Molecular Interactions of Shank Scaffolds Self Association

One of the most prominent interaction partners of Shank proteins is the Shank proteins themselves. The C-terminal sterile alpha motif (the SAM domain), which is required for the correct synaptic targeting of Shank2 and Shank3 (but not

Shank1; Boeckers et al. 2005; Sala et al. 2001), has a strong ability to self-associate. Three-dimensional structures of the recombinant Shank3 SAM domain led to the identification of an extended surface area that is available for polymerization.

The SAM domains first assemble into helical filamentous structures, which then may pack alongside each other, leading to the formation of a two-dimensional array (Baron et al. 2006). Interdomain contacts generate binding sites for Zn2+ that further stabilize the sheet. Site-directed mutagenesis and expression of mutant constructs in cultured neurons indicated that intact Zn2+-binding sites and the ability to polymerize are required for correct postsynaptic targeting of the Shank3 protein. Baron et al. (2006) argued that the arraying capacity of the Shank SAM domains may provide a driving force in PSD assembly and that the Zn2+-dependent addition or removal of Shank monomers might contribute to structural plasticity of the postsynaptic scaffold (see also Gundelfinger et al. 2006). Intriguingly, Jan et al. (2002) have reported that Zn2+ can trigger the reassembly of a denatured PSD preparation in vitro, further supporting the functional relevance of the SAM domain/Zn2+ complex.

Further self-association is provided by the PDZ domain, which, besides binding various membrane receptors and signaling proteins (see below), also has the capacity to form dimers, as observed in the high-resolution structure obtained from the Shank1 PDZ domain in complex with the PDZ ligand motif derived from GKAP/SAPAP (Im et al. 2003). Taken together with the remarkable capacities of the SAM domain, this should already be enough self-association to provide for a complex network of Shank protein molecules. Nevertheless, one more mode of binding between two Shank molecules has been detected in the N-terminal part of the proteins. Direct binding between SH3- and ankyrin-repeat domains also contributes to the ability of Shank to form large molecular aggregates, independent of the C-terminal domains. When expressed recombinantly in COS cells (i.e., in the absence of other synaptic interaction partners of Shank), SH3-Ank concatemerization leads to filamentous aggregates of Shank1 that are deposited in aggresomes, possibly followed by proteasomal degradation (Romorini et al. 2004). Association with GKAP and PSD-95, either from endogenous supplies in neurons or coexpressed in COS cells, prevents aggregation/degradation, leading to formation of Shank/GKAP/PSD-95 clusters instead. Besides pointing to an intricate network of intra- and intermolecular interactions regulating the assembly of Shank-containing clusters, this may also tell us two things: (1) Shank is instable on its own, particularly when not attached to the GKAP/PSD-95 complex; (2) the fact that overexpressed Shank in neurons is not instable suggests that there are enough interaction partners (enough GKAP/PSD-95 in particular) to support it; in turn, the availability of Shank (rather than GKAP/PSD-95) should be the limiting factor for postsynaptic complex formation. This would provide a straightforward explanation for the striking effects of haploinsufficiency for the SHANK3 gene. The lack of one functional copy of the gene would thus lead to severe neurological deficits because individual neurons cannot provide enough Shank protein for synapse development and maintenance. In addition, it makes sense that the amount of Shank protein in dendrites is tightly controlled, most likely by local synthesis derived from dendritically transported mRNAs, and activity-dependent degradation through the ubiquitin proteasome system (Boeckers et al. 2004; Ehlers, 2003; see below).

0 0

Post a comment