalized strategy to reduce aging-related symptoms for humans with ankyrin-B mutations.
This work has been supported in part by the Muscular Dystrophy Association and the National Institutes of Health.
As of March 19, 2010