Human Human Human Human Human Human

(Q15822) (P32297) (P43681) (P30532) (Q15825) (P36544)

ACH9_HUMAN (P43144) ACH10_HUMAN (Q9GZZ6) ACHN_HUMAN (P17787) ACHO_HUMAN (Q05901) ACHP HUMAN (P30926)

*Isoform 1; a splice variant with an additional 25 amino acids is known (isoform 2): this does not form functional channels)

Embryonic

Adult

The human a7 gene is partially duplicated in the same chromosomal region

Figure 11.1. Diagram of topology of a single receptor subunit. Each subunit is thought to cross the cell membrane four times.

used extensively in investigations of receptor structure and function by electron microscopy (see below).

The muscle-type ACh receptor is a glycoprotein complex (-290kDa), which consists of five subunits arranged around a central membrane-spanning pore. Nicotinic subunits are similar in amino acid sequence and have the same topology (Fig. 11.1): each subunit consists of a large extracellular amino-terminal domain, four predicted membrane-spanning segments (M1-M4), and a long cytoplasmic loop between M3 and M4.

These characteristics are shared with sub-units that form other ion channels/receptors and thus define a receptor superfamily, usually referred to as the nicotinic family. All members in this superfamily function as either cation- or anion-selective channels, thereby mediating fast excitatory or inhibitory synaptic transmission. In mammalian cells, the cation-selective members include nicotinic and 5HT3 receptors, whereas the anion-selective members include GABAa, GABAc, and glycine receptors. Anion-selec-tive channels in this family are also found in invertebrates: these channels are gated by glutamate, 5-HT, histidine, and acetylcholine (1).

The muscle-type receptor has the composition (al)2, ft y, and 8 in embryonic (ordener-vated) muscle, but in the adult the ysubunit is replaced by an e subunit. The adult receptor is found, at high density, only in the end-plate region of the muscle fiber, but before innervation embryonic receptors are distributed over the whole muscle fiber. The electric organ contains only the embryonic y-form of the receptor. All the subunits share a high degree cf homology (typically 31 —41 % pairwise identity to the a subunits, depending on the species).

The properties of and interactions between individual subunits have been explored extensively by a range of biochemical, molecular genetic, and electrophysiological techniques [for recent reviews, see Karlin (2); Corringer et al. (3)]. Their order around the pore is most likely to be a, y, a, ft and 5 going in the clockwise sense and viewed from the direction of the synaptic cleft. Opening of the channel occurs upon binding of ACh to both a subunits (a, and a,) at sites that are at, or close to, the interfaces made with neighboring y and 5 sub-units (4-6). These sites are shaped by three separate regions of the polypeptide chain (3) and include the so-called C-loop (see below).

2.2.1 Molecular Architecture. The tubular crystals from the Torpedo ray form the basis of almost all quantitative three-dimensional studies of the whole receptor [e.g., see Kistler and Stroud (7) ;Miyazawaet al. (8)1. Tubesare built from tightly packed ribbons of receptor dimers and intervening lipid molecules (9). They grow naturally from the isolated postsynaptic membranes, retaining a curvature similar to that at the crests of the junctional folds. Apparently, there is a close structural correspondence between the tubes, which are simply elongated protein-lipid vesicles, and the receptor-rich membrane as it exists in vivo.

Ice-embedded tubes, imaged with the electron microscope, can be made to retain their circular cross section and be analyzed as helical particles (10). At low resolution, using this approach, the receptor appears as an approximately 70 x 160-A (diameter x length) cylinder composed of five similar rod-shape sub-units arranged around the central axis and aligned approximately normal to the membrane plane. The ion-conducting pathway, delineated by the symmetry axis, appears as a narrow (unresolved) pore across the mem-

Figure 11.2. Architecture of whole receptor, emphasizing the external surface and openings to the ion-conducting pathway on the outer (extracellular) and inner (cytoplasmic) sides of the membrane. The positions of the two a subunits, the binding pockets (asterisks),gate of the closed channel (upper arrow), and the constrictingpart of the open channel (lower arrow) are indicated. See color insert.

Figure 11.2. Architecture of whole receptor, emphasizing the external surface and openings to the ion-conducting pathway on the outer (extracellular) and inner (cytoplasmic) sides of the membrane. The positions of the two a subunits, the binding pockets (asterisks),gate of the closed channel (upper arrow), and the constrictingpart of the open channel (lower arrow) are indicated. See color insert.

brane, bounded by two large (-20-A diameter) vestibules. Further development of this approach has led to resolutions of 9 A (ll)and, more recently, 4.6 A (8), being achieved.

Figure 11.2 shows the appearance of the whole receptor at 4.6-A resolution. In the extracellular portion, the subunits form a pentagonal wall around the central axis and make the cylindrical outer vestibule of the channel. The outer vestibule is about 20 x 65 A (width length). About halfway up this portion are the ACh-binding regions in the two a subunits (asterisks). In the cytoplasmic portion, the subunits form an inverted pentagonal cone, which comes together on the central axis at the base of the receptor, so shaping a spherical inner vestibule of the channel. This is about 20

A in diameter. The only aqueous links between the inner vestibule and the cell interior are the narrow (<8-9 A wide) ' "windows" between the subunits lying directly under the membrane surface. The gate of the channel, made by the pore-lining segments, M2, is near the middle of the membrane (upper arrow), and the constriction zone (the narrowest part of the open channel) is at the cytoplasmic membrane surface (lower arrow).

2.2,2 The Vestibules. One likely physiological role of the vestibules is to serve as prese-lectivity filters for ions, making use of charged groups at their mouths and on their inner walls, to concentrate the ions they select for (cations), while screening out the ions they discriminate against (anions). In this way, the ionic environment would be modified close to the narrow membrane-spanning pore, increasing the efficiency of transport of the per-meant ions and enhancing the selectivity arising from their direct interaction with residues and/or backbone groups lining the constriction zone. A more direct means of increasing the cation conductance may be achieved by rings of negative charge located at the mouth of the pore. These rings (at positions —4\ -1\ and 20' of M2, using the numbering system for M2 residues defined in Fig. 11.3) are significant in that they have been shown to influence channel conductance (12).

Consistent with a screening role, the cylindrical shape and about 10-A radius of the outer (extracellular) vestibule provide a route that is narrow enough for charged groups on the inner wall to influence ions at the center, but not too narrow to restrict their diffusion. The design of this portion of the receptor might therefore have some parallels with the fast-acting enzyme acetylcholinesterase, where the whole protein surface plays a role in producing an electrostatic field that guides the positively charged ACh substrate to the active site (13). The inner (cytoplasmic)vestibule is architecturally distinct from the outer vestibule, yet presumably plays a similar functional role in concentrating the cations, given that electrophysiological experiments on the muscle-type receptor have shown that there is no marked preference for cations to go in one direction across the membrane (i.e., rectifica-

Figure 11.3. Helical net plot of the amino acid sequence around the membrane-spanning segment 1VE (Torpedo a subunit). The leucine residue near the middle of the membrane (yellow) is the conserved leucine L251 (at the 9' position), which may be involved in forming the gate of the channel. The dots denote other residues that have been shown to affect the binding affinity of an open channel blocker (17,18) and ion flow through the open pore (19). The numbers shown refer to the numbering scheme for M2 residues used in the text. See color insert.

Figure 11.3. Helical net plot of the amino acid sequence around the membrane-spanning segment 1VE (Torpedo a subunit). The leucine residue near the middle of the membrane (yellow) is the conserved leucine L251 (at the 9' position), which may be involved in forming the gate of the channel. The dots denote other residues that have been shown to affect the binding affinity of an open channel blocker (17,18) and ion flow through the open pore (19). The numbers shown refer to the numbering scheme for M2 residues used in the text. See color insert.

tion). Negatively charged groups framing the windows would have a strong local effect, given that the windows are not significantly wider than the diameter of an ion, including its first hydration shell.

The large proportion of mass (-70%) that is not within the membrane, and shapes these vestibules is also needed for other purposes, such as making the (complex) ACh-binding pockets, and providing sites of attachment for regulatory molecules and other proteins (such as rapsyn) that are concentrated at the synapse.

and therefore must correspond to M2, the stretch of sequence shown by chemical labeling (14,15) and by site-directed mutagenesis/ electrophysiology experiments (12, 16, 17) to be lining the pore. In the shut-channel form cf the receptor, this helix is bent inward, toward the central axis, making the lumen of the pore narrowest near the middle of the membrane. This is the most constricted region of the whole ion pathway and therefore presumably corresponds to the gate of the channel.

A tentative alignment can be made between the three-dimensional densities and the amino acid sequence of M2 (Fig. 11.3) (11). This alignment places the charged groups at the ends of M2 symmetrically on either side of the lipid bilayer and a highly conserved leucine residue (Torpedo aLeu251) at the level of the bend. It seems likely that the leucine side chains, by side-to-side interactions with neighboring M2 segments, are involved in making the gate of the channel. Site-directed mutagenesis experiments, combined with electrophysiological study of function, have highlighted the uniqueness of the conserved leucine residue in relation to the gating mechanism. The profound effects of mutating this leucine to a hydrophilic amino acid on the agonist sensitivity of the receptor and its desen-sitization properties were first reported for the recombinant homomerica7 neuronal nicotinic receptor by Revah et al. (20). In the muscle-type receptor, progressive replacement of leucines by serines (21) or by threonines (22) increases, by roughly uniform increments, the sensitivity of the channel (i.e., decreases the EC,, value for ACh). A similar effect is seen in neuronal nicotinic receptors that contain a3, P4, and j83 subunits (23). However, other experiments [e.g., Wilson and have been interpreted to indicate that the gate is located closer to the cytoplas-mic membrane surface.

2.2.3 Membrane-SpanningPore. The membrane-spanning portion of the receptor has not yet been completely resolved by direct structural methods, although the pore-lining segments are partially visible as a ring of five rod-shape densities, consistent with an a-heli-cal configuration. This helix is the part of the structure closest to the axis of the receptor

2.2.4 The Snail ACh-Binding Protein. Before going on to discuss the agonist binding site, we next discuss the snail acetylcholine-binding protein (AChBP). Glial cells in the snail, Lymnea stagnalis, produce and secrete this protein, which is a homopentamer having structural homology with the large N-termi-nal portion of the extracellular domain of ion

«G153S

«G153S

ffYl98

£D175N

Cys M2-M3

loop loop

(membrane)

Figure 11.4. Three subunits of the Lymnaea AChBP viewed perpendicular to the fivefold axis of symmetry and from the outside of the pentamer. The inner and outer sheets of the j8-sandwich are blue and red, respectively, whereas the putative ligand HEPES is purple. The approximate positions of the a carbons of residues discussed in the text are marked with arrows on the foremost subunit. The approximate positions of the cell membrane and of the M2-M3 loop are shown diagrammatically. The inner j3-sheet (blue) is thought to rotate after agonist binding and to interact with the M2-M3 loop (as indicated by the asterisk). See color insert.

ffYl98

£D175N

Cys M2-M3

loop loop

(membrane)

Figure 11.4. Three subunits of the Lymnaea AChBP viewed perpendicular to the fivefold axis of symmetry and from the outside of the pentamer. The inner and outer sheets of the j8-sandwich are blue and red, respectively, whereas the putative ligand HEPES is purple. The approximate positions of the a carbons of residues discussed in the text are marked with arrows on the foremost subunit. The approximate positions of the cell membrane and of the M2-M3 loop are shown diagrammatically. The inner j3-sheet (blue) is thought to rotate after agonist binding and to interact with the M2-M3 loop (as indicated by the asterisk). See color insert.

channels in the nicotinic superfamily. The protomer of AChBP is composed of 210 amino Is and has 20-23% sequence identity with muscle-type ACh receptor subunits. It tains most of the residues that were previ-ly suspected to be involved in ACh binding the receptor. Its crystal structure was ed recently to 2.7-A resolution (25), re-vealling the protomer to be organized around sets of j3-strands, forming Greek key-like ;ifs, folded into a curled /3-sandwich. The indwich can be divided into inner- and jr-sheet parts, shown respectively in blue and Ired in Figure 11.4, which are covalently linked together through a disulfide bond. The -loop11 disulfide bond (C128-C142 in Tor> and human al subunits) plays an important! structural role in stabilizing the three-dimensional fold (25) and is absolutely conserved among all members of the ion channel superfamily.

AChBP has been crystallized only with HEPES, rather than ACh, bound, and "owing to low occupancy and limited resolution, the precise orientation of the HEPES molecule cannot be definitely resolved" (25). In overall appearance, it is very similar to the extracellular domain of the receptor (Fig. 11.5). The C-loop is particularly prominent both in AChBP (Fig. 11.4) and in the receptor, where it is shown as the projection labeled C in Fig. 11.5. The C-loop contains several conserved residues that are thought to be part of the acetylcholine binding site: two adjacent cysteines that are characteristic of a subunits, homologs of o:C192, «C193, two tyrosines (c*Y190, <*Y198), and an aspartic acid (aD200). It is orientated more tightly against the neighboring

Figure 11.5. Wooden model of the extracellular part of the ACh receptor, based on the 4.6-A map of the shut channel (8). The membrane surface is at the bottom of the figure. C denotes the C-loop region of the receptor (see also AChBP, Figs, 11.4 and 11.6), which makes part of the ACh-binding site. The width of each wooden slab corresponds to 2 A. [From Unwin et al. (30).]

Figure 11.5. Wooden model of the extracellular part of the ACh receptor, based on the 4.6-A map of the shut channel (8). The membrane surface is at the bottom of the figure. C denotes the C-loop region of the receptor (see also AChBP, Figs, 11.4 and 11.6), which makes part of the ACh-binding site. The width of each wooden slab corresponds to 2 A. [From Unwin et al. (30).]

subunit in AChBP than it is in the (unligan-ded) receptor. These residues on the C-loop, and other nearby aromatic residues (see below) that form part of the binding site in AChBP, are homologs of residues that have been postulated to form the binding site of nic-otinic receptors, on the basis of mutational studies and/or photoaffinity labeling studies. One exception is H145 (see Fig. 11.6a),which aligns with oY151 and which had not been thought to be important. Also, some residues postulated to be part of the binding site by other methods do not appear to be in the AChBP binding site (e.g., aY86).

2.3 ACh Binding Region of the Receptor

The two ACh binding sites are located in the extracellular domain, about 30 A from the membrane surface, or 45 A from the gate. Although the actual ligand-binding site has not yet been identified definitively within the three-dimensional structure of the receptor, is expected to bind through cation-a interactions, where the positive charge of its quaternary ammonium moiety interacts with electron-rich aromatic side chains (26). The recently solved structure of AChBP (25), discussed above, shows that the "signature"

aromatic residues lie in a pocket next to tht interface with the anticlockwise-positionec protomer, as seen from the "synaptic cleft.' The pocket identified in AChBP would lie be hind the protruding densities, labeled C in Fig 11.5, near the a/y and a/8 interfaces. The den sities at C can also be identified with the C loop structure in AChBP (see Fig. 11.4), al though they do not curve around so tightly toward the neighboring subunits, making i more open cleft in the (unliganded) receptor.

The key aromatic residues at the binding site are most likely aY93, aW149, aY190, anc aY198. These residues are located in thre< separate loops of the polypeptide chain (3) designated A (Y93), B (W149), and C (Y19( and Y198). All were identified as being neai the agonist binding site by labeling with i small photoactivatable ligand that covalentlj reacts with the receptor upon UV-irradiatior and acts as a competitive antagonist (27), anc the first three are highly conserved in alignec positions of all muscle and neuronal a sub units. Experiments in which a series of unnat ural tryptophan derivatives were substitutec in place of the natural residues have suggestec that the side chain of aW149 is in van dei Waals contact with the quaternary ammo nium group of ACh in the bound state of thi receptor (26). Chemical labeling has alsc shown that the pair of adjacent cysteines (a;C192 and <xC193) is likely to be close to the binding site (28).

Figure 11.6a shows the binding site regior for the AChBP with the ligand (HEPES), tc show the position of the residues mentionec above. The "plus" side of the interface (Fig 11.6a) is analogous with the a subunit of the nicotinic receptor.

The most important residues in neighbor. ing subunits that influence ACh binding art: W57 of the 5 subunit and the homologous W5£ of the y subunit (6, 29). The "minus" side ol the AChBP interface, shown in Fig. 11.6b would be the y/e or the 8 subunit in the receptor, but in the AChBP it is another identica" subunit. The residues shown to be in contaci with the ligand, labeled in Fig. 11.6b, mostlj have no obvious analog in the ACh receptor The one important exception is W53, which corresponds to the tryptophan residues men. tioned above. Consequently, the snail protein

Figure 11.6. The binding site of AChBP. The ligand (HEPES) is in yellow. Blue and red regions denote the inner- and outer-sheet parts of the j3-sandwich (30). The views in a and b are with the fivefold axis vertical, and the "'membrane" at the bottom. (a) The structure shown is analogous to the binding site of the the Torpedo or human ctl subunit, to which the numbering of identified residues refers. Numbers in parentheses refer to AChBP. (b) Surface of the neighboring protomer that faces the binding site of AChBP. In the receptor this would be part of the y, e> or 6 subunit. Numbers in parentheses refer to the AChBP. W53 aligns with W55 in mouse y, e, and 6 nicotinic receptor subunits, but most of the residues have no obvious analog in the y, e, or 6 subunits of the nicotinic receptor. For example, Q55 aligns with mouse 7E57, eG57, and SD57; LI 12 aligns with mouse ?Y117 or ST119, and Y164 is A or G in the nicotinic subunits. See color insert.

model is rather less helpful about the non-a side of the receptor interface.

Was this article helpful?

0 0

Post a comment