Conformationally Biased Mutants

Arrestin conformation is the key determinant of its functional capabilities (Gurevich and Gurevich 2003). Thus, mutations that affect the flexibility of the molecule and/ or limit the conformational space it can "explore" would enhance or reduce its ability to interact with some partners without affecting the binding to others, thereby dramatically shifting arrestin-mediated signaling. Crystal structure of arrestins (Hirsch et al. 1999; Han et al. 2001; Sutton et al. 2005) and functional studies (Vishnivetskiy et al. 2002; Hanson et al. 2006a,b) identify two significant conformational rearrangements in the molecule: receptor- and microtubule-binding induced the release of arrestin C-tail and the movement of the two arrestin domains relative to each other (Fig. 2). Destabilization of intramolecular interactions that hold arrestin in the basal conformation, the polar core and the one that anchors the C-tail to the body of the N-domain (Fig. 4), yield mutants that are more flexible than the wild type (Carter et al. 2005). These "pre-activated" arrestins demonstrate dramatically enhanced binding to active unphosphorylated receptor (Gurevich 1998), shut off the signaling even without receptor phosphorylation (Gray-Keller et al. 1997; Kovoor et al. 1999; Celver et al. 2002), and change the pattern of receptor trafficking, greatly reducing its down-regulation and facilitating recycling (Pan et al. 2003). Interestingly, in cell-based assays these mutant forms of all arrestin subtypes bind JNK3 normally, but show reduced affinity for Mdm2 (Song et al. 2006, 2007), demonstrating that mutations designed to change arrestin conformation differentially affect its interactions with receptors and non-receptor partners. Extensive deletions in the inter-domain hinge region (Fig. 4) restrict domain mobility, limiting their ability to move "forward," in the direction of receptor-binding concave sides (Fig. 2). These mutations severely reduce receptor binding (Vishnivetskiy et al. 2002), considerably enhance the binding to microtubules (Hanson et al. 2007b) and ubiquitin ligase Mdm2, but do not appreciably affect JNK3 interaction (Song et al. 2006, 2007). Mutations that drastically change arrestin conformation and therefore simultaneously affect its interactions with multiple partners demonstrate the feasibility of constructing arrestins with "biased" signaling capabilities, but are not likely to yield proteins with therapeutic potential. Mutations that change arrestin flexibility in more subtle ways and affect its interactions with just one or very few partners appear more promising in this regard. Arrestin structure identifies the part of the molecule that must be targeted to achieve this goal: the extensive hydrophobic inter-domain surface along which the domains "slide" relative to each other (Fig. 4) (Sutton et al. 2005). The introduction of appropriately positioned opposite charges in the two domains can "fix" their relative orientation, "freezing" the arrestin molecule in any conformation between the two extremes, receptor-bound-like and microtubule-bound-like (Fig. 2). Recent findings suggest that a similar strategy is actually used by cells, where arrestin bound to differentially phosphorylated receptor demonstrates a distinct signaling bias: it does or does not stabilize agonist-receptor interaction (Key et al. 2003), and does or does not promote ERK1/2 activation (Kim et al. 2005; Ren et al. 2005), while enhancing receptor internalization in both cases.

Obviously, binding partners that interact with both arrestin domains, such as receptors (Vishnivetskiy et al. 2004), microtubules (Hanson et al. 2007b), calmodulin (Wu et al. 2006), JNK3, Mdm2 (Song et al. 2007), and probably a number of others, are sensitive to these conformational manipulations, in contrast to the partners with the binding site localized on a single domain. Subtle manipulation of the conforma-tional state of arrestin has a potential to change its signaling capabilities to a great extent. Moreover, the mutations affecting interactions with individual partners (described in Sect. 6.2) will likely produce a more dramatic effect in the context of conformationally restricted arrestins. Thus, the combination of both approaches has a better chance of yielding "designer arrestins" with precisely fine-tuned signaling bias specifically crafted for particular therapeutic purposes. Further development of vectors targeting individual tissues and cell types is needed to enable the delivery of these molecular tools to their intended targets.

0 0

Post a comment