Structural basis of arrestin-3 activation and signaling

Chen, Qiuyan; Perry, Nicole A.; Vishnivetskiy, Sergey A.; Berndt, Sandra; Gilbert, Nathaniel C.; Zhuo, Ya; Singh, Prashant K.; Tholen, Jonas; Ohi, Melanie D.; Gurevich, Eugenia V.; Brautigam, Chad A.; Klug, Candice S.; Gurevich, Vsevolod V.; Iverson, T. M.

doi:10.1038/s41467-017-01218-8

Download PDF

Article
Open access
Published: 10 November 2017

Structural basis of arrestin-3 activation and signaling

Qiuyan Chen¹,
Nicole A. Perry¹,
Sergey A. Vishnivetskiy¹,
Sandra Berndt¹,
Nathaniel C. Gilbert¹^nAff9,
Ya Zhuo²,
Prashant K. Singh³,
Jonas Tholen⁴,
Melanie D. Ohi^3,5,6,
Eugenia V. Gurevich¹,
Chad A. Brautigam⁷,
Candice S. Klug²,
Vsevolod V. Gurevich ORCID: orcid.org/0000-0002-3950-5351¹ &
…
T. M. Iverson ORCID: orcid.org/0000-0001-8816-6352^1,5,6,8

Nature Communications volume 8, Article number: 1427 (2017) Cite this article

7015 Accesses
78 Citations
6 Altmetric
Metrics details

Subjects

Abstract

A unique aspect of arrestin-3 is its ability to support both receptor-dependent and receptor-independent signaling. Here, we show that inositol hexakisphosphate (IP₆) is a non-receptor activator of arrestin-3 and report the structure of IP₆-activated arrestin-3 at 2.4-Å resolution. IP₆-activated arrestin-3 exhibits an inter-domain twist and a displaced C-tail, hallmarks of active arrestin. IP₆ binds to the arrestin phosphate sensor, and is stabilized by trimerization. Analysis of the trimerization surface, which is also the receptor-binding surface, suggests a feature called the finger loop as a key region of the activation sensor. We show that finger loop helicity and flexibility may underlie coupling to hundreds of diverse receptors and also promote arrestin-3 activation by IP₆. Importantly, we show that effector-binding sites on arrestins have distinct conformations in the basal and activated states, acting as switch regions. These switch regions may work with the inter-domain twist to initiate and direct arrestin-mediated signaling.

Giant organelle vesicles to uncover intracellular membrane mechanics and plasticity

Article Open access 04 May 2024

The intrinsic substrate specificity of the human tyrosine kinome

Article Open access 08 May 2024

Extensive protein pyrophosphorylation revealed in human cell lines

Article Open access 25 April 2024

Introduction

Arrestins modulate G protein-coupled receptor (GPCR) signaling in two ways. First, arrestins bind to activated, phosphorylated receptor and sterically block G protein coupling, terminating G protein activation¹. Second, receptor-bound non-visual arrestins (arrestin-2 and -3, a.k.a. β-arrestin-1 and -2) initiate G protein-independent signaling² via >100 proteins³. This arrestin-mediated signaling regulates cell proliferation and apoptosis².

GPCR-dependent arrestin signaling occurs in the context of a complex between arrestin and phosphorylated, activated GPCR. In classic biochemical work on rhodopsin and arrestin-1, this arrestin–receptor interaction was found to rely on arrestin “sensors” that detect the phosphorylation and activation of the receptors. The phosphate sensor is formed by a group of positively charged side chains on the concave side of the N-domain that directly bind receptor-attached phosphates and is relatively unselective for the underlying sequence⁴. The activation sensor distinguishes between active and inactive GPCRs. To our knowledge, neither the identity of the activation sensor nor how it recognizes >800 distinct GPCRs have been proposed. These phosphate and activation sensors are believed to act synergistically, as the simultaneous triggering of both sensors elicits the highest-affinity receptor binding⁵.

Structural studies in the context of biochemical work suggest that triggering the phosphate and activation sensors of arrestin promotes a conformational change. Structures of all four vertebrate arrestins in the basal state^{4, 6,7,8} showed that arrestin is a two-domain protein with limited interaction between domains. Two inter-domain interactions have been suggested as key for stabilizing the basal conformation. The first is a group of buried charged side chains known as the “polar core”⁷. The second is a sequence in the C-terminus of arrestin termed the “C-tail” (residues 385–393 of arrestin-3) that binds to the N-domain in a way that both contributes a charge to the polar core and blocks access to the phosphate sensor⁷. Perturbation of either the polar core or the C-tail shifts the arrestin equilibrium to favor activation, suggesting that the inter-domain arrangement changes during activation⁹.

Recently, several strategies to stabilize active arrestin allowed structural characterization. These included the use of: (1) a more easily activated (termed “pre-activated”) splice variant of visual arrestin-1 called p44¹⁰, which is truncated before the C-tail; (2) an antibody to stabilize arrestin-2 bound to a vasopressin receptor phosphopeptide¹¹; (3) the R175E mutant of arrestin-1 which has a destabilized polar core¹²; and (4) arrestin-1 with activating mutations tethered to constitutively active rhodopsin¹³. In the structures of p44, the phosphopeptide-bound arrestin-2, and the receptor-bound arrestin-1, a ~ 20° inter-domain rotation was observed, identifying inter-domain rotation as a hallmark of arrestin activation^{10, 11, 13}. In the R175E variant, a ~ 7.5° inter-domain rotation suggests that this structure represents an activation intermediate¹².

Despite these advances, many fundamental questions in arrestin-mediated signaling remain unanswered. Non-visual arrestins bind >800 distinct GPCRs, yet it is not clear how this broad receptor specificity is achieved. It is also unclear how the phosphate and activation sensors elicit the inter-domain rotation that accompanies arrestin activation, or how the active arrestin conformation initiates signaling. Finally, the arrestin-3 isoform is uniquely able to signal independently of GPCRs^{14,15,16,17,18}. Previous studies identified that receptor-independent arrestin-3 activation biases signaling toward the activation of c-Jun N-terminal Kinase-3 (JNK3)^{14, 16}. However it is neither clear how arrestin-3 is activated in the absence of GPCRs nor are there compelling hypotheses for why receptor-independent arrestin-3 activation preferentially initiates the JNK3 signaling cascade. Indeed, it is not clear how arrestin activated by any input correctly directs signaling toward one out of >100 downstream effectors.

To address these questions, we determined the 2.4 Å resolution crystal structure of arrestin-3 in complex with the non-receptor activator inositol hexakisphosphate (IP₆). We demonstrate that the same sensors critical for detecting the phosphorylated and activated state of GPCRs are triggered during receptor-independent activation, and reveal how the triggering of these sensors elicits inter-domain rearrangements. Moreover, we identify properties of the arrestin activation sensor that allow broad specificity for GPCRs. Finally, we propose a mechanism for arrestin-dependent signal initiation and bias. Collectively, these findings address many outstanding questions in the arrestin field.

Results

Structure of IP₆-activated arrestin-3

GPCR-independent arrestin-3 initiation of the JNK3 cascade is established in the literature^{14,15,16,17,18}. Although a physiological non-receptor activator of arrestin-3 has not been unambiguously identified, inositol phosphates and heparin have been suggested as possibilities¹⁹. To test whether IP₆ could activate arrestin-3, we determined the structure to 2.4 Å resolution (Fig. 1a; Table 1), which shows that the IP₆-arrestin-3 complex is a trimer (Fig. 1a). In the structure, we identified electron density consistent with two bound IP₆ molecules per arrestin (Supplementary Fig. 1a, b); each IP₆ molecule is located between the protomers of the arrestin trimer (Fig. 1a). The structure shows that IP₆ binds to arrestin-3 at the same location as C-tail in basal arrestin such that IP₆ displaces the C-tail (Supplementary Fig. 1c). The IP₆-bound structure exhibits an inter-domain twist of 17.7° with respect to the basal conformation (Fig. 1b). Moreover, the IP₆ binding sites in the N-domain overlap with the binding sites for phosphorylated receptor (Fig. 1c, d). These observations are consistent with this structure representing activated arrestin-3.

Table 1 Summary of crystallographic data collection and refinement statistics^a

Full size table

IP₆ binding supports GPCR-independent activation in cells

As IP₆-bound arrestin-3 adopts an active conformation, we tested whether IP₆ binding enhances effector binding in vitro and induces arrestin-3-dependent JNK3 activation^{14,15,16,17,18} in cells. To identify whether IP₆ enhances JNK3 binding, we compared the capacity of column-immobilized His₆-JNK3 to bind arrestin-3 in the presence or absence of IP₆ (Fig. 1e). Even in the absence of IP₆, arrestin-3 binds to JNK3, as described in^{17, 20}. This may reflect either low-affinity binding of JNK3 to the basal arrestin-3, or the ability of arrestin-3 to spontaneously sample the active conformation in the absence of activator. We observed an approximate doubling of the arrestin-3 binding to JNK3 in the presence of IP₆ (Fig. 1e). This is consistent with IP₆ shifting the conformational equilibrium of arrestin-3 to an active state that binds JNK3.

We next mutagenized the IP₆ binding residues in the N-domain (ΔN_IP6; Fig. 1c) or the C-domain (ΔC_IP6, Supplementary Fig. 1d), then tested the impact on receptor-independent JNK3 activation in HEK293 cells (Fig. 1f). Co-expression of wild-type arrestin-3 with ASK1 (an upstream kinase in the JNK3 cascade²¹) and JNK3 resulted in robust JNK3 phosphorylation. This was significantly attenuated in cells expressing the ΔN_IP6 or ΔC_IP6 arrestin-3 variants, providing strong evidence that IP₆ activates arrestin-3 in cells, and facilitates arrestin-dependent JNK3 activation.

To rule out alternative interpretations for the loss of function in the arrestin-3 variants, we performed two types of assays. First, we tested whether the affinity of arrestin-3 for IP₆ would allow binding at physiological concentrations of IP₆. Second, we ensured that the ΔN_IP6 and ΔC_IP6 variants specifically reduced receptor-independent signaling and had no global signaling defect.

To determine whether arrestin-3 binds IP₆ at the concentrations found in cells, we used microscale thermophoresis (MST) to measure binding affinity. We labeled wild-type arrestin-3 with AlexaFluor C5 maleimide dye, titrated IP₆ into fluorophore-labeled arrestin-3, activated with an infrared laser, and monitored the fluorescence intensity. MST allowed us to obtain binding curves and calculate equilibrium dissociation constants. Wild-type arrestin-3 exhibits two binding affinities for IP₆ (K_D = 57 nM and 90 μM; Supplementary Fig. 1e). As intracellular IP₆ concentrations range between 35 and 105 µM^{22, 23}, these affinities are consistent with IP₆ being able to occupy both binding sites in cells.

To ensure that that the mutants had no global signaling deficit, we evaluated folding and rhodopsin binding of purified wild-type and mutant arrestin-3 in vitro. As the IP₆ binding sites in the arrestin-3 N-domain overlap with the binding sites for phosphorylated receptor (Fig. 1c, d), we anticipated that the ΔN_IP6 variant would exhibit dramatically reduced receptor binding. However, the binding sites in the C-domain are unique to IP₆, and if this variant is correctly folded and functional, the ΔC_IP6 mutations should only have a minor impact on receptor association. As anticipated, the ΔC_IP6 arrestin-3 variant retained significant receptor binding, while ΔN_IP6 arrestin-3 mutant showed a substantial reduction (Supplementary Fig. 1f) but was properly folded. Collectively, these experiments support the importance of the IP₆-binding sites in receptor-independent arrestin-3 signaling.

IP₆ triggers the arrestin-3 phosphate sensor

The IP₆ molecules make extensive contacts with arginine and lysine side chains on the arrestin-3 N-domain (Fig. 1c). Critically, phosphate binding recruits Lys295 of the lariat loop, which contains two out of five charged residues in the polar core (Fig. 2a, b). This interaction likely elicits a conformational change that disrupts the polar core and triggers the phosphate sensor (Fig. 2a, b). A comparison of the IP₆-bound structure to the reported structure of arrestin-2 with a vasopressin receptor phosphopeptide shows that the positions of the IP₆ phosphates closely resemble those of the peptide-attached phosphates (Fig. 1c, d)¹¹. The structure of arrestin-3 in the basal conformation⁶ shows that the C-tail binds to the N-domain, sterically occluding the phosphate binding sites (Supplementary Fig. 1c), but does not recruit Lys295. As a result, the polar core remains intact (Fig. 2b). Thus, IP₆ and polyphosphated receptors bind to the same arrestin elements and trigger the phosphate sensor via the same mechanism by: (1) recruiting phosphate-binding side chains, (2) altering the conformation of the lariat loop, and (3) disrupting the polar core (Fig. 2a, b). This helps explain how relatively disparate phosphorylated species promote arrestin activation and helps clarify how the C-tail stabilizes the basal conformation.

Intriguingly, the Benovic and Brenner groups showed that soaking crystals of arrestin-2 with IP₆ did not trigger the phosphate sensor²⁴. Comparing the binding locations of the IP₆ in arrestin-2 vs. arrestin-3²⁴ shows that one of the IP₆ molecules binds arrestin-2 in a different location (Fig. 1c, Supplementary Fig. 1d, g, h). As a result, IP₆ binding to the arrestin-2 does not recruit Lys294 (equivalent to arrestin-3 Lys295). This leaves the polar core intact despite strict conservation of phosphate binding residues. While it is not clear how IP₆ binding is directed differently in these isoforms, it is consistent with reports that arrestin-2 does not support receptor-independent signaling^{15, 16, 21}.

Engagement of the arrestin-3 activation sensor

We wanted to explore whether the activation sensor is also triggered in IP₆-activated arrestin-3. The activation sensor of arrestins is proposed to distinguish between active and inactive GPCRs⁵, although its identity has not been suggested in the literature. To offer insight into the identity of the activation sensor, we analyzed differences between active and inactive receptors. The most striking structural difference is the accessibility of a hydrophobic pocket on the intracellular side of the protein unique to active GPCRs²⁵. This pocket represents a major site of interaction with both G proteins^{26, 27} and arrestins^{13, 28}. In fact, it contributes 596 Å² out of a total of 1362 Å² of buried surface area in the rhodopsin-arrestin complex¹⁵. Cocrystal structures of opsin with pre-activated arrestin-1 or an arrestin-derived peptide suggest that the hydrophobic pocket binds to an α-helical conformation of the finger loop^{13, 28}. However, there is controversy on this because of the quality of the electron density for this helix in these structures (Supplementary Fig. 2a, b). The α-helix of the finger loop contains a ED(I/L)D motif (residues 68–71 of arrestin-3)^{13, 28} that lacks secondary structure in basal arrestins^{4, 6,7,8} and is not altered in partially activated arrestin²⁹. Helix formation of the finger loop presents several hydrophobic residues to the intracellular hydrophobic pocket of activated receptor^{13, 28}.

Similarly, in the IP₆-mediated trimer the finger loop is presented as an α-helix (Fig. 2c, d). In contrast with the receptor-bound structure¹³, this helix is associated with clear electron density (Supplementary Fig. 2c). In this conformation, hydrophobic residues would be exposed to solvent if the IP₆-activated arrestin-3 were a monomer, but in the context of the trimer, these form Van der Waals (<4 Å) self-contacts around the molecular three-fold axis (Fig. 2c) and are shielded from solvent. This suggests that receptor-independent trimerization and the GPCR hydrophobic pocket stabilize the triggered activation sensor in the same way (Fig. 2c, d).

IP₆-mediated arrestin-3 trimerization and activation

If trimerization stabilizes the triggered activation sensor, then the trimer should be quite stable in the presence of IP₆. It would also be critical for receptor-independent activation of arrestin-3. Importantly, the IP₆-bound arrestin-3 trimer buries 3053 Å² of surface area per protomer, a value indicative of a biologically relevant oligomer³⁰. The trimer does not appear to be influenced by crystal contacts (Supplementary Fig. 3a); however, it does require the inter-domain twist of active arrestin. Indeed, modeling the basal conformation causes the misalignment of the IP₆ binding sites (Supplementary Fig. 3b). This would be predicted to reduce IP₆ affinity and therefore disfavor trimerization. We therefore next explored the stability of the arrestin-3 trimer in solution.

Using analytical ultracentrifugation, we found that the addition of IP₆ to purified arrestin-3 converts the observed molar mass from 43 ± 1 kD (monomer) to 134 ± 5 kD (trimer) (Fig. 3a, b). Size exclusion chromatography (Fig. 3c, Supplementary Table 1) is also consistent with IP₆-dependent trimerization of arrestin-3. We then explored the effect of arrestin-3 concentration on IP₆-dependent trimerization and found that the trimer is stable at the lowest arrestin-3 concentration detectable by our instruments (~1 µM; Supplementary Fig. 3c).

We next assessed whether trimerization contributes to receptor-independent activation in cells using a trimerization deficient arrestin-3 variant. To design a trimerization deficient variant, we identified residues that are surface exposed in basal (monomeric) arrestin-3, buried in the trimer interface, and not associated with known biochemical functions of arrestin. This suggested Cys17 as a candidate side chain for targeting. We used a fully Cys-less variant of arrestin-3 because it has previously been shown to bind receptors normally³¹. This indicates that the Cys-less variant is folded and rules out many alternative interpretations of any results.

We first assessed whether the Cys-less variant lost the ability to form trimers, using size exclusion chromatography. We found that the Cys-less variant does not form stable trimers in response to IP₆, but appears to shift to a molecular weight consistent with a dimer (Supplementary Fig. 3d). Comparing the arrestin-3 trimer to the arrestin-2 oligomers suggests that while the trimer can stabilize the inter-domain twist (Fig. 1a, b), other arrestin oligomers cannot. This makes Cys-less arrestin-3 suitable for measuring how trimerization impacts receptor-independent signaling.

To quantify IP₆-dependent activation of Cys-less arrestin-3 in vitro, we employed double electron electron resonance (DEER) spectroscopy. One of the hallmarks of arrestin activation is the displacement of the C-tail from the N-domain upon the binding of phosphates to the phosphate sensor. We monitored this process via attached spin labels to cysteines replacing Ser13 in the N-domain and Ala392 in the C-tail. At these positions, the spin labels are separated by 22 Å in basal arrestin-3, when the C-tail is bound to the N-domain. Upon arrestin activation and C-tail release, these convert to a wide distribution of longer inter-spin distances. After titration with IP₆, we found that only ~ 30% of the 22 Å distance converted to longer distances (Fig. 3d, Supplementary Fig. 3e), a ~ 70% loss in IP₆-dependent activation. Cys-less arrestin-3 was reported to exhibit nearly 100% release of the C-tail in response to phosphorhodopsin in the same assay³¹, which rules out the possibility that Cys-less arrestin-3 is signaling deficient.

We then compared the ability of wild-type and Cys-less arrestin-3 to mediate receptor-independent JNK3 activation in cells. We observed a 65% reduction in JNK3 activation in cells expressing the Cys-less mutant as compared to wild-type arrestin-3 (Fig. 3e). This correlates with the loss of IP₆-dependent activation measured by DEER.

Our observed trimer contrasts with reported multi angle laser light scattering (MALLS) of IP₆-bound arrestin-3, which were explained by a dimer³². To investigate this discrepancy, we evaluated differences in the experimental design and identified that a protease inhibitor, benzamidine, used in the MALLS studies induces heterogeneous oligomerization (Supplementary Fig. 4a–c, details in the legend), with oligomers having an average molecular weight consistent with a dimer. As benzamidine is not present in cells, it is likely that these oligomers are non-physiological.

The phosphate and activation sensors are intimately linked

With the phosphates of IP₆ apparently triggering the phosphate sensor and trimerization likely triggering the activation sensor, we wanted to explore the relationship between IP₆-binding and trimerization. Since IP₆ molecules mediate trimer formation, we propose that IP₆-binding and trimerization are intertwined. This would suggest that triggering the phosphate and activation sensors are linked during receptor-independent arrestin-3 activation.

To test this, we first evaluated the ability of the ΔN_IP6 and ΔC_IP6 variants of arrestin-3 to form IP₆-dependent trimers. As described above, the IP₆-binding sites are altered by mutagenesis in these variants, and receptor-independent JNK3 activation in cells is compromised (Fig. 1f). Size exclusion chromatography showed that these variants no longer trimerize in the presence of IP₆, instead exhibiting mobility consistent with a monomer (Fig. 4a, Supplementary Table 1).

We next tested how loss of trimerization impacts IP₆ affinity using MST. The Cys-less variant is a valuable tool for this measurement because trimerization is disrupted; however, our experimental design used a maleamide-conjugated fluorophore, which interacts with cysteines. Thus, we re-introduced a cysteine residue to allow labeling. The T222C variant was considered suitable for this purpose because T222 is distal from all characterized functional regions. This mutant is also monodisperse, consistent with correct folding (Supplementary Fig. 3c). When compared to wild-type arrestin-3 (K_D = 57 nM and 90 μM; Supplementary Fig. 1e), the T222C-Cys-less variant shows substantially reduced IP₆ binding affinity (K_D = 1.1 µM and 1.2 mM, Fig. 4b), consistent with trimerization contributing to IP₆ affinity.

Collectively, these data strongly suggest IP₆-binding and trimerization work together to support receptor-independent arrestin-3 activation of JNK3. Because IP₆ appears to directly bind the phosphate sensor, and trimerization appears to stabilize a triggered activation sensor, we conclude that during receptor-independent activation, triggering of these sensors is intimately linked. This mirrors the observed synergy of these two sensors in receptor-dependent signaling⁵.

Arrestin-3 sensors in broad receptor selectivity

There are >800 GPCRs, but only two non-visual arrestins (arrestin-2 and arrestin-3) that recognize the phosphorylated and activated forms of these receptors; arrestin-3 also recognizes non-receptor activators. Intuitively, the mechanism of broad receptor specificity of the phosphate sensor is straightforward, as the N-domain of arrestin can bind receptor-attached phosphates with the correct spacing. In contrast, it is less clear how the activation sensor might recognize this large number of receptors.

If our assignment of the finger loop as a part of the activation sensor is correct, this element must interact with the hydrophobic pocket on >800 GPCRs with limited sequence similarity. We therefore analyzed the finger loop to identify properties that could contribute to broad receptor recognition. An overlay of available arrestin structures suggests that the finger loop is on a flexible tether (Fig. 4c) and can be presented at many angles to an interaction partner. This could allow the finger loop to adapt to the different hydrophobic pockets on receptors. If so, the helicity, hydrophobicity, and flexibility of the finger loop would be predicted to contribute to the broad receptor specificity of the activation sensor. Hence, perturbation of these properties would reduce activation sensor-dependent receptor binding. We therefore designed mutations (Fig. 4d) that introduced (a) a helix-breaking proline (D68P), (b) a flexibility-reducing proline terminating the helix (D70P), or (c) a charge within the hydrophobic region (L69E/R). Modeling based on available structures^{13, 28} suggests that neither introduced proline should directly interact with the receptor.

We measured the binding of these variants to M2 muscarinic and D2 dopamine receptors, which we selected because they are more dependent upon the activation sensor than on receptor-attached phosphates for arrestin binding³³. We used a bioluminescence resonance energy transfer (BRET) assay in COS7 cells cotransfected with luciferase-tagged receptors and Venus-tagged arrestin-3¹⁴ (Fig. 4e, f, Supplementary Fig. 5a, b). Wild-type arrestin-3 shows a robust increase in BRET signal upon agonist stimulation, whereas arrestin-3-KNC (a negative control with 12 substitutions of key receptor-binding residues that binds poorly to receptors¹⁴) does not^{33, 34}. Mutants designed to perturb helix formation or flexibility had substantially reduced binding to both receptors (Fig. 4e, f). Interestingly, charged residues had different effects on binding to the M2 and D2 receptors. This variability may reflect the presence or absence of complementary charges in the inter-helical cavity of different GPCRs, or the strength of other interactions between receptor and arrestin-3.

As the finger loop is also a major contact for IP₆-dependent arrestin-3 trimerization, we tested if finger loop variants affect receptor-independent activation and signaling. Alteration of the finger loop prevented IP₆-dependent trimerization of the D68P variant, reduced trimerization of the L69R and D70P mutations, and had no detectable impact in the L69E variant (Supplementary Fig. 5c, d). JNK3 activation in cells was not affected in the D69R and D70P mutations (Supplementary Fig. 5e). One interpretation of this finding is that IP₆-induced arrestin-3 activation is more dependent upon the phosphate sensor than the activation sensor, a property shared with many GPCRs³³. However, given our data suggesting that triggering the phosphate and activation sensors during receptor-independent activation is closely linked, it is more likely that the finger loop of our variants is presented in a way that allows trimerization.

Switch regions in arrestin-mediated signaling

A major functional consequence of non-visual arrestin activation is the engagement of downstream effectors, yet how activated arrestin supports signaling has never been explained. In the IP₆-activated arrestin-3 structure we identified structural changes in known effector binding elements that may promote signal initiation. Activation-induced conformational changes in these regions have not been previously reported, although inspection of the coordinates for the active arrestin-1 and arrestin-2 indicates that similar changes accompany activation of other arrestins^{10, 11, 13}. We suggest that these conformational changes act as molecular switches functionally analogous to the switch regions of G proteins [reviewed in 35], and therefore term these “arrestin switch” regions. In conjunction with the inter-domain twist, these conformational changes would create effector-binding sites and turn on signaling when arrestin is activated.

Arrestin switch I (aSwI; residues 89–97; Fig. 5a) is strongly conserved in arrestin-3 homologs from other species, but not in the other three arrestin isoforms (Fig. 5a, inset). In arrestin-3, this nine amino acid segment contains seven prolines, including two PPXP motifs that may be recognized by SH3 domains^{35, 36}. Comparison of the basal⁶ and active conformations of aSwI reveals a maximal displacement of 5.8 Å and includes an unusual pair of tandem cis bonds (Pro94-Pro95 and Pro95-Arg96) in the best-fitting model. This rare structural feature is associated with conformational change^{37, 38}. However, the electron density of this region is difficult to interpret, and not all residues could be modeled with confidence (Supplementary Fig. 6a). The flexibility of this switch may be important for adapting to different effectors.

Arrestin switch II (aSwII; residues 176–191; Fig. 5a) contains two distinct parts: IIa (residues 176–183) and IIb (184–191). ASwIIa connects the N- and C-domain, and has been proposed to act as a hinge between domains^{39, 40}. ASwIIa also contains a polyproline motif (PQP residues 181–183) suggested as a non-canonical SH3 binding site in arrestin-2³⁵. The aSwIIa shows a nearly identical conformation in all available active arrestin structures, but variable conformations in basal arrestins (Fig. 5a, inset).

ASwIIb includes the entirety of the first β-strand of the C-domain (residues 184–191) and superimposes in all active arrestin structures. As compared to most arrestins in the basal state, the β-strand of aSwII is register-shifted, which moves it one position in arrestin-2 and -3 and two positions in arrestin-1 (Supplementary Fig. 6b, c), although the active position of aSwIIb is observed in one crystal form of basal arrestin-2^{24, 35} (Supplementary Fig. 6b). Register-shifted β-strands are unusual, but have proposed roles in regulating signaling^{41, 42} and forming protein binding sites⁴³.

Arrestin switch III (aSwIII, residues 307–316; Fig. 5a) is an extension of the lariat loop⁷, which is a part of the polar core that stabilizes the orientation of the two domains in the basal state. ASwIII becomes disordered in both activated arrestin-3 and arrestin-2¹¹, a property commonly associated with the ability to mediate protein-protein interactions^{44, 45}.

Using the same terminology, the arrestin C-tail may be considered aSwIV. In basal arrestins, the C-tail is anchored to the N-domain^{4, 6,7,8, 35}. Receptor binding induces release of the C-tail, which becomes unstructured^{31, 46, 47}. The binding sites for clathrin⁴⁸ and clathrin adapter AP-2⁴⁸ are localized in the C-tail of non-visual arrestins, and both become fully accessible upon C-tail release⁴⁹.

To test the role of the switch regions, we focused on aSwIIb, and used disulfide trapping to stabilize the register-shifted β-strand (Supplementary Fig. 6d–f). We used the Cys-less arrestin-3 as background to avoid unintended disulfide bonds with native cysteines. We introduced a cysteine at position 338 and either position 186 (stabilizing the basal conformation) (Supplementary Fig. 6d), or position 187 (stabilizing the active conformation) (Supplementary Fig. 6e). Electrophoretic mobility suggests that disulfides formed at statistically identical levels in these variants (Supplementary Fig. 6f).

We then assessed the binding of each of these disulfide-stabilized versions of arrestin-3 to JNK3. In the presence of IP₆, column-immobilized JNK3 exhibits increased binding to arrestin-3 with aSwII trapped in the active conformation (Fig. 5b). In the absence of IP₆, the binding is lower for both variants, suggesting that the disulfide-trapped aSwII does not fully shift the arrestin equilibrium to an active state in the absence of IP₆.

Discussion

We used IP₆-bound arrestin-3 to identify the structural requirements for receptor-independent arrestin activation, and to suggest how activation results in arrestin-mediated signaling. These data inform on several aspects of arrestin activation that have been enigmatic: how an arrestin can interact with many disparate receptor and non-receptor activators, adopt an active conformation, then initiate a variety of downstream signaling pathways.

Arrestin-3 is activated by >800 GPCRs plus non-receptor activators, including IP₆. When coupling to receptors, arrestin acts as a coincidence detector that binds with high affinity to phosphorylated and activated receptors, as mediated by two independent sensors⁵. We show that the IP₆ triggers the same two sensors. Indeed, the binding site for IP₆ sterically overlaps with the binding site for receptor-attached phosphates (Fig. 1c, d). Similarly, the surface of the finger loop that interacts with the hydrophobic pocket of receptor also interacts with sister protomers in the trimer^{13, 28} (Fig. 2c, d). Because the same surfaces support receptor-independent trimerization and receptor binding, receptor-dependent and receptor-independent arrestin-3 activation appear to be mutually exclusive, i.e., the trimer does not bind receptors. The use of the same sensors and surfaces for trimerization and receptor binding also suggests that receptor-dependent and receptor-independent arrestin activation occur via a similar mechanism.

Structurally, arrestin activation involves several conformational rearrangements. First, negatively charged phosphates bind to the N-domain, displacing the C-tail and disrupting the polar core¹¹. Our data and previous reports suggest the disruption of the polar core by phosphate binding is a hallmark of a triggered phosphate sensor^{7, 50, 51}. Second, a hydrophobic environment induces the external presentation of hydrophobic residues of the finger loop on the central crest of the arrestin molecule, which folds into a short α-helix^{13, 28}. Our data suggest that this helical finger loop is a hallmark of the triggered activation sensor. Finally, both receptor-dependent and receptor-independent arrestin activation induce an inter-domain twist in IP₆-bound arrestin-3 (Fig. 1b) and all other activated arrestin structures^10,11,12,13.

Molecular mechanisms connecting the phosphate and activation sensors to the domain twist have not been proposed. Our data suggest that each sensor acts via a distinct pathway. Triggering the phosphate sensor likely facilitates the inter-domain rotation in two ways. First, IP₆ phosphates bind Lys295 on the lariat loop (Fig. 2a). This changes the loop position, removing two negative charges from the polar core (Fig. 2a, b). Second, IP₆ directly displaces the C-tail, which contributes Arg393 to the polar core (Fig. 2b). Both changes break the polar core, making domain rotation possible.

Triggering the activation sensor likely promotes inter-domain rotation by a complementary mechanism. An interaction between the finger loop and a hydrophobic environment can induce helix formation. The finger loop is part of the N-domain, but in basal arrestin, the hydrophobic residues are shielded from solvent via an interaction with a loop in the C-domain (residues 243–247, Fig. 6a, b). Activation-associated rearrangement of the finger loop alters this hydrophobic core, so that the β-strands surrounding the finger loop in the basal state shift in a manner promoting the domain rotation (Fig. 6a).

Based on this model, we propose how arrestins have achieved an inter-domain twist in reported structures (Supplementary Table 2). Our analysis suggests that in the IP₆-bound arrestin-3, a physiologically relevant activator triggers both sensors, which appear to be highly interdependent. During receptor-induced activation, the sensors may act more independently, with some GPCRs relying more heavily on one sensor³³. However, the two sensors likely act synergistically, as arrestins exhibit the highest binding when receptor is both activated and phosphorylated⁵.

The comparison of IP₆-activated arrestin-3 and other active arrestin structures suggests a general model explaining how activated arrestins initiate and direct signaling. First, the inter-domain twist and the new positions of switches likely work in conjunction to form effector-binding sites. Because receptor-independent activation apparently biases signaling toward the JNK3 cascade^{15,16,17,18, 21}, IP₆-activated arrestin-3 is the first case where the signaling bias can be correlated with the structure. The inspection of other active arrestin structures suggests an intriguing possible mechanism of signal bias. The inter-domain twist is of different magnitude in available active structures, consistent with the proposed ability of active arrestin to adopt a range of conformations^{9, 52, 53}. Moreover, conformations of the switch regions in our structure differ from those observed in other active arrestins. This ability to adopt a range of active conformations agrees with electron microscopy images of a chimeric β₂-adrenergic receptor with arrestin-2⁵⁴, fluorescence quenching binding studies of rhodopsin-arrestin-1⁵⁵ and NMR spectroscopy of rhodopsin-arrestin-1²⁹, all of which suggest heterogeneity of the receptor-arrestin complex.

Conceivably, signaling toward different effectors can be directed by combining distinct conformations of the arrestin switch regions with different inter-domain rotation angles (Fig. 6c–e, Supplementary Movies 1 and 2). The magnitude of each conformational change likely depends on the identity of the activator. In case of GPCRs, the conformations may be further influenced by the phosphorylation pattern of the receptor^{52, 53, 56, 57}. Structures of arrestin in complex with a range of activators, including receptors with different phosphorylation patterns, are required to test this idea.

Methods

Materials

DNA modifying enzymes were from New England Biolabs (Ipswich, MA). DNA purification kits were from Zymo Research (Irvine, CA). HEK293 and COS7 cells are from ATCC. Cell culture reagents and media were from Mediatech (Manassas, VA) or Invitrogen (Carlsbad, CA). The luciferase substrate coelenterazine-h was from NanoLight (Pinetop, AZ). All other reagents were from Amresco (Solon, OH) or Sigma-Aldrich (St. Louis, MO). The primers used in this study are listed in Supplementary Table 3.

Arrestin-3 expression and purification

Bovine arrestin-3 was cloned into the pTrcHisB vector (Invitrogen; Carlsbad, CA) with the codon for L394 mutagenized to a stop codon (TGA). This created a version of arrestin-3 truncated immediately after the C-tail attachment point (encoding residues 1–393 and deleting residues 394–408). This version of arrestin-3 has activation propensity indistinguishable from wild-type and was used to produce protein for the structure of the basal conformation of arrestin-3⁶. Because it has superior properties after purification, arrestin-3 (1–393) was used for the in vitro experiments in the manuscript, with the exception of a subset of the replicates that tested the effects of benzamidine. Only for these benzamidine experiments, one of the replicates arrestin-3 with a stop codon introduced at position R393. This variant of arrestin (1–392) is more easily activated. Experiments in cells used full-length arrestin-3 (1–408).

The pTrcHisB plasmid containing arrestin-3 (1–393) was transformed into E. coli BL21 Gold and grown in 1 L cultures in LB medium supplemented with 100 mg/L ampicillin. Cells were grown at 30 °C overnight with shaking at 250 r.p.m., then protein expression was induced with the addition of 35 µM IPTG for 4 h. Cells were collected by centrifugation and the pellet was stored at −80 °C.

Arrestin-3 was purified using a modification of a previously described protocol⁵⁸. Briefly, cell pellets were resuspended in buffer containing 50 mM MOPS, pH 7.2, 5 mM EGTA, 2 mM tris-(2-carboxyethyl) phosphine (TCEP), and two protease inhibitor cocktail tablets (Sigma). Cells were disrupted by sonication at 4 °C. The lysate was clarified by centrifugation at 20,800×g (SLA 3000 rotor, RC-5B Plus centrifuge) for 60 min, and arrestin was precipitated by the addition of (NH₄)₂SO₄ to a final concentration 0.32 mg/mL. Precipitated arrestin-3 was collected by centrifugation at 20,800×g (SLA 3000 rotor, RC-5B Plus centrifuge) for 90 min, and dissolved in buffer containing 10 mM MOPS pH 7.2, 2 mM EGTA, and 1 mM TCEP, then centrifuged again at 20,800×g (SLA 3000 rotor, RC-5B Plus centrifuge) for 60 min to remove particulates. The supernatant containing soluble arrestin-3 was applied onto a heparin column and eluted with a linear NaCl gradient. Fractions containing arrestin-3 were identified by SDS-PAGE and Western and combined. The salt concentration of the pooled fractions was adjusted to 100 mM, and the solution was loaded onto a linked HiTrap Q HP (GE healthcare) and HiTrap SP HP (GE healthcare) column. At a NaCl concentration of 100 mM, arrestin-3 flows through the Q column (while most contaminants bind), but binds the SP column. The columns were uncoupled and a linear NaCl gradient was used to elute arrestin-3 from the SP column. The fractions containing arrestin-3 were identified by SDS-PAGE and combined, concentrated with a 30 kDa cutoff concentrator, then further purified using a Superdex 200 increase 10/300 GL column (GE healthcare) equilibrated with 20 mM MOPS pH 7.2, 150 mM NaCl, and 2 mM TCEP. Folding and trimerization were monitored by size exclusion chromatography.

Crystallography

Purified arrestin-3 (residues 1–393) was concentrated to 5 mg/mL and incubated with IP₆ at 1:20 molar ratio for 30 min on ice. Crystals in the hexagonal space group P6₃ were grown using the sitting drop vapor diffusion method by combining 2 µl of arrestin-3-IP₆ and 2 µl of reservoir solution (100 mM Succinate/Phosphate/Glycine pH 8.5 and 25% PEG 1500). Microcrystals appeared within 24 h and were used for seeding. Crystals were harvested after 7 days, cryoprotected in 50% w/v glycerol and cryocooled by plunging into liquid nitrogen.

Diffraction data (Table 1) were collected at the Advanced Photon Source LS-CAT beamline 21-ID-D. Data were processed using HKL2000⁵⁹. The structure was determined by molecular replacement in PHASER⁶⁰ using isolated domains of a pruned version of arrestin-2 (PDB entry 1G4M⁴) as the search model. The best solution was associated with an initial R/R_free of 0.34/0.37 and was improved by model building in COOT⁶¹ and refinement in PHENIX⁶². The myoD conformation of IP₆ was placed into difference electron density manually using the conformation found in PDB entry 4HNW. Figures were prepared in PYMOL.

JNK3 activation measurements in HEK293 and COS7 cells

HA-tagged ASK1 and Flag-tagged JNK3 were co-transfected with either: (1) empty vector; (2) wild-type arrestin-3 (residues 1–408); or the indicated arrestin-3 mutants (residues 1–408). Lipofectamine2000 (Thermo Fisher) was used for transfection. After 48-hours, cells were incubated with phosphatase inhibitors (50 mM NaF and 10 mM Na₃VO₄) in PBS for 15 min at 37 °C and lysed with lysis buffer containing 50 mM Tris pH7.8, 2 mM EDTA, 250 mM NaCl, 10% glycerol, 0.5% NP-50, 20 mM NaF, 1 mM Na₃VO₄, 1 mM phenylmethanesulfonyl fluoride (PMSF) and 2 mM benzamidine. Sonication was used to further lyse the cells (60 Sonic Dismembrator, Fisher Scientific). The whole cell lysate were centrifuged at 10,000×g for 15 min and the supernatant was used for Western analysis. Activation of JNK3 was assessed using a phospho-JNK specific antibody (Cell Signaling #9251, 1:1000 dilution). The expression level of HA-ASK1, Flag-JNK3 and arrestin-3 were assessed by Western analysis using antibodies against the HA tag (Cell Signaling #C29F4, 1:1000 dilution), the Flag tag (Sigma #F3165, 1:500 dilution), arrestin (F4C1)⁶³ or GFP (JL-8, Choltech #632381, 1:2000 dilution), respectively. The results were quantified using VersaDoc and QuantityOne software (Bio-Rad). The uncropped blots of one representative experiment are shown in Supplementary Fig. 8.

DEER distance measurements

Cysteine substitutions of Ser13 and Ala392 were introduced into otherwise Cys-less arrestin-3 (residues 1–393). This variant was purified and spin-labeled with MTSL, as described previously³¹. Double electron electron resonance (DEER) spectroscopy data were collected using a Bruker E580 operating at Q-band and equipped with an EN5107D2 resonator. Samples contained 20% deuterated glycerol as a cryoprotectant and 100 µM protein with varying amounts of IP₆ were run at 80 K following flash freezing in a dry ice and acetone mixture. Acquired raw dipolar evolution data were phase and background corrected, plotted and analyzed for distance distributions in the same way for each data set using the algorithms included in the LongDistances software program⁶⁴ written by C. Altenbach (University of California-Los Angeles, CA). The upper reliable distance limit of 65 Å was determined based on the maximum data collection time (t = 4.5 µs) of the DEER experiments according to the equation d ≈ 5(t/2)1/3⁶⁵ and is reflected in the x-axis of the distance distribution plot. The release of the C-tail causes the distances to spread out and possibly become longer than the observable range of our data.

Microscale thermophoresis

Microscale thermophoresis (MST) was conducted using a NT.115 MST instrument (NanoTemper Technologies GmbH) equipped with green and blue filter sets. All arrestin-3 variants (residues 1–393; 20 µM in MST buffer (20 mM 3-morpholinopropane-1-sulfonic acid (MOPS) pH 7.5, 150 mM NaCl, and 1 mM TCEP) were labeled by adding 200 μL of 20 μM arrestin-3 to 1 μL of 40 mM Alexa Fluor C5 maleimide dye (final dye concentration, 200 µM; Molecular Probes, Eugene, OR) in DMSO and incubating the mixture at room temperature in the dark for 30 min. Free dye was separated from the protein-dye conjugate using a PD10 G25 column (GE Healthcare Bio-Sciences, Piscataway, NJ) equilibrated in MST buffer. Lauryl maltose neopentyl glycol (MNG-3; Anatrace, Maumee, OH) at 1% (w/v) in all solutions prevented the adherence of protein to the plasticware. The labeling stoichiometry was 0.95–2.0 dye molecules per arrestin-3 as estimated by spectrophotometry.

Titrations were usually accomplished by preparing 15 samples of IP₆ in a 1:1 dilution scheme. A sixteenth sample with no IP₆ established the baseline fluorescence or temperature jump. Labeled arrestin-3 (50 nM) and the reaction mixtures were equilibrated for 45 min then loaded into premium coated capillary tubes (NanoTemper). Data were acquired using 40% MST and 50% LED settings. A 5 s pre-IR phase was recorded, followed by a 60 sec phase with the IR laser on and a 5 s post-IR phase. In the case of the wild-type arrestin-3 construct, the pre-IR fluorescence varied strongly (~ 30% difference between the maximum and minimum intensity values) as a function of IP₆ concentration, allowing this signal to be analyzed directly. The temperature-jump methodology was used to analyze data from the ΔN_IP6 or T222C variants⁶⁶. To prevent aggregation of the latter mutant, we included 0.5 mg/mL soybean trypsin inhibitor (Worthington Biochemical) in the MST buffer; this mutant also required an adjustment of the illumination protocol, featuring 75% LED power and an IR-on phase of only 30 s. Data were the average of at least two replicates. Data were analyzed in a version of PALMIST⁶⁶ modified to include two-site binding models (manuscript submitted). MST figures were rendered using GUSSI⁶⁷.

Analytical ultracentrifugation

Purified arrestin-3 (residues 1–393) with or without IP₆ (100 µM) was analyzed in an Optima XLI ultracentrifuge (Beckman Coulter, Brea, CA) equipped with a four-hole An-60 Ti rotor at 142,000×g at 4 °C. Samples were loaded into double-sector cells (path length of 1.2 cm) with charcoal-filled Epon centerpieces and sapphire windows. Sedfit (version 12.0) was used to analyze velocity scans using every scan from a total of between 250–300 scans⁶⁸. Approximate size distributions were determined for a confidence level of p = 0.95, a resolution of n = 300, and sedimentation coefficients between 0.1 and 15 S. The frictional ratio was allowed to float.

BRET measurements of arrestin-3 binding to receptors

Interactions between N-terminally Venus-tagged arrestin-3 (residues 1–408) and C-terminally RLuc8-tagged M2 muscarinic receptor or D2 dopamine receptor were determined by BRET, as described, using the highest (saturating) arrestin-3 concentrations¹⁴. Absolute levels of luminescence were used as the measure of the expression levels of RLuc-tagged receptors, whereas direct fluorescence was used to determine the expression of Venus-tagged arrestins, as described¹⁴. BRET was measured 15 min after the addition of 10 µM M2 agonist carbachol or 10 µM D2 agonist quinpirole, both of which were added along with 5 µM of the luciferase substrate coelenterazine-h. Luminescence and fluorescence were measured using Infinite F500 multimode plate reader (Tecan). Net BRET was calculated as the difference between BRET signal in the presence and absence of agonist to determine the expression of Venus-tagged arrestins, as described¹⁴. BRET was measured 15 min after the addition of 10 µM M2 agonist carbachol or 10 µM D2 agonist quinpirole, both of which were added in conjunction with 5 µM of the luciferase substrate coelenterazine-h. Luminescence and fluorescence were measured using Infinite F500 multimode plate reader (Tecan). Net BRET was calculated as the difference between BRET signal in the presence and absence of agonist.

JNK3 binding of arrestin-3 variants

His tagged JNK3α2 (10 µg) was incubated with Ni-NTA resin for 2 h and wild-type or disulfide-containing arrestin-3 (10 µg) was added to the mixture with and without 100 µM IP₆. The samples were washed with buffer (20 mM MOPS, pH 7.4, 150 mM NaCl, 25 mM imidazole) and eluted with 100 µl of the same buffer containing 250 mM imidazole. The eluate was methanol precipitated, and the JNK3 bound arrestin-3 (residues 1–393) was visualized by both Coomassie blue staining and Western blot, quantified using the QuantityOne software, and the data were analyzed using Prism. For the disulfide-linked bands, measurements were made in quadruplicate and one-way ANOVA was used to compare JNK3 binding to disulfide-containing arrestin-3. The uncropped blots of one representative experiment are shown in Supplementary Fig. 8.

Data availability

The authors declare that all data supporting the findings of this study are available within the article and its Supplementary Information files. Coordinates and structure factors have been deposited in the Protein Data Bank with accession code 5TV1. Raw diffraction data are deposited with SBGrid with the accession code 10.15785/SBGRID/330. The additional data that support the findings of this study are available from the corresponding authors upon request.

References

Wilden, U., Hall, S. W. & Kuhn, H. Phosphodiesterase activation by photoexcited rhodopsin is quenched when rhodopsin is phosphorylated and binds the intrinsic 48-kDa protein of rod outer segments. Proc. Natl Acad. Sci. USA 83, 1174–1178 (1986).
Article ADS CAS PubMed PubMed Central Google Scholar
DeWire, S. M., Ahn, S., Lefkowitz, R. J. & Shenoy, S. K. Beta-arrestins and cell signaling. Annu. Rev. Physiol. 69, 483–510 (2007).
Article CAS PubMed Google Scholar
Xiao, K. et al. Functional specialization of beta-arrestin interactions revealed by proteomic analysis. Proc. Natl Acad. Sci. USA 104, 12011–12016 (2007).
Article ADS CAS PubMed PubMed Central Google Scholar
Han, M., Gurevich, V. V., Vishnivetskiy, S. A., Sigler, P. B. & Schubert, C. Crystal structure of beta-arrestin at 1.9 A: possible mechanism of receptor binding and membrane translocation. Strucutre 9, 869–880 (2001).
Article CAS Google Scholar
Gurevich, V. V. & Benovic, J. L. Visual arrestin interaction with rhodopsin. Sequential multisite binding ensures strict selectivity toward light-activated phosphorylated rhodopsin. J. Biol. Chem. 268, 11628–11638 (1993).
CAS PubMed Google Scholar
Zhan, X., Gimenez, L. E., Gurevich, V. V. & Spiller, B. W. Crystal structure of arrestin-3 reveals the basis of the difference in receptor binding between two non-visual subtypes. J. Mol. Biol. 406, 467–478 (2011).
Article CAS PubMed PubMed Central Google Scholar
Hirsch, J. A., Schubert, C., Gurevich, V. V. & Sigler, P. B. The 2.8 A crystal structure of visual arrestin: a model for arrestin’s regulation. Cell 97, 257–269 (1999).
Article CAS PubMed Google Scholar
Sutton, R. B. et al. Crystal structure of cone arrestin at 2.3 A: evolution of receptor specificity. J. Mol. Biol. 354, 1069–1080 (2005).
Article CAS PubMed Google Scholar
Gurevich, V. V. & Gurevich, E. V. The structural basis of arrestin-mediated regulation of G-protein-coupled receptors. Pharmacol. Ther. 110, 465–502 (2006).
Article CAS PubMed PubMed Central Google Scholar
Kim, Y. J. et al. Crystal structure of pre-activated arrestin p44. Nature 497, 142–146 (2013).
Article ADS CAS PubMed Google Scholar
Shukla, A. K. et al. Structure of active beta-arrestin-1 bound to a G-protein-coupled receptor phosphopeptide. Nature 497, 137–141 (2013).
Article ADS CAS PubMed PubMed Central Google Scholar
Granzin, J., Stadler, A., Cousin, A., Schlesinger, R. & Batra-Safferling, R. Structural evidence for the role of polar core residue Arg175 in arrestin activation. Sci. Rep. 5, 15808 (2015).
Article ADS CAS PubMed PubMed Central Google Scholar
Kang, Y. et al. Crystal structure of rhodopsin bound to arrestin determined by femtosecond X-ray laser. Nature 523, 561–567 (2015).
Article ADS CAS PubMed PubMed Central Google Scholar
Breitman, M. et al. Silent scaffolds: inhibition OF c-Jun N-terminal kinase 3 activity in cell by dominant-negative arrestin-3 mutant. J. Biol. Chem. 287, 19653–19664 (2012).
Article CAS PubMed PubMed Central Google Scholar
McDonald, P. H. et al. Beta-arrestin 2: a receptor-regulated MAPK scaffold for the activation of JNK3. Science 290, 1574–1577 (2000).
Article ADS CAS PubMed Google Scholar
Miller, W. E. & Lefkowitz, R. J. Expanding roles for beta-arrestins as scaffolds and adapters in GPCR signaling and trafficking. Curr. Opin. Cell. Biol. 13, 139–145 (2001).
Article CAS PubMed Google Scholar
Zhan, X., Kaoud, T. S., Kook, S., Dalby, K. N. & Gurevich, V. V. JNK3 enzyme binding to arrestin-3 differentially affects the recruitment of upstream mitogen-activated protein (MAP) kinase kinases. J. Biol. Chem. 288, 28535–28547 (2013).
Article CAS PubMed PubMed Central Google Scholar
Zhan, X., Kook, S., Gurevich, E. V. & Gurevich, V. V. Arrestin-dependent activation of JNK family kinases. Handb. Exp. Pharmacol. 219, 259–280 (2014).
Article CAS PubMed PubMed Central Google Scholar
Palczewski, K., Pulvermuller, A., Buczylko, J. & Hofmann, K. P. Phosphorylated rhodopsin and heparin induce similar conformational changes in arrestin. J. Biol. Chem. 266, 18649–18654 (1991).
CAS PubMed Google Scholar
Zhan, X., Perez, A., Gimenez, L. E., Vishnivetskiy, S. A. & Gurevich, V. V. Arrestin-3 binds the MAP kinase JNK3alpha2 via multiple sites on both domains. Cell. Signal. 26, 766–776 (2014).
Article CAS PubMed PubMed Central Google Scholar
Song, X., Coffa, S., Fu, H. & Gurevich, V. V. How does arrestin assemble MAPKs into a signaling complex? J. Biol. Chem. 284, 685–695 (2009).
Article CAS PubMed PubMed Central Google Scholar
Bunce, C. M. et al. Comparison of the levels of inositol metabolites in transformed haemopoietic cells and their normal counterparts. Biochem. J. 289, 667–673 (1993). (Pt 3).
Article CAS PubMed PubMed Central Google Scholar
Sasakawa, N., Sharif, M. & Hanley, M. R. Metabolism and biological activities of inositol pentakisphosphate and inositol hexakisphosphate. Biochem. Pharmacol. 50, 137–146 (1995).
Article CAS PubMed Google Scholar
Milano, S. K., Kim, Y. M., Stefano, F. P., Benovic, J. L. & Brenner, C. Nonvisual arrestin oligomerization and cellular localization are regulated by inositol hexakisphosphate binding. J. Biol. Chem. 281, 9812–9823 (2006).
Article CAS PubMed Google Scholar
Farrens, D. L., Altenbach, C., Yang, K., Hubbell, W. L. & Khorana, H. G. Requirement of rigid-body motion of transmembrane helices for light activation of rhodopsin. Science 274, 768–770 (1996).
Article ADS CAS PubMed Google Scholar
Scheerer, P. et al. Crystal structure of opsin in its G-protein-interacting conformation. Nature 455, 497–502 (2008).
Article ADS CAS PubMed Google Scholar
Rasmussen, S. G. et al. Crystal structure of the beta2 adrenergic receptor-Gs protein complex. Nature 477, 549–555 (2011).
Article ADS CAS PubMed PubMed Central Google Scholar
Szczepek, M. et al. Crystal structure of a common GPCR-binding interface for G protein and arrestin. Nat. Commun. 5, 4801 (2014).
Article CAS PubMed PubMed Central Google Scholar
Zhuang, T. et al. Involvement of distinct arrestin-1 elements in binding to different functional forms of rhodopsin. Proc. Natl Acad. Sci. USA 110, 942–947 (2013).
Article ADS CAS PubMed Google Scholar
Krissinel, E. B. & Henrick, K. Inference of macromolecular assemblies from crystalline state. J. Mol. Biol. 372, 774–797s (2007).
Article CAS PubMed Google Scholar
Zhuo, Y., Vishnivetskiy, S. A., Zhan, X., Gurevich, V. V. & Klug, C. S. Identification of receptor binding-induced conformational changes in non-visual arrestins. J. Biol. Chem. 289, 20991–21002 (2014).
Article CAS PubMed PubMed Central Google Scholar
Chen, Q. et al. Self-association of arrestin family members. Handb. Exp. Pharmacol. 219, 205–223 (2014).
Article CAS PubMed PubMed Central Google Scholar
Gimenez, L. E. et al. Role of receptor-attached phosphates in binding of visual and non-visual arrestins to G protein-coupled receptors. J. Biol. Chem. 287, 9028–9040 (2012).
Article CAS PubMed PubMed Central Google Scholar
Gimenez, L. E., Babilon, S., Wanka, L., Beck-Sickinger, A. G. & Gurevich, V. V. Mutations in arrestin-3 differentially affect binding to neuropeptide Y receptor subtypes. Cell Signal. 26, 1523–1531 (2014).
Article CAS PubMed PubMed Central Google Scholar
Milano, S. K., Pace, H. C., Kim, Y. M., Brenner, C. & Benovic, J. L. Scaffolding functions of arrestin-2 revealed by crystal structure and mutagenesis. Biochemistry 41, 3321–3328 (2002).
Article CAS PubMed Google Scholar
Luttrell, L. M. et al. Beta-arrestin-dependent formation of beta2 adrenergic receptor-Src protein kinase complexes. Science 283, 655–661 (1999).
Article ADS CAS PubMed Google Scholar
Vey, J. L. et al. Structure and mechanism of ORF36, an amino sugar oxidizing enzyme in everninomicin biosynthesis. Biochemistry 49, 9306–9317 (2010).
Article CAS PubMed PubMed Central Google Scholar
Bruender, N. A., Thoden, J. B. & Holden, H. M. X-ray structure of kijd3, a key enzyme involved in the biosynthesis of D-kijanose. Biochemistry 49, 3517–3524 (2010).
Article CAS PubMed Google Scholar
Vishnivetskiy, S. A., Hirsch, J. A., Velez, M. G., Gurevich, Y. V. & Gurevich, V. V. Transition of arrestin into the active receptor-binding state requires an extended interdomain hinge. J. Biol. Chem. 277, 43961–43967 (2002).
Article CAS PubMed Google Scholar
Hanson, S. M. et al. Arrestin mobilizes signaling proteins to the cytoskeleton and redirects their activity. J. Mol. Biol. 368, 375–387 (2007).
Article CAS PubMed PubMed Central Google Scholar
Johnson, E., Nguyen, P. T., Yeates, T. O. & Rees, D. C. Inward facing conformations of the MetNI methionine ABC transporter: implications for the mechanism of transinhibition. Protein Sci. 21, 84–96 (2012).
Article CAS PubMed Google Scholar
Renault, L., Guibert, B. & Cherfils, J. Structural snapshots of the mechanism and inhibition of a guanine nucleotide exchange factor. Nature 426, 525–530 (2003).
Article ADS CAS PubMed Google Scholar
Westblade, L. F. et al. Structural basis for the bacterial transcription-repair coupling factor/RNA polymerase interaction. Nucleic Acids Res. 38, 8357–8369 (2010).
Article CAS PubMed PubMed Central Google Scholar
Shoemaker, B. A., Portman, J. J. & Wolynes, P. G. Speeding molecular recognition by using the folding funnel: the fly-casting mechanism. Proc. Natl Acad. Sci. USA 97, 8868–8873 (2000).
Article ADS CAS PubMed PubMed Central Google Scholar
Sugase, K., Dyson, H. J. & Wright, P. E. Mechanism of coupled folding and binding of an intrinsically disordered protein. Nature 447, 1021–1025 (2007).
Article ADS CAS PubMed Google Scholar
Hanson, S. M. et al. Differential interaction of spin-labeled arrestin with inactive and active phosphorhodopsin. Proc. Natl Acad. Sci. USA 103, 4900–4905 (2006).
Article ADS CAS PubMed PubMed Central Google Scholar
Vishnivetskiy, S. A. et al. The role of arrestin alpha-helix I in receptor binding. J. Mol. Biol. 395, 42–54 (2010).
Article CAS PubMed Google Scholar
Laporte, S. A. et al. The beta2-adrenergic receptor/betaarrestin complex recruits the clathrin adaptor AP-2 during endocytosis. Proc. Natl Acad. Sci. USA 96, 3712–3717 (1999).
Article ADS CAS PubMed PubMed Central Google Scholar
Kang, D. S. et al. Structure of an arrestin-2-clathrin complex reveals a novel clathrin binding domain that modulates receptor trafficking. J. Biol. Chem. 284, 29860–29872 (2009).
Article CAS PubMed PubMed Central Google Scholar
Vishnivetskiy, S. A. et al. How does arrestin respond to the phosphorylated state of rhodopsin? J. Biol. Chem. 274, 11451–11454 (1999).
Article CAS PubMed Google Scholar
Vishnivetskiy, S. A. et al. An additional phosphate-binding element in arrestin molecule. Implications for the mechanism of arrestin activation. J. Biol. Chem. 275, 41049–41057 (2000).
Article CAS PubMed Google Scholar
Kim, J. et al. Functional antagonism of different G protein-coupled receptor kinases for beta-arrestin-mediated angiotensin II receptor signaling. Proc. Natl Acad. Sci. USA 102, 1442–1447 (2005).
Article ADS CAS PubMed PubMed Central Google Scholar
Ren, X. R. et al. Different G protein-coupled receptor kinases govern G protein and beta-arrestin-mediated signaling of V2 vasopressin receptor. Proc. Natl Acad. Sci. USA 102, 1448–1453 (2005).
Article ADS CAS PubMed PubMed Central Google Scholar
Shukla, A. K. et al. Visualization of arrestin recruitment by a G-protein-coupled receptor. Nature 512, 218–222 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
Sinha, A., Jones Brunette, A. M., Fay, J. F., Schafer, C. T. & Farrens, D. L. Rhodopsin TM6 can interact with two separate and distinct sites on arrestin: evidence for structural plasticity and multiple docking modes in arrestin-rhodopsin binding. Biochemistry 53, 3294–3307 (2014).
Article CAS PubMed PubMed Central Google Scholar
Yang, F. et al. Phospho-selective mechanisms of arrestin conformations and functions revealed by unnatural amino acid incorporation and (19)F-NMR. Nat. Commun. 6, 8202 (2015).
Article PubMed PubMed Central Google Scholar
Nobles, K. N. et al. Distinct phosphorylation sites on the beta(2)-adrenergic receptor establish a barcode that encodes differential functions of beta-arrestin. Sci. Signal. 4, ra51 (2011).
Article CAS PubMed PubMed Central Google Scholar
Vishnivetskiy, S. A., Zhan, X., Chen, Q., Iverson, T. M. & Gurevich, V. V. Arrestin expression in E. coli and purification. Curr. Protoc. Pharmacol. 67, Unit 2 11(2014).
Otwinowski, Z. & Minor, W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276, 307–326 (1997).
Article Google Scholar
McCoy, A. J. et al. Phaser crystallographic software. J. Appl. Crystallogr. 40, 658–674 (2007).
Article CAS PubMed PubMed Central Google Scholar
Emsley, P. & Cowtan, K. Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 (2004).
Article PubMed CAS Google Scholar
Adams, P. D. et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta. Crystallogr. D Biol. Crystallogr. 66, 213–221 (2010).
Article CAS PubMed PubMed Central Google Scholar
Donoso, L. A. et al. S-antigen: preparation and characterization of site-specific monoclonal antibodies. Curr. Eye Res. 9, 343–355 (1990).
Article CAS PubMed Google Scholar
Toledo Warshaviak, D., Khramtsov, V. V., Cascio, D., Altenbach, C. & Hubbell, W. L. Structure and dynamics of an imidazoline nitroxide side chain with strongly hindered internal motion in proteins. J. Magn. Reson. 232, 53–61 (2013).
Article ADS CAS PubMed Google Scholar
Jeschke, G. DEER distance measurements on proteins. Annu. Rev. Phys. Chem. 63, 419–446 (2012).
Article ADS CAS PubMed Google Scholar
Scheuermann, T. H., Padrick, S. B., Gardner, K. H. & Brautigam, C. A. On the acquisition and analysis of microscale thermophoresis data. Anal. Biochem. 496, 79–93 (2016).
Article CAS PubMed Google Scholar
Brautigam, C. A. Calculations and publication-quality illustrations for analytical ultracentrifugation data. Methods Enzymol. 562, 109–133 (2015).
Article PubMed Google Scholar
Schuck, P. Size-distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and lamm equation modeling. Biophys. J. 78, 1606–1619 (2000).
Article ADS CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank Drs B. Spiller, J. York, H. Hamm, C. Sanders, W. Hubbell, C. Goodman, and K. McCulloch for critical reading and insightful discussions, H.E. Xu for an advance on the rhodopsin-arrestin-1 coordinates, S. Prokop, I. Yamakawa, M. Schoenberger, K. McCulloch, X. Zhan. M. Tyska, H. McDonald, and K. Schultz for experimental assistance. This work was supported by NIH grants to VVG (GM077561, GM109955 (these two R01 grants were replaced by R35 GM122491), EY011500) and TMI (GM095633, GM120569), and TMI/VVG (DA043680), a Vanderbilt Discovery Grant (TMI/VVG), and a Vanderbilt Vision Research Center NEI Core Grant (EY008126). Q.C. was supported by the Vanderbilt International Scholars Program. N.A.P. was supported by T32 GM007628 and the American Heart Association (16PRE30180007). NCG was supported by the American Heart Association (13POST16910057). J.T. was supported by a DAAD RISE fellowship. The Vanderbilt crystallization facility is supported by S10 RR026915. A.U.C. is supported by the Vanderbilt Center for Structural Biology. The Advanced Photon Source, a User Facility operated for the U.S. DOE Office of Science, was supported under Contract DE-AC02-06CH11357. LS-CAT Sector 21 is supported by the Michigan Economic Development Corporation and the Michigan Technology Tri-Corridor (085P1000817). DEER instrumentation is supported by S10RR022422 and S10OD011937, use of the National Biomedical EPR Center is supported by P41EB001980.

Author information

Nathaniel C. Gilbert
Present address: Louisiana State University, Baton Rouge, LA, 70803, USA

Authors and Affiliations

Department of Pharmacology, Vanderbilt University, Nashville, TN, 37232, USA
Qiuyan Chen, Nicole A. Perry, Sergey A. Vishnivetskiy, Sandra Berndt, Nathaniel C. Gilbert, Eugenia V. Gurevich, Vsevolod V. Gurevich & T. M. Iverson
Department of Biophysics, Medical College of Wisconsin, Milwaukee, WI, 53226, USA
Ya Zhuo & Candice S. Klug
Department of Cell and Developmental Biology, Vanderbilt University, Nashville, TN, 37232, USA
Prashant K. Singh & Melanie D. Ohi
University of Applied Sciences Emden/Leer, Emden, 26723, Germany
Jonas Tholen
Department of Biochemistry, Vanderbilt University, Nashville, TN, 37232, USA
Melanie D. Ohi & T. M. Iverson
Center for Structural Biology, Vanderbilt University, Nashville, TN, 37232, USA
Melanie D. Ohi & T. M. Iverson
Departments of Biophysics and Microbiology, The University of Texas Southwestern Medical Center, Dallas, TX, 75390, USA
Chad A. Brautigam
Vanderbilt Institute of Chemical Biology, Vanderbilt University, Nashville, TN, 37232, USA
T. M. Iverson

Authors

Qiuyan Chen
View author publications
You can also search for this author in PubMed Google Scholar
Nicole A. Perry
View author publications
You can also search for this author in PubMed Google Scholar
Sergey A. Vishnivetskiy
View author publications
You can also search for this author in PubMed Google Scholar
Sandra Berndt
View author publications
You can also search for this author in PubMed Google Scholar
Nathaniel C. Gilbert
View author publications
You can also search for this author in PubMed Google Scholar
Ya Zhuo
View author publications
You can also search for this author in PubMed Google Scholar
Prashant K. Singh
View author publications
You can also search for this author in PubMed Google Scholar
Jonas Tholen
View author publications
You can also search for this author in PubMed Google Scholar
Melanie D. Ohi
View author publications
You can also search for this author in PubMed Google Scholar
Eugenia V. Gurevich
View author publications
You can also search for this author in PubMed Google Scholar
Chad A. Brautigam
View author publications
You can also search for this author in PubMed Google Scholar
Candice S. Klug
View author publications
You can also search for this author in PubMed Google Scholar
Vsevolod V. Gurevich
View author publications
You can also search for this author in PubMed Google Scholar
T. M. Iverson
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

Q.C. performed protein purification, crystallography, SEC measurements, and IP₆-dependent JNK3 activation. N.C.G. assisted with crystallography. N.A.P. and J.T. performed BRET assays. S.B. performed JNK3 binding assays, E.V.G. performed JNK3 activation assays. Y.Z. and C.S.K. performed DEER experiments. S.A.V. and Q.C. performed rhodopsin binding. C.A.B. performed MST analysis. P.K.S. and M.D.O. performed AUC. V.V.G. and T.M.I. developed the study, performed molecular biology and protein purification, and guided the experiments and analysis. Q.C., V.V.G., and T.M.I. wrote the paper with input from the other authors.

Corresponding authors

Correspondence to Vsevolod V. Gurevich or T. M. Iverson.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Additional information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Supplementary Information

Description of Additional Supplementary Files

Supplementary Movie 1

Supplementary Movie 2

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Chen, Q., Perry, N.A., Vishnivetskiy, S.A. et al. Structural basis of arrestin-3 activation and signaling. Nat Commun 8, 1427 (2017). https://doi.org/10.1038/s41467-017-01218-8

Download citation

Received: 04 January 2017
Accepted: 29 August 2017
Published: 10 November 2017
DOI: https://doi.org/10.1038/s41467-017-01218-8

This article is cited by

Disordered-to-ordered transitions in assembly factors allow the complex II catalytic subunit to switch binding partners
- Pankaj Sharma
- Elena Maklashina
- T. M. Iverson
Nature Communications (2024)
Dynamic Nature of Proteins is Critically Important for Their Function: GPCRs and Signal Transducers
- Vsevolod V. Gurevich
- Eugenia V. Gurevich
Applied Magnetic Resonance (2024)
Structure of the neurotensin receptor 1 in complex with β-arrestin 1
- Weijiao Huang
- Matthieu Masureel
- Brian K. Kobilka
Nature (2020)
Many faces of the GPCR-arrestin interaction
- Kiae Kim
- Ka Young Chung
Archives of Pharmacal Research (2020)
Distinct G protein-coupled receptor phosphorylation motifs modulate arrestin affinity and activation and global conformation
- Daniel Mayer
- Fred F. Damberger
- Dmitry B. Veprintsev
Nature Communications (2019)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.