Vanderbilt Institute of Chemical Biology



Discovery at the VICB







Structural Basis of Arrestin Activation


By: Carol A. Rouzer, VICB Communications
Published:  November 21, 2017



New data provide key insights that explain how arrestin-3 is activated by a diverse range of G protein-coupled receptors and nonreceptor ligands.  


Arrestins are proteins discovered for their ability to block the interaction of G protein-coupled receptors (GPCRs) with G-proteins, thereby preventing or stopping G protein-mediated signaling. Prior work has shown that arrestins bind primarily to phosphorylated and activated GPCRs through two distinct sensors. The phosphorylation sensor recognizes and interacts with receptor-bound phosphate groups, with very little dependence on the underlying amino acid sequence of the receptor. The activation sensor detects the activation state of the receptor; however, the mechanism by which this sensor is able to carry out its function for over 800 structurally distinct GPCRs is a long-standing mystery. Another important question relates to the ability of some arrestins to initiate signaling independently of any receptor interaction. The desire to better understand the mechanism by which this receptor-independent signaling occurs led Vanderbilt Institute of Chemical Biology members Tina Iverson and Seva Gurevich along with their collaborators Candice Klug (Medical College of Wisconsin) and Chad Brautigam (University of Texas - Soutwestern) to investigate the structure of arrestin-3 in complex with a small molecule activator, inositol hexakisphosphate (IP6) (Q. Chen et al. Nat. Commun., 2017, 8:1426 DOI: 10.1038/s41467-017-01218-8). Their results provide important insights into the mechanisms of both receptor-dependent and receptor-independent arrestin function.


The investigators initiated their studies by obtaining a 2.4 Å resolution crystal structure of an arrestin-3-IP6 complex. The structure revealed that arrestin-3 had trimerized during complex formation and that two molecules of IP6 were bound per arrestin monomer. All of the IP6 molecules were located at the points of interaction between two monomers of arrestin-3 (Figure 1). Like all members of its class, the arrestin-3 monomer comprises two relatively independent domains, designated the N-domain and the C-domain. The basal state is characterized by the presence of a "polar core" formed by hydrogen bonds between Asp-27 and Arg-170 in the N-domain, Asp-298 and Asp-291 in a part of the C-domain known as the lariat loop, and Arg-393 in another portion of the C-domain designated the C-tail (Figure 2). Disruption of these interactions accompanies arrestin activation as does a rotation of the two domains relative to one another. The extent of rotation varies depending on the means of activation, but values as high as 20o have been reported. The crystal structure revealed that IP6 binds to arrestin-3 at the same location as the phoephosrylated GPCR, disrupting key interactions normally found in the basal state (Figure 3). Furthermore, a 17.7o rotation of the domains relative to their positions in basal arrestin-3 (Figure 4) indicated that the protein had adopted an active conformation in the IP6 complex.


FIGURE 1. Diagrammatic representation of the crystal structure of the arrestin-3-IP6 complex. Each arrestin-3 monomer is shown in a different color (blue, gold, and gray). The IP6 molecules are shown in stick format. The circle highlights the finger loops. Figure reproduced under the Creative Commons Attribution 4.0 International License from Q. Chen et al. Nat. Commun., 2017, 8:1426 DOI: 10.1038/s41467-017-01218-8).




FIGURE 2. Close-up view of the polar core of basal arrestin-3. The lariat loop is shown in green, and the C-tail is in bronze. Key residues forming hydrogen bonds in the core are labeled. Figure reproduced under the Creative Commons Attribution 4.0 International License from Q. Chen et al. Nat. Commun., 2017, 8:1426 DOI: 10.1038/s41467-017-01218-8).



FIGURE 3. Close-up view of the polar core of the arrestin-3-IP6 complex. The lariat loop is shown in green, and the C-tail is in pink. Key residues forming hydrogen bonds in the core are labeled. Note the disruption of interactions shown in Figure 2. Figure reproduced under the Creative Commons Attribution 4.0 International License from Q. Chen et al. Nat. Commun., 2017, 8:1426 DOI: 10.1038/s41467-017-01218-8).


FIGURE 4. Diagrammatic representation of the N-domain of basal arrestin-3 (gray with pink C-tail) overlaid on the N-domain of arrestin-3 as seen in the complex with IP6 (blue). Figure reproduced under the Creative Commons Attribution 4.0 International License from Q. Chen et al. Nat. Commun., 2017, 8:1426 DOI: 10.1038/s41467-017-01218-8).



Prior evidence had suggested that IP6 mediates receptor-independent arrestin-3 signaling that modulates the activity of the c-Jun N-terminal Kinase-3 (JNK3) pathway. To test whether IP6 can mediate activation of JNK3 via arrestin-3 the researchers evaluated the ability of JNK3 to interact with arrestin-3 in an in vitro binding assay. The results showed that arrestin-3 binds to JNK3 even in the absence of IP6. However, addition of IP6 to a fixed amount of arrestin-3 doubled the amount of protein-bound JNK3. Further studies, using microscale thermophoresis, enabled a measurement of the affinity of the arrestin-3-IP6 binding interaction, revealing two KD values of 57 nM and 90 μM. These findings were consistent with the presence of two IP6 molecules per arrestin-3 monomer and supported the hypothesis that the binding interaction can occur in intact cells, which typically exhibit IP6 concentrations between 35 μM and 105 μM. The investigators then devised an assay to measure activation of JNK3 by arrestin-3 in intact cells, showing that JNK3 activation (phosphorylation) was substantially increased when both arrestin-3 and ASK1 (an upstream JNK3 kinase) were expressed. However, if wild-type arrestin-3 was replaced by a mutant arrestin-3 protein in which the IP6 binding sites in either the C-domain (ΔCIP6) or the N-domain(ΔNIP6) had been removed, JNK3 activation did not occur.

To further establish the physiological relevance of the arrestin-3-IP6 complex, the researchers investigated the importance of trimer formation. They first added IP6 to solutions of arrestin-3 in vitro and showed that the protein formed a trimer. Then they idnetified Cys17 as important for trimer formation. They supported this prediction by showing that a mutant arrestin-3 in which all cysteine residues had been replaced with valine, leucine, or serine did not form trimers upon IP6 addition. This finding was particularly important because this Cys-less mutant retains the ability to bind to activated and phosphorylated GPCRs, indicating that the mutations had not resulted in a global disruption of the protein's function. Direct binding studies revealed a marked reduction in binding affinity of the Cys-less mutant for IP6 as indicated by KD values of 1.1 μM and 1.2 mM. Further work, using double electron-electron resonance, demonstrated a 70% reduction in the movement of the C-tail in response to IP6 in the Cys-less mutant as compared to the wild-type protein. Consistently, this mutant also exhibited a 65% reduction in its capacity to activate JNK3 in the cellular assay. These findings suggested a strong correlation between trimer formation and both IP6 binding and arrestin-3-dependent JNK3 activation.

The crystal structure of the arrestin-3-IP6 complex showed that IP6 binds to the protein at essentially the same sites where the phosphate groups of phosphorylated GPCRs bind. Thus, it was easy to imagine how IP6 can trigger the phosphate sensors of arrestin-3. However, the mechanism by which IP6 triggers the activation sensor was not so readily apparent. Prior work had suggested that a portion of arrestin called the finger loop acquires an α-helical conformation upon activation, exposing a hydrophobic surface that binds to a complementary hydrophobic pocket on the intracellular side of activated GPCRs. The investigators noted that, in the arrestin-3-IP6 complex, the finger loop was present as an α-helix (Figure 5) and that the loops from all three subunits had established hydrophobic interactions with each other at the center of the trimer (Figure 1). These observations offered strong support for the hypothesis that the finger loop was the key activation sensor trigger.



FIGURE 5. (a) Overlay of the finger loop and a second structure, the C-loop (green), as seen in the arrestin-3-IP6 complex in contrast to the comparable structures in basal arrestin-3. (b) Structure of the finger loop and C-loop in basal arrestin-3. Figure reproduced under the Creative Commons Attribution 4.0 International License from Q. Chen et al. Nat. Commun., 2017, 8:1426 DOI: 10.1038/s41467-017-01218-8).




Arrestin-3 can bind to over 800 different activated and phosphorylated GPCRs despite significant structural differences between them. A careful analysis of the crystal structure of the arrestin-3-IP6 complex and those of various arrestin-GPCR complexes revealed that the finger loop is connected to the protein via a flexible tether that would allow it to adopt multiple conformations for interaction with many different receptors. This led the investigators to hypothesize that helicity, hydrophobicity, and flexibility of the finger loop were key to its ability to serve as the activation sensor. They tested this hypothesis by creating mutant arrestin-3 proteins, each of which would disrupt one of these properties, and they then evaluated the ability of the mutants to bind to the M2 muscarinic receptor or the D2 dopamine receptor, two GPCRs noted for a high dependence on the activation sensor for arrestin binding. The results revealed that perturbation of helix formation or flexibility of the finger loop markedly reduced receptor binding. Adding charged residues to the loop resulted in receptor-specific effects. Finger loop alteration also reduced trimer formation in response to IP6 and JNK3 activation in intact cells, although the magnitude of the effects varied depending on the precise mutation.

The crystal structure of the arrestin-3-IP6 complex revealed four "switch" regions in the protein that undergo conformational changes as a result of IP6 binding (Figure 6). The investigators hypothesized that these could be binding sites for effector proteins that are important in IP6-dependent signaling. To assess the importance of one of these regions (amino acids 184-191), the investigators introduced cysteine residues at position 338 and either 186 or 187 in the Cys-less mutant. These mutations allowed disulfide bond formation that locked the protein into either a basal or active conformation. They found that both protein forms bound modest amounts of JNK3 in vitro; however the protein locked into the active conformation exhibited a much greater increase in JNK3 binding in the presence of IP6 than did the protein locked in the basal conformation.




FIGURE 6. Diagrammatic representations of the structure of arrestin-3 highlighting the switch regions (bold). The spheres designate the locations of identical amino acids in each case. Gray - basal arrestin-3. Blue - arrestin-3 in complex with IP6. Gold - preactivated arrestin-3. The structures highlight the rotation of the two domains and the resulting relative movement of the switch regions.  Creative Commons Attribution 4.0 International License from Q. Chen et al. Nat. Commun., 2017, 8:1426 DOI: 10.1038/s41467-017-01218-8).



Together the results strongly support the conclusion that IP6 activates both the phosphate sensor and the activation sensor of arrestin-3. Phosphate sensor triggering occurs via interaction of the phosphates of IP6 with residues of the N-domain, displacing the C-tail and disrupting the polar core. Activation sensor triggering results from exposure of a hydrophobic environment that promotes formation of an α-helix by the finger loop. The two processes then work synergistically to induce the inter-domain twist. The findings are particularly important because they provide strong support for the role of the finger loop in the activation sensor, and they help to explain how the activation sensor can respond to so many structurally different GPCRs. The data also reveal interesting switch regions in the arrestin-3 protein. Exploring the roles of these regions in IP6-mediated signaling and other arrestin functions will be a fascinating goal for future research.




View Nature Communications article: Structural basis of arrestin-3 activation and signaling







The Vanderbilt Institute of Chemical Biology, 896 Preston Building, Nashville, TN 37232-6304, phone 866.303 VICB (8422), fax 615 936 3884
Vanderbilt University is committed to principles of equal opportunity and affirmative action. Copyright © 2014 by Vanderbilt University Medical Center