CrossTalk proposal: TARPs modulate AMPA receptor gating transitions
Year: 2013
Type: article
Abstract: Synaptic transmission is the foundation of information processing in the CNS and glutamate mediates the vast majority of fast excitatory synaptic transmission in the human brain and spinal cord. Although a number of mechanisms influence synaptic signalling, one key determinant is the properties of the postsynaptic ionotropic receptors (AMPA, NMDA and kainate subtypes) to which glutamate binds. In addition to the core GluA subunits, AMPA receptors (AMPARs) interact directly with a number of other proteins, and reconstitution of several properties of native receptors requires association with transmembrane AMPA receptor regulatory proteins (TARPs). Not only do TARPs play an important role in AMPAR trafficking, but a variety of studies have demonstrated that they are auxiliary subunits of neuronal AMPARs and that TARPs modulate several features of AMPAR gating (Coombs & Cull-Candy, 2009; Kato et al. 2010b; Jackson & Nicoll, 2011; Yan & Tomita, 2012). Thus understanding the impact of TARPs on receptor properties is essential to understanding how those properties influence synaptic transmission. At the level of population currents, co-expression of TARPs has been shown to slow the decay of AMPA receptor responses to both brief and long pulses of glutamate (deactivation and desensitization), as well as speed recovery from desensitization (Priel et al. 2005; Tomita et al. 2005; Turetsky et al. 2005; Cho et al. 2007; Kott et al. 2007; Milstein et al. 2007). Based on TARP effects on ensemble currents, as well as single-channel data showing that TARPs do not alter mean open time but do increase burst durations (Tomita et al. 2005), we argued that the effects of TARPs can best be accounted for by an effect of TARPs to increase the rate constant for channel opening, which also accounts for increases in peak Popen (Cho et al. 2007). Figure 1 shows a simple gating scheme for AMPA receptors where glutamate binding first results in closing of the clamshell-like ligand binding domain, resulting in a brief transition state in which the channel pore remains closed (Armstrong & Gouaux, 2000; Jin et al. 2003; Zhang et al. 2008). Activation and desensitization proceed in parallel (Raman & Trussell, 1995; Partin et al. 1996; Robert & Howe, 2003), and the instability of this closed-clamshell/closed-pore state may be relieved either by the clamshell opening, the pore opening, or the dimer interface separating and the receptor desensitizing (Sun et al. 2002). As described in the figure legend, a modest increase in β accounts for the quantitatively similar slowing of both deactivation and desensitization produced by TARPs (Cho et al. 2007), as well as the effect of TARPs to increase peak Popen. In contrast, reducing δ will slow desensitization and increase peak Popen but not slow deactivation substantially, whereas reducing CO will slow deactivation, but will have little effect (or even speed) desensitization and will not increase peak Popen substantially. Both the proposed increase in β, and an increase in γ that accounts for the faster recovery from desensitization seen with TARPs (Priel et al. 2005), shift the equilibrium occupancy toward the open state. Together these TARP-mediated changes increase steady-state currents. Simple AMPA receptor gating mechanism The ligand-binding domain, linker regions, and transmembrane domains of one receptor dimer of the tetrameric receptor are shown in schematic form. Glutamate (black circle) binds to the ligand binding domain (LBD) of one monomer in the receptor dimer. The LBD closes (violet, at a rate determined by CC), trapping glutamate in the binding pocket and resulting in a brief transition state in which one LBD and the channel pore are closed. The receptor may transition out of this state in three ways: the LBD may open (at a rate determined by CO), the channel pore may open, or the dimer interface may separate resulting in receptor desensitization. When a population of receptors is exposed to a rapid application of a saturating concentration of glutamate, the decay of the ensemble current is determined by the three sets of rate constants governing the transitions in and out of state 2. Rate constants that give population responses similar to those seen experimentally are: k1= 1 × 107 m−1s−1, k−1= 10,000 s−1, CC = 50,000 s−1, CO = 3000 s−1, β= 5000 s−1, α = 3000 s−1, δ = 1500 s−1, γ= 10 s−1. By increasing β, TARPs slow the decay of responses to both short and long pulses of glutamate (deactivation and desensitization) because TARP modulation increases the average number of receptor openings before the LBD opens and releases glutamate or before the dimer interface separates and the receptor enters long-lived desensitized states. Because activation and desensitization occur in parallel, some receptors desensitize without ever opening. The effect of TARPs to increase β reduces the proportion of such receptors and therefore increases peak Popen. A number of groups have reported that TARPs also produce leftward shifts in glutamate EC50 values (Priel et al. 2005; Tomita et al. 2005; Turetsky et al. 2005; Kott et al. 2007; Morimoto-Tomita et al. 2009), and recently it was suggested that TARPs decrease the rate constant for ligand dissociation (Maclean & Bowie, 2011). I have always rejected the idea that TARPs have an effect on the true affinity of glutamate binding for two reasons. First, given the generally accepted idea that activation and desensitization occur in parallel from a common set of closed states, this would not be expected to slow desensitization. Second, it is well known that EC50 values measured under steady-state conditions reflect both binding and gating equilibria (Colquhoun, 1998), and shifting occupancy toward the open state by increasing β is predicted to left shift agonist concentration–response curves, which I assumed accounted for the observed shifts in EC50 values. Interestingly, TARPs have little effect on the concentration dependence of peak currents evoked by rapid agonist applications. The EC50 values we measured for peak currents, either in GluA/TARP co-expression experiments or with GluA-TARP tandem constructs, differed from those for GluA homomers by only 20–25% (Morimoto-Tomita et al. 2009). These latter findings appear to argue against an effect of TARPs to slow glutamate dissociation by stabilizing the closed binding domain (Zhang et al. 2008). There are at least two important caveats to the above argument that the effect of TARPs is primarily on activation gating. First, although it has the merit of parsimony, there is no a priori reason why TARPs cannot alter multiple rate constants. Indeed at least both β and γ appear to be changed. Second, our previous conclusions, which attempt to ascribe changes in observed quantities (like the rate at which ensemble currents decay) to changes in individual rate constants, are model dependent. The mechanism in Fig. 1 incorporates several key features of AMPA receptor gating that are generally agreed upon, but it is clearly too simple to account for other features known to be true, for example that the receptors are tetramers with four glutamate binding sites and that they exhibit multiple open states and conductance levels (Rosenmund et al. 1998; Smith & Howe, 2000; Sobolevsky et al. 2009). Unfortunately, mechanisms that incorporate such features have a very large number of states and potentially very many free parameters. Importantly, however, the appealing simplicity of our conclusions has yet to be verified by direct experimental evidence. I would emphasize too that we have studied the so-called type I TARPs (γ-2, γ-3, γ-4, γ-8), and whether the effects of other TARPs on receptor gating can be accounted for by similar mechanisms remains to be seen (Kato et al. 2010a; Jackson & Nicoll, 2011; Bats et al. 2012; Yan & Tomita, 2012). Our proposed mechanism also does not account for the effects of TARPs on ion permeation (Turetsky et al. 2005; Cho et al. 2007; Soto et al. 2007; Jackson et al. 2011; Shelley et al. 2012) or explain how these effects relate, if at all, to alterations in receptor gating behaviour. Readers are invited to give their views on this and the accompanying CrossTalk articles in this issue by submitting a brief comment. Comments may be posted up to 6 weeks after publication of the article, at which point the discussion will close and authors will be invited to submit a ‘final word’. To submit a comment, go to http://jp.physoc.org/letters/submit/jphysiol;591/7/1581 This work was supported by USPHS grant R56-NS057725. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
Source: The Journal of Physiology
Author: James R. Howe
Institution: Yale University
Language: English
Subfield: Cellular and Molecular Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W2165460659