Desensitization at the Interface - ACS Chemical Biology (ACS

Nov 17, 2006 - Department of Pharmacology, Yale University School of Medicine, SHM B-251, 333 Cedar Street, New Haven, Connecticut 06520. ACS Chem...
0 downloads 0 Views 779KB Size
Point of

VIEW

Desensitization at the Interface James R. Howe*

Department of Pharmacology, Yale University School of Medicine, SHM B-251, 333 Cedar Street, New Haven, Connecticut 06520

N

eurons make specialized contacts with their target cells called synapses, sites where small-molecule neurotransmitters are released in response to electrical activity. Once released, the neurotransmitters bind to the extracellular domains of receptor proteins in the cell membrane of the postsynaptic cell. At most excitatory synapses in the brain, glutamate is the neurotransmitter, and rapid neuronto-neuron communication is mediated by ␣-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA) receptors, a subtype of ionotropic glutamate receptor (iGluR). The binding of glutamate, AMPA, or other agonists to these receptors initiates a series of conformational changes (“gating”) that allow cations to flow through the channels down their electrochemical gradients. The resultant transmembrane (TM) current initiates a regenerative electrical signal called an action potential, but the channels quickly become unresponsive during sustained exposure to glutamate. A recent paper by Armstrong et al. (1) provides the first highresolution views of receptor conformations adopted during desensitization. Both structural and functional data indicate that iGluRs are tetrameric assemblies, organized as dimers of dimers, with each subunit containing a binding site for glutamate (2). Individual subunits are composed of four domains, with the extracellular ligand binding domain (LBD) and the pore domain (composed of membrane helices) being sufficient for channel gating. X-ray crystallographic structures of the isolated LBD show that it is a clamshell-like structure and that

www.acschemicalbiology.org

glutamate and other ligands bind within a deep cleft between the two halves of the clamshell (LBD domains 1 and 2). In agonist-bound structures, domain 2 is displaced ⬃20° relative to its position in apo structures, closing the clamshell and trapping agonists (3, 4). Domain 2 is connected to the extracellular ends of the first two TM helices, and the movement of domain 2 is thought to result in rearrangements of the helices that open the channel. The available evidence, although circumstantial, supports the conclusion that closing of the LBD is the initial large-scale conformational change that triggers subsequent gating rearrangements at the level of the pore (3, 5). Most receptors desensitize upon continued exposure to neurotransmitter, and desensitization is especially prominent for AMPA receptors. The current generated when channels formed from the AMPA receptor subunit GluR2 were rapidly exposed to a high concentration of glutamate is illustrated (Figure 1, panel a). The diffusion-limited binding of glutamate results in the near-simultaneous activation of hundreds of receptors in the patch, generating a rapidly developing inward current, which then decays to 1–2% of its peak value in 15–20 ms. The decay of the current results from entry of the channels into desensitized states, and the accumulation of channels in these unresponsive states limits the size of the current evoked by a second application of glutamate (which progressively increases as channels recover from desensitization; Figure 1, panel a). Desensitization predominates at low micro-

A B S T R A C T Normal brain function requires the faithful transmission and integration of information on a timescale of milliseconds, and this rapid signaling is mediated by cell membrane receptors that are ligand-gated ion channels. Fast excitatory transmission occurs when synaptically released glutamate opens channels in neighboring neurons, but these channels desensitize rapidly during sustained high-frequency firing. Recent structural data have begun to provide important insights into the molecular mechanisms that underlie channel activation and desensitization.

*Corresponding author, [email protected].

Published online November 17, 2006 10.1021/cb6004382 CCC: $33.50 © 2006 by American Chemical Society

VOL.1 NO.10 • ACS CHEMICAL BIOLOGY

623

Figure 1. Probing the GluR2 channel. a) Inward currents evoked by the rapid application (bars) of 10 mM glutamate in a patch of membrane that contains hundreds of recombinant GluR2 channels. The first application of glutamate causes a rapidly activating current that quickly fades as the channels desensitize. The responses to a second application of glutamate (made at three different intervals) are reduced. b) Ribbon diagram of the wild-type GluR2-S1S2 structure. The LBD crystallizes as a dimer in which contacts are made between domain 1 residues in each monomer across the twofold axis of symmetry. This dimer interface is in blue, and the yellow balls show positions at which cysteines were introduced by sitedirected mutagenesis. The red spheres at the bottom of domain 2 indicate the position where the LBD would be connected to the first and second TM helices in the full-length receptor. c) Reagents used to cross-link nonnative cysteines across the dimer interface. Adapted with permission from Elsevier, copyright 2006 (1) and the Biophysical Society, copyright 2006 (11).

molar concentrations of glutamate and appears to have evolved as a mechanism to protect neurons from excitotoxic damage when ambient levels of glutamate are elevated (e.g., during cerebral ischemia). This protection comes, however, with a cost. Although at many synapses glutamate clearance mechanisms remove glutamate before most channels desensitize during a single synaptic event, high-frequency trains of impulses can lead to the accumulation of 624

VOL.1 NO.10 • 623–626 • 2006

channels in desensitized states and ultimately result in the failure of transmission. Because information processing in the brain is frequency-encoded, the rate of desensitization (and recovery from it) significantly impact synaptic signaling. Previous work that combined crystallographic and biochemical studies of the isolated GluR2 LBD with mutagenesis and electrophysiology on full-length receptors led to the conclusion that desensitization occurred when monomer–monomer contacts along the two-fold axis of symmetry in each receptor dimer (the “dimer interface”) slipped, rendering closure of the LBD less effective at promoting conformations of the membrane helices that allow ion flux (6). Building on this previous work, Armstrong and colleagues (1) introduced cysteines into positions along the dimer interface that the prior GluR2 crystal structures predicted would be largely inaccessible to bulk solvent in resting and activated channel states (Figure 1, panel b). Using voltage-clamp recording of currents through recombinant channels expressed in oocytes, they then measured the rate at which these cysteines were modified by sodium (2-sulfonatoethyl)methane thiosulfonate (MTSES) under conditions that favored resting, active, or desensitized channel states. (To minimize potentially confounding MTSES modification of native cysteines, the authors employed a recombinant GluR2 construct from which the amino terminal domain had been removed and in which most native cysteines were replaced by conservative mutation.) They found that MTSES reaction rates were ⬃2 orders of magnitude faster for desensitized channels, the first direct evidence that desensitization is indeed accompanied by rearrangement of the dimer interface at the level of the LBD. To begin to quantify the extent of the rearrangement, the authors used bifunctional thiol-directed reagents that cross-link nearby cysteine residues (Figure 1, panel c). Cysteines were introduced at various depths along the interface, and four linkers that HOWE

ranged in length from 8.6 to 25.1 Å were used to cross-link cysteines located on opposing sides of the dimer interface. Each reagent would be predicted to cross-link cysteine pairs in resting and activated states, whereas cross-linkers whose fully extended length was less than the separation of the residues following rearrangement of the interface would be expected to reduce desensitization. Results with these molecular rulers indicated that during desensitization the two monomers separate at the top and bottom of the V-shaped interface by 16.2 and 12.4 Å, respectively. While characterizing the various cysteine mutants, the authors identified one that gave very small steady-state currents in normal conditions but much larger currents after the channels were exposed to the reducing agent dithiothreitol (DTT). Western blotting demonstrated that the mutant channels formed spontaneous inter-subunit disulfide bonds, although previous crystal structures suggested that the non-native cysteines were much too far apart to do so in conformations adopted by non-desensitized channels. This led Armstrong and coworkers to hypothesize that the disulfide bond was formed when the LBD adopted a conformation not observed in previous crystal structures, perhaps one preferentially associated with desensitization. Studies on the DTT sensitivity of additional cysteine mutants highlighted the importance of a region on the surface of domain 2 (below the interface) that contains a loop pointing across the two-fold axis to residues on the adjacent monomer. Binding studies on the isolated LBD, as well as functional measurements made with a domain 2 histidine mutant (G725) in the absence and presence of Zn2⫹, supported the view that tethering the relevant domain 2 residues across the two-fold axis promoted desensitization (rather than stabilizing a resting closed state). Armed with this additional information, the authors introduced the S729C mutation into the GluR2-S1S2 LBD and solved the www.acschemicalbiology.org

Point of

VIEW Figure 2. Structural changes in the desensitized GluR2 LBD structure. a) Superposition of GluR2 LBD structures putatively corresponding to conformations adopted by active (L483Y, gray) and desensitized (S729C, blue and green) channels. The view is roughly down the twofold axis with the linkers that replace residues connected to the channel pore shown as pink or red spheres (bottom). The arrows indicate displacements of the linkers and residues along the dimer interface that characterize the transition from active to desensitized conformations. Domain 1 residues along the dimer interface separate, whereas separation of the linkers decreases. b) Close-up view of the dimer interface in the S729C (desensitized) structure reveals new interactions formed between amino acids comprising the loop between helices F and G (R661, K663, and A665) and residues in helix K of the adjacent monomer. c) Cartoon showing the LBD and TM helices of an AMPA receptor dimer in resting, bound, open, and desensitized states. Glutamate binds to the open LBD. Movement of domain 2 in each monomer closes the LBD, creating what is likely an unstable transition state in which both the LBD and channel pore are closed (not shown). Movements of the TM helices partially relieve this instability and open the channel (on average 2–3 times). Eventually, however, separation of the dimer interface dissipates the work initially done by closure of the LBD, and the channel adopts a stable state in which glutamate binding and activation gating are uncoupled. Adapted with permission from Elsevier, copyright 2006 (1).

crystal structure at 2.3 Å resolution. Like previous structures of the isolated GluR2 binding core, the S729C mutant crystallized as a dimer (Figure 2). However, the dimer interface is separated substantially, as suggested by the oocyte experiments on functional receptor mutants. The distances characterizing this rearrangement in the crystal www.acschemicalbiology.org

structure agree well with those estimated in the cross-linking experiments; this supports the conclusion that the structure is similar to the conformation of the LBD adopted by desensitized channels. Unlike previous GluR2-S1S2 structures, the two LBDs are asymmetric, a possible explanation for electrophysiological evi-

dence that during recovery from desensitization the first two bound glutamates (one in each dimer) dissociate rapidly, whereas the last two glutamates dissociate much more slowly (7). The structure also suggests that the asymmetry of the amino terminal domains observed previously in cryo-electron microscopy images (8) results from asymmetric rearrangement of the LBD dimer interface. The X-ray crystallographic data, as well as the new functional results, speak directly to stable conformations adopted under equilibrium conditions but do not provide information on receptor kinetics. In this sense, the available crystal structures are somewhat like a set of vacation snapshots, which reveal where you were but not how you got there (or how long it took). However, although the crystallographic data do not discriminate between alternative kinetic mechanisms, they do provide a clear roadmap for additional structure–function work on full-length receptors. For example, kinetic studies done before structural data were obtained suggested that the rate at which channels recover from desensitization was primarily determined by the rate at which agonists dissociated from desensitized channels (9, 10), a proposal that nicely accounts for the well-known correlation between agonist affinity and the rate of recovery (11). In contrast, more recent studies concluded that the rate-determining steps during recovery are sequential reassembly of the two dimer interfaces (12). Interestingly, the new LBD structure shows that separation of the dimer interface results in additional interactions across the binding cleft that would be predicted to stabilize the closed-cleft conformation, whereas the new inter-subunit interactions between domain 2 residues are relatively minimal. The newly identified interactions in these two regions are obvious targets for kinetic studies that might further illuminate the sequence of events during recovery. Ultimately, structural data on receptor constructs that include both the LBD and the VOL.1 NO.10 • 623–626 • 2006

625

channel pore are required if the mechanisms of activation and desensitization are to be understood, likely in combination with high-resolution electrophysiology and spectroscopic measurements to provide realtime correlations between ion flux and protein movements (13). Nevertheless, the recent work of Armstrong and colleagues represents another significant advance in unraveling how a major class of synaptic proteins works.

13. Blunck, R., Cordero-Morales, J. F., Cuello, L. G., Perozo, E., and Bezanilla, F. (2006) Detection of the opening of the bundle crossing in KcsA with fluorescence lifetime spectroscopy reveals the existence of two gates for ion conduction, J. Gen. Physiol. 128, 569–581.

REFERENCES 1. Armstrong, N., Jasti, J., Beich-Frandsen, M., and Gouaux, E. (2006) Measurement of conformational changes accompanying desensitization in an ionotropic glutamate receptor, Cell 127, 85–97. 2. Mayer, M. L. (2006) Glutamate receptors at atomic resolution, Nature 440, 456–462. 3. Armstrong, N., and Gouaux, E. (2000) Mechanisms for activation and antagonism of an AMPA-sensitive glutamate receptor: crystal structures of the GluR2 ligand binding core, Neuron 28, 165–181. 4. Jin, R., Horning, M., Mayer, M. L., and Gouaux, E. (2002) Mechanism of activation and selectivity in a ligand-gated ion channel: structural and functional studies of GluR2 and quisqualate, Biochemistry 41, 15635–15643. 5. Jin, R., Banke, T. G., Mayer, M. L., Traynelis, S. F., and Gouaux, E. (2003) Structural basis for partial agonist action at ionotropic glutamate receptors, Nat. Neurosci. 6, 803–810. 6. Sun, Y., Olson, R., Horning, M., Armstrong, N., Mayer, M., and Gouaux, E. (2002) Mechanism of glutamate receptor desensitization, Nature 417, 245–253. 7. Robert, A., and Howe, J. R. (2003) How AMPA receptor desensitization depends on receptor occupancy, J. Neurosci. 23, 847–858. 8. Nakagawa, T., Cheng, Y., Ramm, E., Sheng, M., and Walz, T. (2005) Structure and different conformational states of native AMPA receptor complexes, Nature 433, 545–549. 9. Partin, K. M., Fleck, M. W., and Mayer, M. L. (1996) AMPA receptor flip/flop mutants affecting deactivation, desensitization, and modulation by cyclothiazide, aniracetam, and thiocyanate, J. Neurosci. 16, 6634–6647. 10. Raman, I. M., and Trussell, L. O. (1995) The mechanism of alpha-amino-3-hydroxy-5-methyl-4isoxazolepropionate receptor desensitization after removal of glutamate, Biophys. J. 68, 137–146. 11. Zhang, W., Robert, A., Vogensen, S. B., and Howe, J. R. (2006) The relationship between agonist potency and AMPA receptor kinetics, Biophys. J. 91, 1336–1346. 12. Robert, A., Armstrong, N., Gouaux, J. E., and Howe, J. R. (2005) AMPA receptor binding cleft mutations that alter affinity, efficacy, and recovery from desensitization, J. Neurosci. 25, 3752–3762.

626

VOL.1 NO.10 • 623–626 • 2006

HOWE

www.acschemicalbiology.org