The authors have solved the structure of L-selectin, which can now be added to previous structures of P- and E-selectin. The extracellular domains of all three have an N-terminal lectin domain and an EGF-like domain that are separated by a hinge. In L- and E-selectin, the two sides of this hinge are connected by a hydrogen bond between a tyrosine at residue 37 and an asparagine at residue 138.
In the absence of force, the lectin domain of L-selectin flipped between closed and open angles relative to the EGF domain, flexing at the hinge. But with increasing force, L-selectin was more often in an open conformation. A mutant L-selectin without the hydrogen bond between Y37 and N138 took less than half the force to be opened than did the wild type, suggesting that the hinge is held shut by this hydrogen bond.
Maximal tethering of microspheres with this mutant L-selectin also occurred at a lower flow. The change was due to increased rotational freedom of the mutant's hinge, which increased the likelihood of protein–ligand contact.
After the L-selectin bound its ligand, the authors found, dissociation also slowed as force increased (up to an optimal level, at which point the selectins are fully open). Again the optimal force was lower for the hydrogen bond mutant than for wild-type L-selectin. The group proposes that an open hinge, which aligns selectin's binding interface with the direction of flow, might allow a ligand that would otherwise let go to slide along its selectin until it can rebind to a new site.
P-selectin has a glycine at position 138 and thus does not form a hinge-closing hydrogen bond. The increased flexibility of the P-selectin hinge is consistent with its ability to form catch bonds under smaller shear forces than those of L-selectin.