Spinophilin Is Indispensable for the α2B Adrenergic Receptor-Elicited Hypertensive Response

The α2 adrenergic receptor (AR) subtypes are important for blood pressure control. When activated, the α2A subtype elicits a hypotensive response whereas the α2B subtype mediates a hypertensive effect that counteracts the hypotensive response by the α2A subtype. We have previously shown that spinophilin attenuates the α2AAR-dependent hypotensive response; in spinophilin null mice, this response is highly potentiated. In this study, we demonstrate that spinophilin impedes arrestin-dependent phosphorylation and desensitization of the α2BAR subtype by competing against arrestin binding to this receptor subtype. The Del301-303 α2BAR, a human variation that shows impaired phosphorylation and desensitization and is linked to hypertension in certain populations, exhibits preferential interaction with spinophilin over arrestin. Furthermore, Del301-303 α2BAR-induced ERK signaling is quickly desensitized in cells without spinophilin expression, showing a profile similar to that induced by the wild type receptor in these cells. Together, these data suggest a critical role of spinophilin in sustaining α2BAR signaling. Consistent with this notion, our in vivo study reveals that the α2BAR-elicited hypertensive response is diminished in spinophilin deficient mice. In arrestin 3 deficient mice, where the receptor has a stronger binding to spinophilin, the same hypertensive response is enhanced. These data suggest that interaction with spinophilin is indispensable for the α2BAR to elicit the hypertensive response. This is opposite of the negative role of spinophilin in regulating α2AAR-mediated hypotensive response, suggesting that spinophilin regulation of these closely related receptor subtypes can result in distinct functional outcomes in vivo. Thus, spinophilin may represent a useful therapeutic target for treatment of hypertension.


Introduction
The discovery of a plethora of G protein-coupled receptor (GPCR) interactions with non-G protein partners has led to the view that GPCRs do not function in isolation, but in complex protein networks which impact receptor trafficking, signaling and pharmacology. Among GPCR interacting proteins, G protein-coupled receptor kinases (GRKs) and arrestins are considered "universal regulators" and are the most extensively studied (e.g. recently reviewed in [1][2][3][4][5][6]). GRK-catalyzed phosphorylation, which leads to subsequent arrestin binding, represents a major mechanism for homologous desensitization of GPCRs. In addition to terminating G protein coupling, β-arrestins (arrestin 2 and 3) mediate receptor trafficking and scaffold cellular signaling cascades. Our previous studies identified spinophilin as an endogenous antagonist of arrestin, impeding multiple arrestin-mediated regulations of the receptor in cultured cells and in vivo [7]. Spinophilin is a ubiquitously expressed protein, and competes against arrestin for binding to the α 2A adrenergic receptor (AR) third intracellular (3i) loop [8]. Interestingly, multiple α 2A AR-elicited central responses are dampened in arrestin 3 deficient mice, but these responses are potentiated in spinophilin deficient mice where arrestin functions are unimpeded [7,9,10]. This suggests that arrestin 3 promotes, whereas spinophilin attenuates, α 2A AR-dependent processes in vivo.
In the present study, we demonstrated that spinophilin impeded β arrestin-dependent α 2B AR phosphorylation and desensitization by competing against arrestin binding to the receptor. Compared to the wild type (WT) α 2B AR, the Del301-303 α 2B AR exhibited diminished binding affinity to arrestin 3 but enhanced interaction with spinophilin. Moreover, ERK signaling induced by this polymorphic variant was prolonged. Intriguingly, Del301-303 α 2B AR-induced ERK signaling was quickly desensitized in cells without spinophilin expression, showing a profile similar to that induced by the WT receptor in these cells. Together, these data suggest a critical role of spinophilin in sustaining α 2B AR signaling. Furthermore, the α 2B AR-elicited hypertensive response is diminished in spinophilin deficient mice, but the same response is enhanced in arrestin 3 deficient mice where the α 2B AR has a stronger binding to spinophilin. These data strongly suggest that the interaction with spinophilin is indispensable for α 2B AR to elicit the hypertensive response.

Animals
Spinophilin deficient (Sp -/-), arrestin 3 deficient (Arr3 -/-) and their respective corresponding wild type (WT) mice in the same genetic background were obtained and maintained as described previously [7]. Mice were housed in the Association for Assessment and Accreditation of Laboratory Animal Care-accredited Animal Resources Program at the University of Alabama at Birmingham. Experimental procedures are in accordance with Animal Welfare Act and the 1989 amendments to this Act. All protocols were approved by University of Alabama Institutional Animal Care and Use Committee.

Cells
HEK293 and CosM6 cells were originally from American Type Culture Collection (ATCC). Immortalized arrestin 2 and 3 double knockout (Arr2,3 -/-) and the corresponding WT (Arr2,3 +/+ ) mouse embryonic fibroblasts (MEFs) were generated [30] and generously provided by Dr. Lefkowitz's laboratory. Sp -/and the corresponding Sp +/+ MEFs were generated previously as described in [31]. Cells were cultured in 5% CO 2 at 37°C in DMEM (Invitrogen) supplemented with 10% fetal bovine serum (FBS) and 100 units/ml of penicillin and 10 μg/ml of streptomycin (Invitrogen). For mouse embryonic fibroblasts (MEFs), 2mM glutamine was added to the culture medium. Parental cell lines used in this study have no detectable endogenous expression of α 2 AR subtypes.

Transfection
Cells were transfected with indicated plasmid using Lipofectamine 2000 (invitrogen) according to manufacturer's instruction. MEFs were transduced with a retroviral construct encoding HAα 2B AR as described previously [31,34].

ERK1/2 activation
Kinase activation for ERK1/2 was determined by measuring the level of phosphorylated kinase and normalizing the value to the total protein level of ERK1/2. Phospho-and total-ERK1/2 were detected and analyzed by LI-COR odyssey Fc dual-mode western blot system. Mouse anti-phospho-p44/42 MAPK (T202/Y204) (Cell signaling) and donkey anti-mouse IRDye 800CW (LI-COR) were used to detect activation of ERK1/2 by green fluorescence channel. Rabbit anti-p44/42 MAPK (Cell signaling) and goat anti-rabbit IRDye 680RD (LI-COR) were used to detect total ERK1/2 by red fluorescence channel.

Measurement of cardiovascular responses
Measurement of cardiovascular responses was performed as described previously [9]. Mice were anesthetized with 100mg/kg ketamine and 10 mg/kg xylazine. Catheterized left femoral artery was used to measure arterial pressure while the right jugular vein was used for anesthetic administration. Arterial blood pressure was recorded with a pressure transducer (BIOPAC's AcqKnowledge 3.8.2, BioPac, Goleta, CA) continually in conscious and free moving mice 24 hr after the surgery. Twenty minutes after the baseline measurement, 0.1mg/kg UK14,304 was administered through bolus injection into the right jugular vein.

Data analysis
Data are expressed as mean ± SEM. Unpaired Student's t-tests were performed to determine differences between two groups. All plots were generated using GraphPad Prism.

Results
Interactions of the α 2B AR with spinophilin and β-arrestins are mutually exclusive We first confirmed that β-arrestins and spinophilin compete for interaction with the α 2B AR in cells. MEFs express endogenous arrestins and spinophilin, and we readily detected interactions of the α 2B AR with endogenous arrestin 3 (Fig 1A, left) and spinophilin (Fig 1C, left), which were enhanced by epinephrine stimulation. In spinophilin deficient (Sp -/-) MEFs, the epinephrine-promoted interaction between the α 2B AR and arrestin 3 was markedly increased, as compared to that in the corresponding Sp +/+ MEFs prepared from WT mice with the same genetic background (Fig 1A, right, and Fig 1B). Similarly, the association of the α 2B AR with spinophilin in response to epinephrine treatment was significantly enhanced in MEFs with no β-arrestin expression (Arr2,3 -/-), as compared to that in the corresponding Arr2,3 +/+ MEFs (Fig 1C, right, and Fig 1D). We obtained similar results with other α 2 agonists, including clonidine and UK14,304 (data not shown). Additionally, in the absence of arrestin, the basal interaction between α 2B AR and spinophilin was also dramatically enhanced (Fig 1C and 1D). Together, these data demonstrate that interactions of the α 2B AR with spinophilin and β-arrestins are mutually exclusive. Interaction between α 2B AR and the endogenous arrestin 3 was enhanced in Sp -/-MEFs. Sp -/and corresponding WT (Sp +/+ ) MEFs expressing HA-α 2B AR were stimulated with 100μM epinephrine (plus 1 μM propranolol to block βARs) for indicated time points. Cell lysates were subjected to IP assays using an HA antibody. (B) Quantitation of the fold change of arrestin 3 in the IP complex isolated from cells with or without stimulation. Data were mean ± SEM. n = 4 for each condition. *, p<0.05 by unpaired Student's t test, Sp -/vs. Sp +/+ . (C) Interaction between α 2B AR and the endogenous spinophilin was enhanced in Arr2,3 -/-MEFs. Arr2,3 -/and corresponding WT (Arr2,3 +/+ ) MEFs expressing HA-α 2B AR were stimulated with 100μM epinephrine (plus 1μM propranolol to block βARs). Lane C (control) refers to MEFs (Arr2,3 +/+ or Arr2,3 -/-) without HA-α 2B AR overexpression. (D) Quantitation of the fold change of spinophilin in the IP complex isolated from cells with or without stimulation. Data were mean ± SEM. n = 4 for each condition. **, p<0.01, Arr2,3 -/vs. Arr2,3 +/+ .

Spinophilin attenuates α 2B AR phosphorylation through competition against arrestin in cells
We previously demonstrated that stable phosphorylation of the α 2B AR requires β-arrestins [7]. Consistently, when arrestin 3 was overexpressed in CosM6 cells, which have a relatively low level of endogenous arrestins compared to other cell lines such as HEK cells and MEFs [7,36,37], the levels of α 2B AR phosphorylation following epinephrine stimulation were dramatically enhanced compared to the control (Fig 2A and 2B). On the other hand, when spinophilin was overexpressed in HEK293 cells, epinephrine-induced α 2B AR phosphorylation was markedly reduced compared to the control (Fig 2C and 2D). This result is consistent with the notion that spinophilin competes against β-arrestins for binding to the α 2B AR and impedes the arrestin effect in promoting receptor phosphorylation.
However, spinophilin may regulate α 2B AR phosphorylation through mechanisms other than competing against arrestin, given that it contains multiple functional domains in addition to the receptor binding region [38,39]. To address this, we examined whether the receptor binding region of spinophilin (Sp156-444) alone can sufficiently regulate α 2B AR phosphorylation. In HEK293 cells overexpressing Myc-Sp156-444, the levels of α 2B AR phosphorylation in response to agonist stimulation were significantly reduced compared to those in control cells expressing the empty vector (Fig 3A and 3B). This effect on α 2B AR phosphorylation caused by the receptor binding region of spinophilin is comparable to that caused by the full length spinophilin (comparing Fig 2D and Fig 3B). Furthermore, in CosM6 cells (which express a low level of endogenous β-arrestins), overexpression of Myc-Sp156-444 failed to alter agonist-induced α 2B AR phosphorylation (Fig 4C). This suggests that the inhibitory effect of the receptor binding domain of spinophilin on α 2B AR phosphorylation requires a relatively high level of arrestin expression to be detected. Taken together, these data strongly support that spinophilin attenuates α 2B AR phosphorylation through competition against β-arrestins in cells.
Spinophilin counteracts arrestin-dependent desensitization of α 2B ARinduced ERK1/2 activation We next examined the effect of spinophilin and β-arrestins on α 2B AR signaling. In MEFs expressing both spinophilin and arrestins, activation of the α 2B AR by clonidine induced transient ERK1/2 activation, which was desensitized after 20 min of stimulation (Fig 4). In MEFs without β-arrestin expression (Arr2,3 -/-), ERK1/2 signaling was prolonged after a 30-min treatment (Fig 4A and 4B) as compared to that in the corresponding Arr2,3 +/+ MEFs, suggesting that β-arrestins are required for terminating α 2B AR signaling. On the other hand, in cells without spinophilin expression (Sp -/-), α 2B AR-induced ERK1/2 activation was quickly desensitized at the 10 min time point (Fig 4C and 4D). This desensitization of ERK1/2 signaling is much faster than that in the corresponding Sp +/+ MEFs, suggesting an opposing effect of spinophilin on arrestin-dependent desensitization. We obtained similar results with other α 2 agonists, including epinephrine and UK14,304 (data not shown). The activation rate of α 2B AR-induced ERK1/2 signaling seemed not altered in Arr2,3 -/or Sp -/cells. This is different from what we have previously observed for the α 2A AR-induced ERK1/2 activation, which was accelerated in Sp -/cells but slowed in Arr2,3 -/cells [7]. Taken together, our data suggest that reciprocal regulation by spinophilin and β-arrestins can have differential impacts on signaling evoked by the closely related α 2 AR subtypes. The Del301-303 α 2B AR exhibits preferential interaction with spinophilin over arrestin The human variation Del301-303 α 2B AR displays reduced phosphorylation and desensitization following agonist treatment [27]. Given the importance of β-arrestins for α 2B AR phosphorylation and desensitization demonstrated above, we predicted that this receptor would show impaired interaction with β-arrestins. Indeed, while epinephrine markedly enhanced the amount of arrestin 3 associated with the WT α 2B AR, such treatment failed to increase the association of arrestin 3 with the Del301-303 α 2B AR (Fig 5A and 5B). Impaired arrestin binding to the Del301-303 α 2B AR may be a consequence of the decreased phosphorylation level of this receptor. Alternatively, Del301-303 may cause conformational changes that reduce its binding to arrestins independent of receptor phosphorylation. To address this possibility, we examined the ability of Del301-303 α 2B AR to interact with the phosphorylation-insensitive arrestin 3 (Arr3R170E). Replacement of Arg170 with a Glu results in constitutive binding of arrestin 3 to agonist-activated GPCRs even in the absence of receptor phosphorylation [40,41]. Epinephrine treatment significantly enhanced interaction of Arr3R170E with the WT α 2B AR (Fig 5C  and 5D). However, such treatment had no effect on Arr3R170E interaction with the Del301-303 α 2B AR (Fig 5C and 5D). These data suggest that Del301-303 changes the conformation of the receptor leading to a diminished affinity for β-arrestin binding.
Based on the reciprocal effect of spinophilin and arrestin on receptor interaction, we predicted that the Del301-303 α 2B AR would have a higher affinity for spinophilin binding. As expected, interaction of spinophilin with the Del301-303 α 2B AR was significantly increased when compared to that with the WT α 2B AR (Fig 6). Taken together, our data suggest biased interaction of the Del301-303 α 2B AR with spinophilin.
Spinophilin is essential for sustaining the prolonged ERK1/2 signaling elicited by the Del301-303 α 2B AR The above data suggest that the diminished affinity of the Del301-303 α 2B AR for β-arrestin binding likely underlies impaired desensitization of signaling elicited by this receptor variant. Since the Del301-303 α 2B AR showed increased interaction with spinophilin, we further sought to address whether spinophilin binding to this receptor plays a role in sustaining its prolonged signaling. We therefore examined the kinetics of ERK1/2 signaling elicited by the WT or Del301-303 α 2B AR in Sp -/and the corresponding Sp +/+ MEFs. Consistent with the reduced desensitization of the Del301-303 α 2B AR signaling reported previously [27], ERK1/2 activation elicited by the Del301-303 α 2B AR was prolonged when compared to that elicited by the WT α 2B AR in Sp +/+ MEFs (Fig 7A and 7B). Strikingly, in Sp -/-MEF, we failed to observe any difference in ERK1/2 activation kinetics induced by the Del301-303 versus the WT α 2B AR (Fig 7C  and 7D). In both cases, ERK1/2 signaling was quickly desensitized (Fig 7C and 7D). These data suggest that spinophilin is required for sustaining the prolonged ERK1/2 signaling elicited by the Del301-303 α 2B AR.  The α 2B AR-elicited hypertensive response is enhanced in arrestin 3 null mice but diminished in spinophilin null mice Given the reciprocal effects of spinophilin and β-arrestins in regulating α 2B AR phosphorylation and desensitization, we sought to further determine the role of these proteins in regulating the α 2B AR-elicted hypertensive response. We first compared the level of increase in blood pressure (a response known to be elicited by the α 2B AR [18]) following administration of an α 2 agonist, UK14,304, in Arr3 -/and the corresponding WT (Arr3 +/+ ) mice in the same genetic background. In Arr3 -/mice, UK14,304 induced a significantly higher increase in the mean arterial pressure (MAP) compared to that in Arr3 +/+ mice (Fig 7A and 7B). Furthermore, this α 2B ARelicited hypertensive response appeared to last for a longer time in Arr3 -/mice compared to WT mice (Fig 7A), and the peak area under the hypertensive curve (AUC) was more than twice of that in WT mice (Fig 7C). These data suggest that the α 2B AR-elicited hypertensive response is enhanced and prolonged in the absence of arrestin 3, supporting an in vivo role of arrestin 3 in desensitizing this effect. Regulation of α 2B AR Responsiveness by Spinophilin We then compared the level of increase in blood pressure induced by UK14,304 in Sp -/and the corresponding WT (Sp +/+ ) mice in the same genetic background. UK14,304-induced change in MAP in Sp +/+ mice were somewhat higher than that in Arr3 +/+ mice (7.4 ± 2.5 vs 5.4 ± 2.9, p = 0.28), likely due to the difference in genetic background between these lines. In Sp -/mice, the α 2B AR-elicited hypertensive response was greatly diminished (Fig 8A and 8B) and the AUC in these mice was dramatically reduced as compared to those in Sp +/+ mice ( Fig  8C). These data strongly suggest that the in vivo α 2B AR responsiveness requires the presence of Agonist treatment failed to promote interaction of the Del301-303 α 2B AR with arrestin 3. Cells co-expressing GFP-tagged arrestin3 (GFP-Arr3) with HA-tagged WT α 2B AR or Del301-303 α 2B AR were stimulated with 100μM epinephrine (plus 1μM propranolol to block βARs), and the interaction between arrestin and either WT or Del301-303 α 2B AR was examined by co-IP assays. (B) Quantitation of the agonist-induced fold change of GFP-Arr3 in the IP complex with the WT or Del301-303 α 2B AR. n = 3-4 for each condition. **, p<0.01, WT vs. Del301-303. (C) Del 301-303 α 2B AR was unable to interact with constitutively active mutant arrestin3 R170E following agonist stimulation. Cells co-expressing GFP-Arr3R170E together with WT or Del301-303 α 2B AR were stimulated with 100μM epinephrine (plus 1μM propranolol). (D) Quantitation of the agonist-induced fold change of GFP-Arr3R170E in the IP complex with the WT or Del301-303 α 2B AR. n = 3-4 for each condition. *, p<0.05, WT vs. Del301-303.

Discussion
Using the α 2A AR subtype as a model, we previously identified competition between spinophilin and β-arrestins for interaction with the 3i loop of the receptor [8]. In this study, we confirmed the mutually exclusive binding of these proteins to the 3i loop of the α 2B AR subtype   (Fig 1), and validated that β-arrestins promoted, whereas spinophilin attenuated, agonistinduced α 2B AR phosphorylation (Fig 2). In addition, we demonstrated that the receptorbinding domain of spinophilin sufficiently attenuated α 2B AR phosphorylation (Fig 3), which further supports the notion that spinophilin regulates receptor phosphorylation through competition against arrestin binding to the receptor.
Consistent with a role of β-arrestins in desensitizing receptor signaling, we found that α 2B AR-mediated ERK1/2 activation was prolonged in cells without β-arrestins expression (Fig 4A and 4B). In cells without spinophilin expression, this signaling process was quickly desensitized (Fig 4C and 4D), presumably due to enhanced arrestin binding to the receptor in these cells. In our previous studies, activation of ERK1/2 signaling by the α 2A subtype was accelerated in spinophilin deficient cells but slowed in arrestin deficient cells [7]. However, we did not observe these changes for α 2B AR-induced ERK1/2 activation, suggesting that the signaling profiles induced by closely related α 2 AR subtypes are differentially regulated by spinophilin and β-arrestins.
The α 2B AR mediates the hypertensive response to α 2 ligands [18]. The human Del301-303 α 2B AR variant, which exhibits reduced phosphorylation and desensitization profiles in vitro [27] (also Fig 7A and 7B) and in vivo [42], has been associated with early onset hypertension in a Swedish population [23,24]. We found that this variant showed impaired interaction with arrestin 3 (Fig 5A). In particular, the Del301-303 α 2B AR failed to interact with the constitutive form of arrestin 3 that is insensitive to receptor phosphorylation (Fig 5B), suggesting that the conformational change caused by Del301-303 alters the intrinsic affinity of the receptor to arrestins. By contrast, the Del301-303 α 2B AR showed a much stronger interaction with spinophilin (Fig 6). Strikingly, in cells without spinophilin expression, ERK1/2 activation induced by the Del301-303 α 2B AR was quickly desensitized with a profile similar to that induced by the WT receptor (Fig 7C and 7D). These data suggest that spinophilin interaction is essential for maintaining the prolonged signaling profile induced by this receptor variant, and further indicate that spinophilin may represent an attractive target in manipulating functions of this polymorphic variant.
We previously found that multiple in vivo responses elicited by the α 2A AR subtype are potentiated in spinophilin deficient mice, but dampened in arrestin 3 deficient mice where spinophilin binding to the receptor is enhanced [7,9,10]. Particularly, we have found that spinophilin attenuates the α 2A AR-dependent hypotensive response; in spinophilin null mice, this response is highly potentiated [9]. This is the opposite of what we have observed for responses elicited by the α 2B AR subtype in the current study. In spinophilin null mice, the α 2B AR-elicited hypertensive response was nearly abolished (Fig 9), whereas in arrestin 3 deficient mice, this response was enhanced and prolonged (Fig 8). Our current data suggest that interaction with spinophilin is indispensable for α 2B AR to elicit the hypertensive response. Collectively, our previous and current studies suggest that spinophilin regulation of the closely related α 2 AR subtypes can result in distinct functional outcomes in vivo. Diminished regulation by spinophilin enhances the hypotensive effect elicited by the α 2A subtype [9] while reducing the counteracting hypertensive effect by the α 2B subtype (Fig 9). Hence, reducing spinophilin binding to the α 2 AR subtypes may represent a useful therapeutic strategy for treatment of hypertension. This strategy may be particularly beneficial to the hypertensive population with the spinophilinbiased variation of α 2B AR, Del301-303, given the essential role of spinophilin in sustaining signaling by this receptor variant (Fig 7C and 7D).
We tested multiple α 2 ligands in this study. Although all these ligands promoted binding of both arrestin and spinophilin and gave similar results in our experimental readouts, it should be noted that these ligands likely exhibit different biases for the arrestin pathway, as described for the α 2C AR subtype previously [43]. Further investigation is needed to quantitatively compare the functional selectivity of these ligands to the α 2B AR, including arrestin and spinophilin recruitment. Our studies suggest that ligands that lead to stronger biased interaction of the α 2 AR subtypes with β-arrestins over spinophilin would be more beneficial for hypertension treatment than traditional ligands that can enhance binding of both proteins to the receptor. Identifying such ligands may represent a new direction of therapeutic development for treatment of hypertension.