Extended Data Figure 10: Conserved motifs of residues can be used for structural comparison.

a, Zoomed view of two archetypal motifs, conserved throughout the Cys-loop receptor family: the ‘stabilizing box’ (top) and the ‘Cys-loop region’ (bottom). The stabilizing box is made of residues belonging to or neighbouring loops defining the orthosteric binding site: motifs WXPD (loop A), S-Φ-X-H/Y (loop B; Φ being a tryptophan and H/Y a histidine in the 5-HT₃ receptor), W (loop D) and GXV (loop E). The ‘Cys-loop region’ is made of residues belonging to or neighbouring loops defining the extracellular domain (ECD)–transmembrane domain (TMD) interface: Φ-P-Φ-D (loop β6–β7), R (pre-M1) and Φ (loop β8–β9); the latter being member of the GEW motif in nACh and 5-HT₃ receptors. These sets of residues, shown as yellow sticks, define a core with little configurational variation among known structures of Cys-loop receptors, and can thus be considered as defining a ‘rigid rod’ (the ‘rod’ can be seen as the union of the Van der Waals balls shown as dotted spheres). b, The two rigid rods are represented as yellow surfaces for the 5-HT₃ receptor, along with another conserved residue, F242, located close to the intracellular end of the M1 segment. Notably, at the quaternary level, this last residue seems to be superimposable from one structure to the other. It can thus be used to define a common framework into which structures from different species can be compared. The segment joining this residue to the conserved arginine of the pre-M1 is shown as a broken line, while the tetragon joining αC atoms of D97, V123, F142 and W187 is shown with plain lines. c, Rotated view of the 5-HT receptor, now viewed from the extracellular medium, further defining the geometric elements used to compare structures. Additional segments joining the five αCs of F242 are shown as plain lines, along with a broken line representing the segment joining the conserved Leu 9′ of M2 to the conserved Pro of loop M2–M3. d, In the framework defined above, the tetragons of the 5-HT₃ receptor, GluCl, GLIC (in its apparently open form) and ELIC, are represented viewed parallel to the membrane (left) and perpendicular to the membrane (right). It appears from this representation that there is a gradual motion when going from ELIC (closed) to GLIC (open), GluCl (open) and finally to the 5-HT₃ receptor. This motion can be approximated as an anti-clockwise rotation of the ECD with respect to the lower TMD when viewed from the extracellular medium. Notably, the segment used to represent the transmembrane pore status seems to follow the motion of the ECD, but in a more all-or-nothing way since ELIC is more distinguishable from the other structures than the latter between themselves. Overall, the reductionist representation of structures by their conserved motifs minimizes the bias of inter-species comparison and may present an interesting way to classify conformations. It suggests implicitly that the motion sampled by the structures compared here results both in different shapes for the neurotransmitter binding pocket and in different states for the transmembrane channel pore.