Defining the MO's of RGK proteins

Members of the RGK (Rad-Rem-Rem2-Gem/Kir) family of small GTP binding proteins profoundly inhibit high voltage-activated Ca_V1.X and Ca_V2.X Ca²⁺ channels. The ability of RGK proteins to interact with these channels was revealed when Gem was fished out of a yeast-2-hybrid screen which used the channel β subunit as bait.¹ The functional significance of this interaction was immediately clear as Ca²⁺ currents were more-or-less absent in cells coexpressing Gem with either Ca_V1.2, Ca_V1.3 or Ca_V2.1. Since the ablation of Ca²⁺ currents by Gem, and later Rad, Rem and Rem2, required coexpression of a β subunit isoform, two reasonable assumptions arose regarding the nature of RGK protein-mediated inhibition. The first assumption was that RGK proteins blocked channel membrane expression by interfering with the β subunit's well-known ability to promote channel trafficking. The second assumption was that RGK proteins interacted only with the β subunit.

The first indication that first assumption might be a bit naïve became apparent when Chen et al.² reported that HEK 293 coexpressing Ca_V2.2 channels and Rem2 displayed virtually no difference in maximal ω-conotoxin GVIA binding relative to cells expressing just the channel. However, other evidence, such as the reduction of intramembrane gating charge movement in cardiomyocytes by virally-overexpressed Gem,³ still supported the idea that RGK proteins reduce channel membrane expression. These conflicting viewpoints were reconciled by Yang and colleagues⁴ who, using HEK293 cells coexpressing Ca_V1.2, β_2a and Rem, found that Rem can support three distinct modes of inhibition: 1) reduction of channel membrane expression, 2) immobilization of the voltage-sensors, and 3) reduction of channel P_o without impaired voltage-sensor movement (see Figure 1).

Figure 1.

Three in one: inhibition of L-type Ca_V1.2 channels by Rem. Rem utilizes three distinct, but overlapping, mechanisms to inhibit Ca_V1.2 channel heterologously-expressed in HEK293 cells.⁴ Two of these modes, a reduction of channel membrane expression (N; red circle) and inhibition of pore opening (P_o; blue circle), are dependent on an interaction with the channel β subunit.⁶ The third mode, in which the voltage-sensing elements of the channel are immobilized (Q_max; green circle), is facilitated by an interaction between the proximal amino-terminus (encompassing residues 93–153) of the α_1C subunit of Ca_V1.2 and the distal carboxyl-terminus of Rem (residues 265–285).⁷ This latter form of inhibition specifically requires the Ras-like G domain of Rem, although the Rad G domain is likely capable of substituting for the Rem G domain. Gem and Rem2 operate only through β-dependent mechanisms.^6,7

Two relatively recent studies have challenged the second assumption that RGK proteins exert their influence on channel activity solely via interactions with the β subunit. Fan et al. washed away β subunits from inside-out membrane patches to demonstrate that application of purified Gem can directly inhibit Ca_V2.1 channels,⁵ while Yang et al. showed a substantial component of the total inhibition of Ca_V1.2 effected by Rem or Rad, but not by Gem or Rem2, remains intact in HEK293 cells coexpressing a mutantβ_2a subunit that does not bind RGK proteins in vitro.⁶ The latter study also showed that this newly uncovered form of inhibition, was reliant on the proximal amino-terminus of the principal α_1C subunit of the Ca_V1.2 heteromultimer.

In the highlighted study, the Colecraft laboratory set out to identify the cognate elements of Rem that are essential for α_1C subunit-binding-dependent (ABD) inhibition.⁷ To accomplish this goal, they employed 3-cube FRET and co-immunoprecipitations to demonstrate that the α_1C amino-terminus interacts with residues 265–285 in the distal carboxyl-terminus of Rem. ABD inhibition of L-type current was blunted in HEK293 cells expressing Rem constructs lacking this portion of the carboxyl-terminus in manner consistent with the aforementioned binding assays, but expression of the Rem carboxyl-terminus (residues 224–297) alone was insufficient for ABD inhibition. However, ABD inhibition was fully restored with the addition of the G Domain (Ras-like core) of Rem. By contrast, the Rem carboxyl-terminus did not support ABD inhibition when fused to the Gem G Domain, despite the high level of conservation between the G domains of Gem and Rem. The authors integrated their observations in a model where voltage-sensor translocation is partially impeded by the binding of the Rem carboxyl-terminus to the amino-terminus of Ca_V1.2, which in turn is dependent on the specific anchoring of the Rem G domain to an unknown moiety. Interestingly, the linear sequences required for ABD inhibition overlap with those required for β subunit-binding-dependent (BBD) inhibition, whereby the G domain and carboxyl-terminus interact with the β subunit and the plasma membrane, respectively, to generate a steric tension that prevents channel opening without great effect on movement of the voltage sensors.⁸

In addition to defining the structural elements of Rem that are important for ABD inhibition of Ca_V1.2, Puckerin and colleagues showed that this form of inhibition is physiologically relevant in cardiomyocytes as adenoviral expression of the α_1C amino-terminus significantly reduced inhibition of native L-type currents by coexpressed Rem.⁷ Since ABD inhibition has only been demonstrated to be a property of Ca_V1.2 channels so far, one can envision that specific therapeutic strategies based on the Rem-Ca_V1.2 interaction could be developed.^7,8 Still, more investigation of these fundamental mechanisms is necessary since it is now clear that RGK protein-mediated inhibition of Ca_V channels is far, far more complex than we first had thought.

Abbreviations

ABD

α_1C subunit-binding-dependent

BBD

β subunit-binding-dependent

RGK

Rad-Rem-Rem2-Gem/Kir

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.