Stoichiometry of the cardiac I_Ks complex

Nothing is easy for ion channels that assemble with membrane-embedded regulatory subunits. Something as trivial yet as fundamental as counting the number of subunits in an ion channel complex can be challenging and marred with controversy. The most infamous stoichiometric debate in the ion channel field has been over the potassium (K⁺) channel complex that generates the repolarizing cardiac I_Kslow (I_Ks) current (1–5). This union between a tetrameric voltage-gated K⁺ channel (KvLQT1, KCNQ1, Kv7.1) and type I transmembrane regulatory subunit (minK, KCNE1) is unquestioned; however, the last stoichiometric salvo in the debate proposed a haphazard coassembly of cardiac I_Ks complexes, containing one to four regulatory subunits (4). Determining the exact number of regulatory subunits in the cardiac I_Ks complex is vital because mutations in either the ion conducting or regulatory subunit that reduce potassium flow give rise to Long QT and Jervell and Lange-Nielsen syndromes—two diseases where individuals are prone to adrenaline-induced, life-threatening cardiac arrhythmias (6, 7). Attempts to find small molecule openers that enhance cardiac repolarization by targeting the K⁺ channel (8, 9) have been ironically thwarted by the presence of the regulatory subunit whose stoichiometry in the complex is continually in question.

Let's Go Photobleaching

In PNAS, Plant et al. (10) warm up their laser beams and take aim at this cardiac potassium channel complex using fluorescent subunit photobleaching. This single molecule spectroscopy approach relies on the mathematically predictable photobleaching of a fluorophore (11), usually monomeric GFP (or a color variant), which is appended to each subunit of interest in an ion channel complex. Thus, the technique can be used for any membrane-embedded complex as long as enough fluorescent spots are sampled to make an accurate stoichiometric determination. To zap as many spots as possible, Plant et al. photobleach the cardiac I_Ks complex expressed in mammalian cells instead of the traditional Xenopus oocytes. This switch also avoided the endogenous xKCNQ and xKCNE subunits that are responsible for both the discovery of the KCNE1 subunit (12) and much of the confusion that permeates the cardiac I_Ks field.

Because the human cardiac I_Ks complex (like most protein complexes) is not inherently fluorescent, fluorescent protein tags needed to be added to each subunit in the complex. Normally, slapping a tag (or two) onto your favorite protein of interest is trivial; however, the KCNE1 regulatory subunit is

The 4:2 KCNQ1:KCNE1 subunit stoichiometry of the cardiac IKs complex has several implications for our heart's tremolo.

tiny (129 amino acids) and behaves erratically in cells when tags, proteins or otherwise, are added willy-nilly to its water-soluble domains. Therefore, Plant et al. carefully characterize their monomeric fluorescent protein-tagged KCNQ1 and KCNE1 subunits using electrophysiology, ensuring that these fluorescent constructs had WT biophysical properties where paired with untagged and fluorescent protein-tagged subunits.

Given the longstanding debate, Plant et al. use three spectroscopic counting approaches. The first approach was standard sequential photobleaching of a fluorescent protein-tagged KCNQ1 or KCNE1 subunit partnered with an untagged counterpart. As expected for a voltage-gated K⁺ channel (13), four KCNQ1 subunits were present in the complex. For KCNE1, only two photobleaching steps were observed. In fact, no particle showed more than two bleaching steps. Because asynchronous photobleaching is prone to particles moving in and out of the field of view and experimental cherry picking, Plant et al. performed simultaneous, two-color subunit counting of fluorescently tagged KCNQ1 and KCNE1 subunits. Again, a 4:2 stoichiometry was observed, and no protein complexes possessed more than two KCNE1 subunits. Their last approach exploited an untagged KCNQ1 channel engineered to bind the peptide toxin, agitoxin-2, in a 1:1 stoichiometry with high affinity when paired with fluorescently labeled KCNE1. Disulfide bond-stabilized toxins, such as agitoxin-2, were first used in the 20th century to show that there were four subunits in a voltage-gated K⁺ channel (13); thus, their utility and credibility in the field is unassailable. Simultaneous photobleaching of this cardiac I_Ks complex in the presence of bound, rhodamine-labeled agitoxin-2 showed one bleaching step for the toxin, and again, no more than two bleaching steps for the KCNE1 subunits.

Counting Subunits Is Hard

So there are two, and only two, KCNE1 subunits in the cardiac I_Ks complex? Is that possible—it breaks the revered fourfold symmetry of the KCNQ1 channel? It can’t be two; it shouldn’t be two. What about those papers that showed there were four? I know what most of you are thinking, “This debate has lasted longer than any of the names given to describe either subunit in the cardiac I_Ks complex! It can’t be resolved that easily.” Well, because you asked…

One reason for the debate’s longevity has been the multiple suppositions that the stoichiometry of the cardiac I_Ks complex can be variable and depends on the ratio of KCNQ1 and KCNE1 subunits expressed in a cell (3, 4). To rebut this notion, Plant et al. performed the simultaneous photobleaching experiments with cells transfected with limiting amounts of KCNQ1 or KCNE1 cDNA. In cells with limiting KCNQ1, the stoichiometry was still 4:2 KCNQ1:KCNE1. In contrast, two stoichiometries were observed with cells with limiting KCNE1: a population of KCNQ1 channels with no KCNE1 subunits and a population of KCNQ1 channels with two KCNE1 subunits. Because the activation kinetics of homotetrameric KCNQ1 channels are much faster than the KCNQ1-KCNE1 complex, a quick electrophysiological examination of the currents confirmed a mixture of unpartnered and KCNE1-partnered KCNQ1 channels at the cell surface. Thus, KCNQ1 channels do have two assembly options: (i) KCNQ1 subunits can coassemble with two KCNE1 subunits to form the cardiac I_Ks complex, or (ii) they can go it alone in the cell and function as a traditional homotetrameric voltage-gated K⁺ channel (Kv-type).

Arrhythmic Implications

The 4:2 KCNQ1:KCNE1 subunit stoichiometry of the cardiac I_Ks complex has several implications for our heart’s tremolo. First and foremost, it means the human cardiac I_Ks current emanates from a K⁺ channel complex that is not fourfold symmetric. Thus, previously described small molecule activators of homotetrameric KCNQ1 channels that are less effective on the cardiac I_Ks complex are being desensitized by two and only two KCNE1 subunits (8, 9). Conversely, any pharmacological activity observed in cardiomyocytes (9) indicates that either unpartnered KCNQ1 channels are present and can contribute to cardiac repolarization or the efficacy arises from cross-reactivity with another cardiac ion channel complex.

The apparent loss of fourfold symmetry also impacts how our beloved cardiac I_Ks complex assembles, traffics, and functions in the heart. Assembly of the cardiac I_Ks complex occurs early in the secretory pathway: in the ER and/or cis-Golgi (14). If the complex forms early in the ER, one KCNE1 subunit could form a heterotrimer with two KCNQ1 subunits and then dimerize to form the 4:2 complex via a dimerization mechanism that has been previously observed for archetypical Kv-type channels (15). Although the experiments by Plant et al. are performed in Chinese hamster ovary (CHO) cells, the cell surface densities of the KCNQ1 and KCNE1 subunits were significantly greater when coexpressed in the same cell, indicating that coassembly mediates trafficking to the plasma membrane (14). Photobleaching fluorescently tagged KCNQ1 and KCNE1 subunits expressed in cardiomyocytes is needed to determine whether unpartnered KCNQ1 and KCNE1 subunits can traffic independently, as a recent study suggests (16). In addition, these single molecule approaches may also shed some light on the K⁺ channel–KCNE partnerships and trafficking patterns in the heart, where all five KCNE regulatory subunits are endogenously expressed (17) with several different cardiac K⁺ channels (including KCNQ1) that can coassemble with more than one KCNE partner (18). Last, how do two tiny KCNE1 subunits regulate KCNQ1 function? After all, it is the strikingly slow opening and closing kinetics of the cardiac I_Ks complex that keeps our hearts beating about once per second. Several recent studies suggest that KCNE1 modulates the movements of the four KCNQ1 voltage sensors (19, 20); however, we know now that the KCNE1 subunits are outnumbered two to one, requiring some reinterpretation of previous data and design of new experiments to look for potential asymmetric modulation of KCNQ1 channels. Plant et al. strongly argue that all KCNE peptides will modulate KCNQ1 and other K⁺ channels with a similar 4:2 stoichiometry. Will this stimulate new KCNE controversies? You can count on it.