Why Does Synergistic Activation of WASP, but Not N-WASP, by Cdc42 and PIP2 Require Cdc42 Prenylation?

Human Wiskott-Aldrich syndrome protein (WASP) and neural WASP (N-WASP) are homologous proteins that contain intrinsically disordered regions throughout their sequences [1]. Typically of intrinsically disordered proteins (IDPs), WASP and N-WASP engage in numerous intramolecular and intermolecular interactions. Intramolecular interactions result in autoinhibition 2, 3, 4, 5 whereas intermolecular interactions result in activation and subsequent engagement in cellular activities. A major activity of WASP and N-WASP is to stimulate the initiation of branched actin polymerization [6]. WASP is expressed only in hematopoietic cells [7], whereas N-WASP, despite its prefix, is ubiquitously expressed [8]. Mutations in WASP are responsible for its namesake, an X-linked immunological disorder [7]. The fact that N-WASP in the hematopoietic cells of patients does not compensate for the defects caused by WASP mutations points to nonredundant roles of these two proteins. Indeed, different effects of WASP and N-WASP have been reported in filopodium formation [9], podosome-mediated extracellular matrix degradation [10], and T cell chemotaxis [11]. Phylogenetic analysis suggested a direct link between N-WASP and ancestral WASPs, whereas hematopoietic WASP branched off during the onset of vertebrates [12]. WASP and N-WASP are excellent models for studying how a single IDP binds multiple regulators to achieve full activation as well as for deciphering the differences between closely related IDPs in processes ranging from activation to cellular activities.

WASP and N-WASP share ∼50% sequence homology [8] and have the same domain organization with one exception 1, 13 (Figure 1A). The functional element resides toward the C-terminus, where the verprolin-central-acidic (VCA) region binds G-actin (via the V motif) and Arp2/3 (via the CA motifs) to initiate actin polymerization 2, 3, 14, 15. The exception to the shared domain organization is an additional V motif in N-WASP, which increases the activity for stimulating actin polymerization [16]. Upstream of the functional element are regulatory elements. At the N-terminus is a WASP homology 1 (WH1; also known as EVH1) domain that binds WASP-interacting proteins [17]. Following a linker of ∼65 (or ∼35 in N-WASP) residues, a basic region (BR) is present to bind the acidic lipid phosphatidylinositol 4,5-bisphosphate (PIP2) 3, 5, 18, 19, 20. Higgs and Pollard [18] suggested that the WH1 domain of WASP may also participate in PIP2 binding. Immediate after the BR is a GTPase-binding domain (GBD), which binds Cdc42 and other small GTPases 9, 21. In WASP, several Lys residues in the basic region, including K230KK232, are involved in electrostatic interactions with Cdc42 [22] and contribute to their binding rate 23, 24, 25. Lastly, a proline-rich region (PRR) connects the GBD and the VCA region, and binds SH3-containing proteins such as Nck [26].

Under resting conditions, WASP and N-WASP are autoinhibited by intramolecular interactions, between the GBD and the C motif 3, 4, 5, 9 and between the BR and the A motif 1, 25, 27. Binding to PIP2 and Cdc42 relieve N-WASP of autoinhibition, and the two regulators have synergistic effects on activation [3]. Endogenous Cdc42 is prenylated at the C-terminus and then tethered to membranes [28], but a follow-up study reported no difference between prenylated and soluble Cdc42 in their synergistic effects with PIP2 on N-WASP activation [19]. Prehoda et al. [5] also observed synergistic activation of N-WASP by PIP2 and soluble Cdc42, though a follow-up study found higher activity when Cdc42 was prenylated and colocalized with PIP2 [20]. In contrast, synergistic activation of WASP by PIP2 and Cdc42 required Cdc42 prenylation [18]. Indeed, instead of doubling the effect of PIP2 when prenylated Cdc42 was added, a 30% decrease was observed when soluble Cdc42 was added. Both the synergistic effect of soluble Cdc42 on N-WASP activation and the negative effect of soluble Cdc42 on WASP activation were reproduced by Tomasevic et al. [29], but the reason for this contrast has remained a mystery.

The structures of the WASP GBD bound to the C motif (modeling the autoinhibited state) and to Cdc42 (modeling the activated state) have been determined by NMR spectroscopy 4, 22. In the autoinhibited state, the C motif folds into an α-helix and docks to a three-helix platform formed by the GBD (Figure S1A). In the activated state, the platform falls apart as one of these helices, along with upstream segments, binds to Cdc42 (Figure S1B). Very little structural information is available for the binding of WASP or N-WASP with the other activator, i.e., PIP2. Binding studies [20] have shown that the N-WASP BR binds multiple PIP2 molecules and it is the number of basic residues, not the precise sequence, that determines the PIP2 binding affinity, typical of IDP binding to acidic membranes [30]. Structural characterization of IDP-membrane binding, due to its dynamic nature, presents challenges for experimental techniques, but this problem can now be addressed by molecular dynamics (MD) simulations 31, 32, 33, 34.

Here we report MD simulation results on the binding of WASP and N-WASP with PIP2 and soluble and prenylated Cdc42. The simulations provided an opportunity to visualize how WASP and N-WASP simultaneously bind with two regulators, PIP2 and Cdc42, thereby fully releasing the VCA region for stimulating actin polymerization. For N-WASP, simultaneous binding can occur whether Cdc42 is soluble or prenylated. However, WASP can do so only when Cdc42 is prenylated. This result explains the previous puzzling observation that synergistic activation of WASP, but not N-WASP requires Cdc42 prenylation. The origin of this contrast lies in the differences between the BRs of WASP and N-WASP, in particular regarding the number of basic residues and their spacing from the GBD.

留言 (0)

沒有登入
gif