Fig. 5. Proposed role of the salt bridge in the strand-swap mechanism in mutant and wild-type N-cadherin. A full explanation is offered in the text. (a) Domain 1 in isolation. An equilibrium between docked and undocked Trp2 favours the docked form because the salt bridge between Glu89 and the N-terminus (shown as a dotted line) stabilises Trp2 insertion. (b) Adhesion between wild-type molecules. A transition state in which Trp2 is undocked is sampled from either side and is depicted as an `unshaded' tryptophan. Adhesion is moderate. (c) The salt bridge on one side is prevented by extension of the N-terminus. (d) The salt bridge is prevented by the mutation Glu89Ala. In both c and d, adhesion is very weak. Although one strand can cross-intercalate, the process competes unfavourably with intramolecular docking of Trp2 into the wild-type domain. (e) The two mutations form a complementary pair. Intramolecular docking is prevented and therefore the activation barrier for strand exchange is lowered. Exchange of one strand is possible and cross-intercalation of Trp2 is stabilised by the salt bridge. Adhesion is enhanced. (f) The double mutation is present on each side. The salt bridge cannot form to support strand exchange so there is no adhesion.