Role of Ca2+ and L-Phe in Regulating Functional Cooperativity of Disease-Associated “Toggle” Calcium-Sensing Receptor Mutations

The Ca2+-sensing receptor (CaSR) regulates Ca2+ homeostasis in the body by monitoring extracellular levels of Ca2+ ([Ca2+]o) and amino acids. Mutations at the hinge region of the N-terminal Venus flytrap domain (VFTD) produce either receptor inactivation (L173P, P221Q) or activation (L173F, P221L) related to hypercalcemic or hypocalcemic disorders. In this paper, we report that both L173P and P221Q markedly impair the functional positive cooperativity of the CaSR as reflected by [Ca2+]o–induced [Ca2+]i oscillations, inositol-1-phosphate (IP1) accumulation and extracellular signal-regulated kinases (ERK1/2) activity. In contrast, L173F and P221L show enhanced responsiveness of these three functional readouts to [Ca2+]o. Further analysis of the dynamics of the VFTD mutants using computational simulation studies supports disruption in the correlated motions in the loss-of-function CaSR mutants, while these motions are enhanced in the gain-of-function mutants. Wild type (WT) CaSR was modulated by L-Phe in a heterotropic positive cooperative way, achieving an EC50 similar to those of the two activating mutations. The response of the inactivating P221Q mutant to [Ca2+]o was partially rescued by L-Phe, illustrating the capacity of the L-Phe binding site to enhance the positive homotropic cooperativity of CaSR. L-Phe had no effect on the other inactivating mutant. Moreover, our results carried out both in silico and in intact cells indicate that residue Leu173, which is close to residues that are part of the L-Phe-binding pocket, exhibited impaired heterotropic cooperativity in the presence of L-Phe. Thus, Pro221 and Leu173 are important for the positive homo- and heterotropic cooperative regulation elicited by agonist binding.


Introduction
The human calcium (Ca 2+ )-sensing receptor (CaSR) is a seven transmembrane, G protein-coupled receptor (GPCR) that is expressed at the highest levels in the parathyroid glands and kidneys [1]. The principal role of CaSR is to sense alterations of the extracellular calcium concentration ([Ca 2+ ] o ) and to maintain Ca 2+ homeostasis by regulating parathyroid hormone (PTH) secretion as well as renal Ca 2+ reabsorption. Like other members of family C in the GPCR superfamily, the CaSR possesses a large extracellular domain (ECD) consisting of more than 600 amino acids [2]. Ca 2+ , the principal physiological agonist for CaSR, is thought to bind to the ECD of the dimeric form of the receptor resulting in activation of phospholipase C (PLC), with an attendant accumulation of inositol 1,4,5-triphosphate (IP 3 ) followed by a resultant increase in the cytosolic calcium ([Ca 2+ ] i ) concentration [3].
Fluctuations of the plasma levels of amino acids can regulate the rate of hormone synthesis and secretion as well as Ca 2+ metabolism, among other processes [4]. CaSR is present throughout the gastrointestinal tract [5,6]. Lamino acids, especially aromatic amino acids, are known to enhance the sensitivity of CaSR to [Ca 2+ ] o , which could be one potential explanation for how dietary protein modulates [Ca 2+ ] o homeostasis in normal individuals as well as in patients with chronic renal failure [4,7].
Mutations of the CaSR can perturb intracellular signaling events (e.g., intracellular calcium responses) and disrupt regulation of PTH secretion from the parathyroid chief cell and Ca 2+ reabsorption in the renal tubule. Mutations that inactivate CaSR (i.e., result in loss-of-function) cause familial hypocalciuric hypercalcemia (FHH) and neonatal severe primary hyperparathyroidism (NSHPT) [8,9]. On the other hand, activating mutations of the CaSR lead to autosomal dominant hypocalcemia with hypercalciuria (ADHH) [10]. These mutations, including the four studied here that involve the two ''toggle'' residues, Leu 173 and Pro 221 , where mutations of the same residue can either activate or inactivate the CaSR, serve as a rich source of structure-function information.
Interestingly, among thirty-four identified ECD missense mutations in patients with FHH, NSHPT and ADHH, 18 were located within 10 Å of one or more of the five Ca 2+ -binding sites predicted in our previous studies [11], particularly site 1, which is proposed to be the principal site for Ca 2+ -binding during receptor activation as well as for regulating both the homotropic and heterotropic cooperativity of CaSR [12] [13]. Our group has also reported a putative L-Phe binding site composed of residues S170, Y218, L51, S272 and T145, which are located adjacent to the predicted calcium binding pocket 1 at the hinge region of the CaSR ECD [12]. The location of the L-Phe binding site is pivotal in cooperatively regulating the activity of the CaSR [12].
Our previous computational studies showed that movements of Ca 2+ -binding site 1 are dynamically correlated with those of the other predicted calcium-binding sites. Mutations in Ca 2+ -binding site 1 could, therefore, affect both the homotropic and heterotropic cooperativity of the CaSR. These results led us to hypothesize that residues 173 and 221 are essential for the functional cooperativity orchestrated by calcium and L-Phe binding, and that related disease-associated mutations are the result of alteration or perturbation of the receptor's associated molecular connectivity.
In the present studies, we show that the loss-of-function mutations, L173P and P221Q, but not the gain-of-function mutations, L173F and P221L, alter the cooperativity of the CaSR though analyzing changes in functional readouts in response to alterations in [Ca 2+ ] o . Furthermore, L173P and P221Q disrupt the strong correlated motions among the various calcium-binding sites as demonstrated by molecular dynamics (MD) simulations. L-Phe induced positive heterotropic cooperativity of the P221Q mutant, but had only a limited effect in potentiating the functional activity of L173P due to its restricted molecular dynamical features. These in vitro and in silico results provide important insights into how Ca 2+ and L-Phe interact at their respective binding sites in the cleft of the VFTD, which are essential for the maintenance of calcium homeostasis that is required for normal physiological function.

Cell culture and transfection
Human embryonic kidney cells (HEK293) (ATCC) were cultured under standard condition (5% CO 2 , 37˚C) in High Glucose Dulbecco's modified Eagle's medium (DMEM) (Sigma Chemicals) containing 10% fetal bovine serum with 100 mg/ml penicillin-streptomycin. The CaSR mutations were introduced using site-directed mutagenesis (QuikChange, Stratagene, La Jolla, CA, USA) [14]. Cells were transfected with either pEGFP-N1-WT-CaSR or mutant CaSRs using Lipofectamine 2000 in reduced serum Opti-MEM medium following the manufacturer's instructions (Invitrogen). After 4,6 hours, the medium was changed to High Glucose DMEM, and the cells were incubated for an additional 48 hours to increase receptor expression. Transfection efficiency and expression levels were confirmed by analyzing the fluorescence intensity of the EGFP-tagged CaSR by fluorescence microscopy (Leica, DMI600B).

Measurement of [Ca 2+ ] i responses in single cells transfected with WT or mutant CaSRs
with Fura-2 by incubation with 4 mM Fura-2 AM in 2 mL physiological saline buffer (10 mM HEPES, 140 mM NaCl, 5 mM KCl, 1.0 mM MgCl 2 , 1 mM CaCl 2 and pH 7.4). The coverslips were mounted in a bath chamber on the stage of a Leica DM6000 fluorescence microscope. The cells were alternately illuminated with 340 or 380 nm light, and the fluorescence at an emission wavelength of 510 nm was recorded in real time as the concentration of [Ca 2+ ] o was increased in a stepwise manner in the presence or absence of 5 mM L-Phe. The ratio of emission fluorescence intensity from both excitation wavelengths was then monitored as a function of [Ca 2+ ] o and utilized as a parameter reflecting changes in [Ca 2+ ] i All experiments were performed at room temperature. In addition to the use of the cell population assay as a means of calculating EC 50 for Ca 2+ o or L-Phe, the signals from 30 to 60 single cells were recorded for each measurement in a single cell [Ca 2+ ] i assay. In the latter assay, [Ca 2+ ] i oscillations in individual cells were defined as three or more successive fluctuations in [Ca 2+ ] i from the baseline after the initial peak.
Immunostaining and Western blotting of CaSR pcDNA3.1-CaSRs were used in the immunostaining and Western blotting experiments. This CaSR construct contains a flag-tag between Asp 371 and Thr 372 . For immunostaining, 48 hours after transfection, cells were fixed with 3.7% formaldehyde for 15 min at room temperature, and subsequently labeled with mouse anti-Flag monoclonal antibody (1:3000) overnight at 4˚C. The next day cells were washed with PBS and stained with goat anti-mouse Alexa488conjugated secondary antibodies. Images were collected on a Zeiss LSM700 confocal microscope (Carl Zeiss, German). For Western blotting, cells were lysed with RIPA buffer (Millipore, MA). Protein concentrations were determined using a Bradford assay (Bio-Rad laboratories, Hercules, CA). Lysates containing 40 mg of protein were separated by 7.5% SDS-PAGE. After gel electrophoresis, the proteins were transferred to a nitrocellulose membrane, probed with anti-flag monoclonal antibody (1:3000) in 3% milk/TBST overnight at 4˚C. Blots were subsequently probed with alkaline phosphatase-conjugated secondary antibody (1:3000) for 1 hour at room temperature. After washing, the signals were detected by standard enhanced chemiluminescence. The signals were quantitated using ImageJ.

Determination of ERK 1/2 phosphorylation
Thirty-six hours post transfection of monolayers of HEK293 cells with CaSR or its mutants, cells were incubated in serum-free high glucose DMEM medium supplemented with 0.2% w/v BSA at 37˚C overnight. On the following day, cells were first incubated with HBSS for 30 min, followed by stimulation with varying levels of CaCl 2 (0-20 mM) with or without L-Phe (5 mM) for 10 min. At the end of the [Ca 2+ ] o stimulation, cells were lysed with RIPA lysis buffer (Millipore, CA, USA). In total, 150 mg aliquots of lysate protein were loaded into either a 12.5% SDS-gel or a 4%-12.5% gradient gel for PAGE and analyzed by western blotting with an anti-phospho-p44/42 ERK polyclonal antibody (Cell Signaling Technology, Beverly, MA, USA) diluted (1:2000

Computational analysis and Molecular Dynamics (MD) simulation of CaSR
CLUSTALW was used to align the sequences of the human CaSR ECD (residues 25-530) and the mGluR1 ECD [16]. The structure of the ECD of CaSR was modeled based on the crystal structure of mGluR1 (1EWT, 1EWK and 1ISR) using SWISS-MODEL [17] and MODELLER software [18], and the potential Ca 2+ -binding sites in the CaSR ECD were predicted using MetalFinder [19], The electrostatic potentials were calculated using Pymol.
MD simulation provides an approach complementary to experiments in live cells for understanding biomolecular structure, dynamics, and function. The initial coordinates for all the simulations were taken from a 2.20 Å resolution xray crystal structure with PDBID 1EWK [20]. The AMBER 10 suite of programs [21] was used to carry out all of the simulations in an explicit TIP3P water model [22], using the modified version of the all-atom Cornell et. al. [23] force field and the re-optimized dihedral parameters for the peptide v-bond [24]. An initial 2 ns simulation was performed using NOE restraint during the equilibration in order to reorient the side chains of residues in the Ca 2+ -binding site, but no restraints were used during the actual simulation. A total of three MD simulations were carried out for 50 ns each on the apo-form and the ligand-loaded forms. During the simulations, an integration time step of 0.002 ps was used to solve the Newton's equation of motion. The long-range electrostatic interactions were calculated using the Particle Mesh Ewald method, [25], and a cutoff of 9.0 Å was applied for non-bonded interactions. All bonds involving hydrogen atoms were restrained using the SHAKE algorithm [26]. The simulations were carried out at a temperature of 300 K and a pressure of 1 bar. A Langevin thermostat was used to regulate the temperature with a collision frequency of 1.0 ps-1. The trajectories were saved every 500 steps (1ps). The trajectories were analyzed using the ptraj module in AMBER 10.

Statistics
The data are presented as means ¡ SE of the indicated number of experiments. One-way repeated measures analysis of variance (ANOVA), followed by the Dunnett post hoc test was performed using SPSS (SPSS Inc, Chicago, IL, USA) for multiple comparisons to compare the differences between groups without adding L-Phe. Two-way ANOVA was used for comparison among different groups in the presence of L-Phe. A P value of , 0.05 was considered to indicate a statistically significant difference.

Results
The disease-associated CaSR mutations alter the [Ca 2+ ] otriggered [Ca 2+ ] i oscillation pattern The two residues (Leu 173 , Pro 221 ) involved in the four disease-associated mutations studied here are located in the hinge region near the predicted calciumbinding site 1, which is located between lobe 1 and lobe 2. Meanwhile they are also adjacent (within 10 Å ) to the reported L-Phe binding pocket formed by residues K47, L51, W70, T145, G146, S169, S170, I187, Y218, S272, H413 and R415, which can exert effects over a large region of the ECD through positive heterotropic cooperativity as shown in Figure 1a [12]. The immunostaining as well as the Western blotting results suggest that all of the CaSR mutants are expressed at the cell surface at similar levels ( Figure 1b, 1d, 1e).
Analysis of the [Ca 2+ ] o -triggered [Ca 2+ ] i signaling (Figure 1c), especially the pattern of [Ca 2+ ] i oscillations, is a straightforward method to depict the potential alterations in the homotropic cooperativity induced by mutations. Three parameters were employed to analyze the [Ca 2+ ] i oscillation patterns: the starting point, which refers to the [Ca 2+ ] o at which a given cell starts to show at least three continuous, sequential peaks; the frequency, which is defined as the number of Figure 1. The disease-related mutations are located in the hinge region near Ca 2+ -binding site 1. a. Model structure of CaSR ECD based on the mGluR1 crystal structure (PDB entry: 1ISR) was generated using MODELER 9v4 and Swiss-Model. The global view of the ECD is shown in the left panel. Spheres highlighted in red: Ca 2+ ; Purple: residues involved in predicted Ca 2+ -binding sites. L-phenylalanine (in yellow) is positioned at the hinge region between the two lobes by Autodock vina. Right panel: A zoomed in view of the Ca 2+ -binding pocket of site 1. Residues involved in Ca 2+ -binding site 1are highlighted in purple; residues with disease-related mutations are highlighted in pink; residues predicted to interact with L-Phe are presented in green. b. Immunofluorescence analysis of surface expressed WT CaSR and its mutants in HEK293 cells. Immunostaining was done with anti-flag monoclonal antibody, and detection was carried out with Alex Fluor 488-conjugated, goat anti-mouse secondary antibody. Green: CaSR. The images were taken using equal exposure times. c. Schematic figure of calcium and L-Phe induced downstream signaling changes and the principle for the measurement of IP 1 accumulation. Red dots represent calcium ions. d. total proteins (40 mg) were applied to SDS-PAGE under non-reducing conditions and blotted with anti-flag antibody. GAPDH was used as an internal control. e. The signals were analyzed using ImageJ and the normalized intensities were compared with WT CaSR.  Table 1). The two inactivating mutants (L173P and P221Q) responded to changes in [Ca 2+ ] o over a substantially different range. Therefore, the frequencies were measured at the level of [Ca 2+ ] o at which the majority of the cells (.50%) started to oscillate-15 mM for L173P and 5 mM for P221Q. For L173F and P221L, the frequencies were measured at 2.5 mM [Ca 2+ ] o . The oscillation frequency for the two inactivating mutants did not differ from that of the WT (1.3¡0.1 peaks/min) and were 1.3¡0.1 peaks/min for L173P and 1.5¡0.1 peaks/min for P221Q, but there were slight but significant increases for L173F (1.7¡0.1 peaks/min) and P221L (1.6¡0.1 peaks/min).
We also studied the oscillation ending point. About Table 1). Notably, the two inactivating mutations showed increases in EC 50 and disruption of the CaSR's  Table 1).
Activation of the G q/11 pathway by GPCRs activates phospholipase C (PLC), which induces an increase in [Ca 2+ ] i caused by the associated accumulation of inositol phosphates (IPs) [3]. Thus the activation of PLC in response to increases in [Ca 2+ ] o in cells transiently transfected with WT CaSR were assessed by  Table 2). In contrast, the L173F and P221L mutants showed substantial decreases in their EC 50 s to 1.2¡0.1 mM and 1.1¡0.1 mM, respectively ( Table 2).
We further examined the effects of [Ca 2+ ] o on mitogen-activated protein kinases (MAPK) (ERK 1/2 ) activation. Exposure of WT CaSR-transfected HEK293       Table 2). Thus results from the IP 1 experiments were consistent with the [Ca 2+ ] i readouts in terms of EC 50 . L-Phe also had an impact on the [Ca 2+ ] o -triggered ERK 1/2 activities of WT and mutant CaSRs in CaSR transfected cells. As shown in Figure 6 and Table 2 Table 3).

Computational simulations reveal different dynamic behaviors of the four mutants
The correlated motions in Figure 8 illustrate the molecular dynamic motions of WT CaSR and its mutants obtained from computational simulations. In the modeled WT structure, Site 1 has strong correlated motions with sites 2, 3 and 4, shown in blue (negative correlated motions between a pair of residues signify motions in opposite direction) and red (positive correlations, which are movements in the same direction) (Figure 8, top two panels). We also have reported in our earlier work that the presence of both L-Phe and Ca 2+ in site 1 produces greater correlated motions among the Ca 2+ -binding sites compared with L-Phe alone in site 1 [12]. Interestingly, the loss-of-function mutations, L173P and P221Q, exhibit less correlated motions compared to the WT CaSR as demonstrated by a decrease in the negative (blue) and/or positive correlated motions (Figure 8, bottom panel). On the other hand, the correlated motions of the gain-of-function mutants, L173F and P221L, are dramatically increased in regions similar to those of the WT correlated motions in the presence of L-Phe, indicating an enhancement of correlated motions between the respective residues in these two mutants (Figure 8, middle panel). A closer analysis of the correlated motions of L173F and P221L reveals that the negative correlations between Site 1 and Site 3 and between Site 1 and Site 2 observed in the WT correlation map become positive correlations in the gain-of-function mutants indicating that the mutations might have profound impacts on the dynamic properties of the CaSR-ECD. Especially for P221L, strong negative correlations between lobe 1 and lobe 2 are observed as indicated by the abundant blue area between residues 200-300 and residues 24-170. These results indicate that residues Leu 173 and Pro 221 play pivotal roles in modulating the molecular connectivity between the binding sites for  Figure S1a), while L173P/P221L exhibited oscillation patterns similar to those of the L173P mutant, suggesting that the activating mutant P221L cannot overcome the inactivating effect of L173P ( Figure  S1b). These results suggest that mutations close to calcium binding Site 1 may play key roles in modulating correlated motions of the CaSR. The docking of L-Phe to CaSR mutant P221Q engendered a more correlated molecular dynamics features similar to WT CaSR ECD, suggesting that the receptor is converted into an active conformation. Intriguingly, the other inactive mutant L173P did not exhibited dramatic alterations in its correlated motion in response to L-Phe compared to P221Q. This result suggests that the difference in the influence of L-Phe on potentiating the activity of the CaSR mutants may rely on the intrinsic molecular dynamic connectivity that is disrupted by mutant L173P. It is worth noting that the activating mutant L173F also showed enhanced correlated motions, which could be an explanation for the left shift curve of the ERK activity. The addition of L-Phe was unable to make the profile of mutant P221L into a more dramatically active state, which correlated well with the in vitro studies, suggesting the mutation is already in a ''fully active'' state upon the binding with Ca 2+ similar as WT CaSR with both Ca 2+ and L-Phe.
Principal component analysis (PCA) separates out the protein motions into principal modes ranked according to their relative contributions [27]. The analysis was done on the trajectories of the four mutations and the WT CaSR from the molecular dynamics simulations in order to predict the effect of the mutations on the response of L-Phe and Ca 2+ at the atomic level ( Figure 9).
Projection of the trajectories of the CaSR mutants in both the apo forms (upper panel) and the holo forms (lower panel) onto the first three modes that accounted for the majority of the total fluctuations is shown in Figure 9. In the absence of Ca 2+ and L-Phe, the conformations sampled by the CaSR mutants are similar to that of the WT CaSR. However, in the L-Phe and Ca 2+ -loaded forms, the conformations of the CaSR mutants L173P and L173F are distinguishable from the other mutants and WT CaSR (Figure 9). The results suggest that Ca 2+ and L- Phe might shift the population of conformational ensembles of mutant L173P and L173F in a different way than in WT CaSR or in the other two mutants, supporting the essential role for the residue Leu 173 in the molecular connectivity underlying the Ca 2+ -and Phe-mediated functional cooperativity of CaSR.

Discussion
Loss-of-function mutations impair the functional cooperativity of the CaSR The CaSR hinge region is considered to be crucial for the sensing of agonists (e.g., polyvalent cations and amino acids) [19,28]. Mutations near Ca 2+ -binding site 1 are hypothesized to influence the functional cooperativity of the entire receptor.
In vitro studies showed that the L173P and P221Q mutants produce a receptor with reduced positive homotropic cooperativity and impaired capability to sense [Ca 2+ ] o , while the L173F and P221L mutants enhance the sensitivity of the CaSR to [Ca 2+ ] o compared to the WT receptor, but barely change the cooperativity of the receptor. The results with the loss-of-function mutants indicate that Leu 173 and Pro 221 are important in maintaining the positive homotropic cooperativity among the receptor's calcium-binding sites. Conversely, the gain-of-function mutants show that these residues also are important for restraining the receptor from assuming its most active form (e.g., that seen in the WT CaSR in the presence of both high [Ca 2+ ] o and L-Phe). Thus unlike mutations of residues that cause only gain-or loss-of-function, Leu 173 and Pro 221 serve dual functions, perhaps reflecting their key locations in the cleft between the receptor's two lobes and their proximities to the binding sites for both [Ca 2+ ] o and L-Phe.
[Ca 2+ ] i oscillations have been postulated to be the result of complex responses from multiple signaling pathways. The [Ca 2+ ] o -triggered [Ca 2+ ] i oscillation pattern is believed to be modulated by the activity of phosphoinositide pathway [29] as well as a negative feedback loop involving the inhibitory effect of protein kinase C (PKC) on IP 3 production [30]. Downstream intracellular signaling responses, such as the production of IP 1 and the activity of ERK 1/2 , which will be discussed later, can be reflected in the oscillation patterns associated with the different disease-related mutations. Although most research has focused on intracellular calcium mobilization in HEK293 cells heterologously expressing the CaSR, CaSR-induced [Ca 2+ ] i oscillations have also been found in cells with endogenously expressed CaSR, for instance in parathyroid cells [31], bovine anterior pituitary cells [32], opossum kidney (OK) cells [33] and medullary thyroid carcinoma cells [34]. Oscillations in [Ca 2+ ] i modulate not only the rate of parathyroid hormone (PTH) secretion, but also gene expression and other processes [35][36][37][38] ] o -evoked increases in ERK 1/2 activities of WT-CaSR and its four mutants were in accordance with the accumulation of IP 1 . Stimulation of the ERK 1/2 activity in CaSR-transfected HEK293 cells has been reported to involve PKC-mediated as well as a PTX-sensitive, tyrosine kinase-dependent pathways [39]. Although it is believed that the carboxyl terminus of the protein is involved in the activation of MAPK signaling by interacting with the scaffold protein, filamin A [40,41], other data showed that that alterations in the extracellular domain can also affect the ERK cascade, which is in agreement with previous studies [42]. Since all the mutations studied here were expressed at similar levels on the cell membrane (Figure 1b), these downstream signaling changes may at least partially be contributed by the disruption of cooperativity among different calcium binding sites introduced by different mutations. On the other hand, the gain-of-function mutations may increase the stability the receptor as suggested by other studies [43], or they possibly induce more correlated motions among the calcium-binding sites as suggested by the dynamic cross-correlation map.

L-Phe-induced heterotropic cooperativity rescues activity of the CaSR
In the current study, we showed that L-Phe could affect one or two of the three parameters depicting the [Ca 2+ ] o triggered [Ca 2+ ] i oscillation patterns in all of the mutants as measured in individual cellular responses (e.g., [Ca 2+ ] i oscillation starting points, ending points and [Ca 2+ ] i oscillation frequencies). The alterations in the oscillation patterns after the introduction of the allosteric activator could be the result of the activation of signaling downstream of G a12/13 or a combined effect with the positive heterotropic cooperativity between the multiple Ca 2+ binding sites induced by L-Phe. The interaction between CaSR and L-Phe has been reported to modulate the G a12/13 -RhoA pathway [44]; however, L-Phe could also potentially enhance [Ca 2+ ] o -induced alterations in intracellular signaling pathways through G q/11 pathway as reflected by the IP 1 accumulation assay.
The present study also confirms that the ERK signaling pathway in CaSRtransfected HEK293 cells can be modulated not only by [Ca 2+ ] o but also by the positive allosteric modulator, L-Phe [45]. It should be noted that the ability of L-Phe to enhance the intracellular responses of the various signaling pathways varied between different mutations. L-Phe enhanced the maximum responses of the ERK signaling pathway to a greater extent than it altered the apparent affinities in the two gain-of-function mutants, although it barely showed any influence on the [Ca 2+ ] o -triggered [Ca 2+ ] i and IP 1 responses for these mutants, which may explain, in part, the dramatic changes in correlated motions from the computational simulation results using molecular dynamics. As a signaling messenger further downstream than either IP 1 or [Ca 2+ ] i , the phosphorylation of ERK 1/2 may not merely reflect activation of the G q/11 pathway, but could be contributed to by other G-proteins (e.g., G i ). In agreement with the in silico results, L-Phe exhibited less effect on the regulation of the ERK 1/2 pathway in mutant L173P. The predicted L-Phe-binding site has been reported to be located near the hinge region between lobe 1 and lobe 2 and to involve residues S169A/S170A/S171A [46]. Leu173 is located within 5 Å of this region. The Leu to Pro change could potentially generate additional steric effects on the closure of lobe 1 and lobe 2 so that addition of exogenous L-Phe has relatively little impact on the CaSR's response to changes in [Ca 2+ ] o . Thus, the hetero-cooperativity between the allosteric modulator and calcium binding sites could be disturbed, resulting in the lack of a left-shifted concentration-response curve in response to [L-Phe].
Disease-associated mutations influence the molecular connectivity between the Ca 2+ -binding sites of the CaSR as revealed by molecular dynamic simulations Based on all of the in vitro results and the in silico simulations, we propose a model to illustrate the possible mechanism by which the disease-associated mutations affect the function of CaSR and how their activities can be modulated by Ca 2+ and L-Phe through the molecular connectivity that is encoded at the hinge region of the ECD of the protein. As shown in Fig 10, there are four functional states of the CaSR. In the absence of Ca 2+ , WT CaSR is at the basal state can be converted to the active state upon Ca 2+ binding to site 1 at the hinge region. Subsequent binding of L-Phe at the adjacent hinge region further strengthens molecular connectivity and elevates the receptor to a ''fully active'' status in the presence of high [Ca 2+ ] o by a heterotropic positive cooperative mechanism. Mutations close to the hinge region such as those at Leu 173 and Pro 221 affect the overall dynamic correlated motions of the respective mutant receptors and the Ca 2+ -induced homotropic cooperativity, which could further influence the conformation of the receptor. The inactivating mutation L173P impairs the molecular connectivity in the WT CaSR and cannot be rescued by L-Phe. On the other hand, the inactivating mutation P221Q weakens the connectivity but can be partially rescued by binding Phe with a resultant activity similar to the basal activity of the WT CaSR. In contrast, the gain-of-function mutants L173F and P221L may already have been converted to the active state with dynamic properties similar to that of Ca 2+ -loaded WT CaSR. They can be further activated upon binding extracellular Ca 2+ to reach a state close to the ''fully active'' state (Phe + [Ca 2+ ] o in the WT) that is insensitive to further The impaired receptor function and the cross-talk between Ca 2+ -binding sites can be at least partially rescued for some mutants by L-Phe (e.g.P221Q). However, if the mutation interferes the interaction between CaSR and L-Phe, the function of the receptor may not be fully recovered (e.g.L173P). CaSR gain-of-function mutants (left) exhibit enhanced correlated motions (double line arrows) and their activity is not further potentiated by L-Phe, potentially due to a ceiling effect. activation by L-Phe achieving an apparent ''ceiling'' represented by the maximal achievable activity of the CaSR.
The results from PCA may also provide an explanation as to why L-Phe interacts preferably with the P221Q mutant. L-Phe could not effectively shift the inactive population of conformational ensembles of L173P to an active ensemble of conformations as it did with the other inactivating mutant and WT. However, whether the in silico simulations provide a novel method to predict mutational effects on CaSR will ultimately require the kind of structural detail available for the mGluRs, whose structures have been solved. Here, the model structure does not take into consideration the possible interactions between the ECD and the extracellular loops and/or transmembrane domain as well as the intracellular domains. Thus, caution is merited in over-interpreting the in silico results. Nevertheless, the predicted structural motions may help to identify the mutations or ligands that produce increased or diminished coupling among specific structural elements within the CaSR ECD. While this type of analysis is limited at the moment to the environment of the orthosteric site, this strategy could facilitate drug design for modulating the functions of disease-related mutations affecting the CaSR, as well as other members of the family C GPCRs.
The distinctive combinatorial effects of L-Phe and [Ca 2+ ] o on different mutant CaSRs also suggests that ligand selectivity at the orthosteric site and receptor activity at non-orthosteric sites could both contribute to the function of the receptor. This feature of the adjacent orthosteric binding sites and allosteric regulatory sites that cooperatively regulate the molecular connectivity may be shared by other members of the family C GPCRs [47,48]. In addition to the Ca 2+sensing receptor, [Ca 2+ ] o regulates 14 of the other members of the family C G protein-coupled receptors (GPCRs), including the metabotropic glutamate receptors (mGluR), c-aminobutyric acid GABA B receptors and receptors for pheromones, amino acids and sweet substances [1,[49][50][51][52][53][54][55]. The observed molecular modulatory mechanism may be shared by other members of the family C GPCRs [47,48]. For instance, the analyses of multiple point mutations in the taste receptors (T1R1/T1R3) suggested that the combination of the two distinct determinants selectivity at the ligand binding site and receptor activity at allosteric sites may mediate the ligand specificity of T1R1/T1R3 [56].
In conclusion, through analysis of [Ca 2+ ] i oscillations and the [Ca 2+ ] o -and L-Phe triggered downstream signaling changes, functional positive cooperativity of CaSR was revealed to be disrupted in the identified FHH-associated CaSR mutations, L173P and P221Q, but not in the ADHH related mutations, L173F and P221L. The addition of L-Phe rescued the function of P221Q by inducing heterotropic positive cooperativity. The distinctive correlated motions of the gainof-function mutations, on the one hand, and the loss-of-function mutations, on the other, and the potential utility revealed by the molecular dynamic simulation in providing insights into the mechanism for the disease-associated functional alterations in the receptor may also apply to other GPCR proteins.