Distinct phosphorylation sites on the ghrelin receptor, GHSR1a, establish a code that determines the functions of ß-arrestins

The growth hormone secretagogue receptor, GHSR1a, mediates the biological activities of ghrelin, which includes the secretion of growth hormone, as well as the stimulation of appetite, food intake and maintenance of energy homeostasis. Mapping phosphorylation sites on GHSR1a and knowledge of how these sites control specific functional consequences unlocks new strategies for the development of therapeutic agents targeting individual functions. Herein, we have identified the phosphorylation of different sets of sites within GHSR1a which engender distinct functionality of ß-arrestins. More specifically, the Ser362, Ser363 and Thr366 residues at the carboxyl-terminal tail were primarily responsible for ß-arrestin 1 and 2 binding, internalization and ß-arrestin-mediated proliferation and adipogenesis. The Thr350 and Ser349 are not necessary for ß-arrestin recruitment, but are involved in the stabilization of the GHSR1a-ß-arrestin complex in a manner that determines the ultimate cellular consequences of ß-arrestin signaling. We further demonstrated that the mitogenic and adipogenic effect of ghrelin were mainly dependent on the ß-arrestin bound to the phosphorylated GHSR1a. In contrast, the ghrelin function on GH secretion was entirely mediated by G protein signaling. Our data is consistent with the hypothesis that the phosphorylation pattern on the C terminus of GHSR1a determines the signaling and physiological output.


Results
Identification of phosphorylation sites in GHSR1a by mass spectrometry. Following stimulation with ghrelin (100 nM, 5 min), phosphorylation of GHSR1a was enhanced as monitored by increased incorporation of 32 P into a protein with an apparent molecular mass ∼ 100 KDa (~2.8 ± 0.2-fold; Fig. 1A). To identify the precise phosphorylation sites, a mass spectrometry-based proteomics study of tryptic peptides generated from the isolated GHSR1a was conducted ( Fig. 1B-E). These studies revealed 3 serine (Ser 349 , Ser 362 and Ser 363 and 2 threonine (Thr 350 and Thr 366 ) phospho-acceptor sites at the C-terminal tail. The tryptic peptides generated by digestion of GHSR1a included peptides originating from the third intracellular loop, however, there was no indication that any of the peptides from this region were phosphorylated. HEK cells stably expressing GHSR1a were generated in which these phosphorylation sites were mutated to Ala residues. In a double mutant of GHSR1a, designated GHSR1a-DM, in which Thr 350 and Ser 349 were mutated to Ala, phosphorylation in response to ghrelin was reduced by 57± 3% (Fig. 2B). Triple mutation of GHSR1a, designated GHSR1a-TM, in which Ser 362 , Ser 363 and Thr 366 were mutated to Ala, phosphorylation was reduced by 58± 2% (Fig. 2B). A further mutant was generated in which all of the residues identified by mass spectrometry to be phosphorylated were substituted by Ala (Ser 349 , Ser 362 , Ser 363 , Thr 350 and Thr 366 ), and was designated GHSR1a-Total. The phosphorylation status of GHSR1a-Total was significantly less than that of either GHSR1a-DM or GHSR1a-TM indicating that the sites of ghrelin-regulated phosphorylation in the GHSR1a were mainly Ser 349 , Ser 362 , Ser 363 Thr 350 and Thr 366 .
Phosphorylation of the C-terminal tail of the GHSR1a regulates receptor endocytosis and ß-arrestin 1 and 2 recruitment. To determine the importance of the receptor C-termini phosphorylation in directing specific signaling events, we first compared ghrelin-induced receptor endocytosis (100 nM) by confocal microscopy in HEK 293 cells expressing GHSR1a-WT or GHSR1a mutants. In the resting cells, fluorescence associated with the receptor was predominantly localised to the plasma membrane (Fig. 2C). A slight fluorescence was also associated with the Golgi apparatus, even after treatment with cycloheximide. After exposure to ghrelin for 20 and 60 minutes, the GHSR1a-WT-associated fluorescence almost completely disappeared from the plasma membrane to become redistributed to a population of intracellular vesicles distributed throughout the cytoplasm (Fig. 2C). In cells expressing GHSR1a-DM, the receptor was primarily distributed throughout the cytoplasm after 20 and 60 minutes of agonist treatment although the population of intracellular vesicles appeared to be reduced (Fig. 2C). By contrast, very little redistribution of the fluorescent labeling could be observed in cells expressing GHSR1a-TM or GHSR1a-Total after 20 minutes and even after 60 minutes of agonist treatment (Fig. 2C). In order to determine whether this change in the patterns of endocytosis displayed by the mutant Scientific RepoRts | 6:22495 | DOI: 10.1038/srep22495 receptors was due to differences in their ability to interact with ß-arrestins, cells were transiently co-transfected with RFP-tagged ß-arrestin 1 or m-cherry-tagged ß-arrestin 2. As shown in Fig. 3A,B, the receptors (shown in green) were located at the cell surface, whereas RFP-ß-arrestin 1 or m-cherry-ß-arrestin 2 was uniformly distributed in the cytoplasm in unstimulated cells (shown in red). In response to 20-minute stimulation with ghrelin (100 nM), GHSR1a-WT appeared to colocalize with both ß-arrestin 1 and 2 in endocytic vesicles (shown in yellow; Fig. 3A,B). This colocalization is consistent with the assembly of a protein complex containing ß-arrestin and the receptor and appeared to be more robust after 60 minutes of agonist treatment. Similarly, stimulation of GHSR1a-DM induced colocalization of the receptor with both ß-arrestins (Fig. 3A,C). However, this colocalization was rather more evenly distributed in the cytoplasm in a diffuse granular pattern, with no apparent enhancement of localization in endocytic vesicles. In contrast, in the case of GHSR1a-TM and GHSR1a-Total, the receptors remained localized at plasma membrane whilst ß-arrestin remained evenly distributed in the cytoplasm after agonist stimulation (Fig. 3A,C). An examination of two of the coefficients used to quantify the degree of colocalization between fluorophores, the Pearson correlation coefficient (PCC) and the Mander's overlap coefficient (MOC), supported the assembly of a protein complex containing ß-arrestin and GHSR1a or GHSR1a-DM, and ruled out a complex for GHSR1a-TM or GHSR1a-Total (Fig. 3B,D). To examine in more detail the contributions of agonist dependent phosphorylation in the C-terminal tail of GHSR1a to recruitment of ß-arrestin 1 and 2, BRET assays were performed which enabled association of an eYFP tagged GHSR1a and an Rluc-ß-arrestin to be measured in real time in living cells (Fig. 4A,B, respectively). GHSR1a-WT recruited both ß-arrestin 1 and 2 in an agonist dependent manner. Concentration response curves for ß-arrestin 1 recruitment to each receptor were determined, which revealed that the GHSR1a-DM and GHSR1a-TM showed a subtle but significant decrease in Figure 1. Mass spectrometry identifies five distinct sites of phosphorylation in the GHSR1a. HEK 293 cells transiently expressing C-terminally EGFP-tagged GHSR1a were either labeled with 32 P (A) or used to immunoprecipitate and then digest the receptor for analysis using mass spectrometry (C-E). In the 32 P labeling studies, cells were treated with the agonist ghrelin (100 nM) or vehicle for 5 min prior to sample preparation. (A) Left panel, autoradiograph and loading control (EGFP immunoblot) is shown. Right panel, levels of 32 P were quantified by densitometry, normalized to GHSR1a-EGFP, and expressed as fold increase relative to the control cells. Immunoblot is representative of five independent experiments. The data are expressed as the mean ± SE (*p < 0.05). (B) Amino acid sequence of the GHSR1a indicating in red the amino acids identified as being phosphorylated. (C-E) representative mass spectra and associated fragmentation tables are shown for the five phosphorylated amino acid residues, three serine (Ser 349 , Ser 362 and S 363 ) and 2 threonine (Thr 350 and Thr 366 ), which were identified. Green text denotes amino acid residues identified from eGFP. (F) Summary of the overall data set. potency compared to GHSR1a-WT (GHSR1a-WT, pEC 50 = 7.54; GHSR1a-DM, pEC 50 = 7.10; and, GHSR1a-TM, pEC 50 = 7.13; p < 0.05). The GHSR1a-Total mutant showed the greatest reduction in potency, which was substantially decreased, compared to GHSR1a-DM (pEC 50 = 6.01). In addition, there was a significant reduction in efficacy (GHSR1a-DM = 81.3%, GHSR1a-TM = 36.8%, GHSR1a-Total = 31.8% of GHSR1a-WT response; p < 0.05). In the case of ß-arrestin 2, a significant decrease in both potency (GHSR1a-WT, pEC 50 = 7.30, GHSR1a-DM, pEC 50 = 7.14; GHSR1a-TM pEC 50 = 6.97; and, GHSR1a-Total pEC 50 = 6.04; p < 0.05) and efficacy was also observed (GHSR1a-DM = 82.4%, GHSR1a-TM = 24.6%, GHSR1a-Total = 18.0% of GHSR1a-WT response; p < 0.05). These observations are consistent with the absence of receptor/ß-arrestin colocalization observed by confocal analysis. The BRET values for ß-arrestin 2 recruitment to GHSR1a-WT were greater than those for ß-arrestin 1, which might reflect differences in the receptor/ß-arrestin conformation resulting in a greater distance between the luciferase and YFP tags or difference in affinity of both ß-arrestins.
To gain further insight into the role of receptor phosphorylation in regulating ß-arrestin recruitment, we monitored BRET as a function of the acceptor/donor ratio (eYFP-receptor/luciferase-ß-arrestin) and determined the acceptor-donor ratio at which half-maximal BRET (BRET 50 ) is observed (Fig. 3E,F). BRET 50 values for GHSR1a-DM/ß-arrestin interactions were higher than those for GHSR1a-WT, suggesting that GHSR1a-WT has a higher relative affinity for both ß-arrestins than GHSR1a-DM. The BRET max value was also decreased, indicating that the nature of the receptor/ß-arrestin interactions were different such that the acceptor and donor tags were in closer proximity with the GHSR1a-WT. These data might suggest that the binding of ß-arrestins to the GHSR1a involves two separate sets of interactions with the phosphorylated carboxyl-terminus of the receptor, one with showing the Ser and Thr residues that were found to be phosphorylated and that were subsequently mutated to alanine to generate GHSR1a-DM, GHSR1a-TM and GHSR1a-Total mutants. The mutated residues are shown in red. (B) Left panel, 32 P labeling studies were performed using HEK 293 cells transiently expressing either the EGFP-tagged GHSR1a-WT or the mutants. Right panel, levels of 32 P were quantified by densitometry, normalized to GHSR1a-EGFP, and expressed as fold increase relative to the control cells expressing the GHSR1a-WT. Immunoblots are representative of five independent experiments. The data are expressed as the mean ± SEM ( *,# p < 0.05). (C) Analysis of GHSR1a endocytosis by confocal microscopy. EGFP-tagged GHSR1a-WT or C-terminal tail Ala mutants were transiently expressed in HEK 293 cells and stimulated with ghrelin (100 nM) for stated times at 37 °C. In absence of ligand, the fluorescent labeling appeared at the cell surface for the GHSR1a-WT and mutants. After a 20 and 60 minutes stimulation with ghrelin, the extent of receptor endocytosis is substantially reduced in the GHSR1a-DM, GHSR1a-TM and GHSR1a-Total mutants compared to the GHSR1a-WT (lower panel). Confocal images are representative of three independent experiments. The data are expressed as the mean ± SEM ( *,# p < 0.05).
Scientific RepoRts | 6:22495 | DOI: 10.1038/srep22495 the phosphorylation sites Ser 362 , Ser 363 and Thr 366 that serve as essential phosphate recognition elements for the ß-arrestin recruitment, and the other with the phosphorylation sites Thr 350 and Ser 349 that might stabilize active conformation of ß-arrestins. ß-arrestin signaling is determined by the phosphorylation of the C-terminal tail of the GHSR1a: ERK1/2 and Akt activation. The role of the two phospho-acceptor regions at the GHSR1a C-terminal tail (Ser 362 , Ser 363 and Thr 366 , or Thr 350 and Ser 349 ) was first evaluated on ghrelin-induced phosphorylation of ERK1/2 [pERK1/2(T202/Y204)] by transient transfection of GHSR1a-WT, GHSR1a-DM and GHSR1a-TM in HEK 293 cells. In GHSR1a-WT cells stimulated with ghrelin (100 nM), pERK1/2(T202/Y204) was resolved into two components dependent, respectively, on G protein or ß-arrestin signaling as we previously described 11 . G protein-dependent activity was rapid, peaking within ~5 min, followed by a ß-arrestin-dependent activation that was slower in onset, peak ~20 min, and sustained (Fig. 5A). This sustained pERK1/2(T202/Y204) signal declined with C-terminal tail mutations GHSR1a-DM and GHSR1a-TM, whilst the fast and transient G protein-dependent activation was maintained. The decrease in the ERK1/2 activation correlated with the decline of the receptor/ß-arrestin signaling complex formation (Fig. 5A).
These mechanistic differences were confirmed in three types of MEF cells: MEF cells from wild-type mice (MEF WT), MEF cells from ß-arrestin1 null mice (ß-arrestin 1 −/− ) and MEF cells from ß-arrestin 2 null mice (ß-arrestin 2 −/− ) (Fig. 5B). In the MEF WT cells expressing GHSR1a, stimulation with ghrelin led to an early peak phase of ERK activation followed by a sustained plateau phase (Fig. 5B). Whilst the peak ERK activation in response to GHSR1a stimulation in MEF ß-arrestin 1 −/− and 2 −/− cells was unchanged compared to WT-MEFs, the sustained plateau phase was lost indicating that the plateau phase of the ERK response to GHSR1a was ß-arrestin-dependent. Consistent with this was the observation that expression of the GHSR1a-TM which recruits ß-arrestin poorly resulted in a significant decrease in the plateau phase compared to the wild type GHSR1a and to the mutant GHSR1a-DM, which showed relatively robust coupling to ß-arrestin (Fig. 5B).
We next sought to determine the functional consequences of phosphorylation/ß-arrestin-dependent ERK1/2 activation by the analysis of the mitogenic activity associated with the GHSR1a mutants in HEK 293 cells. Whereas ghrelin-treated GHSR1a-WT cells (100 nM) incorporated BrdU at a ~2-fold over control, cells expressing GHSR1a-DM or GHSR1a-TM failed to incorporate BrdU (Fig. 5D). Whereas the results obtained with the GHSR1a-TM mutant excludes the role of G protein-dependent signaling on ghrelin-activated proliferation, the lack of GHSR1a-DM mitogenic effect might be related to the stabilization of the active conformation of ß-arrestins exerting spatial control over MAPK events. We therefore examined the subcellular location of pERK1/2(T202/Y204) in cells expressing GHSR1a-WT and GHSR1a-DM EGFP-tagged receptors after ghrelin stimulation (100 nM) by confocal microscopy. In cells expressing GHSR1a-DM, pERK1/2(T202/Y204) was primarily observed in the cytoplasm after activation, whereas in cells expressing GHSR1a-WT, most of the pERK1/2(T202/Y204) translocated into the nucleus (Fig. 5C). Thus, the phospho-acceptor sites at the GHSR1a C-terminal tail appeared to influence the ultimate cellular consequence of ß-arrestin recruitment.
The model for the activation of Akt by ghrelin involves the interplay of an early G i/o protein-dependent pathway and a late pathway mediated by ß-arrestins 12,13 . Certainly, in HEK 293 cells transiently transfected with the GHSR1a-WT ghrelin (100 nM), ghrelin-activated pAkt (S473) (100 nM) was resolved into two components: an initial rapid G protein-dependent activation which peaks within ∼ 10 min, this is followed by a ß-arrestin-dependent activation which is sustained over time (Fig. 6A). This sustained pAkt(S473) signal decreased with C-terminal tail mutations GHSR1a-DM and GHSR1a-TM, correlating with the decline of the receptor/ß-arrestin signaling complex formation, this effect was also observed in MEF ß-arrestin 1 −/− and 2 −/− cells (Fig. 6B). To determine the importance of the GHSR1a C termini in directing specific Akt signaling events, the effects of siRNA-mediated suppression of ß-arrestins were examined on the ghrelin-induced intracellular lipid storage in 3T3-L1 cells. This approach was chosen based on the endogenous GHSR1a expression in either undifferentiated (preadipocytes) or differentiated 3T3-L1 (adipocytes), which would make it difficult to discern the differences among GHSR1a mutants. Thus, 3T3-L1 preadipocyte cells were induced to differentiate into adipocytes using a standard adipogenic induction cocktail of IBMX, DEX and ghrelin for 72 h (early differentiation), followed by suppression of ß-arrestin 1 and 2 with specific siRNAs during terminal differentiation. Oil Red O staining was performed to monitor intracellular ghrelin-induced lipid storage at day 6 after the initiation of differentiation. Efficiency of ß-arrestin 1 and 2 siRNA depletion was confirmed by immunoblot analysis after differentiation (65 ± 5% and 69 ± 2%, respectively). For the ghrelin-induced adipogenesis, depletion of ß-arrestin 1 or 2 caused a substantial inhibition of fat droplet accumulation when compared to siRNA control (61 ± 13% and 73 ± 16%, respectively; Fig. 6C). The ß-arrestin signal complex determines the adipogenic functions of ghrelin highlighting the importance of the phospho-acceptor sites at the GHSR1a C-terminal tail as molecular determinants for the formation of the receptor/ß-arrestin complex and the ultimate signaling outcomes. . Because the lifetime of IP 3 is extremely short, G αq/11 -dependent GHSR1a activation can be followed by monitoring IP 3 degradation products, such as inositol 1-phosphate (IP 1 ), which accumulates in the cell in the presence of lithium chloride. As shown in Fig. 7A, HEK 293 cells transiently expressing the C-terminal tail mutations, GHSR1a-DM, and GHSR1a-TM, revealed similar IP 1 accumulation in relation to GHSR1a-WT in response to ghrelin (100 nM). It is well established that the GHSR-1a stimulates GH release through intracellular Ca 2+ concentration via IP 3 . To further investigate the role of the GHSR1a phosphorylation sites and ß-arrestin signaling, we tested the effect siRNA knockdown of ß-arrestins on the GH release in GC cells. As with 3T3-L1 cells, this approach was selected based on the endogenous GHSR1a expression in GC cells. Transfection of the cells with siRNAs directed against ß-arrestin 1 or 2, which decreased ß-arrestin 1 and 2 expression by 50 ± 1% and 80 ± 3% respectively, did not significantly alter ghrelin-activated GH release compared to cells treated with control siRNA (Fig. 7B). These results demonstrate the contribution of individual G protein and ß-arrestin pathways in the GHSR1a signaling.

Discussion
In the present study we used mass spectrometry-based proteomic approach to map five phosphorylation sites that can be divided into 2 regions, region 1 (Thr 350 and Ser 349 ) and region 2 (Ser 362 , Ser 363 and Thr 366 ). These regions appear to contribute equally to the overall phosphorylation of the receptor. However we found that the region 2 was primarily responsible for ß-arrestin 1 and 2 binding, receptor internalization, ß-arrestin-mediated ERK and Akt activation. In contrast, region 1 appeared to play a more subtle role of stabilizing the interaction between the receptor and ß-arrestins. In this way our data suggest a differential impact of phosphorylation sites on ß-arrestin recruitment and ß-arrestin-dependent signaling and is consistent with a model in which different phosphorylation pattern (barcode) on the GHSR1a can induce distinct ß-arrestin interactions that determine the ultimate cellular consequences of ß-arrestin signaling.
An intriguing observation was the fact that mutation of the phosphorylation sites Thr 350 and Ser 349 , GHSR1a-DM, subtly reduced the potency and efficacy of ß-arrestin binding, which might suggest the implication of these phosphorylation elements in the fine-tune of their interactions with the GHSR1a. This modifying role of region 1 on responses such as internalization appeared to be a unique feature of this study. Phosphorylation within region 2 primarily mediated ß-arrestin recruitment and receptor internalization as revealed by the mutant GHSR1a-TM, analysis of ß-arrestin interactions with the region 1 mutant, GHSR1a-DM, determined that ß-arrestin 1 and 2 are both recruited to this mutant receptor but with a lower apparent affinity. These data suggest that although phosphorylation of Ser 362 , Ser 363 and Thr 366 can promote ß-arrestin recruitment with the GHSR1a, it is phosphorylation of Thr 350 and Ser 349 that is required to stabilize the interaction between the receptor and ß-arrestin. That this stabilization might have functional significance was evidenced by the lack of nuclear translocation of ERK1/2 and cell proliferation observed for the GHSR1a-DM. Importantly, previous studies have determined these GHSR1a functional responses to be dependent on ß-arrestin signaling 11 . Since differences in GPCR-mediated ERK nuclear signaling have previously been linked to the stability of the receptor:ß-arrestin complex [21][22][23] , it is possible to speculate that the phosphorylation barcode (particularly phosphorylation within region 1) on GHSR1a contributes to the stability of the GHSR1a:ß-arrestin complex in a manner that impacts on ERK nuclear signaling 16,24,25 .
How might phosphorylation mediate changes in the affinity and stability of the GHSR1a:ß-arrestin complex is not currently clear; however, recent studies using a ß-arrestin biosensor to detect gross changes in conformation suggest phosphorylation of the different sets of sites engenders the distinct functionality of ß-arrestin by inducing different conformations of the receptor-bound ß-arrestin 16,24 . These findings are consistent with a model where the patterning of receptor phosphorylation sites establishes a code that determines the conformation of the bound ß-arrestins and subsequently its functional capabilities. Thus, differences in ß-arrestin binding in response to recruitment to GHSR1a versus GHSR1a-DM might underlie the differences in signaling we observed.
The interaction between ß-arrestin and activated GPCRs is proposed to involve a biphasic mechanism 26-28 . The first step comprises an interaction between the phosphorylated C-terminal tail of the receptor and the N-terminal domain of arrestin followed by the insertion of the finger loop of ß-arrestin within the receptor core that engages additional binding sites resulting in a longitudinal arrangement on the receptor. Following this model, the phospho-acceptor sites operate in concert with structural elements within the transmembrane core of the receptor. Indeed, recent studies demonstrated that mutations of the contiguous conserved amino acids Pro 148 and Leu 149 in the GHSR1a intracellular second loop generate receptors with a strong bias to G protein and ß-arrestin respectively, supporting a role for conformation-dependent signaling bias in the wild-type receptor 29 . Thus, the nature of the active conformation of ß-arrestin and the signaling outcome is determined by the complex ensemble of the GHSR1a phosphorylation sites within the C-terminal tail in combination with structural elements within intracellular loops that confer the functional selectivity.
An interesting aspect is the functional selectivity associated with the differential GHSR1a-stimulated G protein-and ß-arrestin-mediated signaling to control particular cellular response. Our findings from ß-arrestin knockdown indicates the GH-releasing activity results from G protein signaling with no implication of ß-arrestin signaling because this action occurs upon activation of ß-arrestin-impaired GHSR1a. Following the mutation of phospho-acceptor sites, the proliferative effect of ghrelin was impaired implicating ß-arrestin-mediated ERK1/2 pathway in this response. Furthermore, the ß-arrestin-scaffolded complex positively determines Akt activity and adipocyte differentiation. We have previously demonstrated the importance of these scaffolding proteins during ghrelin-induced adipogenesis in 3T3-L1 cells, which determined the expression levels of master regulators of early, the CCAAT/enhancer-binding protein ß (C/EBPß) and the CCAAT/enhancer-binding protein δ (C/EBPδ ), and terminal, the peroxisome proliferator-activated receptor (PPARγ ) and the CCAAT/enhancer-binding protein α (C/EBPα ), adipogenesis 14 . Initially, these findings imply the existence of independent G protein-and ß-arrestin-mediated pathways. However, this does not appear to be the case since ß-arrestin signaling is dependent on the G protein activation 11,12 . This fact suggests that certain key components of the G protein-dependent signaling pathways are required to determine ß-arrestin recruitment and signaling. In fact, our previous works demonstrated that ghrelin leads to the activation of Akt through an early G i/o -protein-dependent pathway and a late pathway mediated by ß-arrestins [12][13][14] . The starting point is the G i/o -protein dependent PI3K activation that leads to the membrane recruitment of Akt, which becomes tyrosine phosphorylated by c-Src with the subsequent phosphorylation in both the activation loop within the kinase domain [A-loop (T308)] and the hydrophobic motif in the C-terminal region [HM (S473)] by PDK1 and mTORC2, respectively. Once the receptor is activated, a second signaling pathway is mediated by ß-arrestin 1 and 2, involving the recruitment of ß-arrestins, c-Src and Akt. This ß-arrestin-scaffolded complex leads to full activation of Akt. In agreement with these results, assays performed in 3T3-L1 preadipocyte cells indicate that ß-arrestins and c-Src are implicated in the activation of Akt in response to ghrelin through the GHSR1a 12,13 . These results support the notion that key components of the G protein-dependent signaling pathways, such as Gi/o-protein dependent Src activation, trigger signaling pathways mediated by ß-arrestins, i.e. full activation of Akt. Additionally, the impact of GRKs or second messenger kinases (i.e. PKCs) on the receptor phosphorylation and consequent ß-arrestin recruitment might be proposed as the missing link between G protein and ß-arrestins. Our data highlight the possibility that the functions of ß-arrestins may be pre-specified by GRK/PKC-receptor interaction. This is consistent with previous works on other GPCRs,, which demonstrate a requirement for GRKs to activate specific transducers as well as to affect transducer functionality in a selective manner 16,30 . Thus, the patterning of receptor phosphorylation sites, barcode, could engender subtle differences in ß-arrestin/receptor interactions that lead to divergent ß-arrestin-dependent signaling events.
One of the possible physiological implications of our data is that the signaling outcome of GHSR1a might be determined in a cell type specific manner by differential phosphorylation. For the M3-muscarinic receptor, for example, we have demonstrated that the pattern of receptor phosphorylation varies between different cell types in a manner that might contribute to cell type specific signaling [17][18][19][20] . The fact that we identify two distinct regions of phosphorylation on GHSR1a that contribute differentially to arrestin-dependent signaling means that differential cell type specific phosphorylation might result in different signaling outcomes. We are currently testing this hypothesis by determining if GHSR1a is differentially phosphorylated in different cell types. Ultimately, the fact that GHSR1a can direct functionality via different pathways unveils a tremendous potential for new approaches in developing therapeutics at this receptor particularly taking into account the physiological and pathophysiological effects in both neural and peripheral tissue.

Experimental Procedures
Materials. Human ghrelin was obtained from California Peptides (CA, US). Anti-Akt, anti-ERK1/2, anti-pAkt(S473) and anti-pERK1/2(T202/Y204) antibodies were from Cell Signaling Technology (MA, US). Anti-ßarrestin 1 antibody was obtained from BD Biosciences (CA, US). Anti-EGFP and anti-ß-arrestin 2 antibodies were from Abcam (Cambridge, UK). Secondary antibodies and enhanced chemiluminescence detection system were from GE-Amersham (Buckinghamshire, UK). Radioisotope [ 32 P]-orthophosphate (specific activity 8500-9120 Ci/mmol), were from PerkinElmer Life Sciences. Unless otherwise stated, all biochemical and reagents were from Sigma (Mo, US). Plasmids, mutagenesis and cell transfection. The wild-type GHSR1a fused at its C terminus to enhanced green fluorescent protein (EGFP) (GHSR1a-WT) in pEFGP-N1 (Clontech, Palo Alto, CA, US) was provided by Prof. Catherine Llorens-Cortes (Institut National de la Sante et de la Recherche Medicale, College de France, Chaire de Medecine Experimentale, Paris, France). EGFP tagged GHSR1a receptor used in the present work showed to have cellular location, endocytosis and G-protein associated signaling similar to the native GHSR1a protein 31 . Mutations to the GHSR1a sequence were incorporated using the QuikChange method (Stratagene, Cheshire, UK), and the identities of all plasmids generated were confirmed through sequencing. The cells were transiently transfected with the GHSR1a-WT and its three mutants [Double mutant (GHSR1a-DM): Thr350Ala, Ser349Ala; Triple mutant (GHSR1a-TM): Ser362Ala, Ser365Ala, Thr366Ala; and, Total mutant (GHSR1a -Total): Thr350Ala, Ser349Ala, Ser362Ala, Ser365Ala, Thr366Ala] using Lipofectamine 2000 (Life Technologies, Invitrogen; Gran Island, NY, US), according to the manufacturer's instructions. The cell lines expressing the GHSR1a-WT and its three mutants were cultured as above described. Pure cell lines were selected on the basis of resistance to geneticin (G418; 500 μg/mL; maintenance antibiotic). Resistant cells were further Scientific RepoRts | 6:22495 | DOI: 10.1038/srep22495 selected using flow-assisted cell sorting, after which they were maintained in complete DMEM, supplemented with G418.
ß-arrestin 1 tagged with red fluorescent protein (RFP; ß-arrestin 1-RFP) was provided by Prof. Robert J. Lefkowitz (Duke University Medical Center, Durham, NC, US) through Addgene (Cambridge, MA, US). ß-arrestin 2 tagged with m-cherry or Rluc at its C-terminus (ß-arrestin 2-mcherry) was generated by PROTEX (Protein Expression Laboratory) at the University of Leicester (http://www2.le.ac.uk/departments/biochemistry/ facilities/protex). Briefly, ß-arrestin 2 was amplified by PCR using primers, which removed the stop codon and was subcloned into pLeics-30 or pLeics-85 expression vectors. To generate the GHSR1a WT and three mutants fused to enhanced yellow fluorescent protein (eYFP; GHSR1a), receptors were amplified by PCR using primers which removed the stop codon and introduced a 5′ HinDIII and 3′ Kpn I restriction sites. The template for the reaction was GHSR1a WT in pcDNA3.1 vector. The resulting PCR products were subcloned into pcDNA3.1 upstream of full-length eYFP to generate C-terminal eYFP fusion constructs. Mutations were introduced into the C-terminus of the resulting fusion protein using the QuikChange method (Stratagene, Cheshire, UK). Renilla luciferase-tagged ß-arrestin 1 (Rluc-ß-arrestin 1) was provided by Prof. Mark Scott (Institut Cochin, Paris, FR). All cellular assays involving transient transfections were only carried out if the transfection efficiency was 80% or higher.

GHSR1a Receptor Purification and Mass Spectrometry. For GHSR1a purification, the
GHSR1a-EGFP cell line was harvested (10 confluent T175 flasks), resuspended in Krebs/HEPES buffer and stimulated with ghrelin (100 nM, 5 min). Membranes were then prepared and solubilized by addition of 5 mL of TE buffer plus a mixture of protease and phosphatase inhibitors (Roche Applied Science). After centrifugation at 20,000 × g, the resulting supernatant was diluted 1:1 with PBS, and the receptor was then purified on GFP-trap (Chromotek, DE). After extensive washing with solubilization buffer containing 0.5% Nonidet P-40, the resin was resuspended in 2 × SDS-PAGE sample buffer. The sample was resolved by SDS-PAGE on 10% gels and stained with colloidal Coomassie Blue. Purified GHSR1a receptor was excised from the polyacrylamide and washed three times for 5 min with 50 mM ammonium bicarbonate. Reduction and alkylation of cysteines were performed by addition of 10 mM dithiothreitol (DTT) in 50 mM ammonium bicarbonate at 55 °C for 30 min followed by addition 100 mM iodoacetamide in 50 mM ammonium bicarbonate for 30 min in the dark. Gel slices were washed three times for 5 min with 50 mM ammonium bicarbonate containing 50% acetonitrile and incubated overnight at 37 °C in 50 mM ammonium bicarbonate containing 10% (v/v) acetonitrile and 1 μg of sequencing grade trypsin (Promega, Southampton, UK). After tryptic digestion, phosphopeptides were enriched using PHOS-Select TM iron affinity resin.
LC-MS/MS was carried out on each sample using an LTQ OrbiTrap mass spectrometer (Applied Biosystems, Warrington, UK). Peptides resulting from in-gel digestion were loaded at a high flow rate onto a reverse-phase trapping column (0.3 mm inner diameter × 1 mm), containing 5 μm of C18 300 Å Acclaim PepMap media (Dionex, UK) and eluted through a reverse-phase capillary column (75 μm inner diameter × 150 mm) containing Symmetry C18 100 Å media (Waters) that was self-packed using a high pressure packing device (Proxeon Biosystems, Odense, Denmark). The output from the column was sprayed directly into the nanospray ion source of an LTQ Orbital mass spectrometer. The resulting spectra were searched against the UniProtKB/SwissProt data base using MASCOT software (Matrix Science Ltd.) with peptide tolerance set to 5 ppm and the MS/MS tolerance was set to 0.6 Da. Fixed modifications were set as carbamidomethyl cysteine with variable modifications of phosphoserine, phosphothreonine, phosphotyrosine, and oxidized methionine. The enzyme was set to trypsin/ proline, and up to two missed cleavages was allowed. Peptides with a Mascot score greater than 20 and where the probability (p) that the observed match was a random event was < 0.05 were included in the analysis. The spectra of peptides reported as being phosphorylated were interrogated manually to confirm the precise sites of phosphorylation.