Minisymposium 16: Cell walls

Abs # 31001: What the crystal structure of expansin tells us about its mechanism of action

Presenter:	Cosgrove, Daniel J, fsl@psu.edu
Authors	Yennawar, Neela  (A)   Li, Lian Chao  (B)   Yennawar, Hemant  (A)   Cosgrove, Daniel J (B)
Affiliations:	(A): Department of Biochemistry and Molecular Biology, Penn State University (B): Department of Biology, Penn State University
Web Site:	http://www.bio.psu.edu/expansins/

Expansins are proteins that loosen cell walls in unique ways: they rapidly induce extension and stress relaxation of heat-inactivated cell walls, but they do not increase wall plasticity or elasticity and they do not hydrolyze wall polymers. Expansins are thought to dissociate lateral binding of wall polysaccharides, thereby allowing polymer creep and turgor-driven wall extension. Members of two expansin families, now called EXPA and EXPB (see http://www.bio.psu.edu/expansins/), have characteristic, though distinctive, wall extension activity. To investigate expansin’s mechanism of wall loosening, we have crystallized native Zea m 1 and determined its structure by X-ray crystallography. Zea m 1 is a member of the EXPB subfamily known as group-1 grass pollen allergens. It has wall-loosening activity with high specificity to grass cell walls. Zea m 1 has two small domains (~15 kD and 11 kD) connected by a short linker. Domain 1 (ZeaDom1) has substantial structural similarity to the catalytic domain of a family-45 endoglucanase (Humicola EGV). Residues His105, Thr25, Asp107 and Asn97 in Zea m 1 form a hydrogen-bonded network identical to the EGV active site. However, the second Asp residue that is necessary for EGV enzymatic activity is missing in ZeaDom1. In-vitro assays with wall polysaccharides did not detect endo- or exo-glycanase activity, indicating that Zea m 1 does not loosen walls by hydrolytic action and so it must act by another mechanism. Consistent with this conclusion, EGV lacks the wall extension activity characteristic of expansins. Compared with EGV which has a deep glucan-binding cleft, the homologous site in ZeaDom1 is much more open, reducing steric hindrance for access to large polysaccharide structures, such as a cellulosic surface. The surface of Zea m 1 is also much more open than that of xyloglucan endotransglycosylase. Domain 2 (ZeaDom2) is an Ig-like beta sandwich, essentially identical to that of Phl p 2, a group-2/3 grass pollen allergen. ZeaDom2 has aromatic and polar residues that form potential surfaces for polysaccharide binding. Consistent with this idea, we find that group-2/3 allergens bind to cell walls and can induce cell wall extension, albeit with low efficiency. We propose that the two expansin domains act cooperatively to disrupt glucan-glucan binding and that the consequent reduction in strain energy of the cell wall polymers drives this action. The result is cell wall polymer creep.